Adélie penguin dietary remains reveal Holocene environmental changes in the western Ross Sea (Antarctica)

Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Adélie penguin dietary remains reveal Holocene environmental changesin the western Ross Sea (Antarctica)

Sandra Lorenzini a,⁎, Carlo Baroni a,b, Ilaria Baneschi b, Maria Cristina Salvatore a,Anthony E. Fallick c, Brenda L. Hall d

a Dipartimento di Scienze della Terra, Università degli Studi di Pisa, Via S. Maria 53, 56126 Pisa, Italyb Istituto di Geoscienze e Georisorse, Consiglio Nazionale delle Ricerche, Via G. Moruzzi 1, 56124 Pisa, Italyc Scottish Universities Environmental Research Centre, East Kilbride G75 0QF, Scotland, UKd School of Earth and Climate Sciences and the Climate Change Institute, University of Maine, Orono, ME 04469 USA

⁎ Corresponding author. Tel.: +39 0502215738, fax: +E-mail address: [email protected] (S. Lorenzini).

0031-0182/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.palaeo.2013.12.014

a b s t r a c t

Article history:Received 8 June 2013Received in revised form 28 November 2013Accepted 8 December 2013Available online 18 December 2013

Keywords:Adélie penguinOrnithogenic soilStable isotopePaleoenvironmentPaleoceanography

Carbon and nitrogen stable isotope analyses were performed on modern and Holocene Adélie penguin guanosamples collected from ornithogenic soils along the Scott Coast (Southern Victoria Land, Antarctica), from CapeIrizar to Dunlop Island, and at Cape Bird (Ross Island). Guano samples also were sieved and sorted under stereo-microscope in order to select penguin dietary remains, such asfish bones and otoliths. Carbon andnitrogen stableisotope composition, coupled with the taxonomic identification of fish otoliths from Scott Coast Holocene sam-ples, indicated a mainly fish-based diet for this area, with Pleuragramma antarcticum as the most eaten preythroughout the investigated period (from 390 cal BP to ca 7300 cal BP). The isotopic values of Ross Island sam-ples (from modern to 3850 cal BP) showed a krill consumption increase in the samples younger than 2000 calBP, with the maximum in modern samples. Scott Coast and Ross Island Holocene samples showed δ13C andδ15N trends similar to those previously published fromTerraNovaBay (northernVictoria Land),whereasmodernsamples from Ross Island have similar δ15N composition but different δ13C values. This δ13C divergence started atca 2000 BP and follows the abandonment of the Scott Coast colonies.The δ13C trend observed in Ross Island and Terra Nova Bay samples and the abandonment of the Scott Coastcolonies could suggest the stability and the persistence of the previous oceanographic conditions (i.e. polynya)for the Terra Nova Bay area and the establishment of new conditions for water circulation in the Southern RossSea since ~2000 BP when persistent sea-ice sealed the Scott Coast.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

A fundamental idea in biology is that climatic and environmentalchanges are the driving force behind biodiversity adaptation throughecological and evolutionary responses (Walther et al., 2002; Parmesan,2006; Lambert et al., 2010). This assumption is especially true for Adéliepenguins (Pygoscelis adeliae) which, because of their extraordinary envi-ronmental sensitivity, have becomewell known as a ‘bellwether’ of Ant-arctic climate change (Ainley, 2002). Given the significant role that thisspecies plays in the understanding of how the Antarctic environment ischanging, Adélie penguins represent one of the best-studied animal spe-cies in the world. Moreover, since the description of ornithogenic soils inthe Ross Sea area (multi-millennia penguin-guano deposits, Ugolini,1972), Adélie penguins have also provided valuable new insights intothe late Pleistocene and Holocene glaciological history of Antarctica. Infact, the pristine cold and dry Antarctic environment have led to a highconcentration of well-preserved penguin remains (bones, eggshells,guano and dietary remains) in the ornithogenic soils dated back to the

39 0502215800.

ghts reserved.

late Pleistocene and underlying both existing and abandoned colonies(Ugolini, 1972; Baroni and Orombelli, 1994). These remains represent aunique paleoecological heritage in the context of Antarctic paleoclimaticand paleoenvironmental research. Therefore, since the 1990s,multidisci-plinary studies have intensively investigated Adélie penguin paleoecolo-gy and their colonization history, contributing important information forunderstanding the Holocene history of key areas of Antarctica (Baroniand Orombelli, 1994; Lambert et al., 2002; Baroni and Hall, 2004; Hallet al., 2004, 2006; Emslie and Woehler, 2005; De Bruyn et al., 2009;Hall, 2009). Moreover, investigation has demonstrated that over atleast the last 55,000 yr, local and regional environmental conditions fre-quently changed and Adélie penguins have survived by adapting boththeir ecological behavior (i.e. foraging and feeding behavior) and thegeographic distribution of their colonies (Baroni and Orombelli, 1994;Lambert et al., 2002; Baroni and Hall, 2004; Hall et al., 2004, 2006;Emslie and Woehler, 2005; De Bruyn et al., 2009; Hall, 2009; Lorenziniet al., 2010). The Adélie penguin colonization history of the SouthernVictoria Land Coast indicates that abrupt environmental changesoccurred during the Late Holocene when the persistence of year-roundfast ice caused abandonment of the Scott Coast's colonies, whichhave not been reoccupied since. This abandonment occurred almost

22 S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

concurrently with evidence for the first occupation by breeding pen-guins on Ross Island where suitable conditions for Adélie's settlementhave persisted to the present (Baroni and Orombelli, 1994; Emslieet al., 2003; Hall et al., 2006; De Bruyn et al., 2009).

The guano deposit serves as a unique natural archive of ecologicallyrelevant indicators of diet composition by preserving fish vertebrae andotoliths, and stable isotopes of carbon (δ13C) and nitrogen (δ15N) thatprovide insight into possible food web shifts. In particular, this build-up of material at bird nesting sites represents thousands of years of oc-cupancy, with stable-isotope analysis revealing colonization patternsand changes in prey selection and foraging habitat over long periodsof time (e.g. snow petrels Pagodroma nivea—Ainley et al., 2006; Adéliepenguins Pygoscelis adeliae—Emslie and Patterson, 2007; gyrfalconsFalco rusticolus—Burnham et al., 2009).

Stable isotopes in feces, which are composed of undigested foodmat-ter, gut microbes, digestive secretions, uric acid, and sloughed epithelialtissues, have been demonstrated in experimental andwild settings to re-flect ingested diet in mammalian herbivores and primates (Van derMerwe, 1986; Codron et al., 2005; Sponheimer et al., 2009). Stable-carbon isotope composition (δ13C) provide information on the funda-mental carbon source of the foodweb and regional origin of the primaryproducers at the base of the food chain (Post, 2002; McCutchan et al.,2003). Hence, δ13C values are used to characterize the foraging habitatsof predators, including penguin (Cherel and Hobson, 2007; Cherelet al., 2011). Stable-nitrogen isotope composition (δ15N) is usedmainly to establish trophic relationships (Rau et al., 1982; Hobson andMontevecchi, 1991; Hobson and Welch, 1992).

To date, several studies have shown δ13C and δ15N values of animalfeces to be a faithful tracer of dietary sources, in particular of bats(Mizutani et al., 1992a,b; Wurster et al., 2010 and references therein),birds and also modern penguins (Mizutani and Wada, 1988).

Stable isotope studies applied to Adélie penguin remains collected onthe Antarctica Peninsula and along the Victoria Land coast have recentlyadded considerable information about the paleoecology of this speciesand are a valuable tool for examining biodiversity history and its devel-opment through time (Emslie and Patterson, 2007; Lorenzini et al.,2009, 2012). Particularly, stable carbon (δ13C) and nitrogen (δ15N) iso-tope analyses of Adélie penguin eggshell and guano samples have dem-onstrated to give remarkable temporally different paleoenvironmentalinformation (Emslie and Patterson, 2007; Lorenzini et al., 2010). In fact,the comparison between eggshell and guano isotopic composition hashighlighted that the isotopic ratios of eggshell and guano samples pro-vide different kinds of information on Adélie penguin diet, correspond-ing to short and long periods, respectively: the egg-laying event foreggshell and the summer season for the second (Emslie and Patterson,2007; Lorenzini et al., 2010).

In this paper, we present the results of stable carbon (δ13C) and nitro-gen (δ15N) isotope analyses performed onAdélie penguin guano samplescollected from abandoned penguin colonies distributed along the ScottCoast from Cape Irizar (75°34′S) to Dunlop Island (77°14′S), and onRoss Island, at the extant Cape Bird colony and its outskirts at McDonaldBeach (77°13′E). Dietary data obtained from microscopic analyses andthe isotopic signature of penguin guano are also discussed. Furthermore,comparison with published carbon and nitrogen isotope composition ofguano collected from the Terra Nova Bay area (Lorenzini et al., 2010) en-able the interpretation of Holocene environmental and oceanographicconditions of different sectors of the Ross Sea Embayment, especiallythrough the use of carbon isotopes.

2. Study area

2.1. The Scott Coast

The Scott Coast lies to the south of theDrygalski Ice Tongue (75°30′S)extending south toMinna Bluff (78°31′S) (Fig. 1). Characterized by year-round fast ice, the Scott Coast is a key area for determining both the

extent and the chronology of grounded ice in the Ross Sea Embaymentat the Last Glacial Maximum (LGM), because it displays raised beachescritical for developing relative sea-level (RSL) curves (Hall and Denton,1999; Hall et al., 2004). Several ice-free promontories and islands, withwell-developed raised beaches, occur between Explorers Cove andDepot Island (Fig. 1). North of Depot Island to the Drygalski Ice Tongue,the icebound coast is heavily glaciated, with rare ice-free areas andvery reduced Holocene raised beaches. Abandoned colonies have beenidentified, described and sampled on beach deposits, abrasion platforms,glacial deposits, regolith, and bedrock at Cape Irizar, Prior Island, CapeHickey, Cape Day, Depot Island, Depot Island Peninsula, Cape Ross,Cape Roberts, Dunlop Island, Spike Cape, and Marble Point (Baroni andOrombelli, 1994; Hall and Denton, 1999; Lambert et al., 2002; Hallet al., 2004).

2.2. Ross Island

Ross Island is separated by McMurdo Sound from the Scott Coastmainland (Fig. 1). The northernmargin of Ross Ice Shelf abuts the south-ern coast of the island. Ross Island is almost entirely ice-covered withonly a few ice-free coastal areas available to penguins. Despite thisfact, more than 20% of all nesting pairs breeding in the Ross Sea Embay-ment occur on Ross Island, distributed in just three colonies: Cape Bird,Cape Crozier and Cape Royds, the latter being the smallest colony onRoss Island and the southernmost penguin colony in the world(77°34′S. 166°11′E) (Ainley, 2002; Ballard et al., 2010).

3. Materials and methods

3.1. Field survey and sample collection

Aerial photograph analysis, several field surveys, and detailed geo-morphologic analysis of key sites have been conducted along the VLcoast and on Ross Island. The accurate field survey of ice-free areasallowed the discovery of tens of abandoned Adélie penguin coloniesfrom Cape Adare to Ross Island (Baroni and Orombelli, 1991, 1994;Baroni and Hall, 2004; Hall et al., 2004). Well-sorted pebbles selectedby penguins for building their nests characterize abandoned nestingsites: after the abandonment of nests, pebbles are concentrated at thesurface by wind deflation and prevent the erosion of the lower guano.Pebbly patches hide and protect the lower organic soil formed by birdejecta and remains.

After identifying abandoned colonies, test pits (generally 1–2 m2)were established. Cleaning the entire surface from the top, ornithogenicsoilswere excavated layer-by-layer using techniques commonly used inarcheological research until the underlying bedrock or undisturbedma-rine/glacial sediments were reached (Baroni and Orombelli, 1994;Lambert et al., 2002). The thickness of ornithogenic soils ranges fromsome centimeters to some decimeters. Through the accurate strati-graphic excavation of penguin settlements, multi-occupational phasesof the colonies (eventually separated by mineral layers such as sandand gravel of colluvial or periglacial origin, aeolian deposits, etc.) havebeen identified (Lambert et al., 2002; Baroni and Hall, 2004; Hall et al.,2004). This careful fieldwork allowed collection at distinct organiclayers of several well-preserved and age-constrained Holocene penguinremains and leavings, including penguin bones, feathers, eggshell frag-ments. The good correspondence among multiple radiocarbon datesobtained from different remains collected from the same layer and theaccurate stratigraphic control of sampling, show that no significantcontamination or mixing occurred since the ornithogenic soils formed(Lambert et al., 2002; Baroni and Hall, 2004; Hall et al., 2004). Toprevent the complete destruction of the abandoned colonies, whichmust be regarded as a unique record of the heritage of penguin settling,we collected soil samples weighing (as a mean) only some hundreds ofgrams.

Fig. 1.Mapof Victoria Land showing the location of themodern (asterisk) and abandoned (circles) Adélie penguin colonies sampled in thiswork. Triangles indicate other colonies occupiedat present in the Ross Sea.

23S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

For this study, we analyzed penguin guano samples collected overseveral field surveys between 1994 and 2007. Samples from eightabandoned penguin colonies along the Scott Coast (Cape Irizar,Prior Island, Cape Hickey, Cape Day, Depot Island Peninsula, CapeRoss, Cape Roberts, and Dunlop Island) and from the modern colonyof Cape Bird and its outskirts at McDonald Beach (Ross Island) wereinvestigated. Fresh guano samples were collected at the outskirts ofthe occupied nests; Holocene guano was sampled from each excavatedornithogenic level.

3.2. Stable isotope analyses

We performed organic carbon and nitrogen stable isotope analyseson 81 Holocene and 5 modern penguin guano samples at the IGG-CNRlaboratories (Pisa, Italy). For each sample we selected the b63 μmgrain size fraction that quantitative analyses carried out in the laborato-ry for previous studies have indicated to be the richest in organic carboncontent (Lorenzini et al., 2010, 2012). Samples were treated with 10%HCl to remove inorganic carbon and then washed several times with

deionized water to neutral pH and dried. Isotope ratios of organic car-bon and nitrogen were analyzed using the same sample aliquot. Sub-samples of 2–20 mg, depending on nitrogen and carbon content, wereweighed into tin capsules and analyzed using a Thermo Finnigan Ele-mental Analyzer EA-1108 interfaced with an isotope ratio mass spec-trometer (Finnigan DeltaPlus) via the Conflo II interface (FinniganMAT, Bremen, Germany). Isotope ratios were normalized to VPDB forδ13C and AIR for δ15N using International Atomic Energy Agency(IAEA) NO-3, IAEA CH-6, and internal standards. Precision was betterthan 0.2‰ at one sigma for both carbon and nitrogen isotope values.

3.3. Microscopic analyses

We washed and sieved 47 Holocene and 5 modern guano samplesthrough seven nested screens with square mesh sizes ranging from2 mm to 63 μm. To avoid bias due to variable amounts of sorted sedi-ments, we selected and processed 150 g of dry mass for each sample(when available). We report results as weight percent that is the num-ber of dietary remains normalized for 100 g of soil. The matrix from

24 S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

each screen was dried at low temperature and subsequently sortedunder a low power (5–10×) stereomicroscope to separate eggshellfragments, feathers, and dietary remains (fish vertebrae and otoliths).We classified samples according to the presence/absence of dietaryremains.

We separated otoliths from fish bones in order to carry out taxonom-ic identification. Otolithswere identified using anatomic-comparative ta-bles proposed by Williams and McEldowney (1990) for Antarctic fishtaxa and the Anfibo Base software (Busekist et al., 2007). Fish otolithswere well-preserved and showed erosion signals ranging from 1 to 2according to Leopold et al. (1998), with most of them still preservingthe morphological parameters that are required for taxon identification.

3.4. Radiocarbon dating and statistical analysis

Accelerator mass spectrometry (AMS) dating of penguin guano wasperformed by the NOSAMS laboratory of Woods Hole, USA. All AMS ra-diocarbon dates were corrected and calibrated for the marine-carbonreservoir effect and upwelling of old water in the Southern Ocean byusing aΔR = 791 ± 121 yr based on 14C of the solitary coral, Gardineriaantarctica collected from debris bands on the McMurdo and Hell's GateIce Shelves (Hall et al., 2010). The uncorrected conventional dates ±SDare given with the 2σ calibrated range in calendar years before present(BP) determined from the Calib 6.0.1 software program and INTCAL 09marine calibration curve (Reimer et al., 2009). This range provides 95%confidence that the true age of the specimen falls within the specifiedtime period. A few samples also were dated by association with otherpenguin remains of known age (i.e. penguin bones and eggshells) col-lected from the same stratigraphic level.

Samples younger than 560 cal BP all were collected fromRoss Island,with the exception of one sample from the Scott Coast. By contrast,almost all samples older than 560 cal BP were from Scott Coast.

All statistical analyses were performed using the software Statistica,version 7.1 for Windows (StatSoft, Inc., 2005).

4. Results

4.1. Stable isotope data

The δ13C and δ15N data obtained from modern and fossil penguinguano are given in Appendix 1. The δ13C values of guano samplesrange from −28.9‰ to −24.0‰. Modern samples have a meanvalue of −27.3 ± 0.4‰, whereas Holocene samples show a gradualtendency toward more positive values. In particular, the samplesolder than 560 BP show an average δ13C of −25.5 ± 0.9‰, whereasthemean value for the younger samples, excluding themodern ones,is −27.0 ± 0.7‰. Samples younger than 560 BP are not statisticallydifferent from each other (t-test, t = 1.14, P = 0.265). This supportsthe contention that no diagenetic post-depositional effects have in-fluenced the sample isotopic composition. The samples older than560 BP show an interquartile range of 1.0‰, whereas that for youngersamples is 0.3‰, suggesting more variable feeding in older than inmore recent times. Comparison of samples older and younger than560 BP, includingmodern ones, using an F-test, shows that the standarddeviations of the two groups are not different (F = 2.28, P = 0.04).Hence, a t-test can be used to compare the mean values of the twogroups and this reveals that δ13C mean values of younger samples arestatistically more negative than those for older ones (t = 10.22,P b b0.05).

The δ15N values of guano samples range from +7.7‰ to +27.5‰.Fresh excreta have a mean value of +16.0 ± 2.2‰, not statisticallydifferent from the mean value of Holocene samples younger than560 cal BP (+16.1 ± 1.4‰). Samples older than 560 BP have an aver-age of +19.3 ± 2.2‰ and are all higher than +16.4‰, except for twosamples at around 2900 and 3500 cal BP.

4.2. Microscopic analyses

Guano screening and sorting allowed us to recover well-preserved biological material, including penguin bones, eggshellfragments and hard parts of dietary remains, such as fish bones(vertebrae) and fish otoliths. Of 52 samples, 20 did not containdietary remains and 18 contained only unidentifiable fish vertebraeand fish teeth. Squid beaks were recovered only in one sample atCape Roberts. The presence of eggshell fragments in those sampleswhere neither fish bones nor otoliths were recovered still testifiedto their ornithogenic nature.

We found26otoliths in 14 samples (9 fromRoss Island and5 from theScott Coast) and analyzed them to identify prey taxa. Taxonomic identifi-cation indicated that all fish prey belonged to the family Nototheniidaeand consisted mostly (81%) of Pleuragramma antarcticum. Only a verysmall percentage of the recovered otoliths (5.2%) belonged to fish indi-viduals different from P. antarcticum (identified as Trematomus sp. and,in one case at Dunlop Island, as Trematomus bernacchii) or were remainsthat could be identified only to the Nototheniidae family (13.8%). Basedon our new data and those previously published for the Scott Coast(Lorenzini et al., 2009), the Scott Coast samples showed a higher inter-specific richness than the Ross Island samples that, on the contrary, ex-hibited a paucity both of otoliths and fish bones that persists until thepresent time (Table 1).

5. Discussion

The data obtained from this study will be discussed in conjunc-tion with previously published results (Lorenzini et al., 2009, 2012)in order to describe in detail the paleodietary record of Adélie pen-guins along the Scott Coast to Terra Nova Bay, as well as to providenew late Holocene dietary information about this species at RossIsland.

5.1. Paleodietary reconstruction from microscopic analyses

Adélie penguins primarily feed on krill (Euphausia superba,Euphausia cristallorophias) and fish (mainly the silverfish, Pleuragrammaantarcticum) (Ainley, 2002; Ainley et al., 2003). Krill plays a key role in thepresent-day penguin diet but unlike fish, krill exploitation leaves novisible record in Holocene sediments.

Taxonomic identification of fish otoliths indicated Pleuragrammaantarcticum as the most eaten fish prey throughout the investigated pe-riod both along the Scott Coast and at Ross Island. Very few otolithswererecognized as Trematomus sp. genus or remained unidentified. The abun-dance of remains of Antarctic silverfish observed in this study agreeswith previous paleodietary investigations in the Antarctica Peninsula,East Antarctica, and in the Ross Sea region (Polito et al., 2002; Emslieand Woehler, 2005; Lorenzini et al., 2009), as well as with present-daydata, since this species today accounts for more than 90% of the localfish community in the Ross Sea (Vacchi et al., 2004) and represents akey species in the diet of Antarctic apex predators (La Mesa et al.,2004). In particular, during the chick-rearing period, P. antarcticum cancontribute up to 50–75% by mass to the Adélie penguin diet, even ex-ceeding crystal krill (Euphausia cristallorophias) consumption in thesouthern Ross Sea (Ainley, 2002; Ainley et al., 2003). Samples youngerthan 3000 cal BP collected along the Scott Coast show a general decreasein the number of dietary remains. In addition, even though P. antarticumis the most represented fish taxon, Scott Coast samples also showed va-riety in the prey fish taxa, with a contribution from Pagothenia sp. andTrematomus sp. similar to that at Terra Nova Bay (Lorenzini et al.,2009) (Table 1). In contrast, with the exception of two samples, non-euphausiid paleodiet at Ross Island appears to be almost monospecificand entirely based on silverfish consumption.

Table1

Num

bera

ndtaxo

nomiciden

tification

ofdietaryremains

recove

redin

ornithog

enicsoilsamples

from

theScottC

oastan

dRo

ssIsland

basedon

new

data

repo

rted

here

andthoseprev

ious

lypu

blishe

d(Loren

zini

etal.,20

09).Ana

lyzedgu

anosamples

aregrou

pedby

colle

ctionsite

withtheag

erang

e(cal

yrBP

,using

aΔR=

791±

121yr;H

alle

tal.,20

10)an

dtotald

rymass(g)of

sorted

sedimen

ts.

Colle

ctionsite

Latitude

Radiocarbo

nda

tarang

e(cal

yrsBP

)

Totalm

ass

(g)

No.

ofreco

vered

fish

otoliths

No.

ofreco

vered

fish

bone

sNo.

ofrecove

red

squidbe

aks

Pleu

ragram

ma

antarctic

um(%

)

Trem

atom

usbernacchii

(%)

Trem

atom

ussp

.(%

)Pa

gothen

iasp

.(%

)n.d.

(%)

Cape

Irizar

75°34′S

839–

4702

2111

348

386

.49

0.29

0.29

1.44

9.48

PriorIs.

75°41′S

391–

5374

1127

.663

888

.89

11.11

Cape

Hicke

y76

°05′S

2,15

1–72

8817

62.4

282

92.86

7.14

NCa

peDay

76°15′S

3470

226

4190

.24

2.44

7.32

Dep

otIs.P

eninsu

la76

°42′S

2192

–68

2376

69

110

0.00

Cape

Ross

76°44′S

2983

–37

0811

4834

3594

.12

5.88

Cape

Robe

rts

77°00′S

2902

290

02

2Ca

peBird

77°13′S

0–38

5253

72.3

1949

89.47

10.53

Dun

lopIs.

77°14′S

1248

–51

2615

48.6

812

075

.00

12.50

12.50

25S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

5.2. Paleodietary reconstruction from C and N stable isotope analyses

Although direct information about past krill consumption cannot beobtained by guano screening, because krill and their fragile carapacesare not preserved in the soils, the isotopic approach enables us to iden-tify the dietary contribution of krill and to reconstruct a more detailedpaleodietary record as krill is metabolized and contributes to carbonand nitrogen isotopic composition of tissues, including guano.

We reported the δ13C values for guano samples from the three differ-ent sites (Southern Victoria Land— SVL, Ross Island—RI and Terra NovaBay— TNB) (Fig. 2). Samples from TNB and SVL show similar values andinsignificant variation through the Holocene period. However, δ13Cvalues of samples fromRI and TNByounger than 560BP are significantlydifferent (t-test, t = −8.06, P b b0.05). This implies a different balanceof sources of carbon.

Fish and krill belong to two different trophic levels and are character-ized by different isotopic composition (δ13C = −24.7‰, δ15N = 10.6‰,and δ13C = −31.4‰, δ15N = 5.2‰, respectively; Emslie and Patterson,2007). Thus, major krill consumption implies lower isotopic values inpenguin tissues. Furthermore, δ13C may provide valuable informationon the carbon source of food web and the regional origin of the primaryproducers. The δ13C signature of particulate organic matter decreasesfrom inshore to offshore waters (Hill et al., 2006), including Antarcticwaters (Trull and Armand, 2001), and from surface waters to sea ice(Henley et al., 2011). Moreover, phytoplankton blooms, characteristicof high productivity water, lead to more positive δ13C values of particu-late organic carbon of waters. Various authors also have reported thatδ13C values of fish should be higher in benthic than in pelagic species, be-cause benthic organisms have higher 13C values when compared to pe-lagic organisms (France, 1995; Kaehler et al., 2000; Dunton, 2001;Pruell et al., 2003; Cherel et al., 2011). Furthermore, the δ13C values offish caught in northern waters should be higher than that of southernspecies, because marine plankton δ13C, and thus consumer δ13C, varieswith latitude in oceanic waters (Rau et al., 1982), including the SouthernOcean (Trull and Armand, 2001; Cherel and Hobson, 2007; Jaeger et al.,2010; Quillfeldt et al., 2010).

In our case, these oceanographic features, and in particular the δ13Clatitudinal gradient and the higher δ13C composition of benthic taxa,such as Trematomus bernacchii, could then reasonably explain thehigher δ13C values recorded at Terra Nova Bay and along the ScottCoast, compared to those found on Ross Island.

δ15N values combined with dietary remains afford further informa-tion about penguin diet. Screening of modern samples resulted in veryfew dietary remains (less than 10 in 150 g of matrix sorted). Most die-tary remains (up to 67 in 150 g) were found in Holocene samples thatalso showed higher δ15N values. The number of dietary remains repre-sents a discriminating factor between the Scott Coast and Ross Island.Samples from Ross Island that are dated older than 560 BP containless than 15 prey items (fish bones and/or otoliths) in 150 g of matrix,whereas dietary remains found in the Scott Coast samples can varyfrom 2 to 85 per 150 g. Moreover, the Scott Coast stable isotope datashowed higher δ15N values typical of a predominantly fish-based diet.

Previous studies carried out in different areas of the Victoria Landcoast documented the same correlation between δ15N values andthe number of dietary remains. At Terra Nova Bay, for example, δ15Nb+12‰ was measured in samples without dietary remains, whereassamples with a higher δ15N also provided a great number of dietary re-mains (N20 in 150 g) (Lorenzini et al., 2009, 2010).

5.3. C and N isotopes

Between ca. 8000 and ca. 1500 cal BP, Scott Coast samples show agenerally constant carbon isotope signature of −25.3 ± 0.7‰. Al-though some older Ross Island samples lie in that range, those youngerthan 1500 cal BP tend towardsmore negative δ13C values. As illustratedin Fig. 2, modern samples from Ross Island and Terra Nova Bay show a

Fig. 2. δ13C values from guano samples during the investigated period. Squares and circlesindicate samples from Ross Island and Scott Coast, respectively. Diamonds indicate previ-ously published data from Terra Nova Bay (Lorenzini et al., 2010). Dashed lines indicateδ13C mean values of potential dietary sources (Emslie and Patterson, 2007).

Fig. 3. δ15N values from guano samples during the investigated period. Squares and circlesindicate samples from Ross Island and Scott Coast, respectively. Diamonds indicate previ-ously published data from Terra Nova Bay (Lorenzini et al., 2010). Dashed lines indicateδ15N mean values of potential dietary sources (Emslie and Patterson, 2007).

26 S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

divergence in δ13C with the latter values in general more positive than−26‰, with the exception of only two samples that show more nega-tive values,while no valuemore positive than−27‰has been recordedin Ross Island samples. This divergence seems to have started at least by2000 cal BP and progressively increased towards the modern samples.This could be explained by the different δ13C composition of POC atlower latitude due to the latitudinal gradient (Norkko et al., 2007)and/or to waters consequently more productive at Terra Nova Bay be-cause of the presence of the polynya off Inexpressible Island. In fact, alatitudinal δ13C gradient in Southern Ocean has been reported due towater masses and fronts with different physical and biological charac-teristics encircling the Antarctic continent (Orsi et al., 1995; Cherelet al., 2011). Consequently, bulk δ13C POC values in Southern Ocean sur-face waters are more negative (−33 compared to −21‰) relative toPOC from lower-latitude surface waters (Rau et al., 1991a,b; Dehairset al., 1997; Villinski et al., 2000). Significant enrichment in δ13C ofPOC from the Southern Ocean has been documented in the water col-umn and in habitats associated with sea ice and seasonally during sum-mer (Fischer, 1991; Rau et al., 1991b; Dunbar and Leventer, 1992;McMinn et al., 1999; Villinski et al., 2000; Kennedy et al., 2002; Arrigoet al., 2003). Other factors that may be responsible for isotopic enrich-ment include high growth rates of primary producers during periodsof high productivity, species-specific isotope fractionation, increasedheterotrophic recycling in the upper water column, and bloom-relateddrawdown of [CO2(aq)] in stratified surface waters (Villinski et al.,2000, 2008).

Radiocarbon dating indicates a narrow temporal window between1000 and 2000 cal BP when Ross Island and Scott Coast guano samplescoexisted. During this period, Ross Island generally records δ15N valuesrelatively lower than those measured in the Scott Coast samples, eventhough thedifference is not large. These lowvalues persist in Ross Islandsamples up to the present time (Fig. 3). Furthermore, as is the case forRoss Island, Terra Nova Bay guano samples record a shift towardlower δ15N values in the modern samples (Lorenzini et al., 2010). OnRoss Island the low percentage of dietary remains and the lower δ15Nvalues, especially in modern samples, probably indicate a mixed diet,but one predominantly based on krill exploitation. However, comparedto δ13C values, dietary inferences obtained by analyzing δ15N values of

penguin guano have to be considered as a touchier issue. As reportedin previous studies (Lorenzini et al., 2010), the gap between old andmodern nitrogen isotope ratios is too high (_δ15Nold-new ≅ +20‰) tobe interpreted simply as a result of nutrient uptake. In fact, a changein the diet could explain an isotopic increase of about 3.4‰ with eachtrophic level (Minagawa and Wada, 1984; Vander Zanden andRasmussen, 2001). Mizutani and Wada (1988) reported an increase inδ15N of residual nitrogen, up to 10‰ caused by a nitrogen isotope frac-tionation associatedwith the volatilization of ammonia after excreta de-position. Thus, the very high δ15N values of ornithogenic samplescompared to modern guano and the shift observed between fossil andmodern guano samples could be reasonably explained by adducing anammonia volatilization effect. On the other hand, it is important to con-sider that the ammonia volatilization effect happens shortly after excre-ta deposition (Mizutani and Wada, 1985) but in our case not onlymodern but also older samples show low δ15N values: the shift towardlower δ15N started at least by about 2000 cal BP.

6. Conclusion

This study provides new insights into the Holocene paleoecologicaland paleoenvironmental evolution of the Scott Coast and McMurdoSound area. Together with the screening of guano samples, C and N iso-tope analyses of penguin guano allowed us to reconstruct a detailed andcomprehensive paleodietary record. Taxonomic identification of fishotoliths indicated Pleuragramma antarcticum as the most eaten non-euphausiid prey throughout the investigated period (ca 7300 cal BP topresent). Only a very low percentage of remains belongs to Trematomussp. genus. Compared to guano samples collected along the Scott Coast,samples collected at Cape Bird on Ross Island (3850 cal BP to present)generally show fewer fish bones and otoliths and a lower diversity ofpreyed fish taxa, with a non-euphausiid diet almost entirely based onsilverfish consumption.

Holocene samples from Terra Nova Bay and Scott Coast showed analmost perfect overlap for both carbon and nitrogen isotopic ratios.Ross Island Holocene samples generally showed lower δ13C values butδ15N values similar to those of Terra Nova Bay and Scott Coast. Com-pared to Terra Nova Bay, modern samples from Ross Island show the

27S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

same gradual shift toward lower δ15N but a divergence in δ13C values,which started at least by about 2000 cal BP. Terra Nova Bay samples dis-play values higher than−26‰ (only two samples showmore negativevalues), while no value N−27‰ has been recorded at Terra Nova Bay.

The δ15N trend of Ross Island samples agreeswith the dietary changein modern samples also recorded in Terra Nova Bay and indicates anincrease in the krill consumption in themodern samples. The strong di-vergence apparent in the δ13C values between the two areas could beexplained by (paleo)oceanographic features and physical and biologicalcharacteristics of the Ross Sea water masses. In particular, the presenceof more productive water and of krill specimens with consequentlyhigher δ13C values at Terra Nova Bay could reasonably explain thehigher δ13C values recorded in this area. Also the higher contributionof fish remains belonging to benthic fish taxa, such as Trematomusbernacchii, could contribute to an increase in the δ13C values (France,1995; Kaehler et al., 2000; Dunton, 2001; Pruell et al., 2003; Cherelet al., 2011).

Less certain and more complex is the issue regarding thepaleoenvironmental and paleoecological evolution along the ScottCoast. The Adélie penguin colonization history of this region documentsa rapid abandonment of the Scott Coast colonies during the Late Holo-cene (Baroni and Orombelli, 1994; Hall et al., 2006). However, beforethis abandonment, the C and N isotopic composition of guano samplesindicate the persistence along the Scott Coast of paleoecological condi-tions similar to those recorded at Terra Nova Bay. Interestingly, we notethat the δ13C divergence between Terra Nova Bay and Ross Island thatoccurred at least by 2000 cal BP postdates the abandonment of theScott Coast colonies (Hall et al., 2006). The (paleo)oceanographic fea-tures of the Ross Sea circulation could explain the different evolutionsduring the Late Holocene of these two regions of Victoria Land Coast.Norkko et al. (2007) indicated that the Scott Coast is influenced byless-productive water due to a plankton-depleted current from under-neath the Ross Ice Shelf. The persistence of cold stenothermal condi-tions caused the progressive cooling that induced expansion andpersistence of sea ice coverage. So, the divergence in δ13C observed inRoss Island and Terra Nova Bay samples and the concurrent abandon-ment of the Scott Coast might suggest the stability and the persistenceof the polynya for the Terra Nova Bay area, but the establishment ofnew water circulation in western McMurdo Sound. In contrast toTerra Nova Bay, which experienced a longer period with persistentopen water and a large productive polynya (Faranda et al., 2000) thewater offshore the Scott Coast today is oligotrophic (Dayton andOliver, 1977),which is extremely disadvantageous for penguin survival.We suggest that onset of these conditions probably started earlier than2000 cal BP could explain the abandonment of the Scott Coast in theLate Holocene.

This study highlights the usefulness of developing long-term guanosampling and data bases on isotopic composition of key marine organ-isms to track potential changes in their isotopic niches and in the carry-ing capacity of the environment.

Acknowledgment

Thisworkwas funded by thefinancial support of the Italian NationalProgram on Antarctic Research (PNRA) and executed in the frameworkof the PNRA and the US National Science Foundation (NSF). Isotopicanalyses were performed at the IGG-CNR of Pisa (Italy). New radiocar-bon dates on penguin remains used in this work were performed atthe National Ocean Sciences Accelerator Mass Spectrometry Facility(NOSAMS) at the Woods Hole Oceanographic Institution (USA).

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.palaeo.2013.12.014.

References

Ainley, D.G., 2002. The Adélie Penguin: Bellwether of Climate Change. Columbia Universi-ty Press, New York.

Ainley, D.G., Ballard, G., Barton, K.J., Karl, B.J., Rau, G.H., Ribic, C.A., Wilson, P.R., 2003.Spatial and temporal variation of diet within a presumed metapopulation of AdéliePenguins. Condor 105, 95–106.

Ainley, D.G., Hobson, K.A., Crosta, X., Rau, G.H., Wassenaar, L.I., Augustinus, P.C., 2006. Ho-locene variation in the Antarctic coastal food web: linking dD and d13C in snow pe-trel diet and marine sediments. Mar. Ecol. Prog. Ser. 306, 31–40.

Arrigo, K.R., Robinson, D.H., Dunbar, R.B., Leventer, A.R., Lizotte, M.P., 2003. Physical con-trols of chlorophyll a, POC, and TPN distributions in the pack ice of the Ross Sea,Antarctica. J. Geophys. Res. 108, 3316.

Ballard, G., Dugger, K.M., Nur, N., Ainley, D.G., 2010. Foraging strategies of Adéliepenguins: adjusting body condition to cope with environmental variability. Mar.Ecol. Prog. Ser. 405, 287–302.

Baroni, C., Hall, B.L., 2004. A new Holocene relative sea-level curve for Terra Nova Bay,Victoria Land, Antarctica. J. Quat. Sci. 19, 377–396.

Baroni, C., Orombelli, G., 1991. Holocene raised beaches at Terra Nova Bay, Victoria Land,Antarctica. Quat. Res. 36, 157–177.

Baroni, C., Orombelli, G., 1994. Abandoned penguin rookeries as Holocene paleoclimaticindicators in Antarctica. Geology 22, 23–26.

Burnham, K.K., Burnham, W.A., Newton, I., 2009. Gyrfalcon Falco rusticolus post-glacialcolonization and extreme long-term use of nest-sites in Greenland. Ibis 151,514–522.

Busekist, J., Vacchi, M., Albertelli, G., 2007. Anfibo Base — a computer-based system foridentification of fish bones from Antarctic waters. Programma Nazionale di Ricerchein Antartide.

Cherel, Y., Hobson, K.A., 2007. Geographical variation in carbon stable isotope signaturesof marine predators: a tool to investigate their foraging areas in the Southern Ocean.Mar. Ecol. Prog. Ser. 329, 281–287.

Cherel, Y., Koubbi, P., Giraldo, G., Penot, F., Tavernier, E., et al., 2011. Isotopic niches of fish-es in coastal, neritic and oceanic waters of Adélie Land, Antarctica. Polar Sci. 5,286–297.

Codron, D., Codron, J., Lee-Thorp, J.A., Sponheimer, M., de Ruiter, D., 2005. Animal diets inthe Waterberg based on stable isotopic composition of faeces. S. Afr. J. Wildl. Res. 35(1), 43–52.

Dayton, P.K., Oliver, J.S., 1977. Antarctic soft-bottom benthos in oligotrophic and eutro-phic environments. Science 197, 55–58.

De Bruyn, M., Hall, B., Chauke, L., Baroni, C., Koch, P., Hoelzel, A.R., 2009. Rapid response ofa marine mammal species to Holocene climate and habitat change. PLoS Genet. 5 (7),1–11.

Dehairs, F., Shopova, D., Ober, S., Veth, C., Goeyens, L., 1997. Particulate barium stocks andoxygen consumption in the Southern Ocean mesopelagic water column duringspring and early summer: relationship with export production. Deep Sea Res. PartII 44 (1–2), 497–516.

Dunbar, R.B., Leventer, A., 1992. Seasonal variation in carbon isotopic compositionof Antarctic sea ice and open-water plankton communities. Antarct. J. US 7,79–81.

Dunton, K.H., 2001. δ15N and δ13C measurements of Antarctic Peninsula fauna: trophic re-lationships and assimilation of benthic seaweeds. Am. Zool. 41, 99–112.

Emslie, S.D., Patterson, W.P., 2007. Abrupt shift in δ13C and δ15N values in Adélie penguineggshell in Antarctica. Proc. Natl. Acad. Sci. U. S. A. 104 (28), 11666–11669.

Emslie, S.D., Woehler, E.J., 2005. A 9000-year record of Adélie penguin occupation anddiet in the Windmill Islands, East Antarctica. Antarct. Sci. 17, 57–66.

Emslie, S.D., Berkman, P.A., Ainley, D.G., Coats, L., Polito, M., 2003. Late-Holocene initiationof ice-free ecosystems in the southern Ross Sea, Antarctica. Mar. Ecol. Prog. Ser. 262,19–25.

Faranda, F.M., Guglielmo, L., Ianora, A., 2000. Ross Sea Ecology. Springer-Verlag, Berlin,Germany.

Fischer, G., 1991. Stable carbon isotope ratios of plankton carbon and sinking organic mat-ter from the Atlantic sector of the Southern Ocean. Mar. Chem. 35, 581–596.

France, R.L., 1995. Carbon-13 enrichment in benthic compared to planktonic algae:foodweb implications. Mar. Ecol. Prog. Ser. 124, 307–312.

Hall, B., 2009. Holocene glacial history of Antarctica and the subantarctic islands. Quat. Sci.Rev. 28, 2213–2230.

Hall, B.L., Denton, G.H., 1999. New relative sea-level curves for the southern Scott Coast,Antarctica: evidence for Holocene deglaciation of the western Ross Sea. J. Quat. Sci.14, 641–650.

Hall, B.L., Baroni, C., Denton, G.H., 2004. Holocene relative sea-level history of the SouthernVictoria Land Coast, Antarctica. Glob. Planet. Chang. 42, 241–263.

Hall, B.L., Hoelzel, A.R., Baroni, C., Denton, G.H., Le Boeuf, B.J., Overturf, B., Töpf, A.L., 2006.Holocene elephant seal distribution implies warmer-than-present climate in the RossSea. Proc. Natl. Acad. Sci. U. S. A. 103, 10213–10217.

Hall, B.L., Henderson, G.M., Baroni, C., Kellogg, T.B., 2010. Constant Holocene Southern-Ocean 14C reservoir ages and ice-shelf flow rates. Earth Planet. Sci. Lett. 296,115–123.

Henley, S.F., Annett, A.L., Ganeshram, R.S., Carson, D.S., Weston, K., Crosta, X., Tait, A.,Dougans, J., Fallick, A.E., Clarke, A., 2011. Factors influencing the stable carbon isotopiccomposition of suspended and sinking organic matter in the coastal Antarctic sea iceenvironment. Biogeosci. Discuss. 8, 11041–11088.

Hill, J., McQuaid, C., Kaehler, S., 2006. Biogeographic and nearshore/offshore trends in iso-tope ratios of intertidal mussels and their food sources around the coast of southernAfrica. Mar. Ecol. Prog. Ser. 318, 63–73.

Hobson, K.A., Montevecchi, W.A., 1991. Stable isotopic determination of trophic relation-ships of great auks. Oecologia 87, 528–531.

28 S. Lorenzini et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 395 (2014) 21–28

Hobson, K.A., Welch, H.E., 1992. Determination of trophic relationships within a high Arc-tic marine food web using δ13C and δ15N analysis. Mar. Ecol. Prog. Ser. 84, 9–18.

Jaeger, A., Lecomte, V.J., Weimerskirch, H., Richard, P., Cherel, Y., 2010. Seabird satellitetracking validates the use of latitudinal isoscapes to depict predators' foraging areasin the Southern Ocean. Rapid Commun. Mass Spectrom. 24, 3456–3460.

Kaehler, S., Pakhomov, E.A., McQuaid, C.D., 2000. Trophic structure of the marine foodweb at the Prince Edward Islands (Southern Ocean) determined by δ13C and δ15Nanalysis. Mar. Ecol. Prog. Ser. 208, 13–20.

Kennedy, H., Thomas, D.N., Kattner, G., Haas, C., Dieckmann, G.S., 2002. Particulate organicmatter in Antarctic sea ice: concentration and stable isotopic composition. Mar. Ecol.Prog. Ser. 238, 1–13.

La Mesa, M., Eastman, J.T., Vacchi, M., 2004. The role of notothenioid fish in the food webof the Ross Sea shelf waters: a review. Polar Biol. 27, 321–338.

Lambert, D.M., Ritchie, P.A., Millar, C.D., Holland, B., Drummond, A.J., Baroni, C., 2002.Rates of evolution in ancient DNA from Adélie penguins. Science 295, 2270–2273.

Lambert, D.M., Millar, C., Swaminathan, S., Baroni, C., 2010. Evolution on a frozen continent:ancient-DNA studies of Adélie penguins combined with a detailed picture of a remark-able continent's geological past provide a window on evolution. Am. Sci. 98, 386–393.

Leopold,M., VanDamme, C., VanDerVeer,H., 1998.Diet of cormorants and the impact of cor-morant predation on juvenile flatfish in the DutchWadden Sea. J. Sea Res. 40, 93–107.

Lorenzini, S., Olmastroni, S., Pezzo, F., Salvatore, M.C., Baroni, C., 2009. Holocene Adéliepenguin diet in Victoria Land, Antarctica. Polar Biol. 32, 1077–1086.

Lorenzini, S., Baroni, C., Fallick, A.E., Baneschi, I., Salvatore, M.C., Zanchetta, G., Dallai, L.,2010. Stable isotopes reveal Holocene changes in the diet of Adélie Penguins inNorthern Victoria Land (Ross Sea, Antarctica). Oecologia 164, 911–919.

Lorenzini, S., Baneschi, I., Fallick, A.E., Salvatore, M.C., Zanchetta, G., Dallai, L., Baroni, C.,2012. Insights into the Holocene environmental setting of Terra Nova Bay region(Ross Sea, Antarctica) from oxygen isotope geochemistry of Adélie penguin eggshells.The Holocene 22 (1), 63–69.

McCutchan, J.H., Lewis, W.M., Kendall, C., McGrath, C.C., 2003. Variation in trophic shift forstable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102, 378–390.

McMinn, A., Skerratt, J., Trull, T., Ashworth, C., Lizotte, M., 1999. Nutrient stress gradient inthe bottom 5 cm of fast ice, McMurdo Sound, Antarctica. Polar Biol. 21, 220–227.

Minagawa, M., Wada, E., 1984. Stepwise enrichment of δ15N along food chains: furtherevidence and the relation between δ15N and animal age. Geochim. Cosmochim.Acta 48, 1135–1140.

Mizutani, H., Wada, E., 1985. Ammonia volatilization and high 15N/14N ratio in a penguinrookery in Antarctica. Geochem. J. 19, 323–327.

Mizutani, H., Wada, E., 1988. Nitrogen and carbon isotope ratios in sea bird rookeries andtheir ecological implications. Ecology 69, 340–349.

Mizutani, H., McFarlane, D.A., Kabaya, Y., 1992a. Carbon and nitrogen isotopic signaturesof bat guanos as a record of past environments. Mass Spectrosc. 40, 67–82.

Mizutani, H., McFarlane, D.A., Kabaya, Y., 1992b. Nitrogen and carbon isotope studies of abat guano core from Eagle Creek Cave, Arizona, USA. Mass Spectrometry 40, 57–65.

Norkko, A., Thrush, S.F., Cummings, V.J., Gibbs, M.M., Andrew, N.L., Norkko, J., Schwarz, A.-M.,2007. Trophic structure of coastal Antarctic foodwebs associatedwith changes in sea iceand food supply. Ecology 88 (11), 2810–2820.

Orsi, A.H., Whitworth III, T., Nowlin Jr., W.D., 1995. On the meridional extent and fronts ofthe Antarctic Circumpolar Current. Deep Sea Res. Part I 42, 641–673.

Parmesan, C., 2006. Ecological and evolutionary responses to recent climate change.Annu. Rev. Ecol. Evol. Syst. 37, 637–669.

Polito, M., Emslie, S.D., Walker, W., 2002. A 1000-year record of Adélie penguin diets inthe southern Ross Sea. Antarct. Sci. 14, 327–332.

Post, D.M., 2002. The long and short of food-chain length. Trends Ecol. Evol. 17, 269–277.Pruell, R.J., Taplin, B.K., Cicchelli, K., 2003. Stable isotope ratios in archived striped bass

scales suggest changes in trophic structure. Fish. Manag. Ecol. 10, 329–336.Quillfeldt, P., Masello, J.F., McGill, R.A.R., Adams, M., Furness, R.W., 2010. Moving pole-

wards in winter: a recent change in the migratory strategy of a pelagic seabird?Front. Zool. 7, 15.

Rau, G.H., Sweeney, R.E., Kaplan, I.R., 1982. Plankton 13C:12C ratio change with latitude:differences between northern and southern oceans. Deep-Sea Res. 29, 1035–1039.

Rau, G.H., Sullivan, C.W., Gordon, L.I., 1991a. δ13C and δ15N variations in Weddell Sea par-ticulate organic matter. Mar. Chem. 35, 355–369.

Rau, G.H., Takahashi, T., Des Marias, D.J., Sullivan, C.W., 1991b. Particulate organicmatter δ13C variations across the Drake Passage. J. Geophys. Res. 96 (C8),15,131–15,135.

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C.,Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P.,Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac,F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S.,Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C.E., 2009. IntCal09 andMarine09 radiocarbon age calibration curves, 0–50,000 years cal BP. Radiocar-bon 51 (4), 1111–1150.

Sponheimer, M., Codron, D., Passey, B., de Ruiter, D., Cerling, T., Lee-Thorp, J., 2009. Usingcarbon isotopes to track dietary change in modern, historical, and ancient primates.Am. J. Phys. Anthropol. 140, 661–670.

StatSoft, Inc., 2005. STATISTICA (Data analysis software system). Version 7.1 .Trull, T.W., Armand, L., 2001. Insights into Southern Ocean carbon export from the δ13C of

particles and dissolved inorganic carbon during the SOIREE iron fertilisation experi-ment. Deep Sea Res. Part II 48 (11/12), 2655–2680.

Ugolini, F.C., 1972. Ornithogenic soils of Antarctica. In: Llano, G.A. (Ed.), Antarctic ter-restrial biology. American Geophysical Union, Antarctic Research Series, 20,pp. 181–193.

Vacchi, M., La Mesa, M., Dalu, M., Macdonald, J., 2004. Early life stages in the life cycle ofAntarctic silverfish, Pleuragramma antarcticum in Terra Nova Bay, Ross Sea. Antarct.Sci. 16, 299–305.

Van der Merwe, N.J., 1986. Carbon isotope ecology of herbivores and carnivores. Palaeo-ecology of Africa and the Surrounding Islands 17, 123–131.

Vander Zanden, M.J., Rasmussen, J.B., 2001. Variation in δ15N and δ13C trophic frac-tionation: implications for aquatic food web studies. Limnol. Oceanogr. 46,2061–2066.

Villinski, J.C., Dunbar, R.B., Mucciarone, D.A., 2000. Carbon 13/Carbon 12 ratios of sedi-mentary organic matter from the Ross Sea, Antarctica: a record of phytoplanktonbloom dynamics. J. Geophys. Res. 105, 14163–14172.

Villinski, J.C., Hayes, J.M., Brassell, S.C., Riggert, V.L., Dunbar, R.B., 2008. Sedimentarysterols as biogeochemical indicators in the Southern Ocean. Org. Geochem. 39,567–588.

Walther, G.R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T.J.C., Fromentin, J.M.,Hoegh-Guldberg, O., Bairlein, F., 2002. Ecological responses to recent climate change.Nature 416, 389–395.

Williams, R., McEldowney, A., 1990. A guide to the fish otoliths from waters off theAustralian Antarctic Territory, Heard and Macquire Islands. ANARE Res. Notes, 75.Antarctic Division, Australia.

Wurster, C.M., McFarlane, D.A., Bird, M.I., Ascough, P., Athfield, N.B., 2010. Stable isotopesof subfossil bat guano as a long-term environmental archive: insights from a GrandCanyon cave deposit. J. Cave Karst Stud. 72 (2), 111–121.