Louisiana State University
LSU Digital Commons
LSU Master's Theses
Graduate School
2016
Investigating Species and Population Level
Foraging Variation and Individual Specialization in
Pygoscelis Penguins Using Stable Isotope Analysis
Rachael W. Herman
Louisiana State University and Agricultural and Mechanical College
Follow this and additional works at: http://digitalcommons.lsu.edu/gradschool_theses
Recommended Citation
Herman, Rachael W., "Investigating Species and Population Level Foraging Variation and Individual Specialization in Pygoscelis
Penguins Using Stable Isotope Analysis" (2016). LSU Master's Theses. 315.
http://digitalcommons.lsu.edu/gradschool_theses/315
This Thesis is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU
Master's Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected].
INVESTIGATING SPECIES AND POPULATION LEVEL FORAGING
VARIATION AND INDIVIDUAL SPECIALIZATION IN PYGOSCELIS
PENGUINS USING STABLE ISOTOPE ANALYSIS
A Thesis
Submitted to the Graduate Faculty of the
Louisiana State University and
Agricultural and Mechanical College
in partial fulfillment of the
requirements for the degree of
Masters of Science
in
The College of Oceanography and Coastal Sciences
by
Rachael W. Herman
B.S., University of Massachusetts, 2009
August 2016
ACKNOWLEDGEMENTS
First and foremost, I would like to thank my academic advisor, Dr. Michael Polito, for
the incredible opportunity of not only being his first graduate student here at LSU, but also
bringing me all the way across the Drake Passage to Antarctica to conduct research, something I
thought would remain a pipe dream. He has provided me with tremendous guidance and support,
while pushing me to become an independent researcher. I would also like to thank my committee
members, Dr. Kenneth Rose and Dr. Kanchan Maiti, who have provided feedback and guidance
on my project and thesis over the past two years.
I am grateful to all of my friends at LSU and everyone else who allowed me to lean on
them when times were stressful. I want to especially thank Michael Harvey for letting me bounce
ideas off of him, and for providing invaluable feedback, support, and companionship while I
composed my thesis. I would like to acknowledge everyone who helped collect samples for me
in Antarctica, including Dr. Tom Hart, Jonathon Handley, and Jefferson Hinke. I would also like
to thank Quark Expedition crew for generously provided us with transportation to the Antarctic.
I want to thank my family, especially my parents Mary Alice and Franklin, for all of their
love and support over the years while I traveled throughout the world, disappearing to remote
islands with little to no communication, during which I know they were silently worried sick
about me.
Finally, I would like to thank the LSU’s Department of Oceanography and Coastal
Sciences for providing me with a graduate research assistantship for this study.
ii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS……………………………………………………………………....ii
LIST OF TABLES…………………………………………………..……………………………iv
LIST OF FIGURES…..…………………………….……………….……………………………vi
ABSTRACT……..………………………………………………………..……………………..vii
CHAPTER 1: GENERAL INTRODUCTION……….…………………………………………...1
LITERATURE CITED……………………………….……..…………………………….6
CHAPTER 2: SPECIES AND POPULATION LEVEL DIFFERENCES IN SEASONAL
CONSISTENCY AND INDIVIDUAL VARIATION OF PYGOSCELIS PENGUIN
FORAGING STRATEGIES……………………………….……….……………………………11
INTRODUCTION…...……………………………………..……………………………11
METHODS……...…...……………………………………..……………………………13
RESULTS..……...…...……………………………………..……………………………19
DISCUSSION…...…...……………………………………..……………………………24
LITERATURE CITED……………………………………..……………………………32
CHAPTER 3: GEOGRAPHIC VARIATION IN THE DEGREE OF INDIVIDUAL
SPECIALIZATION IN GENTOO PENGUIN (PYGOSCELIS PAPUA) POPULATIONS
THROUGHOUT THE SCOTIA ARC.….…………………………………………...………….37
INTRODUCTION…...……………………………………..……………………………37
METHODS……...…...……………………………………..……………………………40
RESULTS..……...…...……………………………………..……………………………46
DISCUSSION…...…...……………………………………..……………………………51
LITERATURE CITED……………………………………..……………………………57
CHAPTER 4: GENERAL SUMMARY AND CONCLUSIONS…….…………………………64
LITERATURE CITED……………………………………..……………………………67
APPENDIX: ESTIMATING THE TIMING OF TAIL FEATHER GROWTH..………….……69
VITA……………………………………………………………………………………………..73
iii
LIST OF TABLES
Table 2.1: Foraging “types” and interpretations based on four possible combinations of the
paired t-test/Wilcoxon signed rank test and Pearson correlation/Kendall rank correlation
results for TL and prey δ13C...…………………….…………….……………………….18
Table 2.2: Mean and standard deviation, range (in parentheses), and coefficient of variation
(CV%) of TL, δ15N, and prey δ13C of non-breeding and breeding seasons for all three
species at King George Island and three Gentoo populations at Wiencke, Rongé and
Elephant Island….…………………….…………….…….….…………………………..20
Table 2.3: Results for paired t-test, Wilcoxon signed rank test, Pearson correlation, Kendall rank
correlations, and diet/foraging category based on table 2.1 of TL and prey δ13C for all
three species at King George Island, South Shetland Islands and three Gentoo populations
at Wiencke, Rongé and Elephant Island. Bolded paired t-test and Pearson correlation
results are presented in the results and discussion due to sample distributions that were
both normal and had equal variance for the given species/population. Bolded Wilcoxon
signed rank test and Kendall rank correlation results are presented in the results and
discussion due to sample distributions that were both non-normal and had non-equal
variance for the given species/population….…………….….………………….………..21
Table 3.1: Summary of the differences in ecological and physical characteristics between the
four major study regions. Southern Antarctic Circumpolar Convergence Front (SACCF),
Polar Front (PF), Sub-Antarctic Front (SAF). Prey resources based on previous dietary
studies of Gentoo Penguins in each region……….…………….……………….……….42
Table 3.2: Individual Specialization niche parameters for δ15N and δ13C per region. WIC =
within-individual component, BIC = between-individual component, TNW = total niche
width, WIC/TNW = ratio approximating degree of individual specialization within
populations..….….…………………….…………….…….……………………………..47
Table 3.3: Population level metrics evaluating differences in bivariate niche parameters. TA =
total niche area based on individual averages, MDC = mean distance to centroid, NND =
mean nearest neighbor distance, Mean SEAb = mean of posterior probability distributions
for Bayesian standard ellipse area estimates including 95% credibility intervals. Groups
that do not share at least 1 superscripted letter within a column are significantly different
for the given variable at the 0.05 level. Ranges of values are in brackets……….………49
Table 3.4: Individual relative niche and overlap metrics based on total niche area (TA).
Population TA = total niche area based on all individual values. Asterisk (*) indicates a
significant difference from all other populations for the given variable at the 0.05 level.
Ranges of values are in brackets……….…………….…….…………………………….50
iv
Table A.1: Parameter estimates, 95% confidence intervals (in parentheses), residual mean square
error (MSE) and pseudo R2 values estimated from fitted von Bertalanffy growth equations of
total feather length (cm) relative to time (days) elapsed since the onset of feather molt………..70
Table A.2: Estimated timing of growth for tail feather sections using the formula:
%
,
! = !# − × ln 1+ - . Asterisk (*) indicates tail feather sections used for analysis…….…..71
&
,.
v
LIST OF FIGURES
Figure 2.1: Map of study sites: Wiencke Island, Rongé Island, King George Island, and Elephant
Island. Dashed lines indicate major currents and fronts: Southern Antarctic Circumpolar
Current Front (SACCF) and Southern Boundary (SB)..……….………………..……….14
Figure 2.2: Trophic level (TL) and prey δ13C values for Adélie, Chinstrap, and Gentoo Penguins
at King George Island, South Shetland Islands, Antarctica, 2010. Lines connect nonbreeding season (feather) values and breeding season (blood) for each sampled
individual.……….…………….…………….…………….…………….………….……22
Figure 2.3: Trophic level (TL) and prey δ13C values Gentoo Penguins at three breeding sites in
the South Shetland Island and Western Antarctic Peninsula. Lines connect non-breeding
season (feather) values and breeding season (blood) for each sampled individual……...23
Figure 3.1: Map of the four Gentoo Penguin breeding locations sampled in the study. Major
currents and fronts are indicated by dotted lines: Sub-Antarctic Front (SAF), Polar Front
(PF), and Antarctic Circumpolar Convergence (ACC)...….…………….………….……41
Figure 3.2: Standard ellipse areas and convex hull areas (TA) representing niches of each
population. Points represent individual averaged δ15N and δ13C values. Black = Western
Antarctic Peninsula, red = South Shetland Islands, green = South Georgia, blue =
Falkland Islands……………….…………….…………….…………….………….……48
Figure 3.3: Convex hull areas (TA) for each individual within the four sample populations.
Dotted line represents TA for entire population.………….…………….………….……51
Figure A.1: Fitted von Bertalanffy growth curve (solid line) and 95% confidence intervals
(dotted lines) of total tail feather length relative to time since the onset of feather molt in
a captive population of twelve adult (5 male and 5 females) Gentoo Penguins maintained
at SeaWorld in Orlando, Florida during 2012…………….…………….………….……72
vi
ABSTRACT
Gentoo Penguins (Pygoscelis papua) are known to be generalist foragers, while Adélie
(P. adeliae) and Chinstrap (P. Antarctica) tend to specialize on krill within the Western Antarctic
Peninsula and South Shetland Islands, particularly during the breeding season. However, little is
known on temporal consistency in diet and foraging habitat of these species, particularly at the
individual level. We used stable isotope analysis (SIA) of blood and feathers to evaluate seasonal
and individual foraging consistency within Adélie, Chinstrap and Gentoo Penguins breeding in
the South Shetland Islands, as well as among three Gentoo Penguins’ populations in the Western
Antarctic Peninsula and South Shetland Islands. Our results suggest that Pygoscelis penguins can
differ in foraging ecology not only at the population level among species, sites and seasons, but
also in the level of individual variation within populations, and in the degree of seasonal
consistency within individuals.
Previous dietary analyses suggest Gentoo penguins have a generalist foraging niche,
which may help buffer them from recent climate-driven declines in key prey species, such as
Antarctic krill (Euphausia superba). Ecological theory indicates that generalist populations fall
under two different categories: Type A generalist populations exhibit large variation within
individuals, and little variation between individuals, where Type B generalist populations are
comprised of individual specialists, with large variation between individuals. We conducted SIA
using tail feathers from Gentoo penguins at four geographically isolated breeding sites across the
Scotia arc to assess individual variation in winter diets and determine the type of generalist
strategies that Gentoo penguins utilize. Our results indicate the presence of individual
specialization (type B generalism) within all four geographically distinct breeding colonies, with
lower degrees of individual specialization in southern populations and higher degrees of
vii
individual specialization in northern populations. In addition, our results also suggest that
individual specialization may be driven by prey abundance and diversity, as foraging habitat in
the southern populations are marked by high abundance of Antarctic krill and low prey diversity,
while the northern populations forage on a wider diversity of prey.
viii
CHAPTER 1: GENERAL INTRODUCTION
In the last four decades, the Antarctic Peninsula and South Shetland Islands have
experienced substantial climate-driven ecosystem changes, in particular a decline in the amount
and seasonal duration of sea-ice coverage (Stammerjohn et al. 2008, Pritchard et al. 2012). Many
studies suggest these changes are the major cause of decline in the abundance of Antarctic Krill
(Euphausia superba), a keystone species for the Antarctic marine food web (Atkinson et al.
2004; Ducklow et al. 2007, Trivelpiece et al. 2011). Concurrently, the three species of Pygoscelis
penguins, which consume krill, have experienced substantial changes in population numbers in
the Antarctic Peninsula. Both Chinstrap (Pygoscelis antarctica) and Adélie Penguins (P. adeliae)
in this region are declining, and the prevailing hypothesis suggests that decreases in Antarctic
krill abundance due to climate change and krill fisheries are contributing to their declines
(Trivelpiece et al. 2011). However, Gentoo Penguins (P. papua), which also forage on krill,
have not declined, and some populations are even increasing in numbers and colonizing
previously uninhabited areas south of their historical range (Lynch, et al. 2012).
Gentoo Penguins appear to show resilience to changes in krill abundance and other
climate-driven changes in their ecosystem. One hypothesis is that, relative to other penguin
species, they have a more flexible diet and generalist foraging niche, which could help to buffer
them from recent declines in krill (Miller et al. 2009, Polito et al. 2015). In contrast, Chinstrap
and Adélie Penguins appear to be more dependent on krill and may be unable to adapt to
declines in this principle prey item (Trivelpiece et al. 2011, Polito et al. 2015).
However, much of what we know about the foraging ecology of penguins in the Antarctic
Peninsula as well as throughout the Scotia Arc is based almost exclusively on stomach content
analyses restricted to the breeding season and at very few sites. Unfortunately, stomach content
1
analyses provide only a snapshot of recent diets and can have significant preservation biases
(Polito et al. 2011). Therefore, hypotheses regarding the dietary niche of Gentoo Penguins
require more rigorous testing to assess temporal dietary consistency outside of the breeding
season.
Initially, population changes in Gentoo, Adélie, and Chinstrap Penguin along the
Antarctic Peninsula and South Shetland Islands were attributed to climate driven decline in sea
ice in the region (Montes-Hugo et al. 2009). Gentoo Penguins prefer habitat free of ice in which
to forage, while Adélie Penguins rely on the presence of sea ice for foraging. However,
Chinstrap Penguins also prefer ice-free habitat, so the drastic declines in their population
numbers are unexplained. This suggests an alternative driving force in the substantial changes of
these three species’ populations.
Reduction in the amount and seasonal duration of sea-ice has caused a marked decline in
the overall biomass of Antarctic krill in the Antarctic Peninsula as well as in the southern Scotia
Sea (Trivelpiece et al. 2011). Antarctic krill are largely dependent on sea-ice during a crucial
stage of their life cycle, and loss of this habitat may be causing a decrease in reproductive
success and recruitment (Smetacek et al. 1990). As krill are a major prey item for penguins and
other predatory vertebrates, the current hypothesis is that the changes in penguin populations are
the product of changes in krill availability and abundance (Trivelpiece et al. 2011). This provides
an alternative explanation for the declining populations of Chinstrap and Adélie Penguins, whose
diets are dominated by Antarctic krill. However, it is thought that the more variable diets of
Gentoo Penguins may buffer them against declines resulting from decreases in krill (Polito et al.
2015). Observed increases in populations and range expansion in Gentoo Penguins may further
2
be due to an interspecific ecological release as the other Pygoscelis penguins decline in the
Antarctic Peninsula and South Shetland Island region (Miller et al. 2009, 2010).
Gentoo Penguins are one of the most widespread penguin species, with a circumpolar
breeding distribution and a wide latitudinal range stretching from 46°00’ S in the Crozet Islands
south to 65°16’ S on the Antarctic Peninsula (Ainley et al. 1995). Gentoo Penguins are generalist
which exhibit substantial variation in their diet across breeding locations in the Scotia Arc.
Stomach content analyses of Gentoo Penguins of the South Shetland Islands show diets
composed of mostly Antarctic krill (Euphasia superba) (Miller et al. 2009), while studies
conducted on populations in South Georgia have found that diets there consist of a mix of
Antarctic Krill and fish. In the Falkland Islands, where there are no Antarctic krill, Gentoo
Penguin diets include mostly fish, as well as some crustaceans such as lobster krill (Munida
gregaria) and cephalopods (Putz et al. 2001; Clausen and Putz 2002, 2003).
Sub-Antarctic populations of Gentoo Penguins in South Georgia and the Falkland Islands
have been relatively stable (Baylis et al. 2013), while populations in the South Sandwich Islands
have undergone increases in population size (Lynch et al. 2012; Forcada and Trathan 2009;
Convey et al. 1999). Additionally, there is recent evidence of population increases and southward
range expansion in the Antarctic Peninsula, while the populations of Adélie and Chinstrap
Penguins are drastically declining in these areas (Lynch et al. 2013).
Generalist populations fall under two different categories: Type A generalist populations
exhibit extensive variation within individuals and little variation between individuals, whereas
Type B generalist populations are composed of individuals that specialize, with large variation
between individuals (Bolnick et al. 2003). Quantifying the degree of individual variation within
Gentoo Penguin populations will allow for the detection of the type of generalist strategy that a
3
population employs (Bearhop et al. 2004). It is important to assess which type of generalist
population Gentoo Penguins fall under, as these strategies may impart differing ecological and
evolutionary responses under times of environmental change. Individuals in a Type A generalist
species would be able to respond to change quickly, whereas many individuals in Type B
generalists would not successfully adapt to shifts in prey availability even if proportions of the
populations survive. For example, individual specialization in bluegill sunfish has been shown to
produce a delayed response to fluctuations in prey availability (Werner et al. 1981). VillegasAmntmann et al. (2008) suggested that Galapagos sea lion populations decline during times of El
Nino when prey availability is significantly lower due to a lack of dietary plasticity in individual
specialists. Environmental change could result in significant population declines and lower
genetic diversity in populations composed of Type B generalists.
Foraging strategies employed by a species may differ among geographically distinct
populations. A comparison of multiple studies of Galapagos sea lions (Zalophus wollebaeki)
throughout the Galapagos Archipelago suggests a high degree of variation in foraging strategies
between populations driven by variation in prey availability and physical oceanographic
characteristics (Villegas-Amtmann et al. 2008; Salazar, 2005; Dellinger and Trillmich, 1999;
Kooyman and Trillmich, 1986). Quantifying variation in diet among individuals and through
time would allow for a more robust measure of foraging niche width that can be compared
between populations whose foraging strategies may differ in habitat and trophic composition.
Stable isotope analysis (SIA) provides an effective method for identifying diet
composition and foraging niches penguins (Polito et al 2011, Polito et al. 2015). Nitrogen stable
isotope values (δ15N) of marine consumers exhibits a strong linear relationship with trophic level
(Hobson et al. 1994). This is due to “trophic enrichment” in which an organism’s body
4
selectively sequesters more 15N than 14N from the food it consumes, while excreting more 14N,
resulting in an enrichment of 15N in the organism’s tissue (Hobson et al. 1994). The accumulated
effects of enrichment result in higher relative levels of 15N in organisms foraging at higher
trophic levels. Analysis of individual δ15N values enables the detection of variation in trophic
food choice between and within species that may go undetected through other methods (Bearhop
et al, 2006).
Carbon stable isotope values (δ13C) differs very little between trophic levels and instead
can provide a means of detecting whether carbon acquired during foraging comes from offshore
or inshore locations. Primary producers in inshore (benthic) systems are more efficient at 13C
uptake during photosynthesis than those in offshore (pelagic) systems (France, 1995). Higher
carbon isotopic values (δ13C) therefore indicate that individuals have been focusing foraging at
inshore locations (Cherel and Hobson, 2007). In addition, lower latitude phytoplankton and
particulate organic matter (POM) are more enriched in 13C compared to higher latitudes, which
results in an 13C “isoscape” that increases towards lower latitudes (Rau et al 1991). This is due to
high levels of dissolved CO2 in cold southern oceans, which reduce the presence of organic δ13C
(Rau et al. 1997, Popp et al. 1999). The latitudinal gradient of δ13C can be used to infer
latitudinal foraging positions of seabirds (Quillfeldt et al. 2005; Cheryl and Hobson 2007; Jaeger
et al. 2010).
Together nitrogen and carbon isotopic values have been used to determine the 2dimensional parameters of an isotopic niche size, an index of ecological or foraging niche that
can be used to identify specialist vs. generalist strategies in many top marine predators (Bearhop
et al. 2004, Newsome et al. 2007, Jaeger et al. 2009, Huckstadt et al. 2012). Moreover, this
method has been effectively used to determine the specialist foraging niches of Chinstrap and
5
Adélie Penguins and to broadly characterize the generalist foraging strategies of Gentoo
Penguins, which exhibit a much wider isotopic niche with a higher degree of variation in both
prey choice and location compared to sister species (Miller et al. 2009, 2010; Polito et al. 2015).
Stable isotopes have also been used to identify type A vs. type B generalists in many marine
predators by determining and comparing the degree of individual variation within a population
(Newsome et al. 2009; Huckstadt et al. 2012; Kernaleguen et al. 2015). My proposed research
will build on these existing studies to detect individual variation and determine whether Gentoo
Penguins exhibit Type A or Type B generalist strategies using stable isotope analysis.
In this thesis I conduct stable isotope analysis of three Pygoscelis penguin species to
detect temporal consistency in foraging ecology, identify the level of variation in foraging
strategies among individuals, and assess temporal and eco-geographical variability in generalist
strategies. In Chapter 2, I compare seasonal consistency in foraging strategies between three
Pygoscelis penguin species as well as between three separate Gentoo Penguin populations within
the Western Antarctic Peninsula and South Shetland Islands. In Chapter 3, I determine the
prevalence of Type A vs. Type B foraging niche generalization occurring in Gentoo Penguin
populations across a large portion of their breeding distribution.
LITERATURE CITED
Ainley, D. G., N. Nur, and E. J. Woehler. 1995. Factors affecting the distribution and size of
pygoscelid penguin colonies in the Antarctic. The Auk 112:171-182.
Atkinson, A., V. Siegel, E. Pakhomov, and P. Rothery. 2004. Long-term decline in krill stock
and increase in salps within the Southern Ocean. Nature, 432(7013): 100-103.
Baylis, A. M., S. Croft, and A.C. Wolfaardt. 2013. Population trends of Gentoo Penguins
Pygoscelis papua breeding at the Falkland Islands. Marine Ornithology 41(1): 1-5.
6
Bolnick, D. I., R Svanback, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L.
Forister. 2004. The ecology of individuals: Incidence and implications of individual
specialization. The American Naturalist 161:1-28.
Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. Macleod. 2004. Determining trophic
niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology
73:1007-1012.
Bearhop, S., R. A. Phillips, R. McGill, Y. Cherel, D. A. Dawson, and J. P. Croxall. 2006. Stable
isotopes indicate sex-specific and long-term individual foraging specialisation in diving
seabirds. Marine Ecology Progress Series 311:157-164.
Cherel, Y., and K.A. Hobson. 2007. Geographical variation in carbon stable isotope signatures of
marine predators: a tool to investigate their foraging areas in the Southern Ocean. Marine
Ecology Progress Series 329: 281–287.
Clausen, A. P., and K. Putz. 2002. Recent trends in diet composition and productivity of Gentoo,
Magellanic and rockhopper Penguins in the Falkland Islands. Aquatic Conservation:
Marine and Freshwater Ecosystems 12(1):51-61.
Clausen, A., and K. Pütz. 2003. Winter diet and foraging range of Gentoo Penguins (Pygoscelis
papua) from Kidney Cove, Falkland Islands. Polar Biology 26(1):32-40.
Convey, P., A. Morton, and J. Poncet. 1999. Survey of marine birds and mammals of the South
Sandwich Islands. Polar Record 35(193): 107-124.
Dellinger, T., and F. Trillmich. 1999. Fish prey of the sympatric Galapagos fur seals and sea
lions: seasonal variation and niche separation. Canadian Journal of Zoology 77:1204–
1216.
Ducklow, H. W., K. Baker, D. G. Martinson, L. B. Quetin, R. M. Ross, R. C. Smith, S. E.
Stammerjohn, M. Vernet, and W. Fraser. 2007. Marine pelagic ecosystems: the West
Antarctic Peninsula. Philos T Roy Soc B 362: 67–94.
Forcada, J., and P. N. Trathan. 2009. Penguin responses to climate change in the Southern
Ocean. Global Change Biology 15(7): 1618-1630.
France, R. L. 1995. Carbon-13 enrichment in benthic compared to planktonic algae: foodweb
implications. Marine Ecology Progress Series 124:307–312.
Hobson, K. A., and R. G. Clark. 1992. Assessing avian diets using stable isotopes I: turnover of
13C in tissues. Condor 94:181-188.
Hückstädt, L. A., P. L. Koch, B. I. McDonald, M. E. Goebel, D. E. Crocker, and D. P. Costa.
2012. Stable isotope analyses reveal individual variability in the trophic ecology of a top
marine predator, the southern elephant seal. Oecologia 169(2):395–406.
7
Jaeger, A., P. Blanchard, R. Richard, and Y. Cherel. 2009. Using carbon and nitrogen isotopic
values of body feathers to infer inter- and intra-individual variations of seabirds feeding
ecology during moult. Marine Biology 156:1233-1240.
Jaeger, A., V. Lecomte, H. Weimerskirch, P. Richard. And Y. Cherel. 2010. Seabird satellite
tracking validates the use of latitudinal isoscapes to depict predators' foraging areas in the
Southern Ocean. Rapid Communications in Mass Spectrometry 24(23): 3456-3460.
Kernaléguen, L, N. Dorville, D. Ierodiaconou, A. J. Hoskins, A. M. M. Baylis, M.A. Hindell, J.
Semmens, K. Abernathy, G. J. Marshall, Y. Cherel, and J. P. Y. Arnould. 2015. From
video recordings to whisker stable isotopes: a critical evaluation of timescale in assessing
individual foraging specialization in Australian fur seals. Oecologia 180(3):657-70.
Kooyman, G. L. and F. Trillmich. 1986. Diving behanior of Galapagose fur seals. In Furl seals:
maternal strategies on land and at sea. Princeton University Press, Princeton N.J. pp 186219.
Lynch, H. J., R. Naveen, P. N. Trathan, and W. F. Fagan. 2012. Spatially integrated assessment
reveals widespread changes in penguin populations on the Antarctic Peninsula. Ecology
93(6):1367-1377.
Lynch, H. J., R. Naveen, and P. Casanovas. 2013. Antarctic site inventory breeding bird survey
data, 1994-1013. Ecology 94:2653.
Miller, A. K., N. J. Karnovsky, and W. Z. Trivelpiece. 2009. Flexible foraging strategies of
Gentoo Penguins Pygoscelis papua over 5 years in the South Shetland Islands, Antarctica.
Marine Biology 156:2527–2537.
Miller, A. K., M. A. Kappes, S. G. Trivelpiece, and W. Z. Trivelpiece. 2010. Foraging-Niche
Separation of Breeding Gentoo and Chinstrap Penguins, South Shetland Islands,
Antarctica. Condor 112(4):683–695.
Montes-Hugo, M., S. C. Doney, H. W. Ducklow, W. Fraser, W. Martinson, S. E. Stammerjohn,
and O. Schofield. 2009. Recent changes in Phytoplankton communities associated with
rapid regional climate change along the Western Antarctic Peninsula. Science 323:14711473.
Newsome, S. D., C. Martínez del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic
ecology. Frontiers in Ecology and Environment 5:429–436.
Newsome, S. D., M. T. Tinker, D. H. Monson, O.T. Oftedal, K. Ralls, M. M. Staedler, M. L.
Fogel, and J. A. Estes. 2009. Using stable isotopes to investigate individual diet
specialization in California sea otters (Enhydra lutris nereis). Ecology 90(4): 961-974.
8
Polito, M. J., W. Z. Trivelpiece, N. J. Karnovsky, E. Ng, W. P. Patterson, and S. D. Emslie.
2011. Integrating stomach content and stable isotope analyses to quantify the diets of
Pygoscelid penguins. PLoS ONE 6(10):e26642.
Polito, M. J., W. Z. Trivelpiece, W. P. Patterson, N. J. Karnovsky, C. S. Reiss, and S. D. Emslie.
2015. Contrasting specialist and generalist patterns facilitate foraging niche partitioning
in sympatric populations of Pygoscelis penguins. Marine Ecology Progress Series 519:
332-337.
Popp, B.N. T. Trull, F. Kenig, S.G. Wakeham, T. M. Rust, B. Tilbrook, amd E. A. Laws. 1999.
Controls on the carbon isotopic composition of Southern Ocean phytoplankton. Global
Biogeochemical Cycles 13 (4): 827-843.
Prichard, H. D., S. R. M. Ligtenberg, H. A. Fricker, D. G. Vaughan, M. R. van den Broke, and L.
Padman. 2012. Antartic ice-sheet loss driven by basal melting of ice shelves. Nature
484:502-505.
Putz, K., R. J. Ingham, J. G. Smith, and J. P. Croxall. 2001. Population trends, breeding success
and diet composition of Gentoo Pygoscelis papua, magellanic Spheniscus magellanicus
and rockhopper Eudyptes chrysocome penguins in the Falkland Islands. A review. Polar
Biology 24(11):793-807.
Quillfeldt, P., R. A. McGill, and R. W. Furness. 2005. Diet and foraging areas of Southern Ocean
seabirds and their pre inferred from stable isotopes: review and case study of Wilson’s
storm-petrel. Marine Ecology Progress Series 295:295-304.
Rau, G. H., T. Takahashi, D. J. Des Marais, and C. W. Sullivan. 1991. Particulate organic matter
δ13C variations across the Drake Passage. J Geophys Res 96:15131–15135.
Rau, G. H., U. Riebesell, and D. Wolf- Gladrow. 1997. CO2aq- dependent photosynthetic 13C
fractionation in the ocean: A model versus measurements. Global Biogeochemical
Cycles, 11(2): 267-278.
Salazar, S. K. 2005. Variación temporal y espacial del espectrotrófico del lobo marino de
Galápagos. Master’s thesis, Instituto Politécnico Nacional, La Paz.
Smetacek, V., R. Scharek, and E. M. Nöthig. 1990. Seasonal and regional variation in the
pelagial and its relationship to the life history cycle of krill. In Antarctic Ecosystems (pp.
103-114). New York. Springer Berlin Heidelberg.
Stammerjohn, S.E., D. G. Martinson, R. C. Smith, and R. A. Iannuzzi. 2008. Sea ice in the
western Antarctic Peninsula region: Spatio-temporal variability from ecological and
climate change perspectives. Deep-Sea Res II 55:2041–2058.
9
Trivelpiece, W.Z., J. T. Hinke, A. K. Miller, C. S. Reiss, S. G. Trivelpiece, and G.M. Watters.
2011. Variability in krill biomass links harvesting and climate warming to penguin
population changes in Antarctica. Proceedings to the National Academy of Science USA
108:7625–7628.
Villegas-Amtmann, S., D. P. Costa, Y. Tremblay, S. Salazar, and D. Aurioles-Gamboa. 2008.
Multiple foraging strategies in a marine apex predator, the Galapagos sea lion Zalophus
wollebaeki. Marine Ecology Progress Series 363:299-309.
Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in sizestructured populations. Annual review of ecology and systematics, 15:393-425.
10
CHAPTER 2:
SPECIES AND POPULATION LEVEL DIFFERENCES IN SEASONAL
CONSISTENCY AND INDIVIDUAL VARIATION OF PYGOSCELIS
PENGUIN FORAGING STRATEGIES
INTRODUCTION
Research into the diets and foraging ecologies of Antarctic penguins can provide vital
information about their vulnerability to ecological pressures and help explain declines in certain
species or populations. For example, the diet of Adélie (Pygoscelis adeliae) and Chinstrap
Penguin (P. antarctica) is well documented at the population level, particularly in the South
Shetland Islands and Western Antarctic Peninsula, where past studies indicate that both species
forage primarily on Antarctic Krill in offshore areas (Volkman et al. 1980; Trivelpiece et al.
1987; Karnovsky 1997). This relatively narrow foraging niche and dietary focus on Antarctic
krill, which has declined over the past 30 years (Atkinson et al. 2004), is suggested to drive
recent populations decline of Adélie and Chinstrap Penguins in this region (Trivelpiece et al.
2011). In contrast, Gentoo Penguin (Pygoscelis papua) populations are stable and even
increasing in this same region (Lynch et al. 2013). One possible explanation for this trend is the
more generalist foraging strategy employed by Gentoo Penguins, which may allow for flexibility
in diet and foraging habitat (Miller et al. 2009, Polito et al. 2015).
While the diets and foraging ecology of Pygoscelis penguins have been well studied in
the Antarctic Peninsula region, many studies overlook diet and niche variation between
individuals within populations, as well as variation within individuals. In addition, past studies
focuses primarily on the breeding season and little is known about the diets of these species
during the non-breeding period (Miller et al. 2009, 2010, Juáres et al. 2016). Even so, existing
studies do indicate that Gentoo Penguin populations can exhibit substantial variation in their
11
foraging ecology across breeding locations (Miller et al. 2009; Putz et al. 2001; Clausen and Putz
2002, 2003; Polito et al. 2015). However, it has not been possible to assess if and how
individuals within these populations vary in diets and assess the degree of individual
specialization and foraging constancy within these generalist populations (Croxall et al, 1988;
Wilson et al, 1998; Clausen and Putz, 2003; and Polito et al. 2015). Specifically, these past
studies lack analyses of dietary and foraging habitat consistency of individuals through time,
which is necessary for a robust assessment of individual specialization (Bolnick et al, 2003;
Bearhop et al. 2004). Quantifying the degree of niche variation and consistency at the individual
level is critical because individual variation within populations has the potential to determine a
species population level response to changes in food availability or other environmental change
(Bolnick et al. 2003).
Although the foraging ecology of individual marine predators can be challenging to track
over time, stable isotope analyses (SIA) provides a robust tools for measuring animal diets
through time from a single capture event. Prior studies have demonstrated that SIA is an
effective method for quantifying the diets and foraging niches of Pygoscelis penguins (Polito et
al 2011, Polito et al. 2015). Nitrogen stable isotope values (δ15N) in consumers exhibits a strong
correlation with trophic level (Minagawa & Wada 1984, Hobson et al. 1994). Carbon stable
isotope values (δ13C) differs little between trophic levels but instead provides a proxy of marine
foraging habitat use due to differences in baseline δ13C values between inshore/benthic and
offshore/pelagic habitats (France 1995, Hobson & Cherel 2011). In addition, the δ13C and δ15N
values of specific tissues reflect the diets of consumers at the time of synthesis such that
comparing tissues synthesized at different times from a single individual can assess foraging
niche consistency over time (Newsome et al. 2009; Hückstädt et al. 2012; Kernaleguen et al.
12
2015). For example, Bearhop et al. (2006) compared the δ13C and δ15N values of blood (i.e.
breeding season) and feathers (i.e. non-breeding season) in four diving seabird species at Bird
Island, South Georgia to investigate the degree of foraging specialization and individual
consistency within species. However, to our knowledge a similar analysis has not been
conducted on sympatric Pygoscelis populations. Such an analysis would help to quantifying
seasonal diet variation and the degree of consistency at the individual level and help to inform
these species responses to recent changes in krill availability in the Antarctic Peninsula region.
To address this gap, we use SIA of blood and feathers to evaluate seasonal and
individual foraging consistency within Pygoscelis species (Adélie, Chinstrap and Gentoo
Penguin) breeding in sympatry in the South Shetland Islands, Antarctica. In addition, we also
investigate these same parameters among three Gentoo Penguins’ populations in the Western
Antarctic Peninsula and South Shetland Islands, Antarctica, as this species is predicted to be a
generalist forager at the population level. Our goals are to: 1) determine if and how population
level diet and foraging habitat use shifts seasonally and, 2) assess the degree which individual
variation and foraging consistency within populations influence the population level foraging
niches of Pygoscelis penguins.
METHODS
Study Site and Sample Collection
In December 2010, we collected samples from breeding Adélie, Chinstrap, and Gentoo
Penguins at Admiralty Bay, King George Island, South Shetland Islands (62.1°S, 58.25°W)
(Figure 1.1). In December 2014 we collected samples from breeding Gentoo Penguins at three
sites in the Western Antarctic Peninsula and South Shetland Islands: Damoy Point, Wiencke
Island (6-4.8°S, 63.5°W), Georges Point, Rongé Island (64.7°S, 62.7°W), and Stinker Point,
13
Elephant Island (61.1°S, 55.2°W) (Figure 2.1). At each breeding site, we collected 3 body
60°S
feathers and 1ml of whole blood from 15-20 individuals per species.
Elephant Island
61
CF
SAC
62
SB
King George Island
64
63
SOUTH SHETLAND
ISLANDS
Wiencke Island
65
Rongé Island
0
100
58
56
200 km
66
WESTERN
ANTARCTIC
PENINSULA
64
62
60
54
52°W
Figure 2.1: Map of study sites: Wiencke Island, Rongé Island, King George Island, and Elephant
Island. Dashed lines indicate major currents and fronts: Southern Antarctic Circumpolar Current
Front (SACCF) and Southern Boundary (SB).
In penguins, blood contains isotopic signatures of prey items consumed within
approximately 20 days prior to sampling, thus in this study provides a proxy for dietary
information during the late incubation period when individuals were sampled (Barquete et al.
14
2013). Body feathers are metabolically inert after synthesis and reflect isotopic signatures of prey
items consumed post-breeding and prior to molting in the previous year (Polito et al. 2011a).
Combined these two tissues provide proxies of individual’s diet and foraging habitat usage
during the non-breeding (feathers) and breeding (blood) seasons that can be compared to
evaluate seasonal consistency.
Sample Preparation and Isotopic Analysis
We soaked body feathers in a mixture of 2:1 chloroform-methanol for 24 hours to remove
lipids, then rinsed each segment in 2:1 chloroform-methanol and allowed them to air dry for 24
hours (Cherel et al. 2005). We then subsampled pieces of the vane of equal size from three body
feathers to obtain a total of 0.5-0.6mg per individual. We dried blood samples to constant weight
at 50°C in an analytical oven for 48 hours and then ground the dried blood into a powder and
sub-sampled 0.6mg of blood per individual for analysis. Samples were analyzed using a PDZ
Europa ANCA-GSL and Costech ECS4010 elemental analyzers interfaced to a PDZ Europa 2020 and Thermo Finnigan Delta Plus XP continuous flow stable isotope ratio mass spectrometers.
Raw δ values were normalized using glutamic acid, bovine liver, and nylon 5 as reference
materials (USGS-40: δ13C = -16.65‰, δ15N = -6.8‰; USGS-41: δ13C = -37.63‰, δ15N = 47.6‰;
bovine liver: δ13C = -21.69‰, δ15N = 7.72‰; nylon 5: δ13C = -27.72‰, δ15N = -10.31‰).
Sample precision based on internal repeats and duplicate standard reference materials was 0.1‰,
for both δ15N and δ13C. Stable isotope ratios are expressed in the δ notation in per mil units (‰)
according to the following equation:
δX = [(Rsample/ Rstandard) − 1] × 1000
where X is 13C or 15N and R is the corresponding ratio 13C/12C or 15N/14N. The R standard
values were based on the Peedee belemnite (VPDB) for 13C and atmospheric N2 for 15N.
15
Trophic Level and Prey δ13C Values
Dietary isotopic discrimination, which is the difference between the isotopic values of
diet and consumers, can vary significantly across tissue types even when synthesized under the
same diet (Hobson and Clark, 1998; Bond and Jones, 2009; Polito et al. 2009). Therefore, in
order to directly compare blood and feather δ15N values, we converted these values to trophic
level using the formula below (Hobson et al. 1994, 2002, Hobson & Bond 2012; Brasso and
Polito, 2013):
15
15
TLbird = 3 + (δ N bird - ΔN avian tissue – δ Nprimary consumer)/ΔNfood web
This model uses an individual bird’s δ15N value to estimate its trophic level (TL) in relation to
the mean δ15N values of a food web-specific primary consumer (δ15Nprimary consumer) and the mean
δ15N food web trophic discrimination per trophic transfer (ΔNfood web), while accounting for
tissue-specific δ15N discrimination factors (ΔNavian tissue). For our analyses, we used mean
δ15Nprimary consumer values of salps (Salpa thompsoni) collected from the Antarctic Peninsula region
(2.7‰; Stowasser et al. 2012) and a mean ΔNfood web value of 3.4‰, which is a robust value
across multiple food webs (Deniro and Epstein, 1981; Minagawa and Wada, 1984; Post, 2002;
Søreide et al., 2006; Brasso and Polito, 2013). We incorporated mean blood ΔNavian tissue values
derived from captive studies of four piscivorous birds (+2.7; Cherel et al. 2005) and mean feather
ΔNavian tissue values discrimination factor from a captive feeding study of Pygoscelis penguins
(+3.5; Polito et al. 2011a).
In order to directly compare blood and feather δ13C values, we applied tissue-specific
discrimination values to penguin tissue δ13C values in order to derive expected prey δ13C values
following the methods of Hobson & Bond 2012, which correspond to the prey habitat in which
the penguins are foraging. Once again based on the studies by Polito et al. (2011a) and Cherel et
16
al. (2005), we subtracted the corresponding discrimination factors from feather (+1.3‰) and
blood (+0.0‰) δ13C values prior to analysis.
Statistical Analysis
Prior to analysis, we examined for normality in all populations using the Shapiro-Wilks
test. We also examined for homogeneity of variance using Bartlett’s test for normally distributed
populations and Levene’s test for non-normally distributed populations. In order to determine
whether populations are more generalized or specialized relative to one another, we calculated
coefficient of variance (CV; σ//) as a proxy for individual variation within populations of
species for non-breeding (feather) season and breeding (blood) season TL and prey δ13C values.
Differences in CV were assessed qualitatively as there are no robust statistical analyses that can
directly test for significant differences between CVs (Donnelly and Kramer, 1999).
To test for population level diet and foraging habitat consistency of the three penguin
species sampled from King George Island in 2010, we compared population means of TL and
prey δ13C values between non-breeding (feather) and breeding (blood) seasons using paired ttests for normally distributed populations with equal variance and Wilcoxon signed rank tests for
populations that were non-normal.
To test whether individual diet and foraging habitat is consistent between seasons and/or
individual diet and foraging habitat are consistent relative to each other, we tested for
relationships between individual’s breeding and non-breeding season TL and δ13C values using
Pearson Correlation for normally distribute populations with equal variance and Kendall rank
correlations for populations that were non-normal or had unequal variance. We conducted these
same analyses for three Gentoo Penguin populations sampled in 2014 in order to test for
geographic variation in the presence of population level and individual level diet and foraging
17
consistency. All statistical analyses were performed in R software ver. 3.2.1 (R Core Team
2015). Significance was assumed at the α = 0.05 level and all means are presented ±SD.
To aid in the interpretation and discussion of our analyses, we developed a framework in
which each sample population was categorized into one of four “types” based on four possible
combinations of the paired t-test/Wilcoxon signed rank test and Pearson correlation/Kendall rank
correlation results (Table 2.1).
Table 2.1: Foraging “types” and interpretations based on four possible combinations of the
paired t-test/Wilcoxon signed rank test and Pearson correlation/Kendall rank correlation results
for TL and prey δ13C.
Paired t-test/
Wilcoxon
signed rank
test
TL/
Prey
δ13C
Significant
Not
Not
Significant
Pearson
correlation/
Kendall rank
correlation
Not
Not
Significant
Significant
Foraging
Type
Interpretation
Type 1
Population diets/habitats are not
seasonally consistent; individuals are not
seasonally consistent relative to each
other.
Type 2
Type 3
Type 4
18
Population diets/habitats are seasonally
consistent; individuals are not seasonally
consistent relative to each other.
Population diets/habitats are seasonally
consistent; individuals are seasonally
consistent relative to each other.
Population diets/habitats are not
seasonally consistent; individuals are
seasonally consistent relative to each
other.
RESULTS
Comparison of three species at King George Island in 2010
Gentoo Penguins had the highest CVs for breeding season (blood) and non-breeding
season (feather) TL, followed by Adélie penguins and Chinstrap Penguins (Table 2.2). Adélie
and Chinstrap Penguins had nearly twice the CV for non-breeding season prey δ13C values
compared to Gentoo Penguins. All three species displayed a much lower and similar CV for
breeding season prey δ13C values.
Gentoo Penguins TL and prey δ13C values did not differ significantly between the nonbreeding season (feather) and breeding season (blood; Table 2.3). While Adélie penguin TL did
not differ significantly between the non-breeding and breeding seasons, Adélie penguin prey
δ13C values were significantly lower during the non-breeding season relative to the breeding
season by 1.0‰. Chinstrap Penguin TL during the non-breeding season was significantly lower
relative to the breeding season by 0.15 trophic levels, but there was no significant difference
between the non-breeding and breeding seasons in prey δ13C values (Table 2.3). There was a
significant correlation in individual Gentoo Penguin TL and prey δ13C values between the two
seasons examined (Table 2.3; Figure 2.2). In contrast, there was no correlation in individual TL
or prey δ13C values between seasons for either Chinstrap or Adélie Penguins (Figure 2.2).
Comparison of three Gentoo Penguin populations in 2014
Gentoo Penguins at Elephant Island had the highest CV for TL during both the nonbreeding and breeding season by more than double that of the populations at Wiencke and Rongé
Island (Table 2.2). The population at Rongé Island had the highest CV for prey δ13C during the
non-breeding season, followed by Wiencke and Elephant Island. Wienke Island population had
the highest prey δ13C CV during the breeding season, followed by Elephant and Rongé Island.
19
Table 2.2: Mean and standard deviation, range (in parentheses), and coefficient of variation
(CV%) of TL, δ15N, and prey δ13C of non-breeding and breeding seasons for all three species at
King George Island and three Gentoo populations at Wiencke, Rongé and Elephant Island.
Non-breeding (feather)
mean ± SD
CV
Breeding (blood)
mean ± SD
CV
TL
Adélie
Chinstrap
Gentoo
3.84 ± 0.12 (3.53 to 4.15)
3.83 ± 0.06 (3.71 to 3.91)
3.94 ± 0.17 (3.71 to 4.50)
3.2
1.7
4.4
3.90 ± 0.10 (3.71 to 4.08)
3.98 ± 0.05 (3.85 to 4.09)
3.92 ± 0.18 (3.76 to 4.53)
2.5
1.3
4.7
δ15N
Adélie
Chinstrap
Gentoo
9.1 ± 0.4 (8 to 10.1)
9.0 ± 0.2 (8.6 to 9.3)
9.4 ± 0.6 (8.6 to 11.3)
4.6
2.4
5.9
8.5 ± 0.3 (7.8 to 9.1)
8.7 ± 0.2 (8.3 to 9.0)
8.5 ± 0.6 (8.0 to 10.6)
4.1
2.0
7.3
Prey
δ13C
Adélie
Chinstrap
Gentoo
-25.4 ± 1.1 (-28.7 to -23.7)
-24.6 ± 1.0 (-25.8 to -22.8)
-24.8 ± 0.7 (-25.1 to -23.1)
-4.4
-4.1
-2.7
-24.5 ± 0.3 (-25.1 to -24.1)
-24.9 ± 0.2 (-25.4 to -24.7)
-24.8 ± 0.2 (-25.8 to -23.1)
-1.2
-0.8
-1.8
TL
Wiencke
Rongé
Elephant
4.10 ± 0.10 (3.94 - 4.29)
4.14 ± 0.10 (3.97 - 4.32)
4.10 ± 0.22 (3.56 - 4.39)
2.1
2.4
5.4
4.05 ± 0.06 (3.97 to 4.18)
4.08 ± 0.07 (3.94 to 4.20)
4.52 ± 0.21 (3.87 to 4.70)
1.4
1.7
4.6
δ15N
Wiencke
Rongé
Elephant
9.60 ± 0.3 (9.4 to 10.6)
10.1 ± 0.3 (9.5 to 10.7)
10.0 ± 0.7 (8.1 to 10.9)
2.9
3.4
7.5
9.0 ± 0.2 (8.7 to 9.4)
9.1 ± 0.2 (8.6 to 9.5)
10.6 ± 0.7 (8.4 to 11.2)
2.1
2.6
6.6
Prey
δ13C
Wiencke
Rongé
Elephant
-25.6 ± 0.7 (-26.6 to -24.0)
-25.4 ± 0.9 (-26.3 to -23.7)
-25.6 ± 0.6 (-26.2 to -24.4)
-2.7
-3.5
-2.3
-24.2 ± 0.5 (-24.9 to -22.5)
-24.8 ± 0.2 (-25.1 to -24.6)
-25.0 ± 0.2 (-25.5 to -24.8)
-2.1
-0.6
-0.8
Gentoo Penguins at Wiencke Island had a very small (0.05 of a trophic level), yet
significant difference in their TL between the non-breeding and breeding seasons. In addition,
the prey δ13C values of this population were significantly lower during the non-breeding season
relative to the breeding season by 0.7‰. While the TL of Gentoo Penguins at Rongé Island did
not differ significantly between the non-breeding and breeding seasons, prey δ13C during the
non-breeding season were significantly higher relative to the breeding season by 0.6‰.
20
Table 2.3: Results for paired t-test, Wilcoxon signed rank test, Pearson correlation, Kendall rank correlations, and diet/foraging
category based on table 2.1 of TL and prey δ13C for all three species at King George Island, South Shetland Islands and three Gentoo
populations at Wiencke, Rongé and Elephant Island. Underlined paired t-test and Pearson correlation results are presented in the
results and discussion due to sample distributions that were both normal and had equal variance for the given species/population.
Underlined Wilcoxon signed rank test and Kendall rank correlation results are presented in the results and discussion due to sample
distributions that were both non-normal and had non-equal variance for the given species/population.
Paired t-test
Wilcoxon signed rank test
Pearson correlation
Kendall rank correlation
t
p-value
W
p-value
r
p-value
tau
p-value
Diet/
foraging
category
TL
Adélie
Chinstrap
Gentoo
-2.0
-8.9
1.1
0.06
< 0.001
0.27
158.5
10.5
257.5
0.98
1
0.18
0.09
0.18
0.8
0.68
0.44
< 0.001
0.06
0.02
0.5
0.75
0.95
0.002
1
2
3
Prey
δ13C
Adélie
Chinstrap
Gentoo
-3.3
1.4
-1.1
0.003
0.17
0.28
32.5
135
35.5
0.003
0.27
0.1
-0.08
-1.6
0.66
0.71
0.13
0.001
-0.08
-0.21
0.17
0.61
0.24
0.006
2
1
3
TL
Wiencke
Rongé
Elephant
2.2
2.0
-0.5
0.04
0.07
<0.001
269.5
259.5
18
0.03
0.05
1
0.2
0.01
0.59
0.4
0.98
0.02
0.15
-0.03
0
0.42
0.87
1
2
1
4
Prey
δ13C
Wiencke
Rongé
Elephant
-8.0
-2.9
-3.4
< 0.001
0.009
0.004
33
43
15.5
1
0.02
0.009
0.22
0.46
0.17
0.35
0.04
0.54
0.2
0.33
0.35
0.25
0.07
0.109
2
2
2
21
Chinstrap
Adélie
non-breeding
−27
−28
−28
3.6
3.4
−22
Gentoo
Chinstrap
Gentoo
non-breeding
non-breeding
−22
−23
−22
breeding
−24
−26
−27
non-breeding
breeding
Gentoo
non-breeding
breeding
breeding
non-breeding
breeding
breeding
breeding
strap
4.8
Chinstrap
Gentoo
non-breeding
non-breeding
−22
−23
−24
−25
−26
−27
non-breeding
breeding
δ13 C
−26
−27
−27
non-breeding
−22
−23
−24
δ13 C
−23
−25
−26
δ13 C
4.0
−27
−24
4.4
−25
non-breeding
3.4
3.4
non-breeding
Gentoo
breeding
breeding
non-breeding
breeding
Gentoo
non-breeding
breeding
non-breeding
−28
non-breeding
3.6
3.4
3.6
breeding
3.8
−28
-28
−28
−28
3.8
3.6
3.6
−27
4.0
-27
TL
3.8
4.2
−26
−26
4.2
TL
−22
4.8
−23
4.0
TL
−25
4.4
-26
δ13 C
4.6
−24
−24
4.6
4.2
δ C
-25
13
Prey δ
13C
4.4
-24
Chinstrap
−25
4.6
−23
4.8
-23
Gentoo
−28
−22
Chinstrap
4.6
−22
4.8
breeding
−25
δ13 C
−24
−25
δ13 C
−26
non-breeding
−28
breeding
−28
breeding
non-breeding
3.6
−27
−27
−28
3.8
−27
4.0
−26
4.2
−25
4.4
δ13 C
TL
−23
−24
4.6
−22
−23
−24
−25
δ13 C
3.8
−28
non-breeding
breeding
4.8
-22
4.4
4.2
TL
Gentoo
Gentoo Chinstrap
3.4
3.4
3.4
élie
breeding
−28
non-breeding
3.4
3.6
3.4
3.8
3.6
3.6
−26
3.6
3.8
non-breeding
breeding
−23
4.8
−22
4.6
−24
δ13 C
TL
4.0
−27
4.2
TL
4.0
3.6
3.4
TL
4.0
breeding
non-breedingChinstrap
breeding
non-breeding
4.8
−23
4.6
TL
4.0
3.8
4.2
4.6
4.4
TL
4.2
4.4
−25
3.8
4.4
non-breeding
breeding
Chinstrap
Chinstrap Adélie
4.2
−26
4.0
4.4
4.2
4.6
Adélie
non-breeding
4.8
4.4
4.0
Chinstrap
4.8
4.8
4.6
Adélie
Adélie
4.8
3.8
breeding
−27
−
3.
3.4
3.6
3.4
Adélie
3.4
non-breeding
breeding
breeding
breeding
non-breeding
breeding
−22
−22
−23
−23
−24
−25
22
δ13 C
−25
−26
−27
δ13 C
−24
−24
4.6
δ C
−25
4.4
−26
4.2
TL
−27
4.0
3.8
4.0
13
TL
4.2
4.4
−23
4.8
4.6
−22
Figure 2.2: Trophic level (TL)
and prey δ13C values for Adélie, Chinstrap,
and Gentoo Penguins at King George Island, South
Gentoo
Gentoo
Shetland Islands, Antarctica, 2010. Lines connect non-breeding season (feather) values and breeding season (blood) for each sampled
individual.
non-breeding
non-breeding
breeding
3.4
3.4
−28
breeding
−28
3.6
Rongé Island
Wiencke Island
Wiencke Island
−22
−22
−23
−23
3.6
−26
−27
−27
−28
3.8
−26
−25
δ13 C
−24
−24
−25
4.4
δ13 C
TL
−27
4.0
−26
4.2
−24
δ13 C
−23
−24
4.6
−22
−22
−23
4.8
Elephant Isl
3.4
−28
−28
breeding
Elephant
Islandbreeding
non-breeding
non-breeding
breeding
breeding
Elephant Island
−24
−25
−27
−26
−27
δ13 C
−26
−24
−25
−23
−24
−22
−23
−22
non-breeding
non-breeding
breeding
non-breeding
breeding
breeding
non-breeding
−28
non-breeding non-breeding
breeding
breeding
3.4
Rongé Island breeding
Elephant
Island
non-brereding
4.8
−26
3.4
breeding
breeding
3.4
breeding
−28
3.6
non-brereding non-breeding
non-breeding
gé Island
−27
−28
3.8
-28
breeding
δ13 C
3.6
−23
−24
4.6
−27
4.0
TL
3.8
TL
−26
4.2
-26
−25
4.4
δ13 C
4.2
4.0
−25
4.4
TL
-25
−28
3.8
3.6
−22
4.4
−23
4.8
−24
4.6
δ13C
−26
4.2
Prey δ13C
4.2
TL
4.0
3.8
-24
breeding
Rongé Island
Elephant Island
−23
4.8
-23
-27
3.4
non-breeding
−28
−22
breeding
Rongé Island
non-breeding
breeding
breeding
δ13 C
4.8
non-breeding
−27
4.0
breeding
Elephant
Island
Rongé
Island
non-breeding
non-brereding
breeding
3.6
4.4
−22
4.6
Wiencke Island
Rongé
Island
non-breeding
4.6
4.8
-22
breeding
3.4
3.4
3.4
e Island
−25
non-brereding non-breeding
−25
breeding
−28
non-breeding
3.6
3.6
3.6
3.4
3.8
3.8
−28
3.8
3.6
3.6
3.4
4.0
4.0
δ13C
3.8
−27
4.0
TL
TL
4.2
4.2
non-breeding
breeding
Elephant
Rongé Island
ElephantIsland
Island
−27
−26
δ13C
TL
TL
4.4
TL
4.4
4.0
4.0
4.2
3.8
4.2
4.6
4.6
−25
4.4
4.4
−24
4.6
4.4
4.8
−23
4.8
4.6
4.6
4.8
Rongé Island breeding
non-brereding non-breeding
breeding
Rongé
Wiencke Island
RongéIsland
Island
−22
Wiencke
Island
Wiencke Island
−26
4.2
4.8
4.8
non-breeding
−25
3.6
−27
−27
−3.
3.
3.4
non-breeding
Elephant
Islandbreeding
non-breeding
non-breeding
breeding
non-breeding
breeding
−22
−24
−25
δ13 C
23
−23
−22
−23
−25
−26
27
δ13 C
−24
−24
4.6
−25
4.4
−26
4.2
27
.0
δ13 C
TL
3.8
4.0
TL
4.2
4.4
−23
4.8
4.6
−22
Figure 2.3: Trophic level (TL) and prey δ13C values Gentoo Penguins at three breeding sites in the South Shetland Island and Western
Elephant Island
Elephant Island
Antarctic Peninsula. Lines connect non-breeding season (feather) values and breeding season (blood) for each sampled individual.
non-breeding
The TL and prey δ13C values of Gentoo Penguins at Elephant Island were both
significantly lower during the non-breeding season relative to the breeding season by 0.42 of a
trophic level, and 0.6‰ respectively. Individual TL or prey δ13C values were not correlated
between seasons at Wiencke Island or Rongé Island (Table 2.3). In contrast, Gentoo Penguins at
Elephant Island exhibited a significant correlation in individual TL values between seasons, but
no correlation in individual prey δ13C values.
DISCUSSION
Gentoo Penguins are often regarded as generalist foragers, whereas Adélie and Chinstrap
are suggested to specialize primarily on krill within the Western Antarctic Peninsula and South
Shetland Islands, particularly during the breeding season (Volkman et al. 1980, Lishman, 1985;
Karnovsky 1997, Miller et al. 2008, 2009, 2010; Polito et al. 2011b; Polito et al. 2015).
However, very few studies have investigated foraging variation within populations of Pygoscelis
species (Polito et al. 2015) or between different times of year (e.g. Polito et al. 2011, Juáres et al.
2016). Our results suggest that Pygoscelis penguins can differ in foraging ecology not only at
the population level among species, sites and seasons, but also in the level of individual variation
within populations, and in degree of foraging consistency within individuals.
Comparison of individual variation and foraging strategies of species at King George Island
Qualitative comparisons of CVs among the three species provide additional support for
differences in foraging strategies as indicated by previous studies (Volkman et al. 1980, Miller et
al. 2010, Polito et al. 2015). CV values comparing the level of individual variability in trophic
level (TL) among the three species at King George Island showed that Gentoo Penguins had the
highest CV, Chinstrap Penguins had the lowest CV, and Adélie penguin CV were intermediate of
24
their congeners during both seasons examined (Table 2.2; Figure 2.2). While qualitative, our
results suggest that Adélie Penguins are more generalized in their trophic level of diet relative to
Chinstrap Penguins, while Gentoo Penguins are more generalized in their trophic level of diet
relative to both Adélie and Chinstrap Penguins. Previous studies support our findings that
Gentoo Penguins have greater variability in diet than Adélie and Chinstrap Penguins in the South
Shetland Islands during the breeding season (Volkman et al. 1980, Polito et al. 2015). However,
while past studies indicate that Adélie Penguins diets are more diverse in regions outside of the
Antarctic Peninsula (e.g. Ainley 2002) to our knowledge there are no previous studies showing
that Adélie Penguins are more generalist in diet relative to Chinstrap Penguins.
Examining the CV of prey δ13C values as a proxy of foraging habit use among the three
species at King George Island suggest that during the non-breeding period individual Adélie and
Chinstrap Penguins use a wider range of foraging habitats relative to individual Gentoo Penguins
(Table 2.2). This finding is similar to a previous study by Juáres et al. (2016) at Stranger Point,
King George Island, in which they found Adélie penguins had a greater dispersion of body
feather δ13C values in comparison to Gentoo Penguins. In addition, these results agree with
tracking studies which indicate that while Adélie and Chinstrap Penguins disperse broadly during
the non-breeding period (Hinke et al 2015), Gentoo Penguins remain close to their breeding
colonies and forage in open-water, near-shore habitats (Wilson et al. 1998).
Seasonal consistency in three species at King George Island
We observed differences across species in population and individual-level trophic level
and foraging habitat use consistency at King George Island. Adélie penguins exhibited a type 2
pattern for TL, suggesting that population diets are seasonally consistent but individual are not
seasonally consistent relative to each other. Past dietary studies during the chick rearing period
25
using stomach content analyses of Adélie penguins have found that their diets comprise
primarily Antarctic Krill, with very low percentages of fish (Volkman et al. 1980; Karnovsky,
1997). Our results are focused at two other times of the year when adults are not provisioning
chicks, late incubation and post-breeding, and when combined with their CV suggests that
Adélie penguins in the South Shetland Islands may have a more diverse diet that is consistent
between these two seasons. This compliments a prior study by Polito et al. (2011b) in which they
found that Adélie penguins in the South Shetland Islands have a heterogeneous diet of fish and
krill during the pre-breeding seasons. Adélie penguins demonstrated a type 1 pattern for prey
δ13C in which population foraging habitat use is not seasonally consistent and individual are also
not seasonally consistent relative to each other. The low values of non-breeding season prey δ13C
and relatively high CV suggest that Adélie penguins disperse widely and forage in more
pelagic/offshore habitats, as δ13C values are generally lower in offshore pelagic systems than
inshore benthic systems (France 1995, Hobson & Cherel 2011). However, relatively higher
breeding season δ13C values and lower CV suggest Adélie penguins forage in more constrained
and relatively more inshore locations at that time. This might be evidence that adults are
restricted to inshore and/or shallow locations during incubation and chick rearing, but have more
foraging range flexibility outside the breeding season once chicks have fledged, a pattern also
observed in diet analyses and tracking studies (Trivelpiece et al. 1987; Polito et al. 2015).
Chinstrap Penguins demonstrated a type 1 pattern for TL, suggesting that population diet
is not seasonally consistent, and individuals are not seasonally consistent relative to each other.
However, Chinstraps TL CVs were the lowest observed and the range of TL values was less than
one quarter of a trophic level for both breeding and non-breeding seasons (Table 2.2, Figure 2.2).
This suggests that the Chinstrap Penguin population has very low variation between individuals.
26
TL values indicate that all individuals forage within one quarter of a trophic level from one
another. As prey items such as Antarctic krill and fish differ by at least one trophic level, this
would suggest all individuals have a relatively similar diet (Polito et al 2011). Therefore,
Chinstrap Penguins likely consistently forage within the same general trophic level between
seasons, even though population may shift within the trophic level overall. Our findings are
consistent with many studies that have found Chinstrap Penguin diets in the South Shetland
Islands are composed primarily of Antarctic krill during the chick rearing season (Volkman et al.
1980; Trivelpiece et al. 1987; Miller et al. 2008; Miller et al. 2010; Polito et al. 2015). Chinstrap
Penguin exhibited a type 1 pattern for prey δ13C, suggesting population foraging habitat use is
seasonally consistent and individual are not seasonally consistent relative to each other.
Combined with a large range in non-breeding season prey δ13C values and a relatively large CV
value, Chinstrap Penguins appear to have considerable seasonal variability in individual foraging
habitat, which is supported by a previous study of both tracking and δ13C data that found
Chinstrap Penguins at King George Island exhibit individual variation in movement patterns and
the population generally occupies a broad geographic range of foraging habitats (Hinke et al
2015). Although results from the Wilcoxon signed rank test suggest the Chinstrap population
does not shift its foraging habitat seasonally, prey δ13C values appear to reflect the same pattern
seen in Adélie penguins. Breeding season prey δ13C values are substantially narrower and less
variable than prey δ13C non-breeding values, suggesting a similar foraging restriction due to
incubation/chick-rearing responsibilities during the breeding season.
Gentoo Penguins exhibited a type 3 pattern for TL, suggesting that population diets are
seasonally consistent, and individual are seasonally consistent relative to each other. This is
supported by one study investigating seasonal consistency of diet in a population of Gentoo
27
Penguins at King George Island, South Shetland Islands, in which δ15N values of Gentoo
Penguins during the non-breeding season not differ from the following breeding season (Juàres
et al. 2016). However, this previous study did not examine consistency at the individual level.
Gentoo Penguins also demonstrated a type 3 pattern for prey δ13C, which indicates that
population foraging habitat is seasonally consistent, and individual are seasonally consistent
relative to each other. Similar to Adélie and Chinstrap Penguins, examination of prey δ13C values
reveals a slight truncation during the breeding season, which may also suggest some foraging
restrictions due to incubation/chick-rearing responsibilities.
Foraging strategies of Pygoscelis penguins at King George Island
Overall, our results indicate that Chinstrap Penguins are the most specialized in their diet
compared to both Adélie and Gentoo Penguins, with little variation between individuals. In
contrast, Adélie penguins appear more generalist in diet relative to Chinstrap Penguins but less
generalist than Gentoo Penguins, and individuals do not specialize due to a lack of individual
consistency between seasons. Furthermore, our results suggest that Gentoo Penguins may be
dietary generalists at the population level due to a higher degree of individual variation. Our
results also indicate that individual Gentoo Penguins may be more specialized within the
population, as they exhibit individual consistency between seasons. Though generalist
populations comprising individual specialists have been documented in other top marine
predators, no previous studies have examined this in a penguin species (Woo et al. 2008;
Newsome et al. 2009; Hückstädt et al. 2012; Kernaleguen et al. 2015).
28
Comparison of seasonal consistency and individual variation among Gentoo populations:
Even though we found evidence of seasonal and individual consistency in diet and
foraging habitat use of Gentoo Penguins at King George Island, these patterns varied
considerably between Gentoo Penguins populations in the Western Antarctic Peninsula and the
South Shetland Islands. Comparisons among the three Gentoo populations provide support for
differences in the degree of individual variation in diet and foraging habitat. Assessments of CV
values between breeding locations indicated that Gentoo Penguins at Elephant Island
demonstrate higher individual variation in TL relative to Gentoo Penguins at Wiencke and
Rongé Island during both the breeding and non-breeding periods (Table 2.2; Figure 2.3).
Elephant Island’s TL CVs were more than twice as high as the Wiencke and Rongé populations,
indicating a wider dietary diversity and variation between individuals when compared to the
Western Antarctic Peninsula populations. A comparison of prey δ13C values CV among the
populations suggest that all populations demonstrate a similar level of variation in foraging
habitat use during the non-breeding season, while Gentoo Penguins at Wiencke Island exhibit a
higher degree of foraging habitat use variation during the breeding season relative to Gentoo
Penguins at Rongé and Elephant Island (Table 2.2; Figure 2.3)
The Gentoo Penguin population at Wiencke Island exhibited a type 1 pattern for TL,
suggesting that the population’s diet is not seasonally consistent, and that individuals were not
seasonally consistent relative to each other. However, the t-test result for this population was
only weakly significant (p=0.04). Furthermore, TL range values and CVs, as well as individual
patterns look noticeably similar to the population at Rongé Island, which exhibited a type 2
pattern for TL, suggesting that population diet is seasonally consistent, and individuals are not
seasonally consistent relative to each other (Table 2.3; Figure 2.3). These two populations
29
appear to have similar dietary trends, with no individual consistency between seasons. Elephant
Island demonstrated a type 4 pattern for TL, suggesting the population diet is not seasonally
consistent, and individuals are consistent relative to each other. All three populations exhibited a
type 2 for prey δ13C, suggesting population foraging habitats shifts seasonally, and individuals
are not consistent relative to each other. Overall prey δ13C values were higher during the
breeding season at all sites, similar to those at King George Island, suggesting individuals are
also restricted to more inshore and/or shallow locations during incubation.
The within-species variation in foraging strategy we observed may be due to regional
differences in productivity and prey availability. Studies on Galapagos sea lions (Zalophus
wollebaeki) found a high degree of variation in foraging strategies and the level of individual
specialization between populations throughout the Galapagos Archipelago whose habitats vary in
prey availability and physical oceanographic characteristics (Villegas-Amtmann et al. 2008;
Salazar, 2005; Dellinger and Trillmich, 1999; Kooyman and Trillmich, 1986). It is possible that
prey abundance and availability may differ between the foraging habitats in the Western
Antarctic Peninsula and South Shetland Islands. The Western Antarctic Peninsula is known for a
relatively year round abundance of Antarctic krill (Lascara et al. 1999). Gentoo Penguin
populations at Wiencke and Rongé Island are most likely foraging primarily on krill, which
would explain their relatively narrow variation in TL. The South Shetland Islands are in close
proximity to an ocean convergence zone, resulting in the Southern Antarctic Convergence Front
(SACCF) and Southern Boundary of the ACC (SB) (Figure 2.1). These fronts are associated with
high densities of Antarctic krill, but with high variability between seasons (Dietrich et al. 2014).
Variability in the abundance and availability of krill may select for greater diversity in diet
between individual penguins in order to reduce competition, which could explain the high
30
diversity in TL for Gentoo Penguins at Elephant Island as well as King George Island. This also
suggests that studies examining multiple Adélie and Chinstrap Penguins populations in the
Western Antarctic Peninsula and South Shetland Islands are merited to determine if similar
trends in within-species variation in foraging strategies exist.
Stable isotope analysis in Pygoscelis penguins
Because there are no published discrimination factors for Pygoscelis penguin blood, we
used δ15N and δ13C discrimination factors for whole blood derived from captive studies of four
piscivorous birds for our TL and prey δ13C value calculations, as suggested by Cherel et al.
(2005) and used in other studies on Pygoscelis penguins (Brasso et al. 2013, Juáres et al 2016). It
is possible that the absolute values of calculated TL and prey δ13C values would differ if speciesspecific discrimination factors were available. However, correlation results did not differ for
15
δ N and TL or δ13C and prey δ13C comparisons, suggesting that the observed relationships are
not an artifact of the chosen discrimination factors. Nonetheless, we suggest that future research
should aim to include controlled dietary studies of the target species in order to obtain more
15
refined δ N and δ13C discrimination factors for whole blood.
Conclusions
This study is the first to use stable isotope analysis of blood and feather of all three
Pygoscelis penguins to compare temporal foraging consistency within individuals and between
populations. Our results indicate species-specific differences at King George Island: Gentoo
Penguins exhibit a generalist foraging strategy with individual specialization, Adélie Penguins
exhibit an intermediate generalist foraging strategy with little individual consistency, and
Chinstrap Penguins exhibit a specialized diet with little variation between individuals. However,
31
unlike trophic variation, we found similarities in foraging habitat across all three species, with
greater variation in foraging habitat at the population level during the non-breeding season
compared to the breeding season. Finally, we found variation in trophic level between Gentoo
Penguin populations, with the two populations in the Western Antarctic Peninsula exhibiting a
narrow range of trophic level with little individual variation, relative to the population on
Elephant Island, South Shetland Islands, which had a large range in trophic level, and evidence
of individual foraging consistency. Future research on individual foraging specialization in
Pygoscelis penguins should focus on Gentoo Penguins across their larger breeding distribution to
identify the environmental mechanisms that may act to mediate the degree of individual variation
within these generalist populations. Identifying and understanding these mechanisms that drive
variation in foraging strategy in Pygoscelis penguins may aid in our understanding of how these
species will respond to recent climate driven changes in the Antarctic ecosystem.
LITERATURE CITED
Ainley, D. G., N. Nur, and E. J. Woehler. 1995. Factors affecting the distribution and size of
pygoscelid penguin colonies in the Antarctic. The Auk 112:171-182.
Ainley, D. 2002. The Adélie Penguin: Bellwether of Climate Change. Columbia University
Press.
Altabet, M. A., and R. Francois. 1994. Sedimentary nitrogen isotope ratio as a recorder from
surface ocean nitrate utilization. Global Biogeochemical Cycles 8:103-116.
Araújo, M. S., P. R. Guimaraes Jr, R. Svanbäck, A. Pinheiro, P. Guimarães, S. F. D. Reis, and D.
I. Bolnick. 2008. Network analysis reveals contrasting effects of intraspecific competition
on individual vs. population diets. Ecology 89:1981-1993.
Barquete, V., V. Strauss, P. G. Ryan. 2013. Stable isotope turnover in blood and claws: a case
study in captive African penguins. Journal of Experimental Marine Biology and Ecology
448: 121-127.
32
Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. Macleod. 2004. Determining trophic
niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology
73:1007-1012.
Bearhop, S., R. A. Phillips, R. McGill, Y. Cherel, D. A. Dawson, and J. P. Croxall. 2006. Stable
isotopes indicate sex-specific and long-term individual foraging specialisation in diving
seabirds. Marine Ecology Progress Series 311:157-164.
Bolnick, D. I., R Svanback, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L.
Forister. 2004. The ecology of individuals: Incidence and implications of individual
specialization. The American Naturalist 161:1-28.
Brinton, E., M. D. Ohman, A. W. Townsend, M. D. Knight, and A. Bridgeman. 1999.
Euphausiids of the world ocean. (Expert-center for Taxonomic Identification, Scripps
Institution of Oceanography, Eds.)., Amsterdam: University of Amsterdam, ETI.
Cherel, Y., and K.A. Hobson. 2007. Geographical variation in carbon stable isotope signatures of
marine predators: a tool to investigate their foraging areas in the Southern Ocean. Marine
Ecology Progress Series 329: 281–287.
Clausen, A. P., and K. Putz. 2002. Recent trends in diet composition and productivity of Gentoo,
Magellanic and rockhopper Penguins in the Falkland Islands. Aquatic Conservation:
Marine and Freshwater Ecosystems 12(1):51-61.
Clausen, A., and K. Pütz. 2003. Winter diet and foraging range of Gentoo Penguins (Pygoscelis
papua) from Kidney Cove, Falkland Islands. Polar Biology 26(1):32-40.
Croxall, J. P., and P. A. Prince. 1980. The food of Gentoo Penguins Pygoscelis papua and
Macaroni penguins Eudyptes chrysolophys at South Georgia. Ibis 122: 245-253.
Croxall, J. P., R. W. Davis, and M. J. O'Connell. 1988. Diving patterns in relation to diet of
Gentoo and macaroni penguins at South Georgia. Condor 90:157-167.
Dellinger T., and F. Trillmich. 1999. Fish prey of the sympatric Galapagos fur seals and sea
lions: seasonal variation and niche separation. Canadian Journal of Zoology 77:1204–
1216.
Donnelly, S. M., and A. Kramer. 1999. Testing for multiple species in fossil samples: an
evaluation and comparison of tests for equal relative variation. American Journal of
Physical Anthropology 108(4): 507-529.
Fligner, M. A., and T. J. Killeen. 1976. Distribution-free two-sample tests for scale. Journal of
the American Statistical Association 71(353): 210-213.
France, R. L. 1995. Carbon-13 enrichment in benthic compared to planktonic algae: foodweb
implications. Marine Ecology Progress Series 124:307–312.
33
Frédérich, B., O. Lehanse, P. Vandewalle, and G. Lepoint. 2010. Trophic niche width, shift, and
specialization of Dascyllus aruanus in Toliara lagoon, Madagascar. Copeia 2010: 218226.
Hobson, K. A., and R. G. Clark. 1992. Assessing avian diets using stable isotopes I: turnover of
13C in tissues. Condor 94:181-188.
Hobson, K. A., and A. L. Bond. 2012. Extending an indicator: year-round information on seabird
trophic ecology from multiple-tissue stable-isotope analyses. Marine Ecology Progress
Series 461: 233-243.
Hückstädt, L. A., P. L. Koch, B. I. McDonald, M. E. Goebel, D. E. Crocker, and D. P. Costa.
2012. Stable isotope analyses reveal individual variability in the trophic ecology of a top
marine predator, the southern elephant seal. Oecologia 169(2):395–406.
Jackson, A., R. Inger, A. Parnell, and S. Bearhop. 2011. Comparing isotopic niche widths among
and within communities: SIBER – Stable Isotope Bayesian ellipses in R. Journal of
Animal Ecology 30:595-602.
Jablonski, B. 1985. The diet of penguins on King George Island, South Shetland Islands. Acta
Zool Cracov 29:117–186.
Juáres, M. A. 2016. Diet composition and foraging habitats of Adélie and Gentoo penguins in
three different stages of their annual cycle. Marine Biology 163:105.
Jensen, J. K., P. L. Boveng, J. L. Bengston. 1998 Foraging modes of Chinstrap Penguins:
Contrasts between day and night. Marine Ecology Progress Series 165:161–172.
Karnovsky, N. J. 1997. The fish component of Pygoscelis penguin diets (Doctoral dissertation,
Montana State University-Bozeman, College of Letters & Science)
Kernaléguen, L, N. Dorville, D. Ierodiaconou, A. J. Hoskins, A. M. M. Baylis, M.A. Hindell, J.
Semmens, K. Abernathy, G. J. Marshall, Y. Cherel, and J. P. Y. Arnould. 2015. From
video recordings to whisker stable isotopes: a critical evaluation of timescale in assessing
individual foraging specialization in Australian fur seals. Oecologia 180(3):657-70.
Lishman, G.S. 1985 The food and feeding ecology of Adélie penguins (Pygoscelis adeliae) and
Chinstrap Penguins (Pygoscelis antarctica) at Signy Island, South Orkney Islands.
Journal of Zoology 205:245–263.
Lynch, H. J., R. Naveen, and P. Casanovas. 2013. Antarctic site inventory breeding bird survey
data, 1994-1013. Ecology 94:2653.
Lynnes, A.S., K. Reid, and J. P. Croxall. 2004. Diet and reproductive success of Adélie and
Chinstrap Penguins: Linking responses of predators to prey population dynamics. Polar
Biology 27:544–554.
34
Miller, A. K., M. A. Kappes, S. G. Trivelpiece, and W. Z. Trivelpiece. 2010. Foraging-Niche
Separation of Breeding Gentoo and Chinstrap Penguins, South Shetland Islands,
Antarctica. Condor 112(4):683–695.
Miller, A. K., N. J. Karnovsky, and W. Z. Trivelpiece. 2009. Flexible foraging strategies of
Gentoo Penguins Pygoscelis papua over 5 years in the South Shetland Islands, Antarctica.
Marine Biology 156:2527–2537.
Miller, AK, and W. Z. Trivelpiece. 2008. Chinstrap Penguins alter foraging and diving behavior
in response to the size of their principal prey, Antarctic krill. Marine Biology 154: 201208.
Minagawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further
evidence and the relation between δ15N and animal age. Geochimica et Cosmochimica
Acta 48(5):1135-1140.
Newsome, S. D., C. Martínez del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic
ecology. Frontiers in Ecology and Environment 5:429–436.
Polito, M.J., S. Abel, C. R. Tobias, S. D. Emslie. 2011a. Dietary isotopic discrimination in
Gentoo Penguin (Pygoscelis papua) feathers. Polar Biol 34(7):1057–1063.
Polito, M.J., S. Abel, H. J. Lynch, R. Naveen, and S. D. Emslie. 2011b. Stable isotopes reveal
regional heterogeneity in the pre-breeding distribution and diets of sympatrically
breeding Pygoscelis spp. Penguins. Marine Ecology Progress Series 421: 265-277.
Polito, M. J., W. Z. Trivelpiece, W. P. Patterson, N. J. Karnovsky, C. S. Reiss, and S. D. Emslie.
2015. Contrasting specialist and generalist patterns facilitate foraging niche partitioning
in sympatric populations of Pygoscelis penguins. Marine Ecology Progress Series 519:
332-337.
Putz, K., R. J. Ingham, J. G. Smith, and J. P. Croxall. 2001. Population trends, breeding success
and diet composition of Gentoo Pygoscelis papua, magellanic Spheniscus magellanicus
and rockhopper Eudyptes chrysocome penguins in the Falkland Islands. A review. Polar
Biology 24(11):793-807.
Rombola, E., E. Marschoff, and N. Coria. 2006. Interannual study of Chinstrap Penguin’s diet
and reproductive success at Laurie Island, South Orkney Islands, Antarctica. Polar
Biology 29:502–509.
Salazar, S. K. 2005. Variación temporal y espacial del espectrotrófico del lobo marino de
Galápagos. Master’s thesis, Instituto Politécnico Nacional, La Paz.
Trivelpiece, W.Z., S. G. Trivelpiece, and N. J. Volkman. 1987. Ecological segregation of Adélie,
gentoo, and Chinstrap Penguins at King George Island, Antarctica. Ecology 68: 351–361.
35
Trivelpiece, W.Z., J. T. Hinke, A. K. Miller, C. S. Reiss, S. G. Trivelpiece, and G.M. Watters.
2011. Variability in krill biomass links harvesting and climate warming to penguin
population changes in Antarctica. Proceedings to the National Academy of Science USA
108:7625–7628.
Villegas-Amtmann, S., D. P. Costa, Y. Tremblay, S. Salazar, and D. Aurioles-Gamboa. 2008.
Multiple foraging strategies in a marine apex predator, the Galapagos sea lion Zalophus
wollebaeki. Marine Ecology Progress Series 363:299-309.
Volkman, N. J., P. Presler, and W. Z. Trivelpiece. 1980. Diets of pygoscelid penguins at King
George Island, Antarctica Condor:373-378.
Wilson, R. P., B. M. Culik, P. Kosiorek, and D. Adelung. 1998. The over-winter movements of a
Chinstrap Penguin (Pygoscelis antarctica). Polar Record 34:107-112.
Wilson, R.P., B. Alvarrez, L. Latorre, D. Adelung, B. Culik, and R. Bannasch. 1998. The
movements of gentoo penguins Pygoscelis papua from Ardley Island, Antarctica. Polar
Biology 19:407–413.
36
CHAPTER 3:
GEOGRAPHIC VARIATION IN GENTOO PENGUIN (PYGOSCELIS
PAPUA) GENERALIST FORAGING STRATEGIES THROUGHOUT THE
SCOTIA ARC
INTRODUCTION
While traditionally ecological studies have treated conspecific individuals as functionally
equivalent, there is growing evidence to suggest that individual specialization and individual
niche variation is widespread with important implications for ecology and evolutionary dynamics
(Bolnick et al. 2003, Sargeant 2007). For example, specialist individuals can have an advantage
over generalist individuals within populations if they are better able to exploit a particular highprofit resource. Pigeon guillemots (Cepphus columba) that specialized on lipid rich fish species
have higher fledging rates than their generalist conspecifics (Golet et al. 2000). Likewise,
spectacled salamanders with specialized foraging niches (Salamandrina perspicillata)
demonstrate better physiological condition than individuals with a broader trophic foraging niche
(Costa et al. 2015).
Individual specialization and individual niche variation is frequently found within
generalist populations (Bolnick et al. 2003; Sargeant 2007). Identifying individual specialization
within generalist populations requires determining the degree of foraging variation within
individuals relative to variation among individuals, and if these variations persist through time.
Using this apporoach, generalist populations can be broadly grouped into two categories: Type A
generalist populations exhibit extensive niche variation within individuals and little niche
variation between individuals, whereas type B generalist populations are composed of
individuals with specialized niches, that are smaller than the population niche as a whole
(Bolnick et al. 2003; Bearhop et al. 2004). Due to the ecological importance of individual niche
37
specialization there is a growing interest to better understand what factors drives individuals
within generalist populations to specialize. There is evidence to suggest that individual
specialization may be a density-dependent strategy, in which individuals restrict their foraging
niche to reduce intraspecific competition as population size increases and prey resources
decrease (Parent et al. 2014). For example, Svanbäck and Bolnick (2004) found that increases in
population density of Eurasian perch (Perca fluviatilis) were accompanied by an expansion of
the population’s total niche width, while within-individual niche width remained small and
specialized on a single habitat. Therefore, individual specialization and niche variation within a
population may actually act as a buffer against environmental pressures affecting prey resources
and habitats by enabling a population to adapt to such changes through individual foraging
diversity (Durell 2000; Bolnick et al. 2003).
Stable isotope analysis (SIA) provides an effective way to assess the diet and foraging
niches of individuals through time (Bearhop et al. 2004; Newsome et al. 2007, Jaeger et al. 2009,
Vander Zaden et al. 2010; Hückstädt et al. 2012, Kernaleguen et al. 2015). Nitrogen isotopic
values (δ15N) increase with trophic level and can be used as a proxy of a consumer’s diet, while
carbon isotopic values (δ13C) in marine consumers vary with latitude and can be used to indicate
differences in foraging habitat with respect to inshore versus offshore locations (France 1995;
Cherel and Hobson, 2007). Together, δ15N and δ13C values can be used to determine the 2dimensional parameters of an isotopic niche space, a comparable measurement of an ecological
or foraging niche (Newsome et al. 2007). Comparing isotopic niches between and within species
has allowed researchers to identify specialist vs. generalist foraging strategies at both the
population (Polito et al. 2015) and individual level (Newsome et al. 2009). However, while
informative, studies of individual foraging specialization often focus on a single population or
38
are limited in spatial or temporal scope (Vander Zaden et al. 2010; Hückstädt et al. 2012;
Kernaleguen et al. 2015).
To our knowledge, no past studies have examined individual foraging specialization
across a species’ distributional range to determine if populations can exhibit either type A or type
B generalist strategies in response to differing environmental and density dependent conditions.
We seek to address this knowledge gap, by investigating foraging niche specialization in four
geographically distinct Gentoo Penguin (Pygoscelis papua) populations spanning the latitudinal
extent of their breeding range. Gentoo Penguins are an ideal model species for a large-scale
comparative study of individual specialization across generalist populations due to their large
breeding distribution and evidence as generalist foragers at the population level (Miller et al.
2009; Lynch 2013; Polito et al. 2015). While there is evidence of considerable variation in this
species’ population level foraging niches across breeding populations (Croxall et al, 1988;
Wilson et al, 1998; Miller et al. 2009; Putz et al. 2001; Clausen and Putz; and Polito et al. 2015),
past studies have not assessed individual diet and foraging consistency through time, which is
necessary for a robust assessment of individual specialization (Bolnick et al, 2003; Bearhop et al.
2004, Jaeger et al 2009). Furthermore, there are large latitudinal differences in factors such as
population sizes, physical oceanographic features, and prey assemblage and abundance across
this species breeding range, which are likely to influence the use of type A vs. type B generalist
strategies.
We hypothesize that northern populations of Gentoo Penguins that are relatively large in
size and whose foraging habitat comprises a high diversity of prey resources will favor
individual foraging specialization (type B strategies) and lower niche overlap between
individuals. In contrast, smaller southern populations whose prey resources are dominated by
39
Euphaisa superba (Antarctic krill) are more likely to exhibit type A generalist strategies with
greater niche overlap between individuals. To test these hypotheses, we examine isotopic niche
metrics within four geographically isolated Gentoo Penguin populations at both the population
and individual levels to quantify the degree to which individual niches are specialized relative to
their population’s total niche (i.e. type A vs. type B generalists; Bearhop et al. 2004; Sargeant
2007). In addition, we compare metrics of individual isotopicniche overlap within the four
populations to reveal the degree to which individuals differentiate their foraging niches relative
to one another within each population (Sargeant 2007).
METHODS
Study area and sample collection
Our research focuses on Gentoo Penguin breeding colonies from four geographically
isolated regions in the Scotia Arc: Cuverville Island in the Western Antarctic Peninsula
(64.6875°S, 62.6219°W); Cape Shirreff on Livingston Island in the South Shetland Islands
(62.4589°S, 60.7886°W); Start Point on South Georgia (54.0472°S, 37.3564°W); and Cow Bay
on the Falkland Islands (51.4335°S, 57.8564°W) (Figure 3.1). These oceanographic regions
differ considerably both physically and ecologically (Table 3.1). The Western Antarctic
Peninsula and South Shetland Islands are found just south of the Southern Antarctic Circumpolar
Convergence Front (SACCF) and Southern Boundary of the ACC (SB), where there is an
abundance of Antarctic krill (Euphaisa superba), a primary food source for Pygoscelis penguins
(Volkman et al. 1980; Miller et al. 2008, 2010). South Georgia is located north of the SACCF
and within the flow of the ACC, which transports Antarctic krill from areas around the Antarctic
continent, resulting in high krill abundance in the surrounding waters as well. However, this
influx recruitment can be ephemeral and vary considerably from year to year (Murphy et al.
40
2013). The Falkland Islands are found in shallower warmer waters north of the Polar Front and
the Sub-Antarctic Front (SAF) where there is no Antarctic krill, but a wide diversity of prey
items (Clausen and Pütz, 2003). In addition, Gentoo Penguin population sizes are much larger in
the Falkland Islands and South Georgia at the northern extent of their range relative to the South
Shetland Islands and Western Antarctic Peninsula at the southern extent of their range (Table
45°
3.1).
50°
SAF
PF
Falkland Islands
South Georgia
South Shetland
Islands
Western Antarctic
Peninsula
70°S
65°
60°
55°
AC
C
70°W
60°
50°
40°
30°
Figure 3.1: Map of the four Gentoo Penguin breeding locations sampled in the study. Major
currents and fronts are indicated by dotted lines: Sub-Antarctic Front (SAF), Polar Front (PF),
and Antarctic Circumpolar Convergence (ACC).
41
Table 3.1: Summary of the differences in ecological and physical characteristics between the
four major study regions. Southern Antarctic Circumpolar Convergence Front (SACCF), Polar
Front (PF), Sub-Antarctic Front (SAF). Prey resources based on previous dietary studies of
Gentoo Penguins in each region.
Ecological &
Physical
Characteristic
Western Antarctic South Shetland
Peninsula
Islands
South Georgia
Falkland Islands
Population Size
63,897
36,450
98,867
132,321
Currents and
Fronts
South of SACCF
South of
SACCF
Between
SACCF and
PF
North of PF and
SAF
Seasonal Sea Ice
Presence
yes
yes
no
no
Antarctic krill
fish (Polito et al.
2011)
Antarctic krill
fish (Volkman
et al. 1980)
Antarctic krill
fish
squid (Croxall
and Prince,
1980;
Williams,
1991)
Cephalopods
fish
lobster krill
(Clausen and Putz,
2003
Prey Resources
In the austral summer of 2011/12, we captured 16-20 individual breeding adults from
active nests at each colony. Sex was not determined. For each individual we collected a single
central tail feather (i.e. the longest), which was stored at room temperature prior to analysis.
Gentoo penguins tail feathers are grown over a period of approximately 100 days during the
post-molt season. They are metabolically inert and can be sampled for SIA at discrete intervals
to reflect the early winter diets of penguins throughout the tail feather growth period and allow
for serial sampling to detect temporal variability or consistency in foraging (see Appendix 1; also
Hinke et al. 2015).
42
Sample preparation and isotopic analysis
We cut tail feathers into 1cm segments and rinsed segment in 2:1 chloroform:methanol
and allowed them to air dry for 24 hours (Cherel et al. 2005). We subsampled 0.5mg of feather
vane material from four 1cm sections between 5.5-6.5cm, 7.5-8.5cm, 9.5-10.5cm, and 11.512.5cm for a total of four sections per individual feather. This sampling design allowed for the
analysis of individual penguin isotopic niche during the late fall /early winter period at during
four discrete 5-7 day long foraging periods distributed across 33 days of continuous feather
growth during the post-molt, non-breeding season (Appendix A). Samples were analyzed using a
PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ Europa 20-20 isotope ratio
mass spectrometer. Results are in the format of stable isotope ratios using del notation in per mil
units derived from the equation: δX = [(Rsample/ Rstandard) − 1] × 1000. Rstandard is the ratio 13C/12C
or 15N:14N. Rstandard values are based off of Vienna PeeDee Belemnite for δ13C and atmospheric
N2 for δ15N. Raw δ values were normalized using glutamic acid, bovine liver, and nylon 5 as
reference material (USGS-40: δ13C = -16.65‰, δ15N = -6.8‰; USGS-41: δ13C = -37.63‰, δ15N
= 47.6‰; bovine liver: δ13C = -21.69‰, δ15N = 7.72‰; nylon 5: δ13C = -27.72‰, δ15N = 10.31‰). Sample precision based on internal repeats and duplicate standard reference materials
was 0.1‰ for both δ15N and δ13C.
Niche analysis at the population level
For population level analysis the isotopic values (δ15N and δ13C) of tail feather sections
were first averaged the four feather values within individuals to produce a single mean isotopic
value per individual. We used individual mean values for all of the population level comparisons
described below. We compared the mean individual isotopic values between populations for both
δ15N and δ13C values using one-way ANOVAs and Tukey's HSD (honest significant difference)
43
test to examine differences in all pairwise comparisons of populations. To test for variation in
population-level isotopic niche size, we used the R package SIAR to calculate standard ellipse
areas corrected for sample size (SEAC) for each population, which provide a measure of the 2dimensional core isotopic niche area of a population (Jackson et al. 2011). We then compared all
posterior draws from the Bayesian approximation of this metric (SEAb) between populations to
test for significant differences in core isotopic niche area. We also constructed convex hulls for
each population, which act as proxies for total niche area (TA; Layman & Allgeier 2012).
Finally, we calculated each population’s mean distance to centroid (MDC) and mean nearest
neighbor distance (NND), which can be interpreted as a metric for a population’s trophic
diversity and the density of individual packing within a population (Laymen et al. 2007, Turner
et al. 2010)
Niche analysis at the individual level
For individual level analysis of isotopic niches we used the δ15N and δ13C values of all
individual tail feather sections from each individual. We first calculated the degree of individual
specialization found in each population using the R-package RInSp (Zaccarelli et al. 2013). This
package applies the method of Bolnick et al. (2002) to calculate an individual specialization
index for a population using total niche width index for continuous data proposed by
Roughgarden (1972). Specifically, the total niche width of a population (TNW) can be
partitioned into two components: the variation within individuals, or the within-individual
component (WIC), and the variation between-individual in a population, or the betweenindividual component (BIC) so that TNW = WIC + BIC (Roughgarden, 1972). The degree of
individual specialization of a population is represented by the ratio of WIC/TNW so that ratios
approaching zero suggest high degrees of individual specialization (Type B population) and
44
ratios approaching one suggest lower degrees of individual specialization or more generalism at
the individual level (Type A population). For each population we calculated separate TNW,
WIC, BIC and WIC/TNW metrics based on tail feather section δ15N and δ13C values. We
assessed significance of WIC/TNW using a nonparametric Monte Carlo technique to generate
10,000 replicate datasets under the null hypothesis that all individuals are generalists (Type A
population) and used these distributions to calculate P-values (Zaccarelli et al. 2013).
To compliment the above univariate analyses, we also calculated the total isotopic niche
areas (TA) for each individual using their respective δ15N and δ13C values of individual tail
feather sections. We divided individual’s TA values by their respective population’s TA to
obtain an index of relative niche area to quantify the degree to which an individual’s niche is
specialized relative to the population. Values approaching zero indicate individuals have smaller
isotopic niches than their population’s total niche and thus a higher degrees of individual
specialization (Type B population). In contrast, values approaching one indicate individual’s
isotopic niche is similar in size to their population’s total niche and thus a more generalist at the
individual level (Type A population). In addition we used TA to obtain an index of relative
niche overlap to quantify the degree to which individuals within populations are specialized
relative to each other (Sergeant, 2007). We calculated relative niche overlap as the number of
individual TAs that overlapped with any single individual’s TA, divided by the number of total
individuals within each population. Values approaching zero suggest a low degree of overlap
between individual’s niches with individuals differentiating in resource use, while values
approaching one indicate greater overlap between individual’s niches and the use of common
resources among individuals. We then used ANOVA (for normally distributed data) or the
Kruskal-Wallis H test (for non-normally distributed data) to test for differences in individual’s
45
relative niche area and relative niche overlap across the Gentoo Penguin populations. All
statistical analyses were performed within the program R (Ver. 3.2.1; R Core Team 2015).
RESULTS
Population Level Niche Metrics
Tail feather δ15N and δ13C values differed significantly among populations (Table 3.2;
δ15N: F3,296 = 2035, p < 0.0001, δ13C: F F3,296 = 5635, p < 0.001). All four populations differed
significantly from one another for both δ15N and δ13C and did not overlap. Isotopic values were
correlated with the latitudinal position of each population, with the Western Antarctic Peninsula
having the lowest δ15N and δ13C values, followed by the South Shetland Islands, South Georgia,
and the Falkland Islands (Figure 3.2). Trophic diversity measured by MDC was significantly
higher in South Georgia compared to all other sites (p <0.05). MDC was also significantly higher
in the South Shetland Islands compared to the Western Antarctic (p = 0.014). Mean nearest
neighbor distance NND was significantly lower in the Western Antarctic Peninsula than South
Georgia (p = 0.009). Total and core isotopic niche areas also differed across sites (Table 3.3).
South Georgia had the largest total niche area (TA) followed by the South Shetland Islands, the
Falkland Islands and the Western Antarctic Peninsula. Similarly, core niche area (SEAb) showed
the same spatial pattern as TA with significantly larger SEAb in South Georgia relative to the
Falkland Islands and the Western Antarctic Peninsula (p < 0.01, p <0.001, respectively) and
significantly larger SEAb in the South Shetland Islands relative to the Western Antarctic
Peninsula (p<0.01).
46
Table 3.2: Individual Specialization niche parameters for δ15N and δ13C per region. WIC = within-individual component, BIC =
between-individual component, TNW = total niche width, WIC/TNW = ratio approximating degree of individual specialization within
populations.
Isotope
values
δ15N
δ13C
Population
n
Western Antarctic
20
Peninsula
South Shetland
19
Islands
Mean ± SD
CV (%)
WIC
BIC
TNW
WIC/TNW
p-value
9.0 ± 0.3 (8.3 to 9.8)
3.3
0.045
0.037
0.083
0.55
0.003
9.6 ± 0.6 (8.5 to 11.8)
6.3
0.19
0.21
0.39
0.47
0.001
South Georgia
16
12.0 ± 1.2 (10.2 to 15.2)
10.0
0.36
1.02
1.38
0.26
0.001
Falkland Islands
20
16.8 ± 0.5 (15.7 to 18.5)
3.0
0.09
0.18
0.27
0.35
0.001
24.7 ± 0.4 (-25.8 to -23.9)
2.0
0.09
0.04
0.12
0.67
0.11
23.3 ± 0.7 (-24.8 to -22.0)
3.0
0.28
0.22
0.5
0.56
0.003
Western Antarctic
20
Peninsula
South Shetland
19
Islands
South Georgia
16
18.7 ± 0.7 (-20.4 to -17.2)
3.7
0.15
0.31
0.45
0.33
0.001
Falkland Islands
20
14.3 ± 0.4 (-15.3 to -13.4)
2.8
0.09
0.11
0.2
0.47
0.001
47
20
18
14
12
South
Georgia
South
Shetland Islands
Western
Antarctic Peninsula
6
8
10
δ
15
N (‰)
16
Falkland
Islands
−26
−24
−22
−20
−18
−16
−14
−12
13
δ C (‰)
Figure 3.2: Standard ellipse areas and convex hull areas (TA) representing niches of each
population. Points represent individual averaged δ15N and δ13C values. Black = Western
Antarctic Peninsula, red = South Shetland Islands, green = South Georgia, blue = Falkland
Islands.
Individual niche area
WIC/TNW for δ15N was highest in the Western Antarctic Peninsula followed by the
South Shetland Islands and the Falkland Islands and lowest at South Georgia (Table 3.2). All
four regions had WIC/TNW based on δ15N values consistent with significant individual
specialization within populations relative to a null distribution (p<0.005).
48
Table 3.3: Population level metrics evaluating differences in bivariate niche parameters. TA =
total niche area based on individual averages, MDC = mean distance to centroid, NND = mean
nearest neighbor distance, Mean SEAb = mean of posterior probability distributions for Bayesian
standard ellipse area estimates including 95% credibility intervals. Groups that do not share at
least 1 superscripted letter within a column are significantly different for the given variable at the
0.05 level. Ranges of values are in brackets.
Population
TA
MDC
NND
SEAb
Western Antarctic
Peninsula
0.39
0.26a
0.10a
0.49 [0.34-0.70] a
South Shetland
Islands
1.82
0.61b
0.23ab
1.05 [0.71-1.52] bc
South Georgia
2.06
0.97
0.32b
1.69 [1.12-2.47] b
Falkland Islands
1.11
0.44ab
0.18ab
0.75 [0.52-1.07] ac
WIC/TNW for δ13C were also the highest in the Western Antarctic Peninsula, followed by the
South Shetland Islands, Falkland Islands, and South Georgia (Table 3.4). The South Shetland
Islands, Falkland Islands, and Georgia showed significant individual specialization for δ13C
WIC/TNW relative to null distributions (p<0.005), while the Western Antarctic Peninsula did
not. We found no significant differences in individual’s relative niche area among Gentoo
Penguin populations. All four populations had mean relative niche area values below 7% (0.321.0%; Table 3.4).
Individual niche overlap
We found significant differences in relative niche overlap between populations (Figure
3.3; Table 3.4; F3/71 = 11.84, p < 0.0001). Western Antarctic Peninsula population had a
significantly higher relative niche overlap compared to all other sites (Table 3.4; South Shetland
Islands, p < 0.0001; South Georgia, p = 0.004; Falkland Islands, p = 0.006).
49
Table 3.4: Individual relative niche and overlap metrics based on total niche area (TA). Population TA = total niche area based on all
individual values. Asterisk (*) indicates a significant difference from all other populations for the given variable at the 0.05 level.
Ranges of values are in brackets.
Population
Population TA
(‰2)
Individual TA (‰2)
Mean ± SD
Relative niche area (%)
Mean ± SD
Relative niche overlap
Mean ± SD
Western Antarctic
Peninsula
2.18
0.15 ± 0.05 [0.01-0.35]
6.7 ± 4.5 [0.3-16.1]
0.66 ± 0.16* [0.25-0.85]
South Shetland
Islands
5.83
0.35 ± 0.05 [0.05-0.95]
5.9 ± 4.0 [0.9-16.3]
0.31 ± 0.08 [0.16-0.42]
South Georgia
5.66
0.31 ± 0.04 [0.04-0.97]
6.1 ± 5.7 [0.7-17.1]
0.44 ± 0.23 [0.06-0.69]
Falkland Islands
3.04
0.19 ± 0.06 [0.03-0.64]
5.5 ± 4.0 [1.0-21.0]
0.46 ± 0.22 [0.00-0.75]
50
13
9
8
7
−26
−25
−24
13
−23
−22
−21
−26
−25
−24
C (‰)
20
δ
South
Georgia
−23
13
−22
−14
−13
−21
−20
−12
−11
C (‰)
Falkland
Islands
18
17
14
10
15
11
16
12
δ
δ
15
13
14
N (‰)
19
15
16
δ
N (‰)
11
12
15
δ
8
7
6
−27
15
South Shetland
Islands
10
N (‰)
10
9
δ
15
N (‰)
11
12
Western Antarctic
Peninsula
−22
−21
−20
−19
δ
13
−18
−17
−16
−17
C (‰)
−16
−15
δ
13
C (‰)
Figure 3.3: Convex hull areas (TA) for each individual within the four sample populations.
Dotted line represents TA for entire population.
DISCUSSION
We originally hypothesized that Gentoo Penguins populations in the northern Scotia Arc,
which are relatively larger in size and have a higher diversity of prey resources, will favor
individual foraging specialization (type B strategies) and lower niche overlap between
individuals. Conversely, we hypothesized Gentoo Penguins populations in the southern Scotia
51
Arc, which are relatively smaller in size and have prey resources dominated by Antarctic krill
(Euphaisa superb), would exhibit type A generalist strategies with higher niche overlap between
individuals. Contrary to these hypotheses, our results indicate that all four Gentoo Penguins
populations we examined exhibited clear evidence of individual foraging specialization (e.g. type
B generalist foraging strategies) based on their isotopic niches. However, consistent with our
predictions, the relative degree of individual niche specialization varied between populations.
Gentoo Penguin populations in the north had a higher degree of individual niche specialization
relative to southern populations and the southernmost population examined (Western Antarctic
Peninsula) exhibited the highest degree of foraging niche overlap among individuals.
Population-level isotopic niches
The linear trend observed in both δ15N and δ13C values corresponding to the latitudinal
position of each population (the lowest being the Western Antarctic Peninsula, followed by the
South Shetland Islands, South Georgia, and the Falkland Islands) is consistent with the
latitudinal gradient in δ15N and δ13C due to baseline differences in productivity within frontal
zones in which both isotopic values increase from south to north (Francois et al. 1993; Altabet
and Francois, 1994). These gradients create a geographical isoscape that has been documented in
other seabird species using stable isotope analysis (Quillfeldt et al. 2005, Cherel and Hobson
2007; Jaegar et al. 2010)
MDC results suggest a larger population niche in South Georgia than all other
populations, and a smaller population niche in the Western Antarctic Peninsula than in the South
Shetland Islands (Table 3.2). These results are further supported by SEAb values, where the core
isopotic niche area in the South Georgia population is significantly larger than the Falkland
Islands and the Western Antarctic Peninsula, and the Western Antarctic Peninsula is significantly
52
smaller than the South Shetland Islands. NND results suggest that individual niches are farther
apart from one another in South Georgia than in the Western Antarctic Peninsula. Population
niche size appears to be positively correlated with a greater degree of individual niche diversity,
which may act to reduce intraspecific competition. Our results agree with several empirical
studies on fish species where high population density was associated with greater variation in
individual niche diversity (Svanbäck & Persson 2004; Svanbäck et al. 2008; Araújo et al. 2008;
Svanbäck & Persson 2009; Frederich et al. 2010).
Individual niche specialization relative to the population
Contrary to our hypotheses, we found consistent evidence of individual specialization and
type B generalism within all four of our study populations. Individual specialization indices of
δ15N suggest that individuals in all populations tend to specialize trophically, or on particular
prey items, consistent with a type B generalist strategy. Furthermore, all four populations had
mean individual relative niche areas that were below 0.07, meaning that individuals in all
populations on average utilize under 7% of the total population niche, indicative of type B
generalism (Table 3.4). The degree of specialization, however, varies by population. Individual
specialization indices of δ13C suggest Gentoo Penguin populations found in the South Shetland
Islands, South Georgia, and Falkland Islands specialize in foraging location, consistent with type
B generalism, whereas individuals in the Western Antarctic Peninsula population do not (Table
3.2). In addition, both individual specialization indices of δ15N and δ13C exhibited the same trend
of increasing specialization with the lowest in the Western Antarctic Peninsula, followed by the
South Shetland Islands, the Falkland Islands, and South Georgia. This trend somewhat reflects
our hypotheses with regards to the northern populations with larger populations and wider
diversity of prey resources (South Georgia and the Falkland Islands) being more specialized
53
compared to the smaller southern populations with dominant prey sources (Western Antarctic
Peninsula and South Shetland Islands). Therefore, although Gentoo Penguins appear to be type B
generalists overall, differences in the degree of individual specialization observed in populations
of this species may be driven by geographic differences in prey availability and population size.
Studies of Galapagos sea lions (Zalophus wollebaeki) throughout the Galapagos
Archipelago suggest a high degree of variation in foraging strategies and individual
specialization between populations driven by variation in prey availability and physical
oceanographic characteristics (Villegas-Amtmann et al. 2008; Salazar, 2005; Dellinger and
Trillmich, 1999; Kooyman and Trillmich, 1986). The waters surrounding the Western Antarctic
Peninsula and South Shetland Islands are known for high abundance of Antarctic krill due to the
presence of sea ice, which is crucial for the juvenile life stage of krill (Ducklow et al. 2007). It is
possible that the presence of a highly available dominant prey source may exert less pressure on
individuals to specialize, which would explain the lowest degrees of individual specialization we
found in these regions, particularly in the Western Antarctic Peninsula population. Previous
dietary studies support these findings that Gentoo Penguin diets are dominated by Antarctic krill
in these regions (Miller et al. 2008, 2010). On the contrary, dietary studies of Gentoo Penguins
in South Georgia have found variation in both prey choice and foraging habitat in this region.
(Tanton et al. 2004; Williams et al 2008; Croxall and Prince 1980). Waters around South
Georgia are also known to be abundant with krill, but this is highly variable both seasonally and
year to year, which may act as a strong selective pressure for individuals to specialize given the
ephemeral nature of prey abundance (Tarling et al. 2007). Therefore, individuals may have to
rely on other prey items during times of low krill abundance, resulting in a wider population
niche and higher levels in individual specialization, as indicated in our results.
54
Population size also appears correlated with degree of individual specialization, as
individuals in larger populations have been shown to become more specialized, while their
population niche increases (Svanbäck and Bolnick 2004; Svanbäck and Persson 2004; Parent et
al. 2014). Both the Western Antarctic Peninsula and South Shetland Islands have significantly
smaller populations of Gentoo Penguins with corresponding lower levels of individual
specialization found in our study. South Georgia and the Falkland Islands have the largest
population sizes, with corresponding high indices of individual specialization. This supports
previous studies that found increases in population size results in larger population niches and
higher degrees of individual specialization to reduce intraspecific competition (Svanbäck &
Persson 2004; Svanbäck et al. 2008; Svanbäck & Persson 2009; Frederich et al. 2010). Our
results provide evidence to suggest that individual specialization relative to the population may
be a density-dependent strategy, in which individuals restrict their foraging niche to reduce
intraspecific competition as population size increases and prey resources decrease.
Individual niche specialization relative to other individuals
We hypothesized that the two southern populations of Gentoo Penguins would have
greater niche overlap between individuals relative to two northern populations. However, only
the Western Antarctic Peninsula had a significantly greater degree of individual niche overlap
compared to all other populations based on relative niche overlap (Table 3.4; Figure 3.3). This
suggests that individuals in the Western Antarctic Peninsula exhibit more overlap in their
foraging niches, resulting in less specialization among individuals compared to other three
populations. This could be explained by prey resource clustering, or less pressure for resource
competition between individuals specific to this region compared to other populations (Sargeant
2007). Although both the Western Antarctic Peninsula and South Shetland Islands are known for
55
high abundances of krill relative to northern population habitats, prey abundance and availability
does differs between the foraging habitats in the Western Antarctic Peninsula and South Shetland
Islands. The Western Antarctic Peninsula is known for a high year round abundance of Antarctic
krill (Lascara et al. 1999) which may result in individuals clustering around resources and
minimize the need to reduce competition through niche partitioning, as are results demonstrate.
The South Shetland Islands are in close proximity to an ocean convergence zone, resulting in the
Southern Antarctic Convergence Front (SACCF) and Southern Boundary of the ACC (SB)
(Figure 3.1). These fronts are associated with high densities of Antarctic krill, but with high
variability between seasons (Dietrich et al. 2014). Variability in the abundance and availability
of krill may select for greater diversity in diet between individual penguins in order to reduce
competition, which would explain the lower niche overlap among individuals in the South
Shetland Islands. Furthermore, δ13C values vary substantially north and south of convergence
zones and fronts (Francois et al. 1993; Altabet and Francois, 1994). This may explain why
individuals in the South Shetland Islands exhibit less individual niche overlap in foraging habitat,
as some individuals foraging north of these frontal zones would have higher δ13C values than
individuals foraging primarily south of the frontal zones, which has been documented in other
seabirds (Quillfeldt et al. 2005, Cherel and Hobson 2007; Jaegar et al. 2010)
Significance
Our study is the first to compare and detect population level differences in the degree of
individual specialization within a species. Determining the generalist strategy of Gentoo
Penguins and detecting any variation in individual specialization between geographically distinct
populations may have implications for their responses to environmental change and prey
availability in the Scotia Arc. The Antarctic Peninsula and South Shetland Islands are currently
56
undergoing a drastic climate driven decline in the amount and seasonal duration of sea ice
(Stammerjohn et al. 2008, Pritchard et al. 2012). Concurrently, Antarctic krill is also declining in
this region due to the loss of sea ice, which is crucial habitat for the juvenile life stage (Ducklow
et al. 2007; Trivelpiece et al. 2011). As this principle prey item continues to decline, Gentoo
Penguin populations may become more specialized in these regions, similar to those populations
in South Georgia and the Falkland Islands. However, Gentoo Penguins in the South Georgia and
the Falkland Islands are known to forage on a wider diversity of prey items and its unclear
whether the Western Antarctic peninsula and South Shetland Islands harbor a comparable range
of prey resources. Therefore, populations and individuals that specialize on krill in South
Shetland Islands, Western Antarctic Peninsula may be at higher risk, a risk intensified because
their foraging ranges overlap with both commercial krill fisheries and sea ice loss (Miller et al.
2009, 2010; Nicol and Foster 2002; Trivelpiece et al. 2011).
LITERATURE CITED
Ainley, D. G., N. Nur, and E. J. Woehler. 1995. Factors affecting the distribution and size of
pygoscelid penguin colonies in the Antarctic. The Auk 112:171-182.
Altabet, M. A., and R. Francois. 1994. Sedimentary nitrogen isotope ratio as a recorder from
surface ocean nitrate utilization. Global Biogeochemical Cycles 8:103-116.
Araújo, M. S., P. R. Guimaraes Jr, R. Svanbäck, A. Pinheiro, P. Guimarães, S. F. D. Reis, and D.
I. Bolnick. 2008. Network analysis reveals contrasting effects of intraspecific competition
on individual vs. population diets. Ecology 89:1981-1993.
Baylis, A. M., S. Croft, and A.C. Wolfaardt. 2013. Population trends of Gentoo Penguins
Pygoscelis papua breeding at the Falkland Islands. Marine Ornithology 41(1): 1-5.
Barquete, V., V. Strauss, P. G. Ryan. 2013. Stable isotope turnover in blood and claws: a case
study in captive African penguins. Journal of Experimental Marine Biology and Ecology
448: 121-127.
57
Bolnick, D. I., R Svanback, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L.
Forister. 2004. The ecology of individuals: Incidence and implications of individual
specialization. The American Naturalist 161:1-28.
Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. Macleod. 2004. Determining trophic
niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology
73:1007-1012.
Brinton, E., M. D. Ohman, A. W. Townsend, M. D. Knight, and A. Bridgeman. 1999.
Euphausiids of the world ocean. (Expert-center for Taxonomic Identification, Scripps
Institution of Oceanography, Eds.)., Amsterdam: University of Amsterdam, ETI.
Cherel, Y., and K.A. Hobson. 2007. Geographical variation in carbon stable isotope signatures of
marine predators: a tool to investigate their foraging areas in the Southern Ocean. Marine
Ecology Progress Series 329: 281–287.
Clausen, A. P., and K. Putz. 2002. Recent trends in diet composition and productivity of Gentoo,
Magellanic and rockhopper Penguins in the Falkland Islands. Aquatic Conservation:
Marine and Freshwater Ecosystems 12(1):51-61.
Clausen, A., and K. Pütz. 2003. Winter diet and foraging range of Gentoo Penguins (Pygoscelis
papua) from Kidney Cove, Falkland Islands. Polar Biology 26(1):32-40.
Convey, P., A. Morton, and J. Poncet. 1999. Survey of marine birds and mammals of the South
Sandwich Islands. Polar Record 35(193): 107-124.
Costa, A., S. alvidio, M. Posillico, G. Matteucci, B. De Cinti, A. Romano. 2015. Generalization
within specialization: inter-individual diet variation in the only specialized salamander in
the world. Scientific Reports 5.
Croxall, J. P., and P. A. Prince. 1980. The food of Gentoo Penguins Pygoscelis papua and
Macaroni penguins Eudyptes chrysolophys at South Georgia. Ibis 122: 245-253.
Croxall, J. P., R. W. Davis, and M. J. O'Connell. 1988. Diving patterns in relation to diet of
Gentoo and macaroni penguins at South Georgia. Condor 90:157-167.
Dellinger, T., and F. Trillmich. 1999. Fish prey of the sympatric Galapagos fur seals and sea
lions: seasonal variation and niche separation. Canadian Journal of Zoology 77:1204–
1216.
Ducklow, H. W., K. Baker, D. G. Martinson, L. B. Quetin, R. M. Ross, R. C. Smith, S. E.
Stammerjohn, M. Vernet, and W. Fraser. 2007. Marine pelagic ecosystems: the West
Antarctic Peninsula. Philos T Roy Soc B 362: 67–94.
58
Durell, S.E.L.V.D. 2000. Individual feeding specialization in shorebirds: population
consequences and conservation implications. Biological Reviews of the Cambridge
Philosophical Society 75(04): 503-518.
Ehlinger, T.J., D.S. Wilson. 1988. Complex foraging polymorphism in bluegill sunfish.
Proceedings to the National Academy of Science 85:1878-1882.
Forcada, J., and P. N. Trathan. 2009. Penguin responses to climate change in the Southern
Ocean. Global Change Biology 15(7): 1618-1630.
France, R. L. 1995. Carbon-13 enrichment in benthic compared to planktonic algae: foodweb
implications. Marine Ecology Progress Series 124:307–312.
Francois, R., M. A. Altabet, R. Goericke, D. C. McCorkle, C. Brunet, and A. Poisson. 1993.
Changes in the δ13C of surface water particulate organic matter across the subtropical
convergence in the SW Indian Ocean. Global Biogeochemical Cycles 7(3): 627-644.
Frédérich, B., O. Lehanse, P. Vandewalle, and G. Lepoint. 2010. Trophic niche width, shift, and
specialization of Dascyllus aruanus in Toliara lagoon, Madagascar. Copeia 2010: 218226.
Golet, G. H., K. J. Kuletz, D. D. Roby, and D. B. Irons. 2000. Adult prey choice affects chick
growth and reproductive success in pigeon guillemots. The Auk 117(1): 82-91.
Grey, J. 2001. Ontogeny and dietary specialization in brown trout (Salmo trutta L.) from Loch
Ness, Scotland, examined using stable isotopes of carbon and nitrogen. Ecology of
Freshwater Fish 10:168-176.
Hinke, J.T., M. J. Polit, C. S. Reiss, S. G. Trivelpiece, and W. Z. Trivelpiece. 2012. Flexible
reproductive timing can buffer reproductive success of Pygoscelis spp. penguins in the
Antarctic Peninsula region. Marine Ecology Progress Series 454:91–104.
Hinke, J. T., M. J. Polito, M, M. E. Goebel, S. Jarvis, C. S. Reiss, S. R. Thorrold, S, W. Z.
Trivelpiece, and G. M. Watters. 2015. Spatial and isotopic niche partitioning during
winter in Chinstrap and Adélie Penguins from the South Shetland Islands. Ecosphere,
6(7): 1-32.
Hobson, K. A., and R. G. Clark. 1992. Assessing avian diets using stable isotopes I: turnover of
13C in tissues. Condor 94:181-188.
Hückstädt, L. A., P. L. Koch, B. I. McDonald, M. E. Goebel, D. E. Crocker, and D. P. Costa.
2012. Stable isotope analyses reveal individual variability in the trophic ecology of a top
marine predator, the southern elephant seal. Oecologia 169(2):395–406.
59
Jackson, A., R. Inger, A. Parnell, and S. Bearhop. 2011. Comparing isotopic niche widths among
and within communities: SIBER – Stable Isotope Bayesian ellipses in R. Journal of
Animal Ecology 30:595-602.
Jaeger, A., P. Blanchard, R. Richard, and Y. Cherel. 2009. Using carbon and nitrogen isotopic
values of body feathers to infer inter- and intra-individual variations of seabirds feeding
ecology during moult. Marine Biology 156:1233-1240.
Jaegar, A., V. Lecomte, H. Weimerskirch, R. Richard, and Y. Cherel, 2010. Seabird satellite
tracking aclidates the use of latitudinal isoscapes to depict predators’ foraging areas in the
Southern Ocean. Rapid Communications in Mass Spectrometry 24:3456-3460.
Kernaléguen, L., B. Cazelles, J. P. Arnould, P. Richard, C. Guinet, and Y. Cherel. 2012. Longterm species, sexual and individual variations in foraging strategies of fur seals revealed
by stable isotopes in whiskers. PloS one 7(3): e32916.
Kernaléguen, L, N. Dorville, D. Ierodiaconou, A. J. Hoskins, A. M. M. Baylis, M.A. Hindell, J.
Semmens, K. Abernathy, G. J. Marshall, Y. Cherel, and J. P. Y. Arnould. 2015. From
video recordings to whisker stable isotopes: a critical evaluation of timescale in assessing
individual foraging specialization in Australian fur seals. Oecologia 180(3):657-70.
Layman, C.A., and J. E. Allgeier. 2011. Characterizing trophic ecology of generalist consumers:
a case study of the invasive lionfish in The Bahamas. Marine Ecology Progress Series
448: 131-141.
Lynch, H. J. 2013. Gentoo Penguins in Penguins: natural history and conservation (pp. 73-88).
Seattle, WA: University of Washington Press.
Lynch, H. J., R. Naveen, and P. Casanovas. 2013. Antarctic site inventory breeding bird survey
data, 1994-1013. Ecology 94:2653.
McMahon, K. W., M. J, Polito, S. Abel, M. D. McCarthy, S. R. Thorrold. 2015. Carbon and
nitrogen isotope fractionation of amino acids in an avian marine predator, the Gentoo
Penguin (Pygoscelis papua). Ecology and Evolution 5(6): 1278-1290.
Miller, AK, and W. Z. Trivelpiece. 2008. Chinstrap Penguins alter foraging and diving behavior
in response to the size of their principal prey, Antarctic krill. Marine Biology 154: 201208.
Miller, A. K., N. J. Karnovsky, and W. Z. Trivelpiece. 2009. Flexible foraging strategies of
Gentoo Penguins Pygoscelis papua over 5 years in the South Shetland Islands, Antarctica.
Marine Biology 156:2527–2537.
Miller, A. K., M. A. Kappes, S. G. Trivelpiece, and W. Z. Trivelpiece. 2010. Foraging-Niche
Separation of Breeding Gentoo and Chinstrap Penguins, South Shetland Islands,
Antarctica. Condor 112(4):683–695.
60
Minagawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further
evidence and the relation between δ15N and animal age. Geochimica et Cosmochimica
Acta 48(5):1135-1140.
Montes-Hugo M., S. C. Doney, H. W. Ducklow, W. Fraser, W. Martinson, S. E. Stammerjohn,
and O. Schofield. 2009. Recent changes in Phytoplankton communities associated with
rapid regional climate change along the Western Antarctic Peninsula. Science 323:14711473.
Murpy, E.J., E. E. Hofmann, J. L. Watkins, N. M. Jornston, A. Piñones, T. Ballerini, S. L. Hill,
P. N. Trathan, G. A. Tarling, R. A. Cavanagh, E. F. Young, S. E. Thorpe, and P. Fretwell.
2013. Comparison of the structure and function of Southern Ocean regional ecosystems:
The Antarctic Peninsula and South Georgia. Journal of Marine Systems 109-110:22-42.
Newsome, S. D., C. Martínez del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic
ecology. Frontiers in Ecology and Environment 5:429–436.
Newsome, S. D., M. T. Tinker, D. H. Monson, O.T. Oftedal, K. Ralls, M. M. Staedler, M. L.
Fogel, and J. A. Estes. 2009. Using stable isotopes to investigate individual diet
specialization in California sea otters (Enhydra lutris nereis). Ecology 90(4): 961-974.
Nicol, S., A. J. Constable, and T. Pauly. 2000. Estimates of circumpolar abundance of Antarctic
krill based on recent acoustic density measurements. CCAMLR Science 7:87-99.
Nicol, S. and J. Foster. 2003. Recent trends in the fishery for Antarctic krill. Aquatic Living
Resources, 16(1):42-45.
Parent, C. E., D. Agashe, D. I. Bolnick. 2014. Intraspecific competition reduces niche width in
experimental populations. Ecology and Evolution 20:3978-3990.
Polito, M. J., W. Z. Trivelpiece, N. J. Karnovsky, E. Ng, W. P. Patterson, and S. D. Emslie.
2011a. Integrating stomach content and stable isotope analyses to quantify the diets of
Pygoscelid penguins. PLoS ONE 6(10):e26642.
Polito, M.J., S. Abel, C. R. Tobias, S. D. Emslie. 2011b. Dietary isotopic discrimination in
Gentoo Penguin (Pygoscelis papua) feathers. Polar Biol 34(7):1057–1063.
Polito, M. J., W. Z. Trivelpiece, W. P. Patterson, N. J. Karnovsky, C. S. Reiss, and S. D. Emslie.
2015. Contrasting specialist and generalist patterns facilitate foraging niche partitioning
in sympatric populations of Pygoscelis penguins. Marine Ecology Progress Series 519:
332-337.
Prichard, H. D., S. R. M. Ligtenberg, H. A. Fricker, D. G. Vaughan, M. R. van den Broke, and L.
Padman. 2012. Antartic ice-sheet loss driven by basal melting of ice shelves. Nature
484:502-505.
61
Putz, K., R. J. Ingham, J. G. Smith, and J. P. Croxall. 2001. Population trends, breeding success
and diet composition of Gentoo Pygoscelis papua, magellanic Spheniscus magellanicus
and rockhopper Eudyptes chrysocome penguins in the Falkland Islands. A review. Polar
Biology 24(11):793-807.
Quillfeldt, P., R. A. McGill, and R. W. Furness. 2005. Diet and foraging areas of Southern Ocean
seabirds and their pre inferred from stable isotopes: review and case study of Wilson’s
storm-petrel. Marine Ecology Progress Series 295:295-304.
Rau, G. H., T. Takahashi, D. J. Des Marais, and C. W. Sullivan. 1991. Particulate organic matter
δ13C variations across the Drake Passage. J Geophys Res 96:15131–15135.
Salazar, S. K. 2005. Variación temporal y espacial del espectrotrófico del lobo marino de
Galápagos. Master’s thesis, Instituto Politécnico Nacional, La Paz.
Sargeant, B.L. 2007. Individual foraging specialization: niche width versus niche overlap. Oikos
116:1431-1437.
Schindler, D.E., J. R. Hodgson, and J. F. Kitchell. 1997. Density-dependent changes in
individual foraging specialization of largemouth bass. Oecologia 110:592-600.
Stammerjohn, S.E., D. G. Martinson, R. C. Smith, and R. A. Iannuzzi. 2008. Sea ice in the
western Antarctic Peninsula region: Spatio-temporal variability from ecological and
climate change perspectives. Deep-Sea Res II 55:2041–2058.
Svanbäck, R., and D. I. Bolnick. 2007. Intraspecific competition drives increased resource use
diversity within a natural population. Proceedings of the Royal Society of London B:
Biological Sciences 274: 839-844.
Tanton, J. L., K. Reid, J. P. Croxall, and P. N. Trathan. 2004. Winter distribution and behaviour
of Gentoo Penguins Pygoscelis papua at South Georgia. Polar Biology27: 299-303.
Tarling, G.A., J. Cuzin-Roudy, S. E. Thorpe, R. S. Shreeve, P. Ward, and E. J. Murphy. 2007.
Recruitment of Antarctic Krill Euphausia superba in the South Georgia region: adult
fecundity and the fate of larvae. Marine Ecology Progress Series 331:161-179.
Svanbäck, R., and L. Persson. 2004. Individual diet specialization, niche width and population
dynamics: implications for trophic polymorphisms. Journal of Animal Ecology 73: 973982.
Svanbäck, R., P. Eklöv, R. Fransson, K. Holmgren. 2008. Intraspecific competition drives
multiple species resource polymorphism in fish communities. Oikos 117:114-124.
Svanbäck, R., and L. Persson. 2009. Population density fluctuations change the selection
gradient in Eurasian perch. The American Naturalist 173: 507-516.
62
Trivelpiece, W.Z., J. T. Hinke, A. K. Miller, C. S. Reiss, S. G. Trivelpiece, and G.M. Watters.
2011. Variability in krill biomass links harvesting and climate warming to penguin
population changes in Antarctica. Proceedings to the National Academy of Science USA
108:7625–7628.
Turner, T. F., M. L. Collyer, and T. J. Krabbenhoft. 2010. A general hypothesis-testing
framework for stable isotope ratios in ecological studies. Ecology 91: 2227-2233.
Vander Zaden, H.B., K. A. Bjorndal, K. J. Reich, and A. B. Bolten. Individual specialists in a
generalist population: results from a long-term stable isotope series. Biology Letters 6:
711-714.
Villegas-Amtmann, S., D. P. Costa, Y. Tremblay, S. Salazar, and D. Aurioles-Gamboa. 2008.
Multiple foraging strategies in a marine apex predator, the Galapagos sea lion Zalophus
wollebaeki. Marine Ecology Progress Series 363:299-309.
Volkman, N. J., P. Presler, and W. Z. Trivelpiece. 1980. Diets of pygoscelid penguins at King
George Island, Antarctica Condor:373-378.
Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in sizestructured populations. Annual review of ecology and systematics 15: 393-425.
Wilson, R. P., B. M. Culik, P. Kosiorek, and D. Adelung. 1998. The over-winter movements of a
Chinstrap Penguin (Pygoscelis antarctica). Polar Record 34:107-112.
Zaccarelli, N., D. Bolnick, G. Mancinelli. 2013. RInSp: an R Package for the analysis of
individual specialization in resource use. Methods in Ecology and Evolution 4:10181023.
63
CHAPTER 4:
GENERAL SUMMARY AND CONCLUSIONS
Populations of Chinstrap and Adélie Penguins have declined significantly in the Western
Antarctic region over the past few decades, while Gentoo Penguin population numbers have
remained stable and even increased in this region (Lynch, et al. 2012). Many studies have
indicated that, during the breeding season, Gentoo Penguins are generalist foragers, while Adélie
and Chinstrap tend to specialize on krill within the Western Antarctic Peninsula and South
Shetland Islands (Volkman et al. 1980, Lishman, 1985; Karnovsky 1997, Miller et al. 2008,
2009, 2010; Polito et al. 2011b; Polito et al. 2015). These differences in foraging strategies may
help to explain differences in population trends among these three species. Decreases in
Antarctic krill abundance may be causing the declines we see in Adélie and Chinstrap Penguins,
while a generalist foraging strategy exhibited by Gentoo Penguins may act as a buffer to the
decline in this principle prey item. However, little is known on temporal consistency in diet and
foraging habitat of these species, particularly at the individual level, which can further reveal a
more detailed understanding of foraging ecology, as well as provide further information as to
how these species may react to ecological pressures. The overall goal of this thesis was to
investigate dietary and foraging habitat consistency in Pygoscelis species using stable isotope
analysis and to identify the drivers of variation in foraging strategies across populations.
In chapter two, we found significant variation in foraging strategies among Pygoscelis
penguin species in the South Shetland Islands. Gentoo Penguins demonstrated a generalist
foraging strategy marked by significant individual specialization and seasonal consistency,
Adélie Penguins exhibited an intermediate generalist foraging strategy with little seasonal
consistency between individuals, and Chinstrap Penguins appear to be highly specialized in
64
trophic level and consistent between seasons. These results suggest that Chinstrap Penguins may
be at the greatest risk to the decline in Antarctic krill, as they appear to consistently specialize on
this prey species between seasons. Although Adélie Penguins exhibited a more variable diet than
Chinstrap Penguins, the decrease in Antarctic krill abundance may still be a driving factor in
their decline in population numbers, although more research is needed. The seasonal consistency
in generalist diet with significant individual variability that we found in Gentoo Penguins may
provide further evidence as to why this species appears to be resilient to declines in Antarctic
krill.
In contrast, we found similarities in foraging habitat across all three species, with greater
variation in foraging habitat at the population level during the non-breeding season compared to
the breeding season. This could be due to adult penguins having a more restricted foraging range
during the breeding season while they are incubating eggs. We also found evidence of significant
variation in foraging strategy between populations of Gentoo Penguins in the Western Antarctic
Peninsula and South Shetland Islands, suggesting that foraging ecology may be a product of
local geographic and environmental conditions.
In chapter three, we further explored this by comparing foraging strategies of four
geographically isolated Gentoo Penguin populations throughout the Scotia Arc, which vary in
population size and prey variability and abundance. We found that all four populations exhibited
type B generalism, but the degree of individual specialization varied geographically. The smaller
southern populations in the Western Antarctic Peninsula and South Shetland Islands had lower
degrees of individual specialization than the larger northern populations in South Georgia and
the Falkland Islands. This suggests that individual specialization may be mediated by population
size as a way to minimize competition between conspecifics. In addition, our results also suggest
65
that individual specialization may be driven by prey abundance and diversity, as foraging habitat
in the southern populations are marked by high abundance of Antarctic krill and low prey
diversity, while the northern populations forage on a wider diversity of prey.
Investigating population level foraging strategies in penguins using SIA of multiple
tissues representing different time scales has only recently been used (Juàres et al. 2016).
However, we are the first to use this method to look at individual consistency in foraging
strategies, which provides a more robust understanding of a species’ and population’s ecology.
Furthermore, our studies are the first to compare and detect population level differences in the
degree of individual specialization within a species. Investigating individual specialization using
SIA with serial sampling of metabolically inert tissue has only recently been conducted in very
few studies with the vibrissae of marine mammals (Newsome et al 2009, Huckstadt et al 2012,
Kerguelan et al 2015; Jaegar et al. 2009). Our results suggest that this method might be useful in
other organisms in which serial sampling is possible to quantify temporal changes and confirm
generalist strategies through time.
Given the current pressures on Pygoscelis penguins in the Scotia arc such as climate
change and the decline in principle prey items such as Antarctic krill (Atkins et al. 2004;
Ducklow et al. 2007, Trivelpiece et al. 2011), it would be important to investigate how species
may be affected by these large scale changes, and whether they can adapt over time. Future
directions of research should include temporal analysis of Pygoscelis penguin foraging strategies
by combining the SIA sampling methods from chapter two and three, as well as repeated
sampling over multiple years. This would help to understand whether populations and
individuals vary or maintain foraging consistency over a larger time scale, as well as help to
identify populations that are more at risk to significant environmental changes.
66
LITERATURE CITED
Atkinson, A., V. Siegel, E. Pakhomov, and P. Rothery. 2004. Long-term decline in krill stock
and increase in salps within the Southern Ocean. Nature, 432(7013): 100-103.
Ducklow, H. W., K. Baker, D. G. Martinson, L. B. Quetin, R. M. Ross, R. C. Smith, S. E.
Stammerjohn, M. Vernet, and W. Fraser. 2007. Marine pelagic ecosystems: the West
Antarctic Peninsula. Philos T Roy Soc B 362: 67–94.
Hückstädt L. A., P. L. Koch, B. I. McDonald, M. E. Goebel, D. E. Crocker, and D. P. Costa.
2012. Stable isotope analyses reveal individual variability in the trophic ecology of a top
marine predator, the southern elephant seal. Oecologia 169(2):395–406.
Jaeger, A., P. Blanchard, R. Richard, and Y. Cherel. 2009. Using carbon and nitrogen isotopic
values of body feathers to infer inter- and intra-individual variations of seabirds feeding
ecology during moult. Marine Biology 156:1233-1240.
Juáres, M. A. 2016. Diet composition and foraging habitats of Adélie and Gentoo penguins in
three different stages of their annual cycle. Marine Biology 163:105.
Karnovsky, N. J. 1997. The fish component of Pygoscelis penguin diets (Doctoral dissertation,
Montana State University-Bozeman, College of Letters & Science)
Kernaléguen, L, N. Dorville, D. Ierodiaconou, A. J. Hoskins, A. M. M. Baylis, M.A. Hindell, J.
Semmens, K. Abernathy, G. J. Marshall, Y. Cherel, and J. P. Y. Arnould. 2015. From
video recordings to whisker stable isotopes: a critical evaluation of timescale in assessing
individual foraging specialization in Australian fur seals. Oecologia 180(3):657-70.
Lishman, G. S. 1985. The food and feeding ecology of Adélie penguins (Pygoscelis adeliae) and
Chinstrap penguins (P. antarctica) at Signy island, South Orkney Islands. Journal of
Zoology 205: 245-263.
Lynch, H. J., R. Naveen, P. N. Trathan, and W. F. Fagan. 2012. Spatially integrated assessment
reveals widespread changes in penguin populations on the Antarctic Peninsula. Ecology
93(6):1367-1377.
Miller, A. K., M. A. Kappes, S. G. Trivelpiece, and W. Z. Trivelpiece. 2010. Foraging-Niche
Separation of Breeding Gentoo and Chinstrap Penguins, South Shetland Islands,
Antarctica. Condor 112(4):683–695.
Miller, A. K., N. J. Karnovsky, and W. Z. Trivelpiece. 2009. Flexible foraging strategies of
Gentoo Penguins Pygoscelis papua over 5 years in the South Shetland Islands, Antarctica.
Marine Biology 156:2527–2537.
67
Miller, AK, and W. Z. Trivelpiece. 2008. Chinstrap Penguins alter foraging and diving behavior
in response to the size of their principal prey, Antarctic krill. Marine Biology 154: 201208.
Newsome, S. D., M. T. Tinker, D. H. Monson, O.T. Oftedal, K. Ralls, M. M. Staedler, M. L.
Fogel, and J. A. Estes. 2009. Using stable isotopes to investigate individual diet
specialization in California sea otters (Enhydra lutris nereis). Ecology 90(4): 961-974.
Polito, M. J., W. Z. Trivelpiece, N. J. Karnovsky, E. Ng, W. P. Patterson, and S. D. Emslie.
2011a. Integrating stomach content and stable isotope analyses to quantify the diets of
Pygoscelid penguins. PLoS ONE 6(10):e26642.
Polito, M. J., W. Z. Trivelpiece, W. P. Patterson, N. J. Karnovsky, C. S. Reiss, and S. D. Emslie.
2015. Contrasting specialist and generalist patterns facilitate foraging niche partitioning
in sympatric populations of Pygoscelis penguins. Marine Ecology Progress Series 519:
332-337.
Trivelpiece, W.Z., J. T. Hinke, A. K. Miller, C. S. Reiss, S. G. Trivelpiece, and G.M. Watters.
2011. Variability in krill biomass links harvesting and climate warming to penguin
population changes in Antarctica. Proceedings to the National Academy of Science USA
108:7625–7628.
Volkman, N. J., P. Presler, and W. Z. Trivelpiece. 1980. Diets of pygoscelid penguins at King
George Island, Antarctica Condor:373-378.
68
APPENDIX:
ESTIMATING THE TIMING OF TAIL FEATHER GROWTH
We investigated tail feather growth in a captive population of Gentoo Penguins at
SeaWorld in Orlando, Florida in 2012. We monitored ten individual adults (5 males and 5
females) from the start of molt. Following the beginning of molt, we measured the exposed
length (cm) of the largest newly growing central tail feather of each individual at approximately
20 day intervals until 100 days after molt when tail feather growth was near completion. We
assumed that the duration of feather growth is similar between captive and wild Pygoscelis.
We then used PROC NLIN with the Marquardt method in SAS (SAS Institute Inc. 1999)
to fit von Bertalanffy growth curves to this dataset. The von Bertalanffy growth curve has been
commonly used to model growth in large seabirds (Ricklefs 1967) and is formulated as:
!" = !
1 − & '(("'"*
where Lt is the predicted length (in cm) at time t (in days), L∞ is the mean length that
would be reached if feathers grew indefinitely, k is a growth parameter of dimension time−1, and
t0 is the hypothetical time (in days) when feather length is zero. We modeled growth curves and
compared 95% confidence intervals of parameter estimates, residual mean square error (MSE)
and pseudo-R2 as measures of goodness of fit. MSE indicated that the von Bertalanffy growth
curve fit our data for males slightly better than for females and both sexes combined, while
pseudo R2 indicated that the curve fit slightly better for females than males and both sexes
combined. However, 95% confidence intervals (CI) of parameter estimates overlapped
considerably between all three models (Table A.1). Because of this overlap and the small
differences between sexes, we used the parameter estimates derived from both sexes combined to
estimate the timing of tail feather growth based on length:
69
1
!"
+ = +, − × ln 11
!
where t is the predicted time (in days) since the start of molt that a section of tail feather at length
Lt was synthesized based on the predicted von Bertalanffy growth parameters (L∞ = 27.2, k =
0.01, t0 = 1.7; Figure A.1).
Table A.1: Parameter estimates, 95% confidence intervals (in parentheses), residual mean square
error (MSE) and pseudo R2 values estimated from fitted von Bertalanffy growth equations of
total feather length (cm) relative to time (days) elapsed since the onset of feather molt.
Parameter Estimates
Group
Fit Statistics
MSE
Pseudo R2
L∞
k
t0
Males
24.8 (18.6-30.9)
0.01 (0.01-0.02)
5.4 (0.08-10.8) 0.948
Female
30.5 (19.2-41.7)
0.01 (0.003-0.01) -2.7 (-9.7-4.3)
0.631
0.978
Combined
27.2 (21.5-32.9)
0.01 (0.002-0.01) 1.7 (-2.6-6.0)
0.820
0.918
0.965
We applied the above formula to estimate timing of growth for discrete sections of tail
feathers that we obtained from individual wild Gentoo Penguins (Table A.2). We sampled
discrete sections of the tail feathers for stable isotope analysis by using stainless steel scissors to
cut four 1.0cm sections of feather shaft starting at the distal portion of each feather between 5.56.5cm, 7.5-8.5cm, 9.5-10.5cm, and 11.5-12.5cm. These sections represent growth occurring
between 21-25, 29-33, 38-43, and 48-54 days following the onset of molt. We chose to sample
sections starting at 5.5cm in order to avoid sampling feather material grown within the molt
period, during which Gentoo Penguins fast for 19.5 days on average (Adams and Brown, 1990).
70
Table A.2. Estimated timing of growth for tail feather sections using the formula:
2
3
+ = +, − × ln 11 4 . Asterisk (*) indicates tail feather sections used for analysis.
(
3
Tail feather
length (cm)
Day that feather
reaches length
0
-
0.5
3
1.5
6
2.5
10
3.5
13
4.5
17
5.5
21
6.5
25
7.5
29
8.5
33
9.5
38
10.5
43
11.5
48
12.5
54
13.5
60
14.5
66
15.5
73
16.5
81
17.5
89
18.5
98
19.5
109
20.5
120
Period of growth per 1cm
feather section
3 days
5 days
5 days
5 days*
6 days*
6 days*
7 days*
7 days
9 days
10 days
13 days
71
20
15
10
5
Total feather length (cm)
20
40
60
80
100
Time since start of molt (days)
Figure A.1 Fitted von Bertalanffy growth curve (solid line) and 95% confidence intervals (dotted
lines) of total tail feather length relative to time since the onset of feather molt in a captive
population of twelve adult (5 male and 5 females) Gentoo Penguins maintained at SeaWorld in
Orlando, Florida during 2012.
72
VITA
Rachael was born in Florida in 1987. She grew up in Sutton, Massachusetts, and received
her B.S. in Biology with Departmental Honors from the University of Massachusetts at Amherst
in 2009. In 2011, she lived and worked on the remote island atoll of Laysan, located in the
Northwestern Hawaiian Islands, for seven months, during which she monitored the reproductive
ecology of albatross and developed her passion for seabirds and oceanography. Following this
formative experience, she continued to follow her passion for seabirds and worked many
positions including Island Manager for the Roseate Tern Recovery program for Massachusetts
Division of Fish and Wildlife, avian technician for Alaska Maritime National Wildlife Refuge,
and Research Coordinator for the California Least Tern and Snowy Plover Recovery Program
with the San Diego Zoo Institute for Conservation Research. She returned to school in 2014 to
join Dr. Michael Polito’s lab at Louisiana State University, where she had the incredible
opportunity to study penguin foraging ecology in the Antarctic for her Master’s Thesis. She is
expected to graduate in 2016.
73