Journal of Experimental Marine Biology and Ecology,
241 (1999) 107–136
L
Settlement-driven, multiscale demographic patterns of large
benthic decapods in the Gulf of Maine
Alvaro T. Palma a , *, Robert S. Steneck b , Carl J. Wilson b
a
´ , Pontificia Universidad Catolica
´
Departamento Ecologıa
de Chile, Alameda 340, Casilla 114 -D, Santiago,
Chile
b
Ira C. Darling Marine Center, School of Marine Sciences, University of Maine, Walpole, ME 04573, USA
Received 3 November 1998; received in revised form 30 April 1999; accepted 5 May 1999
Abstract
Three decapod species in the Gulf of Maine (American lobster Homarus americanus Milne
Edwards, 1837, rock crab Cancer irroratus Say, 1817, and Jonah crab Cancer borealis Stimpson,
1859) were investigated to determine how their patterns of settlement and post-settlement
abundance varied at different spatial and temporal scales. Spatial scales ranged from centimeters to
hundreds of kilometers. Abundances of newly settled and older (sum of several cohorts)
individuals were measured at different substrata, depths, sites within and among widely spaced
regions, and along estuarine gradients. Temporal scales ranged from weekly censuses of new
settlers within a season to inter-annual comparisons of settlement strengths. Over the scales
considered here, only lobsters and rock crabs were consistently abundant in their early postsettlement stages. Compared to rock crabs, lobsters settled at lower densities but in specific
habitats and over a narrower range of conditions. The abundance and distribution of older
individuals of both species were, however, similar at all scales. This is consistent with previous
observations that, by virtue of high fecundity, rock crabs have high rates of settlement, but do not
discriminate among habitats, and suffer high levels of post-settlement mortality relative to lobsters.
At settlement, large, habitat-scale differences exist for lobsters but not for rock crabs; these are
probably the result of larval settling behavior. In contrast, patterns at the largest, inter-regional,
spatial scales suggest oceanographic control of larval delivery. Increased mobility and vagility
with greater body size for both species reduces demographic differences among older individuals
over a range of spatial scales.  1999 Elsevier Science B.V. All rights reserved.
Keywords: Settlement; Post-settlement; Newly settled individuals; Spatial scale; Temporal scale;
Predation; Subtidal; Gulf of Maine; Cancer irroratus; Cancer borealis; Homarus americanus
*Corresponding author. Tel.: 156-2-686-2971; fax: 156-2-686-2621.
E-mail address: [email protected] (A.T. Palma)
0022-0981 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved.
PII: S0022-0981( 99 )00069-6
108
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
1. Introduction
Ecologists seeking to understand factors that affect patterns of distribution and
abundance of organisms must consider the problem over a daunting range of temporal
and spatial scales. Understanding the appropriate scale in ecological research presents
both practical and theoretical challenges (Levin, 1992, 1993). Central among these is the
challenge of linking the scale of demographically important patterns with the scale of the
pattern-driving processes (Horne and Schneider, 1994).
Dispersal among marine organisms with pelagic larvae is scale-dependent. The
abundance of larval stages (Roughgarden et al., 1991), early post-larval stages (Jones,
1991) or a combination of both (Roughgarden et al., 1988) has a major impact on the
demography of these of organisms. If population densities of benthic marine organisms
are regulated by settlement success, as has been shown for reef fish (Doherty and
Fowler, 1994), barnacles (Connell, 1985; Gaines and Roughgarden, 1985), and many
other benthic and demersal marine organisms (Underwood and Fairweather, 1989), then
factors contributing to successful settlement into appropriate nursery grounds may be as
important as broodstock abundance in determining the size of future populations.
However, larvae may select particular habitats at the scale of millimeters when settling,
while the ocean currents may have transported those larvae hundreds of kilometers.
Similarly, larvae may settle in regular patterns within a season but show significant
inter-annual variability.
In general, successful settlement, hence recruitment to the benthos, requires (1)
available competent larvae, which depends on both sufficient egg and larval production
and oceanographic dispersal (Underwood and Fairweather, 1989), (2) the propensity to
settle [sounding behavior and appropriate tactile, visual or chemical cues (Boudreau et
al., 1992)] and (3) available nursery ground habitats where post-settlement survivorship
is high (Gotceitas and Brown, 1993).
Identical patterns in post-settlement abundance can result from very different
settlement and early post-settlement processes, all dependent on the scale of observation
(Jones, 1997). For example the abundance of newly settled individuals in nursery
grounds can result from (1) individuals actively selecting certain habitats (Steger, 1987;
Wahle and Steneck, 1991; Fernandez et al., 1993; Eggleston, 1995; Eggleston and
Armstrong, 1995; Morgan et al., 1996), or (2) individuals settling indiscriminately but
surviving in nursery grounds through differential mortality between nursery grounds and
other habitats (Lough et al., 1989; Gotceitas and Brown, 1993; Palma et al., 1998).
Far too many studies in marine ecology have one or a few study sites. From those
studies, generalizations are made that relate to the entire region – sometimes to the
entire field of marine ecology. In studies ranging over wide geographic ranges it has
been found that oceanographic differences can affect the fundamental structure of
marine communities (e.g., Connolly and Roughgarden, 1998). This opens a large set of
questions to the study of ecology; namely, at what scales and under what conditions do
ecological patterns and processes react? While most modern ecological studies take great
pains to replicate sites within a region, not many of them have looked at other sources of
spatial variability. In this study we advance the general hypothesis that very small-scale
variability can result from organismic behaviors but large-scale variability can result
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
109
from oceanographic forcing functions. Since the former is invariably controlled by the
latter, the problem is general. However, determining the scale of such interactions
requires considerable exploring research focused on the question of relevant scale for
ecological processes. Our study is one of the first in the western North Atlantic to
consider this question.
In the Gulf of Maine three large decapods, the American lobster Homarus americanus, the rock crab Cancer irroratus, and the Jonah crab Cancer borealis, are
ecologically and economically important as large-bodied benthic invertebrates (Krouse
and Cowger, 1980; Ojeda and Dearborn, 1990; Fogarty, 1995; Lawton and Lavalli,
1995). The occurrence of these species was recorded in several sites within four
widely-separated coastal regions in the Gulf of Maine. Individuals were quantified on
different substrate types, on habitats of different exposure relative to prevailing currents,
and along bathymetric and estuarine gradients.
The purpose of this study was to consider the scale at which patterns of abundance
occur among decapods in the Gulf of Maine and integrate this with recent scaledependent process-level surveys. Different age groups, newly settled vs. older individuals, were studied at spatial scales ranging from different microhabitats centimeters
apart to regions hundreds of kilometers apart. Temporal scales ranged from weekly to
inter-annual comparisons. One goal of this study is to consider how and at what scales
the abundance of settlers translates into the demography of older segments of
populations. This study elucidates the emergent demography of these species at all these
scales.
The approach followed throughout this study considered the comparison of the
number of individuals at these various spatial and temporal scales. For each case the
underlying hypothesis is that their abundances do not differ. The acceptance or rejection
of these multiple hypotheses will allow us to discuss our findings within the scope of the
more general hypothesis stated above.
2. Materials and methods
2.1. Survey sites and methods
Field surveys were made at several sites within four regions along the coast of Maine,
from west to east: York, Pemaquid, Penobscot Bay, and Mount Desert Island (Fig. 1A).
In all sites, the abundance of the three study species, Homarus americanus, Cancer
irroratus, and Cancer borealis (hereafter also referred to as lobster, rock crab, and Jonah
crab, respectively), was quantified through visual and air-lift suction sampling SCUBA
surveys. The former method is primarily used to estimate the abundance of larger, less
cryptic individuals, normally older than 1 yr. (hereafter also referred to as 1YR 1 ). Size
criteria for the 1YR 1 group were: . 10 mm CW (carapace width) crabs and CL
(carapace length) lobsters. In each site over the same depth range (around 10 m below
MLW), the visual surveys consisted of 1.0 m 2 quadrats that were haphazardly tossed on
different substrate types, and all individuals therein counted and measured in situ, then
released. Air-lift suction sampling was used to estimate the abundance of the less-
110
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Fig. 1. Four coastal study regions and sites within (A) the Gulf of Maine: (B) York (A, B, C, D), (C)
Pemaquid and Damariscotta River (GL, Glidden Ledge; McGP, McGuire Point; PI, Peters Island; JC, Jones
Cove; HIC, Holly Inn Cove; KRE, Kresge; XMAS, Christmas Cove; TOL, Thread of Life; FIW, Fisherman Is.;
DIW, Damariscove Is. West Side; DIE, Damariscove Is. East Side), (D) Penobscot Bay (SP, Squaw Point; OC,
Orr Cove; DT, Ducktrap; TH, Thomaston; HI, Hurricane Is.; IAH, Isle Au Haut; AI, Allen Is.; RI1, Ragged Is.
1; RI2, Ragged Is. 2), and (E) Mount Desert Island (BI, Baker Is.; BLI, Black Is.; GDI, Great Duck Is.; LIS,
Long Is. South Side).
apparent, newly settled or young-of-the-year (hereafter also referred to as YOY)
individuals (crabs and lobsters , 10 mm CW and CL, respectively). Despite the
similarity between newly settled rock and Jonah crabs, they were positively identified
based on previously described differences (Williams and Wahle, 1992). For each census,
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
111
0.25 m 2 size quadrats were used. For details on the air-lift suction sampling procedure,
see Wahle and Steneck (1991). The abundance of YOY and 1YR 1 individuals along
depth gradients and along estuarine conditions was estimated employing 0.24 m 2
decapod-settler collectors constructed of PVC (polyvinyl chloride) pipe. Collectors were
SCUBA deployed before the settlement season started (mid-July) and retrieved after the
season was over (end September). Collectors were square structures (60 3 40 3 9 cm)
made of 3 cm diameter grey PVC pipes spaced 1 cm apart and stacked three rows deep
in alternating tiers. The stack was placed on AstroturfE inside a wire basket lined with 1
mm mesh nylon screen. The basket was made of 3.5 cm mesh vinyl-coated wire
commonly used for lobster traps. For more details on collector description and similar
general field methodology, see Palma et al. (1998).
2.2. Patterns between substrata (centimeter to meter scale)
The use of substrate by newly settled and older individuals of all species was studied
through seasonal surveys made in the seven outer-coastal sites in the Pemaquid region
between 1994 and 1997 (Fig. 1C). In each site suction samples were taken from specific
substrate types (sand and cobble) considered extremes in a continuum of grain sizes.
This choice was made also based on the high preference for cobble habitats by the early
benthic phase of lobsters (Wahle and Steneck, 1991; 1992) and the fact that at least rock
crabs had been found capable of settling on coarse sand as well as on cobble with
comparable densities (Palma et al., 1998). Estimates of the abundance of older
individuals were more efficient by visual surveys since they are more conspicuous. In
addition, the densities of older stages of either decapod species are normally lower than
that of settlers, and also because boulders were added as another type of substrate
utilized by these larger individuals, not efficiently surveyed using the air-lift method.
The comparisons of abundance between substrata and among species corresponded to
the average densities found at seven sites during 4 years, although not all sites were
surveyed every year. Thus, the sample size considered here was: eight and 11 averages
on sand and cobble, respectively, for newly settled individuals and 15 and 16 on sand
and cobble 1 boulder, respectively, for older individuals.
2.3. Patterns among depths (meters to tens of meters scale)
The use of PVC collectors allowed us to standardize the substrate at different depths
and locations since the presence of suitable settlement substrata for lobsters (i.e. cobble)
at the study site (DIW) tends to be most common at shallow depths (5 m), rapidly
decreasing with depth (C. Wilson, personal communication). Previous tests showed that
these collectors capture post-larval lobsters and crabs as effectively as cobble substrata
(Steneck et al., unpublished). At each depth (5, 10, and 20 m) 20, eight, and eight
collectors were randomly placed at least 1.0 m apart along the respective isobaths,
normally parallel to the coastline. This experiment was deployed between mid-July and
mid-September of 1995.
112
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
2.4. Predator exclusion experiment (tens to hundreds of meters scale)
To test the effect of benthic predators on newly settled rock crabs in sites separated by
hundreds of meters, we deployed a meso-scale exclusion experiment on opposite sides of
Damariscove Island (DIW and DIE) during the settlement season of 1997 (August–
September) (Fig. 1C). The difference in abundance of new settlers inside and out of
cages would also indirectly indicate potential differences in physical factors (i.e. wind,
currents) that could have influenced these differences. The island is approximately 3 km
long by less than 1 km wide and is aligned nearly north–south. Twenty exclusion and
open cages were placed at each side of the island near a narrow isthmus (less than 50 m)
that separates the two experimental sites. The experiment was deployed at 10 m below
MLW on similar sand patches. In each site 10 open and 10 predator-exclusion cages
were randomly interspersed and placed 1.5 m apart in a square design. Each unit
consisted of a 0.5 m 2 (71 3 71 cm) wooden frame filled with sand. The sand, obtained
days earlier in the same area, was washed and dried in the laboratory to ensure no
artificial addition of already settled individuals then brought to the study site. The total
exclusion treatment consisted of frames covered with 4 mm size mesh (for more details
on design, see Palma et al., 1998). Previous tests with partial cages (roofs but not sides)
showed no cage effect. All frames (open and caged) were suction-sampled completely
before starting the experiment 1 week later. This experiment was designed only to
quantify newly settled rock crabs; lobsters were not expected to be found because the
small size of the mesh and the type of substrate (sand) was not appropriate for this
species. The experiment started on August 21 and continued through September 10. In
this period all cages were surveyed weekly (three times) using the same air-lifting
technique as described above. From a total of 30 plots considered in this design (30 for
each treatment and island side), 28 and 29 corresponded to the exclusion treatment on
the west and east side of the island, respectively, and 27 and 28 for the open treatment
on the west and east side of the island, respectively. The missing data corresponded to
cages that broke open or plots that were lost. After each survey the cages were filled
with new sand and the experiment restarted. This was a two-factor experimental design
with island side (west and east) and cage treatment (full cage and no cage) as
independent variables. The response variable was number of newly settled rock crabs /
0.5 m 2 . It was possible to pool the three separate surveys since their mean square errors
were very similar (54.43, 48.13, 44.72), so the data were analyzed with a two-factor
ANOVA.
2.5. Intra-regional patterns ( hundreds of meters to tens of kilometers scale)
The comparison of abundances within regions was made in the Pemaquid and Mount
Desert regions, both with the longest uninterrupted data record (1994–97). In this
comparison, only open-coast sites with the same westward exposure were considered.
The selected sites in Pemaquid were: DIW, FIW, and KRE (Fig. 1C), and in Mount
Desert Island: LIS, BLI, GDI, and BI (Fig. 1E). The sampling procedure was the same
as described above, combining air-lift suction samples for the estimation of new settlers
and visual surveys for the estimation of older individuals. Surveys for lobster settlers
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
113
were only performed in cobble beds, while sand was also considered for crabs. For each
case, the sample size corresponds to the number of quadrats surveyed at each site during
these years combined.
2.6. Patterns up estuaries (tens of kilometers scale)
In the summer of 1995, PVC collectors were placed at 10 m below MLW in five sites
several kilometers apart up the Damariscotta estuary. Collectors were randomly placed
from the outer-coastal DIW to four sites along the river (JC, PI, McGP and GL, Fig. 1C)
in numbers of 8, 16, 14, 12, and 14, respectively. In addition to ensuring the
standardization of substratum with collectors, all sites had a westward exposure. The
salinity gradient along the Damariscotta River is minimal, varying during the summer
from 31‰ at the open coast and river mouth, to about 29‰ at Glidden ledge (McAlice,
1977). Despite little change in the overall salinity, the Damariscotta River has a typical
estuarine circulation, however poorly stratified, particularly during the summer
(McAlice, 1977; Mayer et al., 1996). A similar pattern exists in Penobscot Bay, where
despite a greater freshwater input, salinity values only fluctuate between 27 and 31‰ at
the head and mouth of the bay, respectively (N. Pettigrew, personal communication).
Because these salinities are well within the tolerance units of the larval and adult phases
of these species, this study represents a useful test of how settlement patterns may be
influenced by estuarine transport processes. Estimates of the abundance of newly settled
and older individuals in Penobscot Bay were made at the end of the settlement season of
1997 (September) in nine sites throughout the Bay (Fig. 1D). Estimations of abundance
were made by surveying only natural cobble substrata. The abundance of new settlers
was estimated by haphazardly tossing twelve 0.25 m 2 quadrats at each site and
collecting all decapods using the air-lift. The quantification of older individuals on
cobble substrata at the same sites was made by surveying a similar number of 1.0 m 2
haphazardly tossed quadrats. The selected sites had equivalent coverage of cobble
substrata and were scattered throughout Penobscot Bay.
2.7. Inter-regional patterns ( hundreds of kilometers scale)
The same sites considered for the intra-regional comparisons (within Pemaquid and
Mount Desert) were combined to compare abundances among regions. Two more
regions were added to this analysis (York and Penobscot Bay). Data from York came
from four sites with similar exposure (Fig. 1B), and data from Penobscot Bay came from
only the six, more exposed, sites near the mouth of the Bay (Fig. 1D). In the case of the
last two regions, suction sample data were only available for 1997; visual surveys for
1995 and 1997 in York; and only 1997 in Penobscot Bay. The sample size for each
region is indicated in the legend of Fig. 9.
2.8. Temporal patterns (weekly and annual scales)
The change in abundance of newly settled lobsters and crabs was quantified during the
late-summer period in weekly censuses for seven consecutive weeks in 1994. One site
114
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
(DIW, Fig. 1C) was chosen for its long history of high post-larval supply and
recruitment (Incze and Wahle, 1991; Wahle and Incze, 1997; Palma et al., 1998; Incze et
al., 1999). Four 1.0 m 2 previously placed square plots filled with cobble were surveyed
every week using the air-lift (see Palma et al., 1998, for more details on the experimental
design).
The inter-annual comparison of abundance was made only with data from the sites in
Pemaquid between 1994 and 1997. The methodology was the same as described earlier.
The sample size for each year is indicated in the legend of Fig. 11.
The data were analyzed using parametric approaches, and given the correlational
nature of the data (primarily, measurements of the abundance of individuals in numbers /
m 2 ), the predominant method for overall comparison was by means of ANOVAs for
fully factorial models (SYSTAT 1992). The identification of lower-level differences was
always done (although not always shown) mainly using Tukey’s HSD and Fisher’s LSD
post-hoc tests for simultaneous pairwise mean comparison, with significances chosen so
that overall probabilities were equal to the critical value of 0.05. Assumptions of
normality, homogeneity of variance, and independence of observations were tested and
the appropriate transformations performed when necessary (Zar, 1974; Sokal and Rohlf,
1981).
To investigate how abundance varied among replicate quadrats and what proportion of
its overall variation was explained at this scale, a subset of abundance data were
analyzed at three different spatial scales (quadrat, sites, and regional scales). This
analysis was done for newly settled individuals (YOY) and older stages (1YR 1 ) of H.
americanus and C. irroratus. The data were analyzed using a nested analysis of
variance, where the F-ratios were obtained considering the mean square estimates (see
Underwood and Chapman, 1996; Underwood, 1997).
3. Results
3.1. Patterns between substrata (centimeters to meters scale)
In the surveys performed in the Pemaquid region from 1994 to 1997, newly settled
lobsters were found only on cobble substrata (F1,17 5 10.84, P 5 0.004, Fig. 2A). Similar
surveys on adjacent finer-sediment substrata (sand of varying coarseness) yielded a
conspicuous lack of newly settled lobsters, or densities below detection. In comparison,
rock crabs settled at significantly higher densities, and with no significant differences
between cobble and sand (F1,17 5 0.017, P 5 0.899, Fig. 2A). Newly settled Jonah crabs
were never observed in the Pemaquid region surveys. Older individuals of the three
species occurred on both types of substrata, and Jonah crabs were significantly less
abundant than the other two decapods (F2,90 5 28.1, P , 0.001). The abundance of
lobsters and rock crabs older than 1 year on cobble and boulder bottoms was higher than
on sand, but that difference did not exist for Jonah crabs (F1,29 5 32.07, P , 0.001;
F1,29 5 8.10, P 5 0.008; F1,29 5 3.94, P 5 0.057, respectively, Fig. 2B). For all species,
the population densities were different between newly settled and older individuals. For
lobsters and rock crabs, the density of settlers was greater than that of older individuals
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
115
Fig. 2. Centimeter to meter scale substrate-related patterns of abundance for (A) newly settled (YOY), and (B)
older (1YR1) decapods in the Pemaquid region. Number of 0.25 m 2 quadrats used to quantify YOY and 1.0
m 2 for 1YR1 individuals (61 SE) for each substrata shown inside legend frame. Dashed horizontal lines
represent comparison between species and continuous horizontal lines represent comparison within species by
substrate. Interrupted lines represent significant differences.
(lobsters: F1,33 5 3.85, P 5 0.046; rock crabs: F1,33 5 5.60, P 5 0.024), but for Jonah
crabs it was lower (F1,33 5 11.46, P 5 0.002).
3.2. Patterns among depths (meters to tens of meters scale)
Settlement of lobsters into PVC collectors was restricted to shallow depths without
116
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
significant difference between 5 and 10 m, and below detectable levels at 20 m
(F2,33 5 2.33, P 5 0.113, Fig. 3A). In comparison, newly settled rock crabs were more
abundant at all three depths, although densities were significantly higher, but equal, at 5
and 10 m (F2,33 5 8.01, P 5 0.001, Fig. 3B). The abundance of older individuals that
walked into the collectors was similar throughout the depth range for both species (Fig.
3C,D). Only older lobsters showed significantly higher densities at 5 compared to 20 m
(lobsters: F2,33 5 4.50, P 5 0.019; rock crabs: F2,33 5 2.24, P 5 0.122).
3.3. Intra-regional patterns and predator exclusion experiment (tens of meters to tens
of kilometers scale)
From data obtained during the settlement season of 1994, the abundance of both
newly settled and older lobsters and rock crabs was estimated in two nearby sites on
opposite sides of Damariscove Island (DIW and DIE). These sites are separated by only
a few hundred meters across an island isthmus, but more than 2 km along shore (see Fig.
1C). In those observations the ratio of newly settled lobsters was six-fold (4.561.5 at
DIW versus DIE 0.860.5 indiv. / m 2 61 SD) compared to more similar densities for
newly settled rock crabs (12.368.1 at DIW versus DIE 9.464.0 indiv. / m 2 61 SD).
Fig. 3. Meters to tens of meters scale depth-related pattern of abundance of (A,B) newly settled (YOY) and
(C,D) older decapods (1YR1) (61 SD) inside PVC collectors at different depths (5, 10 and 20 m). Number
of collectors used were 20, 8, and 8 at 5, 10, and 20 m, respectively. Different letters above bars mean
significant differences among depths. (*) No individuals were detected.
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
117
However, the abundance of older individuals of both species was similar on the two
sides of the island (lobsters: 4.561.1 and 2.560.6; rock crabs: 6.061.9 and 8.262.2
indiv. / m 2 61 SD at DIW and DIE, respectively).
The above data are consistent with those found for lobsters from the same two sites by
Wahle and Incze (1997). This suggests that the supply of larvae at this scale is different
for these two species. Wahle and Incze (1997) proposed that different supply regimes
(wind-driven currents) were the main factor responsible for such difference in postsettlement abundances of lobsters. The results obtained from the exclusion experiment
on the two sides of the island in 1997 strongly suggest that this is not the case for rock
crabs. Abundance of newly settled rock crabs inside the exclusion cages (less than 1
week after they reach the benthos) was not significantly different between sides of the
island (Table 1). Although the difference in absolute abundance between the total
exclusion and the open treatments was significant, the differences were proportional for
the two sides of the island (Fig. 4). Thus, collections made on a weekly basis still appear
to reflect the supply signal on both sides.
The abundance of newly settled lobsters was similar when three sites with identical
south-westward exposure in the Pemaquid region were compared (F2,88 5 1.32, P 5
0.273, Fig. 5A), however newly settled rock crabs showed more variability and an up to
five-fold difference among sites was detected (e.g., FIW vs. KRE, F2,157 5 7.10,
P 5 0.001, Fig. 5B). The abundance of older individuals was lower but within the same
magnitude for both species in all sites (,1.0 indiv. / m 2 ). Values were significantly lower
for lobsters at KRE and significantly higher for rock crabs at DIW (lobsters: F2,991 5
8.86, P , 0.001; rock crabs: F2,991 5 4.90, P 5 0.008, Fig. 5C,D). For each site in the
Pemaquid region the difference between the abundance of newly settled and older
individuals was always significant for both species (see Fig. 5).
At the four sites in Mount Desert region, the abundance of newly settled lobsters
occurred at levels below detection. The abundance of older individuals, although with
slight differences among sites (Fig. 6A) (F3,1139 5 10.67, P , 0.001), was significantly
lower than that recorded in the Pemaquid region (0.89 vs. 0.25 indiv. / m 2 , F1,6 5 6.47,
P , 0.001). Density of newly settled rock crabs, although lower than that reported for
Pemaquid (22.61 vs. 4.57 indiv. / m 2 , F1,216 5 18.044, P 5 0.001), was quantifiable in
three of the four sites with no significant difference among them despite a high density
of settlers in LIS (Fig. 6B) (F3,71 5 1.71, P 5 0.173). Similarly, the abundance of older
individuals was remarkably similar among all sites (F3,1139 5 0.86, P 5 0.459). Despite
the overall greater abundance of settlers compared to older individuals, there was no
Table 1
Two-factor ANOVA to compare the abundance of newly settled C. irroratus at the east and west sides of
Damariscove Island (Island side) under open or exclusion of predators (Treatment) conditions. The response
variable is number of new settlers 0.5 m 2 . Sample sizes as shown in Fig. 4
Source
df
SS
MS
F-value
P-value
Island side
Treatment
Island side3Treatment
Error
1
1
1
108
86.620
5477.121
46.554
10 443.199
86.620
5477.121
46.554
96.696
0.896
56.643
0.481
0.346
,0.0001
0.489
118
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Fig. 4. Cancer irroratus. Abundance of newly settled individuals inside predator-free exclusion and open
control cages on the east and west side of Damaricove Island (61 SE). Experiment was conducted during the
settlement season of 1997. Continuous horizontal lines represent comparison by treatment and dashed
horizontal lines represent comparison by island side. Interrupted lines represent significant differences.
significant difference between their densities when the data of the four sites were pooled
and compared (T 5 1.537, df574, P 5 0.129).
3.4. Patterns up estuaries (tens of kilometers scale)
The abundance of settling lobsters inside PVC collectors significantly decreased from
an outer coastal site (DIW) up the Damariscotta River estuary. At DIW, the abundance
of older lobsters was greater than that of newly settled ones, and although equal between
age groups, significantly lower at JC and below detectable levels in any of the three sites
further up the estuary (Fig. 7). When all sites were considered, there was a significant
decrease of all lobsters up the river, with a significant interaction between site and age
group (Table 2). Similarly, the abundance of newly settled rock crabs decreased with
increasing distance from DIW, however settlers were still observed much higher up the
estuary (McGP, 18.3 km from DIW). Except for JC, no site showed a significant
difference between settlers and older rock crabs (Fig. 7, Table 2).
In Penobscot Bay, newly settled lobsters were found in a few of the southern study
sites (Allen to Hurricane Island), all near the mouth of the bay (Fig. 1D). Densities were
low to below detection and highly variable, no significant differences were detected
(F8,109 5 1.38, P 5 0.212). On the other hand, older individuals were present in all nine
sites at highly variable densities, with the lowest densities up the bay (F8,708 5 23.06,
P , 0.001). In only two of the sites (Ragged Is.1 and Thomaston) was the difference
between YOY and 1YR1 significant (Fig. 8, Table 3). The abundance of newly settled
rock crabs was significantly higher in the seaward sites, gradually decreasing towards the
head of the bay, falling below detectable levels in the three sites farthest from the open
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
119
Fig. 5. Hundreds of meters to tens of kilometers scale among-site pattern of abundance in the Pemaquid
region. Abundance of (A,B) newly settled (YOY) and (C,D) older decapods (1YR1) (61 SE). Values
represent the averages of all quadrats surveyed. The number of 0.25 m 2 quadrats used to quantify YOY and 1.0
m 2 for 1YR1 individuals is shown inside bars. For each species and age group, different letters above bars
mean significant differences among sites.
sea (F8,109 5 18.511, P , 0.001). In contrast, and like lobsters, older rock crabs were
recorded in most sites at low and quite variable densities (F8,708 5 8.505, P , 0.001).
Although the site-specific differences between YOY and 1YR1 were not significant
(except Ragged Is. 1 and 2), the significances were marginal, and most likely due to very
low values with large variance. The overall significance was due to comparatively high
densities of settlers in two sites, with significant interaction between site and age group
(Ragged Is. 1 and 2; Fig. 8, Table 3).
3.5. Inter-regional patterns ( hundreds of kilometers scale)
The combined average density of newly settled lobsters was significantly greater in
the two western-most regions (York and Pemaquid) compared to the two eastern-most
regions (Penobscot Bay and Mount Desert) and was even below detectable levels in the
Mount Desert region (F3,216 5 14.56, P , 0.001, Fig. 9A). Although more abundant, the
densities of newly settled rock crabs exhibited a similar trend (F3,216 5 11.24, P , 0.001,
Fig. 9B). The abundance of newly settled Jonah crabs was the lowest of all, with
measurable levels only in the York region (Fig. 9C). The abundance of older individuals
of the three species was more homogeneous and within the same order of magnitude
120
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Fig. 6. Hundreds of meters to tens of kilometers scale among-site pattern of abundance in the Mount Desert
region. Mean number of newly settled (YOY, black bars) and older decapods (1YR1, open bars) (61 SE).
Numbers of 1.0 m 2 quadrats for quantifying 1YR1 individuals were 189, 201, 320, and 433 and the numbers
of 0.25 m 2 quadrats for quantifying YOY individuals were 24, 24, 4, and 23 in LIS, BLI, GDI, and BI,
respectively. For each species, different letters above bars mean significant differences among sites. Prime
letters identify 1YR1 comparisons for C. irroratus.
among regions (,1.0 indiv. / m 2 ). Older lobsters occurred at significantly lower densities
in Mount Desert (F3,63 5 15.15, P , 0.001, Fig. 9D) and older rock crabs at significantly
higher densities in Pemaquid (F3,63 5 7.62, P , 0.001, Fig. 9E). Jonah crabs occurred at
similar densities in all the regions (F3,63 5 3.37, P 5 0.064, Fig. 9F). There was an
overall tendency for the density of settlers to be greater than that of older individuals
(although not always significant). This tendency was particularly strong for rock crabs in
York (F1,18 5 19.92, P , 0.001) and Pemaquid (F1,107 5 6.26, P 5 0.014), and to a lesser
extent for lobsters in Pemaquid (F1,107 5 4.06, P 5 0.046).
3.6. Temporal patterns (weekly and annual scales)
Although the abundance of newly settled lobsters and rock crabs in DIW was variable
over time, when seven consecutive weeks were considered, the temporal variability for
the two species followed similar trends, with a positive and significant relationship
(r 2 5 0.515, P 5 0.005, Fig. 10). Both species started settling within the same first 2
weeks of the surveying period, peaked within the same 2 weeks, and only diverged at
the end of the experimental period when crab settlement started to decline while
lobster’s was still high.
On an inter-annual scale the abundance of newly settled lobsters in PEM was not
significantly different among years (1994–97) (F3,12 5 0.177, P 5 0.910, Fig. 11A). In
contrast, newly settled rock crabs showed much higher temporal variability, with
significantly greater abundance in 1997 compared with the two preceding years (F3,12 5
4.235, P 5 0.029, Fig. 11B). The abundance of older lobsters was variable over time
with values for 1994 higher than the two following years. The abundance of older rock
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
121
Fig. 7. Tens of kilometers scale up-estuary pattern of abundance in five sites along the Damariscotta River
[Damariscove Island west (DIW), Jones Cove (JC), Peters island (PI), McGuire Point (McGP), and Glidden
Ledge (GL)]. Mean number of newly settled (YOY) and of older individuals (1YR1) (61 SD) inside (8, 16,
14, 12, and 14, respectively) 0.24 m 2 PVC collectors from the open coastal site (DIW) up the river. For each
species and site, horizontal lines represent comparisons between age groups. Interrupted lines mean significant
differences.
crabs was also variable, with 1994 having the highest abundance (lobsters: F3,14 5 6.88,
P 5 0.004; rock crabs: F3,14 5 14.95, P , 0.001, Fig. 11C,D).
4. Discussion
4.1. Spatial small- and meso-scale-related patterns
Patterns of distribution and abundance for each decapod species were different at
122
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Table 2
Damariscotta estuary. Two-factor ANOVA to compare the abundance of different age groups (YOY vs.
1YR1) of decapods at five different sites (DIW, JC, PI, McGP, and GL). Data are number of individuals
inside PVC collectors. Sample sizes as shown in Fig. 7
Source
df
H. americanus
Site
Age group
Site3Age group
Error
SS
4
1
4
118
101.00
9.14
21.00
201.00
C. irroratus
Site
Age group
Site3Age group
Error
4
1
4
118
4390.81
90.28
1958.14
4377.05
MS
25.250
9.14
5.25
1.70
1097.70
90.28
489.54
37.09
F-value
P-value
14.82
5.37
3.08
,0.0001
0.022
0.019
29.59
2.43
13.20
,0.0001
0.121
,0.0001
varying spatial and temporal scales. The largest differences between lobsters and rock
crabs occurs at the smallest spatial scales and at the time of settlement. Settlement
densities of H. americanus are consistently lower than those of C. irroratus, and lobster
settlement is restricted only to rocky, cobble bottoms. Newly settled lobsters were
conspicuously absent in sand, and given the number of quadrats surveyed, temporal
coverage (1994–97), and techniques used (exhaustive air-lift suction sampling and
visual surveys), it is unlikely that the sampling effort was not large enough to detect any
(but see Barshaw and Rich, 1997). Rock crabs, on the other hand, settled indiscriminately on rocky and sandy bottoms. The absence or extremely low densities of newly settled
Jonah crabs does not allow further conclusions about their potential settlement-substrate
preferences. Newly settled Jonah crabs were found only in the southern-most region of
York in 1997. In contrast, the abundances of older individuals of the three species
(particularly in the case of rock crabs) were within the same order of magnitude (,1.0
indiv. / m 2 ). Overall, the abundances of older lobsters and rock crabs (on hard and soft
substrata) were similar to each other and higher than that of Jonah crabs. The strong
pattern of cobble habitat selection in newly settled lobsters is lost with age, as older
lobsters become more common on soft bottoms. Rock crabs that settled indistinctly on
both substrata showed a slight preference for hard substrata as they became larger,
whereas Jonah crabs showed no clear preference
Along a bathymetric gradient, the reduction in the number of settlers with depth was
significant for both lobsters and rock crabs. However, lobsters were restricted to
shallower depths; no settlers were found in collectors placed at 20 m. In contrast, newly
settled rock crabs were more abundant; almost twice the abundance of lobsters at 5 m
depth, and present at 20 m (although less dense than at 5 m). The possible mechanisms
involved in this shallow settlement pattern are not completely understood. Settlement
cues for megalopae of C. irroratus have been advanced, involving the presence of
hydrocarbons (Bigford, 1977) and light intensity (positive phototaxis for zoea stages
Bigford, 1979a). Compared to the zoea stages, brachyuran megalopae in general have a
more precise regulation and perception of depth. These could represent tactics that favor
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
123
Fig. 8. Tens of kilometers scale up-estuary pattern of abundance in Penobscot Bay. Mean number of newly
settled (YOY) and of older individuals (1YR1) (61 SD) from surveys using twelve 0.25 m 2 and 1.0 m 2
quadrats, respectively. The number of quadrats surveyed at each site is indicated by the numbers at the bottom.
For each species and site, horizontal lines represent comparison between age groups. Interrupted lines mean
significant differences.
dispersal by early stages and recruitment to favorable adult habitats by later instars
(Sulkin, 1984). A more homogeneous pattern of abundance among depths for older
individuals suggests a reduced habitat selection behavior and increasing propensity to
124
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Table 3
Penobscot Bay. Two-factor ANOVA to compare the abundance of different age groups (YOY vs. 1YR1) of
H. americanus and C. irroratus at nine different sites. Data are number of individuals within 0.25 m 2 (for
YOY) and 1.0 m 2 (for 1YR1) quadrats. Sample sizes as shown in Fig. 8
Source
df
SS
H. americanus
Site
Age group
Site3Age group
Error
8
1
8
817
44.47
30.81
9.54
670.18
C. irroratus
Site
Age group
Site3Age group
Error
8
1
8
817
855.37
184.82
689.33
961.14
MS
F-value
P-value
5.56
30.81
1.19
0.82
6.78
37.56
1.45
,0.001
,0.001
0.171
106.92
184.82
86.17
1.18
90.89
157.10
73.24
,0.001
,0.001
,0.001
Fig. 9. Hundreds of kilometers scale among-regions (separated by 20 to 230 km and arranged from west to
east) pattern of abundance along coastal Gulf of Maine. Abundance of (A,B,C) newly settled (YOY) (61 SD),
and (D,E,F) older individuals (1YR1) (61 SE). Values for YOY represent the average abundances in 0.25 m 2
quadrats: 12 in York region in 1997, 91 for H. americanus and 160 for C. irroratus and C. borealis between
1994 and 1997 in Pemaquid region, 82 in Penobscot Bay in 1997, and 35 in Mount Desert Island between
1995 and 1996. For 1YR1 the numbers of 1.0 m 2 quadrats were: 284 in York in 1995 and 1997, 994 in
Pemaquid between 1994 and 1997, 72 in Penobscot Bay in 1997, and 1143 in Mount Desert Island between
1994 and 1997. These data were pooled and the final design had a sample size of 8, 18, 9, and 32 for each
region, respectively. In (A) Mount Desert Island is assumed to have the smallest next variance. For each
species and age group, different letters above bars mean significant differences between regions.
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
125
Fig. 10. Weekly scale pattern of settlement at a single site (DIW) in the Pemaquid region. Settlement densities
of H. americanus and C. irroratus during the late-summer of 1994 on four standardized 1.0 m 2 natural cobble
plots (61 SE). The relationship between the numbers of lobsters and crabs is significant: r 2 5 0.515,
P 5 0.005.
move among older individuals of both species. The greater densities of older individuals
inside collectors, compared to adjacent natural substrata (Fig. 2B), was likely due to (1)
the tendency of relatively small-sized juveniles to colonize unoccupied and shelterproviding habitat, and (2) the fact that these observations were made only during one
settlement season (1995) and compared to the average densities from 4 years of data for
older individuals outside collectors, which exhibited greater variability.
Spatially and temporally the pattern of abundance of newly settled lobsters was
conservative among sites within a region, at 1 to 15 km apart. The average abundances
over a 4 year period at three sites with similar exposure were identical. The south-west
exposed shores that face prevailing summer winds tend to have the highest rates of
settlement (Wahle and Incze, 1997). Average abundance of older lobsters followed a
similar trend, probably reflecting previous similar settlement events. This is consistent
with the idea that lobster recruits are relatively sedentary for at least the first 2 years
after settlement to the benthos (Incze et al., 1999). The abundance of newly settled rock
crab was more variable than that of older individuals in the same sites, without a clear
connection between them. This suggests the linkage between settlement and subsequent
recruitment into older segments of the population is not tightly coupled at the intraregional scale. In similar annual censuses in the Mount Desert region with similarly
spaced sites, the general abundance of settlers (for either species) was low, to below
detectable levels in the case of lobsters. Although the abundance of older individuals for
both species was lower at Mount Desert than that in the Pemaquid region, it occurred at
measurable levels, and without a clear relationship to the abundance of settlers. This
pattern suggests that local populations of older individuals of both species could be the
126
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Fig. 11. Annual scale pattern of settlement at three sites (DIW, FIW, KRE) in the Pemaquid region.
Abundance of (A,B) newly settled (YOY) and of (C,D) older individuals (1YR1) (61 SE). Values for YOY
represent the combined average abundances in 0.25 m 2 quadrats surveyed at different numbers of sites between
1994 and 1997: 2, 7, 4, and 3 for each respective year. For 1YR1 the numbers of 1.0 m 2 quadrats were 3, 5, 5,
and 5, respectively. Different letters above bars mean significant differences between years.
result of immigration of rather large individuals into this region. This is indirectly
supported by the presence of significantly larger older individuals (lobsters and both crab
species) in Mount Desert compared to the Pemaquid region (A.P., personal observation).
Meso-scale patterns in settlement between lobsters and crabs contrast strongly. At the
spatial scale of hundreds of meters, the post-settlement abundance pattern for lobsters
can be different among nearby sites having different larval delivery, as on the east and
west sides of Damariscove Island (Wahle and Incze, 1997). These authors showed that
the observed pattern for lobsters related to prevailing summer wind conditions,
concentrated postlarvae on the west side of the island. Other research in the same
location (Palma et al., 1998) suggested that the abundance of newly settled rock crabs
could also be affected by these physical factors. However, the similar densities recorded
for both sides of the island under open and predator-exclusion conditions are not
consistent with the wind-driven hypothesis (Fig. 6). Instead, local differences in
abundance of newly settled rock crabs may be the result of strong post-settlement
mortality. Clancy and Cobb (1997) found that rock crab megalopae were very patchy at
the hundreds of meters scale. This discrepancy might lie in the fact that these authors
quantified the abundance of megalopae before settlement occurred, whereas in this study
newly settled post-megalopae were quantified. The latter approach considers important
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
127
post-settlement events that might overwhelm signals obtained only by measuring the
abundance of individuals before they settle. On the other hand, for each species the
abundance of older individuals on either side of the island was similar, without
necessarily showing a clear link with new settlers. This suggests that, at this spatial
scale, the abundance of older rock crabs is likely to be the result of events occurring
after settlement such as post-settlement mortality or migration. Evidence of seasonal
short- and long-distance movements have been described for lobsters and rock crabs
(Krouse, 1980, 1981; Ennis, 1984; Pezzack and Duggan, 1986; Rebach, 1987).
The abundances of newly settled lobsters and rock crabs decreased up the estuary. An
interpretation of this pattern should consider the increasing distance from the potential
off-shore larval source together with factors such as larval mobility and physiological
tolerance. On the other hand, barnacles, also with planktotrophic larvae, show significant
recruitment densities even in the upper-most sections of the same river (Leonard et al.,
1998), suggesting that other crustaceans can have local settlement. The influence of
onshore wind-driven currents as well as tidal amplitude has been shown to have an
important influence on the entry of larvae of other decapods into estuaries (Eggleston
and Armstrong, 1995; Mense et al., 1995; Morgan et al., 1996). Consequently, the
narrowness and almost straight south–north orientation of the Damariscotta River may
represent a barrier affecting the supply of rock crab and lobster larvae. As with the
depth-related experiment, the use of collectors in this experiment permitted a more
objective measure of the settlement pattern by standardizing the substratum, especially
since most of the river bed is composed of soft bottom. Lobsters clearly settled only in
the southern-most sites, close to the mouth of Penobscot Bay, whereas older individuals
were distributed throughout with similar densities, despite some local differences. This
coincides with the higher abundances of newly settled rock crab near the mouth, which
gradually disappear further up the bay. Older rock crabs were also distributed more
evenly throughout the bay. Hence, and despite the differences in size and shape of these
two geographic features, the pattern of up-estuary and up-bay decrease in settlement
densities of the two species was comparable.
The transport mechanism for larvae of these species in estuaries and bays is unknown
in coastal Maine, however Wahle (1993) described a similar pattern of abundance of
newly settled and older H. americanus in estuarine systems in Narragansett Bay. The
patterns we have found, along with several other studies on different species (e.g.,
Eggleston and Armstrong, 1995; Mense et al., 1995; Morgan et al., 1996), have shown
how different physical factors such as wind-driven currents and tidal amplitude (Mense
et al., 1995) can influence settlement rates in estuarine brachyuran postlarvae. Particularly for blue crabs, Callinectes sapidus, passive wind-driven delivery of larvae and
habitat preferences determine geographical settlement patterns, and some megalopae
even settle as far as the head of an estuary in the northern Gulf of Mexico (Morgan et
al., 1996). The re-invasion of blue crab megalopae along Mid-Atlantic States is effected
by fall southward wind events (Epifanio, 1995). The shoreward advection also depends
on the vertical position of larvae in the water column (Blanton et al., 1995). The
presence of older individuals, particularly of C. irroratus, further up the estuary and
throughout the Bay could be directly related to broader in situ settlement, but also to the
tendency of older individuals to migrate into these environments. The latter has been
128
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
shown in Dungeness crabs, Cancer magister (Smith and Jamieson, 1991), and green
crabs, Carcinus meanas (Abello´ et al., 1997), under similarly protected conditions.
4.2. Spatial large-scale patterns
At the largest, inter-regional, spatial scale considered in this study, where regions are
25 to 235 km apart, the abundance of newly settled individuals of all species was highest
in the two western-most regions, or only farthest west for Jonah crabs. This pattern is
only proportional to the abundance of older individuals for lobsters in all regions, except
in Penobscot Bay. The low abundance of older rock crabs had no relationship with the
pattern of settlement abundance at this scale; the same is true for Jonah crabs, since in
three of the four regions settlement for this species was not detected.
The correlation between the abundance of settlers and that of older individuals was in
general much stronger for lobsters than for any of the crab species at most of the scales
considered in this study. In studies at broader geographic scales (i.e., the Gulf of Maine),
offshore sources of lobster larvae may largely contribute to coastal patterns of
recruitment (Harding and Trites, 1988; Steneck and Wilson, 1998). For lobsters, subsidy
from nearby offshore sources have been suggested as the main mechanism for larval
supply, as opposed to surface currents and wind-induced transport (Katz et al., 1994). In
an intra-regional scale study in Rhode Island, Clancy and Cobb (1997) described how
wind- and tidal-driven directional transport appeared to be the mechanism of larval
delivery for C. irroratus. In their study, a significant relationship between transport
direction and collection dates was detected for megalopae. Megalopae collected at fine
(meters) scales were found to be similarly distributed, compared to those collected at
broader (hundreds of meters) scale.
As indicated by Caley et al. (1996), the majority of marine populations are
demographically open; therefore, understanding the dynamics of larval delivery will help
understand the dynamics of local populations and harvested stocks. Our observations
agree with this view; however, we emphasize the importance of perspectives that include
other potential limiting phases in the life history of organisms. Finally, these observations should be sensitive to the wide spectrum of spatial and temporal scales involved.
4.3. Temporal scale-related patterns
At the different time scales considered here, the variation in abundance of lobsters and
rock crabs was not always proportional. Similar settlement pulses for the two species in
one location during one summer may reflect similar oceanographic conditions to which
post-larvae of both species are exposed (Fig. 10). Although this was not a direct
measurement of larvae present in the water column, newly settled individuals were
quantified within a week after reaching the bottom, and possible negative effects (such
as substrate quality, or presence of already-settled individuals on the bottom) were
reduced by periodically surveying the same standardized 1.0 m 2 cobble plots. When the
abundance of newly settled individuals of the two species was measured annually over a
4 year period, the temporal pattern was different. The abundance of newly settled
lobsters did not differ among years, while that of rock crabs did (Fig. 11). Variable
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
129
inter-annual abundance of older individuals may not necessarily reflect past settlement
events, since the estimations of abundance took more than one single cohort into
account. Therefore, it is likely that more than one single settlement season was
quantified.
4.4. Multiscale patterns
Although a multiscale view was taken, the whole distribution range or all possible
habitats for these species and their different life stages were not explored. No other study
describes the combined patterns of distribution and abundance of these sympatric
decapods. Most studies are largely restricted to a single area (within a region) (Ojeda and
Dearborn, 1989, 1990; Palma et al., 1998) or discuss behavioral interactions (Richards et
al., 1983; Hudon and Lamarche, 1989; Richards, 1992; Miller and Addison, 1995). Most
ecological studies, however, have been focused on the American lobster (for a review,
see Elner and Campbell, 1991), describing patterns at a local (intra-regional) scale (but
see Harding and Trites, 1988; Pezzack, 1989), emphasizing processes of post-larvae
habitat selection (Barshaw and Rich, 1997), the importance of appropriate nursery
grounds for post-larvae (Boudreau et al., 1990; Wahle and Steneck, 1991), the effect of
physical factors affecting recruitment (Aiken and Waddy, 1986; Katz et al., 1994; Wahle
and Incze, 1997), or the effect of predation on early benthic stages (Wahle, 1992; Wahle
and Steneck, 1992; Barshaw et al., 1994; Spanier et al., 1998). In comparison, our
knowledge of essential ecological factors that affect populations of Cancer irroratus and
C. borealis is less complete, focusing on general biological (Winget et al., 1972; Sastry,
1977a,b; Reilly and Saila, 1978; Bigford, 1979b) or behavioral aspects of older
individuals (Rebach, 1987; Richards, 1992).
A graphic representation (Fig. 12) describes how the above estimations of abundance
vary at different scales of observation. The percentage variability corresponds to the
quotient between the difference, and the sum, of adjacent measurements of abundance
(e.g., adjacent depths, neighboring sites within a region). These values were then
averaged for each of the spatial variables considered (e.g., substrata, estuary). This was
done separately for newly settled and for older individuals. Overall, variability was
greater at most scales considered near the time of settlement than for older individuals.
Lobsters showed the greatest variability at the smallest scale, decreasing at intermediate
scales, and increasing again at the largest scales. The pattern of variability for older
individuals was similar and low for the two species at all scales. However, lobsters are
consistently more variable than crabs (except at the largest scale). Several different
processes are proposed that could account for the observed pattern at each scale. For
newly settled individuals these processes are: differential substratum- and depth-related
habitat selection at the smallest scales, differential site-specific settlement success or
differential mortality at the intra-regional scale, and up-estuary and longshore transport
processes for estuarine and inter-regional, respectively. The equally low variability
recorded for older individuals of both species, particularly at intermediate spatial scales,
suggests that processes operating at or near the time of settlement continue to dominate
demographic patterns throughout older stages. However, the variability measured for
both species at finest and broadest spatial scales could be attributed to increasing
130
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Fig. 12. Graphic representation of the spatial scale-dependent (log-transformed) variability in settlement and
post-settlement patterns for H. americanus and C. irroratus. The percent variability corresponds to the quotient
between the difference and the sum of adjacent measurements of abundance (see Conclusions). Potentially
important processes are shown for each figure within the frames. (*) Correspond to data taken from Palma et
al. (1998) for opposite sides of Damariscove Island.
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
131
size-related mobility within a given habitat at the local scales (substratum and depth).
Such a local scale pattern was shown by Ennis (1984) where bathymetric migration in
American lobsters occurred in relation to changes in the seasonal thermocline. Seasonal
migrations of rock crabs have been attributed to unidentified exogenous cues (Rebach,
1987).
In order to determine how much of the variation in abundance belongs to each spatial
scale, the data of abundance for the different life stages of H. americanus and C.
irroratus were analyzed at different spatial scales (Table 4). The greatest variability for
newly settled lobsters occurred at the largest spatial scale (regional) while the opposite
was true for older individuals, where significant variability was only observed among
quadrats. Rock crabs, on the other hand, showed no significant differences in the
variance of their abundance at any of the spatial scales considered here. Although with a
slight significance for older lobsters at the smallest scale, the percentage contribution of
variation for older individuals was greater at the quadrat and site scales. These additional
results corroborate the previously discussed results, where different processes can be
invoked for the observed patterns of these two species at different life stages.
Several studies in marine systems have shown the existence of scale-dependent
patterns and associated processes for a variety of organisms on a wide range of scales.
For example, inter-regional comparisons of variability in density estimates and morphometric relationships in subtidal kelp stands (Camus and Ojeda, 1992), fine-scale
gregarious settlement patterns mediated by larval behavior (Hills and Thomason, 1996)
Table 4
Results of the nested analysis of variance for the abundance of different life stages (YOY and 1YR1) of
lobsters and crabs. These correspond to a subset of data from two regions (PEM and MDI), three sites within
each region and six quadrats within each site only on cobble substrata surveyed at the same depth (10 m
BMLW). (a) Significance of F-ratios and (b) percentage contribution to the variance of several spatial scales
from that of replicate quadrats to sites in regions hundreds of kilometers apart (procedure after Underwood and
Chapman, 1996). ns P . 0.05, *P , 0.05, ***P , 0. 001
Species
YOY
(a) Significance of F-ratios
H. americanus
C. irroratus
(b) Percentage contribution to the variance
H. americanus
C. irroratus
1 YR 1
(a) Significance of F-ratios
H. americanus
C. irroratus
(b) Percentage contribution to the variance
H. americanus
C. irroratus
Source of variation
Quadrat
(cm to m)
Sites
(100s m to 10s km)
Regions
(100s km)
ns
ns
ns
ns
***
ns
0
11
1
13
99
76
*
ns
ns
ns
ns
ns
62
30
38
54
0
16
132
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
or multiscale patterns resulting from a combination of larval supply and behavior in
barnacles (Raimondi, 1990), different post-settlement use of space by older stages of
demersal fish (Levin, 1994a,b), spatial- and temporal-scale dependent variability in
larval supply of tropical shorefish (Thorrold et al., 1994), and kilometer-scale densityindependent recruitment patterns in bivalves (David et al., 1997). McConaugha (1992)
points to different spatial and temporal scales where diverse biological and physical
processes affect decapods, particularly at their larval stages.
The differences described here, between the two most common large benthic decapod
species in coastal Maine, suggest that even though rock crabs are more capable of
settling at higher densities under more variable conditions (i.e., substratum, depth,
distance up-estuaries), high post-settlement losses are responsible for significant decreases in numbers, ultimately yielding similar adult population densities to those of
lobsters. The wide scale ranges at which rock crabs settle and the important postsettlement losses that they undergo (Clancy, 1995; Palma et al., 1998) agree with
predictions from the general life history strategy differences between lobsters and crabs
(Cobb et al., 1997), that is, rock crabs mature earlier and are more fecund (Bigford,
1979a,b) than lobsters (Krouse, 1973; Waddy et al., 1995).
In our study we emphasize how the abundance of these species shows a scaledependent variability for different ontogenetic stages. Furthermore, we propose several
processes that may explain this variability, along with the acceptance of the hypothesis
that very small-scale variability can result from organismic behaviors, in contrast to
large-scale variability that can result from oceanographic conditions.
Especially for the American lobster, early life-history characteristics that may affect
the number of recruits, such as fecundity and duration of the larval period, could
ultimately affect their population dynamics (Cobb and Wahle, 1994; Cobb et al., 1997).
Furthermore, for commercially valuable species like this, knowledge about the link
between settlement and older (harvestable) stages of the population, as well as the scales
at which processes operate, is of the utmost importance. The necessity of integrating the
different scales that operate throughout the fishery into management schemes has, for
example, been emphasized for groundfish in the western north Atlantic (Langton et al.,
1995). Moreover, having a better understanding of the population dynamics of other
similar, less exploited, species should help in considering what common and unique
processes, at what scales, are particularly important to consider for management or
conservation.
Acknowledgements
We want to thank the enthusiastic field assistance of H. Peckham, L. Van Bergen, K.
Dearborn, C. Rigaud, B. McMillan, and E. Eisenhardt. We also want to thank P. White
(Maine Lobstermen Association) for support in the York region and D. Cousins in the
Thomaston area, and The Island Institute for support with the Penobscot Bay project.
This manuscript was greatly improved by the helpful comments of M. Fogarty, M.
˜ and
Bertness, R. Wahle, L. Watling, L. Mayer, S. Zimsen, A. Underwood, J.M. Farina,
two anonymous reviewers. This research was also supported by NOAA / NURC grant
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
133
UCAZP 94-121 and U. Maine / U. New Hampshire Sea Grant Program to R.S.S. A.T.P.
acknowledges support by the Association of Graduate Students at the University of
Maine. Thanks to the Darling Center for great facilities and working atmosphere. This is
contribution No. 331 to the Darling Marine Center.
References
´ P., Aagaard, A., Warman, C.G., Depledge, M.H., 1997. Spatial variability in the population structure of
Abello,
the shore crab Carcinus meanas (Crustacea: Brachyura) in a shallow-water, weakly tidal fjord. Mar. Ecol.
Prog. Ser. 147, 97–103.
Aiken, D.E., Waddy, S.L., 1986. Environmental influence on recruitment of the American lobster, Homarus
americanus: a perspective. Can. J. Fish. Aquat. Sci. 43, 2258–2270.
Barshaw, D.E., Able, K.W., Heck, K.L., 1994. Salt marsh peat reefs as protection for postlarval lobsters
Homarus americanus from fish and crab predators: comparison with other substrates. Mar. Ecol. Prog. Ser.
106, 203–206.
Barshaw, D.E., Rich, D.R., 1997. An analysis of substrate selection by postlarval American lobsters, Homarus
americanus, using a dynamic optimization model. Oikos 80, 554–564.
Bigford, T.E., 1977. Effects of oil on behavioral responses to light, pressure, and gravity in larvae of the rock
crab Cancer irroratus. Mar. Biol. 43, 133–148.
Bigford, T.E., 1979a. Ontogeny of light and gravity responses in rock crab larvae (Cancer irroratus). Mar.
Biol. 52, 69–76.
Bigford, T.E., 1979b. Synopsis of biological data on the rock crab, Cancer irroratus Say. NOAA Technical
Report NMFS Circular 426, 26 pp.
Blanton, J., Wenner, E., Werner, F., Knott, D., 1995. Effects of wind-generated coastal currents on the transport
of blue crab megalopae on a shallow continental shelf. Bull. Mar. Sci. 57, 739–752.
Boudreau, B., Bourget, E., Simard, Y., 1990. Benthic invertebrate larval response to substrate characteristics at
settlement: shelter preferences of the American lobster Homarus americanus. Mar. Biol. 106, 191–198.
Boudreau, B., Simard, Y., Bourget, E., 1992. Influence of a thermocline on vertical distribution and settlement
of post-larvae of the American lobster Homarus americanus Milne-Edwards. J. Exp. Mar. Biol. Ecol. 162,
35–49.
Caley, M.J., Carr, M.H., Hixon, M.A., Hughes, T.P., Jones, G.P., Menge, B.A., 1996. Recruitment and the local
dynamics of open marine populations. Annu. Rev. Ecol. Syst. 27, 477–500.
Camus, P.A., Ojeda, F.P., 1992. Scale-dependent variability of density estimates and morphometric relationships in subtidal stands of the kelp Lessonia trabeculata in northern and central Chile. Mar. Ecol. Prog. Ser.
90, 193–200.
Clancy, M., Cobb, J.S., 1997. Effects of wind and tidal advection on distribution patterns of rock crab Cancer
irroratus megalopae in Block Island Sound, Rhode Island. Mar. Ecol. Prog. Ser. 152, 217–225.
Clancy, M., 1995. Recruitment of the rock crab, Cancer irroratus. Ph.D. Thesis, University of Rhode Island,
Providence, RI, USA.
Cobb, J.S., Wahle, R.A., 1994. Early life history and recruitment processes of clawed lobsters. Crustaceana 67,
1–25.
Cobb, S.J., Booth, J.D., Clancy, M., 1997. Recruitment strategies in lobsters and crabs: a comparison. Mar.
Freshwater Res. 48, 797–806.
Connell, J.H., 1985. The consequences of variation in initial settlement vs. post-settlement mortality in rocky
intertidal communities. J. Exp. Mar. Biol. Ecol. 93, 11–45.
Connolly, S.R., Roughgarden, J., 1998. A latitudinal gradient in Northeast Pacific intertidal community
structure: evidence for an oceanographically based synthesis of marine community theory. Am. Nat. 151,
311–326.
David, P., Berthou, P., Noel, P., Jarne, P., 1997. Patchy recruitment patterns in marine invertebrates: a spatial
test of the density-dependent hypothesis in the bivalve Spisula ovalis. Oecologia 111, 331–340.
134
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Doherty, P., Fowler, T., 1994. An empirical test of recruitment limitation in a coral reef fish. Science 263,
935–939.
Eggleston, D.B., 1995. Recruitment in Nassau grouper Epinephelus striatus: post-settlement abundance,
microhabitat features, and ontogenetic habitat shifts. Mar. Ecol. Prog. Ser. 124, 9–22.
Eggleston, D.B., Armstrong, D.A., 1995. Pre- and post-settlement determinants of estuarine Dungeness crab
recruitment. Ecol. Monogr. 65, 193–216.
Elner, R.W., Campbell, A., 1991. Spatial and temporal patterns in recruitment for American lobster, Homarus
americanus, in the northwestern Atlantic. Mem. Queensland Museum 31, 349–363.
Ennis, G.P., 1984. Small-scale seasonal movements of the American lobster Homarus americanus. Trans. Am.
Fish. Soc. 113, 336–338.
Epifanio, C.E., 1995. Transport of blue crab (Callinectes sapidus) larvae in the waters off Mid-Atlantic States.
Bull. Mar. Sci. 57, 713–725.
Fernandez, M., Iribarne, O., Armstrong, D., 1993. Habitat selection by young-of-the-year Dungeness crab
Cancer magister and predation risk in intertidal habitats. Mar. Ecol. Prog. Ser. 92, 171–177.
Fogarty, M.J., 1995. Population, fisheries, and management. In: Factor, J.R. (Ed.), Biology of the Lobster
Homarus americanus, Academic Press, San Diego, pp. 111–137.
Gaines, S., Roughgarden, J., 1985. Larval settlement rate: a leading determinant of structure in an ecological
community of the marine intertidal zone. Proc. Natl. Acad. Sci. USA 82, 3707–3711.
Gotceitas, V., Brown, J.A., 1993. Substrate selection by juvenile Atlantic cod (Gadus morhua): effect of
predation risk. Oecologia 93, 31–37.
Harding, G.C., Trites, R.W., 1988. Dispersal of Homarus americanus larvae in the Gulf of Maine from Browns
Bank. Can. J. Fish. Aquat. Sci. 45, 416–425.
Hills, J.M., Thomason, J.C., 1996. A multi-scale analysis of settlement density and pattern dynamics of the
barnacle Semibalanus balanoides. Mar. Ecol. Prog. Ser. 138, 103–115.
Horne, J.K., Schneider, D.C., 1994. Analysis of scale-dependent processes with dimensionless ratios. Oikos 70,
201–211.
Hudon, C., Lamarche, G., 1989. Niche segregation between American lobster Homarus americanus and rock
crab Cancer irroratus. Mar. Ecol. Prog. Ser. 52, 155–168.
Incze, L.S., Wahle, R.A., 1991. Recruitment from pelagic to early benthic phase in lobsters Homarus
americanus. Mar. Ecol. Prog. Ser. 79, 77–87.
Incz, L.S., Wahle, R.A., Palma, A.T., 1999. Advection and settlement rates in a benthic invertebrate:
recruitment to first benthic stage in Homarus americanus. ICES J. Mar. Sci. (in press).
Jones, G.P., 1991. Post-recruitment processes in the ecology of coral reef fish populations: a multifactorial
perspective. In: Sale, P.F. (Ed.), The Ecology of Fishes On Coral Reefs, Academic Press, San Diego, pp.
294–328.
Jones, G.P., 1997. Relationships between recruitment and postrecruitment processes in lagoonal populations of
two coral reef fishes. J. Exp. Mar. Biol. Ecol. 213, 231–246.
Katz, C.H., Cobb, J.S., Spaulding, M., 1994. Larval behavior, hydrodynamic transport, and potential offshore
recruitment in the American lobster, Homarus americanus. Mar. Ecol. Prog. Ser. 103, 265–273.
Krouse, J.S., 1973. Maturity, sex ratio, and size composition of the natural population of the American lobster,
Homarus americanus, along the Maine coast. Fish. Bull. 71, 165–173.
Krouse, J.S., 1980. Summary of lobster, Homarus americanus, tagging studies in American waters (1898–
1978). Can. Tech. Rep. Fish. Aquat. Sci. 932, 135–140.
Krouse, J.S., 1981. Movement, growth and mortality of American lobsters, Homarus americanus, tagged along
the coast of Maine. U.S. Department of Commerce, NOAA Tech. Rep. NMFS SSRF-747, 12 pp.
Krouse, J.S., Cowger, J., 1980. The Maine Cancer crab fishery. Maine Department of Marine Resources
Lobster Informational Leaflet [8, 18 pp.
Langton, R.W., Auster, P.J., Schneider, D.C., 1995. A spatial and temporal perspective on research and
management of groundfish in the Northwest Atlantic. Rev. Fish. Sci. 3, 201–229.
Lawton, P., Lavalli, K.L., 1995. Postlarval, juvenile, adolescent, and adult ecology. In: Factor, J.R. (Ed.),
Biology of the Lobster Homarus americanus, Academic Press, San Diego, pp. 47–88.
Leonard, G.H., Levine, J.M., Schmidt, P.R., Bertness, M.D., 1998. Flow-driven variation in intertidal
community structure in a Maine estuary. Ecology 79, 1396–1411.
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
135
Levin, P.S., 1994a. Fine-scale temporal variation in recruitment of a temperate demersal fish: the importance of
settlement versus post-settlement loss. Oecologia 97, 124–133.
Levin, P.S., 1994b. Small-scale recruitment variation in a temperate fish: the roles of macrophytes and food
supply. Environ. Biol. Fish. 40, 271–281.
Levin, S.A., 1992. The problem of pattern and scale in ecology. Ecology 73, 1943–1967.
Levin, S.A., 1993. Concepts of scale at the local level. In: Ehleringer, J.R., Field, C.B. (Eds.), Scaling
Physiological Processes: Leaf to Globe, Academic Press, San Diego, pp. 7–19.
Lough, R.G., Valentine, P.C., Potter, D.C., Auditore, P.J., Boltz, G.R., Neilson, J.D., Perry, R.I., 1989. Ecology
and distribution of juvenile cod and haddock in relation to sediment type and bottom current on eastern
Georges Bank. Mar. Ecol. Prog. Ser. 56, 1–12.
Mayer, L.M., Townsend, D.W., Pettigrew, N.R., Loder, T.C., Wong, M.W., Kistner, D., Laursen, A.K.,
Schoudel, A.D., Conairis, C., Brown, J., Newell, C., 1996. The Kennebec, Sheepscot and Damariscotta
River etuaries: seasonal oceanographic data. University of Maine Tech. Rep. No. 9601, 110 pp.
McAlice, B.J., 1977. A preliminary oceanographic survey of the Damariscotta River Estuary, Lincoln County,
Maine. Maine Sea Grant Tech. Rep. No. 43.
McConaugha, J.R., 1992. Decapod larvae: dispersal, mortality, and ecology. A working hypothesis. Am. Zool.
32, 512–523.
Mense, D.J., Posey, M.H., West, T., Kincheloe, K., 1995. Settlement of brachyuran postlarvae along the North
Carolina coast. Bull. Mar. Sci. 57, 793–806.
Miller, R.J., Addison, J.T., 1995. Trapping interactions of crabs and American lobster in laboratory tanks. Can.
J. Fish. Aquat. Sci. 52, 315–324.
Morgan, S.G., Zimmer-Faust, R.K., Heck, Jr. K.L., Coen, L.D., 1996. Population regulation of blue crabs
Callinectes sapidus in the northern Gulf of Mexico: postlarval supply. Mar. Ecol. Prog. Ser. 133, 73–88.
Ojeda, F.P., Dearborn, J.H., 1989. Community structure of macroinvertebrates inhabiting the rocky subtidal
zone in the Gulf of Maine: seasonal and bathymetric distribution. Mar. Ecol. Prog. Ser. 57, 147–161.
Ojeda, F.P., Dearborn, J.H., 1990. Diversity, abundance, and spatial distribution of fishes and crustaceans in the
rocky subtidal zone of the Gulf of Maine. Fish. Bull. 88, 403–410.
Palma, A.T., Wahle, R.A., Steneck, R.S., 1998. Different early post-settlement strategies between American
lobsters Homarus americanus and rock crabs Cancer irroratus in the Gulf of Maine. Mar. Ecol. Prog. Ser.
162, 215–225.
Pezzack, D.S., Duggan, D.R., 1986. Evidence of migration and homing of lobsters (Homarus americanus) on
the Scotian Shelf. Can. J. Fish. Aquat. Sci. 43, 2206–2211.
Pezzack, D.S., 1989. Comments on ‘Dispersal of Homarus americanus larvae in the Gulf of Maine from
Browns Bank’ by G.C. Harding and R.W. Trites. Can. J. Fish. Aquat. Sci. 46, 1077–1078.
Raimondi, P.T., 1990. Patterns, mechanisms, consequences of variability in settlement and recruitment of an
intertidal barnacle. Ecol. Monogr. 60, 283–309.
Rebach, S., 1987. Entrainment of seasonal and nonseasonal rhythms by the rock crab Cancer irroratus. J.
Crustacean Biol. 7, 581–594.
Reilly, P.N., Saila, S.B., 1978. Biology and ecology of the rock crab, Cancer irroratus Say, 1817, in southern
New England waters (Decapoda, Brachyura). Crustaceana 34, 121–140.
Richards, R.A., 1992. Habitat selection and predator avoidance: ontogenetic shift in habitat use by the Jonah
crab Cancer borealis (Stimpson). J. Exp. Mar. Biol. Ecol. 156, 187–197.
Richards, R.A., Cobb, J.S., Fogarty, M.J., 1983. Effects of behavioral interactions on the catchability of
American lobster, Homarus americanus and two species of Cancer crab. Fish. Bull. 81, 51–60.
Roughgarden, J., Gaines, S., Possingham, H., 1988. Recruitment dynamics in complex life cycles. Science 241,
1460–1466.
Roughgarden, J., Running, S.W., Matson, P.A., 1991. What does remote sensing do for ecology? Ecology 72,
1918–1922.
Sastry, A.N., 1977a. The larval development of the rock crab, Cancer irroratus Stimpson, 1859, under
laboratory conditions (Decapoda Brachyura). Crustaceana 32, 155–168.
Sastry, A.N., 1977b. The larval development of the Jonah crab, Cancer borealis Stimpson, 1859, under
laboratory conditions (Decapoda Brachyura). Crustaceana 32, 290–303.
Smith, B.D., Jamieson, G.S., 1991. Movement, spatial distribution, and mortality of male and female
Dungeness crab Cancer magister near Tofino, British Columbia. Fish. Bull. 89, 137–148.
136
A.T. Palma et al. / J. Exp. Mar. Biol. Ecol. 241 (1999) 107 – 136
Sokal, R.R., Rohlf, F.J. (Eds.), 1981. Biometry: the Principles and Practice of Statistics in Biological Research,
2nd ed., W.H. Freeman, San Francisco, pp. 859.
Spanier, E., McKenzie, T.P., Cobb, J.S., Clancy, M., 1998. Behavior of juvenile American lobsters, Homarus
americanus, under predation risk. Mar. Biol. 130, 397–406.
Steger, R., 1987. Effects of refuges and recruitment on gonodactylid stomatopods, a guild of mobile prey.
Ecology 68, 1520–1533.
Steneck, R.S., Wilson, C., 1998. Why are there so many lobsters in Penobscot Bay? In: Platt, D.D. (Ed.), Rim
of the Gulf: Restoring Estuaries in the Gulf of Maine, Island Institute, Rockland Maine, pp. 68–71.
Sulkin, S.D., 1984. Behavioral basis of depth regulation in the larvae of brachyuran crabs. Mar. Ecol. Prog.
Ser. 15, 181–205.
Thorrold, S.R., Shenker, J.M., Wishinski, E., Mojica, R., Maddox, E.D., 1994. Larval supply of shorefishes to
nursery habitats around Lee Stocking Island, Bahamas. I. Small-scale distribution patterns. Mar. Biol. 118,
555–566.
Underwood, A.J., Fairweather, P.G., 1989. Supply-side ecology and benthic marine assemblages. Trends Ecol.
Evol. 4, 16–20.
Underwood, A.J., Chapman, M.G., 1996. Scales of spatial patterns of distribution of intertidal invertebrates.
Oecologia 107, 212–224.
Underwood, A.J., 1997. Experiments in Ecology: Their Logical Design and Interpretation Using Analysis of
Variance, 1st ed., Cambridge University Press.
Waddy, S.L., Aiken, D.E., De Kleijn, D.P.V., 1995. Control of growth and reproduction. In: Factor, J.R. (Ed.),
Biology of the Lobster Homarus americanus, Academis Press, San Diego, pp. 217–266.
Wahle, R.A., 1992. Body-size dependent anti-predator mechanisms of the American lobster. Oikos 65, 52–60.
Wahle, R.A., 1993. Recruitment to American lobster populations along an estuarine gradient. Estuaries 16,
731–738.
Wahle, R.A., Incze, L.S., 1997. Pre- and post-settlement processes in recruitment of the American lobster. J.
Exp. Mar. Biol. Ecol. 217, 179–207.
Wahle, R.A., Steneck, R.S., 1991. Recruitment habitats and nursery grounds of the American lobster Homarus
americanus: a demographic bottleneck? Mar. Ecol. Prog. Ser. 69, 231–243.
Wahle, R.A., Steneck, R.S., 1992. Habitat restrictions in early benthic life: experiments on habitat selection
and in situ predation with the American lobster. J. Exp. Mar. Biol. Ecol. 157, 91–114.
Williams, A.B., Wahle, R.A., 1992. Distinguishing juvenile stages of Jonah and Atlantic rock crabs, Cancer
borealis and C. irroratus (Decapoda: Cancridae). J. Crustacean Biol. 12, 464–466.
Winget, R.R., Maurer, D., Seymour, H., 1972. Occurrence, size composition and sex ratio of the rock crab,
Cancer irroratus Say and the spider crab, Libinia emarginata Leach in Delaware Bay. J. Nat. Hist. 8,
199–205.
Zar, J.H., 1974. Biostatistical Analysis, Prentice-Hall, Englewoods Cliffs, NJ, pp. 718.