Lobsters: ocean icons in changing times

ICES Journal of
Marine Science
Journal du Conseil
Volume 72 Supplement 1 July 2015
http://icesjms.oxfordjournals.org/
Lobsters in a Changing Climate
Proceedings of the 10th International Conference and Workshop on
Lobster Biology and Management
Conveners: Patricia Briones-Fourzán (Mexico), Enrique Lozano-Álvarez
(Mexico)
ICES Journal of
Marine Science
Journal du Conseil
Volume 72 Supplement 1 July 2015
http://icesjms.oxfordjournals.org/
Contents
Introduction
P. Briones-Fourzán & E. Lozano-Álvarez
Lobsters: ocean icons in changing times
i7
Reviews
E. Spanier, K. L. Lavalli, J. S. Goldstein, J. C. Groeneveld, G. L. Jordaan, C. M. Jones,
B. F. Phillips, M. L. Bianchini, R. D. Kibler, D. Dı́az, S. Mallol, R. Goñi, G. I. van Der Meeren,
A-L. Agnalt, D. C. Behringer, W. F. Keegan, & A. Jeffs
A concise review of lobster utilization by worldwide human populations from prehistory
to the modern era
i22
J. W. Penn, N. Caputi, & S. de Lestang
A review of lobster fishery management: the Western Australian fishery for Panulirus
cygnus, a case study in the development and implementation of input and output-based
management systems
i35
C. D. Ellis, D. J. Hodgson, C. L. Daniels, D. P. Boothroyd, R. C. A. Bannister, &
A. G. F. Griffiths
European lobster stocking requires comprehensive impact assessment to determine fishery
benefits
i49
Original Articles
S. de Lestang, N. Caputi, M. Feng, A. Denham, J. Penn, D. Slawinski, A. Pearce, & J. How
What caused seven consecutive years of low puerulus settlement in the western rock lobster
fishery of Western Australia?
i59
I. A. Hinojosa, B. S. Green, C. Gardner, & A. Jeffs
Settlement and early survival of southern rock lobster, Jasus edwardsii, under climatedriven decline of kelp habitats
i69
R. A. Wahle, L. Dellinger, S. Olszewski, & P. Jekielek
American lobster nurseries of southern New England receding in the face of climate change
i79
B. K. Quinn & R. Rochette
Potential effect of variation in water temperature on development time of American
lobster larvae
i91
C. D. Ellis, H. Knott, C. L. Daniels, M. J. Witt, & D. J. Hodgson
Geographic and environmental drivers of fecundity in the European lobster (Homarus
gammarus)
i101 M. Butler, IV, R. Bertelsen, & A. MacDiarmid
Mate choice in temperate and tropical spiny lobsters with contrasting reproductive systems
i115 M. J. Butler, IV, A. Macdiarmid, & G. Gnanalingam
The effect of parental size on spermatophore production, egg quality, fertilization success,
and larval characteristics in the Caribbean Spiny lobster, Panulirus argus
i1
i124
i128
i139
i147
i155
i164
i170
i177
i185
i199
i211
i225
i232
i244
i252
i260
R. O’Rorke, S. D. Lavery, M. Wang, R. Gallego, A. M. Waite, L. E. Beckley, P. A. Thompson, &
A. G. Jeffs
Phyllosomata associated with large gelatinous zooplankton: hitching rides and stealing
bites
C. E. Davies, A. F. Johnson, E. C. Wootton, S. J. Greenwood, K. F. Clark, C. L. Vogan, &
A. F. Rowley
Effects of population density and body size on disease ecology of the European lobster in a
temperate marine conservation zone
A. S. Kough, C. B. Paris, D. C. Behringer, & M. J. Butler, IV
Modelling the spread and connectivity of waterborne marine pathogens: the case of PaV1
in the Caribbean
J. S. Goldstein, E. A. Dubofsky, & E. Spanier
Into a rhythm: diel activity patterns and behaviour in Mediterranean slipper lobsters,
Scyllarides latus
P. Briones-Fourzán, R. Domı́nguez-Gallegos, & E. Lozano-Álvarez
Aggressive behaviour of spotted spiny lobsters (Panulirus guttatus) in different social
contexts: the influence of sex, size, and missing limbs
R. I. Candia-Zulbarán, P. Briones-Fourzán, E. Lozano-Álvarez, C. Barradas-Ortiz, &
F. Negrete-Soto
Caribbean spiny lobsters equally avoid dead and clinically PaV1-infected conspecifics
M. J. Childress, K. A. Heldt, & S. D. Miller
Are juvenile Caribbean spiny lobsters (Panulirus argus) becoming less social?
B. C. Gutzler, M. J. Butler, IV, & D. C. Behringer
Casitas: a location-dependent ecological trap for juvenile Caribbean spiny lobsters,
Panulirus argus
C. B. Butler & T. R. Matthews
Effects of ghost fishing lobster traps in the Florida Keys
I. D. Tuck, D. M. Parsons, B. W. Hartill, & S. M. Chiswell
Scampi (Metanephrops challengeri) emergence patterns and catchability
H. A. Lizárraga-Cubedo, I. Tuck, N. Bailey, G. J. Pierce, A. F. Zuur, & D. Bova
Scottish lobster fisheries and environmental variability
G. Ewing & S. Frusher
New puerulus collector design suitable for fishery-dependent settlement monitoring
E. A. Babcock, W. J. Harford, R. Coleman, J. Gibson, J. Maaz, J. R. Foley, & M. Gongora
Bayesian depletion model estimates of spiny lobster abundance at two marine protected
areas in Belize with or without in-season recruitment
Z. Kordjazi, S. Frusher, C. D. Buxton, & C. Gardner
Estimating survival of rock lobsters from long-term tagging programmes: how survey
number and interval influence estimates
C. Gardner, K. Hartmann, A. E. Punt, & E. Hoshino
Fewer eggs from larger size limits: counterintuitive outcomes in a spatially heterogeneous
fishery
S. Lee, N. D. Hartstein, & A. Jeffs
Modelling carbon deposition and dissolved nitrogen discharge from sea cage aquaculture
of tropical spiny lobster
Published by Oxford University Press for
International Council for the Exploration of the Sea
Conseil International pour l’Exploration de la Mer
H. C. Andersens Boulevard 44-46, DK-1553 Copenhagen V, Denmark
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i1 –i6. doi:10.1093/icesjms/fsv111
Introduction to the Supplement: ‘Lobsters in a Changing Climate’
Introduction
Lobsters: ocean icons in changing times
Patricia Briones-Fourzán* and Enrique Lozano-Álvarez
Instituto de Ciencias del Mar y Limnologı́a, Unidad Académica de Sistemas Arrecifales Puerto Morelos, Universidad Nacional Autónoma de México,
Quintana Roo, México
*Corresponding author: tel: + 52 998 8710009; e-mail: [email protected]
Briones-Fourzán, P., and Lozano-Álvarez, E. Lobsters: ocean icons in changing times. – ICES Journal of Marine Science, 72: i1– i6.
Received 10 April 2015; revised 26 May 2015; accepted 28 May 2015; advance access publication 23 June 2015
.
The 10th International Conference and Workshop on Lobster Biology and Management was held in Cancún, Mexico, in May 2014. The papers
included in this supplementary issue of the ICES Journal of Marine Science are a sample of the multidisciplinary nature of the conference and
provide new knowledge of the biology, ecology, fisheries, and management and aquaculture of clawed, spiny, and slipper lobsters. The emphasis
of the conference was climate change and its consequences for lobster biology, population dynamics, ecology, and fisheries. As noted in several
papers, climate change is already affecting different lobster species by altering growth rates, sizes at maturity, the timing of reproductive processes,
duration of larval development, and the timing and levels of settlement; by affecting key benthic habitat-forming species in settlement habitats; by
increasing the risk of disease and impacting the behavioural ecology of lobsters, and by changing the spatial distribution of the stocks and, hence,
affecting catches and the territorial behaviour of fishers. Other issues addressed at the conference included aquaculture and enhancement—the
holy grails of lobster management—sustainable management strategies, and a fascinating review of the use of lobsters through human history. In
addition to their economic importance, lobsters continue to provide valuable information to understand different marine environments in a
changing climate.
Keywords: biology, clawed lobsters, climate change, fisheries, management, slipper lobsters, spiny lobsters.
Introduction
Marine lobsters are a diverse group of large, abundant, long-lived,
crustaceans that inhabit a wide range of habitats. Lobsters are
highly valued and sustain some of the most profitable fisheries in
all tropical, subtropical, and temperate-cool waters of the world. In
2012, worldwide landings of marine lobsters were 294 000 metric
tons (mt) with a value of around US $2800 million, with an additional
2000 mt produced through aquaculture (FAO, 2014). Chan (2010)
recognized 248 lobster species in 55 genera and 6 families from 4 infraorders; however, most commercially important species are members of
the families Nephropidae (clawed lobsters and scampi), Palinuridae
(spiny or rock lobsters), and Scyllaridae (slipper lobsters).
Apart from their economic importance, lobsters play key roles in
the maintenance of healthy and diverse marine ecosystems, given
their generally high local abundances and trophic position as
benthic consumers. Lobsters exhibit different types of lifestyles and
reproductive strategies—even within the same genus—and display
complex behaviours, yet are relatively easy to keep in the laboratory
and to use in controlled experiments. Thus, lobsters provide valuable
# International
information to understand different marine environments in a changing world and constitute useful subjects, and even model organisms,
for many types of biological and ecological studies (Cobb, 2006).
The International Conference and Workshop
on Lobster Biology and Management
In January 1977, a group of 34 scientists from six countries met in
Perth, Australia, to discuss issues on lobster ecology and physiology.
Over time, the scope widened and the meeting became the
“International Conference and Workshop on Lobster Biology and
Management” (or ICWL for short). The aims of the ICWL are to
review recent advances in the study of all aspects of the biology,
ecology, fisheries, and aquaculture of clawed, spiny, and slipper lobsters; to identify gaps in current knowledge and future research priorities, and to encourage collaborative studies for future research.
Established and new generations of lobster scientists, academics, students, managers, and industry representatives attend the ICWLs
to exchange knowledge and mingle in a friendly atmosphere. After
Perth, subsequent ICWLs were held in Saint Andrews, Canada
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i2
P. Briones-Fourzán and E. Lozano-Álvarez
(1985), La Habana, Cuba (1990), Sanriku, Japan (1993), Queenstown,
New Zealand (1997), Key West, USA (2000), Hobart, Australia (2004),
Charlottetown, Canada (2007), and Bergen, Norway (2011).
The 10th International Conference and Workshop on Lobster
Biology and Management was held in Cancún, Mexico, in May
2014 and was attended by 182 registered participants from 21 countries (Figure 1). To celebrate the fact that this was the first ICWL
held in Mexico, the logo of the 10th ICWL depicted a stylized
“pre-Columbian” spiny lobster (Figure 2). Over the 5-day conference, 200 presentations were delivered: 133 oral and 67 posters, including keynote presentations by Bruce Phillips, Richard Wahle,
Ehud Spanier, James Penn, and Juan Carlos Seijo (see the Book of
Abstracts in Supplementary Material). The emphasis of the 10th
ICWL was “Lobsters in a Changing Climate” and presenters were
encouraged to address this issue from their particular field of
study, but there were in all 14 theme sessions: Climate change;
Behavioural ecology; Diseases and parasites; Population and
community ecology; Lobsters in antiquity; Reproduction, development, and physiology; Connectivity and larval studies; Genetics;
Aquaculture, nutrition, and population enhancement; Fisheries
and fishing technologies; Marine protected areas; Habitat and ecosystem issues for fisheries management; Stock assessment; and
Management. The conference ended on May 23 with the announcement and presentation of the next venue. The 11th ICWL will be held
in Portland, Maine, USA, in 2017.
Lobsters: the state of the art
The 26 papers (3 reviews and 23 original articles) published in this
special issue of the ICES Journal of Marine Science exemplify the diversity of topics addressed at the 10th ICWL. The conference featured a special session on the use of lobsters through the history of
Figure 2. The logo of the 10th International Conference and
Workshop on Lobster Biology and Management featured a
“pre-Columbian” spiny lobster (designed by Jorge Quiñonez-Alvarado).
Figure 1. Participants in the 10th International Conference and Workshop on Lobster Biology and Management (Cancún, Mexico, 18– 23 May
2014).
Lobsters in changing times
humans with eight presentations from different geographic regions
and civilizations, and all the presenters from that session joined their
efforts to produce the comprehensive and fascinating first review
(Spanier et al., 2015). The review shows that lobsters have been
iconic creatures for thousands of years, that humans have used lobsters for food and trade essentially always, and that many indigenous
peoples appear to have harvested lobsters at a sustainable rate.
However, as the demand for lobsters increased, fishing methods
became more efficient and preservation techniques increased the
potential for international trade, eventually leading to the present
status in which most lobster stocks are either fully exploited or overexploited (e.g. Jeffs et al., 2013; Phillips et al., 2013; Wahle et al.,
2013). Therefore, present-day lobster populations are more vulnerable to environmental changes and in dire need of sound and sustainable management strategies.
In the second review, Penn et al. (2015) give an interesting
account of the evolution of the management of the fishery for the
western rock lobster (Panulirus cygnus) in Western Australia,
which is considered one of the best-managed fisheries in the
world. The review shows that the initial use of biological controls
enhanced with input controls (limited entry) was necessary to accumulate sufficient fishery-based data for effective management decision processes that could be understood by fishers. However, fishers
always find ways to maximize capture, which will eventually counteract input controls and impact the stocks, thus leading to the
need for output controls such as total allowable catch with individual transferrable quotas. The management of the fishery for P. cygnus
appears an effective model that may be widely applicable.
Aquaculture and population enhancement remain the holy grails
of lobster management. Clawed lobsters have short larval phases,
but may take years to reach marketable size, so in many European
countries stocking populations of Homarus gammarus with
hatchery-reared juveniles is a common activity believed to provide
a viable tool for sustainable management. However, as noted in
the review by Ellis et al. (2015a), evidence of success (or failure) is
sorely lacking. In particular, there is a dearth of basic knowledge
of ecology and population dynamics of wild European lobsters, especially the elusive early life stages, and little is known of the fitness
of reared lobsters and the impact of stocking on wild lobster population genetics, for which the authors advocate the use of modern
genetic approaches and stock assessment methods.
Four original articles are directly related to the main topic of the
conference: climate change. Indeed, some species that support
major fisheries (e.g. P. cygnus, Jasus edwardsii, and Homarus
americanus) occur in areas that are hotspots for increasing water
temperature (Hobday and Pecl, 2014), which appears particularly
affecting the early life stages of these species. For example, in
Western Australia, very low settlement levels of postlarvae of the
western rock lobster (P. cygnus) were recorded for seven consecutive
years (2006/2007–2012/2013), raising concern in this wellmanaged fishery. By analysing an impressive amount of biological
and environmental data, de Lestang et al. (2015) discovered that
these low settlement levels are related to an earlier seasonal spawning
and higher water temperatures at the time of the onset of spawning,
thus potentially causing a mismatch with other environmental
factors that are related to growth of early-stage larvae and the
return of postlarvae to the coast several months later. Identifying
the factors that affect puerulus settlement will improve assessment
models and support management planning. Along the east coast
of Tasmania, coverage of brown algae has decreased substantially
over the last few decades due to climate change, with potentially
i3
negative consequences for local lobster productivity, as Hinojosa
et al. (2015) found that coverage of brown algae increases settlement
of postlarvae and reduces predation rates on early benthic juveniles
of southern rock lobster (J. edwardsii).
Climate change is having complex effects on populations of
American lobsters (H. americanus) along the east coast of North
America. For example, in Narragansett Bay, Rhode Island, Wahle
et al. (2015) found significant declines in the abundance, distribution, and size composition of juvenile lobsters during 2011– 2012
compared with data from 1990, when the lobster fishery and
larval settlement were at historical highs. Moreover, in 2011–12, juvenile lobsters were restricted to the outer coast and deeper water at
the mouth of the bay, whereas in 1990 they extended from the outer
coast well into the mid-sections of the bay. During this period, water
temperature has increased by 28C for several weeks during summer,
exceeding the thermal physiological threshold for lobsters and
turning semi-enclosed embayments increasingly inhospitable to lobsters in this region. Quinn and Rochette (2015) used different thermal
scenarios to model development time of American lobster larvae. The
model revealed that variability in temperature can also have strikingly
different effects on larval duration of H. americanus depending on
whether the larvae originated in “warm” vs. “cold” waters along the
geographic distribution of the species.
Ten articles deal with biological aspects of lobsters such as reproduction, larval behaviour, diseases, and behavioural ecology, some
of them from the perspective of climate change. In exploited
species such as lobsters, a good understanding of reproductive
biology provides insights into the suitability of management strategies throughout the range of that species and the potential consequences of fishing practices. Using datasets from multiple
locations throughout Europe, Ellis et al. (2015b) found that fecundity at mean size of female European lobsters (H. gammarus) did not
correlate positively with either mean water temperature or latitude,
but with longitude and greater annual ranges in sea surface temperature. The authors identified proximity to stable Atlantic currents as
the most likely driver of the relationship with longitude. Lobsters
have mating behaviours that appear tuned to the cost of reproduction, as suggested by Butler et al. (2015a), who experimentally compared mate choice among males and females of different sizes
between J. edwardsii, a temperate species, and the Caribbean spiny
lobster (Panulirus argus), a tropical species. During the corresponding reproductive season, most females of J. edwardsii chose to
cohabit with large males, whereas large males tended to cohabit
with large females. In contrast, males and females of P. argus associated freely with no preference for mate size. Therefore, compared
with P. argus, the variance in mate quality appears higher in
J. edwardsii and the cost of poor mate choice greater, especially for
larger females. However, as in many other exploited lobsters, large
males of P. argus are being disproportionally extracted from populations, potentially leading to sperm limitation. Butler et al. (2015b)
examined the effect of male size on fertilization success and the role
that parental size plays in gamete and larval quality of P. argus.
Maternal size had no effects on egg size or quality, or larval quality,
but fertilization success declined with spermatophore size, suggesting
that sperm limitation occurs in this species.
Spiny lobsters have a long larval phase (5 – .20 months, depending on the species) that develops in oligotrophic oceanic waters, and
how the phyllosoma larvae survive in these waters has always been a
puzzling question. O’Rorke et al. (2015) report finding several
phyllosomata of P. cygnus attached to a large salp zooid. Using highthroughput sequencing analyses of DNA, they determined that the
i4
larvae were indeed feeding on the salp. Thus, by attaching themselves to large gelatinous prey items, phyllosomata may both
survive better in oligotrophic waters and conserve energy by
riding along with salps.
Climate change is expected to increase the occurrence of diseases
in marine organisms, and levels of disease may further increase in
no-take zones (NTZs) as a consequence of increased density of
hosts. However, in the UK, Davies et al. (2015) found that, although
the abundance and mean size of H. gammarus were significantly
higher in an NTZ relative to a nearby fished zone, endozootic
shell disease (caused by chitinolytic bacteria) was more prevalent
in large lobsters, injured lobsters, and in males irrespective of
zone. The authors make a strong point for the need to monitor
the health status of target species in NTZs both before and after implementation. However, marine pathogens tend to spread more
rapidly than their terrestrial counterparts, and little is known
about their mechanisms of dispersal. Kough et al. (2015) used a
modelling approach to test the potential for dispersal of Panulirus
argus Virus 1 (PaV1), a pathogen that attacks Caribbean spiny lobsters, via two mechanisms (i) as a passive waterborne particle and
(ii) within infected lobster postlarvae. Their results suggest that
the latter mechanism would better explain the current geographic
and genetic distribution of PaV1.
Some slipper lobsters are economically important, but scyllarids
in general remain poorly studied. Goldstein et al. (2015) tested the hypothesis that Scyllarides latus (the Mediterranean slipper lobsters)
would exhibit a circadian rhythm as spiny lobsters do. Using timelapse video and accelerometers, they assessed the activity of lobsters
kept under simulated 12 : 12 light : dark periods for several days,
then under constant conditions of darkness. The results confirmed
the hypothesis and will be useful for investigating seasonal migrations
of S. latus and its responses to changing environmental conditions.
Social behaviour may influence the patterns of population
distribution, and spiny lobsters vary widely in behaviours such
as gregariousness and aggressiveness. Using video techniques,
Briones-Fourzán et al. (2015) examined the relative importance of individual traits in the shelter-related aggressive behaviour of spotted
spiny lobsters (Panulirus guttatus) held in the laboratory. In
same-sex groups, size was an important predictor of aggressiveness,
but the number of missing limbs significantly impacted the degree
of aggressiveness in males. In mixed-sex groups, males were substantially more aggressive than females, and when the largest individual
was a female, it took longer for smaller males to assert themselves, increasing aggressiveness levels. These findings are relevant in the
context of fishing practices (which selectively remove the largest lobsters, mostly males) and increasing loss of complexity in Caribbean
coral reefs (which may reduce availability of suitable shelters).
Panulirus argus is a more gregarious spiny lobster, a behaviour that
is mediated by conspecific chemical cues, but PaV1 alters this behaviour. Using lobsters screened for PaV1 by PCR in Y-maze trials,
Candia-Zulbarán et al. (2015) found that uninfected lobsters
avoided to a similar degree chemical cues emanating from PaV1diseased conspecifics and from dead conspecifics, whereas their
response to chemical cues emanating from subclinically infected or
uninfected conspecifics did not differ from random. The results
suggest that diseased lobsters emit chemicals that are repellent to
healthy conspecifics, and that subclinically infected lobsters either
do not emit these chemicals or do so at subthreshold levels.
Childress et al. (2015) tested the hypothesis that juvenile P. argus
are becoming less social via a meta-analysis of six Y-maze chemical
cue choice experiments conducted between 1996 and 2012. Results
showed that conspecific attraction has indeed decreased since 2010,
P. Briones-Fourzán and E. Lozano-Álvarez
with a significantly lower value in 2012. Some potential explanations
for these results include the current widespread presence of PaV1, and
a fishery-induced selection on attraction behaviour resulting from the
reliance of current fishing methods on aggregation behaviour.
Indeed, in several Caribbean countries, large artificial shelters
(“casitas” or “pesqueros”) are used to facilitate aggregation of
P. argus lobsters, whereas in Florida, lobster traps are “baited” with
live sublegal lobsters that function as chemical attractants for conspecifics. However, both casitas and traps have downsides, as shown in
two articles that examine these particular fishing methods. Gutzler
et al. (2015) found that casitas deployed in nursery areas rich in
natural shelters may increase predation-induced mortality of small
juvenile lobsters—which are not the target of the fisheries—that
may be chemically attracted to the casitas by their larger counterparts.
Therefore, deployment of commercial casitas in nursery areas or
where natural shelter abounds should be avoided. Butler and
Matthews (2015) found that tens of thousands of lobster traps are
lost every year in the Florida fishery, and that these “ghost traps”
remain intact and continue to capture lobsters, fish, and other organisms, for up to 1–2 years, thus having a pervasive negative effect on
many components of the local communities. On average, about
630 000 of P. argus are estimated to die in ghost traps annually in
the Florida fishery. Attempts to address these issues include provisions requiring that all traps have a degradable wooden panel.
Six articles examine different aspects related to the fisheries and
stock assessment of clawed and spiny lobsters. Deep-sea New
Zealand scampi (Metanephrops challengeri) occupy burrows and are
only available to trawl fisheries when emerged on the seabed. Using
acoustic tags and hydrophone receivers moored close to the seabed,
Tuck et al. (2015) discovered strong emergence cycles in relation to
tidal current and time of day. This information was incorporated
into a stock assessment to reduce uncertainty in burrow occupancy
and catchability estimates. In Scotland, Lizárraga-Cubedo et al.
(2015) used 30-year long time-series to analyse the relationship
between some components of the fishery for H. gammarus and
several environmental variables for two stocks, the Hebrides and
Southeast, to explain variability in the fluctuations of both lobster
populations. They found that sea surface temperature, windspeed,
and sea level pressure strongly affected the capture of lobsters
throughout the year, although at different levels between geographic
areas and at different spatial scales.
In some fisheries for spiny lobsters, the use of artificial collectors for
postlarvae allows for long-term settlement monitoring programmes
that provide important data for stock assessment and management.
In Tasmania, “crevice” collectors used to monitor pueruli settlement
of J. edwardsii are typically deployed and serviced by divers in waters
,10 m, although the fishery typically operates in waters .50 m in
depth. Ewing and Frusher (2015) adapted the collectors to fit within
a lobster trap frame and tested these collectors in depths of 50–60
and 90–105 m for 6 months. The collectors caught pueruli at both
depth ranges, but mostly at 50–60 m. Commercial fishing vessels
were successfully used to service these collectors, which is expected
to both reduce the costs of research and promote co-management.
Stock assessment, a vital tool for fisheries management, strongly
relies on modelling approaches. In Belize, Babcock et al. (2015) used
Bayesian depletion models with in-season recruitment and several
lobster catch per unit effort (cpue) indices to estimate the abundance and fishing mortality of adult P. argus lobsters at two
marine reserves: Glover’s Reef (an oceanic reserve surrounded by
deep wall reefs with an NTZ of 20% of its area) and Port
Honduras (a coastal reserve with shallow habitats and an NTZ of
5% of its area). Abundance was estimated at 66 –79 000 lobsters
i5
Lobsters in changing times
in Glover’s Reef, of which 60 –85% were present at the beginning of
the season and the rest recruited into the fished population later in
the season, and at 12 000 lobsters in Port Honduras, where in-season
recruitment was not supported by the data. The models estimated a
harvest fraction of 70% in both reserves. In Tasmania, Kordjazi
et al. (2015) investigated how the number and duration of surveys
affect the precision of survival estimates for male and female J.
edwardsii lobsters in long-term tagging studies. Using Cormack–
Jolly –Seber models and model selection, they found that a
minimum of five surveys was generally required to obtain stable estimates, with the most parsimonious model being gender-dependent
irrespective of the periodicity of intervals between surveys.
Therefore, they suggest that tagging programmes should aim to be
longer than the typical 3-year duration of most research projects.
Also in Tasmania, Gardner et al. (2015) used a bioeconomic model
to investigate probable outcomes of increasing the legal minimum
size (LML) of female J. edwardsii across this island state, which
could presumably increase egg production and prevent low levels of
post-larval recruitment such as occurred in recent years. However,
as both the stock and fishery exhibit strong spatial patterns around
Tasmania due to variable abiotic factors, they found that raising the
LML statewide would not improve management of egg production
because it would increase protection of females in southern areas
where egg production is already high, and decrease egg production
in the northern areas due to displacement of effort.
The final article deals with environmental impacts of spiny
lobster farms. Culture from the egg is still not economically viable
for spiny lobsters due to their long larval phase, but these lobsters
grow rapidly. Therefore, in several Asian countries, wild-caught
small juveniles, especially of the ornate spiny lobster, Panulirus
ornatus, are grown into marketable sizes in sea cage farms. The potential environmental impacts of these farms were investigated via a
modelling approach by Lee et al. (2015). The authors concluded that
the levels of deposition of particulate organic carbon and release of
dissolved inorganic nitrogen from typical spiny lobster farms are
low compared with levels from the aquaculture of finfish species,
and advocate locating spiny lobster farms in areas with some
levels of hydrodynamical activity and away from sensitive habitats
such as seagrass and coral reefs.
Conclusions
In the context of global climate change, new gaps in knowledge constantly emerge. This is especially true for shallow-water lobster
species that inhabit coastal zones where increasing temperatures,
in conjunction with other anthropogenic pressures such as overfishing and habitat degradation, are resulting in rapid changes in water
conditions and loss of essential benthic habitats (Caputi et al., 2013).
As shown in papers presented at the 10th ICWL, climate change is
already affecting some lobster species by changing the spatial distribution of the stocks and hence affecting catches and the territorial
behaviour of fishers; by affecting growth rates, sizes at maturity,
the timing of reproductive processes, duration of larval development, and the timing and levels of settlement; by affecting key
habitat-forming species in settlement habitats, especially in temperate zones, and by increasing the risk of disease and impacting the behavioural ecology of lobsters. In addition, there are still many
technological challenges to develop effective aquaculture and enhancement methods that may allow a sustainable supply of lobsters
separate from the harvest of wild populations, as well as big holes in
the basic knowledge of many species—including some quite heavily
exploited—that preclude evaluating the success or failure of those
methods.
On the other hand, research on lobsters has grown exponentially
during the 37 years that have elapsed between the first and the 10th
ICWL. For example, a Web of Science search (excluding patent databases) for articles with “lobster” in the title yielded 980 entries for all
years before 1977 (the year of the first ICWL), but 14 574 entries for
the period 1977–2014. There has been considerable progress in the
understanding of life history strategies and the relationships
between environmental factors and spawning, fecundity, post-larval
recruitment, and distribution of juveniles and adults. The advent of
molecular techniques is broadly expanding the scope of studies on
genetics, phylogeny, connectivity, and trophic ecology of lobsters.
Technological advances are enabling more in-depth studies on
lobster behaviour and interspecific relationships. Improved modelling approaches are being developed that take into account the
biology and behaviour of the exploited species. The accumulation
of biological data and the lengthening of environmental and fisheries time-series are favouring the development of better management
strategies, including co-management, and advances are being
achieved in the investigation of nutrition and culture conditions
of different lobster species. The generation and search for knowledge
will continue as long as there remain lobsters and lobster fisheries.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
We thank our director, Elva Escobar-Briones, for the institutional
support, and the keynote speakers (Bruce Phillips, Richard Wahle,
Ehud Spanier, Jim Penn, and Juan Carlos Seijo), members of the organizing committee, session organizers, volunteer assistants, and all
presenters for a wonderful conference. Our special appreciation to
Bruce Phillips (Curtin University, Perth, Australia) for establishing
the ICWLs back in 1977 and attending all 10 conferences to date, and
for delivering the main keynote presentation titled “Lobsters in a
Changing Climate”. Grant Stentiford suggested the title for this
Introduction. Our sincere thanks to Gabriela Macedo, Fernando
Llano, and the staff of DMC-Cancún for managing the conference
and providing logistic support. We especially thank Fernando
Negrete-Soto and Cecilia Barradas-Ortiz for their invaluable technical support through the conference preparation and during its
development, and Jorge Quiñonez-Alvarado for designing the colourful logo for the 10th ICWL. The 10th ICWL received much
appreciated sponsorships from several sources. We are especially
grateful for the generous financial support provided by the
National Council of Science and Technology (Consejo Nacional
de Ciencia y Tecnologı́a, México) and the Coordination of
Scientific Research-National Autonomous University of Mexico
(Coordinación de la Investigación Cientı́fica—UNAM), which
was instrumental not only for the development of the conference,
but also for the publication of this issue of the ICES Journal of
Marine Science. Substantial support for the conference was also provided by several organizations from Mexico (Instituto Nacional de
Pesca, Chaktunché Art and Crafts, Colectividad Razonatura,
Comisión Nacional de Áreas Naturales Protegidas, Universidad
Marista de Mérida, El Colegio de la Frontera Sur, Mayan Star
Fisheries, Snorkeling Adventure, Integradora de Pescadores de
Quintana Roo, Academia Mexicana de Ciencias, and Asociación
Latinoamericana de Carcinologı́a) and from the USA
(Wiley-Blackwell), South Africa (ORLAC SPS), and Canada
(Clearwater Seafood).
i6
References
Babcock, E. A., Harford, W. J., Coleman, R., Gibson, J., Maaz, J., Foley,
J. R., and Gongora, M. 2015. Bayesian depletion model estimates of
spiny lobster abundance at two marine protected areas in Belize
with or without in-season recruitment. ICES Journal of Marine
Science, 72(Suppl. 1): i232–i243.
Briones-Fourzán, P., Domı́nguez-Gallegos, R., and Lozano-Álvarez, E.
2015. Aggressive behaviour of spotted spiny lobsters (Panulirus guttatus) in different social contexts: The influence of sex, size, and
missing limbs. ICES Journal of Marine Science, 72(Suppl. 1):
i155 – i163.
Butler, C. B., and Matthews, T. R. 2015. Effects of ghost fishing lobster
traps in the Florida Keys. ICES Journal of Marine Science,
72(Suppl. 1): i185 – i198.
Butler, M., IV, Bertelsen, R., and MacDiarmid, A. 2015a. Mate choice in
temperate and tropical spiny lobsters with contrasting reproductive
systems. ICES Journal of Marine Science, 72(Suppl. 1): i101 – i114.
Butler, M. J., IV, MacDiarmid, A., and Gnanalingam, G. 2015b. The
effect of parental size on spermatophore production, egg quality, fertilization success, and larval characteristics in the Caribbean spiny
lobster, Panulirus argus. ICES Journal of Marine Science,
72(Suppl. 1): i115 – i123.
Candia-Zulbarán, R. I., Briones-Fourzán, P., Lozano-Álvarez, E.,
Barradas-Ortiz, C., and Negrete-Soto, F. 2015. Caribbean spiny lobsters equally avoid dead and clinically PaV1-infected conspecifics.
ICES Journal of Marine Science, 72(Suppl. 1): i164 – i169.
Caputi, N., de Lestang, S., Frusher, S., and Wahle, R. A. 2013. The impact
of climate change on exploited lobster stocks. In Lobsters: Biology,
Management, Aquaculture and Fisheries, pp. 84 – 112. Ed. by B. F.
Phillips. Wiley-Blackwell, Oxford.
Chan, T. -Y. 2010. Annotated checklist of the world’s marine lobsters
(Crustacea: Decapoda: Astacidea, Glypheidea, Achelata,
Polychelida). The Raffles Bulletin of Zoology, (Suppl. 23): 153 – 181.
Childress, M. J., Heldt, K. A., and Miller, S. D. 2015. Are juvenile
Caribbean spiny lobsters (Panulirus argus) becoming less social?
ICES Journal of Marine Science, 72(Suppl. 1): i170 – i176.
Cobb, J. S. 2006. The Lobster Conferences: Origin, development, and
impact. Journal of Crustacean Biology, 26: 550– 554.
Davies, C. E., Johnson, A. F., Wootton, E. C., Greenwood, S. J., Clark,
K. F., Vogan, C. L., and Rowley, A. F. 2015. Effects of population
density and body size on disease ecology of the European lobster
in a temperate marine conservation zone. ICES Journal of Marine
Science, 72(Suppl. 1): i128 – i138.
de Lestang, S., Caputi, N., Feng, M., Denham, A., Penn, J., Slawinski, D.,
Pearce, A., et al. 2015. What caused seven consecutive years of low
puerulus settlement in the western rock lobster fishery of Western
Australia? ICES Journal of Marine Science, 72(Suppl. 1): i149 – i158.
Ellis, C. D., Hodgson, D. J., Daniels, C. L., Boothroyd, D. P., Bannister,
R. C. A., and Griffiths, G. F. 2015a. European lobster stocking
requires comprehensive impact assessment to determine fishery
benefits. ICES Journal of Marine Science, 72(Suppl. 1): i35– i48.
Ellis, C. D., Knott, H., Daniels, C. L., Witt, M. J., and Hodgson, D. J.
2015b. Geographic and environmental drivers of fecundity in the
European lobster (Homarus gammarus). ICES Journal of Marine
Science, 72(Suppl. 1): i91– i100.
Ewing, G., and Frusher, S. 2015. New puerulus collector design suitable
for fishery-dependent settlement monitoring. ICES Journal of
Marine Science, 72(Suppl. 1): i225 – i231.
FAO. 2014. FAO yearbook. Fishery and Aquaculture Statistics 2012.
Statistics and Information Branch of the Fisheries and Aquaculture
Department. FAO, Rome, 76 pp.
Gardner, C., Hartmann, K., Punt, A. E., and Hoshino, E. 2015. Fewer
eggs from larger size limits: Counterintuitive outcomes in a spatially
heterogeneous fishery. ICES Journal of Marine Science, 72(Suppl. 1):
i252 – i259.
P. Briones-Fourzán and E. Lozano-Álvarez
Goldstein, J. S., Dubofsky, E. A., and Spanier, E. 2015. Into a rhythm:
Diel activity patterns and behaviour in Mediterranean slipper lobsters, Scyllarides latus. ICES Journal of Marine Science, 72(Suppl.
1): i147 – i154.
Gutzler, B. C., Butler, M. J., IV, and Behringer, D. C. 2015. Casitas: A
location-dependent ecological trap for juvenile Caribbean spiny lobsters, Panulirus argus. ICES Journal of Marine Science, 72(Suppl. 1):
i177 – i184.
Hinojosa, I. A., Green, B. S., Gardner, C., and Jeffs, A. 2015. Settlement
and early survival of southern rock lobster, Jasus edwardsii, under
climate-driven decline of kelp habitats. ICES Journal of Marine
Science, 72(Suppl. 1): i59– i68.
Hobday, A. J., and Pecl, G. T. 2014. Identification of global marine hotspots: Sentinels for change and vanguards for adaptation action.
Reviews in Fish Biology and Fisheries, 24: 415 – 425.
Jeffs, A. G., Gardner, C., and Cockcroft, A. 2013. Jasus and Sagmariasus
species. In Lobsters: Biology, Management, Aquaculture and
Fisheries, pp. 259–288. Ed. by B. F. Phillips. Wiley-Blackwell, Oxford.
Kordjazi, Z., Frusher, S., Buxton, C. D., and Gardner, C. 2015.
Estimating survival of rock lobsters from long-term tagging programmes: How survey number and interval influence estimates.
ICES Journal of Marine Science, 72(Suppl. 1): i244 – i251.
Kough, A. S., Paris, C. B., Behringer, D. C., and Butler, M. J., IV. 2015.
Modelling the spread and connectivity of waterborne marine pathogens: the case of PaV1 in the Caribbean. ICES Journal of Marine
Science, 72(Suppl. 1): i139 – i146.
Lee, S., Hartstein, N. D., and Jeffs, A. 2015. Modelling carbon deposition
and dissolved nitrogen discharge from sea cage aquaculture of tropical spiny lobster. ICES Journal of Marine Science, 72(Suppl. 1):
i260 – i275.
Lizárraga-Cubedo, H. A., Tuck, I., Bailey, N., Pierce, J. C., Zuur, A. F.,
and Bova, D. 2015. Scottish lobster fisheries and environmental
variability. ICES Journal of Marine Science, 72(Suppl. 1): i211 – i224.
O’Rorke, R., Lavery, S. D., Wang, M., Gallego, R., Waite, A. M., Beckley,
L. E., Thompson, P. A., et al. 2015. Phyllosomata associated with
large gelatinous zooplankton: Hitching rides and stealing bites.
ICES Journal of Marine Science, 72(Suppl. 1): i124 – i127.
Penn, J. W., Caputi, N., and de Lestang, S. 2015. A review of lobster
fishery management: The Western Australian fishery for Panulirus
cygnus, a case study in the development and implementation of
input and output based management systems. ICES Journal of
Marine Science, 72(Suppl. 1): i22 – i34.
Phillips, B. F., Melville-Smith, R., Kay, M. C., and Vega-Velázquez, A.
2013. Panulirus species. In Lobsters: Biology, Management,
Aquaculture and Fisheries, pp. 289– 325. Ed. by B. F. Phillips.
Wiley-Blackwell, Oxford.
Quinn, B. K., and Rochette, R. 2015. Potential effect of variation in water
temperature on development time of American lobster larvae. ICES
Journal of Marine Science, 72(Suppl. 1): i79– i90.
Spanier, E., Lavalli, K. L., Goldstein, J. S., Groeneveld, J. C., Jordaan, G.
L., Jones, C. L., Phillips, B. F., et al. 2015. A concise review of lobster
utilization by worldwide human populations from pre-history to the
modern era. ICES Journal of Marine Science, 72(Suppl. 1): i7– i21.
Tuck, I. D., Parsons, D. M., Hartill, B. W., and Chiswell, S. M. 2015.
Scampi (Metanephrops challengeri) emergence patterns and catchability. ICES Journal of Marine Science, 72(Suppl. 1): i199 – i210.
Wahle, R. A., Castro, K. M., Tulle, O., and Cobb, J. S. 2013. Homarus.
In Lobsters: Biology, Management, Aquaculture and Fisheries,
pp. 221 –258. Ed. by B. F. Phillips. Wiley-Blackwell, Oxford.
Wahle, R. A., Dellinger, L., Olszewski, S., and Jekielek, P. 2015. American
lobster nurseries of southern New England receding in the face of
climate change. ICES Journal of Marine Science, 72(Suppl. 1):
i69– i78.
Handling editor: Howard Browman
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i7 –i21. doi:10.1093/icesjms/fsv066
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Reviews
A concise review of lobster utilization by worldwide human
populations from prehistory to the modern era
Ehud Spanier 1, Kari L. Lavalli 2 *, Jason S. Goldstein 1,3, Johan C. Groeneveld 4, Gareth L. Jordaan 4,
Clive M. Jones 5, Bruce F. Phillips 6, Marco L. Bianchini 7, Rebecca D. Kibler 1, David Dı́az 8,9, Sandra Mallol 9,
Raquel Goñi 9, Gro I. van Der Meeren10, Ann-Lisbeth Agnalt 11, Donald C. Behringer 12,13,
William F. Keegan 14, and Andrew Jeffs 15
1
The Leon Recanati Institute for Maritime Studies and Department of Maritime Civilizations, The Leon H. Charney School for Marine Sciences,
University of Haifa, Haifa, Israel
2
Division of Natural Sciences and Mathematics, College of General Studies, Boston University, Boston, MA, USA
3
Department of Biological Sciences, University of New Hampshire, Durham, NH, USA
4
Oceanographic Research Institute, PO Box 10712, Marine Parade, Durban 4056, South Africa
5
James Cook University, School of Marine and Tropical Biology, Cairns, Queensland, Australia
6
Curtin University, Environment and Agriculture, Perth, Western Australia, Australia
7
Istituto Ambiente Marino e Costiero (IAMC-CNR), Italian National Research Council, 91026 Mazara del Vallo TP, Italy
8
Departament de Biologia Marina i Oceanografia, Institut de Ciències del Mar—CSIC, Passeig Marı́tim de la Barceloneta, 37-49, 08003 Barcelona, Spain
9
Instituto Español de Oceanografı́a—Centro Oceanográfico de Baleares, Muelle de Poniente, s/n Apdo 291. 07015 Palma de Mallorca, Spain
10
Institute of Marine Research, Austevoll Research Station, NO-5392 Storebø, Norway
11
Institute of Marine Research, PB 1870 Nordnes, NO-5817 Bergen, Norway
12
School of Forest Resources and Conservation—Program in Fisheries and Aquatic Sciences, Gainesville, FL, USA
13
Emerging Pathogens Institute, University of Florida, Gainesville, FL, USA
14
Florida Museum of Natural History, University of Florida, Gainesville, FL, USA
15
Leigh Marine Laboratory, Institute of Marine Science, University of Auckland, Warkworth 0985, New Zealand
*Corresponding author: tel: +1 774 287 7550; fax: +1 617 353 5868; e-mail: [email protected].
Spanier, E., Lavalli, K. L., Goldstein, J. S., Groeneveld, J. C., Jordaan, G. L., Jones, Clive M., Phillips, B. F., Bianchini, M. L., Kibler, R. D., Dı́az,
D., Mallol, S., Goñi, R., van Der Meeren, G. I., Agnalt, A-L., Behringer, D. C., Keegan, William F., and Jeffs, A. A concise review of lobster
utilization by worldwide human populations from prehistory to the modern era. – ICES Journal of Marine Science, 72: i7– i21.
Received 9 November 2014; revised 18 March 2015; accepted 25 March 2015; advance access publication 7 May 2015
.
Lobsters are important resources throughout the world’s oceans, providing food security, employment, and a trading commodity. Whereas marine
biologists generally focus on modern impacts of fisheries, here we explore the deep history of lobster exploitation by prehistorical humans and
ancient civilizations, through the first half of the 20th century. Evidence of lobster use comprises midden remains, artwork, artefacts, writings
about lobsters, and written sources describing the fishing practices of indigenous peoples. Evidence from archaeological dig sites is potentially
biased because lobster shells are relatively thin and easily degraded in most midden soils; in some cases, they may have been used as fertilizer
for crops instead of being dumped in middens. Lobsters were a valuable food and economic resource for early coastal peoples, and ancient
Greek and Roman Mediterranean civilizations amassed considerable knowledge of their biology and fisheries. Before European contact, lobsters
were utilized by indigenous societies in the Americas, southern Africa, Australia, and New Zealand at seemingly sustainable levels, even while other
fish and molluscan species may have been overfished. All written records suggest that coastal lobster populations were dense, even in the presence
of abundant and large groundfish predators, and that lobsters were much larger than at present. Lobsters gained a reputation as “food for the poor”
in 17th and 18th century Europe and parts of North America, but became a fashionable seafood commodity during the mid-19th century.
# International
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i8
E. Spanier et al.
High demand led to intensified fishing effort with improved fishing gear and boats, and advances in preservation and long-distance transport. By the
early 20th century, coastal stocks were overfished in many places and average lobster size was significantly reduced. With overfishing came attempts
to regulate fisheries, which have varied over time and have met with limited success.
Keywords: Africa, Americas, ancient era, Australia, Europe, fisheries, human utilization, lobsters, Mediterranean, middle ages, modern era,
New Zealand, prehistory, zoo-archaeology.
Introduction
Lobsters have been a part of the world’s seascape for millions of years
with their origins (likely Polychelida) appearing in the Devonian
[409–372 million years ago (MYA)]. Early forms were probably
deep water species, much as living polychelids are today
(Bracken-Grissom et al., 2014). The Achelata appeared 391–351
MYA, but did not diverge into the palinurid and scyllarid lineages
until the Permian (250 MYA) (Patek et al., 2006; Bracken-Grissom
et al., 2014). Palinurids underwent a rapid diversification in the
Upper Jurassic (200 –160 MYA), which coincided with the
breakup of Pangea and formation of many new habitats (George
and Main, 1967; Patek and Oakley, 2003; Bracken-Grissom et al.,
2014). Fossil remains of scyllarids date back to the midCretaceous (100–120 MYA; Glaessner, 1969; Webber and Booth,
2007) and those of nephropids to the Lower Cretaceous (Wahle
et al., 2012). Nephropids may have been more common in shallow
shelf waters during the Cretaceous than at present, because of the
subsequent invasion of these environments by predatory fish and
competing brachyuran crabs (Vermeij, 1977; Wahle et al., 2012).
Most lobster stocks are currently exploited by humans, some to
the point of overexploitation, and management schemes often
seek to restore natural populations and their ecosystems. Therefore,
understanding the history of human impacts on lobster stocks
might provide valuable information to guide restoration and
fishery practices. Erlandson and Rick (2010) suggested that modern
views of ancient human impacts on marine fauna are biased by the
belief that humans were not populous enough, nor technologically
advanced enough to significantly deplete exploited species. Yet
more recent case studies from around the world (reviewed in
Erlandson and Rick, 2010) demonstrate that humans modified prehistorical marine ecosystems by removing keystone predators, reducing the size of finfish and shellfish, and causing local extinctions of
seabirds and marine mammals. This review seeks to reconstruct
how humans came to exploit lobsters as they dispersed throughout
the world and tries, when possible, to point out areas where fishing
practices were seemingly sustainable and where overexploitation
happened early on, before more modern fishing techniques. It
follows the chronological migratory patterns of humans into specific
regions to discuss the practices of the indigenous peoples and then
covers how European colonization altered traditional fishing
methods, the local seascapes, and, ultimately, lobster populations.
The time line is divided into sections for prehistoric, ancient,
middle age, and up to the introduction of modern fisheries. This is
done by examining archeological data in the form of midden findings,
ancient artwork and rituals, oral traditions, and written records.
A brief review of human dispersal
Current dispersal models based on dated remains suggest that
archaic, anatomically modern humans (AMHs) left the African
Rift Valley in several pulses and migrated west- and southward to
the coasts of Africa, and then eastward to the Levant (Armitage
et al., 2011; Figure 1). From there, migrants continued eastward
along the Indian coast to colonize various Asian Pacific islands,
Australia, and the Chinese coast. Others migrated inland to colonize
Figure 1. Migratory pattern of AMHs (arrows with dates based on DNA evidence) overlaid with coastal species of lobster distribution (dark grey
area, present-day depth distribution of 10 m or less of commercial species of lobster according to Holthuis, 1991). Stars indicate areas where we have
evidence of use of lobster by indigenous peoples.
i9
Historical use of lobsters by humans worldwide
the interior regions of China and Siberia. Migrating Levantine
peoples also hugged the Mediterranean coastline to colonize modernday Turkey, Romania, and southern Europe. The Americas were only
recently colonized when ice sheets extending over much of Canada
receded. Peoples from Siberia and Asia moved across Beringia,
with some moving along the northwest coast of North America
to Central and South America, and others spreading over the
North American plains (Eriksson et al., 2012). Movements in
South America followed the Pacific coastline, until reaching the
Patagonian glaciers, whereupon migrants crossed the Andes and
continued southward along the Atlantic coast (de Sainte Pierre
et al., 2012).
Coastlines at the time of the first migrations out of Africa were
more extensive than today because of the last Ice Age that ended
18 000 years ago. Migrants using land-bridges to move from continent to continent or islands and those using coastlines as pathways
would have used marine resources as a relatively accessible food
source, if fishing techniques and/or swimming had been mastered.
Evidence of fishing technology used by Homo erectus has been found
in Zaire (Yellen et al., 1995) and Southeast Asia (Morwood et al.,
1998). More evidence comes from Middle Stone Age sites in
South Africa (Henshilwood et al., 2004; Marean et al., 2007),
Neanderthal sites off Gibraltar and the Mediterranean (125 000–
30 000 years ago; Garrod et al., 1928; Stringer et al., 2008), and
Australia (40 000 –30 000 years ago; Jones, 1992). Boating technology was more advanced than previously thought since pelagic
fish were exploited by some Pacific islanders 40 000 years ago
(O’Connor et al., 2011). Based on midden analysis, fish, molluscs,
sea turtles, and marine mammal (dugongs, seals) were seemingly
more commonly exploited than lobsters (Klein et al., 2004).
However, the absence of lobster remains in middens may be due
not to lack of use, but rather to differences in preservation of
lobster tissue and/or differences in how lobster remains were used
compared with fish, bird, turtle, mollusc, or marine mammal
remains.
Sampling issues
Many extant spiny (Palinuridae), clawed (Nephropidae), and
slipper (Scyllaridae) lobsters overlap the migratory pathways
taken by AMHs (Figure 1). However, determining the diets of migratory hunter-gatherers is difficult, and evidence of shellfish consumption (incl. lobsters) dates only from the development of
relatively permanent settlements with midden heaps (“kitchen
remains”). Fish bones and mollusc shells are common in prehistorical and historical middens and have often been identified and
analysed to species level. Taxa with more fragile shells, such as echinoderms and crustaceans, have been largely ignored in most of the
zoo-archaeological literature (e.g. Zugasti, 2011). One reason for
this is the composition of the relatively thin exoskeletons (compared
with thicker mollusc shells), which degrade and dissolve over time,
leaving no recognizable remains in archaeological digs. Carbonic
acid, formed by the interaction of rainwater and CO2, removes all
but the thickest shell fragments, leaving behind only parts of
lobster claws, mandibles, and possibly gastric mill teeth. Middens
in dry sites, such as caves protected from rain, may retain fragments
for thousands of years, although it is sometimes difficult to separate
lobster from crab fragments without DNA analysis. Lobster mandibles also appear to be well preserved in alkaline soils, such as those
found in ash layers and shell middens (Leach and Boocock, 1993).
Because of rapid decomposition in non-optimal soils, lobster
remains will be rare in typical middens; however, an additional
Figure 2. Example of calcified lobster mouthpart (opposed left and
right grinding plates). These are one of the few elements of the
exoskeleton that survive in the archaeological record.
bias can occur during screening when objects are removed from
the sieved matrix. If objects are not recognized as potentially identifiable, fragments may be discarded, and thus shell bits may fail
to be set aside for further examination or even considered relevant.
Leach and Boocock (1993) argue that lack of knowledge on the part
of archaeologists has led to the discard of lobster shell fragments at
many sites in the past. In addition, the remains may differ among
lobster families. For example, slipper lobsters (Scyllaridae) have
thicker exoskeleton than spiny lobsters (Palinuridae) and especially
clawed lobsters (Nephropidae; e.g. Barshaw et al., 2003; Tarsitano
et al., 2006), but they lack heavily calcified mandibles of palinurids
or the claws of nephropids. Thus, evidence of their use may
be difficult to assess from midden samples alone. Remains of
lobster mandibles in middens can date to over 11 000 years ago
(Jerardino and Navarro, 2002; Figure 2), which is far less than that
of fish bones that can date up to 75 000 years ago (Yellen et al.,
1995). Also, the assumption that lobster remains should be found
only in middens is not necessarily correct, especially for indigenous
peoples who practiced agriculture. In those cultures, lobster exoskeletons may have been spread over fields as fertilizer, in which case
they would have completely degraded. Furthermore, some indigenous peoples had strictures of how to treat the remains of animals
from whence they obtained food, and often for marine species,
this meant returning the remains to the sea (Maillard, 1863).
Hence, it remains unclear whether the scarcity (or absence) of lobsters in some archaeological records is caused by their inaccessibility
to ancient fishers, or avoidance by ancient peoples, or decomposition of exoskeletons, or sampling bias. This means that for some
parts of the world, evidence of use may focus more strongly on
written and oral records and/or artwork.
Use of lobsters worldwide
Southern Africa
The Cape west coast is influenced by the northwards flowing cooltemperate Benguella Current and nutrient-rich upwelling systems,
which support abundant marine life. Rocky shores and shallow subtidal kelp forests shelter an abundance of shellfish, including dense
mussel beds and the endemic Cape rock lobster, Jasus lalandii. These
i10
E. Spanier et al.
shellfish occur in inter- and shallow subtidal habitats, where they
can easily be collected by hand, without the use of specialized tools.
Islanders (from 3000 years ago), and other groups that settled
after 1788, when the first European colony was established.
Prehistorical era
Prehistorical and historical eras
Early human use of marine resources in South Africa date from
165 000 years ago, during the Middle Pleistocene (Marean et al.,
2007). Middle Stone Age sites along the Cape west coast also
confirm that subsistence harvesting of coastal resources is very
ancient, based on well-developed shell middens dated to 70 000
years ago. Preserved calcareous mandibles in middens confirm
that lobsters were exploited by early hunter-gatherers over the past
15 000 years (Jerardino et al., 2008). Mandible size has been used
to estimate lobster carapace length and past exploitation rates
(Jerardino et al., 2001, 2008; Jerardino and Navarro, 2002;
Jerardino, 2010a). Middens show increasing human settlement in
west coast cave sites and shelters that gathered deposits between
5000 and 3000 years ago (Jerardino, 2010b). A megamidden
period (large shell middens; many marine shells; low densities of
artefacts) between 3000 and 2000 years ago show an intensified reliance on marine resources, including lobster, and a full range of
settlement, subsistence, and new cultural, social, and economic
developments (Jerardino, 2010a, b). Human habitation in this
area decreased between 2000 and 500 years ago, thus concluding
the Late Stone Age lifestyle before the arrival of European settlers
(Jerardino et al., 2009).
Aboriginals possessed no literacy, and their history was communicated to successive generations through stories, art, and dance.
They lived with a strong dependence on the land and water,
within groups or tribes that developed skills (usually hunting,
fishing, or gathering) for their specific environment. Most
Aboriginals inhabited southern and eastern Australia, particularly
along the larger permanent rivers, and fewer lived along the coast
or gained sustenance from it. Archaeological and ethnographic
records indicate that lobsters formed part of the diet of coastal
Aboriginals (Anonymous, 2004). The few references that mention
lobster use or European paintings depicting fishing by Aborigines
suggest that fishing was typically conducted along the coastal
fringe from rock platforms or shallow rocky reefs, by diving or
using small canoes.
Aboriginal middens older than 4000 years in Tasmania (Flood,
1995) contained species that could be collected along rocky shores
or by wading. Substantial consumption of subtidal shellfish and
crustaceans (including lobsters) began when people started to dive
and swim, around 3500– 3000 years ago. The most significant
animals to Aboriginal people were often depicted in paintings rendered on bark or rock surfaces, but there are few records of lobsters
being depicted in such paintings. Only Sonkkila (2013) mentions
rock art in northwestern Australia that includes images of crayfish
(sic) among many other animals.
The inhabitants of Torres Strait Islands, located between Cape
York, in northern Queensland, and Papua New Guinea, are a “sea
people”, without a recorded history until European contact.
Historical accounts suggest that lobsters formed an integral part
of their diet (primarily seafood), equal to other marine species.
McNiven et al. (2009) recorded evidence of lobster (possibly
Panulirus ornatus) in rock art on Dauan Island, but such paintings
are rare, known from only two sites. Europeans settled at Cape
York in 1863. Trade with native Torres Strait Islanders was mostly
for pearls and beche-de-mer, with lobsters being less important
(Loban, 2007).
Introduction of modern fisheries
The Cape was first colonized by Dutch settlers in 1652, bringing
them into contact with the local Khoi-San, and “strandlopers”
(beach-wanderers) or “strand bosjemans” (beach bushmen;
Brink, 2004). Pappe (1853) reported that rock lobster “was easily
caught in vast numbers all the year round” and that they had a proclivity for massing at the sewer outfalls in Table Bay. Colonists considered lobster to be a “regular godsend” to misers and the poor
classes of the community (Melville-Smith and Van Sittert, 2005),
and this negative connotation persisted until lobster meat became
popular with the bourgeois in Europe. Commercial exploitation
of J. lalandii began in 1875, when a processing plant was established
in Cape Town to can lobsters for export to Europe. Weights of
canned, frozen, tailed, and live lobster exported from South Africa
have been reported since 1891, leaving a record of most commercial
landings for over a century (Melville-Smith and Van Sittert, 2005).
Growth and modernization of the fishing fleet and processing plants
during and after World War II led to peak landings of 17
000 t year21 in the early 1950s. Since then commercial landings
have declined to between 10 and 15% of what they were at the
height of the fishery. The biomass of lobsters above the legal
minimum size is currently at 3.5% of pristine levels (DAFF, 2012).
Australia
Because Australia is bounded by the Indian, Pacific, and Southern
oceans, its oceanography is quite complex. Currents running
along the west and east coasts bring warm, nutrient-poor waters
that flow south; these currents mean that the west and east coasts
of Australia have a warm climate with coastal waters that support
tropical species at what would normally be considered a temperate
latitude. Hence, coral reefs are well supported along both coasts,
bringing with them a rich diversity of species, including a variety
of lobsters: at least 11 palinurids, 14 scyllarids, and 8 nephropids
(Holthuis, 1991). The human population of around 23 million
comprises Aboriginals (from 40 000 years ago), Torres Strait
Introduction of modern fisheries
Lobster fishing in Australia developed as an industry during the late
1800s, simultaneously in southwestern Australia for Panulirus
cygnus, South Australia, Victoria, and Tasmania for Jasus edwardsii,
and New South Wales for Sagmariasus verreauxi. Regulations accompanied the earliest developments of lobster fishing, and although they varied among the different jurisdictions and species,
they generally included size limits, prohibition against taking ovigerous females, and some spatial and/or temporal closures. For recreational fishers, the regulations focused on gear and bag limits per
person. From the late 1880s through to the mid-1900s, the lobster
fisheries remained relatively small and non-mechanised, but after
World War II, the broad adoption of diesel engines and development of refrigeration allowed for expanded local and foreign
markets. Consequently, fishing effort and catches increased steeply,
necessitating fisheries management controls such as individual
transferable quotas (Phillips, 2013).
Mediterranean basin
The Mediterranean is mostly landlocked with only a narrow connection to the Atlantic Ocean. The cold, low salinity water from the
Historical use of lobsters by humans worldwide
Atlantic flows eastward and is warmed along the way. Evaporation in
this region exceeds precipitation, so the surface water increases in
salinity as it moves and eventually sinks when it reaches the
Levant and then travels westward at depth, exiting back into the
Atlantic at the bottom of the Strait of Gibraltar. Continental
shelves in the basin are narrow and the basin is thus mostly deep
sea and nutrient poor compared with the Atlantic. Nevertheless, it
supports a high diversity of species, many of which are endemic to
this sea. While the Mediterranean today is connected to the Red
Sea via the Suez Canal, these seas were separated for most of
human history. The Red Sea is an inlet of the Indian Ocean and is
extremely salty and warm, being situated between desert and semidesert land with high evaporation rates. Water exchange occurs
between the Indian Ocean and Arabian Sea which helps to ameliorate the effects of high salinity and allows for extensive reef development. Partly due to the species richness of the Mediterranean and
Red Seas, several ancient coastal and maritime civilizations originated along their coasts. Edible lobster species important to inhabitants along these coasts included at least two Mediterranean clawed
lobsters, five spiny lobsters (two in the Mediterranean; three in the
Red Sea), and five slipper lobsters (three in the Mediterranean;
two in the Red Sea; Holthuis, 1991).
Prehistorical and ancient eras
The earliest depiction of a lobster is from a mural on a temple wall in
Deir el-Bahari, Egypt, commemorating the trade voyage of Egyptian
Queen Hatshepsut to the southern Red Sea 3500 years ago (e.g.
Glenister, 2008). It appears to be a spiny lobster (possibly
Panulirus pencillatus) carved together with other Red Sea marine
animals. Recently, only a partial claw (chela) of Homarus gammarus
(Figure 3) was found among other kitchen refuse (mostly marine),
dating 2700 years ago (early Iron Age) in northern Sardinia
(B. Wilkens, pers. comm.). The considerable size of the crushing
teeth on this claw indicates a sizeable lobster. Perhaps large
European lobsters were common in shallow Mediterranean waters
during ancient times, and were exploited by coastal peoples.
Patchy archaeological and historical information of lobster utilization by Mediterranean and Red Seas civilizations indicate a
Figure 3. The end of a claw of H. gammarus found in an excavation of a
seventh century BCE nuragic village of Sant’Imbenia near Alghero,
northern Sardinia. Unearthed by Stefano Masala (under the
supervision of Dr Barbara Wilkens, Department of Environmental
Sciences, University of Sassari, Italy). Used by special permission of the
copyright owners and the Soprintendenza ai Beni Archeologici di
Sassari e Nuoro, Italian Ministry for the Cultural Heritage.
i11
variable attitude towards lobster consumption, ranging from complete religious prohibition for Jews (and Christians) to epicurean
status in the Hellenistic-Roman world. Consequently, lobsters are
either completely absent from artwork and literary expression, or
they appear quite prominently. While rare, lobsters were depicted
on Roman coins, rings, terracotta vessels, tableware, wall drawings,
and reliefs. However, the bulk of the information on Mediterranean
lobsters comes from written text and mosaics (Charmantier, 2014).
Decorative mosaics contribute to understanding the role of
marine organisms to humans in ancient times. Mosaics are often
not accurate enough to allow for identification to species level, and
in some cases, only parts of animals are depicted. A good example
is a 2300-year-old illustration of Pontus (“Oceanus”, a preOlympian sea-god) on Roman mosaic from Neapolice, Tunisia,
that shows antennae and claws (presumably of a lobster) emerging
from his forehead (Fradier, 1982). Another mosaic from the same
period, the “Unswept Floor”, from Pergamon depicts legs and an
antenna of a spiny lobster (Orange and Nordhagen, 1966). The accuracy of images increased during the Hellenistic period, such that a
lobster appearing in a Roman mosaic from Thugga (Dougga of
today) in the “House of Dionysus and Ulysses” (Poinssot, 1965)
some 600 years later (Figure 4) is easily identified as Palinurus
elephas. Other mosaics where lobsters can be recognized are associated with representations of fishing activities.
Greeks and Romans were knowledgeable about the biology of
Mediterranean lobsters and also understood fishing techniques.
Hippocrates writing around 2460–2370 years ago distinguished
among clawed, spiny, and slipper lobster forms (Lloyd and
Chadwick, 1984). From 2400 to 1800 years ago, Aristotle (d’ArcyThompson, 1952), Pliny the Elder (Rakham, 1962), and Aelian
(Scholfield, 1958) classified lobsters as “animals without blood”,
and supplied detailed information about their morphology, recognizing them as decapods. They provided detailed descriptions of the
claws (the “astakós”), spines, thorny antennae (the “káravos”), tail,
eyes, colors, and other external features. Classical naturalists such as
Aristotle and Pliny the Elder described the habitats, behaviour, predators, territoriality, homing, and movements of several lobster
“types” from shallow to deeper waters in winter and summer, respectively. Aspects of lobster reproduction, life cycle, moulting,
and physiology were also known. Aristotle even described tiny
“káravos” that were “smaller than a finger” (possibly the first
benthic stage), and with Pliny the Elder and Aelian detailed the backward escape swimming (tail-flip) of lobsters when in danger.
Plutarch noted that octopus prey on lobsters some 1900 years ago
(Babbit, 1957). Interactions among the common octopus (Octopus
vulgaris), a predator of lobsters, the Mediterranean moray eel
(Muraena helena), a predator of octopus, and lobsters were not
only documented by the above naturalists, but were depicted in
ancient Roman mosaics. A notable mosaic from Pompeii from
1900 years ago (Ward-Perkins and Claridge, 1978; Figure 5)
shows an octopus attacking P. elephas while being threatened by a
moray eel.
Roman and Greek fishers knew the habitats, seasons, baits, and
methods to catch lobsters. The Greek poet Theocritus remarked
2400 years ago that lobsters were a superb food that could be
caught with pots (traps) made of weaved reed (Holden, 1974),
and pots appeared in Roman mosaics of fishing activities (e.g.
Fradier, 1982). Similar weaved pots were used in Norway well into
the modern era, and are known from Africa, Brazil, and Sardinia.
The Romans were the first to transfer preserved fish far inland,
and they built “cetariae ”, to hold live lobsters, that were similar to
i12
E. Spanier et al.
Figure 4. The common spiny lobster, P. elephas, in a third century CE Roman mosaic known as “Ulysses and the Sirens” from the “House of Dionysus
and Ulysses” in Thugga, Tunisia. Photo by Dennis Jarvis, used with permission.
today’s holding facilities (André, 1961). “Cetariae” were built close
to the sea to take advantage of natural coastal ponds fed by waves.
Wealthy Roman city dwellers even had their own facilities.
Lobsters appeared in classical comedies, tragedies, and poems of
Greek and Roman writers, and were described as an excellent food
by the dramatist Epicharmus (Rodrı́guez-Noriega Guillén, 1996).
They occur in recipes in cook books, including the Deipnosophistae
(“Dinner-Table Philosophers”) by Athenaeus written between
1700 and 1800 years ago (Gulick, 1957). Lobsters were also
thought to possess various medicinal properties (King, 2011) inclusive of aphrodisiacs (e.g. Pajovic et al., 2012, and references therein).
Middle ages
Few artefacts show lobster use in the Byzantine period, primarily
because of the decline of non-religious art. One pavement mosaic
from a Byzantine church dating to 1400 years ago from Bet
Guvrin, southern Israel (Ovadiah and Ovadiah, 1987), depicts two
clawed lobsters, possibly H. gammarus, but Israel lies outside the
present distribution range of any clawed lobster species (Holthuis,
1991). This discrepancy may be explained by H. gammarus extending
into the southeastern Mediterranean during the Little Ice Age (2000–
500 years ago), or the artist may have been from the western
Mediterranean where this species was common or may have copied
a model while visiting there.
In the Middle Ages, mosaic representations were replaced by
paintings, some of which accurately depicted lobsters (e.g.
Charmantier, 2014). The popularity of lobsters continued into
the Renaissance and early modern era, as expressed in many paintings illustrating fish markets and luxurious table settings (see Table
1 in Charmantier, 2014) as well as in the literature (e.g. Crane,
2009).
lobster fisheries in the Mediterranean started relatively late in the
1900s, but at-sea enforcement was difficult with the large number
of vessels in the basin. Non-uniform regulations across the basin
combined with poor enforcement have been cited for severe
overfishing.
Europe
The North Atlantic current extends the Gulf Stream towards the
northwest coast of Europe and brings warm surface water to that
region, resulting in a highly productive marine ecosystem with
high biodiversity. While several lobster species occur in European
waters, the two main species of interest to humans are H. gammarus
from coastal and shallow rocky habitats along the coasts of Norway
south to the Azores and Morocco (and historically in parts of the
Mediterranean and Black Sea), and P. elephas from rocky and coralligenous habitats of the shallow and continental shelf from Norway
to Morocco and in the west-central Mediterranean. These two
species have been exploited by humans for centuries, but extensive
histories are only available for some regions covered herein.
Norway
Prehistorical era
Remnants of Neolithic living dating from 10 000 to 12 000 years ago
are scattered along the Norwegian coast (A.K. Hufthammer, Bergen
Zool. Museum, pers. comm.). Marine invertebrate remains in dry
caves consist mostly of mollusc shells, with some decapod claw
fragments, but it is unclear whether they represent middens from
human activities or caching by animals. Stone carvings display
fish and whales, but no lobsters are present in art or have been identified in middens.
Middle ages
Introduction of modern fisheries
While the yields of Mediterranean lobster declined in Roman times,
they accelerated over the last two centuries, with the introduction of
modern technologies and effort intensification. Fishing techniques
changed from the use of pots or diving to more efficient trammelnets, resulting in rapid stock declines during the last century
(Goñi and Latrouite, 2005; Groeneveld et al., 2013). Regulation of
Norse sagas and songs from 700 to 800 years ago refer to lobsters,
and in medieval times, the Norse word “humarr”, today recognized
as “hummer”, was acknowledged as a poetic metaphor (“kenning”)
for the sea/seabed (Sturlason 1222–1231a, b; referred to in a list of
sagas, e.g. Herrick, 1909 and references therein). Although the use of
lobsters was not mentioned in literature at the time, the word
“hummer” (which means to move or row backwards) is the root
i13
Historical use of lobsters by humans worldwide
Figure 5. Part of a Roman mosaic from Pompeii (house n. 16, insula 2, Regio VIII; first century CE) that described predator– prey interactions
between an octopus, a spiny lobster, and a moray eel (Soprintendenza Speciale per i Beni Archeologici di Napoli e Pompei).
of European lobster names in 20 countries, from Scandinavia,
throughout northern Europe, and as far as Russia and Armenia.
The widespread use of the name “hummer” suggests that lobsters
may have been traded among countries, since shared names
usually imply shared goods (E. Ellingsve, NTNU, pers. comm.).
Such trade does not exist in modern times, but the sharing of the expression may indicate an ancient trade, perhaps with lobsters sold
and transported as frozen goods from Scandinavian middle age
winter markets (Magnus, 1555).
In his Carta Marina map of northern Europe (1539) (Figure 6a
and b), giant lobster-like monsters catch a sailor and fight a
“rhinoceros”. These lobster illustrations were later copied by other
naturalists (Gesner, 1558). Magnus (1555) also mentions lobsters
in his descriptions of the natural history and social structure of
Scandinavia. Lobsters (carabii) are here connected to octopuses
and moray eels, as depicted in Greek and Roman mosaics. Since
moray eels and large octopuses are not present in Scandinavia,
these observations may have been adopted from Pliny the Elder,
whom Magnus mentions. However, Magnus (1555) describes
cephalopods (Di Polypis) in a chapter illustrated by man-hunting
lobster-like creatures, so it is unclear if his illustrations actually represent lobster or octopus. The first actual mention of lobsters
and crabs for human consumption was made by Clausson Friis
(1599, published by Storm, 1881) who described how lobsters
could be caught in the intertidal zone, how cooking or drying
changed their colour, and that they avoided traps containing live
crabs.
Introduction of modern fisheries
During the Enlightenment in Europe (1650 –1750), wealthy and
royal classes could afford lobsters, as depicted in still life paintings
(details in Charmantier, 2014). In Norway, lobsters were not commonly accepted as food by privileged classes; instead, “heaps of
oyster and lobster shells laying outside their shacks” showed the
popularity (or the necessity of consumption) among the poor
(Pontoppidan, 1752–1753). Lobsters and their shells were also
used for fertilizer, and not served to farmhands and servants more
than thrice weekly up until the last century (E. Nøstvold, Kvitsøy,
pers. comm.).
While not popular with Norwegians, the privileged classes of
England and the Netherlands created a strong demand for lobsters
400 – 500 years ago. Dutch traders established a trade for
Norwegian lobsters around 1650, but were eventually replaced by
a British trading monopoly. By 1700, the export industry expanded
from a few southwestern fishing communities to cover most of
southern Norway. Local reports on social and biological trends
relevant to lobster exploitation and export exist since the late 1600s
and trade is described in contemporary books on Norwegian
natural resources and social life in the 1700s, including that by
Pontoppidan (1752 –1753).
i14
E. Spanier et al.
Figure 6. (a) The Olaus Magnus Carta Marina (1539 version) and the (b) and (c) “lobsters”, which may or may not be actual lobsters, but
misrepresentation of cephalopods. High-resolution copy kindly provided by the Norwegian National Library, Oslo.
As lobster export became more lucrative in Norway, landowners
extended their claims into the sea, attempting to privatize access
rights (Sunde, 1991, and references therein). Strong disputes overfishing rights arose, culminating in a royal decree on 23 April
1728, stating that lobsters and all other marine organisms, except
salmon, were common property. Increased exploitation since
1650s rapidly led to reductions in local lobster stocks, which subsequently fluctuated as a result of the wars in Europe, climate changes,
market demands, and technological developments. Fishing regulations were first introduced in 1849 with little effect. At present, the
Norwegian lobster stock is at a record low, showing little or no signs
of recovery (Kleiven et al., 2012).
Spain
The first spiny lobster fishing grounds exploited by Spanish boats for
export markets were along the northern Iberian coast, in the late 1800s.
Live P. elephas and P. mauritanicus were transported in boats with seawater tanks to cities in France and the UK, where profits from their
sales were four times higher than if sold domestically (Montesinos,
1905). Spanish vessels also caught clawed lobster, H. gammarus, in
the Atlantic and Mediterranean Sea, but not as a target species.
When yields of spiny lobster from northern Spain began to
decline, fleets moved to fishing grounds in West Africa. The
evidence of overexploitation noticed by then triggered the implementation of seasonal restrictions and a special fishing tax.
Despite these measures, overfishing worsened in the 1900s with dramatic changes in the mean lobster weight declining from 2 kg per
lobster at the start of the century to 0.6 kg today.
Traditional fishing gear consisted of cotton nets until Basque
fishers introduced lobster pots (Montesinos, 1905). Both gears
were used simultaneously until the late 1980s, when pots were virtually abandoned for lack of effectiveness (Goñi and Latrouite, 2005).
Targeted lobster fisheries are now restricted to archipelagos and
islands in the Western Mediterranean, such as the Balearic Islands in
Spain, where trammel-nets are used. Here, artisanal fishers have
exploited lobsters for over 150 years, and trade with the Iberian
Peninsula dates back to the mid-1800s. Official statistics show
much lower landings at present compared with 100 years ago.
The Americas
Pacific coast of North America
The Pacific coast of North America supports the spiny lobster
Panulirus interruptus only at its southern-most end, from San Luis
Obispo Bay, California, through the length of the Baja Peninsula
(Holthuis, 1991). Paleo-Indians occupied this region around
13 000 years ago and expanded to the Northern Channel Islands
i15
Historical use of lobsters by humans worldwide
across a 7 –20 km wide strait, and to the more dispersed and distant
Southern Channel Islands further from the mainland when sea levels
were lower than at present (Inman, 1983; Rick et al., 2005). The
Channel Islands and Santa Barbara Channel lie at the confluence
of two major ocean currents (the cool California Current and
warm Southern California Countercurrent) in a region of persistent
upwelling that supports a rich diversity of habitats composed of kelp
forests, rocky reefs, and eelgrass beds.
Prehistorical and historical eras
Coastal middens from the Early Holocene demonstrate that
Paleo-Indians harvested a variety of marine organisms, including
red abalone, giant chitons, mussels, turban snails, crabs, geese, cormorants, albatross, pinnipeds, sea otter, and a variety of finfish
(Erlandson et al., 2011). Active searches in well-preserved ancient
middens found no lobster remains (J.M. Erlandson, pers. comm.).
Middle Holocene (around 6000 years ago) middens contain a
more diverse mixture of marine taxa (Rick et al., 2005) from indigenous Chumash Indians (northern islands) and Tongva peoples
(southern islands; Jones and Klar, 2005; Rick et al., 2005). The
Chumash harvested shellfish so intensively that middens illustrate
declines in shell sizes of some species, suggestive of overfishing.
These anthropogenic effects seem focused on abalone, but
Spanish explorers in the 1700s remarked that lobsters were
common in Indian villages (Fray Juan Crespi, 1769–1774; reported
in McArdle, 2008), abundant in coastal waters (Sebastian-Vizcaino,
1602; reported in McArdle, 2008), and fished with devices (pots)
similar to Spanish ones (Hudson and Blackburn, 1982). Less information is available for the Tongva, so it remains unclear if they used
lobsters.
Introduction of modern fisheries
Chinese and Caucasian fishers replaced the island Chumash during
the 1800s and fished for lobsters off Santa Barbara and Santa Cruz
Island. They delivered dried tails to San Francisco’s Chinatown for
export to China (Rathbun, 1887; McArdle, 2008), and live lobsters
for local use. Rathbun (1887) reported that lobsters were so abundant in the Santa Barbara Channel that a single person could
capture around 230 kg in the space of 2 h, from the nearshore
with dipnets. Spiny lobster became popular as a substitute for dwindling eastern clawed lobster, and canneries sprung up to provide a
product that could withstand long-term storage and be shipped
worldwide. Canneries used their own engine-powered ships to
move lobster fishers between islands and along the coast, and transport catches. Industrialization increased fishing effort (both in
terms of fishers and traps deployed) and landings and gradually
reduced the size and weight of landed lobsters. By the early 1900s,
lobster stocks were considered to be overfished and regulations
were imposed. The regulations have resulted in a relatively stable
catch in modern times, but at the cost of increased effort per
fisher, and in a small size of fishable and reproductive lobster
(McArdle, 2008).
Atlantic coast of North America
The northeastern American coast lies at the junction of the
nutrient-rich, warm northward flowing Gulf Stream and the
oxygen-rich, cold southward flowing Labrador Current. An abundance of fish and marine mammals have historically thrived here, including the American clawed lobster, Homarus americanus, which
occupies a range from Labrador and Newfoundland to the shelf
waters of North Carolina (Holthuis, 1991), though historically,
it has been far more abundant from northern New Jersey to
Newfoundland.
Prehistorical and historical eras
Aboriginal peoples were present in Labrador some 8500 years ago,
and reached Newfoundland some 4500–5000 years ago, when the
climate was warmer than today (Marshall, 1996; Renouf, 1999).
The Beothuk Indians occupied Newfoundland at the time of
European contact in the 1400s (Tuck, 1976), although other groups
may have travelled there for particular resources (Marshall, 1996).
Faunal analysis suggests that Beothuk ate a wide variety of marine
and terrestrial animals, including lobster, recognized from claw
fragments (Howley, 1915; Marshall, 1996; Pastore, 1997). The
Beothuk harvested lobster during summer (Marshall, 1996) and
could store lobster tails (Tuck and Pastore, 1985) and dry the
meat (Anonymous, 1987). Europeans arriving in Newfoundland in
1578 commented on the great abundance of lobsters near the shore
(Parkhurst, 1578)—thus the Beothuk appear not to have fished populations down greatly.
The Mi’kmaq Indians (Canadian Maritime provinces, Gulf of
St. Lawrence to the Bay of Fundy, Nova Scotia, Cape Breton,
Prince Edward Island) fished for lobster when migrating from
inland winter quarters in spring and summer. They used handcollecting, spears, and hooks while wading at low tide, but also captured lobsters incidentally in fish weirs (Wallis and Wallis, 1955;
Anonymous, 2012). Lobsters were used as food, fertilizer, decorations, totems, tobacco pouches, and pipes (Ganong, 1968;
Anonymous, 2012).
English explorers and chroniclers of early colonies recounted
how natives in Maine (probably Eastern Abenaki or Mi’kmaq)
and Connecticut (Mohegan, Quinnipiac, and Pequot Indians)
used lobsters (Rosier, 1605; Morton, 1632). Natives used lobster
for bait to catch eel and cod, and dried them for food in winter
(Wood, 1639). In Connecticut, natives were described as converging
in large numbers for a month or more to harvest lobsters at low tide
and dry their meat for winter (Morton, 1632). Lobster remains have
been found in a Manissean seafood processing village midden at
Block Island (Rhode Island) and date to 2350– 2500 years ago
(K. McBride, pers. comm.; Jaworski, 1990); these may be the
oldest surviving remnants in New England. Other island tribes
(Wampanoags) have legends describing feasts where lobsters were
boiled in pots with succotash and fish, and served with steamed
clams, roast pheasant, and rabbit stew (Scoville, 1970).
Early European chroniclers also remarked on the abundance of
lobsters inshore, usually by recounting how quickly lobsters could
be caught via different fishing methods. For instance, Parkhurst
(1578) described how with just a single gaff, he could capture
enough lobsters to feed 300 men in less than a half a day in the
waters of Newfoundland. Others recounted how men, with only
one haul of a small draw net, caught more than 140 lobsters off
Nova Scotia in 1597 (published by Brown, 1869). Wood (1639)
remarked that lobsters were so plentiful that they (meat and shell)
were used as fertilizer instead of food when fish were readily available. Nearly 250 years later, hoop netters were able to capture 500
lobsters in 8 –10 h in waters off New Brunswick (Simmonds, 1859).
Introduction of modern fisheries
Many settlers in the 1600 and 1700s considered lobster as a suitable
stand-in when other foodstuffs were unavailable, as a fertilizer for
crops, or as a food for the poor (Rowan, 1876; Wells, 1986).
Lobsters were often used as bait to capture the more highly prized
i16
cod. By the late 1700s, lobster became a commodity in urban centres
(Boston, New York). High demand was initially met from nearby
waters, but by the early 1800s, southern New England fishers
could no longer keep up (Martin and Lipfert, 1985). During this
period of growing demand, the indigenous peoples were mostly
excluded from lobster fishing and did not regain guaranteed
fishing rights until the late 1900s (Davis et al., 2004). Live shipment
gave rise to the “smack vessel” (similar to the “hummer-busser”
used in Norway), and the “car” (floating box to hold lobster for
sale to the smack vessel). Hand-held gaffs that typically killed lobsters were replaced by hoop nets that required constant monitoring
by fishers. The inefficient hoop net gave way to the wooden trap
(“pot”) that became more sophisticated in design to prevent
lobster escape.
Lobster canning grew during the mid-1800s, and by the 1880s,
there were 23 plants in Maine and over 40 in Canada (Rathbun,
1887). The canneries required millions of pounds of lobster per
year to supply demand, and mounting fishing pressure caused
a decline in lobster size and weight (from 1.8– 4.5 to ,0.9 kg).
The live lobster industry also impacted stocks, because fishing rowboats became larger, engine powered, then gave way to large
mechanized boats, permitting fishing further from the shore.
Increased demand and the industrialization of the lobster fishery
led initially to regulations stipulating who could fish local waters,
but by the late 1800s these stipulations included a variety of measures that differed from state to state. Some of the stricter regulations
caused a decline in the Maine canneries with many closing; by the
end of the 1800s, all Maine canneries had closed or relocated to
the Canadian Maritimes (Martin and Lipfert, 1985).
Massachusetts and Rhode Island (later joined by Maine,
Newfoundland, and other Canadian regions) opened up state
hatcheries by the end of the 1800s, in an attempt to restock state
waters with more than 800 million larvae hatched from egg-bearing
females (Martin and Lipfert, 1985). Despite these “seeding” efforts
and the imposition of additional regulations, lobster stocks continued to experience dramatic drops in abundance and shifts to smaller
individuals. Today’s stocks are highly depleted in all but their most
northern range (Gulf of Maine northward), and even those stocks
only provide high yields with enormous fishing effort.
South Florida
The Caribbean spiny lobster, Panulirus argus, occurs in shallow
coastal waters of southeastern Florida, along the Keys archipelago,
and throughout the Caribbean Sea (Holthuis, 1991). Humans
settled the area during the Archaic Period (9000 –3000 years ago)
and gave rise to the Glades cultures (2750– 500 years ago). By the
time of the European contact (2500 years ago), the descendants
of the Glades culture had given rise to the Tequesta (near modernday Miami and the Upper Florida Keys) and the Calusa (west coast
of south Florida, from Fort Myers to Naples and the Ten Thousand
Islands area). Both tribal societies were based on marine and estuarine resource exploitation.
Prehistorical and historical eras
Middens indicate that the tribes consumed more than 50 species of
fish and 20 species of molluscs and crustaceans, sea and land
turtles, ducks, and some land mammals, although these latter items
accounted for ,7% of the their animal-based energy (MacMahon
and Marquardt, 2004). The Calusa had little contact with Europeans,
but one shipwreck survivor observed that lobsters were consumed
regularly by both the Calusa and Tequesta (Fontaneda, 1575).
E. Spanier et al.
Introduction of modern fisheries
Wars between Spain and England devastated the South Florida
tribes. The British took over Florida in the mid-1700s, and reported
that lobsters were plentiful and quite large, but they were not highly
prized as food, and were instead used as bait for fish such as snapper
and grouper (De Brahm, 1772; Moe, 1991). For the next centuryand-a-half, lobsters were captured by south Floridians with castnets,
gillnet, haul seines, and spears, and were easily captured en masse
when roaming over grass flats. A commercial fishing industry for
lobsters arose when the Florida Keys were connected to the mainland via the Overland Railroad in 1912, but no regulations on
fishing were imposed until the 1920s. Nevertheless, regulations
have failed to prevent overfishing and landings have declined dramatically in the present time.
Caribbean and Mesoamerica
The tropical Caribbean Sea basin (2 754 000 km2) supports five
spiny and four slipper lobsters species. The Caribbean spiny
lobster P. argus is the most abundant, and has been identified in
pre-Columbian archaeological deposits throughout the Caribbean
(Keegan, 1982, 2000).
Prehistorical and historical eras
The earliest migrants (ca. 6000 years ago) were hunter-gatherers,
who transitioned from tropical forest life to coastal living, shifting
from subsistence hunting to gathering shellfish (Rouse, 1989;
Wing, 1989; Keegan, 2000). Archaeological evidence for this shift
comes from sites in Cuba, Haiti, the Dominican Republic, Puerto
Rico, and the Lesser Antilles (see review in Keegan et al., 2013).
For example, middens from the pre-ceramic period (5000 –3000
years ago) in the Lesser Antilles contain thousands of mollusc
shells and fish bones (Wilson, 2007), demonstrating that subsistence
hunting had been replaced by coastal living.
Saladoid cultures, also known as ceramic-making farmer-fishers
(around 2400– 1100 years ago), originated along the Orinoco River
of lowland Amazonia (present-day Venezuela) and maintained
subsistence strategies that included shellfish gathering (Newsom
and Wing, 2004). Archaeological sites on Montserrat (2400 –2200
years ago), Nevis (965 –735 years ago), Turks and Caicos Islands
(915–715 years ago), eastern Puerto Rico and Anguilla (815 –615
years ago), and Saint Lucia (1115 –515 years ago) suggest that the
consumption of lobster was widespread and covered the entire
Ceramic Age occupation (de France et al., 1996). However, lobster
remains occurred in low frequencies at all sites, suggesting that
lobsters were not commonly harvested or remains were not well
preserved.
The Taino (Arawakan) civilization (6000 –500 years ago) occupied the Greater Antilles and some of the northern Lesser Antilles.
They were seafaring people who harvested conch, oysters, lobsters,
and crabs using various methods (Rouse, 1992). The sizes and
types of fish in archaeological deposits throughout the Caribbean
indicate that basket traps were probably used to harvest lobsters.
A test of this hypothesis by Keegan (1982) using 2 m long × 1 m
wide Haitian-style basket traps (Figure 7), similar to those used by
the Taino, caught lobsters of relatively small size at low frequency.
This matched archaeological evidence for ubiquitous, but limited
use of lobsters by the Taino (Keegan, 1982).
The Maya (Yucatan peninsula; 3800 – 1100 years ago) comprised a large and pervasive civilization extending from Mexico
into Mesoamerica (present-day Honduras, Belize, Guatemala, El
i17
Historical use of lobsters by humans worldwide
comes from reports by European explorers, such as Theodor de
Bry’s 450-year-old engraving of Incas bringing treasures, including
a lobster, as ransom for their leader, Atahualpa, held by the Spanish
(King, 2011). The effigies and engravings may, however, represent
the freshwater prawn, Macrobrachium brasiliense, which is widely
distributed in South America.
The spiny lobster Jasus frontalis is known from the Juan
Fernández and Desventuradas Islands in the southeastern Pacific
Ocean (Holthuis, 1991). Andrew Selkirk, a Scottish sailor, was marooned on Robinson Crusoe Island for 4 years. When he was rescued,
the ship’s captain described his men bringing back an “abundance of
Craw-fish” along with Selkirk (King, 2011). Selkirk’s biographer
wrote that he had often captured huge lobsters that he boiled or
grilled (Howell, 1829). Later travellers would write that this
lobster was large (weighing 3.5 or 4 kg each), highly flavourful,
and incredibly abundant (e.g. Anson, 1853).
Figure 7. Haitian-style basket trap used to capture fish and lobsters
(purchased in Cap-Haı̈tien, 1995).
Salvador, and Costa Rica). Mayans consumed many marine
animals, including lobsters, which were often dried, roasted, and
salted (Lang, 1971). Lobsters form part of Mayan aquatic iconography (Bonnafoux, 2008), and represented the expression of maritime trade (Lang, 1971). Effigy vessels were widespread and an
effigy of a Caribbean spiny lobster excavated in 2007 from Belize is
one of the oldest (http://www.belize.com/maya).
Introduction of modern fisheries
Commercial fishing for J. frontalis began in the late 1800s, when lobsters were commercially harvested from shallow waters (,15 m),
but as fishing pressure increased, lobsters were harvested from
deeper and deeper water (Holthuis, 1991). Assessments show that
the population near the island is overexploited, but fishing has
expanded to new grounds (Jeffs et al., 2013). The fishery continues
today and provides employment for a large proportion of the archipelago inhabitants.
New Zealand
At the time of the Spanish conquest in 1492, the Taino were cultivating crops and had evolved individual specialities including fishing
and trapping. New technologies would substantially increase
lobster harvests in the coming centuries. The 1600 and 1700s were
marked by increased trading between Europe and the New World.
New World fishing techniques that targeted both fish and lobsters
included bottom netting, trapping with wooden pots, harpooning,
torch fishing, and diving (Price, 1966).
New Zealand straddles the boundary of the Pacific Ocean and the
Tasman Sea midway between Antarctica and the tropics. The
islands lie at the confluence of two current systems: the warm
Subtropical Gyre (northwest and northeast coasts) and the cooler
Antarctic Circumpolar Current (southern islands). Upwellings
occur at the convergence zones of the two systems, providing nutrients for rich fishing grounds. Two spiny lobster species inhabit New
Zealand waters: the widespread red spiny lobster, J. edwardsii, and
the packhorse lobster, S. verreauxi, that is restricted to the northeast
North Island (Kensler, 1967).
South America and the Juan Fernandez archipelago
Prehistorical and historical eras
It is unknown how early South America was colonized after peoples
migrated through Beringia; however, human occupation sites, particularly Monte Verde in Chile and Quebrada Jaguay and Quebrada
Tacahuay in Peru, have dated to around 14 600 years ago (Dillehay
et al., 2008) and around 12 700 –12 900 years ago (Keefer et al.,
1998), respectively.
New Zealand was first settled 1200 years ago by Polynesian seafarers
who initially lived along the coast and relied heavily on gathering
seafood from shallow waters (Best, 1986). Archaeological evidence
(mandibles) from multiple shell middens date to 700 –900 years
ago and show that lobsters were exploited (Anderson, 1979; Leach
and Anderson, 1979; Leach and Boocock, 1993). Written records
from European explorers and Maori oral history indicate that lobsters were important seafood (Best, 1986). Apart from hand gathering in shallow waters, Maori fishers also used plant-based traps and
fishing lines baited with abalone (Best, 1986). Thus far, only J.
edwardsii mandibles have been found at archaeological sites
(Leach and Boocock, 1993).
Lobsters were eaten fresh, dried, and prepared by soaking in
freshwater so that the flesh separated easily from the shell without
cooking (Best, 1986). Coastal Maori traded sun- and oven-dried
lobsters with inland tribes (Best, 1986). Lobster fishing was highly
regulated within Maori communities through cultural rituals and
taboos. Chants and charms related to the process of catching lobsters
were integral to fishing, reflecting a spiritual perspective towards the
natural world. Tribal groups administered spatial, seasonal, or rotational harvesting strategies (Best, 1986). Even with these measures,
Introduction of modern fisheries
Prehistorical and historical eras
Faunal evidence suggests a wide use of resources, both coastal and
mountainous. There is clear evidence that marine seaweeds from
sandy and rocky shorelines were exploited (Dillehay et al., 2008)
where marine shellfish and crustaceans would also be present. In
Quebrada Jaguay and Quebrada Tacahuay, knotted cordage, fish
bones, and crustacean and mollusc shells have been recovered (e.g.
Keefer et al., 1998). Indications of lobster use on the mainland
include lobster effigies from ancient Peru (e.g. of the Chimu-Inca
Culture, Moche Valley, north coast of Peru, dating 585–475 years
ago, Mingei Museum, CA, USA; http://www.mingei.org/collection
object/ceremonial-sprinkler-with-lobster-effigy-2/). This is peculiar, because no shallow water marine clawed or spiny lobsters are
present along the Pacific coast of South America. More information
i18
changes in mandible size in some middens strongly suggest that the
Maori had negative impacts on J. edwardsii populations and drove
the species to smaller sizes over several centuries (Leach and
Anderson, 1979). Some archaeologists suggest that the Maori may
have practiced intensive use of a resource for a time, followed by a
period where they allowed the resource to recover (Leach and
Boocock, 1993), but it does not appear that intensively fished populations of J. edwardsii rebounded.
Introduction of modern fisheries
There is now a significant commercial fishery in New Zealand, with
around 2700 t of lobsters harvested a year, and with significant
stakeholding by Maori descendants.
Conclusions
It is undisputed that ancient and more recent peoples around the
world used lobsters for food and trade. In many cultures, lobsters
were a valuable and highly prized commodity; in others, they were
considered food for the poor, or a replacement when other foods
were not available. However, before the modern era, knowledge of
lobster use is based mainly on archaeological evidence, artwork,
or historical accounts. These sources are less exact and often biased
by ancient and sometimes lost cultural insights. Nevertheless, they
do give important information, as long as their interpretation
accounts for the cultures from which they derive. One should
keep in mind that when Europeans provided descriptions of
lobster stocks in new lands, they did so within the context of
their own often already overfished home stocks. Therefore, when
Europeans described new found stocks of lobsters as highly abundant, we should take them at their word that lobsters were as abundant as described and that these written descriptions were not
exaggerations (Bolster, 2012).
Apart from the Mediterranean civilizations and possibly some
Maori tribes, indigenous peoples appear to have harvested lobsters
at a sustainable rate, even when other taxa (marine mammals,
pelagic fish, land mammals, shellfish) in the same area may have
been overexploited. Traditional hand collection methods, small
baskets (pots), and gaffs limited the numbers of lobsters that could
be captured at one time. As colonists increased their demand for
lobster, new technologies were developed that increased the efficiency
of capture.
Overexploitation typically followed a specific pattern. For
example, when lobsters became sought-after seafood commodities
in European and other markets during the mid-1800s, fisheries
became industrialized, extracting greater quantities. International
trade followed the development of preservation techniques (frozen,
canned, or live lobsters) and long-distance transport of products.
Some of the modern fishing methods and gear, such as trammel-nets,
led to precipitous declines in lobster abundance. In some cases, indigenous peoples were pushed out of lobster fisheries during the period
of economic expansion; in countries such as New Zealand and
Canada, they now have traditional fishing rights.
Lobster life history, which typically progresses from pelagic and
well-hidden juvenile stages to more accessible adolescent and adult
stages, has allowed them to retain a part of the recruiting population
out of reach of fishing gear. Nevertheless, the long lifespans, slow
growth, and variable recruitment success leave many lobster
species vulnerable to continued overexploitation. Despite regulatory attempts, lobster stocks around the world are currently
heavily or overexploited, and it is unrealistic to expect a recovery
E. Spanier et al.
to the near-pristine levels reported by explorers some 400– 500
years ago.
High landings of some stocks today come as a result of a highly
mechanized fishing fleet, use of a large number of pots or nets,
smaller sizes of landed lobster, and reductions in the age of reproduction. In addition, lobsters inhabit ecosystems where large
groundfish have themselves been overfished, and where some of
their prey (e.g. urchins, oysters, scallops, clams, mussels) may also
have been overfished. Hence, today’s lobster populations may be
much more vulnerable to perturbations, such as fishing pressure
and climate change than in the past. Recognizing how anthropogenic selection pressures imposed upon lobster populations have
altered their demography, genetic diversity, and ecosystems in
which they live is critical for future management schemes.
Historical analyses, similar to those conducted by McArdle (2008)
for California spiny lobster and Bolster (2012) for groundfish,
may prove useful to managers and fishers alike in furthering an
understanding of human impacts on worldwide lobster stocks.
References
Anderson, A. J. 1979. Prehistory exploitation of marine resources at
Black Rocks Point, Palliser Bay. National Museum Bulletin, 21:
49 – 65.
André, J. 1961. L’alimentation et la cuisine a Rome. Librairie
Klincksieck, Parı́s. 262 pp.
Anonymous. 1987. The Spirit Sings, Artistic Traditions of Canada’s First
Peoples, p. 22. McLelland and Stewart and the Glenbow Museum,
Toronto. 264 pp.
Anonymous. 2004. Rock Lobster Fishery Management Strategy—
Assessment of Indigenous Issues and Historic Heritage Issues.
NSW Department of Primary Industries, Sydney, NSW, Australia.
145 pp.
Anonymous. 2012. A Brief History of the Lobster Fishery in the
Southern Gulf of St. Lawrence. Gulf Region, Fisheries and Oceans
Canada, Moncton, New Brunswick. 47 pp.
Anson, G. 1853. AVoyage Round the World in the Years 1740, 41, 42, 43,
44, p. 39. Ingram, Cooke & Co., London. 128 pp.
Armitage, S. J., Jasim, S. A., Marks, A. E., Parker, A. G., Usik, V. I., and
Uerpmann, H-P. 2011. The southern route “Out of Africa”: evidence
for an early expansion of modern humans into Arabia. Science, 331:
453– 456.
Babbit, F. C. 1957. Plutarch’s Moralia. William Heinemann, London.
457 pp.
Barshaw, D. E., Lavalli, K. L., and Spanier, E. 2003. Is offence the best
defence: the response of three morphological types of lobsters to predation. Marine Ecology Progress Series, 256: 171– 182.
Best, E. 1986. (3rd print, first published in 1929). Fishing methods and
devices of the Maori. Published by National Museum, Wellington,
Dominion Museum Bulletin, No. 12. 264 pp.
Bolster, J. 2012. The Mortal Sea: Fishing the Atlantic in the Age of Sail.
Bellnap Press, Cambridge. 416 pp.
Bonnafoux, P. 2008. Etude iconographique des céramiques du Classique
ancient dans les basses terres mayas. Unpublished PhD dissertation,
Université Paris 1 Panthéon—Sorbonne, Paris. 336 pp.
Bracken-Grissom, H. D., Ahyong, S. T., Wilkinson, R. D., Feldmann, R.
M., Schweitzer, C. E., Breinholt, J. W., Bendall, M., et al. 2014. The
emergences of the lobsters: phylogenetic relationships, morphological evolution and divergence time comparisons of an ancient
group (Decapoda: Achelata, Astacidea, Glypheidea, Polychelidae).
Systematic Biology, 63: 457 –479.
Brink, Y. 2004. The transformation of indigenous societies in the southwestern Cape during the rule of the Dutch East India Company
(1652– 1795). In The Archaeology of Contact in Settler Societies,
pp. 91– 108. Ed. by T. Murray. Cambridge University Press,
Cambridge.
Historical use of lobsters by humans worldwide
Brown, R. 1869. A History of the Island of Cape Breton with Some
Account of the Discovery of Canada, Nova Scotia, and
Newfoundland. Sampson Low, Son and Marstow, London. 44 pp.
Charmantier, G. 2014. Crustaceans in art. In The Crustacea, Traite de
Zoologie 4A, Chapter 27B, pp. 139 – 187. Ed. by J. C. von Vaupel
Klein, M. Charmantier-Daures, and F. R. Schram. Koninklijke
Brill, Leiden, Boston. 462 pp.
Crane, M. T. 2009. Illicit privacy and outdoor spaces in early modern
England. Journal for Early Modern Cultural Studies, 9: 4 – 22.
DAFF. 2012. Status of the South African marine fishery resources.
Department of Agriculture, Forestry and Fisheries, Cape Town,
South Africa. 67 pp.
d’Arcy-Thompson, W. 1952. The Works of Aristotle. Vol. 4, Historia
Animalium. Clarendon Press, Oxford. 508 pp.
Davis, A., Wagner, J., Prosper, K., and Paulette, M. J. 2004. The
Paq’tnkek Mi’kmaq and Ka’t (American eel): a case study of cultural
relations, meanings, and prospects. The Canadian Journal of Native
Studies XXIV, 2: 359 – 390.
De Brahm, J. G. W. 1772. The Atlantic Pilot. T. Spisbury, London. 25 pp.
de France, S. D., Keegan, W. F., and Newsom, L. A. 1996. The archaeobotanical, bone isotope, and zooarchaeological records from
Caribbean sites in comparative perspective. In Case Studies in
Environmental Archaeology, pp. 289– 304. Ed. by E. J. Reitz, L. A.
Newsom, and S. J. Scudder. Plenum Press, New York. 463 pp.
de Sainte Pierre, M., Bravi, C. M., Motti, J. M. B., Fuku, N., Tanaka, M.,
Llop, E., Bonalto, S. L., et al. 2012. An alternative model for the
early peopling of southern South America revealed by analysis of
three mitochondrial DNA haplogroups. PLoS One, 7: e43486.
doi:10.1371/journal.pone.0043486.
Dillehay, T. D., Ramı́rez, C., Pino, M., Collins, M. B., Rossen, J., and
Pino-Navarro, J. D. 2008. Monte Verde: seaweed, food, medicine,
and the peopling of South America. Science, 320: 784– 786.
Eriksson, A., Betti, L., Friend, A. D., Lycett, S. J., Singarayer, J. S., von
Cramon-Taubadel, N., Valdes, P. J., et al. 2012. Late Pleistocene
climate change and the global expansion of anatomically modern
humans. Proceedings of the National Academy of Science of the
USA, 109: 16089– 16094.
Erlandson, J. M., and Rick, T. C. 2010. Archaeology meets marine
ecology: the antiquity of maritime cultures and human impacts on
marine fisheries and ecosystems. Annual Review of Marine
Science, 2: 165– 185.
Erlandson, J. M., Rick, T. C., Braje, T. J., Casperson, M., Culleton, B.,
Fulfrost, B., Garcia, T., et al. 2011. Paleoindian seafaring, maritime
technologies, and coastal foraging on Californias’s Channel
Islands. Science, 331: 1181– 1185.
Flood, J. 1995. Archaeology of the Dreamtime: The Story of Prehistoric
Australia and Its People. Australia: Harper Collins, Sydney. 328 pp.
Fontaneda, H. D’E. 1575. Fontaneda’s Memoir. Translation by
Buckingham Smith, 1854. http://www.keyshistory.org/Fontenada.
html.
Fradier, G. 1982. Mosaiquees Romaines de Tunisie. Editions Ceres
Productions, Tunis. 63 pp.
Ganong, W. F. (Ed.) 1968. Denys: Description and Natural History of the
Coasts of North America (Acadia) 1672, pp. 424– 425. Greenwood
Press, New York.
Garrod, D. A. E., Buxton, L. H. D., Smith, G. E., and Bare, D. M. A. 1928.
Excavation of a Mousterian rockshelter at Devil’s Tower, Gilbraltar.
Journal of the Royal Anthropological Institute of Great Britain and
Ireland, 58: 33– 113.
George, R. W., and Main, A. R. 1967. The evolution of spiny lobsters
(Palinuridae): a study of evolution in the marine environment.
Evolution, 21: 803 – 820.
Gesner, C. 1558. Historia Animalium Liber III. Qui est de Pisicum &
Aquatilium animantium natura. Zürich: apud Christophorum
Froschoverum. British Library, 37. g. 7.
Glaessner, M. F. 1969. Decapoda. In Treatise on Invertebrate
Paleontology, R, Arthropoda, (4.2), pp. R262– R286. Ed. by R. C.
i19
Moore. University of Kansas and Geological Society of America,
Inc., Lawrence, Kansas, USA. 398 pp.
Glenister, C. L. 2008. Profiling Punt: using trade relations to locate
“God’s Land”. Master of Arts Thesis, University of Stellenbosch,
South Africa. 140 pp.
Goñi, R., and Latrouite, D. 2005. Review of the biology, ecology and fisheries of Palinurus species of European waters: Palinurus elephas
(Fabricius, 1787) and Palinurus mauritanicus (Gruvel, 1911).
Cahiers de Biologie Marine, 46: 127 – 142.
Groeneveld, J. C., Goñi, R., and Diaz, D. 2013. Palinurus species. In
Lobsters: Biology, Management, Aquaculture and Fisheries, 2nd
edn, Chapter 11, pp. 326 – 356. Ed. by B. F. Phillips. Blackwell
Scientific Publications, Oxford, UK. 488 pp.
Gulick, C. B. 1957. Athenaeus: “The Deipnosophists” LCL, 1. The Loeb
Classical Library. William Heinemann, London. 484 pp.
Henshilwood, C. S., d’Errico, F., Yates, R., Jacobs, Z., Tribolo, C., Duller,
G. A. T., Mercier, N., et al. 2004. Emergence of modern human behavior: Middle Stone Age engravings from South Africa. Science,
295: 1278– 1280.
Herrick. 1909. Natural history of the American lobster. Bulletin of the
Bureau of Fisheries, Bulletin of the U.S. Bureau of Fisheries, 29:
149– 408.
Holden, A. 1974. Theocritus: “The Idylls”. Harmondsworth, Penguin,
UK. 250 pp.
Holthuis, L. B. 1991. FAO Species Catalogue, No. 125, 13. Marine
Lobsters of the World. Food and Agricultural Organisation, Rome.
292 pp.
Howell, J. 1829. The Life and Adventures of Andrew Selkirk; Containing
the Real Incidents upon Which the Romance of Robinson Crusoe Is
Founded, pp. 72, 131 – 132. Oliver & Boyd, Tweeddale-Court, and
Geo. B. Whittaker, London, Edinburgh. 196 pp.
Howley, J. P. 1915. The Beothuks or Red Indians: The Aborginal
Inhabitants of Newfoundland. Cambridge University Press,
Cambridge. 348 pp. http://www.mun.ca/rels/native/beothuk/
beo2gifs/texts/shana2.html.
Hudson, T., and Blackburn, T. C. 1982. Food procurement and transportation. In The Material Culture of the Chumash Interactions
Sphere, 1. Ed. by T. C. Blackburn. Ballena Press Anthropological
Papers No. 25. Ballena Press, Los Altos, and Santa Barbara
Museum of Natural History, Santa Barbara, California. 387 pp.
Inman, D. L. 1983. Application of coastal dynamics to the reconstruction of paleocoastlines in the vicinity of La Jolla, California.
Quaternary Coastlines and Marine Archaeology, 1983: 1 – 49.
Jaworski, C. 1990. Discovery on Block Island: 2,500-year-old village predates agriculture. Nor’easter: Magazine of the Northeast Sea Grant
Programs 2: 32 –37.
Jeffs, A. G., Gardner, C., and Cockroft, A. 2013. Jasus and Sagmariasus
species. In Lobsters: Biology, Management, Aquaculture and
Fisheries, pp. 259– 288. Ed. by B. F. Phillips. Blackwell Scientific
Publications, Oxford, UK. 488 pp.
Jerardino, A. 2010a. Large shell middens in Lamberts Bay, South Africa:
a case of hunter– gatherer resource intensification. Journal of
Archaeological Science, 37: 2291– 2302.
Jerardino, A. 2010b. Prehistoric exploitation of marine resources in
Southern Africa with particular reference to shellfish gathering: opportunities and continuities. Pyrenae, 41: 7– 52.
Jerardino, A., Branch, G. M., and Navarro, R. 2008. Human impact on
precolonial west coast marine environments of South Africa. In
Human Impacts on Ancient Marine Ecosystems: a Global
Perspective, pp. 279– 296. Ed. by T. C. Rick, and J. M. Erlandson.
University of California Press, Berkeley. 336 pp.
Jerardino, A., Horwitz, L. K., Mazel, A., and Navarro, R. 2009. Just before
Van Riebeeck: glimpses into terminal LSA lifestyle at Connies Limpet
Bar, west coast of South Africa. The South African Archaeological
Bulletin, 64: 75 – 86.
Jerardino, A., and Navarro, R. 2002. Cape rock lobster (Jasus lalandii)
remains from South African west coast shell middens: preservational
i20
factors and possible bias. Journal of Archaeological Science, 29:
993– 999.
Jerardino, A., Navarro, R., and Nilssen, P. 2001. Cape rock lobster (Jasus
lalandii) exploitation in the past: estimating carapace length from
mandible sizes: research letter. South African Journal of Science,
97: 59 – 62.
Jones, R. 1992. The human colonization of the Australian continent. In
Continuity or Replacement: Controversies in Homo sapiens
Evolution, pp. 289– 301. Ed. by G. Bräuer, and F. H. Smith. A.A.
Balkema, Rotterdam. 315 pp.
Jones, T. L., and Klar, K. A. 2005. Diffusionism reconsidered: linguistic
and archaeological evidence for prehistoric Polynesian contact with
southern California. American Antiquity, 70: 457– 484.
Keefer, D. K., deFrance, S. D., Moseley, M. E., Richardson, J. B., III,
Satterlee, D. R., and Day-Lewis, A. 1998. Early maritime economy
and El Niño events at Quebrada Tacahuay, Peru. Science, 281:
1833– 1835.
Keegan, W. F. 1982. Lucayan fishing practices: an experimental
approach. The Florida Anthropologist, 35: 146 – 161.
Keegan, W. F. 2000. West Indian archaeology. Part 3. Ceramic age.
Journal of Archaeological Research, 8: 135– 167.
Keegan, W. F., Hofman, C. L., and Rodriguez Ramos, R. 2013. The
Oxford Handbook of Caribbean Archaeology. Oxford University
Press, New York. 624 pp.
Kensler, C. B. 1967. The distribution of spiny lobsters in New Zealand
waters (Crustacea: Decapoda: Palinuridae). New Zealand Journal
of Marine and Freshwater Research, 1: 412– 420.
King, R. J. 2011. Lobster. Reaction Books, London. 216 pp.
Klein, R. G., Avery, G., Cruz-Uribe, K., Halkett, D., Parkington, J. E.,
Steele, T. E., Volman, T. P., et al. 2004. The Ysterfontein 1 Middle
Stone Age site, South Africa, and early human exploitation of
coastal resources. Proceedings of the National Academy of
Sciences of the USA, 101: 5708– 5715.
Kleiven, A. R., Olsen, E. M., and Vølstad, J. H. 2012. Total catch of a redlisted marine species is an order of magnitude higher than official
data. PLoS ONE, 7: e31216. doi:10.1371/journal.pone.0031216.
Lang, F. W. 1971. Marine resources: a viable subsistence alternative for
the prehistoric lowland Maya. American Anthropologist, New
Series, 73: 619 – 639.
Leach, B. F., and Anderson, A. J. 1979. Prehistoric exploitation of
crayfish in New Zealand. In Birds of a Feather—Osteological
and Archaeological Papers from the South Pacific in Honour of
R. J. Scarlett, pp. 141 – 164. Ed. by A. J. Anderson. BAR
International Series 62, N.Z.A.A. Monograph 11.
Leach, B. F., and Boocock, A. S. 1993. Prehistoric fish catches in New
Zealand. Tempus Reparatum, BAR International Series, 584: 1 – 305.
Lloyd, G. E. R., and Chadwick, J. 1984. Hippocrates Writings. Penguin
Classics, London. 384 pp.
Loban, F. 2007. Ngalpun adhabath a goeygayil bangal (our sea, our
future): an examination of Torres Strait Protected Zone Joint
Authority principles and Torres Strait Islander needs and aspirations
for the Torres Strait fisheries, from a Torres Strait Islander perspective. Master’s thesis, James Cook University, Townsville, Australia.
176 pp.
MacMahon, D. A., and Marquardt, W. H. 2004. The Calusa and Their
Legacy: South Florida People and Their Environments (Native
Peoples, Cultures, and Places of the Southeastern United States).
University Press of Florida, Gainseville, FL. 240 pp.
Magnus, Olaus. 1539. Carta Marina. Stored at Carolina Redviva,
Uppsala University Library, Sweden (Map of the Sea) (in Latin).
Magnus, Olaus. 1555. Historia de gentibus septentrionalibus. Rome,
Italy (About the Nordic countries and people) (in Latin). 815 pp.
Maillard, A. S-P. 1863. Lettre à Madame de Drucourt. In Les Soirées
Canadiennes. Brousseau Freres, Québec. 304 pp.
Marean, C. W., Bar-Matthews, M., Bernatchez, J., Fisher, E., Goldberg,
P., Herries, A. I. R., Jacobs, Z., et al. 2007. Early human use of marine
E. Spanier et al.
resources and pigment in South Africa during the Middle
Pleistocene. Nature, 449: 905 – 908.
Marshall, I. 1996. A History and Ethnography of the Beothuk.
McGill-Queen’s University Press, Montreal. 640 pp.
Martin, K. R., and Lipfert, N. R. 1985. Lobstering and the Maine Coast.
Maine Maritime Museum, Bath, ME. 143 pp.
McArdle, B. A. 2008. Integrating ecology and history to understand historical marine population dynamics: a case study of the California
spiny lobster. Doctoral thesis, University of California, Santa
Barbara, CA. 193 pp.
McNiven, I. J., Brady, L. M., and Barham, A. J. 2009. KABADUL KULA
and the antiquity of Torres Strait rock art. Australian Archaeology,
69: 29 – 40.
Melville-Smith, R., and Van Sittert, L. 2005. Historical commercial West
Coast rock lobster Jasus lalandii landings in South African waters.
African Journal of Marine Science, 27: 33 – 44.
Moe, M. A., Jr 1991. Lobsters: Florida—Bahamas—the Caribbean.
Green Turtle Publications, Plantation, FL. 510 pp.
Montesinos, E. 1905. La pesca de la langosta. Notas de verano.
LaVanguardia 7 pp.
Morton, T. 1632. New English Canaan, or New Canaan. The Second
Book Containing a Description of the Natural Endowments of the
Countrie, and what Staple Commodities it yeeldeth. Chapter VII.
p. 61. Of the Fishes, and What Commodity They Prove. Charles
Green. 128 pp.
Morwood, M. J., O’Sullivan, P. B., Aziz, F., and Raza, A. 1998.
Fission-track ages of stone tools and fossils on the East Indonesian
island of Flores. Nature, 392: 173 – 176.
Newsom, L. A., and Wing, E. S. 2004. On Land and Sea. Native American
Uses of Biological Resources in the West Indies. University of
Alabama Press, Tuscaloosa, AL. 272 pp.
O’Connor, S., Ono, R., and Clarkson, C. 2011. Pelagic fishing at 42,000
years before the present and the maritime skills of modern humans.
Science, 333: 1117– 1121.
Orange, H. P. L., and Nordhagen, P. J. 1966. Mosaics: From Antiquity to
the Early Middle Ages. Methuen, London. 92 pp.
Ovadiah, R., and Ovadiah, A. 1987. Hellenistic, Roman and early
Byzantine mosaic pavements in Israel. Rome, “L’erma di
Bretschneider”. 276 pp.
Pajovic, B., Radosavljevic, M., Radunovic, M., Rdojevic, N., and
Bjelogrlic, B. 2012. Arthropods and their products as aphrodisiacs—review of literature. European Review for Medical and
Pharmacological Sciences, 16: 539– 547.
Pappe, K. W. L. 1853. Synopsis of the Edible Fishes of the Cape of Good
Hope. Ward, London. 57 pp.
Parkhurst, M. A. 1578. A letter written to M. Richard Hakluyt of the
middle Temple, conteining a report of the true state and commodities of Newfoundland, by M. Anthonie Parkhurst Gentleman,
1578. In The Principal Navigations, Voyages, Traffiques, and
Discoveries of the English Nations Made by Sea or Ouer-land, to
the Remote and Farthest Distant Quarters of the Earth, at Any
Time within the Compasse of these 1600 Yeres. Ed. by R. Hakluyt.
http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%
3A1999.03.0070%3Anarrative%3D635.
Pastore, R. T. 1997. The Beothuk. In Heritage Site of Newfoundland and
Labrador. Memorial University of Newfoundland. http://www.
heritage.nf.ca/aboriginal/beothuk.html.
Patek, S. N., Feldmann, R. M., Porter, M., and Tshudy, D. 2006.
Phylogeny and evolution of lobsters. In Lobsters: Biology,
Management, Aquaculture and Fisheries, pp. 113 –145. Ed. by B. F.
Phillips. Blackwell Publishing, Oxford. 506 pp.
Patek, S. N., and Oakley, T. H. 2003. Comparative tests of evolutionary
tradeoffs in a palinurid lobster acoustic system. Evolution, 57:
2082– 2100.
Phillips, B. F. 2013. Lobsters: Biology, Management, Aquaculture and
Fisheries, 2nd edn. Wiley-Blackwell, Oxford. 474 pp.
Historical use of lobsters by humans worldwide
Poinssot, C. 1965. Quelques remarques sur les mosaı̈ques de la Maison
de Dionysos et d’Ulysse à Thugga (Tunisie). In La Mosaı̈que
Gréco-Romaine I. Paris 29/08-03/09/1963, pp. 219 – 232.
Pontoppidan, E. 1753. Det første Forsøg paa Norges naturlige Historie,
forestillende dette Kongeriges Luft, Grund, Fielde, Vande, Væxter,
Metaller, Mineralier, Steen-Arter, Dyr, Fugle, Fiske og omsider
Indbyggernes Naturel, samt Sædvaner og Levemaade. Copenhagen
(in Danish). 464 pp.
Price, R. 1966. Caribbean fishing and fishermen: a historical sketch.
American Anthropologist, 68: 1363 – 1383.
Rakham, H. 1962. (Translator). Pliny the Elder: Natural History, 3. The
Loeb Classical Library, William Heinemann, London. 308 pp.
Rathbun, R. 1887. The crab, lobster, crayfish, rock lobster, shrimp, and
prawn fisheries. In The Fisheries and Fishing Industries of the United
States, Section V, History and Methods of the Fisheries, 11, pp.
627– 810. Ed. by G. B. Goode. United States Commission of Fish
and Fisheries. U.S. Government Printing Office, Washington, DC
881 pp.
Renouf, M. A. P. 1999. Prehistory of Newfoundland hunter-gatherers:
extinctions or adaptations? World Archaeology, 30: 403– 420.
Rick, T. C., Erlandson, J. M., Vellanoweth, R. L., and Braje, T. J. 2005.
From Pleistocene mariners to complex hunter-gathers: the archaeology of the California Channel Islands. Journal of World
Prehistory, 19: 169– 228.
Rodrı́guez-Noriega Guillén, L. 1996. Epicarmo de Siracusa: testimonios
y fragmentos. Edición crı́tica bilingüe. Universidad de Oviedo,
Servicio de Publicaciones, Oviedo. 247 pp.
Rosier, J. 1605 (reprinted in 1860). Rosier’s Narrative or Waymouth’s
Voyage to the Coast of Maine, in 1605. Complete with Remarks by
George Prince, Showing the River Explored to Have Been the
Georges River, pp. 21, 27. Eastern Times Press, Bath, ME. 45 pp.
Rouse, I. 1989. People and cultures of the saladoid frontier. In Early
Ceramic Population Lifeways and Adaptive Strategies in the
Caribbean, pp. 282 –289. Ed. by P. E. Siegel. British Archaeological
Reports, BAR International Series 506, Oxford. 418 pp.
Rouse, I. 1992. The Tainos: Rise and Decline of the People Who Greeted
Columbus. Yale University Press, New Haven, CT. 213 pp.
Rowan, J. J. 1876. The Emigrant and Sportsman in Canada. Some
Experiences of an Old Country Settler. p. 154. Edward Stanford,
London. 440 pp.
Scholfield, A. F. 1958. Aelian: On the Characteristics of Animals, 2.
Harvard University Press, Cambridge, MA. 359 pp.
Scoville, D. R. 1970. Indian Legends of Martha’s Vineyard. Martha’s
Vineyard Printing Company, Oak Bluffs, MA. 28 pp.
Simmonds, P. L. 1859. The Curiosities of Food; or the Dainties and
Delicacies of Different Nations Obtained from the Animal
Kingdom, p. 324. Richard Bentley Publishing, London. 372 pp.
Sonkkila, C. 2013. Rock art may last up to 60,000 years. http://www.
cosmosmagazine.com/news/rock-art-may-last-up-to-60000-years/.
Storm, G. 1881. Samlede skrifter af Peder Clausson Friis. Den Norske
Historiske Forening, Kristiania (Brøgger). 493 pp. [in Danish: The
Collection of the writings by Peder Clausson Friis; The lobster, and
the Crabs (from The Animals and Plants of Norway (1599)].
Stringer, C. B., Finlayson, J. C., Barton, R. N. E., Fernández-Jalvo,
Cáceres, I., Sabin, R. C., Rhodes, E. J., et al. 2008. Neanderthal
i21
exploitation of marine mammals in Gilbraltar. Proceedings of the
National Academy of Science of the USA, 105: 14319– 14324.
Sturlason, S. 1222– 1231a. Heimskringla eða Sögur Noregs konunga,
Reykjaholt, Iceland. (Heimskringla or the saga of Norwegian
kings) [In old-Norse, source for numerous later interpretations in
Icelandic (used here: F. Jónsson, Copenhagen 1893, reprinted
Nabu Press 2013)]. 534 pp.
Sturlason, S. 1222 –1231b. Skáldskaparmála-quædi. In Snorra-Edda,
Reykjaholt, Iceland [In old-Norse, source for numerous later interpretations in Swedish (A. Uppström, Uppsala 1859, used here),
Danish, Norwegian and English]. 79 pp.
Sunde, J. Ø. 1991. “Fiske Grunden (. . .) utdgjør landets fortog og en del
af dets ejendom”—Den rettslege karakteren av og innhaldet i sedvanebasert fiskerettar i saltvann (Fishing grounds as private proterty—
the legal character and content of custom-based fishing rights in salt
water. Historisk tidskrift, 88: 7 – 39 (in Norwegian, abstract in
English).
Tarsitano, S. F., Lavalli, K. L., Horne, F., and Spanier, E. 2006. The constructional properties of the exoskeleton of homarid, scyllarid, and
palinurid lobsters. Hydrobiologia, 557: 9– 20.
Tuck, J. A. 1976. Newfoundland and Labrador Prehistory. Van Nostrand
Reinhold, Ltd., London. 127 pp.
Tuck, J. A., and Pastore, R. T. 1985. Nice place to visit, but. . .. prehistoric
human extinctions on the island of Newfoundland. Canadian
Journal of Archaeology 9: 69 –80.
Vermeij, G. J. 1977. The Mesozoic marine revolution: evidence from
snails, predators and grazers. Paleobiology, 3: 245 – 258.
Wahle, R. A., Tshudy, D., Cobb, J. S., Factor, J., and Jaini, M. 2012.
Infraorder Astacidea Latreille, 1802 P. P.: the marine clawed lobsters.
In The Crustacea: Traite de Zoologie, 9B, pp. 3 – 108. Ed. by F. R.
Schram, and J. C. von Vaupel Klein. Koninklijke Brill, Leiden. 362 pp.
Wallis, W. D., and Wallis, R. S. 1955. The Micmac Indians of Eastern
Canada. University of Minnesota Press, Minneapolis, MN. 530 pp.
Ward-Perkins, J., and Claridge, A. 1978. Pompeii A. D. 79: Essay and
Catalogue. Knopf, New York. 215 pp.
Webber, W. R., and Booth, J. D. 2007. Taxonomy and evolution. In The
Biology and Fisheries of the Slipper Lobster, pp. 25 – 52. Ed. by K. L.
Lavalli, and E. Spanier. CRC Press, USA.
Wells, K. 1986. The Fishery of Prince Edward Island. Ragweed Press,
Charlottetown, PEI. 211 pp.
Wilson, S. M. 2007. The Archaeology of the Caribbean. Cambridge
University Press, Cambridge. 207 pp.
Wing, E. S. 1989. Human exploitation of animal resources in the
Caribbean. In Biogeography of the West Indies, pp. 137 – 152. Ed.
by C. A. Woods. Sandhill Crane Press, Inc., Gainesville. 582 pp.
Wood, W. 1639. New England’s Prospect, pp. 29 – 30, 81. a John
Dawson, London. 88 pp.
Yellen, J. E., Brooks, A. S., Cornelissen, E., Mehlman, M. J., and Stewart,
K. 1995. A middle Stone Age worked bone industry from Katanda,
Upper Semliki Valley, Zaire. Science, 268: 553 – 556.
Zugasti, F. I. G. 2011. The use of echinoids and crustaceans as food
during the Pleistocene – Holocene transition in Northern Spain:
methodological contribution and dietary assessment. Journal of
Island and Coastal Archaeology, 6: 115– 133.
Handling editor: Carmel Finley
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i22 –i34. doi:10.1093/icesjms/fsv057
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Review
A review of lobster fishery management: the Western Australian
fishery for Panulirus cygnus, a case study in the development and
implementation of input and output-based management systems
J. W. Penn*, N. Caputi, and S. de Lestang
Western Australian Fisheries and Marine Research Laboratories, PO Box 20, North Beach, WA 6920, Australia
*Corresponding author: tel: +61 8 92030184; fax: +61 8 92030199; e-mail: jim.penn@fish.wa.gov.au
Penn, J. W., Caputi, N., and de Lestang, S. A review of lobster fishery management: the Western Australian fishery for Panulirus
cygnus, a case study in the development and implementation of input and output-based management systems. – ICES
Journal of Marine Science, 72: i22 – i34.
Received 30 August 2014; revised 12 March 2015; accepted 16 March 2015; advance access publication 12 April 2015
.
Lobster stocks around the world support high-value fisheries with production currently about 260,000 tonnes annually. The largest fisheries harvest
Homarus, Nephrops and Panulirus species with smaller production from the Jasus, Palinurus and Scylarid species groups. The majority of larger
industrial fisheries have systems limiting fishing effort or catches, while many of the smaller fisheries remain open access and have yet to implement
basic management controls. The review uses the Western Australian fishery for Panulirus cygnus, valued between AUS$200 – 400 million annually
with a long history of successful management, as a case study for the consideration of lobster fisheries management systems more generally. The
conclusions from the review suggest that an evolutionary approach to management with biological controls as a precursor to input-based controls
is necessary to allow sufficient fishery-based data to be accumulated for management decision processes to be effective. The case study experience
suggests that well-defined fishing rights leading to an input-based total allowable effort system with individually transferable effort (ITE) units can
provide efficient mechanisms for the reduction of latent effort, which characterises most lobster fisheries with open access or basic limited entry.
Further the system has been shown to be capable of generating significant license values for fishermen while maintaining owner-operators as the
dominant group in the fishery. The ITE system was also used effectively to adjust fishing to compensate for a severe environmentally-driven downturn in recruitment, but resulted in highly complex management rules. In 2010 the fishery moved seamlessly to a total allowable catch with individually transferable quotas which removed the complexity of management, further increased the catch value and reduced costs of fishing. Price/
earnings (P/E) ratios have been used to track trends in license values which highlight the industry’s increasing economic viability over time under
both input and output based management.
Keywords: input controls, licence valuations, lobster fisheries, management, output controls.
Introduction
Lobster stocks around the world support catches in the order of 260
000 t annually (FAO STATS, 2011) which are taken by a range of
fishing activities from artisanal fishing in many tropical regions to
industrial fisheries in temperate and cold waters (Phillips et al.,
2013). Lobsters are commercially taken mostly using traps, but
also by trawling and diving, are highly valued in the world seafood
market and also prized by recreational fishers.
Most large commercial fisheries initially relied on the biological control measures of minimum size limits and protection of
egg-bearing females to maintain breeding stocks and many
later developed forms of effort or catch controls to maintain stock
levels. In contrast, many of the smaller fisheries, particularly in
tropical Africa and Asia, remain open access, have limited capacity
to control fishing and have yet to implement basic biological
protection measures. Stocks in these fisheries are likely to be
suffering from growth and possibly recruitment overfishing, but
can continue to operate at depressed catch rates because of the
high value of lobsters and low costs of fishing in developing
economies.
# International Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i23
Review of lobster fishery management
The first recorded direct control on lobster catches was in South
Africa during the 1940s, when lobster tail quotas were set for the industrial fishery for Jasus lalandii (Melville-Smith and van Sittert,
2005). Later in the mid-1950s, the alternative approach of limiting
fishing effort to indirectly control catches of Panulirus argus began
in Mexico, when fishers’ cooperatives along the coast were issued
with exclusive lobster-fishing rights (Briones-Fourzan and LozanoAlvarez, 2000). In the early 1960s, the first control on fishing vessel
and pot (trap) numbers occurred when limited entry management
was introduced in the Western Australian (WA) fishery for
Panulirus cygnus (Bowen and Hancock, 1989), an approach later
adopted by many fisheries in developed economies. From these
initial measures to control lobster catches, more market-based management systems using freely tradable pot quotas in Western
Australia (de Lestang et al., 2010) and catch quotas in the New
Zealand and now most Australian fisheries have been developed
and are providing significant improvements in economic performance (Gardner et al., 2013).
Against this background, it is notable that most of the larger trapbased fisheries (see reviews in Phillips, 2013) continue to rely heavily
on biological measures to maintain their lobster breeding stocks. In
terms of controls on lobster catches, basic limits on effort remain
the most common form of management being applied in Canada,
the United States, Mexico, Cuba, Chile, and Japan, where about
two-thirds of the world lobster harvest occurs. The largest lobster
fishery where output controls, i.e. a total allowable catch (TAC),
are used is the multinational European trawl fishery for Nephrops
(Bell et al., 2013) and this system is also now applied to most
lobster trap fisheries in New Zealand and Australia and South Africa.
For most fisheries, management is focused on sustaining
maximum catches rather than on the economic performance with
the main exceptions being the fisheries in New Zealand and
Australia, where fully market-based versions of the total allowable
effort (TAE) [with individually transferable effort (ITEs)] and
TAC [with individually transferable quotas (ITQs)] management
have been implemented with a strong emphasis on maximizing economic performance (Gardner et al., 2013; Reid et al., 2013; Caputi
et al., 2015).
The primary purpose of this review is to assess the performance
of the alternative input and output-based management systems for
lobster fisheries using the Western Australian fishery for P. cygnus as
a case study. This is Australia’s most valuable single-species fishery
generating between AUS$200 and 400 million annually (Reid
et al., 2013) and has been the subject of comprehensive biological research and management since 1963 (Bowen and Hancock, 1989).
Over this period, the initial limited-entry system evolved into a
sophisticated TAE control system with ITE units issued to each
vessel (de Lestang et al., 2012), which allowed the fishery to
operate close to maximum sustainable yield (MSY) levels and to
generate significant licence values for fishers. More recently
(2010), this fishery was converted to TAC management with a strategy of targeting maximum economic yield (MEY; Caputi et al.,
2014b, 2015). This provides a unique opportunity to compare the
operation of the two alternative control systems on the same
fishery and also illustrates the way these more market-based management strategies impact both lobster stocks and industry economic performance. For this latter purpose, a new approach for valuing
commercial fishing rights, using a licence transfer price to earnings
(P/E) ratio method, has been developed and used to track licence
values through several decades of the effort management and
since the implementation of a TAC with ITQs.
Western Australian lobster fishery case study
Background
The western rock lobster P. cygnus is an endemic Australian species
occurring along the west coast where it supports a large fishery operating from Shark Bay (268S) in the north to Cape Leeuwin (348S)
in the south and around the Abrolhos Islands (28– 29.58S) on the
continental shelf edge (Figure 1). This fishery produced around
11 000 t annually from the 1980s to early 2000s, although the
catches have been significantly reduced by management action to
around 6000 t since 2008/2009 to adjust for lower recruitment
(puerulus settlement) resulting from environmental changes
(Caputi et al., 2014c; de Lestang et al., 2015). The stock also supports
an important licensed recreational fishery taking several hundred
tonnes annually (Melville-Smith et al., 2004) under catch share
arrangements (de Lestang et al., 2012).
Since 1949, the commercial fishery has been documented
through industry landing records and detailed research logbooks
since 1963. Additional information from puerulus settlement
(1968 onwards), observer programmes (starting 1971), and fisheryindependent spawning stock surveys (1991 onward) have provided
particularly detailed databases for the fishery. Stock assessments
using these data underpin management processes and have contributed to the fishery being the first certified by the Marine Stewardship
Council in 2000 (de Lestang et al., 2012).
Figure 1. The Western Australian coastline, showing the three western
rock lobster fishery management zones (A– C) and the puerulus
collector sites (filled circle) used in the catch forecasting process.
i24
Early management development (pre-1993–1994)
The history of the fishery records that minimum size limits and
spawner protection measures were first introduced in the late
1890s (Bowen and Hancock, 1989). By the 1930s, annual catches
were around 250 t and reached 1000 t in the late 1940s after a
market was developed for frozen lobster tails in the United States
(Brown, 1991). This led to a rapid expansion of fishing through
the 1950s and early 1960s, until some levelling out of catches and
the threat of an influx of vessels from outside of the state, led to
a WA Government decision to limit vessels numbers in 1963
(Bowen, 1980). These initial controls on fishing effort occurred
through duplicated State and Federal Government legislation;
however, since 1980, the fishery has operated solely under WA
(State) control following an Offshore Constitutional Settlement
(Anon., 1980) which simplified the management processes.
While the initial decision to limiting vessels to 858 was quickly
implemented, it immediately raised concerns that pot (trap)
numbers would increase and an additional rule limiting pots to
three per foot of boat length (10 m21) with a maximum holding
of 200 pots was introduced. Although this capped pot numbers at
76 000, fishers then responded by trying to replace their existing
vessels with larger ones to increase their pot holdings and led to a
policy (1965) only allowing the same length replacement vessels.
A simultaneous decision to phase out offshore processing vessels
suspected of undertaking illegal activities occurred and was facilitated by allowing them to redistribute their pot entitlements in
J. W. Penn et al.
small parcels to other fishers wanting to operate larger replacement
vessels.
The administrative process to allow replacement vessels and purchase of additional pots involved the buyer and seller submitting
evidence of the sale to the Fisheries Agency, which would then recognize the ownership change and link the associated pot entitlements to the new vessel. This by default led to the “transferability”
of the licence or “right to fish” for lobsters. Although these pot
licences initially had little value, the vessel sale prices soon began
to involve an additional “goodwill” payment of around the value
of 1 year’s catch by the number of pots being transferred (Bowen,
1980).
Although the vessel and pot numbers were controlled, nominal
fishing effort measured as “pot lifts” continued to increase
through the 1970s and1980s and the effectiveness of fishing gear
was also increasing (Brown, 1991). This led to additional controls
on pot size and a requirement for multiple escape gaps. Despite
these measures, effort and catches continued to increase partly
due to the increasing efficiency (Brown et al., 1995), but also due
to a series of high catch years and improved juvenile survival from
better handing of released undersize lobsters (Brown and Caputi,
1986).
While these additional controls slowed the rate of increase in the
catch, the number of pot lifts continued to expand (Figure 2) as
larger replacement vessels were able to work more days during the
fishing season. To specifically address this issue of “effort creep”,
Figure 2. The history of catches, fishing effort, and TAE in pot-lifts for the western rock lobster fishery, from the start of significant commercial
fishing in the 1940s to 2009/2010. From 2010/2011 to 2014/2015, the fishery operated under TAC management and a 14-month quota season
(with proportional TAC increase) occurred in 2011 –2013 to reset the start date for the quota year.
i25
Review of lobster fishery management
the fishing season was shortened by 6 weeks in 1977 and in response
to very low puerulus settlement in 1982/1983, pot entitlements were
also temporarily reduced by 10% for the 1986/1987 season. This pot
reduction, although supported by industry leaders, was resisted by
most fishers, but found to be economically beneficial and made permanent at 2% per year over 5 years from 1987 to 1988 (Brown,
1991).
Development of sophisticated effort management systems
During the late 1980s and early 1990s, there were considerable
improvements in fish-finding equipment (colour echosounders
and GPS plotters) that enabled fishers to better target lobsters in
deep water where most of the breeding stock resides (Hall and
Chubb, 2001). To compensate for the resulting decline in breeding
stocks possibly related to lower puerulus settlements, a controversial
management package was introduced in 1993/1994. This increased
the minimum size by 1 mm, extended protection from berried
females to all mature females exhibiting egg-bearing setae, and
“temporarily” reduced the number of pots that could be used by
each licensed vessel by 18%. This move to “variable pot usage”
was the effective start of the full TAE with ITEs management
system which was formalized in 2001, when each vessel’s pot
holding was converted to ITE “units”. The 68 961 units issued corresponding to the number of pots in use before the 1993/1994
package and effectively fixed the “share” of the fishery that each
vessel operator would have access to in the future. The 18% pot reduction was then applied as a “pot usage ratio” (0.82) to each vessel’s
ITE unit holding, to determine the number of pots which could be
deployed for that season. (Note: reference to TAE or TAC management also implies individually transferable units of effort or catch
are in use.)
The TAE at that time then became the maximum pot lifts permitted, i.e. the total number of pots in use (68 961 units×0.82) multiplied by the fishing days during the season (227) or 12.8 million pot
lifts. This annual TAE since the start of limited entry (Figure 2) compared with the actual effort in pot-lifts each year shows that the
fleet had considerable scope to increase effort under the rules,
when the initial controls were enacted in 1963. This “latent” or unutilized effort capacity (Hall and Brown, 2000) was mostly created by
the increased seagoing capacity of replacement vessels and the
advent of technology for offshore fishing not available to the original
fleet. While the series of management decisions to reduce effort
through the 1970s and 1980s had some impact on this increasing
fleet capacity, it was not until the 1993/1994 management
changes that latent effort was effectively removed from the fishery
(Figure 2).
A further small adjustment to the TAE (Figure 2) occurred in
2005/2006 before significant reductions were implemented in
2008/2009 and 2009/2010 to adjust for an unprecedented environmentally caused decline in puerulus settlement (Caputi et al., 2014c;
de Lestang et al., 2015). These latter effort reductions (.70%)
represented a proactive management response, being introduced
before the poor puerulus year classes reached the legal size for
fishing, i.e. 3 –4 years later. As a result of this emergency process,
the TAE was reduced to around 2.2 million pot lifts with the objective of achieving a target catch of 5500 t in future years (Reid et al.,
2013). The rules negotiated with industry to achieve this outcome
were, however, very complex and varied between management
zones. These included different pot usage rates, closed periods,
limits on fishing days each week, and alterations to minimum and
maximum sizes, together with uniform changes to escape gaps
(Department of Fisheries, 2011). Predictably, the large reduction in
TAE also had a significant impact on vessel numbers which declined
from around 460 to 280 vessels over the 2 years (Figure 3) and contributed significantly to the decrease (70%) since the start of limited
entry.
The reduction in TAE was successful in lowering the catch to
7470 and then 5770 t in the 2008/2009 and 2009/2010 seasons, respectively (Figure 2), well down from the forecast catches from the
three zones (Figure 4) in aggregate totalling 9900 and 8900 t had
the effort continued at 2007/2008 levels. This resulted in a significant carry-over of legal stock into the years affected by the very
low recruitment, a substantial increase in catch rate from around
1 to 2.5 kg pot lift21 and breeding stocks reaching historically
high levels (Caputi et al., 2014c). In addition, the effort of
,3 million pot lifts had the fortuitous effect of moving the fishery
close to the MEY level previously projected (Reid et al., 2013).
This lower effort and fleet size automatically reduced operating
costs and the smaller lobster supply increased prices such that
total fishery revenue did not decrease substantially and profitability
increased, despite the lower catches (Reid et al., 2013).
While the TAE changes implemented in 2008/2009 and 2009/
2010 achieved their objectives, the industry negotiation process to
tailor management to regional needs proved particularly difficult
and the resulting fishing arrangements were complex to administer
and enforce. Some of these controls also required frequent stock
assessments, real-time monitoring of catch, and adjustments midseason to ensure the fishery achieved the nominal target catch.
For these reasons, the TAE changes created significant industry
debate, uncertainty, and costs which ultimately contributed to the
Government’s unilateral decision to simplify management by directly setting a TAC and adopting ITQs. This occurred despite a
lack of overall industry support at the time, although there had
been industry consultation and debate in the preceding years;
J. Kennedy, pers. comm. (Fisheries Department, Manager West
Coast Lobster Fishery). In fact, first detailed consideration of a
TAC with ITQs had started in the early 1990s (Bowen, 1994)
without ever achieving majority industry support.
Conversion of fishery from TAE to TAC management
The management change to a TAC with ITQs occurred at the start of
the 2010/2011 season (de Lestang et al., 2012) and was simplified by
direct conversion of the ITE units to ITQ units, so that the share of
the fishery held by each owner remained the same. While the
one-to-one conversion disadvantaged the more efficient fishers
who had historically achieved above-average catch per ITE unit,
some of these operators were supportive and accepted the situation
to enable a quicker transition to the new system and expected economic benefits. The existing target catch (notional TAC) used to
set the TAE was also adopted as the actual TAC for the first 2 years
of the new system. The smooth transition was also assisted by retaining most of the existing gear controls to minimize disruption to
vessel operations, although pot usage rates were relaxed back to
50% of the unit holding and the season was extended by 6 weeks.
In addition to the anticipated reduction in operating costs, the
TAC with ITQ system had a significant impact on catch value.
While the previous catch reduction to 5500 t under the TAE
system generated a useful price increase (Figure 5) of around
US$5– 7 kg21 resulting from essentially all of the reduced catch
being exported live at premium prices, the move to a TAC had a
more significant impact (Reid et al., 2013; Caputi et al., 2014b).
This event corresponded to an approximate US$10 kg21 additional
i26
J. W. Penn et al.
Figure 3. Fleet numbers and effort in pot lifts in the West Australian rock lobster fishery in three periods: 1963/1964 – 1992/1993 when limited
entry operated, 1993/1994 – 2009/2010 when ITE management was developed, and 2010/2011 onwards following adoption of ITQ management.
increase in price (Figure 5), as the longer season allowed fishers to
reduce catches during high abundance periods with depressed
prices and target periods of high demand. Highgrading to select preferred market sized also contributed to the price increase with
limited negative impact as lobsters handled correctly suffer little
discard mortality (Brown and Caputi, 1986). The full 12-month
season introduced in 2013 also corresponded to further price rises
(Figure 5) of about US$7 –8 kg21 (Caputi et al., 2014b), although
this may have also been influenced by greater demand for lobsters.
[Note: Reid et al. (2013) reported beach (dockside) prices converted
to US dollars to minimize the impact of Australian dollar exchange
rate variations, as essentially all of the catch is exported and marketed under contracts in US$.]
Overall, the conversion of the fishery from a TAE (with ITEs) to a
TAC (with ITQs) has had a significant economic benefit and simplified the management arrangements in the fishery. MEY has been formally added to the management plan for fishery in 2014, in addition
to sustainability objectives. Second, the very conservative TAC
adopted initially for sustainability reasons, has also minimized the
inherent risks of this control system where the catch available can
be highly variable, and overestimates of stock abundance can lead
to breeding stock depletion. The implementation process for the
TAC (with ITQs) in the fishery also benefited greatly from the longstanding catch prediction system (Caputi et al., 2014a), which had
also allowed industry to be restructured in advance of the forecast
low catch period from 2010 to 2011 onwards. Further, the historical
reliability of the forecasts, which could only have been developed
under the TAE system, provided industry with confidence that
their catch quotas would be achieved and the fishery would
remain sustainable under the new TAC system.
Fishery licence valuations
The early adoption of licence transferability resulted in pot licences
having a sale value and created a small “windfall” profit for the original vessel operators. This generated considerable debate at the
time about the associated capital costs and particularly about their
inhibiting new fishers from entering the fishery (Bowen and
Hancock, 1989). Despite these concerns, lobster pot values continued to increase over time from several hundred AUS dollars in the
early 1970s to around AUS$10 000 per pot unit by 1992/1993
(Penn et al., 1997) with industry adapting to the changed situation.
Prices further increased after the 1993/1994 management package
to around AUS$26 000 in 1995/1996 when the effort reduction
improved catch rates and created additional demand for pots.
Review of lobster fishery management
i27
Figure 4. Catches and catch forecasts (based on puerulus settlement 3 – 4 years prior) taking into account effort levels for the three fishing zones in
the western rocklobster fishery from 1986/1987 to 2007/2008. Forecasts from 2008/2009 onward assume the effort levels of 2007/2008 continued;
however, the catches achieved reflect the significant management changes to the TAE in 2008/2009, 2009/2010, and the implementation of a TAC
in 2010/2011.
Prices later peaked at AUS$36 000 during early 2000s before falling
to around AUS$10 000 when forecast catches were in steep decline
and before the change to ITQs, when values again returned to the
previous highs, i.e. above AUS $30 000. While the overall increase
in pot or unit prices has been significant over time, the values
obtained have also been affected by other factors unrelated to the
management performance of the fishery, e.g. inflation rates, large
exchange variations relative to export market destinations, and increasing fuel costs. These outside influences on pot unit prices
make it difficult to relate the values achieved to the economic performance of the fishery over time and also make comparisons
with other lobster fisheries less reliable.
In a more general context, these prices paid for pot units traded
through the licence register set up under the Fisheries Act are an equivalent to “stock market” prices where the fishery as a whole is equivalent
to a listed “company” and the ITE or ITQ units are effectively “shares”
in that company. On this basis, the more conventional approach of
valuing company shares (in this case, fishery units) has been adopted
by using the ratio of the prices paid for a unit relative to the expected
earnings from a unit. The use of a purchase price to earnings (P/E)
ratio better reflects the underlying performance of the fishery
(company) and allows easier comparison with other fisheries.
This valuation approach is essentially the same basic method
used by financial analysts (Campbell and Shiller, 1988) to value
listed corporations, except for the use of income from the season following purchase rather than the reported income for the previous
period. This variation to the conventional method was adopted
because the catch prediction system for the fishery (Figure 4) has
provided the purchaser with a particularly reliable forecast of the
catch by a unit and lobster processors provide some guidance
on market prices before the season start when sales typically
occur. Purchasers are more likely to use these forecasts to guide
their decisions than rely solely on the previous season’s catch or
price outcomes. In this context, the P/E ratio valuations for
lobster fishery units are an indication of the “market” view of the
future profit stream from ownership of the units and the long-term
sustainability of a fishery. This method also formalizes the system
historically used by the WA Fishery Agency when comparing the
performance of different fisheries and by some bankers when
assessing fishers’ loan applications.
The time-series of P/E ratios created using this method for the
fishery (Figure 6) show that valuations were initially relatively low
(,1) through the early 1970s reflecting the uncertainty about the
fishery at that time. However, by the early 1980s, when illegal
i28
J. W. Penn et al.
Figure 5. The relationship between catch volume and average beach (dockside) price for the WA lobster fishery from 1991/1992 to 2014 under
both TAE and TAC management systems (updated from Reid et al., 2013).
activity, including the sale of undersize lobsters and use of pots more
than licence numbers (de Lestang et al, 2012) had been effectively
eliminated and the underlying economics of the fishery became
better known, the P/E ratios had moved up to the 2 –3 range.
(Note: the P/E ratios reflect the expectation of cumulative future
profits flowing from the ownership of the fishing rights in perpetuity.) The first pot reduction programme (1986/1987–1991/1992)
further improved the fishery’s economic performance and created
additional demand for pots from fishers wanting to bring their
vessels back to full design capacity, i.e. add back the 10% of pots
lost which effectively increased their share of the fishery. A further
major increase in value occurred after the 1993/1994 management
changes which reduced pot usage by 18%, and corresponded to the
P/E ratio reaching new highs between 6 and 8 from 1995 to 1997
(Figure 6). This increase was again the result of market demand
from successful and “efficient” operators buying additional ITE
units and others taking the opportunity of high pot values to
retire from the fishery. These generally higher P/E ratios also corresponded to improved puerulus settlement (Caputi et al., 2014c)
which indicated increasing catches over the next 3 –4 years and
influenced the ability of fishers to borrow to purchase additional
units. The P/E ratios from the late 1990s onwards became more
volatile as the fishers responded more directly to changes in puerulus
settlement (Figure 7) based forecasts (Figure 4) and the lobster
supply–export price relationship (Figure 5) became better
known. While the P/E decline in 2007/2008 and 2008/2009 corresponded to the record lows in puerulus settlement, it was notable
that the radical management decision to cut the TAE from 2008 to
2009 onwards and carry forward catches into years affected by low
puerulus settlement had an immediate impact on the industry confidence, with the effect that the P/E ratio stabilized (Figure 6).
Despite the ongoing stock uncertainties, the resulting improvement
in catch rates, lower fishing costs and increased lobster export prices,
corresponded to a slight upturn in values, which continued after the
change to ITQ management in 2010/2011 season (Figure 5). The
associated lobster price increases and further lowering of costs
under ITQs all appear to have contributed to the P/E ratios returning to the previous highs of 6– 8 (under the TAE) and then moving
to above 9 (Figure 6). The P/E ratio is expected to remain at these
higher levels, given the increased certainty in catches and ongoing
improvements in lobster price under the more flexible ITQ fishing
system.
Ownership of fishing rights in Western Australian fishery
The issue of lobster licence ownership became of concern to both
Western Australian fisheries administrators and industry leaders,
when limited-entry licences began to assume significant values
(Bowen and Hancock, 1989). However, despite the lack of specific
controls on who could own lobster fishing licences, ownership
outside of the Western Australian fishing community has not
become a feature of the fishery to date. While licence ownership controls were not imposed, there was until recently a legislated requirement that lobster processing companies (also historically subject to
limited licences) were majority (80%) Australian owned to control
offshore ownership and the associated possibility for transfer
pricing to minimize Australian taxes.
This lack of corporate ownership persisted through the effort
management period, despite the presence of one multinational
i29
Review of lobster fishery management
Figure 6. The TAE and unit licence valuations (P/E ratios) for the western rock lobster fishery from the early 1970s to the present, including the
periods under basic limited entry management (1970/1971 – 1992/1993), TAE with ITE unit management (1993/1994– 2009/2010), and TAC
with ITQ management (2010/2011 onwards). The 2014/2015 estimates are preliminary. Note: The P/E ratio uses the catch value achieved by a unit
in the season following purchase.
corporation processing lobsters and operating vessels, when limited
entry was first introduced. This company was allocated vessel
licences in the initial 1963 allocation, but was using employee skippers and found its fleet operation was unprofitable. The company
later sold the vessels and ultimately its processing licence. Since
then, larger corporations and public companies have not become
involved in the fishery through the effort-management period
(to 2010), despite the fishery’s history of profitability. More recently,
some private family-based companies have purchased additional
vessels and successfully operated small fleets, mostly using family
or independent skippers on equitable share fishing arrangements.
Overall, the majority of licences have remained under individual
or owner operator control under small family-based company
structures.
This situation may be gradually changing due to the practise of
leasing of fishing licences which was formalized under the
Fisheries Act in 1994 (FRMA, 1994) and became more prevalent following the 1993/1994 management changes which resulted in a significant decrease in vessel numbers. As a result of this ongoing fleet
reduction process, many owners retiring their vessels from the
fishery now prefer to lease (rather than sell) their units so as to maintain an income stream and retain the option to re-enter the fishery in
the future. This trend to lease rather than sell units has continued
under the ITQ system and is expected to increase based on the
scope for further fleet reductions and increases in unit values
under this system. While many ex-fishers currently retain the
fishing rights through leasing, ultimately this process is likely to
result in more non-fisher ownership of the lobster fishing rights
through inter-generational transfers.
Discussion
Application of input and output-based management
to lobster fisheries
For input-based management, the catchability and behaviour of the
species being harvested can influence the directness of the relationship between the fishing effort and catch achieved and is a primary
consideration for management. While for most lobster fisheries including the case study fishery, short-term variations in catchability
related to moulting and lunar phase can be expected (Wright et al.,
2006); there is typically a reasonable relationship between annual
effort applied and catch achieved, such that input-based systems
can generally operate effectively. An exception to this is where
lobster species undergo mass migrations, e.g. P. cygnus (de Lestang
et al., 2012) and P. argus (Kanciruk and Herrnkind, 1978), and
during those events, lobsters can be highly aggregated, such that
the catches obtained are poorly related to the effort being applied
at that time. During these short periods, e.g. the “whites” migration
in the case study fishery (Melville-Smith et al., 1998), input-based
management is less effective in controlling catches and this was
one of the factors which influenced the case study fishery decision
to move to output-based management.
i30
J. W. Penn et al.
Figure 7. The average standardized annual settlement of P. cygnus pueruli on collectors along the coastline from Port Gregory in the north to Cape
Mentelle in the south. Note: before 1982/1983, only three puerulus collector sites were available in the centre of the fishery.
The degree of variability in annual catch available or “recruitment to the fishery” also affects the way the alternative management
control systems operate. For example, where there is high variability
in recruitment to the fishery, the setting of a precise catch quota
(TAC) is inherently more difficult and errors in the forecast of the
annual catch available are more likely. Further, as in the case study
fishery, variability in recruitment to the fishery can also be significant between regions, i.e. is much higher in the south (C Zone,
Figure 4), which further complicated management decision processes. Where these more variable recruitment to the fishery situations occur, input (effort-based) controls provide a lower risk
management strategy as the catch achieved tends to self-correct
when errors in stock abundance estimates occur. That is, with
effort and exploitation rate relatively fixed, the catch achieved
more closely reflects the underlying abundance, and leaves a relatively constant proportion of the stock after fishing each year.
Conversely, the application of output-based (catch) controls to fisheries with high catch variability is a higher risk strategy whenever
stock levels are overestimated, as uncontrolled effort can expand
to achieve a TAC set high and reduce residual stock with potential
impacts on breeding stocks.
For these reasons, tropical Panulirus lobster stocks with typically
higher variability in recruitment to the fishery (due to faster growth
and shorter life cycles often resulting in a single year class dominating the catch each year) are generally better suited to the “selfcorrecting” effort-based management systems. This is particularly
relevant where stocks are heavily fished or a strategy targeting
MSY is adopted, as was historically the situation in the case study
fishery.
Alternatively, the longer-lived, slower-growing lobster species
occurring in temperate to cold waters can be expected to have less
variable recruitment to the fishery (i.e. more year classes contribute
to each year’s catch) such that errors in assessing the stock levels
available for TAC setting purposes are less likely. Because such longlived species have lower natural mortalities, there is also little penalty
from setting conservative TACs and allowing fishable stock to carryover to future years. From this perspective, output-based management is likely to be more suitable for the cold-water fisheries for
the Homarus, Palinurus, and Jasus lobster groups. Although catch
quotas are also applied to sections of the European Nephrops trawl
fishery (Bell et al., 2013), this shorter-lived group is likely to be suitable for either form of management.
The third factor influencing the effectiveness of the management
strategy adopted is the ability to collect data from the fishery to
support the management decision processes. For example, in an
open access fishery situation, excess fishing capacity is likely to
have developed (Gardner et al., 2013) and there are often insufficient
fishery catch data for a manager to do other than implement some
basic input controls to stabilize the situation. In practical terms,
the initial use of input controls as occurred in the case study
fishery resulted in fishing effort applied consistently to the stock
and enabled a comprehensive time-series of low cost fisherydependent catch and effort data to be developed. The availability
of these data was critical to decisions to limit entry (Bowen and
Hancock, 1989) and later assessments of fishing power (Brown
et al., 1995) needed to adjust the effort levels. These data issues generally make input controls and fishery-dependent data gathering a
useful precursor to the implementation of output-based control
i31
Review of lobster fishery management
systems. Conversely, ITQ management designed to allow individual
vessels to fish to market demand usually leads to the inconsistent application of effort, highgrading, and the need for higher cost fisheryindependent survey data to support decision processes.
These issues suggest that input controls are more suitable for
most lobster stocks initially, particularly those with open access
and excess fishing capacity. Similarly, data-poor lobster fisheries
are likely to benefit from an initial input management phase as a
precursor to more sophisticated controls being considered.
Management implementation in lobster fisheries
A reason commonly given for lack of basic management in many
lobster fisheries is the limited community support and associated
difficulties in compliance. This is often associated with a lack of
ongoing fishing or exclusive rights (Gardner et al., 2013) which
would ensure that the fishers being asked to suffer short-term reductions in catch will be the beneficiaries in the longer term. Based on
the positive Mexican experience in applying management controls
in the 1950s, Seijo (1993) has noted that “a useful first step in
these situations is for governments to provide some basic, ongoing
community fishing rights at the community level”. These issues of
lack of support for management and illegal activity (taking of
undersize and berried lobsters) were also evident in the case study
fishery (de Lestang et al., 2012) through the 1960s, before exclusive
fishing rights were issued and continued until transferability was
well established.
While the lobster fisheries in the developed economies mostly
now have rights-based management in place with biological and
some effort controls, there has been limited uptake of the more
complex market-based management systems which have now developed, despite their potential to significantly improve economic performance (Gardner et al., 2013; Caputi et al., 2015). These more
complex fishery management systems have however mostly developed in single legal jurisdictions (e.g. the Australian states, New
Zealand) and their implementation in fisheries managed under
more complex federal and state legal systems or multinational jurisdictions is likely to be more difficult. Nonetheless, the management
effectiveness and improved economic outcomes under these control
systems suggest they have general applicability to most lobster
fisheries.
In many of the established lobster fisheries, the debate around
these more active management strategies continues to be socially
based and often relates to competition between fishing groups and
potential impacts on younger fishers wishing to enter commercial
fisheries (Acheson and Gardner, 2011). Fishing communities are
also concerned about loss of control over the fishery, creation of
windfall profits for historical operators, and the potential for aggregation of fishing rights to create monopoly ownership, particularly
under ITQ management (Copes and Charles, 2004). While some of
these concerns stem from local legal and social complexities, others
have not been borne out by the experience in the case study fishery,
particularly under the effort-control system.
While these issues were also initially of concern in the WA lobster
fishery (Bowen and Hancock, 1989), the concerns raised decreased
as the fishers began to understand that the increasing transfer values
reflected the underlying profitability of their “lobster fishing businesses” and they were able to borrow to fund unit purchases in
line with normal business practice. Further, under the TAE with
ITEs system, new operators wanting to enter the fishery would typically build up a small unit holding through purchases then lease
additional units sufficient to operate a small vessel. This process
was easier for skilled skippers, who could then achieve aboveaverage catches, a process not legally possible under an ITQ system.
This history of ownership of the fishing licences in the case study
fishery is also relevant to the wider issue of loss of fishing community
control over the fisheries (Copes and Charles, 2004). That is, the
transfer of licences to corporate bodies, apart from private fishing
family companies, has not been evident through the TAE management period, despite the obvious profitability demonstrated by
the P/E ratios being achieved. Historically, some of the processing
companies have entered this market by purchasing units for lease
back to active fishers in an attempt to guarantee their supply of lobsters; however, this did not become a significant activity under the
TAE system. The apparent reason for this is that a fisher using
leased pots does not have to legally land the product to the ITE
owner (Processor). That is, the Government does not need to directly track individual vessel catches or landings, as the only practical
compliance concern is to ensure the correct licensed fishing
vessels and gear are operating through the season. Second, the
major long-standing and most successful processor in the Western
Australian fishery has historically been a large fishers’ cooperative,
which distributes its profits back to fishers and has had no incentive
to accumulate fishing rights on its own behalf and compete with its
shareholders.
While the previous TAE with ITEs avoided many of the concerns
about new entrants and fishery ownership or control, the new ITQ
system now in place is not likely to have these same social characteristics. That is, the experience with ITQs generally is that transfer to
corporate ownership is more likely and leasing becomes common
place (Gardner et al., 2013). For these reasons, limits on aggregation
of ownership to prevent monopolies developing are typically
required. The experience in the New Zealand lobster fishery
(Donohue and Barker, 2000) is that despite a maximum quota
holding limit of 10%, fisher ownership diminished rapidly with corporate and processing companies controlling about half of the
quota, 10 years after the introduction of ITQs. The contracting
out of the catching of quotas has also led to concerns that fishers
on small margins are less likely to show custodianship of the resource compared with owner operators. In addition, new social
issues of greater risk taking at sea are now emerging in the
Tasmanian fishery linked to non-fishing quota ownership and
leasing of quotas (Emery et al., 2014). This practice of leasing
under ITQs also changes the relationship between quota owner
and a lessee, because the Government compliance system tracks
all catches and can ensure that all legal catches are landed to the
quota holder. In this situation, quota-owners (including former
fishers) naturally seek to maximize their return from their ownership of the fishing rights and lease fees can reach 50 –60% of the
value of the catch, in effect making the vessel operator an employee
skipper.
Performance of the effort-control system
The potential for effort increases under the early limited entry
system was recognized early in the WA fishery and measures to
reduce latent effort and control “effort creep” were instituted.
Through this period, the competitive nature of fishing sometimes
resulted in limited adherence to the fishing rules, with fishers repeatedly trying to capitalize on any loophole to increase their effort
(Bowen and Hancock, 1989). Further, their report noted “it is therefore imperative that a mechanism be established at the start of effort
limitation which sets out a method for fishing capacity reduction to
counterbalance the increases that inevitably occur”. A second factor
i32
that contributed to effort creep under the input control system was
that as the fishery became more profitable through the 1980s and
1990s, many competitive fishers invested in bigger and faster
vessels in an attempt to gain a greater share of the fishery output.
This process, known as “capital stuffing” (Gardner et al., 2013),
results in a “rush to fish” (Emery et al., 2013) in which all operators
are forced to compete or lose catch share. These larger more powerful vessels were able to achieve above-average catches by being able to
move their full pot holding to areas of high abundance daily. While
this factor was initially controlled to some extent by the rule that
vessels had to have between 7 and 10 pots m21 of vessel length to
operate, it was later removed after political pressure from affected
industry operators. While the capital stuffing in the case study
fishery has largely been eliminated by the change to ITQs, the previous controls linking gear entitlements to vessel size remain relevant
to other lobster fisheries using effort-based management systems.
While a series of adjustments to the fishery were made to overcome “effort creep” through the early development period, the
first major test of the capacity of the limited-entry regime to actively
manage the stock occurred when exploitation rates in the early
1990s (Wright et al., 2006) had reached high levels (70–75% annually) following the introduction of electronic technology (including
GPS plotters, etc.). This caused significant depletion of breeding
stocks, possibly below 20% of virgin levels (Hall and Chubb,
2001) and raised the possibility of recruitment overfishing.
Despite the potentially high risks to the fishery, there was intense industry opposition to the planned management changes to recover
breeding stock levels. While the package of new management measures in 1993/1994 proved successful and licence values quickly
recovered, this experience illustrated the difficulties associated
with achieving industry consensus in large fisheries when significant
effort adjustments are required. In contrast, the more radical effort
adjustments made in response to the puerulus settlement decline
starting in 2007/2008 required similarly difficult negotiations, but
were implemented quickly, with reasonable industry leadership
support, due to the concerns about heavy fishing pressure continuing on the projected record-low stock levels.
These difficulties in making effective management changes
under effort controls when high levels of latent effort and significant
excess fishing capacity exist are likely to apply to many other effortmanaged commercial lobster fisheries and in most artisanal situations. The long-established Homarus fisheries in the United States,
where catches and fishing effort have continually increased over a
long period but are now suffering decreased recruitment and survival in some areas under climate change (Wahle et al., 2013), are
also likely to be in the above situation. In these situations, the case
study experience strongly supports the need in all input control fisheries for automatic effort adjustments, so that management action
to deal with significant changes in stock abundance is not delayed
by having to first remove excess capacity.
Benefits from conversion to TAC with ITQ management
The case study experience from moving to ITQs is that this system
can provide significant additional increases in economic performance. This occurred through the virtual elimination of the rush to
fish, which reduced vessel operating costs and considerably
improved the prices being achieved for lobsters (Caputi et al.,
2014b, 2015). This was illustrated by the P/E valuations which
were low at the time of TAC (ITQ) implementation, but recovered
quickly back to the levels previously achieved under TAE system
before the puerulus settlement collapse.
J. W. Penn et al.
This experience suggests that many long-standing lobster fisheries that have stable catches, particularly if they can predict recruitment, could achieve much better economic performance under an
ITQ system without compromising sustainability. However, such
conversions from basic limited entry to more market-based
system with ITEs (or ITQs) in fisheries and particularly where
there is considerable excess fishing capacity can be expected to
result in significant reductions in vessel numbers. Because these
changes and the expected increase in licence values are likely to
have significant socio-economic consequences for fishing communities and affect the future ownership of the fishery rights, the legislation to make these changes needs to anticipate and address these
issues.
Lobster fishing rights and management
The creation of exclusive fishing rights in the WA lobster fishery
flowed from limiting entry were ultimately formalized under the
Fisheries Act (FRMA, 1994) and enhanced by the licence register
where contract arrangements around ownership could be recorded.
Notably, these “access rights” and associated administrative systems
developed for the WA commercial fishery have most of the traits
suggested for fishing rights by Gardner et al. (2013) and led to
increasing industry support in the form of lower compliance and
research costs which have improved management outcomes
(Bowen, 1994).
While the nature of the commercial access rights are specific to
the Western Australian legal system, it has been noted that they
are in effect very similar to the franchisee rights under franchise contract (Penn et al., 1997) which operate successfully in most countries
where lobsters are fished. That is, the franchisee contract arrangements have a long legal history and because they operate in a wide
variety of jurisdictions can provide a useful starting point where
improved fishing rights are proposed, but locally contentious.
A key aspect of franchisee “contracts” is that there are typically provisions for buyback of the individual enterprises, usually at market
value, under this system. This aspect of the franchise contract equating to the “obligations” of Government after the fishing rights are
issued is not often included directly in Fisheries management plans
(Penn et al., 1997), but would better protect all fisher’s rights if significant restructuring of fisheries is needed to deal with common
issues such as catch share changes between sectors (Gardner et al.,
2013) or stock shifts under climate change. Second, while the franchise contracts usually have buy-back provisions, the secondary
leasing of the operator rights is not generally a feature of these management systems. Noting that leasing is expected to increasingly
transfer fishery income away from fishing communities and can
have other negative social impacts (Emery et al., 2013), the incorporation of the franchise contract principles in fisheries legislation may
provide an effective legal basis to control these potentially negative
effects and provide better outcomes for the wider community as
the ultimate “owners” of the lobster resources. In addition to the
franchisee contract matching commercial lobster fishing access
rights, the model is also relevant to defining the rights of other
fishers (sport and artisanal) and conservation groups seeking catch
shares for inclusion in “not take” areas.
Conclusions
The review suggests that early adoption of tradable “access rights”
rights similar to franchisee contracts in the case study fishery was a
critical first step in gaining fisher support for management and the
licence values created provide a useful indicator of fishery viability.
Review of lobster fishery management
The historical process where basic biological controls are
enhanced with limited entry before developing sophisticated ITE
or ITQ-based systems appears an effective and efficient pathway
for fishery development with wider relevance. The case study experience has however shown that ITE and ITQ-based systems are inherently different in their operation and socio-economic consequences.
The ITE system appears to avoid many of the negative consequences
often attributed to ITQs and suggests that they could be usefully
applied where fishing community concerns are impeding management implementation. Notably, both ITE and ITQ systems have
proved capable of significant fleet reductions to overcome the
excess fishing capacity without the need for government financial
involvement. However, the review shows that ITQ management
has additional economic benefits, but involves a higher risk strategy
to be applied with caution, where fisheries are operating at high
levels of exploitation or stock estimates are unreliable.
While the case study experience with ITE and particularly
ITQ-based management suggests that both are capable of generating
significant economic benefits, the longer term impacts of these
control systems on ownership of the fishing rights, the flow of benefits from fishing, and the resulting manageability of lobster stocks
are questions yet to be answered.
Acknowledgements
The authors thank Dr Alex Hesp, Mr M. Rossbach, and Mr E. Barker
for their help with the licence information and analysis, and contribution to the many discussions about the management of the WA
lobster fishery. Mrs J. Moore provided much appreciated assistance
with the manuscript preparation.
References
Acheson, J., and Gardner, R. 2011. The evolution of the Maine lobster
V-notch practice: cooperation in a prisoner’s dilemma game.
Ecology and Society, 16: 41.
Anon. 1980. Offshore constitutional settlement: a milestone in cooperative federalism. Attorney-General’s Department, Australian Government Publishing Service, Canberra.
Bell, M., Tuck, I., and Dobby, H. 2013. Nephrops species. In Lobsters:
Biology, Management, Aquaculture & Fisheries, 2nd edn, pp.
357– 413. Ed. by B. F. Phillips. Wiley-Blackwell, Oxford, UK.
Bowen, B. K. 1980. Spiny lobster fisheries management. In The Biology
and Management of Lobsters, Vol. 2 Ecology and Management, pp.
243– 264. Ed. by J. S. Cobb, and B. F. Phillips. Academic Press,
New York.
Bowen, B. K. 1994. Long term management strategies for the Western
Rock Lobster Fishery (4 volumes). Evaluation of management
options Vol.1, Fisheries Management Paper No. 66, Fisheries WA,
Perth, Australia. 67 pp.
Bowen, B. K., and Hancock, D. A. 1989. Effort limitation in the
Australian rock lobster fisheries. In Scientific Approaches to
Management of Shellfish Resources, pp. 375– 396. Ed. by J. Caddy.
John Wiley and Sons, New York.
Briones-Fourzan, P., and Lozano-Alvarez, E. 2000. The spiny lobster
fisheries in Mexico. In Spiny Lobsters: Fisheries and Culture, 2nd
edn, pp. 169– 188. Ed. by B. F. Phillips, and J. Kittaka. Fishing
News Books, Oxford, UK.
Brown, R. S. 1991. A decade (1980– 1990) of research and management
of the western rock lobster (Panulirus cygnus) fishery of Western
Australia. Revista de Investigaciones Marinas (Cuba), 12: 204 – 222.
Brown, R. S., and Caputi, N. 1986. Conservation and recruitment of the
western rock lobster (Panulirus cygnus) by improving survival and
growth of undersize rock lobsters captured and returned by fishermen to the sea. Canadian Journal of Fisheries and Aquatic
Sciences, 43: 2236– 2242.
i33
Brown, R. S., Caputi, N., and Barker, E. 1995. A preliminary assessment
of increase in fishing power on stock assessment and fishing effort
expended in the western rock lobster, Panulirus cygnus, fishery.
Crustaceana, 68: 227 – 237.
Campbell, J. Y., and Shiller, R. J. 1988. Stock prices, earnings, and
expected dividends. Journal of Finance, 43: 661 –676.
Caputi, N., de Lestang, S., Hart, A., Kangas, M., Johnston, D., and Penn,
J. 2014a. Catch predictions in stock assessment and management of
invertebrate fisheries using pre-recruit abundance; case studies from
Western Australia. Reviews in Fisheries Science, 22: 36 – 54.
Caputi, N., de Lestang, S., Reid, C., Hesp, A., and How, J. 2014b.
Decision-support tools for economic optimization of western rock
lobster fishery of Western Australia. Fisheries Research Report,
257, Department of Fisheries, Western Australia. 96 pp.
Caputi, N., de Lestang, S., Reid, C., Hesp, A., and How, J. 2015.
Maximum economic yield of the western rock lobster fishery of
Western Australia after moving from effort to quota control.
Marine Policy, 51: 452 – 464.
Caputi, N., Feng, M., de Lestang, S., Denham, A., Penn, J., Slawinski, D.,
and Pearce, A. 2014c. Identifying factors affecting the low western
rock lobster puerulus settlement in recent years. Final FRDC
Report—Project 2009/18. Fisheries Research Report, 255.
Department of Fisheries, Western Australia. 144 pp.
Copes, P., and Charles, A. 2004. Socio-economics of individual transferable quotas and community-based fishery management.
Agricultural and Resource Economics Review, 33: 171– 181.
de Lestang, S., Caputi, N., Feng, M., Denham, A., Penn, J., Slawinski, D.,
Pearce, A., et al. 2015. What caused seven consecutive years of
low puerulus settlement in the western rock lobster fishery of
Western Australia? ICES Journal of Marine Science, 72(Suppl. 1):
i49– i58.
de Lestang, S., Caputi, N., How, J., Melville-Smith, R., Thomson, A., and
Stephenson, P. 2012. Stock assessment for the west coast rock lobster
fishery. Fisheries Research Report, 217. Department of Fisheries,
Western Australia. 200 pp.
de Lestang, S., Thompson, A., and Rossbach, M. 2010. West coast rock
lobster fishery status report. In State of the Fisheries and Aquatic
Resources Report 2009/10, pp. 28 – 38. Ed. by W. J. Fletcher, and
K. Santoro. Department of Fisheries, Western Australia.
Department of Fisheries. 2011. State of the Fisheries and Aquatic
Resources Report 2010/11. Ed. by W. J. Fletcher, and K. Santoro.
Department of Fisheries, Western Australia. 359 pp.
Donohue, K., and Barker, E. 2000. Information on Quota Management
of Rock Lobster Fisheries in South Australia, Tasmania and New
Zealand. Fisheries Management Paper, 138, Department of
Fisheries, Western Australia. 60 pp.
Emery, T. J., Hartmann, K., Green, B. S., Gardner, C., and Tisdell, J. 2013.
Does “race to fish” behaviour emerge in an individual transferable
quota fishery when the total allowable catch becomes non-binding?
Fish and Fisheries, 15: 151– 169.
Emery, T. J., Hartmann, K., Green, B. S., Gardner, C., and Tisdell, J. 2014.
Fishing for revenue: how leasing quota can be hazardous to your
health. ICES Journal of Marine Science, 71: 1854– 1865.
FAO. 2011. Fisheries and Aquaculture Information and Statistical
Service—2009. http://www.fao.org/fishery/statistics/globalproduction/query/en (date last accessed, August 2011).
FRMA. 1994. Fisheries Resources Management Act. 1994. Government
of Western Australia. 226 pp. www.slp.wa.gov.au.
Gardner, C., Larkin, S., and Seijo, J. C. 2013. Systems to maximize
economic benefits in lobster fisheries. In Lobsters: Biology,
Management, Aquaculture & Fisheries, 2nd edn, pp. 113– 138. Ed.
by B. F. Phillips. Wiley-Blackwell Publishing, Oxford, UK.
Hall, N., and Brown, R. S. 2000. Modelling for management: the western
rock lobster fishery. In Spiny Lobsters: Fisheries and Culture, 2nd
edn, pp. 386– 399. Ed. by B. F. Phillips, and J. Kittaka. Fishing
News Books, Oxford, UK.
i34
Hall, N., and Chubb, C. 2001. The status of the western rock lobster
Panulirus cygnus fishery and the effectiveness of management controls in increasing the egg production of the stock. Marine and
Freshwater Research, 52: 1657 –1667.
Kanciruk, P., and Herrnkind, W. 1978. Mass migration of spiny lobster,
Panulirus argus (Crustacea: Palinuridae): behaviour and environmental correlates. Bulletin of Marine Science, 28: 601– 623.
Melville-Smith, R., Chubb, C., Caputi, N., Cheng, Y. W.,
Christianopoulos, D., and Rossbach, M. 1998. Fishery independent
survey of the breeding stock and migration of the western rock
lobster (Panulirus cygnus). Final Report, Fisheries Research and
Development Corporation Project 96/108, Canberra. 47 pp.
Melville-Smith, R., Thomson, A., and Caputi, N. 2004. Improved forecasts of recreational western rock lobster (Panulirus cygnus) catches
in Western Australia, by predicting license usage. Fisheries Research,
68: 203 –208.
Melville-Smith, R., and van Sittert, L. 2005. Historical commercial West
Coast rock lobster Jasus lalandii landings in South African waters.
African Journal of Marine Science, 27: 33 – 44.
Penn, J. W., Morgan, G. R., and Millington, P. J. 1997. Franchising fisheries resources: an alternative model for defining access rights in
Western Australian fisheries. In Developing and Sustaining World
Fisheries Resources: the State of Science and Management,
pp. 383– 390. Ed. by D. A. Hancock, D. C. Smith, A. Grant, and
J. W. Penn et al.
J. P. Beumer. 2nd World Fisheries Congress Proceedings, Brisbane,
Australia, July 1996.
Phillips, B. F. (Ed.). 2013. Lobsters: Biology, Management, Aquaculture
& Fisheries, 2nd edn. Wiley-Blackwell Publishing, Oxford, UK.
Phillips, B. F., Melville-Smith, R., Kay, M. C., and Vega-Valazquez, A.
2013. Panulirus species. In Lobsters: Biology, Management,
Aquaculture & Fisheries, 2nd edn, pp. 289– 325. Ed. by B. F.
Phillips. Wiley-Blackwell, Oxford, UK.
Reid, C. R., Caputi, N., de Lestang, S., and Stephenson, P. 2013.
Assessing the effects of moving to maximum economic yield effort
level in the western rock lobster fishery of Western Australia.
Marine Policy, 39: 303 – 313.
Seijo, J. C. 1993. Individually transferable grounds in a community
managed artisanal fishery. Marine Resource Economics, 25:
281– 293.
Wahle, R., Castro, K., Tully, O., and Cobb, S. 2013. Homarus. In
Lobsters: Biology, Management, Aquaculture & Fisheries, 2nd
edn, pp. 221 – 247. Ed. by B. F. Phillips. Wiley-Blackwell, Oxford,
UK.
Wright, I., Caputi, N., and Penn, J. 2006. Depletion -based population
estimates for western rock lobster (Panulirus cygnus) fishery in
Western Australia. New Zealand Journal of Marine and Freshwater
Research, 40: 107– 122.
Handling editor: Sarah Kraak
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i35 –i48. doi:10.1093/icesjms/fsu196
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Review
European lobster stocking requires comprehensive impact
assessment to determine fishery benefits
Charlie D. Ellis 1,2*, David J. Hodgson 3, Carly L. Daniels 2, Dominic P. Boothroyd 2, R. Colin A. Bannister 4,
and Amber G. F. Griffiths1
1
Environment and Sustainability Institute, University of Exeter, Penryn Campus, Penryn, Cornwall TR10 9FE, UK
National Lobster Hatchery, South Quay, Padstow, Cornwall PL28 8BL, UK
3
Centre for Ecology and Conservation, University of Exeter, Penryn Campus, Penryn, Cornwall TR10 9FE, UK
4
Centre for Environment, Fisheries and Aquaculture Science, Lowestoft, Suffolk NR33 0HT, UK
2
*Corresponding author: tel: +44 7940 316348; fax: +44 2326 254243; e-mail: [email protected]
Ellis, C. D., Hodgson, D. J., Daniels, C. L., Boothroyd, D. P., Bannister, R. C. A., and Griffiths, A. G. F. European lobster stocking
requires comprehensive impact assessment to determine fishery benefits. – ICES Journal of Marine Science, 72: i35 – i48.
Received 15 July 2014; revised 13 October 2014; accepted 14 October 2014; advance access publication 18 November 2014.
Historically, hatcheries in Europe and North America attempted to contribute to the conservation and enhancement of clawed lobster stocks, but
lacked monitoring programmes capable of assessing success. In the 1990s, this perspective was changed by the results of restocking and stock
enhancement experiments that inserted microwire tags into hatchery-reared juvenile European lobsters (Homarus gammarus) before release.
This allowed recapture in sufficient numbers to prove that lobsters had survived and recruited to the mature fishable stock. However, evidence
of recruitment still failed to answer key questions about the ultimate ecological and economic benefits. As a result, a growing number of
lobster stocking ventures remain hindered by a lack of clear evidence of the effects of their stocking schemes. This review evaluates these experiments and related studies on other fished species, summarizes key findings, and identifies data and knowledge gaps. Although studies of fitness in
cultured lobsters provide some of the most encouraging results from the wider field of hatchery-based stocking, the limitations of physical tagging
technology have significantly hindered appraisals of stocking impacts. We lack basic knowledge of lobster ecology and population dynamics,
especially among prerecruits, and of the impact of stocking on wild lobster population genetics. We advocate the use of genetic methods to
further our understanding of population structure, rearing processes, and stocking success. We also recommend that more focused and comprehensive impact assessments are required to provide a robust endorsement or rejection of stocking as a viable tool for the sustainable management
of lobster fisheries.
Keywords: crustacea, genetics, hatchery, Homarus gammarus, mark – recapture, population structure, restocking, stock enhancement,
tagging.
Introduction
Capture fisheries make crucial contributions to the world’s wellbeing and prosperity. The global value of fisheries was estimated
at over E65 billion per annum in 2010, ca. 10% of the world’s population are dependent on fish-related jobs, and seafood products
are a vital source of protein and micronutrients for 3 billion
people (Fisheries and Aquaculture Department, 2014). Commonly,
however, conventional management fails to prevent the overexploitation of stocks. Interventions that use hatchery technology to
improve or re-establish the productivity and sustainability of
# International
capture fisheries, which can be categorized as “stocking”, are, therefore, worth considering. For many aquatic species, the survival of
juveniles in aquaculture facilities is several orders of magnitude
higher than in the wild, allowing increased recruitment above
natural levels (Lorenzen, 2005). Stocking schemes aim to improve
and sustain capture fisheries and are categorized as either “restocking”
(the release of cultured juveniles to restore spawning biomass) or
“stock enhancement” (the recurrent release of cultured juveniles
to overcome recruitment limitations) (Bell et al., 2006). Lorenzen
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i36
(2008) advocates that aquaculture-based enhancement of stocks
ranks alongside regulation of fishing effort and restoration of key
habitats as a principal means by which wild fisheries can be sustained
and improved.
With many capture fisheries under intense pressure, aquaculture
technologies have become an increasingly important means of
seafood production, largely through the full grow out of marketable
fish, but also by restocking and stock enhancement of wild populations. Hatchery stocking is undertaken worldwide and has been
most successful in large-scale schemes coordinated and funded by
government or industry. For example, the government-financed
programme in Japan alone involves the enhancement of .80
marine species (Kitada, 1999) and is estimated to account for 90%
of the chum salmon (Oncorhynchus keta) fishery, 50% of the
kuruma prawn (Penaeus japonicus) and red sea bream (Pagrus
major) catch, 30% of the flounder (Paralichthys olivaceus), and
almost all the scallop harvest (Kitada et al., 1992; Kitada and
Kishino, 2006). However, the contribution of stock enhancement
to global fisheries production has remained small (2%), and few
case studies have been declared outright successes (Lorenzen,
2008). Overall, the available literature appraising the impact of
stocking is heavily biased towards certain finfish; Araki and
Schmid (2010) found that 62% of genetic-based stocking impact
studies evidenced salmonids, flatfish, and bream, despite these
groups accounting for only 5% of the catch tonnage of enhanced
fisheries.
For many years, the progress of stocking enterprises was hindered by a lack of appropriate research into wild life histories
and by a lack of effective methods for distinguishing released individuals from wild conspecifics. As a result, robust evaluation of the
economic and ecological benefits of stocking has been impeded,
restricting impetus within the industry. Extensive knowledge of
the ecosystem, species biology, and population-specific data is
required for the design of successful stocking programmes. For
example, of eight species across a variety of taxa cultured in
Japan reviewed by Kitada (1999), six showed significant variation
in the effectiveness of stocking with differing release locations and/
or release densities. The method, timing, and recipient habitat of
releases and the density, size, and conditioning of released
animals can all have significant effects on survivability (van der
Meeren, 2000; Ball et al., 2001; Stunz and Minello, 2001; Svåsand
et al., 2004; Leber et al., 2005; Hamasaki and Kitada, 2008a;
Ochwada-Doyle et al., 2010).
The focus on this review is the European lobster (Homarus
gammarus L.), an ecologically and economically important decapod
crustacean ranging from northern Norway to Morocco and the
eastern Mediterranean (Triantafyllidis et al., 2005). Global catches
of the European lobster have been increasing since the 1980s, with
recent recorded pot-caught landings reaching 5913 t in 2011
(Fisheries and Aquaculture Department, 2014). Compared with
many finfish or the recent very large landings of the American
lobster (Homarus americanus) in North America [e.g. 50 000 t in
Maine (Steneck and Wahle, 2013)], European lobster landings are
small and come from sparse stocks. The species is of very high
value, however, fetching an average market price of E12.50 kg21
at the time of writing (Fish Information and Services, 2014).
Therefore, lobster populations are disproportionately important
to local fishing communities and regional economies as well as fulfilling key roles in the maintenance of healthy and diverse marine
ecosystems (Mann and Breen, 1972; Breen and Mann, 1976).
Aquaculture-based augmentation of wild Homarid lobster
C. D. Ellis et al.
populations has been attempted on both sides of the North
Atlantic for over 150 years using many release strategies and life
history stages (Nicosia and Lavalli, 1999). Because enhancement
of existing populations was difficult to identify, few of these experiments have been assessed in terms of benefits to fisheries (Addison
and Bannister, 1994; Nicosia and Lavalli, 1999). Lobster hatcheries
have provided most of the recorded information on clawed lobster
life history (Nicosia and Lavalli, 1999), but significant voids still
exist in our understanding of the species’ basic ecology.
The basic technology to rear lobsters through the planktonic
phases has long been available. This lifestage is presumed to be an
important recruitment bottleneck due to predation in the wild
(Richards and Wickins, 1979; Bannister and Addison, 1998).
However, efforts to trial the stocking of lobsters were renewed in
Europe throughout the 1980s–1990s in response to three key
drivers. First was a severe collapse of the fishery throughout
Scandinavia from 1930 to 1970 due to overexploitation and inadequate management, which saw landings decline 99% in
Denmark, 92% in Norway, and 90% in Sweden, all but wiping out
a once-thriving export commodity (Dow, 1980; Agnalt et al.,
1999; Fisheries and Aquaculture Department, 2014). This led to
aspirations to restock depleted populations as well as to enhance
stocks where uncapped potting effort rose in response to new continental export opportunities, such as the United Kingdom
(Bannister, 1986). Second, it was demonstrated that hatcheryreared lobsters acquired benthic, shelter-seeking behaviours
(Cobb, 1971; Cooper and Uzmann, 1980; Botero and Atema,
1982) that might decrease their vulnerability to wild predators
and hence improve survival (Howard, 1980, 1988). Third, the development of coded microwire tagging (CWT) technology (Jefferts
et al., 1983) allowed cultured juvenile lobsters to be distinguished
from wild conspecifics after release (Wickins et al., 1986;
Bannister and Addison, 1998).
Experimental lobster stock enhancement programmes were
launched to release large numbers of juvenile lobsters onto known
lobster grounds at a range of sites in France (Henocque, 1983;
Latrouite and Lorec, 1991), the United Kingdom (Burton, 1992;
Bannister et al., 1994; Cook, 1995), and Norway (Agnalt et al.,
1999, 2004; Agnalt, 2008). Coded microwire tags were inserted
into late-stage juveniles before release, and their recapture provided
the first definitive evidence that cultured lobsters were able to
survive in the wild. In both UK stock enhancement and
Norwegian restocking trials, cultured lobsters were shown to
attain adult sizes (Bannister et al., 1994; Agnalt et al., 1999) and
add to spawning –stock biomass (Bannister et al., 1994; Agnalt,
2008). Restocking also showed that released lobsters could
augment rather than simply displace natural stocks (Agnalt et al.,
1999, 2004). Although most of these studies declared the renewed
lobster stocking efforts as tentatively successful, it was also proposed
that production costs and lobster market values did not make the
observed recapture rates economically viable (Whitmarsh, 1994;
Moksness et al., 1998).
In this review, we summarize current practices in lobster stocking, and reappraise the measurement of stocking success and the
practices of monitored stocking trials. We then highlight critical issues for lobster stocking, including hatchery production
methods, understanding the ecology of lobsters in the wild to optimize success of released lobsters, and genetic considerations.
Finally, we address the problem of comparing stocking with alternative management strategies and conclude by suggesting future
research directions and hatchery protocols.
i37
Lobster stocking requires impact assessment
Hatchery rearing of European lobsters
The rationale for current European lobster cultivation is typical of
hatchery enterprises. Fishery stakeholders are attracted to stocking
where other management options are limited or unappealing.
Intensive developments in husbandry, infrastructure, and stakeholder engagement are required to establish a lobster hatchery,
and significant gaps remain in our understanding of aspects of the
biology and ecology of H. gammarus. Nevertheless, severe stock
depletions, high market value, and well-functioning rearing technology continue to encourage new lobster stocking efforts in
Europe (Svåsand et al., 2004).
Female lobsters, bearing eggs fertilized naturally in the wild, are
typically bought or loaned from fishers or merchants and are held
until the larvae have hatched. Larvae are normally reared communally through the planktonic lifestages (Zoea larval stages I –III
and post-larval stage IV) in tapered hoppers or Hughes/Kreisel
cones in which upwelling air and/or water reduces settling and cannibalization (Richards and Wickins, 1979; Beard et al., 1985;
Grimsen et al., 1987; Beard and Wickins, 1992; Burton, 1992;
Cook, 1995; Nicosia and Lavalli, 1999; Daniels et al., 2010).
Survival of the planktonic phase is highly sensitive and variable
even in the captive environment, and although individual batches
may attain survival .50%, typically 10– 15% of stage I larvae
reach the onset of benthic behaviours a few days after moulting to
stage IV (Burton, 1992; Nicosia and Lavalli, 1999; Daniels et al.,
2010). The absence of interspecific predation suggests that cultured
larval survival is likely to far exceed that of wild larvae, although the
scarcity with which wild conspecifics are found (Nichols and
Lovewell, 1987) means that no reliable estimates of natural survival
exist for comparison. Once they attain stage IV, post-larvae have a
much greater swimming ability and are generally then separated
into individual holding compartments for on-growing before
being released into wild environments at an early benthic juvenile
phase.
Over 1.4 million cultured juvenile European lobsters have been
released by known stocking programmes between 1983 and 2013.
Of these releases, 90% can be classified as stock enhancement of
existing commercial fisheries around the United Kingdom,
Ireland, and France, and 10% as restocking heavily depleted populations in Norway, Germany, and Italy (Table 1). Approximately
255 000 released lobsters (mostly in Norway and the United
Kingdom) were grown on to late juvenile stages [12 –21 mm
carapace length (CL), Latrouite and Lorec, 1991; Burton, 1992;
Cook, 1995; Bannister and Addison, 1998; Agnalt et al., 1999;
Schmalenbach et al., 2011] and tagged to allow wild survival to be
monitored. More recently, stock enhancement programmes in
Orkney, Scotland, and Cornwall, England, and restocking trials in
Lazio, Italy, have released some 900 000 untagged juveniles at earlier lifestages (stage V+, .5 mm CL; D. Shearer and G. Nascetti,
pers. comm.).
Assessments of lobster stocking success
Monitored stocking trials
Long after the development of the requisite technology to rear lobsters through the larval phases for release as juveniles, the success of
early stocking programmes still could not be formally evaluated
(Addison and Bannister, 1994). Ecdysis (exoskeletal moulting) precludes the use of externally fixed markers in lobsters, particularly
juveniles which moult frequently. As a result, there was no lasting
method to discriminate between hatchery-reared and wild individuals. Whether released animals survived and actually enhanced
natural stocks (instead of displacing them) were unproven, proponents of stocking were unable to demonstrate whether the method
provided any benefits to fisheries (Addison and Bannister, 1994).
Flawed attempts to recognize recaptured hatchery-reared individuals led to the trial release of 1300 H. gammarus × H. americanus
hybrid juveniles in France during the 1970s (Latrouite and Lorec,
1991), despite no evidence of their ecological suitability and the
Table 1. Summary of major and/or widely reported stock enhancement projects for European lobsters 1972– 2013.
Location (hatchery—area)
France (Ile de Sein; Ile d’Yeu; Ile
de Houat)
France (Ile de Sein; Ile d’Yeu; Ile
de Houat)
United Kingdom
(MAFF—Bridlington;
NWSFC—Aberystwyth;
SFIA—Ardtoe; Orkney)
France (Ile de Sein; Ile d’Yeu; Ile
de Houat)
Norway (Kvitsøy)
Ireland (Galway; Wexford)
Germany (Helgoland)
United Kingdom
(OSFH—Orkney)
United Kingdom
(NLH—Cornwall)
Italy (CISMAR—Viterbo)
Total
a
Release
Monitoring Release
years
years
age/stage
1972 –1977
–
Stage 5–1
year
1978 –1983 1980 –1983 1 year
Number
released
265 000
Number Recapture ratio
recapture (% recaptured) Source reference
–
–
Henocque (1983)
1300b
0
Latrouite and Lorec (1991)
1 : 62 (1.6%)
1 : 1 158 (0.1%)
Bannister et al. (1994);
Cook (1995); Burton
(1993); Bannister and
Addison (1998)
Latrouite and Lorec (1991)
1983 –1990 1985 –1994
1 year
90 925a
1471
1984 –1987 1987 –1989
1 year
25 480a
22
1990 –1994 1992 –2001
1993 –1997
–
2000 –2005 2001 –2009
2000 –2013
–
1 year
127 945a
Stages 4–5 292 000
1 year
5400a
Stages 4–10 747 000
7950
–
487
–
1 : 16 (6.2%)
–
1 : 11 (9.0%)
–
Agnalt et al. (2004)
Browne and Mercer (1998)
Schmalenbach et al. (2011)
D. Shearer, pers. comm.
2002 –2013
Stages 5–10 150 000
–
–
This paper
–
9930
–
1 : 25 (4.0%)
G. Nascetti, pers. comm.
–
–
2010 –2013
–
1983 –2013 1985 –2009
Stage 4+
Stage 4–
1 year
10 000
1 714 947
(249 750
tagged)
Tagged.
Homarus gammarus × H. americanus hybrids; “phenotypically marked”, but omitted from tagged release total.
b
–
i38
scheme relying on local fishers identifying precise morphological
variations in surviving hybrids. Extensive interannual fluctuations
in landings inhibited the usefulness of fishery capture statistics in
quantifying stocking success (Le Gall et al., 1983), but the advent
of the first suitable internal tagging methods in the early 1980s
encouraged three groups in France, Norway, and the United
Kingdom to commit significant resources to new experimental
stocking programmes (Bannister and Addison, 1998). These projects (Table 1, entries 3 –5) reared and released in all 244 350 latestage juveniles. The insertion of magnetized, batch-coded CWTs
offered the prospect of detecting survivors and evaluating the contribution of stocking to fisheries. Since these experiments concluded
in the late 1980s or early 1990s, only one further scientific assessment
of H. gammarus stocking has been reported: 5400 one-year-old
lobsters were tagged with visible implant elastomers (VIEs—
Uglem et al., 1996) and released during 2000–2005 on the German
island of Helgoland (Schmalenbach et al., 2011). Monitoring of these
projects has enabled the identification of cultured lobsters upon recapture several years after wild release, and currently provides all the
data available with which to assess the effectiveness of lobster stocking in Europe (Latrouite and Lorec, 1991; Burton, 1992; Bannister
et al., 1994; Cook, 1995; Bannister and Addison, 1998; Agnalt
et al., 1999; Agnalt et al., 2004; Schmalenbach et al., 2011).
There were many differences of detail in relation to release sites
and methods, local fishing effort and legislations, and monitoring
patterns both within and among the groups undertaking European
stocking trials. However, hatchery rearing protocols were largely
shared and, with little information about the habitat requirements
of prerecruit lobsters, all groups released juveniles into areas populated by adults. Release numbers were maximized but dispersed in
relatively small batches to reduce potential competitive interactions.
Each group released a succession of annual juvenile cohorts over 4
or more years and, usually, monitored stocks and landings for at
least a comparable period to estimate survival and the proportion
of tagged lobsters in the fishable stock (Latrouite and Lorec, 1991;
Burton, 1992; Bannister et al., 1994; Cook, 1995; Bannister and
Addison, 1998; Agnalt et al., 1999, 2004; Schmalenbach et al., 2011).
In France, Norway, and the United Kingdom, recaptured lobsters
fitted with CWTs were detected using magnetic detectors on board
potting vessels or at quayside landing stations (Bannister and
Addison, 1998), while VIE-tagged lobsters in Germany were identified visually by fishers and divers (Schmalenbach et al., 2011). The
recapture profiles of release cohorts typically illustrated common
sequences of growth, accumulation, and decay over the monitoring
period. Annual recaptures were largely on the scale of tens to hundreds, cumulating to a total of 9930 individuals across all monitored
projects, mostly recaptured 3 –10 years after release as subadults or
adults in the size range 50 –120 mm CL (Burton, 1992; Bannister
et al., 1994; Cook, 1995; Bannister and Addison, 1998; Agnalt
et al., 1999, 2004; Schmalenbach et al., 2011). Released lobsters generally showed high site fidelity (e.g. recaptured within 6 km of
release sites; Bannister and Howard, 1991), and many of the adult
females carried fertilized eggs, although whether these were sired
by wild or cultivated males was not assessed.
Recapture rates
Monitoring of hatchery-reared European lobster recruitment has
shown that releases in the order of 100 tagged juveniles have typically
yielded single-figured numbers of recaptures (Table 1; Bannister
and Addison, 1998; Agnalt et al., 2004; Schmalenbach et al.,
2011). These nominal recovery rates were regarded as indicators
C. D. Ellis et al.
of the potential contribution to the local fishery, but also of the potential economic rates of return (Whitmarsh, 1994).
Stocking trials in France provided the least encouraging total recapture figures (Table 1), although these results can be somewhat
discounted due to deficiencies in their monitoring programmes.
Although CWTs were implanted into 24 500 juveniles released
around the French Atlantic coast during 1984–1987, monitoring
began 2 years after the first releases, but lasted only 3 years
(Latrouite and Lorec, 1991). Only 22 lobsters were recaptured, but
the maximum recapture window (2–5 years for different release
cohorts) appears insufficient in light of the recapture profiles of
later trials elsewhere. At that same time in the United Kingdom,
almost 91 000 one-year-old juveniles were tagged and released in
four areas—Bridlington in England, Aberystwyth in Wales, and
Ardtoe and Orkney in Scotland—where the natural stocks were
depleted (though still more abundant than in Norway). Total
recaptures were 1471 over the 5- to 8-year monitoring period,
with the regional recovery rates ranging from 1.3 to 2.4% (Bannister
and Addison, 1998).
Higher recapture results came several years later from the
heavily depleted lobster stock in the Norwegian archipelago of
Kvitsøy. By 2001, 6.2% of the 128 000 coded-wiretagged year-old
juveniles released during 1990 – 1994 had been recaptured, and
released lobsters outnumbered wild conspecifics among the legalsized catch (Agnalt et al., 2004). Importantly, both the proportion
of hatchery-reared lobsters in the fishable stock and catch per unit
effort increased over the monitoring period, suggesting that cultured lobsters had enhanced existing stocks rather than replacing
them (Agnalt et al., 1999; Svåsand et al., 2004). Most recently, off
Helgoland, .9% of the 2000 – 2005 release cohorts had been
recaptured by 2009, when 8% of the total landings comprised
hatchery-reared lobsters (Schmalenbach et al., 2011). Of those lobsters released in 2001, 1 in 7 was recaptured, the highest rate
recorded for any stocked H. gammarus cohort (Schmalenbach
et al., 2011).
Projections and perceptions of success
The results of European projects have produced very different perceptions about the potential worth of lobster stocking. In France, the
small number of recaptures caused an abrupt and premature termination of the monitoring programme (Latrouite and Lorec,
1991). In the United Kingdom, the results were welcomed as the
first definitive proof of successful survival and recruitment of cultivated lobsters in the wild (Addison and Bannister, 1994; Bannister
and Addison, 1998). However, modelling showed that recovery
rates were too low to generate a positive net value to the fishery,
even when offsetting the costs of building a hatchery over a
25-year release period (Whitmarsh, 1994). In Norway, the large proportional contribution to the depleted stock was viewed positively
(Agnalt et al., 1999; Svåsand et al., 2004), though production costs
exceeded the value of recaptured lobsters here too (Moksness
et al., 1998). In a global context, lobster stocking in Norway gave
more efficient fishery yields than those of prawn or crab enhancement in the Far East (Hamasaki and Kitada, 2008b).
Although none of these monitored European stocking trials generated total recapture rates of even 10% of the number of lobsters
released (Bannister and Addison, 1998; Agnalt et al., 1999; Nicosia
and Lavalli, 1999; Agnalt et al., 2004; Schmalenbach et al., 2011),
some studies have estimated more encouraging survival rates from
speculative calculations of capture probability. For hatchery-reared
lobsters in Helgoland, the survival rate to the fishery minimum
i39
Lobster stocking requires impact assessment
landing size (MLS) was estimated to be 30 –40% using the Lincoln–
Peterson method (Schmalenbach et al., 2011). When converted via
an independent estimate of trap catchability, recapture numbers
produced very high survival estimates of 50 –84% for individual
release sites in northeast England (Bannister et al., 1994).
Norwegian recaptures provided more tangible evidence of success
by showing that cultured lobsters contributed significantly to
spawning biomass. Within 4 –10 years of release, cultured females
were estimated to account for 27% of egg production within the
Kvitsøy population and showed no difference to wild females in
measures of fecundity or egg development (Agnalt et al., 2007;
Agnalt, 2008).
Fitness of hatchery-reared lobsters
Studies from stocked populations in Norway provide the only direct
evidence of the fitness of cultured H. gammarus in the wild, with
ecological and genetic indicators used to assess pre- and post-release
fitness. Mature cultured females appear to perform as well as wild
equivalents in terms of size-specific fecundity, weight of egg mass,
egg size, and embryonic development (Agnalt, 2008), a crucial
finding rarely achieved among other stocked species. Results have
been less conclusive when rearing the offspring of wild and cultured
broodstock together in competitive, “common garden” environments. The progeny of cultured females recaptured around
Kvitsøy, Norway, experienced only 60% of the survival of the offspring of local wild females through both the larval and juvenile
phases (Jørstad et al., 2005a, 2009). While in isolation, this represents a damaging assessment of the fitness of cultured lobsters,
results were confused by the performance of a second group of wild
females that originated just 12 km away, but whose offspring were
similarly outperformed by those of local natural females. Perhaps
most tellingly though, the authors acknowledged that both wild
and cultured males had access to mate with either cohort of females
(Jørstad et al., 2005a, 2009), which may have significantly biased
the categorization of offspring as wild- or hatchery-derived, particularly within a population where natural and cultured lobsters were
fairly evenly represented (Agnalt et al., 2004).
Limitations of existing impact assessments
Existing assessments of lobster stocking success are susceptible to
caveats and assumptions. The recapture numbers cited in Table 1
were not corrected for (i) tag loss, which would yield false negatives
and underestimates of survival among tagged lobsters (Agnalt et al.,
2004); (ii) emigration to adjacent areas, which would reduce the
number of marked lobsters available for recapture (Cook, 1995);
(iii) spatial mismatch between release and resampling sites; and
(iv) imperfect recapture sampling by quayside monitoring teams.
These issues have not been factored into the lobster survival estimates of any impact assessment, suggesting that the recorded recovery rates cited in Table 1 were almost certainly underestimates. As
such, pessimistic assessments of the economic viability of lobster
stocking by Whitmarsh (1994) and Moksness et al. (1998) were
probably based on pessimistic estimates of the survival of cultured
lobsters. More basically, these economic assessments evaluated the
viability of stocking programmes to be run purely as self-financing
businesses and failed to account for the long-term potential for
hatchery-reared lobsters to boost or restore local recruitment.
Additionally, this appraisal technique fails to account for any potential benefits of raising the profile of lobsters and sustainable fishing
among the public.
Summary of stocking performance and current hatcheries
Stocking has been proved to be a potentially effective method of fisheries remediation (Bannister and Addison, 1998; Svåsand et al.,
2004). Despite uncertainties in the magnitude of the recovery
rates, monitoring of H. gammarus releases have shown that
hatchery-reared lobsters have survived, grown, and mated in the
wild in considerable numbers and in multiple locations and ecotypes. However, there remains a considerable scope to improve
our knowledge of the ecological dynamics influencing stocked
lobster survival and to standardize methods of lobster stocking
and assessments of its impact.
Interest in undertaking European lobster stocking has soared in
recent years as a tool to conserve and improve fisheries and even to
mitigate proposed offshore developments (e.g. pipe-laying, wind
farms, and spoil dumping). Currently, there are two established
hatcheries in the United Kingdom undertaking stock enhancement
on a relatively significant scale. These programmes operate in the
Orkney Islands and Cornwall (Table 1), where the continued pressure on lobster stocks and the economic importance of the fishery
justify the concept of engaging in stock enhancement. They are responsible for over half of the reported releases of cultured lobsters
into European waters in the past four decades, but neither programme has ever undertaken routine monitoring of their effects.
This is mostly due to the prohibitive costs incurred in growing juveniles to sizes suitable for physical tagging and subsequent monitoring of the wild population for recaptures (D. Shearer, Orkney
Lobster Hatchery, pers. comm.). For scientific support, they refer
to the basic impact assessments already described; Orkney was
one location of the 1980s mark– recapture trials, whereas Cornish
enhancement endeavours are based entirely on the experimental
results from outside Cornwall.
Both hatcheries have been active in undertaking research and
developing technical innovations to more effectively and economically rear lobsters. They are aware that reducing expenditure per
juvenile produced is a principal method of increasing their economic viability, alongside increasing the survival probability of
hatchery-reared lobsters in the wild. These hatcheries also accept
their obligation to validate the impact of their stocking programmes, but have been unable to self-subsidize the comprehensive
ecological research and monitoring required. What follows is a
summary of several aspects of marine stocking that are critical to
resolve to improve or perhaps even disprove the value of releasing
cultured lobsters for stock management.
Critical issues for lobster stocking
Understanding lobster ecology
Knowledge gaps regarding the ecology and population dynamics of
H. gammarus significantly obstruct the unbiased assessment of the
performance of hatchery stocking. The most serious of these is the
continued absence of methodologies for locating or capturing
wild post-larvae and juveniles, despite coordinated efforts (e.g.
Linnane et al., 2001; Mercer et al., 2001). As a result, it is unknown
whether recruitment is density-dependent and, therefore, limited
by habitat-specific carrying capacities (as it is in H. americanus—
Wahle and Steneck, 1991, 1992; Wahle and Incze, 1997; Steneck
and Wahle, 2013), and we have no understanding of how cultured
lobsters compare with wild equivalents in basic behavioural, physiological, and morphological traits. Almost all published information
on the biology of early benthic phase H. gammarus emerges from
studies based on cultured lobsters, most of which have occurred
i40
in aquaria environments (e.g. Wickins et al., 1996; Linnane et al.,
2000). Even when based in the wild (e.g. van der Meeren, 2000,
2005), observations of the behaviour and performance of
hatchery-reared juveniles still may not accurately reflect the
biology of natural juveniles in wild ecosystems.
Similarly, the planktonic larval phases are rarely collected in the
wild, even in areas high in abundance of reproductively mature
adults (S. Clark, Devon and Severn IFCA, pers. comm.). Light
traps have proved useful for surveying wild larvae in Scandinavian
fjords, which exhibit considerable water retention (Øresland and
Ulmestrand, 2013), but have had limited success within the
Bristol Channel in the United Kingdom due to strong tides and currents (S. Clark, Devon and Severn IFCA, pers. comm.). Elsewhere,
continuous plankton recorder samples provide temporally and spatially extensive datasets of planktonic abundance, but decapod
larvae are not routinely identified to species level (Richardson
et al., 2006). The absence of basic data on natural larvae and juveniles
has inhibited the creation of demographic models that have been
useful to predict the effect of stocking in other species (e.g.
Lorenzen, 2005, 2006; Hervas et al., 2010).
There is a dearth of studies dedicated to operational variables and
their influence on settlement success in hatchery-reared lobsters,
and the lack of standardization in existing stocking trials makes
their data unsuitable for analysis. Comparisons of different methodological aspects are likely to be biased by the presence of many
uncontrolled covariates throughout the culture, release, and monitoring processes. Experimental features such as release methods
have varied extensively within and among individual projects,
with juveniles variously delivered onto benthic habitats by divers
or water flume (Bannister et al., 1994; Burton, 2001), released offshore at the sea surface at night (Schmalenbach et al., 2011), and
even released during the day into shallow waters off boats or along
the intertidal shoreline (Agnalt et al., 1999). In isolation, the lower
recapture rates recorded in the United Kingdom compared with
Norway and Germany could, therefore, be interpreted as a sign
that benthic releases yield lower settlement success than surface
and shore releases. However, this is counterintuitive to our expectation that delivering lobsters onto shelter-providing benthic substrates, avoiding pelagic predators, should increase settlement
success. It is more likely that the lower UK recapture results arise
from the higher abundance of the wild stock, as enhancing productive stocks has been less effective than restocking depleted populations in other decapod crustaceans (Hamasaki and Kitada,
2008b). However, this cannot be evaluated using existing data and
should be investigated.
Improving tagging technology
Existing monitored stocking experiments have depended on the use of
physical tags to detect recaptured lobsters, with first the CWT in the
1980s and later the VIE from the late 1990s. These assessments provided the first empirical evidence of the performance of hatcheryreared lobsters in the wild, but there are important limitations to the
use and effectiveness of these tags. Both tag types are normally injected
into ventral tissues of the upper abdomen, from where VIE tags have
been shown not to alter behaviour or growth (Neenan et al., 2014). VIE
tags are logged visually through translucent tissues (Uglem et al., 1996;
Neenan et al., 2014), whereas CWTs must be retrieved by dissection
after initial detection by magnetometer (Burton, 1992; Bannister
et al., 1994). Large juveniles (7 months; 12–16 mm CL) show high
tag retention (99%) and survival (97%) over 3 months when tagged
with CWT and VIE and reared in aquaria (Uglem et al., 1996;
C. D. Ellis et al.
Linnane and Mercer, 1998). Modern hatcheries typically release
younger H. gammarus juveniles, however (post-larval stages V–VI,
4–6 weeks old, 5–8 mm CL), which show reduced survival after
tagging (83% for CWT; 68% for VIE) and significant tag migration
(Uglem et al., 1996; Linnane and Mercer, 1998).
The lack of a suitable tag with which to mark juveniles from the
first post-larval instar has prohibited any assessment of whether the
considerable investment required to grow juveniles to sizes facilitating tagging is reflected in increased recruitment. Since the founding
principle of stocking is to culture vulnerable lifestages in captivity, it
is conceivable that lobster survival is suitably optimized at the onset
of benthic settlement behaviours (i.e. post-larval stages IV– V). This
principle, plus the opportunity to maximize numerical release
outputs and avoid on-growing expenses, has meant that most
active European hatcheries now release early juvenile stages as standard, although the only evidence for the effectiveness of this strategy is
inferred from localized increases in abundance of H. americanus in
eastern Canada following releases of cultured post-larvae (e.g.
Comeau, 2006; Côté and Cloutier, 2014). These results were
obtained by the utility of before-after-control-impact (BACI)
methods, where lobster abundance in release areas is compared
with that, in similar, unenhanced habitats over several years. BACI
methods have proved useful in implying enhancement effects
where hatchery-reared lobsters are not tagged (Comeau, 2006;
Côté and Cloutier, 2014), although this style of monitoring produces data that lack the definitive evidence provided by the recapture of tagged individuals. Ideally, a new physical tag is required
that is cheap and easy to apply, is capable of marking lobsters
from the first post-larval phase to adulthood, and is visually detectable by fishers. This would enable a large number of juveniles to be
tagged as standard release procedure and facilitate assessments of
optimal stocking protocols via low-cost and widespread monitoring
by fishery stakeholders, who may be positively motivated by a visible
tag. However, such a development is unlikely to be forthcoming,
given the regular turnover of sclerotized body parts at ecdysis and
the vast discrepancy in size between post-larvae and adults.
Attention is, therefore, turning to the potential for polymorphic
genetic markers to assign parentage and replace or augment physical
tags in future assessments of lobster stocking impact. Methods of
genetic profiling can assign hatchery origin with a high degree of certainty (e.g. Jones and Arden, 2003) and have important advantages
over established internal tags (Table 2). Tag loss can be effectively
eliminated, individuals can be sampled sublethally on multiple
occasions, and there are no restrictions on the release size of juveniles (Neenan et al., 2014). Genetic profiling can allow assessments
of the recruitment performance of different groups, families, or even
genotypes (Sekino et al., 2005; Tringali, 2006) and the extent to
which wild and cultured animals integrate and interbreed in the
environment. With genetic markers of sufficient quantity and variation, hatchery-derived lineages may even be tracked beyond the
released generation by identifying the wild-born offspring of
hatchery-reared parents, potentially enabling multigenerational
assessments of stocking (Letcher and King, 2001; Blouin, 2003).
Employing genetic methods has already proved successful in the
detection of hatchery-reared fish among enhanced wild populations
of steelhead trout (Oncorhynchus mykiss; Christie et al., 2012a,b)
and black sea bream (Acanthopagrus schlegelii; Jeong et al., 2007)
and has been proposed as a method of establishing traceability for
aquaculture-derived fish at the marketplace (Hayes et al., 2005).
In one of the most positive impact assessments of fishery enhancement, microsatellite-based pedigree reconstructions showed that
i41
Lobster stocking requires impact assessment
Table 2. Summary of the expected performance of different tag types for use in impact assessments of European lobster stock enhancement.
Tag performance criteria
Individual
Tag type ID
CWT
Yes
VIE
No
Genotype Nob
No min.
juvenile
size
No
No
Yes
No
tag
loss
No
No
Yesc
Sublethal
sampling
No
Yes
Yes
Stakeholder
independent
monitoring
Noa
Yes
Noa
Stakeholder
social impact
Noa
Yes
Noa
Multiple
generations
traceable
No
No
Yes
Genetic
fitness
impact
No
No
Yes
Stock
integration
testable
No
No
Yes
CWT and VIE performance is based on reported performance in previous uses, whereas genotype tag performance is based on theoretical performance and
reports from other stocked species.
a
Stakeholders may be utilized and socially impacted by monitoring, but cannot readily identify released individuals as part of routine fishing activities.
b
Individual identification is possible, but often requires a much larger panel of genetic markers than is required to establish hatchery origin via parentage
assignment, the most commonly used genotype-based method.
c
No tag “loss”, but similar error can be introduced by genotyping errors (e.g. flawed tissue collection or processing, the presence of null alleles, the interpretation
of results, etc.). Repeat sample processing and analysis of data can be used to estimate and/or correct this error rate.
stocked A. schlegelii suffered no loss of heterozygosity, integrated
with wild schools, and contributed 59% of individuals to an important fishery in Japan (Jeong et al., 2007). Similarly, thorough evaluation is required to elucidate the long-term impact of stocking
H. gammarus, although such investigations are not cheap or accomplishable without archived tissues from which the genotypes of
hatchery progeny can be deduced (i.e. maternal and egg samples).
The type and quantity of markers required for parentage assignments to accurately detect hatchery-reared lobsters from large-scale
surveys of wild populations would be largely dependent on the
population’s genetic diversity, effective size, and gene flow, the
broodstock turnovers and recapture survey methods employed,
and whether multiple paternity frequently exists among individual
broods [as has been found in H. americanus (Gosselin et al., 2005)].
Sampling only landed lobsters that are destined for the market may
be a more practical survey method than in situ, on-board sampling
of the catch (including undersized lobsters destined for return to the
sea). The latter could be biased by the inclusion of single individuals
sampled on multiple occasions, which would be indistinguishable
from multiple individuals possessing genotypes that are identical
by descent, although this approach does lend itself well to obtaining
recapture data that could reveal the movements of stocked lobsters
and the spatial impacts of stocking. Simulations and case studies
have shown that parentage can be accurately assigned, even where
systems boast hundreds or thousands of candidate parents, using
as few as 60 –100 single-nucleotide polymorphisms SNPs (Hayes
et al., 2005; Anderson and Garza, 2006) or 7 –15 microsatellites
(Bernatchez and Duchesne, 2000; Letcher and King, 2001; Hayes
et al., 2005; Jeong et al., 2007; Christie et al., 2012a), although this
is also dependent on the overall power provided by the number
and frequency of alleles (Bernatchez and Duchesne, 2000).
For H. gammarus, it may well be possible to base such parentage
assignments on established and available genetic markers, such as
the 12 microsatellites published by André and Knutsen (2010).
However, where spatial population genetic structuring is minimal,
hatchery broodstock turnovers are high, and multiple paternity
occurs frequently within individual broods (all of which are possibly
the case for H. gammarus), the number of markers required to
resolve parentage may rise to become prohibitively costly. Nextgeneration genotyping resources, such as restriction site associated
DNA (RAD) markers and larger panels of SNPs, offer the resolution
to overcome such obstacles (Baird et al., 2008; Hohenlohe et al.,
2010), and for species such as Atlantic salmon (Salmo salar), microarray genotyping chips featuring many thousands of SNPs are now
widely available (Affymetrix, 2014). The development and widespread utilization of such technology is likely to be beyond the financial means and expertise of independent lobster hatchery ventures,
however. Still, there is a significant time lag between captive rearing
and potential recapture in the wild, and many universities and research facilities are now equipped with the capabilities to carry
out a range of molecular genetic analyses. Therefore, even where
no immediate plans exist to assess stocking, all lobster hatcheries
should routinely archive tissue and several fertilized eggs from
every brood female for potential future collaborative research opportunities.
Improving hatchery production
All hatcheries require the stable production of juveniles to enable
release numbers to achieve stocking targets. Because facilities culturing lobster have experienced prolonged and sometimes unexplained periods of production failure, stabilizing juvenile output
is required. Where cultured juveniles have no reduction in fitness,
increasing both the quantity released and their chances of wild
establishment can improve the effectiveness of stocking. Some
significant biotechnical advances have been made in recent years
that improve lobster hatchery production and cost-effectiveness.
While ovigerous females are plentiful in spring and summer, the
separation of some at reduced water temperatures (68C) slows
egg development and allows the rearing season to be extended.
Anecdotally, this has been more effective and reliable than the
upward manipulation of egg development and raises the possibility
that stable year-round production may be possible. In trials of the
so-called green water technique, utilizing algal cultures and
enriched live feed more than doubled survival to the first post-larval
instar compared with standard rearing protocols (Browne et al.,
2009). The larval and post-larval stages are particularly vulnerable
to the effects of nutrient limitation; therefore, nutritional enrichments improve growth and survival, even in standard culture
environments (Daniels et al., 2010; Schoo et al., 2014). Further
improvements have arisen from the long-awaited innovation of
multilayered juvenile rearing systems, which increase hatchery
capacity 40-fold compared with traditional single-layer vessels
(Gowland, 2013). As advancements continue, hatcheries are able
to increase production and the overall economic viability of
lobster stocking. For example, one H. americanus hatchery more
than doubled its production costs from 2002 to 2013, although
this enabled technical advances that increased annual production
i42
from 1500 to 417 000 juveniles, slashing the investment per juvenile
from over US$33 to just US$0.26 (Haché et al., 2014).
As well as ensuring they can produce the quantity of juveniles
required, stocking projects must aim to ensure that the quality of
cultured lobsters is sufficient to achieve long-term population
enhancement. In Norway, the performance of recaptured lobsters
has been promising in basic fitness traits, such as reproductive
potential (Agnalt, 2008). Nevertheless, juveniles reared in captive
conditions are frequently shown to have reduced suitability to the
demands of life in natural ecosystems (e.g. Davis et al., 2004, 2005;
Castro and Cobb, 2005). Ecological naivety is evident in the higher
predation vulnerability of cultured H. americanus juveniles compared with wild conspecifics (Castro and Cobb, 2005). For
H. gammarus, the continued failure to locate wild juveniles has
prevented comparisons of fitness to that of cultured equivalents,
an approach used widely for other stocked decapods (e.g. Davis
et al., 2004, 2005; Castro and Cobb, 2005; Ochwada-Doyle et al.,
2010). Even so, studies have shown that juveniles reared in competitive communal environments grow faster than those raised
in isolation (Jørstad et al., 2001), while previous exposure to
predator odours gives cultured juveniles a superior ability to outcompete untreated cohorts for limited shelter spaces (Trengereid,
2012).
Although some cultured decapod juveniles have matched the
predator avoidance of wild conspecifics regardless of acclimation
regimes (Ochwada-Doyle et al., 2010), innate behaviours are likely
to be complemented by targeted ecological conditioning before
wild release. In hatchery-reared blue crabs (Callinectes sapidus),
conditioning via controlled predator exposure significantly increases carapace spine length and subsequent post-release survival
(Davis et al., 2004, 2005). The traditional hatchery culture of
H. gammarus juveniles is isolated and largely devoid of environmental enrichment, but in recent years, attempts have been
made to on-grow juveniles in sea-based submerged containers.
This semi-wild environment appears to promote traits that are
likely to have a positive impact on settlement success and adaption
to the natural environment and offers significant potential as an
acclimation step before the release of cultured lobsters. Survival
often exceeds that of hatchery-reared cohorts (Beal et al., 2002;
Benavente et al., 2010), and container-reared lobsters typically
demonstrate altered behavioural responses and improved growth
and pigmentation (Figure 1). Overall, the unnatural selection pressures of culture environments are a fitness concern that remains
largely unaddressed in lobster hatcheries, and significant adjustments to existing rearing and conditioning protocols may well be
required to increase the viability of current lobster stocking ventures
(van der Meeren, 2005; Trengereid, 2012).
Ensuring effective genetic management
Poorly regulated fishing throughout most of the range of
H. gammarus is likely to have seriously impacted the status of
benthic ecosystems and significantly influenced the population genetics of European lobsters. Genetic management of the species has
rarely been prioritized or even considered by fishery managers,
and the pressures of intensive commercial fishing activities are
likely to have impacted the genetics of lobster populations more profoundly than the limited activity of stocking schemes to date.
However, mismanagement of lobster fisheries in general should
not mean that ventures aiming to enhance and conserve these fisheries via hatchery stocking should not be expected to pursue rigorous standards of ecological accountability. While stocking is
C. D. Ellis et al.
Figure 1. Cultured juvenile European lobsters on-grown in sea
containers in an open bay (c) and estuary (b) show increased growth
and pigmentation compared with equivalents reared only in the
hatchery (a).
generally expected to increase short-term abundance of populations, troubling recent data in other species suggest that negative
genetic impacts may arise in target stocks, undermining fishery conservation objectives (Sekino et al., 2002, 2003; Bert et al., 2007;
Kitada et al., 2009; Rourke et al., 2009; Hamasaki et al., 2010;
Christie et al., 2012a,b; Satake and Araki, 2012). It is increasingly
apparent that the dual goals of short-term productivity and longterm conservation are not usually complementary and are difficult
to achieve simultaneously (Satake and Araki, 2012).
Many authors have proposed ways in which stocking schemes
can limit negative genetic impacts, and routinely comparing the
genetic diversity and relative fitness of wild and cultured fish is commonly recommended (e.g. Blankenship and Leber, 1995; Shaklee
and Bentzen, 1998; Bell et al., 2006; Gaffney, 2006; Bert et al.,
2007; Tringali et al., 2008; Laikre et al., 2010; Lorenzen et al.,
2010). For example, Bert et al. (2007) suggest that stocking enterprises should study the species’ regional population genetics,
genotype broodstock at a resolution sufficient to distinguish
their offspring, monitor the genetic variation of cultured juveniles
and incoming broodstock, and use genetic assays to scan the wild
population for both hatchery progeny and any flux in the larger
gene pool. Many independent hatcheries are unable to fund such
research or have prioritized investing in biotechnical innovations
though, so genetic aspects of management have often been ignored
(Bell et al., 2006). This is largely the case among organizations
stocking H. gammarus and requires rectifying to ensure that
heavily exploited lobster fisheries are not subject to any deleterious
effects via stocking.
Maintaining fitness and genetic diversity
Attaining long-term population growth and simultaneous conservation of the regional gene pool is unlikely where stocked animals
have fitness disadvantages (Satake and Araki, 2012). Fitness disadvantages can arise in cultured individuals as a consequence of
i43
Lobster stocking requires impact assessment
narrow genetic make-up or via inadvertent selection processes occurring in the hatchery environment that make cultured juveniles
ill-suited to their natural ecosystem. Where released animals introduce heritable reductions in fitness, stocking has the potential to
have negative impacts on wild stocks. This is reported most often
where target populations are small and/or show high levels of adaptation to local conditions (Lorenzen et al., 2012). Released animals
often have reduced fitness for the natural environment compared
with wild conspecifics; Araki and Schmid (2010) reviewed 39
studies that assessed fitness effects, of which 22 found that survival,
growth, or reproductive success were reduced by hatchery rearing.
Given the dissimilarities between hatchery and wild environments,
traits that lead to high fitness in one may reduce fitness in the other.
Trout (O. mykiss) raised in captivity have nearly double the reproductive success of wild-born fish when spawned in a hatchery, but
their offspring suffer greatly reduced performance in the wild,
where survival is less than a third that of the wild-origin cohort
(Christie et al., 2012a).
A key principle of stocking is that offspring survival is relatively
increased in the captive environment, which means that many
released individuals may be closely related. Increasing the number
of related individuals in a population generally decreases the
overall genetic diversity and effective population size and increases
the potential for inbreeding depression (Ryman and Laikre, 1991).
Cultured individuals often show reduced genetic diversity (e.g.
Sekino et al., 2002) and have low effective population sizes, especially where broodstock are captive-reared, are used to rear multiple
generations of offspring, or where competitive processes lead to
highly skewed reproductive success (Sekino et al., 2003; Shishidou
et al., 2008). Parentage assignments in hatchery-reared flounder
(P. olivaceus) revealed that almost all the offspring were sired by
one of six males, and that half of the 12 spawning females yielded
no surviving juveniles at all (Sekino et al., 2003). Although the influence of stocking on population genetic diversity may be trivial compared with that caused by environmental or fishing pressures
(Sugaya et al., 2008; Kitada et al., 2009), in some cases, it can be
extremely damaging; stocking doubled the number of adult trout
(O. mykiss) on spawning grounds in Oregon, United States, but actually cut the total effective population size by two-thirds (Christie
et al., 2012b).
Wild-mated females have typically been utilized for H. gammarus
stocking, with several hundred new broodstock sourced for each
production season. Where broodstock are marketed for human
consumption upon return to their donors, their repeated use is
prevented. Whether achieved via the ease of accessing readilymated females or by enlightened genetic practices, these methods should have contributed to ensuring relatively high genetic
diversity among progeny. However, family contributions have
been found to be skewed in H. gammarus culture (Jørstad et al.,
2005a), and how the genetic diversity of released lobsters compares
with that within target populations requires evaluation using
modern techniques.
Consideration of population structure
and local adaptation
Genetic diversity is the principal origin of adaptive evolutionary
potential (Frankham et al., 2011), so populations are increasingly
vulnerable to environmental change where genetic diversity is
eroded by the release of cultured individuals (Laikre et al., 2010).
Where cultured animals lack hereditary adaptations to their
release environment and interbreed with wild fish that are more suitably adapted, adaptive traits crucial to the species’ fitness in that environment are likely to be eroded, reducing the overall fitness of
the population. In recent studies on wild marine fish, molecular
markers have helped reveal previously unforeseen levels of population structure and local adaptation to environmental heterogeneity (e.g. temperature and salinity), even at small geographical
scales [e.g. Atlantic cod (Gadus morhua)—Knutsen et al., 2003,
2011; Jorde et al., 2007; Atlantic herring (Clupea harengus)—
Lamichhaney et al., 2012; Limborg et al., 2012; Teacher et al.,
2013; sticklebacks (Pungitius pungitius, Gasterosteus aculeatus)—
Shikano et al., 2010; Shimada et al., 2011; Bruneaux et al., 2013;
Atlantic salmon—Griffiths et al., 2010)]. Although genotyping
only 25 individuals per population can often provide accurate estimates of population-level differences in allele frequencies (Hale
et al., 2012), even relatively basic genetic studies are generally
complex and expensive. As a result, population genetic data are frequently absent or outdated for stocked marine species, and such
studies on H. gammarus provide somewhat contradictory evidence
or lack peer review.
Investigations on genetic structure and diversity in H. gammarus
populations using polymorphic microsatellites, allozymes, and
mitochondrial DNA (e.g. Jørstad and Farestveit, 1999; Jørstad
et al., 2004, 2005b; Triantafyllidis et al., 2005; Huserbråten et al.,
2013) have attempted to delineate populations and estimate gene
flow within and between them. Observed restrictions in adult
migration give the potential for considerable genetic isolation
between H. gammarus subpopulations (Øresland and Ulmestrand,
2013), although the most recent research suggests that high genetic
connectivity exists over relatively large spatial scales (≈400 km),
even among semi-enclosed habitats (Huserbråten et al., 2013).
These results, obtained via microsatellite DNA analysis of heavily
depleted Scandinavian Skagerrak populations, suggest that larval dispersal must be high and must be the primary origin of gene flow
(Huserbråten et al., 2013). Where larvae are distantly dispersed and
cultured lobsters add significantly to the spawning biomass, the longterm impacts of stocking could extend far beyond the spatial boundaries over which releases occur.
Spatial heterogeneity in H. gammarus population genetic variation has been detected, however, particularly in regions isolated
by oceanographic and topographic conditions, such as northern
Norway and throughout the Mediterranean (Jørstad and Farestveit,
1999; Ulrich et al., 2001; Jørstad et al., 2004, 2005b; Triantafyllidis
et al., 2005) and even among populations from the comparatively
unrestricted Atlantic coasts of Ireland, France, and Portugal (Ulrich
et al., 2001). There appears an overall association between geographic distance and genetic variation (Ulrich et al., 2001), although considerable genetic differences can be found over modest spatial scales
(e.g. 142 km between fjords, Jørstad et al., 2004). Rapid recent developments in whole-genome genotyping methodologies and the field
of bioinformatics now offer greater resolution and deeper insights
into the extent of population structure and local adaptation. Studies
utilizing these technologies throughout the range of H. gammarus
will be critical for understanding the spatial scales that stocking
may be expected to impact and for ensuring that lobster releases
are non-detrimental.
Stocking vs. alternative management strategies
To date, lobster stocking in Europe has always been practiced in addition to legislative fishery management measures such as closed
i44
seasons, closed areas, gear restrictions, and landing bans on undersized, v-notched, or ovigerous lobsters. However, assessments of
the relative effectiveness of lobster stock enhancement and other
alternative fishery management tools are either lacking or are too
ambiguous to allow formative comparison between methods. Several
conservation methods are often applied concurrently, which potentially gives a greater chance of safeguarding stocks, but it becomes
difficult to appraise the relative strengths and limitations of individual components. The need for rigorous analysis of lobster
stocking is particularly urgent, but so too is the analysis of other
management measures to enable comparative assessments of
fishery conservation tools.
Our understanding of the effects of most fishery management
options is poor, but marine protected areas (MPAs) have recently
demonstrated potential in sustaining exploited lobster populations.
In the United Kingdom, the closure of waters off Lundy Island to all
fishing activities led to a rapid increase in lobster abundance and
mean body size (Hoskin et al., 2011), whereas in Norway, MPA designation increased lobster cpue by 245% over 4 years, far beyond the
87% increase in control areas (Moland et al., 2013). Over 95% of
lobsters caught, tagged, and rereleased into both Norwegian and
Swedish MPAs remained within or very near to reserve boundaries
in multiannual mark– recapture analyses (Moland et al., 2011;
Øresland and Ulmestrand, 2013), while high genetic connectivity
between these MPAs suggests that larval dispersal benefits may be
extensive and far-reaching (Huserbråten et al., 2013). Arguably,
thoughtfully designated MPAs have offered more conclusive stock
conservation benefits than hatchery stocking to date, although
MPAs do have an immediate negative economic impact on displaced fishers. However, employing the two methods simultaneously (i.e. releasing cultured lobsters into MPAs) may offer a
powerful stock conservation method and provide quicker enhancement of adjacent fisheries.
Conclusions
The regulation of European lobster stocking has been largely ad hoc
and lacks alignment with the robust frameworks established for
the informed management of marine stocking ventures (e.g.
Blankenship and Leber, 1995; Lorenzen et al., 2010). Given that
recent findings from the wider field of aquatic stocking show
that the successful integration of cultured individuals into
dynamic wild populations is a highly complex process, this is
clearly unsatisfactory. While some deviation from best practice
may have been the result of insufficient and fragmented planning
by regulatory managers or hatchery operators, much more has
been unavoidable. The inconclusive performance of previous
lobster stocking projects in providing economically viable benefits
to lobster fisheries has made it hard for active hatcheries to attract
significant financial backing and industry support. However,
the exhaustive monitoring and technical developments required
to evidence economic viability are often economically unviable
in their own right; as our understanding of the potential ecological considerations mounts, the costs associated with piloting
a stocking programme increase (Blankenship and Leber, 1995;
Lorenzen et al., 2010). In the absence of focused guidelines or coordinated investment from industry or government, active hatcheries have been largely unable to address significant gaps in our
scientific understanding of lobster biology that are integral to the
informed management of stocking ventures and lobster fisheries
themselves. As a result, hatcheries have been forced to focus on
C. D. Ellis et al.
advancing production and revenue, conducting ecological research where possible along the way.
From existing studies designed to assess the potential for stocking H. gammarus, a proof-of-concept has been demonstrated.
Based on recaptures of hatchery-reared lobsters achieving fishery
MLSs and reproductive maturity in multiple locations, conclusions have been generally positive that stocking could represent a
worthwhile fishery conservation method. However, these conclusions are undermined by a lack of consistent evidence that benefits
are universal and cost-effective and by a series of inconclusive or
damaging reports into the effects of stocking in other marine
species. Nevertheless, in the wake of increased pressures on some
fisheries and the regional collapse of others, interest in stocking
programmes aimed at restoring or enhancing lobster populations
has only increased in recent years. The societal decision whether to
pursue stocking of European lobster populations requires evidence of both positive and negative impacts of hatchery releases,
so a renewed evaluation of lobster stocking, utilizing the more
thorough assessment tools now available, is required to limit the
ambiguity of that decision.
Impact assessments attempting to appraise the effect of European
lobster stocking are significantly hindered by the elusive nature of
wild juveniles and scarcity of other basic information on the
ecology of natural populations and have, so far, been restricted to
unfavourable juvenile tagging methods. Genetic methods should
be employed to improve wider understanding of lobster biology
and population ecology as well as to deliver assessments on the evolutionary fitness of cultured lobsters and the likelihood of their
release to cause negative effects on natural populations. Genetic
resources also require testing for their effectiveness in identifying
cultured lobsters in the wild. Recent improvements in the quality
and cost-efficiency of juvenile production could help make stocking
a viable tool for improving the productivity and sustainability of
lobster fisheries, although this requires a thorough and strategic
evaluation.
Overall, our understanding of the dynamics and potential for
lobster stocking remains limited, and further research using contemporary methods is required to deliver informative impact assessments. Ideally, all lobster hatcheries should implement the following
initiatives: (i) archive maternal and progeny tissues from all broodstock; (ii) establish a management strategy that will limit negative
impacts of releases in the presence of population structure and
local adaptation; (iii) conduct controlled temporal studies of
lobster abundance in release areas, both before and after stocking;
and (iv) link with a research institute or university to enable collaborative research. Implementation of these procedures would help
raise the ethical and ecological standards of stocking ventures,
would provide basic evidence of the effect of stocking on local abundance, would lay the foundations for more comprehensive assessments of the performance of stocked lobsters, and would facilitate
partnerships with organizations capable of assessing population
structure and stock boundaries throughout the species’ range, as
well as driving efforts to locate wild juveniles to resolve associated
knowledge gaps.
Acknowledgements
The authors are grateful to the European Social Fund and the
Fishmongers’ Company, United Kingdom, for funding this research
and to the insightful comments of two anonymous reviewers who
helped improve this paper. We are also indebted to the authors of
the many studies we have referenced in this review and to those
Lobster stocking requires impact assessment
lobster fishers and hatchery operators who contributed to much of
their research.
References
Addison, J. T., and Bannister, R. C. A. 1994. Restocking and enhancement of clawed lobster stocks—a review. Crustaceana, 67: 131 – 155.
Affymetrix. 2014. http://www.affymetrix.com. (last accessed 9 September 14).
Agnalt, A-L. 2008. Fecundity of the European lobster (Homarus gammarus) off southwestern Norway after stock enhancement: do cultured females produce as many eggs as wild females? ICES Journal
of Marine Science, 65: 164 – 170.
Agnalt, A-L., Jørstad, K. E., Kristiansen, T. S., Nøstvold, E., Farestveit, E.,
Næss, H., Paulsen, O. I., et al. 2004. Enhancing the European lobster
(Homarus gammarus) stock at Kvitsøy Islands; Perspectives of
rebuilding Norwegian stocks. In Stock Enhancement and Sea
Ranching: Developments, Pitfalls and Opportunities, pp. 415–426.
Ed. by K. M. Leber, S. Kitada, H. L. Blankenship, and T. Svåsand.
Blackwell Publishing Ltd, Oxford.
Agnalt, A-L., Kristiansen, T. S., and Jørstad, K. E. 2007. Growth, reproductive cycle, and movement of berried European lobsters (Homarus
gammarus) in a local stock off southwestern Norway. ICES Journal of
Marine Science, 64: 288– 297.
Agnalt, A-L., van der Meeren, G. I., Jørstad, K. E., Næss, H., Farestveit,
E., Nøstvold, E., Svåsand, T., et al. 1999. Stock enhancement of
European lobster (Homarus gammarus): a large-scale experiment
off South-western Norway (Kvitsøy). In Stock Enhancement
and Sea Ranching, pp. 401– 419. Ed. by B. Howell, E. Moksness,
and T. Svåsand. Fishing News Books Ltd, Farnham.
Anderson, E. C., and Garza, J. C. 2006. The power of single-nucleotide
polymorphisms for large-scale parentage inference. Genetics, 172:
2567– 2582.
André, C., and Knutsen, H. 2010. Development of twelve novel microsatellite loci in the European lobster (Homarus gammarus).
Conservation Genetics Resources, 2: 233 – 236.
Araki, H., and Schmid, C. 2010. Is hatchery stocking a help or harm?
Evidence, limitations and future directions in ecological and
genetic surveys. Aquaculture, 308: S2 – S11.
Baird, N. A., Etter, P. D., Atwood, T. S., Currey, M. C., Shiver, A. L.,
Lewis, Z. A., Selker, E. U., et al. 2008. Rapid SNP discovery and
genetic mapping using sequenced RAD markers. PLoS ONE, 3.
doi:10.1371/journal.pone.0003376.
Ball, B., Linnane, A., Munday, B., Browne, R., and Mercer, J. P. 2001. The
effect of cover on in situ predation in early benthic phase European
lobster Homarus gammarus. Journal of the Marine Biological
Association of the United Kingdom, 81: 639 – 642.
Bannister, R. C. A. 1986. Assessment and population dynamics of commercially exploited shellfish in England and Wales. In North Pacific
Workshop on Stock Assessment and Management of Invertebrates,
pp. 182– 194. Ed. by G. S. Jamieson, and N. Bourne. Canadian
Special Publication on Fisheries and Aquatic Sciences, Department
of Fisheries and Oceans, Ottawa, 92.
Bannister, R. C. A., and Addison, J. T. 1998. Enhancing lobster stocks: a
review of recent European methods, results, and future prospects.
Bulletin of Marine Science, 62: 369 – 387.
Bannister, R. C. A., Addison, J. T., and Lovewell, S. R. J. 1994. Growth,
movement, recapture rate and survival of hatchery-reared lobsters
(Homarus gammarus, L.) released into the wild on the English east
coast. Crustaceana, 67: 156 –172.
Bannister, R. C. A., and Howard, A. E. 1991. A large-scale experiment to
enhance a stock of lobster (Homarus gammarus (L.)) on the English
east coast. ICES Marine Science Symposia, 192: 99– 107.
Beal, B. F., Mercer, J. P., and O’Conghaile, A. 2002. Survival and growth
of hatchery-reared individuals of the European lobster, Homarus
gammarus (L.), in field-based nursery cages on the Irish west coast.
Aquaculture, 210: 137– 157.
i45
Beard, T. W., Richards, P. R., and Wickins, J. F. 1985. The techniques and
practicability of year-round production of lobsters, Homarus gammarus (L.), in laboratory recirculation systems. Lowestoft, UK.
MAFF Directorate of Fisheries Research, Fisheries Research
Technical Report, 79. 22 pp.
Beard, T. W., and Wickins, J. F. 1992. Techniques for the production of
juvenile lobsters (Homarus gammarus (L.)). Lowestoft, UK. MAFF
Directorate of Fisheries Research, Fisheries Research Technical
Report, 92. 22 pp.
Bell, J. D., Bartley, D. M., Lorenzen, K., and Loneragan, N. R. 2006.
Restocking and stock enhancement of coastal fisheries: potential,
problems and progress. Fisheries Research, 80: 1 –8.
Benavente, G. P., Uglem, I., Browne, R., and Balsa, C. M. 2010. Culture of
juvenile European lobster (Homarus gammarus L.) in submerged
cages. Aquaculture International, 18: 1177– 1189.
Bernatchez, L., and Duchesne, P. 2000. Individual-based genotype analysis in studies of parentage and population assignment: how many
loci, how many alleles? Canadian Journal of Fisheries and Aquatic
Sciences, 57: 1 – 12.
Bert, T. M., Crawford, C. R., Tringali, M. D., Seyoum, S., Galvin, J. L.,
Higham, M., and Lund, C. 2007. Genetic management of hatcherybased stock enhancement. In Ecological and Genetic Implications of
Aquaculture Activities, pp. 123– 174. Ed. by T. M. Bert. Springer,
Dordrecht.
Blankenship, H. L., and Leber, K. M. 1995. A responsible approach to
marine stock enhancement. American Fisheries Society Symposium,
15: 167–175.
Blouin, M. S. 2003. DNA-based methods for pedigree reconstruction
and kinship analysis in natural populations. Trends in Ecology and
Evolution, 18: 503 – 511.
Botero, L., and Atema, J. 1982. Behaviour and substrate selection during
larval settling in the lobster, Homarus americanus. Journal of
Crustacean Biology, 2: 59 – 69.
Breen, P. A., and Mann, K. H. 1976. Changing lobster abundance and
the destruction of kelp beds by sea urchins. Marine Biology, 34:
137– 142.
Browne, R., Benavente, G. P., Uglem, I., and Marino Balsa, J. C. 2009. An
illustrated hatchery guide for the production of clawed lobsters using
a green water technique. Aquaculture Explained, 23: 1 – 36.
Browne, R., and Mercer, J. P. 1998. The European clawed lobster
(Homarus gammarus) stock enhancement in the Republic of Ireland.
In Proceedings of a Workshop on Lobster Stock Enhancement held
in the Magdalen Islands (Quebec) from October 29 to 31, 1997,
pp. 33 – 41. Ed. by L. Gendron. Canadian Industry Report of
Fisheries and Aquatic Sciences, 244.
Bruneaux, M., Johnston, S. E., Herczeg, G., Merila, J., Primmer, C. R.,
and Vasema, A. 2013. Molecular evolutionary and population
genomic analysis of the nine-spined stickleback using a modified
restriction-site-associated DNA tag approach. Molecular Ecology,
22: 565– 582.
Burton, C. A. 1992. Techniques of lobster stock enhancement. Sea Fish
Industry Authority, Marine Farming Unit, Ardtoe, UK. Report 36.
Burton, C. A. 1993. The United Kingdom lobster stock enhancement
experiments. In Proceedings of a Conference, 12 September 1992,
Stromness, Orkney Islands, UK, pp. 22 – 35. Ed. by M. Baine.
International Centre for Island Technology, Stromness.
Burton, C. A. 2001. The role of lobster (Homarus sp.) hatcheries in
ranching, restoration and remediation programmes. Hydrobiologia,
465(1–3): 45–48.
Castro, K. M., and Cobb, J. S. 2005. Behaviour of hatchery-reared and
wild-caught 4th and 5th stage American lobsters, Homarus americanus. New Zealand Journal of Marine and Freshwater Research,
39: 963– 972.
Christie, M. R., Marine, M. L., French, R. A., and Blouin, M. S. 2012a.
Genetic adaptation to captivity can occur in a single generation.
Proceedings of the National Academy of Sciences of the USA, 109:
238– 242.
i46
Christie, M. R., Marine, M. L., French, R. A., Waples, R. S., and Blouin,
M. S. 2012b. Effective size of a wild salmonid population is greatly
reduced by hatchery supplementation. Heredity, 109: 254 – 260.
Cobb, J. S. 1971. The shelter related behaviour of the lobster, Homarus
americanus. Ecology, 52: 108 – 115.
Comeau, M. 2006. Lobster (Homarus americanus) enhancement project
in the southwestern Gulf of St. Lawrence, Canada. Alaska Crab Stock
Enhancement and Rehabilitation: Workshop Proceedings, Alaska
Sea Grant College Program, AK-SG-06-04: 71– 78.
Cook, W. 1995. A lobster stock enhancement experiment in Cardigan
Bay. Final report. North Western North Wales Sea Fisheries
Committee, University of Lancaster, UK. 33 pp.
Cooper, R. A., and Uzmann, J. R. 1980. Ecology of adult and juvenile
Homarus. In The Biology and Management of Lobsters, Vol. II,
Ecology and Management, pp. 97 – 142. Ed. by J. S. Cobb, and
B. F. Phillips. Academic Press, New York.
Côté, J., and Cloutier, O. 2014. An overview of the American lobster
(Homarus americanus) stock enhancement program in southern
Gaspé, QC, Canada. In Proceedings of the 10th International
Conference and Workshop on the Biology and Management
of Lobsters, Cancún, Mexico. May 18 – 23, 2014. Puerto Morelos,
Quintana Roo, Mexico, Institute of Ocean Sciences and
Limnology, 61.
Daniels, C. L., Merrifield, D. L., Boothroyd, D. P., Davies, S. J., Factor,
J. R., and Arnold, K. E. 2010. Effect of dietary Bacillus spp. and
mannan oligosaccharides (MOS) on European lobster (Homarus
gammarus L.) larvae growth performance, gut morphology and
gut microbiota. Aquaculture, 304: 49– 57.
Davis, J. L. D., Eckert-Mills, M. G., Young-Williams, A. C., Hines, A. H.,
and Zohar, Y. 2005. Morphological conditioning of a
hatchery-raised invertebrate, Callinectes sapidus, to improve field
survivorship after release. Aquaculture, 243: 147 – 158.
Davis, J. L. D., Young-Williams, A. C., Aguilar, R., Carswell, B. L.,
Goodison, M. R., Hines, A. H., Kramer, M. A., et al. 2004.
Differences between hatchery-raised and wild blue crabs: implications for stock enhancement potential. Transactions of the
American Fisheries Society, 133: 1 – 14.
Dow, R. L. 1980. The clawed lobster fisheries. In The Biology and
Management of Lobsters, Vol. II, Ecology and Management, pp.
265– 306. Ed. by J. S. Cobb, and B. F. Phillips. Academic Press,
New York.
Fisheries and Aquaculture Department. 2014. www.fao.org. (last
accessed 15 January 14).
Fish Information and Services. 2014. http://www.fis.com/. (last
accessed 15 January 14).
Frankham, R., Ballou, J. D., and Briscoe, D. A. 2011. Genetic diversity. In
Introduction to Conservation Genetics, 2nd edn, pp. 41– 65. Ed. by
R. Frankham. Cambridge University Press, Cambridge.
Gaffney, P. M. 2006. The role of genetics in shellfish restoration. Aquatic
Living Resources, 19: 277– 282.
Gosselin, T., Sainte-Marie, B., and Bernatchez, L. 2005. Geographic variation of multiple paternity in the American lobster, Homarus americanus. Molecular Ecology, 14: 1517– 1525.
Gowland, D. 2013. Shellfish Hatchery Systems Ltd, (last accessed 19
November 2013). http://aquahive.co.uk.
Griffiths, A. M., Machado-Schiaffino, G., Dillane, E., Coughlan, J.,
Horreo, J. L., Bowkett, A. E., Minting, P., et al. 2010. Genetic stock
identification of Atlantic salmon (Salmo salar) populations in the
southern part of the European range. BMC Genetics, 11: 31.
Grimsen, S., Jaques, R. N., Erenst, V., and Balchen, J. G. 1987. Aspects of
automation in a lobster farming plant. Modelling, Identification and
Control, 8: 61 – 68.
Haché, R., Daoud, D., Hébert, Y., and Dumas, A. 2014. New developments in lobster resource enhancement in Atlantic Canada
through stage IV larvae release. In Proceedings of the 10th
International Conference and Workshop on the Biology and
Management of Lobsters, Cancún, Mexico. May 18 – 23, 2014.
C. D. Ellis et al.
Puerto Morelos, Quintana Roo, Mexico, Institute of Ocean
Sciences and Limnology, 60.
Hale, M. L., Burg, T. M., and Steeves, T. E. 2012. Sampling for
microsatellite-based population genetic studies: 25 to 30 individuals
per population is enough to accurately estimate allele frequencies.
PLoS ONE, 7: e45170.
Hamasaki, K., and Kitada, S. 2008a. The enhancement of abalone stocks:
lessons from Japanese case studies. Fish and Fisheries, 9: 243 – 260.
Hamasaki, K., and Kitada, S. 2008b. Potential of stock enhancement for
decapod crustaceans. Reviews in Fisheries Science, 16: 164 – 174.
Hamasaki, K., Toriya, S., Shishidou, H., Sugaya, T., and Kitada, S. 2010.
Genetic effects of hatchery fish on wild populations in red sea bream,
Pagrus major, inferred from a partial sequence of mitochondrial
DNA. Journal of Fish Biology, 77: 2123– 2136.
Hayes, B., Sonesson, A. K., and Gjerde, B. 2005. Evaluation of three strategies using DNA markers for traceability in aquaculture species.
Aquaculture, 250: 70– 81.
Henocque, Y. 1983. Lobster aquaculture and restocking in France. In
Proceedings of the First International Conference on Warm Water
Aquaculture—Crustacea: Held on the Brigham Young University
Hawaii Campus, February [sic] 9 – 11, 1983, pp. 235 – 237. Ed. by
G. L. Rogers. Brigham Young University Hawaii Campus, Office of
Continuing Education. 584 pp.
Hervas, S., Lorenzen, K., Shane, M. A., and Drawbridge, M. A. 2010.
Quantitative assessment of a white seabass (Atractoscion nobilis)
stock enhancement program in California: Post-release dispersal,
growth and survival. Fisheries Research, 105: 237– 243.
Hohenlohe, P. A., Bassham, S., Etter, P. D., Stiffler, N., Johnson, E. A.,
and Cresko, W. A. 2010. Population genomics of parallel adaptation
in threespine stickleback using sequenced RAD tags. PLoS Genetics,
6: e1000862.
Hoskin, M. G., Coleman, R. A., von Carlshausen, E., and Davis, C. M.
2011. Variable population responses by large decapod crustaceans
to the establishment of a temperate marine no-take zone.
Canadian Journal of Fisheries and Aquatic Sciences, 68: 185– 200.
Howard, A. E. 1980. Substrate controls on the size composition of
lobster (Homarus gammarus) populations. ICES Journal of Marine
Science, 39: 130– 133.
Howard, A. E. 1988. Lobster behaviour, population structure and enhancement. Symposium of the Zoological Society of London, 59:
355– 364.
Huserbråten, M. B. O., Moland, E., Knutsen, H., Olsen, E. M., Andre, C.,
and Stenseth, N. C. 2013. Conservation, spillover and gene flow
within a network of Northern European marine protected areas.
PLoS ONE, 8: e73388.
Jefferts, K. B., Bergman, P. K., and Fiscus, H. F. 1983. A coded wire identification system for macro-organisms. Nature, 198: 460– 462.
Jeong, D. S., Gonzalez, E. B., Morishima, K., Arai, K., and Umino, T.
2007. Parentage assignment of stocked black sea bream
Acanthopagrus schlegelii in Hiroshima Bay using microsatellite
DNA markers. Fisheries Science, 73: 823 – 830.
Jones, A. G., and Arden, W. R. 2003. Methods of parentage analysis in
natural populations. Molecular Ecology, 12: 2511– 2523.
Jorde, P. E., Knutsen, H., Espeland, S. C., and Stenseth, N. C. 2007.
Spatial scale of genetic structuring in coastal cod Gadus morhua
and geographic extent of local populations. Marine Ecology
Progress Series, 343: 229– 237.
Jørstad, K. E., Agnalt, A-L., Kristiansen, T. S., and Nøstvold, E. 2001.
High survival and growth of European lobster juveniles (Homarus
gammarus) reared communally on a natural-bottom substrate.
Marine and Freshwater Research, 52: 1431 –1438.
Jørstad, K. E., and Farestveit, E. 1999. Population genetic structure of
lobster (Homarus gammarus) in Norway, and implications for enhancement and sea-ranching operation. Aquaculture, 173: 447 – 457.
Jørstad, K. E., Farestveit, E., Kelly, E., and Triantafyllidis, C. 2005b.
Allozyme variation in European lobster (Homarus gammarus)
Lobster stocking requires impact assessment
throughout its distribution range. New Zealand Journal of Marine
and Freshwater Research, 39: 515 –526.
Jørstad, K. E., Kristiansen, T. S., Farestveit, E., Agnalt, A-L., Prodöhl,
P. A., Hughes, M., Ferguson, A., et al. 2009. Survival of
laboratory-reared juvenile European lobster (Homarus gammarus)
from three brood sources in southwestern Norway. New Zealand
Journal of Marine and Freshwater Research, 43: 59 – 68.
Jørstad, K. E., Prodöhl, P. A., Agnalt, A-L., Hughes, M., Apostolidis,
A. P., Triantafyllidis, A., Farestveit, E., et al. 2004. Sub-arctic populations of European lobster, Homarus gammarus, in northern Norway.
Environmental Biology of Fishes, 69: 223– 231.
Jørstad, K. E., Prodöhl, P. A., Kristiansen, T. S., Hughes, M., Farestveit,
E., Taggart, J. B., Agnalt, A-L., et al. 2005a. Communal larval rearing
of European lobster (Homarus gammarus): family identification by
microsatellite DNA profiling and offspring fitness comparisons.
Aquaculture, 247: 275– 285.
Kitada, S. 1999. Effectiveness of Japan’s stock enhancement programmes: Current perspectives. In Stock Enhancement and Sea
Ranching, pp. 103 – 131. Ed. by B. R. Howell, E. Moksness, and
T. Svåsand. Blackwell Science, Oxford.
Kitada, S., and Kishino, H. 2006. Lessons learned from Japanese marine
finfish stock enhancement programmes. Fisheries Research, 80:
101– 112.
Kitada, S., Shishidou, H., Sugaya, T., Kitakado, T., Hamasaki, K., and
Kishino, H. 2009. Genetic effects of long-term stock enhancement
programs. Aquaculture, 290: 69– 79.
Kitada, S., Taga, Y., and Kishino, H. 1992. Effectiveness of a stock enhancement program evaluated by a two-stage sampling survey of
commercial landings. Canadian Journal of Fisheries and Aquatic
Sciences, 49: 1573– 1582.
Knutsen, H., Jorde, P. E., Andre, C., and Stenseth, N. C. 2003.
Fine-scaled geographical population structuring in a highly mobile
marine species: the Atlantic cod. Molecular Ecology, 12: 385– 394.
Knutsen, H., Olsen, E. M., Jorde, P. E., Espeland, S. H., Andre, C., and
Stenseth, N. C. 2011. Are low but statistically significant levels of
genetic differentiation in marine fishes “biologically meaningful”?
A case study of coastal Atlantic cod. Molecular Ecology, 20: 768 – 783.
Laikre, L., Schwartz, M. K., Waples, R. S., Ryman, N., and The GeM
Working Group. 2010. Compromising genetic diversity in the
wild: unmonitored large-scale release of plants and animals.
Trends in Ecology and Evolution, 25: 520– 529.
Lamichhaney, S., Barrio, A. M., Rafati, N., Sundstrom, G., Rubin, C. J.,
Gilbert, E. R., Berglund, J., et al. 2012. Population-scale sequencing
reveals genetic differentiation due to local adaptation in Atlantic
herring. Proceedings of the National Academy of Sciences of the
USA, 109: 19345– 19350.
Latrouite, D., and Lorec, J. 1991. L’experience Francaise de forcage du
recruitement du homard Europeen (Homarus gammarus) resultats
preliminaries. ICES Marine Science Symposia, 192: 93 – 98.
Leber, K. M., Cantrell, R. N., and Leung, P. S. 2005. Optimizing costeffectiveness of size at release in stock enhancement programs.
North American Journal of Fisheries Management, 25: 1596– 1608.
Le Gall, J. Y., Jezequel, M., Lorec, J., and Henocque, Y. 1983. Evaluation
de la variabilité annuelle de recrutement d’une pècherie littorale de
homard, (Homarus gammarus) en Bretagne (Isle d’Houat) 1973–
1980. La Mer, 21: 102 – 112.
Letcher, B. H., and King, T. L. 2001. Parentage and grandparentage assignment with known and unknown matings: application to
Connecticut River Atlantic salmon restoration. Canadian Journal
of Fisheries and Aquatic Sciences, 58: 1812– 1821.
Limborg, M. T., Helyar, S. J., de Bruyn, M., Taylor, M. I., Nielsen, E. E.,
Ogden, R., Carvalho, G. R., et al. 2012. Environmental selection on
transcriptome-derived SNPs in a high gene flow marine fish, the
Atlantic herring (Clupea harengus). Molecular Ecology, 21:
3686– 3703.
Linnane, A., Ball, B., Mercer, J. P., Browne, R., van der Meeren, G.,
Ringvold, H., Bannister, C., et al. 2001. Searching for the early
i47
benthic phase (EBP) of the European lobster: a trans-European
study of cobble fauna. Hydrobiologia, 465: 63– 72.
Linnane, A., Mazzoni, D., and Mercer, J. P. 2000. A long-term mesocosm
study on the settlement and survival of juvenile European lobster
Homarus gammarus (L.) in four natural substrata. Journal of
Experimental Marine Biology and Ecology, 249: 51 – 64.
Linnane, A., and Mercer, J. P. 1998. A comparison of methods for
tagging juvenile lobsters (Homarus gammarus) reared for stock enhancement. Aquaculture, 163: 195 – 202.
Lorenzen, K. 2005. Population dynamics and potential of fisheries stock
enhancement: practical theory for assessment and policy analysis.
Philosophical Transactions of the Royal Society—Biological
Sciences, 360: 171– 189.
Lorenzen, K. 2006. Population management in fisheries enhancement:
gaining key information from release experiments through use of a
size-dependent mortality model. Fisheries Research, 80: 19 – 27.
Lorenzen, K. 2008. Understanding and managing enhancement fisheries systems. Reviews in Fisheries Science, 16: 10 – 23.
Lorenzen, K., Beveridge, M. C. M., and Mangel, M. 2012. Cultured fish:
integrative biology and management of domestication and interactions with wild fish. Biological Reviews, 87: 639 – 660.
Lorenzen, K., Leber, K. M., and Blankenship, H. L. 2010. Responsible approach to marine stock enhancement: an update. Reviews in
Fisheries Science, 18: 189 – 210.
Mann, K. H., and Breen, P. A. 1972. The relation between lobster abundance, sea urchins and kelp beds. Journal of the Fisheries Research
Board of Canada, 29: 603 – 605.
Mercer, J. P., Bannister, R. C. A., van der Meeren, G. I., Debuse, V.,
Mazzoni, D., Lovewell, S., Browne, R., et al. 2001. An overview of
the LEAR (Lobster Ecology and Recruitment) project: results of
field and experimental studies on the juvenile ecology of Homarus
gammarus in cobble. Marine and Freshwater Research, 52:
1291– 1301.
Moksness, E., Støle, R., and van der Meeren, G. I. 1998. Profitability analysis of sea ranching with Atlantic salmon (Salmo salar), Arctic char
(Salvelinus alpinus) and European lobster (Homarus gammarus) in
Norway. Bulletin of Marine Science, 62: 689 – 699.
Moland, E., Olsen, E. M., Andvord, K., Knutsen, J. A., and Stenseth,
N. C. 2011. Home range of European lobster (Homarus gammarus)
in a marine reserve: implications for future reserve design. Canadian
Journal of Fisheries and Aquatic Sciences, 68: 1197 – 1210.
Moland, E., Olsen, E. M., Knutsen, H., Garrigou, P., Espeland, S. H.,
Kleiven, A. R., Andre, C., et al. 2013. Lobster and cod benefit from
small-scale northern marine protected areas: inference from an empirical before-after control-impact study. Proceedings of the Royal
Society—Biological Sciences, 280: 20122679.
Neenan, S. T. V., Hodgson, D. J., Tregenza, T., Boothroyd, D., and Ellis,
C. D. 2014. The suitability of VIE tags to assess stock enhancement
success in juvenile European lobsters (Homarus gammarus).
Aquaculture Research. doi: 10.1111/are.12445.
Nichols, J. H., and Lovewell, S. R. J. 1987. Lobster larvae (Homarus gammarus L.) investigations in Bridlington Bay. Can quantitative sampling be confined to the neuston layer? Journal of Natural History,
21: 825– 841.
Nicosia, F., and Lavalli, K. 1999. Homarid lobster hatcheries: their
history and role in research, management, and aquaculture.
Marine Fisheries Review, 61: 1 – 57.
Ochwada-Doyle, F., Gray, C. A., Loneragan, N. R., and Taylor, M. D.
2010. Using experimental ecology to understand stock enhancement: comparisons of habitat-related predation on wild and
hatchery-reared Penaeus plebejus (Hess). Journal of Experimental
Marine Biology and Ecology, 390: 65– 71.
Øresland, V., and Ulmestrand, M. 2013. European lobster subpopulations from limited adult migration and larval retention. ICES
Journal of Marine Science, 70: 532– 539.
i48
Richards, P. R., and Wickins, J. F. 1979. Lobster culture research.
Ministry of Agriculture, Fisheries and Food, MAFF Directorate of
Fisheries Research, Lowestoft. MAFF Laboratory Leaflet, 47. 33 pp.
Richardson, A. J., Walne, A. W., John, A. W. G., Jonas, T. D., Lindley, J.
A., Sims, D. W., Stevens, D., et al. 2006. Using continuous plankton
recorder data. Progress in Oceanography, 68: 27 – 74.
Rourke, M. L., McPartlan, H. C., Ingram, B. A., and Taylor, A. C. 2009.
Polygamy and low effective population size in a captive Murray cod
(Maccullochella peelii peelii) population: genetic implications for
wild restocking programs. Marine and Freshwater Research, 60:
873– 883.
Ryman, N., and Laikre, L. 1991. Effects of supportive breeding on the
genetically effective population size. Conservation Biology, 5:
325– 329.
Satake, A., and Araki, H. 2012. Stocking of captive-bred fish can cause
long-term population decline and gene pool replacement: predictions from a population dynamics model incorporating densitydependent mortality. Theoretical Ecology, 5: 283 – 296.
Schmalenbach, I., Mehrtens, F., Janke, M., and Buchholz, F. 2011. A
mark-recapture study of hatchery-reared juvenile European lobsters, Homarus gammarus, released at the rocky island of
Helgoland (German Bight, North Sea) from 2000 to 2009.
Fisheries Research, 108: 22 – 30.
Schoo, K. L., Aberle, N., Malzahn, A. M., Schmalenbach, I., and
Boersma, M. 2014. The reaction of European lobster larvae
(Homarus gammarus) to different quality food: effects of ontogenetic shifts and pre-feeding history. Oecologia, 174: 581– 594.
Sekino, M., Hara, M., and Taniguchi, N. 2002. Loss of microsatellite and
mitochondrial DNA variation in hatchery strains of Japanese flounder, Paralichthys olivaceus. Aquaculture, 213: 101– 122.
Sekino, M., Saitoh, K., Yamada, T., Hara, M., and Yamashita, Y. 2005.
Genetic tagging of released Japanese flounder (Paralichthys olivaceus) based on polymorphic DNA markers. Aquaculture, 244:
49 – 61.
Sekino, M., Saitoh, K., Yamada, T., Kumagai, A., Hara, M., and
Yamashita, Y. 2003. Microsatellite-based pedigree tracing in a
Japanese flounder (Paralichthys olivaceus) hatchery strain: implications for hatchery management related to stock enhancement
program. Aquaculture, 221: 255– 263.
Shaklee, J. B., and Bentzen, P. 1998. Genetic identification of stocks of
marine fish and shellfish. Bulletin of Marine Science, 62: 589 – 621.
Shikano, T., Ramadevi, J., and Merila, J. 2010. Identification of localand habitat-dependent selection: scanning functionally important
genes in nine-spined sticklebacks (Pungitius pungitius). Molecular
Biology and Evolution, 27: 2775 – 2789.
Shimada, Y., Shikano, T., and Merila, J. 2011. A high incidence of selection on physiologically important genes in the three-spined stickleback, Gasterosteus aculeatus. Molecular Biology and Evolution, 28:
181– 193.
Shishidou, H., Kitada, S., Sakamoto, T., and Hamasaki, K. 2008. Genetic
variability of wild and hatchery-released red sea bream in Kagoshima
Bay, Japan, evaluated by using microsatellite DNA analysis. Nippon
Suisan Gakkaishi, 74: 183– 188.
Steneck, R. S., and Wahle, R. A. 2013. American lobster dynamics in a
brave new ocean. Canadian Journal of Fisheries and Aquatic
Sciences, 70: 1612– 1624.
Stunz, G. W., and Minello, T. J. 2001. Habitat-related predation on juvenile wild-caught and hatchery-reared red drum Sciaenops ocellatus
(L.). Journal of Experimental Marine Biology and Ecology, 260:
13 – 25.
C. D. Ellis et al.
Sugaya, T., Sato, M., Yokoyama, E., Nemoto, Y., Fujita, T., Hamasaki, K.,
and Kitada, S. 2008. Population genetic structure and variability of
Pacific herring, Clupea pallasii, in the stocking area along the
Pacific coast of northern Japan. Fisheries Science, 74: 579– 588.
Svåsand, T., Agnalt, A-L., Skilbrei, O. T., Borthen, J., and Heggberget, T.
2004. An integrated development programme for marine stocking:
the Norwegian example. FAO Fisheries Technical Paper, 429: 19– 72.
Teacher, A. G. F., Andre, C., Jonsson, P. R., and Merila, J. 2013.
Oceanographic connectivity and environmental correlates of
genetic structuring in Atlantic herring in the Baltic Sea.
Evolutionary Applications, 6: 549 –567.
Trengereid, H. 2012. Shelter seeking and competitive behaviour in
hatchery reared juvenile European lobster (Homarus gammarus)
exposed to predator odours. Thesis for MSc in Aquaculture
Biology, University of Bergen, Norway. 83 pp.
Triantafyllidis, A., Apostolidis, A. P., Katsares, V., Kelly, E., Mercer, J.,
Hughes, M., Jørstad, K., et al. 2005. Mitochondrial DNA variation
in the European lobster (Homarus gammarus) throughout the
range. Marine Biology, 146: 223– 235.
Tringali, M. D. 2006. Bayesian approach for the genetic tracking of cultured and released individuals. Fisheries Research, 77: 159– 172.
Tringali, M. D., Seyoum, S., Wallace, E. M., Higham, M., Taylor, R. G.,
Trotter, A. A., and Whittington, J. A. 2008. Limits to the use of contemporary genetic analyses in delineating biological populations for
restocking and stock enhancement. Reviews in Fisheries Science, 16:
111– 116.
Uglem, I., Næss, H., Farestveit, E., and Jørstad, K. E. 1996. Tagging of
juvenile lobsters (Homarus gammarus) with visible implant fluorescent elastomer tags. Aquacultural Engineering, 15: 499 – 501.
Ulrich, I., Muller, J., Schutt, C., and Buchholz, F. 2001. A study of population genetics in the European lobster, Homarus gammarus
(Decapoda, Nephropidae). Crustaceana, 74: 825 –837.
van der Meeren, G. I. 2000. Predation on hatchery-reared lobsters
released in the wild. Canadian Journal of Fisheries and Aquatic
Sciences, 57: 1794– 1803.
van der Meeren, G. I. 2005. Potential of ecological studies to improve
survival of cultivated and released European lobsters, Homarus gammarus. New Zealand Journal of Marine and Freshwater Research, 39:
399– 424.
Wahle, R. A., and Incze, L. S. 1997. Pre- and post-settlement processes in
recruitment of the American lobster. Journal of Experimental
Marine Biology and Ecology, 217: 179– 207.
Wahle, R. A., and Steneck, R. S. 1991. Recruitment habitats and nursery
grounds of the American lobster, Homarus americanus: a demographic bottleneck. Marine Ecology Progress Series, 69: 231 – 243.
Wahle, R. A., and Steneck, R. S. 1992. Habitat restrictions in early
benthic life: experiments on habitat selection and in situ predation
with the American lobster. Journal of Experimental Marine
Biology and Ecology, 157: 91 – 114.
Whitmarsh, D. 1994. Economic analysis of lobster stock enhancement.
Briefing Paper. CEMARE. University of Portsmouth, UK. 12 pp.
Wickins, J. F., Beard, T. W., and Jones, E. 1986. Microtagging cultured
lobsters, Homarus gammarus (L.), for stock enhancement trials.
Aquaculture and Fisheries Management, 17: 259 – 265.
Wickins, J. F., Roberts, J. C., and Heasman, M. S. 1996. Within-burrow
behaviour of juvenile European lobsters Homarus gammarus (L).
Marine and Freshwater Behaviour and Physiology, 28: 229– 253.
Handling editor: Emory Anderson
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i49 –i58. doi:10.1093/icesjms/fsu177
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Articles
What caused seven consecutive years of low puerulus settlement
in the western rock lobster fishery of Western Australia?
Simon de Lestang1 *, Nick Caputi 1, Ming Feng2, Ainslie Denham 1, James Penn 1, Dirk Slawinski2,
Alan Pearce1,3, and Jason How1
1
Western Australian Fisheries and Marine Research Laboratories, PO Box 20, North Beach, WA 6920, Australia
CSIRO Marine and Atmospheric Research, Private Bag No. 5, Wembley, WA 6913, Australia
3
Curtin University, GPO Box U1987, Perth, WA 6845, Australia
2
*Corresponding author: tel: +61 892030174; fax: +61 892030199; e-mail: simon.delestang@fish.wa.gov.au
Present address: West Australian Fisheries and Marine Research Laboratories 39 Northside Drive, Hillarys. 6025. Western Autsralia.
‡
de Lestang, S., Caputi, N., Feng, M., Denham, A., Penn, J., Slawinski, D., Pearce, A., and How, J. What caused seven consecutive years of
low puerulus settlement in the western rock lobster fishery of Western Australia?. – ICES Journal of Marine Science, 72: i49 – i58.
Received 4 August 2014; revised 17 September 2014; accepted 18 September 2014; advance access publication 19 October 2014.
Puerulus settlement in the western rock lobster fishery has remained below average for seven consecutive years (2006/2007 –2012/2013), with
2008/2009 being the lowest in over 40 years. Examination of the timing of the start of spawning using fishery-independent data since the mid2000s indicated that spawning has been occurring earlier. The low settlement appears related to higher water temperatures at the time of the
onset of spawning (October) since the mid-2000s. Statistical analysis shows that the most (71%) of the variation in puerulus settlement was
explained by the timing of spawning, storm activity during autumn/spring, and offshore water temperatures in February. Earlier spawning may
cause a mismatch with other environmental factors such as peaks in ocean productivity and/or storms that assist the larvae return to the
coast and offshore water temperatures that help the early stage larval growth. These variables produced a plausible hypothesis to explain the
decline in puerulus settlement for these 7 years, including the recruitment failure of 2008/2009. They also predicted the substantial improvement
in settlement for 2013/2014. Egg production levels did not to have a significant relationship with puerulus settlement levels after taking environmental variables into account. Further verification with additional years is required to see whether this relationship is maintained. Global climate
change may influence these environmental factors: the timing of spawning is influenced by water temperature and there has been a reduced trend
of autumn to spring storms off southwest Australia.
Keywords: climate change, environmental effects, puerulus, rainfall, storms, timing of spawning, water temperature, western rock lobster.
Introduction
Puerulus settlement of the western rock lobster (Panulirus cygnus)
fishery of Western Australia (WA; Figure 1) has remained well
below average for seven consecutive years (2006/2007– 2012/
2013), with 2008/2009 being the lowest in over 40 years over the
whole area of the fishery (Caputi et al., 2010a). The puerulus settlement is important in the stock assessment and management of the
fishery as it is a reliable predictor of recruitment to the fishery and
catch 3 –4 years later (Caputi et al., 1995a; de Lestang et al., 2009).
Previous studies had shown that environment factors such as the
strength of the Leeuwin Current and storms in late winter/spring
# International
typically affect the abundance and spatial distribution of puerulus
settlement (Pearce and Phillips, 1988; Caputi et al., 2001; Caputi,
2008). A strong Leeuwin Current (associated with warm water temperatures and influenced by La Niña conditions in the Pacific) has
historically been associated with above-average settlement.
However, during the recent series of low recruitments, the low settlement has occurred despite a strong Leeuwin Current in 2008 and
2011. It was therefore important to identify other environmental
and/or biological influences that may have contributed to the atypically low recruitment and whether any factors identified were
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i50
Figure 1. Location of current (2013/2014) puerulus settlement sites
along the coast and the different fishery zones (A, B, and C) operating in
the fishery.
exhibiting long-term trends, which could explain the new lower
settlement levels.
One of the possible causes of the decline in recruitment was
whether the breeding stock had declined to levels low enough to
cause recruitment overfishing. During the recent period of low
settlement, the spawning stock was within historical levels at all
sites that had been surveyed since the early 1990s which covered
the main part of the fishery (de Lestang et al., 2012). However,
there were concerns about the status of the stock in the northern
part of the fishery (e.g. northern Abrolhos Islands and Big Bank),
which did not represent a major area of the fishery and therefore
had not been regularly surveyed. Significant effort reductions were
introduced in 2008/2009 as a result of the poor puerulus settlement
before these year classes recruited to the fishery. These reductions
were designed to increase residual stock abundance that could be
fished in future poor catch years and to maintain the spawning
stock at safe levels, i.e. above threshold reference levels. The effort
reductions resulted in the average catch declining from the longterm average of 11 000 –5500 t in recent years (de Lestang et al.,
2012).
The lower west coast of WA has been identified as a hotspot for
increasing water temperature (Pearce and Feng, 2007), particularly
in the autumn –winter period (Caputi et al., 2009). There has also
been a significant decline in storm activity and rainfall in the
region during winter (IOCI, 2012). These factors have been previously identified as affecting different aspects of the life history of
the western rock lobster (Caputi and Brown, 1993; Caputi et al.,
2001) and include the impact of water temperature and winter
storms during the phyllosoma larval phase, Leeuwin Current
S. de Lestang et al.
effect on the spatial distributions of puerulus settlement, reduced
migration of white lobsters to northern breeding stocks, and
decreases in the size at maturity and migration (Melville-Smith
and de Lestang, 2006; Caputi et al., 2010b).
The western rock lobster has a relatively long spawning phase of
5 months and larval phase of 9 –11 months. Changes in environmental conditions during any phase of the spawning and larval life
could have contributed to a mismatch between environmental conditions and a later larval or puerulus stage. An oceanographic larval
model has recently identified the timing of spawning as an important factor affecting puerulus settlement, with early egg releases providing the best fit between the model settlement and actual
settlement (Caputi et al., 2014).
Understanding the cause(s) of recruitment variability and identifying any potential long-term trends has important implications in
the stock assessment and management of the fishery. That is, the
management response would be significantly different if the major
cause of the series of low recruitments was due to reduced egg production rather than variations in environmental factors outside of
management control.
A previous assessment of the stock –recruitment –environment
relationship (SRR-E) for western rock lobster showed that environmental conditions explained most of the variation in puerulus
settlement at the coastal sites with spawning stock not being significant (Caputi et al., 1995b); however, this relationship does not
explain the variation in more recent years of puerulus settlement.
This study examines whether the magnitude of the spawning
stock and the timing of spawning are significant factors in the
recent decline in puerulus settlement. It also examines previously
identified key environmental factors that may also be influencing
the puerulus settlement to determine an SRR-E.
Methods
Biological indices
Puerulus settlement
The puerulus settlement data are currently collected monthly from
nine locations (Figure 1) throughout the rock lobster fishery by the
Western Australian Department of Fisheries (DoF; de Lestang et al.,
2012), with Dongara being the first location sampled in 1968. A
standardized annual mean of fishery-wide puerulus settlement
index was obtained from nine locations (Port Gregory, Horrocks,
Abrolhos Is., Dongara, Jurien Bay, Lancelin, Alkimos, Warnbro
Sound, and Cape Mentelle). The index represented an annual
mean settlement rate per collector standardized for location and
month of sampling using a generalized linear model (GLM) and
the “lsmeans” package in R (R Development Core Team, 2014),
and ranged between about 1 and 200. The timing of peak settlement
was determined by fitting a normal distribution to the relationship
between the date and the sum of all puerulus collected (locations
pooled excluding the Abrolhos Islands) in each month of the settlement season (May –April) and ranged between late September and
early January. A 5-year centred moving average was fitted to the
time-series to emphasize its long-term underlying trend.
Egg production
Annual egg production in the fishery was determined using data
from the annual Independent Breeding Stock Survey (IBSS). This
survey is conducted by the DoF at three locations every year
(Lancelin, Dongara, and the Abrolhos Islands) and intermittently
at four other locations (Fremantle, Jurien, Kalbarri, and Big Bank;
Low puerulus settlement in the western rock lobster fishery of Western Australia
de Lestang et al., 2012). The IBSS is conducted over a 10-d period
during the last new moon before mid-November. This period is
just before the annual peak of egg production, which occurs in
November/December (Chubb, 1991). The egg production index
is based on the catch rate of mature female lobsters and a fecundity–
size relationship (de Lestang et al., 2012) at each of the locations
sampled. It represents a mean catch rate of eggs per pot lift standardized for location of capture, month of capture, water depth, ocean
swell, and soak time of the pot using a GLM. The index ranged
between about 12 and 15, and was estimated using the “lsmeans”
package. For more information, see de Lestang et al. (2012).
Timing of spawning
The timing of the start of spawning of P. cygnus varies between years,
most likely in relation to interannual variation in winter and spring
water temperatures. The general pattern of reproduction is for
mating to start around June and fertilization to begin in
September before reaching its maximum with the most mature
females being ovigerous (egg bearing) in November and
December (Chubb, 1991; de Lestang et al., 2012). The numbers of
ovigerous mature females then decline to only a few still being
present by March.
The timing of the onset spawning was examined using the reproductive state of mature females lobsters caught during the IBSS
(October/November). Data from the Kalbarri and Big Bank
regions were not used in constructing the index of onset of spawning
as the timing of some of these surveys were too early for any significant interannual variation the proportion of spawning females to be
detected (September). The annual estimate of the onset of spawning
was a mean spawning stage standardized (“lsmeans”) for location of
capture, year of sampling, the logarithm of timing (days since 1
September) of the survey, and the logarithm of carapace length
using a GLM. Logarithmic transformations associated with the
timing of the survey and carapace length were based on Tukey’s
Bulging rule (Tukey, 1977).
Spawning stage was assigned according to the presence and developmental stage of eggs and the condition of a spermatophoric
mass on mature females. Females were designated as being mature
if they had mated (possess an external spermatophore) or had developing gonads. The stages assigned were: 0 ¼ mature female with no
eggs and no used spermatophoric mass; 1 ¼ recently fertilized eggs;
2 ¼ mid-way through egg development; 3 ¼ eggs fully developed;
4 ¼ residual egg remnants from larval release; 5 ¼ a used spermatophoric mass but no eggs (i.e. the lobster had spawned). The resultant
index ranged between just above zero when the most mature females
had not yet extruded their eggs by mid-October (representing a late
start to spawning) to above one when the most females had extruded
their eggs by mid-October (representing an early start to spawning).
Environmental indices
Leeuwin Current and sea surface temperature
The sea surface temperature (SST) off the lower west coast of WA’s
continental shelf is influenced by the strength of the Leeuwin
Current and direct air sea fluxes in the region (Feng et al., 2008).
The Fremantle sea level has also been used as an indicator of the
strength of the Leeuwin Current and an indicator of the strength
of the eddy structure associated with the current (Feng et al.,
2009). The Leeuwin Current is influenced by El Niño Southern
Oscillation (ENSO) events in the Pacific through the oceanic waveguide, and the Southern Oscillation Index is an indicator of the
strength of the ENSO events. Levels of puerulus settlement have
i51
historically shown a strong relationship with offshore water temperatures in an area west of the fishery (25–288S, 109–1128E)
during the early larval stages (February), so this relationship has
been updated with additional years (de Lestang et al., 2012). The
temperature data used were the ‘NOAA Optimum Interpolation
SST V2’ dataset provided by the NOAA website at “http://www
.esrl.
noaa.gov/psd/data/gridded/data.noaa.oisst.v2.html” (Reynolds
et al., 2002).
Winter storms
Monthly rainfall data were collated from five coastal sites from
Jurien (30.18S) in the north to Fremantle (32.08S) in the south
(http://www.bom.gov.au/climate/cdo/about/cdo-rainfall-feature.
shtml). Rainfall was used as a proxy for the effects of winter storms
on the ocean environment (e.g. swell and waves action) and the
westerly winds associated with storms crossing the coast, which is
crucial to the onshore movement of late stage larvae (Feng et al.,
2011b). Individual months of rainfall data were examined to determine what periods the storms having the biggest impact on puerulus
settlement. Monthly rainfall data ranged from 0 to 342 mm.
Bottom water temperature
The effect of bottom water temperature on the timing of spawning
was examined as it is the most likely cue for the initiation of spawning (Chittleborough, 1976). An empirical time-series of seabed
water temperatures in the spawning grounds was not available;
therefore, model-derived estimates of seabed water temperatures
were collated by combining outputs from two of the Australian
Bureau of Meteorology/CSIRO’s data assimilating oceanographic
models, BRAN and OceanMaps v1 for the periods prior and
during the spawning period (Oke et al., 2008; Brassington et al.,
2012). Outputs from the two models overlapped for a period of
14 months, i.e. BRAN spanned January 1993–April 2008 and
OceanMaps October 2007–November 2011. This overlap period
was used to develop a correction factor to scale the outputs from
the OceanMaps v1 model to that from the BRAN model.
Modelled water temperatures ranged from 17.4 to 20.38C.
Statistical analysis
Stock –recruitment– environment relationship
Whether an SRR-E existed and in what form was examined by comparing annual variation in puerulus settlement, with a combination
of indices, each previously identified as being capable of impacting
on spawning or larval stages. A GLM was employed that utilized a log
transformation for the puerulus and egg production indices to the
skewed nature of their abundance distributions.
The relationship between puerulus settlement and the factors
was each examined individually and in combination. Whether the
addition of factors contributed significantly to the overall model
was assessed using an ANOVA. Multicoefficient models were simplified by removing components (main effects and/or interactions) if
they contributed ,1% to the model’s overall R 2.
Results
Puerulus settlement
The puerulus settlement time-series emphasizes the protracted
period of very low settlement from 2006/2007 to 2012/2013, with
the lowest level of settlement being recorded in 2008/2009 in all
regions (Figure 2). Since this record low settlement, levels of settlement have increased, with the greatest increase occurring in the
i52
S. de Lestang et al.
Figure 2. Standardized mean puerulus settlement time-series from 1968/1969 to 2012/2013 in four regional areas [Dongara/Port Gregory (a),
Abrolhos (b), Jurien/Lancelin (c), and Alkimos/Warnbro/Mentelle (d)] and all regions combined (e). The horizontal dotted line represents each
time-series median between 1984/1985 and 2005/2006 season.
central and northern coastal regions (Figure 2a), where settlement
levels have increased to well above their historical medians in
2013/2014 for the first time in 8 years. The weakest of these increases
has been in the offshore Abrolhos region, where levels are still just
below the median level (Figure 2b).
The timing of peak settlement varied markedly between seasons
ranging from late September during the 1982/1983 season to early
January in the 2012/2013 season (Figure 3). A smoothing of the
time-series displayed a progressive trend, with the timing of settlement changing from being delayed (December) during the late
Low puerulus settlement in the western rock lobster fishery of Western Australia
i53
1960s/early 1970s, to variable between mid-October and midNovember in the 1980s then back to being late post 2010 (Figure 3).
An important aspect of the recent decline in settlement has been the
particularly low settlement during the early part of the season,
August–October, which has historically (before 2008) been a period
of good settlement.
Puerulus–environment relationships
Figure 3. Mean timing of peak puerulus settlement (black line) each
season from 1968/1969 until 2013/2014. A five-point centred
smoothed time-series is shown in grey.
The timing of spawning recorded during the IBSS (measured by egg
stage development) increased during the mid-2000s, with this index
reaching its maximum in 2007 due to the presence of large numbers
of mature females with eroded (used) spermatophoric masses and
no external eggs (Figure 4a). This state indicates that many
females had started reproduction very early and spawned before
the survey. This 2007 spawning season corresponds to the 2008/
2009 puerulus settlement, which was the lowest on record. Apart
from 1994, the most years when the timing of spawning was early
Figure 4. (a) Standardized mean (and 95% C.L.) timing of spawning between 1992 and 2014. (b) The relationship between the timing of spawning
and the logarithm of puerulus settlement, with the years representing the puerulus settlement season. (c) Standardized mean (and 95% C.L.) egg
production from the IBSSs. (d) Relationship between mean egg production and the logarithm of puerulus settlement, with the dotted line
representing the fitted relationships between the two factors and puerulus settlement.
i54
have occurred since 2004 (Figure 4a). This timing of spawning index
displayed a significant (R 2 ¼ 0.23, d.f. ¼ 19, p ¼ 0.026) negative relationship with variation in the puerulus settlement during the subsequent settlement period (Figure 4b). Surveys conducted in years
when the most mature females were yet to initiate spawning (as
represented by a low index value of 0.2) generally coincided
with years of above-average puerulus settlement.
Egg production showed a general increase during the 1990s
of 8%, followed by a similar decline down in the mid-2000s
before increasing by 16% to record levels by 2012 (2013/2014
puerulus settlement year; Figure 4c). The most variation has been
caused by major management changes introduced in 1993 and
2008, which reduced exploitation on the lobster stock, especially
the spawning stock. The relationship between the magnitude of
egg production and that of puerulus settlement in the following
settlement season was very poor (R 2 ¼ 0.02, d.f. ¼ 18, p ¼ 0.60),
with settlement remaining below average in 2011/2012 and
2012/2013 despite the record-high egg production at the time
(Figure 4d).
Since the timing of spawning showed a significant relationship
with puerulus settlement, this index was then modelled in combination with environmental factors previously identified as being
related to puerulus settlement (offshore SST in February and rainfall
levels; Caputi et al., 2001) as well as the egg production index.
S. de Lestang et al.
The addition of rainfall data (as a main effect and interaction)
into the model significantly improved its ability to fit the settlement
data (Figure 5a). The months of rainfall data that most improved the
model were May and October, i.e. at the start and end of the rainy
season (Figure 5a). Since most months between these 2 months
also increased the overall fit of the model (although not significantly), the entire period from May to October was combined to
produce an annual index to represent the onshore storms fronts
that the late larval stages and the puerulus stage would experience.
This combined index significantly (Dd.f. ¼ 2, F ¼ 10.83, p , 0.001)
increased the overall model fit, to a greater extent than any single
month (Figure 5a).
This mid-autumn to mid-spring period (May – October) was
then used to develop a long-term time-series of rainfall (Figure 5b).
This series showed a relatively large interannual variation, ranging
from a maximum average monthly rainfall of 181 mm month21
in 1890 to a minimum of 47 mm month21 in 2006. The years containing the 20 highest average monthly rainfalls all occurred before
1965, whereas 9 of the 20 lowest rainfalls have occurred since 2000
(Figure 5b). The 5-year smoothed average applied to the index
remained relatively steady between 1880 until 1960 with a monthly
average of 100 mm month21. From 1970 through until the end of
the century, this smoothed index dropped by 10% to 90 mm
month21 before declining progressively from the early 2000s until
Figure 5. (a) R 2 values from a model between the timing of spawning/mean coastal rainfall in each month (and a grouping of months
May – October) and the puerulus settlement in that same year. The grey area represents the months chosen to develop a long-term rainfall index of
May – October. *, **, and *** identify in which models the addition of rainfall data significantly (p , 0.05, 0.01, and 0.001, respectively) increased the
models relationship. (b) Long-term time-series of coastal May – October rainfall between 1880 and 2013 with a 5-year rolling average shown.
Low puerulus settlement in the western rock lobster fishery of Western Australia
the current year of 2013 to its record low of 70 mm month21
(Figure 5b).
SST during the early larval phase (February) has historically
(1982– 2007) been a good indicator of strength of puerulus settlement that occurs later in the year, explaining 58% of the variation
during this 26-year period. This correlation declined markedly
with the addition of recent puerulus settlement data, explaining
only 1% of the total variation when the years 2008/2009–2012/
2013 were added. When both breeding time and SST in February
(as a main effect and interaction) were combined into the same
model, that model’s ability to fit the observed variation in puerulus
settlement increased, but not significantly (DR 2 ¼ 0.135; Dd.f. ¼ 2,
p ¼ 0.195) from a model with breeding time alone. When SST in
February was added to the more complex model containing both
timing of spawning and rainfall in May–October, there was a 7% increase in model fit from this addition, although it was just not significant (DR 2 ¼ 0.07; Dd.f. ¼ 1, p ¼ 0.09). Since the addition of
SST was marginally significant in its improvement to the overall
model and SST alone displayed such a strong correlation with puerulus settlement before 2008, even when the dataset was extended
back to 1968 (Caputi et al., 2001), we considered it valid to retain
this factor in the optimal model.
Egg production levels (as a main effect and interaction) were also
re-examined in combination with breeding time (DR 2 ¼ 0.03;
Dd.f. ¼ 1, p ¼ 0.39), breeding time and rainfall in May– October
(DR 2 ¼ 0.12; Dd.f. ¼ 4, p ¼ 0.20) and breeding time, rainfall in
May–October, and SST in February (DR 2 ¼ 0.06; Dd.f. ¼ 3, p ¼
0.33). Egg production levels did not significantly improve the
model description in any of these combinations.
The optimum model (breeding time, rainfall, and SST)
explained 71% of the variation in puerulus settlement, including
two marked increases that occurred in 1995/1996 and 2000/2001,
the historical minimum in 2008/2009, and the dramatic increase
in 2013/2014 (Table 1 and Figure 6a). The model contained a
two-way interaction term between the timing of spawning and
SST and a three-way interaction between all three main effects.
These two interaction terms indicated that both rainfall and SST
were positive influences under most typical spawning stage conditions, but had little impact when the spawning stage was early.
The relationship between SST and settlement was greater in
concert with high rainfall and dampened under conditions of low
rainfall (Figure 6b).
Variation in timing of spawning
The timing of spawning (log-transformed) displayed a significant
(R 2 ¼ 0.30, d.f. ¼ 17, p , 0.015) positive relationship with bottom
water temperatures throughout the breeding stock areas (Figure 7).
Table 1. Summary of the optimal model used to describe the
variation in (log) puerulus settlement.
Breeding time
SST
Btime:SST
Btime:Rain:SST
Constant
Observations
R2
Residual s.e.
F-statistic
*p , 0.001.
Independent variable coefficients (s.e.):
15.713 (30.395)
0.577 (0.442)
21.382 (1.364)
0.006* (0.001)
28.183 (9.640)
21
0.707
0.637 (d.f. ¼ 16)
9.640 (d.f. ¼ 4; 16)
i55
The coolest water temperature in the time-series was predicted in
2001 which coincided with the latest onset of spawning, whereas
the most warm years displayed average to early timings for the
onset of spawning. All 7 years (2005–2011) associated with a low
puerulus settlement period, including the year when spawning
started the earliest in 2007, coincided with above-average bottom
water temperatures (Figure 7).
Discussion and Conclusions
Although levels of puerulus settlement have historically varied
markedly between years, in the mid to late 2000s, the western rock
lobster fishery received an extended period of below average levels
of settlement including the lowest settlement on record in 2008/
2009. This study showed that the timing of spawning (which is
related to bottom ocean temperatures), coastal rainfall levels during
late larval stages to peak settlement, and offshore water temperatures
during the early larval phase collectively explain the most variation
in the puerulus settlement. This relationship emphasizes the climate
sensitivity of this species during its spawning and larval phase.
The early spawning combined with the reduced storm activity
(particularly during early winter) during the larval phase may
have created a mismatch with the peak in food availability and/or
larval transport mechanism back to the coast, as per the ‘match –
mismatch’ hypothesis (Cushing, 1990). Under normal conditions,
the enhanced ocean production in austral autumn/winter, the enhanced onshore flow that feeds the strengthened Leeuwin Current
in austral winter, as well as winter storm activity support the onshore
movement of the larvae. The strengthening Leeuwin Current in
February –April and winter/spring storms that have historically
been identified as affecting the level of settlement for over 35 years
(Caputi et al., 2001) may both affect larval food availability and
transport, so that a mismatch may be occurring due to the early
spawning. Schmalenbach and Franke (2010) identified that an increasing decoupling of the larval peak from optimal external conditions that affected food availability would result in a serious problem
for the European lobsters in the warming North Sea. This hypothesis
needs to be explored further for the western rock lobster using the
oceanographic larval model and environmental data.
Timing of spawning
Demersal water temperatures throughout the spawning grounds of
P. cygnus during the start of spawning (about October) were identified as a factor that may be influencing the timing of the onset of
spawning. Increases in water temperature in October since the
mid-2000s may have resulted in an earlier onset of spawning that
has coincided with the decline in puerulus settlement since 2006/
2007. Studies on historical water temperature trends in WA have
identified the lower west coast of Australia as a hotspot for water
temperature increases over the last 40 –50 years (Pearce and Feng,
2007). Increases were found to be particularly high in the austral
autumn –winter period with little or no increases evident in the
spring –summer period (Caputi et al., 2009). However, the
bottom water temperature in the lobster spawning grounds
during October obtained from oceanographic models shows a possible increasing trend since the early 1990s. This increase may reflect
that in the strength of the Leeuwin Current since the early 1990s
which has been described as a result of decadal climate variation
in the Pacific (Feng et al., 2010), further enhancing the general increase in water temperature that has been occurring off the lower
west coast of Australia (Pearce and Feng, 2007; Caputi et al.,
2009). These two drivers of ocean temperature may have important
i56
S. de Lestang et al.
Figure 6. (a) Observed (black) and estimated +1 s.d (grey) standardized puerulus settlement between 1993/1994 and 2013/2014 seasons. (b) The
relationship between the timing of spawning (year t) and the subsequent level of puerulus settlement (year t + 1/t + 2) under good
environmental conditions (high rainfall and SST in year t + 1) and poor environmental conditions. Seasons shown represent the puerulus
settlement seasons in high (grey) and low (black) rainfall years, with high SST years identified by grey circles.
long-term implications for the rock lobster fishery since the
increases in water temperature under the decadal climate variation
may be expected to revert back to previous levels under periods of
weaker Leeuwin Current, and the future projections are for a longterm weakening of the current (Feng et al., 2012). However under
the greenhouse gas-induced global warming, long-term increases
of water temperature would not be expected to be reversed soon.
Winter storms
Figure 7. Relationship between the modelled bottom water
temperatures throughout the breeding stock areas of the fishery in
October and the index of timing of spawning in that same year.
Storms (measured by rainfall and associated with westerly winds) in
late winter/spring have historically been shown to be positively
related to puerulus settlement (Caputi and Brown, 1993). They
assessed the rainfall for the 22 years, 1969/1970–1990/1991, although the current study has examined the most recent 21 years,
1993/1994–2013/2014, as onset of spawning is only available for
this period. The significance of winter storms in these two separate
periods covering over 40 years confirms the importance of this
factor on the settlement. Westerly wind patterns during the 2010/
2011 settlement season have been unusual, with the westerly component being far weaker than average. This is reflected by the second
lowest winter/spring rainfall on record in 2010, with 2006 having
Low puerulus settlement in the western rock lobster fishery of Western Australia
the lowest rainfall on record. It is likely that these conditions were
adverse for the 2006/2007 and 2010/2011 puerulus settlement.
Previous studies have identified rainfall during July to November
as being a significant factor affecting puerulus settlement (Caputi
and Brown, 1993; Caputi et al., 2001). However, in this study, the
rainfall between May and October when combined with the breeding time index provided a better fit to the variation in puerulus
settlement since the early 1990s including the recent years of low
settlement. Rainfall represents an index of storm activity affecting
the lower west coast of WA, which influences the water vapour transport and is generally associated with westerly winds that may help
transport the larvae back to the coast. The rainfall time-series for
May–October shows a declining trend since the early 2000s with
many of the lowest values occurring since this period. This decline
is part of a long-term trend of declining rainfall in the southwest
of WA that has intensified in the last 10 years and is expected to continue (IOCI, 2012). The IOCI study has demonstrated that storm
development has reduced in the mid-latitudes but increased in the
higher latitudes (50–708S) with the net result of fewer storms affecting southwest WA.
Egg production
The early management intervention in 2008, before the low puerulus settlement year-classes reached legal-size, has resulted in a record
level of breeding stock in recent years (2011 –2013) as measured by
the fishery-independent spawning stock survey, which is also supported by the model assessment of egg production. The effect of
this level of egg production on the poor 2012/2013 and very high
2013/2014 puerulus settlements provides some insights on the
effect of breeding stock on settlement. This relationship indicates
that spawning stock is not a dominant factor, providing little contribution when modelled against puerulus settlement singularly or in
combination with other factors.
Relationship with decadal and longer term climate trends
The Leeuwin Current has historically been identified as a key factor
affecting the level of puerulus settlement, therefore understanding
its long-term trend is important. The Leeuwin Current variability
is essentially driven by the variations and changes in Pacific equatorial easterly winds: the Leeuwin Current has experienced a strengthening trend during the past two decades, which has almost reversed
the weakening trend during 1960s to early 1990s (Feng et al., 2010,
2011a). Currently, most climate models project a weakening trend of
the Pacific trade winds and a reduction in the Leeuwin Current
strength (as well as the Indonesian Throughflow) in response to
greenhouse gas forcing over the next century. Whereas the greenhouse gas forcing induced changes may be obvious in the long-term
climate projection (such as 2100), natural decadal climate variations
still need to be taken into account for assessment of short-term
climate projection, e.g. 2030s (Feng et al., 2012).
The strengthening of the Leeuwin Current in recent decades may
have increased the southward warm water transport and bottom
water temperature on the continental shelf off the west coast of
Australia. In addition, the expansion of the tropics due to global
warming may have pushed mid-latitude storm activity southward,
reducing wave-induced onshore movement off the west coast in
austral winter. Taking into account the status of the breeding
stock and variations in the Leeuwin Current and the fact that the
settlement has remained low for an extended period, a long-term
change in the environmental conditions may be the main cause of
the recent levels of poor puerulus settlement. The increasing
i57
bottom water temperature that may be influencing the start of
spawning and the decline in winter storms are probably both influenced by long-term climate change trends that may continue to
affect the settlement into the future. It will be interesting to see
whether the very good settlement that has occurred in 2013/2014
heralds the return to an improved settlement period or whether it
represents a short-term spike. These will need to be closely monitored.
From a downscaling model simulation, SST may increase by
18C in the next 50 years; in the meantime the strength of the
Leeuwin Current may reduce by 15% (Chamberlain et al. 2012;
Sun et al., 2012). An examination of the eddy energetics in the downscaling model suggests that the eddies will be less energetic in the
future climate, due to the reduction in the Leeuwin Current
strength, which may also affect the ocean primary production in
the region (R. Matear, pers. comm.; Matear et al., 2013). Climate
model projections suggest that precipitation in southwestern
Australia will continue to fall in the future due to further expansion
of the tropics and poleward shift of the storm activity (IPCC, 2013).
While the reduction in the Leeuwin Current may cancel some of the
effects of the increasing ocean temperature, the reduction in the
Leeuwin Current eddy energetics and storm activity off the coast
may still cause the mismatch between the spawning/settlement
and the physical environment. Continued monitoring of the physical environment, including shelf bottom temperature and settlement numbers, is crucial to understanding the sustainability of
the fishery.
Conclusions
The onset of spawning combined with rainfall during May–October
and offshore SSTexplains a significant proportion (0.71) of the variation in puerulus settlement since the early 1990s and provides a
plausible hypothesis to explain the decline in puerulus settlement
in recent years. This relationship and that of the extended model
that includes egg production and SST and their interactions need
to be verified with additional years of data to see whether the relationships are maintained.
The assessment of the possible contribution of a long-term environmental factor(s) to the low puerulus settlement enables the stock
assessment model to factor the low settlement into its assessment, so
that management and industry can take this into account in future
planning. The fishery has been proactive in its adaptation to the low
puerulus settlement by implementing early management actions to
reduce fishing effort and catch before the low puerulus year classes
recruited to the fishery. This has resulted in a lower exploitation
rate on the spawning stock as well as achieving a carry-over of
legal lobsters into the poor year-class years.
Acknowledgements
The authors thank the Fisheries Research and Development corporation for their financial support of this project as well as: CSIRO
Wealth from Oceans Flagship for their support; reviewers and workshop participants; internal reviewers at CSIRO and Department of
Fisheries (Western Australia); and Jenny Moore for assistance
with editing.
References
Brassington, G., Freeman, J. W., Huang, X., Alves, O., Pugh, T., and Oke,
P. R. 2012. Ocean Model, Analysis and Prediction System: version 2.
Centre for Australian Weather and Climate Research CAWCR
Technical Report No. 052. 103 pp.
i58
Caputi, N. 2008. Impact of the Leeuwin Current on the spatial distribution of the puerulus settlement of the western rock lobster (Panulirus
cygnus) fishery of Western Australia. Fisheries Oceanography, 17:
147– 152.
Caputi, N., and Brown, R. S. 1993. The effect of environment on puerulus settlement of the western rock lobster (Panulirus cygnus) in
Western Australia. Fisheries Oceanography, 2: 1 – 10.
Caputi, N., Brown, R. S., and Chubb, C. F. 1995a. Regional prediction of
the western rock lobster, Panulirus cygnus, catch in Western
Australia. Crustaceana, 68: 245– 256.
Caputi, N., Chubb, C. F., and Brown, R. S. 1995b. Relationships between
spawning stock, environment, recruitment and fishing effort for the
western rock lobster, Panulirus cygnus, fishery in Western Australia.
Crustaceana, 68: 213 – 226.
Caputi, N., Chubb, C., and Pearce, A. 2001. Environmental effects on the
recruitment of the western rock lobster, Panulirus cygnus. Marine
and Freshwater Research, 52: 1167– 1174.
Caputi, N., de Lestang, S., Feng, M., and Pearce, A. 2009. Seasonal variation
in the long-term warming trend in water temperature off the Western
Australian coast. Marine and Freshwater Research, 60: 129–139.
Caputi, N., Feng, M., de Lestang, S., Denham, A., Penn, J., Slawinski, D.,
Pearce, A., et al. 2014. Identifying factors affecting the low western
rock lobster puerulus settlement in recent years. Fisheries Research
Report 255, Department of Fisheries, Western Australia. 144 pp.
Caputi, N., Feng, M., Penn, J. W., Slawinski, D., de Lestang, S., Weller, E.,
and Pearce, A. 2010a. Evaluating source-sink relationships of the
western rock lobster fishery using oceanographic modelling
(FRDC Project 2008/087). Fisheries Research Report 209,
Department of Fisheries, Western Australia. 82 pp.
Caputi, N., Melville-Smith, R., de Lestang, S., Feng, M., and Pearce, A.
2010b. The effect of climate change on the western rock lobster
fishery. Canadian Journal of Fisheries and Aquatic Sciences, 67:
85 – 96.
Chamberlain, M., Sun, C., Matear, R. J., Feng, M., and Phipps, S. J. 2012.
Downscaling the climate change for oceans around Australia.
Geoscientific Model Development, 5: 1177– 1194.
Chittleborough, R. G. 1976. Breeding of Panulirus longipes cygnus
George under natural and controlled conditions. Australian
Journal of Marine and Freshwater Research, 27: 499 – 516.
Chubb, C. F. 1991. Measurement of spawning stock levels for the western
rock lobster Panulirus cygnus. Revista de Investigaciones Marinas,
12: 223 –233.
Cushing, D. H. 1990. Plankton production and year-class strength in
fish populations: an update of the match/mismatch hypothesis.
Advances in Marine Biology, 26: 249 – 293.
de Lestang, S., Caputi, N., How, J., Melville-Smith, R., Thomson, A., and
Stephenson, P. 2012. Stock assessment for the west coast rock lobster
fishery. Fisheries Research Report No. 217. Department of Fisheries,
Western Australia. 200 pp.
de Lestang, S., Caputi, N., and Melville-Smith, R. 2009. Using fine-scale
catch predictions to examine spatial variation in growth and catchability of Panulirus cygnus along the west coast of Australia. New
Zealand Journal of Marine and Freshwater Research, 43: 443 –455.
Feng, M., Biastoch, A., Boning, C., Caputi, N., and Meyers, G. 2008.
Seasonal and interannual variations of upper ocean heat balance
off the west coast of Australia. Journal of Geophysical Research,
113: 1 – 16.
Feng, M., Boning, C., Biastoch, A., Behrens, E., Weller, E., and Masumoto,
Y. 2011a. The reversal of the multi-decadal trends of the equatorial
S. de Lestang et al.
Pacific easterly winds, and the Indonesian Throughflow and Leeuwin
Current transports. Geophysical Research Letters, 38: L11604.
Feng, M., Boning, C., Biastoch, A., Behrens, E., Weller, E., and
Masumoto, Y. 2011b. Ocean circulation, Stokes drift and connectivity of western rock lobster population. Canadian Journal of Fisheries
and Aquatic Sciences, 68: 1182– 1196.
Feng, M., Caputi, N., and Pearce, A. 2012. Leeuwin Current. In Marine
Climate Change in Australia, Impacts and Adaptation Responses.
2012 Report Card. Ed. by E. S. Poloczanska, A. J. Hobday, and
A. J. Richardson. CSIRO, Australia. ISBN 978-0-643-10927-8.
Feng, M., McPhaden, M. J., and Lee, T. 2010. Decadal variability of the
Pacific subtropical cells and their influence on the southeast Indian
Ocean. Geophysical Research Letters, 37: L09606.
Feng, M., Waite, A., and Thompson, P. 2009. Climate variability and
ocean production in the Leeuwin Current system off the west coast
of Western Australia. Journal of the Royal Society of Western
Australia, 92: 67 – 81.
Indian Ocean Climate Initiative (IOCI). 2012. Western Australia’s
weather and climate: A synthesis of Indian Ocean Climate
Initiative Stage 3 Research. In SWWA Rainfall: Intensified and
expanded decline. Ed. by B. Bates, C. Frederiksen, and J.
Wormworth. CSIRO and BoM, Australia. 117 pp.
IPCC. 2013. Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of
the Intergovernmental Panel on Climate Change. In Near-term
climate change: Projections and predictability. Ed. by T. F. Stocker, D.
Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y.
Xia, V. Bex, and P. M. Midgley. Cambridge University Press,
Cambridge, New York, NY, UK, USA. 1535 pp.
Matear, R. J., Chamberlain, M. A., Sun, C., and Feng, M. 2013. Climate
change projection of the Tasman Sea from an eddy-resolving ocean
model. Journal of Geophysical Research: Oceans, 118: 2961 –2976.
Melville-Smith, R., and de Lestang, S. 2006. Spatial and temporal variation in the size at maturity of the western rock lobster Panulirus
cygnus George. Marine Biology, 150: 183– 195.
Oke, P. R., Brassington, G. B., Griffin, D. A., and Schiller, A. 2008. The
Bluelink ocean data assimilation system (BODAS). Ocean Modelling,
21: 46–70.
Pearce, A., and Feng, M. 2007. Observations of warming on the Western
Australian continental shelf. Marine and Freshwater Research, 58:
914– 920.
Pearce, A. F., and Phillips, B. F. 1988. ENSO events, the Leeuwin Current,
and larval recruitment of the western rock lobster. Journal du
Conseil, 45: 13 – 21.
R Development Core Team. 2014. R: A language and environment for
statistical computing. Version 2.14.1. R Foundation for Statistical
Computing, Vienna, Austria. http://www.R-project.org.
Reynolds, R. W., Rayner, N. A., Smith, T. M., Stokes, D. C., and Wang, W.
2002. An improved in situ and satellite SST analysis for climate.
Journal of Climate, 15: 1609 –1625.
Schmalenbach, I., and Franke, H. 2010. Potential impact of climate
warming on the recruitment of an economically and ecologically
important species, of the European lobster (Homarus gammarus)
at Helgoland, North Sea. Marine Biology, 157: 1127– 1135.
Sun, C., Feng, M., Matear, R. J., Chamberlain, M. A., Craig, P., Ridgway,
K. R., and Schiller, A. 2012. Marine Downscaling of a future climate
scenario for Australian Boundary Currents. Journal of Climate, 25:
2947– 2962.
Tukey, J. W. 1977. Exploratory Data Analysis. Addison-Wesley, Reading,
MA. ISBN 0-201-07616-0. OCLC 3058187. 688 pp.
Handling editor: Claire Paris
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i59 –i68. doi:10.1093/icesjms/fsu199
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Settlement and early survival of southern rock lobster, Jasus
edwardsii, under climate-driven decline of kelp habitats
Iván A. Hinojosa 1*, Bridget S. Green1, Caleb Gardner 1, and Andrew Jeffs 2
1
Institute for Marine and Antarctic Studies, University of Tasmania, Nubeena Crescent, Taroona, TAS 7053, Australia
Leigh Marine Laboratory, Institute of Marine Science, University of Auckland, P.O Box 349, Warkworth, New Zealand
2
*Corresponding author: tel: +61 3 6227 7277; fax: +61 3 6227 8035 e-mail: [email protected], [email protected]
Hinojosa, I. A., Green, B. S., Gardner, C., and Jeffs, A. Settlement and early survival of southern rock lobster, Jasus edwardsii, under
climate-driven decline of kelp habitats. – ICES Journal of Marine Science, 72: i59 – i68.
Received 28 August 2014; revised 14 October 2014; accepted 16 October 2014; advance access publication 14 November 2014.
Kelp habitats provide food, refuge, and enhance the recruitment of commercially important marine invertebrates. The southern rock lobster, Jasus
edwardsii, supports valuable fisheries in southern Australia and New Zealand. Kelp habitats once covered large areas of inshore reef around
Tasmania, Australia, but coverage has reduced over the last few decades due to climate change, especially off the eastern coast of the island.
We investigated whether the kelp influences the settlement of lobster post-larvae to artificial collectors and how the presence of kelp affected
the overnight predation on the early benthic phase (EBP). Settlement of lobster was tracked over 6 months using crevice collectors
that had either natural or artificial giant kelp, Macrocystis pyrifera attached, or nothing attached (control). Collectors with natural kelp had
higher catches than those with artificial kelp or controls (p ¼ 0.003), which suggested enhanced settlement through chemical attraction.
Additionally, we measured overnight predation of the EBP in barren and kelp habitats individually tethered to artificial shelters. The kelp
habitat was dominated by brown macroalgal species of Ecklonia radiata, Phyllospora comosa, and M. pyrifera, while the barren was devoid of macroalgae. Survival of the EBP was higher (40%) in the kelp habitat than in the barren habitat (10%) due to differences in predation (p ¼ 0.016).
These results suggest that the kelp habitat improves the recruitment of J. edwardsii and that decline in this habitat may affect local lobster productivity along the east coast of Tasmania.
Keywords: climate change, early survival, Jasus edwardsii, kelp forest, Macrocystis, settlement.
Introduction
The availability of suitable habitats for settlement in coastal waters
can be a bottleneck in recruitment of important fisheries species
(Wahle and Steneck, 1991; Sheaves et al., 2014). On temperate
marine reefs, the loss of productive kelp forest is one of the most dramatic ecosystem shifts reported (Filbee-Dexter and Scheibling,
2014). This ecosystem change affects local species diversity, produces changes in foodwebs, reduces available habitat for small
organisms, and can also have localized impacts on the productivity
of important fishery resources (Byrnes et al., 2011; Harley et al.,
2012).
The southern rock lobster, Jasus edwardsii, supports valuable
fisheries in southern Australia and in New Zealand. Females
release between 50 000 and 500 000 larvae per year (Green et al.,
# International
2009), which spend 18 –24 months as phyllosoma stages and are
widely dispersed by ocean currents (Booth and Phillips, 1994;
Bruce et al., 2007). The final stage phyllosoma metamorphose to a
puerulus (post-larva), which can actively migrate into the coastal
habitat (Jeffs et al., 2005). The mechanism by which the pueruli
move inshore to settle is unclear, but it is thought to involve a combination of natural onshore advection and active onshore migration
(Jeffs et al., 2005; Linnane et al., 2010a). Puerulus monitoring of J.
edwardsii on artificial collectors has been undertaken across their
distribution range for 30 –40 years and their settlement trends
have been correlated with the fishable biomass in some specific
regions (Gardner et al., 2001; Booth and McKenzie, 2009; Linnane
et al., 2010b, 2014). The first few days immediately post-settlement
in J. edwardsii (early benthic phase, EBP) is a period of high
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i60
mortality due to predation by fish, octopus, crabs, and other lobsters
(Edmunds, 1995; Mills et al., 2006, 2008). Therefore, the availability
of suitable habitats and shelters for EBP during this time could
greatly affect their survival and recruitment, as occurs in other
lobster species (Butler et al., 2006).
Southeastern Australia is one of the fastest warming coastal areas
in the southern hemisphere, warming three to four times faster than
the global average (Hobday and Pecl, 2014). An intensification of the
South Pacific Gyre has been suggested as the main cause (Ridgway,
2007), where the warm and nutrient-poor East Australian Current
(EAC) is extending further south towards the eastern Tasmanian
coast (Figure 1) (Johnson et al., 2011). The intensification of the
EAC to Tasmania has resulted in the southward range extension
of some marine species as well as benthic habitat changes, such as
the depletion of kelp habitats and changes in their algal composition
(Johnson et al., 2011).
Beds of giant kelp, Macrocystis pyrifera, off eastern Tasmania were
once extensive enough to be harvested commercially (Sanderson,
1987), but have declined dramatically over the last 30 years
(Figure 1; Johnson et al., 2011). This decline in giant kelp is believed
to have been caused by the increasing influence of the EAC with
nutrient-poor and warm waters (Johnson et al. 2011). Kelp habitats
on reefs of eastern Tasmania have also been affected by grazing of the
Australian long-spined sea urchin, Centrostephanus rodgersii, which
has extended its range southward as the EAC has strengthened (Ling
and Johnson, 2009). This range extension is the result of pulses of
larval input from populations to the north of Tasmania over the
past 30 years (Ling et al., 2008; Johnson et al., 2011). At high
density, C. rodgersii overgrazes the productive macroalgal habitat,
including reef previously dominated by M. pyrifera, so that areas
of bare rock reef are created, which are known as barrens habitat
(e.g. Flukes et al., 2012). In northeast Tasmania these urchin
I. A. Hinojosa et al.
barrens can cover 50% of nearshore rocky reef, although the
barrens are not occurring yet towards the south or on the west
coast of Tasmania (Johnson et al., 2011). In northeastern
Tasmania there is also a negative relationship between the abundance of the urchin, C. rodgersii, and the abundance of the subadult
lobsters of J. edwardsii, which is causing concerns for the potential
impact of this habitat change on lobster productivity (Johnson
et al., 2011).
The EBPs of J. edwardsii actively search for food and shelter
(horizontal holes and crevices) in rocky reef at night when they
are active (Butler et al., 1999; Booth, 2001; Hayakawa and
Nishida, 2002). Also, the EBP eat small invertebrates that commonly
inhabit marine vegetation (Edmunds, 1995; Mills et al., 2006). In
this context, kelp could also offer protection from predation
during periods of nocturnal activity, as well as a potential source
of food because of their associated fauna. However, the relative importance of kelp habitat depletion on the recruitment of J. edwardsii
has not been tested. Here we examine whether the giant kelp attracts
settling pueruli and whether the kelp habitat offers protection from
predation during the nocturnal activity of EBP.
Material and methods
We undertook two field experiments to examine the importance of
the kelp habitat on settlement and nocturnal predation of southern
rock lobster, J. edwardsii. The first experiment tested whether the
presence of M. pyrifera influences settlement of pueruli. The
second experiment compared the overnight survival of the EBPs
in a habitat with kelp species (i.e. Ecklonia radiata plus Phyllospora
comosa and M. pyrifera) contrasted to a habitat that was barren of
all macroalgal species due to overgrazing by urchins.
Figure 1. The main oceanographic currents around Tasmania (during winter when most settlement occurs) and the change in the extent of the
visible surface canopy of M. pyrifera in five regions (a) – (e) on the east coast of Tasmania, 1940 – 2007. Bicheno and Fortescue Bay are the sites where
the settlement and the predation experiments were done, respectively. Oceanographic currents based on Cresswell (2000). Visible surface canopy of
M. pyrifera as reported by Johnson et al. (2011).
Settlement and early survival of southern rock lobster, Jasus edwardsii
i61
Giant kelp and J. edwardsii settlement: experiment design
sampled separately using individual mesh bags. No pueruli or
EBP were found on kelp, or kelp holdfast, so sampling was simplified
to include the kelp and the collector in the same bag.
Pueruli and EBP of different developmental stages were found on
the collectors. The first three pueruli stages were categorized as S1–S3,
and the first and the second instar juveniles were categorized as J1 and
J2 (or EBP) according to the schema of Booth (2001): S1’s are completely clear pueruli (without any pigmentation on the body); S2’s
are clear but with a visible hepatopancreas (white); and S3’s have pigmented exoskeletons. J1 and J2’s occur after the moult from puerulus
which is apparent by the change in morphology of the pleopods,
which are shorter as they are no longer used for forward swimming.
J1’s and J2’s are the first clearly benthic phase which are fully pigmented and the two categories were separated by size, as there is a moult
between these stages (J2’s .15 mm cephalothorax length).
To evaluate whether M. pyrifera attracts pueruli of J. edwardsii, we
conducted an experiment on settlement onto crevice collectors at
Bicheno (41852.3′ S; 148818.1′ E) (Figure 1). The crevice collectors
were similar in design to that described and illustrated by Booth
and Tarring (1986) and are used in Australia for monthly pueruli
monitoring programmes (e.g. Linnane et al., 2014). A total of 18
crevice collectors were used with three treatments: six with natural
giant kelp attached—testing for visual, chemical, and structural
cues; six with artificial giant kelp (http://www.pangea.dk/)—
testing for visual and structural cues; and six collectors with no
attachments, as controls. The artificial kelp is used for large public
aquariums. The fronds were made of an extruded polyethylene
plastic and the stipes consisted in a mix of 50% of polyethylene
and 50% ethylene vinyl acetate, both of food-grade standard.
The artificial kelp was created by computer scanning a natural
M. pyrifera from California, USA, which was used for preparing
the production moulds that replicated the structure and colour of
the upper parts of the plant but not the holdfast. Two complete
natural young sporophytes of giant kelp including their holdfasts,
or two artificial giant kelp, were attached onto each collector
using cable ties (two stipes per collector). Both natural and artificial
kelp were 150 cm long and were replaced every month by fresh
natural kelp or conditioned artificial kelp. Artificial kelp was conditioned by leaving it for at least 1 month in an indoor tank with flowthrough seawater.
Collectors were checked each month and kelp was replaced with
new young sporophytes of giant kelp. These were removed using
dive knives from natural reef at 3 – 5 m depth via snorkelling at
Blackmans Bay (4380.5′ S; 147819.7′ E; Figure 1). In the laboratory
they were rinsed with seawater to remove epiphytes and grazers,
and were maintained in running seawater for 15 –20 h before to
transport in a plastic drum with 80 l seawater to the experimental
collectors at Bicheno.
Crevice collectors were arranged on a sandy bottom at 5 m depth
in three parallel rows or lines (A, B, and C) and six columns
(Figure 2A). Each line (row) and column had the three treatments
(natural kelp, artificial kelp, and control, Figure 2A). Two collectors
with the same treatment were placed next to each other (3 m apart)
in the same line. Collectors with different treatments were spaced
4 m apart (Figure 2A). The treatments on collectors were repositioned in a sequence to moving to the left each month after servicing
to randomize any potential spatial effect from their position in the
grid or from the individual collector (Figure 2A). This spatial
design of three lines (A, B, and C) and six collectors per treatment
followed the guidance of Phillips et al., (2001, 2005), which controlled for the spatial effect on pueruli catches on collectors (i.e.
corner and line effect). This included following the recommendation that at least two lines of collectors and at least twice that the
number of columns were required (see statistical analysis).
Pueruli catches were recorded from 16 July to 18 December 2013,
commencing 1 month after the collectors and treatments were first
installed (see Gardner et al., 2001). Collectors were serviced by divers
who first placed a mesh bag around each collector including the kelp
treatment. The collectors were then hauled to the surface for counting the catch of pueruli and EBP and to clean fouling. Each sample
was conducted during the third week of the calendar month, which
was normally close to the full moon in this experiment. In a pilot experiment conducted in the previous year, the kelp (natural
and artificial) attached to the collectors and the collectors were
Giant kelp and J. edwardsii settlement: statistical analysis
We primarily examined whether there were differences in lobster
catches (pueruli and EBP) on collectors due to their spatial arrangement using three independent one-way ANOVA analyses
(Figure 2A). These tests were performed to identify whether
spatial distribution of collectors could confound the statistical analysis of treatment. In these exploratory tests of spatial layout, catches
were analysed independently of collector treatments and months,
but these factors were included in the final analysis. ANOVAs
treated the collector lines (A, B and C) and columns (1–6) as
factors (Figure 2A). We also replicated the approach used by
Phillips et al. (2001) to examine the “neighbour effect” with categories of collectors with two (corner collectors), three (collectors in the
outer lines; Line A and C) and with four neighbours (i.e. collectors in
line B). All datasets were transformed using log(x + 1), to achieve
equal variances and normality from this count data.
As the number of lobsters on the collectors varied by line (p ¼
0.002), but not by column (p ¼ 0.663), and the “neighbour
effect” was not detected (p ¼ 0.135) (Table 1) a generalized linear
model (GLM) was used to evaluate differences in lobster catches
among collector treatments including collector line as a factor (see
Randomized Block design in Quinn and Keough, 2002). The full
GLM thus included factors of collector line (A, B, and C), treatment
(natural, artificial kelp, and control), and month (July to
December). ANOVA tests and GLM were conducted using R (R
Development Core Team, 2008).
Differences between treatments and through time in the proportion of the different developmental stages of the pueruli encountered on the collectors was tested using two-way contingency
tables to establish whether the treatment may influence the timing
of development of pueruli or their retention after settlement. One
table considered the treatments on the collectors (natural and artificial kelp, and control) and pueruli stages (S1 –S3, and Juveniles) as
variables. In a second table, month (July to December) and puerulus
stage (S1 –S3, and Juveniles) were considered as variables. The
interactions between these variables were not considered in these
analyses to simplify the results and interpretation. However, a
graphical representation of all data are reported.
Kelp habitat and J. edwardsii early survival
Survival rate of the EBPs of J. edwardsii was compared in kelp and in
urchin barren habitats by tethering juveniles (J1) to artificial shelters
located in both habitats. The experiment was done in August 2013
(Austral winter) at Fortescue Bay (4388.4′ S; 147858.0′ E; Figure 1)
i62
I. A. Hinojosa et al.
Figure 2. Schema of the experimental set-ups: (A) arrays of crevice collectors at Bicheno used to examine the effect of Macrocystis (artificial and
natural Macrocystis, and Control) on the catch rate of lobster (pueruli plus early-stage juveniles). The collectors were arranged in three rows
(collector lines) parallel with the shore line (A, B, and C, respectively) and in six columns. Each collector lines and column had three treatments,
respectively. The position of the treatments was repositioned each month to the left (as indicated by arrow). (B) The camera system and artificial
shelter used to evaluate the survival of the EBP at Fortescue Bay. Insert photo shows the EBP and the position of the monofilament tether.
on rocky reef at 10 m depth. The kelp habitat was dominated
by brown macroalgae species of the common kelp, E. radiata
(70%); the crayweed, P. comosa (20%); and M. pyrifera
(10%), while the urchin barren was devoid of macroalgae.
Artificial shelters were used to standardize variation in natural
crevice structure between habitats following Mislan and Babcock
(2008), Mills et al. (2008), and Weiss et al. (2008) designs. The artificial shelters were constructed with two PVC tubes of 25 mm diameter and a PVC elbow to form an “L” shape with a base of 75 mm
length and a height of 120 mm (Figure 2B). The base tube was
blocked inside with a plastic barrier to create a shelter of 25 mm
diameter × 75 mm length, which was similar in size to holes used
by EBP on natural substrate (Edmunds, 1995; Butler et al., 1999).
The shelter was attached to a 1 kg dive lead with cable ties. A steel
rod was attached to the top of the L-shaped pipe with a swivel at
the tip and secured using electrical insulation tape. Electrical insulation tape was wrapped around the structure to cover any gaps
that could otherwise be used by EBPs. All the artificial shelters
were placed in a tank with running seawater for 1 month before
use to enable surface fouling to develop.
i63
Settlement and early survival of southern rock lobster, Jasus edwardsii
Table 1. Independent ANOVA analyses of the effect of collector orientation (columns and lines), and “neighbour effect” on J. edwardsii
pueruli catch rate (replicates were a single monthly sample from a single collector).
Analysis
ANOVA
Data ¼ log10(x + 1)
Factors
Lines
Source of variation
Between groups
Residual
Total
Tukey test
C vs. A
C vs. B
B vs. A
d.f.
2
105
107
Power ¼ 0.050: 0.842
Diff Means
0.226
0.0533
0.173
SS
1.004
8.298
9.302
MS
0.502
0.079
F
6.349
p
3
3
3
q
4.82
1.137
3.683
P
0.003
0.702
0.028
P
0.002
ANOVA
Data ¼ log10(x + 1)
Columns
Source of variation
Between groups
Residual
Total
d.f.
5
102
107
Power ¼ 0.050: 0.050
SS
0.287
9.015
9.302
MS
0.057
0.088
F
0.649
P
0.663
ANOVA
Data ¼ log10(x + 1)
Neighbour
Source of variation
Between groups
Residual
Total
d.f.
2
105
107
Power ¼ 0.050: 0.218
SS
0.348
8.954
9.302
MS
0.174
0.085
F
2.041
P
0.135
Lines (A, B, and C), columns (1–6), and number of collector next to each other (2, 3, and 4) were considered as factors in three independent analyses. Line A was
close to the shore and line C to seaward (see Figure 2A). Significant values in bold.
Pueruli used for the experiment were collected in July 2013 from
Bicheno using the crevice collectors and the method described
above. The animals were transported in seawater in an 80 l plastic
drum then maintained in a 400 l tank with flow-through seawater.
After the pueruli reached the J1 stage they were fed with open
mussels ad libitum. Only J1 stages were used in this experiment
(n ¼ 70; 11.05 + 0.16 mm CL, mean + SE).
Tethers were attached to the EBP using the method described in
Mislan and Babcock (2008). A 250 mm long monofilament nylon
line (0.25 mm diameter, 2.7 kg breaking strength) was tied to the
cephalothorax of the EBP between the 4th and 5th pairs of periopods
and fixed to the cephalothorax with a drop of cyanoacrylate glue
(Figure 2B). This tether was then attached to the swivel on the rod
at the top of the shelter. Tethering from the top of shelter allowed
the EBP to move between the inside of the shelter and around the
seabed within a circular area of 100 mm of radius.
EBP were attached to shelters in the laboratory at least 24 h before
being transferred to the experimental site at Fortescue Bay in 70 l
containers with seawater. Divers randomly placed the shelters
with the EBP attached into the kelp and urchin barren habitats.
The entrance to each shelter with the EBP inside was blocked with
a piece of E. radiata during deployment and this was removed
around 30 min before sunset at around 5:40 pm. Divers returned
the following morning around 30 min after sunrise at around 6:45
am and recorded whether the EBP remained in the artificial
shelter. Three runs of the experiment were conducted over three
nights during 1 week. On the first night, 15 units were placed in
each habitat treatment; on the second night 8 and 7 units were allocated to kelp and barren, respectively, and on the third night 12 and
13 were allocated, respectively.
Tethers clearly restrict the escape behaviour of the EBP so results
do not provide an estimate of absolute survival but rather an indication of relative differences in survival between habitats. Bias in relative estimates from tethering can occur where a predation event
occurs through facilitation for an uncommon predator. This was
evaluated previously in this species by Mills et al. (2008) who
made a special note of predation by crabs because they used the
tether to “reel in” EBP so these events were excluded from their analyses. In these experiments we reduced the problem of predation by
crabs through the overhead tethering design and also used video to
observe predators as recommended by Mills et al. (2008). This was
conducted using Go-ProTM cameras quipped with extra battery
capacity (www.gopro.com), and illuminated using a dive torch
with a red filter (Figure 2B). The red filter was an INON red Filter
LE or a red cellophane paper. The cellophane filter was based on
the Kodak Wratten Gelatin Filter #29 which transmits light at
.600 nm, which is outside the visible spectrum of lobsters and
octopus (e.g. Weiss et al., 2006). Cameras were set on time lapse
and recorded every 5 s. Fourteen EBPs in the kelp habitat and
11 EBPs in the barren habitat were video monitored (03:26 +
0:24 h × EBP21 and 02:44 + 00:23 h × EBP21, respectively,
mean + SE).
We tested whether the survival of the EBP was different between
the two habitats with a simple goodness-of-fit test using an equal
survival in both habitat as the expected frequency (Quinn and
Keough, 2002). The data were pooled from the three sampling
runs undertaken within the same week and were considered as
the observed frequency in the test (Quinn and Keough, 2002).
The effect of habitat type on the mean per cent of time that the
EBP spent inside the shelter was determined with a t-test using
data derived from analyses of video recordings, with each deployment providing a single sample and considering the total time of
the video or when the EBP was predated. The proportional data
were transformed taking the arcsine of the square root of the data
(Zar, 2010).
Results
Effect of kelp on settlement
A total of 488 pueruli and EBP were caught within 18 crevice collectors over 6 months. The position of collector lines influenced the
number of lobsters caught (p ¼ 0.001, Table 2). Line A, which
i64
I. A. Hinojosa et al.
Table 2. Generalized lineal model on the number of lobsters per collector considering months (July – December), treatments (natural kelp—
NAT; control of no kelp—CON; and artificial kelp—ART) and collector lines (A– C) as factors.
Analysis
GLM (quasi Poisson)
data ¼ log10(x + 1)
Factors
Months
Treatments
Lines
Source of variation
NULL
Months
Treatments
Lines
D.F.
Deviance
0.895
1.348
1.591
d.f.
107
102
100
98
Residual
19.105
18.210
16.862
15.271
5
2
2
Linear hypotheses
NAT—CON
NAT—ART
CON—ART
Estimate
20.317
0.244
20.072
Std. error
0.100
0.098
0.106
z value
23.172
2.498
20.687
Pr(.|z|)
0.004
0.033
0.771
P
0.180
0.003
0.001
Significant values in bold.
was closest to the coast, caught significantly fewer lobsters (3.14 +
0.45, mean + SE) than the line B and C (4.94 + 0.66; 5.47 + 0.59;
p , 0.05, Tables 1 and 2). Treatment had an effect, with significantly
larger numbers of lobsters on collectors with the natural kelp treatment (5.97 + 0.64, mean + SE) than the control collectors (3.47 +
0.44; p ¼ 0.004; Table 2 and Figure 3). The collectors with natural
kelp also had more lobsters than the collectors with artificial kelp
(4.11 + 0.64; p ¼ 0.033; Table 2). These patterns were consistent
during time and there were no differences in catches among
months (p ¼ 0.180; Table 2 and Figure 3). Over the 6 months of
sampling, 215 lobsters were caught in the collectors with natural
kelp treatment (44%), 148 (30%) in artificial kelp, and 125 (26%)
on the control collectors.
Puerulus development stages on collectors
The development stage of pueruli was similar among treatments
(x 2 ¼ 8.73; p ¼ 0.189; Figure 4A and B), but different between sampling months (x 2 ¼ 79.24; p , 0.001; Figure 4A and B). This difference appeared to be mainly driven by changes in the proportion of
the clear pueruli (S1) and juvenile lobsters (J1 and J2). Clear pueruli
were absent in September and their proportion was higher in July
and in November (Figure 4B). The proportion of juvenile lobsters
was lower during July and much higher in September and
October (Figure 4B).
Kelp forest and post-settlement survival
Overnight survival of the EBP in this experiment was very low with
,50% over each single night of the experiment. Tethered EBPs had
higher overnight survival in the kelp environment (42.9%) than in
the urchin barren (11.4%) (x 2 ¼ 6.37; p ¼ 0.016, Figure 5).
The shelter occupancy (proportion of time that the EBP spend
inside the shelter) was similar between habitats (31.4 + 10.6% in
kelp habitat and 61.3 + 14.1% in urchin barren, mean + SE)
(t ¼ 2 1.579; d.f. ¼ 23; p ¼ 0.128). However, the statistical power
of the analysis was 0.203, which was below the desired power of
0.800 (Cohen, 1988). Only one predator in the urchin barren
was clearly identified from video recordings which was a rosy
wrasse (Pseudolabrus rubicundus), with predation occurring
before sunset. Non-lethal interactions with five fish species and
two crab species with tethered EPB were recorded with pueruli successfully using the shelter to avoid predation (Table 3). From these
non-lethal interactions, wrasse species appeared only in daylight,
whereas the cod and the cardinal fish were more active during the
night (Table 3). Additionally, when the EBP were stalked by fish
species, they remained inside the shelter, and when the EBP were
stalked by crab they escaped by a quick backward swim response
Figure 3. Mean (and standard error) of the number of J. edwarsii
caught per collector treatment between July and December of 2013
(see Table 2 for statistical analysis).
(Supplementary material, Video S1). A predation event by
another labridae fish, the yellow-saddled wrasse (Notolabrus fucicola), was identified by the presence of the monofilament tether
line and swivel in their mouth—this was observed by the diver
rather than captured on video (Supplementary material, Photo S2).
Discussion
The results suggest that recruitment of J. edwardsii is enhanced in
kelp habitat. Catches of the pueruli and EBP on collectors with the
giant kelp M. pyrifera were higher than controls, suggesting that
coverage of giant kelp on natural reef would attract pueruli to
settle. Overnight survival of tethered EBP was also higher in kelp environment where an inherent quality that we have not identified,
such as the structural complexity of this habitat, or differences in
habitat-associated predators, reduced the predation events.
Climate change-induced loss of the kelp habitat would thus be
expected to affect settlement and subsequent recruitment of lobsters
to this valuable fishery (Johnson et al., 2011).
Settlement processes in J. edwardsii
The metamorphosis from phyllosoma to puerulus stage occurs far
offshore from coastal reefs and has been detected beyond the continental shelf (Jeffs et al., 2001). The mechanisms that pueruli use to
return to coastal reef is poorly understood, but there is evidence
that they orientate shoreward with long-distance cues and active
swimming in combination with physical transport (Jeffs et al.,
2005; Linnane et al., 2010a). Pueruli of J. edwardsii, as other spiny
lobsters, have sensory mechanisms that allow them to perceive
and use multiple cues from their environment to guide their swimming, including celestial, wind, acoustic, and magnetic cues (Jeffs
Settlement and early survival of southern rock lobster, Jasus edwardsii
i65
Figure 4. (A) Proportion of the different development stages of pueruli and juveniles of J. edwardsii collected between July and December of
2013 from three treatments on crevice collectors. (B) Contingency table results over months [critical value of chi-square (0.05, 15) ¼ 25.0] and
treatments [critical value of chi-square (0.05, 6) ¼ 12.6]. (+) and (2) represent the observed frequencies above and below the expected
frequencies (chi-square values .5, respectively).
et al., 2005; Kough et al., 2014). Of particular interest for this study,
laboratory experiments with P. argus showed that pueruli could
identify coastal waters and they would actively select these waters
in choice experiments suggesting the use of chemical cues
(Goldstein and Butler, 2009).
Once the pueruli of J. edwardsii reach coastal waters, they may
also use local cues to identify the settlement habitat, as do other
several decapod crustacean species (Gebauer et al., 2003; Stanley
et al., 2010). Observed higher pueruli settlement on collectors
with the natural kelp, M. pyrifera, than collectors with plastic kelp
and controls suggested that pueruli were receptive to chemical
cues from the kelp, in addition to the physical structure of kelp.
This local chemical attraction may guide the pueruli to locate
a hole and/or crevice to settle, or attract them from surrounding
habitats.
The time that the J. edwardsii puerulus (S1) takes to moult to juvenile (J1) has been estimated to be between 1 –3 weeks depending
on temperature (Booth and Stewart, 1993; Fitzgibbon et al., 2014),
i66
I. A. Hinojosa et al.
although it could be affected by some chemical and physical cues
from the habitat (e.g. Gebauer et al., 2003; Stanley et al., 2011).
For example, reduction in the time to moult in pueruli of P. argus
has been reported in response to the presence of chemical cues
from the red macroalgae, Laurencia spp., which dominates their
settlement habitat in Florida, USA (Goldstein and Butler, 2009). If
the giant kelp expedites the time to moult in J. edwardsii we could
expect a difference in the proportion of the pueruli stages according
with the treatments in our collectors. However, we did not find any
difference of the pueruli stages among our collector treatments suggesting no effect on the time to moult within a monthly sampling.
Post-settlement predation on J. edwardsii
Structured habitats such as seagrass beds, mangroves, corals,
mussels, and kelp can provide the necessary resources and environmental conditions to facilitate the settlement and survival of many
other species, including those of commercial importance (Grutter
and Irving, 2007; Halpern et al., 2007; Almanza et al., 2012). For
example, it is well established that mangroves enhance the settlement of some fish species where the two most important pressures
governing this habitat selection are the risk of predation and the
availability of food (e.g. Laegdsgaard and Johnson, 2001). In addition, direct predatory avoidance behaviour has been reported as
an important driver to settle in some competent decapod larvae
(Tapia-Lewin and Pardo, 2014). Our results suggest that kelp
habitat may provide some predator protection, or there is a reduction of the potential predators for J. edwardsii.
The EBP of J. edwardsii inhabit small holes and crevices scaled to
their body size, and forage only at night across small areas to reduce
encounter rates with predators (Edmunds, 1995; Butler et al., 1999).
Our experiment mimics nocturnal emergence and offered standardized artificial shelters similar to natural holes to avoid variation in
this factor between kelp and urchin habitats. We found that the overnight predation of tethered EBP was lower in the kelp environment
than in the urchin barren, indicating that the kelp habitat reduces
predation encounters possibly by obstructing prey detection and
also changing the ability of predators to manoeuvre and capture
prey (e.g. Syms and Jones, 2000; Rogers et al., 2014). This pattern
of reduced predation risk with increased structural complexity
appears a general pattern observed in kelp habitats (Villegas et al.,
2008; Kovalenko et al., 2012). The presence of kelp is also likely to
increase the availability of food for juvenile lobsters (Caputi et al.,
2013) and modify the predator composition and abundance
(Pérez-Matus and Shima, 2010), although these factors were not directly examined in our experiments.
Climate changes and fishery implications in Tasmania
Figure 5. Proportion of tethered EBP of J. edwardsii surviving in kelp
and urchin barren habitats. Numbers in parentheses are replicates.
Critical value of chi square (0.05, 1) ¼ 3.8. (+) and (2) represent the
observed frequencies above and below the expected frequencies
(chi-square values .3, respectively).
Our experiments confirm that settlement of J. edwardsii involves
active searching (Booth, 2001) and that local settlement is increased
by the presence of giant kelp. In addition, during the solitary and
cryptic phase of EBP, kelp habitat reduced predation, possibly
through modification of the suite of associated predators or their
ability to successfully capture EBP. This facultative association
between J. edwardsii and kelp habitats may “facilitate” (sensu
Bruno et al., 2003) and enhance the recruitment in this lobster
species.
Our results imply that climate change-driven loss of kelp habitat
may affect local productivity of the J. edwardsii fishery. There was a
period of widespread declines in recruitment of J. edwardsii from
2004 to 2011, including in areas where the loss of kelp had occurred
(Hartmann et al., 2013). Although this decline in recruitment occurred at a broader scale than the loss of kelp and was due to
other processes (Linnane et al., 2010b), the impact of loss of kelp
examined here could be expected to compound the effects of these
low recruitment events. Jasus edwardsii recruit into a range of habitats where holes and crevices are available (Edmunds, 1995; Butler
et al., 2006), however, the kelp habitat “facilitates” the recruitment
Table 3. Potential predators (predation risk) of the EBP of J. edwardsii recorded on video in kelp and urchin barren habitats.
Osteichthyes
Decapoda
Scientific name
Notolabrus fusicola
Pictilabrus laticlavius
Meuschenia sp.
Pseudolabrus rubicundus
Vincentia sp.
Lotella rhacina
Strigopagurus strigimanus
Plagusia chabrus
Common name
Yellow-saddled wrasse
Senator wrasse
Leatherjacket
Rosy wrasse
Cardinal fish
Beardie (cod)
Hermit crab
Speedy crab
Recorded
1
2
1
1
1
2
1
2
Habitat
Kelp
Kelp
Kelp
Barren
Barren
Barren
Barren
Kelp & Barren
Dark/daylight
Daylight
Daylight
Daylight
Daylight
Dark
Dark
Dark
Dark
Settlement and early survival of southern rock lobster, Jasus edwardsii
i67
process. Jasus edwardsii is possibly less vulnerable to a loss of kelp
habitat than some other lobster species (Wahle and Steneck, 1992;
Butler et al., 2006) because a change in kelp abundance would not
be expected to create local extinction of the fishery but rather
reduce productivity due to a reduction in recruitment.
Some quantification and future monitoring of the effect of
habitat change may be possible in combination of data collected
in puerulus settlement monitoring programmes, which have been
conducted across Australia and New Zealand since early 1990s
(Gardner et al., 2001; Booth and McKenzie, 2009; Linnane et al.,
2014). Change in survival of settling lobsters as a consequence of
habitat change is expected to affect the relationship between
pueruli catch data from these artificial collectors and measures of
future recruitment of adult lobsters to the fishery.
Bruno, J. F., Stachowicz, J. J., and Bertness, M. D. 2003. Inclusion of facilitation into ecological theory. Trends in Ecology & Evolution, 18:
119– 125.
Butler, M. J., MacDiarmid, A. B., and Booth, J. D. 1999. The cause and
consequence of ontogenetic changes in social aggregation in New
Zealand spiny lobsters. Marine Ecology Progress Series, 188:
179– 191.
Butler, M. J., Steneck, R. S., and Herrnkind, W. F. 2006. Juvenile and
adult ecology. In Lobsters: Biology, Management, Aquaculture and
Fisheries, pp. 263– 309. Ed. by B. F. Phillips. Blackwell Scientific
Press, Oxford.
Byrnes, J. E., Reed, D. C., Cardinale, B. J., Cavanaugh, K. C., Holbrook,
S. J., and Schmitt, R. J. 2011. Climate-driven increases in storm frequency simplify kelp forest food webs. Global Change Biology, 17:
2513– 2524.
Caputi, N., de Lestang, S., Frusher, S., and Wahle, R. A. 2013. The impact
of climate change on exploited lobster stocks. In Lobsters: Biology,
Management, Aquaculture and Fisheries, pp. 84 – 112. Ed. by B. F.
Phillips. Wiley-Blackwell, New York.
Cohen, J. 1988. Statistical Power Analysis for the Behavioral Science.
Lawrence Erlbaum, New York, NY, USA.
Cresswell, G. 2000. Currents of the continental shelf and upper slope of
Tasmania. In Tasmania and the Southern Ocean, pp. 21– 30. Ed. by
M. R. Banks, and M. J. Brown. Papers and Proceedings of the Royal
Society of Tasmania, Vol. 133(3). ISSN 0080-4703. CSIRO Division
of Oceanography, Hobart, Tasmania, Australia 7001.
Edmunds, M. 1995. The Ecology of the Juvenile Southern Rock
Lobster, Jasus edwardsii (Hutton 1875) (Palinuridae). University
of Tasmania, Hobart, Australia.
Filbee-Dexter, K., and Scheibling, R. E. 2014. Sea urchin barrens as alternative stable states of collapsed kelp ecosystems. Marine Ecology
Progress Series, 495: 1– 25.
Fitzgibbon, Q. P., Jeffs, A. G., and Battaglene, S. C. 2014. The Achilles
heel for spiny lobsters: the energetics of the non-feeding post-larval
stage. Fish and Fisheries, 15: 312 –326.
Flukes, E. B., Johnson, C. R., and Ling, S. D. 2012. Forming sea urchin
barrens from the inside out: an alternative pattern of overgrazing.
Marine Ecology Progress Series, 464: 179– 194.
Gardner, C., Frusher, S. D., Kennedy, R. B., and Cawthorn, A. 2001.
Relationship between settlement of southern rock lobster pueruli,
Jasus edwardsii, and recruitment to the fishery in Tasmania,
Australia. Marine and Freshwater Research, 52: 1271– 1275.
Gebauer, P., Paschke, K., and Anger, K. 2003. Delayed metamorphosis in
decapod crustaceans: evidence and consequences. Revista Chilena de
Historia Natural, 76: 169– 175.
Goldstein, J. S., and Butler, M. J. 2009. Behavioral enhancement of
onshore transport by postlarval Caribbean spiny lobster (Panulirus
argus). Limnology and Oceanography, 54: 1669 – 1678.
Green, B. S., Gardner, C., and Kennedy, R. B. 2009. Generalised linear
modelling of fecundity at length in southern rock lobsters, Jasus
edwardsii. Marine Biology, 156: 1941– 1947.
Grutter, A., and Irving, A. D. 2007. Positive interactions in marine communities. In Marine Ecology, pp. 110 – 137. Ed. by S. D. Connell, and
B. M. Gillanders. Oxford University Press, Oxford.
Halpern, B. S., Silliman, B. R., Olden, J. D., Bruno, J. P., and Bertness,
M. D. 2007. Incorporating positive interactions in aquatic
restoration and conservation. Frontiers in Ecology and the
Environment, 5: 153 – 160.
Harley, C. D. G., Anderson, K. M., Demes, K. W., Jorve, J. P., Kordas, R.
L., Coyle, T. A., and Graham, M. H. 2012. Effects of climate change
on global seaweed communities. Journal of Phycology, 48:
1064– 1078.
Hartmann, K., Gardner, C., and Hobday, D. 2013. Tasmanian rock
lobster fishery 2011/12. Fishery Assessment Report. The Institute
for Marine and Antarctic Studies, University of Tasmania, Hobart,
Australia. 61 pp. http://www.imas.utas.edu.au (last accessed 27
August, 2014).
Supplementary data
Supplementary material is available at ICESJMS online version of
the manuscript.
Acknowledgments
We thank Ruari Colquhoun, David Fallon, Kylie Cahill, Graeme
Ewing, Sarah Pyke, Hugh Jones, John Keane, Luis Henriquez and
Lara Marcus for assistance in the field and collecting data. We also
thank Rafael Leon who helped with the statistical analyses.
Particular thanks to Stewart Frusher who contributed to the original
research ideas. The settlement experiment could not have been conducted without the support of the Melbourne Aquarium which provided the plastic Macrocystis. We also thank two anonymous
reviewers for their contribution. Funding for this research has
been provided by the PhD scholarship BECAS-Chile programme
to I.A.H., the Australian Research Council Linkage project
(Project No. LP120200164) from B.S.G, the ANZ Trustees programme “Holsworth Wildlife Research Endowment”, as well as,
the Australian Research Council’s Industrial Transformation
Research Hub funding scheme (Project number IH 120100032)
from A.J.
References
Almanza, V., Buschmann, A. H., Hernandez-Gonzalez, M. C., and
Henriquez, L. A. 2012. Can giant kelp (Macrocystis pyrifera) forests
enhance invertebrate recruitment in southern Chile? Marine
Biology Research, 8: 855– 864.
Booth, J. D. 2001. Habitat preferences and behaviour of newly settled
Jasus edwardsii (Palinuridae). Marine and Freshwater Research, 52:
1055– 1065.
Booth, J. D., and McKenzie, A. 2009. Strong relationships between levels
of puerulus settlement and recruited stock abundance in the red rock
lobster (Jasus edwardsii) in New Zealand. Fisheries Research, 95:
161– 168.
Booth, J. D., and Phillips, B. F. 1994. Early life history of spiny lobster.
Crustaceana, 66: 271 – 294.
Booth, J. D., and Stewart, R. A. 1993. Puerulus settlement in the red rock
lobster, Jasus edwardsii. New Zealand Fisheries Assesment Research
Document 93/5. http://koha.cloudapp.net/cgi-bin/koha/opacMARCdetail.pl?biblionumber=150026 (for printed copy location).
Booth, J. D., and Tarring, S. C. 1986. Settlement of the red rock lobster,
Jasus edwardsii, near Gisborne, New Zealand. New Zealand Journal
of Marine and Freshwater Research, 20: 291– 297.
Bruce, B. D., Bradford, R., Griffin, D., Gardner, C., and Young, J. 2007.
Larval transport and recruitment processes of southern rock lobster.
FRDC and CSIRO Marine and Atmospheric Research publication
No. 2002/007. Australia. http://www.marine.csiro.au (last accessed
27 August, 2014).
i68
Hayakawa, Y., and Nishida, S. 2002. Diel behavior of the puerulus and
the first instar juvenile of the red rock lobster, Jasus edwardsii.
Fisheries Science, 68: 393 – 396.
Hobday, A. J., and Pecl, G. T. 2014. Identification of global marine hotspots: sentinels for change and vanguards for adaptation action.
Reviews in Fish Biology and Fisheries, 24: 415 – 425.
Jeffs, A. G., Montgomery, J. C., and Tindle, C. T. 2005. How do spiny
lobster post-larvae find the coast? New Zealand Journal of Marine
and Freshwater Research, 39: 605 –617.
Jeffs, A. G., Nichols, P. D., and Bruce, M. P. 2001. Lipid reserves used by
pueruli of the spiny lobster Jasus edwardsii in crossing the continental shelf of New Zealand. Comparative Biochemistry and Physiology
A-Molecular and Integrative Physiology, 129: 305 –311.
Johnson, C. R., Banks, S. C., Barrett, N. S., Cazassus, F., Dunstan, P. K.,
Edgar, G. J., Frusher, S. D., et al. 2011. Climate change cascades: Shifts
in oceanography, species’ ranges and subtidal marine community
dynamics in eastern Tasmania. Journal of Experimental Marine
Biology and Ecology, 400: 17 – 32.
Kough, A. S., Paris, C. B., and Staaterman, E. 2014. In situ swimming and
orientation behavior of spiny lobster (Panulirus argus) postlarvae.
Marine Ecology Progress Series, 504: 207– 219.
Kovalenko, K. E., Thomaz, S. M., and Warfe, D. M. 2012. Habitat complexity: approaches and future directions. Hydrobiologia, 685:
1 – 17.
Laegdsgaard, P., and Johnson, C. 2001. Why do juvenile fish utilise mangrove habitats? Journal of Experimental Marine Biology and
Ecology, 257: 229 – 253.
Ling, S. D., and Johnson, C. R. 2009. Population dynamics of an ecologically important range-extender: kelp beds versus sea urchin
barrens. Marine Ecology Progress Series, 374: 113– 125.
Ling, S. D., Johnson, C. R., Frusher, S., and King, C. K. 2008.
Reproductive potential of a marine ecosystem engineer at the edge
of a newly expanded range. Global Change Biology, 14: 907 – 915.
Linnane, A., Gardner, C., Hobday, D., Punt, A., McGarvey, R., Feenstra,
J., Matthews, J., et al. 2010b. Evidence of large-scale spatial declines in
recruitment patterns of southern rock lobster Jasus edwardsii, across
south-eastern Australia. Fisheries Research, 105: 163 – 171.
Linnane, A., James, C., Middleton, J., Hawthorne, P., and Hoare, M.
2010a. Impact of wind stress anomalies on the seasonal pattern of
southern rock lobster (Jasus edwardsii) settlement in South
Australia. Fisheries Oceanography, 19: 290 –300.
Linnane, A., McGarvey, R., Gardner, C., Walker, T. I., Matthews, J.,
Green, B., and Punt, A. E. 2014. Large-scale patterns in puerulus
settlement and links to fishery recruitment in the southern rock
lobster (Jasus edwardsii), across south-eastern Australia. ICES
Journal of Marine Science, 71: 528– 536.
Mills, D. J., Gardner, C., and Johnson, C. R. 2006. Experimental reseeding of juvenile spiny lobsters (Jasus edwardsii): comparing survival
and movement of wild and naive lobsters at multiple sites.
Aquaculture, 254: 256– 268.
Mills, D. J., Johnson, C. R., and Gardner, C. 2008. Bias in lobster tethering experiments conducted for selecting low-predation release sites.
Marine Ecology Progress Series, 364: 1 – 13.
Mislan, K. A. S., and Babcock, R. C. 2008. Survival and behaviour of juvenile red rock lobster, Jasus edwardsii, on rocky reefs with varying
predation pressure and habitat complexity. Marine and Freshwater
Research, 59: 246– 253.
Pérez-Matus, A., and Shima, J. S. 2010. Disentangling the effects of
macroalgae on the abundance of temperate reef fishes. Journal of
Experimental Marine Biology and Ecology, 388: 1 – 10.
Phillips, B. F., Cheng, Y. W., Cox, C., Hunt, J., Jue, N. K., and
Melville-Smith, R. 2005. Comparison of catches on two types of
I. A. Hinojosa et al.
collector of recently settled stages of the spiny lobster (Panulirus
argus), Florida, United States. New Zealand Journal of Marine and
Freshwater Research, 39: 715 – 722.
Phillips, B. F., Melville-Smith, R., Cheng, Y. W., and Rossbach, M. 2001.
Testing collector designs for commercial harvesting of western rock
lobster (Panilirus cygnus) puerulus. Marine and Freshwater
Research, 52: 1465– 1473.
Quinn, G. P., and Keough, M. J. 2002. Experimental Design and Data
Analysis for Biologists. Cambridge University Press, Cambridge,
NY, UK. xvii, 537 pp.
R Development Core Team. 2008. R: A Language and Environment for
Statistical Computing. R Foundation for Statistical Computing,
Vienna, Austria.
Ridgway, K. R. 2007. Long-term trend and decadal variability of the
southward penetration of the East Australian Current. Geophysical
Research Letters, 34: 1 – 5.
Rogers, A., Blanchard, J. L., and Mumby, P. J. 2014. Vulnerability of coral
reef fisheries to a loss of structural complexity. Current Biology, 24:
1000– 1005.
Sanderson, J. C. 1987. A survey of the Macrocystis pyrifera (L.) C. Agardh
stocks on the east coast of Tasmania (Technical Report 21). Technical
Report. Department of Sea Fisheries, Tasmania, Hobart, TAS. http://
eprints.utas.edu.au/11925/ (last accessed 27 August 2014).
Sheaves, M., Baker, R., Nagelkerken, I., and Connolly, R. M. 2014. True
value of estuarine and coastal nurseries for fish: incorporating
complexity and dynamics. Estuaries and Coasts. Doi:10.1007/
s12237-014-9846-x
Stanley, J. A., Radford, C. A., and Jeffs, A. G. 2010. Induction of settlement in crab megalopae by ambient underwater reef sound.
Behavioral Ecology, 21: 113– 120.
Stanley, J. A., Radford, C. A., and Jeffs, A. G. 2011. Behavioural response
thresholds in New Zealand crab megalopae to ambient underwater
sound. Plos One, 6: 1– 9.
Syms, C., and Jones, G. P. 2000. Disturbance, habitat structure, and the
dynamics of a coral-reef fish community. Ecology, 81: 2714– 2729.
Tapia-Lewin, S., and Pardo, L. M. 2014. Field assessment of the predation risk—food availability trade-off in crab megalopae settlement.
PLoS One, 9(4): e95335. doi: 10.1371/journal.pone.0095335.
Villegas, M. J., Laudien, J., Sielfeld, W., and Arntz, W. E. 2008.
Macrocystis integrifolia and Lessonia trabeculata (Laminariales;
Phaeophyceae) kelp habitat structures and associated macrobenthic
community off northern Chile. Helgoland Marine Research, 62:
S33– S43.
Wahle, R. A., and Steneck, R. S. 1991. Recruitment habitats and nursery
grounds of the American lobster Homarus americanus—a demographic bottleneck. Marine Ecology Progress Series, 69: 231 – 243.
Wahle, R. A., and Steneck, R. S. 1992. Habitat restrictions in early
benthic life—experiments on habitat selection and in situ predation
with the American lobster. Journal of Experimental Marine Biology
and Ecology, 157: 91 – 114.
Weiss, H. M., Lozano-Alvarez, E., and Briones-Fourzan, P. 2008.
Circadian shelter occupancy patterns and predator-prey interactions
of juvenile Caribbean spiny lobsters in a reef lagoon. Marine Biology,
153: 953– 963.
Weiss, H. M., Lozano-Alvarez, E., Briones-Fourzan, P., and
Negrete-Soto, F. 2006. Using red light with fixed-site video
cameras to study the behavior of the spiny lobster, Panulirus argus,
and associated animals at night and inside their shelters. Marine
Technology Society Journal, 40: 86 – 95.
Zar, J. H. 2010. Biostatistical Analysis. Prentice-Hall/Pearson, Upper
Saddle River, NJ. xiii, 944 pp.
Handling editor: Stéphane Plourde
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i69 –i78. doi:10.1093/icesjms/fsv093
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
American lobster nurseries of southern New England
receding in the face of climate change
Richard A. Wahle 1*, Lanny Dellinger 2, Scott Olszewski 3, and Phoebe Jekielek 1
1
School of Marine Sciences, Darling Marine Center, University of Maine, Walpole, ME 04573, USA
Rhode Island Lobstermen’s Association, Providence, RI, USA
3
Rhode Island Division of Fish and Wildlife, Jamestown, RI, USA
2
*Corresponding author: tel: + 1 207 563 8297; fax: +1 207 563 3119; e-mail: [email protected]
Wahle, R. A., Dellinger, L., Olszewski, S., and Jekielek, P. American lobster nurseries of southern New England receding in the face of
climate change. – ICES Journal of Marine Science, 72: i69 – i78.
Received 28 August 2014; revised 24 April 2015; accepted 28 April 2015; advance access publication 27 May 2015.
Historically, southern New England has supported one of the most productive American lobster (Homarus americanus) fisheries of the northeast
United States. Recently, the region has seen dramatic declines in lobster populations coincident with a trend of increasingly stressful summer
warmth and shell disease. We report significant declines in the abundance, distribution, and size composition of juvenile lobsters that have accompanied the warming trend in Narragansett Bay, Rhode Island, since the first comprehensive survey of lobster nurseries conducted there in 1990. We
used diver-based visual surveys and suction sampling in 1990, 2011, and 2012, supplemented by post-larval collectors in 2011 and 2012. In 1990,
lobster nurseries extended from the outer coast into the mid-sections of the bay, but by 2011 and 2012 they were largely restricted to the outer coast
and deeper water at the mouth of the bay. Among five new study sites selected by the lobster fishing industry for the 2011 and 2012 surveys, the
deepest site on the outer coast (15– 17 m depth) harboured some of the highest lobster densities in the survey. Separate fixed site hydrographic
monitoring at 13 locations in the bay by the Rhode Island Division of Fish and Wildlife recorded an approximately 2.08C increase in summer surface
temperatures over the period, with 2012 being the warmest on record. Additional monitoring of bottom temperatures, dissolved oxygen and pH at
our sampling sites in 2011 and 2012 indicated conditions falling below physiological optima for lobsters during summer. The invasion of the Asian
shore crab, Hemigrapsus sanguineus, since the 1990s may also be contributing to declines of juvenile lobster shallow zones (,5 m) in this region.
Because lobster populations appear increasingly restricted to deeper and outer coastal waters of southern New England, further monitoring of
settlement and nursery habitat in deep water is warranted.
Keywords: American lobster, climate change, Narragansett Bay, nursery habitat, settlement.
Introduction
Climate change is having tangible impacts on commercially important fisheries of coastal and shelf waters worldwide (Harley et al.,
2006). This is especially true of the North Atlantic where ocean
warming has outpaced that of other parts of the world, resulting
in altered phenology and geographic range shifts for a host of
marine fish and invertebrates with important implications for the
future of the fisheries and their management (Nye et al., 2009;
Hare et al., 2012; Mills et al., 2013; Pinsky et al., 2013). It is rare,
however, to have the opportunity to evaluate the impact of
climate change on the earliest life stages of marine species since
most long-term datasets are derived from commercial catch or
# International
fishery-independent trawl surveys that selectively sample largerbodied older juveniles and adults and do not target nursery habitat.
Southern New England has experienced dramatic changes in its
population of American lobster, Homarus americanus, over the past
two decades. These changes suggest that summer conditions in
coastal areas are becoming inhospitable to this iconic, historically
abundant and commercially important species. In the mid-1990s,
lobster abundance and harvests were at historical highs in southern
New England. Since then, the onset of shell disease and sharp
declines in larval settlement to the region’s coastal nurseries have
preceded precipitous declines in adult abundance and landings
that are now only a fraction of what they were two decades ago
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i70
(Wahle et al., 2009a). Extreme summer temperatures, storminduced mixing and rainfall during 1999–2000 triggered a
massive die-off in Long Island Sound that resulted in a 75% drop
in lobster landings and a local collapse of the commercial fishery
(Pearse and Balcom, 2005). The Atlantic States Marine Fisheries
Commission’s (ASMFC) Lobster Technical Committee has
deemed the southern New England lobster stock to be in a state of
recruitment failure (ASMFC, 2009). In contrast, further north,
where cooler temperatures prevail, abrupt warming events have
caused shifts in phenology. Most recently, 2012 brought a recordbreaking ocean heat wave that triggered an early molt and inshore
migration in the Gulf of Maine, resulting in an over-supply of
landed lobsters that overwhelmed processors and severely depressed
the price (Mills et al., 2013).
The American lobster is not the only species to have experienced
dramatic changes associated with ocean warming in southern New
England (Keller et al., 1999; Nixon et al., 2004, 2009). In the wellstudied Narragansett Bay and nearby coastal waters, the winter–
spring diatom bloom has all but disappeared (Keller et al., 1999;
Oviatt et al., 2002; Nixon et al., 2009), eelgrass has been on the
decline (Bintz et al., 2003), comb jellies have increased (Sullivan
et al., 2001), the composition of the groundfish assemblage has
changed (Jeffries and Terceiro 1985; Jeffries, 2002), and the invasive
green crab, Carcinus maenas, has been replaced by the equally invasive
Asian shore crab, Hemigrapsis sangineus (Lohrer and Whitlatch, 2002).
Of particular relevance to lobster, summer water temperatures
in southern New England have exceeded a long-recognized 208C
physiological threshold with historical frequency (McLeese and
Wilder, 1958; Pearse and Balcom, 2005; Glenn and Pugh, 2006).
The net result is the northward shift in the lobster’s thermal niche,
a scenario whereby coastal southern New England will no longer
be a hospitable nursery to the American lobster in the coming
decades (Frumhoff et al., 2007). It is, therefore, valuable to
compare current distributions with previous baselines.
In 1990, Wahle (1993) undertook the first comprehensive
survey of Rhode Island lobster nurseries in Narragansett Bay, including 17 sites surveyed visually by divers, 6 of which were also
suction-sampled (Figure 1; Supplementary Table S1). In retrospect,
it is apparent that these surveys were done at a time when the lobster
fishery and larval settlement were at historical highs. After the 1990
survey, a subset of the lower bay sites plus additional outer coastal
sites continued to be monitored by suction sampling under
what would become the American Lobster Settlement Index
(ALSI), a collaborative monitoring programme of lobster nurseries
throughout New England and Atlantic Canada. Larval settlement
in Rhode Island has fallen to historical lows in recent years (Wahle
et al. 2009a, 2012; Supplementary Figure S1). Concurrently, since
1990, summer surface temperatures in the bay have risen at an
average rate of 0.098C per year (Olszewski, 2013; Supplementary
Figure S2). However, the original Narragansett Bay sites have not
been re-surveyed since 1990, despite the dramatic changes reported
over the intervening years. In this study, we repeated and augmented
the Narragansett Bay lobster nursery survey in a collaborative
project among the University of Maine, the Rhode Island lobster
fishing industry, and Rhode Island Division of Fish and Wildlife
(RI DFW) to evaluate how lobster nurseries within the bay had
changed over the two decades since the first survey.
Methods
This study repeated the diver-based visual and suction sampling
surveys at the same sites visited in 1990 and augmented them at
R. A. Wahle et al.
Figure 1. Narragansett Bay and Rhode Island’s outer coast showing the
17 sites surveyed in 1990 and resurveyed in 2011 and 2012 (circles).
Black circles denote sites subject to both visual surveys and suction
sampling by divers; white circles are sites where only visual surveys were
conducted. Crosses indicate industry-selected sites for the 2011– 2012
survey; white crosses denote the site where collectors and visual and
suction samples were done; black cross denotes the sole
industry-selected site where collectors were not deployed. RI, Rhode
Island, MA, Massachusetts. Map includes 20 m isobath. Inset shows
Narragansett Bay on a map of New England, USA.
several locations using passive post-larval collectors to provide
an independent comparison to the diver-based methods. We also
sampled five additional study sites selected by participating
lobster fishers to address industry concerns that previous sampling
was insufficient to characterize current juvenile lobster distributions. Visual surveys and suction sampling were completed and
passive collectors were deployed at these new sites. In addition to
the methodology outlined below, further details for visual and
suction sampling appear in Wahle (1993), and for the use of
passive collectors, Wahle et al. (2009b).
Visual surveys
We conducted visual surveys in July and early August to provide a
rapid assessment of general bay-wide patterns in the benthic
lobster population. Divers haphazardly placed 1-m2 quadrats on
cobble–boulder habitat at each site. Four divers surveyed in all
20 quadrats in 30 min. Divers carefully overturned rocks and
searched crevices, working slowly so as to avoid suspending
sediment that hampers visibility. Data recorded for each quadrat
included cobble and algae coverage and counts of lobster, crab,
and other larger fauna. We also measured lobster carapace length
and crab carapace width to the nearest millimetre.
We conducted visual surveys in 2011 and 2012 at all 17 sites surveyed in 1990 (listed in Supplementary Table S1). Each year, we
included two additional industry-selected sites which were only
i71
American lobster nurseries of southern New England
sampled for 1 year. These sites ranged in depth from 3 to 5 m below
mean low water. In 2012, we added one additional “deep” site, Fort
Wetherill, at 17 m below MLW.
Suction sampling surveys
Suction sampling was necessary because visual methods are less
effective for recently settled young-of-year (YoY) and 1-year-old
lobsters. However, all sizes of lobster were sampled in the process.
We conducted suction sampling from late August to the first week
of September to coincide with RI DFW long-term monitoring of
stations on Rhode Island’s outer coast at the end of the settlement
season. Diver pairs worked together to vacuum the contents of
0.5 m2 quadrats placed in cobble– boulder habitat. Rocks were
removed during the process to expose hidden infauna. Samples
were retained in a mesh bag secured to the top of the sampler;
bags were changed between quadrats. We sampled 20 quadrats per
site. Two diver pairs completed this task in ,1 h, making it feasible
to have completed 3 –4 sites in a day. Sample bags were returned to
the laboratory where they were sorted fresh or were frozen to be
sorted later. Lobsters were measured, sex was determined, and
claws counted. Associated fauna of interest included several
species of native and introduced crabs (C. maenas, Dyspanopeus
sayi, Cancer borealis, and Hemigrapsus sanguineus) and benthic
fish which were also identified to species, counted, and measured
(see Supplementary Tables S2 and S3 for a full list of crabs, lobsters,
and fish collected by suction sampling).
Collector surveys
Passive post-larval collectors were also deployed to provide a
measure of settlement independent of the diver-based method.
Previous studies have documented the comparability of the
collector- and suction sampling-based methods (Wahle et al.,
2009b, 2012). Passive collectors were made of 10-gauge vinyl-coated
wire with a 3.7-cm (1.5 inch) mesh. They measured 61.0 × 91.5 ×
15.0 cm deep, providing 0.55 m2 in the floor area. A cover
made of the same wire mesh served to retain the rocks and allow
the free passage of post-larval lobsters. The floor and walls were
lined inside with 2-mm plastic mesh (PetmeshTM ) to retain newly
settled lobsters, crabs, and other organisms during retrieval.
Collectors were filled with clean, rounded cobbles ranging in diameter from 10 to 15 cm. Collectors filled with rocks weighed 80 kg.
Each collector was fitted with a bridle to enable lifting in a horizontal
position, which is important to retaining collections during retrieval. An earlier pilot study demonstrated that losses of organisms
were negligible when retrieved in this manner (Wahle et al., 2009b).
Collectors were deployed in late May/early June each year.
Twenty collectors were deployed at each of the six original and
two industry-selected sites, comprising in all 160 collectors per
year. We retrieved collectors at the end of the settlement season at
the same time suction sampling was conducted (late August –
early September). Once on deck, lids were removed, rocks rinsed
with a hose to remove sediment, and then carefully removed to
inspect for lobsters and other organisms. All lobsters, crabs, and
fish were measured. Lobsters were further inspected to determine
sex and the number of claws, noted for evidence of shell disease,
and then released.
Water quality
RI DFW provided surface and bottom temperatures from 1990
to 2013 for 13 fixed locations throughout the bay sampled during
its monthly bottom trawl survey (Supplementary Figure S2;
Olszewski, 2013). In addition, we collected water quality data specific to our benthic sampling sites in 2011 and 2012, although no
hydrographic data were collected in 1990. At the subset of six original sites and two industry-selected sites subject to the more intensive
sampling, we deployed continuously recording Tid-BitTM temperature loggers set to record temperature hourly. Temperature loggers
were fastened to post-larval collectors and deployed at 5– 7 m depth.
Furthermore, we conducted SeaBird CTD casts in July and August to
assess thermal profiles at selected deeper mid-channel locations next
to sampling sites. We conducted hydrographic surveys of our
benthic sampling sites during 22 –24 August 2012, by measuring
surface and bottom temperature, salinity, dissolved oxygen (DO),
and pH at using a YSI 556 Multiparameter System.
Statistical analysis
We evaluated the statistical significance of lobster demographic
changes between 1990, 2011, and 2012 using the visual survey dataset
encompassing the 17 sites originally visited in 1990. Although visual
surveys may undersample the smallest size classes of lobster, this
dataset comprised the largest number of lobsters and the greatest
spatial coverage compared with the suction sampling and collectorbased methods. To evaluate changes in lobster population density,
we used a Wilcoxon signed-ranks test to assess the three possible
between-year comparisons of site means. Differences in mean carapace
length among years were tested using a single-factor ANOVA, with each
year treated as a separate level (numbers of lobsters for each year
denoted in Figure 2b). Similarly, we used a single-factor ANOVA to
assess spatial shifts in occupied lobster nurseries from 1990 to 2011
and 2012, by evaluating differences in the latitude for sites where lobsters were present (not weighted by density) in each year’s survey.
Latitude runs perpendicular to the roughly N–S axis of the bay’s
basin, and therefore served as the most useful dimension with
which to assess spatial shifts; the northern and southern most sites
of the 17 are separated by 0.2278 of latitude or 25.2 km (0.2278 ×
111 km degree21 latitude). Finally, to evaluate the link between
hypoxia and acidification at our sampling sites, we used a least-squares
linear regression to test the statistical relationship between dissolved
oxygen concentration and pH measured at the surface and bottom
during our hydrographic survey.
Results
Visual surveys
Diver visual surveys of Narragansett Bay lobster nurseries conducted in 2011 and 2012 revealed a dramatic reduction in lobster
density, distribution, and size composition compared with 1990
levels (Figure 2). In 1990, lobster densities among the 17 sites averaged 0.51 lobsters per m2 (+0.14 SE). By 2011 and 2012, densities
had fallen to 0.13 (+0.05 SE) and0.06 (+0.02 SE) individuals per m2,
respectively, representing a statistically significant nearly 10-fold
overall decline from 1990 to 2012 (Wilcoxon signed-ranks test: 1990
vs. 2011, Z ¼ 22.86, P ¼ 0.002; 1990 vs. 2012, Z ¼ 23.17, P ¼
0.004; 2011 vs. 2012, Z ¼ 22.03, P ¼ 0.04). Notably, most of the
lobsters counted in the 2012 visual survey were from the single deep
site, Fort Wetherill, selected by the fishing industry, that was not
part of the 1990 survey.
The spatial distribution of lobster nurseries also receded south
towards the outer coast from 1990 to 2011 and 2012 (Figure 2a).
Whereas juvenile lobsters in 1990 extended well into the midsection of the bay, they were largely restricted to the lower bay and
outer coastal locations in the recent surveys. In the context of the
i72
R. A. Wahle et al.
Figure 2. Visual surveys of H. americanus nurseries in Narragansett Bay, 1990, 2011, and 2012. (a) Lobster density (n m22 + 1 SE) listed by the site
north-to-south (n ¼ 20 1-m2 quadrats sampled per site), and (b) lobster size composition. Industry selected denoted with asterisks.
sampling domain of 17 visual sampling sites, spanning 0.2278
of latitude (25.2 km), this represents a statistically significant
southward shift of 0.0388 (4.2 km) in the centre of the nursery
population from 41.5068N (+0.0148 SE) in 1990 to 41.4688N
(+0.0068 SE) in both 2011 and 2012 (one-way ANOVA, F ¼
3.94, d.f. ¼ 2, P ¼ 0.03).
We also recorded a significant upwards shift in the size composition of the nursery population in the recent surveys, reflecting the
i73
American lobster nurseries of southern New England
absence of the smaller juveniles that were prevalent two decades
earlier (Figure 2b). Mean carapace length significantly increased
from 31.5 mm (+1.2 mm SE) in 1990 to 36.7 (+1.9 SE) and
41.0 (+2.2 SE) in 2011 and 2012, respectively (one-way ANOVA,
F ¼ 6.74, d.f. ¼ 2, P ¼ 0.001).
mode in the size distribution, but in 2011 they were absent from
the suction samples, and in 2012 only one was found (Figure 3b).
In 2012, all the lobsters collected by suction sampling came from
Fort Wetherill, the single deep site, and the only site, historical or
industry-selected, where lobsters of any size were collected by this
method.
Suction sampling surveys
Suction sampling reinforced evidence of diminishing lobster recruitment of YoY lobsters (Figure 3). In the 1990 survey, YoY lobsters extended from the outer coast to mid-bay locations
(Figure 3a). In 1990, YoY lobsters also represented the dominant
Passive collectors
All collectors were retrieved, except for 15 lost at Southwest Point
and 5 lost at Sachuest Point, as a result of Hurricane Irene in 2011.
All lobsters found in collectors in 2011 came from the lower bay
Figure 3. Suction sampling of H. americanus nurseries in Narragansett Bay 1990, 2011, and 2012. (a) Lobster density (n m22 +1 SE) listed by the site
north-to-south, (n ¼ 200.5-m2 quadrats sampled per site), and (b) lobster size composition. Industry selected denoted with asterisks.
i74
and outer coast, whereas lobsters from 2012 collectors were exclusively from the outer coast. We recorded only one YoY in 2011,
and five in 2012, all at the two outermost coastal sites (Figure 4a).
No YoY lobsters were found in collectors deployed at the industryselected sites. Overall, the size composition of lobsters in collectors
was similar to that of suction samples (Figure 4b).
Although a quantitative assessment of associated fauna is not
included in this report, several points are worth highlighting
on the decapod crustacean and fish assemblages (Supplementary
Table S2). The Asian shore crab (H. sanguineus) is a notable
recent introduction, not reported in the 1990 survey, but counted
at all sites except the deep one (Fort Wetherill) in the recent
survey. At least 16 species of fish were identified and counted in collectors deployed during this study (Supplementary Table S3).
Several species of fish, such as cunner, black sea bass, toad fish,
and eels, were reported at remarkable densities of tens per collector.
Several of these species are known to be predators of juvenile lobster
(Wahle et al., 2013). Despite the high predator diversity and abundance in collectors, it is noteworthy that YoY and juvenile lobster
densities were still found at comparable or even larger numbers
than those in suction samples of the natural nursery habitat at the
same location, which has been the case in previous studies (Wahle
et al., 2009b, 2012).
Water quality
Temperatures at all sites rose well above the thermal threshold of
208C for lobsters during the 2-year study (Figure 5). Coincidentally,
summer of 2012 proved to be the warmest on record since the initiation in 1990 of RI DFW fixed site temperature monitoring at 13
sites in the bay (Supplementary Figure S2; Olszewski, 2013).
Loggers were lost from our outer coastal stations at Sachuest Point
and Southwest Point during the hurricane in 2011, so we used data
from a more protected site nearby, House-on-Rocks. Temperatures
at upper bay sites exceeded 208C 2 weeks ahead of the lower bay
sites, and reached maximum temperatures of 258C (Figure 5).
Representative upper and lower bay profiles for 2011 and 2012
revealed considerable thermal stratification (Figure 6). The entire
water column exceeded 208C down to a depth of 10 m in the
upper bay. At the mouth of the bay, only the first few meters of
depth were .208C, with temperatures at 15 –20 m ranging from
17 to 188C, which approximates the thermal optimum for lobsters.
The hydrographic survey also revealed considerable differences
between surface and bottom conditions and bay-wide gradients in
water quality (Supplementary Figure S3). Salinity ranged from 29
to 31 psu in a down-bay direction, and was slightly higher near
bottom than at the surface, as is typical in estuaries. Temperature
ranged from a high of 258C in the upper bay to a low of 198C on
the outer coast, and was consistently a degree or two warmer at
the surface than at the bottom. Dissolved oxygen levels were generally higher at the surface than at the bottom, ranging from about 6.8
to 8.1 mg l21. There were two sites with exceptionally low DO concentrations: Fogland and Easton Point, both in the Sakonnet River,
with bottom readings below 6 mg l21. Levels of pH varied from
typical ocean levels of 8.1 to ,7.0, and were also generally lower
near the seabed than at the surface. Levels of pH also correlated significantly with DO concentrations, especially near the seabed
(surface: r 2 ¼ 0.39, P ¼ 0.004; bottom: r 2 ¼ 0.83, P , 0.001),
which would be consistent with near-bottom O2 depletion and
CO2 generation and acidification resulting from benthic respiration
(Supplementary Figure S4).
R. A. Wahle et al.
Discussion
These results indicate significant declines in abundance and a receding distribution of juvenile lobsters in Narragansett Bay since the
early 1990s. The three independent and parallel sampling protocols
revealed consistent patterns of declining lobster numbers in shallow
nurseries. The change in lobster abundance we detected within
Narragansett Bay is also consistent with trends observed in juvenile
populations at sites monitored continuously by suction sampling
from 1990 to the present along the Rhode Island’s outer coast
(Wahle et al., 2009a; Pershing et al., 2012; Supplementary Figure
S1). Supplemental sampling of additional industry-selected sites,
with one important exception, agreed with observations at the original sites. Densities of juvenile lobster were as low at these sites as
they were at nearby locations in the bay. The exception was the
deep site (15–17 m) at Fort Wetherill where relatively high densities
of juvenile lobsters were found. Taken together, the three sampling
methods, the added sampling locations, and long-term monitoring
outside the bay, strongly indicate a decline and receding distribution
of juvenile lobster from shallow nurseries. In short, the relative rarity
of YoY lobsters represents a demographic hole in the nursery population, which now comprises only a few remaining juveniles most
likely from previous years of settlement.
Water quality indicators suggest that midsummer conditions
have become increasingly stressful for lobster in the upper and
mid-bay sections of Narragansett Bay. The American lobster
begins to experience physiological stress at temperatures above
208C (McLeese and Wilder 1958; Pearse and Balcom 2005).
Lobsters were absent in the upper bay where water temperatures
ranged above 258C for extended periods during summer and
where salinity was reduced. Independent RI DFW hydrographic
surveys indicate that mean summer surface temperatures in these
same areas of the bay were 2–38C warmer in 2011 and 2012 than
they were in 1990 (Supplementary Figure S2). Coincidently, summer
temperatures in 2012 happened to be the warmest on record, not
only in Narragansett Bay (Olszewski, 2013; Supplementary Figure
S2), but over a large portion of the Northwest Atlantic (Mills et al.,
2013). Despite this warming trend, much of the bay and outer coast
thermal stratification during summer can create a thermal refuge for
lobsters in deeper water that may be within reach of settling post-larvae
or larger juvenile and adult lobsters. These changes in the distribution
of lobster nurseries suggest that favourable conditions have shifted
considerably southward towards the mouth of the bay. Furthermore,
Glenn et al. (2011) recently observed that with increasingly
warm summer temperatures in nearshore shallows, egg bearing
lobsters were being caught at greater depths, and therefore greater
distances, from shore. This suggests that larvae are hatching further
from shore and are more likely to be exported from coastal nurseries,
further exacerbating the effects of a declining broodstock in nearshore
waters.
Since the 1970s, temperatures in Narragansett Bay have been
rising at a rate of 0.38C per decade (Nixon et al., 2004). This is consistent with the long-term trend reported by Mills et al. (2013) for
the Gulf of Maine, although their more recent study also suggests
that the pace of warming has accelerated over the past decade.
Even at the more conservative rate, future warming projected by
the International Panel on Climate Change models is likely to
make southern New England shallow coastal areas increasingly inhospitable to lobsters in the coming decades (Frumhoff et al., 2007).
Abrupt warming events can play an especially important role in
altering species distribution and abundance. The lobster die-off of
American lobster nurseries of southern New England
i75
Figure 4. Passive collectors deployed in H. americanus nurseries of Narragansett Bay, 2011, and 2012. (a) Lobster density (n m22 + 1 SE) listed by
the site north-to-south (n ¼ 20 collectors sampled per site), and (b) lobster size composition. Collectors were not used in 1990. Industry selected
denoted with asterisks.
1999 in Long Island Sound, next to Rhode Island waters, may be a
prime example (Pearse and Balcom, 2005). Mass mortality of
lobster and associated fauna in the Sound resulted from a combination of warmer than average temperatures and wind-induced
turnover of the water column that mixed warm, oxygen-poor
water down to deeper channels that are normally a cool refuge
from the midsummer heat in the shallows. Although the 2012
ocean heat wave was a record-breaker, it was not accompanied
by mass mortality event in southern New England, even though
it dramatically impacted lobster molting phenology further
north (Mills et al., 2013). We speculate that the absence of a late
summer storm-related mixing event, such as occurred in 1999,
may have allowed the southern New England to avert another
catastrophic die-off.
i76
R. A. Wahle et al.
Figure 5. Temperature logger data at upper (Popasquash 41.67438N, 271.30838W) and lower (House on Rocks 41.48158N, 271.35618W and
Sachuest Point 41.47338N, 271.24838W) Narragansett Bay locations during collector deployments in 2011 and 2012. Grey, hourly record; black,
24-h running mean.
Adverse secondary effects of warming, such as hypoxia and
possibly shell disease, may only exacerbate conditions for lobster,
and may also contribute to the receding distribution of lobster nurseries in Narragansett Bay. The onset of the shell disease epizootic in
southern New England, for example, followed almost a decade of
above normal summer temperatures (Glenn and Pugh, 2006),
although the role of toxic pollutants in the epizootic remains in
question (Laufer et al., 2013). At some of our study sites, high temperatures were accompanied by hypoxic and acidified conditions.
The extent to which these conditions also contributed to the
absence of lobsters is unclear. Narragansett Bay is a particularly
urbanized estuary prone to summer hypoxia events related to
local anthropogenic eutrophication (Altieri and Witman, 2006;
Nixon et al., 2009). Reducing nutrient supplies into the bay, therefore, could mitigate adverse effects of hypoxia and low pH.
We also cannot rule out the possibility that the invasion of the
Asian shore crab, H. sanguineus, may have contributed to declining
numbers of juvenile lobsters in Rhode Island’s shallow lobster nurseries. Asian shore crabs were not present in our 1990 survey, but
were abundant in the recent sampling at all sites except the
deepest one at Fort Wetherill in Narragansett Bay. This introduced
species is an aggressive predator, and has excluded or limited distributions of the green crab, C. maenas, from some shores in southern
New England (Lohrer and Whitlatch, 2002). It is not unreasonable
to suspect this crab could have a similar impact on juvenile lobsters
to the extent the two species overlap. On the other hand, the depth
American lobster nurseries of southern New England
i77
Figure 6. Midsummer temperature profiles at upper (41.66478N, 271.39048W) and lower (41.46248N, 271.21688W) Narragansett Bay locations
taken in 2011 and 2012.
range of the Asian shore crab does not extend greatly into the lobster
habitat, so its impact, if any, would be restricted to the shallowest
subtidal locations. Therefore, further monitoring of lobster settlement and nursery habitat in deep water is warranted.
It is likely that with increasingly inhospitable summer temperatures in shallow coastal areas, lobster settlement in southern New
England will be further restricted to deeper zones, provided suitable
shelter-providing habitat is available. To date, there has been only
limited sampling, mostly by passive collectors, to evaluate lobster
settlement in .10 m of water (Glenn et al., 2011; Wahle et al.,
2012). These studies were done during the recent dearth of larval
settlement in southern New England, and not surprisingly have
produced only a few records of YoY lobsters. Collector deployments
in other parts of New England and Atlantic Canada, where settlement has remained strong, have reported settlement at depths as
great as 80 m. Collectors have not been deployed to greater
depths to date. This raises the possibility that settlement may be
occurring at depths greater than previously thought, and that
there is a need for further deep water monitoring. To monitor
lobster settlement in deeper water will likely require the expanded
use of passive collectors. Furthermore, high-resolution mapping
of available nursery habitat will enable extrapolation of estimated
settlement densities to the area of suitable habitat.
In conclusion, this study reports the receding distribution of
American lobster early benthic life stages in previously wellpopulated nurseries of Narragansett Bay. Since 1990, when the
first survey was conducted, lobsters have suffered mass mortalities,
shell disease, and diminishing fishery recruitment on a larger scale
in southern New England. Climate warming has been implicated
in these events. But, coastal ocean warming is not the only stressor
associated with climate change. It will be important to gain a
greater understanding of how hypoxia, ocean acidification, and
the shifting ranges of predators and competitors impact the changing geography of lobster distributions. Monitoring the most vulnerable life stages of commercially important species and their
environment can provide the early warning that coastal and
i78
fishery managers need to anticipate and adapt to the impacts of a
warming climate.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
Special thanks to C. Bergeron, M. Jaini, N. Oppenheim, T. Wahle,
as well as the RI DFW staff, and Rhode Island fishing industry
members for their assistance in collector deployments, data collection, and processing. The project was supported by a grant from the
Commercial Fisheries Research Foundation. Rhode Island Division
of Fish and Wildlife and the University of Maine’s Darling Marine
Center provided facilities and logistic support.
References
Altieri, A., and Witman, J. 2006. Local extinction of a foundation species
in a hypoxic estuary: Integrating individuals to ecosystem. Ecology,
87: 717 –730.
Atlantic State Marine Fisheries Commission (ASMFC) 2009. American
lobster stock assessment report for peer review. Stock Assessment
Report No. 09-01 (Supplement). Atlantic States Marine Fisheries
Commission, Boston, Massachusetts.
Bintz, J., Buckley, B., Granger, S., and Nixon, S. 2003. Impacts of temperature and nutrients on coastal lagoon plant communities.
Estuaries, 23: 765 –776.
Frumhoff, P. C., McCarthy, J. J., Melillo, J. M., Moser, S. C., and
Wuebbles, D. J. 2007. Confronting Climate Change in the U.S.
Northeast: Science, Impacts, and Solutions. Synthesis Report of
the Northeast Climate Impacts Assessment (NECIA). Union of
Concerned Scientists (UCS), Cambridge, MA.
Glenn, R. P., Angell, T., DeCosta, A., Dellinger, L., Gibson, M., Manning,
J., Pugh, T. L., et al. 2011. Examining settlement dynamics of postlarval American lobster (Homarus americanus) in Lobster Management
Area 2. Commercial Fisheries Research Foundation Final Project
Report. 76 pp.
Glenn, R. P., and Pugh, T. L. 2006. Epizootic shell disease in American
lobster (Homarus americanus) in Massachusetts coastal waters:
interactions of temperature, maturity, and intermolt duration.
Journal of Crustacean Biology, 26: 639– 645.
Hare, J.A., Manderson, J. P., Nye, J. A., Alexander, M. A., Auster, P. J.,
Borggaard, D. L., Capotondi, A. M., et al. 2012. Cusk (Brosme
brosme) and climate change: assessing the threat to a candidate
marine fish species under the US Endangered Species Act. ICES
Journal of Marine Science, 69: 1753– 1768.
Harley, C. D. G., Hughes, A. R., Hultgren, K. M., Miner, B. G., Sorte,
C. J. B., Thornber, C. S., Rodriguez, L. F., et al. 2006. The impacts
of climate change in coastal marine systems. Ecology Letters, 9:
228– 241.
Jeffries, P. 2002. Rhode Island’s ever-changing Narragansett Bay.
Maritimes, 41: 1 – 5.
Jeffries, P., and Terceiro, M. 1985. Cycle of changing abundances in the
fishes of the Narragansett Bay area. Marine Ecology Progress Series,
25: 239 –244.
Keller, A., Oviatt, C., Walker, H., and Hawk, J. 1999. Predicted impacts
of elevated temperature on the magnitude of the winter-spring
phytoplankton bloom in temperate coastal waters: A mesocosm
study. Limnology and Oceanography, 44: 344 –356.
Laufer, H., Chen, M., Baclaski, B., Bobbitt, J. M., Stuart, J. D., Zuo, Y.,
and Jacobs, M. W. 2013. Multiple factors in marine environments
R. A. Wahle et al.
affecting lobster survival, development, and growth, with emphasis
on alkylphenols: A perspective. Canadian Journal of Fisheries and
Aquatic Sciences, 70: 1588– 1600.
Lohrer, A. M., and Whitlatch, R. B. 2002. Interactions among aliens:
apparent replacement of one exotic species by another. Ecology,
83: 719– 733.
McLeese, D. W., and Wilder, D. G. 1958. The activity and catchability of
the lobster (H. americanus) in relation to temperature. Journal of the
Fisheries Research Board, Canada, 15: 1345 –1354.
Mills, K. E., Pershing, A. J., Brown, C. J., Chen, Y., Chiang, F. -S.,
Holland, D. S., Lehuta, S., et al. 2013. Fisheries management in a
changing climate: Lessons from the 2012 ocean heat wave in the
Northwest Atlantic. Oceanography, 26: 191– 195.
Nixon, S., Fulweiler, R. W., Buckley, B. A., Granger, S. L., Nowicki, B. L.,
and Henry, K. M. 2009. The impact of changing climate on phenology, productivity, and benthic– pelagic coupling in Narragansett
Bay. Estuarine, Coastal and Shelf Science, 82: 1 – 18.
Nixon, S., Granger, S., Buckley, B., Lamont, M., and Rowell, B. 2004. A
one hundred and sixteen year coastal water temperature record from
Woods Hole, Massachusetts. Estuaries, 27: 397– 404.
Nye, J. A., Link, J. S., Hare, J. A., and Overholtz, W. J. 2009. Changing
spatial distribution of fish stocks in relation to climate and population size on the Northeast United States continental shelf. Marine
Ecology Progress Series, 393: 111– 129.
Olszewski, S. 2013. F-61-R Annual Performance Report, Assessment of
recreationally important finfish stocks in Rhode Island waters.
Rhode Island Division of Fish and Wildlife, Coastal Fishery
Resource Assessment Trawl Survey 2013.
Oviatt, C., Keller, A., and Reed, L. 2002. Annual primary production in
Narragansett Bay with no bay-wide winter-spring phytoplankton
bloom. Estuarine, Coastal and Shelf Science, 56: 1013– 1026.
Pearse, J., and Balcom, N. 2005. The 1999 Long Island Sound lobster
mortality event: Findings of the Comprehensive Research
Initiative. Journal of Shellfish Research, 24: 691– 697.
Pershing, A. J., Wahle, R. A., Meyers, P. C., and Lawton, P. 2012.
Large-scale coherence in New England lobster settlement associated
with regional weather. Fisheries Oceanography, 21: 348 – 362.
Pinsky, M. L., Worm, B., Fogarty, M. J., Sarmiento, J. L., and Levin, S. A.
2013. Marine taxa track local climate velocities. Science, 341:
1239– 1242.
Sullivan, B., van Keuren, D., and Clancy, M. 2001. Timing and size of
blooms of the ctenophore Mnemiopsis ledyi in relation to temperature in Narragansett Bay, R.I. Hydrobiologia, 451: 113– 120.
Wahle, R. A. 1993. Recruitment to American lobster populations along
an estuarine gradient. Estuaries, 16: 731– 738.
Wahle, R. A., Bergeron, C., Tremblay, J., Wilson, C., Burdett-Coutts, V.,
Comeau, M., Rochette, R., et al., 2012. The geography and bathymetry of American lobster benthic recruitment as measured by diverbased suction sampling and passive collectors. Marine Biology
Research, 9: 42– 58.
Wahle, R. A., Bergeron, C., Wilson, C., and Parkhurst, M. 2009b. A
vessel-deployed passive post-larval collector for the American
lobster. New Zealand Journal of Marine and Freshwater Research,
43: 465– 474.
Wahle, R. A., Gibson, M., and Fogarty, M. J. 2009a. Distinguishing
disease impacts from larval supply effects in a lobster fishery collapse.
Marine Ecological Progress Series, 376: 185– 192.
Wahle, R. A., Brown, C., and Hovel, K. 2013. The geography and bodysize dependence of top-down forcing in New England’s lobstergroundfish interaction. Bulletin of Marine Science, 89: 189– 212.
Handling editor: Jonathan Grabowski
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 10(Supplement 1), i79 –i90. doi:10.1093/icesjms/fsv010
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Potential effect of variation in water temperature
on development time of American lobster larvae
Brady K. Quinn* and Rémy Rochette
Department of Biology, University of New Brunswick, 100 Tucker Park Road, Saint John, NB, Canada E2L 4L5
*Corresponding author: tel: +1 506 343 7676; e-mail: [email protected]
Quinn, B. K., and Rochette, R. Potential effect of variation in water temperature on development time of American lobster larvae.
– ICES Journal of Marine Science, 10: i79 –i90.
Received 1 August 2014; revised 19 December 2014; accepted 8 January 2015; advance access publication 7 February 2015.
Studies typically assess the effects of temperature on development time, larval drift, and fisheries recruitment in American lobster at a range of
constant temperatures. However, in nature, lobster larvae are exposed to varying temperatures, which might result in different development
times than would be predicted from mean temperatures alone. To investigate this hypothesis, we conducted a modelling exercise in which we
simulated larval development from hatch through stages I– IV under different combinations of mean and variance in temperature. Two
thermal scenarios were modelled, the first based on estimated (i.e. interpolated by a model from empirical data) recent historical mean and variability of sea surface temperatures (SSTs) experienced by developing larvae in specific parts of the species’ range, and the second based on a broad
range of simulated combinations of mean and variability in temperature, including conditions that may be experienced by larvae in the future. The
model calculated development times using daily SSTs and temperature-dependent development equations from previous studies of warm- and
cold-water origin larvae. For warm-origin larvae, higher variability in temperature resulted in shorter development times at very cold and very warm
mean temperatures, and longer development at intermediate mean temperatures, than lower (or no) variability. For cold-origin larvae, the effect of
variable temperature was overall much smaller, and opposite to that for warm-origin larvae at very cold and very warm mean temperatures. These
results show that lobster larvae experience meaningful variability of water temperature in nature, and that this variability can markedly impact larval
development. Thermal variability therefore should be considered when estimating development and drift of lobster larvae, including under
scenarios of climate change.
Keywords: American lobster, climate change, development, Homarus americanus, larvae, mean, temperature, variability.
Introduction
The American lobster, Homarus americanus (Milne Edwards, 1837),
inhabits coastal marine waters along eastern North America, from
Cape Hatteras, NC (35.258N), to the Strait of Belle Isle, Labrador
(51.738N; Lawton and Lavalli, 1995), and supports major fisheries
(Wahle et al., 2004) between Rhode Island and Newfoundland
(Pezzack, 1992). Lobster larvae are released into the water column
and complete development during summer, typically sometime
between May and October (Aiken and Waddy, 1986). After hatching, lobsters pass through three planktonic larval stages (I –III)
and a swimming post-larval stage (IV) that generally inhabit the
water column above local thermoclines (Factor, 1995; Harding
et al., 2005) and drift with ocean currents (Chassé and Miller,
2010; Incze et al., 2010). Post-larvae sample the seabed and settle
# International
to the benthos once suitable substrate is found (Botero and
Atema, 1982; Factor, 1995). The duration of each larval stage
depends on water temperature, with higher temperature generally
resulting in faster development and shorter stage duration
(Hadley, 1906; Templeman, 1936; MacKenzie, 1988; Quinn et al.,
2013). Drift of larvae can connect different lobster fisheries and
populations, resulting in their interdependence for larval supply
and new recruits (Ennis, 1995; Chassé and Miller, 2010; Incze
et al., 2010; Quinn, 2014), but time and distance drifted depend
heavily on the speed at which larvae complete development
(Quinn, 2014). Specifically, slower-developing larvae spend more
time drifting than faster-developing ones, and are thus less likely
to contribute to benthic recruitment in their place of origin
(e.g. Miller, 1997; Chassé and Miller, 2010; Incze et al., 2010;
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i80
Quinn, 2014) and more likely to experience mortality (Incze and
Naimie, 2000; Kough et al., 2013), for example, by predation
(Robert et al., 2007). Therefore, the relationship between temperature and development time has important implications for the
supply of larvae and potential recruits to lobster populations and
fisheries (Tlusty et al., 2008; Schmalenbach and Franke, 2010;
Boudreau, 2012; Green et al., 2014).
Several studies have examined the relationship between water
temperature and development time of American lobster larvae.
Most of these studies examined larvae originating from relatively
warm-water regions within the lobster’s range, including parts of
the Gulf of Maine and the southern Gulf of St Lawrence (Hadley,
1906; Templeman, 1936; Hughes and Mathiessen, 1962; Hudon
and Fradette, 1988; MacKenzie, 1988). However, a recent study
showed that development time of larvae originating from a relatively
cold-water region in the northern Gulf of St Lawrence (north of
Gaspé, QC) have a different functional relationship with temperature; temperature had a smaller impact on development of these
“cold-origin” larvae than had been observed for “warm-origin”
larvae (Quinn et al., 2013). Cold-origin larvae still developed
faster at warmer than at colder temperatures, but they developed
faster at temperatures ≤108C (and slower at temperatures
≥148C) than warm-origin larvae did in previous studies, suggesting
that these larvae may have undergone counter-gradient adaptation
(e.g. Botta et al., 2014 and references therein) for development in
colder waters (Quinn et al., 2013).
The effect of temperature on biota is generally studied by measuring biological response variables, such as development time or moulting rate (McLaren et al., 1969; Heip, 1974; MacKenzie, 1988), in
different constant-temperature regimes, or by relating observed
responses to mean temperatures experienced in variable-temperature
regimes (e.g. Hughes and Mathiessen, 1962). However, natural variability in temperature results in effects on biota distinct from what can
be inferred from the mean values alone (Anger, 1983; Niehaus et al.,
2012; Studer and Poulin, 2013). For example, variable temperatures
result in more rapid growth and development in larval frogs
(Niehaus et al., 2012) and juvenile crayfish (Thorp and Wineriter,
1981), as well as higher survival of rock crab (Cancer irroratus)
larvae (Sastry, 1979), than constant temperatures, due perhaps to occasional exposure to highly beneficial or “optimal” temperatures
(Thorp and Wineriter, 1981). However, variability in temperature
can also have harmful effects, including slower overall development
of spider crab (Hyas araneus) larvae (Anger, 1983) and relatively
lower feeding and reproductive performance of zebrafish (Condon
et al., 2010). Thermal variability can therefore have significant, but
species- and context-specific impacts on biological processes (e.g.
larval development rates), and its effects should therefore be considered in addition to the effects of average temperatures. Whether
thermal variability impacts development time of lobster larvae has
not yet been examined.
Studies by Templeman (1936), MacKenzie (1988), and Quinn
et al. (2013) observed development times of lobster larvae at constant, controlled water temperatures in the laboratory. Other
studies carried out in hatchery tanks (Hughes and Mathiessen,
1962) or the field (Hadley, 1906; Hudon and Fradette, 1988;
Annis et al., 2007) did expose larvae to thermally variable conditions, but they made inferences based on mean temperatures experienced by the larvae and did not quantify whether and how thermal
variability impacted larval development. Results of these studies are
used to make inferences about larval development times in nature;
for example, to estimate larval supply patterns to local fisheries
B. K. Quinn and R. Rochette
around Prince Edward Island, Miller et al. (2006) estimated larval
development times in summer using the equations of MacKenzie
(1988) and estimates of the mean water temperature. Similar
approaches may also be used to estimate how larval development
patterns could change as a result of future climate change
(Galbraith et al., 2012; Khan et al., 2013; Loder et al., 2013).
However, these estimates are also likely to be affected by variability
in temperature (Easterling et al., 2000; Hordoir and Meier, 2011;
Loder et al., 2013; Ma and Xie, 2013).
The objectives of the present study were to (i) estimate the
thermal variability typically experienced by lobster larvae throughout the species’ range, (ii) contrast larval development time predicted under variable thermal regimes to those predicted at the
same mean temperatures but with no variability, and (iii) develop
a more generalized model of the predicted effect of thermal variability on larval development time, which among other things may
help forecast effects of climate change. We address these objectives
using development equations derived in previous studies, where
American lobster larvae were raised at constant temperature, and
hence do not address the possibility that development functions
themselves may be modulated by thermal variability (i.e. phenotypically plastic). Achieving these objectives will help determine the
extent to which water temperature variability impacts the development time of lobster larvae in nature, and whether this variability
needs to be considered in connectivity and climate change studies
of lobster.
Material and methods
A computer model was created to calculate development times of
American lobster larvae under different thermal regimes. The
model was written in FORTRAN 90/95, compiled using gFortran
4.8 (Free Software Foundation, Inc., 2012), and run in a virtual
ubuntu 10.10 Linux instance (Canonical Ltd, 2012) accessed
through VMware Player 4 for Windows (VMware, Inc., 2012).
Water temperatures experienced by larvae
in model simulations
Sea surface temperature (SST) is generally a good indicator of water
temperature experienced by lobster larvae because these tend to
remain above local thermoclines, in the surface mixed layer
(Factor, 1995; Harding et al., 2005; Quinn et al., 2013). Two different
datasets were used to provide daily SSTs to larval development
models, which were used to calculate total larval development
times: (i) daily SST data derived from a historical dataset (Ouellet
et al., 2003) for specific regions of the lobster’s geographic range,
and (ii) daily SSTs simulated using a randomization algorithm to
create different “thermal regimes” (mean and variability). The
first dataset was used to estimate how the development time of
lobster larvae is affected by values of temperature (mean and variability) these larvae are known to experience in nature, whereas
the second was used to assess development over a wider range of
temperature regimes (means and variances), and then derive a
generalized model that can be used to predict development under
more varied thermal regimes, including different climate change
scenarios.
Historical SSTs
Estimates of SSTs between 1981 and 2000 were extracted from
outputs of the oceanographic model of Ouellet et al. (2003; available
online at http://www.meds-sdmm.dfo-mpo.gc.ca/isdm-gdsi/
azmp-pmza/climat/sst-tsm/sst-tsm-eng.asp). Ouellet et al. (2003)
Potential effect of variation in water temperature on development time
i81
Figure 1. Map of the Atlantic Shelf of eastern North America, including most of the American lobster’s geographic range and the domain of the
oceanographic model of Ouellet et al. (2003) from which daily historical SST data were obtained for this study. Solid black lines encompass different
regions (SGM, southern Gulf of Maine; NGM, northern Gulf of Maine; SGSL, southern Gulf of St Lawrence; NGSL, northern Gulf of St Lawrence)
considered in this study. Numbers in black and white circles indicate the locations of previous studies of the effect of water temperature on
development time of warm- and cold-origin lobster larvae, respectively.
generated daily SST per 0.1758W × 0.1758N (or 13.6 × 19.5 km)
“site” by linear interpolation in time between weekly satellite observations by the Physical Oceanography Archive Centre of the Jet
Propulsion Laboratory. We extracted daily SST data from this
model in the form of harmonic daily means, which Ouellet et al.
(2003) calculated by averaging the average SSTon the same calendar
day across a 20-year period (1981 –2000). Data fell within a physical
domain between 40 –758W and 40 –528N (Ouellet et al., 2003).
Daily summer SSTs (from 1 May to 30 September) were extracted
from the database for all sites within each of four selected geographic
regions in which previous studies have assessed thermal effects on
American lobster larval development time: (i) the southern Gulf
of Maine (SGM, number of 0.1758W × 0.1758N sites ¼ 637),
(ii) northern Gulf of Maine (NGM, n sites ¼ 155), (iii) southern
Gulf of St Lawrence (SGSL, n sites ¼ 294), and (iv) the northern
Gulf of St Lawrence (NGSL, n sites ¼ 270; Figure 1). Daily SSTs
from 1 May to 30 September (153 d) for each site in each region
were passed to a larval development model (described below) and
then used to calculate development times of larvae released at
each site.
When the model was run using 1981–2000 SSTs, it input one
warm-origin and one cold-origin larva (sensu Quinn et al., 2013)
per day at each site from 1 May to 1 September, corresponding to
the period when hatching typically occurs in nature (Aiken and
Waddy, 1986); this resulted in 246 larvae being input at each site.
Once a larva had completed development (moult to stage V
i82
Figure 2. Mean (+95% confidence intervals) SST over the summer
(day 121 ¼ 1 May, day 273 ¼ 30 September) in four geographic regions
(SGM, southern Gulf of Maine; NGM, northern Gulf of Maine; SGSL,
southern Gulf of St Lawrence; NGSL, northern Gulf of St Lawrence)
within the American lobster’s range. The mean values were calculated
from daily SSTs of all sites (n ¼ 155 2 637) within each region,
obtained from the model dataset of Ouellet et al. (2003) (see details in
the Material and methods section). Different colours represent
different geographic regions. Horizontal lines on the plot indicate the
typical period during which larval release (hatch), development, and
drift occur, as estimated based on at-sea sampling of ovigerous female
lobsters in the Gulf of Maine (GM; Incze et al., 2010) and Gulf of St
Lawrence (GSL; Chassé and Miller, 2010 and references therein).
completed, see below), the mean and s.d. of daily SSTs it had experienced were calculated. To simplify data analysis, the mean and s.d. of
SST values were rounded to the nearest 18C; an s.d. value of ,0.58C
was therefore rounded to 08C. To quantify the thermal variability
experienced by larvae within each region, the percentage (%) of
larvae (across all sites and release dates) within each of the four
regions experiencing different combinations of the mean and s.d.
of SSTs was calculated. As these particular results were extremely
similar for warm- and cold-origin larvae, thermal variability experienced by larvae was quantified and illustrated using combined data
for all larvae.
Simulated SSTs
Historical SSTs were further assessed to determine the range of temperatures that could be experienced by lobster larvae in nature. First,
we calculated the mean daily SSTs from historical SSTs (Ouellet
et al., 2003) for all sites within each of the four regions considered
in this study (Figure 1). Second, we determined the range of historical daily SSTs occurring during the period in which larvae are potentially released and develop in these regions (1 May to 30
September), as estimated based on at-sea sampling of ovigerous
females (Chassé and Miller, 2010 and references therein; Incze
et al., 2010; Figure 2). Finally, we defined simulated SST “thermal
regimes” that included all combinations of the mean and s.d. of
these daily temperatures (during the period of larval hatch and development) observed yearly between 1981 and 2000. Simulated SST
data also included combinations of the mean and s.d. of SST not
observed in the historical SST datasets, to explore potential effects
of future climate change. Daily SSTs during larval development
B. K. Quinn and R. Rochette
ranged from 5 to 268C (Ouellet et al., 2003; see also the mean
values in Figure 2); therefore, a total of 154 simulated SST thermal
regimes were defined, each with a specific combination of the
mean SST (5–268C) and s.d. (0 –68C) of SST (Ouellet et al.,
2003), at 18C increments. The range of mean temperatures used
was selected to comfortably encompass temperatures at which
larvae are likely to be hatched and develop in the four regions
under study (Figures 1 and 2). Although lobster spawning is typically thought to occur only when temperature exceeds 10– 128C (e.g.
Harding et al., 1983; MacKenzie, 1988; Ennis, 1995), in some parts of
the NGM and NGSL hatch can occur when SSTs are as low as 58C
(Figure 2), and egg-bearing females are found in some parts of
Newfoundland where SSTs never exceed 5 –108C (Ma et al., 2012;
R. Stanley, Memorial University of Newfoundland, pers. comm.).
When run using simulated SSTs, the larval development models
(described below) generated 100 individual warm-origin larvae and
100 cold-origin larvae (sensu Quinn et al., 2013) to be subjected to
each thermal regime during development, for a total of 30 800
larvae. The model then randomly generated daily SSTs experienced
by each larva based on the thermal regime to which it was assigned.
Daily temperature was always set to the mean temperature if the s.d.
of temperature for a given thermal regime was zero (no variability).
Otherwise (s.d. . 08C), to define the probability of any temperature
occurring on a given day a Gaussian probability curve (for 10 000
temperature values, at 0.018C increments) was generated for each
thermal regime based on its mean temperature and s.d. Then,
each day (i.e. at every model time-step), a random number generator that used RANLUX (James, 1994) selected the temperature to
be experienced by that larva on that day.
Larval development model
Daily SST (historical or simulated) experienced by each larva was
passed on to one of two different larval development models,
which were based on experiments involving warm-origin
(MacKenzie, 1988) and cold-origin (Quinn et al., 2013) lobster
larvae, respectively (Table 1, Figure 3). Each model contained four
equations relating development time through a particular larval
stage (stages I –IV) to water temperature (MacKenzie, 1988; Incze
et al., 2010; Quinn et al., 2013; Quinn, 2014; Table 1). Quadratic
equations were used (Table 1, Figure 3) to account for the fact
that development time of arthropod larvae does not indefinitely decrease with temperature (e.g. Sastry and Vargo, 1977; Brière et al.,
1999; Shi and Ge, 2010; Pakyari et al., 2011). Importantly, equations
for both warm- and cold-origin larvae were derived in their original
studies (MacKenzie, 1988; Quinn et al., 2013) within a more limited
thermal range (10–228C) than that used in this study (5 –268C).
Therefore, development times predicted in this study were extrapolated beyond current knowledge, and thus interpreted with caution.
It is worth noting, however, that an earlier study of warm-origin
larvae by Templeman (1936) used temperatures as low as 68C and
as high as 248C, and observed nearly identical development times
and functions as MacKenzie (1988), providing some confidence in
these extrapolations.
The development models had a time-step of 1 d, during which
temperatures were generated and/or read, and developmental progress of each larva was calculated based on SST. The model initialized larvae at stage 1.0, or the beginning of the first larval instar
(stage I) immediately after hatching and moult from the prezoea
stage (Charmantier et al., 1991). Then, daily SSTs were used in the
appropriate equation for each larval stage to calculate the proportion of the stage through which a larva would advance on that
i83
Potential effect of variation in water temperature on development time
Table 1. Temperature-dependent development equations used to increment development through larval stages I– IV each day, based on
water temperatures.
Stage
I
II
III
IV
Warm-origin larvae
D ¼ 39.558– 3.659T + 0.092T 2
D ¼ 55.670– 5.227T + 0.130T 2
D ¼ 81.298–7.631T + 0.189T 2
D ¼ 156.895 –14.316T + 0.359T 2
Cold-origin larvae
D ¼ 22.704 –1.525T + 0.031T 2
D ¼ 16.469 –0.425T + 0.001T 2
D ¼ 30.219 –1.674T + 0.033T 2
D ¼ 49.368 –2.354T + 0.029T 2
Quadratic equations of the form D ¼ aT 2 + bT + c were used, in which D is the total stage duration (in days) at a certain temperature, T (in 8C), and a, b, and c
are fitted constants. Equations for development through stages I–III for warm-origin larvae were derived in this study from data presented by MacKenzie (1988),
while the equation for stage IV was derived by Incze et al. (2010) based on data from MacKenzie (1988). Equations for cold-origin larvae were obtained from
Quinn et al. (2013) (stages I –III) and Quinn (2014) (stage IV, based on data from Quinn et al., 2013).
Figure 3. Predicted total development time (in days) of warm- (solid
squares and line) and cold-origin (open circles, dashed line) lobster
larvae at different constant temperatures (in 8C), as predicted using
development equations from previous studies (Table 1). These
predicted development times were compared with those predicted by
our model, which included realistic variability in temperature within
each regime (see the Results section, Figures 4 and 5). Note that
predicted development times of warm-origin larvae at temperatures
,6 and .248C, and of cold-origin larvae at temperatures ,10 and
.228C, are extrapolations beyond the range of temperatures used in
the studies upon which the development functions (Table 1) are based.
day. To avoid computational errors in high variability regimes, stage
increment of larvae that experienced daily SST ≤ 08C was set to zero
because lobster larvae are not known (and extremely unlikely) to
continue development at subzero temperatures (Huntsman,
1924). When a larva reached stage V, which is the first benthic juvenile stage and therefore the end of the larval phase (Factor, 1995), the
time elapsed (in days) from the larva’s hatch up to this point was
recorded as the total larval development time.
Analyses
After model simulations were completed, total larval development
time calculated for each individual larva was used for analyses.
Separate analyses were performed on data for cold- and
warm-origin larvae. Total development times calculated by the
model using historical SSTs were analysed on a per-region basis;
in other words, data for larvae generated for all sites and dates
within each of the four regions were pooled [number of warm- or
cold-origin larvae per region ¼ 78 351 (SGM), 19 065 (NGM),
36 162 (SGSL), and 33 210 (NGSL)]. We calculated the difference
between the development time of each larva as predicted under different scenarios of the mean and variance in temperature to that predicted assuming the same mean temperature but no variance
(Figure 3). Differences in development time were then grouped together based on the mean and s.d. of SSTs experienced by each larva
(in 18C steps), and for each combination of the mean and s.d. of SST,
the mean difference (+95% confidence intervals) in development
time was calculated. If the 95% confidence intervals around the
mean difference between development times determined with
and without thermal variability at a particular mean and s.d. of
SST did not overlap with 0, it was concluded to be significant
(p , 0.05; Cummings et al., 2007).
Next, analyses were carried out on total development times calculated by the model using simulated SSTs. For each of the 154 combinations of the mean and s.d. of temperature, means and 95%
confidence intervals of development time were calculated for 100
cold- and 100 warm-origin larvae. Development times calculated
with and without thermal variability were concluded to be significantly different (p , 0.05) if the 95% confidence intervals of a
larval development time in a particular variable thermal regime
(s.d. ¼ 1–68C) did not overlap the mean development time in
the constant (s.d. ¼ 08C) regime at the same mean temperature
(Cummings et al., 2007).
Finally, a nested non-linear regression equation was derived for
both warm- and cold-origin larvae to allow total development
time to be predicted based on the mean and s.d. of temperature,
both within and beyond the ranges of recently recorded mean
and s.d. of SSTs. These equations had a basic quadratic structure
(as did the original development equations in Table 1) relating
total development time (D) to the mean temperature (TM), as
D = A + BTM + CTM2 . However, the magnitude of s.d. affected
this relation. The relationship between s.d. and development
time within each mean temperature appeared to be non-linear
and quadratic, so the equations describing the effects of s.d. of
temperature (s.d.T) on regression coefficients of the mean
temperature-development relationship were structured as follows:
A = a1 + b1 s.d.T + c1 s.d.T2 ; B = a2 + b2 s.d.T + c2 s.d.T2 ; C = a3 +
b3 s.d.T + c3 s.d.T2 .
Results
Thermal regimes experienced by lobster larvae
In general, the lowest mean daily SSTs during larval release and development occurred in the NGM (range of the mean daily SSTs
across all sites in the region ¼ 6.9 –14.68C), and the highest occurred in the SGM (range ¼ 9.5 –19.38C; Figure 2); daily summer
SSTs in the NGSL (range ¼ 9.4 –14.98C) tended to be lower than
those in the SGSL (range ¼ 10.5 –17.58C), though SSTs in both of
these were higher than those in the NGM and lower than those in
i84
B. K. Quinn and R. Rochette
Figure 4. Incidence of different combinations of mean (x-axis) and variability (s.d., indicated by different fill patterns) in historical daily summer SST
(in 8C) experienced by American lobster larvae in four regions within the species’ range. Daily SSTs were harmonic means of historical (1981 – 2000)
SSTs extracted from outputs of the physical oceanographic model of Ouellet et al. (2003) (see details in the Methods section). The mean and s.d. of
SST experienced during development were calculated for individual larvae released at specific sites and on each day over summer, between May and
September. The per cent (%) of all warm- and cold-origin larvae in each region (n ¼ 38 130 –156 702 larvae/region) that experienced each
combination of mean and s.d. is plotted on the y-axis.
the SGM (Figure 2). The mean overall SSTs experienced by lobster
larvae between 1981 and 2000, as well as the variability around
these means, differed among the four regions of the lobster’s
range (Figure 4). The mean SSTs ranged from 7 to 258C, while the
s.d. of SSTs ranged from 0 to 48C (Figure 4). Means as low as 7 –
88C occurred rarely in all four regions, while the highest mean
SST values experienced differed among regions (Figure 4); the
highest mean SST, 258C, was experienced by larvae in SGM, while
the mean SSTs in the NGM, SGSL, and NGSL did not exceed 18,
19, and 158C, respectively (Figure 4). Across all regions, the
highest mean SSTs were accompanied by relatively low variability
(s.d. ¼ 0–18C), with larvae experiencing the mean SST . 168C
never experiencing s.d. . 18C, while conversely lower mean SSTs
tended to coincide spatially with higher levels of variability
(Figure 4). Thermal variability over the larval period was relatively
low in the Gulf of Maine, with only 18.8% of larvae in the SGM
and 22.5% in the NGM experiencing s.d. of SST ≥ 28C (Figure 4),
and was comparatively high in the Gulf of St Lawrence, with
46.5% of larvae in the SGSL and 49.7% in the NGSL experiencing
s.d. of SST ≥ 28C (Figure 4).
Development time of larvae under estimated
natural variation in SST
Warm-origin larvae
In all four study regions, warm-origin larvae that experienced variable SSTs during development had shorter total development times
than predicted based on the mean temperatures ≤14 –168C alone,
and the magnitude of this difference increased with variability in
temperature (Figure 5). In contrast, at higher mean temperatures,
there was little evidence of any effect of variability in temperature
on development time (Figure 5), likely because larvae that experienced higher mean SSTs experienced little to no thermal variability
(s.d. ≤ 18C) during development (Figure 4). On average, larvae at
higher mean SSTs that experienced s.d. ¼ 18C tended to take slightly
longer to develop than those at the same means that experienced constant SST (s.d. ¼ 08C), but these differences were never significant (Figure 5). Conversely, for s.d. ¼ 28C, the difference was always
significantly negative (faster development in variable than in constant temperature) except at mean temperatures .148C, and for
s.d. ¼ 3 and 48C, they were always significantly negative (Figure 5).
The mean of significant deviations from predicted development
i85
Potential effect of variation in water temperature on development time
Figure 5. Effect of variability in 1981 – 2000 SSTs on total development time (from hatch to stage V) of warm-origin lobster larvae. Values are mean
differences (+95% confidence intervals) between predicted total development time (in days) using realistically variable SSTs vs. those predicted
assuming no thermal variability (as per Figure 3), at different mean temperatures (in 8C). Different symbol types and colours represent larvae
experiencing different levels of thermal variability (s.d. of SST, in 8C). Note that development of larvae in the NGSL most likely should be calculated
using cold-origin equations (Quinn et al., 2013), as in Figure 6, and that development at mean SSTs .22 and ,108C is extrapolated beyond the
thermal range thus-far examined for warm-origin larvae (MacKenzie, 1988).
time without thermal variability (for predicted values, see Figure 3)
ranged from 1 to 9 d (2–12%) for s.d. ¼ 28C, 8–23 d (9–20%) for
s.d. ¼ 38C, and 20–31 d (19–24%) for s.d. ¼ 48C (Figure 5).
Cold-origin larvae
In contrast to warm-origin larvae, cold-origin larvae were not strongly affected by thermal variability (Figure 6; note difference in scale of
y-axis relative to Figure 5). Nevertheless, some significant differences
in development times of ,2 d, or ≤4%, from mean-only predictions
were observed (Figure 6). Thermal variability was somewhat associated with the magnitude and direction of this deviation, with development times of larvae experiencing s.d. of SST ≥28C being shorter at
the mean SSTs ≤10–138C, and those of larvae experiencing s.d. of
SST ≤18C being longer in the majority (82%) of cases, than would
be predicted at the same mean temperatures but no variance
(Figure 6); differences were always ≤1 d, however (Figure 6).
variable (s.d. ¼ 1–68C) vs. constant (s.d. ¼ 08C) thermal
regimes, and the magnitude of this difference increased with
thermal variability (Figure 7). Usually, means and 95% confidence
intervals of development times under variable and constant temperatures did not overlap, and thus were significantly different
(p ≤ 0.05, Figure 7). At mean temperatures ≥58C and ≤148C,
and at mean temperatures ≥258C, greater thermal variability was
predicted to result in more rapid development (Figure 7).
Conversely, for mean temperatures ≥158C and ≤248C, greater variability resulted in slower development (Figure 7). The consequence
of these complex interactive effects of the mean and variability in
water temperature is that the parabola relating total development
time of larvae to mean water temperature became less steeply
curved (i.e. development less affected by the mean temperature)
as variability in water temperature increased (Figure 7).
Cold-origin lobster larvae
Development time of larvae under simulated
variation in SST
Warm-origin lobster larvae
Total development time of warm-origin larvae under simulated SST
regimes differed significantly between larvae that experienced
The effect of simulated variable SSTs on development of cold-origin
larvae was markedly smaller and in some regards opposite to that
predicted for warm-origin larvae (Figure 7). In particular, the
curve relating total development time of cold-origin larvae to the
mean SST became somewhat more steeply curved as variability in
i86
B. K. Quinn and R. Rochette
Figure 6. Effect of variability in 1981 – 2000 SSTs on total development time of cold-origin lobster larvae. Values are the same as those plotted in
Figure 5, only for cold-origin larvae. As s.d. values were rounded to the nearest 18C increment, the s.d. ¼ 08C group does not necessarily represent
absolutely zero variability, but rather very low variability; thus, these symbols are not all exactly on the x-axis (difference ¼ 0 d), as would be
expected. Note that development of larvae in all areas but the NGSL most likely should be calculated using warm-origin equations (MacKenzie,
1988), as in Figure 5, and that development at mean SSTs ,10 and .228C in these plots is extrapolated beyond the thermal range thus-far
examined for cold-origin larvae (Quinn et al., 2013).
temperature increased, although this change occurred mainly over
the lower range (5– 88C) of mean temperatures tested (Figure 7).
Nevertheless, in the majority (69%) of cases, development time of
cold-origin larvae differed significantly between larvae that experienced variable (s.d. ¼ 1–68C) vs. constant (s.d. ¼ 08C) SSTs, and
the magnitude of difference increased with greater thermal variability (Figure 7). The magnitude of differences in development time
between different levels of variability was much smaller for
cold-origin larvae than was observed for warm-origin larvae
(Figure 7); for example, the maximum difference in development
time between s.d. ¼ 08C vs. s.d. ¼ 68C was observed at the mean
temperature ¼ 58C for both warm- and cold-origin larvae, but
the difference was 231 d for warm-origin larvae, but only
+15.4 d for cold-origin larvae (Figure 7). In fact, for many mean
temperatures, the maximum difference between constant and variable regimes for cold-origin larvae was extremely small, and often
≤1 d (Figure 7). Significant harmful effects (slower development)
of thermal variability on development of cold-origin larvae were
observed at mean temperatures between 5 –98C and ≥218C,
whereas significant beneficial effects (faster development) were
not observed at any mean temperature (Figure 7).
Discussion
Differential effects of variability depending on larval
origin and mean temperature
In this study, we examined the extent to which larvae of the
American lobster experience variable water temperatures during development in different parts of the species’ range, and we estimated
the effect this variability may have on their development time. In all
four regions that we examined, some larvae were estimated to
experience s.d. of SST as high as 48C during their 21 –200+ d development period (Templeman, 1936; MacKenzie, 1988; Quinn et al.,
2013). Such variability in water temperature greatly affected calculated larval development times, which has important implications
to predictions of larval drift time in nature, including in climate
change scenarios. Interestingly, whether the effect of variability
in water temperature on larval development time was positive or
negative varied in a relatively complex fashion depending on
larval origin and mean water temperature. This was likely because warm- and cold-origin larvae exhibit distinct temperaturedependent development functions, which may be the result of adaptation to different thermal regimes (Quinn et al., 2013), including
i87
Potential effect of variation in water temperature on development time
Figure 7. Mean (+95% confidence intervals) total development time
(in days) from hatch to the moult to stage V of warm- (top) and
cold-origin (bottom) lobster larvae predicted by the model using
simulated SSTs at different combinations of mean (x-axis) and s.d.
(indicated by different symbol types and colours) of temperature
(in 8C). Note that the scales of the y-axes differ between the warm- and
cold-origin plots. Lower and higher mean temperatures in the hatched
regions of both plots represent extrapolations beyond the range of
temperatures used in the original studies (MacKenzie, 1988; Quinn
et al., 2013) from which the development functions were derived. Also
indicated on both graphs are the non-linear regression equations
derived in this study to express how s.d. of temperature (s.d.T) alters the
shape of the relationship between mean temperature (TM) and total
larval development time (D); both regressions were highly significant
(p , 0.001) and had R 2 ≥ 0.99.
of this relation, increased variability in temperature increases development time when temperature is close to the “development
optimum” (208C), but it decreases development time at mean temperatures that are colder (,158C) and warmer (.248C) than this
optimum. The negative effect of variability in temperature near
the optimum is intuitive, whereas the positive effect at less
optimum temperatures arises because greater variance around suboptimum temperature will result in a greater amount of time spent
at more favourable temperatures (e.g. Thorp and Wineriter, 1981;
Niehaus et al., 2012). As warm-origin larvae in the SGM and SGL
mainly experience temperatures between 15 and 248C (near their
thermal optimum; Ouellet et al., 2003; see also Figures 2 and 4), contemporary effects of natural thermal variability on these larvae are
most likely negative.
Variability in water temperature also affected the development
time of cold-origin larvae, but these effects were markedly different
and overall much smaller than those for warm-origin larvae. First,
there was little evidence of positive effects of variability in temperature on development times of cold-origin larvae across the temperature range we examined. Second, significant negative effects were
observed, but usually, these were extremely small, and resulted in
≤1 d longer development in variable vs. constant temperatures.
Somewhat greater lengthening of development times (1–14 d) in
variable vs. constant thermal regimes was observed at lower mean
temperatures, between 5 and 98C, but these effects were still relatively small compared with those observed for warm-origin larvae. The
reason for the relatively limited effect of temperature variability on
development of cold-origin larvae is that the overall shape of their
temperature-development function is much less curvilinear than
that of warm-origin larvae (MacKenzie, 1988; Quinn et al., 2013),
meaning that thermal variability around a particular mean value
results in similar increases and decreases in development time,
and thus variance has less impact on expected (i.e. mean) development times. Cold-origin larvae in the northern Gulf of St Lawrence
mainly experience temperatures ≤158C (Ouellet et al., 2003; see also
Figures 2 and 4), so contemporary effects of natural thermal variability on these larvae are most likely negative, and could be very
small. These considerable differences in how thermal variability
impacted warm- and cold-origin larvae point to the need for
further research into the possibility and scope of geographic variation of temperature-based development functions of American
lobster larvae. In particular, future research should quantify the
effect of water temperature on development of larvae from other
cold-water regions of the lobster’s range, such as the north shore
of Quebec, northern Gulf of St Lawrence (mean summer SSTs
≤148C; Ouellet et al., 2003), or Placentia Bay, Newfoundland
(summer SSTs ≤58C; Ma et al., 2012).
Implications of results to estimates of larval drift
possibly faster development at cold temperatures in regions where
summer/fall temperatures may become too cold for larval development or survival (e.g. northern Gulf of St Lawrence; Figure 2).
For warm-origin larvae, significant beneficial effects of thermal
variability, resulting in much shorter development times, were predicted at very low (5–148C) and very high (≥ 258C) mean SSTs, but
significant negative effects were predicted at intermediate temperatures of 15 –248C. This occurred because the relation between temperature and development time of these larvae resembles a parabola,
with development time increasing as temperature moves away from
an optimum (i.e. where development time is shortest) of 208C
(see the Material and methods section, Figure 3). As a consequence
Results of this study showed that variability in water temperature
could have important effects on the development of lobster larvae
in nature, which is consistent with observations on other species
and systems (e.g. Thorp and Wineriter, 1981; Anger, 1983;
Niehaus et al., 2012; Studer and Poulin, 2013). This may be important to lobster populations and fisheries, as changes to larval development times impact larval drift times and distances (Quinn, 2014),
which in turn could influence connectivity and recruitment patterns
among and within local fisheries (Ennis, 1995; Chassé and Miller,
2010; Incze et al., 2010; Green et al., 2014; Quinn, 2014). For
example, in this study, total development times of larvae in variable
thermal regimes at the mean simulated SST ¼ 208C were as much as
i88
60% longer than those of larvae that experienced a constant temperature of 208C, which could result in up to 500 km longer
larval drift distances (Quinn, 2014). Therefore, our results demonstrate the potential for considerable errors to be made if estimates of
larval development and drift are made without accounting for
effects of variability in temperature. One way to account for
thermal variability while making such estimates would be to use
the complex nested model equations derived in this study
(Figure 7). Assuming one knew the mean and s.d. of water temperature in an area and period of interest, they could apply the appropriate equation for either warm- or cold-origin larvae and estimate
their total development time. These results should be experimentally validated, though, to ensure that different aspects of larval
biology—such as mortality at extreme temperatures or phenotypic
plasticity (discussed below)—do not unduly impact the predictions
of these models.
Future research needed to confirm results
Thermal limits of lobster larval development
As noted in the Material and methods section, projections (of development times) made beyond the thermal ranges (10–228C) used in
the studies on which larval development equations were based
(MacKenzie, 1988; Quinn et al., 2013) must be considered with
caution. Development time of arthropod larvae increases markedly
beyond certain threshold temperatures (Sastry and Vargo, 1977; Shi
and Ge, 2010; Pakyari et al., 2011) until a lethal temperature is
reached beyond which no development occurs (Bĕlehrádek, 1935;
McLaren et al., 1969; Brière et al., 1999). If such thermal limits
occur for lobster larvae within or just outside the range of temperatures (5–268C + 0– 68C s.d.) considered in this study, changes to
the models we used would be required, as our current projections
would underestimate development times at more extreme temperatures. Anecdotal evidence suggests lethal temperatures for lobster
larvae at or below 1 –58C (Huntsman, 1924; MacKenzie, 1988)
and above 28 –358C (Huntsman, 1924; Sastry and Vargo, 1977).
However, there was no clear evidence of thermal threshold(s) (e.g.
Brière et al., 1999) within the range of temperature (10–228C)
examined in the studies from which the development functions
used here were derived (MacKenzie, 1988; Quinn et al., 2013), or
even within the extended range (6– 248C) utilized in an earlier
study (Templeman, 1936). It therefore seems unlikely that development times estimated beyond 10 –228C in the present study (mean
SSTs of 5 –108C and 22 –268C) would have led to markedly erroneous conclusions. In highly variable simulated thermal regimes daily
SSTs further outside this range could be generated, but due to the
way in which our model simulated daily SSTs (see the Material
and methods) these were extremely unlikely to occur, and because
they would have occurred so infrequently their impact on our
results would have been very short term and overall small. But
perhaps more important, if this problem did bias our results, it
would not affect any of our major inferences. Specifically, even
if results based on extrapolated mean SSTs (,108C and .228C)
are excluded, we would still conclude that variable water temperature, compared with constant temperature, results in (i) faster development of warm-origin larvae at means ≤148C, (ii) slower
development of warm-origin larvae at means between 15 and
248C, and (iii) either no effect or slightly (≤1 d) longer development times of cold-origin larvae. We nevertheless recommend
that experiments be undertaken to precisely determine lethal temperatures of lobster larvae and the shape of their development function(s) near these limits.
B. K. Quinn and R. Rochette
Potential metabolic and/or phenotypic plasticity
In addition to the uncertainty of applying development equations at
“extreme temperatures”, other aspects of lobster larval biology may
also need to be considered to fully understand the effect of thermal
variability on their development. In the present study, water temperature was assumed to affect larval development in the same
manner on each day, irrespective of the temperature the larva experienced on previous days. Enzyme kinematics offer a theoretical foundation for this assumption, as enzymes are expected to be influenced
by temperature in the same way over a given period regardless
of temperature in previous periods (so long as damage or denaturation do not occur), and there is evidence this is true for processes
related to development and moulting of arthropod larvae (Thorp
and Wineriter, 1981; Brière et al., 1999). However, if larvae experiencing variable temperatures undergo phenotypic plasticity of
temperature-dependent development rate, for example, through
modification of the enzymes that mediate larval development
(Somero, 2004), then actual development times could differ from
those predicted in the present study. It is unknown whether such
plastic responses occur in American lobster larvae, but in certain
species of arthropods (mosquito, Anopheles quadrimaculatus:
Huffaker, 1944; rock crab, C. irroratus: Sastry, 1979) larvae reared
in fluctuating thermal environments had higher metabolic rates
than conspecifics reared in less variable environments, and these
plastic changes affected future growth and development rates of
larvae at particular temperatures. If lobster larvae display similar
plasticity in development, then our predictions of development
time may be biased, and it is difficult to predict the nature of this
bias. Future research is needed to assess the effect of “thermal
experience” on the development of American lobster larvae at
different temperatures.
Importance of results in relation to climate change
Thermal regimes likely to be experienced by larvae
The range of means and s.d. of SSTs estimated to be experienced by
larvae in this study showed that both warm- and cold-origin larvae
frequently experience variable temperatures during development in
nature. However, thermal regimes including relatively high thermal
variability (s.d. of SST ¼ 2–48C) were confined to relatively low
mean temperatures (≤14 –168C), while larvae experiencing
higher mean temperatures tended to experience much lower or
almost no (s.d. ¼ 0–18C) thermal variability during development.
This pattern appears to have emerged because larvae that experience
higher mean SSTs develop faster, and thus experience relatively little
thermal variability during development, while larvae that experience lower mean SSTs take longer to develop and experience
greater variability. This means that it is unlikely that larvae developing under natural conditions will experience a thermal regime including a high mean SST and high variability. These results also
suggest that the overriding consequence of future climate change
may be that lobster larvae will experience higher mean SSTs and
lower thermal variability, which for the most part may result in
faster development and reduced connectivity via larval dispersal.
That being said, at the mean SSTs ≥258C, there is potential for
slower development due to thermal inhibition, which (as discussed
above) could result in larvae experiencing both high mean and high
variability in temperature. Additionally, larvae developing in simulated SSTs in the second part of our study that were “forced” to experience “unusual” thermal regimes characterized by high mean
and high variability in temperature did develop significantly
i89
Potential effect of variation in water temperature on development time
slower than larvae at the same high means but experiencing little or
no variability. Therefore, if combinations of high means and high
variability of SST do someday occur in some parts of the lobster’s
range, the effects on larval development could be quite large.
Importance of thermal variability to estimating effects
of future climate changes
Empirically derived thermal regimes in this study covered a recent
historical period, from 1981 to 2000, but more contemporary (i.e.
2001–2014) and future patterns of SST are likely to be quite different. In recent years, summer water temperatures in Atlantic Canada
have been warmer on average than in recent history (Galbraith et al.,
2012; Loder et al., 2013) and have potential to become considerably
higher over the next century (Khan et al., 2013). Studies of the effects
of future climate changes often consider how mean annual, seasonal, and/or monthly temperatures are expected to change, and then
use these projections to estimate effects on biota (e.g. Drinkwater,
2005; Khan et al., 2013). However, our results show that the
magnitudes of the effects of future climate change could also
markedly depend on thermal variability, and that making these
inferences based on projected mean temperatures alone may not
be satisfactory.
Projections of specific changes to thermal variability in the ocean
are rare, involve considerable modelling effort, and are limited by
much uncertainty due to interactions between spatial and temporal
variability in physical oceanography, climate, etc. (Ma and Xie,
2013). Thermal variability in certain locations might increase,
for example, due to more frequent and intense storm events
(Easterling et al., 2000), while in other areas, it could decrease due
to increased stratification and reduced mixing of the water
column (Hordoir and Meier, 2011). Preliminary work in eastern
North America suggests that increased stratification (Loder et al.,
2013) and thus less thermal variability (J. Chassé, Fisheries and
Ocean Canada, pers. comm.) may be the dominant change across
much of the lobster’s range in the future. Ongoing modelling
efforts may eventually allow changes to thermal regimes experienced
by lobster larvae to be estimated. Our results show that such efforts
are worthwhile, to accurately estimate future changes to lobster
larval development. For example, if the mean August SST experienced by warm-origin larvae in a given area at present is 158C,
and is expected to increase to 208C by 2100 (Khan et al., 2013),
the actual reduction in larval development times that would result
could differ by a factor of 2 depending on thermal variability; specifically, the change could be as small as 9 d if s.d. of SST ¼ 68C, or as
large as 20 d if s.d. ¼ 08C. Quantifying mean and variability in temperature is therefore crucial to understanding and predicting effects
of future climate changes on the development of lobster larvae,
which in turn will impact their dispersal and contribution to population connectivity.
Acknowledgements
The authors wish to thank Joël Chassé, Steve Gamblin, Raja
Wetuschat, VMware, Inc., and the University of New Brunswick’s
Information Technology Services and Department of Computer
Sciences and Applied Statistics for providing advice, assistance,
and computing resources for this project, as well as the editor of
ICES JMS and two anonymous reviewers for providing helpful comments on the manuscript. Funding was obtained from the NSERC
Canadian Fisheries Research Network and a New Brunswick
Innovation Foundation RIF grant, as well as a NSERC CGS D and
UNB Graduate Research Award held by BKQ.
References
Aiken, D. E., and Waddy, S. L. 1986. Environmental influence on recruitment of the American lobster, Homarus americanus: a perspective.
Canadian Journal of Fisheries and Aquatic Sciences, 43: 2258–2270.
Anger, K. 1983. Temperature and the larval development of Hyas
araneus L. (Decapoda: Majidae); extrapolation of laboratory data
to field conditions. Journal of Experimental Marine Biology and
Ecology, 69: 203 – 215.
Annis, E. R., Incze, L. S., Wolff, N., and Steneck, R. S. 2007. Estimates of
in situ larval development time for lobster, Homarus americanus.
Journal of Crustacean Biology, 27: 454– 462.
Bĕlehrádek, J. 1935. Temperature and living matter. Protoplasma
Monographia, No. 8. Borntraeger, Berlin. 277 pp.
Botero, L., and Atema, J. 1982. Behavior and substrate selection during
larval settling in the lobster Homarus americanus. Journal of
Crustacean Biology, 2: 59 – 69.
Botta, A. M., Rochette, R., Saunders, G. W., Addison, J. A., and Barbeau,
M. A. 2014. Evidence for genotypic differentiation between marine
snails (Littorina sitkana) from the upper- and lower-intertidal
zone in Bamfield Inlet (British Columbia, Canada). Journal of
Experimental Marine Biology and Ecology, 461: 389 – 396.
Boudreau, S. A. 2012. Explaining variation in lobster (Homarus americanus) and snow crab (Chionoecetes opilio) abundance in the northwest
Atlantic Ocean. PhD thesis, Dalhousie University, Halifax, NS, Canada.
Brière, J. F., Pracross, P., Rioux, A. Y., and Pierre, J. S. 1999. A novel rate
model of temperature-dependent development for arthropods.
Environmental Entomology, 28: 22– 29.
Charmantier, G., Charmantier-Daures, M., and Aiken, D. E. 1991.
Metamorphosis in the lobster Homarus (Decapoda): a review.
Journal of Crustacean Biology, 11: 481– 495.
Chassé, J., and Miller, R. J. 2010. Lobster larval transport in the southern
Gulf of St. Lawrence. Fisheries Oceanography, 19: 319– 338.
Condon, C. H., Chenoweth, S. F., and Wilson, R. S. 2010. Zebrafish take
their cue from temperature but not photoperiod for the seasonal
plasticity of thermal performance. Journal of Experimental
Biology, 213: 3705– 3709.
Cummings, G., Fiddler, F., and Vaux, D. L. 2007. Error bars in experimental biology. The Journal of Cell Biology, 177: 7 – 11.
Drinkwater, K. F. 2005. The response of Atlantic cod (Gadus morhua) to
future climate change. ICES Journal of Marine Science, 62: 1327–1337.
Easterling, D. R., Meehl, G. A., Parmesan, C., Changnon, S. A., Karl, T.
A., and Mearns, L. O. 2000. Climate extremes: observations, modeling, and impacts. Science, 289: 2068– 2074.
Ennis, G. P. 1995. Larval and postlarval ecology. In Biology of the Lobster
Homarus americanus, pp. 23– 46. Ed. by J. R. Factor. Academic Press,
Toronto.
Factor, J. R. 1995. Introduction, anatomy, and life history. In Biology of
the Lobster Homarus americanus, pp. 1– 11, Ed. by J. R. Factor.
Academic Press, Toronto.
Galbraith, P. S., Larouche, P., Chassé, J., and Petrie, B. 2012. Sea-surface
temperature in relation to air temperature in the Gulf of
St. Lawrence: interdecadal variability and long term trends. Deep
Sea Research II, 77 – 80: 10 – 20.
Green, B. G., Gardner, C., Hochmuth, D., and Linnane, A. 2014.
Environmental effects on fished lobsters and crabs. Reviews in Fish
Biology and Fisheries, 24: 613– 638.
Hadley, P. B. 1906. Regarding the rate of growth of the American lobster
(Homarus americanus). Annual Report of the Commissioners of
Inland Fisheries made to the General Assembly of Rhode Island,
36: 156– 235.
Harding, G. C., Drinkwater, K. F., Hannah, C. G., Pringle, J. D., Prena, J.,
Loder, J. W., and Pearre, S., Jr. et al. 2005. Larval lobster (Homarus
americanus) distribution and drift in the vicinity of the Gulf of
Maine offshore banks and their probable origins. Fisheries
Oceanography, 14: 112– 137.
Harding, G. C., Drinkwater, K. F., and Vass, W. P. 1983. Factors influencing the size of American lobster (Homarus americanus) stocks along
i90
the Atlantic coast of Nova Scotia, Gulf of St. Lawrence, and Gulf of
Maine: a new synthesis. Canadian Journal of Fisheries and Aquatic
Sciences, 40: 168– 184.
Heip, C. 1974. A comparison between models describing the influence
of temperature on the development rate of copepods. Biologisch
Jaarboek Dodonaea, 42: 121– 125.
Hordoir, R., and Meier, H. E. M. 2011. Effect of climate change on
thermal stratification of the Baltic Sea: a sensitivity experiment.
Climate Dynamics, doi:10.1007/s00382-011-1036-y.
Hudon, C., and Fradette, P. 1988. Planktonic growth of larval lobster
(Homarus americanus) off iles de la Madeleine (Quebec), Gulf of
St. Lawrence. Canadian Journal of Fisheries and Aquatic Sciences,
45: 868 –878.
Huffaker, C. B. 1944. The temperature relations of the immature stages
of the malarial mosquito, Anopheles quadrimaculatus Say, with a
comparison of the developmental power of constant and variable
temperatures in insect metabolism. Annals of the Entomological
Society of America, 37: 1 – 27.
Hughes, J. T., and Mathiessen, G. C. 1962. Observations on the biology
of the American lobster, Homarus americanus. Limnology and
Oceanography, 7: 414 – 421.
Huntsman, A. G. 1924. Limiting factors for marine animals. 2. Resistance
of larval lobsters to extremes of temperature. Contributions to
Canadian Biology and Fisheries, 2: 89–94.
Incze, L., and Naimie, C. E. 2000. Modelling the transport of lobster
(Homarus americanus) larvae and postlarvae in the Gulf of Maine.
Fisheries Oceanography, 9: 99– 113.
Incze, L., Xue, H., Wolff, N., Xu, D., Wilson, C., Steneck, R., Wahle, R.,
et al. 2010. Connectivity of lobster (Homarus americanus) populations in the coastal Gulf of Maine part II. Coupled biophysical dynamics. Fisheries Oceanography, 19: 1 – 20.
James, F. 1994. RANLUX: a Fortran implementation of the high-quality
pseudorandom number generator of Lüscher. Computer Physics
Communications, 79: 111 – 114.
Khan, A. H., Levac, E., and Chmura, G. L. 2013. Future sea surface temperatures in large marine ecosystems of the northwest Atlantic. ICES
Journal of Marine Science, 70: 915– 921.
Kough, A. S., Paris, C. B., and Butler, M. J., IV. 2013. Larval connectivity
and the international management of fisheries. PLoS ONE, 8: e64970.
Lawton, P., and Lavalli, K. L. 1995. Postlarval, juvenile, adolescent, and
adult ecology. In Biology of the Lobster Homarus americanus, pp.
47 – 88. Ed. by J. R. Factor. Academic Press, Toronto.
Loder, J. W., Han, G., Galbraith, P. S., Chassé, J., and van der Baaren, A.
(Eds). 2013. Aspects of climate change in the Northwest Atlantic off
Canada. Canadian Technical Report of Fisheries and Aquatic
Sciences, 3045. x + 190 pp.
Ma, J., and Xie, S-P. 2013. Regional patterns of sea surface temperature
change: a source of uncertainty in future projections of precipitation
and atmospheric circulation. Journal of Climate, 26: 2482– 2501.
Ma, Z., Han, G., and de Young, B. 2012. Modelling temperature, currents
and stratification in Placentia Bay. Atmosphere-Ocean, 50: 244–260.
MacKenzie, B. R. 1988. Assessment of temperature effects on interrelationships between stage durations, mortality, and growth in
laboratory-reared Homarus americanus Milne Edwards larvae.
Journal of Experimental Marine Biology and Ecology, 116: 87 – 98.
McLaren, I. A., Corkett, C. J., and Zillioux, E. J. 1969. Temperature adaptations of copepod eggs from the Arctic to the Tropics. Biological
Bulletin, 137: 486– 493.
Miller, R. J. 1997. Spatial differences in the productivity of American
lobster in Nova Scotia. Canadian Journal of Fisheries and Aquatic
Sciences, 54: 1613– 1618.
Miller, R. J., Mallet, M., Lanteigne, M., and MacMillan, R. 2006.
Bottlenecks to fishery yield for American lobster in a portion of
the Gulf of St. Lawrence. North American Journal of Fisheries
Management, 26: 765– 776.
B. K. Quinn and R. Rochette
Milne Edwards, H. 1837. Histoire naturelle des Crustacés, comprenant
l’anatomie, la physiologie et la classification de ces animaux, 2.
Librairie encyclopédique de Roret, Paris.
Niehaus, A. C., Angilletta, M. J., Jr., Sears, M. W., Franklin, C. E., and
Wilson, R. S. 2012. Predicting the physiological performance of
ectotherms in fluctuating thermal environments. Journal of
Experimental Biology, 215: 694– 701.
Ouellet, M., Petrie, B., and Chassé, J. 2003. Temporal and spatial scales of
sea-surface temperature variability in Canadian Atlantic waters.
Canadian Technical Report of Hydrography and Ocean Sciences,
228. v + 30 pp.
Pakyari, H., Fathipour, Y., and Enkegaard, A. 2011. Estimating development and temperature thresholds of Scholothrips longicornis
(Thysanoptera: Thripidae) on eggs of two-spotted spider mite using
linear and nonlinear models. Journal of Pest Science, 84: 153–163.
Pezzack, D. S. 1992. A review of lobster (Homarus americanus) landing
trends in the northwest Atlantic, 1947 – 86. Journal of Northwest
Atlantic Fishery Science, 14: 115– 127.
Quinn, B. K. 2014. Assessing potential influence of larval development
time and drift on large-scale spatial connectivity of American lobster
(Homarus americanus). MSc thesis, University of New Brunswick,
Saint John, NB, Canada.
Quinn, B. K., Rochette, R., Ouellet, P., and Sainte-Marie, B. 2013. Effect
of temperature on development rate of larvae from cold-water
American lobster (Homarus americanus). Journal of Crustacean
Biology, 33: 527 – 536.
Robert, D., Castonguay, M., and Fortier, L. 2007. Early growth and recruitment in Atlantic mackerel Scomber scombrus: discriminating
the effects of fast growth and selection for fast growth. Marine
Ecology Progress Series, 337: 209– 219.
Sastry, A. N. 1979. Metabolic adaptation of Cancer irroratus developmental stages to cyclic temperatures. Marine Biology, 51: 243 –250.
Sastry, A. N., and Vargo, S. L. 1977. Variations in the physiological
responses of crustacean larvae to temperature. In Physiological
Responses of Marine Biota to Pollutants, pp. 401– 424. Ed. by F. J.
Vernberg, A. Calabrese, F. P. Thurberg, and W. B. Vernberg.
Academic Press, New York, NY.
Schmalenbach, I., and Franke, H-D. 2010. Potential impact of climate
warming on the recruitment of an economically and ecologically important species, the European lobster (Homarus gammarus) at
Helgoland, North Sea. Marine Biology, 157: 1127– 1135.
Shi, P., and Ge, F. 2010. A comparison of different thermal performance
functions describing temperature-dependent development rates.
Journal of Thermal Biology, 35: 225 – 231.
Somero, G. N. 2004. Adaptations of enzymes to temperature: searching
for basic strategies. Comparative Biochemistry and Physiology, Part
B, 139: 321– 333.
Studer, A., and Poulin, R. 2013. Differential effects of temperature variability on the transmission of a marine parasite. Marine Biology, 160:
2763– 2773.
Templeman, W. 1936. The influence of temperature, salinity, light and
food conditions on the survival and growth of the larvae of the
lobster (Homarus americanus). Journal of the Biological Board of
Canada, 2: 485 – 497.
Thorp, J. H., and Wineriter, S. A. 1981. Stress and growth response of
juvenile crayfish to rhythmic and arrhythmic temperature fluctuations. Archives of Environmental Contaminants and Toxicology,
10: 69 – 77.
Tlusty, M., Metzler, A., Malkin, E., Goldstein, J., and Koneval, M. 2008.
Microecological impacts of global warming on crustaceans—
temperature induced shifts in the release of larvae from American
lobster, Homarus americanus, females. Journal of Shellfish
Research, 27: 443– 448.
Wahle, R. A., Incze, L. S., and Fogarty, M. J. 2004. First projections of
American lobster fishery recruitment using a settlement index and
variable growth. Bulletin of Marine Science, 74: 101– 114.
Handling editor: C. Brock Woodson
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i91 –i100. doi:10.1093/icesjms/fsv044
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Geographic and environmental drivers of fecundity
in the European lobster (Homarus gammarus)
Charlie D. Ellis 1,2, Hannah Knott 2,3, Carly L. Daniels 2, Matthew J. Witt 1, and David J. Hodgson 4 *
1
Environment and Sustainability Institute, University of Exeter, Penryn Campus, Penryn, Cornwall TR10 9FE, UK
National Lobster Hatchery, South Quay, Padstow, Cornwall PL28 8BL, UK
3
Marine Biology and Ecology Research Centre, School of Marine Science and Engineering, University of Plymouth, Plymouth, Devon PL3 8AA, UK
4
Centre for Ecology and Conservation, University of Exeter, Penryn Campus, Penryn, Cornwall TR10 9FE, UK
2
*Corresponding author: tel: +44 1326 371829; fax: +44 1326 254243; e-mail: [email protected]
Ellis, C. D., Knott, H., Daniels, C. L., Witt, M. J., and Hodgson, D. J. Geographic and environmental drivers of fecundity
in the European lobster (Homarus gammarus). – ICES Journal of Marine Science, 72: i91 – i100.
Received 31 August 2014; revised 22 February 2015; accepted 24 February 2015; advance access publication 18 March 2015.
Fecundity in the European lobster (Homarus gammarus) has been shown to exhibit extensive spatial variation across northern Europe. Previously,
this has been attributed to a lack of methodological standardization among samples. Instead, we show significant correlations between fecundity
and both geographical and environmental drivers. We use linear mixed-effect models to assess the contribution of latitude, longitude, and measures
of sea surface temperatures on the size– fecundity relationships of 1058 ovigerous females from 11 locations in the UK, Ireland, and Norway. We
include new data for 52 lobsters from Falmouth, UK, the southwest limit of existing samples. Fecundity at mean female size correlated positively with
eastings and greater annual ranges in sea surface temperature, but not with mean temperature or latitude. This contradicts the established latitudinal and mean temperature dependence reported for the closely related H. americanus. We postulate that proximity to stable Atlantic currents
is the most likely driver of the relationship between fecundity and longitude. Mechanisms are discussed by which egg production or retention may
be influenced by temperature range rather than by mean temperature. With further validation, we propose that temperature-correlated fecundity
predictions will provide a valuable tool in ensuring that management thresholds are appropriate for the reproductive characteristics of lobster
populations.
Keywords: crustacea, egg production, environmental driver, fishery management, general linear model, reproductive ecology, reproductive
variation.
Introduction
Measures of egg production are vital parameters for estimating the
reproductive capacity of marine populations, the maintenance of
which is a key objective of fishery management. Knowledge of reproductive capacity is critical for informed management of exploited
populations because it is required for models of stock and recruitment dynamics and can be used to define the maximum threshold
for fishing mortality (Laurans et al., 2009). It is also important to determine the geographic scales over which the reproductive characteristics of managed species vary to apply commensurate stock
conservation measures to each region (Tully et al., 2001;
MacCormack and DeMont, 2003; Currie and Schneider, 2011).
A size-specific fecundity factor is well documented in populations of the European lobster (Homarus gammarus, L.), a prized
# International
decapod crustacean fished extensively throughout its range (e.g.
Hepper and Gough, 1978; Bennett and Howard, 1987; Tully et al.,
2001; Lizarraga-Cubedo et al., 2003; Agnalt et al., 2007; Agnalt,
2008). However, published estimates of mean fecundity have
varied considerably among putative populations throughout northwest Europe (Agnalt, 2008), ranging from 5200 eggs per oviposition in southeast Scotland (Lizarraga-Cubedo et al., 2003) to
12 500 in southern England (Roberts, 1992) and southwest
Norway (Agnalt, 2008), among females of 100 mm carapace
length (CL).
Environmental determinants of fecundity variation have been
identified in many marine species (Wright, 2013), including seawater parameters such as temperature, salinity (e.g. Gomez et al.,
2013), and dissolved oxygen (e.g. Wu et al., 2003). Temperature
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i92
(or latitude, as a proxy) has been found to correlate tightly with the
exponent of size-specific fecundity variation in American lobster
(Homarus americanus; Currie and Schneider, 2011). It also aligns
with reproductive traits in other lobsters, including Southern rock
lobster (Jasus edwardsii; Annala et al., 1980; Gardner et al., 2006),
and in fish inhabiting a similar range throughout the Northeast
Atlantic, such as Atlantic cod (Gadus morhua; Thorsen et al.,
2010; Wright et al., 2011a; Hansen et al., 2012) and Dover sole
(Solea solea; Witthames et al., 1995; Mollet et al., 2013). We aimed
to test associations between H. gammarus fecundity and geographical and environmental factors, to assess whether they may contribute to the observed spatial variation in fecundity. Management has
failed to prevent extensive and enduring stock collapses in the recent
past (e.g. throughout Scandinavia in the mid-20th century; Dow,
1980; Agnalt et al., 1999), and where stock thresholds fail to reflect
regional differences in fecundity, the management of pressured fisheries can be seriously undermined (Lambert, 2008; Morgan, 2008).
Therefore, the identification of drivers that explain reproductive
variation may be important in conserving lobster populations
(Green et al., 2014).
Despite the established influence of ecological drivers in reproductive variation across a range of taxa, whether regional differences
in H. gammarus fecundity may be driven by environmental factors
has not been assessed. Observed variation in clutch size among
clawed lobsters has been attributed to differences in the success of
attaching the externally incubated eggs (Currie and Schneider,
2011), the rate of egg loss over a lengthy incubation of 9 –10
months (Wahle et al., 2013), and the retention of eggs during
capture and subsequent handling and storage (Agnalt, 2008).
Agnalt (2008) hypothesized that a lack of methodological standardization among studies may prevent the detection of population-level
variations, but we aimed to assess whether the influence of thermal
environment might be detectable within the observed variation of
H. gammarus fecundity.
We hypothesized that a relationship would exist between temperature and fecundity among putative populations of H. gammarus. To
test this hypothesis, egg counts of ovigerous females were collated
from existing studies of fecundity in northern Europe. A new fecundity measurement was also made for females from the Atlantic peninsula of Cornwall, UK, an unassessed region at the southwest edge of
the range of available data where the lobster fishery is vital in supporting 370 commercial potting vessels (S. Davies, pers. comm.; Cornwall
IFCA, 2014). Parameters of the size-specific fecundity relationships of
these samples were regressed against geographical and environmental
covariates. We find longitudinal and environmental predictors of
fecundity at mean size and discuss our findings in relation to
lobster physiology, evolutionary ecology, and fishery management.
Material and methods
New samples
Animal acquisition and storage
Ovigerous female lobsters (n ¼ 52) were caught in baited pots and
collected directly from inshore fishers working in Falmouth Bay,
southwest UK in January –March 2013. This was carried out with
permission from the local authority, as the landing of ovigerous
females within inshore waters is normally prohibited (Cornwall
IFCA, 2014). A large and evenly distributed range in female sizes
was requested because this improves the accuracy of estimates of
size –fecundity relationships (Estrella and Cadrin, 1995). A broad
size range was achieved, although legal landing restrictions meant
C. D. Ellis et al.
that no females could be obtained less than the 90-mm CL jurisdictive
minimum landing size. Most females were sampled immediately upon
collection; where this was not possible, females were stored for a
maximum of 3 d in a modern 2000 l recirculation system, where
chilled temperatures (5–68C), shelter provisions, and low stocking
density (maximum 3 m22) ensured egg loss was negligible (daily net
cleaning revealed that egg loss equated to ,10 eggs lobster21 d21).
Physical fecundity estimation
CL was measured using Vernier calipers, rounding down to the nearest
whole millimetre, and the egg mass was collected by hand, as per
Agnalt (2008). A subsample of the eggs was separated and counted
manually, ranging from 517 to 708 individual eggs (mean ¼ 606,
+3.45). No repeat subsamples were taken because Agnalt (2008)
showed that the correlation between two counts was .0.99 using
even smaller subsamples [wet weights of 1–1.5 g, compared with
2.2–3.9 g (mean ¼ 2.97 g, +0.05 g) in this study]. Egg development
was similar among all females, with most clutches being partially
“eyed”, although no formal measurements of development stage
were taken.
Individual fecundity estimates were made by calculating the dry
weight of the egg mass against that of the counted subsample; dry
weight was preferred so that any variation in the amount of seawater
incidentally gathered with the egg mass would not bias the measurement. All egg samples were dried in a drying oven (UT6200, Thermo
Electron LED, Germany) at 1058C for 24 h (+1 h). Samples were
moved into a sealed desiccating cabinet to cool before mass was
measured to the nearest 1 mg by electronic balance (AE240
Balance, Mettler, UK). After an additional hour in the drying
oven, sample mass was remeasured to check that it was stable and
that drying had completed; all samples were deemed fully dried
after this check because the difference in mass between the measurements was ,1% of the total sample mass. The dry mass of the subsample of known egg count was used to determine the mean dry
mass per egg as:
Dry mass per egg (mg) =
Subsample dry mass (mg)
.
Subsample size (n eggs)
(1)
Fecundity estimates for each individual were then obtained from the
total dry mass of eggs as:
Subsample dry mass (mg)+Remaining
sample dry mass (mg)
Fecundity (n eggs) =
.
Dry mass per egg (mg)
(2)
Geographical survey
Data collection and statistical modelling
To test potential geographic and environmental drivers of fecundity
variability in H. gammarus, data were collected from five studies
assessing fecundity among 1009 individuals in 10 areas around
the UK, Ireland, and Norway, plus the 52 individuals from
Falmouth, southwest UK (Figure 1), measured by this study. Each
regional sample location was assigned latitudinal and longitudinal
coordinates from the approximate centre of the spatial range of sampling, as could be best deduced from study methodologies. The
mean sea surface temperature (SST) data were obtained for each location the first day of each month during the year(s) of the study and
one preceding year, since the majority of Homarid lobsters spawn in
Geographic and environmental drivers of fecundity
i93
Figure 1. Map of the UK and Ireland, with continental Europe around the North Sea, showing the locations of regional fecundity samples.
Fecundity in Falmouth (F) was assessed in this study, while other samples used to model correlations with temperature were: Arranmore (A), Galway
(G), Cork (C), and Rosslare (R) from Tully et al. (2001); St Davids (D) from Bennett and Howard (1987); Milford Haven (M), Selsey (S), and Bridlington
(B) from Free (1994); Poole (P) from Roberts (1992); and Kvitsøy (K) from Agnalt (2008). See Table 1 for further information on regional samples.
a biennial cycle (Tully et al., 2001; Comeau and Savoie, 2002; Agnalt
et al., 2007). Using SST data, the mean temperature (mean SST of all
months in all years) and temperature range (the mean difference
between the mean SST of the three coldest months and the mean
SST of the three coldest months of each year) were calculated for
each location. SST data were obtained via AVHRR Oceans
Pathfinder from the Physical Oceanography Distributed Active
Archive Center (PO.DAAC, 2014) for all locations except
Falmouth, UK, for which SST data were only available via MODIS
Aqua EOS-PM from the Goddard Space Flight Center
(OceanColor, 2014) due to the recentness of the sampling.
SST was utilized instead of seabed temperature (SBT) because
SBT was unavailable at the spatial and temporal resolutions
required. While SBT may present a more biologically relevant parameter for benthic lobsters, the use of SST was supported by a
regression of 80 surface (mean ¼ 1.8 m below surface) and
bottom (mean ¼ 3.3 m above seabed) temperature measurements
obtained by depth casts (ICES Data Centre, 2014) taken between
1998 and 2008 at fishable locations (within 15 km of the coast and
,85 m depth) across the geographic range of the study. The relationship showed a highly significant correlation between surface
and bottom temperatures (Pearson’s product-moment correlation,
r 2 ¼ 0.96, p , 0.01).
General linear models (GLMs) were constructed using R (R Core
Team, 2012) to apply power (log –log), log-linear, and linear fits to
the global relationship between fecundity (F ) and female size (CL)
across all 1061 individuals. Analysis of the distribution of residuals
and comparisons of the log-likelihood ratio statistic and Akaike information criterion (AIC) of each model confirmed that the power
fit, log(F) = log (aCL)b , best described this relationship (see
i94
C. D. Ellis et al.
Table 1. Summary of regional samples analysed including: study origin; sample region; sample size (n); central coordinates used for sample SST
data and geographic factors in modelling, SST-derived mean temperature, and temperature range; a and b (Fslope) of the power-fitted
relationship between fecundity and CL (F ¼ aCLb), with r 2 and associated p-values; and Fmean.
Study
Ellis et al. (this study)
Tully et al. (2001)
Tully et al. (2001)
Tully et al. (2001)
Tully et al. (2001)
Bennett and Howard
(1987)
Free (1994)
Free (1994)
Free (1994)
Roberts (1992)
Agnalt (2008)
SST range
(88 C)
6.76
b
a
(Fslope)
0.0066 3.08
Fmean
r 2 p-value (n eggs)
0.68 ,0.001 11 011
8830′ 36′′ W 11.64
4.63
0.0042 3.18
0.81 ,0.001
9835′ 60′′ W
9824′ 0′′ W
6824′ 0′′ W
5819′ 48′′ W
12.41
12.87
12.13
11.10
5.31
5.98
7.40
7.13
0.0017
0.0031
0.0164
0.0003
3.29
3.18
3.01
3.42
0.73
0.57
0.49
0.73
588′ 24′′ W
11.75
7.19
0.0000 3.14
0.48
76 50842′ 36′′ N 0846′ 48′′ W 12.94
177 5484′ 48′′ N 0810′ 12′′ W 10.06
7.81
8.68
0.1827 2.85
0.0344 2.84
0.26 ,0.001 11 622
0.59 ,0.001 11 776
50 50840′ 48′′ N 1858′ 12′′ W 12.49
9.84
217 5983′ 36′′ N 5826′ 24′′ E
7.05
9.38
0.0114 3.03
0.0047 3.11
0.53 ,0.001 11 208
0.85 ,0.001 12 920
Sample region
Falmouth, SW
England
Arranmore, NW
Ireland
Galway, W Ireland
Cork, SW Ireland
Rosslare, SE Ireland
St Davids, SW Wales
n
Lat.
52 5088′ 24′′ N
Milford Haven, SW
Wales
Selsey, S England
Bridlington, NE
England
Poole, S England
Kvitsøy, SW Norway
8 51842′ 0′′ N
73 5580′ 36′′ N
144
70
111
80
5386′ 36′′ N
51827′ 36′′ N
52810′ 12′′ N
51852′ 12′′ N
Long.
581′ 48′′ W
Supplementary material). Power law models have been favoured in
other recent studies (e.g. Tully et al., 2001; Lizarraga-Cubedo et al.,
2003; Agnalt, 2008) because they account for the volumetric nature
by which the brooding capacity of the abdomen increases in length
and width with increasing CL. The outlying data of three individuals
for which fecundity estimates lay beyond 4 s.e. of the allometric relationship were removed from the analysis.
A linear mixed-effects model was constructed using the
R package lme4 (Bates et al., 2014) to test the effect of the sizes of
potential geographical and environmental drivers on lobster
fecundity. Geographical factors assessed were latitude, longitude,
and the interaction between the two, while environmental factors
(analysed separately) were mean temperature, temperature range,
and the interaction between the two. The relative strength of all geographical and environmental covariates was standardized via an adjustment to similar scales (mean ¼ 0; s.d. ¼ 1). From these models,
coefficients of log(fecundity) at the mean size of all sampled females
(Fmean) and the exponent of the size-specific fecundity power relationship (Fslope; b value) were extracted for each regional sample
and then regressed in GLMs against the scaled geographical or environmental covariates. All combinations of models containing the
effects of geographical or environmental factors on Fmean and
Fslope in each regional sample were compared using multimodel inference and model averaging in the R package MuMIn (Barton,
2013). Model-averaged effect sizes and AIC weights (the proportion
of weight accumulated by all models containing the assessed variable) were extracted to evaluate the relative importance of each variable on Fmean and Fslope. Correlation between geographical and
environmental factors was tested by linear regressions. The GLMs
used to regress Fmean and Fslope against geographical and environmental parameters were weighted by the sample sizes studied in
each lobster population to limit the influence of imprecise estimates
on global relationships.
Some existing fecundity samples within the spatial range investigated were not analysed because raw data were unavailable (e.g.
eastern and western Scotland; Lizarraga-Cubedo et al., 2003) or
were collected before the backdated availability of SST measurements (e.g. northwest France; Latrouite et al., 1984). Data from
another sample taken near Whitby in northeast England by
SST mean
(88 C)
11.85
9559
,0.001 9353
,0.001 8947
,0.001 10 105
,0.001 9466
0.02
10 293
Bennett and Howard (1987) were omitted because it was deemed
likely that they were biased by considerable egg loss before fecundity
estimation. The data included extremely low egg counts (e.g. ,750
eggs) and yielded a very low correlation for the power-fitted size –fecundity relationship (r 2 ¼ 0.12). A sample from Milford Haven,
Wales (Free, 1994), was included despite the sample size being
very small (n ¼ 8) because the data exhibited a reasonable correlation for a power-fitted size –fecundity slope (r 2 ¼ 0.62).
Results
Physical fecundity estimation
Among females collected from Falmouth, UK, CL ranged from 90
to155 mm (mean ¼ 110 mm, +1.9 mm), and estimated egg production ranged from 3712 to 35 241 eggs ind.21. The relationship
between fecundity (F ) and female size (CL) was described by F ¼
0.0066 CL3.10 using a power-fitted model (r 2 ¼ 0.68, p , 0.001;
Table 1), or by F ¼ 406.92 CL 229 749 using a linear-fitted model
(r 2 ¼ 0.77, p , 0.001). The mean dry mass egg21 ranged from
1.53 to 2.24 mg among females, but demonstrated no relationship
with overall fecundity (linear fit; r 2 ¼ 0.14, p , 0.01). Mass egg21
appeared to fit a natural logarithm relationship with female size,
as described by Agnalt (2008), although overall correlation of this
model fit was weak (r 2 ¼ 0.29, p , 0.001; see Supplementary material). Compared with the sample from Kvitsøy (K) and pooled Irish
samples (I), the mean dry mass egg21 (mg) at Falmouth (F) was
slightly higher at the lower distribution of female sizes (90 mm
CL: K ¼ 1.3; I ¼ 1.4; F ¼ 1.6), but was comparable at upper size
limits (150 mm CL: K ¼ 1.9; I ¼ 1.9; F ¼ 2.0; Tully et al., 2001;
Agnalt, 2008). Estimates suggest that fecundity among lobsters
from Falmouth is fairly central within the range recorded for the
species across northern Europe, despite the location lying at the
southwest geographical extremity of all samples.
Drivers of fecundity variation
Table 1 shows SSTand fecundity relationship results for each regional sample. North Sea sites at Kvitsøy and Bridlington had both the
lowest mean temperatures (9.84 and 10.068C, respectively) and
highest temperature ranges (9.38 and 8.688C). The mean
i95
Geographic and environmental drivers of fecundity
temperature was highest at sites in the English Channel at Selsey
(12.948C) and Poole (12.498C) and in the Northeast Atlantic off
western Ireland at Cork (12.878C) and Galway (12.418C). Western
Ireland also experienced the smallest temperature ranges, decreasing
northwards from Cork (5.988C) to Galway (5.318C) and being
lowest at Arranmore (4.638C). Across all samples, Fmean corresponded to a female size of 102.8 mm CL. For the log power-fitted
relationship, log(F) = log(aCL)b , b (Fslope) was lowest for the
samples from Bridlington (2.84) and Selsey (2.85), and was
highest for the St Davids sample (3.42). Fmean ranged from 8947
eggs female21 in Cork to 12 920 in Kvitsøy. For all North Sea and
English Channel samples, Fmean exceeded 11 000 eggs female21,
whereas it was below 10 300 eggs for all samples from the Irish Sea
and western Ireland.
We found that increases in Fmean were strongly associated with
increases in both (easterly) longitude and mean annual temperature
range (Figure 2). Each variable had a high cumulative AIC weight
(temperature range ¼ 0.92; longitude ¼ 0.89; Table 2), and a
model-averaged effect size identifiably .0, with 95% confidence
intervals not overlapping zero (Figure 3). The influence of longitude
and temperature range on fecundity also extends to females in other
size classes. These variables also yielded identifiable positive effect
sizes in linear mixed-effect models of fecundity at the current
European Commission minimum landing size of 87 mm CL (data
not presented). Latitude and mean temperature variables, and interactions of these factors, had no influence on fecundity variation,
however. Modelled with Fmean, these variables had low cumulative
model weightings (AIC weights ,0.1) and 95% confidence intervals that spanned an effect-size of zero (Figure 3). We also demonstrated that variation in Fslope could not be attributed to any of the
geographical or environmental variables investigated (Figure 3).
No variable had an identifiable effect upon Fslope, with confidence
intervals spanning zero effect-sizes and low cumulative weighting
(AIC weights ,0.4) for all model factors. Linear regressions
between variables showed a significant positive correlation
between mean annual temperature range and longitude (Pearson’s
coefficient: r 2 ¼ 0.90, p , 0.001; Figure 4), and a significant negative relationship between latitude and mean temperature among regional fecundity samples (r 2 ¼ 20.74, p , 0.01).
Discussion
Figure 2. Plot of the relationship between Fmean and scaled
temperature range, showing that increased Fmean was positively
associated with increased range in annual temperature (r 2 ¼ 0.83,
p , 0.002).
Knowledge of factors contributing to fecundity variation is vital to
ensure that fishery management strategies are suitable for exploited
species throughout their range (Lambert, 2008; Morgan, 2008). We
have demonstrated geographical and environmental factors that
correlate with fecundity variation in H. gammarus across a
portion of its range which has accounted for over 75% of the
species’ recorded landings in recent years (FAO, 2014). Our
results are an important indication that the observed spatial variation may reflect differences between the fecundity of putative
populations, not simply study-level differences in investigative approach, and that environmental temperature is a driver contributing
to variation in the production and/or retention of eggs in H. gammarus. In isolation, the new fecundity sample taken from Falmouth,
Table 2. Summary of candidate linear mixed models with measures of model likelihood and weighting to show the effect of geographical and
environmental covariates on Fmean.
F parameter
Fmean
Factors
Geographical
Environmental
Fslope
Geographical
Environmental
Model variables
Longitudea
Latitude + longitude
Latitude + longitude + latitude:longitude
Latitude
Temperature rangea
Mean temperature + temperature range
Mean temperature
Mean temperature + temperature range + mean
temperature:temperature range
Longitude
Latitude + longitude
Latitude
Latitude + longitude + latitude:longitude
Temperature range
Mean temperature
Mean temperature + temperature range
Mean temperature + temperature range + mean
temperature:temperature range
Factors denoted a were deemed identifiable effects by model-averaging.
d.f.
3
4
5
3
3
4
3
5
3
4
4
5
3
3
4
5
logLik
20.966
21.406
22.075
8.444
14.687
14.816
9.712
14.948
AICc
232.5
228.1
222.2
27.5
219.9
215.0
210.0
27.9
3.454
5.147
2.313
7.036
4.060
2.565
4.680
4.997
2.5
4.4
4.8
7.9
1.3
4.3
5.3
12.0
D AIC
0.00
4.36
10.35
25.04
0.00
4.98
9.95
12.05
AIC weight
0.894
0.101
0.005
0.000
0.915
0.076
0.006
0.002
1.64
3.49
3.92
7.05
0.43
3.42
4.43
11.13
0.251
0.099
0.080
0.017
0.384
0.086
0.052
0.002
i96
C. D. Ellis et al.
Figure 3. Model-averaged effect sizes of geographical and environmental variables modelled against the fecundity parameters Fslope and Fmean.
Variables with effect-sizes that are identifiably different from zero have 95% confidence interval bars that do not overlap the model mean (dashed
vertical line).
Figure 4. Plot of the relationship between longitude and mean
annual temperature range among regional fecundity samples
(r 2 ¼ 0.90, p , 0.001).
the first such assessment in southwest England, can contribute an
important parameter of the reproductive capacity of H. gammarus
in this important regional fishery.
Our most important findings are that fecundity at mean size
improved with increasing range in annual temperature, along a gradient towards more easterly longitudes, and that longitude and temperature range were closely associated. The most obvious underlying
driver linking gradients of longitude and temperature range in the
area of this study is proximity to the North Atlantic Drift of the
Gulf Stream. The North Atlantic Drift brings greater thermal stability to the coastal waters of the immediate Atlantic coast along
western Europe than that experienced by more enclosed shelf sea
areas. By example, among the three most northerly regional
samples we surveyed, the mean annual range in sea temperature
for the Northeast Atlantic at Arranmore was only 4.68C, compared
with 8.78C around Bridlington and 9.48C at Kvitsøy in the North
Sea. Considering the strength of the associations we found
between fecundity at mean female size and both longitude and temperature range, we propose that proximity to currents associated
with the Gulf Stream contributes to the regulation of egg production
i97
Geographic and environmental drivers of fecundity
Table 3. Regional samples ranked via smallest SOM LP50 (the CL at which 50% of females are functionally mature), as physiologically
determined by Free et al. (1992) and Tully et al. (2001), with comparison to fecundity at the global mean female size (Fmean) as calculated in
this study using raw data from Free (1994) and Tully et al. (2001).
Study
Free et al. (1992), Free (1994)
Tully et al. (2001)
Sample region
Selsey, S England
Bridlington, NE England
Galway, W Ireland
Cork, SW Ireland
Rosslare, SE Ireland
Arranmore, NW Ireland
and/or retention in H. gammarus across the northern part of the
species’ distribution.
In contrast to the relationship detected for H. americanus by
Currie and Schneider’s (2011) similar meta-analysis of spatial variation in fecundity, we found no evidence of the slope of size-specific
fecundity increasing with decreased latitudinal gradient. Instead, we
found that fecundity at mean size was increased among regions with
high ranges in annual temperature, irrespective of mean temperature. This finding defies the expectation that mean temperature
drives the reproductive investment of ectotherms (e.g. Ernsting
and Isaaks, 2000; Thorsen et al., 2010; Tobin and Wright, 2011;
Wright et al., 2011a). Currie and Schneider (2011) found that
fecundity-at-size in H. americanus (in this case, 85 mm CL lobsters)
met this expectation, as it aligned closely to latitudinal gradient.
However, the direction of this relationship was unexpected, with
fecundity-at-size found to increase in higher latitudes (Currie and
Schneider, 2011), suggesting that clutch size does not increase
with increasing temperature in either Homarus species.
Rather than being a function of size, Currie and Schneider (2011)
propose that H. americanus fecundity may be age-related, with fewer
growing degree-days (e.g. Neuheimer and Taggart, 2007) at higher
latitudes leading to smaller size at maturity and comparatively
greater clutches at equivalent body sizes. Age-at-size validation
methods remain too unreliable among crustaceans (Hartnoll,
2001) to evidence this, but the proposition is not supported by
Currie and Schneider’s (2011) own assertion of overall increases
in size-specific fecundity slopes towards southerly latitudes, nor
by our finding of a disconnect between fecundity at mean size and
mean temperature in H. gammarus. A comparable pattern to that
which we revealed is shown by sole (S. solea) populations from
colder North Sea environments, whose earlier maturity and
higher reproductive investment compared with conspecifics from
warmer seas to the south and west has been attributed to countergradient environmental adaptation. This suggests that greater fecundity can arise among populations inhabiting colder regions to
compensate for high mortality caused by winter sea temperatures
(Conover, 1992; Mollet et al., 2013), and that similar pressures
could be driving variation in egg production for H. gammarus. In
most studies of fish, spatial and temporal trait adaptations associated
with temperature variation have been attributed to phenotypic plasticity (Crozier and Hutchings, 2014), although evolutionary mechanisms are more commonly proposed to explain counter-gradient
variations (Conover, 1992; Mollet et al., 2013). Compared with
plastic traits, locally adapted fecundity variation is less likely to be flexible to global climate change (Conover et al., 2009), and evidence of
such adaptation to thermal gradients has already been established
among H. americanus populations across the Atlantic, with larval
growth and planktonic duration found to be comparatively shortened
under local sea temperatures (Quinn et al., 2013).
SOM (CL, mm) (rank)
82 (1)
90 (2)
92 (3)
94 (4)
95 (5)
96 (6)
Fmean (n 3 103 eggs) (rank)
11.6 (2)
11.8 (1)
9.35 (5)
8.95 (6)
10.1 (3)
9.56 (4)
Reported variation in size at the onset of maturity (SOM) also
appears to support the suggestion that geographical and environmental factors may influence reproductive ecology atypically in H.
gammarus. Female SOM has been estimated to be generally
smaller in those samples farther from the mild Northeast Atlantic
currents (Table 3), despite an expectation to positively align with
mean temperature as a product of greater energy acquisition and
growth rate (e.g. Zuo et al., 2011; Green et al., 2014), as has been
asserted for H. americanus (Little and Watson, 2003, 2005; Caputi
et al., 2013). Physiological assessments found SOM to be smaller
in Bridlington than at any location around Ireland (Free, 1994;
Tully et al., 2001), and morphologically determined SOM was
lower in the Scottish North Sea than at the Hebridean Atlantic
coast. In both scenarios, lobsters mature at smaller sizes in the
area of greater temperature range, despite those areas experiencing
lower overall mean temperatures. Assessing the relative contributions of environmental, demographic, and genotypic factors can
be extremely challenging (Wright, 2013), but the alignment of multiple traits to gradients of temperature range is a strong indicator
that reproductive variation in H. gammarus is driven by thermal environment.
It is not possible to disentangle whether the observed spatial variation in H. gammarus fecundity arises as a result of differences in the
production of eggs or in the retention of eggs after oviposition, or
both. Agnalt (2008) measured fecundity soon after extrusion and
again soon before hatch, and detected no egg loss across 7 months
among lobsters from Kvitsøy, whereas Latrouite et al. (1984) estimated that 27% of eggs were lost during incubation off the northwest coast of France. Agnalt (2008) sourced lobsters stringently
and argued that the egg loss observed by Latrouite et al. (1984)
could have arisen from handling and inappropriate storage,
factors well known to downwardly bias subsequent egg counts.
Nevertheless, most studies of H. americanus imply that 15% or
more of eggs are lost during incubation (Wahle et al., 2013), and
egg retention could exist as a result of thermal environment, so
egg loss during incubation cannot be discounted as a mechanism
of H. gammarus fecundity variability. Egg loss among communally
captive H. gammarus is dramatically reduced below a thermal
tipping point of 98C (B. Marshall, pers. comm.), with decreased
metabolism and movement inhibiting behaviours and interactions
which otherwise inhibit egg retention. It is also conceivable that the
diversity and abundance of known fungal and nemertean pathogens
of lobster eggs (e.g. Alderman and Polglase, 1986; Campbell and
Brattey, 1986) is influenced by sea temperatures. However, speculative
hypotheses that rate of egg loss may be improved in colder winters are
tempered by the extended duration of the incubation period at
lower temperatures (Charmantier and Mounet-Guillaume, 1992;
Schmalenbach and Franke, 2010) and by our analysis of samples
from Selsey and Poole, which also had high fecundity at mean size,
i98
but where high temperature ranges were driven by warm summers
rather than cold winters.
Although there is a tendency for mass egg21, egg and larval size,
and larval robustness to increase with female size (Tully et al., 2001;
Agnalt, 2008; Moland et al., 2010), scant evidence has been found
of any trade-off between quantity and quality of egg production in
H. gammarus. Investment per egg in terms of dry mass appears consistent between samples from Ireland, Kvitsøy, and Falmouth and
showed no discernible association to clutch size in our Falmouth
sample. In the geographic range of this study, it is also unlikely that
fecundity variation arises as a result of regional differences in spawning frequency, as a biennial reproductive cycle has been recorded for
the majority of lobsters in both Norway and Ireland (Tully et al., 2001;
Agnalt et al., 2007), althoughvariation in spawning strategies is apparent in the genus and is poorly understood (Gendron and Ouellet,
2009). Fishing-induced mortality is another candidate driver of
spatial variation in lobster fecundity. A response to selection pressures
incurred via recruitment overfishing has been proposed to explain
temporal fecundity increases in North Sea populations of cod
(G. morhua), haddock (Melangrammus aeglefinus), and plaice
(Pleuronectes platessa; Yoneda and Wright, 2004; Rijnsdorp et al.,
2005; Stares et al., 2007; Wright et al., 2011b) and was also considered
as a driver of temporal SOM variation in H. americanus (Landers
et al., 2001). Among the samples we investigated, the highest fecundities at mean size were recorded from the post-collapse population at
Kvitsøy (Agnalt et al., 1999) and the samples from Bridlington, Selsey,
and Poole, which are from stocks in the east and south of England that
experience heavier fishing pressure than those of Atlantic coasts
towards the southwest (Cefas, 2011). The status of stocks around
Ireland and Wales are not known. The strong effects of longitude
and temperature range that we identified suggest that any demographic pressure must also align closely with these gradients, although
from the limited information available on current and historical
fishing pressure, this does seem to be the case for H. gammarus in
parts of northern Europe.
The confirmation and elucidation of geographical and/or environmental drivers of fecundity variation would be valuable to the
management of reproductive potential in H. gammarus stocks, especially among unassessed regions in lieu of laborious manual quantifications (Currie and Schneider, 2011). Predictions facilitated via
relationships we have demonstrated with temperature range may
be a suitable method of fecundity estimation among unmeasured
populations, although the associations we found between temperature and fecundity are not as categorical as those offered by Currie
and Schneider (2011) for H. americanus. This may be an artefact
of uncontrolled variation in the effective spatial ranges of the regional samples we analysed. Our findings would be strengthened by the
standardized assessment of H. gammarus fecundity in other regions
within the spatial range encompassed by this study, as well as in areas
such as Subarctic Norway, the Iberian peninsula, Morocco, and the
Mediterranean to determine whether temperature range may be a
driver of clutch size throughout the species’ range. Repeat estimations in regions previously assessed could elucidate whether fecundity varies temporally as well as spatially, and provide further evidence
that the recorded variation in lobster fecundity reflects populationlevel differences in the production and/or retention of eggs, rather
than inherent bias between samples.
Conclusions
We show that the fecundity of European lobsters at mean female size
correlates positively with easterly longitude and annual range in
C. D. Ellis et al.
SSTs across the northern range of this species. Fecundity at mean
size did not correlate with mean temperature or latitude, contradicting the widely assumed temperature dependence of ectotherms. We
propose that the proximity of populations to stable Atlantic currents
is the driver of this variation. With further validation, temperaturecorrelated fecundity predictions would provide a valuable tool in
ensuring that conservation management is suited to the reproductive characteristics of lobster populations.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
We are indebted to Ann-Lisbeth Agnalt and Oliver Tully, for supplying fecundity data and references for this analysis, and to Liz Preston
of the University of Plymouth’s Aquatic Animal Nutrition and
Health Research Group, for methodological assistance. We further
acknowledge the generous help of several Cornish fishers who
donated lobsters for this study and to the Cornwall Inshore
Fisheries and Conservation Authority, for granting permission for
some sampling required for this investigation, and for contributing
useful information to our study. Finally, we are grateful to the
European Social Fund and the Fishmongers’ Company, UK,
whose funding contributed to aspects of this project, and to
Emory Anderson and two anonymous reviewers, whose input
greatly improved the manuscript.
References
Agnalt, A. L. 2008. Fecundity of the European lobster (Homarus gammarus) off southwestern Norway after stock enhancement: do cultured females produce as many eggs as wild females? ICES Journal
of Marine Science, 65: 164 – 170.
Agnalt, A. L., Kristiansen, T. S., and Jørstad, K. E. 2007. Growth, reproductive cycle, and movement of berried European lobsters (Homarus
gammarus) in a local stock off southwestern Norway. ICES Journal of
Marine Science, 64: 288– 297.
Agnalt, A. L., van der Meeren, G. I., Jørstad, K. E., Næss, H., Farestveit,
E., Nøstvold, E., Svåsand, T., et al. 1999. Stock enhancement of
European lobster (Homarus gammarus): a large-scale experiment
off South-western Norway (Kvitsøy). In Stock Enhancement and
Sea Ranching, pp. 401 – 419. Ed. by B. Howell, E. Moksness, and T.
Svåsand. Fishing News Books Ltd, Farnham, England. 616 pp.
Alderman, D. J., and Polglase, J. L. 1986. Are fungal diseases significant
in the marine environment? In The Biology of Marine Fungi, pp.
189– 199. Ed. by S. T. Moss. CUP Archive, Cambridge, England.
400 pp.
Annala, J. H., McKoy, J. L., Booth, J. D., and Pike, R. B. 1980. Size at the
onset of sexual maturity in female Jasus edwardsii (Decapoda:
Palinuridae) in New Zealand. New Zealand Journal of Marine and
Freshwater Research, 14: 217 – 227.
Barton, K. 2013. MuMIn: Multi-model Inference. R package version 1.9.
13. R Foundation for Statistical Computing, Vienna, Austria.
http://CRAN.R-project.org/package=MuMIn.
Bates, D., Maechler, M., Bolker, B., and Walker, S. 2014. lme4: Linear
Mixed-effects Models using Eigen and S4. R package version 1.1-7.
http://CRAN.R-project.org/package=lme4.
Bennett, D. B., and Howard, A. E. 1987. Estimates of lobster (Homarus
gammarus) fecundity from east and west Britain. ICES Document
CM 1987/K: 47.
Campbell, A., and Brattey, J. 1986. Egg loss from the American lobster,
Homarus americanus, in relation to nemertean, Pseudocarcinonemertes
homari, infestation. Canadian Journal of Fisheries and Aquatic
Sciences, 43: 772–780.
Geographic and environmental drivers of fecundity
Caputi, N., de Lestang, S., Frusher, S., and Wahle, R. A. 2013. The impact
of climate change on exploited lobster stocks. In Lobsters: Biology,
Management, Aquaculture and Fisheries, 2nd edn, pp. 98– 126.
Ed. by B. Phillips. John Wiley & Sons, Hoboken. 488 pp.
Cefas. 2011. English regional stock assessments for the European lobster
(Homarus gammarus). http://www.cefas.defra.gov.uk/our-science/
fisheries-information/commercial-species/shellfish.aspx (last accessed
18 January 2015).
Charmantier, G., and Mounet-Guillaume, R. 1992. Temperaturespecific rates of embryonic development of the European lobster,
Homarus gammarus (L.). Journal of Experimental Marine Biology
and Ecology, 160: 61 – 66.
Comeau, M., and Savoie, F. 2002. Maturity and reproductive cycle of the
female American lobster, Homarus americanus, in the southern Gulf
of St. Lawrence, Canada. Journal of Crustacean Biology, 22:
762– 774.
Conover, D. O. 1992. Seasonality and the scheduling of life history at different latitudes. Journal of Fish Biology, 4: 161 – 178.
Conover, D. O., Duffy, T. A., and Hice, L. A. 2009. The covariance
between genetic and environmental influences across ecological gradients. Annals of the New York Academy of Sciences, 1168: 100 – 129.
Cornwall IFCA. 2014. Summary of Cornwall IFCA district bylaws. http://
cornwall-ifca.gov.uk/sitedata/Byelaw_review/CIFCA_byelaws_A5_
bookletv1.pdf (last accessed 1 February 2015).
Crozier, L. G., and Hutchings, J. A. 2014. Plastic and evolutionary
responses to climate change in fish. Evolutionary Applications, 7:
68 – 87.
Currie, J. J., and Schneider, D. C. 2011. Spatial scaling from latitudinal
gradients: size-specific fecundity in the American lobster Homarus
americanus. Marine Ecology Progress Series, 439: 193– 201.
Dow, R. L. 1980. The clawed lobster fisheries. In The Biology and
Management of Lobsters—Volume II; Ecology and Management,
pp. 265 – 306. Ed. by J. S. Cobb, and B. F. Phillips. Academic Press,
London. 390 pp.
Ernsting, G., and Isaaks, A. 2000. Ectotherms, temperature, and tradeoffs: size and number of eggs in a Carabid beetle. The American
Naturalist, 155: 804– 813.
Estrella, B. T., and Cadrin, S. X. 1995. Fecundity of the American lobster
(Homarus americanus) in Massachusetts coastal waters. ICES
Marine Science Symposia, 199: 61– 72.
Fisheries and Aquaculture Department (FAO). 2014. Global capture
production statistics. www.fao.org (last accessed 4 February 2015).
Free, E. K. 1994. Reproductive processes in the European lobster,
Homarus gammarus. PhD thesis, University of Southampton,
England. 447 pp.
Free, E. K., Tyler, P. A., and Addison, J. T. 1992. Lobster (Homarus gammarus) fecundity and maturity in England and Wales. ICES
Document CM 1992/K: 43.
Gardner, C., Frusher, S. D., Barrett, N. S., Haddon, M., and Buxton,
C. D. 2006. Spatial variation in size at onset of maturity of female
southern rock lobster Jasus edwardsii around Tasmania, Australia.
Scientia Marina, 70: 423 – 430.
Gendron, L., and Ouellet, P. 2009. Egg development trajectories of early
and late-spawner lobsters (Homarus americanus) in the Magdalen
Islands, Québec. Journal of Crustacean Biology, 29: 356 – 363.
Gomez, J., Barboza, F. R., and Defeo, O. 2013. Environmental drivers defining linkages among life-history traits: mechanistic insights from a
semi-terrestrial amphipod subjected to macro-scale gradients.
Ecology and Evolution, 3: 3918– 3924.
Green, B. S., Gardner, C., Hochmuth, J. D., and Linnane, A. 2014.
Environmental effects on fished lobsters and crabs. Reviews in Fish
Biology and Fisheries, 24: 613– 638.
Hansen, O. J., Puvanendran, V., and Mortensen, A. 2012. Importance of
broodstock holding temperature on fecundity and egg quality in
three groups of photo-manipulated Atlantic cod broodstock.
Aquaculture Research, 44: 140– 150.
i99
Hartnoll, R. G. 2001. Growth in Crustacea—twenty years on.
Hydrobiologia, 449: 111– 122.
Hepper, B. T., and Gough, C. J. 1978. Fecundity and rate of embryonic
development of the lobster, Homarus gammarus (L.), off the coast of
North Wales. Journal du Conseil International pour l’Exploration de
la Mer, 38: 54– 57.
ICES Data Centre. 2014. ICES Dataset on Ocean Hydrography.
International Council for the Exploration of the Sea, Copenhagen.
Lambert, Y. 2008. Why should we closely monitor fecundity in marine
fish populations? Journal of Northwest Atlantic Fishery Science,
41: 93 – 106. doi:10.2960/J.v41.m628.
Landers, D. F., Keser, M., and Saila, S. B. 2001. Changes in female lobster
(Homarus americanus) size at maturity and implications for the
lobster resource in Long Island Sound, Connecticut. Marine and
Freshwater Research, 52: 1283– 1290.
Latrouite, D., Morizur, Y., and Raguenés, G. 1984. Fecondites individuelle et par recrue du homard european Homarus gammarus (L.)
des cotes françaises. ICES Document CM 1984/K: 38.
Laurans, M., Fifas, S., Demaneche, S., Brerette, S., and Debec, O. 2009.
Modelling seasonal and annual variation in size at functional maturity in the European lobster (Homarus gammarus) from self-sampling
data. ICES Journal of Marine Science, 66: 1892– 1898.
Little, S. A., and Watson, W. H. 2003. Size at maturity of female
American lobsters from an estuarine and coastal population.
Journal of Shellfish Research, 22: 857 – 863.
Little, S. A., and Watson, W. H. 2005. Differences in the size at maturity
of female American lobsters, Homarus americanus, captured
throughout the range of the offshore fishery. Journal of Crustacean
Biology, 25: 585 – 592.
Lizarraga-Cubedo, H. A., Tuck, I., Bailey, N., Pierce, G. J., and Kinnear,
J. A. M. 2003. Comparisons of size at maturity and fecundity of two
Scottish populations of the European lobster, Homarus gammarus.
Fisheries Research, 65: 137 – 152.
MacCormack, T. J., and DeMont, M. E. 2003. Regional differences in
allometric growth in Atlantic Canadian lobster (Homarus americanus). Journal of Crustacean Biology, 23: 258 – 264.
Moland, E., Moland Olsen, E., and Stenseth, N. C. 2010. Maternal influences on offspring size variation and viability in wild European
lobster Homarus gammarus. Marine Ecology Progress Series, 400:
165– 173.
Mollet, F. M., Engelhard, G. H., Vainikka, A., Laugen, A. T., Rijnsdorp,
A. D., and Ernande, B. 2013. Spatial variation in growth, maturation
schedules and reproductive investment of female sole Solea solea in
the Northeast Atlantic. Journal of Sea Research, 84: 109 – 121.
Morgan, M. J. 2008. Integrating reproductive biology into scientific
advice for fisheries management. Journal of Northwest Atlantic
Fishery Science, 41: 37 –51. doi:10.2960/J.v41.m615
Neuheimer, A. B., and Taggart, C. T. 2007. The growing degree-day and
fish size-at-age: the overlooked metric. Canadian Journal of Fisheries
and Aquatic Sciences, 64: 375 –385.
OceanColor. 2014. NASA Goddard Space Flight Centre. Greenbelt, MD,
USA. http://oceancolor.gsfc.nasa.gov (last accessed 24 January
2014.
PO.DAAC. 2014. Physical Oceanography Distributed Active Archive
Center. NASA Jet Propulsion Laboratory, Pasadena, CA, USA. http://
podaac.jpl.nasa.gov (last accessed 24 January 2014).
Quinn, B. K., Rochette, R., Ouellet, P., and Sainte-Marie, B. 2013. Effect
of temperature on development rate of larvae from cold-water
American lobster (Homarus americanus). Journal of Crustacean
Biology, 33: 527 – 536.
R Core Team. 2012. R: a Language and Environment for Statistical
Computing. R Foundation for Statistical Computing, Vienna,
Austria. ISBN 3-900051-07-0. http://www.R-project.org/.
Rijnsdorp, A. D., Grift, R. E., and Kraak, S. B. M. 2005. Fisheries-induced
adaptive change in reproductive investment in North Sea plaice
(Pleuronectes platessa)? Canadian Journal of Fisheries and Aquatic
Sciences, 62: 833– 843.
i100
Roberts, E. A. 1992. Fecundity of the European lobster (Homarus gammarus L.) from the central south coast of England. MSc thesis,
University of Southampton, England. 53 pp.
Schmalenbach, I., and Franke, H-D. 2010. Potential impact of climate
warming on the recruitment of an economically and ecologically important species, the European lobster (Homarus gammarus) at
Helgoland, North Sea. Marine Biology, 157: 1127– 1135.
Stares, J. C., Rideout, R. M., Morgan, M. J., and Brattey, J. 2007. Did
population collapse influence individual fecundity of Northwest
Atlantic cod? ICES Journal of Marine Science, 64: 1338– 1347.
Thorsen, A., Witthames, P. R., Marteinsdottir, G., Nash, R. D. M., and
Kjesbu, O. S. 2010. Fecundity and growth of Atlantic cod (Gadus
morhua L.) along a latitudinal gradient. Fisheries Research, 104:
45 – 55.
Tobin, D., and Wright, P. J. 2011. Temperature effects on female maturation in a temperate marine fish. Journal of Experimental Marine
Biology and Ecology, 403: 9 – 13.
Tully, O., Roantree, V., and Robinson, M. 2001. Maturity, fecundity and
reproductive potential of the European lobster (Homarus gammarus) in Ireland. Journal of the Marine Biological Association of
the United Kingdom, 81: 61– 68.
Wahle, R. A., Castro, K. M., Tully, O., and Cobb, J. S. 2013. Homarus. In
Lobsters: Biology, Management, Aquaculture and Fisheries, 2nd
edn, pp. 2212258. Ed. by B. Phillips. John Wiley & Sons,
Hoboken. 288 pp.
C. D. Ellis et al.
Witthames, P. R., Greer Walker, M., Dinis, M. T., and Whiting, C. L.
1995. The geographical variation in the potential annual fecundity
of dover sole, Solea solea, from European shelf waters during 1991.
Netherlands Journal of Sea Research, 34: 45– 58.
Wright, P. J. 2013. Methodological challenges to examining the causes of
variation in stock reproductive potential. Fisheries Research, 138:
14 – 22.
Wright, P. J., Gibb, F. M., Gibb, I. M., and Millar, C. P. 2011b.
Reproductive investment in the North Sea haddock: temporal and
spatial variation. Marine Ecology Progress Series, 432: 149 – 160.
Wright, P. J., Millar, C. P., and Gibb, F. M. 2011a. Intra-stock differences
in maturation schedules of Atlantic cod, Gadus morhua. ICES
Journal of Marine Science, 68: 1918– 1927.
Wu, R. S. S., Zhou, B. S., Randall, D. J., Woo, N. Y. S., and Lam, P. K. S.
2003. Aquatic hypoxia is an endocrine disruptor and impairs fish reproduction. Environmental Science and Technology, 37:
1137– 1141.
Yoneda, M., and Wright, P. J. 2004. Temporal and spatial variation in reproductive investment of Atlantic cod Gadus morhua in the northern
North Sea and Scottish westcoast. Marine Ecology Progress Series,
276: 237– 248.
Zuo, W., Moses, M. E., West, G. B., Hou, C., and Brown, J. H. 2011. A
general model for effects of temperature on ectotherm ontogenetic
growth and development. Proceedings of the Royal Society B:
Biological Sciences, doi: 10.1098/rspb.2011.2000.
Handling editor: Emory Anderson
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i101 –i114. doi:10.1093/icesjms/fsu227
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Mate choice in temperate and tropical spiny lobsters
with contrasting reproductive systems
Mark Butler, IV 1*, Rodney Bertelsen 2, and Alison MacDiarmid 3
1
Department of Biological Sciences, Old Dominion University, Norfolk, VA 23529, USA
Florida Fish and Wildlife Conservation Commission, Florida Marine Research Institute, South Florida Regional Laboratory, 2796 Overseas Highway,
Suite 119, Marathon, FL 33050, USA
3
National Institute of Water and Atmospheric Research, PO Box 14-901, Wellington, New Zealand
2
*Corresponding author: tel: +1 757 683 3609; fax: +1 757 683 5283; e-mail: [email protected]
Butler, M.IV, Bertelsen, R., and MacDiarmid, A. Mate choice in temperate and tropical spiny lobsters with contrasting reproductive
systems. – ICES Journal of Marine Science, 72: i101– i114.
Received 23 July 2014; revised 11 November 2014; accepted 19 November 2014; advance access publication 22 December 2014.
Sperm limitation of reproductive success is common in decapod crustaceans, favouring mating systems in which females compete for large males of
high reproductive value. We investigated these phenomena in two species of spiny lobsters—one temperate, one tropical—with contrasting reproductive systems: the Southern Rock Lobster (Jasus edwardsii) and the Caribbean Spiny Lobster (Panulirus argus). We hypothesized that female
mate selection should be more pronounced in the temperate J. edwardsii than in the tropical P. argus because J. edwardsii matures later, has a shorter
mating season, and produces just one clutch of eggs per year that benefit from significant maternal investment of resources. As hypothesized,
experiments conducted in large mesocosms revealed that female J. edwardsii cohabited with large males more often than expected by chance
during their receptive period, but not at other times. Large male J. edwardsii cohabited in dens with the largest unmated females, whereas
small males exhibited no mate size preference. In contrast, the proportion of female and male P. argus that co-occupied dens with the opposite
sex was no more than expected by chance. Cohabitation patterns in the wild supported these laboratory findings for both species. Our results
demonstrate the tight connection between contrasting reproductive strategies and the specificity of mate choice in spiny lobsters that are consistent with predictions based on environmental seasonality in temperate vs. tropical ecosystems.
Keywords: Jasus, mate choice, Panulirus, reproduction, spiny lobsters.
Introduction
Latitudinal gradients in animal life history characteristics, such as
the trade-off in maternal investment in egg mass vs. numbers of
eggs (i.e. clutch size), are deeply rooted in evolutionary biology
theory (MacArthur and Wilson, 1967; Pianka, 1970; Reznick et al.,
2002). These differences are paraphrased by the classic r-selected
vs. K-selected life history dichotomy: species that evolved at lower
latitudes display r-selected life history attributes (e.g. many offspring at low “cost” to the parent, thus: high fecundity, early maturity, low parental investment, etc.) and those that evolved at higher
latitudes tend to be K-selected (e.g. fewer and better provisioned offspring: lower fecundity, late maturity, large parental investment,
etc.). Empirical evidence from a variety of taxa, including marine
fish (Vila-Gispert et al., 2002; Foster and Vincent, 2004) and
# International
crustaceans (Anger et al., 2002; Sarma et al., 2005), is consistent
with this theory. Latitudinal differences in life history theory are
also relevant to applied fields such as fishery management, that
to be effective must adopt different management regulations
where latitudinal gradients in life history exist (Fromentin and
Fonteneau, 2001).
Spiny lobsters (Decapoda; Palinuridae) are good models for exploring the implications of latitudinal differences in the evolution of
reproductive life history characteristics such as mate choice and
competition and the potential effects of, or implications for,
human exploitation of those populations. Spiny lobsters have a
global distribution, their patterns of reproduction vary latitudinally,
their reproductive systems are finely tuned with very low sperm:egg
ratios, and their size-dependent breeding systems are sensitive to
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i102
changes caused by overfishing (Lipcius et al., 1983; Lipcius, 1985;
DeMartini et al., 1993; MacDiarmid and Butler, 1999; Kelly et al.,
2000; Melville-Smith and de Lestang, 2006; Butler et al., 2011).
We investigated latitudinal effects on mating systems in two
species of spiny lobsters with contrasting reproductive systems: a
temperate species—the Southern Rock Lobster (Jasus edwardsii)
from New Zealand and south Australia; and a tropical species—
the Caribbean Spiny Lobster (Panulirus argus).
We hypothesized that mate selection should be stronger in
J. edwardsii than in P. argus for at least two reasons: J. edwardsii
has fewer lifetime mating opportunities and the consequences of a
missed mating opportunity are severe. Female J. edwardsii breed
only once per year after a pre-mating molt and during a 3 –5
month autumn mating season repeated over an average of 7 years
(Pollock, 1991). So a single mating opportunity represents 5 –
15% of the average lifetime matings for J. edwardsii. In contrast,
female P. argus mate and spawn up to three times per year depending
on their size (Lyons et al., 1981; Maxwell et al., 2009) and do so over a
prolonged breeding season (6 –12 months depending on latitude)
and for anywhere from 5 to 30 years (Pollock, 1997; Bertelesen
and Mathews, 2001; Ehrhardt, 2008; Maxwell et al., 2009). So a
single mating constitutes as little as 1.5% of all matings by a longlived female P. argus. Moreover, the fecundity of female J. edwardsii
drops 10% each day that mating is delayed once a female becomes
receptive and, if no mate is found, absorption of the egg mass
scars the ovaries resulting in up to a 40% drop in fecundity the following mating season (MacDiarmid and Sainte-Marie, 2006). In
contrast, if female P. argus do not find a mate, they simply extrude
the unused egg mass (Butler et al., 2011). In short, the reproductive
cost of not acquiring a mate for female J. edwardsii (the temperate
species) is far greater than that for P. argus.
The same temperate–tropical differences that give rise to disparate female reproductive patterns and mate selection also influence
the choice of mates by male lobsters. Large male spiny lobster
should preferentially choose and compete for large competent
females with whom to mate, given their greater egg production,
and this preference should be more strongly exhibited by male
J. edwardsii. In both J. edwardsii and P. argus, female fecundity is
size dependent with the largest females producing 9 –19 times
the number of eggs per clutch of the smallest mature females
(MacDiarmid and Sainte-Marie, 2006). However, the short breeding season of J. edwardsii in temperate climates and limited
mating opportunities with females that mate only once per year
should result in stronger male–male competition for mates in
J. edwardsii than for P. argus.
Yet, these highly evolved differences in the reproductive biology
of tropical and temperate species may be similarly affected by overfishing. In intact populations within MPAs where large males are still
present, spiny lobsters exhibit lek-style mating systems in which
large males defend specific dens or groups of dens from other
males and females travel among these males and their territories
(MacDiarmid, 1994; Bertelsen and Cox, 2001; Bertelsen and
Matthews, 2001; Robertson and Butler, 2012). However, this lekstyle mating system breaks down in heavily fished spiny lobster
populations where large lobsters, especially large males, are absent
and smaller individuals tend not to establish or defend mating territories (MacDiarmid, 1994; Bertelsen and Cox, 2001; Bertelsen and
Matthews, 2001).
Here, we describe how we first compared the reproductive
biology of these two species of spiny lobster whose life histories
are indicative of temperate and tropical species. We then conducted
M. Butler et al.
a series of female and male mate choice experiments in large,
outdoor mesocosms and compared those findings with field observations of mating and den co-occupancy. As predicted by life history
theory, we hypothesized that (i) preference for larger mates would
be strongest by females, larger individuals, and in J. edwardsii compared with P. argus, and (ii) male– male competition would also be
most intense among large males, particularly J. edwardsii.
Methods
Southern rock lobster mesocosm experiments
Mature male and pre-molt female J. edwardsii were obtained from
the Chatham Islands and near Wellington, New Zealand, and transferred to a flow-through seawater system (Figure 1). Only J. edwardsii of a size whereby all are certain to be mature (.90 mm CL) were
included in our experiments. Lobsters were measured (carapace
length; CL), individually marked with colour-coded antennae
tags, and distributed among six 1.8 m diameter × 0.6 m deep concrete holding tanks. Sexes were held separately until the start of the
experiments. The lobsters in each holding tank had continuous
access to live blue mussels (Mytilus galloprovincialis) for food and
were shaded from sunlight.
Experiment 1: mate choice by female J. edwardsii
The choice of a mate by small (99 –108 mm CL) and large (127 –
153 mm CL) mature female J. edwardsii were tested at four different
phases of the mating cycle: pre-molt, early post-molt, late post-molt,
and egg-bearing. The designation of reproductive stages that we
tested for J. edwardsii are related to a specific pre-mating molt that
females undergo each year before the breeding season. To define
these stages for each female, we calculated the expected day of
mating from a function incorporating molt-mate interval, day of
molting, and female size (MacDiarmid, 1989b). We then used this
information to subdivide the period between molting and mating
into two phases: early post-molt and late post-molt.
The experiment was conducted in two large outdoor tanks measuring 7 × 5 and 2 m deep (75 000 l) and covered by a plastic mesh
screen that reduced the ambient light level by 75%. In each corner, a
shelter was constructed from two hollow concrete building blocks
supporting a 500 × 500 mm concrete paving slab. Shelters were
large enough to simultaneously house a large male and a large
female. One shelter was left empty, but in the other three shelters
we tethered either a large mature male (180 mm CL), a small
mature male (100 mm CL), or a female of the same size and reproductive stage as the test female. Lobsters were tethered to their
shelter by a 0.5 m length of 40 kg test-strength monofilament
fishing line. The line was attached to the lobster via a swivel on a
plastic cable tie fastened around the cephalothorax between the
second and third walking legs. This fit snugly between the legs and
allowed normal locomotor activity. Tethering is more commonly
used as a means to compare relative rates of predation on marine
animals under differing conditions, and its utility and problems have
been thoroughly debated (Peterson and Black, 1994; Aronson and
Heck, 1995; Aronson et al., 2001). We used it simply to constrain
an individual lobster near a particular den. Perhaps doing so may
result in some behavioural artefact, but we know of no method to
constrain a specific sized lobster within a specific shelter that
would have been less obtrusive. The length of the tether and lack
of any obstructions in each den minimized the potential for tangling
of the tether that might alter lobster behaviour. Otherwise, lobsters
Spiny lobster mate choice
Figure 1. Field study sites in northeast New Zealand and Florida, USA.
i103
i104
appeared to behave normally and remained sheltered within each
den during the day.
The test female was liberated in the centre of the test tank and her
position in the tank recorded 24 h later, after which she was
removed. To eliminate any bias in shelter choice due to the
uneven distribution of natural light penetration or water circulation, the positions of the shelters and tethered lobsters were
rotated around the tanks among replicates. Both experimental
tanks ran simultaneously until 20 or more females had been tested
for each combination of size and molt stage. The choice of the test
females among the different shelters was analysed using a log-linear
goodness-of-fit test with an even distribution of females among
shelters as the expected outcome if shelter choice was random.
Experiment 2: mate choice by male J. edwardsii
The same tanks, shelters, tethering arrangements, and experimental
protocols were used as in Experiment 1. However, in this case,
females of four different sizes (,99, 100 –119, 120–139, and
.140 mm CL) were tethered singly in each shelter. Two sets of replicates were run with either pre-molt or post-molt unmated females
tethered to shelters. In both cases, a single small (99 –112 mm CL)
or large (178–185 mm CL) mature male was liberated in the
centre of the test tank and its position in the tank recorded 24 h
later. Both experimental tanks ran simultaneously until 20 or
more males of each size had been tested for the two female molt
phases. The choice of the test males among the different shelters
was tested using log-linear goodness-of-fit tests with an even distribution of males among shelters as the expected outcome if shelter
choice was random.
Caribbean spiny lobster mesocosm experiments
Mature male and female P. argus were obtained from reef environments in the Florida Keys and the Dry Tortugus National
Sanctuary (USA) by divers (Figure 1). Lobsters were transported
to the Keys Marine Laboratory (Long Key, FL, USA) where they
were measured (CL), individually marked with colour-coded antennae tags, and transferred to four outdoor flow-through holding
tanks (2 m diameter × 1 m depth). Sexes were held separately
until the start of the experiments. The lobsters in each holding
tank were provided shelter and were fed frozen shrimp and squid
daily. Only P. argus of a size whereby all are certain to be mature
(.80 mm CL) were included in our experiments.
Experiment 3: mate choice by female P. argus
This experiment was run outdoor in two large (15 m long by 7 m
wide by 1.5 m deep), oblong concrete channels supplied with flowthrough seawater. Four shelters as described in Experiment 1 were
placed equidistant around the perimeter of each channel. Four alternative shelter conditions were established for this experiment. To
three of the shelters, we tethered either a large mature male
(.120 mm CL), a small mature male (,100 mm CL), or a
female the same size and reproductive condition as the test
female. The fourth shelter was left empty. Lobsters were tethered
as described for Experiment 1. Each morning, a test female was liberated in the centre of the test channel and her position in the tank
recorded 24 h later after which she was removed. We first tested the
shelter (i.e. mate) choice of unmated small (,90 mm CL) and large
(.100 mm CL) females during the reproductive season, and to
contrast these patterns with shelter choice at other times of year,
we repeated the test during the non-reproductive season (n ¼
19 –63 lobsters per experimental condition; mean ¼ 37). Female
M. Butler et al.
choice of the available den partners was analysed using log-linear
goodness-of-fit tests. Four separate analyses were run comparing
mate choice by small and large females during the reproductive
season and during the non-reproductive season; an even distribution of females among the four shelter conditions was considered
the expected outcome if shelter choice was random.
Experiment 4: mate choice by male P. argus
To investigate mate choice by male P. argus, we used a similar experimental protocol as that described in Experiment 3. However, in this
case, a single small (80 –100 mm CL) or large (.120 mm CL) male
lobster (as opposed to a female lobster as in Experiment 3) was
released in each mesocosm to choose among four different shelter
conditions: a small female (,90 mm CL) tethered in a den, a
large female (.100 mm CL) tethered in a den, an equivalent size
male tethered to a den, or an empty den. We ran these trials for
small (n ¼ 26) and large (n ¼ 36) males during the reproductive
season and, as for the female tests described above, repeated the
test during the non-reproductive season (n ¼ 12 –57 for small
and large males, respectively). Again, log-linear goodness-of-fit
tests were used to assess non-random shelter associations for small
and large males during the reproductive and non-reproductive
periods.
Field studies
We examined patterns of male and female dispersion and courting
activity in wild populations of J. edwardsii and P. argus in unfished
marine reserves and nearby fished areas. For J. edwardsii, we studied
four adjacent populations—two protected (Leigh Marine Reserve
and Tawharanui Marine Park) and two fished (eastern side of
Kawau Island and northwest Hen Island)—in northeast New
Zealand (Figure 1). At each of these localities, the abundance, size
frequency, sex ratio, and courting activity of lobsters was estimated
during the peak of mating in June 1995. At each locality, divers
observed lobsters during daylight hours at four sites, two shallow
(1–10 m) and two deep (11– 20 m), situated within areas of
crevice and boulder habitat in which J. edwardsii are primarily
found. At each site, we searched for lobsters within five haphazardly
placed 50 × 10 m transects. Lobsters were counted, their sex determined, and their size estimated using established visual techniques
(see MacDiarmid, 1989a, 1991). Lobsters in each den were recorded
separately. Courting activity was defined by lobsters engaged in
“frontal approach” behaviour, which is unique to pre-copulatory
courtship in spiny lobsters (Lipcius et al., 1983; Lipcius and
Herrnkind, 1985; MacDiarmid and Kittaka, 2000). Before the
surveys, 2 days were spent training divers to reliably estimate CL
to within +5 mm over the whole size range (20 –200 mm CL).
The same three divers were used throughout the survey.
Fished populations of P. argus in four areas of the Florida Keys
(Carysfort Reef in the northeast to Marquesas Rocks in the west)
and a protected population in the Dry Tortugus National Marine
Sanctuary (USA) were also studied in situ by divers (Figure 1).
Every 6 weeks from March to September in 1996 and 1997, two backreef (landward side of reef crest, 1 –6 m), two fore-reef (seaward of
the reef crest, 1 – 10 m), and two deep-reef (.15 m) sites were
chosen for study within each of the regions in the Florida Keys
and at the Dry Tortugus. At each site, divers searched for lobsters
for two 30 min periods; the time needed to capture each lobster
with a hand net was not included in the search period. All lobsters
captured from each den were held in separate mesh bags for later
processing aboard a research vessel. If a lobster evaded capture,
i105
Spiny lobster mate choice
we noted its sex (if known) and estimated its CL. We also noted
whether lobsters were dwelling alone or cohabiting with other individuals in a den. For each captured lobster, we determined its CL and
sex. The presence and condition (i.e. fresh, used, single, multiple) of
spermatophores on females was also noted. Lobsters were released
unharmed after processing.
To determine whether male P. argus patterns of cohabitation
were influenced by male–male competition, we used computer
simulations to determine if the difference in size of the two largest
males in dens was significantly different from that produced when
we randomly resampled the data to repopulate the dens. The data
were blocked by site (Tortugas, Upper Keys, Lower Keys), year,
and month; the number of dens, and den occupants were held
constant in the simulations. Dens were then repopulated using the
lobsters captured within the site–month –year blocks and the
difference in size between the largest and second largest males in
each den was calculated. These results were then compared with
the original field data. If there was little difference between the randomized and observed datasets, we concluded that observed pattern
was due to other factors (e.g. population structure and den availability) other than interactions among males.
Results
Southern rock lobster mesocosm experiments
Experiment 1: mate choice by female J. edwardsii
Before molting, neither large nor small mature female J. edwardsii
showed any particular preference when given a choice of four shelters that were either empty or contained a single tethered mature
female, small mature male, or large mature male (Figure 2).
However, soon after molting the majority of the large (72%), and
many small (46%) females chose to cohabit with a large male
(G ¼ 31.3, d.f. ¼ 3, p , 0.001and G ¼ 17.4, d.f. ¼ 3, p , 0.001,
respectively; Figure 2). This preference by large and small females
for shelters containing the largest male increased to 78 and 71%, respectively, during the 25 –35 days between molting and mating
(Figure 2). After mating had taken place and females were brooding
eggs, both large and small females showed a tendency to cohabit with
other females, though only in larger females was this statistically
significant (G ¼ 15.6, d.f. ¼ 3, p ¼ 0.0014; Figure 2). Never did
females prefer to shelter in any particular corner of the test tanks
(i.e. there was no tank position bias).
Experiment 2: mate choice by male J. edwardsii
Large mature males showed no tendency to shelter with pre-molt
females of any particular size (G ¼ 0.40, d.f. ¼ 3, p ¼ 0.940;
Figure 3). However, they showed an increasing preference for cohabitation with large, unmated post-molt females. This preference
rose steadily from 12% cohabiting with the smallest females to
42% cohabiting with the largest females (Figure 3). When all four
female size classes were included in the analysis, the trend was nonsignificant (G ¼ 5.35, d.f. ¼ 3, p ¼ 0.148). However, pooling of the
results for the two largest and two smallest female size classes indicated there was a significant tendency for large males to shelter with
females larger than 120 mm CL (G ¼ 3.95, d.f. ¼ 1, p ¼ 0.0470).
Small mature males showed no tendency to shelter with any size
of pre-molt (G ¼ 4.50, d.f. ¼ 3, p ¼ 0.210) or post-molt (G ¼
1.31, d.f. ¼ 3, p ¼ 0.724) mature female (Figure 3). Never did
males prefer to shelter in any particular corner of the test tanks.
Caribbean spiny lobster mesocosm experiments
Experiment 3: mate choice by female P. argus
In the non-mating season, large female P. argus preferentially sheltered in empty shelters or with other females, whereas small
females showed no shelter preference (Figure 4). During the reproductive season, the proportion of both large and small mature
females that denned with mature males was significantly higher
(51–52%; Figure 4) than in the non-mating season (19 –40%;
large females: G ¼ 20.05, d.f. ¼ 3, p , 0.001; small females: G ¼
16.63, d.f. ¼ 3, p , 0.001). However, the distribution of females
among the four den types during the reproductive season was no different from expected by chance alone (Figure 4). These results
suggest that female P. argus actively avoid male-occupied dens
most of the year, but this relaxes during the reproductive season
when females freely associate with males with no particular preference for male size.
Experiment 4: mate choice by male P. argus
In the non-mating season, the greatest proportion of large mature
male P. argus denned alone, although this was not statistically significant (Figure 5); small males also showed no den preference
(Figure 5). During the reproductive season, the proportion of
small and large mature male P. argus that denned with mature
females was significantly higher (51 –55%) than at other times of
year (33 –34%; large males: G ¼ 10.94, d.f. ¼ 3, p ¼ 0.012; small
males: G ¼ 30.61, d.f. ¼ 3, p , ,0.001). The distribution of
males among the four den types during the breeding season,
however, was no different from expected by chance alone
(Figure 5). These results suggest that male P. argus, especially large
individuals, are solitary in dens most of the year but freely associate
with females during the reproductive season with no clear preference for females of any size.
Field studies
Jasus edwardsii
We observed 44 courting male and female J. edwardsii at the four
field sites in New Zealand, although most (91%) of these were
observed in the two MPAs. Most males (95%) were larger than the
females they were courting. In only two cases was the male the
same size as the female and never at any locality was the male
smaller than the female he courted. There was a marginally significant, positive relationship between the size of courting pairs in the
2
Leigh Marine Reserve (radj
= 0.086, F1,37 ¼ 4.46, p ¼ 0.042) but
no relationship among the size of courting pairs within the
2
Tawharanui Marine Park (radj
= 0.055, F1,15 ¼ 1.87, p ¼ 0.192).
At both localities, the dispersion of courting among male size
classes was significantly different from expected by the relative
abundance of mature male size classes (Leigh x 2 ¼ 69.15, d.f. ¼
11, p ,,0.001; Tawharanui x 2 ¼ 72.48, d.f. ¼ 10, p ,,0.001).
In both marine reserves, courting was exclusively undertaken by
large mature males: courting males were all ≥130 mm CL at the
Leigh Marine Reserve and ≥120 mm CL at the Tawharanui
Marine Park (Figure 6). Although smaller (90–120 mm CL) individuals comprised 42 and 50% of the mature male populations at
Leigh and Tawharanui, respectively, none was observed courting.
Too few courting pairs were observed at fished localities to determine the mating success of different sized males.
The dispersion of mature post-molt, but as yet still unmated,
female J. edwardsii among dens containing a large male differed
with female size on reefs within the Leigh Marine Reserve (Table 1).
i106
M. Butler et al.
Figure 2. Jasus edwardsii. Patterns of cohabitation of paired (a) small and (b) large mature females. Expected frequencies are based on independent
female choice from Experiment 1. Random frequencies are those that would occur if females neither choose large males nor competed for them.
Spiny lobster mate choice
i107
Figure 3. Jasus edwardsii. Shelter choices by (a and b) large mature males and (c and d) small mature males when dens contained one of four size
classes of tethered pre-molt or post-molt mature females.
The proportion of small (,120 mm CL), post-molt females sheltering alone, or cohabiting in groups with a large male was no different
from that expected under a random model of dispersion (x 2 ¼ 0.609,
d.f. ¼ 3, p ¼ 0.894). Large post-molt females, in comparison, were
significantly dispersed among more males (i.e. fewer male dens had
no females), were much more frequently alone with large males,
and occurred less often in groups with a large male than expected
(x 2 ¼ 18.45, d.f. ¼ 4, p ¼ 0.001).
Panulirus argus
We observed no courtship among P. argus in the field, probably
because most courting activity in this species takes place during
crepuscular periods (Lipcius and Herrnkind, 1985) and at night
(Bertelsen and Horn, 2000), whereas our fieldwork was conducted
during the day. Instead, as a proxy, we used the incidence of
females with new spermatophores cohabiting with a single mature
male to indicate the likely size of mating pairs. Although often
males were larger than the females, in some instances, they were
smaller, especially at the Dry Tortugas. There was no relationship
among the size of females with new spermatophores or the male coha2
= 0.083, F1,19 ¼ 2.86,
biting with them at either the Florida Keys (radj
2
= −0.068, F1,6 ¼
p ¼ 0.110) or the Dry Tortugus Sanctuary (radj
0.552, p ¼ 0.485). There was no evidence of females bearing multiple
spermatophores.
The frequency of cohabitation of male P. argus with like-sized
males reflects the intensity of male– male competition for females
and it varies with male size. At the Dry Tortugus Sanctuary,
mating activity is most intense in February and March and peaks
again in June when some females mate again to fertilize a second
clutch (Bertelsen and Cox, 2001). Large males (.101 mm CL)
were never observed together in dens during the first intense
phase of mating, but male–male cohabitation briefly increased in
May before dropping again in June/July (Figure 7a–c) when
some females mate again. A similar pattern was observed for
smaller mature males except that they began cohabitating earlier
in the year.
In the fished Florida Keys population, males .101 mm CL were
too rarely observed to determine patterns of cohabitation. There the
reproductive season starts in late March/April and lasts until
September without the pronounced peaks in mating activity evident
at the Dry Tortugus (Bertelsen and Cox, 2001). This is also reflected
in the patterns of mature male cohabitation, which was less pronounced than at the Dry Tortugas Sanctuary (Figure 7d and e).
Competition among male P. argus of different sizes is also apparent at the Dry Tortugus Sanctuary when the size of the two largest
males in a den is compared at different times of the year. This analysis excludes those dens in which males were solitary and is thus a
conservative measure of the degree to which large males will tolerate
the presence of other males in a den during the peak of mating in
i108
M. Butler et al.
Figure 4. Panulirus argus. Shelter choice by large and small mature females during the mating (top panels) and non-mating (bottom panels)
seasons.
February –March. At this time of year, cohabiting males differ by
60 mm in CL, which equates to a fivefold difference in body
weight. This difference in body size among cohabiting males
decreased sharply after the mating period. In the fished Florida
Keys population, large males were rare and the difference in size of
the two largest males in a den was lower than found at the Dry
Tortugus and did not show the same sharp peak.
To determine the extent that these patterns of male association
were influenced by the availability of different sized males in the
population, we randomly repopulated the dens in computer simulations. These results show that at the Dry Tortugus Sanctuary,
the observed difference in the size of cohabiting males early in the
year, especially if the den contained a large male, was much larger
than expected due to chance alone. In the fished Florida Keys population where large males were rare, the distribution of males among
dens was not different from that explained by random association.
Discussion
We documented differences in the mating systems of two species
of spiny lobster that are consistent with, and presumably linked
to, the evolution of drastically different reproductive characteristics
representative of temperate and tropical species. The results of our
mesocosm experiments and field observations revealed that mate selection in the temperate species (J. edwardsii) is more precise than
in its tropical dwelling counterpart (P. argus), especially among
larger individuals. Large female J. edwardsii preferentially cohabited
with large males during the mating season and vice versa (Figures 2
and 3). This contrasts with the behaviour of both sexes outside of the
mating season (e.g. before the female pre-mating molt and once eggbearing) when they exhibited no preference for cohabitation with
large individuals of the opposite sex. In contrast, female and male
P. argus did not discriminate among mating partners of different
sizes, although females cohabited with males more frequently
during the mating season (Figures 4 and 5). These species-specific
differences in mate size preferences revealed in mesocosm mate
choice experiments were also borne out in field observations of
mate choice (Figures 6 and 7). Our findings are in keeping with evolutionary theory because in J. edwardsii, the variance in mate quality is
higher and the cost of poor mate choice greater, especially for larger
females. More specifically, there are greater rewards for J. edwardsii
than for P. argus to mate with a larger partner who can provide more
and larger eggs resulting in more robust larvae, or more sperm for
males (Annala and Bycroft, 1987; MacDiarmid and Butler, 1999;
Butler et al., 2015). The consequences of not locating a suitable mate
are also dire for J. edwardsii females, because unmated females incur
long-term ovarian damage and thus a significant reduction in annual
and lifetime egg production (Pollock, 1991; MacDiarmid and Butler,
1999; MacDiarmid et al., 1999).
Spiny lobster mate choice
i109
Figure 5. Panulirus argus. Shelter choice by large and small mature males during the mating (top panels) and non-mating (bottom panels) seasons.
The differences that we observed in mate selectivity by females
were expected because of the differing reproductive strategies of
the lobsters. The number of egg clutches a female produces on
average each year and over her lifetime, and the consequences of
not finding a suitable mate, differ markedly between these species,
whose reproductive dynamics are representative of the two predominate evolutionary courses that the palinurids have taken.
For females of the temperate species, J. edwardsii, there is only
one annual opportunity to breed (MacDiarmid, 1989b) over an
average reproductive life of 7 years (Pollock, 1991). Therefore, on
average each mating constitutes 100% of a female’s annual
matings and 14% of her lifetime matings. In addition, because the
spermatophore is short-lived, mating must take place within a relatively narrow 1 –2 day window coinciding with peak egg fertility,
otherwise the proportion of the clutch able to be successfully
fertilized decreases substantially (MacDiarmid et al., 1999;
MacDiarmid and Kittaka, 2000). Unmated females do not extrude
unfertilized eggs, which typically for larger females (.125 mm
CL) results in severe damage to her ovaries that greatly diminishes
her reproductive output in future years (MacDiarmid et al.,
1999). Thus, there is strong selective pressure on female J. edwardsii
with ripening ovaries to first locate potential mates and then to
choose a male likely to have sufficient sperm to fertilize all of her
eggs in a single mating timed to coincide with her peak fertility.
The strong selective pressures that have presumably driven
the evolution of a precise system for selection of males by female
J. edwardsii are absent for the tropical species, P. argus. Female
P. argus become reproductive in 1.5 –2 years post-settlement and
remain reproductively active throughout their lifetime, which is
unknown but may approach 30 years. Each mating season for
P. argus is several months long and large females produce two
to three clutches a year (Bertelsen and Cox, 2001). In addition,
female P. argus can mate up to 28 days before extruding eggs,
which allows females time to locate a more appropriate male if the
first male is judged too small to provide enough sperm. Finally, if
no mate is located, then female P. argus simply release unfertilized
eggs with no long-term consequences on future matings.
Female mate choice is well described among crabs (Christy, 1987;
Kendall and Wolcott, 1999; Sainte-Marie et al., 1999) and in the
clawed lobsters, Homarus americanus (Atema, 1986; Gosselin
et al., 2003) and Homarus gammarus (Debuse et al., 1999, 2003),
but in few other decapod crustaceans. However, it is likely to
evolve in any species where there is high variance in male quality
coupled with a method for females to distinguish among these
males. Laboratory experiments using female J. edwardsii suggest
that although a chemosensory mechanism is important, visual
and possibly tactile cues also play a role in mate choice (Raethke
et al., 2004).
An alternative explanation for the patterns of association between
post-molt female J. edwardsii and large males is that the males provide
a measure of protection during the vulnerable soft-shell phase after
molting, as occurs in H. americanus (Atema, 1986; Gosselin et al.,
i110
M. Butler et al.
2003). The evidence, however, is counter to this interpretation. First,
if the protection hypothesis is true, then smaller post-molt females
would be more likely to seek shelter with a male for protection after
Figure 6. Jasus edwardsii. The frequency of mature males in 10 mm size
classes and the proportion of each size class observed engaged in
courting activities at (a) Leigh Marine Reserve and (b) Tawharanui
Marine Park.
Table 1. Jasus edwardsii: expected and observed frequencies (%) of
cohabitation with large mature males for two size classes of late
post-molt unmated females in the Leigh Marine Reserve during the
peak of mating in June 1995. Frequencies expected based on
independent or random female association were calculated from a
Poisson model.
Female size class
Number of
females in male
den
0
1
2
3
4+
Mean # of females
per den
<120 mm CL
>120 mm CL
Expected
(%)
42.81
36.8
15.83
4.54
–
0.86
Expected
(%)
25.5
34.93
23.93
10.88
4.77
1.37
Observed
(%)
41.86
34.88
18.6
4.65
–
0.86
Observed
(%)
17.07
51.22
21.95
2.44
7.32
1.37
molting. Moreover, the association would decrease rapidly after
molting as the female’s new shell hardened. In fact, the patterns
were opposite, which suggests that the association between post-molt
females and large males is not associated with protection after
molting. Post-molt protection of females by males is also irrelevant
in P. argus as mature males and females molt well before the beginning
of the reproduction season (Lyons et al., 1981).
The preference by large male J. edwardsii to cohabit with large
females and the absence of any such relationship in P. argus suggests
that this is linked to the increase in egg quality with female size in
female J. edwardsii as has been found in other species (Kraak and
Bakker, 1998). Although there is no relationship between female
size and egg size in P. argus (Butler et al., 2015), large female P.
argus share with female J. edwardsii the desirable attribute of high
egg production. Still, we found no evidence that male P. argus preferentially choose to den with or mate mostly with larger females
(MacDiarmid and Butler, 1999). This makes sense when males
can match the size of the ejaculate with the size of the female and
thus the number of eggs available to fertilize. There is little advantage
to mate with a large female in these circumstances except the additional costs of having to locate and court several times with successive small females to fertilize the equivalent number of eggs that a
large female produces. These costs may be low if the mating
season is long and the probability of a male mating on any specific
day is low.
Theory suggests that when both males and females exercise mate
choice, it gives rise to assortative mating with high-quality females
mating with high-quality males and a decline in this correlation
over the course of the mating season (Johnstone, 1997). We found
little field evidence for size assortative pairing of courting J. edwardsii or P. argus; neither has it been documented in P. guttatus
(Robertson and Butler, 2012), which cohabits coral reefs with
P. argus in the Caribbean. However, two factors not yet included
in reproductive models may play an important role in determining
mating patterns. First, male –male competition when large males
are present prevents smaller mature males from participating in reproduction (MacDiarmid, 1989a), thereby limiting the size range of
males available for females to choose among. Second, the choice by
large male J. edwardsii for large females observed in our laboratory
experiments was much less intense than that of large females for
large males (43 vs. 78%) and small males exhibited no preference
at all. Thus, in J. edwardsii, the size association of males and postmolt females is dominated by the preference by all sizes of post-molt
females for large males. In contrast, male and female P. argus do not
appear to discriminate among size classes of the opposite sex during
the reproductive period.
In both species, competition among males is a major element in
the mating system. This is reflected in the frequency with which large
males segregate during the peak of the mating season. In J. edwardsii
as few as 5% of large males (.140 mm CL) cohabit in dens at
this time of year (MacDiarmid, 1994) and no large mature male
P. argus (.101 mm CL) den with other like-sized males at the
Dry Tortugus Sanctuary during the peak of mating. Smaller male
P. argus are therefore relegated to suboptimal habitats where reproductively active females are scarce. At Looe Key in the Florida Keys,
for example, most reproductive activity is on the fore reef where
female:male sex ratios are 4:1 during the reproductive season
(Hunt et al., 1991), whereas in nearby habitats, sex ratios are near
1:1 and reproductive activity is nil. Male–male competition in
areas where large males are common prevents small male J. edwardsii
from participating in reproduction (MacDiarmid, 1989a, this study).
Spiny lobster mate choice
i111
Figure 7. Panulirus argus. Cohabitation patterns of three sizes of mature males from (a – c) the Dry Tortugus Sanctuary and (d and e) the fished
Florida Keys. The number of males sampled is shown above each bar.
Larger males cohabit with many mature females (MacDiarmid,
1994), a greater proportion engage in courting activity, usually do
not move far from their den, and rarely shift among dens on consecutive days (MacDiarmid et al., 1991). Small mature male J. edwardsii, in
contrast, often share shelters (up to 60% at peak of mating;
MacDiarmid, 1994), cohabit with no or few mature females, are
excluded from courting activities, move greater distances at night
than large males, frequently shift shelters on consecutive days, and
have low levels of fidelity on a reef (MacDiarmid et al., 1991;
MacDiarmid, 1994; Kelly and MacDiarmid, 2003). However, some
smaller mature males are tolerated in the dens of large males.
The patterns of mate selection and competition in J. edwardsii
and P. argus that we found depend on variance in mate quality
and the cost of mating. These patterns are likely to be common
across a range of temperate and tropical decapods. In species—
typically temperate—that have only one brood per year (many
i112
crabs, all temperate clawed and spiny lobsters) or mate only once in
their lifetime and rely on stored sperm supplies thereafter (e.g. snow
crabs, blue crabs; Kendall and Wolcott, 1999; Rondeau and
Sainte-Marie, 2001; Sainte-Marie et al., 2002; Sainte-Marie, 2007),
the cost of mating with an inadequate male is high. Consequently,
the evolutionary pressure for females to develop mate selection
and perhaps competition strategies is likely to be intense. In contrast, in species—often tropical—that mate and brood many
times per year or can sequester sperm from many males, the cost
of mating with a single male with insufficient sperm is much
lower and the selection for mate choice and competition strategies
is also much reduced (Sato et al., 2005).
Many decapods are exploited commerciallyand males in particular
are sometimes heavily fished (Carver et al., 2005; Sato and Goshima,
2006, 2007; Sato et al., 2010). Even with similar levels of male and
female exploitation, male size is reduced disproportionately by
fishing because males grow much larger than females (MacDiarmid
and Sainte-Marie, 2006). For high-value commercially fished
species like spiny lobsters whose populations worldwide are almost
universally considered fully or overexploited, there are few situations
where natural mating dynamics remain intact (Rowe and Hutchings,
2003; Fenberg and Roy, 2008). In rare cases, there are spiny lobster
species whose natural abundance or size render them unacceptable
candidates for widespread commercial fishing (e.g. P. guttatus), providing the opportunity to study natural mating systems (Robertson
and Butler, 2009). Yet for most species of spiny lobster, only in large
and well enforced, unfished marine reserves do there exist populations
whose abundance, size composition, and sex ratio are unaltered by
humans and thus their mating dynamics a true reflection of the
species evolutionary past (Jack and Wing, 2010). Without those vestiges of unspoiled natural populations, our understanding of lobster
biology goes wanting and the consequent management of exploited
populations is flawed being based on targets whose foundation are
“natural baselines” that are unknown or shift (Heino et al., 2013).
For example, it is a common misconception among fishers and managers of P. argus that spawning stocks and mating activities occur in
deeper waters .20 m. This is true in heavily exploited lobster populations where the preponderance of large lobsters are relegated to deeper
waters where fishing intensity is less—but this is a consequence of
fishing, and not the natural system. In large, effectively enforced
MPAs like the Dry Tortugas nearly all of the mating activity occurs
in shallow water ,8 m where large, mature lobsters roam freely.
The extirpation of large male and female lobsters through overfishing
of spiny lobsters worldwide has often dramatically changed mating
systems that evolved based on size-specific male–female relationships.
The mating systems of overfished lobster populations now persist as a
scramble competition for mates among barely mature individuals that
mate at most once before capture by fishers. Overfishing is a strong selective force indeed that can greatly alter the reproductive biology of
species with strong consequences for population sustainability
(Coltman et al. 2003; Garcia et al., 2012; Kuparinen and Hutchings,
2012). Thus, studies like ours that reveal the true mating behaviour
of marine species not only offer insight into the factors shaping the
evolution and maintenance of animal breeding systems, they
remind us of the potential effects of fishing on reproductive dynamics
and population sustainability.
Acknowledgements
We greatly appreciate the help in the field and laboratory given
by S. Kelly, R. Stewart, M. Oliver, J. Heisig, L. Cox, and
D. Behringer. E. Abraham made useful comments on the draft ms.
M. Butler et al.
This research was supported in New Zealand by FRST contracts
CO1408, CO1803, and CO1X0004, and in the United States by
the National Science Foundation Award: OCE—9730195 and by
MARFIN contract # 0518. All laboratory experiments conformed
to the ethical treatment of animals as stipulated by US or NZ law
and required in animal ethics approval NIWA No 33.
References
Anger, K., Moreira, G. S., and Ismael, D. 2002. Comparative size,
biomass, elemental composition (C,N,H), and energy concentration of caridean shrimp eggs. Invertebrate Reproduction and
Development, 42: 83– 93.
Annala, J. H., and Bycroft, B. L. 1987. Fecundity in the New Zealand red
rock lobster, Jasus edwardsii. New Zealand Journal of Marine and
Freshwater Research, 21: 591 – 597.
Aronson, R. B., and Heck, K. L. 1995. Tethering experiments and
hypothesis testing in ecology. Marine Ecology Progress Series, 121:
307– 309.
Aronson, R. B., Heck, K. L., and Valentine, J. F. 2001. Measuring predation with tethering experiments. Marine Ecology Progress
Series, 214: 311 – 312.
Atema, J. 1986. Review of sexual selection and chemical communication
in the lobster Homarus americanus. Canadian Journal of Fisheries
and Aquatic Sciences, 43: 2283– 2290.
Bertelsen, R. D., and Cox, C. 2001. Sanctuary roles in populations
and reproductive dynamics of spiny lobster. In Spatial Processes
and Management of Marine Populations, Alaska Sea Grant
AK-SE-01-02, pp. 591– 605. Ed. by G. H. Kruse, N. Bez, A. Booth,
M. W. Dorn, S. Hills, R. N. Lipcius, D. Pelletier, et al. University of
Alaska Sea Grant, Fairbanks, AK, USA.
Bertelsen, R., and Horn, L. 2000. Night-time courting behaviour of
Panulirus argus in south Florida. The Lobster Newsletter, 13: 12– 13.
Bertelsen, R. D., and Mathews, T. R. 2001. Fecundity dynamics of female
spiny lobster (Panulirus argus) in a south Florida fishery and Dry
Totugas National Park lobster sanctuary. Marine and Freshwater
Research, 52: 1559– 1565.
Butler, M. J., IV, Heisig-Mitchell, J., MacDiarmid, A. B., and Swanson,
R. J. 2011. The effect of male size and spermatophore characteristics
on reproduction in the Caribbean spiny lobster, Panulirus argus.
New Frontiers in Crustacean Biology, 15: 69 – 84.
Butler, M. J., IV, MacDiarmid, A., and Gnanalingam, G. 2015. The effect
of parental size on spermatophore production, egg quality, fertilization success, and larval characteristics in the Caribbean Spiny lobster,
Panulirus argus. ICES Journal of Marine Science, 72(Suppl. 1): i115 –
i123.
Carver, A. M., Wolcott, T. G., Wolcott, D. L., and Hines, A. H. 2005.
Unnatural selection: effects of a male-focused size-selective fishery
on reproductive potential of a blue crab population. Journal of
Experimental Marine Biology and Ecology, 319: 29 – 41.
Christy, J. H. 1987. Competitive mating, mate choice and mating associations of brachyuran crabs. Bulletin of Marine Science, 41: 177–191.
Coltman, D. W., O’Donoghue, P., Jorgenson, J. T., Hogg, J. T., Strobeck,
C., and Festa-Bianchet, M. 2003. Undesirable evolutionary consequences of trophy hunting. Nature, 426: 655 – 658.
Debuse, V. J., Addison, J. T., and Reynolds, J. D. 1999. The effects of
sex ratio on sexual competition in the European lobster. Animal
Behaviour, 58: 973– 981.
Debuse, V. J., Addison, J. T., and Reynolds, J. D. 2003. Effects of breeding
site density on competition and sexual selection in the European
lobster. Behavioural Ecology, 14: 396– 402.
DeMartini, E. D., Ellis, D. M., and Honda, V. A. 1993. Comparisons of
spiny lobster Panulirus marginatus fecundity, egg size, and spawning
frequency before and after exploitation. Fisheries Bulletin US, 91:
1 – 7.
Ehrhardt, N. M. 2008. Estimating growth of the Florida spiny lobster,
Panulirus argus, from molt frequency and size increment data
Spiny lobster mate choice
derived from tag and recapture experiments. Fisheries Research, 93:
332– 337.
Fenberg, P. B., and Roy, K. 2008. Ecological and evolutionary consequences of size-selective harvesting: how much do we know?
Molecular Ecology, 17: 209– 220.
Foster, S. J., and Vincent, A. C. J. 2004. Life history and ecology of seahorses: implications for conservation and management. Journal of
Fish Biology, 65: 1– 61.
Fromentin, J-M., and Fonteneau, A. 2001. Fishing effects and life history
traits: a case study comparing tropical versus temperate tunas.
Fisheries Research, 53: 133 – 150.
Garcia, H. N., Kolding, J., Rice, J., Rochet, M. J., Zhou, S., Arimoto, T.,
Beyer, J. T., et al. 2012. Reconsidering the consequences of selective
fisheries. Science, 335: 1045– 1047.
Gosselin, T., Sainte-Marie, B., and Bernachetz, L. 2003. Patterns of
sexual cohabitation and female ejaculate storage in the American
lobster (Homarus americanus). Behavioural Ecology and
Sociobiology, 55: 151 – 160.
Heino, M., Baulier, L., Boukal, D. S., Ernande, B., Johnston, F. D.,
Mollet, F. M., Pardoe, H., et al. 2013. Can fisheries-induced evolution shift reference points for fisheries management? ICES Journal
of Marine Science, 70: 707 – 721.
Hunt, J. H., Mathews, T. R., Forcucci, D., Hedin, B. S., and Bertelsen,
R. D. 1991. Management implications of trends in the population
dynamics of the Caribbean spiny lobster, Panulirus argus, at
Looe Key National Marine Sanctuary. NOAA Final Report,
#50-DGNC-6-00093. 45 pp.
Jack, L., and Wing, S. R. 2010. Maintenance of old-growth size structure
and fecundity of the red rock lobster Jasus edwardsii among marine
protected areas in Fiordland, New Zealand. Marine Ecology Progress
Series, 404: 161 – 172.
Johnstone, R. A. 1997. The tactics of mutual mate choice and competitive search. Behavioural Ecology and Sociobiology, 40: 51– 59.
Kelly, S., and MacDiarmid, A. B. 2003. Movement patterns of spiny
lobsters, Jasus edwardsii, from a marine reserve. New Zealand
Journal of Marine and Freshwater Research, 37: 149 – 158.
Kelly, S., Scott, D., MacDiarmid, A. B., and Babcock, R. C. 2000. Spiny
lobster, Jasus edwardsii, recovery in New Zealand marine reserves.
Biological Conservation, 92: 359– 369.
Kendall, M. S., and Wolcott, T. G. 1999. The influence of male mating
history on male – male competition and female choice in mating
associations in the blue crab, Callinectes sapidus (Rathbun).
Journal of Experimental Marine Biology and Ecology, 239: 23 – 32.
Kraak, S. B. M., and Bakker, T. C. M. 1998. Mutual mate choice in sticklebacks: attractive males choose big females which lay big eggs.
Animal Behaviour, 56: 859 – 866.
Kuparinen, A., and Hutchings, J. A. 2012. Consequences of
fisheries-induced evolution for population productivity and recovery potential. Proceedings Royal Society B, 279: 2571– 2579.
Lipcius, R. N. 1985. Size-dependent reproduction and molting in spiny
lobsters and other long-lived decapods. In Crustacean Issues
3. Factors in Adult Growth, pp. 129 – 148. Ed. by A. M. Wenner.
A.A. Balkema, Rotterdam.
Lipcius, R. N., Edwards, M. L., Hernkind, W. F., and Waterman, S. A.
1983. In situ mating behaviour of the spiny lobster Panulirus
argus. Journal of Crustacean Biology, 3: 217 – 222.
Lipcius, R. N., and Herrnkind, W. F. 1985. Photoperiodic regulation and
daily timing of spiny lobster mating behaviour. Journal of
Experimental Marine Biology and Ecology, 89: 191 – 204.
Lyons, W. G., Barber, D. G., Foster, S. M., Kennedy, F. S., and Milano,
G. R. 1981. The spiny lobster of the middle and upper Florida
Keys: population structure, seasonal dynamics and reproduction.
Florida Marine Research Publications, 38: 1– 45.
MacArthur, R., and Wilson, E. O. 1967. The Theory of Island
Biogeography. Princeton University Press, Princeton, NJ.
MacDiarmid, A. B. 1989a. Size at onset of maturity and size dependent
reproductive output of female and male spiny lobsters Jasus
i113
edwardsii (Hutton) (Decapoda, Palinuridae) in northern New
Zealand. Journal of Experimental Marine Biology and Ecology,
127: 229– 243.
MacDiarmid, A. B. 1989b. Molting and reproduction of the spiny
lobster Jasus edwardsii (Decapoda, Palinuridae) in northern New
Zealand. Marine Biology, 103: 303– 310.
MacDiarmid, A. B. 1991. Seasonal changes in depth distribution, sex
ratio and size frequency of spiny lobsters Jasus edwardsii
(Decapoda, Palinuridae) on a coastal reef in northern New
Zealand. Marine Ecology Progress Series, 70: 129– 141.
MacDiarmid, A. B. 1994. Cohabitation in the spiny lobster Jasus edwardsii (Hutton 1875). Crustaceana, 66: 341 –355.
MacDiarmid, A. B., and Butler, M. J., IV 1999. Sperm economy and limitation in spiny lobsters. Behavioral Ecology and Sociobiology, 46:
14 – 24.
MacDiarmid, A. B., Hickey, B., and Maller, R. A. 1991. Daily movement
patterns of spiny lobster Jasus edwardsii (Hutton) on a shallow reef in
northern New Zealand. Journal of Experimental Marine Biology and
Ecology, 147: 185 – 205.
MacDiarmid, A. B., and Kittaka, J. 2000. Breeding. In Spiny
Lobsters: Fisheries and Culture, pp. 485– 507. Ed. by B. F. Phillips,
and J. Kittaka. Blackwell Scientific Publications, Cambridge.
MacDiarmid, A. B., and Sainte-Marie, B. 2006. Reproduction.
In Lobsters: Biology, Management, Aquaculture and Fisheries,
pp. 45 –77. Ed. by B. F. Phillips. Blackwell Scientific Publications,
Cambridge.
MacDiarmid, A. B., Stewart, R., and Oliver, M. 1999. Jasus females are
vulnerable to mate availability. The Lobster Newsletter, 12: 1 – 2.
Maxwell, K. E., Matthews, T. R., Bertelsen, R. D., and Derby, C. D. 2009.
Using age to evaluate reproduction in Caribbean spiny lobster,
Panulirus argus, in the Florida Keys and Dry Tortugas, United
States. New Zealand Journal of Marine and Freshwater Research,
43: 139– 149.
Melville-Smith, R., and de Lestang, S. 2006. Spatial and temporal variation in the size at maturity of the western rock lobster Panulirus
cygnus George. Marine Biology, 150:183 – 195.
Peterson, C. H., and Black, R. 1994. An experimentalist’s challenge:
when artifacts of intervention interact with treatment. Marine
Ecology Progress Series, 111: 289– 297.
Pianka, E. R. 1970. On r and k selection. American Naturalist, 104:
592– 597.
Pollock, D. E. 1991. Population regulation and stock recruitment
relationships in some crayfish and lobster populations. In
Crustacean Issues 7: Crustacean Egg Production, pp. 247 – 266. Ed.
by A. Wenner, and A. Kuris. Balkema, Rotterdam.
Pollock, D. E. 1997. Egg production and life-history strategies in some
clawed and spiny lobster populations. Bulletin of Marine Science,
61: 97 – 109.
Raethke, N., MacDiarmid, A. B., and Montgomery, J. C. 2004. The role
of olfaction during mating in the southern temperate spiny lobster
Jasus edwardsii. Hormones and Behaviour, 46: 311– 318.
Reznick, D., Bryant, M. J., and Bashey, F. 2002. r- and K-selection revisited: the role of population regulation in life-history evolution.
Ecology, 83: 1509 – 1520.
Robertson, D. N., and Butler, M. J., IV. 2009. Variable reproductive
success in fragmented populations. Journal of Experimental
Marine Biology and Ecology, 377: 84– 92.
Robertson, D. N., and Butler, M. J., IV. 2012. Mate choice and sperm
limitation in the spotted spiny lobster, Panulirus guttatus. Marine
Biological Research, 9: 69– 76.
Rondeau, A., and Sainte-Marie, B. 2001. Variable mate guarding time
and sperm allocation by male snow crabs (Chionoecetes opilio) in response to sexual competition, and their impact on the mating success
of females. Biological Bulletin, 201: 204 – 217.
Rowe, S., and Hutchings, J. A. 2003. Mating systems and the conservation of commercially exploited marine fish. Trends in Ecology and
Evolution, 18: 567 – 572.
i114
Sainte-Marie, B. 2007. Sperm demand and allocation in decapod crustaceans. In Evolutionary Ecology of Social and Sexual Systems:
Crustaceans as Model Systems. Ed. by J. E. Duffy, and M. Thiel.
Oxford University Press, Oxford.
Sainte-Marie, B., Sévigny, J-M., and Carpentier, M. 2002. Interannual
variability of sperm reserves and fecundity of primiparous females
of the snow crab (Chionoecetes opilio) in relation to sex ratio.
Canadian Journal of Fisheries and Aquatic Sciences, 59: 1932– 1940.
Sainte-Marie, B., Urbani, N., Sévigny, J-M., Hazel, F., and Kuhnlein, U.
1999. Multiple choice criteria and the dynamics of assortative mating
during the first breeding season of female snow crab Chionoecetes
opilio (Brachyura, Majidae). Marine Ecology Progress Series, 181:
141– 153.
Sarma, S., Nandini, S., and Gulati, R. D. 2005. Life history strategies
of cladocerans: comparisons of tropical and temperate taxa.
Hydrobiologia, 542: 315– 333.
Sato, T., Ashidate, M., Wada, S., and Goshima, S. 2005. Effects of male
mating frequency and male size on ejaculate size and reproductive
M. Butler et al.
success of female spiny king crab Paralithodes brevipes. Marine
Ecology Progress Series, 296: 251– 262.
Sato, T., and Goshima, S. 2006. Impacts of male-only fishing and sperm
limitation in manipulated populations of an unfished crab,
Hapalogaster dentata. Marine Ecology Progress Series, 313:
193– 204.
Sato, T., and Goshima, S. 2007. Effects of risk of sperm competition,
female size, and male size on number of ejaculated sperm in the
stone crab Hapalogaster dentata. Journal of Crustacean Biology,
27: 570– 575.
Sato, T., Yoseda, K., Okuzawa, K., and Suzuki, N. 2010. Sperm
limitation: possible impacts of large male-selective harvesting on
reproduction of the coconut crab Birgus latro. Aquatic Biology, 10:
23 – 32.
Vila-Gispert, A., Moreno-Amich, R., and Garcıa-Berthou, E.
2002. Gradients of life-history variation: an intercontinental
comparison of Fishes. Reviews in Fish Biology and Fisheries, 12:
417– 427.
Handling editor: Jonathan Grabowski
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i115 –i123. doi:10.1093/icesjms/fsv015
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
The effect of parental size on spermatophore production,
egg quality, fertilization success, and larval characteristics
in the Caribbean Spiny lobster, Panulirus argus
Mark J. Butler, IV1 *, Alison Macdiarmid 2, and Gaya Gnanalingam1
1
Department of Biological Sciences, Old Dominion University, Norfolk, VA, USA
National Institute for Water and Atmospheric Research, Wellington, New Zealand
2
*Corresponding author: tel: +1 757 683 3609; fax: +1 757 683 5283; e-mail: [email protected]
Butler, M. J.IV, Macdiarmid, A., and Gnanalingam, G. The effect of parental size on spermatophore production, egg quality,
fertilization success, and larval characteristics in the Caribbean Spiny lobster, Panulirus argus. – ICES Journal of Marine
Science, 72: i115 – i123.
Received 24 August 2014; revised 7 January 2015; accepted 11 January 2015; advance access publication 17 February 2015.
The average size of spiny lobsters (Decapoda; Palinuridae) has decreased worldwide over the past few decades. Market forces coupled with
minimum size limits compel fishers to target the largest individuals. Males are targeted disproportionately as a consequence of sexual dimorphism
in spiny lobster size (i.e. males grow larger than females) and because of protections for ovigerous females. Therefore, overexploitation of males has
led to sperm limitation in several decapod populations with serious repercussions for reproductive success. In the Caribbean spiny lobster, Panulirus
argus, little is known about the effect of reduced male size on fertilization success or the role that individual size plays in gamete and larval quality.
We conducted a series of laboratory experiments to test the relationship between male size and spermatophore production over multiple mating
events and to determine whether spermatophore reduction and female size affected fertilization success or larval attributes in P. argus in the Florida
Keys, FL (USA). We found that over consecutive matings, larger males consistently produced spermatophores of a greater weight and area than
smaller males, although size-specific differences in sperm cell density were undetected and probably obscured by high variance in the data.
Where spermatophores were experimentally reduced to mimic the decline in spermatophore size with declining male size, fertilization success
(the number of fertilized eggs/total number of eggs extruded) declined, indicating that sperm availability is indeed limited. No maternal size
effects on egg size or quality (C:N ratio) or larval quality (size, swimming speed, mortality) were observed. Our results demonstrate the importance
of maintaining large males in populations of P. argus to ensure fertilization success and caution against their overexploitation through fishing, which
may severely reduce reproductive success and thus population sustainability.
Keywords: egg quality, larvae, mating, Panulirus argus, reproduction, spermatophore.
Introduction
Fishing significantly alters the size structure of exploited populations with major consequences for mating systems, reproduction,
and the sustainability of exploited populations (Rowe and
Hutchings, 2003; Maxwell et al., 2009; Butler et al., 2011a; Garcia
et al., 2012; Kuparinen and Hutchings, 2012). Typically, the
largest and oldest individuals are removed first, particularly where
minimum size limits are enforced, which reduces the mean size of
individuals in the population (Roberts and Polunin, 1991;
Jennings and Lock, 1996; Heath and Speirs, 2012). Where there is
sexual dimorphism in the size of adults and males grow larger
# International
than females, fishing can skew sex ratios in favour of females.
Prohibitions on the removal of ovigerous females may further
skew sex ratios. As highlighted by a number of decapod species,
these fishing induced changes in size structure, abundance, and
sex ratio can restructure mating systems and reduce the reproductive output of the population (Kendall et al., 2002; Sato and Goshima,
2007; Sato et al., 2010; Robertson and Butler, 2013). Such may be the
case for the Caribbean spiny lobster, Panulirus argus.
Panulirus argus is ubiquitous throughout the Western Atlantic
Ocean, Caribbean Sea, and Gulf of Mexico from NC, USA, to northeast Brazil (Briones-Fourzan et al., 2008). Fisheries for P. argus are
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i116
some of the largest and most economically valuable in the Caribbean
with an estimated annual regional value more than $450 million
USD (CRFM, 2013). As a consequence of its high value and
market demand, many regional populations of P. argus are currently
fully capitalized or overfished (Chavez, 2009; Ehrhardt et al., 2010).
The species displays considerable sexual dimorphism; males grow
more than three times larger (by mass) than females and are distinguishable from females by their broader sternum, elongated walking
legs, and curved dactyls (Holthius, 1991). So although fishing
reduces body size in both sexes, the decline is more pronounced
in males (Bertelsen and Mathews, 2001; Cox and Hunt, 2005).
In unfished populations where large males are still present, lobsters demonstrate a lek-style mating system in which large males
defend a den from other large males and females choose among
males and their dens (Bertelsen and Cox, 2001). In heavily fished
populations where large males have been removed this system
breaks down, although competition among males still remains a
major component of the mating system (Bertelsen and Mathews,
2001; Butler et al., 2015). During mating, male and female P. argus
couple and males deposit an external, tar-like spermatophore on
the sternum of the female (Lipcius et al., 1983; Figure 1). Females
resist further mating attempts by males and scratch open the spermatophore 1–28 d after copulation to expose the non-motile sperm,
which fertilize the eggs externally as they are extruded (Talbot and
Summers, 1978; Butler et al., 2011a). The fertilized eggs attach to
the setose pleopods on the underside of the female’s abdomen, and
they are carried there for 3–4 weeks until they hatch (Saul, 2004).
Throughout much of the Caribbean P. argus spawn year-round
(Butler et al., 2010), although spawning peaks in late spring and in
subtropical areas, the breeding season is constrained to just
spring–summer (Chubb, 2000). In Florida, for example, spawning
generally commences in March and continues through August.
However, the number of clutches and the duration of the spawning
period vary with female size; larger females produce more clutches
and become reproductively active earlier in the spawning period
when compared with small, mature females (Bertelsen and
Mathews, 2001). Mating of all palinurids occurs during the intermoult phase, and there is no clear link between molting and
mating in P. argus, as there is in other spiny lobster species such as
Jasus edwardsii (MacDiarmid and Butler, 1999).
Individual fecundity in female P. argus is size-dependent with
larger individuals producing exponentially more eggs (Bertelsen
and Mathews, 2001; Ehrhardt, 2005; MacDiarmid and
Sainte-Marie, 2006). A reduction in female body size as a result of
fishing, therefore, has clear implications for reproductive capacity
in terms of egg production. What is less clear is the effect of
reduced male size on reproductive dynamics. Decapods may experience sperm limitation if ejaculate size scales with male body size,
mating history, expected female output, or future mating opportunities (reviewed by MacDiarmid and Sainte-Marie, 2006). The probability that populations are sperm limited is therefore closely related
to population structure, mean male size, density, sex ratio, and the
intensity of sexual competition (Sato et al., 2010). Previous
studies of P. argus have demonstrated a clear link between the availability of large males and female fecundity because spermatophore
size, a function of male size, explains nearly half of the variance in
clutch size (MacDiarmid and Butler, 1999) and because sperm:egg
ratios—which are already low (25:1)—are 40% lower in fished
populations (Butler et al., 2011a).
Building on the previous work of MacDiarmid and Butler
(1999), we present the results of a series of laboratory experiments
M. J. Butler et al.
that describe the mating system of P. argus with special reference
to the effect of male size on reproductive success. First, we experimentally manipulated the size of spermatophores on mated
females to determine the potential effects of reduced male size on
fertilization success and egg production by females. We then
assessed the effects of male size on spermatophore production, depletion, and recovery over successive matings. Finally, using data
from the spermatophore reduction and depletion/recharge experiments, we evaluated maternal effects—i.e. the effects of female size
on egg and larval quality.
Methods
Spermatophore reduction experiment
To assess whether sperm availability can limit reproductive success
in P. argus, we manipulated the size of spermatophores on mated
females collected from the field and measured the resultant production of fertilized and unfertilized eggs. Female P. argus with spermatophores were collected by divers from the Florida Keys, FL, USA
(a fished population), and the Dry Tortugas National Park, FL,
USA (a no-take marine protected area; MPA) in February– March
of 1998– 2000. Lobsters were transported in aerated live wells to
the Florida Fish and Wildlife Conservation Commission field
laboratory in Marathon, FL, where the experiments took place.
Females were split into two size classes: small (carapace length
80 –90 mm) and large (carapace length .90 mm) and kept in individual 75 l tanks with aeration and flow through seawater; lobsters
were fed squid and shrimp ad libitum daily. These size classes of
female lobsters represent those that are common in heavily fished
vs. unfished populations of P. argus (respectively); thus, the results
of this study provide specific insight into how current fishing practices potentially impact reproductive dynamics. Spermatophores on
females were then manipulated with a scalpel and flat forceps such
that lobsters belonged to one of three treatment groups: (i) a 50%
reduction in spermatophore mass, (ii) a 75% reduction in spermatophore mass, and (iii) a no reduction control (Figure 1). Sperm
cells are evenly distributed between the left and right halves
(Butler et al., 2011a) of the spermatophore, but are more concentrated in the posterior portion of each paired spermatophore.
Therefore, the removal of one of the paired spermatophores is a
proxy for a 50% reduction in spermatophore mass, which approximates the difference in spermatophore sizes transferred to
females by large and small males (MacDiarmid and Butler, 1999);
a 75% reduction in size was simply chosen as a third, extreme treatment. Although cutting the spermatophore transversely from anterior to posterior would have maintained the total area of the
spermatophore, they were cut longitudinally to preserve as much
as possible the tough outer layer of the spermatophore and
prevent the exposure of the sperm before fertilization. The 50% reduction treatment would not have resulted in any inadvertent
“leakage” of sperm from the remaining spermatophore, given that
each spermatophore half is an independent unit produced by the
paired testes and delivered via paired gonopores. Perhaps, sperm
may have leaked from the remaining portion of the spermatophores
in the 75% reduction treatment, but if so that procedural effect
would be the same for both large and small lobsters.
Lobsters were then monitored daily for the presence of attached
eggs, an indication of spawning and fertilization. After spawning,
unfertilized eggs fail to attach to the female’s pleopods, so we
siphoned them from the tank bottom and counted them to estimate
the abundance of unfertilized eggs (see below). When eyespots
Effect of parental size in the Caribbean Spiny lobster
i117
Figure 1. Visualization of the way in which spermatophore size was diminished in the spermatophore reduction experiment. In each frame is a
ventral view of the same female P. argus with eggs from a previous clutch (visible under the abdomen) and a new spermatophore (black mass) that
will be used to fertilize the next clutch. (a) 50% reduction treatment: white area depicts how half of the spermatophore was removed. (b) 75%
reduction treatment: white areas depict how three-fourth of the spermatophore was removed. (c) Control: no manipulation of the spermatophore.
became visible in brooded eggs (14 d after spawning), a scalpel
was used to gently scrape off egg masses from the pleopods.
Counts of unfertilized (siphoned from tank) and fertilized
(scraped from female pleopods) eggs were made by taking three
2 g subsamples (weighed to the nearest hundredth of a gramme)
of each egg mass. Eggs were counted under a dissecting microscope
to provide an estimate of individual female fecundity (number of
fertilized eggs) and the total number of eggs extruded (number of
fertilized + unfertilized eggs). Fertilization success is defined as
the ratio of fertilized eggs retained on pleopods to the total
number of eggs extruded. To evaluate differences in fertilization
success as a function of the spermatophore reduction treatment,
we used a two-factor ANOVA with female size class and treatment
as factors.
Sperm depletion and recovery
We determined if large and small adult males differ in their production of spermatophores and their ability to recharge sperm stores
after mating, by repeatedly mating large males collected from an
MPA, and small males from a fished population, with several
females in a laboratory experiment. Divers collected males and
females from the Florida Keys fished population and the Dry
Tortugas National Park MPA during February and March of
1999–2001. The absence of large males in the heavily exploited
Florida Keys population necessitated the collection of large individuals only from the Dry Tortugas. All males were collected before
the onset of the breeding season and therefore had not mated
since the previous year’s mating season. A single large male (126 –
160 mm CL; n ¼ 22) or small male (90– 99 mm CL; n ¼ 19)
along with seven females ranging in size from 70 –144 mm CL
were placed into experimental tanks (1.75 m diameter; 1500 l)
that received aerated filtered seawater from a flow-through
system. Ambient seawater temperature and photoperiod were
maintained and lobsters were fed shrimp and squid ad libitum.
These experimental conditions (e.g. sizes and sex ratios) are
typical of conditions in the wild during the reproductive season
when females significantly outnumber males on coral reefs
because large, agonistic males temporarily drive smaller males
into alternative habitats (Bertelsen and Mathews, 2001; Butler
et al., 2015).
Females were checked for spermatophores daily. Flat forceps
were used to remove intact spermatophores after they had hardened
for 24 h. Before their removal, the spermatophore area was estimated. An outline of the mass was traced onto a sheet of a clear
acetate paper, which was scanned and digitally measured in Image J.
Once removed, spermatophores were weighed to the nearest hundredth of a gramme and refrigerated at 58C in 10 ml of filtered seawater until sperm counts could be conducted. Females were
returned to their original experimental tanks after the removal of
the spermatophore so that they could remate. In the Florida Keys
and Dry Tortugas, large P. argus females mate and produce up to
three and perhaps four clutches within a given season, whereas
i118
small females produce only a single clutch per year (Lyons et al.,
1981; Bertelsen and Mathews, 2001). As such, we could rely on
females remaining receptive to courtship and remating during the
course of the experiment. The male mating frequency in the wild
is unknown, but the lek-style mating strategy of large males indicates
that at least large males mate repeatedly during the breeding season.
To conduct sperm counts, spermatophores were laterally sliced
into thin sections (0.5 mm thick) with a scalpel into a Petri
dish. The spermatophore sections were then washed from the
Petri dish into a 25-ml test tube with 10 –15 ml of sterile seawater;
the test tube was capped and then mechanically shaken for 3 min
to liberate sperm from the spermatophore matrix. A haemocytometer was then used to count the number of sperm cells in five separate
10 ml aliquots of each sample. The total sperm number and sperm
density (total sperm number/spermatophore weight) was then calculated. This method for sperm counts has been used previously and
it has been shown to consistently liberate .70% of the sperm from
the spermatophore matrix (Butler et al., 2011a).
Males repeatedly mated with the females in their experimental
tanks and the order in which spermatophores were produced was
recorded. On the cessation of mating (i.e. production of new spermatophores) in each experimental tank, typically 10 d to a month from
first mating, males were removed and isolated in individual tanks for
1, 2, or 3 weeks to potentially “recharge” their sperm stores. After the
recharge period, males were returned to their original experimental
tank with the same set of females to resume mating.
Sperm depletion and recharge rates were assessed as a function of
male size with individuals separated into large (.100 mm CL) and
small (,100 mm CL) size classes. To test whether sperm attributes
(i.e. spermatophore weight, spermatophore area, sperm density,
and total number of sperm) varied relative to male size for the
first deposited spermatophore, a one-factor MANOVA with canonical discriminant function analysis was performed. To determine
whether male size affected sperm attributes over repeated matings
(sperm depletion), we used a split plot MANOVA with male size
as the whole plot factor, ejaculate number as the subplot factor,
and individual as a randomized block. To assess whether male size
affected sperm attributes after the set recharge period, we used a
two-factor MANCOVA with the covariate “time between mating”
and the independent variables male size and recharge period (i.e.
1, 2, or 3 weeks). Finally, paired t-tests were used to compare spermatophores produced initially and following the recharge period
by the same individual.
M. J. Butler et al.
with an ocular micrometre. The C:N content of eggs (a measure
of egg lipid content and thus egg condition or quality; Giminez
and Anger, 2001; Liddy et al., 2003; Bas et al., 2007) was also determined from the subsamples of each egg clutch. Subsamples were
rinsed in sterile seawater, frozen, and stored at –208C, then dried
before C and N were measured with a Carlo –Erba elemental analyser and standard methods.
Larval condition was measured in three ways: larval size, survival,
and swimming speed. The carapace length of 20 first stage phyllosome larvae per clutch was determined with a dissecting microscope
fitted with an ocular micrometre. Larval survival was assessed in a
starvation trial in which 20 first stage phyllosome larvae per clutch
were individually housed in 15 ml Petri dishes filled with sterile seawater but no food at ambient light and temperature (268C). Larval
mortality was assessed daily and sterile seawater was changed daily
until all larvae had expired. The swimming speed of ten individual
phyllosome larvae per clutch was also estimated in a seawater-filled,
black swimming chamber (25 cm long × 10 cm wide × 5 cm deep)
with a 1-cm grid scale on the bottom. Clear holes (0.5 cm dia) at
either end of the plastic chamber permitted light (wavelength:
400–600 nm; intensity: 0.80 –1.00 mmol m22 s21) to enter either
end of the chamber from a 300-W quartz-halogen filament lamp
(Olympus model LGPS) projected from a fibre optic cable and
through a filter. This wavelength and intensity is similar to that
observed in the surface water (,25 m) of the open sea where firststage larvae normally dwell (Butler et al., 2011b). At the start of
each trial, the room was darkened and a single larva was added to
one end of the test chamber, while the chamber was illuminated
from the opposite end. First-stage phyllosome larvae are positively
phototactic, so they swam towards the light at the opposite side of
the chamber. We measured the distance the larvae moved over a
5-s period and then let the larvae swim all the way to the lighted
end of the chamber. Larvae were given a 1-min respite in total darkness before we illuminated the opposite end of the chamber and
repeated the procedure. The swimming speed of each larva was
tested in this manner ten times and the mean swimming speed determined for each larvae. The relationships between female size and
egg characteristics (area, C:N ratio) and larval characteristics
(larval carapace length, days until 100% mortality, and swimming
speed) were analysed with linear regression.
Results
Spermatophore reduction and sperm limitation
Egg quality and larval characteristics
During sperm depletion/recharge experiments, the spermatophore
on six mated females from every treatment group was left intact and
the date of spermatophore deposition and the date of egg extrusion
were recorded. This took anywhere between 0 and 19 d for different
females with a mean time of 4 d. These females were removed from
the experimental tanks and on the 10th day after egg extrusion
random samples of the egg masses were removed, for the assessment
of maternal effects on egg quality and to provide the estimates of fecundity. Females were then returned to holding tanks until their
remaining clutch had hatched. Eggs and newly hatched larvae
from these females in the sperm depletion/recharge experiment
and from control females in the spermatophore reduction experiment were used to assess possible maternal effects on egg and
larval quality.
For egg quality, the egg area was determined by measuring the
diameter of 20 eggs per clutch with a dissecting microscope fitted
When spermatophores were experimentally reduced in area, the
total number of eggs that were released did not differ significantly
among treatments (Figure 2). However, fertilization success
(number of fertilized eggs/total eggs extruded) was lower for
females whose spermatophores had been reduced (Figure 2;
ANOVA: F ¼ 13.59, p , 0.001, d.f. ¼ 1). Post hoc tests indicated
that differences lay between the control and 50% reduction treatments (Tukey HSD, p ¼ 0.030) and the control and 75% reduction
treatments (Tukey HSD, p ¼ 0.003) with no detectable difference
between the reduction treatments. When fertilization success was
assessed for large and small females separately, differences among
treatments were only significant for small females (Figure 2;
ANOVA: F ¼ 13.78, p ¼ 0.001, d.f. ¼ 1); the results were similar
but not significant for large females (ANOVA: F ¼ 1.34, p ¼ 0.27,
d.f. ¼ 1), although the results for large females are tentative given
the small sample sizes, hence lower power of the tests when the
data were divided into two lobster size groups.
Effect of parental size in the Caribbean Spiny lobster
i119
Figure 2. Results of the spermatophore reduction experiment; bars are
the means with standard errors (top panel). The total number of eggs
released (at left) by female P. argus in the three treatments and the total
number of fertilized eggs retained by female lobsters (at right). (bottom
panel) The number of fertilized eggs produced by large (at left) and
small (at right) female lobsters in the three treatment groups. Note that
the number of eggs retained by female lobsters reflects fertilization
success because unfertilized eggs do not attach to the females
pleopods. NS results and significant results (with P-values) among
treatment groups are shown in each panel; treatments sharing a letter
are not significantly different from each other.
Spermatophores depletion and recharge over multiple
matings
Males took 1 –22 d between consecutive matings with a mean time
of 2.5 d between mating events. Attributes of initial spermatophores
varied significantly between large and small males (Wilks’ lambda,
p , 0.0001), with large males typically producing spermatophores
that were heavier and of a larger surface area (Figure 3; spermatophore weight: F ¼ 13.41, p ¼ 0.001, d.f. ¼ 1; spermatophore area:
F ¼ 14.07, p ¼ 0.001, d.f. ¼ 1). However, the number of sperm
cells and sperm density for initial spermatophores did not differ significantly between large and small males (number of cells: F ¼ 3.41,
p ¼ 0.72, d.f. ¼ 1; density: F ¼ 0.01, p ¼ 0.943, d.f. ¼ 1), although
the values for large males are consistently greater than those for small
males and the lack of a difference may very well be due to the high
variance in these data.
Males fertilized between 2 and 27 females consecutively; larger
males (.100 mm) mated with 15 females on average and small
males (,100 mm) mated with 11 females on average over a
period of up to 1 month. This difference was not statistically significant (t-test: t ¼ 0.25, p ¼ 0.80, d.f. ¼ 28). For males of all sizes
Figure 3. Results of the sperm depletion experiment for large (dark
squares; n ¼ 30) and small (grey triangles; n ¼ 11) male P. argus; shown
are the means with 1 s.d. Spermatophore weight (top panel), number of
sperm per spermatophore (second panel), sperm density (third panel),
and spermatophore area (bottom panel) are shown for consecutive
matings and after a non-mating recharge period of 1, 2, or 3 weeks.
i120
combined, the spermatophore weight and area decreased markedly
as the number of spermatophores per individual increased (Figure 3;
spermatophore weight: F ¼ 6.26, p , 0.001, d.f. ¼ 41; spermatophore area: F ¼ 2.82, p ¼ 0.003, d.f. ¼ 41).
For all males combined, no significant differences in the total
number of sperm cells and sperm density were observed over
repeated matings. But there is a large amount of variance in the
data, particularly for the smaller males, which likely obscures a
clear result when all the data are combined. However, when analysed
by size class, spermatophore weight, spermatophore area, and the
total number of sperm cells declined as the number of spermatophores extruded per individual increased (Figure 3; spermatophore
weight: F ¼ 75.18, p , 0.001, d.f. ¼ 1; spermatophore area: F ¼
101.70, p , 0.001, d.f. ¼ ; total number of cells: F ¼ 15.55,
p , 0.001, d.f. ¼ 1).
After a recharge period of 1, 2, or 3 weeks, larger males again had
heavier spermatophores that were larger in area (Figure 3; spermatophore weight: F ¼ 6.73, p ¼ 0.018, d.f. ¼ 1; spermatophore area:
F ¼ 4.65, p ¼ 0.044, d.f. ¼ 1), but there were no differences
between large and small males in the total number of sperm cells
or sperm density (Figure 3; number of cells: F ¼ 1.82, p ¼ 0.19,
d.f. ¼ 1; sperm density: F ¼ 0, p ¼ 0.99, d.f. ¼ 1). This same
result was observed regardless of the length of the recharge period
(Figure 3). There were also no differences in sperm attributes
between initial spermatophores and spermatophores produced following the recharge period (paired t-test, p . 0.05), even when size
classes were analysed separately, with the sole exception of spermatophore weight for the small males (t ¼ 3.8, p , 0.01, d.f. ¼ 7).
Egg quality and larval characteristics
Maternal size did not influence the measured egg characteristics (i.e.
area and C:N ratio), larval carapace length, or larval survival
(Figure 4). The slope of the regression of larval swimming speed
with female carapace length differed significantly from zero and
was unexpectedly negative, but that result was of borderline significance (F ¼ 5.51, p ¼ 0.02, d.f. ¼ 43) and the relationship explained
very little variance in the data, as shown by the low r 2 value (r 2 ¼
0.09; Figure 4).
Discussion
The rationale for this study was to obtain a greater understanding of
the effects of male and female P. argus size on gamete production and
quality, fertilization success, and larval characteristics associated
with survival. Given the dramatic reduction in lobster size in
exploited populations, correlations between offspring quality or
quantity with paternal or maternal size could adversely affect the reproductive potential of those populations. Our results indicate that
smaller males produce lighter and smaller spermatophores over
consecutive mating events. Increased female body size, on the
other hand, had no discernible effect on egg or larval characteristics
beyond the well-known exponential relationship between maternal
size and egg production.
Our finding that reduced spermatophore size did not alter the
number of eggs extruded by females, but instead reduced fertilization success, indicates that smaller clutches produced by females
mated by small males are not a result of reduced egg output by the
female but instead reflect the sperm limitation of fertilization
success. Spermatophore reduction also appeared to have a greater
impact on the egg retention and hence fertilization success of
smaller (,90 mm CL) females, as has been similarly noted for the
Western rock lobster P. cygnus (Kuris, 1991). Large P. argus males
M. J. Butler et al.
actually allocate larger spermatophores to large females
(MacDiarmid and Butler, 1999) and larger spermatophores are
associated with positive residual clutch weights (MacDiarmid and
Sainte-Marie 2006), so perhaps this results in an overapportionment of a viable sperm available for fertilization, which
would explain why our experimental reductions in spermatophore
size were less effective on large females than smaller ones.
Conversely, if males over-apportion sperm to larger females but
are less generous to smaller females, as is true in the American
lobster, Homarus americanus, then smaller females may receive
under-apportioned spermatophores relative to large females
(Gosselin et al., 2003). Females across a range of taxa including
fish (Skinner and Watt, 2007), insects, birds, and amphibians differentially allocate reproductive resources according to the attractiveness of their mate (see review by Sheldon, 2000). In decapods,
evidence of differential allocation of eggs based on mate quality is
limited to freshwater crayfish (Galeotti et al., 2006).
For P. argus, the apparent inability to mediate egg output suggests that females should select larger males as mates so as to maximize fertilization success. Indeed, large females of all palinurid
species thus far tested (P. argus, P. guttatus, Jasus edwardsii) experience lower fertilization success when mated with smaller males
(MacDiarmid and Butler, 1999; Robertson and Butler, 2013;
Butler et al., 2015), implying that all experience strong selection
for larger mates. Although true for some spiny lobsters (e.g.
P. guttatus; Robertson and Butler, 2013; J. edwardsii; Butler et al.,
2015), it has not proven true in experiments with P. argus
(Butler et al., 2015). This discrepancy among species in female selectivity of males by size is difficult to reconcile and bears further
study.
Spermatophores produced by large males were significantly
heavier and larger initially and following multiple matings. This is
a finding in accord with previous studies of sperm limitation in
P. argus (MacDiarmid and Butler, 1999; Butler et al., 2011a). The
number of sperm cells in the spermatophore and sperm density,
however, was not different between the size classes (the exception
being the number of sperm cells within spermatophores following
multiple matings). This despite indications that the mean number
of sperm for large males was almost always higher than for small
males (Figure 3). So perhaps the high variance in these results
obscures a size-specific difference in sperm abundance in addition
to the clear difference in spermatophore size. We also believe that
our data on the size-specific male recovery of sperm stores after depletion are probably inconclusive. Although there was no statistical
difference between spermatophore attributes after recharge periods
of 1 –3 weeks, the sample sizes were small, variances high, and means
inconsistent.
It is also possible that the larger seminal fluid component of the
spermatophore in large males when compared with small males is of
consequence to sperm viability and hence fertilization success.
Similar results have been found for the blue crab Callinectes
sapidus (Kendall et al., 2002). As with blue crabs, however, the consequences of reduced seminal fluid in terms of female reproductive
success are unknown. Accessory fluid has several functions that
include acting as an antibacterial agent or nutritive substance for
sperm, as an aid for storage and retention of sperm (Tram and
Wolfner, 1999), or as a sperm plug (Carver et al., 2005). New
genetic evidence based on paternity analysis (S. Johnson, Univ.
Otago, pers. comm.) suggests that female J. edwardsii may engage
in multiple copulations to fertilize a single egg clutch. This phenomenon appears unlikely in P. argus where more detailed examinations
Effect of parental size in the Caribbean Spiny lobster
i121
Figure 4. The relationships between the size of P. argus females and several egg and larval quality attributes, including: (a) egg area, (b) C:N ratio of
eggs, (c) larval carapace length, (d) larval mortality, and (e) larval swimming speed. Regression summary statistics are shown in each panel (r 2 ¼
variance explained, ANOVA p-value, n ¼ sample size), although a regression line is only plotted for the one instance where there was a significant
regression.
of spermatophore structure (Butler et al., 2011a) indicates that
reports of layered spermatophores (Mota-Alves and Pavia, 1976)
are most likely due to new, unused spermatophores laid upon previously exhausted spermatophores used to fertilize a prior clutch.
Thus, to avert sperm competition, large males may use accessory
fluid to enhance the size of their perceived investment or to act as
a barrier to secondary matings. Indeed, previous experiments with
P. argus demonstrate that spermatophores provide both physical
and chemical cues perceptible by females that inhibit them from
further mating before utilization of the existing spermatophore
(Butler et al., 2011a).
As first articulated by Dewsbury (1982), the cost to males of producing ejaculates is physiologically non-trivial and all the more so
for multiple ejaculates (Wedell et al., 2002). Evidence of depletion
over the production of successive spermatophores in our study
clearly demonstrates this. Large male P. argus produced more than
25 ejaculates during a mating cycle and small males around 15,
but the size of the ejaculates declined precipitously after just a few
repeated matings. Similarly, small male king crabs Paralithodes
camtschaticus experience reduced reproductive success after seven
matings, whereas large crabs can mate up to nine times before experiencing a reduction in fertilization success (Powell et al., 1974).
i122
Large P. argus also require a week or so hiatus from mating to restore
spermatophore size; small males appear to need even longer. This is
similar to observations of blue crabs Callinectes sapidus that require
between 9 and 20 d to fully recover sperm stores (Kendall et al.,
2001). Other crab species, such as the coconut crab Birgus latro
and stone crab Haplogaster dentata, are not thought to recover
sperm stores until the next reproductive season (Sato, 2012).
Maternal effects on offspring fitness have garnered considerable
attention, particularly in relation to exploited marine species
(Berkeley et al., 2004a; Green, 2008; Venturelli et al., 2010). In addition to evidence for exponential increases in fecundity relative to
size or age, positive correlations between maternal size or age and
larval characteristics provide further support for the protection of
mature size and age structure in exploited animal populations
(Berkeley et al., 2004b; Birkeland and Dayton, 2005). In lobsters,
the relationship between maternal size and fecundity is clearly established (Saul, 2004; MacDiarmid and Sainte-Marie, 2006). In comparison, knowledge of maternal influences on egg and larval
characteristics is limited, which is disconcerting given the increasing
truncation of the size range of exploited lobster populations.
Maternal size influences on gamete and/or offspring quality have
been observed in the European clawed lobster Homarus gammarus
(Moland et al., 2010), some populations of the American clawed
lobster Homarus americanus (Attard and Hudon, 1987), and in
the temperate spiny lobster J. edwardsii (A. MacDiarmid, unpub.
data). However, maternal size resulted in no detectable difference
on egg or larval characteristics in P. argus with the possible exception
of larval swimming speed, the result of which we question given its
low r 2 value. According to Marshall et al. (2010), there are few evolutionary arguments for a maternal size–offspring quality relationship unless the physical environment is the primary driver selecting
for maternal effects.
Although studies on the effects of fishing on female size and reproductive patterns are known for many species of spiny lobster
(reviewed in MacDiarmid and Sainte-Marie, 2006), detailed
mating and fertilization dynamics are poorly documented. Our
study reveals that the mating system of P. argus is considerably
more complex than previously believed. Years of continuous
fishing and the consequent human manipulation of lobster demographic structure has added to this complexity. For heavily exploited
populations like those in the Florida Keys, where the largest lobsters
that normally dominate matings are noticeably absent (Bertelsen
and Mathews, 2001; Butler et al., 2015), successful reproduction
now depends on smaller individuals. Smaller males that in healthy
unfished populations would typically be outcompeted for access
to females are afforded the chance to mate repeatedly with available
females in overfished populations. Given both the value of, and the
reliance on, P. argus fisheries in the Western Atlantic, it is imperative
that we incorporate knowledge of mating systems in population
assessments (Rowe and Hutchings, 2003) and undertake research
to better understand the implications of this. This study, like
others before it, clearly demonstrates the importance of maintaining
large individuals in exploited populations if we wish to maintain
healthy breeding populations of P. argus into the future.
Acknowledgements
We are grateful for help with our laboratory experiments by a dedicated group of individuals, including J. Heisig-Mitchell,
D. Behringer, D. Hansen-Behringer, J. Schratwieser, S. Donahue,
and D. Robertson. The use of laboratory space at the Florida Fish
and Wildlife Conservation Commission, Marathon Field
M. J. Butler et al.
Laboratory, is greatly appreciated. This research was supported by
an award to M. Butler from the National Science Foundation
(OCE-9730195).
References
Attard, J., and Hudon, C. 1987. Embryonic development and energetic
investment in egg production in relation to size of female lobster
(Homarus americanus). Canadian Journal of Fisheries and Aquatic
Sciences, 44: 1157– 1164.
Bas, C. C., Spivak, E. D., and Anger, K. 2007. Seasonal and interpopulational variability in fecundity, egg size, and elemental composition
(CHN) of eggs and larvae in a grapsoid crab, Chasmagnathus granulatus. Helgoland Marine Research, 61: 225 – 237.
Berkeley, S. A., Chapman, C., and Sogard, S. M. 2004a. Maternal age as a
determinant of larval growth and survival in a marine fish, Sebastes
melanops. Ecology, 85: 1258 – 1264.
Berkeley, S. A., Hixon, M. A., Larson, R. J., and Love, M. S. 2004b.
Fisheries sustainability via protection of age structure and spatial distribution of fish populations. Fisheries, 29: 23– 32.
Bertelsen, R. D., and Cox, C. 2001. Sanctuary roles in populations and
reproductive dynamics of spiny lobster. In Spatial Processes and
Management of Marine Populations, Alaska Sea Grant
AK-SE-01-02, pp. 591– 605. Ed. by G. H. Kruse, N. Bez, A. Booth,
M. W. Dorn, S. Hills, R. N. Lipcius, and D. Pelletier et al.
University of Alaska Sea Grant, Fairbanks, AK, USA.
Bertelsen, R. D., and Mathews, T. R. 2001. Fecundity dynamics of female
spiny lobster (Panulirus argus) in a south Florida fishery and Dry
Tortugas National Park lobster sanctuary. Marine and Freshwater
Research, 52: 1559– 1565.
Birkeland, C., and Dayton, P. K. 2005. The importance in fishery management of leaving the big ones. Trends in Ecology and Evolution, 20:
356– 358.
Briones-Fourzan, P., Candela, J., and Lozano-Alvarez, E. 2008.
Postlarval settlement of the spiny lobster Panulirus argus along the
Caribbean coast of Mexico: Patterns, influence of physical factors,
and possible sources of origin. Limnology and Oceanography, 53:
970– 985.
Butler, M. J., IV, Bertelsen, R., and MacDiarmid, A. 2015. Mate choice in
temperate and tropical spiny lobsters with contrasting reproductive
systems. ICES Journal of Marine Science, 72(Suppl. 1): i101 – i114.
Butler, M. J., IV, Heisig-Mitchell, J. S., MacDiarmid, A. B., and Swanson,
R. J. 2011a. The effect of male size and spermatophore characteristics
on reproduction in the Caribbean spiny lobster, Panulirus argus.
New Frontiers in Crustacean Biology, 15: 69 – 84.
Butler, M. J., IV, Mojica, A. M., Sosa-Cordero, E., Millet, M.,
Sanchez-Navarro, P., Maldonado, M. A., and Posada, J.et al. 2010.
Patterns of spiny lobster (Panulirus argus) postlarval recruitment
in the Caribbean: a CRTR Project. Proceedings of the 62nd Gulf
and Caribbean Fisheries Institute, 62: 361 –369.
Butler, M. J., IV, Paris, C. B., Goldstein, J. S., Matsuda, H., and Cowen, R.
K. 2011b. Behaviour constrains the dispersal of long-lived spiny
lobster larvae. Marine Ecology Progress Series, 422: 223– 237.
Carver, A. M., Wolcott, T. G., Wolcott, D. L., and Hines, A. H. 2005.
Unnatural selection: effects of a male-focused size-selective fishery
on reproductive potential of a blue crab population. Journal of
Experimental Marine Biology and Ecology, 319: 29 – 41.
Chavez, E. A. 2009. Potential production of the Caribbean spiny lobster
(Decapoda, Palinura) fisheries. Crustaceana, 82: 1393– 1412.
Chubb, C. F. 2000. Reproductive biology: issues for management. In
Spiny Lobsters: Fisheries and Culture, 2nd edn, pp. 245– 276. Ed.
by B. F. Phillips, and J. Kittaka. Blackwell Scientific Press, Oxford.
Cox, C., and Hunt, J. H. 2005. Change in size and abundance of
Caribbean spiny lobsters Panulirus argus in a marine reserve in the
Florida Keys National Marine Sanctuary, USA. Marine Ecology
Progress Series, 294: 227– 239.
CRFM. 2013. CFRM Annual Report April 1 2012-March 31 2013.
CRFM Administrative Report. 43pp.
Effect of parental size in the Caribbean Spiny lobster
Dewsbury, D. A. 1982. Ejaculate cost and male choice. The American
Naturalist, 119: 601– 610.
Ehrhardt, N. M. 2005. Population dynamic characteristics and sustainability mechanisms in key Western Central Atlantic spiny lobsters,
Panulirus argus, fisheries. Bulletin of Marine Science, 76: 501– 525.
Ehrhardt, N. M., Puga, R., and Butler, M. J., IV. 2010. Large ecosystem
dynamics and fishery management concepts: the Caribbean spiny
lobster, Panulirus argus, fisheries. In Towards Marine
Ecosystem-Based Management in the Wider Caribbean. Ed. by L.
Fanning, R. Mahon, and P. McConney. Amsterdam University
Press, Amsterdam.
Galeotti, P., Rubolini, D., Fea, G., Ghia, D., Nardi, P. A., Gheradi, F., and
Fasola, M. 2006. Female freshwater crayfish adjust egg and clutch size
in relation to multiple male traits. Proceedings of the Royal Society of
London, Series B, 273: 1105 – 1110.
Garcia, H. N., Kolding, J., Rice, J., Rochet, M. J., Zhou, S., Arimoto, T.,
and Beyer, J. T.,et al. 2012. Reconsidering the Consequences of
Selective Fisheries. Science, 335: 1045– 1047.
Giminez, L., and Anger, K. 2001. Relationships among salinity, egg size,
embryonic development, and larval biomass in the estuarine crab
Chasmagnathus granulata Dana, 1851. Journal of Experimental
Marine Biology and Ecology, 260: 241– 247.
Gosselin, T., Sainte-Marie, B., and Bernatchez, L. 2003. Patterns
of sexual cohabitation and female ejaculate storage in the
American lobster (Homarus Americanus). Behavioral Ecology and
Sociobiology, 55: 151 – 160.
Green, B. 2008. Chapter 1 Maternal effects in fish populations. Advances
in Marine Biology, 54: 1– 105.
Heath, M. R., and Speirs, D. C. 2012. Changes in species diversity and
size composition in the Firth of Clyde demersal fish community
(1927 –2009). Proceedings of the Royal Society, Series B, 279:
543– 552.
Holthius, L. B. 1991. FAO Species Catalogue. Marine Lobsters of the
World. An annotated and illustrated catalogue of species of interest
to fisheries known to date. FAO Fish Synopsis, 125.
Jennings, S., and Lock, J. M. 1996. Population and ecosystem effects of
reef fishing. In Reef Fisheries, pp. 193 – 218. Fish and Fisheries Series,
20. Chapman & Hall, New York.
Kendall, M. S., Wolcott, D. L., Wolcott, T. G., and Hines, A. H. 2001.
Reproductive potential of individual male blue crabs, Callinectes
sapidus, in a fished population: depletion and recovery of sperm
number and seminal fluid. Canadian Journal of Fisheries and
Aquatic Sciences, 58: 1168 –1177.
Kendall, M. S., Wolcott, D. L., Wolcott, T. G., and Hines, A. H. 2002.
Influence of male size and mating history on sperm content of ejaculates of the blue crab Callinectes sapidus. Marine Ecology Progress
Series, 230: 235 – 240.
Kuparinen, A., and Hutchings, J. A. 2012. Consequences of
fisheries-induced evolution for population productivity and recovery potential. Proceedings of the Royal Society of London, Series B,
279: 2571– 2579.
Kuris, A. M. 1991. A review of patterns and causes of crustacean brood
mortality. In Crustacean Egg Production, pp 117 – 143. Ed. by A.
Wenner, and A. M. Kuris. CRC Press, Boca Raton, FL.
Liddy, G. C., Phillips, B. F., and Maguire, G. B. 2003. Survival and growth
of instar 1 phyllosoma of the western rock lobster, Panulirus cygnus,
starved before or after periods of feeding. Aquaculture International,
11: 53 – 67.
Lipcius, R. N., Edwards, M. L., Hernkind, W. F., and Waterman, S. A.
1983. In situ mating behaviour of the spiny lobster Panulirus
argus. Journal of Crustacean Biology, 3: 217 – 222.
Lyons, W. G., Barber, D. G., Foster, S. M., Kennedy, F. S., and Milano,
G. R. 1981. The spiny lobster of the middle and upper Florida
Keys: population structure, seasonal dynamics and reproduction.
Florida Marine Research Publications, 38: 1– 45.
i123
MacDiarmid, A. B., and Butler, M. J., IV. 1999. Sperm economy and
limitation in spiny lobsters. Behavioural and Ecological
Sociobiology, 46: 14 – 24.
MacDiarmid, A. B., and Sainte-Marie, B. 2006. Reproduction. In
Lobsters: Biology, Management, Aquaculture and Fisheries, pp
45 – 77. Ed. by B. Phillips. Blackwell Publishing, Oxford, UK.
Marshall, D. J., Heppell, S. S., Munch, S. B., and Warner, R. R. 2010. The
relationship between maternal phenotype and offspring quality: do
older mothers really produce the best offspring? Ecology, 91:
2862– 2873.
Maxwell, K. E., Matthews, T. R., Bertelsen, R. D., and Derby, C. D. 2009.
Using age to evaluate reproduction in the Caribbean spiny lobster,
Panulirus argus in the Florida Keys and Dry Tortugas, United
States. New Zealand Journal of Marine and Freshwater Research,
43: 139– 149.
Moland, E., Olsen, E. M., and Stenseth, N. C. 2010. Maternal influences
on offspring size variation and variability in wild European lobster
Homarus gammarus. Marine Ecology Progress Series, 400: 165 – 173.
Mota-Alves, M. I., and Pavia, M. P. 1976. Frequencia de acasalamentos
en langostas do genero Panulirus White (Decapoda, Palinuridae).
Arquiveria Ciênces Marina, 16: 61 – 63.
Powell, G. C., James, K. E., and Hurd, C. L. 1974. Ability of male king
crab, Paralithodes camtschatica to mate repeatedly, Kodiak, Alaska,
1973. Fishery Bulletin US, 72: 171 – 179.
Roberts, C. M., and Polunin, N. V. C. 1991. Are marine reserves effective
in management of reef fisheries? Review in Fish Biology and
Fisheries, 1: 65– 91.
Robertson, D., and Butler, M. J., IV. 2013. Mate choice and sperm limitation in the spotted spiny lobster Panulirus guttatus. Marine Biology
Research, 9: 69– 76.
Rowe, S., and Hutchings, J. A. 2003. Mating systems and the conservation of commercially exploited marine fish. Trends in Ecology and
Evolution, 18: 567 – 572.
Sato, T. 2012. Impacts of large male selective harvesting on reproduction: illustration with large decapod crustacean resources.
Aqua-BioScience Monographs, 5: 67 – 102.
Sato, T., and Goshima, S. 2007. Effects of risk of sperm competition,
female size, and male size, on the number of ejaculated sperm in
the stone crab Hapalogaster dentata. Journal of Crustacean
Biology, 27: 570 – 575.
Sato, T., Yoseda, K., Okuzawa, K., and Suzuki, N. 2010. Sperm limitation: possible impacts of large male selective harvesting on reproduction of the coconut crab Birgus latro. Aquatic Biology, 10: 23– 32.
Saul, S. 2004. A review of the literature and life history study of the
Caribbean spiny lobster, Panulirus argus. Caribbean Southeast
Data Assessment Review Workshop Report, SEDAR-DW-05.
Sustainable Fisheries Division Contribution No. SFD- 2004-048.
Sheldon, B. C. 2000. Differential allocation: tests, mechanisms and
implications. Trends in Ecology and Evolution, 15: 397– 402.
Skinner, A. M. J., and Watt, P. J. 2007. Strategic egg allocation in the
zebra fish, Danio rerio. Behavioural Ecology, 18: 905 – 909.
Talbot, P., and Summers, R. G. 1978. The structure of sperm from
Panulirus, the spiny lobster, with special regard to the acrosome.
Journal of Ultrastructural Research, 64: 341– 351.
Tram, U., and Wolfner, M. F. 1999. Male seminal fluid proteins are essential for sperm storage in Drosophila melanogaster. Genetics, 153:
837– 844.
Venturelli, P. A., Murphy, C. A., Shuter, B. J., Johnston, T. A., Van
Coeverden de Groot, P. J., Boag, P. T., and Casselman, J. M.et al.
2010. Maternal influences on population dynamics: evidence from
an exploited freshwater fish. Ecology, 91: 2003 – 2012.
Wedell, N., Gage, M. J. G., and Parker, G. A. 2002. Sperm competition,
male prudence and sperm-limited females. Trends in Ecology and
Evolution, 17: 313 – 320.
Handling editor: Stéphane Plourde
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i124 –i127. doi:10.1093/icesjms/fsu163
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Phyllosomata associated with large gelatinous zooplankton:
hitching rides and stealing bites
Richard O’Rorke 1*, Shane D. Lavery1,2, Miao Wang 2, Ramón Gallego 1, Anya M. Waite 3,4,
Lynnath E. Beckley 5, Peter A. Thompson 6, and Andrew G. Jeffs 2
1
School of Biological Sciences, University of Auckland, Auckland, New Zealand
Leigh Marine Laboratory, Institute of Marine Science, University of Auckland, Warkworth, New Zealand
3
School of Environmental Systems Engineering and the Oceans Institute, University of Western Australia, Crawley, WA, Australia
4
Alfred Wegener Institute, Helmholz Centre for Polar and Marine Research, Bremerhaven, Germany
5
School of Veterinary and Life Sciences, Murdoch University, Murdoch, WA, Australia
6
Australian Commonwealth Scientific Industrial and Research Organisation, Hobart, TAS, Australia
2
*Corresponding author: e-mail: [email protected]
O’Rorke, R., Lavery, S. D., Wang, M., Gallego, R., Waite, A. M., Beckley, L. E., Thompson, P. A., and Jeffs, A. G. Phyllosomata associated with
large gelatinous zooplankton: hitching rides and stealing bites. – ICES Journal of Marine Science, 72: i124–i127.
Received 25 June 2014; revised 24 August 2014; accepted 27 August 2014; advance access publication 3 October 2014.
During a zooplankton survey 350 km off the coast of Western Australia, we captured a large and robust zooid of a salp (Thetys vagina), to which six
late stage larvae (phyllosomata) of the western rock lobster (Panulirus cygnus) were attached. High-throughput sequencing analyses of DNA
extracts from midgut glands of the larvae confirmed that each phyllosoma had consumed mainly salp tissue (x ¼ 64.5% + 15.9 of DNA reads).
These results resolve long-standing conjecture whether spiny lobster phyllosomata attach to large gelatinous hosts to feed on them.
Keywords: food web, Indian Ocean, PCR, phyllosoma, plankton, pyrosequencing, salp, spiny lobster, Thetys, trophic level.
Introduction
For many years, there has been conjecture that spiny lobster (Family
Palinuridae) phyllosoma larvae might attach themselves to large gelatinous zooplankton for assistance with locomotion and possibly to
feed upon them (Jeffs, 2007). This hypothesis was partly extrapolated from in situ observations of phyllosomata of slipper lobsters
(Family Scyllaridae; a family related to spiny lobsters) attached to,
and most probably consuming, large cnidarian medusae (Shojima,
1963; Thomas, 1963; Hernkind et al., 1976; Ates et al., 2007) as well
as captive feeding trials where the phyllosomata were capable of
clinging to, and consuming the entirety of a wide variety of jellyfish
medusae (Wakabayashi et al., 2012). Spiny lobster phyllosomata
have never been directly observed in the wild due to their small
size, transparent morphology, and far offshore habitat (extending
to over 1500 km offshore, Phillips et al., 1979), which makes direct
observations of the nature of any association with large gelatinous
zooplankton unlikely. Phyllosomata can grip and manipulate other
animals by the use of dactyls at the end of their limbs (Wakabayashi
et al., 2012), and on Panulirus cygnus the dactyl of the second pereiopod is particularly large and suited to this purpose (Braine et al.,
1979: refer video in supplementary files). However, their grip on
their hosts is unlikely to survive sampling, because large gelatinous
pelagic organisms, such as colonial radiolaria and siphonophores,
often disintegrate when sampled in zooplankton net tows. This
means that even if phyllosomata do adhere to them, it will be unlikely to observe them doing so in net hauls.
In this study, we report the capture by net tow of a large and robust
zooid of a salp (Thetys vagina), to which six late stage phyllosomata
were attached (Supplementary Figure 1). To our knowledge, this is
the first record of spiny lobster phyllosomata being captured while
remaining attached to another animal. To determine if the phyllosomata were feeding on the salp, high-throughput sequencing analyses
were performed on DNA extracts from the midgut glands of phyllosomata carefully removed from the salp zooid. DNA methods provide
a highly reliable technique for identifying the diet of pelagic predators
when in situ observations are not possible (O’Rorke et al., 2012a, b).
# International Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i125
Phyllosomata associated with large gelatinous zooplankton
Table 1. Phyllosomata recovered from Thetys salp.
Phyllosoma
1
2
3
4
5a
6a
Size (mm)
27.25
18
18
16.75
19.5
17.5
Stage
9
7
7
7
7
6
Length refers to the distance from the anterior margin of the cephalic shield
(cephalothorax) from between the eyestalks to the posterior tip of the pleon.
a
DNA from these phyllosomata did not amplify.
Methods, results, and discussion
A zooplankton net tow of surface water taken on 29 August 2011
between 02:00 and 02:05 AWST at 31.098S, 111.438E 350 km off
the Western Australian coast following the methods of Wang et al.
(2014) recovered a salp zoid to which six late stage Panulirus
cygnus phyllosomata were attached (Table 1). The phyllosoma
were all in a typical clinging posture adopted by phyllosoma while
feeding with the ventral surface on the cephalic shield, which contains the mouthparts, pressed against the outer surface of the salp
zooid. The salp was 115 × 55 mm at its longest and widest dimensions and was identified by Dr Lisa-ann Gershwin (CSIRO-Marine &
Atmospheric Research, Australia) to be Thetys vagina, based on its
unique morphology. DNA was ChelexTM (Bio-Rad) extracted
from tissue of the salp, and a short 18S rDNA region was PCRamplified and sequenced using the Uni1304F and Uni1670R
primers and protocol from Larsen et al. (2005). The sequenced
Thetys vagina (Genbank accession KM360161) locally aligns to
other salp sequences on the NCBI database. However, it does not
have a perfect match on the database and was only a 91.2% match
to that of the only other 18S sequence of this species previously
reported on Genbank (from the NW Atlantic; Govindarajan et al.,
2011). From examination of the outer surface of the salp upon
landing on the research vessel, it was not possible to determine
whether loss of salp tissue was associated directly with predation
from individual phyllosoma due to the irregularity in the surface
of the salp. Therefore, a DNA approach was adopted to determine
if the phyllosomata had consumed salp material. DNAwas extracted
from the midguts of the phyllosomata using 31 gauge syringes
(O’Rorke et al., 2013a) and also extracted from the codend of the
net from which the salp was recovered, as well as from PCR-grade
water-only negative control taken from water drained from the
codend. PCR amplifications of the v7 and v9 regions of the 18S
rDNA were performed according to O’Rorke et al., (2012a,b) and
sequenced at Macrogen (Korea) on the 454 GS platform using
Titanium chemistry.
Sequenced DNA reads were processed in MOTHUR (Schloss
et al., 2009) using methods outlined in O’Rorke et al. (2013b) and
it was found that DNA was successfully amplified from four of
these six phyllosomata and each of these phyllosomata contained
a majority of DNA fragment sequence reads that were identified as
belonging to salp (Thaliacea) (x + SE ¼ 64.5% + 15.9) (Figure 1).
DNA failing to amplify in approximately one-third of phyllosomata
is not unusual and could be due to these two phyllosomata only recently attaching to their host, perhaps after capture in the net
(O’Rorke et al., 2012). However, it is very unusual to have the
same OTU detected in such abundance in every phyllosoma
sampled from a single site (Suzuki et al., 2008; Chow et al., 2010;
O’Rorke et al., 2012a, b; 2013b). The difference between this study
and previous studies is that the phyllosomata were not randomly
distributed throughout the plankton, but were all adhering to a
single animal.
Two loci of the 18S rRNA gene were amplified and sequenced,
which both showed the same distribution of DNA sequence reads
(Figure 1), which indicates that the dominance of thaliacean reads
was not an artefact of primer bias. In addition to the gut contents
of the phyllosomata being sequenced, a sample of cod-end water
was also collected to control for the remote possibility that we
were detecting DNA that had been passively ingested by the phyllosoma. For this, 5 mL of mixed codend water was passed through a
0.5-mm syringe filter (Millipore) into 10 mL of pre-chilled EtOH
and stored at 2208C. A 0.5-mm filter was used because this is the
exclusion size of the filter press of late stage phyllosomata (Smith
et al., 2009; Simon et al., 2012). Perhaps surprisingly, this control
contained no salp DNA, but was dominated by DNA from colonial
radiolaria. Colonial radiolaria are a significant component of the
zooplankton assemblage in the East Indian Ocean, although they
do disintegrate when sampled by nets (Stemmann et al., 2008),
and it is therefore not surprising that their DNA was detected in
the codend sample.
After salp DNA, the second most abundant organisms in each
phyllosoma were not identical between samples, with phyllosomata
2 and 4 containing 25% of DNA from an anthozoan (Cerianthia)
and a bony fish, respectively. This suggests that each phyllosoma was
either feeding on a different prey item before encountering the salp
or that they were also feeding on other food sources while attached to
the salp. Captive phyllosomata have been observed to feed on a range
of gelatinous zooplankton such as fish larvae, chaetognaths, ctenophores, and cnidarians (Mitchell, 1971; Kittaka, 1997), which are all
a significant part of the plankton assemblage of the East Indian
Ocean (Säwström et al., 2014). In a previous experiment, it has
been determined that P. cygnus phyllosomata are preferential
feeders that consume arrow worms over either salps or krill in a preychoice experimental design (Saunders et al., 2012). This, along with
the present study, indicates that P. cygnus phyllosomata are opportunistic predators, attaching themselves and feeding on large prey
items, such as salps when they are available. This is consistent
with the fact that salps are relatively poor in lipids, protein, and available energy (Wang and Jeffs, 2013; Wang et al., 2013), but they are
large and represent a high level of readily digestible biomass that is
a guaranteed meal in the absence of better prey options. P. cygnus
larvae may not consume the entirety of their hosts, unlike some
slipper lobster larvae (Wakabayashi et al., 2012), because it is not
clear that this is a good strategy for larvae in oceanic waters with
low productivity. However, T. vagina have considerable regenerative
properties (Hirose et al., 2005) and it would be valuable to learn if
salp tissue can regenerate at a rate that matches consumption by
phyllosomata. It has been observed that even after T. vagina
zooids are emptied into “barrels” to become the homes of
Phronimids (a Hyperiid amphipod), that the salp tissue continues
to regenerate (Hirose et al., 2005). Salps have very rapid growth
rates, with several species increasing their body length by over 5%
per hour, and some at a staggering 20% per hour (Madin and
Deibel, 1998). Unfortunately, there are no data on the growth
rates of Thetys, but they are the largest known salp (.300 mm
length, Nakamura and Yount, 1958) and they can rapidly increase
in numbers in response to bursts in productivity (Iguchi, 2006).
Furthermore, phyllosomata that attach themselves to large animals
that undergo vertical migration would be able to track the diurnal
changes in prey density and therefore save considerable energy.
i126
R. O’Rorke et al.
Figure 1. Distribution of reads across samples. The relative proportions of DNA sequence reads for the four phyllosomata that yielded at least 2000
reads for the (a) 18S v7 and (b) 18S v9 rRNA loci. A sample of water was also taken from the net codend to control against the possibility that
phyllosomata guts contain soluble DNA ingested after capture. Both loci are concordant and show that phyllosomata contain a majority of
Thaliacea (salp) DNA. In contrast, the codend water mostly contains DNA from colonial radiolarian (Polycystinea).
This is important because saving energy has considerable implications for the abilities of the spiny lobster post-larvae (pueruli)
to subsequently migrate back onshore (Wilkin and Jeffs, 2011;
Fitzgibbon et al., 2013).
opportunistic feeding strategy that enables phyllosomata to
survive in sparsely populated and oligotrophic oceanic waters.
Importantly, by attaching to a large vertically migrating animal,
the phyllosomata are able to conserve energy by not swimming, as
well as having access to a meal of relatively magnificent size.
Conclusion
The unique discovery of spiny lobster phyllosomata attached and
feeding on a large gelatinous zooplankter resolves conjecture that
these animals share this behaviour of the larvae of their close relatives, the slipper lobsters. Sequencing of the gut contents of these
spiny lobster phyllosomata confirms that they not only adhere to
but also ingest the tissues of their gelatinous hosts. Locating and
attaching themselves to large prey items could be an effective
Acknowledgements
Dr Lisa-ann Gershwin (CSIRO-Marine & Atmospheric Research,
Australia) for salp ID. The RV Southern Surveyor was provided by the Australian Marine National Facility under Grant
SS05-2010. This project was funded by a grant from the Fisheries
Research and Development Council of Australia under FRDC
PROJECT NUMBER: 2010/047. The study was conducted with
Phyllosomata associated with large gelatinous zooplankton
the financial support of the Australian Research Council’s Industrial Transformation Research Hub funding scheme (project
number IH 120,100,032).
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
References
Ates, R., Lindsay, D., and Sekiguchi, H. 2007. First record of an association between a phyllosoma larva and a Prayid siphonophore.
Plankton and Benthos Research, 2: 67 – 69.
Braine, S. J., Rimmer, D. W., and Phillips, B. F. 1979. An illustrated key
and notes on the phyllosoma stages of the western rock lobster
Panulirus cygnus. CSIRO Australian Division of Fisheries and
Oceanography, Perth, 10 pp.
Chow, S., Suzuki, S., Matsunaga, T., Lavery, S., Jeffs, A. G., and
Takeyama, H. 2010. Investigation on natural diets of larval marine
animals using peptide nucleic acid-directed polymerase chain reaction clamping. Marine Biotechnology, 13: 305– 313.
Fitzgibbon, Q. P., Jeffs, A. G., and Battaglene, S. C. 2013. The Achilles
heel for spiny lobsters: the energetics of the non-feeding post-larval
stage. Fish and Fisheries, 15: 312 –326
Govindarajan, A. F., Bucklin, A., and Madin, L. P. 2011. A molecular
phylogeny of the Thaliacea. Journal of Plankton Research, 33:
843– 853.
Hernkind, W, Halusky, J., and Kanciruk, P 1976. A further note on phyllosoma larvae associated with medusae. Bulletin of Marine Science,
26: 110 –112.
Hirose, E., Aoki, M. N., and Nishikawa, J. 2005. Still alive? Fine structure
of the barrels made by Phronima (Crustacea: Amphipoda). Journal
of the Marine Biological Association of the UK, 85: 1435– 1439.
Iguchi, N. 2006. Horizontal distribution of Thetys vagina Tilesius
(Tunicata, Thaliacea) in the Japan Sea during spring 2004. Journal
of Plankton Research, 28: 537 – 541.
Jeffs, A. G. 2007. Revealing the natural diet of the phyllosoma larvae of
spiny lobster. Bulletin of Fisheries Research Agency (Japan), 20:
9 – 13.
Kittaka, J. 1997. Culture of larval spiny lobsters: a review of work done in
northern Japan. Marine and Freshwater Research, 48: 923– 930.
Larsen, J., Frischer, M., Rasmussen, L., and Hansen, B. 2005. Single-step
nested multiplex PCR to differentiate between various bivalve larvae.
Marine Biology, 146: 1119 –1129.
Madin, L. P., and Deibel, D. 1998. Feeding and energetics of Thaliacea. In
The Biology of Pelagic Tunicates, pp. 81 – 103. Ed. by Q. Bone.
Oxford University Press, New York.
Mitchell, J. 1971. Food preferences, feeding mechanisms, and related behavior in phyllosoma larvae of the California spiny lobster, Panulirus
interruptus (Randall). San Diego State College. MSc thesis.
Nakamura, E. L., and Yount, J. 1958. An unusually large salp. Pacific
Science, 12: 181.
O’Rorke, R., Jeffs, A. G., Fitzgibbon, Q. P., Chow, S., and Lavery, S.
2013a. Extracting DNA from whole organism homogenates and
the risk of false positives in PCR based diet studies: a case study
using spiny lobster larvae. Journal of Experimental Marine Biology
and Ecology, 441: 1 – 6.
O’Rorke, R., Lavery, S., Chow, S., Takeyama, H., Tsai, P., Beckley, L. E.,
Thompson, P. A., et al. 2012a. Determining the diet of larvae of
western Rock Lobster (Panulirus cygnus) using high-throughput
DNA sequencing techniques. PLoS ONE, 7: 1 – 10.
O’Rorke, R., Lavery, S. D., Wang, M., Nodder, S. D., and Jeffs, A. G.
2013b. Determining the diet of larvae of the red rock lobster (Jasus
i127
edwardsii) using high-throughput DNA sequencing techniques.
Marine Biology, 161: 551 – 563.
O’Rorke, R., Lavery, S., and Jeffs, A. 2012b. PCR enrichment techniques
to identify the diet of predators. Molecular Ecology Resources, 12:
5 – 17.
Phillips, B. F., Brown, P., Rimmer, D., and Reid, D. 1979. Distribution
and dispersal of the phyllosoma larvae of the western rock lobster,
Panulirus cygnus, in the South-Eastern Indian Ocean. Australian
Journal of Marine and Freshwater Research, 30: 773 –783.
Saunders, M. I., Thompson, P. A., Jeffs, A. G., Säwström, C., Sachlikidis,
N., Beckley, L. E., and Waite, A. M. 2012. Fussy feeders: phyllosoma
larvae of the western rock lobster (Panulirus cygnus) demonstrate
prey preference. PLoS ONE, 7: e36580.
Säwström, C., Beckley, L. E., Saunders, M. I., Thompson, P. A., and
Waite, A. M. 2014. The zooplankton prey field for rock lobster
phyllosoma larvae in relation to oceanographic features of the
south-eastern Indian Ocean. Journal of Plankton Research, 36:
1003– 1016.
Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M.,
Hollister, E. B., Lesniewski, R. A., et al. 2009. Introducing mothur:
open-source, platform-independent, community-supported software for describing and comparing microbial communities.
Applied and Environmental Microbiology, 75: 7537– 7541.
Shojima, Y. 1963. Scyllarid phyllosoma’s habit of accompanying the
jellyfish. Bulletin of the Japanese Society of Scientific Fisheries
(Japan), 29: 349 – 353.
Simon, C. J., Carter, C. G., and Battaglene, S. C. 2012. Development and
function of the filter-press in spiny lobster, Sagmariasus verreauxi,
phyllosoma. Aquaculture, 370 – 371: 68 – 75.
Smith, G., Hall, M. W., and Salmon, M. 2009. Use of microspheres, fresh
and microbound diets to ascertain dietary path, component size, and
digestive gland functioning in phyllosoma of the spiny lobster
Panulirus ornatus. New Zealand Journal of Marine and Freshwater
Research, 43: 205– 215.
Stemmann, L., Youngbluth, M., Robert, K., Hosia, A., Picheral, M.,
Paterson, H., Ibanez, F., et al. 2008. Global zoogeography of fragile
macrozooplankton in the upper 100–1000 m inferred from the underwater video profiler. ICES Journal of Marine Science, 65: 433–442.
Suzuki, N., Hoshino, K., Murakami, K., Takeyama, H., and Chow, S.
2008. Molecular diet analysis of phyllosoma larvae of the Japanese
spiny lobster Panulirus Japonicus (Decapoda: Crustacea). Marine
Biotechnology, 10: 49– 55. 7 pp.
Thomas, L. R. 1963. Phyllosoma Larvæ associated with Medusæ.
Nature, 198: 208 –208.
Wakabayashi, K., Sato, R., Hirai, A., Ishii, H., Akiba, T., and Tanaka, Y.
2012. Predation by the phyllosoma larva of Ibacus novemdentatus on
various kinds of venomous jellyfish. The Biological Bulletin, 222:
1 – 5.
Wang, M., and Jeffs, A. G. 2013. Nutritional composition of potential
zooplankton prey of spiny lobster larvae: a review. Reviews in
Aquaculture, 5: 1 – 30.
Wang, M., O’Rorke, R., Nodder, S. D., and Jeffs, A. G. 2013. Nutritional
composition of potential zooplankton prey of the spiny lobster phyllosoma (Jasus edwardsii). Marine and Freshwater Research, 65:
337– 349.
Wang, M., O’Rorke, R., Waite, A. M., Beckley, L. E., Thompson, P. A.,
and Jeffs, A. G. 2014. Fatty acid profiles of phyllosoma larvae of
western rock lobster (Panulirus cygnus) in cyclonic and anticyclonic
eddies of the Leeuwin Current off Western Australia. Progress in
Oceanography, 122: 153– 162.
Wilkin, J., and Jeffs, A. G. 2011. Energetics of swimming to shore in the
puerulus stage of a spiny lobster: can a postlarval lobster afford the
cost of crossing the continental shelf? Limnology & Oceanography:
Fluids & Environments, 1: 163–175.
Handling editor: Stéphane Plourde
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i128 –i138. doi:10.1093/icesjms/fsu237
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Effects of population density and body size on disease
ecology of the European lobster in a temperate marine
conservation zone
Charlotte E. Davies1*, Andrew F. Johnson 2, Emma C. Wootton 1, Spencer J. Greenwood 3,
K. Fraser Clark 3, Claire L. Vogan4, and Andrew F. Rowley1
1
Department of Biosciences, College of Science, Swansea University, Swansea SA2 8PP, UK
Center for Marine Biodiversity and Conservation, Scripps Institution of Oceanography, 9500 Gilman Drive, La Jolla, CA 92093-0202, USA
3
Department of Biomedical Sciences and AVC Lobster Science Centre, Atlantic Veterinary College, University of Prince Edward Island,
550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3
4
College of Medicine, Swansea University, Swansea SA2 8PP, UK
2
*Corresponding author: tel: +44 7814 161356; e-mail: [email protected]
Davies, C. E., Johnson, A. F., Wootton, E. C., Greenwood, S. J., Clark, K. F., Vogan, C. L., and Rowley, A. F. Effects of population
density and body size on disease ecology of the European lobster in a temperate marine conservation zone. – ICES
Journal of Marine Science, 72: i128 – i138.
Received 28 July 2014; accepted 2 December 2014; advance access publication 29 December 2014.
Marine conservation zones (MCZs) are a form of spatial marine management, increasingly popular since the move towards ecosystem-based fisheries management. Implementation, however, is somewhat contentious and as a result of their short history, their effects are still widely unknown
and understudied. Here, we investigate the population and health of the European lobster (Homarus gammarus) in the Lundy Island Marine
Conservation Zone, Bristol Channel, UK. Using the fished refuge zone (RZ) as a control area, catch per unit effort was calculated for both the
no-take zone (NTZ) and RZ and binomial logistic regression models were used to examine the effects of site, sex, landing size, and loss of chelae
on the probability of shell disease and injury presence in individuals. Lobsters were also tested for the causative agent of gaffkaemia, Aerococcus
viridans var. homari, and white spot syndrome virus (WSSV). The analysis revealed a higher lobster density and larger lobsters in the NTZ compared
with the RZ. Shell disease was present in 24% of lobsters and the probability of shell disease occurrence increased notably for individuals over the
minimum landing size (MLS) of 90 mm carapace length. Shell disease was also more prevalent in lobsters displaying injury, and in males. Injury was
present in 33% of lobsters sampled and prevalence was higher in lobsters in the NTZ compared with the RZ, and in lobsters .MLS. Aerococcus
viridans var. homari was detected in ,1% of individuals, but WSSV was absent from all sampled lobsters. Overall, the study demonstrates both
positive and potentially negative effects of NTZs, methods for effective non-lethal sampling of disease agents, and highlights the need for more
comprehensive, long-term monitoring within highly protected MCZs, both before and after implementation.
Keywords: gaffkaemia, health, Lundy Island, marine monitoring, MPA, no take zone, pathogen entry, shell disease, WSSV.
Introduction
Overfishing persists in many of the world’s oceans. Fish and invertebrate stocks have been overexploited, leaving some species in
dangerous decline (Kleisner et al., 2013), and others in complete collapse (Myers et al., 1997). To protect biodiversity, commercial stocks,
and to aid in the recovery of declining populations, establishment of
marine reserves (no-take zones, NTZs), or marine protected areas
(MPAs) are seen as a new paradigm for spatial management (Kaiser,
2005). These areas of conservation are aimed at protecting habitats
and species and in some cases have been of significant benefit to
local fisheries worldwide (Rosenberg, 2003). The use of MPAs conforms to an ecosystem-based approach to management, as often,
these reserves have benefits for multiple species and ecosystems
(Botsford et al., 2003; Gaines et al., 2010).
# International Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i129
Effects of population density and body size
While protecting habitats or species is often the primary objective of areas closed to fishing, it is of paramount importance to evaluate the success of MPA implementation (Pomeroy et al., 2005). In
general, surveys and monitoring are expensive, and as a result, the
majority of commercial target species are currently inadequately
assessed worldwide (Costello et al., 2012). To provide a “fished”
or “non-protected” comparison, it is necessary to establish a baseline before implementation as well as ongoing evaluations of the
same measures (Pinnegar and Engelhard, 2007). There have been
studies offering methods for evaluating the effectiveness of management of MPAs (Pomeroy et al., 2005; Sciberras et al., 2013). One
example method is the before-after control-impact design, with
data from replicated MPA and control sites both before and after
MPA designation (Fenberg et al., 2012). This has been considered
the optimal way of assessing effects of protection (Moland et al.,
2013a). In cases where there is insufficient baseline information, it
is difficult to ascertain when a healthy population reaches its threshold, both in terms of population density and individual species
health (e.g. pathogens and disease). Subsequent MPA management
measures may therefore be misguided and inappropriate (Hilborn
et al., 2004). Conflicting interests between stakeholder groups
adds additional complexity and can exacerbate the problems
associated with determining effective management strategies (Fox
et al., 2014).
Studies have shown that increased density in marine reserves may
manifest in higher levels of disease (McCallum et al., 2005; Wootton
et al., 2012; Wood et al., 2013). Higher densities are also thought to
increase injury due to more conspecific interaction (Debuse et al.,
2003; Wootton et al., 2012), which may be linked to disease transmission (Vogan et al., 1999; Whitten et al., 2014). In addition, the
number of diseases reported in marine organisms over the past
three decades has risen substantially (Harvell et al., 2004); therefore,
health monitoring of conservation areas, such as MPAs, is vital to
ensure the correct management measures are being assigned on a
case by case basis (Agardy et al., 2003; Pomeroy et al., 2005).
A previous study by Wootton et al. (2012) of the Lundy Island
NTZ, Bristol Channel, UK, demonstrated that the high densities
of European lobster (Homarus gammarus) inside the NTZ (compared with the fished Refuge Zone; RZ) may be subjecting individuals to a higher risk of shell disease. As a consequence, it was
concluded that a further survey of lobster disease ecology in the
Lundy Island NTZ was warranted. The present study extends the original survey to include non-lethal diagnostic tests for the causative
agent of gaffkaemia (a potentially fatal bacterial disease of lobsters),
and white spot disease (WSD) caused by the white spot syndrome
virus (WSSV). WSD is primarily associated with shrimp aquaculture and is rarely seen in the wild; however, it has been suggested
that this may be due to limited monitoring of wild populations
(Chapman et al., 2004).
Shell disease syndrome, also known as enzootic (or “classical”)
shell disease, is a progressive condition of crustaceans whereby
lesions develop on the cuticle and in extreme cases can totally erode
through the carapace, exposing underlying soft tissues (Vogan
et al., 2008). Commonly observed among both European and
American (Homarus americanus) lobsters (Sindermann, 1989), it
must not be confused with the more severe epizootic shell disease, responsible for significant economic losses to American lobster fisheries
(Castro et al. 2006; Wahle et al., 2009).
Gaffkaemia is the disease resulting from an infection by the
Gram-positive bacterium, Aerococcus viridans var. homari (Stewart,
1980). Also known as “red-tail” disease, it is primarily associated
with American lobsters (H. americanus) in tidal pound holding
facilities where the stress induced by high lobster density and increased temperature exacerbates the spread of the disease, resulting
in mortalities (Snesizko and Taylor, 1947). It has also been detected
in both holding facilities and in the wild in European lobsters (Wiik
et al., 1987; Stebbing et al., 2012).
The current study examined the effect of population density
on the presence of disease and injury by comparing sites both
inside and outside the Lundy Island NTZ. We tested the hypotheses
that density of H. gammarus was higher in the NTZ compared with
the fished, RZ, and that individual lobsters from the NTZ had a
higher probability of exhibiting signs of disease (shell disease, gaffkaemia, WSD) and injury. We also tested whether large individuals
were more likely to be diseased and injured than small individuals and whether females had an increased likelihood of disease
and injury than males (due to increased intermoult periods when
older, or when holding eggs). Additionally, as a breached cuticle
increases the potential risk of infection to disease (Davies et al.,
2014; Whitten et al., 2014), the hypothesis that injured individuals
were more likely to be diseased than non-injured individuals was
also tested.
Materials and methods
Study area
Lundy Island, off the North Devon coast, was Britain’s first
MPA. The waters around the island were established as a voluntary
Marine Nature Reserve (MNR) in 1971 and in 1986, it was
designated as England’s first and only statutory MNR. The MNR
consists of an RZ, where pot fisheries (for crabs and lobsters) are
authorized, but trawl and net fisheries prohibited, and in 2003,
a statutory NTZ was implemented within the existing RZ under a
byelaw from the local Sea Fisheries Committee (Devon and Severn
IFCA). Within the NTZ, all fishing, including potting, and removal
of wildlife is forbidden.
As a result of the 2009 Marine and Coastal Access Act, in 2010, the
waters around Lundy Island also became England’s first Marine
Conservation Zone (MCZ). This new designation superseded the
MNR designation and established the site as the foundation of a
new network of MPAs. In November 2013, 27 sites were designated
within the Defra marine area (UK Ministerial Orders, 2013) and
Lundy Island’s NTZ designation still remains.
Lobster collection and general observations
In May 2010 (2 d), July 2010 (3 d), and August 2011 (3 d), the
H. gammarus population was surveyed across six sampling sites (4
in RZ and 2 in NTZ; Figure 1) similar to previous Lundy Island
surveys (Hoskin et al., 2011; Wootton et al., 2012). One string of
baited commercial parlour pots (35 pots with escape gaps closed)
was deployed at each sampling site. Each string was immersed for
24 h, retrieved, emptied of all catch, rebaited, and redeployed.
Animals were sexed and six additional measurements for each individual were taken (Table 1) before they were returned to the water.
Recent, non-melanized breach of the cuticle (injuries), or claw
loss (i.e. those sustained in the pot) were not recorded. Animals
exhibiting exoskeletal abnormalities or severe shell disease were
photographed.
Surveillance of A. viridans and WSSV
For all three surveys (May 2010, July 2010, and August 2011),
haemolymph was drawn into 100% analytical grade ethanol using
i130
C. E. Davies et al.
Figure 1. Map of Lundy Island MCZ, Bristol Channel, southwest UK (adapted from http://www.lundymcz.org.uk/). Latitude and longitude
coordinates represent the MCZ boundary. Note the RZ where pot fisheries are authorized, but trawl and net fisheries are banned and the NTZ where
removal of all wildlife is prohibited. Asterisks represent the sampling sites. Inset: size frequency distributions of male (black) and female (white)
lobsters surveyed from the NTZ and RZ. Broken lines indicate MLS; CL .90 mm.
Table 1. Parameters recorded for each individual H. gammarus caught in the Lundy Island MCZ, used as predictor variables for subsequent
models.
Measurement
Description
Carapace length (CL)
Length from rear of the eye socket to the rear of the carapace
Minimum landing size (MLS)a Whether carapace length was above the 90 mm MLS as designated by The Devon
and Severn IFCA District shore Fisheries and Conservation Authority byelaw
Berried
Presence of eggs attached to pleopods on the abdomen of females
Wounds such as punctures and stress fractures to the cuticle as described in Wootton
Injurya
et al. (2012). Injuries inflicted during captivity within the pot (i.e. recent,
non-melanized breach of the cuticle) were not recorded
Missing cheliped (or dwarfed cheliped)
Claw lossa
Presence of shell disease, as well as severity, was recorded (i.e. “high” or “low”, as
Shell disease presence and
described in Wootton et al., 2012)
severitya
Other
Tagged individuals; those showing abnormalities
Measure
Continuous measure (mm)
Ordered binomial: 0 (no) or
1 (yes)
Presence vs. absence (P vs. A)
(P vs. A)
(P vs. A)
(P vs. A) Ordered trinomial: 0
(none), 1 (low), 2 (high)
Numbered and photographic
a
Parameters used for binomial logistic regression models.
23 G needles and 2 ml syringes, to assess the presence of A. viridans
var. homari, and WSSV.
DNA extraction
DNA was extracted from haemolymph (ntotal ¼ 508/1092) using an
adapted version of Ivanova et al. (2006) (see Supplementary materials) using 96-well filter plates (AcroPrep Advance 1 ml for DNA
binding; Pall Life Sciences, Southampton, UK). DNA was eluted
with water and stabilized with Tris-EDTA buffer (10×) then used
as the template for subsequent polymerase chain reactions (PCR).
Extraction was optimized to ensure detection of all pathogens
using “spiked” haemolymph samples of both the “virulent” and
“avirulent” strains of A. viridans var. homari (1 × 103, 1 × 104, 1 ×
105, and 1 × 106 colony forming units (CFU) ml21; NVI 1032 and
88B; Table 2), and a positive control from WSSV-infected shrimp.
Polymerase chain reaction
All PCRs were carried out using MangoMix (Bioline Ltd, UK), primers synthesized by Eurofins MWG Operon (Ebersberg, Germany),
and a Bio-Rad PTC-100 Peltier Thermal Cycler, before being
i131
Effects of population density and body size
Table 2. Aerococcus viridans var. homari, causative agent of gaffkaemia, isolates from H. gammarus or H. americanus used to verify primers
and positive template controls.
Isolate
NCIMB1121a
NCIMB1119, ATCC29838a
NVI 1032bc
88Bc
Identification
A. viridans var. homari
A. viridans var. homari
A. viridans var. homari
A. viridans-like coccus
Origin
H. gammarus
H. gammarus
H. americanus
H. americanus
Location, year of isolation
Harwich, England, UK, 1962
Southampton, England, UK, 1962
NVI, Norway, 1977
Nova Scotia, Canada, 1962
Virulence status
Avirulent
Virulent
Virulent
Avirulent
a
Isolate deposited in National Collection of Industrial and Marine Bacteria.
National Veterinary Institute, Norway (infected lobsters imported from St Lawrence area, Canada, 1977).
Strains supplied by Dr P. Stebbing, Cefas, Weymouth, UK.
b
c
visualized on a 1.5% agarose gel. Decapod-specific primers 143F 5′ TGCCTTATCAGCTNTCGATTGTAG-3′ and 145R 5′ -TTCAG
NTTTGCAACCATACTTCCC-3′ yielding an 848 bp amplicon (N
represents G, A, T, or C) were used to verify the quality of the
extracted DNA and the integrity of the PCR reaction (Lo, 2014).
Cycling conditions were: 4 min at 948C followed by 40 cycles of
1 min at 938C, 1 min at 558C, and 2 min at 728C, followed by
5 min at 728C.
and sites were visualized in GraphPad Prism 5.0. Differences in
size frequency of populations were tested using a two-sample
Kolmogorov–Smirnov test and Mann–Whitney U-test. To
compare catch data between zones, t-tests were used (mean +
SEM); tests were two-tailed, used a significance level of 0.05, and
tested to follow a Gaussian distribution (using the Kolmogorov–
Smirnov test) before any further analysis. Catch per unit of fishing
effort (CPUE) was calculated as the mean number of lobsters per pot.
Detection of A. viridans var. homari
Disease and injury ecology
Following preliminary evaluations of several candidate primer pairs
developed from A. viridans var. homari GenBank ID: AY707775.1
(Greenwood et al., 2005), the primers Av1F 5′ -TCGGAAACGGG
TGCTAATAC-3′ and Av2R 5′ -TAAGGTTCTTCGCGTTGCTT-3′
were chosen to detect all A. viridans var. homari (Primer 3 design
for Av ATCC29838 16S rRNA, product 837 bp); verified with PCR
of A. viridans colony boils (NCIMB1121, and NCIMB1119;
Table 2). Cycling conditions were optimized to detect as little as
1 × 103 CFU ml21: 3 min at 958C followed by 38 cycles of 30 s at
948C, 1 min 30 s at 648C (20.58C cycle21 for the first 33 cycles
and 47.58C thereafter) and 2 min at 728C, followed by 5 min at
728C. Positive samples were repeated and the PCR product was
purified using the Wizard SV Gel and PCR Clean-Up System
(Promega, Madison, WI, USA) and sequenced by Eurofins MWG
Operon (Ebersberg, Germany). Contigs from sequences were
created using the CAP3 sequence assembly programme (Huang
and Madan, 1999) and matched to positive controls using NCBI
BLAST search.
To determine whether the measured parameters (Table 1) had a significant effect on the presence of shell disease, disease severity, and
injury in the lobster population sampled, binomial logistic regression models were used (MASS library in R; R development Core
Team, 2011). The information theoretical approach was employed
for model selection and assessment of model performance
(Richards, 2005) and initial models included all the binomial parameters highlighted in Table 1 (a). To select the best model among
the entire set of initial models, i.e. the model that best described
the presence of shell disease, shell disease severity, and injury,
Akaike’s information criterion (AIC) was used (Burnham and
Anderson, 1998). The most complex models with full interaction
terms between predictor variables were run first, followed sequentially by models with all combinations of predictor variables as
full and partial interactions until a simple main effects model was
reached. Model selections were based on the lowest AIC value
(Table 3). Once selected, non-significant predictor variables were
removed to produce final, reduced, and simpler models with
increased predictive power (Zuur et al., 2009).
Fitted probability plots were used to visualize the significant
relationships inferred from the reduced models using carapace
length (CL) as the independent variable. The probability of each
of the tested response variables was calculated using the following
equation:
Detection of WSSV
Detection of WSSV was performed using an adaptation of methods
from Clark et al. (2013), using forward primer F1- WSDvp28
5′ -GTGACCAAGACCATCGAAAC-3′ and reverse primer R1WSDvp28 5′ -TGAAGTAGCCTGATCCAACC-3′ which are based
around the VP28 gene of WSSV. Cycling conditions were: 5 min
at 958C followed by 40 cycles of 15 s at 958C, 1 min at 608C, followed
by 5 min at 728C. A positive control of DNA extracted from
WSSV-infected shrimp was used to verify the integrity of the PCR
reaction.
Statistical analysis
Population ecology
For population, shell disease and injury ecology, the 2011 data were
analysed (see Wootton et al., 2012, for 2010 data analyses). So that
individuals were not double sampled (i.e. recaptured after day 1
and considered a unique individual), each was given an identifier
based on the parameters in Table 1 and any individuals sharing
the identifier were removed (a total of four individuals were
removed before analysis). Population distributions between sexes
r=
1
1 + exp−bx
where r is the probability of each response variable and bx the
estimate (slope) for the predictor variable analysed (Table 1).
Results
General population ecology
Catch data from the 2011 survey revealed that significantly more lobsters were caught per string in the NTZ compared with the RZ (p ¼
0.0105, t ¼ 3.030, d.f. ¼ 12; NTZ ¼ 40.00 + 5.58, RZ ¼ 20.25 +
3.81, Figure 2). The CPUE was 2.13 times greater in the NTZ.
A two-sample KS test concluded that the size frequency distributions were different between the NTZ and RZ; and neither followed a
i132
C. E. Davies et al.
Table 3. Full models used to predict response variables of shell disease, shell disease severity, and injury before model reduction.
Model
Model 1: shell disease
Shell disease Site + sex + injury + landing size + claw loss
AIC: 363.37
d.f. ¼ 396
Model 2: shell disease severity
Shell disease severity Site + sex + injury + landing size + claw loss
AIC: 119.22
d.f. ¼ 94
Model 3: injury
Injury site + sex + landing size + claw loss
AIC: 482.62
d.f. ¼ 397
Predictor variable
Site
Sex
Injury
Landing size
Claw loss
Site
Sex
Injury
Landing size
Claw loss
Site
Sex
Landing size
Claw loss
Estimate (slope)
0.34
1.54
1.07
21.40
0.74
0.02
20.81
20.08
0.88
0.70
1.03
0.35
20.76
20.03
p-value
0.254
7.09e208*
9.50e205*
1.72e206*
0.084
0.977
0.235
0.863
0.219
0.339
3.21e205*
0.12
,0.001*
0.93
*Significance (a ¼ 0.05)
p , 0.001). Separating by sex, the mean size of males in the
NTZ was larger than in the RZ (96.6 + 1.6 vs. 84.9 + 1.2 mm;
p , 0.001) and females were also larger in the NTZ than the RZ
(89.8 + 1.3 vs. 85.2 mm + 1.5; p ¼ 0.012). The NTZ population
comprised 60.8% “large” commercially viable lobsters (those over
90 mm minimum landing size; MLS) and the RZ only 32.1%.
It was not possible to test whether there were more ovigerous
(“berried”) females being caught in the NTZ compared with the
RZ, as the number was too small (n ¼ 3). In total, 0.75% of
females caught were berried (all from the NTZ).
Shell disease and injury ecology
Figure 2. Mean numbers of individuals caught per string in the NTZ
and RZ with 95% CI. Independent samples t-test, p ¼ 0.0105, t ¼ 3.030,
d.f. ¼ 12; NTZ ¼ 40.00 + 5.58, RZ ¼ 20.25 + 3.81 (mean + SEM).
normal distribution (p , 0.001 for NTZ; p ¼ 0.003 for RZ). In terms
of population curve shapes, the skewness values were similar (NTZ ¼
1.50, RZ ¼ 1.29), but the kurtosis values very different (NTZ ¼ 2.00,
RZ ¼ 0.45). The difference between the two populations was therefore caused by the shape (kurtosis) of the two population distributions rather than the median values (p ¼ 0.215). These results
highlight that the NTZ population had a higher number of individuals around the median size compared with the RZ, but the
spread from small to large individuals was similar between the two
populations. Lobsters caught in the NTZ were significantly larger
than those in the RZ (93.3 + 1.0 vs. 85.0 + 1.0 mm, respectively;
In the Lundy MCZ as a whole, 24% of lobsters sampled had shell
disease. In the NTZ, 28% had shell disease and in the RZ, 17%. Of
the predictor variable combinations tested, the regression model
that resulted in the best AIC was the main effects model (Model 1;
Table 3). This model showed that there was no significant effect of
site (inside NTZ vs. RZ) or claw loss (absence or presence of missing
chelae) on the presence of shell disease in the lobsters sampled
(Table 3). NTZ lobsters were therefore no more likely to be affected
by shell disease than those caught in the RZ. Removing the nonsignificant predictors of site and claw loss on shell disease presence
from the final main effects model produced a reduced model (Model
4; Table 4). Sex, injury, and landing size all had a significant effect
on the presence of shell disease. Sex had the most significant effect
on explaining the probability of shell disease presence, followed by
landing size and then injury (Table 4). Probability calculations
demonstrated that if a caught lobster was male it was 83% more
likely to have shell disease than a female. If the lobster was injured,
it was 76% more likely to have shell disease and finally, if the
lobster caught was over the MLS of 90 mm, then it was 83% more
likely to have shell disease.
Table 4. Binomial logistic regression Model 4, reduced from the full, main effects model (Model 1) used to explain the effects of variables; sex,
landing size, and injury on the presence of shell disease.
Model
Shell disease sex + injury + landing size
d.f. ¼ 398
AIC: 363.47
*Significance (a ¼ 0.05).
Predictor variable
Sex (male)
Landing size (,90 mm)
Injury (Yes)
Estimate (slope)
1.50
21.46
1.11
+ Standard error
0.28
0.29
0.27
p-value
1.18e207*
3.50e207*
3.59e205*
Effects of population density and body size
i133
Figure 3. Fitted probability plots of shell disease presence against CL, separated by sex (a), injury (b), claw loss (c), and MLS (d). Asterisks denote
significance of the predictor variable in the original full, main effects model and reduced model. The broken line in each plot represents the legal MLS
of 90 mm CL.
Figure 4. Fitted probability plots of injury presence against CL, separated
by site (a) and MLS (b). Asterisks denote significance of the predictor
variable in the original full, main effects model and reduced model. The
broken line in each plot represents the legal MLS of 90 mm CL.
Fitted probability plots, using the model (Disease CL), separated
by the significant predictor variables sex, injury, claw loss, and above
vs. below MLS allowed an in-depth examination of the relationship
between shell disease presence, size, and each of these predictor variables (Figure 3). Separation between the two lines plotted in each
graph indicated differences in the probability of disease between the
categories for each predictor variable (i.e. male vs. female, injured vs.
non-injured, claw loss vs. intact claws, and above vs. below MLS).
Separations of the predicted probability lines for each predictor
variable categories occurred at CLs of 78, 79, and 95 mm for sex,
injury, and claw loss respectively. It is noteworthy that caution
should be taken when reading probabilities towards the larger sizes
as few individuals were caught over 135 mm CL. Outlying individuals
do, however, follow the general sigmoid probability trend well within
each plot (Figure 3 and 4).
To test whether the predictor variables had a significant effect on
the severity of shell disease in infected individuals, Model 2 was run,
substituting presence vs. absence for shell disease severity (high vs.
low) as the response variable (Model 2; Table 3). However, none
of the predictor variables were significant in predicting the severity
(high vs. low) of shell disease within individuals (Table 3; p . 0.05).
Overall, 33% of lobsters sampled were injured. In the NTZ, 41%
were injured and in the RZ, 19%. To analyse the presence of injury in
individuals, we used binomial logistic regression to examine injury
as the response variable (presence vs. absence), which was tested
against the roles of claw loss, landing size, sex, and site in injury presence (Model 3, Table 3). This model showed that there was no significant effect of sex or claw loss on the presence of injury in the
lobsters sampled (Table 3). Therefore, the loss of claws or the sex
of the lobster was not significant in predicting injury. Removing
these non-significant predictors produced a reduced model
(Model 5, Table 5) in which landing size and site both had a significant effect on the presence of injury. Landing size showed the highest
significance followed by site. This indicates that lobsters over MLS,
and those found in the NTZ were more likely to be injured than
i134
C. E. Davies et al.
Table 5. Binomial logistic regression injury Model 5, reduced from the full, main effects model (Model 3) used to explain the effects of
variables; site and landing size on the presence of injury.
Model
Injury site + landing size
d.f. ¼ 399
Predictor variable
Site (NTZ)
Landing size (,90 mm)
Estimate (slope)
1.02
20.78
+ Standard error
0.25
0.23
p-value
3.43e205*
,0.001*
*Significance (a ¼ 0.05).
those under the MLS and those in the RZ (Figure 4). If a lobster was
caught inside the NTZ, it was 71% more likely to be injured than if it
was caught in the RZ; similarly, if a lobster caught was over the MLS
of 90 mm, then it was 71% more likely to be injured.
Detection of A. viridans and WSSV
Out of the 508 haemolymph samples analysed over the three surveys,
one was positive for A. viridans var. homari. The closest BLAST
match (% identity and % coverage) being 99 and 98%, respectively,
was to A. viridans strain 1030 16S rRNA gene, partial sequence
(GenBank ID: AY707775.1; Greenwood et al., 2005). The positive
individual was found in the NTZ during the August 2011 survey,
thus the total gaffkaemia coverage in the Lundy NTZ (2011)
tested, was 1.05% and in the total Lundy MCZ (2011) tested, was
0.52%. All samples tested from 2010 were negative, and all
samples tested for WSSV were negative, regardless of year.
Discussion
The lack of fishing for over 8 years has resulted in an increase in the
number of H. gammarus in the Lundy Island NTZ compared with
the RZ; the higher CPUE in the NTZ was, therefore, to be expected.
This is supported by previous studies which also found a higher
CPUE in the Lundy Island NTZ (Hoskin et al., 2011; Wootton
et al., 2012). However, in the current study, the CPUE was only
2.13 times higher in the NTZ than the RZ, compared with fivefold
higher in 2008 (Hoskin et al., 2011) and 7.7 times higher in 2010
(Wootton et al., 2012). It is well known that there is generally an
initial population expansion after fishery closure followed by movement out of the high-density territory, which is a suggestion for the
decrease in CPUE experienced in this NTZ (e.g. Abesamis and
Russ, 2005; Goñi et al., 2006, 2010; Halpern et al., 2010). The lobsters
in the Lundy Island NTZ were found to be significantly larger than
those in the RZ, a common occurrence in other studies of MPAs
(Hoskin et al., 2011; Wootton et al., 2012; Moland et al., 2013a, b).
These findings of more abundant and larger individuals in the
Lundy NTZ are likely to benefit the reproductive potential within
the MPAs and possibly neighbouring lobster fisheries in terms of
increased mating success and larval supply (Jennings, 2000; Dı́az
et al., 2011; Moland et al., 2013b).
The probability of shell disease, or shell disease severity, was not
dependent upon zone (NTZ vs. RZ), contrasting with the previous
study by Wootton et al. (2012), which found significantly more
shell disease in the Lundy NTZ, compared with the RZ, particularly
in large male lobsters. This may be explained by a number of possible
factors. The area surveyed was relatively small (8 km2), and studies using tagging systems and acoustic telemetry have shown that
H. gammarus migrate to find food and shelter, especially when an
area is highly populated (Steneck, 2006). Mark–recapture studies
have revealed that H. gammarus move several kilometres a year (e.g.
Smith et al., 2001; Agnalt et al., 2007) and new studies using ultrasonic
tracking technology have shown some individuals have home ranges
of 20 000 m or more from their burrows (Moland et al., 2011a), with
increased activity during the summer (Moland et al., 2011b) such as
those surveyed here. It must be noted that the current survey only
provide a “snap-shot” of the population of interest and dynamic environmental conditions drastically affect the behaviour of organisms;
hence, capture of lobsters will widely fluctuate on a temporal basis,
resulting in variable data collection. Therefore, lobsters sampled by
Wootton et al. (2012), and indeed Hoskin et al. (2011), may have
migrated elsewhere by August 2011, possibly from the NTZ into the
RZ and vice versa. Huserbråten et al. (2013) support this suggestion.
They found that a significant portion of H. gammarus migrated to
fishing grounds next to the reserves in which they were originally
tagged. This spillover phenomenon has also been demonstrated by
studies observing species such as the European spiny lobster,
Palinurus elephas (Goñi et al., 2006, 2010), and the squat lobster,
Pleuroncodes monodon (Roa and Bahamonde, 1993). It must also be
noted that the current study undertakes a binomial logistic regression
approach to data analysis, not utilized by Wootton et al. (2012). This
is advantageous in terms of determining detailed population ecology
by examining interactive forces and parameters at a more complex
level and is likely to more realistically reflect the status of the lobster
population sampled as it accounts for multiple factors that may determine disease.
The fact that those lobsters over the MLS were more likely to
be shell diseased was to be expected. Studies have shown that
larger lobsters have longer moult increments (therefore moult less
often) than smaller, younger lobsters (Castro and Angell, 2000),
giving more time for shell disease and lesion progression to manifest
(Smolowitz et al., 1992; Glenn and Pugh, 2006). Shell disease levels
are also usually higher in females, due to the increased moult increments when holding eggs (Glenn and Pugh, 2006). Our result, that
there was a higher probability of shell disease in males, therefore was
somewhat unexpected. This was perhaps because ,1% of females
sampled were ovigerous in this study. It could also be speculated
that males are more likely to take part in conspecific interaction
than females, due to intrasexual selection and mate protection
(Debuse et al., 2003), therefore more likely to be injured, which in
turn may develop into shell disease. Injury presence was indeed a significant factor in predicting shell disease in the current study. In
order for chitinolythic bacteria to enter and progress the development of shell disease, a breached cuticle is required so that the
chitin containing layers beneath the outermost epicuticle may be
reached (Smolowitz et al., 1992; Vogan et al., 1999, Davies et al.,
2014, Whitten et al., 2014). Injury can therefore facilitate the initiation and progression of shell disease.
Although non-significant, results suggested that lobsters with
claw loss were less likely to have shell disease than those with both
claws intact. This may be explained by studies that have shown that
decapod crustaceans experiencing claw loss moult more frequently
in an attempt to regenerate the lost appendages (Skinner and Graham,
1972). This frequent moulting will in turn temporarily rid a lobster of
shell disease (Smolowitz et al., 1992; Glenn and Pugh, 2006).
Injury presence was found to be significantly higher in those lobsters over the MLS. Species of homarid lobsters are known to be solitary, highly territorial, and establish hierarchical ranks within
i135
Effects of population density and body size
populations (Karnofsky et al., 1989; Childress and Jury, 2006; Skog,
2009); they therefore may suffer from combative injuries when faced
with intrasexual selection and shelter choice (O’Neill and Cobb,
1979; Debuse et al., 2003; Skog, 2009). This will be especially prevalent when they are larger and sexually mature (Karnofsky and Price,
1989), such as those found above the MLS. The finding that the
probability of injured lobsters in the NTZ was higher than the RZ
may be explained by the aforementioned higher density of lobsters
in the NTZ. Some studies have shown that when a population
reaches high densities, there is an increase in conspecific interaction
(Lizaso et al., 2000), competition, and therefore fighting and injury
(Castro et al., 2012). This, in conjunction with the more abundant,
and larger lobsters found in the NTZ, means that conflict would be
especially increased, leading to injury.
The increase in injured lobsters in the NTZ may increase the risk
of infection from other pathogens such as A. viridans var. homari.
This bacterium lacks invasive mechanisms and therefore is likely
to enter the lobster only through a damaged carapace, causing septicaemia that eventually results in death (Stewart, 1980). The low
prevalence of A. viridans var. homari detected in the present study
is comparable with other wild population studies, which found gaffkaemia in ,1% of H. gammarus from UK (Stebbing et al., 2012) and
Norwegian (Wiik et al., 1987) waters. It has been noted, however,
that levels of disease in wild populations are probably underestimated as infected lobsters are lethargic and therefore less likely to
enter traps (Lavallée et al., 2001). It is also important to note that
there are “virulent” and “avirulent” strains of A. viridans var.
homari (Stewart et al., 2004); therefore, perhaps the causative
agent may lie dormant in sediment until the optimum temperature
is reached (Stewart et al., 1969).
Current diagnostics for identification of gaffkaemia are
based on time-consuming methods (see Stebbing et al., 2012)
whereby haemolymph must be examined or cultured immediately. The technique developed in this study, utilizing PCRbased methods, whereby as little as 1 × 103 CFU ml21 could be
detected, has potential to be more effective since haemolymph
may be stored in ethanol for an extended period before DNA extraction and screening.
Although there were no WSSV-infected lobsters found in this
study, the European Union EC Directive 2006/88/EC states that
all decapod crustaceans are susceptible to this viral infection (EC
Directive, 2006). Recent studies have shown that both H. gammarus
(Bateman et al., 2012) and H. americanus (Clark et al., 2013) may be
susceptible to this disease. The phenomenon of over-population
and disease is common in aquaculture systems where diseases
must be monitored diligently (Pillay and Kutty, 2005), since populations grow rapidly and unlike “wild” ocean scenarios, the diseased
individuals cannot migrate elsewhere or be “fished out” (Wood
et al., 2010). Although originating in Asia where waters are
warmer, WSSV is not a tropical disease and studies have shown
WSSV infection in water temperatures as low as 158C (Guan et al.,
2003), with WSSV replication possible at 108C (Du et al., 2008).
WSSV-infected shrimp have also been found in European shrimp
farms (Stentiford and Lightner, 2011), and given the mean
summer temperature of UK waters can be as high as 17.98C
(Cefas, 2014), it is viable that WSSV may enter and reside in UK
waters. Thus, threat from this disease is plausible and reinforces
the need for monitoring, especially in an area such as the Lundy
Island NTZ, where population density is thought to have increased
the risk of injury, apertures for pathogen entry (current study), and
prevalence of shell disease (Wootton et al., 2012).
Appropriately managed MPAs have been shown to increase low
population numbers, satisfying both fishing industry and environmental stakeholders, especially those implemented and monitored
over longer periods. For example, Goñi et al. (2010) showed
density spillover of European spiny lobster (P. elephas) into adjacent
fisheries using a decade of tag re-capture data, Aburto-Oropeza et al.
(2011) recorded a 463% increase in fish biomass over a decade of no
take protection, while in Norway, designation of an MPA increased
H. gammarus CPUE by 245% over 4 years, compared with just 87%
in control areas (Moland et al., 2013a).
Of late, the subject of implementation of MPAs in the UK has
caused much controversy (Jones, 2012; Rees et al., 2013), and
often their efficacy is up for debate worldwide (Roberts et al.,
2001; Caveen et al., 2013). The Lundy Island MCZ was created
to meet unspecified conservation benefits rather than verifiable
management objectives (Hoskin et al., 2009). The lack of predesignation data from this MCZ prevents a true historical assessment of changes in population sizes or the spread and evolution
of shell disease since the fishery closure, therefore, promotes a call
for further, long-term monitoring studies to take place. This study
highlights the necessity to monitor MCZs both before and after implementation, with a requirement for regular, standardized sampling protocols to discern and monitor the health status of species
within NTZs and other types of MPA. Data collected can be used
to make predictions for future management scenarios, and ultimately better manage the area under protection. This means that
under changing environmental conditions, managers are better
equipped at predicting resultant effects on the reserve and nearby
fisheries, in some cases mitigating negative impacts such as injury,
and potential increases in disease.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
We would like to thank the skipper, Geoff Huelin, and crew of FV
“Our Jenny” for assistance during sampling, plus Devon and
Severn IFCA and Natural England for permission to sample lobsters
in the NTZ of Lundy Island. We also thank Dr Gethin Thomas, Dr
Richard Unsworth, Dr Ed Pope, Dr Kristina Hamilton, Mr Keith
Naylor, and Mr Ian Tew of Swansea University, and Lundy Island
staff for their support during sampling. We thank Dr Paul
Stebbing (Cefas, Weymouth) for kindly providing some strains
of A. viridans var. homari ; and finally, Dr Andrew Woolmer of
Salacia Marine, Dr Miranda Whitten, and Mrs Carolyn Greig of
Swansea University, Mr Adam Acorn of the AVC Lobster Science
Centre, UPEI, Dr Hilmar Hinz of IEO Baleares and Dr Rob
Thomas of Cardiff University for advice and assistance. This research was funded by SEAFISH to Swansea University and the
ERDF INTERREG IVA, Ireland–Wales programme grant—
SUSFISH (Project No. 042) to AFR. CED was part-funded by a
tuition fee bursary from Swansea University’s Colleges of Science
and Medicine. CED would like to thank the Society of Biology
and the Climate Change Consortium for Wales (C3W) for travel
grants enabling the collaboration with the AVC Lobster Science
Centre, and also the Challenger Society for Marine Science,
British Ecological Society, Society for Experimental Biology
through the Company of Biologists and Educational Charity of
John Mathews for travel grants to attend the 10th ICWL
Conference, for which this work forms a contribution.
i136
References
Abesamis, R. A., and Russ, G. R. 2005. Density-dependent spillover from
a marine reserve: long-term evidence. Ecological Applications, 15:
1798– 1812.
Aburto-Oropeza, O., Erisman, B., Galland, G. R., Mascareñas-Osorio,
I., Sala, E., and Ezcurra, E. 2011. Large recovery of fish biomass in
a no-take marine reserve. PLoS One, 6: e23601.
Agardy, T., Bridgewater, T., Crosby, M. P., Day, J., Dayton, P. K.,
Kenchington, R., Laffoley, D., et al. 2003. Dangerous targets?
Unresolved issues and ideological clashes around marine protected
areas. Aquatic Conservation: Marine and Freshwater Ecosystems,
13: 353 –357.
Agnalt, A. L., Kristiansen, T. S., and Jørstad, K. E. 2007. Growth, reproductive cycle, and movement of berried European lobsters (Homarus
gammarus) in a local stock off southwestern Norway. ICES Journal of
Marine Science, 64: 288– 297.
Bateman, K. S., Munro, J., Uglow, B., Small, H. J., and Stentiford, G. D.
2012. Susceptibility of juvenile European lobster Homarus gammarus to shrimp products infected with high and low doses of
white spot syndrome virus. Diseases of Aquatic Organisms, 100:
169– 184.
Botsford, L. W., Micheli, F., and Hastings, A. 2003. Principles for the
design of marine reserves. Ecological Applications, 13: S25 – S31.
Burnham, K., and Anderson, D. 1998. Model Selection and Inference: a
Practical Information-Theoretic Approach. Springer, New York. pp.
46 – 51.
Castro, K. M., and Angell, T. E. 2000. Prevalence and progression of shell
disease in American lobster, Homarus americanus, from Rhode
Island waters and the offshore canyons. Journal of Shellfish
Research, 19: 691– 700.
Castro, K. M., Cobb, J. S., Gomez-Chiarri, M., and Tlusty, M. 2012.
Epizootic shell disease in American lobsters Homarus americanus
in southern New England: past, present and future. Diseases of
Aquatic Organisms, 100: 149 – 158.
Castro, K. M., Factor, J. R., Angell, T., and Landers, D. F. 2006. The conceptual approach to lobster shell disease revisited. Journal of
Crustacean Biology, 26: 646– 660.
Caveen, A. J., Gray, T. S., Stead, S. M., and Polunin, N. V. 2013. MPA
policy: what lies behind the science? Marine Policy, 37: 3– 10.
Cefas. 2014. Monthly Mean Sea Temperature for Ilfracombe at
518 12′ N, 48 8′ W. Sea temperature and salinity trends monitoring
programmes, http://cefas.defra.gov.uk/our-science/observing-andmodelling/monitoring-programmes/sea-temperature-and-salinitytrends/presentation-of-results/station-27-illfracome.aspx.
Chapman, R. W., Browdy, C. L., Savin, S., Prior, S., and Wenner, E. 2004.
Sampling and evaluation of white spot syndrome virus in commercially important Atlantic penaeid shrimp stocks. Diseases of Aquatic
Organisms, 59: 179 –185.
Childress, M. J., and Jury, S. H. 2006. Behaviour. In Lobsters: Biology,
Management, Aquaculture and Fisheries, pp. 78 – 112. Ed. by B. F.
Phillips. Blackwell Publishing Ltd, Oxford.
Clark, F. K., Greenwood, S. J., Acorn, A. R., and Byrne, P. J. 2013.
Molecular immune response of the American lobster (Homarus
americanus) to the White Spot Syndrome Virus. Journal of
Invertebrate Pathology, 114: 298 – 308.
Costello, C., Ovando, D., Hilborn, R., Gaines, S. D., Deschenes, O., and
Lester, S. E. 2012. Status and solutions for the world’s unassessed
fisheries. Science, 338: 517– 520.
Davies, C. E., Whitten, M. M. A., Kim, A., Wootton, E. C., Maffeis, T. G. G.,
Tlusty, M., Vogan, C. L., et al. 2014. A comparison of the structure of
American (Homarus americanus) and European (Homarus gammarus) lobster cuticle with particular reference to shell disease susceptibility. Journal of Invertebrate Pathology, 117: 33–41.
Debuse, V., Addison, J., and Reynolds, J. 2003. Effects of breeding site
density on competition and sexual selection in the European
lobster. Behavioural Ecology, 14: 396– 402.
C. E. Davies et al.
Dı́az, D., Mallol, S., Parma, A. M., and Goñi, R. 2011. Decadal trend in
lobster reproductive output from a temperate marine protected area.
Marine Ecology Progress Series, 433: 149– 157.
Du, H., Dai, W., Han, X., Li, W., Xu, Y., and Xu, Z. 2008. Effect of low
water temperature on viral replication of white spot syndrome
virus in Procambarus clarkia. Aquaculture, 277: 149– 151.
EC Directive. 2006. Council Directive 2006/88/EC of 24 October 2006
on animal health requirements for aquaculture animals and
products thereof, and on the prevention and control of certain diseases in aquatic animals. http://eur-lex.europa.eu/LexUriServ/
LexUriServ.do?uri=OJ:L:2006:328:0014:0056:en:PDF.
Fenberg, P. B., Caselle, J. E., Claudet, J., Clemence, M., Gaines, S. D.,
Antonio Garcı́a-Charton, J., Gonçalves, E. J., et al. 2012. The
science of European marine reserves: status, efficacy, and future
needs. Marine Policy, 36: 1012 – 1021.
Fox, H. E., Holtzman, J. L., Haisfield, K. M., McNally, C. G., Cid, G. A.,
Mascia, M. B., Parks, J. E., et al. 2014. How are our MPAs doing?
Challenges in assessing global patterns in marine protected area
performance. Coastal Management, 42: 207– 226.
Gaines, S. D., White, C., Carr, M. H., and Palumbi, S. R. 2010. Designing
marine reserve networks for both conservation and fisheries management. Proceedings of the National Academy of Sciences of the
United States of America, 107: 18286– 18293.
Glenn, R. P., and Pugh, T. L. 2006. Epizootic shell disease in American
lobster (Homarus americanus) in Massachusetts coastal waters:
interactions of temperature, maturity, and intermolt duration.
Journal of Crustacean Biology, 26: 639– 645.
Goñi, R., Hilborn, R., Dı́az, D., Mallol, S., and Adlerstein, S. 2010. Net
contribution of spillover from a marine reserve to fishery catches.
Marine Ecology Progress Series, 400: 233– 243.
Goñi, R., Quetglas, A., and Reñones, O. 2006. Spillover of spiny lobsters
Palinurus elephas from a marine reserve to an adjoining fishery.
Marine Ecology Progress Series, 308: 207– 219.
Greenwood, S. J., Keith, I. R., Després, B. M., and Cawthorn, R. J. 2005.
Genetic characterization of the lobster pathogen Aerococcus viridans
var. homari by 16S rRNA gene sequence and RAPD. Diseases of
Aquatic Organisms, 63: 237 – 246.
Guan, Y., Yu, Z., and Li, C. 2003. The effects of temperature on white
spot syndrome infections in Marsupenaeus japonicus. Journal of
Invertebrate Pathology, 83: 257 – 260.
Halpern, B. S., Lester, S. E., and Kellner, J. B. 2010. Spillover from marine
reserves and the replenishment of fished stocks. Environmental
Conservation, 36: 268– 276.
Harvell, D., Aronson, R., Baron, N., Connell, J., Dobson, A., Ellner, S.,
Gerber, L., et al. 2004. The rising tide of ocean diseases: unsolved
problems and research priorities. Frontiers in Ecology and
the Environment, 2: 375– 382.
Hilborn, R., Stokes, K., Maguire, J. J., Smith, T., Botsford, L. W., Mangel,
M., and Walters, C. 2004. When can marine reserves improve fisheries management? Ocean and Coastal Management, 47: 197– 205.
Hoskin, M. G., Coleman, R. A., von Carlshausen, E., and Davis, C. M.
2011. Variable population responses by large decapod crustaceans to
the establishment of a temperate marine no-take zone. Canadian
Journal of Fisheries and Aquatic Sciences, 68: 185 – 200.
Hoskin, M. G., Coleman, R. A., and von Carlshausen, L. 2009. Ecological
effects of the lundy no-take zone: the first five years (2003– 2007).
Report to Natural England, DEFRA and WWF-UK.
Huang, X., and Madan, A. 1999. CAP3: a DNA sequence assembly
program. Genome Research, 9: 868 – 877.
Huserbråten, M. B. O., Moland, E., Knutsen, H., Olsen, E. M., André, C.,
and Stenseth, N. C. 2013. Conservation, spillover and gene flow
within a network of northern European marine protected areas.
PLOS One, 8: e73388.
Ivanova, N. V., Dewaard, J. R., and Hebert, P. D. N. 2006. An inexpensive, automation-friendly protocol for recovering high-quality
DNA. Molecular Ecology Notes, 6: 998– 1002.
Effects of population density and body size
Jennings, S. 2000. Patterns and prediction of population recovery in
marine reserves. Reviews in Fish Biology and Fisheries, 10: 209 – 231.
Jones, P. J. S. 2012. Marine protected areas in the UK: challenges in combining top-down and bottom-up approaches to governance.
Environmental Conservation, 39: 248 – 258.
Kaiser, M. J. 2005. Are marine protected areas a red herring or fisheries
panacea? Canadian Journal of Fisheries and Aquatic Sciences, 62:
1194– 1199.
Karnofsky, E. B., Atema, J., and Elgin, R. H. 1989. Field observations of
social behaviour, shelter use, and foraging in the lobster, Homarus
americanus. Biological Bulletin, 176: 239– 246.
Karnofsky, E. B., and Price, H. J. 1989. Dominance, territoriality and
mating in the lobster, Homarus americanus: a mesocosm study.
Marine and Freshwater Behaviour and Physiology, 15: 101– 121.
Kleisner, K., Zeller, D., Froese, R., and Pauly, D. 2013. Using global catch
data for inferences on the world’s marine fisheries. Fish and
Fisheries, 14: 293 – 311.
Lavallée, J., Hammell, K. L., Spangler, E. S., and Cawthorn, R. J. 2001.
Estimated prevalence of Aerococcus viridans and Anophryoides haemophila in American lobsters Homarus americanus freshly captured
in the waters of Prince Edward Island, Canada. Diseases of Aquatic
Organisms, 46: 231 –236.
Lizaso, J. L. S., Goñi, R., Reñones, O., Charton, J. a. G., Galzin, R., Bayle,
J. T., Jerez, P. S., et al. 2000. Density dependence in marine protected
populations: a review. Environmental Conservation, 27: 144– 158.
Lo, C-F. G. 2014. White spot disease. OIE (Office International des
Epizooties) Manual of Diagnostic Tests for Aquatic Animal Diseases,
pp. 177–190. Office International des Epizooties, Paris, France.
McCallum, H., Gerber, L., and Jani, A. 2005. Does infectious disease influence the efficacy of marine protected areas? A theoretical framework. Journal of Applied Ecology, 42: 688– 698.
Moland, E., Olsen, E., Andvord, K., Knutsen, J., and Stenseth, N. C.
2011a. Home range of European lobster (Homarus gammarus) in a
marine reserve: implications for future reserve design. Canadian
Journal of Fisheries and Aquatic Sciences, 68: 1197 – 1210.
Moland, E., Olsen, E., Knutsen, H., Knutsen, J., Enersen, S., André, C.,
and Stenseth, N. 2011b. Activity patterns of wild European lobster
Homarus gammarus in coastal marine reserves: implications for
future reserve design. Marine Ecology Progress Series, 429: 197 – 207.
Moland, E., Olsen, E. M., Knutsen, H., Garrigou, P., Espeland, S. H.,
Kleiven, A. R., André, C., et al. 2013a. Lobster and cod benefit
from small scale northern marine protected areas: inference from
an empirical before– after control-impact study. Royal Society
Proceedings B Biological Sciences, 280: 1754.
Moland, E., Ulmestrand, M., Olsen, E., and Stenseth, N. 2013b.
Long-term decrease in sex-specific natural mortality of European
lobster within a marine protected area. Marine Ecology Progress
Series, 491: 153 – 164.
Myers, R. A., Hutchings, J. A., and Barrowman, N. J. 1997. Why do fish
stocks collapse? The example of cod in Atlantic Canada. Ecological
Applications, 7: 91 – 106.
O’Neill, D. J., and Cobb, J. S. 1979. Some factors influencing the
outcome of shelter competition in lobsters (Homarus americanus).
Marine Behavior Physiology, 6:33– 45.
Pillay, T. V. R., and Kutty, M. N. 2005. Aquaculture: Principles and
Practices, 2nd edn. Blackwell Publishing Ltd, Oxford.
Pinnegar, J. K., and Engelhard, G. H. 2007. The “shifting baseline” phenomenon: a global perspective. Reviews in Fish Biology and
Fisheries, 18: 1 – 16. doi:10.1007/s11160-007-9058-6.
Pomeroy, R. S., Watson, L. M., Parks, J. E., and Cid, G. A. 2005. How is
your MPA doing? A methodology for evaluating the management
effectiveness of marine protected areas. Ocean and Coastal
Management, 48: 485– 502.
R Development Core Team. 2011. R: a Language and Environment for
Statistical Computing. The R Foundation for Statistical Computing,
Vienna, Austria. ISBN: 3-900051-07-0. http://www.R-project.org/.
i137
Rees, S., Fletcher, S., Glegg, G., Marshall, C., Rodwell, L., Jefferson, R.,
Campbell, M., et al. 2013. Priority questions to shape the marine
and coastal policy research agenda in the United Kingdom. Marine
Policy, 38: 531 –537.
Richards, S. A. 2005. Testing ecological theory using the informationtheoretic approach: examples and cautionary results. Ecology, 86:
2805– 2814.
Roa, R., and Bahamonde, R. 1993. Growth and expansion of an
exploited population of the squat lobster (Pleuroncodes
monodon) after 3 years without harvesting. Fisheries Research, 18:
305 – 319.
Roberts, C. M., Halpern, B., Palumbi, S. R., and Warner, R. R. 2001.
Designing networks of marine reserves: why small, isolated protected areas are not enough. Conservation Biology in Practice, 2:
10 – 17.
Rosenberg, A. A. 2003. Managing to the margins: the overexploitation
of fisheries. Frontiers in Ecology and the Environment, 1:
102 – 106.
Sciberras, M., Jenkins, S. R., Kaiser, M. J., Hawkins, S. J., and Pullin, A. S.
2013. Evaluating the biological effectiveness of fully and partially
protected marine areas. Environmental Evidence, 2: 4.
Sindermann, C. J. 1989. The Shell Disease Syndrome in Marine
Crustaceans. NOAA Technical, Memorandum NMFS-F/NEC-64,
43 pp.
Skinner, D., and Graham, D. 1972. Loss of limbs as a stimulus to ecdysis
in Brachyura (true crabs). The Biological Bulletin, 143: 222– 233.
Skog, M. 2009. Intersexual differences in European lobster (Homarus
gammarus): recognition mechanisms and agonistic behaviours.
Behaviour, 146: 1071– 1091.
Smith, I., Jensen, A., Collins, K., and Mattey, E. 2001. Movement of wild
European lobsters Homarus gammarus in natural habitat. Marine
Ecology Progress Series, 222: 177– 186.
Smolowitz, R. M., Bullis, R. A., and Abt, D. A. 1992. Pathologic cuticular
changes of winter impoundment shell disease preceding and during
intermolt in the American lobster, Homarus americanus. The
Biological Bulletin, 183: 99– 112.
Snesizko, S. F., and Taylor, C. C. 1947. A bacterial disease of the lobster
(Homarus americanus). Science, 105: 500.
Stebbing, P. D., Pond, M. J., Peeler, E., Small, H. J., Greenwood, S. J., and
Verner-Jeffreys, D. 2012. Limited prevalence of gaffkaemia (Aerococcus
viridans var. homari) isolated from wild-caught European lobsters
Homarus gammarus in England and Wales. Diseases of Aquatic
Organisms, 100: 159–167.
Stentiford, G. D., and Lightner, D. V. 2011. Cases of white spot disease
(WSD) in European shrimp farms. Aquaculture, 319: 302– 306.
Steneck, R. S. 2006. Possible demographic consequences of intraspecific
shelter competition among American lobsters. Journal of
Crustacean Biology, 26: 628– 638.
Stewart, J. E. 1980. Diseases. In The Biology and Management of
Lobsters, pp. 301 – 342. Ed. by J. S. Cobb, and B. F. Phillips.
Academic Press, New York.
Stewart, J. E., Cornick, J. W., and Zwicker, B. M. 1969. Influence of temperature on gaffkemia, a bacterial disease of the lobster Homarus
americanus. Journal of the Fisheries Research Board of Canada, 26:
2503– 2510.
Stewart, J. E., Cornick, J. W., Zwicker, B. M., and Arie, B. 2004. Studies
on the virulence of Aerococcus viridans (var.) homari, the causative
agent of gaffkemia, a fatal disease of homarid lobsters. Diseases of
Aquatic Organisms, 60: 149 – 155.
UK Ministerial Orders. 2013. Formal Designation Orders for 27
Marine Conservation Zones. http://www.legislation.gov.uk/
ukmo/2013.
Vogan, C., Llewellyn, P., and Rowley, A. 1999. Epidemiology and dynamics of shell disease in the edible crab Cancer pagurus: a preliminary study of Langland Bay, Swansea, UK. Diseases of Aquatic
Organisms, 35: 81 – 87.
i138
Vogan, C. L., Powell, A., and Rowley, A. F. 2008. Shell disease in
crustaceans—just chitin recycling gone wrong? Environmental
Microbiology, 10: 826– 835.
Wahle, R. A., Gibson, M., and Fogarty, M. 2009. Distinguishing disease
impacts from larval supply effects in a lobster fishery collapse.
Marine Ecology Progress Series, 376: 185– 192.
Whitten, M. M. A., Davies, C. E., Kim, A., Tlusty, M., Wootton, E. C.,
Chistoserdov, A., and Rowley, A. F. 2014. Cuticles of European
and American lobsters harbor diverse bacterial species
and differ in disease susceptibility. MicrobiologyOpen, 3:
395 – 409.
Wiik, R., Egidius, E., and Goksøyr, J. 1987. Screening of Norwegian lobsters Homarus gammarus for the lobster pathogen Aerococcus viridans. Diseases of Aquatic Organisms, 3: 97– 100.
C. E. Davies et al.
Wood, C. L., Lafferty, K. D., and Micheli, F. 2010. Fishing out marine
parasites? Impacts of fishing on rates of parasitism in the ocean.
Ecology Letters, 13: 761 – 775.
Wood, C. L., Micheli, F., Fernández, M., Gelcich, S., Castilla, J. C., and
Carvajal, J. 2013. Marine protected areas facilitate parasite populations among four fished host species of central Chile. The Journal
of Animal Ecology, 82: 1276– 1287.
Wootton, E. C., Woolmer, A. P., Vogan, C. L., Pope, E. C., Hamilton,
K. M., and Rowley, A. F. 2012. Increased disease calls for a cost–
benefits review of marine reserves. PloS One, 7: e51615.
Zuur, A., Ieno, E. N., Walker, N., Saveliev, A. A., and Smith, G. M. 2009.
GLM and GAM for absence– presence and proportional data. Mixed
Effects Models and Extensions in Ecology with R, pp. 245. Springer.
New York.
Handling editor: Rochelle Seitz
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i139 –i146. doi:10.1093/icesjms/fsu209
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Modelling the spread and connectivity of waterborne marine
pathogens: the case of PaV1 in the Caribbean
Andrew S. Kough 1*, Claire B. Paris 1, Donald C. Behringer 2,3, and Mark J. Butler, IV4
1
Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, Miami, FL 33149, USA
Fisheries and Aquatic Sciences Program, University of Florida, Gainesville, FL 32653, USA
3
Emerging Pathogens Institute, University of Florida, Gainesville, FL 32611, USA
4
Department of Biological Sciences, Old Dominion University, Norfolk, VA 23529, USA
2
*Corresponding author: tel: +1 240 447 0426; e-mail: [email protected]
Kough, A. S., Paris, C. B., Behringer, D. C., and Butler, M. J.IV Modelling the spread and connectivity of waterborne marine
pathogens: the case of PaV1 in the Caribbean. – ICES Journal of Marine Science, 72: i139 – i146.
Received 25 August 2014; revised 6 October 2014; accepted 28 October 2014; advance access publication 21 November 2014.
The PaV1 virus infects spiny lobsters (Panulirus argus) throughout most of the Caribbean, where its prevalence in adult lobsters can reach 17% and
where it poses a significant risk of mortality for juveniles. Recent studies indicate that vertical transmission of the virus is unlikely and PaV1 has not
been identified in the phyllosoma larval stages. Yet, the pathogen appears subclinically in post-larvae collected near the coast, suggesting that
lobster post-larvae may harbour the virus and perhaps have aided in the dispersal of the pathogen. Laboratory and field experiments also
confirm the waterborne transmission of the virus to post-larval and early benthic juvenile stages, but its viability in the water column may be
limited to a few days. Here, we coupled Lagrangian modelling with a flexible matrix model of waterborne and post-larval-based pathogen dispersal
in the Caribbean to investigate how a large area with complex hydrology influences the theoretical spread of disease. Our results indicate that if the
virus is waterborne and only viable for a few days, then it is unlikely to impact both the Eastern and Northwestern Caribbean, which are separated by
dispersal barriers. However, if PaV1 can be transported between locations by infected post-larvae, then the entire Caribbean becomes linked by
pathogen dispersal with higher viral prevalence in the North. We identify possible regions from which pathogens are most likely to spread, and
highlight Caribbean locations that function as dispersal “gateways” that could facilitate the rapid spread of pathogens into otherwise isolated areas.
Keywords: disease, lobster, Panulirus argus, physical oceanography, virus.
Introduction
Within the dense and stable medium of seawater, many animals
have evolved complex life cycles that include a distinctive planktonic
larval phase that facilitates population dispersal. Marine larvae can
harness powerful currents and potentially travel long distances
which led to the once widely accepted paradigm that marine populations are “open” (Roughgarden et al., 1985). However, we now know
that many larvae have specializations and behaviours that limit their
dispersal (Paris et al., 2007; Butler et al., 2011; Staaterman et al., 2012).
These adaptations enhance the local retention of larvae (Jones et al.,
1999; Paris and Cowen, 2004; Almany et al., 2007), resulting in a
continuum of dispersal potentials. Although this theory was promulgated on meroplankton, the dispersal of other important ecosystem
components, such as marine pathogens, may be similar but remain
understudied (Harvell et al., 2004).
# International
Pathogens play an important role in shaping marine communities (McCallum et al., 2001), where their dispersal and spread are
quite different from that of the terrestrial environment. Marine
pathogens tend to spread more rapidly than in terrestrial systems
(McCallum et al., 2003, 2004) and can fundamentally alter ecosystems (Harvell et al., 1999). For example, an unknown pathogen
that rapidly spread around the Caribbean in the 1980s killed longspined sea urchins (Diadema antillarium) en masse, removed an
important herbivore from the system, and contributed to a shift
towards macroalgae-dominated reefs (Lessios et al., 1984; Lessios,
1988). Emerging diseases that afflict habitat engineers, such as
corals or seagrasses, may become more prevalent as the climate
shifts (Harvell et al., 1999; Ward and Lafferty, 2004) and could
alter their range in tandem with their hosts. Unfortunately, the dispersal, epidemiology, and population dynamics of most marine
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i140
diseases are unknown, and many diseases are never detected
(Harvell et al., 2004). Many pathogens affect both aquacultured
and wild species of interest to fisheries, so understanding the dynamics of disease in both is crucial (Daszak et al., 2000).
One such pathogen, the virus Panulirus argus Virus 1 (PaV1),
infects the Caribbean spiny lobster P. argus (Shields and
Behringer, 2004) and is found throughout much of the wider
Caribbean (Moss et al., 2013). PaV1 is a large, icosahedral,
non-enveloped, DNA virus, and is primarily pathogenic to juvenile
lobsters (Shields and Behringer, 2004; Butler et al., 2008; Behringer
et al., 2012a, b). The high mortality that juveniles suffer adds another
challenge to managing this fully exploited fishery (Chavez, 2009;
Ehrhardt et al., 2010). Caribbean spiny lobsters have phyllosoma
larvae which have a long pelagic larval duration of 6 months or
more (Goldstein et al., 2008) and thus great dispersal potential.
Research using genetics (Silberman et al., 1994; Naro-Maciel et al.,
2011) and biophysical modelling (Briones-Fourzan et al., 2008;
Butler et al., 2011; Kough et al., 2013) suggest that the international
exchange of larvae is common so that populations around the
Caribbean are linked through larval transport. Likewise, strains of
PaV1 from disparate countries share phenotypes (Moss et al.,
2013) and some post-larvae that arrive onshore in Florida are
infected with PaV1 (Moss et al., 2012), making infected larvae a potential dispersal pathway for the pathogen.
Experiments testing vertical transmission from female to larvae
have proven negative (Behringer and Butler, unpublished data),
so the mechanism or vector of dispersal for PaV1 among distant
populations remains elusive. Thus far, no PaV1-infected phyllosoma
larvae have been detected in field collections (Lozano-Alvarez et al.,
unpublished data). However, some post-larvae collected near the
Florida coast (Moss et al., 2012) and others many kilometres off of
the continental shelf of the Yucatan peninsula (Lozano-Alvarez
et al., unpublished data) were found subclinically infected with
PaV1. Post-larval lobsters are incapable of feeding (Wolfe and
Felgenhauer, 1991) and live off concentrated lipids stored during
the long larval phase (Lemmens, 1994; Lemmens and Knott, 1994),
so ingestion of the PaV1 pathogen by post-larvae is not a viable transmission pathway. Whether late-stage phyllosoma larvae acquire
the virus through ingestion of infected prey or marine aggregates and
then retain the virus after metamorphosis into post-larvae, or if only
post-larvae are susceptible to infection and the virus is acquired
offshore is as yet unknown. Regardless, the presence of PaV1-infected
post-larvae means that the virus can potentially spread over long
distances within infected, planktonic hosts.
Our goal was to compare the genetic relatedness of PaV1 strains
in the Caribbean (Moss et al., 2013) with modelled patterns
of pathogen spread. We used biophysical Lagrangian stochastic
modelling coupled to a stepping-stone model of pathogen spread
to describe the probable dispersal of a waterborne pathogen
compared with behaviourally adroit P. argus post-larvae in the
Caribbean Sea. We tested two hypotheses for how PaV1 may be
spread, as: (i) a passive waterborne particle within the top 10 m of
the ocean, or (ii) a lobster post-larvae infected with PaV1. Our
approach evaluates how hydrology could affect the potential
spread of a marine pathogen throughout the Caribbean.
Methods
We used an advanced multi-scale open-source Lagrangian biophysical model of dispersal, the Connectivity Modeling System (CMS;
Paris et al., 2013), to probabilistically describe Caribbean-wide connections between reef sites for 5 years, from the start of 2004 through
A. S. Kough et al.
the end of 2008. We then used the resulting matrices to randomly
and repeatedly seed an ensemble of simulations with pathogen
emergence and subsequent spread.
Habitat sites
We chose habitat sites containing coral reefs from those identified in
the Millennium Coral Reef Mapping Project (Andréfouët et al.,
2006). We overlaid an 8 × 8 km grid across the Caribbean and
created 3202 sites intersecting the reef extent, following Holstein
et al. (2014). Real-world spatial data do not always translate into operational hydrodynamic models, and can rest on boundaries
between land and water in the model, rendering them ineffective.
We removed 806 sites that were impacted by the structure of the
modelling framework. Further, we restricted our simulations to
sites that both exchanged and received larvae (removing another
714 sites) so that we were working with a theoretical network with
no disconnected sites. The resulting 1682 habitat sites stretched
across the Caribbean, from the Barbados to the northern edge of
the Florida Keys (USA) reef track (Figure 1).
Lagrangian tracking
We coupled and nested offline operational hydrodynamic models
with the CMS. More specifically, we used the 1/128 Hybrid
Coordinate Ocean Model (HyCOM) with NCODA assimilated
and the 1/248 HyCOM Gulf of Mexico models (Bleck, 2002) that
are publicly available and cover our study location. Both models
yield daily output, resolve mesoscale eddy features, and have been
widely applied to many different ocean transport scenarios. The
habitat seascape and hydrodynamic models were integrated
within the CMS for simulations of particle transport. We used highfrequency daily particle releases to model connectivity, corresponding to the daily windforcing of the HyCOM models (Bleck, 2002). A
sensitivity analysis indicated that 100 particles released daily from
each reef habitat site achieved probabilistic saturation at particle
tracking times longer than those used in this study (Kough, 2014).
Simulations
We tested two different hypotheses on how PaV1 virions may be
transported around the Caribbean: in seawater or in infected postlarvae. Our first hypothesis was that the pathogen is passively
advected as a waterborne particle among habitat locations in the
surface (,10 m) water layer (Waterborne simulation: WB). In the
WB scenario, we used a pelagic duration of 5 d, which corresponds
to laboratory estimates of how long PaV1 can be infectious in seawater (Butler and Behringer, unpublished data). Considering
PaV1 virions lack a structural envelope (Shields and Behringer,
2004), they are apt to be more environmentally stable (see
Satheesh Kumar et al., 2013 and references therein), which supports
the possibility that PaV1 spreads in the water column itself. Unlike
non-enveloped viruses such as PaV1, the lipoprotein layer that typically covers enveloped viruses and functions in host cell recognition
and penetration, is fragile, and can limit their persistence. The 5 d
duration in seawater that we modelled for PaV1 also fits well
within the range of viabilities known for several other marine
viruses. The viability of all viruses depends on environmental characteristics, particularly salinity and temperature, but the examples
that follow all come from experiments that were conducted in seawater (36 psu) and under temperatures likely to be experienced
by a virus in the tropical Caribbean Sea (20–338C). Some of these
viruses have envelopes, such as white spot syndrome virus, a large
DNA virus of decapod crustaceans that can remain viable in the
Marine pathogen dispersal
i141
Figure 1. Habitat map and mean pathogen spread times around the Caribbean. The habitat map (a) shows the centroids of the 8 × 8 km square
coral reef habitat locations (n ¼ 1682) used in the simulations. The mean number of steps in the model (equivalent to time) for PaV1 to spread to
each habitat location from random origins is shown for two scenarios: waterborne (WB; b) and post-larvae (PL; c). Colour is used to identify locations
on the map and links the position of the habitat on the x-axis in the panels (b and c) of the mean pathogen spread time to a spatial location (a). The
position of the habitat identified here is consistent throughout the paper.
seawater for 12 d (Satheesh Kumar et al., 2013), or the viral haemorrhagic septicaemia virus, an enveloped RNA virus of fish that can
remain viable in seawater for 4 d (Hawley and Garver, 2008).
Other viruses such as PaV1 are non-enveloped or “naked”, such as
the poliovirus, echovirus, and coxsackie virus; all are RNA mammal
viruses that remain viable in seawater for 1–12 d (Fujioka et al.,
1980; Labelle and Gerba, 1980; Block, 1983), 3 d (Labelle and
Gerba, 1980), and 30 d (Nasser et al., 2003), respectively.
Our second hypothesis was that after metamorphosis from the
phyllosoma larval phase, presumably seaward of the continental
shelf (McWilliam and Phillips, 2007; Phillips and McWilliam,
2009), P. argus post-larvae become infected with PaV1 and the
pathogen spreads among locations during the post-larval competency period (Postlarvae simulation: PL). In the PL scenario, we
used a variable pelagic duration of 5 –14 d based on estimates of
the average duration (Goldstein et al., 2008) and metabolic
i142
A. S. Kough et al.
constraints (Wilkin and Jeffs, 2011) of the post-larval stage. For the
PL, we included a daily vertical migration behaviour from the
surface layer at night to depths of at least 100 m during the day as
hypothesized by Phillips et al. (1978) for P. cygnus and Butler et al.
(2011) for P. argus.
Connectivity probability matrices M(i,j) were constructed for
the WB and PL scenarios using the modelled particles that successfully arrived at a habitat site ( j) from a habitat site (i), after the
pelagic duration.
Particlesij
Mij = .
Particlesi
measure of how important a habitat location is to the entire connectivity network. It was formulated by calculating how many times a
given habitat location is involved in the fastest pathway of pathogen
spread between any other two locations in the network. Thus, we
used the time to infection as an edge weight, to assess how rapidly
pathogens from one location spread throughout the system. In
this context, locations in the Caribbean with the highest betweenness centrality played crucial roles in rapidly sending pathogens to
other areas. Calculations were made in MATLABw using the BGL
toolbox developed by Gleich (2006).
Results
Waterborne
Stepping-stone pathogen spread model
We used the probability connectivity matrices from the Lagrangian
tracking simulations in a MATLABw model (Supplementary material) that seeded a randomly selected location in the Caribbean
with disease and then spread the pathogen elsewhere based on
either the WB or PL simulation scenario (Figure 2). The generality
of the model is improved by incorporating stochasticity in the
spatial origin and connectivity of the pathogen, which is pertinent
for PaV1 as its origin is unknown and the probability of exchange
between any two locations is likely to change from year to year
based on factors such as temporal variability in the prevalence of
PaV1 (Behringer et al., 2011) or changing atmospheric and
oceanic conditions (e.g. ENSO, current positions and meanders,
mesoscale eddy formation, etc.).
Gateway habitats
To evaluate how important any given habitat location is in the
disease network, we considered its potential for facilitating further
spread using betweenness centrality. Betweenness centrality is a
In the WB scenario, the Eastern and Central Caribbean were not
reached by pathogens originating in the North (Figure 3). Indeed,
many regions do not exchange disease with other regions in this
scenario and were isolated. This isolation delayed (i.e. more
model steps and longer mean time for pathogen spread) or prevented the spread of the pathogen to other locations. Spread-time
greatly increased near major oceanographic barriers: for example,
near the Turks and Caicos Islands and the island of Hispanola (see
Figures 1b and 3WB). The locations with the most betweenness centrality were clustered near the northwest coast of Cuba, and the
northeast coast of the Yucatan (Figure 4). The connections at
these sites represent a pathway for disease to spread from the
Yucatan into the Northern Caribbean and would expand the
network.
Post-larvae
The modelled pathogen spread faster throughout the Caribbean in
the PL scenario than in the WB scenario, regardless of origin (see
Figure 1c vs. b). The pathogen spread more widely between the
Northern and the Central and Eastern Caribbean (see Figure 3PL)
and spread more rapidly from the Central and Eastern Caribbean
into the Northwestern Caribbean than vice versa. In both the PL
and WB scenarios, the Lesser Antilles remained isolated, because
it took more model steps for pathogen to spread to them or it
never reached them. However, the Florida Keys took longer to
receive the pathogen in the PL scenario than it did in the WB scenario (see Figure 3; the leftmost columns are darker in WB than in
PL). The locations with the highest betweenness centrality in the
PL scenario were next to oceanographic features such as the
Windward, Mona, and Anegada passages, the Gulf of Colombia
and Gulfo de los Mosquitos, the Nicaraguan Rise, and the
Yucatan Channel (see Figure 4).
Discussion
Figure 2. A flowchart summarizing the steps in the pathogen
stepping-stone simulations. A pathogen was seeded 10 000 times from
a random starting location in the Caribbean. For each seeding event, the
pathogen could spread between locations for 300 model steps. During
each step, the probability of dispersal away from an infected location
was taken from a matrix of the 5-year mean modelled particle transport
specific to either a waterborne or post-larval scenario. This probability
was augmented by a stochastic addition to better capture
unpredictability between years.
Our findings support the expected pattern of linkages among areas
in the Caribbean based on well-studied oceanic circulation. The
Northern, Central, and Eastern Caribbean were isolated from each
other when the pathogen was waterborne and short-lived.
Particles have to travel against the prevailing flow to spread from
the Northern into the Central and Eastern Caribbean, so this was
expected. It is unlikely that the two regions would rapidly share a
disease outbreak that originated in the North, although some exchange is possible through pathways from the Greater into the
Lesser Antilles and on into the gyres off of Colombia and Panama.
Thus, even short-term waterborne transport could spread PaV1
through the Caribbean over long enough periods of time.
However, when we modelled the pathogen travelling within an
Marine pathogen dispersal
i143
Figure 3. Time matrices for the Waterborne (WB) pathogen simulation (left) and Post-larval (PL) pathogen simulation (right) that show how fast
the pathogen spread from one specific location (y-axis) to another specific location (x-axis). The position of habitat on both axes corresponds to
previously used colour (see Figure 1a) and x-axis position (see Figure 1b and c). The shade of grey represents how many model steps it took for the
pathogen to spread from one location to the next. There was no spread of disease among locations shown in white, slow spread between areas
shown in lighter shades of gray, and rapid spread between areas shown in black.
Figure 4. Map showing habitats that functioned as “gateways” from which the modelled pathogens spread most rapidly to other habitats.
Locations denoted with white stars (waterborne scenario) and grey squares (post-larval scenario) were the areas we defined as “gateways” based on
the betweenness centrality values of their pathogen spread time. These were the sites with the highest betweenness centrality (same order of
magnitude) in the networks of each scenario. The prevailing surface currents in the Caribbean are shown with the dotted line, and major
oceanographic features are labelled.
infected post-larval host, there was considerably more exchange
between the Northern and Southern Caribbean. Regardless of the
pathogen transport scenario, it took more model time-steps for
the pathogen to spread to the Lesser Antilles, South American
Coast, and Central America. Thus, a highly virulent disease that is
mainly limited by oceanographic dispersal should spread to
i144
these locations last. We also identified certain locations that acted as
“gateways” through which the pathogen spread rapidly among habitats. Gateways tended to be next to major currents and to other
oceanographic divides. These gateways facilitated connections
between localities that were otherwise isolated, and therefore are
important locations to focus potential management for future
epizootics.
How do the two scenarios that we modelled (e.g. WB vs. PL)
compare with the geographic structure of relatedness between
strains of PaV1 around the Caribbean (Moss et al., 2013)? The
WB scenario better captured the isolation and absence of PaV1 in
the Lesser Antilles, and in the South and Central Caribbean. Yet,
only the PL scenario was consistent with genetic linkages between
the PaV1 virus in Panama and the Northern Caribbean (Moss
et al., 2013), as it created increased exchange of the pathogen from
Central and South America into the Central and Northern
Caribbean. However, the number of infected lobsters found in
Panama and the Southern Caribbean was low and the authors
noted that they had the most diversity where they had the highest
sample size (Moss et al., 2013). Still, the presence of shared variants
between Central America and the Greater Antilles indicates that
the dispersal method for PaV1 requires a waterborne or vectored
pathogen that is viable in seawater for more than a few days.
From the perspective of the pathogen, post-larvae represent an
idealized method of dispersal. Post-larvae use chemical cues
(Herrnkind and Butler, 1986; Goldstein and Butler, 2009) and
tidal signals (Kough et al., 2014) to reach nursery habitat—an
optimal location for a pathogen to reach susceptible juvenile
hosts. Still, how do the non-feeding post-larvae acquire the pathogen, if they are not first infected as larvae? Early benthic juvenile
lobsters are highly susceptible to PaV1 and can be readily infected
through viral particles in the water column (Butler et al., 2008).
Spiny lobster post-larvae come onshore during the dark phase of
the lunar month with nocturnal flood tides (Acosta et al., 1997;
Yeung et al., 2001), navigating from offshore using chemical, tidal,
and probably other cues (Goldstein and Butler, 2009; Kough et al.,
2014). Post-larvae orient towards coastal water masses, which may
expose them to viral particles or other post-larvae already infected
within the nursery habitat. Indeed, the model results suggest that
the pathogen could spread across the Caribbean travelling within
post-larvae if they remain subclinically infected and offshore for
an extended time. In addition, stronger linkages between the
North and South would exist and be maintained if the pathogen
had origins in South or Central America.
The transport of pathogens within marine aggregates is perhaps
another way that coastal communities around the Caribbean may be
linked. Our results in the WB scenario demonstrate that passive particle dispersal connects otherwise disparate areas, even over short
time-scales. Many organisms, chemicals, and gradients are transported throughout the water column (Maddox, 2006) and drifting
debris can harbour infectious pathogens (Lizarraga-Partida et al.,
2011) and a diverse microbial community (Lyons et al., 2010).
Marine microbial communities can transition into opportunistic
pathogens that afflict plants and animals, especially in a changing
ocean (Burge et al., 2013). The floating assemblages of biological
materials such as mucilaginous aggregates and algae offshore of
coastal communities can play an important role for larval recruitment as shelter and perhaps as a signal of nearby benthic habitat
(Wells and Rooker, 2004; Goldstein and Butler, 2009). If such aggregates are used by meroplankton transitioning from the pelagic into
benthic habitats, these could be locations where concentrations of
A. S. Kough et al.
susceptible animals come into contact with pathogens such as
PaV1. Marine aggregates may also function as vessels for keeping
the virus protected (e.g. from UV light) and viable (Bongiorni
et al., 2007), and thus aid in pathogen transport for longer times
and potentially larger distances.
Our discovery of key gateways that facilitate pathogen connectivity among localities that would otherwise be isolated is meaningful
for marine pathogen epidemiology and potential management of
the spread of disease. The Greater Antilles have the highest PaV1
genetic diversity (Moss et al., 2013) and also contained the highest
number of gateway habitats in the PL simulation scenario, supporting the contention that such locations acted as points connecting a
large number of origins and destinations. In the WB scenario, the
Caribbean network was split into two relatively separate areas:
northwest and southeast of the prevailing current. The major gateways in the WB scenario straddled the primary pathway for linkage
between these two areas and occurred at the northeastern tip of
Mexico and on the west coast of Cuba. Very few gateway locations
were shared between the two scenarios, although there were overlaps
at the Cay Sal Bank (western Bahamas) and southeastern Cuba.
The gateways in the Cay Sal Bank facilitated connections into
the Bahamas and Florida, and were an important intermediary location next to the powerful Florida current. Active management of
fishery activities that potentially increase pathogen prevalence or
transmission (Behringer et al., 2012a, b) within gateway sites may
help to forestall the spread of pathogens among regions of the
Caribbean.
As with any study, there are certain caveats to our results that bear
mention. The time-scale of the Lagrangian modelling was 5 years,
which does not encompass the full length of time since PaV1 was
first detected. However, our aim was not to recreate the actual
pattern of the spread of PaV1 disease, which is unknown, but to
instead explore the potential way in which water- or vector-borne
pathogens may transition between locations. By tying a stochastic
stepping-stone model to the connectivity matrices generated by
the Lagrangian particle tracking model, we added variation to simulate periodic, but important events. This construct could miss rare
events that happen with a periodicity of ,5 years, but the mean connectivity measured over shorter periods is comparable to the connectivity over longer time-scales (Berglund et al., 2012). We also
simulated the spread of a waterborne pathogen using only a 5 d duration, a time frame believed relevant to the PaV1 virus. The viability
of many seawater-borne pathogens is poorly known, but their duration will clearly affect patterns of connectivity, as will other factors
that we did not simulate (e.g. alternate hosts, dormant stages, host
population dynamics, etc.). The CMS also used general ocean circulation models that best represented geostrophic flow (Bleck, 2002);
the nearshore tidal and local wind-driven dynamics were not well
represented in the general ocean circulation. However, we were
interested in the transport through pelagic oceans that could
advect a pathogen across greater distances than possible by nearshore processes. When the modelled pathogen originated in
Florida in the PL scenario, it spread in a countercurrent direction
similar to the well-documented invasion of the Caribbean by lionfish (Morris and Whitfield, 2009). Therefore, although the PL scenario used a shorter dispersal time than a lionfish larva, our model’s
results were comparable to a known pattern of Caribbean-wide invasion. In addition, using only open-access tools, such as HyCOM
and CMS, makes our work easily reproducible and accessible for
other researchers to build on while investigating dispersal of other
marine pathogens.
Marine pathogen dispersal
Although we were inspired to model pathogen dispersal by the
discovery and subsequent Caribbean-wide interest in PaV1, our
work is relevant to a myriad of different scenarios and organisms
affected by marine pathogens. Distant pathogen dispersal could be
tied into well-studied disease and ecosystem model systems
(Cáceres et al., 2014; Dolan et al., 2014), therefore linking the
oceanographic environment to population-based models and yielding a more comprehensive epidemiological model (McCallum et al.,
2004). Indeed, considering distant dispersal in tandem with disease
dynamics has better explained the outbreak of Aspergillosis in
seafans throughout the Caribbean (Bruno et al., 2011) and the persistence of PaV1 in lobster populations in Florida (Dolan et al.,
2014). Still, our understanding of the large-scale epidemiology of
marine diseases is in its infancy and will likely require analyses
that integrate physical dynamics, anthropogenic influence, and
disease history so as to identify potential hotspots of marine pathogens in the world’s oceans, similar to work in terrestrial environments (Jones et al., 2008). Benthic marine ecosystems receive new
recruits, and potentially pathogens, from ocean currents that
unite habitat over broad spatial and temporal scales, making
Lagrangian modelling an appropriate tool for examining the potential connectivity of marine diseases.
Supplementary data
Supplementary material is available at ICESJMS online version of
the manuscript.
Acknowledgements
This project greatly benefited from advice and direction given
by Michael Schmale, Tom Dolan, and two anonymous reviewers.
Funding was provided through the National Science Foundation
awards OCE-0928930 (CBP), OCE-0929086 (MJB), and OCE0928398 (DCB).
References
Acosta, C. A., Matthews, T. R., and Butler, M. J., IV. 1997. Temporal patterns and transport processes in recruitment of spiny lobster
(Panulirus argus) postlarvae to South Florida. Marine Biology,
129: 79 –85.
Almany, G. R., Berumen, M. L., Thorrold, S. R., Planes, S., and Jones, G.
P. 2007. Local replenishment of coral reef fish populations in a
marine reserve. Science, 316: 742– 744.
Andréfouët, S., Muller-Karger, F. E., Robinson, J. A., Kranenburg, C. J.,
Torres-Pulliza, D., Spraggins, A., and Murch, B. 2006. In 10th ICRS.
Global Assessment of Modern Coral Reef Extent and Diversity for
Regional Science and Management Applications: a View from
Space, pp. 1732– 1745. Japanese Coral Reef Society, Okinawa, Japan.
Behringer, D. C., Butler, M. J., IV, Shields, J. D., and Moss, J. 2011.
Review of Panulirus argus Virus 1 (PaV1)—a decade after its discovery. Diseases of Aquatic Organisms, 94: 153 – 160.
Behringer, D. C., Butler, M. J., IV, and Stentiford, G. D. 2012a. Disease
effects on lobster fisheries and ecology. Diseases of Aquatic
Organisms, 100: 89 –93.
Behringer, D. C., Moss, J., Shields, J. D., and Butler, M. J., IV. 2012b.
PaV1 infection in the Florida spiny lobster fishery and its effects
on trap function and disease transmission. Canadian Journal
Fisheries and Aquatic Sciences, 69: 136– 144.
Berglund, M., Jacobi, M. N., and Jonsson, P. R. 2012. Optimal selection
of marine protected areas based on connectivity and habitat quality.
Ecological Modeling, 240: 105 – 112.
Bleck, R. 2002. An oceanic general circulation model framed in hybrid
isopycnal-cartesian coordinates. Ocean Modeling, 4: 55 – 88.
i145
Block, J. C. 1983. Viruses in environmental waters. In Viral Pollution of
the Environment. Ed. by G. Berg. CRC Press, Boca Raton.
Bongiorni, L., Armeni, M., Corinaldesi, C., Dell’Anno, A., Pusceddu, A.,
and Danovaro, R. 2007. Viruses, prokaryotes and biochemical composition of organic matter in different types of mucilage aggregates.
Aquatic and Microbial Ecology, 49: 15– 23.
Briones-Fourzan, P., Candela, J., and Lozano-Alvarez, E. 2008.
Postlarval settlement of the spiny lobster Panulirus argus along the
Caribbean coast of Mexico: patterns, influence of physical factors,
and possible sources of origin. Limnology and Oceanography, 53:
970– 985.
Bruno, J. F., Ellner, S. P., Vu, I., Kim, K., and Harvell, C. D. 2011. Impacts
of aspergillosis on sea fan coral demography: modeling a moving
target. Ecological Monographs, 81: 123– 139.
Burge, C. A., Kim, C. J. S., Lyles, J. M., and Harvell, C. D. 2013. Special
issue oceans and humans health: the ecology of marine opportunists.
Microbial Ecology, 65: 869 – 879.
Butler, M. J., IV, Behringer, D. C., and Shields, J. D. 2008. Transmission
of Panulirus argus virus 1 (PaV1) and its effect on the survival of juvenile Caribbean spiny lobster. Diseases of Aquatic Organisms, 79:
173– 182.
Butler, M. J., IV, Paris, C. B., Goldstein, J. S., Matsuda, H., and Cowen, R.
K. 2011. Behavior constrains the dispersal of long-lived spiny lobster
larvae. Marine Ecology Progress Series, 422: 223– 237.
Cáceres, C. E., Tessier, A. J., Duffy, M. A., and Hall, S. R. 2014. Disease in
freshwater zooplankton: what have we learned and where are we
going? Journal of Plankton Research, 36: 1 – 8.
Chavez, E. A. 2009. Potential production of the Caribbean spiny lobster
(Decapoda, Palinura) fisheries. Crustaceana, 82: 1393– 1412.
Daszak, P., Cunningham, A. A., and Hyatt, A. D. 2000. Wildlife
ecology—emerging infectious diseases of wildlife—threats to biodiversity and human health. Science, 287: 443– 449.
Dolan, T. W., Butler, M. J., IV, and Shields, J. D. 2014. Host behavior and
habitat structure alter disease dynamics in a social marine species:
the Caribbean spiny lobster. Ecology, 95: 2346– 2361.
Ehrhardt, N. M., Puga, R., and Butler, M. J., IV. 2010. Large ecosystem dynamics and fishery management concepts: the Caribbean spiny
lobster, Panulirus argus, fisheries. In Towards Marine EcosystemBased Management in the Wider Caribbean. Ed. by L. Fanning, R.
Mahon, and P. McConney. Amsterdam University Press, Amsterdam.
Fujioka, R. S., Loh, P. C., and Lau, L. S. 1980. Survival of human enteroviruses in the Hawaiian ocean environment: evidence for
virus-inactivating microorganisms. Applied and Environmental
Microbiology, 39: 1105– 1110.
Gleich, D. 2006. Matlab BGL v4.0. http://dgleich.github.io/matlabbgl/ (last accessed 18 August 2014).
Goldstein, J. S., and Butler, M. J., IV. 2009. Behavioral enhancement of
onshore transport by postlarval Caribbean spiny lobster (Panulirus
argus). Limnology and Oceanography, 54: 1669 – 1678.
Goldstein, J. S., Matsuda, H., Takenouchi, T., and Butler, M. J. I. V. 2008.
The complete development of larval Caribbean spiny lobster
(Panulirus argus) in culture. Journal of Crustacean Biology, 28:
306– 327.
Harvell, C. D., Kim, K., Burkholder, J. M., Colwell, R. R., Epstein, P. R.,
Grimes, D. J., Hofmann, E. E., et al. 1999. Emerging marine
diseases—climate links and anthropogenic factors. Science, 285:
1505– 1510.
Harvell, D., Aronson, R., Baron, N., Connell, J., Dobson, A., Ellner, S.,
Gerber, L., et al. 2004. The rising tide of ocean diseases: unsolved problems and research priorities. Frontiers in Ecology and the
Environment, 2: 375 – 382.
Hawley, L. M., and Garver, K. A. 2008. Stability of viral hemorrhagic
septicemia virus (VHSV) in freshwater and seawater at various temperatures. Diseases of Aquatic Organisms, 82: 171– 178.
Herrnkind, W. F., and Butler, M. J., IV. 1986. Factors regulating settlement and microhabitat use by juvenile spiny lobsters, Panulirus
argus. Marine Ecology Progress Series, 34: 23 – 30.
i146
Holstein, D. M., Paris, C. B., and Mumby, P. J. 2014. Consistency and
inconsistency in multispecies population network dynamics of
coral reef ecosystems. Marine Ecology Progress Series, 499: 1 – 18.
Jones, G. P., Millicich, M. J., Emslie, M. J., and Lunow, C. 1999. Selfrecruitment in a coral reef fish population. Nature, 402: 802– 804.
Jones, K. E., Patel, N. G., Levy, M. A., Storeygard, A., Balk, D., Gittleman,
J. L., and Daszak, P. 2008. Global trends in emerging infectious
diseases. Nature, 451: 990 – 994
Kough, A. S. 2014. Population connectivity and larval dispersal
of Caribbean spiny lobster. Dissertation, University of Miami,
Miami, FL, 177 pp.
Kough, A. S., Paris, C. B., and Butler, M. J., IV. 2013. Larval connectivity
and the international management of fisheries. PLoS ONE, 8:
e64970.
Kough, A. S., Paris, C. B., and Staaterman, E. 2014. In situ swimming and
orientation behavior of spiny lobster (Panulirus argus) postlarvae.
Marine Ecology Progress Series, 504: 207– 214.
LaBelle, R. L., and Gerba, C. P. 1980. Influence of estuarine sediment
in viral survival under field conditions. Applied Environmental
Microbiology, 39: 749– 755.
Lemmens, J. W. T. J. 1994. Biochemical evidence for absence of feeding
in puerulus larvae of the Western rock lobster Panulirus cygnus
(Decapoda: Palinuridae). Marine Biology, 118: 383 – 391.
Lemmens, J. W. T. J., and Knott, B. 1994. Morphological changes in
external and internal feeding structures during the transition
phyllosoma-puerulus-juvenile in the Western Rock Lobster
(Panulirus cygnus, Decapoda: Palinuridae). Journal of Morphology,
220: 271–280.
Lessios, H. A. 1988. Mass mortality of Diadema antillarum in the
Caribbean: what have we learned? Annual Reviews of Ecological
Systems, 19: 371 –393.
Lessios, H. A., Robertson, D. R., and Cubit, J. D. 1984. Spread of
Diadema antillarum mass mortality through the Caribbean.
Science, 226: 335– 337.
Lizarraga-Partida, M. L., Mendez-Gomez, E., Rivas-Montano, A. M.,
Vargas-Hernandez, E., Portillo-Lopez, A., Gonzalez-Ramirez, A. R.,
Huq, A., et al. 2011. Association of Vibrio cholera with plankton in
coastal areas of Mexico. Environmental Microbiology, 11: 201–208.
Lyons, M. M., Ward, J. E., Gaff, H., Hicks, R., Drake, J., and Dobbs, F. C.
2010. Theory of island biogeography on a microscopic scale:
are organic aggregates islands for aquatic pathogens? Aquatic
Microbial Ecology, 60: 1 – 13.
Maddox, E. D. 2006. Temporal patterns in the ingress of settlementstage gray snapper (Lutjanus griseus) and other fishes to Biscayne
Bay, Florida, and the Caicos Bank, Turks and Caicos Islands.
Dissertations from ProQuest. Paper 2468.
McCallum, H. I., Barlow, N., and Hone, J. 2001. How should pathogen
transmission be modelled? Trends in Ecology and Evolution, 16:
295– 300.
McCallum, H. I., Harvell, C. D., and Dobson, A. P. 2003. Rates of spread
of marine pathogens. Ecology Letters, 6: 1062– 1067.
McCallum, H. I., Kuris, A. M., Harvell, C. D., Lafferty, K. D., Smith,
G. W., and Porter, J. 2004. Does terrestrial epidemiology apply to
marine systems? Trends in Ecology and Evolution, 19: 585 –591.
McWilliam, P. S., and Phillips, B. F. 2007. Spiny lobster development:
mechanisms inducing metamorphosis to the puerulus: a review.
Reviews in Fish Biology and Fisheries, 17: 615 – 632.
Morris, J. A., Jr, and Whitfield, P. E. 2009. Biology, Ecology, Control and
Management of the Invasive Indo-Pacific Lionfish: an Updated
Integrated Assessment. NOAA Technical Memorandum NOS
NCCOS 99. 57 pp.
Moss, J., Behringer, D. C., Shields, J. D., Baeza, J. A., Aguilar-Perera, A.,
Bush, P. G., and Dromer, C., IV.et al. 2013. Distribution, prevalence,
and genetic analysis of Panulirus argus virus I from the Caribbean
Sea. Diseases of Aquatic Organisms, 104: 129– 140.
A. S. Kough et al.
Moss, J., Butler, M. J., IV, Behringer, D. C., and Shields, J. D. 2012.
Genetic diversity of the Caribbean spiny lobster virus, Panulirus
argus virus 1, and the discovery of PaV1 in lobster post-larvae.
Aquatic Biology, 14: 223 – 232.
Naro-Maciel, E., Reid, B., Holmes, K. E., Brumbaugh, D. R., Martin, M.,
and DeSalle, R. 2011. Mitochondrial DNA sequence variation in
spiny lobsters: population expansion, panmixia, and divergence.
Marine Biology, 158: 2027– 2041.
Nasser, A. M., Zaruk, N., Tenenbaum, L., and Netzan, Y. 2003.
Comparative survival of Cryptosporidium, coxsackievirus A9 and
Escherichia coli in stream, brackish and seawaters. Water Science
and Technology, 47: 91 –96.
Paris, C. B., Cherubin, L. M., and Cowen, R. K. 2007. Surfing, spinning,
or diving from reef to reef: effects on population connectivity.
Marine Ecology Progress Series, 347: 285– 300.
Paris, C. B., and Cowen, R. K. 2004. Direct evidence of a biophysical retention mechanism for coral reef larvae. Limnology and
Oceanography, 49: 1964– 1979.
Paris, C. B., Helgers, J., Van Sebille, E., and Srinivasan, A. 2013. The connectivity modeling system: a probabilistic modeling tool for the
multi-scale tracking of biotic and abiotic variability in the ocean.
Environmental Modelling Software, doi:10.1016/j.envsoft.2012.
12.006
Phillips, B. F., and McWilliam, P. S. 2009. Spiny lobster development:
where does successful metamorphosis to the puerulus occur?: a
review. Reviews in Fish Biology and Fisheries, 19: 193 – 215.
Phillips, B. F., Rimmer, D. W., and Reid, D. D. 1978. Ecological investigations of the late-stage phyllosoma and puerulus larvae of the Western
rock lobster Panulirus longipes cygnus. Marine Biology, 45: 347 – 357.
Roughgarden, J., Iwasa, Y., and Baxter, C. 1985. Demographic theory for
an open marine population with space-limited recruitment.
Ecology, 66: 54 – 67.
Satheesh Kumar, S., Ananda Bharathi, R., Rajan, J. J. S., Alavandi, S. V.,
Poornima, M., Balasubramanian, C. P., and Ponniah, A. G. 2013.
Viability of white spot syndrome virus (WSSV) in sediment
during sun-drying (drainable pond) and under non-drainable
pond conditions indicated by infectivity to shrimp. Aquaculture,
402: 119– 126.
Shields, J. D., and Behringer, D. C. 2004. A new pathogenic virus in the
Caribbean spiny lobster Panulirus argus from the Florida Keys.
Diseases of Aquatic Organisms, 59: 109– 118.
Silberman, J. D., Sarver, S. K., and Walsh, P. J. 1994. Mitochondrial DNA
variation and population structure in the spiny lobster Panulirus
argus. Marine Biology, 120: 601 – 608.
Staaterman, E., Paris, C. B., and Helgers, J. 2012. Orientation behavior in
fish larvae: a missing piece to Hjort’s critical period hypothesis.
Journal of Theoretical Biology, 304: 188 – 196.
Ward, J. R., and Lafferty, K. D. 2004. The elusive baseline of marine
disease: are diseases in ocean ecosystems increasing? PLoS Biology,
2: e120. doi:10.1371/journal.pbio.0020120
Wells, R. J. D., and Rooker, J. R. 2004. Spatial and temporal patterns of
habitat use by fishes associated with Sargassum mats in the northwestern Gulf of Mexico. Bulletin of Marine Science, 74: 81 – 99.
Wilkin, J. L., and Jeffs, A. G. 2011. Energetics of swimming to shore in the
puerulus stage of a spiny lobster: can a postlarval lobster afford
the cost of crossing the continental shelf? Limnology and
Oceanography: Fluids and Environments, 1: 163– 175.
Wolfe, S. H., and Felgenhauer, B. E. 1991. Mouthpart and foregut ontogeny in larval, postlarval, and juvenile spiny lobster, Panulirus
argus Latreille (Decapoda, Palinuridae). Zoologica Scripta, 20:
57 – 75.
Yeung, C., Jones, D. L., Criales, M. M., Jackson, T. L., and Richards, W. J.
2001. Influence of coastal eddies and counter-currents on the influx
of spiny lobster, Panulirus argus, postlarvae into Florida Bay. Marine
and Freshwater Research, 52: 1217– 1232.
Handling editor: C. Brock Woodson
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i147 –i154. doi:10.1093/icesjms/fsv008
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Into a rhythm: diel activity patterns and behaviour in
Mediterranean slipper lobsters, Scyllarides latus
Jason S. Goldstein1,2*, Elizabeth A. Dubofsky 3, and Ehud Spanier 1
1
The Leon Recanati Institute for Maritime Studies & Department of Maritime Civilizations, The Leon H. Charney School for Marine Sciences,
University of Haifa, Israel
2
Department of Biology, Eastern Connecticut State University, Willimantic, CT, USA
3
Department of Biological Sciences, New England Institute of Technology, East Greenwich, RI, USA
*Corresponding author: tel: +1 603 892 7135; fax: +1 860 465 5213; e-mail: [email protected]
Goldstein, J. S., Dubofsky, E. A., and Spanier, E. Into a rhythm: diel activity patterns and behaviour in Mediterranean slipper lobsters,
Scyllarides latus. – ICES Journal of Marine Science, 72: i147 – i154.
Received 29 August 2014; revised 2 January 2015; accepted 5 January 2015; advance access publication 30 January 2015.
Although the natural history for Mediterranean slipper lobsters (Scyllarides latus) is well established, there exists a disproportionate lack of important biological and physiological data to verify many key traits, including to what extent endogenous rhythms modulate aspects of their behaviour.
Although Scyllarids appear nocturnally active, few studies exist that quantify this tendency. Our overall objective was to test the hypothesis that
adult slipper lobsters are nocturnal and to determine if their diel activity rhythms are under the influence of an endogenous circadian clock. In the
laboratory, we exposed a total of 16 animals (CLavg ¼ 92.6 + 6.6 mm; CL, carapace length) to a 12 : 12 light : dark (LD) cycle for 7– 10 d, followed by
***constant dark (DD) for 15– 20 d. Activity was assessed using a combination of time-lapse video and accelerometers. Of a total of 16 lobsters, we
analysed data from 15 (one mortality). All 15 lobsters were evaluated using video. Thirteen of these lobsters were also evaluated using accelerometers. All lobsters were more active during night-time than during daytime and synchronized their activity to the LD cycle, expressing a diel
activity pattern (t ¼ 24.04 + 0.13 h). In DD, lobsters maintained a circadian rhythm with a t of 23.87 + 0.07 h. These findings may provide
insight into the behaviour of these animals in their natural habitat and help explain their ability to anticipate dawn and dusk.
Keywords: biological clock, circadian, endogenous rhythms, light– dark cycles, scyllaridae.
Introduction
In many marine crustaceans, rhythmic behaviours such as foraging,
locomotion, and overall activity are often associated with the
synchronization and modulation by circadian clocks (Naylor, 1988,
2005; reviewed in Arechiga et al., 1993). These rhythms are strongly
influenced by environmental factors, most notably light (Arechiga
and Rodriguez-Sosa, 1997; reviewed in Golombek and Rosenstein,
2010). Endogenous rhythms have been described for a variety of
marine crustaceans (reviewed in Arechiga and Rodriguez-Sosa,
1997) including crabs (Palmer, 1995; Forward et al., 2003),
shrimps (Adhub-Al and Naylor, 1975; Matthews et al., 1991),
isopods (Bohli-Abderrazak et al., 2012), amphipods (Rossano
et al., 2008), and copepods (Hough and Naylor, 1992). With
respect to lobsters, there are representative studies of activity
rhythms and locomotion of two general groups: spiny lobsters
(e.g. Panulirus argus; Lipcius and Herrnkind, 1982, 1985) and
# International
clawed lobsters (e.g. Homarus americanus; Jury et al., 2005). For
example, Atkinson and Naylor (1976) and Aguzzi et al. (2003,
2004) established the presence of endogenous rhythms in
Norwegian lobsters, Nephrops norvegicus, and the potential alterations to these activities based on light-limiting habitats (i.e. the continental slope). Jury et al. (2005) demonstrated that American
lobsters, H. americanus, were strongly influenced by an endogenous
circadian clock in both racetrack and running wheel laboratory
trials; however, there was large variability in this pattern. In a subsequent study, Golet et al. (2006) concluded that the variation in the
expression of diel in situ activity patterns might be driven by
changes in natural habitat, altering these predominantly nocturnal
behaviours. Two other spiny lobster species, Jasus edwardsii
and Panulirus japonicus, appear to exhibit strong, persistent
rhythms (under constant conditions) and are likely under the control of an endogenous circadian clock (Williams and Dean, 1989;
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i148
Nagata and Koike, 1997). In the latter species, direct and consistent
exposures to light can override the circadian clock, creating variation in locomotive activity among individuals. This phenomenon
has also been shown in other marine arthropods (e.g. horseshoe
crabs; Chabot et al., 2007). Therefore, while it is assumed that
many lobsters possess a nocturnal bias in their overall activity, the
behavioural output of these patterns can be flexible.
A third and major group of lobsters, the Scyllarids (slipper lobsters), also appear nocturnally active. However, the prevalence of endogenous rhythms in slipper lobsters remains elusive (see Lavalli
et al., 2007). Mediterranean slipper lobsters (Scyllarides latus) exemplify a commercially and biologically important species throughout
their range. Scyllarides latus are formidable, highly mobile marine
decapods that inhabit both shallow and relatively deep areas (4 –
100 m) of the Mediterranean Sea and eastern Atlantic Ocean from
the mainland coasts of Portugal south to Gambia and the islands
of Madeira, the Azores, and Cape Verde (Holthuis, 1991; Spanier
and Lavalli, 1998). Scyllarides latus are capable of attaining sizes
.100 mm carapace length (CL) and weights of 1.5 kg (Martins,
1985) and are the target of a selective, but overexploited, fishery
throughout their range (reviewed in Spanier and Lavalli, 1998).
Recently, this species has been listed under the IUCN Red List of
Threatened Species and is designated as “data deficient” (http://
www.iucnredlist.org/details/169983/0).
Although the natural history for S. latus has been largely disseminated, there remains a lack of important biological and physiological
data to substantiate many key behavioural traits, including to what
extent endogenous rhythms modulate locomotion and activity
(Spanier and Lavalli, 1998; Pessani and Mura, 2007). Tangential
studies suggest these animals display strong nocturnal behaviour indicative of an endogenous circadian rhythm (Spanier et al., 1988;
Spanier and Almog-Shtayer, 1992); however, this has not been
fully substantiated. Adult S. latus are found mainly on hard substrata
and rocky outcroppings in coastal waters; sheltering during the day
on the ceilings of highly recessed caves or crevices with minimal light
(Spanier and Almog-Shtayer, 1992; Barshaw and Spanier, 1994).
Some discrepancies in the display of nocturnal behaviour have
been noted in laboratory settings, especially in the absence of predators, resulting in a shift toward more diurnal activity (Barshaw and
Spanier, 1994). Interestingly, for the 88 species of Scyllarids
known, there is only one study that has formally described endogenous rhythms. Jones (1988) provided the most detailed analysis of
diel activity patterns for two shallow-water Scyllarid species
(Thenus indicus and T. orientalis), which varied from diel to nocturnal behaviours under naturalistic aquarium conditions. Based on
these contrasting patterns, the primary goal of the present study
was to provide better resolution of the activity patterns in S. latus
during LD cycles. Our hypothesis was that S. latus are nocturnal
under simulated LD periods and continue to display this behaviour
under constant conditions, suggestive of a circadian rhythm. We
further postulate that such behaviours are adaptive and advantageous to an animal that relies on cryptic behaviour and low-light
conditions to avoid predation.
Our approach was to use accelerometry as a tool for quantifying
the long-term activity associated with endogenous rhythms in
S. latus. The use of accelerometers is becoming widespread and
serves as a relatively low-cost, but highly effective method and has
been employed for other species of lobsters (clawed, H. americanus,
Lyons et al., 2013 and spiny, P. argus, Gutzler and Butler, 2014) and
allows for the collection of large datasets that can record activity
patterns and movement. We utilized time-lapse video recordings
J. S. Goldstein et al.
to help ground-truth the accelerometer data, as well as to observe
other behaviours associated with endogenous rhythms.
Materials and methods
Animal acquisition and environmental conditions
Adult, intermolt slipper lobsters used for this study (n ¼ 16,
CLavg ¼ 94.6 + 2.0 mm; CL range ¼ 81 – 105 mm) were obtained
two ways: (i) at a local fish market in in Akko (Acre, northern
Israel, N 328.9331; W 358.0827) where they were caught by
fishers*** using trammel nets in shallow (5–10 m) coastal areas
and obtained within 24 h of capture and (ii) using SCUBA in the
same vicinity. The lobsters were captured in the early summer
before the migration to cooler waters offshore (Spanier and
Lavalli, 1998). All lobsters were transported in coolers to the
marine biological laboratory at the Leon Recanati Institute for
Maritime Studies, University of Haifa where they were used in activity assays within 2 –3 weeks. Animals of both sexes (11 males and 5
females) were held communally in tanks with shelters in ambient
seawater (20 –228C, 40 psu salinity) and exposed to a 12 : 12 h
light : dark (LD) cycle and fed weekly a combination of bivalves
(Spondylus spinosus) and limpets (Patella caerulea). Daytime light
levels were produced by full-spectrum lights (Ikea Sparsam model
1230, 7 W, 45 lm/W) at an intensity of 185 + 0.64 lux (mean +
standard error of the mean (s.e.m.), 5.55 mmol photons/m2/s,
n ¼ 750 readings) during the day and 2.4 + 0.21 lux (0.072
mmol photons/m2/s, n ¼ 750 readings,) at night. Light levels were
measured using a Vernier Labquest Mini system with light sensor
(LS-BTA, Vernier Technology Co.). We determined ecologically appropriate light intensities for our holding tanks and trial simulations
in a preliminary study by securing HOBO light loggers in situ (n ¼
3, model UA0200, Onset Computer Corp., Bourne, MA, USA) to the
entrances of lobster dens at a depth of 10 –15 m (close to the lobster
capture site). The sensors were left for two weeks after which they
were retrieved and the data downloaded.
Experimental setup
Two activity arenas (60 × 40 × 24 cm (L × W × H), 57L, one
lobster/arena) were used for all trials and placed next to each
other, separated by an opaque divider, in a light-tight, climatecontrolled room. Before each trial, lobsters were fed to satiation.
Each arena included a round opaque PVC shelter (16 cm diameter)
that provided shelter and shade. Tanks were fitted with a filter system
(Mignon filter 60, AZOO) that allowed each tank (replicate) to run
as a closed system. Full-spectrum lighting was used and temperature
and light levels were maintained at levels described above. For the
first 8 –10 d, lobsters in each trial were subjected to 12 : 12 LD
cycles, controlled by a digital timer (Hyundai BND-50/IS8) programmed to a current exogenous sunrise –sunset schedule. All
light and temperature levels were recorded using HOBO Pendant
temperature–light loggers in 15-min intervals. After this initial
8 –10 d period, lobsters were exposed to constant darkness (DD)
for an average of 17 + 0.63 d. During this time, infrared lighting
(embedded in the camera and chosen for its minimal influence on
the visual sensitivity in crustaceans, inclusive to lobsters; Bruno
et al., 1977) was used to illuminate animal activity.
Activity assessment using video
Overall, lobster activity was recorded using a colour day–night
infrared-enabled CCD camera (PC177IRHR08, SuperCircuits,
Austin, TX, USA) mounted above the activity arenas. The camera
i149
Slipper lobster activity
output was then digitized (Canopus ADV5, Grass Valley, Hillsboro,
OR, USA) and archived to a computer (Macmini, OS 10.5.6,
Cupertino, CA, USA) using video capture software (Gawker,
v. 0.8.4, Seattle, WA, USA) that date–time stamped and recorded
the video at 24 FPS with a picture captured at each 15-s interval
(240 data points per hour). A subset of videos, categorized into
1-h time blocks, were analysed using Windows Media Player
Classic (v.1.2.1008). Activity was defined as forward, backward, or
sideways locomotion (.2 cm) for each lobster. We also observed
neutral behaviours (seen in the video and detected by the accelerometers), such as den-maintenance or antennual flicking.
However, these behaviours were excluded from data analyses as
they did not meet the threshold for activity, which was quantified
as locomotion rather than neutral behaviours. Activity was recorded
per hour by counting each minute during an hour the lobster was
active. Minutes active per hour were averaged over each light and
dark period (per 24-h calendar day) to give an average percentage
of time active per hour. Average time active per light or dark period
(per animal) were compared using a Student’s t-test. During video
analysis, we also made notes of other behaviours (e.g. digging,
climbing, etc.). The video data also served as a backup in the event
of any issues with the accelerometers.
Activity assessment using accelerometers
Concurrent with the video recordings, 13 of the original 16 lobsters
were fitted with an accelerometer (HOBO Pendant G, model:
UA-004-64, Onset Computer Corp., Bourne, MA, USA) capable
of recording and storing movement and activity over a three-axis
range at a high resolution (+3G). Accelerometers were fastened
to each lobster using a custom-built harness consisting of a
curved platform (9.4 cm2) with two embedded cable ties. The
harness was cemented, using a two-part putty-epoxy (AquaMend,
Polymeric Systems Inc., Elverson, PA, USA), to the dorsal carapace
of each animal and then sealed with liquid epoxy (Figure 1). This
configuration allowed the loggers to be easily removed, downloaded, and re-fastened at weekly intervals with minimal disturbance to the animals. Activity data were recorded by each
accelerometer every 30 s. Data were analysed to determine a threshold for activity, as the accelerometer was capable of picking up subtle
Figure 1. An accelerometer unit and mounting harness cemented to
the dorsal side of a Mediterranean slipper lobster, S. latus. The
accelerometer was used for quantifying movement activity during both
light:dark (LD) and constant dark (DD) conditions. Data were collected
at 30 s intervals; the accelerometer was removed and re-attached at
weekly intervals (see Methods for details).
movements in lobsters (e.g. twitching and antennual waving behaviours) that we did not consider activity (see Table 1 in Barshaw
and Spanier, 1994) and therefore filtered out. The accelerometer
data were plotted and analysed as actograms and Lomb-Scargle
periodograms (p , 0.05) using ClockLabw (MatLab, Actimetrics,
Evanston, IL, USA, v. R2009a), as described in Dubofsky et al.
(2013). The evaluation of peaks (10 –30 h) in the periodogram
was used to determine if the animals expressed a circadian
(24 h) rhythm and visual inspection of each actogram was also
used to substantiate data from the periodograms. Significant differences in activity level during the day and night were determined for
animals that expressed a diel rhythm in LD. Cumulative total activity
(per cent active per hour during light and dark hours) was calculated
for the entire trial period and analysed using a Student’s t-test. All
data were analysed using JMP Pro v.11.0 (SAS Institute, Inc.,
Cary, NC, USA) and represented +s.e.m. unless otherwise noted.
Results
Of a total of 16 lobsters, we analysed data from 15 (one mortality).
All 15 lobsters were evaluated using video. Thirteen of these lobsters
were also evaluated using accelerometers. Due to technical difficulties, two of the replicates did not provide accelerometer data.
Furthermore, we also determined if there were any sex-related behavioural differences. We used a series of t-tests to compare males
(n ¼ 11) and females (n ¼ 5) for both overall activity in light
(p ¼ 0.94) and dark (p ¼ 0.55) as well as periodogram data
(t values, LD, p ¼ 0.37; and DD, p ¼ 0.08); no significant differences between males and females were apparent. The variable t
represents the length of the activity rhythm.
Video data
Of the 15 lobsters for which we had video, we analysed a subset
(n ¼ 6) of individuals due to time restraints. However, we felt that
this was a representative sample, especially when we compared
these data with those from our periodogram analyses (completed
for 13 animals). During LD, lobsters were observed to be nocturnal
(Figure 2) which appeared to continue as a circadian rhythm
Figure 2. Average activity per hour over a 24-h period for six lobsters
(S. latus) over 9.33 + 0.56 days (the entire LD period for each lobster).
Data were collected via video. Over the 24 h period, we combined
lobster activity into four periods: (i) 07:00–14:00, (ii) 15:00– 19:00, (iii)
20:00– 03:00, and (iv) 04:00–06:00 corresponding to daytime, sunset,
night, and sunrise. The dashed vertical line “A” indicates sunrise (06:30),
dashed vertical line “B” indicates sunset (18:30). Activity between all
four times was significantly different (ANOVA; p , 0.001).
i150
during the DD period (17 + 0.63 d, active 47% of night-time vs.
14% of daytime).
Over the 24-h cycle, we combined lobster activity into four
periods: (i) 07:00– 14:00, (ii) 15:00 –19:00, (iii) 20:00 –03:00, and
(iv) 04:00 –06:00 corresponding to daytime, sunset, night-time,
and sunrise. Activity during these times was significantly different
(n ¼ 6, ANOVA; F3, 140 ¼ 116.5, MS ¼ 24710.8, p , .001,
Figure 2). Additionally, lobsters displayed markedly different activity levels between any of the given four periods (Tukey HSD; q ¼
2.60, p , 0.02, a ¼ 0.05, Figure 2). For example, between the
hours of 20:00 and 03:00 lobsters were more active (mean ¼
56.1 + 2.6%) compared with activity between 07:00 and 14:00
(mean ¼ 2.6 + 2.1%).
In addition to forward, backward, or sideways locomotion, we
noted a variety of other activities, including grooming, waving
(slow repetitive rocking movements), and antennual flicking (see
Supplementary Video, Barshaw and Spanier, 1994, Table 1). These
neutral behaviours also appeared to increase in frequency as sunset
approached and decreased before sunrise (see Supplementary
Video). When comparing video and accelerometer activity data,
we found no significant difference between the two methods
of collecting activity data (Pearson’s correlation, p , 0.0001,
R2adj = 0.898, Figure 3).
Accelerometer data
All 13 lobsters expressed a diel rhythm of activity when exposed to a
LD cycle (t ¼ 23.96 + 0.318 h, Figure 4). In addition, 12 of the 13
lobsters displayed significantly more activity at night during the
LD cycle (t-test, d.f. ¼ 11, 7, p ≤ 0.01, Figure 5). The remaining
lobster still showed a circadian trend (t ¼ 24.75 h, actogram not
shown); however, this animal appeared to express an increased
level of activity during the day. Overall, lobsters were active
59.05 + 1.37% per hour at night vs. 40.35 + 1.56% per hour
during the day. In constant conditions (DD), all lobsters expressed
a circadian rhythm of activity (t ¼ 23.85 + 0.134 h, Figure 4).
J. S. Goldstein et al.
Because their endogenous rhythms were shorter than 24 h, lobsters
became active earlier with each consecutive day in DD (Figure 4).
Discussion
Our results provide the first evidence that the diel rhythm of activity
in Mediterranean slipper lobsters, S. latus, is controlled by an endogenous clock (Figure 4). After synchronizing their activity patterns to an imposed LD cycle, these lobsters continued to express
the same activity pattern in constant conditions, maintaining a circadian rhythm for the duration of the trial (Figures 4 and 5). Few
studies have examined diel activity patterns in Scyllarid lobsters
despite their economic and ecological importance throughout
their range (Lavalli et al., 2007). Jones (1988) has provided the
only other detailed analysis of diel activity patterns for Scyllarid
species (T. indicus and T. orientalis). Interestingly, our study
animals demonstrated more activity during the day than expected,
some of which was associated with grooming and den-maintenance
behaviours. Both evaluative methods (video and accelerometer)
showed a baseline of activity along with neutral movements
during the day, but were significantly different from activity at
night (Figures 2 and 5).
Our goal was to examine activity rhythms in both LD and DD
without the influence of any other exogenous cues (e.g. predators,
chemical cues, etc.) besides light. Although other laboratory
studies suggest the shifting of S. latus to a diurnal rhythm when predators are not encountered (Barshaw and Spanier, 1994), we did not
see this kind of behaviour. Our trials were focused on assessing individual activity patterns despite the propensity of this species to
cohabitate with conspecifics. Previous studies of group behaviour
may have different outcomes (Barshaw and Spanier, 1994), accounting for results that may differ from this study. In daytime field
surveys by Spanier and Almog-Shtayer (1992), .50% of lobsters
censused were found in dens or artificial reefs cohabitating with conspecifics. The influence of conspecifics and predators and their
effects on circadian rhythms is still not well understood.
Figure 3. A methodological comparison between lobster (S. latus) activity measured using accelerometry and video analysis. The relationship
between video and accelerometer data for an individual lobster over the entire testing period. There was no significant difference between the two
methods of collecting activity data (Pearson’s correlation, p , 0.0001). R2adj = 0.898 (Video ¼ 22.115246 + 0.9779221*Accel).
Slipper lobster activity
i151
Figure 4. Characteristic locomotor activity recorded from two Mediterranean slipper lobsters (S. latus) during LD and DD periods of the study.
Both animals showed a nocturnal pattern of activity including drifting during periods of DD. Each lobster was exposed to an imposed LD followed by
DD, as indicated on left side of the figure. (a) LD ¼ 9 days, DD ¼ 16 days; (b) LD ¼ 10 days, DD ¼ 16 days. The imposed LD cycle is indicated at the
top of the image; dark portions indicate night-time. Images on the left are double-plotted actograms, with periods of activity indicated by vertical
black lines. Horizontal scale is time of day (hours) and activity is indicated in metre per second squared. Images on the right are Lomb-Scargle
periodograms for the same animal during one of the two treatments (LD top, DD bottom). Vertical scale is the relative strength of rhythmicity
(Q(p)); the horizontal scale is the period in hours (5 – 30); the largest peak above the horizontal line of significance (p , 0.05) is indicated by a
numerical value.
Circadian rhythms are driven by internal biological clocks that
have a period of 24 h, allowing animals to anticipate sunset and
sunrise and, as a result, modulate changes in their behaviour or
physiology (Naylor, 2005; Kronfeld-Schor et al., 2013). Circadian
clocks are typically entrained by external stimuli (light); however,
they tend to persist in the absence of external cues. In the present
study, we observed a noticeable increase of activity in our lobsters
before sunset and an opposing decline in activity just before
sunrise (Figure 2).
Adaptation to light
Scyllarides latus appear highly adapted to living in low-light environments and it is well known that these animals are found in
dimly lit conditions, taking refuge in crevices and rocky outcrops
along coastal ridges (Spanier and Almog-Shtayer, 1992). Despite
our simulated light levels being 5 –10% stronger than what we
measured in situ, all our animals exhibited a strong endogenous
rhythm. How these lobsters sense and react to changes in light
levels is not well known, although recent work by Lau et al. (2009)
determined S. latus possess “superposition optics” (i.e. sensitivity
to a light source affected by the number of facets by which light is
collected and focused in the random) with respect to their vision
that allows for a balance between optimal sensitivity and acuity.
This optical mechanism is specifically adapted to dimly lit environments, and it is theorized that organisms equipped with superposition optics are excellent candidates for circadian rhythms (Lau
et al., 2009). We surmise that, like other nocturnal marine crustaceans, there is an increase in visual sensitivity that assists in prey
i152
Figure 5. Comparison of average activity and light levels over 5 d for
one lobster, S. latus (CL ¼ 87 mm). Activity was recorded every 30 s via
accelerometer. Average activity was calculated by summing the two
activity readings per minute then averaging the sum per minute over
five consecutive days. Activity per hour at night averaged 59.05 +
1.37% and was significantly different from daytime activity per hour
(40.35 + 1.56%, p , 0.01). Light levels were recorded every 15 min
(light sensor) and ranged from 0 to 350 lx (0–10.5 mmol photons/m2/s).
Overall, activity levels increased dramatically as light levels decreased.
capture or predator avoidance (Meyer-Rochow, 2001), which would
be advantageous for lobsters like S. latus in their natural habitat.
Furthermore, it is proposed that the synchronization of circadian
rhythms through visual inputs provides an enhanced level of protection aiding in other activities (Meury and Gibson, 1990; Pizzatto
et al., 2008; Lee, 2010). Future work to assess the visual sensitivity
of S. latus in relation to their activity (e.g. ERG assay, Watson
et al., 2008) and examining the role of circadian clocks would
serve as a useful follow-up to this study.
Although the lobsters used in these trials were all obtained from
relatively shallow locations (10 –15 m), Mediterranean slipper lobsters are distributed throughout a gradient of depths, including
areas of the continental slope .50 m. It is not known how S. latus
obtained from greater depths might respond to LD cycles given
the more weakened light intensities to which they are accustomed.
Norwegian lobsters (N. norvegicus) are known to entrain to circadian rhythms (Aguzzi and Sarda, 2008) and have also been shown
to vary their burrow emergence rates between dusk and dawn
(Sbragaglia et al., 2013). However in lab-based trials, exposure to
varying light intensities (representative of depths from slope to
shelf waters, 10– 400 m), resulted in the lobsters switching from
nocturnal to diel activity. This suggests that light intensity and composition help to modulate a “behavioural temporal niche” (Chiesa
et al., 2010).
Behavioural adaptations
Endogenous clocks in mobile but cryptic species, such as S. latus,
provide a selective advantage by enabling them to anticipate environmental change and return to their den before the onset of unfavourable conditions (Naylor, 1988, 2005). It is proposed that,
like spiny (Lipcius and Herrnkind, 1982) and clawed (Karnofsky
et al., 1989) lobsters, S. latus forage exclusively at night to reduce
their encounter with diurnally active predators, including triggerfish and grouper (Martins, 1985; Barshaw and Spanier, 1994).
Maintaining strong, consistent locomotor rhythms is an adaptive
J. S. Goldstein et al.
trait to maximize foraging and predator avoidance. Because
S. latus predominately reside in low-light habitats, possessing an endogenous rhythm is essential to regulate optimal times for foraging
vs. sheltering.
The capacity to alter endogenous clocks has been shown in a
variety of organisms and the timing and type of change depends
on the circumstances that modify the existing rhythm. For
example, age may cause cane toads (Bufo marinus) to switch from
a nocturnal to a diel pattern of activity (Pizzatto et al., 2008) while
amphipods (Orchestoidea tuberculata) make a transition from a
nocturnal to a crepuscular pattern of activity (Kennedy et al.,
2000). Both of these organisms alter activity rhythms to avoid predation, especially by adult conspecifics. These patterns are noticeable in some lobsters as well, for example, ovigerous N. norvegicus
(when compared with non-ovigerous females) appear to decrease
their range and duration of emergence from dens (Aguzzi et al.,
2007).
Ontogenetic changes in the rhythms of S. latus are not known.
However, seasonal movements of coastal populations (Martins,
1985; Spanier et al., 1988) suggest that modifications in activity
may originate from physiological, neurological, or metabolic
changes as opposed to endogenous clocks (Roenneberg and
Merrow, 2002; Helm et al., 2013). The initiation and environmental
trigger(s) responsible for these seasonally-induced movements are
unknown, but are theorized to include changes in temperature
as well as photoperiod (Spanier et al., 1988; Spanier and Lavalli,
1998), similar to what is observed in spiny lobsters (Herrnkind,
1983). Alternatively, circannual clocks may serve as the mechanism
behind changes in temporal and spatial distributions as is evident in
other species such as migratory birds (Gwinner, 1996). We collected
and tested for evidence of circadian rhythms in S. latus over nonmigratory periods and purposefully used lobsters of intermolt
phase to avoid potential biases. However, it would be of interest to
investigate circannual rhythms and their zeitgebers that may be
associated with migration, molting, or mating.
The establishment of a circadian rhythm in Mediterranean
slipper lobsters is an important descriptor of behaviour in this
species. Our results may be useful for investigating the timing of biological processes, seasonal migrations, and responses to changing
environmental conditions due to global climate change.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
This manuscript was submitted as a contribution to the Proceedings
of the 10th International Conference and Workshop on Lobster
Biology and Management, held in Cancun, Mexico, May 2014.
We sincerely thank J.J. Gottlieb, Mia Elsar, and Joseph Pacheco for
their help with husbandry, maintenance, and video analyses
throughout this study. We thank Rachael Pollack for her illustration
of Figure 1. We are also grateful to Amir Yurman and Eyal Miller for
their assistance in finding and collecting live lobsters in the field.
A special thanks to Rebecca Kibler and Win Watson for helpful
edits and comments that improved this manuscript. This research
project was supported by a Sir Morris Hatter Research Grant and
a United States-Israel Fulbright Postdoctoral Fellowship to J.S.G.
Slipper lobster activity
References
Adhub-Al, A. H. Y., and Naylor, E. 1975. Emergence rhythms and tidal
migrations in the brown shrimp Crangon crangon (L.). Journal of the
Marine Biological Association of the UK, 55: 801 –810.
Aguzzi, J., Company, J. B., and Abello, P. 2004. Locomotor activity
rhythms of continental slope Nephrops norvegicus (Decapoda:
Nephropidae). Journal of Crustacean Biology, 24: 282 – 290.
Aguzzi, J., Company, J. B., and Sarda, F. 2007. The activity rhythm of
berried and unberried females of Nephrops norvegicus (Decapoda,
Nephropidae). Crustaceana, 80: 1121– 1134.
Aguzzi, J., and Sarda, F. 2008. A history of recent advancements on
Nephrops norvegicus behavioural and physiological rhythms.
Reviews in Fish Biology and Fisheries, 18: 235 – 248.
Aguzzi, J., Sarda, F., Abello, P., Company, J. B., and Rotllant, G. 2003.
Diel and seasonal patterns of Nephrops norvegicus (Decapoda:
Nephropidae) catchability in the western Mediterranean. Marine
Ecology Progress Series, 258: 201– 211.
Arechiga, H., Fernandez-Quiroz, F., Fernandez de Miguel, F., and
Rodriguez-Sosa, L. 1993. The circadian system of crustaceans.
Chronobiology International, 10: 1 – 19.
Arechiga, H., and Rodriguez-Sosa, L. 1997. Coupling of environmental
and endogenous factors in the control of rhythmic behaviour in
decapod crustaceans. Journal of the Marine Biology Association
UK, 77: 17 – 29.
Atkinson, R. J., and Naylor, E. 1976. An endogenous activity rhythm and
the rhythmicity of catches of Nephrops norvegicus (L). Journal of
Marine Biology and Ecology, 25: 95– 108.
Barshaw, D. E., and Spanier, E. 1994. The undiscovered lobster – a first
look at the social behavior of the Mediterranean slipper lobster,
Scyllarides latus Decapoda, Scyllaridae. Crustaceana, 67: 187– 197.
Bohli-Abderrazak, D., Ayari, A., Morgan, E., and Nasri-Ammar, K.
2012. Towards a characterization of the locomotor activity rhythm
of the supralittoral isopod Tylos europaeus. Chronobiology
International, 29: 166 – 174.
Bruno, M. S., Barnes, S. N., and Goldsmith, T. H. 1977. The visual
pigment and visual cycle of the lobster, Homarus. Journal of
Comparative Physiology, 120: 123– 142.
Chabot, C. C., Betournay, S. H., Braley, N. R., and Watson, W. H., III.
2007. Endogenous rhythms of locomotion in the American horseshoe crab, Limulus polyphemus. Journal of Experimental Marine
Biology and Ecology, 345: 79 – 89.
Chiesa, J. J., Aguzzi, J., Garcia, J. A., Sarda, F., and de la Iglesia, H. O.
2010. Light intensity determines temporal niche switching of behavioral activity in deep-water Nephrops norvegicus (Crustacea:
Decapoda). Journal of Biological Rhythms, 25: 277 – 287.
Dubofsky, E. A., Simpson, S. D., Chabot, C. C., and Watson, W. H., 3rd.
2013. Patterns of activity expressed by juvenile horseshoe crabs.
Biological Bulletin, 225: 42– 49.
Forward, R. B., Jr., Tankersley, R. A., and Pochelon, P. N. 2003. Circatidal
activity rhythms in ovigerous blue crabs, Callinectes sapidus: implications for ebb-tide transport during the spawning migration.
Marine Biology, 142: 67 – 76.
Golet, W. J., Scopel, D. A., Cooper, A. B., and Watson, W. H., 3rd. 2006.
Daily patterns of locomotion expressed by American lobsters
(Homarus americanus) in their natural habitat. Journal of
Crustacean Biology, 26: 610– 620.
Golombek, D. A., and Rosenstein, R. E. 2010. Physiology of circadian
entrainment. Physiological Review, 90: 1063– 1102.
Gutzler, B. C., and Butler, M. J., IV. 2014. Accelerometry as a tool for
studying lobster behavior: preliminary results from the Florida
Keys, FL (USA). The Lobster Newsletter, 27: 8 – 9.
Gwinner, E. 1996. Circadian and circannual programmes in avian migration. The Journal of Experimental Biology, 199: 39 – 48.
Helm, B., Ben-Shlomo, R., Sheriff, M. J., Hut, R. A., Foster, R., Barnes,
B. M., and Dominoni, D. 2013. Annual rhythms that underlie
i153
phenology: biological time-keeping meets environmental change.
Proceedings of the Royal Society B, 280: 20130016.
Herrnkind, W. F. 1983. Movement patterns and orientation. In The
Biology of Crustacea, III, pp. 41 –105. Ed. by D. E. Bliss. Academic
Press, New York. 479 pp.
Holthuis, L. B. 1991. Marine lobsters of the world. FAO Species
Catalogue, FAO Fisheries Synopsis No. 125, 13, Rome: FAO, 292 pp.
Hough, A. R., and Naylor, E. 1992. Endogenous rhythms of circatidal
swimming activity in the estuarine copepod Eurytemora affinis
(Poppe). Journal of Experimental Marine Biology and Ecology,
161: 27– 32.
Jones, C. 1988. The biology and behavior of Bay lobsters, Thenus spp.
(Decapoda: Scyllaridae), in Northern Queensland, Australia. PhD
dissertation. University of Queensland, Brisbane, Australia.
Jury, S. H., Chabot, C. C., and Watson, W. H., III. 2005. Daily and circadian rhythms of locomotor activity in the American lobster,
Homarus americanus. Journal of Experimental Marine Biology and
Ecology, 318: 61 – 70.
Karnofsky, E. B., Atema, J., and Elgin, R. H. 1989. Field observations of
social behavior, shelter use, and foraging in the lobster, Homarus
americanus. Biological Bulletin, 176: 239 – 246.
Kennedy, F., Naylor, E., and Jaramillo, E. 2000. Ontogenetic differences
in the circadian locomotor activity rhythm of the talitrid amphipod
crustacean Orchestoidea tuberculata. Marine Biology, 137: 511 – 517.
Kronfeld-Schor, N., Bloch, G., and Schwartz, W. J. 2013. Animal clocks:
When science meets nature. Proceedings of the Royal Society B., 280:
20131354.
Lau, S. T., Spanier, E., and Meyer-Rochow, V. B. 2009. Anatomy and
ultrastructural organization of the eye of the Mediterranean
slipper lobster, Scyllarides latus: preliminary results. New Zealand
Journal of Marine and Freshwater Research, 43: 233 –245.
Lavalli, K. L., Spanier, E., and Grasso, F. 2007. Behavior and sensory
biology of slipper lobsters. In The Biology and Fisheries of the
Slipper Lobster, pp. 133 – 181. Ed. by K. Lavalli, and E. Spanier.
CRC Press, Boca Raton, FL. 416 pp.
Lee, W. 2010. Intensive use of an intertidal mudflat by foraging adult
American horseshoe crabs Limulus polyphemus in the Great Bay
estuary, New Hampshire. Current Zoology, 56: 611– 617.
Lipcius, R. N. M., and Herrnkind, W. F. 1982. Molt cycle alterations in
behavior, feeding and diel rhythms of a decapod crustacean, the
spiny lobster Panulirus argus. Marine Biology, 68: 241 –252.
Lipcius, R. N. M., and Herrnkind, W. F. 1985. Photoperidoic regulation
and daily timing of spiny lobster mating behavior. Journal of
Experimental Marine Biology and Ecology, 89: 191 – 204.
Lyons, G. N., Halsey, L. G., Pope, E. C., Eddington, J. D., and Houghton,
J. D. R. 2013. Energy expenditure during activity in the American
lobster Homarus americanus: Correlations with body acceleration.
Comparative Biochemistry and Physiology, Part A, 166: 278– 284.
Martins, H. R. 1985. Biological studies of the exploited stock of the
Mediterranean locust lobster Scyllarides latus (Latreille, 1803)
(Decapoda: Scyllaridae) in the Azores. Journal of Crustacean
Biology, 5: 294 – 305.
Matthews, T. R., Schroeder, W. W., and Stearns, D. E. 1991. Endogenous
rhythm, light and salinity effects on postlarval brown shrimp
Penaeus aztecus Ives recruitment to estuaries. Journal of
Experimental Marine Biology and Ecology, 154: 177 – 189.
Meury, T. W., and Gibson, D. G., III. 1990. Force generation in juvenile
Limulus polyphemus: effects on mobility in the intertidal environment. Bulletin of Marine Science, 47: 536– 545.
Meyer-Rochow, V. B. 2001. The crustacean eye: dark/light adaptation,
polarization sensitivity, flicker fusion frequency, and photoreceptor
damage. Zoological Science, 18: 1175– 1197.
Nagata, Y. U., and Koike, T. 1997. Collapse of the diurnal variation
pattern of lobster activity and its causes. Bulletin of Marine
Science, 61: 129– 138.
Naylor, E. 1988. Rhythmic behavior of decapod crustaceans. Symposia
of the Zoological Society of London, 59: 177– 199.
i154
Naylor, E. 2005. Chronobiology: implications for marine resource exploitation and management. Scientia Marina, 69: 157– 167.
Palmer, J. D. 1995. The Biological Rhythms and Clocks of Intertidal
Animals. Oxford University Press, New York.
Pessani, D., and Mura, M. 2007. The biology of the Mediterranean
Scyllarids. In The Biology and Fisheries of the Slipper Lobster, pp.
264– 286. Ed. by K. Lavalli, and E. Spanier. CRC Press, Boca
Raton, FL. 416 pp.
Pizzatto, L., Child, T., and Shine, R. 2008. Why be diurnal? Shifts in activity time enable young cane toads to evade cannibalistic conspecifics. Behavioral Ecology, 19: 990– 997.
Roenneberg, T., and Merrow, M. 2002. Life before the clock: modeling
circadian evolution. Journal of Biological Rhythms, 17: 495– 505.
Rossano, C., Morgan, E., and Scapini, F. 2008. Variation of the locomotor activity rhythms in three species of talitrid amphipods,
Talitrus saltator, Orchestia montagui, and O. gammarellus, from
various habitats. Chronobiology International, 25: 511– 532.
Sbragaglia, V., Aguzzi, J., Garcia, J. A., Chiesa, J. J., Angelini, C., and
Sarda, F. 2013. Dusk but not dawn burrow emergence rhythms of
J. S. Goldstein et al.
Nephrops norvegicus (Crustacea: Decapoda). Scientia Marina, 77:
641– 647.
Spanier, E., and Almog-Shtayer, G. 1992. Shelter preferences in the
Mediterranean slipper lobster: effect of physical properties.
Journal of Experimental Marine Biology and Ecology, 164: 103 – 116.
Spanier, E., and Lavalli, K. L. 1998. Natural history of Scyllarides latus
(Crustacea Decapoda): a review of the contemporary biological
knowledge of the Mediterranean slipper lobster. Journal Natural
History, 32: 1769– 1786.
Spanier, E., Tom, M., Pisanty, S., and Almog, G. 1988. Seasonality and
shelter selection by the slipper lobster Scyllarides latus in the southeastern Mediterranean. Marine Ecology Progress Series, 42:
247– 255.
Watson, W. H., III, Bedford, L., and Chabot, C. 2008. Rhythms of locomotion expressed by Limulus polyphemus, the American Horseshoe
crab: II. Relationship to circadian rhythms of visual sensitivity.
Biological Bulletin, 215: 46– 56.
Williams, B. G., and Dean, I. C. 1989. Timing of locomotor activity in
the New Zealand rock lobster, Jasus edwardsii. New Zealand
Journal of Marine and Freshwater Research, 23: 215 –224.
Handling editor: Howard Browman
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i155 –i163. doi:10.1093/icesjms/fsu219
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Aggressive behaviour of spotted spiny lobsters (Panulirus guttatus)
in different social contexts: the influence of sex, size,
and missing limbs
Patricia Briones-Fourzán*, Roberto Domı́nguez-Gallegos, and Enrique Lozano-Álvarez
Instituto de Ciencias del Mar y Limnologı́a, Unidad Académica de Sistemas Arrecifales, Universidad Nacional Autónoma de México, Puerto Morelos,
Quintana Roo 77580, Mexico
*Corresponding author: tel: +52 998 8710009; fax: +52 998 8710138; e-mail: [email protected]
Briones-Fourzán, P., Domı́nguez-Gallegos, R., and Lozano-Álvarez, E. Aggressive behaviour of spotted spiny lobsters (Panulirus
guttatus) in different social contexts: the influence of sex, size, and missing limbs. – ICES Journal of Marine Science, 72: i155 –i163.
Received 14 August 2014; revised 31 October 2014; accepted 3 November 2014; advance access publication 26 November 2014.
Panulirus guttatus is a sedentary spiny lobster that exhibits cryptic behaviour and a low degree of gregariousness. Because these lobsters are
obligate coral reef-dwellers and avoid sandy expanses, they are potentially distributed in relatively small, discrete populations with variable
social contexts, which can strongly influence the expression of aggression. The present study examined the relative importance of sex, size,
and the number of missing limbs in the shelter-related aggressive behaviour of replicated groups of four lobsters that differed in social
context (i.e. same-sex and mixed-sex groups). Each group was held in a seawater tank with a single artificial cave-like shelter. The interior of
the cave was video-recorded for 72 h and the number of aggressions performed by each individual was quantified in a 10-min segment of
video per hour. Most aggressions were related to occupancy of the shelter inner space and tended to end when individuals were sufficiently
spaced out. In general, per-capita rates of aggression were higher at night and size was an important predictor of aggressiveness among individuals of the same sex. Males were substantially more aggressive than females, but the number of missing limbs significantly impacted
the degree of aggressiveness in males. In mixed-sex groups, fewer aggressions occurred when the largest individual was a male than when it
was a female, suggesting that it may take longer for smaller males to assert themselves. Our results provide insights into some potential consequences of increase in fishing pressure and loss of habitat complexity in Caribbean reefs for the social behaviour and population dynamics
of these lobsters.
Keywords: aggression, agonistic interactions, Caribbean coral reefs, social behaviour, social context, spiny lobsters.
Introduction
In many animal systems, aggressive behaviour is an important
component of agonistic and social behaviour that includes actual
attacks or any threats of attack (Huber and Kravitz, 2010).
Although aggression and dominance are not necessarily correlated
(Hand, 1986; Drews, 1993), large crustacean males are typically
more aggressive and tend to dominate smaller conspecifics, often
monopolizing resources such as food, mates, or shelter (Lee and
Fielder, 1983; Thorpe et al., 1994; Edsman and Jonsson, 1996;
Beattie et al., 2012). In clawed decapods (e.g. crabs, crayfish, and
clawed lobsters), the claws constitute potentially deadly weapons
and hence agonistic interactions involve ritualized displays that
serve as tests of strength rather than actual attacks (Neil, 1985;
# International
Huber and Kravitz, 1995; Davis and Huber, 2007). In contrast, in
the clawless spiny lobsters (Decapoda: Achelata: Palinuridae),
ritualized threat behaviours have not evolved and agonistic behaviour can involve a great deal of contact and struggle, but is less
likely to result in injury or death (Atema and Cobb, 1980; Cobb,
1981).
Spiny lobsters are large, nocturnally active crustaceans that
depend on available structured shelter to hide from predators. Spiny
lobsters exhibit gregarious sheltering, a behaviour mediated by conspecific chemical communication (Childress, 2007). Although the
degree of gregariousness varies with species, even in highly gregarious species individuals can behave aggressively towards conspecifics
attempting to share shelters and may form dominance hierarchies
Council for the Exploration of the Sea 2014. All rights reserved.
For Permissions, please email: [email protected]
i156
that determine their position within the shelter (Fielder, 1965;
Berrill, 1976; McKoy, 1979; Cobb, 1981; Shabani et al., 2009).
Panulirus guttatus, the Caribbean spotted spiny lobster, is a relatively small, coral reef-obligate palinurid (Sharp et al., 1997). Individuals
of P. guttatus have a limited home range (Lozano-Álvarez et al., 2002;
Negrete-Soto et al., 2002) and hide deeply in reef crevices from which
they emerge only for brief periods at night to forage on the reef itself
(Wynne and Côté, 2007). Although P. guttatus lobsters are attracted
to conspecific odours, especially when reproductive activity is high
(Briones-Fourzán and Lozano-Álvarez, 2005), they exhibit a low
propensity to aggregate, and when multiple individuals co-occupy
a reef crevice, they tend to space out in separate cavities or locations
within the crevice (Childress, 2007; Briones-Fourzán et al., 2013).
Similarly, in a laboratory setting using artificial cave-like shelters,
large males aggressively displaced smaller individuals or evicted
them altogether from the cave, but aggressions subsided once each
lobster secured a separate location within the cave (Lozano-Álvarez
and Briones-Fourzán, 2001; Segura-Garcı́a et al., 2004). This spacing
behaviour of P. guttatus contrasts with other spiny lobsters, wherein
individuals sharing crevices typically maintain close contact with
each other. Therefore, the distribution of P. guttatus lobsters over
the reef habitat is modulated by a complex interplay between chemical attraction and aggressive behaviour (Briones-Fourzán and
Lozano-Álvarez, 2005).
Although large males of P. guttatus are generally more aggressive
than smaller conspecifics (Lozano-Álvarez and Briones-Fourzán,
2001; Segura-Garcı́a et al., 2004), this does not imply that large
males should dominate all agonistic encounters because all individuals, irrespective of sex and size, have the potential to express
aggression and the observed aggressiveness is a circumstantial
product of the cost/benefit ratio of using it (Drews, 1993; Huber
and Kravitz, 2010). Moreover, aggressiveness can be strongly influenced by social context (Karnofsky et al., 1989; Graham and
Herberholz, 2009; Beattie et al., 2012), which is particularly relevant
to P. guttatus because these lobsters are restricted to a highly patchy
habitat and hence potentially distributed in fragmented populations
with variable social contexts (Sharp et al. 1997; Negrete-Soto et al.,
2002; Robertson and Butler, 2009; Briones-Fourzán et al., 2013).
Given the great amount of time that P. guttatus lobsters spend sheltered, we examined in the laboratory the influence of social context
(i.e. same-sex and mixed-sex groups) on their shelter-related aggressive behaviour. In particular, we were interested in determining
the effect of social context on the rate of aggression, whether individual traits affect aggressiveness of lobsters in a given social context,
and whether there is a significantly more aggressive individual in
every lobster group irrespective of social context. Following the
observations of Lozano-Álvarez and Briones-Fourzán (2001) and
Segura-Garcı́a et al. (2004), we predicted that the rate of aggression
would be significantly affected by social context, with males exhibiting higher levels of aggressiveness than females, but that the largest
individuals would exhibit a higher degree of aggressiveness regardless of social context.
Material and methods
Lobster collection and holding
Lobsters were collected from the Puerto Morelos coral reef (eastern
margin of the Yucatan peninsula, Mexico), put into a seawater
container in a boat and transported to our laboratory, located at
Puerto Morelos (20852.1′ N, 86852.1′ W), within an hour of collection. Lobsters were allowed to acclimate for up to 8 d in large
P. Briones-Fourzán et al.
tanks fitted with plenty of individual shelters to reduce the probability of shelter-related fights. The shelters were segments of
PVC pipe sealed on one end, both to make them darker and to
allow the lobsters to fully retreat into them (Briones-Fourzán and
Lozano-Álvarez, 2005). The holding tanks (and the experimental
units mentioned below) received seawater from the Puerto
Morelos reef lagoon via an open-flow system and were in the open
but under a translucent shade and hence subject to ambient light.
During the holding period, lobsters were fed ad libitum every
other day with frozen crabs and mussels, previously thawed. All
experimental lobsters were adults.
Experimental design
The experimental units were three fibreglass tanks (2 m in diameter
and 0.9 m in depth, with 2500 l of seawater), each containing a
single artificial shelter in the form of a cave with two low entrances.
The cave was constructed with an inverted plastic trash bin measuring 0.45 m in diameter and 0.75 m in height with the rim weighed to
increase stability. Because P. guttatus lobsters tend to cling to the
walls or ceiling of reef crevices, the interior of the cave was lined
with wire mesh slightly separated from the walls to provide gripping
surface for the lobsters. The mesh reduced the inner height of
the cave to 0.50 m. The top of the cave had two holes. One hole
held a surveillance video camera (Sony CC-2) encased in an underwater Plexiglas housing (Ikelite 5210). The other hole held an underwater lamp that produced a dim red light that provided sufficient
illumination for nocturnal recordings of the behaviour of lobsters
inside the cave without disturbing the lobsters (Weiss et al., 2006).
Although this shelter design is rather crude, it provides important
den requirements for P. guttatus [height, gripping surface, and darkness, see Lozano-Álvarez and Briones-Fourzán (2001)] while allowing the unhindered video-recording of all lobsters in its interior
(Segura-Garcı́a et al., 2004; Weiss et al., 2006). Lobsters were distinctively marked with strips of reflecting tape glued to the carapace
to provide individual recognition.
The experiment consisted of four treatments that differed in
social context and relative size of lobsters. The treatments denoted
2MF and 2FM consisted of mixed-sex groups (two males and two
females of varied sizes), but in treatment 2MF the males were
larger than the females, whereas in treatment 2FM one of the
females was at least 5 mm carapace length (CL) larger than the
largest male. The treatments denoted 4M and 4F were same-sex
groups consisting of either four males (4M) or four females (4F)
of varied sizes. Each treatment was replicated three times. The
experiment was conducted over winter–spring of 2003, but due
to the limited number of experimental units, the treatments were
not run simultaneously (Table 1). To ensure that no lobster
odours remained between treatments, at the end of each treatment
all shelters were extracted from the tanks, brushed, rinsed, and sun
dried, and the tanks were drained and thoroughly brushed.
Table 1. Dates and relevant characteristics of experimental periods
for each treatment.
Water
temperature
Treatment Dates
Sunrise (h) Sunset (h) (88 C, mean + SD)
2MF
04 –06 December 06:10
17:05
25.0 + 1.9
4M
21 –23 April
05:22
18:09
26.1 + 1.0
4F
28 –30 April
05:17
18:11
28.0 + 0.6
2FM
09 –11 June
06:05
19:29
27.8 + 0.6
i157
Behaviour of lobsters in different social contexts
To avoid the potentially confounding effects of moult cycle and
prior residence on aggressive behaviour (Atema and Cobb, 1980;
Lipcius and Herrnkind, 1982; Peeke et al., 1998), we used exclusively lobsters in intermoult as determined through microscopic
observation of the tip of a pleopod (Lyle and MacDonald, 1983),
and introduced the four lobsters in each group into the corresponding tank at the same time. All lobsters quickly took shelter
in the cave, which was sufficiently large so as to loosely accommodate the entire group. Also, to avoid potential competition for
food, which can occur among lobsters in captivity (Thomas
et al., 2003), lobsters were not fed for the duration of the experiment (72 h, see below). However, upon collection many lobsters
were missing one or more limbs and several females were ovigerous
(Table 2), as individual females of P. guttatus produce multiple
clutches per year (Negrete-Soto et al., 2002). Although inclusion
of these factors was not contemplated originally, we included
them a posteriori because aggressive behaviour can be affected by
the lack of limbs (Juanes and Smith, 1995; Edsman and Jonsson,
1996) and, in females, by reproductive state (i.e. ovigerous vs. nonovigerous; Figler et al., 1995; Mello et al., 1999).
After the lobsters acclimated to the experimental tanks for 24 h,
the inside of the caves was continuously recorded for 72 h starting at
00:00 h of Day 1. A multiplexor (Panasonic WJ-FS216) allowed the
simultaneous recording of the interior of the three caves on a timelapse video recorder (Sony SVT-L400). We randomly chose 10 min
from each hour of recording (one 5-min segment per each halfhour) to quantify the number of aggressions performed by each
lobster (Martin and Bateson, 1993). Aggressions are defined as
those offensive behaviours (i.e. attacks or threats) that elicit avoidance or escape behaviours in the opponent (Edwards and
Herberholz, 2005; Table 3). All lobsters were used in one treatment
only and were released back to the coral reef habitat at the end of the
treatments.
Table 2. Summary of data for lobsters used in experiment.
Replicate
Treatment unit
2MF
A
B
C
4M
B
C
4F
A
B
C
Data analyses
We were interested in determining (i) the effect of social context
(treatment) and time of day (i.e. night vs. day) on the per-capita
rate of aggression (i.e. aggressions per hour per individual, Johnson
et al., 2004), (ii) whether individual traits (size, sex, number of
missing limbs, and reproductive state) affected aggressiveness of individuals in a given social context, and (iii) whether irrespective of social
context there was a significantly more aggressive individual in every
group.
We did a factorial general linear model to examine the effects of
social context (four levels corresponding to the four treatments) and
time of day (two levels: day and night) on the per-capita rate of
aggression (aggressions per individual per hour, Johnson et al.,
2004; response variable). We included time of day as a fixed
factor because agonistic encounters tend to increase with the level
of activity (Capelli and Hamilton, 1984), which for P. guttatus is substantially higher during the night (Lozano-Álvarez and BrionesFourzán, 2001). For every treatment, we averaged the data from
each unit across the night (from the hour of sunset to the hour of
sunrise) and the day (from 1 h after sunrise to 1 h before sunset).
Before analysis, the data were transformed to ln (no. of aggressions
+1) to increase homogeneity of variances as assessed with Levene’s
tests. For significant effects, differences in means were assessed with
Tukey’s HSD post hoc test for unequal sample sizes.
To examine whether individual traits (sex, size, number of
missing limbs, and reproductive state) affected aggressiveness of
lobsters in a given social context, we used a series of general
A
2FM
A
B
C
Lobster
ID (sex
and no.)
M1
M2
F3
F4
M1
M2
F3
F4
M1
M2
F3
F4
M1
M2
M3
M4
M1
M2
M3
M4
M1
M2
M3
M4
F1
F2
F3
F4
F1
F2
F3
F4
F1
F2
F3
F4
F1
M2
M3
F4
F1
M2
M3
F4
F1
M2
M3
F4
CL
(mm)
71.4
69.0
53.3
50.0
61.0
58.0
56.0
54.5
63.2
59.7
43.7
40.8
75.0
61.7
59.4
43.0
75.2
60.7
47.0
40.9
68.5
64.7
59.0
51.4
60.5
52.3
50.0
42.8
56.2
55.0
53.0
49.0
61.2
47.3
45.8
41.9
55.9
50.7
45.0
45.7
60.4
53.4
53.0
48.2
55.0
49.0
44.0
40.3
No. of
missing
limbs
1
0
0
1
0
2
0
1
1
0
1
1
3
1
2
3
5
1
2
0
1
3
2
1
2
1
0
4
3
0
1
2
1
1
2
2
0
0
0
4
1
1
0
1
2
1
1
0
Reproductive
state
Non-ovigerous
Non-ovigerous
Non-ovigerous
Ovigerous
Non-ovigerous
Non-ovigerous
Non-ovigerous
Ovigerous
Ovigerous
Ovigerous
Non-ovigerous
Non-ovigerous
Non-ovigerous
Non-ovigerous
Ovigerous
Ovigerous
Ovigerous
Non-ovigerous
Ovigerous
Ovigerous
Non-ovigerous
Ovigerous
Ovigerous
Non-ovigerous
M, male; F, female; CL, carapace length.
regression analyses. In all models, the dependent variable was the
total number of aggressions performed by each individual throughout the experimental period, but the independent variables differed
with social context. Thus, for the all-male context, we subjected
the data from treatment 4M to a multiple regression analysis with
male size and the number of missing limbs as continuous predictors.
For the all-female context, we subjected the data from treatment 4F
to an ANCOVA with reproductive state as a categorical factor (two
levels: ovigerous and non-ovigerous), and female size and number
of missing limbs as covariates. To examine the effect of sex in a
i158
P. Briones-Fourzán et al.
Table 3. Behaviours performed by lobsters involved in aggressive
encounters [after Atema and Cobb (1980); Cobb (1981);
Segura-Garcı́a et al. (2004)].
Behaviour
Attack or threat
Antenna point
Description
Direction of one or both antennae towards another
lobster.
Approach
Movement towards another lobster (forwards,
backwards, and sideways).
Clasp
To straddle another lobster and hug it using all
pereopods.
Chase
Rapid approach towards a retreating animal.
Fight
Prolonged and intense aggressive contact.
Frontal
Two lobsters face to face in a high posture and with
approach
abdomens tucked.
Grasp
To reach out with pereopods, grab and hold onto
part of another lobster.
Probe
Active probing or poking at another lobster with one
or more pereopods.
Push
Body thrust at another lobster, resulting in its
displacement.
Rush
Rapid approach, often with first two or three pairs
of pereopods raised.
Escape or avoidance
Avoid
Alteration of direction of motion, or reorientation
of body because of an action by another lobster.
Crouch
To lower body so that ventral surface touches or
almost touches the substrate.
Flee
Rapid movement away from another lobster.
Tail flip
Backward withdrawal by the abdomen flex escape
response.
Walk away
Retreat from another lobster (backward, sideward,
or forward).
mixed-sex context, we pooled the data from treatments 2MF and
2FM. Pooling the data from different treatments is not warranted,
but we did this to increase the sample size to do an ANCOVA with
sex as a categorical factor and size and the number of missing limbs
as covariates. Before all analyses, the size data were ln-transformed,
and the dependent variable was transformed to ln (no. of aggressions
+1) to increase homogeneity of variances.
Finally, we used goodness-of-fit x 2 analyses to test the null hypothesis that all four lobsters in each experimental unit of a given treatment
were equally aggressive. If the test was significant, we then proceeded
with subdivided x 2 analyses (Zar, 1999) to determine whether or not
there was a significantly more aggressive individual in that unit and
interpreted the results based on individual traits and the corresponding social context.
Results
Time of day and aggressiveness
The per-capita rate of aggression (mean + SE) varied significantly
with time of day (night: 1.05 + 0.11 aggressions h21; day: 0.53 +
0.06 aggressions h21; F1, 226 ¼ 20.456, p , 0.001) and treatment,
with higher values in treatments 4M and 2FM (1.03 + 0.12 and
0.98 + 0.15 aggressions h21, respectively) than in treatments 4F
and 2MF (0.51 + 0.08 and 0.42 + 0.09 aggressions h21, respectively; F3, 226 ¼ 7.984, p , 0.001). However, the interaction termwas also
significant (F3, 226 ¼ 4.204, p ¼ 0.006), indicating that the effect of
time of day on aggressiveness varied with social context. Indeed,
only in treatments 2MF and 2FM was the per-capita rate of aggression
Figure 1. Night (filled circles) vs. day (open circles) per-capita rate of
aggression (no. of aggressions per individual per hour) in groups of four
P. guttatus held in different social contexts (treatments): 2MF, two
males and two females and the males were larger than the females; 2FM,
two males and two females but the largest individual was a female; 4M,
four males of different sizes; 4F: four females of different sizes. Error bars
denote 95% confidence interval. Different letters indicate statistically
different groups (treatment × day/night interaction).
significantly higher at night than during the day (Figure 1). In treatments 4M and 4F, in contrast, the rate of aggression did not differ
significantly between night and day, but was higher for both periods
in treatment 4M relative to treatment 4F (Figure 1).
Effects of individual traits and social context
on aggressive behaviour
The regression model for the data from lobsters in an all-male context
(treatment 4M) was significant (R2adj = 0.449; F2, 9 ¼ 5.484, p ¼
0.028). Partial correlations revealed that the number of aggressions performed by individual males increased with size (b ¼ 0.626, SE ¼ 0.253,
t ¼ 2.477, p ¼ 0.035; Figure 2a), but decreased with an increasing
number of missing limbs (b ¼2 0.783, SE ¼ 0.253, t ¼ 3.097, p ¼
0.013; Figure 2b). In contrast, the regression model for the data from
lobsters in an all-female context (treatment 4F) was not significant
(R2adj = 0.327; F3, 8 ¼ 2.780, p ¼ 0.110), although partial correlations showed that the number of aggressions performed by individual females also increased with size (b ¼ 0.685, SE ¼ 0.273, t ¼
2.507, p ¼ 0.036; Figure 2c), but had no apparent relationship
with either the number of missing limbs (b ¼ 2 0.091, SE ¼
0.267, t ¼2 0.341, p ¼ 0.741; Figure 2d) or maternal condition
(b ¼ 2 0.042, SE ¼ 0.257, t ¼2 0.162, p ¼ 0.875).
The regression model for the data from the mixed-sex context
was significant (R2adj = 0.290; F3, 20 ¼ 4.134, p ¼ 0.0197). Partial
correlations showed that neither the number of missing limbs
(b ¼ 0.137, SE ¼ 0.180, t ¼ 0.761, p ¼ 0.455) nor size (b ¼ 0.047,
SE ¼ 0.190, t ¼ 0.246, p ¼ 0.808) had a significant effect on the
number of aggressions, which was expected due to the difference in
relative size of males and females between treatments 2MF and 2FM.
However, sex did have a significant effect on the number of aggressions
(b ¼ 0.611, SE ¼ 0.191, t ¼ 3.204, p ¼ 0.004), with males performing, on average, four times as many aggressions (15.92 + 4.79 aggressions lobster21) as females (3.67 + 1.09 aggressions lobster21).
Was there a more aggressive individual in each group?
In all three replicate units (denoted A, B, and C) of all treatments,
the number of aggressions varied significantly among individuals
i159
Behaviour of lobsters in different social contexts
Figure 2. Partial residual plots for (a and c) carapace length, CL (ln-transformed) and (b and d) number of missing limbs vs. total number of
aggressions performed by (a and b) individual males of P. guttatus held in an all-male context (treatment 4M, black dots), and (c and d) individual
females held in an all-female context (treatment 4F, white dots). Dotted lines are 95% confidence interval. Plots for treatment 4M were obtained
from a general regression model (GRM) evaluating the effects of both covariates on the number of aggressions performed by each male. Plots for
treatment 4F were obtained from a GRM evaluating the effects of size, number of missing limbs, and reproductive state on the total number of
aggressions performed by each female. The effects on the number of missing limbs and reproductive state were not significant.
(x 2 tests, all p-values ,0.001), except in unit A of treatment 4F
(x23 = 3.514, p ¼ 0.173). Subdivided x 2analyses revealed that, in
treatments 2MF and 4M, there was a clearly more aggressive individual in each unit, whereas in treatments 4F and 2FM there was wide
variability among units. In all units of treatment 2MF, the most aggressive individual was the largest male (Figure 3a). In treatment
4M, it was the largest male in unit C, but the second largest male
in units A and B (Figure 3b). However, the second largest males in
units A and B were missing one limb each, whereas the largest
males were missing three and five limbs, respectively, and were
not observed performing aggressions (Figure 3b). In treatment 4F,
all four females in unit A exhibited a similar, relatively low level
of aggressiveness, but the most aggressive individual in unit C was
the largest female and in unit B it was the second largest female
(Figure 3c). In treatment 2FM, despite the largest individual in all
units being a female, the more aggressive individual in unit A was
the largest male, whereas the two males in unit B exhibited a
similar level of aggressiveness, higher than that of both females.
In contrast, in unit C, the two males and the largest female performed a similar number of aggressions (Figure 3d).
Discussion
Our results provide further insights into the shelter-related aggressive behaviour of P. guttatus lobsters in different social contexts.
Aggressions were generally of low intensity, with only occasional
bouts of strong intensity. Males exhibited a higher level of aggressiveness than females, but the number of missing limbs significantly
impacted the degree of aggressiveness in males, and size was an
important predictor of aggressiveness among individuals of the
same sex. Because the treatments were not run concurrently, there
is some possibility that the results may be related with the sequence
in which the treatments were run. However, Lipcius and Herrnkind
(1987) found that photoperiod and temperature did not significantly influence levels of aggression in Panulirus argus, a species that
co-occurs with P. guttatus, probably because aggression is expressed
in many activities that tropical spiny lobsters perform year-round
(e.g. territorial disputes and reproductive activities; Berry, 1970;
Lipcius and Herrnkind, 1982).
Consistent with previous studies on P. guttatus (Lozano-Álvarez
and Briones-Fourzán, 2001; Segura-Garcı́a et al., 2004), most
aggressions were related to the occupancy of the shelter space
i160
P. Briones-Fourzán et al.
Figure 3. Percentage of the total number of aggressions performed by each lobster (n ¼ 4) in each replicate unit (a, b and c) per treatment.
Numbers in parentheses on the x-axes are the total number of aggressions recorded in that unit. Numbers above columns denote the number of
limbs that each lobster was missing. In the graph legends, F is female, M is male, and the numbers 1 – 4 indicate the size of lobsters in a decreasing
order (i.e. 1 is largest and 4 is smallest). Asterisks denote individuals that performed significantly more aggressions in a given unit.
(with the vault tending to be the preferred spot), but tended to
subside when individuals were sufficiently spaced out. Spacing is
an important mechanism to avoid future agonistic interactions in
other decapods (Clayton, 1988; Fero and Moore, 2008; Daws
et al., 2011) and possibly also in P. guttatus. Spacing could also
reflect the dependence of P. guttatus lobsters on cryptic behaviour as
a mechanism to avoid detection by predators, as opposed to the communal defensive behaviour exhibited by other, more social spiny lobsters such as P. argus (Briones-Fourzán et al., 2006; Lozano-Álvarez
et al., 2007).
We expected the per-capita rate of aggression to be greater at
night but this only occurred in the mixed-sex groups, although
less markedly in treatment 2MF (wherein the two largest individuals
were males) relative to treatment 2FM (wherein the largest individual was a female). This result likely reflects the occurrence of a clearly
more aggressive individual (the largest male) in each replicate unit
of treatment 2MF. In many decapods, once the dominant individual
in a group asserts itself, aggressive encounters tend to become fewer
and less intense because dominants and subordinates are identified
through status-related cues (Capelli and Hamilton, 1984; Huber
and Kravitz, 1995; Karavanich and Atema, 1998; Edwards and
Herberholz, 2005). In an experiment with P. argus, for example,
once some individuals became dominants and others subordinates,
urine-borne signals released by the dominants induced avoidance
behaviour from subordinates, reducing the general level of aggression (Shabani et al., 2009). In treatment 2FM, in contrast, the
per-capita rate of aggression was higher and there was great variability in the number of aggressions within and among replicate units.
These findings suggest that when the largest individuals in mixedsex groups are males, the largest male will probably assert itself
rather quickly, reducing the general level of aggression in the group,
but when the largest individual is a female, it may take longer for
any one lobster to assert itself.
The per-capita rate of aggression did not vary significantly with
time of day in the same-sex treatments, but was twice as high in treatment 4M as in treatment 4F, reflecting the generally higher degree of
aggressiveness in males. Size was an important predictor of aggressiveness in both these treatments, indicating that, in the absence of
males, large females can also behave aggressively towards smaller
females, as has been observed in other decapods (Lee and Fielder,
1983; Thorpe et al., 1994; Edsman and Jonsson, 1996). In crayfish
Procambarus clarkii and clawed lobsters Homarus americanus, ovigerous females are more aggressive than non-ovigerous females,
presumably to ensure offspring survival (Figler et al., 1995; Mello
et al., 1999), but reproductive state did not affect aggressiveness in
females of P. guttatus. However, as ovigerous and non-ovigerous
females were not equally present in all replicates of treatment 4F,
this result should be considered as preliminary.
Antennal contact is important in the social behaviour of many
decapods (Solon and Cobb, 1980; Mello et al., 1999; Vickery et al.,
2012), but its relative importance in aggressive encounters varies
with taxa. For example, crayfish and clawed lobsters use their
i161
Behaviour of lobsters in different social contexts
antennae to whip and attack conspecifics (Solon and Cobb, 1980;
Edsman and Jonsson, 1996; Smith and Dunham, 1996), whereas
spiny lobsters use their antennae to threaten rather than to attack
conspecifics (Atema and Cobb, 1980; Cobb, 1981; Shabani et al.,
2009). We observed that the rapid pointing of antennae by a large
lobster towards a smaller conspecific occasionally sufficed to elicit
the retreat of the latter.
In decapods, the degree to which individual performance in agonistic encounters is affected by the lack of limbs varies with the type
and number of missing limbs (Neil, 1985; Juanes and Smith, 1995).
In the clawless spiny lobsters, agonistic interactions largely involve
the use of pereopods (Table 3). This, in conjunction with a higher
level of aggressiveness of males of P. guttatus relative to females,
could explain why, in an all-male context, aggressiveness decreased
with an increasing number of missing limbs. In male spiny lobsters,
the first three pairs of pereopods are longer and thicker than in
females and play an important role in agonistic and mating behaviour (McKoy, 1979; Lipcius et al., 1983). Therefore, large intact males
would have a competitive advantage over smaller males or males
lacking limbs. However, individuals that lose limbs may benefit
from becoming subordinate by avoiding costs such as increased
energy expenditure, further injury from a conspecific, or increased
predation risk (Edsman and Jonsson, 1996; Fleming et al., 2007).
For example, Berry (1970) observed that the two largest males in a
group of Panulirus homarus lobsters sharing a tank were missing
several limbs and were subordinate to a smaller but intact male. We
found that, in an all-male context, aggressiveness of males decreased
with an increasing number of missing limbs to the point of neutralizing the advantage conferred by a large size. Thus, the most aggressive
individual in all replicate units of treatment 4M was the male that had
fewer missing limbs in conjunction with a relatively large size.
In females, the apparent lack of relationship between the number
of missing limbs and aggressiveness must be taken with caution
because, unlike in treatment 4M wherein one-third of the test
males were severely handicapped (i.e. missed three limbs or more)
and these males were present in all three replicates, in treatment
4F only 2 of the 12 test females were severely handicapped and no
such females were present in one replicate. In addition, reproductive
condition of females, which varied widely within and between replicates, may have interacted in complex ways with the number of
missing limbs of individual females, obscuring the relative importance of these two factors.
Our results also provide insights into some potential consequences of the loss of habitat complexity and increasing fishing
pressure for these obligate reef-dwelling lobsters. Density of
P. guttatus increases with habitat quality (i.e. coral reef complexity), because complex reefs provide a wide variety of refuges and
foraging spaces (Wynne and Côté, 2007; Robertson and Butler,
2009). However, Caribbean coral reefs are losing structural complexity and becoming increasingly flatter, due in part to the loss
of key reef-building, morphologically complex corals that are
being replaced by faster-growing species that are more resistant
to stress but structurally simpler (Álvarez-Filip et al., 2011). The
tendency of P. guttatus lobsters to space out in separate cavities
within larger crevices presumably reduces agonistic interactions
and detection by predators. Therefore, habitat constraints and
loss of complexity could increase local levels of aggressiveness of
these lobsters, as has been observed for other species (Basquill and
Grant, 1998; Corkum and Cronin, 2004), particularly in periods
of high reproductive activity when their tendency to aggregate
increases (Briones-Fourzán et al., 2013).
In Caribbean locations where P. guttatus is heavily exploited,
fishing pressure can reduce the mean size of lobsters and even alter
sex ratios by disproportionally extracting the largest individuals,
which are mostly males (Evans et al., 1996; Wynne and Côté, 2007).
Because large females of P. guttatus prefer to mate with larger males
(Magallón-Gayón et al., 2011) and our results suggest that, in the
absence of large males, it can take longer for smaller males to assert
themselves, fishing pressure may also indirectly lead to an increased
level of aggressiveness among these spiny lobsters.
Acknowledgements
We greatly appreciate the highly skilled technical support in field
and laboratory activities provided by F. Negrete-Soto and
C. Barradas-Ortiz. Additional help was provided by A. OsorioArciniegas, E. Ramı́rez-Zaldı́var, M. Pérez-Ortiz, and J. ValladárezCob. Three anonymous reviewers greatly improved the manuscript.
This study was funded by Universidad Nacional Autónoma de
México and Consejo Nacional de Ciencia y Tecnologı́a (Mexico)
through grant 40159-Q and a B.Sc. scholarship for RDG. Annual
permits to capture lobsters for experimental purposes were issued
by Comisión Nacional de Acuacultura y Pesca. Experiments complied
with the current laws of Mexico.
References
Álvarez-Filip, L., Dulvy, N. K., Côté, I. M., Watkinson, A. R., and Gill, J.
A. 2011. Coral identity underpins reef complexity on Caribbean
reefs. Ecological Applications, 21: 2223– 2231.
Atema, J., and Cobb, J. S. 1980. Social behavior. In The Biology and
Management of Lobsters, Vol. 1, pp. 409– 450. Ed. by J. S. Cobb,
and B. F. Phillips. Academic Press, New York. 463 pp.
Basquill, S. P., and Grant, J. W. A. 1998. An increase in habitat complexity reduces aggression and monopolization of food by zebra fish
(Danio rerio). Canadian Journal of Zoology, 76: 770– 772.
Beattie, C. L., Pitt, K. A., and Connolly, R. M. 2012. Both size and gender of
mud crabs influence the outcomes of interference interactions.
Journal of Experimental Marine Biology and Ecology, 434–435: 1–6.
Berrill, M. 1976. Aggressive behavior of post-puerulus larvae of the
western rock lobster Panulirus longipes (Milne-Edwards). Australian
Journal of Marine and Freshwater Research, 27: 83–88.
Berry, P. F. 1970. Mating behavior, oviposition and fertilization in the
spiny lobster Panulirus homarus (Linneaus). Oceanographic
Research Institution, (Durban), Investigational Report 24: 1 – 16.
Briones-Fourzán, P., and Lozano-Álvarez, E. 2005. Seasonal variations in
chemical response to conspecific scents in the spotted spiny lobster,
Panulirus guttatus (Latreille). New Zealand Journal of Marine and
Freshwater Research, 39: 383–390.
Briones-Fourzán, P., Magallón-Gayón, E., and Lozano-Álvarez, E. 2013.
Increased reproductive opportunity: a potential benefit of seasonal
aggregation for a little-gregarious and highly sedentary spiny
lobster. Marine Biology Research, 9: 77– 87.
Briones-Fourzán, P., Pérez-Ortiz, M., and Lozano-Álvarez, E. 2006.
Defense mechanisms and anti-predator behavior in two sympatric
species of spiny lobsters, Panulirus argus and P. guttatus. Marine
Biology, 149: 227– 239.
Capelli, G. M., and Hamilton, P. A. 1984. Effects of food and shelter on
aggressive activity in the crayfish Orconectes rusticus (Girard).
Journal of Crustacean Biology, 4: 252– 260.
Childress, M. J. 2007. Comparative sociobiology of spiny lobsters. In
Evolutionary Ecology of Social and Sexual Systems: Crustaceans as
Model Organisms, pp. 271 – 293. Ed. by J. M. Duffy, and M. Thiel.
Oxford University Press, Oxford. 520 pp.
Clayton, D. A. 1988. Hood construction as a spacing mechanism in
Cleistostoma kuwaitense (Crustacea: Ocypodidae). Marine Biology,
99: 57 – 61.
i162
Cobb, J. S. 1981. Behaviour of the Western Australian spiny lobster,
Panulirus cygnus George, in the field and laboratory. Australian
Journal of Marine and Freshwater Research, 32: 399 – 409.
Corkum, L. D., and Cronin, D. J. 2004. Habitat complexity reduces aggression and enhances consumption in crayfish. Journal of Ethology,
22: 23 – 27.
Davis, K. M., and Huber, R. 2007. Activity patterns, behavioural repertoires, and agonistic interactions of crayfish: a non-manipulative
field study. Behaviour, 144: 229– 247.
Daws, A. G., Hock, K., and Huber, R. 2011. Spatial structure of hierarchical groups: testing for processes of aggregation, clustering, and
spatial centrality in crayfish (Orconectes rusticus). Marine and
Freshwater Behaviour and Physiology, 44: 209– 222.
Drews, C. 1993. The concept and definition of dominance in animal behavior. Behaviour, 125: 283– 313.
Edsman, L., and Jonsson, A. 1996. The effect of size, antennal injury,
ownership, and ownership duration on fighting success in male
signal crayfish, Pacifastacus leniusculus (Dana). Nordic Journal of
Freshwater Research, 72: 80 – 87.
Edwards, D. H., and Herberholz, J. 2005. Crustacean models of aggression. In The Biology of Aggression, pp. 38– 61. Ed. by R. J. Nelson.
Oxford University Press, Oxford. 528 pp.
Evans, C. R., Lockwood, A. P. M., and Evans, A. J. 1996. Field studies of
the population dynamics of the spotted spiny lobster Panulirus guttatus (Latreille) at Bermuda. Gulf of Mexico Science, 14: 55– 65.
Fero, K., and Moore, P. A. 2008. Social spacing of crayfish in natural
habitats: what role does dominance play? Behavioral Ecology and
Sociobiology, 62: 1119 –1125.
Fielder, D. R. 1965. A dominance order for shelter in the spiny lobster
Jasus lalandei (H. Milne-Edwards). Behaviour, 24: 236 – 245.
Figler, M. H., Twum, M., Finkelstein, J. E., and Peeke, H. V. S. 1995.
Maternal aggression in red swamp crayfish Procambarus clarkii
Girard: the relation between reproductive status and outcome of aggressive encounters with male and female conspecifics. Behaviour,
132: 107– 125.
Fleming, P. A., Muller, D., and Bateman, P. W. 2007. Leave it all behind: a
taxonomic perspective of autotomy in invertebrates. Biological
Reviews, 82: 481 – 510.
Graham, M. E., and Herberholz, J. 2009. Stability of dominance relationships in crayfish depends on social context. Animal Behaviour,
77: 195 –199.
Hand, J. L. 1986. Resolution of social conflicts: dominance, egalitarianism, spheres of dominance, and game theory. Quarterly Review of
Biology, 61: 201 – 220.
Huber, R., and Kravitz, E. 1995. A quantitative analysis of agonistic behavior in juvenile American lobsters (Homarus americanus). Brain,
Behavior and Evolution, 46: 72 – 83.
Huber, R., and Kravitz, E. 2010. Aggression: towards an integration of
gene, brain and behaviour. In Social Behaviour: Genes, Ecology
and Evolution, pp. 165– 180. Ed. by T. Székely, A. J. Moore, and J.
Komdeur. Cambridge University Press, Cambridge. 576 pp.
Johnson, C. A., Grant, J. W. A., and Giraldeau, L. A. 2004. The effect of
patch size and competitor number on aggression among foraging
house sparrows. Behavioral Ecology, 15: 412 – 418.
Juanes, F., and Smith, L. D. 1995. The ecological consequences of limb
damage and loss in decapod crustaceans: a review and prospectus.
Journal of Experimental Marine Biology and Ecology, 193: 197 – 223.
Karavanich, C., and Atema, J. 1998. Individual recognition and memory
in lobster dominance. Animal Behaviour, 56: 1553– 1560.
Karnofsky, E. B., Atema, J., and Elgin, R. H. 1989. Field observations of
social behavior, shelter use, and foraging in the lobster, Homarus
americanus. Biological Bulletin, 176: 239– 246.
Lee, C. L., and Fielder, D. R. 1983. Agonistic behavior and the development
of dominance hierarchies in the freshwater prawn Macrobrachium
australiense Holthuis 1950 (Crustacea: Palaemonidae). Behaviour,
83: 1–17.
P. Briones-Fourzán et al.
Lipcius, R. N., Edwards, M. L., Herrnkind, W. F., and Waterman, S. A.
1983. In situ mating behavior of the spiny lobster Panulirus argus.
Journal of Crustacean Biology, 3: 217– 222.
Lipcius, R. N., and Herrnkind, W. F. 1982. Molt cycle alterations in behavior, feeding, and diel rhythms of a decapod crustacean, the spiny
lobster Panulirus argus. Marine Biology, 68: 241– 252.
Lipcius, R. N., and Herrnkind, W. F. 1987. Control and coordination of
reproduction and molting in the spiny lobster Panulirus argus.
Marine Biology, 96: 207 – 214.
Lozano-Álvarez, E., and Briones-Fourzán, P. 2001. Den choice and occupation patterns of shelters in two sympatric lobster species,
Panulirus argus and Panulirus guttatus, under experimental conditions. Marine and Freshwater Research, 52: 1145 – 1155.
Lozano-Álvarez, E., Briones-Fourzán, P., Osorio-Arciniegas, A.,
Negrete-Soto, F., and Barradas-Ortiz, C. 2007. Coexistence of congeneric spiny lobsters on coral reefs: differential use of shelter
resources and vulnerability to predators. Coral Reefs, 26: 361– 373.
Lozano-Álvarez, E., Carrasco-Zanini, G., and Briones-Fourzán, P. 2002.
Homing and orientation in the spotted spiny lobster, Panulirus guttatus (Decapoda: Palinuridae), towards a subtidal coral reef habitat.
Crustaceana, 75: 859 – 873.
Lyle, W. G., and MacDonald, C. D. 1983. Molt stage determination in the
Hawaiian spiny lobster Panulirus marginatus. Journal of Crustacean
Biology, 3: 208 – 216.
Magallón-Gayón, E., Briones-Fourzán, P., and Lozano-Álvarez, E. 2011.
Does size always matter? Mate choice and sperm allocation in Panulirus
guttatus, a highly sedentary, habitat-specialist spiny lobster. Behaviour,
148: 1333–1358.
Martin, P., and Bateson, P. 1993. Measuring Behaviour: An Introductory
Guide, 2nd edn. Cambridge University Press, Cambridge. 222 pp.
McKoy, J. L. 1979. Mating behaviour and egg laying in captive rock
lobster, Jasus edwardsii (Crustacea: Decapoda: Palinuridae). New
Zealand Journal of Marine and Freshwater Research, 13: 407– 413.
Mello, J. J., Cromarty, S. I., and Kass-Simon, G. 1999. Increased aggressiveness in gravid American lobsters, Homarus americanus. Aggressive
Behaviour, 25: 451–472.
Negrete-Soto, F., Lozano-Álvarez, E., and Briones-Fourzán, P. 2002.
Population dynamics of the spiny lobster Panulirus guttatus in a
coral reef on the Mexican Caribbean. Journal of Shellfish Research,
21: 279– 288.
Neil, S. J. 1985. Size assessment and cues: studies of hermit crab contests.
Behaviour, 92: 22 – 38.
Peeke, H. V. S., Figler, M. H., and Chang, E. S. 1998. Sex differences and
prior residence effects in shelter competition in juvenile lobsters,
Homarus americanus Milne-Edwards. Journal of Experimental
Marine Biology and Ecology, 229: 149– 156.
Robertson, D. N., and Butler, M. J. 2009. Variable reproductive success
in fragmented populations. Journal of Experimental Marine Biology
and Ecology, 377: 84 – 92.
Segura-Garcı́a, I., Lozano-Álvarez, E., and Briones-Fourzán, P. 2004.
Within-shelter behaviour of the spotted spiny lobster, Panulirus guttatus (Latreille), in simulated communal dens: an exploratory study.
Marine and Freshwater Behaviour and Physiology, 37: 17– 30.
Shabani, S., Kamio, M., and Derby, C. D. 2009. Spiny lobsters use urineborne olfactory signaling and physical aggressive behaviors to influence social status of conspecifics. Journal of Experimental Biology,
212: 2464– 2474.
Sharp, W. C., Hunt, J. H., and Lyons, W. G. 1997. Life history of the
spotted spiny lobster, Panulirus guttatus, an obligate reef-dweller.
Marine and Freshwater Research, 48: 687 – 698.
Smith, M. R., and Dunham, D. W. 1996. Antennae mediate agonistic
physical contact in the crayfish Orconectes rusticus (Girard, 1852)
(Decapoda, Cambaridae). Crustaceana, 69: 668 – 674.
Solon, M. H., and Cobb, J. S. 1980. Antennae whipping behavior in the
American lobster Homarus americanus (Milne-Edwards). Journal of
Experimental Marine Biology and Ecology, 48: 217 – 224.
Behaviour of lobsters in different social contexts
Thomas, C. W., Carter, C. G., and Crear, B. J. 2003. Feed availability and
its relationship to survival, growth, dominance and the agonistic behavior of the southern rock lobster, Jasus edwardsii in captivity.
Aquaculture, 215: 45 –65.
Thorpe, K., Huntingford, F., and Taylor, A. 1994. Relative size and agonistic behavior in the female velvet swimming crab, Necora puber (L.)
(Brachyura, Portunidae). Behavioral Processes, 32: 235– 246.
Vickery, R., Hollowell, K., and Hughes, M. 2012. Why have long antennae? Exploring the function of antennal contact in snapping shrimp.
Marine and Freshwater Behaviour and Physiology, 45: 161– 176.
i163
Weiss, H. M., Lozano-Álvarez, E., Briones-Fourzán, P., and Negrete-Soto,
F. 2006. Using red light with fixed-site video cameras to study the
behavior of spiny lobsters, Panulirus argus, and associated animals
at night and inside their shelters. Marine Technology Society
Journal, 40: 86–95.
Wynne, S. P., and Côté, I. M. 2007. Effects of habitat quality and fishing
on Caribbean spotted spiny lobster populations. Journal of Applied
Ecology, 44: 488 – 494.
Zar, J. H. 1999. Biostatistical Analysis, 4th edn. Prentice-Hall, Upper
Saddle River, NJ. 929 pp.
Handling editor: Howard Browman
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i164 –i169. doi:10.1093/icesjms/fsu249
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Caribbean spiny lobsters equally avoid dead and clinically
PaV1-infected conspecifics
Rebeca I. Candia-Zulbarán1,2*, Patricia Briones-Fourzán 2, Enrique Lozano-Álvarez2,
Cecilia Barradas-Ortiz 2, and Fernando Negrete-Soto 2
1
Universidad Nacional Autónoma de México, Posgrado en Ciencias del Mar y Limnologı́a, Ciudad Universitaria, México DF, Mexico
Universidad Nacional Autónoma de México, Instituto de Ciencias del Mar y Limnologı́a, Unidad Académica de Sistemas Arrecifales, Puerto Morelos,
Quintana Roo, Mexico
2
*Corresponding author: tel: +52 998 8710009; e-mail: [email protected]
Candia-Zulbarán, R. I., Briones-Fourzán, P., Lozano-Álvarez, E., Barradas-Ortiz, C., and Negrete-Soto, F. Caribbean spiny lobsters
equally avoid dead and clinically PaV1-infected conspecifics. – ICES Journal of Marine Science, 72: i164 – i169.
Received 30 August 2014; revised 15 December 2014; accepted 20 December 2014; advance access publication 11 January 2015.
Social behaviour in Caribbean spiny lobsters (Panulirus argus) is mediated by conspecific chemical cues. These lobsters can be attracted to shelters
emanating chemical cues from conspecifics but tend to avoid shelters emanating chemical cues from injured conspecifics, dead conspecifics, and
conspecifics with visible signs of a potentially lethal disease caused by the pathogenic Panulirus argus virus 1 (PaV1). However, previous studies have
not controlled for the presence of PaV1 (i.e. subclinical infection) in grossly “healthy” lobsters, although visible signs of disease do not appear until
several weeks after infection. We conducted a controlled experiment using a set of 2 m-long Y-mazes to examine and contrast the response of
P. argus lobsters to shelters emanating chemical cues from conspecifics in four different conditions: uninfected, subclinically PaV1-infected (i.e.
infected but not diseased), clinically PaV1-infected (i.e. infected and diseased), and dead. Using polymerase chain reaction, we tested for PaV1
in all grossly healthy lobsters and used exclusively uninfected lobsters in intermolt as focal lobsters. Focal lobsters similarly avoided shelters emanating chemical cues from clinically infected (80% avoidance) and from dead conspecifics (85% avoidance), but their response to chemical
cues from uninfected and from subclinically infected conspecifics did not differ significantly from random. These results indicate that PaV1diseased lobsters produce chemical cues that are as repellent to conspecifics as are chemicals emanating from dead conspecifics, and that subclinically
infected lobsters either do not emit the repellent chemicals or they do so at sub-threshold levels. However, the nature of the repellent chemicals and
whether they originate from the pathogen or the host remains to be determined.
Keywords: avoidance behaviour, chemical cues, disease, Panulirus argus, Panulirus argus virus 1.
Introduction
The Caribbean spiny lobster, Panulirus argus (Latreille, 1804) is one
of the most valuable fishing resources in the Western Central
Atlantic and constitutes by itself over 50% of the world catch of
spiny lobsters (Phillips et al., 2013). Panulirus argus has a complex
life cycle and undergoes several habitats shifts during its benthic
life. The post-larvae settle in vegetated habitats where the small juveniles remain for a few months and exhibit asocial behaviour.
However, larger juveniles eventually shift from the vegetation to occupying structured crevice-type shelters, a habitat shift that coincides with a change from asocial to social behaviour (Childress
and Herrnkind, 2001a, b). The most ubiquitous example of sociality
in P. argus and other spiny lobsters is gregarious sheltering, which is
mediated by intraspecific chemical cues released in the urine
(Ratchford and Eggleston, 1998; Horner et al., 2006, 2008). For an
individual seeking shelter, following conspecific scents into a den
both reduces its time of exposure and allows it to assess the
quality of the den (Nevitt et al., 2000; Childress and Herrnkind,
2001a), while congregating in dens can increase per capita survival
through either a “dilution effect” or “group defense behavior”
(Eggleston et al., 1990; Childress and Herrnkind, 2001a, b;
Briones-Fourzán et al., 2007). However, gregariousness is not generated by chemical attraction per se unless the benefits of aggregation
outweigh the costs (Loehle, 1995; Childress, 2007). Benefits of
# International Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
Panulirus argus avoid dead and clinically PaV1-infected conspecifics
i165
aggregation for P. argus would include efficient use of available shelters, decreased predation risk, and increased survival, persistence,
foraging ranges, and reproductive opportunities (Childress and
Herrnkind, 2001a; Dolan and Butler, 2006; Briones-Fourzán et al.,
2007; Childress, 2007). Costs would include an increase in competition and intraspecific aggression as well as in transmission of parasites and pathogenic disease (Loehle, 1995; Lafferty et al., 2004;
Childress, 2007).
Catches of P. argus have been declining steadily since 2000,
mostly due to a combination of overexploitation, changes in environmental and ecological conditions impacting the lobster
habitats (Ehrhardt et al., 2011). However, another potential
cause for this decline is the disease caused by a pathogenic virus,
P. argus virus 1 (PaV1; Shields and Behringer, 2004), which has
become an important source of mortality for the juveniles of
P. argus since its detection in 1999 – 2000 (Moss et al., 2013). In
the laboratory, transmission of PaV1 was found to occur by
contact between healthy and diseased lobsters and through
water, at least over short distances (Butler et al., 2008). However,
experiments conducted by Behringer et al. (2006) showed that
healthy lobsters avoided sharing shelters with visibly diseased conspecifics and that the latter were avoided before becoming infectious, which might help reduce contact transmission rates of PaV1
and keep prevalence levels relatively low. More recently, Anderson
and Behringer (2013) showed that avoidance was chemical in
nature. This would appear akin to avoidance of conspecific alarm
odours (haemolymph-borne chemicals emanating from injured conspecifics), which is strongly exhibited by P. argus (Briones-Fourzán
et al., 2008; Shabani et al., 2008).
Aggregation behaviour is partly communicated by chemical
cues released in the urine of conspecifics, which also communicates social status. Thus, urine-borne signals released by P. argus
lobsters at short distances from conspecifics elicit avoidance
behaviour (similar to haemolymph, Shabani et al., 2008) but
attract conspecifics when released at greater distances (Horner
et al., 2006; Briones-Fourzán et al., 2008); therefore, the choice
of shelter will greatly depend on the context on which chemical
cues are transmitted from conspecifics (Shabani et al., 2009).
Also, the progression of PaV1 infection can affect mobility of lobsters, with heavily infected lobsters becoming lethargic (Shields
and Behringer, 2004; Behringer et al., 2008), but to our knowledge,
previous studies with PaV1 have not tested for the presence of the
virus in presumably healthy individuals. This is important because
visible signs of disease do not appear until many weeks after infection (Behringer et al., 2006). Moreover, subclinically infected
lobsters (i.e. positive for PaV1 by polymerase chain reaction
(PCR) but with no visible signs of disease) can be as abundant in
natural habitats as visibly diseased lobsters (Huchin-Mian et al.,
2013) or even more so (Behringer et al., 2011), and hence uninfected lobsters might not be able to avoid being near potentially infectious conspecifics.
Therefore, the aim of the present study was to contrast the behavioural response of healthy, uninfected individuals of P. argus lobsters
to shelters emanating chemical scents from conspecifics in four
conditions: uninfected with PaV1, subclinically infected with
PaV1, clinically infected with PaV1, and dead (freshly killed and
crushed), controlling for the presence of PaV1 in focal lobsters.
Unlike previous studies wherein experimentally infected lobsters
were used, we used exclusively lobsters that were infected due to
viral exposure under natural conditions. We expected focal lobsters
to exhibit attraction to uninfected lobsters, avoidance of clinically
infected and dead lobsters, and a neutral response to subclinically
infected lobsters, as we have found uninfected lobsters very close
to subclinically infected lobsters in field observations (HuchinMian et al., 2013; RIC-Z, unpublished data).
Material and methods
Lobster collection
The experiments were conducted in the Unidad Académica de
Sistemas Arrecifales, Universidad Nacional Autónoma de México,
at Puerto Morelos, Mexico (20854′ N, 86854′ W). Using scuba
diving, we collected grossly healthy and visibly diseased juvenile lobsters by hand over the Puerto Morelos reef lagoon and transferred
them to the laboratory within 1 h of capture. Visibly diseased lobsters (i.e. clinically infected with PaV1) exhibit a milky-white
haemolymph that is clearly visible through the transparent membrane between the cephalothorax and abdomen. Diseased and
“healthy” lobsters were held in separate tanks provided with multiple hollow concrete blocks for shelter. During the holding period
(≤1 week), lobsters were fed every other day with mussels. The
holding tanks and all experimental units (see below) received seawater from an open flow system and were in the open but under
shade. The seawater was pumped from the Puerto Morelos reef
lagoon and was treated with ozone before passing to an elevated reservoir for distribution to the tanks and also after being used in the
tanks.
Experimental set-up
We used a set of four fibreglass Y-mazes 2 m in length, which has
been estimated as the distance over which spiny lobster behaviours
driven by chemoreception of urine probably function (Horner et al.,
2006, 2008; Shabani et al., 2009). Each Y-maze (Figure 1) had two
independent head tanks (Briones-Fourzán et al., 2008). All tanks
Figure 1. Schematic representation of experimental Y-mazes used to
test for response by uninfected P. argus lobsters to shelters emanating
chemical cues from conspecifics in four different conditions.
i166
were filled to a standpipe height of 0.3 m. Thus, each head tank
(0.6 m long × 0.5 m wide × 0.5 m tall) held 90 l of water,
whereas the Y-maze (2.0 m long × 0.8 m wide × 0.6 m tall) contained 500 l. A panel (1.0 m long × 0.6 m tall) divided half the
length of the Y-maze into two equal arms. Seawater flowed into
the head tanks and then from each head tank to an arm of the
Y-maze at a rate of 2 l min – 1. The water then mixed in an open
area before flowing out through the standpipe located behind the
start area of the Y-maze. Using dye visualizations, we estimated
the downstream velocity as 0.52 + 0.03 cm s – 1 (mean + SD, n ¼ 3).
Two identical shelters were placed one in each arm of the Y-maze
at a distance of 1.8 m from the start area. One of the shelters received
water that had flowed through the head tank containing the stimulus
(see below), whereas the other received plain seawater that had
flowed through the head tank that held no stimulus (control). A
semicircular wire-mesh screen was positioned at the start area
(Figure 1) to prevent lobsters from using the corners in the start
area of the Y-maze as refuge. The head tanks and Y-mazes were completely opaque to preclude visual contact between lobsters and were
mounted on different surfaces to eliminate the transference of vibrations. Also, to prevent transference of acoustic cues potentially produced by the stimuli, the water from the head tanks fell from a height
of 5 cm above the water surface of the Y-maze. Before the experiment, 16 lobsters (50.0 + 15.4 mm CL, mean + SD) were individually tested in the Y-mazes with no chemical stimulus (i.e. the two
shelters receiving plain seawater from the head tanks) to test for potential errors in the Y-maze design or orientation that may bias the
choice trials. Each shelter was selected by 50% of the lobsters, indicating no potentially confounding effects of device orientation or
design.
Testing for DNA of PaV1 by PCR
We tested the haemolymph of all grossly healthy lobsters for the
presence of DNA of PaV1 via PCR assays. After swabbing the exoskeleton with 70% ethanol, 300 ml of haemolymph was withdrawn from the base of one of the fifth pereopods using a sterile
1 ml disposable syringe fitted with a 30 G needle and immediately
fixed in 96% ethanol and stored at –208C. DNA was extracted
from haemolymph samples following salt precipitation protocols
similar as described by Aljanabi and Martı́nez (1997). DNA precipitation was achieved by adding 200 ml sodium acetate 3 M, pH 5.2,
instead of NaCl. DNA integrity was assessed by electrophoresis in
1% agarose gels.
The DNA of PaV1 was amplified by PCR in a 25 ml reaction containing 1 ml extracted DNA, 0.33 mM of each primer 45aF and
543aR (Montgomery-Fullerton et al., 2007), 2.5 mM MgCl2
(Promega), 0.6× reaction buffer (Promega), 0.4 mM dNTP
mixture (Promega), and 0.75 U Taq DNA polymerase (Promega).
The thermal cycling conditions were 1 cycle for 948C for 10 min followed by 30 cycles of 948C for 30 s, 638C for 30 s, 728C for 1 min;
followed by 728C for 10 min. The presence of the expected 499 bp
PaV1 amplicon was determined by resolving 5 ml of the PCR
product and 3 ml of loading buffer in a 2% agarose gel containing
0.1% ethidium bromide and DNA visualization using UV illumination (MiniBis Prow). As negative and positive controls, we used
ultrapure water and haemocyte DNA extracted from lobsters
heavily infected with PaV1, respectively (Huchin-Mian et al., 2013).
Experimental treatments
Based on the PCR results, grossly healthy lobsters were categorized
into uninfected (i.e. negative to PaV1 by PCR) and subclinically
R. I. Candia-Zulbarán et al.
infected lobsters (i.e. positive to PaV1 by PCR but with no visible
signs of disease). The experiment consisted of four treatments that
differed in the condition of the lobsters used as sources of chemical
stimuli. These lobsters were denoted “stimulus”. The stimuli in three
treatments consisted of uninfected lobsters, subclinically infected
lobsters, and clinically infected lobsters, whereas the stimuli in the
fourth treatment consisted of one half of a dead lobster, freshly
killed lobster, and bisected lengthwise. We used only one-half of a
dead lobster per trial to minimize the sacrifice of animals. Only lobsters that tested negative for PaV1 by PCR were used as focal lobsters
and all focal lobsters were used only once.
Trials were conducted overnight. In each trial, the stimulus was
randomly assigned to the left or right head tank and placed into
the head tank 30 min before dark, and the focal lobster was then
placed in the start area of the Y-maze, where it was allowed to acclimatize for 2 h within a mesh cylinder (0.45 m in diameter, 0.40 m in
height). The cylinder was then removed, leaving the focal lobster free
to roam the Y-maze. Between 09:00 and 10:00 h the following
morning, we recorded the position of the focal lobster in the
Y-maze and checked the flow rate into the head tanks. The focal
lobster was then removed, measured with calipers (carapace
length, CL, from between the rostral horns to the posterior
margin of the carapace, +0.1 mm), and moult-staged by microscopic examination of the tip of one pleopod (see Lyle and
MacDonald, 1983). To ensure that no scents remained after each
trial, the head tanks and Y-mazes were drained and thoroughly
brushed, and then water was allowed to flow at a high rate until
the beginning of the next trial.
We ran at least 20 replicate trials in each treatment but discarded
those in which the water flow at the end of the trial differed between
the head tanks by more than +0.50 l min21 or the focal lobster was
not in intermolt (Briones-Fourzán et al., 2008). We also discarded
six trials in which the stimuli (clinically infected lobsters) died
during the night. To avoid potentially confounding effects of
season, we interspersed trials from the four treatments across the
experimental period.
Irrespective of the stimulus, the result of a trial was denoted
“attraction” if the focal lobster chose the shelter subjected to chemical cues from the stimulus or “avoidance” if the focal lobster chose
the shelter receiving plain seawater; thus, results are expressed as
percentages of lobsters exhibiting attraction vs. avoidance. We
used the score method with continuity correction (Newcombe,
1998) to compute 95% confidence intervals for percentages.
Results from each treatment were subjected to a two-tailed binomial
test wherein the probability of choosing either shelter was 50% (p ¼
0.5, a ¼ 0.05), and results from the four treatments were contrasted
in a 2 × 4 contingency table.
Results
The overall mean size (+SD) of focal lobsters from all valid trials
was 48.8 + 13.3 mm CL (n ¼ 77) and did not vary significantly
with treatment (one-way analysis of variance, F ¼ 1.418, d.f. ¼ 3,
74; p ¼ 0.244). The focal, uninfected lobsters were not significantly
attracted to shelters with chemical cues emanating from uninfected
conspecifics (63.2% attraction, n ¼ 19, p ¼ 0.359) and did not significantly avoid shelters with chemical cues emanating from subclinically infected conspecifics (44.7% avoidance, n ¼ 18, p ¼ 0.814;
Figure 2). In contrast, focal lobsters significantly avoided shelters
emanating chemical cues from clinically PaV1-infected conspecifics
(80% avoidance, n ¼ 20, p ¼ 0.012) as well as shelters emanating
Panulirus argus avoid dead and clinically PaV1-infected conspecifics
Figure 2. Results of Y-maze experiment investigating the effects of
chemical cues from P. argus lobsters in four different conditions on
shelter choice by uninfected conspecifics. Horizontal dotted line
denotes random shelter choice (50%). Error bars denote 95%
confidence intervals. p-values were based on two-tailed binomial
tests (a ¼ 0.05).
scents from dead conspecific (85% avoidance, n ¼ 20, p ¼ 0.003;
Figure 2).
The full contingency table analysis contrasting the four treatments yielded a significant result (x2 ¼ 14.748, d.f. ¼ 3, p ¼
0.002). A subdivided x2 analysis (Zar, 1999) showed no significant
difference in the response of focal lobsters to uninfected vs. subclinically infected conspecifics (x2 ¼ 0.222, d.f. ¼ 1, p ¼ 0.638), or
in the response of focal lobsters to clinically infected vs. dead conspecifics (x 2 ¼ 0.173, d.f. ¼ 1, p ¼ 0.667).
Discussion
We contrasted the response of focal lobsters uninfected with PaV1 to
chemical scents emanating from uninfected, subclinically infected
(i.e. infected but not diseased), clinically infected (i.e. infected and
diseased, and infectious), and dead conspecifics in a single experiment. Focal lobsters significantly avoided shelters with chemical
cues from clinically infected conspecifics and from dead conspecifics, consistent with the results of Anderson and Behringer (2013)
and Briones-Fourzán et al. (2008), respectively, but their response
to shelters with scents from subclinically infected lobsters and
from uninfected conspecifics did not differ from random. The
latter result was unexpected, as in previous experiments with
P. argus in Y-mazes, the focal lobsters had been significantly
attracted to shelters emanating chemical cues from healthy conspecifics (Ratchford and Eggleston, 1998; Briones-Fourzán et al., 2008).
However, more recently, Anderson and Behringer (2013) also
obtained a non-significant result testing attraction to chemical
cues from non-diseased lobsters. Moreover, according to
Childress et al. (this issue), conspecific attraction in P. argus
appears declining Caribbean-wise, the causes for which remain to
be determined.
Spiny lobsters have been extensively studied regarding the
mechanisms of aggregation and avoidance, and these studies have
shown that in general, aggregation is mediated by urine-borne chemicals and avoidance is mediated by haemolymph-borne chemicals
(Horner et al., 2006, 2008; Shabani et al., 2008, 2009; Aggio and
Derby, 2011). Yet, results from Anderson and Behringer (2013)
and from our study clearly show that avoidance of diseased lobsters
by healthy conspecifics is driven by chemoreception of chemicals in
i167
the urine. The avoidance of diseased lobsters was similar in magnitude to avoidance of “alarm” chemicals in the haemolymph of
injured conspecifics, which is an effective anti-predator strategy
for gregarious species (Dicke and Grostal, 2001; Briones-Fourzán
et al., 2008). However, haemolymph-borne alarm chemicals elicit
aversion for a limited period after injury because they degrade
within a few hours (Ferner et al., 2005); thus, according to Aggio
and Derby (2011), in overnight experiments using dead specimens
as stimuli, the actual chemical stimulus may be changing over time.
Yao et al. (2009) suggest that some studies attributing avoidance to
alarm odours emanating from dead or injured conspecifics may in
fact involve “necromones”, i.e. biochemicals given off by a decomposing organism and hence reliably associated with death, potential
sources of infections, and perhaps other pathologies. Unlike alarm
odours, which degrade rapidly and are species-specific, necromones
(e.g. certain unsaturated fatty acids such as oleic and linoleic acid)
can persist for days (Diez et al., 2013) and are highly conserved, as
they have been shown to repel animals in many diverse taxa, including crustaceans (Yao et al., 2009). However, given the conserved
nature of necromones, death recognition is generally coupled to
additional cues conveying relatedness (Yao et al., 2009). Therefore,
whether the avoidance of dead conspecifics shown by focal lobsters
in our experiment reflects avoidance to alarm odours or aversion to
necromones remains unclear.
Similarly, it remains to be determined whether the aversive chemicals emanating from PaV1-diseased lobsters originate from the
pathogen or from the host (Aggio and Derby, 2011). Some hosts
can chemically detect certain pathogens, but examples from marine
host-pathogen systems are scant (Toth et al., 2004; Wisenden et al.,
2009). Alternatively, PaV1 may alter the chemical cues emanating
from overtly diseased lobsters by altering their body chemistry in a
specific or in a non-specific manner. In the first case, the chemicals
produced by diseased lobsters would convey a specific signal for the
PaV1 disease. In the latter case, the chemicals would convey a nonspecific signal of “sickness” and the uninfected lobsters may simply
discriminate against “sick” conspecifics, and not against a specific
disease. For example, some fish discriminate against conspecifics
infested with parasitic trematodes, although these parasites cannot
be transmitted from fish to fish (see Krause et al., 1999; Wisenden
et al., 2009). However, when clinical signs of PaV1 appear, the infection has already become systemic, extending to the hepatopancreas,
gill, heart, hindgut, glial cells around the ventral nerves, cuticular
epidermis, and foregut (Li et al., 2008). In heavily infected lobsters,
the hepatopancreas exhibits focal necrosis, ischaemia, and atrophy,
which leads to metabolic wasting (Shields and Behringer, 2004).
Changes in haemolymph constituents in lobsters experimentally
infected (Li et al., 2008) and naturally infected with PaV1
(Pascual-Jiménez et al., 2012) reflect increasing tissue degradation
and catabolism of the hepatopancreas. Therefore, it is also possible
that visibly diseased lobsters exude chemicals associated with decomposition that are repelling to conspecifics.
Behringer et al. (2006) found that experimentally infected lobsters were avoided even before they became infectious, which
would appear at odds with our finding that subclinically infected
lobsters were not significantly avoided. However, although the
PCR assay that we used serves to detect the DNA of PaV1 in the
haemolymph of lobsters, it does not allow determining the viral
load or the progression of the disease in infected lobsters.
Therefore, whether subclinically infected lobsters do not emit repelling chemicals or emit these chemicals at subthreshold levels may
depend on the progression of the infection process.
i168
In summary, healthy lobsters do avoid visibly diseased conspecifics overnight in laboratory conditions and their response is similar
to that exhibited towards dead conspecifics. However, this does not
imply that all healthy lobsters will invariably avoid sharing shelters
with diseased conspecifics in the wild. For example, healthy lobsters
often cohabit with diseased conspecifics in “casitas”, large artificial
shelters deployed over otherwise shelter-poor habitats to increase
biomass of P. argus lobsters in some Caribbean countries
(Lozano-Álvarez et al., 2008; Briones-Fourzán et al., 2012), an occurrence for which Lozano-Álvarez et al. (2008) offered two potential explanations. One explanation is that the large shelter area
provided by casitas (1 –2 m2) may allow cohabitation of healthy
and diseased lobsters without physical contact. However, this explanation is not supported by results of our experiment wherein
16 of 20 focal lobsters avoided chemical cues from diseased lobsters
from a distance of up to 2 m (the length of the Y-mazes). Similarly, in
a field experiment testing distribution of juvenile lobsters in a radial
array of shelters located at 0.5, 1, and 2 m from a centrally located
diseased conspecific, lobsters moved on average 1 m away from
the diseased lobsters over time (Anderson and Behringer, 2013).
The other explanation is that in shelter-poor areas, where the risk
of predation for juvenile lobsters is higher compared with shelterrich areas, healthy lobsters make a trade-off between avoiding
sharing shelters with diseased conspecifics and avoiding predation
risk (Lozano-Álvarez et al., 2008). Indeed, avoidance behaviours
may be costly, both energetically and in the form of trade-offs
with conflicting demands (Loehle, 1995; Wisenden et al., 2009)
and, as noted by Shabani et al. (2009), the choice of shelter is
complex and will depend heavily on the context on which chemical
cues are transmitted from conspecifics. More experiments are
needed to examine the trade-off between disease avoidance and predation avoidance under different ecological contexts.
Acknowledgements
We greatly acknowledge the help provided by J. P. Huchin-Mian,
I. Segura-Garcı́a, and T. Islas-Flores in laboratory activities and
PCR assays. A. Espinosa-Magaña, R. Martı́nez-Calderón,
R. Muñoz de Cote-Hernández, and M. Ortiz-Matamoros helped
to collect and maintain the experimental lobsters and/or with the
experimental trials. M. Villanueva-Méndez allowed us the use of
his laboratory facilities. Comments by three anonymous reviewers
greatly enhanced this manuscript. Funds for this study were provided by Universidad Nacional Autónoma de México and
Consejo Nacional de Ciencia y Tecnologı́a, México (CONACYT,
grants no. 82724 and 101200, and a Doctoral studentship for
RIC-Z). Annual permits to collect lobsters were issued by
Comisión Nacional de Acuacultura y Pesca.
References
Aggio, J., and Derby, C. D. 2011. Chemical communication in lobsters.
In Chemical Communication in Crustaceans, pp. 239 – 256. Ed. by
T. Breithaupt, and M. Thiel. Springer, New York. 565 pp.
Aljanabi, S., and Martı́nez, I. 1997. Universal and rapid salt-extraction of
high quality genomic DNA for PCR-based techniques. Nucleic Acids
Research, 25: 4692– 4693.
Anderson, J. R., and Behringer, D. C. 2013. Spatial dynamics in the social
lobster Panulirus argus in response to diseased conspecifics. Marine
Ecology Progress Series, 474: 191– 200.
Behringer, D. C., Butler, M. J., and Shields, J. D. 2006. Avoidance of
disease by social lobsters. Nature, 441: 421.
Behringer, D. C., Butler, M. J., and Shields, J. D. 2008. Ecological and
physiological effects of PaV1 infection on the Caribbean spiny
R. I. Candia-Zulbarán et al.
lobster (Panulirus argus, Latreille). Journal of Experimental
Marine Biology and Ecology, 359: 26– 33.
Behringer, D. C., Butler, M. J., IV, Shields, J. D., and Moss, J. 2011.
Review of Panulirus argus virus 1—a decade after its discovery.
Diseases of Aquatic Organisms, 94: 153– 160.
Briones-Fourzán, P., Candia-Zulbarán, R. I., Negrete-Soto, F.,
Barradas-Ortiz, C., Huchin-Mian, J. P., and Lozano-Álvarez, E.
2012. Influence of local habitat features on disease avoidance by
Caribbean spiny lobsters in a casita-enhanced bay. Diseases of
Aquatic Organisms, 100: 135 – 148.
Briones-Fourzán, P., Lozano-Álvarez, E., Negrete-Soto, F., and
Barradas-Ortiz, C. 2007. Enhancement of juvenile Caribbean spiny
lobsters: an evaluation of changes in multiple response variables
with the addition of large artificial shelters. Oecologia, 151:
401– 416.
Briones-Fourzán, P., Ramı́rez-Zaldı́var, E., and Lozano-Álvarez, E.
2008. Influence of conspecific and heterospecific aggregation cues
and alarm odors on shelter choice by syntopic spiny lobsters.
Biological Bulletin, 215: 182– 190.
Butler, M. J., Behringer, D. C., and Shields, J. D. 2008. Transmission of
Panulirus argus virus 1 (PaV1) and its effect on the survival of juvenile Caribbean spiny lobster. Diseases of Aquatic Organisms, 79:
173– 182.
Childress, M. J. 2007. Comparative sociobiology of spiny lobsters. In
Evolutionary Ecology of Social and Sexual Systems: Crustaceans as
Model Organisms, pp. 271 – 293. Ed. by J. M. Duffy, and M. Thiel.
Oxford University Press, Oxford. 520 pp.
Childress, M. J., and Herrnkind, W. F. 2001a. The guide effect influence
on the gregariousness of juvenile Caribbean spiny lobsters. Animal
Behavior, 62: 465– 472.
Childress, M. J., and Herrnkind, W. F. 2001b. Influence of conspecifics
on the ontogenetic habitat shift of juvenile Caribbean spiny lobsters.
Marine and Freshwater Research, 52: 1077 –1084.
Dicke, M., and Grostal, P. 2001. Chemical detection of natural enemies
by arthropods: an ecological perspective. Annual Reviews in Ecology
and Systematics, 32: 1 – 23.
Diez, L., Moquet, L., and Detrain, C. 2013. Post-mortem changes in
chemical profile and their influence on corpse removal in ants.
Journal of Chemical Ecology, 39: 1424– 1432.
Dolan, T. W., and Butler, M. J. 2006. The adaptive value of aggregation
among juvenile Caribbean spiny lobster: an evaluation using
individual-base modeling. Journal of Crustacean Biology, 26:
565– 578.
Eggleston, D. B., Lipcius, R. N., Miller, D. L., and Cobá-Cetina, L. 1990.
Shelter scaling regulates survival of juvenile Caribbean spiny lobster
Panulirus argus. Marine Ecology Progress Series, 62: 79 – 88.
Ehrhardt, N. M., Puga, R., and Butler IV, M. J. 2011. Implications of the
ecosystem approach to fisheries management in large ecosystems:
the case of the Caribbean spiny lobster. In Towards Marine
Ecosystem-Based Management in the Wider Caribbean, pp.
157– 175. Ed. by L. Fanning, R. Mahon, and P. McConney.
Amsterdam University Press, Amsterdam. 425 pp.
Ferner, M. C., Smee, D. L., and Chang, Y. P. 2005. Cannibalistic crabs
respond to the scent of injured conspecifics: danger or dinner?
Marine Ecology Progress Series, 300: 193– 200.
Horner, A. J., Nickles, S. P., Weissburg, M. J., and Derby, C. D. 2006.
Source and specificity of chemical cues mediating shelter preference
of Caribbean spiny lobsters (Panulirus argus). Biological Bulletin,
211: 128– 139.
Horner, A. J., Weissburg, M. J., and Derby, C. D. 2008. The olfactory
pathway mediates sheltering behavior of Caribbean spiny lobsters,
Panulirus argus, to conspecific urine signals. Journal of
Comparative Physiology A, 194: 243– 253.
Huchin-Mian, J. P., Rodrı́guez-Canul, R., Briones-Fourzán, P., and
Lozano-Álvarez, E. 2013. Panulirus argus virus 1 (PaV1) infection
prevalence and risk factors in a Mexican lobster fishery employing
casitas. Diseases of Aquatic Organisms, 107: 87 – 97.
Panulirus argus avoid dead and clinically PaV1-infected conspecifics
i169
Krause, J., Ruxton, G. D., and Godin, J-G. J. 1999. Distribution of
Crassiphiala bulboglossa, a parasitic worm, in shoaling fish. Journal
of Animal Ecology, 68: 27 – 33.
Lafferty, K. D., Porter, J. W., and Ford, S. E. 2004. Are diseases increasing
in the ocean? Annual Review of Ecology, Evolution and Systematics,
35: 31 – 54.
Li, C., Shields, J. D., Ratzlaff, R. E., and Butler, M. J. 2008. Pathology and
hematology of the Caribbean spiny lobster experimentally infected
with Panulirus argus Virus 1 (PaV1). Virus Research, 132: 104 – 113.
Loehle, C. 1995. Social barriers to pathogen transmission in wild animal
populations. Ecology, 76: 326– 335.
Lozano-Álvarez, E., Briones-Fourzán, P., Ramı́rez-Estévez, A.,
Placencia-Sánchez, D., Huchin-Mian, J. P., and Rodrı́guez-Canul,
R. 2008. Prevalence of Panulirus argus Virus 1 (PaV1) and habitation
patterns of healthy and diseased Caribbean spiny lobsters in shelterlimited habitats. Diseases of Aquatic Organisms, 80: 95– 104.
Lyle, W. G., and MacDonald, C. D. 1983. Moult stage determination in
the Hawaiian spiny lobster Panulirus marginatus. Journal of
Crustacean Biology, 3: 208 –216.
Montgomery-Fullerton, M. M., Cooper, R. A., Kauffman, K. M.,
Shields, J. D., and Ratzlaff, R. E. 2007. Detection of Panulirus argus
Virus 1 in Caribbean spiny lobsters. Diseases of Aquatic
Organisms, 76: 1 – 6.
Moss, J., Behringer, D., Shields, J. D., Baeza, A., Aguilar-Perera, A., Bush,
P. G., Dromer, C., et al. 2013. Distribution, prevalence, and genetic
analysis of Panulirus argus virus 1 (PaV1) from the Caribbean Sea.
Diseases of Aquatic Organisms, 104: 129– 140.
Nevitt, G., Pentcheff, N. D., Lohmann, K. J., and Zimmer, R. K. 2000.
Den selection by the spiny lobster Panulirus argus: testing attraction
to conspecific odors in the field. Marine Ecology Progress Series, 203:
225– 231.
Newcombe, R. G. 1998. Two-sided confidence intervals for the single
proportion: comparison of seven methods. Statistics in Medicine,
17: 857 –872.
Pascual-Jiménez, C., Huchin-Mian, J. P., Simões, N., Briones-Fourzán,
P., Lozano-Álvarez, E., Sánchez-Arteaga, A., Pérez-Vega, J. A., et al.
2012. Physiological and immunological characterization of
Caribbean spiny lobsters naturally infected with Panulirus argus
Virus 1 (PaV1). Diseases of Aquatic Organisms, 100: 113 – 124.
Phillips, B. F., Melville-Smith, R., Kay, M. C., and Vega-Velázquez, A.
2013. Panulirus species. In Lobsters: Biology, Management,
Aquaculture, and Fisheries, 2nd edn, pp. 289– 325. Ed. by B. F.
Phillips. Wiley-Blackwell, Oxford. 474 pp.
Ratchford, S. G., and Eggleston, D. B. 1998. Size- and scale-dependent
chemical attraction contribute to an ontogenetic shift in sociality.
Animal Behaviour, 56: 1027 – 1034.
Shabani, S., Kamio, M., and Derby, C. D. 2008. Spiny lobsters
detect conspecific blood-borne alarm cues exclusively through
olfactory sensilla. Journal of Experimental Biology, 211:
2600 – 2608.
Shabani, S., Kamio, M., and Derby, C. D. 2009. Spiny lobsters use urineborne olfactory signaling and physical aggressive behaviors to influence social status of conspecifics. Journal of Experimental Biology,
212: 2464– 2474.
Shields, J. D., and Behringer, D. C., Jr. 2004. A new pathogenic virus in
the Caribbean spiny lobster Panulirus argus from the Florida Keys.
Diseases of Aquatic Organisms, 59: 109– 118.
Toth, G. B., Norén, F., Selander, E., and Pavia, H. 2004. Marine dinoflagellates show induced life-history shifts to escape parasite infection
in response to water-borne signals. Proceedings of the Royal
Society of London, Series B, 271: 733– 738.
Wisenden, B. D., Goater, B., and James, C. T. 2009. Behavioral defenses
against parasites and pathogens. In Fish Defenses, Vol. 2: Pathogens,
Parasites and Predators, pp. 151– 168. Ed. by G. Zaccone, C. Perriere,
A. Mathis, and B. G. Kapoor. Science Publishers, Enfield, NH.
411 pp.
Yao, M., Rosenfeld, J., Attridge, S., Sidhu, S., Aksenov, V., and Rollo,
C. D. 2009. The ancient chemistry of avoiding risks of predation
and disease. Evolutionary Biology, 36: 267 – 281.
Zar, J. H. 1999. Biostatistical Analysis, 4th edn. Prentice-Hall, Upper
Saddle River, NJ 929 pp.
Handling editor: Howard Browman
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i170 –i176. doi:10.1093/icesjms/fsv045
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Are juvenile Caribbean spiny lobsters (Panulirus argus)
becoming less social?
Michael J. Childress*, Katherine A. Heldt, and Scott D. Miller
Department of Biological Sciences, Clemson University, Clemson, SC 29634-0314, USA
*Corresponding author: tel: +1 864 985 2384; fax: +1 864 656 0413; e-mail: [email protected]
Childress, M. J., Heldt, K. A., and Miller, S. D. Are juvenile Caribbean spiny lobsters (Panulirus argus) becoming less social? – ICES
Journal of Marine Science, 72: i170 – i176.
Received 7 September 2014; revised 19 January 2015; accepted 24 February 2015; advance access publication 19 March 2015.
Caribbean spiny lobsters are one of the most commercially important fisheries due in large part to their highly gregarious nature that facilitates their
harvest by the use of traps or aggregation devices containing conspecifics. Aggregation in this species has been shown to be due to strong attraction
to conspecific chemical cues that influence movement rates, discovery of crevice shelters, and den sharing behaviours. Although aggregation has
been shown to have many potential benefits (reduction in exposure time and predation risk), it may also have significant costs as well (increase in
predator encounters, disease transmission, and fishing mortality). We compared the results of three published and three unpublished Y-maze
chemical cue choice experiments from 1996 to 2012 to determine if there has been a decrease in conspecific attraction by early benthic juvenile
Caribbean spiny lobsters (15 – 55 mm carapace length, CL). We found that attraction to conspecific chemical cues decreased since 2010 and was
significantly lower in 2012. Lobsters showed individual variation in conspecific attraction but this variation was unrelated to size, sex, or dominance
status. We also found localized regional variation in conspecific attraction with lobsters from high shelter/high disease areas showing significantly
lower conspecific attraction than those from low shelter/low disease areas. Given that conspecific attraction varies among individuals and potentially increases mortality through either natural (increased disease transmission) or fishery-induced (attraction to traps) mechanisms, we should
play close attention to this loss of conspecific attraction in juvenile lobsters. Future studies should investigate both the causation and the ecological
significance of changes in conspecific attraction in regions that vary in intensity of disease (PaV1) and fishing pressure.
Keywords: behaviour, Caribbean spiny lobster, disease, dominance, fishery-induced behavioural change, gregariousness, habitat loss, odour
attraction, Panulirus argus, sociality.
Introduction
Caribbean spiny lobsters, Panulirus argus, are one of the most important commercial crustacean fisheries in the Caribbean with estimated landings of 36 938 metric tons per year (SEDAR, 2005). There
are multiple life history traits that contribute to their success including a rapid growth rate, enormous fecundity, long-lived larval stage
with high dispersal potential, and the ability to use multiple habitat
types throughout their development (Butler et al., 2006; Childress
and Jury, 2006). Settlement of P. argus postlarvae (puerulus stage)
occurs in shallow water seagrass, hardbottom, or mangrove habitats
after a 5- to 7-month larval period (Butler et al., 2006). Newly settled
benthic juveniles are solitary and widely dispersed in seagrass or
macroalgae (Marx and Herrnkind, 1985) where they avoid predators
primarily by crypsis (Butler et al., 1997; Anderson et al., 2013). They
transition from macroalgae to natural crevice shelters at 15 mm
carapace length (CL; Butler and Herrnkind, 1997; Childress and
Herrnkind, 1997) and begin orienting toward conspecific chemical
cues at this size (Childress and Herrnkind, 1996; Ratchford and
Eggleston, 1998).
Another important characteristic that P. argus and most other
commercially important Palinurids share is a gregarious temperament mediated by attraction to chemosensory cues from conspecifics (Childress, 2007). Conspecific attraction plays a key role in
several ontogenetic habitat transitions of P. argus. The attractive
chemical is released in the urine (Horner et al., 2006) and detected
by the antennules (Horner et al., 2008). Chemosensory cues potentially influence the coastal orientation of postlarvae (Goldstein and
Butler, 2009), the choice of settlement habitat by postlarvae
# International Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i171
Caribbean spiny lobsters becoming less social
(Livingston-Zito and Childress, 2009), hasten the transition
between macroalgal, seagrass, and hardbottom habitats (Childress
and Herrnkind, 2001a), facilitates the choice of crevice shelters
(Nevitt et al., 2000), and decreases predation risk by reducing
search time to find shelters (Childress and Herrnkind, 2001b).
Shelter sharing may further decrease predation risk through dilution or group defence of shelters (Mintz et al., 1994; BrionesFourzán and Lozano-Álvarez, 2008) and coordination of mass
migrations (Herrnkind et al., 2001).
Despite the wealth of potential benefits for conspecific attraction,
there are also potential costs to aggregation. Juvenile lobsters may
choose to defend shelters rather than sharing them, showing significant aggression against intruding conspecifics (Berrill, 1975; Heldt
et al., 2015). Lobsters avoid the chemical cues of injured conspecifics
presumably to reduce the risk of predator encounters (Parsons and
Eggleston, 2005; Briones-Fourzán et al., 2008). Conspecific attraction can bring small lobsters to shelters co-occupied by large lobsters
as well as lobster predators such as snappers, groupers, and nurse
sharks. While the large lobsters are not at risk from these predators,
small lobsters attracted to large shelters may be at increased risk of
predation (Schratwieser, 1999). Conspecific attraction may also increase exposure to diseases, such as Panulirus argus virus 1 (PaV1), a
lethal virus transmitted by close contact with infected individuals
(Shields and Behringer, 2004; Butler et al., 2008), although this
may be mediated by the ability to healthy lobsters to detect and
avoid infected individuals (Behringer et al., 2006; Anderson and
Behringer, 2013). Furthermore, fishers exploit conspecific attraction by using sublegal lobsters (called “shorts” by the fishers) as
bait in commercial traps to attract legal-sized adults (Heatwole
et al., 1988), or by the addition of large aggregation devices (i.e.
casitas) that draw sublegal and legal-sized lobsters together for
easier harvest (Eggleston et al., 1990; Briones-Fourzán et al., 2000).
In this paper, we compare three experiments of conspecific
attraction of juvenile Caribbean spiny lobsters from Florida
Bay, FL (USA) to three previous studies on conspecific attraction
in Florida (Anderson and Behringer, 2013), Mexico (BrionesFourzán et al., 2008), and the Bahamas (Ratchford and Eggleston,
1998). We also examine how covariates such as sex, size, dominance
status, and regional quality (low and high disease) influence conspecific attraction. Specially, we hypothesize that individuals differ in
their degree of conspecific attraction and these differences correlate
with differences in size, sex, or dominance status. Furthermore, we
hypothesize that conspecific attraction has decreased, and is lowest
in regions with high disease prevalence.
Methods
Collection and housing conditions
Juvenile (15 –55 mm CL) Caribbean spiny lobsters were collected by
divers using hand nets from Florida Bay, FL (USA). Collections of
juvenile lobsters in 2004, 2010, and 2011 were made on the
bayside of Long Key. Collections in 2012 were made at one region
with low disease (0–1% PaV1 visibly infected individuals) bayside
of Lower Matecumbe Key and one region with high disease
(5– 10% PaV1 visibly infected individuals) bayside of Grassy Key.
We recorded the sex, CL in mm, number of missing limbs, and
the visible PaV1 status (disease present, disease absent, following
the methods of Shields and Behringer, 2004) of every individual.
Any animal showing visible signs of PaV1 infection or having
been in contact with a PaV1 visibly infected individual while in captivity were excluded from further analysis. To accurately track
individuals, we attached a uniquely color-coded antennae band.
Lobsters were shipped to Clemson University in Clemson, SC
(USA) via overnight express service and were assigned to 120 l
aquaria in size-matched pairs (max size difference 2.4 mm CL).
Each housing tank included a single crevice shelter made from
stacked patio pavers, a submersible aquarium filter, and a 10-cm
air stone. Artificial seawater (Instant Ocean#) conditions were
monitored daily and 25% water changes were performed weekly.
Tanks were placed in a research greenhouse under naturally
varying photoperiod from 10L : 14D to 14L : 10D depending on
the season. Temperature varied from 22 to 288C, salinity varied
from 32 to 38 psu, and pH varied from 7.8 to 8.3. Each individual
was fed ad libitum frozen pink shrimp daily and all uneaten food
was removed the following day. Newly molted individuals were
re-measured, re-tagged with an individually color-coded antenna
band, and excluded from behavioural or chemical attraction observations for 72 h. There was no evidence of visible PaV1 infection for
any of the individuals analysed in this study during the entire period
of captivity (2 –6 months).
Dominance behaviour observations
We used direct behavioural observations to determine an individual’s
dominance status to examine if dominance increases or decreases
conspecific attraction. Each pair of size-matched (+2.4 mm CL) lobsters housed together were observed for a minimum of 14 focal animal
observations of 10 min duration, half shortly after sunrise (0700–
1000 h) and half shortly after sunset (1900–2200 h) under low level
(25 W) red light illumination. All acts of aggression were summed
across all observations for each individual, including antenna
whips, antenna pushes, and body pushes. The individual within
each pair having initiated the most aggressive acts was considered
the dominant (D) while its partner was considered the subordinate
(S). Observations conducted in consecutive months found dominance status remained highly stable (,3% reversals) for up to 6
months in captivity (Heldt, 2013).
Conspecific attraction experiments
Conspecific chemical cue choice tests were conducted using a
Y-maze wet table similar to the one depicted in Briones-Fourzan
et al. (2008) (Figure 1) with the following differences. The dimensions of the Y-maze were 38 cm width × 18 cm depth × 126 cm
length with a 100-cm long central divider and a volume of 40 l.
The Y-maze contained an overflow drain pipe at the release end
that emptied into a 600-l reservoir tank. Seawater from this tank
was pumped through a filter and returned to the opposite end of
the Y-maze via a PVC spray bar with valves to control the flow of
water to each arm of the maze. Flow rates were set to 2 l per
minute for each arm of the Y-maze. Each arm of the Y-maze contained a single crevice shelter made up of patio pavers. Behind
each crevice shelter was an upstream compartment separated by a
perforated opaque divider where a single conspecific could be
held while releasing urine-born chemical cues.
Chemical cue trials were initiated at 1000–1400 h and were
run for 24 h. For each trial, a single test subject was chosen and
single source lobster was selected from the same housing tank.
The source lobster was placed in either the right or left compartment
behind the shelters as determined by a flip of a coin, while the other
compartment remained empty. Thus one shelter contained conspecific chemical cues while the other shelter contained a seawater
control. A single test subject was transferred from its housing tank
and placed at the choice end of the Y-maze opposite the shelters.
i172
M. J. Childress et al.
A perforated partition was positioned such that the test subject was
blocked from entering the arms of the Y-maze but could sense chemical cues within the water coming down each arm of the maze. The
test subject was allowed a 10-min acclimation period in this choice
end of the Y-maze before the partition was removed allowing access
to arms of the Y-maze. After 12 h and again after 24 h, the position of
the test subject was recorded as either in the chemical cue shelter, in
the control shelter, or in the no-choice end of the Y-maze. Very few
trials resulted in no choice (,4%) and those individuals were
retested at a later date up to three times until a choice preference
was recorded. Conspecific attraction data were graphed as the per
cent of individuals that chose the chemical cue side of the Y-maze
and the 95% confidence interval was calculated using the modified
Wald method (Agresti and Coull, 1998).
Meta-analysis of annual variation in conspecific attraction
Chemical cue preference test results from three previous studies
(Ratchford and Eggleston, 1998; Briones-Fourzán et al., 2008;
Anderson and Behringer, 2013) were compared with three chemical
cue preference tests reported here (Table 1). Although the methodologies and dimensions of their Y-maze choice tanks differed slightly,
their designs were qualitatively the same (see Supplementary materials for Y-maze dimensions and specific differences in the methods
of these previous studies). In all these studies, juvenile lobsters were
Figure 1. Conspecific attraction (mean + 95% CI) measured as per
cent preference for conspecific odour over control odour in a Y-maze
choice test vs. year the study was conducted. Numbers inside of bars
represents the number of lobsters tested. Each bar represents a different
conspecific attraction study; 1996 is from Ratchford and Eggleston
(1998), 2004 is from this study, 2006 is from Briones-Fourzán et al.
(2008), 2010 is from Anderson and Behringer (2013), 2011 is from this
study, and 2012 is from this study. Asterisks above each bar indicate a
significant preference (p , 0.05) for conspecific chemical cues. Solid
lines above the bars indicate results that are statistically similar using
contingency table post hoc tests. Conspecific attraction significantly
decreased beginning in 2010.
allowed the opportunity to choose among two arms of a Y-maze
with either control water lacking conspecific chemical cues or
water containing conspecific chemical cues. Choice was indicated
by the occupation of a crevice shelter on either the conspecific or
control sides of the Y-maze. To assure independence of observations, only the results of a single choice test for each individual
were analysed.
The raw data from each of these three previous studies (number
choosing control and number choosing conspecific cues) and the
raw data from our trials conducted in 2004, 2011, and 2012 were
analysed individually by year using a log-linear contingency test
against the expected random distribution. Then the ratios of individuals attracted to conspecifics vs. control were compared with each
other by log-linear contingency test. Post hoc comparisons were
made by pairwise contingency tests to specifically identify those
years that statistically differed from each other.
Influence of dominance status on conspecific attraction
This experiment was specifically designed to test whether individuals differ in their attraction to conspecific chemical cues and
whether these differences can be explained by sex, size, or dominance status. Sixteen juvenile lobsters (31.3–54.1 mm CL) from
Long Key, FL were held in captivity for 6 months in winter of 2010
and ten juvenile lobsters (32.8– 38.3 mm CL) from Long Key were
held in captivity for 6 months in autumn 2011 while being tested
for conspecific chemical cue attraction. Testing began immediately
and was completed within the first 3 months of captivity. We found
no change in conspecific attraction with time in captivity over this
testing period (Heldt, 2013).
Conspecific chemical cue choice tests were then conducted using
a Y-maze wet table as described above, but with the following modifications. Every individual was tested a total of six times with source
lobsters from one of three dominance status classifications: familiar,
unfamiliar dominant, and unfamiliar subordinate. Familiar source
lobsters were individuals from the test subjects’ housing tank.
Unfamiliar dominant source lobsters were dominant individuals
from a different housing tank while unfamiliar subordinate source
lobsters were subordinate individuals from a different housing
tank. Both unfamiliar dominant and unfamiliar subordinate
source lobsters were selected to be close in size to the test subject
lobster (+7.8 mm CL). Each test subject was tested twice with
each of these three types of conspecific cues for a total of six
choice tests. Since the dominance status of the source lobster was
not found to influence the choice of an individual, we calculated
the proportion of the six trials where an individual chose the conspecific odour as a measure of an individual’s level of conspecific attraction. Analysis of covariance was used to analyse the % conspecific
attraction after arcsine square-root transformation. Size was the
covariate with sex and dominance status as fixed effects.
Table 1. Comparison of six conspecific attraction experiments.
Authors (year published)
Ratchford and Eggleston (1998)
This study
Briones-Fourzán et al. (2008)
Anderson and Behringer (2013)
This study
This study
Journal
Animal Behaviour
ICES JMS
Biological Bulletin
MEPS
ICES JMS
ICES JMS
Year tested
1996
2004
2006
2010
2010 –11
2012
Collection location
Bahamas
Florida Keys
Mexico
Florida Keys
Florida Keys
Florida Keys
Sample size (n)
29
23
20
30
26
78
Size min
49.2
19.5
38.8
25.0
31.3
15.5
Size max
97.7
42.6
75.9
50.0
54.1
42.4
Size mean
69.5
33.4
34.1
38.4
35.1
i173
Caribbean spiny lobsters becoming less social
Influence of region on conspecific attraction
This experiment was specifically designed to test whether individuals differ in their attraction to conspecific chemical cues and
whether these differences can be explained by the local prevalence
of disease from where they were collected. Seventy-eight juvenile
lobsters (15.5 –42.4 mm CL) from two regions of Florida Bay were
held in captivity for 4 weeks in May and June 2012 while being
tested for conspecific attraction. Thirty-six individuals came from
the region near Lower Matecumbe Key. This region is characterized
by a recent mass mortality of sponges, lower juvenile density
(2–20 lobsters/625 m2), and low levels of PaV1 infection (0–1% of
individuals visibly infected). Forty-two individuals came from the
region near Grassy Key. This region is characterized by an intact
sponge community, higher juvenile density (15–50 lobsters/
625 m2), and high levels of PaV1 infection (5–10% of individuals
visibly infected).
These chemical cue choice tests were conducted at the Keys
Marine Laboratory, Long Key, FL (USA). Individuals were housed
in 50 l tanks with flow-through filtered seawater from Florida Bay
under natural photoperiod (13L : 11D), salinity (36–40 psu), and
temperature (26 –308C). Conspecific attraction choice tests were
conducted using a Y-maze wet table as described above with the following modifications. Trials were only 1 h duration beginning at
1900–2300 hours and ending at 2000–0000 hours. While this was
a considerably shorter duration that our previous studies, it was necessary to complete the number of trials needed within a 1-week
period for each cohort of lobsters. Analysis of video-taped trials in
the laboratory have shown that Y-maze side association changes
very little from the first to the last hour over a 24-h period (see
Supplementary materials).
The number of individuals choosing control and number choosing conspecific cues were analysed by region to determine if conspecific attraction was random using a log-linear contingency test. Then
the ratios of individuals attracted to conspecifics vs. control for the
low and high disease regions were compared by log-linear contingency test.
Anderson and Behringer (2013). However, the combined results
from 1996 to 2006 (58 of 72 choosing conspecific chemical cues)
vs. the combined results from 2010 to 2012 (79 of 134 choosing conspecific chemical cues) were significantly different (n ¼ 206, G ¼
10.33, p ¼ 0.0013).
Influence of dominance status on conspecific attraction
Juvenile lobsters establish dominance ranks with similar size conspecifics with one individual usually showing 10–50% more aggressive
acts than its partner (Heldt et al., 2015). But an individual’s dominance status did not influence its preference for conspecific chemical
cues (d.f. ¼ 1, 22, F ¼ 0.3613, p ¼ 0.5539). Nor did an individual’s
sex (d.f. ¼ 1, 22, F ¼ 0.0340, p ¼ 0.8554) or CL (d.f. ¼ 1, 22, F ¼
0.2801, p ¼ 0.6019) influence its conspecific attraction (Figure 2).
Furthermore, there was no significant effect of the dominance
status of the individual used as the chemical cue source on conspecific
attraction (d.f. ¼ 1, 22, F ¼ 1.160, p ¼ 0.2922). There were, however,
individual differences in conspecific attraction with some juveniles
choosing conspecific chemical cues consistently more than other
individuals.
Influence of region on conspecific attraction
Lobsters collected from the Grassy Key region of Florida Bay
(sponges present, high juvenile lobster densities, high prevalence
of PaV1 visibly infection) had significantly lower conspecific attraction than those from the Lower Matecumbe Key region of Florida
Bay (sponges absent, low juvenile lobster density, and low prevalence of PaV1 visibly infection). This regional effect was statistically
significant (n ¼ 78, d.f. ¼ 1, G ¼ 4.687, p ¼ 0.0304, Figure 3).
However, since these regions differ in multiple factors, it is unclear
if this difference is influenced by the prevalence of the disease, the
availability of sponge crevice shelters, and/or the density of juvenile
lobsters.
Discussion
Meta-analysis of annual variation in conspecific attraction
Juvenile Caribbean spiny lobsters historically have demonstrated a
significant preference for conspecific chemical cues when tested in
Y-maze choice tests (Ratchford and Eggleston, 1998; BrionesFourzán et al., 2008). However, in our conspecific attraction
Juvenile spiny lobster conspecific attraction has changed over the
years (n ¼ 206, d.f. ¼ 5, G ¼ 12.64, p ¼ 0.0270, Figure 1). Juvenile
spiny lobsters showed a significant preference for conspecific chemical cues in 1996 (n ¼ 29, G ¼ 5.665, p ¼ 0.0173, Ratchford and
Eggleston 1998), 2004 (n ¼ 23, G ¼ 5.53, p ¼ 0.0186, this study),
and 2006 (n ¼ 20, G ¼ 4.051, p ¼ 0.0441, Briones-Fourzan et al.,
2008). However, studies conducted since 2010 have not found a significant preference for conspecific chemical cues. Studies in 2010
(n ¼ 30, G ¼ 1.724, p ¼ 0.1892, Anderson and Behringer, 2013),
2011 (n ¼ 26, G ¼ 0.310, p ¼ 0.5778, this study), and 2012 (n ¼
78, G ¼ 0.644, p ¼ 0.4222, this study) suggest a gradual decrease in
the attractiveness of conspecific chemical cues (Figure 1). An analysis
of other possible variables that differed across these experiments including location, maze dimensions, flow rates, housing conditions,
and experimental methods were no better explaining these differences
than the year the study was conducted (see Supplementary Table S1).
Post hoc comparisons (Figure 1) found that our conspecific attraction results from 2004 were not significantly different from
those measured by Ratchford and Eggleston (1998) or BrionesFourzán et al. (2008). Similarly, our conspecific attraction results
from 2011 were not significantly different from those measured by
Figure 2. Individual conspecific attraction measured as per cent
preference for conspecific odour over control odour in six Y-maze
choice tests vs. the size (mm CL) of the individual. Squares are males and
circles are females. Individuals with shaded symbols are dominant and
individuals with open symbols are subordinate. Although conspecific
attraction varied among individuals, this variation was not related to
size, sex, or dominance status.
Results
i174
Figure 3. Conspecific attraction (mean + 95% CI) measured as per
cent preference for conspecific odour over control odour in a Y-maze
choice test vs. Florida Bay region where the lobsters were collected.
Numbers inside of bars represent the number of lobsters tested.
Matecumbe Key had lower lobster density, lower disease prevalence,
and no sponge shelters. Grassy Key had higher lobster density, higher
disease prevalence, and large sponge shelters. Neither of these groups of
lobsters showed a statistically significant preference for conspecific
chemical cues, but they did differ significantly in their ratio of
individuals choosing conspecific vs. control sides of the Y-maze.
experiments, we demonstrate that juvenile Caribbean spiny lobster
attraction to conspecific chemical cues has decreased through time.
Individuals collected in 2004, significantly preferred familiar conspecific chemical cues in 82.6% of trials. However, individuals
collected in 2011 and 2012 showed no significant preference for conspecific chemical cues (57.7 and 66.5% preferences, respectively).
Since these experiments were conducted using the same Y-maze,
and similar laboratory conditions, it cannot be dismissed as differences due to the experimental methodologies (see Supplemental
materials). Furthermore, our observations corroborate the findings
of Anderson and Behringer (2013) that also found no significant conspecific attraction for Florida Bay lobsters collected in 2010.
There are several possible explanations that we investigated to
potentially explain this loss of conspecific attraction since 2010.
First, we found that when individuals were tested multiple times,
there were differences between individuals. These differences were
not due to the individual’s dominance status, sex, or size. Nor
were these differences due to the identity or dominance status of
the individual used as the source. However, individual variation
indicates that a juvenile’s attraction to conspecifics shows a degree
of repeatability. This is what one would expect if the response to conspecific chemical cues has some level of heritability (Dinglemanse
and Dochtermann, 2013). Heritable behavioural traits may
respond to natural or artificial selection leading to changes in the
mean and variance of the trait in future generations (Bell et al.,
2009).
Previous studies have found that larger juvenile P. argus tend to
be more aggressive and dominant over smaller juveniles (Berrill,
1975; Heldt et al., 2015) and that dominance status is also communicated through the release and detection of urine-born cues
(Shabani et al., 2008, 2009). Therefore, it is interesting that
conspecific attraction did not differ based on the familiarity or the
dominance status of the source individual. Perhaps this is because
the source lobster was separated from the choosing lobster in
the Y-maze and thus, was not in direct physical contact. Previous
experiments have not repeatedly measured the response of the
same individual multiple times, so the degree in which individual
M. J. Childress et al.
repeatability in chemical cue preference has changed is unknown
and should be investigated.
Second, variation in the conspecific chemical cue attraction
across years and regions may reflect decreases in crevice shelter
availability or increases in shelter competition. Historically, large
sponges (.20 cm diameter) were the primary natural shelter occupied by juvenile spiny lobsters (Butler and Herrnkind, 1997;
Forcucci et al., 1994; Herrnkind et al., 1997; Childress and
Herrnkind, 1997). Beginning in 1991 and again in 2007, large
regions of Florida Bay experienced wide-spread cyanobacteria
blooms that killed nearly all sponges .20 cm diameter (Butler
et al., 1995). Our 2004 collections of juvenile lobsters from the
bayside hardbottom of Long Key, FL, occurred before the mass
sponge mortality of fall 2007. However, after the mass sponge mortality, juvenile lobsters were largely restricted to areas with natural
solution holes, artificial structures, or boat channels with exposed
rock ledges and rubble piles like our Long Key collecting location.
Therefore, lobsters collected after 2007 may have experienced
fewer natural crevice shelters and elevated shelter competition.
The loss of natural crevice shelters may have increased competition
for those few remaining shelters. Although juvenile spiny lobsters
are known to share shelters, there are only a finite number of lobsters
that can receive protection from any crevice shelter. Decreased
shelter number and increased shelter competition might explain
why lobsters are less attracted to conspecific chemical cues, but
our results did not support this prediction. Lobsters from regions
hard hit by the mass sponge mortality actually showed higher attraction than lobsters from regions with an intact sponge community.
They also show higher den fidelity (Heldt et al., 2015). Thus,
where shelters were most abundant, lobsters were least attracted to
the chemical cues of conspecifics. While habitat loss might have
influenced the decrease in conspecific attraction in Florida Bay juveniles, it should be rigorously examined for other regions of the
Caribbean where conspecific attraction has also decreased.
Third, concomitant with decreasing water quality and declining
shelter availability, juvenile lobsters from Florida Bay are exposed to
PaV1, a lethal viral pathogen that is transmitted by close contact with
infected individuals. Although the origin of PaV1 in the Florida Keys
is unknown (Shields and Behringer, 2004), and its clinical prevalence in the population (5– 9%) has remained relatively stable
since its discovery (Behringer et al., 2011), exposure to the pathogen
may still be having a significant effect on variability between individuals in conspecific attraction. Behringer et al. (2006) have demonstrated that healthy lobsters can detect and avoid the chemical cues
of visibly infected individuals but infected lobsters cannot.
Therefore, one possible explanation for the decline in conspecific attraction may be an increased number of individuals with subclinical
PaV1 infection within the populations of juvenile lobsters we tested.
Moss et al. (2012) found that locations in the Florida Keys vary
widely in the prevalence of individuals that test positive for PaV1
and that individuals may indeed be positive for PaV1 virus via
PCR assay but not clinically infected (Behringer et al., 2012).
However, this explanation is unlikely to explain the variation we
observed in attraction to conspecific chemical cues because first,
visibly infected lobsters were excluded from study; second, lobsters
were held in captivity for 3 –6 months after collection; and third,
variation in attraction was not found to be related to the source
individual.
Fourth, disease may have played a more subtle role in explaining
our results. If individuals have variation in their attraction to conspecifics as demonstrated in dominance status experiment and
i175
Caribbean spiny lobsters becoming less social
PaV1 eliminates those individuals that are most attracted to conspecific chemical cues, then those that survive may be less attracted to
conspecifics. This explanation might be supported by our data
since the prevalence of visibly infected individuals near Grassy
Key (4.64%) was higher than the prevalence of visibly infected individuals near Lower Matecumbe Key (1.64%). Thus, where disease
was relatively high, conspecific attraction was relatively low
(45.2%) whereas where disease was relatively low, conspecific attraction was significantly higher (69.4%). Clearly, more research is
needed on the localized effect of disease prevalence and its effect
on the conspecific attraction of those that survive.
So what can explain a decrease in conspecific attraction?
Certainly, the presence of PaV1 is widespread and could be
causing a wide-spread decline in conspecific attraction (Behringer
et al., 2011). Researchers from Mexico also report that levels of
juvenile conspecific attraction from recent Y-maze trials are
showing similar declines (P. Briones-Fourzán, pers. comm.). This
suggests that this pattern is not unique to juvenile lobsters from
Florida. Alternatively, since spiny lobsters are fished very heavily
throughout the Caribbean using baited traps or artificial shelters
where those individuals most attracted to conspecifics are those
most likely removed by the fishery. Is it possible that the decrease
in conspecific attraction is due to fishery-induced selection on attraction behaviour? Other life history traits have been significantly
impacted by intensive fishing including size at first reproduction,
sperm limitation, and sex ratios (Kuparinen and Merilä, 2007).
Like these life history traits, behaviours are also subject to the intensive selection the fishery exerts on the population (Uusi-Heikkila
et al., 2008). Given the importance of conspecific attraction to
growth, survival, and reproduction in Caribbean spiny lobsters, it
would make sense that fishery-induced selection against gregariousness may be having a significant impact on lobster behaviour. This
hypothesis will be very hard to test given that fishing occurs across
the entire geographic range of this species and the dispersal of the
larvae leads to genetic panmixis. Furthermore, why is the decline
in conspecific attraction showing up now, when fishing practices
have been removing the most gregarious individuals for several
decades? These are the important questions that we should focus
on in the future to better understand the impact of human activities
on this commercially important species.
Supplementary data
Supplementary material is available at the ICESJMS online version
of the manuscript.
Acknowledgements
This study was conducted in the Florida Keys National Marine
Sanctuary under research perm its (FKNMS-2005-038, FKNMS2008-035, and FKNMS-2009-118) and the collections of juvenile lobsters were permitted by the Florida Fish and Wildlife Conservation
Commission (SAL-05-1071-SR, SAL-08-1071-SR, and SAL-111071-SR). Experiment D was conducted at the Keys Marine
Laboratory. We thank Lonny Anderson, Cindy Lewis, Kylie Smith,
Frank Kolencik, Kelsey McClellan, and Ken Sercy for assistance in
the collection of animals and the collection of behavioural data. We
thank J. Antonio Baeza, Donald Behringer, Rodney Bertelsen,
Margaret Ptacek, Kylie Smith, and the Conservation of Marine
Resources creative inquiry team for helpful suggestions on earlier versions of the manuscript. Funding was provided by the Clemson
University SC Life and Creative Inquiry Programs (MJC), Sigma Xi
Grant in Aid of Research, (KAH), Lerner Gray Fund for Marine
Research (KAH), Slocum-Lunz Foundation (KAH), and Clemson
University Calhoun Honors Grant (SDM).
References
Agresti, A., and Coull, B. A. 1998. Approximate is better than exact for
interval estimation of binomial proportions. American Journal of
Statistics, 52: 119– 126.
Anderson, J. R., and Behringer, D. C. 2013. Spatial dynamics in the social
lobster Panulirus argus in response to diseased conspecifics. Marine
Ecology Progress Series, 474: 191– 200.
Anderson, J. R., Spadaro, A. J., Baeza, J. A., and Behringer, D. C. 2013.
Ontogenetic shifts in resource allocation: colour change and allometric growth of defensive and reproductive structures in the
Caribbean spiny lobster Panulirus argus. Biological Journal of the
Linnaean Society, 108: 87– 98.
Behringer, D. C., and Butler, M. J., IV. 2010. Disease avoidance influences shelter use and predation in Caribbean spiny lobster.
Behavioural Ecology Sociobiology, 64: 747– 755.
Behringer, D. C., Butler, M. J., IV, Moss, J., and Shields, J. D. 2012. PaV1
infection in the Florida spiny lobster (Panulirus argus) fishery and its
effects on trap function and disease transmission. Canadian Journal
of Fisheries Aquatic Sciences, 69: 136– 144.
Behringer, D. C., Butler, M. J., IV, and Shields, J. D. 2006. Avoidance of
disease by social lobsters. Nature, 441: 421.
Behringer, D. C., Butler, M. J., IV, Shields, J. D., and Moss, J. 2011.
Review of Panulirus argus virus 1—a decade after its discovery.
Diseases of Aquatic Organisms, 94: 153– 160.
Bell, A. M., Hankison, S. J., and Laskowski, K. L. 2009. The repeatability
of behaviour: a meta-analysis. Animal Behaviour, 77: 771 – 783.
Berrill, M. 1975. Gregarious behaviour of juveniles of the spiny lobster,
Panulirus argus (Crustacea: Decapoda). Bulletin of Marine Science,
25: 515– 522.
Briones-Fourzán, P., and Lozano-Álvarez, E. 2008. Coexistence of congeneric spiny lobsters on coral reefs: differences in conspecific aggregation patterns and their potential antipredator benefits. Coral
Reefs, 27:275 – 287.
Briones-Fourzán, P., Lozano-Álvarez, E., and Eggleston, D. B. 2000. The
use of artificial shelters (casitas) in research and harvesting of
Caribbean spiny lobsters in Mexico. In Spiny Lobsters: Fisheries
and Culture, 2nd edn, pp. 420 –446. Ed. by B. F. Phillips, and J.
Kittaka. Blackwell Science, Abingdon.
Briones-Fourzán, P., Ramı́rez-Zaldı́var, E., and Lozano-Álvarez, E.
2008. Influence of conspecific and heterospecific aggregation cues
and alarm odors on shelter choice by syntopic spiny lobsters.
Biological Bulletin, 215: 182– 190.
Butler, M. J., IV, Behringer, D. C., and Shields, J. D. 2008. Transmission
of Panulirus argus virus 1 (PaV1) and its effect on the survival of juvenile Caribbean spiny lobster. Disease Aquatic Organisms, 79:
173– 182.
Butler, M. J., IV, and Herrnkind, W. F. 1997. A test of recruitment limitation and the potential or artificial enhancement of spiny lobster
populations in Florida. Canadian Journal of Fisheries and Aquatic
Sciences, 54: 452– 463.
Butler, M. J., IV, Herrnkind, W. F., and Hunt, J. H. 1997. Factors affecting the recruitment of juvenile Caribbean spiny lobsters dwelling in
macroalgae. Bulletin of Marine Science, 61: 3 – 19.
Butler, M. J., IV, Hunt, J. H., Herrnkind, W. F., Matthews, T., Childress,
M., Bertelsen, R., Sharp, W., et al. 1995. Cascading disturbances in
Florida Bay, USA: cyanobacteria blooms, sponge mortality, and
implications for juvenile spiny lobster Panulirus argus. Marine
Ecology Progress Series, 129: 119– 125.
Butler, M. J., IV, Steneck, R. S., and Herrnkind, W. F. 2006. Juvenile and
adult ecology. In Lobster Biology, Management, Aquaculture and
Fisheries, pp. 263 – 309. Ed. by B. F. Phillips. Blackwell Publishing,
London, UK.
Childress, M. J. 2007. Comparative sociobiology of spiny lobsters. In
Evolutionary Ecology of Social and Sexual Systems: Crustaceans as
i176
Model Organisms, pp. 271 –293. Ed. by E. Duffy, and M. Thiel.
Oxford University Press, Oxford, UK.
Childress, M. J., and Herrnkind, W. F. 1996. The ontogeny of social behaviour among juvenile Caribbean spiny lobsters. Animal
Behaviour, 51: 675– 687.
Childress, M. J., and Herrnkind, W. F. 1997. Den sharing by juvenile
Caribbean spiny lobsters (Panulirus argus) in nursery habitat: cooperation or coincidence? Marine and Freshwater Research, 48:
751– 758.
Childress, M. J., and Herrnkind, W. F. 2001a. Influence of conspecifics
on the ontogenetic habitat shift of juvenile Caribbean spiny lobsters.
Marine Freshwater Research, 52: 1077 – 1084.
Childress, M. J., and Herrnkind, W. F. 2001b. The guide effect influence
on the gregariousness of juvenile Caribbean spiny lobsters. Animal
Behaviour, 62: 465– 472.
Childress, M. J., and Jury, S. H. 2006. Behaviour. In Lobsters: Biology,
Management, Aquaculture and Fisheries, pp. 78 – 112. Ed. by B. F.
Phillips. Blackwell Publishing, Oxford, UK.
Dinglemanse, N. J., and Dochtermann, N. A. 2013. Quantifying individual variation in behaviour: mixed effect modelling approaches.
Journal of Animal Ecology, 82: 39– 54.
Eggleston, D. B., Lipcius, R. N., Miller, D. L., and Coba-Cetina, L. 1990.
Shelter scaling regulates survival of juvenile Caribbean spiny lobster
Panulirus argus. Marine Ecology Progress Series, 62: 79 – 88.
Forcucci, D., Butler, M. J., IV, and Hunt, J. H. 1994. Growth and population dynamics of juvenile Caribbean spiny lobster, Panulirus argus,
in Florida Bay, FL (USA). Bulletin of Marine Science, 54: 805 – 818.
Goldstein, J. S., and Butler, M. J., IV. 2009. Behavioral enhancement of
onshore transport by postlarval Caribbean spiny lobster (Panulirus
argus). Limnology and Oceanography, 54: 1669 – 1678.
Heatwole, D. W., Hunt, J. H., and Kennedy, F. S., Jr. 1988. Catch efficiencies of live lobster decoys and other attractants in the Florida spiny
lobster fishery. Florida Marine Research Publications, 44: 1– 15.
Heldt, K. A. 2013. Individual behavioural variation of juvenile spiny
lobster (Panulirus argus) denning behaviours and the role it plays
in shelter competition during habitat loss. MS thesis, Clemson
University, Clemson, SC, USA, 126 pp.
Heldt, K. A., Bridges, W. C., Jr, and Childress, M. J. 2015. Behavioral response to habitat loss in juvenile spiny lobsters. Marine Ecology
Progress Series, 521: 117– 128.
Herrnkind, W. F., Butler, M. J., IV, Hunt, J. H., and Childress, M. J. 1997.
Role of physical refugia: implications from a mass sponge die-off in a
lobster nursery in Florida. Marine and Freshwater Research, 48:
759– 769.
Horner, A. J., Nickles, S. P., Weissburg, M. J., and Derby, C. D. 2006.
Source and specificity of chemical cues mediating shelter preference
of Caribbean spiny lobsters (Panulirus argus). Biological Bulletin,
211: 128– 139.
Horner, A. J., Weissburg, M. J., and Derby, C. D. 2008. The olfactory
pathway mediates sheltering behaviour of Caribbean spiny lobsters,
M. J. Childress et al.
Panulirus argus, to conspecific urine signals. Journal of Comparative
Physiology A: Neuroethology Sensory Neural and Behavioural
Physiology, 194: 243– 253.
Kuparinen, A., and Merilä, J. 2009. Determining and managing
fisheries-induced evolution. Trends in Ecology and Evolution, 22:
652– 659.
Livingston-Zito, A. N., and Childress, M. J. 2009. Does conspecific
density influence the settlement of Caribbean spiny lobster
Panulirus argus postlarvae? New Zealand Journal of Marine and
Freshwater Research, 43: 313 – 325.
Marx, J. M., and Herrnkind, W. F. 1985. Macroalgae (Rhodophyta,
Laurencia spp.) as habitat for young juvenile spiny lobsters,
Panulirus argus. Bulletin of Marine Science, 36: 423– 431.
Mintz, J. D., Lipcius, R. N., Eggleston, D. B., and Seebo, M. S. 1994.
Survival of juvenile Caribbean spiny lobster: effects of shelter size,
geographic location and conspecific abundance. Marine Ecology
Progress Series, 112: 255– 266.
Moss, J., Butler, M. J., IV, Behringer, D. C., and Shields, J. D. 2012.
Genetic diversity of the Caribbean spiny lobster virus, Panulirus
argus virus 1, and the discovery of PaV1 in lobster post-larvae.
Aquatic Biology, 14: 223 – 232.
Nevitt, G., Pentcheff, N. D., Lohmann, K. J., and Zimmer, R. K. 2000.
Den selection by the spiny lobster Panulirus argus: testing attraction
to conspecific odors in the field. Marine Ecology Progress Series, 203:
225– 231.
Parsons, M. D., and Eggleston, D. B. 2005. Indirect effects of recreational
fishing on behaviour of the spiny lobster Panulirus argus. Marine
Ecology Progress Series, 303: 235– 244.
Ratchford, S. G., and Eggleston, D. B. 1998. Size- and scale-dependent
chemical attraction contribute to an ontogenetic shift in sociality.
Animal Behaviour, 56: 1027 – 1034.
Schratwieser, J. 1999. The impact of resident and transient predators on
the population dynamics of juvenile Caribbean spiny lobster in
Florida. MS thesis, Old Dominion University, Norfolk, VA, USA.
SEDAR. 2005. SEDAR 08, Stock Assessment Report 3, Southeastern US
Spiny Lobster. NOAA-SEFSC, 319 pp.
Shabani, S., Kamio, M., and Derby, C. D. 2008. Spiny lobsters detect
conspecific blood-borne alarm cues exclusively through olfactory
sensilla. Journal of Experimental Biology, 211: 2600– 2608.
Shabani, S., Kamio, M., and Derby, C. D. 2009. Spiny lobsters use urineborne olfactory signaling and physical aggressive behaviours to influence social status of conspecifics. Journal of Experimental
Biology, 212: 2464– 2474.
Shields, J. D., and Behringer, D. C. 2004. A new pathogenic virus in the
Caribbean spiny lobster Panulirus argus from the Florida Keys.
Diseases of Aquatic Organisms, 59: 109– 118.
Uusi-Heikkila, S., Wolter, C., Klefoth, T., and Arlinghaus, R. 2008. A behavioural perspective on fishing-induced evolution. Trends in
Ecology and Evolution, 23: 419– 421.
Handling editor: Jonathan Grabowski
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i177 –i184. doi:10.1093/icesjms/fsv041
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Casitas: a location-dependent ecological trap for juvenile
Caribbean spiny lobsters, Panulirus argus
Benjamin C. Gutzler 1, Mark J. Butler, IV 1*, and Donald C. Behringer 2,3
1
Department of Biological Sciences, Old Dominion University, Norfolk, VA 23529, USA
Fisheries and Aquatic Sciences Program, University of Florida, Gainesville, FL 32653, USA
3
Emerging Pathogens Institute, University of Florida, Gainesville, FL 32611, USA
2
*Corresponding author: tel: +1 757 683 3609; fax: +1 757 683 5283; e-mail: [email protected]
Gutzler, B. C., Butler, M. J.IV, and Behringer, D. C. Casitas: a location-dependent ecological trap for juvenile Caribbean spiny
lobsters, Panulirus argus. – ICES Journal of Marine Science, 72: i177– i184.
Received 25 August 2014; accepted 23 February 2015; advance access publication 15 March 2015.
Casitas are artificial shelters used by fishers to aggregate Caribbean spiny lobsters (Panulirus argus) for ease of capture. However, casitas may function as an ecological trap for juvenile lobsters if they are attracted to casitas and their growth or mortality is poorer compared with natural shelters.
We hypothesized that juvenile lobsters may be at particular risk if attracted to casitas because they are less able than larger individuals to defend
themselves, and do not forage far from shelter. We compared the nutritional condition, relative mortality, and activity of lobsters of various sizes in
casitas and natural shelters in adult and juvenile lobster-dominated habitats in the Florida Keys (United States). We found that the ecological effects
of casitas are complex and location-dependent. Lobsters collected from casitas and natural shelters did not differ in nutritional condition. However,
juvenile lobsters in casitas experienced higher rates of mortality than did individuals in natural shelters; the mortality of large lobsters did not differ
between casitas and natural shelters. Thus, casitas only function as ecological traps when deployed in nursery habitats where juvenile lobsters are
lured by conspecifics to casitas where their risk of predation is higher. These results highlight the importance of accounting for animal size and
location-dependent effects when considering the consequences of habitat modification for fisheries enhancement.
Keywords: aggregation device, casita, ecological trap, Panulirus argus, spiny lobster.
Introduction
Ecological traps, a subset of the broader concept of an evolutionary
trap (Schlaepfer et al., 2002; Robertson et al., 2013), occur when a
maladaptive and often novel habitat is chosen by an animal over
more typical habitats where their fitness is greater (Robertson and
Hutto, 2006). This is generally a result of anthropogenic change
and occurs when the cues animals use to judge habitat suitability,
are decoupled from habitat quality (Gilroy and Sutherland, 2007).
Often the costs of ecological traps to animal fitness are associated
with recruitment failures or decreased breeding success (Dwernychuk
and Boag, 1972; Kriska et al., 1998; Weldon and Haddad, 2005),
decreased physical condition (Hallier and Gaertner, 2008; Semeniuk
and Rothley, 2008), or increased mortality (Hawlena et al., 2010;
Ekroos et al., 2012). Although the concept of an ecological trap was
first described in the 1970s (Dwernychuk and Boag, 1972; Gates and
# International
Gysel, 1978), few cases have been rigorously tested (Schlaepfer et al.,
2002; Pärt et al., 2007) and most come from terrestrial systems. For
example, passerine birds often prefer habitats with heterogeneous vegetation, and thus preferentially nest along human-created habitat edges
where predators of nestlings are concentrated (Gates and Gysel, 1978;
Weldon and Haddad, 2005). Mayflies lay their eggs on asphalt roads
that reflect polarized light at a greater intensity than do the streams
where eggs are normally laid (Kriska et al., 1998). Only a few ecological
traps are documented in marine environments. For example, tuna
that are attracted to fish aggregation devices are in poorer nutritional
condition than free-schooling tuna, and so are more vulnerable to exploitation (Hallier and Gaertner, 2008). Another potential marine
ecological trap is one posed by casitas—artificial structures placed
on the seabed to aggregate Caribbean spiny lobster (Panulirus
argus) for ease of capture.
Council for the Exploration of the Sea 2015. All rights reserved.
For Permissions, please email: [email protected]
i178
Casitas as possible ecological traps
Caribbean spiny lobsters support the most valuable fishery in the
Caribbean (Chávez, 2009) where fishers use a variety of methods
to capture them, including casitas (also known as pesqueros or
condos) that mimic the natural crevice shelters used by lobsters.
Casitas vary in size and construction, but are usually flat, rectangular
structures 4 m2 in area with at least two open sides ,10 cm in
height. Lobsters aggregate in casitas, increasing their ease of capture
by fishers. However, the potential effect of casitas on lobster population biology and productivity is contentious because of gaps in our
understanding of their effects on lobsters of all sizes (Herrnkind
and Cobb, 2007).
Generally, only a few lobsters share a natural den at a time, especially in nursery habitats (Childress and Herrnkind, 1997).
However, casitas can contain dozens of individuals at concentrations far exceeding those found in most natural shelters (Mintz
et al., 1994). When placed in environments with few natural shelters,
casitas are very effective at aggregating lobsters, which are attracted
to conspecifics via chemosensory cues (Zimmer-Faust et al., 1985;
Ratchford and Eggleston, 1998). In nature, this behaviour aids
them in finding scattered shelter and in group defence against predators (Ratchford and Eggleston, 2000; Childress and Herrnkind,
2001). Unlike larger lobsters, small juvenile lobsters are usually solitary and too small to deter predators, even when in groups, so they
rely instead on camouflage for protection; thus, over-aggregation of
small juvenile lobsters may be maladaptive (Anderson et al., 2013).
Most research on casitas has focused on their effects on large,
fishery-sized lobsters targeted by fishers (de la Torre and Miller,
1987; Sosa-Cordero et al., 1998; Nizinski, 2007). Other studies
have examined the potentially beneficial effects of small shelters
on juvenile lobster populations in nursery habitats where shelters
may be scarce (Eggleston et al., 1992; Arce et al., 1997; Butler and
Herrnkind, 1997; Briones-Fourzán et al., 2007) and post-settlement
predation on juvenile lobsters is a major population bottleneck
(Herrnkind and Butler, 1986; Butler and Herrnkind, 1997).
Depending on the habitat where casitas are placed, they aggregate juvenile, adult, or both lobster stages. It is thus crucial to understand
the effects of casitas on lobsters of all sizes (ages), especially because
casitas are currently deployed throughout the Caribbean and their
use is expanding.
Many predators of juvenile lobsters (e.g. grouper, Epinephelinae;
snapper, Lutjanidae; triggerfish, Balistidae; nurse sharks, Ginglymostoma
cirratum) also are attracted to casitas (Eggleston et al., 1992; Mintz et al.,
1994). These gape-limited piscine predators pose little threat to
legal-sized lobsters, but small juvenile lobsters that occupy casitas
may not be able to avoid or fend off such predators. The concentration of lobsters and other animals attracted to casitas also depletes
food resources in the area immediately surrounding the casita
(Nizinski, 2007). Large lobsters are highly mobile and can forage
beyond the prey-depleted zone near casitas, whereas less mobile
juvenile lobsters are perhaps unable to reach prey-rich areas
outside this zone (Butler and Herrnkind, 2000). Increased predation
pressure near casitas may also indirectly limit the movement and
foraging of small lobsters (Weiss et al., 2008).
Like other artificial structures deployed in the sea to attract fish
and invertebrates for ease of capture (e.g. fish aggregating devices,
artificial “reefs”), it remains unclear whether casitas enhance
lobster populations (the “Production” hypothesis) or merely redistribute lobsters by concentrating them for easier exploitation (the
“Attraction” hypothesis; Pickering and Whitmarsh, 1997; Wilson
B. C. Gutzler et al.
et al., 2001; Granneman and Steele, 2014). Some contend that
adding casitas to habitats where a lack of structure limits lobster
abundance may reduce predator-induced mortality (BrionesFourzán and Lozano-Álvarez, 2001) or permit lobsters to exploit
underutilized areas, thus expanding their range, population size,
and “production” (Sosa-Cordero et al., 1998). But the “population
expansion” hypothesis assumes that P. argus populations are normally limited by density-dependent growth, for which evidence
is currently lacking (Butler and Herrnkind, 1997; Behringer and
Butler, 2006). Alternately, the “Attraction” hypothesis suggests
that casitas simply change the spatial distribution of lobsters by aggregating them (Davis, 1985) without a commensurate increase in
survival, growth, or reproduction. In addition, the unnatural aggregation of lobsters in casitas may subject them to more predators,
pathogens, and perhaps greater intraspecific competition for food.
We hypothesized that casitas act as an ecological trap by decreasing
foraging success and increasing the risk of predation on small juvenile lobsters. Our study tested the potential effect of casitas on the
Caribbean spiny lobster in two distinctly different habitats:
shallow nursery areas and deeper adult lobster habitat.
Methods
Study area
Our study sites were located in the Florida Bay north of the Middle
Florida Keys and in the Gulf of Mexico north of the Lower Florida
Keys (FL, United States; Figure 1). The sites that we studied and
the casitas we used in the Middle Keys were those originally deployed
in 1990 by Mintz et al. (1994) near the Arsnicker Keys and Twin Keys
within Everglades National Park. These sites were 2 –3 m deep
within mixed hardbottom and seagrass habitat typical of lobster
nursery areas in the Florida Keys. Each site contained an array of
16 casitas made of concrete and PVC spaced 30 m apart; most
of them were still functional nearly 25 years after their initial deployment by Mintz et al. (1994). As described by Mintz et al. (1994), two
types of casitas of slightly different dimensions were deployed at the
two sites: eight “large” casitas (177 × 118 × 6 cm) and eight
Figure 1. Study areas in the Florida Keys, FL, United States. (Top inset)
Location of the Florida Keys at the southern terminus of the Florida
peninsula. (Bottom inset) Photo of a casita. (a) Adult habitat study
region in the Lower Florida Keys. (b) Nursery habitat study region in the
Middle Keys.
Location-dependent ecological trap for juvenile Caribbean spiny lobsters
“medium” casitas (157 × 105 × 4 cm). These casitas are roughly
equivalent in size and shape to the casitas used in fisheries throughout the Caribbean, especially in the most important dimension of
opening height, which is often 6–15 cm (Lozano-Álvarez et al.,
2003; Ley-Cooper et al., 2013). However, the casitas used by fishers
and those used in this study are much larger than the small, artificial
structures employed in some experimental studies and referred to by
various names (e.g. “small casitas”, sensu Eggleston et al., 1990; “shelters” sensu Butler and Herrnkind 1997). At the Lower Keys sites, we
used existing casitas deployed by fishers on sand habitat in 10 m
of water several km north of the islands. These casitas varied in size
and construction, but were generally 200 cm long × 150 cm
wide × 6 cm tall with two open sides, and thus similar in size to
those on our Middle Keys sites and those used Caribbean-wide by
fishers.
In the Middle Keys, we compared the lobster populations from
casitas to those from natural shelters (e.g. solution holes, coral
heads, and large sponges) found in nearby hardbottom areas. At
the study site in the Lower Keys, an eroded reef line between the
Content Keys and Mud Keys as well as scattered coral heads are the
dominant natural shelters used by lobsters in the area. Lobsters
obtained from those shelters were therefore used for comparison to
lobsters collected from casitas.
We handled and processed all lobsters humanely and in accord
with applicable animal care regulations; in the United States,
studies involving decapods are not subject to the same animal care
regulations as vertebrates.
Effect of shelter type on nutritional condition
From May to August of 2012 and 2013, we compared the population
structure and nutritional condition of lobsters captured by divers
using hand nets and tail snares from a haphazard subset of casitas
and natural shelters (e.g. coral heads, solution holes, and large
sponges) at the two Middle Keys sites (Arsnicker Keys and Twin
Keys) and in the Lower Keys. We also estimated the occupancy of
casitas by surveying all lobsters present under 16 casitas (n ¼ 7 in
the Middle Keys and n ¼ 9 in the Lower Keys), a subset of those
that we visited during the entire study. After capture, we recorded
the carapace length (CL), sex, injuries, and molt stage (based on
microscopic analysis of pleopods; Lyle and MacDonald 1983) of
each lobster (n ¼ 669). We retained a subset of the lobsters for analyses of nutrition (n ¼ 325), euthanizing them by rapid freezing and
then dissecting out their hepatopancreas, which was then preserved
in 95% ethanol. Later, the dry weight of each hepatopancreas was
determined by drying at 608C for 72 h and computing the dry
weight index of the hepatopancreas (DWI; Bryars and Geddes,
2005). The DWI was superior to four other commonly used nutritional indices that we tested (Gutzler, 2014).
We analysed the data in a one-way ANCOVA with four levels
to test the effects of casitas and natural shelters in the Middle
Keys nursery habitat and Lower Keys adult habitat on DWI,
using lobster CL as a covariate. We followed up by testing for differences between treatment group means using a difference contrast.
We rank-transformed the data to avoid violating parametric
assumptions.
Effect of shelter type on relative predation-induced
mortality
We used tethering to assess differences in the relative mortality of
large and small lobsters in casitas and natural shelters in the
two Florida Keys study regions. Lobsters were tethered by tying
i179
monofilament fishing line around their carapace between the
second and third periopods, and tying on a snap swivel on the
dorsal side of the carapace before sealing the knots with cyanoacrylate superglue. The snap swivel was clipped to a brick with 30 cm of
monofilament for small (CL ≤35 mm) lobsters, or two bricks with
50 cm of monofilament for large (CL ≥60 mm) lobsters. Tethered
lobsters were placed by divers at the entrance of casitas or next to appropriately sized natural shelters such as solution holes, large
sponges, and coral heads. In both cases, we used shelters already occupied by lobsters whenever possible. Each tethered lobster was thus
able to freely move in and out of the shelter, and could retreat well
within it. After 24 h, we returned and assessed mortality. Missing
lobsters were considered killed if the tether was torn or there were
remnants of carapace still attached to it; otherwise, we considered
missing lobsters to have escaped and excluded them from the analyses. Less than 10% of the lobsters escaped their tethers. We tethered
184 lobsters at casitas and 147 lobsters at natural shelters during
summer of 2012 and 2013. In the Middle Keys, where two distinct
sizes of casita were used, we obtained usable data (excluding any potential escapes) from 58 lobsters at 13 different casitas of the “large”
design and 25 lobsters under seven different casitas of the “medium”
design.
Although the survival of tethered lobsters may be correlated with
lobster density within casitas, it explains ,10% of the variance in
survival (Mintz et al., 1994). Therefore, we did not determine
lobster density within each casita before deploying tethered lobsters,
so as to avoid disturbing the occupants. We never tethered lobsters
in casitas that were unoccupied or sparsely occupied. We also did
not quantify predator abundance around casitas for two reasons.
First, measures of daytime predation only cover a fraction of the
24-h period, the very fraction when predation on lobsters is often
low. Diver observations, especially daytime observations, also miss
most of the transient predators that consume lobsters (Smith and
Herrnkind 1992). Second, previous results revealed only a weak relationship between lobster survival and predator abundance near
casitas (Mintz et al. 1994, Eggleston et al., 1997).
We analysed the tethering data using a four-way log-linear contingency analysis to compare the effects of region, shelter type, and
size on lobster survival. We followed up with both visual inspection
of the data and subdivision of the contingency table to determine
which comparisons between the sites, lobster size classes, and
shelter types were responsible for differences between observed
and expected frequencies.
Assessment of relative activity patterns between shelter
types and size classes
We used accelerometry to test for differences in lobster activity
between those dwelling in natural shelters vs. casitas, following protocols described in Gutzler and Butler (2014). We captured lobsters
from casitas (n ¼ 7) and natural shelters (n ¼ 13) in the Middle
Keys and affixed to them an electronic backpack that transmitted
accelerometry data (AT-82 coded transmitter tag with a miniSUR
receiver, Sonotronics, Inc.); lobsters were then immediately released
back into their original shelter. After 24 h, we returned to the site and
used a directional hydrophone (DH-4 Underwater Diver Receiver,
Sonotronics, Inc.) to relocate and recapture the lobster. After recapture, we removed the backpack and released the lobster. The accelerometry data were processed in MATLABw, then analysed using
a split-plot ANOVA, with the fraction of data points spent active
each hour as the factor of interest, shelter type as the whole-plot
factor, hour of day as the subplot factor, and individual lobsters as
i180
B. C. Gutzler et al.
Table 1. Results of a one-way ANCOVA testing the effect of the four region × shelter combinations (Lower Keys natural habitat, Lower Keys
casita, Middle Keys natural habitat, and Middle Keys casita) on the nutritional condition (hepatopancreas DWI) of lobsters, with CL as a
covariate. Group means of DWI are shown below, with homogeneous subsets underlined as determined by a deviation contrast.
Source
Carapace length (covariate)
Shelter type + region
Error
Total
d.f.
1
3
320
325
MS
252525.304
18422.838
6755.488
F
37.381
2.727
p
,0.001
0.044
Group
Mean DWI
Lower Keys casita
1.096
Lower Keys natural
0.953
Middle Keys casita
0.756
Middle Keys natural
0.581
blocks. All data were rank-transformed because they violated parametric assumptions.
Results
Effect of shelter type on nutritional condition
Mean lobster size was 50.8 + 22.2 mm CL (mean + 1 SD) in the
Middle Keys (n ¼ 249) and 84.0 + 11.4 mm CL in the Lower Keys
(n ¼ 76). Sex ratios did not differ between regions or shelter
types, and none of the lobsters collected appeared reproductively
active. The lobsters captured in casitas were generally larger than
those captured in natural shelters (mean size 66.0 + 26.1 mm CL
in casitas and 51.6 + 21.1 mm CL in natural shelters, mean + 1
SD). We found a mean occupancy of 40.6 + 27.1 lobsters
(mean + 1 SD) per casita in the Middle Keys and 22.1 + 14.9
(mean + 1 SD) in the Lower Keys. The fewest lobsters found
under a casita were seven in the Lower Keys, whereas up to 83 lobsters were observed under a single casita in the Middle Keys.
Shelter type had no effect on lobster nutritional condition as
measured by DWI (Table 1 and Figure 2). The covariate of lobster
size (CL) explained the most variance in the data. Although the
ANCOVA revealed significant differences between groups (p ¼ 0.044),
a deviation contrast (Middle Keys natural shelters as reference
category) found no distinct homogeneous subsets. Most of the
differences in the lobster nutritional condition were due to region,
with lobsters in the Lower Keys having a higher DWI than those in
the Middle Keys. The lobsters sampled for DWI were selected to
ensure that the full size range of animals was sampled, and we did not
include pre-molt or post-molt individuals, as these conditions can
skew nutritional condition measurements (Oliver and MacDiarmid
2001, Behringer and Butler 2006).
Effect of shelter type on relative predation-induced
mortality
There were no significant interactions between the effects of region,
shelter type, and size on lobster survival (Chi-squared ¼ 0.179,
d.f. ¼ 1, p ¼ 0.672; Figure 3). Subdivision of the main contingency
table and subsequent inspection of the data revealed that the only
significant differences in survival occurred in small lobsters in the
Middle Keys, where lobsters in casitas were killed at higher rates
than those in natural shelters (Fisher’s exact test; p ¼ 0.023). The
small lobsters in Middle Keys casitas experienced the lowest overall
survival after 24 h (57.9%), whereas large lobsters in casitas in the
Lower Keys had the highest survival (97.4%). Small lobsters were
subject to higher rates of mortality than large lobsters always.
A log-linear contingency table analysis examining the effects of
lobster size and casita size (“large” vs. “medium”; sensu Mintz
et al., 1994) on lobster survival after 24 h revealed no significant
Figure 2. Nutritional condition (hepatopancreas DWI) of lobsters as a
function of CL compared between natural shelters and casitas.
Figure 3. Percent of tethered lobsters of two size classes (small
≤35 mm CL; large ≥60 mm CL) surviving after 24 h in different shelter
types and regions. Significant differences in survival between natural
shelters and casitas are marked by an asterisk. Sample size is shown at
the base of the bars. Error bars ¼ 95% confidence intervals.
association (Chi-squared ¼ 0.59, d.f. ¼ 1, p ¼ 0.443). A study conducted in the same area on the same casitas nearly 25 years before
ours (Mintz et al., 1994) also found no significant difference in
lobster survival between large- and medium-sized casitas, which is
not surprising given that the two types of casitas are only 20% different in area (2.0 vs. 1.6 m2) and the difference in the heights of
their openings (2 cm) is likely too small to substantially change
their value as shelters for lobsters in the broad size range we encountered (25 –75 mm CL).
i181
Location-dependent ecological trap for juvenile Caribbean spiny lobsters
Assessment of relative activity patterns across shelter
types and size classes
Small lobsters showed very little diurnal activity, regardless of shelter
type. However, we found a significant interaction between shelter
type and time of day on lobster activity. Small lobsters sheltering
in casitas were more active at night relative to those in natural
shelters (Table 2 and Figure 4).
Discussion
The effects of casitas on lobster populations are complex and depend
both on lobster size and the habitat within which casitas are
deployed. Lobsters collected from casitas and natural shelters did
not differ in nutritional condition. However, lobsters in the deeper
adult habitats of the Lower Keys were in better nutritional condition
than those in the Middle Keys nursery habitats, regardless of the
shelter type from which they were collected. Small juvenile lobsters
tethered in casitas in nursery habitats experienced significantly
greater predation than those in natural shelters, whereas predation
on large lobsters did not differ between shelter types or regions.
These results indicate that casitas have minimal effects on large lobsters. However, if deployed in nursery habitats where small lobsters
are abundant, casitas act as ecological traps for juvenile lobsters by
aggregating them in inappropriately large structures where they
suffer higher predation.
The casita– habitat interaction
Lobsters of all sizes are commonly found in shallow habitats, many
of which are nursery areas. Casitas have become increasingly
common throughout the Caribbean and are often used in shallow
Table 2. Results of a split-plot ANOVA testing the effect of shelter
type and time of day (h) on the proportion of time small lobsters
(CL ≤45 mm) spent active, as determined by accelerometry.
Source
Shelter type
Individual (shelter type)
Hour
Shelter type × hour
Error
Total
d.f.
1
18
23
23
351
479
MS
40868.32
44340.04
42337.03
19264.62
10408.62
F
0.922
p
4.067
1.851
,0.001
0.011
0.350
nursery habitats where our results indicate that they have the most
potential to cause harm. Fishers usually locate casitas in shallow
(,5 m) areas that are close to shore and where they can harvest lobsters by free diving rather than scuba (de la Torre and Miller, 1987;
Lozano-Álvarez and Briones-Fourzán, 1991; Ley-Cooper et al.,
2013). Yet, when casitas are deployed in nursery habitat, the increased
rate of predation on small juvenile lobsters that are drawn to them by
the odour of other lobsters within them produces an ecological trap
acting on the very life stage considered to be the demographic bottleneck to the population (Butler and Herrnkind, 1997).
Predator concentrations near casitas are often greater than those
near the more widely dispersed natural shelters used by juvenile lobsters (Eggleston and Lipcius, 1992; Mintz et al., 1994). Most of these
piscine predators are gape-limited, and thus more of a threat to
smaller juvenile lobsters. Indeed, predation on tethered small lobsters was high (40–45% killed in 24 h) whether in casitas or in
natural shelters. The scaling of a shelter relative to the size of a
lobster plays an important role in how effective it can be in providing
refuge from predation (Eggleston et al., 1990; Mintz et al., 1994).
Because casitas are sized to accommodate legal-sized lobsters for
capture by fishers, they generally have a much wider opening than
the shelters where small juvenile lobsters are normally found. In
our study, we often observed predators of lobsters such as nurse
sharks, red grouper, and loggerhead turtles (Caretta caretta)
nearby or resting inside casitas, with lobsters either notably absent
or occupying parts of the casita as far from the predator as possible.
Thus, if predators can fit under the shelter, it is clearly not scaled
appropriately for small lobsters and is unlikely to serve as a viable
refuge.
In Florida, an illegal casita fishery is present in deeper water
(10 m). It is therefore one of the few areas in the Caribbean
where it is possible to examine the effects of casitas in deeper habitats
where subadult and adult lobsters are common. Our results suggest
that casitas emplaced there have very different effects than those
located in nursery habitats. Indeed, we found very few lobsters
,50 mm CL at our Lower Keys site and most (76%) were larger
than the minimum legal size limit of 76 mm CL. In this deeper
region, large lobsters experienced no differences in mortality or
nutrition between natural shelters and casitas. Thus, unlike casitas
in nursery habitats where both small and large lobsters are
present, casitas in deeper habitats have neutral effects on the
lobster population.
Casitas and the “attraction vs. production” debate
Figure 4. Activity patterns of small lobsters (CL ≤45 mm) in natural
shelters and casitas as determined by accelerometry. The shaded area
approximates hours of darkness. Error bars are +1 standard error.
Whether the aggregations of animals observed around artificial structures (e.g. artificial reefs, fish aggregation devices, and casitas) are the
result of enhanced production of biomass or merely, the attraction of
individuals from elsewhere has been debated for decades (e.g.
Bohnsack, 1989; Pickering and Whitmarsh, 1997; Brickhill et al.,
2005). The attractive value of casitas for lobsters is undoubted, and
accounts for the adoption of casitas as a fishing gear (Puga et al.,
1996). Yet, whether fishing devices like casitas increase the production
of lobster biomass is uncertain. Lobsters are usually rare in areas such
as seagrass where their prey is abundant, but structures suitable as
shelter are scarce. Advocates of casitas argue that casitas deployed in
these habitats can enhance lobster production by allowing exploitation of food resources that lobsters otherwise rarely access
(Eggleston et al., 1990; Lipcius et al., 1998; Briones-Fourzán and
Lozano-Álvarez, 2001). This ignores the possible ecological ramifications of increasing predatory pressure in seagrass communities.
Moreover, for casitas to enhance lobster production, they must
i182
B. C. Gutzler et al.
increase lobster biomass in the population, whether by increased
growth or decreased mortality (Bohnsack and Sutherland, 1985).
However, other studies indicate that the growth of P. argus in the
wild is not density-dependent and does not appear food limited
(Behringer and Butler 2006, Nizinski 2007). Indeed, we found that
the nutritional condition of both juvenile and adult lobsters were
similar whether residing in natural shelters or casitas.
Accelerometry also gave us insights into the movement of small
and large lobsters residing in casitas and natural shelters. We had
hypothesized that the increased risk of predation for small lobsters
sheltering in casitas would lead to a shift in their activity patterns,
causing them to spend less time foraging away from casitas compared with larger lobsters or those in natural shelters (Weiss et al.,
2008). Our results showed that small lobsters in casitas were more
active overnight than lobsters in natural shelters, but we could not
discern how far they moved. The greater activity of small lobsters
near casitas could result from more intraspecific behavioural interactions near crowded casitas, or perhaps reflects greater foraging activity in the prey-depleted feeding halo that surrounds casitas
(Nizinski, 2007).
misnomer. Small, experimental structures that house small juvenile
lobsters should not be called “casitas” because they do not function
the same as the larger structures used by fishers. Smaller shelters
properly scaled for juvenile lobsters enhance survival (Eggleston
et al., 1990; Behringer and Butler, 2006), whereas larger casitas increase mortality on small juvenile lobsters. We therefore urge
caution and the consideration of local habitat and lobster population features before implementing casita-based fisheries, so as to
reduce the possibility of creating an ecological trap for juveniles.
Globally, marine fisheries are experiencing severe declines and
collapses (Worm et al., 2009) from which the Caribbean spiny
lobster is not exempt (Winterbottom et al., 2012). In the United
States, fishery managers are legally obligated to identify and maintain essential habitats for preservation and conservation of fisheries
resources (Rosenberg et al., 2000). Shallow nursery habitats are essential habitats because they harbour animals at their most vulnerable life stages (Lindeman et al., 2000). Casitas only operate as
ecological traps in areas where small lobsters are present, so management that restricts their use in nursery habitats should be encouraged in the interest of preserving this valuable marine resource.
Casitas as an ecological trap for lobsters
Acknowledgements
Our results indicate that casitas are an ecological trap for juvenile
lobsters when deployed in the shallow nursery habitats like those
in the Florida Keys. The trap is set when small juvenile lobsters
follow their normally beneficial shelter-seeking instincts and are
attracted to a casita by the odour of conspecifics (Nevitt et al.,
2000). Large lobsters—whose size lies beyond the gape limit of most
piscine predators, and who thus suffer no increase in mortality—are
present in casitas in aggregations much denser than found in natural
shelters (Mintz et al., 1994; Briones-Fourzán et al., 2007). Odours emanating from such dense lobster aggregations are very attractive to
small, roaming lobsters (Childress and Herrnkind, 1997; Ratchford
and Eggleston, 1998). However, the casita environment teems with
predators of small lobsters, increasing the rate of mortality on this particularly vulnerable size class. An analogous situation occurs naturally.
Large solution holes contain aggregations of large lobsters that are very
attractive to small lobsters, whose abundance in the surrounding
natural shelters is depleted by the presence of predatory groupers
also living in the solution hole (Schratwieser, 1999).
The results of this research showing increased mortality of small
lobsters in casitas contradict some previous studies, which concluded that casitas increase lobster production by enhancing
survival (Briones-Fourzán and Lozano-Álvarez, 2013). However,
of the studies purporting to demonstrate enhanced production
of lobsters due to casitas, many did not directly test for differences
in predation (Sosa-Cordero et al., 1998; Briones-Fourzán and
Lozano-Álvarez, 2001; Briones-Fourzán et al., 2007). Those
studies also often used smaller “mini-casitas” better scaled to suit juvenile lobsters, rather than the larger casitas used by fishers and
which are at the centre of the attraction – production debate (Arce
et al., 1997; Briones-Fourzán and Lozano-Álvarez, 2001;
Briones-Fourzán et al., 2007; Lozano-Álvarez et al., 2009). Fishers
use casitas with opening heights generally ranging from 6 –8 cm
(Lozano-Álvarez et al., 2003) to 15 cm or more (Ley-Cooper et al.,
2013), although no uniform size or design exists. Some studies
have cautioned against using casitas for fishing in nursery habitats
and suggest only using smaller-scale shelters as enhancement
devices (Sosa-Cordero et al., 1998). However, using the term
“casita” or even “mini-casita” to describe smaller lobster aggregation devices designed for population enhancement is a risky
We thank the many students and interns from Old Dominion
University and the University of Florida who assisted with this
project. The accelerometry component was supported by a Special
Equipment Grant from Sonotronics, Inc. This work was funded
by Florida Sea Grant project R/LR-B-65 (MJB and DCB).
References
Anderson, J. R., Spadaro, A. J., Baeza, J. A., and Behringer, D. C. 2013.
Ontogenetic shifts in resource allocation: colour change and allometric growth of defensive and reproductive structures in the
Caribbean spiny lobster Panulirus argus. Biological Journal of the
Linnean Society, 108: 87 – 98.
Arce, A. M., Aguilar-Dávila, W., Sosa-Cordero, E., and Caddy, J. F. 1997.
Artificial shelters (casitas) as habitats for juvenile spiny lobsters
Panulirus argus in the Mexican Caribbean. Marine Ecology
Progress Series, 158: 217– 224.
Behringer, D. C., and Butler, M. J. 2006. Density-dependent population
dynamics in juvenile Panulirus argus (Latreille): The impact of artificial density enhancement. Journal of Experimental Marine Biology
and Ecology, 334: 84 – 95.
Bohnsack, J. A. 1989. Are high densities of fish at artificial reefs the result
of habitat limitation or behavioral preference? Bulletin of Marine
Science, 44: 631– 645.
Bohnsack, J. A., and Sutherland, D. L. 1985. Artificial reef research: a
review with recommendations for future priorities. Bulletin of
Marine Science, 37: 11 – 39.
Brickhill, M. J., Lee, S. Y., and Connolly, R. M. 2005. Fishes associated
with artificial reefs: attributing changes to attraction or production
using novel approaches. Journal of Fish Biology, 67: 53– 71.
Briones-Fourzán, P., and Lozano-Álvarez, E. 2001. Effects of artificial
shelters (casitas) on the abundance and biomass of juvenile spiny
lobsters Panulirus argus in a habitat-limited tropical reef lagoon.
Marine Ecology Progress Series, 221: 221– 232.
Briones-Fourzán, P., and Lozano-Álvarez, E. 2013. Essential habitats for
Panulirus spiny lobsters. In Lobsters: Biology, Management,
Aquaculture and Fisheries, 2nd edn, pp. 186– 220. Ed. by B. F.
Phillips. Wiley-Blackwell, Chichester. 474 pp.
Briones-Fourzán, P., Lozano-Álvarez, E., Negrete-Soto, F., and
Barradas-Ortiz, C. 2007. Enhancement of juvenile Caribbean spiny
lobsters: an evaluation of changes in multiple response variables
with the addition of large artificial shelters. Oecologia, 151:
401– 416.
Location-dependent ecological trap for juvenile Caribbean spiny lobsters
Bryars, S. R., and Geddes, M. C. 2005. Effects of diet on the growth, survival, and condition of sea-caged adult southern rock lobster, Jasus
edwardsii. New Zealand Journal of Marine and Freshwater
Research, 39: 251– 262.
Butler, M. J., and Herrnkind, W. F. 1997. A test of recruitment limitation
and the potential for artificial enhancement of spiny lobster
(Panulirus argus) populations in Florida. Canadian Journal of
Fisheries and Aquatic Sciences, 54: 452– 463.
Butler, M. J., and Herrnkind, W. F. 2000. Puerulus and juvenile ecology.
In Spiny Lobsters: Fisheries and Culture, 2nd edn, pp. 276 – 301. Ed.
by B. F. Phillips, and J. Kittaka. Fishing News Books, Oxford. 679 pp.
Chávez, E. 2009. Potential production of the Caribbean spiny lobster
(Decapoda, Palinura) fisheries. Crustaceana, 82: 1393– 1412.
Childress, M. J., and Herrnkind, W. F. 1997. Den sharing by juvenile
Caribbean spiny lobsters (Panulirus argus) in nursery habitat: cooperation or coincidence? Marine and Freshwater Research, 48:
751– 758.
Childress, M. J., and Herrnkind, W. F. 2001. The guide effect influence
on the gregariousness of juvenile Caribbean spiny lobsters. Animal
Behaviour, 62: 465– 472.
Davis, G. E. 1985. Artificial structures to mitigate marine construction
impacts on spiny lobster, Panulirus argus. Bulletin of Marine
Science, 37: 151– 156.
de la Torre, R., and Miller, D. L. 1987. Update on the Mexican
Caribbean’s artificial habitat-based spiny lobster (Panulirus argus)
fishery: the evaluation of design, material and placement optimums.
Proceedings of the Gulf and Caribbean Fisheries Institute, 38:
582– 589.
Dwernychuk, L. W., and Boag, D. A. 1972. Ducks nesting in association
with gulls: an ecological trap? Canadian Journal of Zoology, 50:
559– 563.
Eggleston, D. B., and Lipcius, R. N. 1992. Shelter selection by spiny
lobster under variable predation risk, social conditions, and shelter
size. Ecology, 73: 992– 1011.
Eggleston, D. B., Lipcius, R. N., and Grover, J. J. 1997. Predator and
shelter-size effects on coral reef fish and spiny lobster prey. Marine
Ecology Progress Series, 149: 43– 59.
Eggleston, D. B., Lipcius, R. N., and Miller, D. L. 1992. Artificial shelters
and survival of juvenile Caribbean spiny lobster Panulirus argus:
spatial, habitat and lobster size effects. Fishery Bulletin, 90: 691 – 702.
Eggleston, D. B., Lipcius, R. N., Miller, D. L., and Coba-Cetina, L. 1990.
Shelter scaling regulates survival of juvenile Caribbean spiny lobster
Panulirus argus. Marine Ecology Progress Series, 62: 79 – 88.
Ekroos, J., Öst, M., Karell, P., Jaatinen, K., and Kilpi, M. 2012.
Philopatric predisposition to predation-induced ecological traps:
habitat-dependent mortality of breeding eiders. Oecologia, 170:
979– 986.
Gates, J. E., and Gysel, L. W. 1978. Avian nest dispersion and fledging
success in field-forest ecotones. Ecology, 59: 871– 883.
Gilroy, J. J., and Sutherland, W. J. 2007. Beyond ecological traps: perceptual errors and undervalued resources. Trends in Ecology and
Evolution, 22: 351 – 356.
Granneman, J. E., and Steele, M. A. 2014. Fish growth, reproduction,
and tissue production on artificial reefs relative to natural reefs.
ICES Journal of Marine Science, 71: 2494 –2504.
Gutzler, B. 2014. The ecological impact of casitas on Caribbean spiny
lobster (Panulirus argus) populations in the Florida Keys, FL
(USA). MS thesis, Old Dominion University, Norfolk, VA, USA.
Gutzler, B. C., and Butler, M. J. 2014. Accelerometry as a tool for studying lobster behavior: preliminary results from the Florida Keys, FL
(USA). Lobster Newsletter, 27: 8– 9.
Hallier, J.-P., and Gaertner, D. 2008. Drifting fish aggregation devices
could act as an ecological trap for tropical tuna species. Marine
Ecology Progress Series, 353: 255– 264.
Hawlena, D., Saltz, D., Abramsky, Z., and Bouskila, A. 2010. Ecological
trap for desert lizards caused by anthropogenic changes in habitat
i183
structure that favor predator activity. Conservation Biology, 24:
803– 809.
Herrnkind, W. F., and Butler, M. J. 1986. Factors regulating postlarval
settlement and juvenile microhabitat use by spiny lobsters
Panulirus argus. Marine Ecology Progress Series, 34: 23 – 30.
Herrnkind, W. F., and Cobb, J. S. 2007. Artificial shelters for clawed and
spiny lobsters: a critical review of enhancement efforts. American
Fisheries Society Symposium, 49: 925– 932.
Kriska, G., Horváth, G., and Andrikovics, S. 1998. Why do mayflies lay
their eggs en masse on dry asphalt roads? Water-imitating polarized
light reflected from asphalt attracts Ephemeroptera. Journal of
Experimental Biology, 201: 2273– 2286.
Ley-Cooper, K., De Lestang, S., Phillips, B. F., and Lozano-Álvarez, E.
2013. Estimates of exploitation rates of the spiny lobster fishery for
Panulirus argus from tagging within the Bahı́a Espı́ritu Santo “Sian
Ka’an” Biosphere Reserve, Mexican Caribbean. Marine Biology
Research, 9: 88– 96.
Lindeman, K. C., Pugliese, R., Waugh, G. T., and Ault, J. S. 2000.
Developmental patterns within a multispecies reef fishery: management applications for essential fish habitats and protected areas.
Bulletin of Marine Science, 66: 929 – 956.
Lipcius, R. N., Eggleston, D. B., Miller, D. L., and Luhrs, T. C. 1998. The
habitat-survival function for Caribbean spiny lobster: an inverted
size effect and non-linearity in mixed algal and seagrass habitats.
Marine and Freshwater Research, 49: 807 – 816.
Lozano-Álvarez, E., and Briones-Fourzán, P. 1991. Fishery characteristics, growth, and movements of the spiny lobster Panulirus argus in
Bahia de la Ascension, Mexico. Fishery Bulletin, 89: 79– 89.
Lozano-Álvarez, E., Briones-Fourzán, P., and Ramos-Aguilar, M. E.
2003. Distribution, shelter fidelity, and movements of subadult
spiny lobsters (Panulirus argus) in areas with artificial shelters
(casitas). Journal of Shellfish Research, 22: 533 – 540.
Lozano-Álvarez, E., Meiners, C., and Briones-Fourzán, P. 2009.
Ontogenetic habitat shifts affect performance of artificial shelters
for Caribbean spiny lobsters. Marine Ecology Progress Series, 396:
85 – 97.
Lyle, W. G., and MacDonald, C. D. 1983. Molt stage determination in the
Hawaiian spiny lobster Panulirus marginatus. Journal of Crustacean
Biology, 3: 208 – 216.
Mintz, J. D., Lipcius, R. N., Eggleston, D. B., and Seebo, M. S. 1994.
Survival of juvenile Caribbean spiny lobster: effects of shelter size,
geographic location and conspecific abundance. Marine Ecology
Progress Series, 112: 255– 266.
Nevitt, G., Pentcheff, N. D., Lohmann, K. J., and Zimmer, R. K. 2000.
Den selection by the spiny lobster Panulirus argus: testing attraction
to conspecific odors in the field. Marine Ecology Progress Series, 203:
225– 231.
Nizinski, M. S. 2007. Predation in subtropical soft-bottom systems:
spiny lobster and molluscs in Florida Bay. Marine Ecology
Progress Series, 345: 185– 197.
Oliver, M. D., and MacDiarmid, A. B. 2001. Blood refractive index and
ratio of weight to carapace length as indices of nutritional condition
in juvenile rock lobsters (Jasus edwardsii). Marine and Freshwater
Research, 52: 1395– 1400.
Pärt, T., Arlt, D., and Villard, M.-A. 2007. Empirical evidence for ecological traps: a two-step model focusing on individual decisions.
Journal of Ornithology, 148: S327– S332.
Pickering, H., and Whitmarsh, D. 1997. Artificial reefs and fisheries exploitation: a review of the “attraction versus production” debate, the
influence of design and its significance for policy. Fisheries Research,
31: 39 – 59.
Puga, R., Estela de Leon, M., and Cruz, R. 1996. Catchability for the main
fishing methods in the Cuban fishery of the spiny lobster Panulirus
argus (Latreille, 1804), and implications for management
(Decapoda, Palinuridea). Crustaceana, 69: 703 – 718.
i184
Ratchford, S. G., and Eggleston, D. B. 1998. Size- and scale-dependent
chemical attraction contribute to an ontogenetic shift in sociality.
Animal Behaviour, 56: 1027 – 1034.
Ratchford, S. G., and Eggleston, D. B. 2000. Temporal shift in the presence of a chemical cue contributes to a diel shift in sociality. Animal
Behaviour, 59: 793– 799.
Robertson, B. A., and Hutto, R. L. 2006. A framework for understanding
ecological traps and an evaluation of existing evidence. Ecology, 87:
1075– 1085.
Robertson, B. A., Rehage, J. S., and Sih, A. 2013. Ecological novelty and
the emergence of evolutionary traps. Trends in Ecology and
Evolution, 28: 552 – 560.
Rosenberg, A., Bigford, T. E., Leathery, S., Hill, R. L., and Bickers, K.
2000. Ecosystem approaches to fishery management through essential fish habitat. Bulletin of Marine Science, 66: 535– 542.
Schlaepfer, M. A., Runge, M. C., and Sherman, P. W. 2002. Ecological and
evolutionary traps. Trends in Ecology and Evolution, 17: 474–480.
Schratwieser, J. 1999. The impact of resident and transient predators on
the population dynamics of juvenile Caribbean spiny lobster in
Florida. MS thesis, Old Dominion University, Norfolk, VA, USA.
Semeniuk, C. A. D., and Rothley, K. D. 2008. Costs of group-living for a
normally solitary forager: effects of provisioning tourism on southern stingrays Dasyatis americana. Marine Ecology Progress Series,
357: 271– 282.
Smith, K. N., and Herrnkind, W. F. 1992. Predation on early juvenile
spiny lobsters Panulirus argus (Latreille): influence of size and
shelter. Journal of Experimental Marine Biology and Ecology, 157:
3 – 18.
B. C. Gutzler et al.
Sosa-Cordero, E., Arce, A. M., Aguilar-Dávila, W., and
Ramı́rez-González, A. 1998. Artificial shelters for spiny lobster
Panulirus argus (Latreille): an evaluation of occupancy in different
benthic habitats. Journal of Experimental Marine Biology and
Ecology, 229: 1 – 18.
Weiss, H. M., Lozano-Álvarez, E., and Briones-Fourzán, P. 2008.
Circadian shelter occupancy patterns and predator-prey interactions
of juvenile Caribbean spiny lobsters in a reef lagoon. Marine Biology,
153: 953– 963.
Weldon, A. J., and Haddad, N. M. 2005. The effects of patch shape on
indigo buntings: evidence for an ecological trap. Ecology, 86:
1422– 1431.
Wilson, J. A., Osenberg, C. W., St Mary, C. M., Watson, C. A., and
Lindberg, W. J. 2001. Artificial reefs, the attraction-production
issue, and density dependence in marine ornamental fishes.
Aquarium Sciences and Conservation, 3: 95– 105.
Winterbottom, M., Haughton, M., Mutrie, E., and Grieve, K. 2012.
Management of the spiny lobster fishery in CARICOM countries:
status and recommendations for conservation. Proceedings of the
Gulf and Caribbean Fisheries Institute, 64: 456– 462.
Worm, B., Hilborn, R., Baum, J. K., Branch, T. A., Collie, J. S., Costello,
C., Fogarty, M. J., et al. 2009. Rebuilding global fisheries. Science,
325: 578– 585.
Zimmer-Faust, R., Tyre, J. E., and Case, J. F. 1985. Chemical attraction
causing aggregation in the spiny lobster, Panulirus interruptus
(Randall), and its probable ecological significance. Biological
Bulletin, 169: 106– 118.
Handling editor: Rochelle Seitz
ICES Journal of
Marine Science
ICES Journal of Marine Science (2015), 72(Supplement 1), i185 –i198. doi:10.1093/icesjms/fsu238
Contribution to the Supplement: ‘Lobsters in a Changing Climate’
Original Article
Effects of ghost fishing lobster traps in the Florida Keys
Casey B. Butler* and Thomas R. Matthews
Florida Fish and Wildlife Conservation Commission, Fish and Wildlife Research Institute, 2796 Overseas Hwy, #119, Marathon, FL 33050, USA
*Corresponding author: tel: +1 305 289 2330; fax: +1 305 289 2334; e-mail: [email protected]
Butler, C. B., and Matthews, T. R. Effects of ghost fishing lobster traps in the Florida Keys. – ICES Journal of Marine Science, 72:
i185 –i198.
Received 26 August 2014; revised 19 November 2014; accepted 3 December 2014; advance access publication 7 January 2015.
Ghost fishing is the capacity of lost traps to continue to catch and kill animals. In the spiny lobster (Panulirus argus) fishery in Florida, the effects of
ghost fishing are of particular concern, given the estimated 10s of 1000s of traps lost annually. We distributed 40 each of the three types of lobster
traps (wire, wood – wire hybrid, and wood slat) at three locations in the Florida Keys to simulate ghost fishing. Divers monitored these traps biweekly
for 1 year then monthly for two additional years, recording the time ghost traps remained intact and continued to fish, as well as the number of live
and dead lobsters and other animals in each trap. Wood slat and hybrid traps remained intact and fished for 509 + 97 (median + median absolute
deviation) and 480 + 142 d, respectively. Wire traps fished significantly longer (779.5 + 273.5 d, p , 0.001), and several fished until the end of the
experiment (1071 d). Traps in Florida Bay fished longer (711.5 + 51.5 d) than traps inshore (509 + 94.5 d) and offshore (381 + 171 days;
p , 0.001) in the Atlantic Ocean. More lobsters were observed in hybrid traps (mean ¼ 4.81 + 0.03 s.e.) than in wood slat (3.85 + 0.16) or
wire traps (3.17 + 0.03; F ¼ 40.15, d.f. ¼ 2, p , 0.001). Wire traps accounted for 83% of fish confined overall and 74% of the dead fish observed
in traps. Ghost traps in Florida Bay and Atlantic inshore killed 6.8 + 1.0 and 6.3 + 0.88 lobsters per trap annually, while Atlantic offshore traps killed
fewer (3.0 + 0.69) lobsters, likely as a result of lower lobster abundance in traps. The combined effects of greater lobster mortality and greater
abundance of lost traps in inshore areas account for the majority of the estimated 637 622 + 74 367 (mean + s.d.) lobsters that die in ghost
traps annually.
Keywords: bycatch, Caribbean spiny lobster, ghost fishing, lobster traps, marine debris, Panulirus argus.
Introduction
The Caribbean spiny lobster (Panulirus argus) supports one of
Florida’s most important commercial and recreational fisheries,
with an annual ex-vessel value of between $18 and 35 million
USD (Florida Fish and Wildlife Conservation Commission, unpublished data; Kildow, 2006). The majority (90%) of the commercial
spiny lobster catch in Florida comes from the waters surrounding
the Florida Keys. Lobsters have been fished from these waters
since the early 1800s, but it was not until the 1940s that the
fishery shifted almost exclusively to trap fishing (Labisky et al.,
1980). The number of traps increased from 250 000 in the
1970s to 936 000 traps by 1991, with no commensurate increase
in landings. Excess traps in the fishery were an economic concern
of the fishery during this period (Milon et al., 1999; Larkin et al.,
2002), but the primary incentive to reducing the number of traps
was to decrease mortality in undersize lobsters intentionally
retained in traps as live bait (Hunt et al., 1986; Matthews, 2001).
# Public
domain, 2015
It would also reduce the loss of coral reef habitat and entanglement
of protected species by traps and trap rope (Lewis et al., 2009;
Adimey et al., 2014).
Trap fishers in Florida primarily use rectangular traps (81 ×
61 × 46 cm) to catch lobsters. Approximately 90% of these traps
are wooden-slat traps composed of a wood frame and six wooden
slats on all sides (Matthews et al., 2005), but wire is used in a
variety of construction methods to facilitate fishing in deep water,
reducing the need for repairs, and increasing trap longevity. On
the Atlantic Ocean side of the Florida Keys, wooden traps are occasionally reinforced with wire to reduce damage to the traps by sea
turtles and dolphins that break the wood as they try to forage on
the lobsters in the traps; these traps represent 8% of traps in
the fishery. All wire traps represent 1% of traps and are used in
deeper water (.20 m) to facilitate trap pulling and reduce trap
movement from storm surge. Plastic traps represent 1% of lobster
traps (Matthews et al., 2005).
i186
Commercial fishers reported an average annual loss of 18% of
lobster traps, equivalent to 90 000 –100 000 traps when 500 000–
530 000 traps were used in the fishery (Matthews and Uhrin, 2009;
Arthur et al., 2014). During years with hurricanes, trap losses have
reached 65% (FWC, unpublished data). Fishing traps become derelict (lost or abandoned) due to hurricanes and winter storms, entanglement with benthic features, detachment of marker buoys by
boat propellers, and theft (Laist, 1995). Derelict traps and ropes
pose similar threats as fished traps to the marine ecosystem as well
as some unique threats. These threats include the entanglement of
trap lines in protected corals (Chiappone et al., 2005), marine
mammals, and sea turtles (Guillory et al., 2001; Adimey et al.,
2014; National Oceanic and Atmospheric Administration Marine
Debris Program, 2014); destruction of coral (Chiappone et al.,
2002; Lewis et al., 2009) and seagrass (Uhrin et al., 2005); and persistent marine debris in the water and on the shoreline (Ocean
Conservancy, 2014). When derelict traps are intact, they have the
capacity to ghost fish, i.e. they can continue to confine animals, possibly resulting in death by starvation or injury (Smolowitz, 1978;
Hunt et al., 1986; Matsuoka et al., 2005).
Research on the amount of trap debris and the number of ghost
traps in the Florida Keys has raised concerns among fishery managers that lost traps continue to capture lobsters and fish (Uhrin
et al., 2014). When ghost fishing results in the death of lobsters
and fish, it reduces landings in the commercial and recreational
spiny lobster fisheries and negatively affects the ecological role of
these marine animals. Although Florida’s Caribbean spiny lobster
fishery has been the subject of research for decades, no information
is available regarding how long derelict traps can continue to ghost
fish or the magnitude of fish and lobster mortality that results. The
objectives of our study were to: (i) determine how long lost traps
ghost fish, (ii) compare the confinement and mortality of lobsters
and fish in different types of ghost traps, (iii) compare the confinement and mortality of lobsters and fish in ghost traps in three
popular lobster-fishing areas in the Florida Keys, and (iv) estimate
the number of lobsters killed by ghost traps.
Methods
Field observations
Three types of lobster traps were constructed for this study: wooden
traps, wire traps (composed mostly of plastic-coated wire), and
wood –wire hybrid traps (combination of wood and plastic-coated
wire; from here on termed hybrid traps; Figure 1). All traps were of
C. B. Butler and T. R. Matthews
the same rectangular shape and size as those typically used in the
fishery (81 × 61 × 46 cm). Wooden traps were constructed entirely of wooden slats (6 per side) spaced 4 cm apart or less. Wire
traps were constructed with 14 G plastic-coated wire mesh (2.5 ×
5.1 cm). Hybrid traps were composed of three wire sides and one
wooden side. All traps had similar wooden tops with one 10 ×
13-cm rectangular plastic funnel entrance.
Three locations in the Florida Keys were chosen for the ghost
fishing experiment: (i) Florida Bay, 4 nm north of the Middle
Keys in Florida Bay, (ii) Atlantic inshore, 0.5 nm south of the
Middle Keys in the Atlantic Ocean, and (iii) Atlantic offshore,
4 nm south of the middle Keys in the Atlantic, shoreward of
Sombrero Reef (Figure 2). These locations were chosen because
they are popular fishing areas in which traps are commonly lost
(Uhrin et al., 2014). They differ in terms of habitat (Shomer and
Drew, 1982; Bertelsen et al., 2009), as well as in the size and abundance of lobsters. Florida Bay has smaller, generally immature, but
more abundant lobsters. Atlantic inshore generally lacks the smallest
juvenile lobsters, while Atlantic offshore generally has lower lobster
density and more reproductive adults (Lyons et al., 1981). Water
depth was 4 m in Florida Bay and 6 m at both Atlantic inshore
and offshore sites. Two lines of traps, running parallel east to west
(250 m apart), were deployed at each location in mixed sand
and seagrass habitat. The traps were placed 30 m apart and linked
by a line running along the bottom to help divers navigate
between traps when monitoring them. A submerged marker buoy
was located at each end of the trap lines. A numbered plastic tag
was attached to each trap for identification. All traps were deployed
unbaited in September 2009, just after the start of the fishing season.
Thirteen traps of each type were deployed at Atlantic inshore and
Atlantic offshore locations, and 14 of each type were deployed in
Florida Bay (n ¼ 120 traps total). Traps of each type were positioned
randomly along each trap line.
Traps were monitored every 10 –14 d during the first year of the
study, then once a month thereafter, until all traps had been rendered non-fishing. Divers documented whether traps were fishing
(intact) or non-fishing (broken), the number and estimated size
of each live or dead confined lobster (in 5-mm size bins), fish
(and determined total length in cm), and other species trapped.
Small fish (,5 cm TL) that could have moved freely in and out of
the trap were not recorded as confined by the trap. All animals
were left in traps and so may have been observed repeatedly.
Lobster shell remnants were temporarily removed from traps for inspection, so that the remains could be identified as exuviae or
Figure 1. Types of lobster traps used in our study; all traps were standard size with a plastic trap entrance funnel (where lobsters crawl into the trap)
attached to a wooden slat top. Standard wooden traps were composed of wooden slats on all four sides; wire traps were composed of plastic-coated
wire mesh on four sides; and hybrid wood– wire traps were composed of three sides plastic-coated wire mesh and one side wooden slats.
Effects of ghost fishing lobster traps
i187
Figure 2. Map of three locations (Florida Bay, Atlantic inshore, and Atlantic offshore) where two parallel lines of 20 experimental ghost fishing
lobster traps were deployed for this study.
remnants of carcasses. Estimation of lobster and fish size was facilitated by comparison to known slat and wire mesh dimensions and
periodically verified by reviewing photographs, by an estimate by
a second diver, or for lobster by measuring exuviae.
Traps were repaired during the first year of the study to maintain
equal numbers of each trap type at each location for collection of
data to assess animal confinement in ghost traps. Some traps (27
traps) had been damaged by humans; usually a lid had been
opened. In these cases, we closed the trap lid and recorded the
trap as still fishing. Traps that appeared to have been broken by a
turtle or dolphin, i.e. that were missing slats, latches, lids, or
throats (24 instances, 16 traps), were considered to be disabled as
a result of the natural processes that render traps non-fishing and
were repaired but recorded as non-fishing at the time the damage
occurred.
Analysis of ghost trap observations
The length of time a trap remained fishing (days ghost fishing) was
the number of days from deployment to the last time it was observed
intact. Comparisons of “days ghost fishing” among trap types and
location treatments were made using failure-time analyses (Cox’s
proportional hazards model; Kleinbaum and Klein, 2005). “Days
ghost fishing” was substituted for “time to event” in the analysis,
and the resulting “hazard function” for each treatment group was
the probability that a trap would be rendered non-fishing during
the next time interval. Experiments continued for as long as
1071 d, at which point the wire traps that were still fishing (n ¼ 4)
were treated as right-censored data in the analysis (Kleinbaum
and Klein, 2005). Comparisons of the “days ghost fishing” for
traps at various treatments were made using a Wald x 2 test
(Allison, 1995) and by comparing the estimated hazard ratios for
the treatment groups, which indicate the probability of a trap’s becoming non-fishing by the next observation. Because the Cox proportional hazards model offers a non-parametric analysis, median
days ghost fishing and the median absolute deviation (MAD) were
used to present measures of central tendency and deviation.
Since lobsters were not tagged and were left in the traps at
each observation, lobsters in traps were potentially measured repeatedly during serial observations. For this reason, a repeatedmeasures analysis of variance (ANOVAR) was used to determine
the effect of trap type and location on the number of confined
lobsters per trap (Quinn and Keough, 2006). The ANOVAR was
run on the ranked data, because they did not conform to tests of
normality or heterogeneity of variances (Conover and Iman,
1981). Frequencies of size distributions (5-mm size bins) of lobsters
were compared via Kolmogorov–Smirnov’s two-sample tests to
i188
determine whether lobster size distributions varied across locations
or trap types (Sokal and Rohlf, 1995). Effects of trap type and location on the number of fish per trap observation and the per cent
lobster mortality were also analysed via an ANOVAR on ranked
data. Per cent lobster mortality was used in the analysis instead of
the number of dead lobsters observed to account for differences in
the abundance of lobsters between locations. For graphing purposes
only, per cent mortality was not represented on the graph if fewer
than 20 live lobsters were present at any location during a sampling
period to prevent skewing of the graph by the presence of a few dead
lobsters in samples with low lobster abundance.
Lobster mortality estimation
The number of dead lobsters observed per ghost trap and the time to
carcass decay were used to estimate unobserved lobster mortality occurring between trap observations. To eliminate the possibility that
a lobster carcass would be counted more than once, and to account
for unobserved mortalities resulting from rapid carcass decay, we
conducted experiments to determine the time required for a lobster
carcass to decay.
Freshly killed lobsters (n ¼ 20 carcasses) were placed in our ghost
fishing lobster traps and their condition evaluated every 24 h by
divers. Euthanasia for this experiment was humane and was effected
by immersion in 148C seawater for 5 min until movement and reactivity to stimuli ceased and then killing by cooling at 2208C for
30 min (American Veterinary Medical Association, 2013). The
lobster carcasses in traps were considered to have decayed when
the exoskeleton remains no longer contained flesh and were indistinguishable from exuviae. The carcass decay experiment was
conducted 17– 24 November 2011, with seawater temperatures
averaging 258C, the average annual temperature at the study locations. Two or three lobster carcasses were placed into one of eight experimental ghost traps (wood, wire, or hybrid) at our Florida Bay
location. Additionally, to evaluate any effects of other confined
animals on the carcass decay time, the species and quantity of live
animals confined in traps used in the carcass decay experiment
were recorded and regression lines fit to the data (days to decay vs.
maximum number of live lobsters, live fish, or live crabs).
The average number of dead lobsters per ghost trap per year
(+s.d.) was estimated from the bimonthly observations. The
observed number of dead lobsters in each trap was transformed to
an estimated number of dead lobsters in a ghost trap per day by dividing the number of dead lobsters observed in the previous 5.1 d (the
time it took for a lobster carcass to decay) by the total number of lobsters present in the trap. The daily death rate was then multiplied by
the number of days between observations, and the resulting counts
of dead lobsters for 1 year were summed for each trap. The number
of dead lobsters in each of the 13 or 14 traps of each type at the three
study locations was then used to calculate the mean number of dead
lobsters per ghost trap per year (+s.d.).
We estimated the number of dead lobsters from ghost fishing
traps based on our observations of dead lobsters in this study and
the number of ghost traps observed by Uhrin et al. (2014). After
we observed the difference in the number of dead lobsters in ghost
traps between inshore and offshore locations, we used Cochran’s
(1963) stratified random design to recalculate the number of
ghost traps in inshore and offshore Atlantic locations (again using
the data of Uhrin et al., 2014). Uhrin et al. (2014) stratified their
sampling area within the Florida Keys National Marine Sanctuary
into six zones based on historical commercial spiny lobster trap
use zones: Atlantic and Florida Bay zones for the Upper, Middle,
C. B. Butler and T. R. Matthews
and Lower Keys. Using the locations of their towed-diver surveys,
we categorized each of their tows as taking place in Florida Bay,
Atlantic inshore, or Atlantic offshore and calculated an estimated
number of ghost traps for each location using their method.
The estimated number of ghost traps was used to extrapolate,
again using Cochran’s (1963) method, the mean and standard
error of the number of dead lobsters resulting from ghost fishing lobster traps per year and to estimate the number of animals
dying in all ghost traps in Florida Bay, Atlantic inshore, and
Atlantic offshore locations around the Florida Keys. Since the
lobster carcasses potentially decayed and disappeared before they
could be observed, we measured the days required for a carcass to
decay and established that number of days as the period that each
of our observations effectively observed. Establishing the period
that our observation could “effectively observe” if a lobster had
died in a trap reduced the potential for false observations of zero
mortality for times between observations. For each survey, the
time effectively observed was the average time required for a
lobster carcass to decay (from lobster carcass decay experiment).
For example, if the actual time between observations was 14 d,
and the carcass decay time was estimated to be 5 d, then the time effectively observed was the 5 d leading up to the subsequent observation. The number of potential days each year all lost traps were
fishing was set at the number of lost traps estimated by Uhrin
et al. (2014) times 365 d.
Results
Ghost trap fishing duration
Most ghost fishing lobster traps continued to fish for more than 1
year. The length of time that traps ghost fished varied with trap
type and location (Wald x 2 ¼ 29.7, d.f. ¼ 2, p , 0.001; Figure 3).
Hybrid and wooden traps decayed at similar rates (Wald x 2 ¼
3.29, d.f. ¼ 1, p ¼ 0.070). The median number of days traps ghost
fished is expressed as TD50, i.e. the number of days it takes 50% of
traps to decay, +MAD. For hybrid traps, this value was 480 +
142 d; for wooden traps, it was 509 + 97 d. Wire traps took significantly longer to decay, 779.5 + 273.5 d (Wald x 2 ¼ 17.99, d.f. ¼ 1,
p , 0.001). Three wire traps were still fishing when the experiment
ended, at 1071 d. The number of days traps ghost fished also differed
among locations (Wald x 2 ¼ 50, d.f. ¼ 2, p , 0.001). Atlantic offshore traps had a TD50 of 381 + 171 d, Atlantic inshore traps of
509 + 94.5 d, and Florida Bay traps of 711.5 + 51.5 d.
Bycatch
Ghost fishing lobster traps contained fish at the first observation and
continued to contain fish over the duration of the 3-year study. We
observed 66 fish species (5909 total fish) and 13 invertebrate
species (407 total invertebrates) in our ghost fishing lobster traps
(Figure 4, Supplementary Appendix SI). Ten double-crested cormorants (Phalacrocorax auritus) were found dead in wire traps.
Observations during the first year of the project that included
equal numbers of traps of each type at each location indicated that
there was a significant effect of trap type on the number of fish confined in lobster traps (F ¼ 59.77, d.f. ¼ 2, p , 0.001; Figure 5). Wire
traps confined the greatest number of species (wire: n ¼ 102, hybrid:
n ¼ 58, wood: n ¼ 50; Figure 4a) and the greatest number of fish
per trap (wire ¼ 4547 total fish, mean fish per trap ¼ 2.88 + 0.03
s.e.; hybrid ¼ 374 total fish, mean fish per trap ¼ 0.300 + 0.01 s.e.;
wood ¼ 163 total fish, mean fish per trap ¼ 0.167 + 0.01 s.e.;
REGW F-tests, p , 0.05; Figures 4 and 5). The most common fish
i189
Effects of ghost fishing lobster traps
Figure 3. Length of time (days) that wooden, hybrid, and wire lobster traps ghost-fished at three locations (a – c) in the Florida Keys.
observed in traps were white grunts (Haemulon plumieri, n ¼ 2333),
saucereye porgies (Calamus calamus, n ¼ 512), scrawled cowfish
(Acanthostracion quadriornis, n ¼ 345), and yellowtail snapper
(Ocyurus chrysurus, n ¼ 306). Wire traps confined 98% of all white
grunts observed during the study, and white grunts made up 13%
of the finfish bycatch during year 1. Non-lobster invertebrates occurred in fewer than 10% of trap observations. There were 99 invertebrates in hybrid traps, 85 in wire traps, and 69 in wooden traps. Stone
crabs (Menippe spp.) accounted for 84% of the invertebrate bycatch
during the first year of study, spider crabs (Mithrax spp.) accounted
for 6%, and 11 other invertebrate taxa made up the remaining 10%
(Supplementary Appendix SI).
More finfish species were observed in traps at the Atlantic offshore location (n ¼ 84) than in Florida Bay (n ¼ 67) and at our
Atlantic inshore location (n ¼ 59). Bycatch abundance also varied
with trap location (F ¼ 615.33, d.f. ¼ 2, p , 0.001; Figure 5). The
abundance of finfish in traps was significantly greater at the
Atlantic offshore location (n ¼ 1278, mean fish per trap ¼ 2.55 +
0.04 s.e.) than at the Atlantic inshore location (n ¼ 803, mean fish
per trap ¼ 1.30 + 0.02 s.e.) and the Florida Bay location (n ¼ 498,
i190
C. B. Butler and T. R. Matthews
Figure 4. Species richness (a) and total abundance (b) of bycatch confined in hybrid, wire, and wooden ghost fishing lobster traps in the Florida
Keys during year 1 of our study (n ¼ 40 traps for each trap type).
mean fish per trap ¼ 0.953 + 0.03 s.e.; REGW F post hoc tests, p ,
0.05). A significant interaction effect was seen between location
and trap type (F ¼ 30.04, d.f. ¼ 4, p , 0.001), likely driven by
greater fish abundance and species richness observed in Atlantic
offshore wire traps. Fish were more abundant in wire traps
during the beginning of the study (months 1 – 3, Figure 5) at the
Atlantic inshore and Atlantic offshore locations. Another peak
in the number of fish confined in Atlantic offshore wire traps
was observed from late April through July; it corresponded to
abundance of dead lobsters (months 7 – 10, Figures 5c and 7c).
Fish counts in Florida Bay were lower (,5 fish confined per
trap, Figure 5a) than those at other locations throughout the
study. Invertebrates were more abundant in Atlantic inshore
traps (n ¼ 179) than in Florida Bay (n ¼ 46) and Atlantic offshore
traps (n ¼ 28). There were 100 observations of dead fish and 15
observations of dead invertebrates in traps (Supplementary
Appendix SII). The number of bycatch mortalities was greatest
for wire traps, while wooden traps had the fewest mortalities.
For almost half of the dead fish (n ¼ 45), it was not possible to
visually identify species or estimate size.
Effects of ghost fishing lobster traps
i191
Figure 5. Mean fish bycatch confined in ghost fishing hybrid, wire, and wooden lobster traps located at three locations (a – c) in the Florida Keys
during a 3-year study. Bars represent standard error.
i192
Lobster confinement and mortality
We observed 12 813 lobsters confined in our traps during the
first year (25 September 2009 – 11 October 2010) and 16 049 lobsters over the duration of the study (120 traps distributed over
three locations). An unknown number of lobsters were likely
observed multiple times during sequential observations of traps,
making comparisons to fishery trap catch rates inappropriate.
Observations of lobsters during the first year (September 2009 –
October 2010) indicated that hybrid traps confined significantly
more lobsters per trap (4.81 + 0.03, mean + s.e.) than did wood
(3.85 + 0.16) and wire traps (3.71 + 0.03; F ¼ 40.15, d.f. ¼ 2,
p , 0.001; Figure 6). Trap location also affected the number of
lobsters per trap (F ¼ 208.6, d.f. ¼ 2, p , 0.001; Figure 6). Traps
in Florida Bay confined significantly more lobsters per trap
(5.26 + 0.04) than did those at the Atlantic inshore (4.37 +
0.03; post hoc tests, p , 0.05) and Atlantic offshore locations
(1.74 + 0.03; p , 0.05). The interaction term between trap type
and location was significant (F ¼ 3.03, d.f. ¼ 4, P ¼ 0.017);
thus, the number of confined lobsters was a function of location
and trap type.
When deployed, our traps did not contain bait or live, undersized
lobsters, which are typically used in the fishery (Heatwole et al.,
1988); our traps required 2–4 weeks to begin to confine lobsters.
The ghost fishing lobster traps continued to confine lobsters throughout the duration of our study, but the number of lobsters confined
cycled seasonally. The number of lobsters confined generally
increased for the first 4 –6 months for all trap types, after which
they tended to decrease to the lowest levels during summer
(Figure 6). The number of lobsters in traps increased again by
August, and the cycle was repeated for those traps that remained
intact during year 2. The number of lobsters confined in hybrid
and wire traps in Florida Bay peaked in March, with averages of
14.9 + 0.49 (mean + s.e.) per hybrid trap and 12.8 + 0.55 per
wire trap; wooden traps in Florida Bay had the most lobsters in
April, with 11.0 + 0.49 lobsters per trap (Figure 6a). We were
unable to observe traps at the Atlantic inshore location for
4 weeks during February –March due to high turbidity. Atlantic
offshore traps had lower but more consistent confinement rates in
the first year, but the reduced number of intact traps in the second
year likely resulted in the increased variability in confinement
rates and may have obscured any seasonal confinement patterns.
There was no difference in size frequency distributions of the lobsters confined in traps during the first year of study among locations
or among trap types (K –S tests, p . 0.05). Lobster size ranged from
16 to 121 mm carapace length (CL), with the median size of lobsters
falling in the 71 –75 mm CL bin for all trap types and location combinations, except for Atlantic inshore wire and Atlantic offshore
wooden traps, where the median size was in the 76 –80 mm CL
bin. Lobsters confined in ghost traps in Florida Bay ranged in size
from 31 to 111 mm CL; lobsters confined in traps at the Atlantic
inshore location ranged in size from 16 to 121 mm CL; and lobsters
confined in traps at the Atlantic Offshore location ranged in size
from 46 to 111 mm CL.
Observed lobster mortality
During the first year of study, we observed 229 dead lobsters in
ghost traps, but other lobsters probably died in traps between observations as well. Of the observed mortalities, 95 (41%) were legal-size
lobsters (≥76.2 mm CL). There was no difference in the proportion
of dead lobsters by trap type (hybrid ¼ 1.6%, wire ¼ 1.6%, wood ¼
C. B. Butler and T. R. Matthews
2.4%; F ¼ 1.06, d.f. ¼ 2, p ¼ 0.35). There was also no difference in
proportion of dead lobsters per trap among locations (Florida
Bay ¼ 1.5%, Atlantic inshore ¼ 2.1%, Atlantic offshore ¼ 2.5%;
F ¼ 0.157, d.f. ¼ 2, p ¼ 0.86). But more mortalities were observed
in the Florida Bay and Atlantic inshore locations due to the
greater number of lobsters in traps in those areas (Figure 7).
Lobster mortality was high for several months in the late spring
and summer in Florida Bay and Atlantic inshore. Lobster mortality
was also high at the Atlantic offshore location in April, and few lobsters remained in traps through August. The number of lobsters and
fish confined varied over time, as did the number of observations of
lobster mortality (Figures 5 and 7). To determine whether these occupancy rates were correlated, we ran Spearman’s rank correlation
analyses on the non-normal data. The number of lobsters observed
confined in traps was negatively correlated with the number of
fish observed per trap in Florida Bay and Atlantic offshore
locations (p , 0.001, Figures 5 and 7) and was negatively correlated
with per cent lobster mortality at the Atlantic inshore location
(p , 0.05; Figure 7).
Carcass decay
Lobster carcasses decayed in 2 –7 d. The results of a linear regression
indicated a slight negative relationship between the maximum
number of live lobsters present in the traps during the study and
the time to lobster carcass decay (B ¼ 20.9404, R 2 ¼ 0.3931).
There was no relationship between time to carcass decay and fish
(B ¼ 20.0351, R 2 ¼ 0.0292) or crabs (B ¼ 20.0525, R 2 ¼ 0.0015)
present in the traps. During the carcass decay experiment, the abundance of lobsters and fish in traps varied minimally, allowing us to use
the maximum number of lobsters or fish as the dependent variable in
the regressions.
Estimation of lobster mortality in ghost traps
After stratifying the Uhrin et al. (2014) data into inshore and offshore Atlantic locations and recalculating the number of ghost
traps, ghost trap abundance estimates indicated that the largest proportion of ghost traps were located in the Atlantic inshore region
(Atlantic inshore ¼ 72%, Atlantic offshore ¼ 21%, Florida Bay ¼
7%; Table 1). The estimated number of traps increased from the
85 548 + 23 387 s.d. reported by Uhrin et al. (2014) to 112 567 +
57 281 s.d. (after stratifying into inshore and offshore). The standard
deviation for the estimates of ghost traps in each Atlantic area
increased because in the stratified data, there were fewer transects
per area.
Our estimates of the number of dead lobsters per ghost trap,
combined with the revised estimate of the number of traps in the
same areas (Uhrin et al., 2014), suggest that ghost fishing lobster
traps kill 637 622 + 74 367 s.d. lobsters each year. The estimated
number of lobsters dying per ghost trap varied with location.
Florida Bay and Atlantic inshore locations had 6.8 + 1.0 (mean +
s.e.) and 6.3 + 0.88 dead lobsters per trap per year, respectively,
and Atlantic offshore traps had 3.0 + 0.69 dead lobsters per trap
per year (all three trap types combined, as there was no significant
difference in lobster mortality among trap types). The combined
effects of the greatest number of ghost traps estimated at Atlantic
inshore location and the relatively high lobster mortality rate in
the Atlantic inshore location resulted in the greatest portion of
lobster ghost fishing deaths occurring at Atlantic inshore location
(Table 2; inshore ¼ 80%, offshore ¼ 11%, Florida Bay ¼ 8%).
Effects of ghost fishing lobster traps
i193
Figure 6. The mean number of spiny lobsters confined per trap observation in ghost fishing lobster traps located at three locations (a – c) in the
Florida Keys. Three types of traps (n ¼ 120): wood, hybrid, and wire (n ¼ 13). Error bars represent standard error.
i194
C. B. Butler and T. R. Matthews
Figure 7. Mean number of lobsters and per cent dead lobsters observed in traps at three locations (a –c) in the Florida Keys for all three trap types
combined. Error bars represent standard error; bars with error ,0.3 were removed from the graph to best illustrate trends.
i195
Effects of ghost fishing lobster traps
Table 1. Re-evaluation of Uhrin et al. (2014), separating out the
number of ghost fishing traps in inshore and offshore zones.
Zone
Upper Keys Bay
Upper Keys
inshore
Upper Keys
offshore
Middle Keys Bay
Middle Keys
inshore
Middle Keys
offshore
Lower Keys Bay
Lower Keys
inshore
Lower Keys
offshore
Total
Transects
14
7
30
Mean (s.e.) number
of ghost traps per
transect (0.8 ha)
0.071 (0.071)
1.0 (1.0)
0.16 (0.084)
Estimated total
number of ghost
traps
1462
55 725
9288
9
9
0 (0)
0.444 (2.42)
0
18 126
28
0.393 (0.173)
7892
21
7
0.048 (0.048)
0.143 (0.143)
6467
7578
26
0.231 (0.101)
6030
112 567
(s.d. ¼ 57 281)
Table 2. Lobster mortality estimates resulting from ghost fishing
lobster traps in the waters surrounding the Florida Keys.
Zone
Upper Keys Bay
Upper Keys inshore
Upper Keys offshore
Middle Keys Bay
Middle Keys inshore
Middle Keys offshore
Lower Keys Bay
Lower Keys inshore
Lower Keys offshore
Total
Estimated total
number of ghost
traps
1462
55 725
9288
0
18 126
7892
6467
7578
6030
112 567 (s.d. ¼ 57 281)
Estimated total
number of dead
lobsters
9934
352 322
27 572
0
114 603
23 428
43 949
47 913
17 900
637 622 (s.d. ¼ 74 367)
Estimates are based on the number of mortalities observed, the rate of carcass
decay, and the number of ghost fishing traps in each location (Table 1,
modified from Uhrin et al., 2014).
Discussion
Landings of the Caribbean spiny lobster have declined throughout
the Caribbean, and the consensus is that the population is in
decline (FAO, 2007). Landings have also declined in the Florida
Keys, and a recent stock assessment indicates that spiny lobster
abundance has been lower since 2000 (Gulf of Mexico Fishery
Management Council, 2010). The decline in lobster landings may
result from a Caribbean-wide population phenomenon, such as a
decline in larval recruitment (Briones-Fourzán, 1994; Eggleston
et al., 1998; Cruz et al., 2001); from combined local declines from
postsettlement processes such as a infection with the PaV1 virus
(Shields and Behringer, 2004; Behringer et al., 2009); or from
Florida fishing practices that heighten population impacts. While
confinement of Caribbean spiny lobsters in fished traps has been
well studied, the magnitude of lobster mortality occurring in
ghost traps in this fishery has been unidentified until now.
We demonstrated that the typical wooden and hybrid traps,
which account for 90 and 8%, respectively, of the traps used in the
fishery (Matthews et al., 2005), remained intact from 1 to 2 years,
and wire traps were intact for 2 or more years. The estimated time
that a trap remained intact likely represents the maximum length
of time a trap could ghost fish, because these traps were new when
deployed. But the length of time new and used traps can ghost fish
if lost may not differ substantially. Fishers repair used traps by replacing decayed wood and re-nailing or re-stapling the traps to
ensure they function for the duration of the next fishing season
(TRM, pers. obs.). Additionally, 34 of the traps in our experiment
succumbed not to decay of the wood, but to damage by dolphins,
turtles, or stone crabs, and 10 traps were disabled due to breakage
of wire components. Trap location seems to be an important criterion for the time a trap remained intact. Fishers report more rapid
decay of wood traps in the Gulf of Mexico for traps fished farther
from shore than our study location in Florida Bay. In addition to
the often observed high degree of biofouling organisms observed
on traps fished in the Gulf, substantial movement of traps due to
periodic winter winds is a likely cause of trap damage in the open
Gulf (Lewis et al., 2009). Conversely, fishers reported that traps
take a longer time to decay in deeper, offshore waters similar to
our Atlantic offshore location. The 18 traps damaged by humans
were not included in the estimate for the length of time required
for traps to become disabled because the location of the traps near
popular scuba-diving locations and the potential for people to
swim the lines that we had connecting the traps, in this case, likely
facilitated trap molestation. Molesting traps—even derelict traps—
during the closed fishing season is illegal, but the frequency with
which it was observed in this study suggests that divers may have a
substantial role in disabling ghost traps.
Ghost traps continued to capture lobsters and fish for as long as
the traps were intact. Once traps began capturing lobsters (2 –4
weeks after deployment for our unbaited traps), lobster abundance
in traps generally increased through April. Panulirus argus is known
to aggregate with conspecifics (Marx and Herrnkind, 1985;
Herrnkind and Butler, 1986), which is the reason live lobsters are
used as bait in traps. This aggregative behaviour resulted in ghost
traps confining many more lobsters than is typically observed in
commercial catches (Matthews, 2001). But the number of lobsters
in ghost traps also depended on location and trap type. In particular,
hybrid and wire traps in Florida Bay captured their first lobster
sooner than did wooden traps, which may be the reason that there
were more lobsters in these traps throughout the fishing season.
Commercial fishers report lower initial catch of lobsters in new
wooden traps than in traps that have been previously fished, presumably because of the chemicals in pressure-treated wood (TRM, pers.
comm.).
The effects of confinement of lobsters in ghost traps appear
similar to those previously observed for lobsters used as live attractants in traps (Hunt et al., 1986; Heatwole et al., 1988). Confining
lobsters for long durations can subject them to starvation, injury,
predation, or death (Lyons and Kennedy, 1981; Lyons et al., 1981;
Kennedy, 1982; Hunt and Lyons, 1986; Hunt et al., 1986; Lyons,
1986; Lyons and Hunt, 1991). In fished traps, these effects are generally restricted to undersize lobsters, which are most often used as
bait, while legal-size lobsters are removed from traps and so have low
confinement-related mortality (Matthews, 2001). The size distribution of lobsters was the same for ghost traps as for fished traps (Lyons
et al., 1981). Thus, the lethal and sublethal effects of long periods
of confinement extend to the legal-size lobsters confined in ghost
traps, and mortality of legal-size lobsters constituted 41% of the
total lobster mortalities in our ghost traps in the first year. Yet
i196
confinement of lobsters as bait or in ghost traps does not always
result in their death. Approximately 0.8 –1.8% of lobsters in
traps escape each day (Yang and Obert, 1978; Davis and Dodrill,
1980; Lyons and Kennedy, 1981), but there are also sublethal
effects of prolonged confinement, such as slowed growth rates,
weight loss, and moulting with no increase in size (Marshall,
1948; see Wilson et al., 2014, for review). The negative relationship
between the time required for a lobster carcass to decay and the
presence of live lobsters in ghost traps suggests that these normally
highly social lobsters may be resorting to cannibalism to alleviate
starvation.
Our estimate of the number of lobsters dying in ghost traps was
dependent on the total number of ghost traps and the observed mortality in each trap. We found no difference among the three trap
types or locations in the proportion of lobsters dying. But since
more lobsters were confined in Florida Bay and Atlantic inshore
traps, ghost traps in these areas had significantly more dead lobsters.
Given the differences between the number of dead lobsters in
Atlantic inshore and Atlantic offshore ghost traps, we developed
corresponding estimates for the number of ghost traps in these
areas based on survey data from Uhrin et al. (2014). We re-evaluated
Uhrin et al.’s (2014) estimate of the number of ghost traps in the
Atlantic by defining inshore and offshore areas to correspond to
the ghost trap study locations. The re-evaluation of the Uhrin
et al. (2014) data identified that the majority of ghost traps were
located inshore, particularly in the Upper and Middle Keys. The estimated number of ghost traps also increased in this re-evaluation,
to 112 567 + 57 281 s.d. from the previous estimate of 85 548 +
23 387 s.d., but the standard deviation around the new estimate
also increased, at least in part due to the smaller number of transects
in each area of the re-evaluation (Table 1). The concentration of
ghost traps inshore is consistent with the premise that boat traffic,
which is highest inshore, is a primary contributor to trap loss.
Despite the decreased confidence in the estimated number of
ghost traps inshore and offshore, overall utilization of locationspecific estimates of ghost trap mortality improved the estimate
of the number of dead lobsters in ghost traps. Multiplying the
observed number of dead lobsters in each trap in Florida Bay,
Atlantic inshore, and Atlantic offshore by the revised estimates of
the number of ghost traps in the same areas indicates that overall
an estimated 637 622 + 74 367 (mean + s.d.) lobsters die in ghost
traps annually in those areas (Table 2).
Dead fish also were observed in ghost traps. Wire traps confined
more species and greater numbers of fish in all locations than did
wooden traps. Wire traps are illegal in Florida waters because they
resemble illegal fish traps composed primarily of wire mesh, but
they are allowed in adjacent federal waters and in the Caribbean.
In our study, the most fish, living or dead, were observed in wire
ghost traps. Similar results were described by Matthews and
Donahue (1997) for catches in wire and wooden traps fishing in
the Florida Keys. In the present study, it was not unusual to
observe fish in wire traps with abrasions, which were likely caused
by contact with the wire mesh (Sutherland and Harper, 1983).
Most injured fish appeared relatively healthy, but our biweekly
observations of traps were insufficient to evaluate whether these
abrasions caused death. Additionally, we may have so infrequently
observed dead fish in ghost traps because the time between observations of the traps exceeded the time required for a fish carcass to
decay or disappear. In a study by Matthews and Donahue (1997),
most fish carcasses in spiny lobster traps had decomposed within
24 h.
C. B. Butler and T. R. Matthews
Location and time of year influence species composition and
catch abundance (Munro et al., 1971; Guillory, 1993; Wolff et al.,
1999; Garrison et al., 2004). We also observed variation in the
species composition of animals confined in traps at the various locations over time. But catch also depends on the species of fish that encounter the traps and the behavioural characteristics of those species
(Renchen et al., 2012). In this study, fish abundance decreased as the
number of lobsters in traps increased, consistent with the findings of
Matthews et al

Download Report

Lobsters: ocean icons in changing times

Paperzz.com

Your Paperzz