MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 497: 259–272, 2014doi: 10.3354/meps10606

Published February 5

INTRODUCTION

Like all seabirds, albatrosses have evolved fromterrestrial origins and, although now highly special-ized to forage at sea, they are wholly dependent

on land to breed. In the Southern Hemisphere, theoceans are vast compared with available breedingspace, and the nesting grounds of seabirds are there-fore typically characterized by important speciesdiversity and very high nesting densities. This results

© Inter-Research 2014 · www.int-res.com*Corresponding author: maelle.connan@gmail.com

Combined stomach content, lipid and stable isotopeanalyses reveal spatial and trophic partitioning amongthree sympatric albatrosses from the Southern Ocean

Maëlle Connan1,2,7,*, Christopher D. McQuaid1, Bo T. Bonnevie3, Malcolm J. Smale4,5, Yves Cherel6

1Rhodes University, Department of Zoology and Entomology, PO Box 94, Grahamstown 6140, South Africa2Percy FitzPatrick Institute of Africa Ornithology, DST/NRF Centre of Excellence, University of Cape Town,

Rondebosch 7701, South Africa3Rhodes University, Information Technology Division, PO Box 94, Grahamstown 6140, South Africa

4Port Elizabeth Museum at Bayworld, PO Box 13147, Humewood 6013, South Africa5Zoology Department, Nelson Mandela Metropolitan University, PO Box 77000, Port Elizabeth 6031, South Africa

6Centre d’Etudes Biologiques de Chizé, Centre National de la Recherche Scientifique, Villiers-en-Bois, 79360 Beauvoir-sur-Niort, France

7Present address: Zoology Department, Nelson Mandela Metropolitan University, PO Box 77000, Port Elizabeth 6031, South Africa

ABSTRACT: A combination of dietary techniques that integrate data on food and feeding habitsover days, weeks and months was used to investigate resource partitioning among 3 sympatricalbatrosses, namely the grey-headed Thalassarche chrysostoma (GHA), light-mantled sootyPhoebetria palpebrata (LMSA) and sooty Phoebetria fusca (SA) albatrosses. These medium-sizealbatrosses typically breed every 2 yr, and Marion Island (southern Indian Ocean) is the onlybreeding site where the 3 species are accessible. Stomach content analysis provided dietary infor-mation about the most recent meal, analysis of fatty acids in stomach oils about the last foragingtrip, and carbon and nitrogen stable isotope values of blood and feathers about the chick-rearing(breeding) and moulting periods, respectively. The combination of techniques highlighted a com-plex pattern regarding the spatial and trophic segregation between the 3 species. During bothseasons, SA were spatially segregated from LMSA and GHA, foraging farther north (in subantarc-tic and subtropical areas) than the 2 other species (subantarctic and Antarctic waters). When feed-ing for themselves during the breeding season (blood isotopic signatures), adults showed a clearspatial segregation. When bringing back food for their chicks (stomach contents), trophic segrega-tion became obvious, with the 2 Phoebetria species specializing mostly on squids. The results illus-trate how sympatrically breeding birds can show niche partitioning through both spatial segrega-tion and prey specialization.

KEY WORDS: Phoebetria palpebrata · Phoebetria fusca · Thalassarche chrysostoma · Trophic segregation · Spatial segregation · Moulting period · Breeding season

Resale or republication not permitted without written consent of the publisher

FREEREE ACCESSCCESS

Mar Ecol Prog Ser 497: 259–272, 2014

in increased foraging pressures on marine resourcesin the vicinity of the breeding grounds at preciselythe time when energetic requirements are greatestdue to breeding.

Subantarctic Marion Island hosts at least 25 speciesof seabirds, including 4 species of albatrosses, namelythe wandering Diomedea exulans, grey-headed Tha-lassarche chrysostoma (GHA), sooty Phoebetria fusca(SA) and light-mantled sooty Phoebetria palpebrata(LMSA) albatrosses (Ryan & Bester 2008). These 4species are biennial breeders, but unlike the wander-ing albatross, which has a breeding season lasting~1 yr, the 3 medium-size albatrosses breed in sum-mer, returning to colonies in August and September,and fledging chicks in May and June (Ryan & Bester2008). While seabirds are restricted to their breedingsites during the breeding season, individuals areable to increase their foraging range during the non-breeding season.

Many methods are used to study seabird diets, andthe combination of techniques with different tempo-ral resolutions allows dietary investigations acrossvarying time scales (Karnovsky et al. 2012). Conven-tional stomach content analysis (i.e. identification ofprey remains) provides information about the mostrecent meals of an individual. Procellariiform speciesare unique in that most of them are able to store oil ofdietary origin in their proventriculus (Warham et al.1976) and the fatty acid signatures of these stomachoils provide dietary information about the last forag-ing trip (Connan et al. 2005, 2007). Carbon (δ13C) andnitrogen (δ15N) stable isotope ratios have been usedas indicators of foraging areas and trophic level,respectively (e.g. Hobson & Clark 1992). Avian wholeblood has a rapid turnover and represents the dietintegrated over a period of a few weeks prior to sam-pling, and for birds sampled during the breeding sea-son, this will give information about the diet duringthat period (Hobson & Clark 1992). The full comple-ment of adult body feathers are replaced graduallyover the whole interbreeding period (Warham 1996),with a body feather of 6 to 10 cm typically takingabout 2 to 3 wk to grow (Jaeger et al. 2010a). Feath-ers remain isotopically inert once fully grown (Mizu-tani et al. 1990). δ13C shows well-defined latitudinalisoscapes within the Southern Ocean, decreasingtowards higher latitudes (e.g. Cherel & Hobson 2007,Jaeger et al. 2010b). Foraging during the interbreed-ing period of albatrosses can thus be examinedthrough the stable isotope carbon signatures of theirfeathers (Phillips et al. 2009, Cherel et al. 2013).

Given the constraints on breeding sites, resourceoverlap between land-breeding marine predators

might be enhanced during the breeding season andrelaxed during the non-breeding season. In an envi-ronment with limited resources, this overlap mayresult in competition, and it is then reasonable toassume that (1) sympatric species will be more likelyto show trophic segregation during the breeding season (Croxall et al. 1999), and (2) each species willreturn to its preferred foraging areas outside thebreeding season. On the other hand, if there wereunlimited resources, competition would not be ex -pected, and, therefore, neither would dietary segre-gation. The aim of the present work was to studyresource partitioning within the community of me -dium-size albatrosses (GHA, SA and LMSA) at Mar-ion Island to estimate the degree of dietary overlapand trophic segregation during the breeding andmoulting seasons using a combination of stomachcontent analysis, with lipids and stable isotopes astrophic markers (Karnovsky et al. 2008, 2012, Con-nan et al. 2010b).

MATERIALS AND METHODS

Study site and species

Fieldwork was carried out on Marion Island (PrinceEdward Islands; 46° 52’ S, 37° 51’ E). The island lies inthe subantarctic zone sensu lato of the SouthernOcean, between the Subtropical Front to the northand the Antarctic Polar Front to the south (Lutjeharms& Ansorge 2008; Fig. 1). Marion Island supports 11%of the world’s population of SA (1400 breeding pairs),8% of its GHA population (7500 breeding pairs), and3% of its LMSA population (600 breeding pairs; Ryanet al. 2009), and, with the Crozet Islands, is one of only2 places where both species of sooty albatrosses (SAand LMSA) breed in substantial numbers. Breedingsites of GHA are restricted to the south coast ofMarion Island, while SA and LMSA nest on cliffsaround the island perimeter, as well as on inland cliffsin the case of LMSA. The 3 species are of similar sizewith a wing span around 1.8 to 2.2 m (Harrison 1983),and their masses range from 2.5 kg for SA to 3.6 kg forGHA (Hockey et al. 2005).

Sampling

Samples were collected in April and May 2009from breeding adults and chicks of the 3 species.Field identification of SA and LMSA chicks was possible using the size and the colour of the ring of

260

Connan et al.: Spatial and trophic segregation among albatrosses

feathers immediately posterior to the eye; molecularidentification was used to confirm borderline cases(Connan et al. 2011).

The prey brought back by the adults during thechick-rearing period is destined for their chicks; con-sequently no distinction was made between thestomach contents of the 2 age classes. Stomach con-tents were collected when either adults or chicksregurgitated spontaneously during handling or aftergentle stomach massage. Stomach oils were sepa-rated from the prey remains when present andimmediately placed in 2 ml chloroform with 0.01%butylated-hydroxytoluene in a nitrogen-saturatedatmo sphere to avoid lipid oxidation. Both stomachcontents and oils were then stored at −20°C until subsequent analyses. For each bird, a blood samplewas taken from the tarsus vein and preserved in 70%ethanol for subsequent stable isotope (adults) orgenetic analysis (chicks). Preservation in 70% etha -nol does not significantly change the stable carbon ornitrogen isotope ratios of samples (Hobson et al.1997). Four feathers were removed from the back ofeach bird’s neck and stored in plastic bags for sub -sequent processing.

Laboratory analyses

Stomach contents. Each stomach content wasthawed and drained by gravity overnight to separate

the solid items from the residual liquidfraction. Fresh remains were separatedfrom accumulated items (mainly squidbeaks with no flesh attached), and di-vided into broad prey classes (cepha -lopods, crustaceans, fish and carrion)that were weighed to estimate their pro -portion by wet mass in the diet. Then,prey items (fresh and accumulated)were counted and identified to the low-est possible taxon, using published keysand our own reference collection. Foraccumulated squid beaks, all lower andupper beaks were identified (Xavieret al. 2011). In addition, the lower ros-tral length of beaks was measured to0.1 mm with Ver nier callipers and allo-metric equations were used to estimatesquid mantle length (see Xavier &Cherel 2009).

Lipids. Since lipid composition ofstomach oils is similar for adults andchicks (Connan et al. 2007), no age dis-

tinction was made in the present study. Total lipids ofstomach oils were extracted following Allan et al.(2010). Lipid classes were then isolated by prepara-tive thin-layer chromatography, and bands of waxesters, triacylglycerols and diacylglycerol etherswere scraped off, eluted and the 3 lipid classesweighed; stomach oils are typically rich in neutrallipids (Warham et al. 1976, Connan et al. 2007). Fattyacid methyl esters were prepared from triacylgly -cerol fractions using methods detailed in Allan et al.(2010). The fatty acid methyl ester composition ofeach sample was determined by gas chromatographyanalysis using an Agilent Technologies 7890A gaschromatographic System with a ZB-Waxplus 320 col-umn (30 m length, 0.25 mm internal diameter,0.25 µm film thickness; Zebron Corporation) withhelium as the carrier gas. The injection temperaturewas 250°C, the flame ionization detector was set at260°C, and the oven was initially set at 150°C. After5 min, the oven temperature was raised to 225°C at arate of 2.5°C min−1 and held for 9 min. Peaks wereintegrated using GC ChemStation software (AgilentTechnologies, version B.04.02), identified by com -parison with retention times of external standards(37 component fatty acid methyl ester mix Supelco,marine PUFA no. 1 Supelco, bacterial acid methylesters mix Supelco), as well as by mass spectrometryanalysis (Agilent Technologies 7000 GC/MS TripleQuad; Agilent MassHunter Qualitative Analysis, version B.03.01).

261

500 km Antarctica

South Africa

Polar Front

Subtropical Front

Subantarctic Front

60°

55°

50°

45°

40°

35°

30°S

65°

20°E 60° 65° 70°45° 50° 55°35°30°25° 40°

Marion Island

Fig. 1. Location of Marion Island in the southwestern Indian Ocean. Front positions follow Park et al. (1993)

Mar Ecol Prog Ser 497: 259–272, 2014

Each fatty acid was measured as a proportion of thetotal array of fatty acids detected (%TFAs).

Stable isotopes. Four feathers per bird were indi-vidually cleaned in a 2:1 chloroform:methanol solu-tion placed in an ultrasonic bath for 2 min, rinsed insuccessive baths of methanol and deionised water,and then dried (50°C, 24 h). Each whole body featherwas homo genised by fine cutting with stainless steelscissors. Whole blood samples were dried at 50°C for24 h and finely ground to a homogenous powder.Typically, avian whole blood does not require lipidextraction (Cherel et al. 2005). The relative iso -topic abundances of carbon (13C/12C) and nitrogen(15N/14N) were determined on 0.5 to 0.6 mg subsam-ples of materials with a Thermo Finnigan Delta XPPlus mass spectrometer interfaced with a Conflo IIIdevice to a Thermo Flash EA 1112 elemental ana-lyzer (Stable Light Isotope Unit, University of CapeTown, South Africa). Results are presented in theusual δ notation relative to Vienna PeeDee Belemniteand atmospheric N2 for δ13C and δ15N, respectively.Replicate measurements of internal laboratory stan-dards indicate measurement errors

Connan et al.: Spatial and trophic segregation among albatrosses

Polar and Subtropical fronts (Cherel & Hobson 2007,Jaeger et al. 2010b).

All statistical analyses were performed using R sta-tistical Software (R Development Core Team 2009)and PAST (Hammer et al. 2001).

RESULTS

Stomach contents

Stomach contents were collected from 39 birds (19GHA, 10 LMSA, 10 SA), with 20 of them containingoil (51% of stomach contents). Food samples ofLMSA and SA were dominated by squid remains(77 and 78% by fresh mass, respectively—squidremains dominated in 8 out of 10 LMSA samples and7 out of 10 SA samples), whereas GHA food sampleswere mostly composed of a mix of fish and squid(46% and 36% by fresh mass, respectively—fish andsquid remains dominated in 11 and 5 out of 19 con-tents, respectively; Fig. 2). Carrion was found in 3stomach contents (1 GHA, 1 LMSA, 1 SA), with thestomach content of 1 LMSA being entirely composedof penguin remains. A total of 1905 prey items wererecovered in samples from the 3 albatross species(Table 1). The digested state of many items pre-cluded their identification to species level. Crus-taceans formed a small part by mass for the 3 species(15%, 2% and 3% of fresh items for GHA, LMSA andSA, respectively), but dominated by number. Fifty-seven percent of the prey items were the subant -arctic krill Euphausia vallentini, recovered from 1 SAsample, followed by the Antarctic krill Euphausiasuperba (27%) and the amphipod Themisto gaudi -chaudii (7%) Antarctic krill were recovered in 20%of LMSA samples and 11% of GHA samples, butnever in SA samples. Most of the fish remains couldnot be identified to species level because otolithswere too eroded. The fish family Macrouridae wasthe most abundant followed by Gempylidae in GHAstomach contents. Although cephalopods dominated

by fresh mass, only 4 species could be identified fromintact buccal masses: Kondakovia longimana, Histio-teuthis eltaninae, Alluroteuthis antarcticus and Bato-teuthis skolops (Table 1).

Accumulated squid beaks were found in 58%, 60%and 70% of the GHA, LMSA and SA stomach con-tents, respectively. The 3 species Histioteuthis eltan-inae, Galiteuthis glacialis and Kondakovia longi-mana were the most abundant cephalopod prey forall 3 albatross species (Table 2). K. longimana wasthe most abundant species in GHA samples, andH. eltaninae in LMSA and SA samples. Estimatedsizes of cephalopods ranged from Mastigoteuthis sp.A (Clarke) with an estimated mantle length of 5.7 cmto an individual of K. longimana with a mantle lengthof 60.5 cm. For the latter species, estimated mantlelengths in GHA stomachs were all be tween 20.6 and39.9 cm, however 75% of K. longimana beaks foundin GHA stomachs came from juveniles and could notbe measured because they were too eroded.

Fatty acids

Triacyglycerols and wax esters were recovered inall 20 stomach oil samples (9 each from GHA andLMSA, 2 from SA), while diacylglycerol ethers werefound in only 30% of the oils (11% of GHA, 44% ofLMSA and 50% of SA samples) (Fig. 3).

Twenty-one fatty acids were detected at >0.5% ofTFAs in the triacylglycerol fractions (Table 3). Half ofthe fatty acids were monounsaturated with oleic acid(18:1(n-9)) being dominant, followed by palmitic acid(16:0) and eicosapentaenoic acid (20:5(n-3)). Overall,differences in the fatty acid compositions between the3 albatross species were marginally non-significant(ANOSIM: global R = 0.157, p = 0.053). This result wasconfirmed with a non-metric multidimensional scalingplot analysis (data not shown; 2D stress value = 0.14).

The optimisation model estimates were variablebetween individuals and no clear distinction couldbe made among the 3 albatross species (ANOSIM:

263

GHA(n = 19)

LMSA(n = 10)

SA(n = 10)

Others

Cephalopods

Crustaceans

Fish

45.8%

36.5%

14.5%

3.2%

77.2%

8.9%11.7%2.2%

78.5%

17.5%

3.5%0.5%

Fig. 2. Broad prey class composition (% byfresh mass) of the stomach contents for the 3albatross species (grey-headed [GHA], light-mantled sooty [LMSA] and sooty [SA] alba-

trosses)

Mar Ecol Prog Ser 497: 259–272, 2014

global R = 0.024, p = 0.346). This result was con-firmed with a non-metric multidimensional scalingplot analysis (data not shown; 2D stress value = 0.04).Overall, triacylglycerol fatty acids originated fromfish species for 90% of the oils followed by cepha lo -

pods and crustaceans (Fig. 4). When looking at preyspecies, the fish Champsocephalus gunnari andCory phaenoides armatus, the cephalopod Gonatusantarcticus and the crustacean Euphausia superbawere found to be the main prey of the albatrosses.

264

GHA (n = 19) LMSA (n = 10) SA (n = 10)Occurrence Number Occurrence Number Occurrence Number

N (%) N (%) N (%) N (%) N (%) N (%)

CrustaceansAmphipodaEurytheneidaeEurythenes obesus 1 (5) 1 (0.1) 1 (10) 1 (1.6)

Gammaridea sp. 1 (5) 1 (0.1)HyperiidaeHyperiella antarctica 1 (5) 2 (0.3)Themisto gaudichaudii 7 (37) 122 (16.6) 1 (10) 1 (1.6) 2 (20) 7 (0.6)

VibiliidaeCyllopus magellanicus 1 (5) 1 (0.1)

CirripediaLepas australis (cypris larvae) 3 (16) 10 (1.4)Lepas sp. 7 (37) 11 (1.5)

DecapodaPasiphaea scotiae 1 (5) 2 (0.3) 1 (10) 1 (1.6)

EuphausiaceaEuphausia superba 2 (11) 474 (64.4) 2 (20) 46 (73.0)Euphausia vallentini 1 (5) 8 (1.1) 1 (10) 1081 (97.7)Euphausia sp. 7 (37) 45 (6.1) 3 (30) 3 (4.8) 1 (10) 7 (0.6)

MysidaNeognathophausia gigas 1 (5) 1 (0.1) 1 (10) 1 (0.1)

Unidentified Isopoda 1 (5) 1 (0.1)Unidentified crustacean 3 (16) 4 (0.5) 1 (10) 1 (1.6)

GastropodaClione sp. 5 (26) 26 (3.5)

FishPhosichthyidaePhosichthys argenteus 1 (5) 1 (0.1)

MyctophidaeGymnoscopelus piabilis 1 (10) 1 (0.1)

MoridaeAntimora rostrata 1 (10) 1 (0.1)

Macrouridae?Coryphaenoides armatus 2 (11) 3 (0.4)

GempylidaeParadiplospinus gracilis 2 (11) 2 (0.3)

Unidentified fish 10 (53) 13 (1.8) 5 (50) 5 (7.9) 1 (10) 1 (0.1)

CephalopodsOnychoteuthidaeKondakovia longimana 1 (10) 1 (0.1)

HistioteuthidaeHistioteuthis eltaninae 1 (5) 1 (0.1) 1 (10) 1 (1.6) 2 (20) 3 (0.3)

NeoteuthidaeAlluroteuthis antarcticus 1 (10) 1 (1.6)

BatoteuthidaeBatoteuthis skolops 1 (5) 1 (0.1)

Unidentified squid 6 (32) 6 (0.8) 3 (30) 3 (4.8) 3 (30) 3 (0.3)

OthersCarrion 1 (5) 1 (10) 1 (10)

Table 1. Frequency of occurrence (% of stomach contents) and number of prey items (% of total number of prey items) recov-ered in the fresh fraction of stomach contents in the 3 albatross species (grey-headed [GHA], light-mantled sooty [LMSA] and

sooty [SA] albatrosses)

Connan et al.: Spatial and trophic segregation among albatrosses 265

GH

A (

n =

19)

LM

SA

(n

= 1

0)S

A (

n =

10)

Occ

urr

ence

Nu

mb

erS

qu

id s

ize

Occ

urr

ence

Nu

mb

erS

qu

id s

ize

Occ

urr

ence

Nu

mb

erS

qu

id s

ize

N(%

)N

(%)

(cm

)N

(%)

N(%

)(c

m)

N(%

)N

(%)

(cm

)

Om

mas

trep

hid

ae s

p.

1(1

0)1

(1)

On

ych

oteu

thid

aeM

orot

euth

is i

ng

ens

2(1

1)2

(4)

2(2

0)6

(5)

Mor

oteu

this

kn

ipov

itch

i2

(11)

2(4

)38

.32

(20)

3(2

)38

.7−

39.6

1(1

0)2

(2)

38.5

−38

.6K

ond

akov

ia l

ong

iman

a9

(47)

20(4

3)28

.7 ±

8.2

5(5

0)11

(9)

42.9

2(2

0)6

(5)

43.4

± 1

8.5

Psy

chro

teu

thid

aeP

sych

rote

uth

is g

laci

alis

2(2

0)9

(7)

31.9

± 4

.1

Gon

atid

aeG

onat

us

anta

rcti

cus

1(1

0)1

(1)

19.9

An

cist

roch

eiri

dae

An

cist

roch

eiru

s le

sueu

ri2

(20)

2(2

)31

.1−

35.2

Oct

opot

euth

idae

Tan

ing

ia d

anae

1(5

)1

(2)

49.2

Oct

opot

euth

issp

.1

(10)

1(1

)18

.1

His

tiot

euth

idae

His

tiot

euth

is a

tlan

tica

3(3

0)6

(5)

11.6

± 1

.3H

isti

oteu

this

bon

nel

lii

corp

usc

ula

2(2

0)10

(9)

6.4

± 0

.2H

isti

oteu

this

elt

anin

ae4

(21)

6(1

3)8.

2 ±

0.5

6(6

0)69

(53)

8.7

± 0

.65

(50)

34(3

1)8.

7 ±

0.6

His

tiot

euth

is m

iran

da

3(3

0)8

(7)

9.8

± 0

.4H

isti

oteu

this

spp

.3

(30)

13(1

2)

Neo

teu

thid

aeA

llu

rote

uth

is a

nta

rcti

cus

2(1

1)2

(4)

17.9

−19

.94

(40)

10(8

)18

.5 ±

1.2

2(2

0)3

(3)

19.2

−19

.6N

otot

euth

is d

imeg

acot

yle

1(1

0)1

(1)

Cyc

lote

uth

idae

Cyc

lote

uth

is a

kim

ush

kin

i1

(10)

1(1

)15

.8

Mas

tig

oteu

thid

aeM

asti

got

euth

is p

sych

rop

hil

a1

(10)

1(1

)1

(10)

1(1

)11

.8M

asti

got

euth

issp

. A (

Cla

rke)

1(1

0)1

(1)

5.7

Ch

irot

euth

idae

Ch

irot

euth

is v

eran

yi1

(10)

1(1

)17

.72

(20)

2(2

)16

.2−

19.1

Bat

oteu

thid

aeB

atot

euth

is s

kol

ops

1(1

0)1

(1)

Cra

nch

iid

aeG

alit

euth

is g

laci

alis

5(2

6)10

(21)

44.7

± 2

.84

(40)

13(1

0)45

.1 ±

2.3

3(3

0)17

(15)

43.5

± 3

.1M

eson

ych

oteu

this

ham

ilto

ni

1(1

0)1

(1)

Tao

niu

ssp

. B (

Vos

s)1

(10)

1(1

)

Un

iden

tifi

ed b

eak

s3

(16)

4(9

)2

(20)

2(2

)1

(10)

2(2

)5.

9

Tot

al47

129

111

Tab

le 2

. Fre

qu

ency

of

occu

rren

ce (

% o

f st

omac

h c

onte

nts

) an

d n

um

ber

of

accu

mu

late

d s

qu

id b

eak

s (%

of

tota

l ac

cum

ula

ted

bea

ks)

rec

over

ed i

n t

he

stom

ach

con

ten

ts

in t

he

3 al

bat

ross

sp

ecie

s (g

rey-

hea

ded

[G

HA

], l

igh

t-m

antl

ed s

ooty

[L

MS

A]

and

soo

ty [

SA

] al

bat

ross

es).

Siz

e d

ata

are

mea

n ±

SD

or

ran

ge

Mar Ecol Prog Ser 497: 259–272, 2014

Stable isotopes

During the breeding season, blood δ13C valueswere higher in SA than in GHA and LMSA (H = 28.2,p < 0.001, SA > GHA = LMSA; Table 4, Fig. 5) andblood δ15N values were lower in GHA than in SA (H =26.7, p < 0.001, SA ≥ LMSA = GHA; Table 4, Fig. 5A).This result was illustrated by the corrected standardellipse areas for the 3 albatross species that differedin size, shape and position (Fig. 6A). GHA had thesmallest corrected standard ellipse area (0.75‰2) fol-lowed by SA and LMSA (0.86 and 2.55‰2, respec-tively). Differences be tween spe cies were also high-lighted with the absence of overlap in the 40%credibility intervals between GHA and SA correctedstandard ellipse areas as well as between SA andLMSA, and only 0.14‰2 overlap between GHA andLMSA corrected standard ellipse areas.

Feather isotopic signatures of both δ13C and δ15Nshowed that SA is segregated from GHA and LMSA(δ13C: ANOVA F = 35.7, p = < 0.001, SA > GHA =LMSA; δ15N: Wald test = 31.3, p < 0.001, SA >LMSA > GHA; Table 4, Fig. 5B). δ15N and δ13C valueswere positively linearly correlated for LMSA (Pear-son’s correlation, r = 0.908, p < 0.001) and SA (Pear-

266

GHA (n = 9) LMSA (n = 9) SA (n = 2)

14:0 5.01 ± 1.70 4.96 ± 2.00 2.71−3.7816:0 14.72 ± 2.81 13.53 ± 2.96 10.87−15.7318:0 1.87 ± 0.43 1.86 ± 0.67 1.57−3.4416:1(n-9) 0.23 ± 0.16 0.52 ± 0.22 1.13−0.7916:1(n-7) 6.88 ± 0.79 7.45 ± 1.67 5.08−5.3418:1(n-9) 20.31 ± 3.77 22.36 ± 2.95 25.77−28.3218:1(n-7) 6.28 ± 1.51 7.04 ± 1.17 5.57−4.8518:1(n-5) 0.62 ± 0.19 0.59 ± 0.15 0.60−0.5020:1(n-11) 0.42 ± 0.21 0.71 ± 0.53 0.53−0.5320:1(n-9) 7.17 ± 2.42 7.26 ± 2.73 7.07−8.9320:1(n-7) 0.64 ± 0.13 0.70 ± 0.18 0.45−0.6722:1(n-11) 2.21 ± 1.04 2.51 ± 1.37 1.33−3.4822:1(n-9) 1.55 ± 0.43 1.55 ± 0.55 0.99−2.0624:1(n-9) 0.85 ± 0.28 0.91 ± 0.40 0.66−1.4618:2(n-6) 1.44 ± 0.30 1.42 ± 0.35 1.18−0.7918:4(n-3) 0.88 ± 0.21 0.65 ± 0.16 0.19−0.2620:4(n-6) 0.61 ± 0.14 0.64 ± 0.11 1.06−0.5920:4(n-3) 0.72 ± 0.18 0.74 ± 0.09 0.76−0.5020:5(n-3) 10.36 ± 2.27 7.86 ± 1.52 10.81−3.6622:5(n-3) 1.04 ± 0.30 1.34 ± 0.68 1.44−0.6822:6(n-3) 9.99 ± 1.88 8.99 ± 1.60 15.78−7.19Others 6.17 ± 0.90 6.41 ± 0.46 4.44−6.43SFA 23.36 ± 4.98 22.10 ± 5.59 16.11−25.29MUFA 47.97 ± 6.58 52.45 ± 7.62 49.57−57.58PUFA 28.67 ± 4.07 25.45 ± 3.37 34.32−17.13

Table 3. Mean values (±SD) or ranges for fatty acid compo-sition of stomach oil triacylglycerol fractions (detected at>0.5% of total fatty acids) for the 3 species (grey-headed[GHA], light-mantled sooty [LMSA] and sooty [SA] alba-trosses). SFA: saturated fatty acids; MUFA: monounsaturated

fatty acids; PUFA: polyunsaturated fatty acids

0

40

20

60

80

100

Lip

id c

lass

es (%

)

TAG WE DAGE

GHA (n = 9)

LMSA (n = 9)

SA (n = 2)

Fig. 3. Neutral lipid class composition of stomach oils (tri-acylglycerols [TAG], wax esters [WE] and diacylglycerolethers [DAGE]) of the 3 albatross species (grey-headed[GHA; black], light-mantled sooty [LMSA; dark-grey] andsooty [SA; light grey and white for samples a and b, respec-

tively] albatrosses)

1 3 5 7 9 1 3 5 7 9 1 20

100

20

40

60

80

GHA LMSA SA

Per

cent

age

of p

rey

clas

ses

Fish Cephalopod Crustacean Unknown

Fig. 4. Model estimates of the contri-bution of broad prey classes to thefatty acid fraction of stomach oil triacylglycerols for the 3 albatrossspecies (grey-headed [GHA], light- mantled sooty [LMSA] and sooty

[SA] albatrosses)

Connan et al.: Spatial and trophic segregation among albatrosses

son’s correlation, r = 0.846, p < 0.001) feathers, butnot for GHA feathers (Pearson’s correlation, r =−0.083, p = 0.422). Feathers from the 3 speciesshowed high variability in both their δ13C and δ15Nvalues, with the isotopic niche being the widest forLMSA and GHA samples followed by SA (Table 4).Corrected standard ellipse areas and convex hull

areas reflected these results, with GHA and LMSAindices being 2 to 3 times the size of SA (centredand rescaled data: corrected standard ellipse areas2.28‰2, 1.64‰2, 0.69‰2 and convex hull areas9.69‰2, 5.27‰2, 2.74‰2 for GHA, LMSA and SA,respectively; Fig. 6B). The isotopic overlap betweenGHA and SA was almost non-existent (< 0.0001‰2)for the 40% credibility interval. On the contrary, halfof the LMSA area was shared with the GHA area and

Mar Ecol Prog Ser 497: 259–272, 2014

in the proximity of the Polar Front, while SA foragedmainly in subantarctic waters (Table 4, Fig. 5A & 6A).To our knowledge, no tracking data of LMSA and SAfrom Marion Island have been published, but trackedGHA forage mainly to the southwest of the islandduring chick rearing (Nel et al. 2001). This patternwas also supported by dietary analysis, with theidentification of some high-Antarctic species only inGHA and LMSA food samples: the Antarctic krillEuphausia superba and the squid Psychroteuthis

glacialis. Noticeably, food samples represent chickfood and not the food of the adults when they feed forthemselves. Hence, it is likely that both GHA andLMSA adults include a higher proportion of Antarcticprey in their diet.

Interestingly, the identification of Phosichthys ar -genteus in GHA samples (Nel et al. 2001, Richoux etal. 2010, this study) suggests that from time to time,foraging trips are also undertaken in subtropicalwaters. P. argenteus has never been caught south of49° S (Froese & Pauly 2012), and is regularly identi-fied in the diet of the yellow-nosed albatross Thalas-sarche chlororhynchos, which forages in subtropicalwaters (Pinaud et al. 2005). In the same way, thepresence of beaks from squid with different biogeo-graphic affinities in food samples suggested that theforaging range of SA is not restricted to subantarcticwaters but also includes northern waters of the sub-tropical zone. For example, Histioteuthis mirandaand Ancistrocheirus lesueuri were identified in 30and 20% of the SA contents, respectively, and aremainly found north of the Subtropical Front (Xavier &Cherel 2009).

Blood δ15N values revealed that SA and LMSAwere feeding at a slightly higher trophic level thanGHA during chick rearing (Table 4, Fig. 5A & 6A),which is in agreement with a higher proportion ofAntarctic krill, a lower trophic level prey in the foodsamples of GHA. Moreover, δ15N values are overallhigher in squid than in fish, from 10.0 to 10.9‰ andfrom 7.3 to 10.2‰, respectively (Cherel et al. 2008,2010), and squid formed higher percentages by massin LMSA and SA than in GHA food samples. Interest-ingly, di acylglycerol ethers, which have been sug-gested as originating from squid species (Connan etal. 2007), were identified in 1 out of 2 SA oils, 4 out of9 LMSA oils and only 1 out of 9 GHA oils.

Triacyglycerol origins were more variable betweenindividuals than between the 3 species of albatrossesand even for the 2 Phoebetria species, most of the tri-acyglycerols seemed to originate from fish species.Surprisingly, the icefish Champsocephalus gunnariwas identified as one of the species from which theoils originated. This species has been identified inGHA stomach content from South Georgia (Cherel &Klages 1998); however it does not seem to occur inthe vicinity of the Prince Edward Islands (Kock &Everson 1997). This result probably reflects the factthat, despite gathering as many potential prey spe-cies as possible, some were absent from the data-base. Hence, instead of feeding on icefish, the alba-trosses could possibly have foraged on a prey speciesor a mix of prey showing similarities in fatty acid

268

–24–25 –23 –22 –21

9

10

11

12

14

Cen

tred

δ15

N (

‰)

δ15 N

(‰

)

Polar Front

Subtropical Front

A

B

–3

–2

–1

0

1

–2–3 –1 0 1

Centred δ13C (‰)

δ13C (‰)

2

2 Polar Front

Subtropical Front

–20 –19

13

Fig. 6. Stable carbon and nitrogen values of blood (A) andindividual feathers (centred and rescaled data; B) of adultgrey-headed (black), light-mantled sooty (grey) and sooty(light grey) albatrosses. Standard ellipse areas (solid lines)and convex hull areas (dashed lines) were estimated usingStable Isotope Bayesian Ellipses in R (Jackson et al. 2011).Isotopic values referring to the position of the Polar and Sub-tropical fronts follow Cherel & Hobson (2007) and Jaeger et

al. (2010b)

Connan et al.: Spatial and trophic segregation among albatrosses

triacyglycerol composition with that of icefish. Thepresence of oils derived from Antarctic krill Euphau-sia superba in all 3 species suggests that SA some-times also feed in Antarctic waters. Indeed, previousdietary investigations identified Antarctic krill in afew SA food samples (Ridoux 1994, Cooper & Klages1995).

In brief, the combination of dietary techniquesshowed that complex spatial and trophic segrega-tions exist between the 3 albatross species. Whenfeeding for themselves (blood isotopic signatures),adult SA showed a clear spatial segregation with the2 other species. When bringing back food for theirchicks (stomach contents), trophic segregation be -came obvious between GHA, feeding on a mix of fishand squids, and the 2 Phoebetria species, specializ-ing mostly on squids.

Interbreeding (moulting) period

Due to the high energetic cost of growing newfeathers, many bird species dissociate moulting andbreeding, at least at the individual level. Some alba-tross species even skip a breeding season when theirmoult has not been completed (Rohwer et al. 2011).

When taking into account the latitudinal δ13C gra-dient from the Southern Ocean (Jaeger et al. 2010b),feather δ13C values showed that, when not breeding,SA were spatially segregated from LMSA and GHAby foraging mostly in northern areas (northern sub-antarctic and subtropical waters), while the 2 otherspecies were mainly concentrated in subantarcticwaters. In addition, feather δ15N isotope signaturesshowed that the albatross species foraged at differenttrophic levels (Table 4, Figs. 5B & 6B).

During moult, GHA mainly foraged in subantarcticwaters (>70% of feathers), where they preyed at sev-eral trophic levels (7.9‰ < δ15NGHA < 14.7‰). δ15Nvalues suggested that GHA would have preyed onfish and squid (Cherel et al. 2008, 2010), as well aslower trophic level prey such as crustaceans or moreprobably jellyfish and/or salps. Black-browed alba-trosses Thalassarche melanophris have been docu-mented feeding on jellyfish umbrellas (Suazo 2008),and, using stomach temperature recorders, Catry etal. (2004) showed that GHA foraged on rapidly di -gested prey (e.g. salps and jellyfish). Taken together,these studies suggest that the role of gelatinous spe-cies has been overlooked in the diet of albatrosses.

Moulting SA foraged in northern waters (mainly inthe subtropical zone) rather than the subantarcticwaters used during the breeding season. Individual

feather data showed that all birds foraged at thesame trophic level (low BIC), and that each bird wasfaithful to that trophic level throughout moulting (lowWIC; Table 4, Fig. 6B). The positive linear correlationbetween δ15N and δ13C values detected in SA (andLMSA) feathers has been found in several other studies (e.g. birds: Phillips et al. 2009, Jaeger et al.2010a,b; squids: Cherel & Hobson 2005), suggestingnot only that δ13C decreases with latitude, but alsothat the δ15N baseline exhibits latitudinal changeswithin the Southern Ocean (Jaeger et al. 2010a).

In contrast, individual feather analysis of LMSAshowed that 3 feathers had been moulted in high-Antarctic waters (δ13C

Mar Ecol Prog Ser 497: 259–272, 2014

localities, reflecting the abundance of Antarctic krillin the waters surrounding the island. At Marion andCrozet islands, a synthesis of prey biogeographyshowed that GHA, LMSA and SA are partially spa-tially segregated during chick rearing, with SA for-aging in warmer, more northerly waters than LMSAand GHA, with a substantial overlap in subantarcticwaters (Berruti & Harcus 1978, Ridoux 1994, Cooper& Klages 1995, Richoux et al. 2010, this study).

Previously, 12 GHA oils were collected in 2006 onMarion Island and 1 from an LMSA chick on Camp-bell Island, but none from SA (Warham et al. 1976,Richoux et al. 2010). The distinct fatty acid composi-tions of GHA between the samples collected in 2006and 2009 might be explained by (1) the protocolsused (Richoux et al. [2010] analysed total fatty acidcompositions, i.e. mix of lipid classes from differentprey origins) and/or (2) different prey ingested by thebirds. The LMSA oil collected on Campbell Islandshowed strong similarities to those collected in ourstudy, with triacylglycerols representing >80% oflipid classes and similar fatty acid compositions.Unfortunately, the precision given in Warham et al.(1976; group of fatty acids such as 20:1, 22:1) pre-cludes deeper comparisons of prey origins.

GHA adult blood δ13C values (Marion Island;Richoux et al. 2010, this study) together with chickfeather δ13C values (Crozet Islands; Jaeger et al.2010a) indicate foraging in cold southern waters ofthe southern Indian Ocean during the chick-rearingperiod. Adult blood δ13C values were higher for birdsfrom South Georgia, but are in agreement with birdsforaging mostly to the north of the island during earlychick rearing (Phillips et al. 2004, Anderson et al.2009, Phillips et al. 2011). No adult blood δ13C andδ15N values of LMSA and SA are available, but chickfeather δ13C values from the Crozet Islands againshow different isotopic niches for the 2 species, withSA foraging further north and at a higher trophicposition than LMSA during the chick-rearing period(Jaeger et al. 2010a).

During the interbreeding (moulting) period, featherδ13C values of adult GHA were remarkably similar inall localities and years so far investigated, with birdsforaging primarily in subantarctic waters (Andersonet al. 2009, Phillips et al. 2009, Cherel et al. 2013, thisstudy). Results are more variable for the 2 Phoebetriaalbatrosses. Depending on localities and years,feather δ13C values of LMSA ranged from −19.5 to−21.2‰, with very large variances (Jaeger et al.2010a, Cherel et al. 2013, this study), except at SouthGeorgia (Phillips et al. 2009). The low standard devi-ation at the latter locality is likely the result of meas-

uring stable isotopes on a pool of body feathersinstead of a single feather, thus blurring any intra-individual variation (Jaeger et al. 2010a). Otherwise,the interbreeding strategy of LMSA is consistentamong colonies, with birds moulting from high-Antarctic to subtropical waters with a focus in thesubantarctic zone. Finally, the feather δ13C values ofSA depicted 2 moulting strategies, with birds fromMarion, Crozet and Amsterdam islands moultingmostly north of the Subtropical Front, whereas birdsfrom Gough Island moult mainly in subantarcticwaters (Jaeger et al. 2010a, Cherel et al. 2013, thisstudy).

In summary, by combining dietary techniques thatintegrate the food and feeding habits over days,weeks and months, the present work highlights acomplex pattern of spatial and trophic segregationsamong the medium-size albatross species that breedbiennially on Marion Island (only island with an easyaccess to the 3 sympatric species). The next stepshould be to track individual birds using satellite tagsand geolocation loggers during the breeding andinterbreeding periods, respectively, to complete thepicture. Both biologging and the stable isotopemethod can also help to investigate the foragingecology of juveniles and immature birds, whichremains poorly known.

Acknowledgements. The authors thank P. G. Ryan, M. Davies-Coleman, E. L. Allan, A. L. Jackson, M. G. W. Jones, N. T. W.Klages, J. Kolasinski, P. Pistorius, N. Richoux, S. Sunassee,C. Tambling, B. Dyer and L. Clokie for their valuable adviceat different stages of this study. The manuscript benefitedfrom the constructive comments of 4 anonymous reviewersand the Editor. We acknowledge the Fatty Acid Facility atRhodes University funded through the National ResearchFoundation and Rhodes University. Isotope samples wereanalyzed by I. Newton at the Stable Light Isotope Labora-tory of the University of Cape Town under the supervision ofJ. Lanham. Funding and logistical support was provided bythe Department of Environmental Affairs and Tourismthrough the South African National Antarctic Program andadministered by the National Research Foundation. M.C.acknowledges funding from the University of Cape Town’sMarine Research Institute. This work is based on researchsupported by the South African Research Chairs Initiative ofthe Department of Science and Technology and the NationalResearch Foundation. This work was undertaken under anethics permit granted by Rhodes University.

LITERATURE CITED

Allan EL, Ambrose ST, Richoux NB, Froneman PW (2010)Determining spatial changes in the diet of nearshore sus-pension-feeders along the South African coastline: stableisotope and fatty acid signatures. Estuar Coast Shelf Sci87: 463−471

270

http://dx.doi.org/10.1016/j.ecss.2010.02.004

Connan et al.: Spatial and trophic segregation among albatrosses

Anderson ORJ, Phillips RA, McDonald RA, Shore RF, McGillRAR, Bearhop S (2009) Influence of trophic position andforaging range on mercury levels within a seabird com-munity. Mar Ecol Prog Ser 375: 277−288

Berruti A, Harcus T (1978) Cephalopod prey of the sootyalbatrosses Phoebetria fusca and P. palpebrata at MarionIsland. S Afr J Antarct Res 8: 99−103

Bolnick DI, Yang LH, Fordyce JA, Davis JM, Svanbäck R(2002) Measuring individual-level resource specializa-tion. Ecology 83: 2936−2941

Catry P, Phillips RA, Phalan B, Silk JRD, Croxall JP (2004)Foraging strategies of grey-headed albatrosses Thalas-sarche chrysostoma: integration of movements, activityand feeding events. Mar Ecol Prog Ser 280: 261−273

Cherel Y (2008) Isotopic niches of emperor and Adélie pen-guins in Adélie Land, Antarctica. Mar Biol 154: 813−821

Cherel Y, Hobson KA (2005) Stable isotopes, beaks andpredators: a new tool to study ecology of cephalopods,including giant and colossal squids. Proc R Soc Lond BBiol Sci 272: 1601−1607

Cherel Y, Hobson KA (2007) Geographical variation in car-bon stable isotope signatures of marine predators: a toolto investigate their foraging areas in the SouthernOcean. Mar Ecol Prog Ser 329: 281−287

Cherel Y, Klages N (1998) A review of the food of alba-trosses. In: Robertson G, Gales R (eds) The albatross: biology and conservation. Surrey Beatty & Sons, Chip-ping Norton, p 113−136

Cherel Y, Hobson KA, Hassani S (2005) Isotopic discrimina-tion between food and blood and feathers of captive pen-guins: implications for dietary studies in the wild. PhysiolBiochem Zool 78: 106−115

Cherel Y, Ducatez S, Fontaine C, Richard P, Guinet C (2008)Stable isotopes reveal the trophic position and mesopela-gic fish diet of female southern elephant seals breedingon the Kerguelen Islands. Mar Ecol Prog Ser 370: 239−247

Cherel Y, Fontaine C, Richard P, Labat JP (2010) Isotopicniches and trophic levels of myctophid fishes and theirpredators in the Southern Ocean. Limnol Oceanogr 55: 324−332

Cherel Y, Jaeger A, Alderman R, Jaquemet S and others(2013) A comprehensive isotopic investigation of habitatpreferences in nonbreeeding albatrosses from the South-ern Ocean. Ecography 36: 277−286

Connan M, Mayzaud P, Boutoute M, Weimerskirch H,Cherel Y (2005) Lipid composition of stomach oil in a pro-cellariiform seabird Puffinus tenuirostris: implications forfood web studies. Mar Ecol Prog Ser 290: 277−290

Connan M, Cherel Y, Mayzaud P (2007) Lipids from stomachoil of procellariiform seabirds document the importanceof myctophid fish in the Southern Ocean. LimnolOceanogr 52: 2445−2455

Connan M, Mayzaud P, Duhamel G, Bonnevie BT, Cherel Y(2010a) Quantitative fatty acid signature analysis docu-ments the diet of five myctophid fish from the SouthernOcean. Mar Biol 157: 2303−2316

Connan M, Mayzaud P, Hobson KA, Weimerskirch H,Cherel Y (2010b) Food and feeding ecology of the Tas-manian short-tailed shearwater (Puffinus tenuirostris,Temminck): insights from three complementary meth-ods. J Oceanogr 3: 19−32

Connan M, Kelly CMR, McQuaid CD, Bonnevie BT, BarkerNP (2011) Morphological versus molecular identificationof Sooty and Light-mantled albatross chicks. Polar Biol

34: 791−798Cooper J, Klages NTW (1995) The diets and dietary segre-

gation of sooty albatrosses (Phoebetria spp.) at sub-antarctic Marion Island. Antarct Sci 7: 15−23

Croxall JP, Reid K, Prince PA (1999) Diet, provisioning andproductivity responses of marine predators to differencesin availability of Antarctic krill. Mar Ecol Prog Ser 177: 115−131

Froese R, Pauly D (eds) (2012) FishBase. www.fishbase.org,accessed 10 Oct 2012

Hammer Ø, Harper DAT, Ryan PD (2001) PAST: paleonto-logical statistics software package for education and dataanalysis. Palaeontol Electronica 4(1):art4

Harrison P (1983) Seabirds: an identification guide. CroomHelm, Beckenham

Hobson KA, Clark RG (1992) Assessing avian diets usingstable isotopes I: turnover of 13C in tissues. Condor 94: 181−188

Hobson KA, Gibbs HL, Gloutney ML (1997) Preservation ofblood and tissue samples for stable-carbon and stable-nitrogen isotope analysis. Can J Zool 75: 1720−1723

Hockey PAR, Dean WRJ, Ryan PG (2005) Roberts birds ofSouthern Africa, 7th edn. The Trustees of the John Voel-cker Bird Book Fund, Cape Town

Iverson SJ, Field C, Bowen WD, Blanchard W (2004) Quan-titative fatty acid signature analysis: a new method ofestimating predator diets. Ecol Monogr 74: 211−235

Jackson AL, Inger R, Parnell AC, Bearhop S (2011) Com -paring isotopic niche widths among and within com -munities: SIBER — Stable Isotope Bayesian Ellipses in R.J Anim Ecol 80: 595−602

Jaeger A, Connan M, Richard P, Cherel Y (2010a) Use of stable isotopes to quantify seasonal changes of trophicniche and levels of population and individual specialisa-tion in seabirds. Mar Ecol Prog Ser 401: 269−277

Jaeger A, Lecomte VJ, Weimerskirch H, Richard P, Cherel Y(2010b) Seabird satellite tracking validates the use of latitudinal isoscapes to depict predators’ foraging areasin the Southern Ocean. Rapid Commun Mass Spectrom24: 3456−3460

Karnovsky NJ, Hobson KA, Iverson SJ, Hunt GL (2008) Sea-sonal changes in diets of seabirds in the North WaterPolynya: a multiple-indicator approach. Mar Ecol ProgSer 357: 291−299

Karnovsky NJ, Hobson KA, Iverson SJ (2012) From lavage tolipids: estimating diets of seabirds. Mar Ecol Prog Ser451: 263−284

Kock KH, Everson I (1997) Biology and ecology of mackerelicefish, Champsocephalus gunnari: an Antarctic fishlacking hemoglogin. Comp Biochem Physiol A 118: 1067−1077

Lutjeharms JRE, Ansorge IJ (2008) Oceanographic settingsof the Prince Edward Islands. In: Chown SN, FronemanPW (eds) The Prince Edwards Archipelago: land-seainteractions in a changing ecosystem. Sun Media, Stel-lenbosch, p 17−38

Mizutani H, Fukuda M, Kabaya Y, Wada E (1990) Carbonisotope ratio of feathers reveals feeding behavior of cor-morants. Auk 107: 400−403

Nel DC, Lutjeharms JRE, Pakhomov EA, Ansorge IJ, RyanPG, Klages NTW (2001) Exploitation of mesoscaleoceanographic features by grey-headed albatross Tha-lassarche chrysostoma in the southern Indian Ocean.Mar Ecol Prog Ser 217: 15−26

Park YH, Gambéroni L, Charriaud E (1993) Frontal struc-

271

http://dx.doi.org/10.1029/93JC00938http://dx.doi.org/10.3354/meps217015http://dx.doi.org/10.2307/4087626http://dx.doi.org/10.1016/S0300-9629(97)86795-3http://dx.doi.org/10.3354/meps09713http://dx.doi.org/10.3354/meps07295http://dx.doi.org/10.1002/rcm.4792http://dx.doi.org/10.3354/meps08380http://dx.doi.org/10.1111/j.1365-2656.2011.01806.xhttp://dx.doi.org/10.1890/02-4105http://dx.doi.org/10.1139/z97-799http://dx.doi.org/10.2307/1368807http://dx.doi.org/10.3354/meps177115http://dx.doi.org/10.1017/S0954102095000046http://dx.doi.org/10.1007/s00300-010-0933-6http://dx.doi.org/10.1007/s00227-010-1497-2http://dx.doi.org/10.4319/lo.2007.52.6.2445http://dx.doi.org/10.3354/meps290277http://dx.doi.org/10.1111/j.1600-0587.2012.07466.xhttp://dx.doi.org/10.4319/lo.2010.55.1.0324http://dx.doi.org/10.3354/meps07673http://dx.doi.org/10.1086/425202http://dx.doi.org/10.3354/meps329281http://dx.doi.org/10.1098/rspb.2005.3115http://dx.doi.org/10.1007/s00227-008-0974-3http://dx.doi.org/10.3354/meps280261http://dx.doi.org/10.1890/0012-9658(2002)083[2936%3AMILRS]2.0.CO%3B2http://dx.doi.org/10.3354/meps07784

Mar Ecol Prog Ser 497: 259–272, 2014272

ture, water masses, and circulation in the Crozet Basin.J Geophys Res 98: 12361−12385

Phillips RA, Silk JRD, Phalan B, Catry P, Croxall JP (2004)Seasonal sexual segregation in two Thalassarche alba-tross species: Competitive exclusion, reproductive rolespecialization or foraging niche divergence? Proc SocLond B Biol Sci 271: 1283−1291

Phillips RA, Bearhop S, McGill RAR, Dawson DA (2009) Sta-ble isotopes reveal individual variation in migrationstrategies and habitat preferences in a suite of seabirdsduring the nonbreeding period. Oecologia 160: 795−806

Phillips RA, McGill RAR, Dawson DA, Bearhop S (2011) Sex-ual segregation in distribution, diet and trophic level ofseabirds: insights from stable isotope analysis. Mar Biol158: 2199−2208

Pinaud D, Cherel Y, Weimerskirch H (2005) Effect of envi-ronmental variability on habitat selection, diet, provi-sioning behaviour and chick growth in yellow-nosedalbatrosses. Mar Ecol Prog Ser 298: 295−304

R Development Core Team (2009) R: a language and envi-ronment for statistical computing. R Foundation for Statis-tical Computing, Vienna, available at www.r-project.org

Richoux NB, Jaquemet S, Bonnevie BT, Cherel Y, McQuaidCD (2010) Trophic ecology of Grey-headed albatrossesfrom Marion Island, Southern Ocean: insights from stom-ach contents and diet tracers. Mar Biol 157: 1755−1766

Ridoux V (1994) The diets and dietary segregation of sea-birds at the sub-Antarctic Crozet Islands. Mar Ornithol22: 1−192

Rohwer S, Viggiano A, Marzluff JM (2011) Reciprocal trade-offs between molt and breeding in albatrosses. Condor113: 61−70

Ryan PG, Bester MN (2008) Pelagic predators. In: ChownSL, Froneman PW (eds) The Prince Edward Islands: land-sea interactions in a changing ecosystem. SunMedia, Stellenbosch, p 121−164

Ryan PG, Jones MGW, Dyer BM, Upfold L, Crawford RJM(2009) Recent population estimates and trends in numbers of albatrosses and giant petrels breeding at thesub-Antarctic Prince Edward islands. Afr J Mar Sci 31: 409−417

Suazo CG (2008) Black-browed albatross foraging on jelly-fish prey in the southeast Pacific coast, southern Chile.Polar Biol 31: 755−757

Syväranta J, Lensu A, Marjomäki TJ, Oksanen S, Jones RI(2013) An empirical evaluation of the utility of convexhull and standard ellipse areas for assessing populationniche widths from stable isotope data. PLoS ONE 8: e56094

Thomas G (1982) The food and feeding ecology of the Light-mantled albatross at South Georgia. Emu 82: 92−100

Warham J (1996) The behaviour, population biology andphysiology of the petrels. Academic Press, London andSan Diego, CA

Warham J, Watts R, Dainty RJ (1976) The composition,energy content and function of the stomach oils ofpetrels (order, Procellariiformes). J Exp Mar Biol Ecol23: 1−13

Xavier JC, Cherel Y (2009) Cephalopod beak guide for theSouthern Ocean. British Antarctic Survey, Cambridge

Xavier JC, Phillips RA, Cherel Y (2011) Cephalopods in ma rine predator diet assessments: why identifying up -per and lower beaks is important. ICES J Mar Sci 68: 1857−1864

Editorial responsibility: Jacob González-Solís, Barcelona, Spain

Submitted: October 12, 2012; Accepted: October 15, 2013Proofs received from author(s): January 10, 2014

http://dx.doi.org/10.1093/icesjms/fsr103http://dx.doi.org/10.1016/0022-0981(76)90081-2http://dx.doi.org/10.1071/MU9820092http://dx.doi.org/10.1371/journal.pone.0056094http://dx.doi.org/10.1007/s00300-008-0412-5http://dx.doi.org/10.2989/AJMS.2009.31.3.13.1001http://dx.doi.org/10.1525/cond.2011.100092http://dx.doi.org/10.1007/s00227-010-1448-yhttp://dx.doi.org/10.3354/meps298295http://dx.doi.org/10.1007/s00227-011-1725-4http://dx.doi.org/10.1007/s00442-009-1342-9http://dx.doi.org/10.1098/rspb.2004.2718

cite43: cite28: cite5: cite14: cite42: cite3: cite13: cite1: cite26: cite41: cite39: cite12: cite40: cite25: cite38: cite11: cite10: cite8: cite23: cite36: cite49: cite22: cite4: cite48: cite21: cite34: cite19: cite2: cite47: cite33: cite18: cite46: cite32: cite17: cite31: cite16: cite44: cite29: cite7: cite30: cite15: