Migration and breeding biology of Arctic terns in Greenland

MIGRATION AND BREEDING BIOLOGY OF 

ARCTIC TERNS IN GREENLAND 

PhD Thesis 2010 

Carsten Egevang 

AU 

Greenland Institute of Natural Resources 

NATIONAL ENVIRONMENTAL RESEARCH INSTITUTE 

center for macroecology, evolution 

a nd climate 

AARHUS UNIVERSITY university of copenhagen

AU 



PhD Thesis 2010 


Greenland Institute of Natural Resources 

NATIONAL ENVIRONMENTAL RESEARCH INSTITUTE 

AARHUS UNIVERSITY 

center for macroecology, evolution 

and climate 

university of copenhagen

Title: Migration and breeding biology of Arctic terns in Greenland 

Subtitle: PhD Thesis 

Author: Carsten Egevang 

Affi liation: Greenland Institute of Natural Resources 

Dep. of Arctic Environment, National Environmental Research Institute, Aarhus University 

Department of Biology, Center for Macroecology, Evolution and Climate, University of Copenhagen 

Publisher: Greenland Institute of Natural Resources and National Environmental Research Institute (NERI)© 

Aarhus University – Denmark 

URL: http://www.neri.dk 

Date of publication: March 2010 

Financial support: Commission for Scientifi c Research in Greenland (KVUG) 

Please cite as: Egevang, C. 2010: Migration and breeding biology of Arctic terns in Greenland. PhD thesis. Greenland Institute 

of Natural Resources, Dep. of Arctic Environment, NERI, Aarhus University & Department of Biology, 

Center for Macroecology, Evolution and Climate, University of Copenhagen. Greenland Institute of Natural 

Resources & National Environmental Research Institute, Aarhus University, Denmark. 104 pp. 

Reproduction permitted provided the source is explicitly acknowledged. 

Abstract: This thesis presents novel fi ndings for the Arctic tern in Greenland. Included is a study on Arctic tern migration 

– the longest annual migration ever recorded in any animal. The study documented how Greenland and Iceland 

breeding terns conduct the roundtrip migration to the Weddell Sea in Antarctica and back. Although the sheer 

distance (71,000 km on average) travelled by the birds is interesting, the study furthermore showed how the 

birds depend on high-productive at-sea areas and global wind systems during their massive migration. 

Furthermore included is the fi rst quantifi ed estimate of the capacity to produce a replacement clutch in Arctic 

terns. We found that approximately half of the affected birds would produce a replacement clutch when the 

eggs were removed late in the incubation period. 

At a level of more national interest, the study produced the fi rst estimates of the key prey species of the Arctic 

tern in Greenland. Although zooplankton and various fi sh species were present in the chick diet of terns 

breeding in Disko Bay, Capelin was the single most important prey species found in all age groups of chicks. 

The thesis also includes a study on the fl uctuating breeding found in Arctic terns. Breeding birds showed a 

considerable variation in colony size between years in the small and mid sized colonies of Disko Bay. These 

variations were likely to be linked to local phenomena, such as disturbance from predators, rather than to 

large-scale occurring phenomena. 

Included is also a study that documents the breeding association between Arctic terns and other breeding 

waterbirds. At Kitsissunnguit a close behavioural response to tern alarms could be identifi ed. These fi ndings 

imply that the altered distributions of waterbirds observed at Kitsissunnguit were governed by the distribution 

of breeding Arctic terns. 

Included in the thesis are furthermore results with an appeal to the Greenland management agencies. Along 

with estimates of the Arctic tern population size at the two most important Arctic tern colonies in West Greenland 

and East Greenland, the study produced recommendations on how a potential future sustainable egg 

harvest could be carried out, and on how to monitor Arctic tern colonies in Greenland. 

Keywords: At-sea hot-spots, Arctic, breeding association, breeding ecology, breeding fl uctuation, colony attendance, 

diet, egg harvest, global wind systems, long-distance migration, nature management, Sterna paradisaea, 

replacement clutch. 

Supervisors: Professor Carsten Rahbek, Department of Biology, Center for Macroecology, 

Evolution and Climate, University of Copenhagen. 

Senior advisor Dr. Morten Frederiksen, Dep. of Arctic Environment, 

National Environmental Research Institute, Aarhus University. 

Dr. Norman Ratcliffe, British Antarctic Survey, Natural Environment Research Council. 

Layout: Tinna Christensen 

Frontpage: Carsten Egevang 

ISBN: 87-91214-46-7 

ISBN (DMU): 978-87-7073-166-9 

Printed by: Rosendahls – Schultz Grafi sk a/s 

Number of pages: 104 

Circulation: 100 

Data sheet 

Internet version: The report is available in electronic format (pdf) at NERI’s website 

http://www.dmu.dk/Pub/PHD_CEP.pdf

Content 

List of publications/manuscripts included in the thesis 5 

Preface 6 

Acknowledgements 8 

Summary 9 

Dansk resumé 11 

Eqikkaaneq 13 

1 Synopsis 15 

1.1 Introduction 15 

1.2 Objectives 16 

1.3 Distribution 16 

1.4 Population status in Greenland 16 

1.5 Population status at Kitsissunnguit 17 

1.6 Fluctuating breeding 17 

1.7 Plausible cause of decline in population size 18 

1.8 Renesting in Arctic terns 19 

1.9 Variation in breeding phenology 20 

1.10 Clutch size 21 

1.11 Hatching and fl edging success 21 

1.12 Growth rate and chick survival 22 

1.13 Feeding 23 

1.14 Breeding association 24 

1.15 Migration 25 

2 Conclusion 29 

3 Future research 30 

4 References 31 

Manus I 35 

Manus II 41 

Manus III 57 

Manus IV 85 

Manus V 97

LIST OF PUBLICATIONS/MANUSCRIPTS 

INCLUDED IN THE THESIS 

Egevang, C., I.J. Stenhouse, R.A. Phillips, A. Petersen, J.W. Fox & J.R.D. 

Silk (2010). Tracking of Arctic terns Sterna paradisaea reveals longest 

animal migration. Proceedings of the National Academy of Sciences of 

the United States. Vol. 107: 2078-2081 (doi:10.1073/pnas.0909493107). 

Egevang, C., N. Ratcliffe & M. Frederiksen. Sustainable regulation of 

Arctic tern egg harvesting in Greenland. Manuscript. 

Egevang, C., M. W. Kristensen, N. Ratcliffe & M. Frederiksen. Food-related 

consequences of relaying in the Arctic Tern: Contrasting prey availability 

within a prolonged breeding season. Manuscript submitted IBIS. 

Egevang, C. & M. Frederiksen. Fluctuating breeding of Arctic Terns (Sterna 

paradisaea) in Arctic and High-arctic colonies in Greenland. Manuscript 

submitted WATERBIRDS. 

Jørgensen, P. S., Kristensen, M. W. & Egevang, C. 2007. Red Phalarope 

Phalaropus fulicarius and Red-necked Phalarope Phalaropus lobatus 

behavioural response to Arctic Tern Sterna paradisaea colonial alarms. 

Dansk Ornitologisk Forenings Tidsskrift 101 (3): 73-78, 2007. 

5

6 

Preface 

This thesis is the result of my PhD conducted at the University of Copenhagen, 

Center for Macroecology, Evolution and Climate (UC), Greenland 

Institute of Natural Resources (GINR) and National Environmental Research 

Institute, Department of Arctic Environment (NERI). The PhD was 

fi nancially supported by Greenland Institute of Natural Resources (GNIR) 

and the Commission for Scientifi c Research in Greenland (KVUG). The 

principle supervisor was Carsten Rahbek (UC) whereas Morten Frederiksen 

(NERI) and Norman Ratcliffe (British Antarctic Survey) functioned as 

external supervisors. 

The pre-2002 level of knowledge on the Arctic tern in Greenland was rather 

low and even rather basic biological information was unavailable for 

the species. On the other side there’s an urgent need for more information 

in the Greenland management of the living resources and protected areas. 

This lack of knowledge has been decisive in my cho ice of study species, 

study topic and study location at the planning of my PhD. It has been both 

a strong motivation factor for me throughout my PhD, but also frustrating 

at times where the lack of information made it diffi cult to approach topics 

from a theoretical point of view. 

Although my PhD offi cially started mid 2004, my work on Arctic terns 

in Greenland is really a continuation of fi eldwork conducted in 2002 and 

2003. Concurrently with my PhD I have worked as seabird researcher at 

GNIR, where I have been responsible for monitoring seabirds in Greenland 

and ad-hoc advising the Greenland Government in a sustainable 

use of the living resources. For this reason the delivery date of my PhD 

has been prolonged to March 2010 – more than one year later than fi rst 

scheduled. Along the course of my PhD the topic included in the PhD 

got extended. The technology in tracking devices for small-sized seabirds 

advanced and from dealing only with breeding ecology and censusing 

methods, my PhD got extended to include migration ecology as well. 

During my PhD I have been physically placed in Greenland at GINR until 

May 2008, where my family and I moved to Denmark and NERI was kind 

enough to offer me a work seat. 

The fi eldwork in connection with my PhD was conducted at two very 

different fi eld sites in Greenland: Kitsissunnguit (Grønne Ejland) during 

2002 to 2006 in the southern part of Disko Bay in the Arctic zone in West 

Greenland and at Sand Island (Sandøen) during 2007 and 2008 in high 

Arctic Northeast Greenland (see fi gure 1-3). 

During my PhD I have produced papers, reports or book contributions as 

a spin-off from the work on Arctic terns conducted during my PhD. These 

are not included in my PhD thesis but are listed here: 

Egevang, C. & D. Boertmann 2008. “Ross’s Gulls (Rhodostethia rosea) breeding 

in Greenland: A Review, with Special Emphasis on Records from 

1979 to 2007.” Arctic 61(3): 322-328.

Egevang, C. Kampp, K. & Boertmann D. 2006. ”Declines in breeding 

waterbirds following a redistribution of Arctic Terns Sterna paradisaea in 

West Greenland” Waterbirds around the World. Eds. G. C. Boere, C. A. 

Galbraith & D. A. Stroud. The stationery Offi ce, Edinburgh, UK. P. 154 

(http://www.jncc.gov.uk/PDF/pub07_waterbirds_part3.3.1.6.pdf). 

Egevang, C. , K. Kampp & D. Boertmann 2004 “The Breeding Association 

of Red Phalaropes (Phalaropus fulicarius) with Arctic Terns (Sterna 

paradisaea): Response to a Redistribution of Terns in a Major Greenland 

Colony” Waterbirds 27 (4): 406-410. 

Egevang, C., I. J. Stenhouse, L. M. Rasmussen, M. W. Kristensen and F. 

Ugarte 2009. ”Breeding and foraging ecology of seabirds on Sandøen 

2008”. in Jensen, L.M. & Rasch, M. (eds.) 2009: Zackenberg Ecological 

Research Operations, 14 th Annual Report, 2008. National Environmental 

Research Institute, Aarhus University, Denmark. 116 pp. 

Egevang, C., I. J. Stenhouse, Lars Maltha Rasmussen, Mikkel Willemoes 

& Fernando Ugarte 2008. “Field report from Sand Island, Northeast 

Greenland, 2008. http://www.natur.gl/UserFiles/File/feltrapporter/ 

Fieldreport_%20SandIsland_2008_fi nal.pdf. 

Egevang, C. 2008. ”Forstyrrelser i grønlandske havfuglekolonier, med speciel 

fokus på ynglende havterner på Kitsissunnguit (Grønne Ejland), 

Disko Bugt”. Teknisk rapport nr. 71, Grønlands Naturinstitut, 21 sider. 

http://www.natur.gl/UserFiles/File/Publlikationer/2008-02_GN_teknisk_rapport_71_fi 

nal_forstyrrelser_fuglekolonier.pdf 

Egevang, C. & Stenhouse, I. J. 2008. Mapping long-distance migration in 

two Arctic seabird species p. 74-76. In: Klitgaard, A.B. and Rasch, M. 

(eds.) 2008. Zackenberg Ecological Research Operations, 13 th Annual Report, 

2007. Danish Polar Center, Danish Agency for Science, Technology 

and Innovation, Ministry of Science, Technology and Innovation, 2008. 

http://www.zackenberg.dk/graphics/Design/Zackenberg/Publications/English/Zero%202007.pdf 

Egevang, C. & I. J. Stenhouse 2007. ”Field report from Sand Island, Northeast 

Greenland, 2007”. http://www.natur.gl/UserFiles/File/feltrapporter/Fieldwork_Sand_Island_2007_1.pdf 

Egevang, C., D. Boertmann & O. S. Kristensen 2005 ”Monitering af 

havternebestanden på Kitsissunnguit (Grønne Ejland) og den sydlige 

del af Disko Bugt, 2002-2004”. Teknisk rapport nr. 62, Grønlands Naturinstitut, 

41 p. http://www.natur.gl/fi ler/Havternemonitering.pdf 

The study on the migration of the Arctic tern included novel fi ndings and 

had appeal to a broad public audience. Therefore I build a web site (www. 

arctictern.info) especially dedicated to the communication of the results 

originating from our study. 

My thesis deals with an array of different topics within the fi eld of ecology. 

From small-scale issues, like egg harvesting by local Inuit at the breeding 

grounds, to large-scale occurring phenomena, such as global wind systems 

and global marine biological productivity. Although, it may seem 

like a confl icting cocktail of topics, it is all focused on the same species – 

the gracious Arctic tern. 

7

8 

Acknowledgements 

I owe tremendous thanks to my external supervisor Morten Frederiksen 

(NERI) who has provided excellent guidance and supervision – not once 

did I experience that Morten couldn’t fi nd the time to support me in technical 

matters and give me input instantly. Also thanks to Norman Ratcliffe 

(British Antarctic Survey), my other external supervisor for help modelling 

the data obtained over the fi eld seasons. 

The Greenland Institute of Natural Resource provided the framework to 

work with my thesis on the Arctic tern – both in Greenland and Denmark. 

For that I’m most grateful to the director Klaus Nygaard and head of department 

Fernando Ugarte. Also thanks to Jesper Madsen, head of department 

and Anders Mosbech, head of section for their willingness to 

provide me with a disk at the crowed Arctic Environment department at 

NERI. 

Special thanks goes to Scottish friend and colleague Iain Stenhouse (National 

Audubon Society) who endured cold and stormy days with me in 

a tent at Sand Island, and gave invaluable language comments to this 

synopsis. 

My thanks goes to the people who assisted me in the fi eld over the years: 

David Boertmann and Morten Bjerrum (National Environmental Research 

Institute, Denmark), Anders Tøttrup, Mikkel Willemoes and Peter S. Jørgensen 

(University of Copenhagen), Lars Maltha Rasmussen, Fernando 

Ugarte, Ole S. Kristensen, Lars Witting, Kate Skjærbæk Rasmussen and 

Rasmus Lauridsen (GNIR), Bjarne Petersen, Qasigiannguit and Norman 

Ratcliffe and Stuart Benn (Royal Society for Protection of Birds). Also 

thanks to Finn Steffens (Qeqertarsuaq) and Elektriker-it (Aasiaat) for 

logistic support in Disko Bay, and to the Danish Polar Centre, Zackenberg 

Ecological Research Operations, POLOG, and Sledge Patrol Sirius 

and Nette Levermann (Greenland Government) for logistic assistance in 

Northeast Greenland (Sand Island). 

I will also like this opportunity to thank the local inhabitants of Disko Bay. 

Through arranged meetings in Aasiaat and multiple opportunistic visits 

at our fi eld camp at Kitsissunnguit, they have thought me another view at 

harvesting natural resources and the signifi cance of this. 

Last, but certainly not least, I would like to thank my family, my wife Inge 

Thaulow and my two boys Emil and Peter, for patience, understanding 

and absents during fi eldwork – THANKS!

Summary 

This PhD thesis presents the results of a study performed on the Arctic 

tern (Sterna paradisaea) in the period 2002-2008. Data in the study were obtained 

from fi eldwork conducted at two study sites in Greenland: Kitsissunnguit 

(Grønne Ejland), Disko Bay in Arctic West Greenland and Sand 

Island (Sandøen) in high-Arctic Northeast Greenland. 

The level of knowledge of the Arctic tern in Greenland before 2002 was to 

a large extent poor, with aspects of its biology being completely unknown 

in the Greenland population. This thesis presents novel fi ndings for the 

Arctic tern, both on an international scale and on a national scale. 

The study on Arctic tern migration (Manus I) – the longest annual migration 

ever recorded in any animal – is a study with an international 

appeal. The study documented how Greenland and Iceland breeding 

terns conduct the roundtrip migration to the Weddell Sea in Antarctica 

and back. Although the sheer distance (71,000 km on average) travelled 

by the birds is interesting, the study furthermore showed how the terns 

depend on high-productive at-sea areas during their massive migration. 

On the southbound migration, the birds would stop for almost a month 

(25 days on average) in the central part of the North Atlantic Ocean before 

continuing south. Close to Equator (~10º N) a divide in the migration path 

way occurred: seven birds migrated along the coast of Africa, while four 

birds crossed the Atlantic Ocean to follow the coast of South America. The 

northbound migration from the winter quarters to the breeding grounds 

was performed particular fast (520 km per day on average) following a 

route of favourable wind systems. 

Included in this thesis is also the fi rst quantifi ed estimate of the capacity to 

produce a replacement clutch in Arctic terns (Manus II and III). Although 

it is often mentioned in the literature that the species willingly relays, this 

study is the fi rst where the reproductive response of experimentally manipulated 

breeding pairs is monitored. We found that approximately half 

(53.3 %) of the affected birds would produce a replacement clutch when 

the eggs were removed late in the incubation period. Surprisingly, growth 

and survival rates in chicks from these clutches did not differ from chicks 

reared four weeks later in the breeding season, although a shift in foraging 

pattern and prey size was apparent (Manus III). 

At a level of more national interest, the study produced the fi rst estimates 

of the key prey species of the Arctic tern in Greenland. Although zooplankton 

and various fi sh species were present in the chick diet of terns 

breeding in Disko Bay, Capelin (Mallotus villosus) was the single most important 

prey species found in all age groups of chicks (Manus III). 

The thesis also includes a study on the fl uctuating breeding found in Arctic 

terns (Manus IV). Although the Arctic tern may show an overall fi delity 

to the breeding region, our study showed a considerable variation in colony 

size between years in the small and mid sized colonies of Disko Bay. 

These variations were likely to be linked to local phenomena, such as disturbance 

from predators, rather than to large-scale occurring phenomena. 

From the long periods spent in Arctic tern colonies it was furthermore 

possible to document the breeding association between Arctic terns and 

9

Figure 1. An Arctic tern nest at 

Kitsissunnguit, Disko Bay, Arctic 

West Greenland where fi eldwork 

was conducted in 2002-2006. 

Note the lush vegetation surrounding 

the nest. 

Photo: Carsten Egevang. 

10 

other breeding waterbirds. In a study conducted on breeding phalaropes 

(Phalaropus fulucarius and P. lobatus) at Kitsissunnguit (Manus V) a close 

behavioural response to tern alarms could be identifi ed. These fi ndings 

imply that the altered distributions of waterbirds observed at Kitsissunnguit 

were governed by the distribution of breeding Arctic terns as 

suggested by Egevang et al. (2004). 

Included in the thesis are furthermore results with an appeal to the Greenland 

management agencies. Along with estimates of the Arctic tern population 

size at the two most important Arctic tern colonies in West Greenland 

and East Greenland, the study produced recommendations on how 

a potential future sustainable egg harvest could be carried out (Manus II), 

and on how to monitor Arctic tern colonies in Greenland (Manus IV).

Dansk resumé 

Denne afhandling præsenterer resultater fra studier udført på havterne 

(Sterna paradisaea) i perioden 2002-2008. Data i forbindelse med afhandlingen 

er indsamlet gennem feltarbejde på to lokaliteter i Grønland: Kitsissunnguit 

(Grønne Ejland), Disko Bugt i arktisk Vestgrønland samt på 

Sandøen i høj-arktisk Nordøstgrønland. 

Vidensgrundlaget for havternen i Grønland før 2002 var gennemgående 

ringe, med aspekter af artens ynglebiologi ubeskrevet for den grønlandske 

bestands vedkommende. Denne afhandling præsenterer ny viden for 

havternen både på en international skala, såvel som på en national skala. 

Først og fremmest er studiet af havternens træk (Manus I) – det længste 

træk registreret hos noget dyr – et studie med international appel. Studiet 

dokumenterer hvorledes terner, der yngler i Nordøstgrønland og Island, 

gennemfører trækket til Weddell Havet ved Antarktis, og tilbage igen. 

Alene den tilbagelagte distance (gennemsnitlig 71.000 km) er interessant, 

men studiet kunne også dokumentere, hvordan ternerne under deres lange 

træk er afhængige af havområder med særlig høj biologisk produktion. 

På det sydgående træk stopper fuglene trækket og opholder sig i næsten 

en måned (25 dage i gennemsnit) i den centrale del af Nordatlanten, før 

de fortsætter mod syd. Tæt ved Ækvator (~10º N) opstår en opdeling af 

trækvejen: Af 11 mærkede individer fortsætter syv langs Afrikas kyst, 

mens fi re individer krydser Atlanten, for at fortsætte langs østkysten af 

Sydamerika. Det nordgående træk fra vinterkvarteret til ynglepladserne 

blev gennemført med særlig høj hastighed (gennemsnitligt 520 km per 

dag) og fulgte en rute med favorable vindretninger. 

Denne afhandling indeholder også det første kvantifi cerede estimat af 

havternens evne til at producere et såkaldt omlægskuld – et erstatningskuld 

hvis første kuld går tabt (Manus II og III). På trods af, at det ofte nævnes 

i litteraturen at havternen villigt omlægger, er dette studie det første 

der overvåger det reproduktive respons på eksperimentalt manipulerede 

ynglepar. Vi fandt, at omtrent halvdelen (53,3 %) af fuglene producerede 

et omlægskuld, når det første kuld blev fjernet i den sidste del af rugeperioden. 

Overraskende viste det sig, at vækstrater og overlevelse hos unger 

fra omlægskuldene ikke var forskellige fra unger opfostret fi re uger tidligere, 

på trods af, at et skifte i både fourageringsmønster og størrelse af 

byttedyr kunne identifi ceres (Manus III). 

Af mere national interesse producerede studierne i forbindelse med afhandlingen 

de første undersøgelser af havternens fødevalg i Grønland. 

På trods af at fl ere forskellige fi ske- og zooplanktonarter var til stede i den 

føde der blev bragt til ungerne i Disko Bugt, Vestgrønland, viste det sig, 

at lodde (Mallotus villosus) var nøglefødearten for alle aldersgrupper af 

unger (Manus III). 

Afhandlingen omhandler også et studie af den fl uktuerende yngleforekomst 

som forekommer hos havternen (Manus IV). Selv om havternen 

udviser en overordnet trofasthed mod yngleregionen, viste vores studier, 

at der forekommer en betydelig variation i antallet af ynglefugle i kolonien 

mellem sæsonerne i de små og mellemstore kolonier i Disko Bugtområdet. 

Disse variationer er sandsynligvis koblet til lokalt forekommen- 

11

Figure 2. An Arctic tern nest at 

the barren Sand Island, High-arctic 

Northeast Greenland, where 

fi eldwork was conducted in 2007 

and 2008. 

12 

de fænomener, eksempelvis forstyrrelse fra prædatorer, mere end de er 

forklaret af forhold der forekommer på større skala. 

Gennem de lange perioder, tilbragt i havternekolonierne i forbindelse 

med feltarbejdet, var det muligt at dokumentere det positive yngleforhold 

mellem havterner og andre ynglende vandfugle. I et studie udført 

på ynglende thors- og odinshane (Phalaropus fulucarius og P. lobatus) på 

Kitsissunnguit (Manus V), kunne et tæt adfærdsmæssigt respons på havternens 

advarselskald identifi ceres. Disse resultater sandsynliggør, at den 

observerede ændrede fordeling af svømmesnepper på Kitsissunnguit følger 

havternens udbredelse på øerne, som foreslået af Egevang et al. 2004. 

Afhandlingen indeholder desuden resultater møntet på de grønlandske 

forvaltningsmyndigheder. Foruden estimater af ynglebestanden i de to 

vigtigste havternekolonier i Vest- og Nordøstgrønland, præsenteres anbefalinger 

til hvordan en fremtidig bæredygtig udnyttelse af havterneægsamling 

kan tænkes udført (Manus II), og konkrete forslag til hvordan 

overvågningen af havternekolonierne i Grønland kan udføres (Manus IV).

Eqikkaaneq 

Ilisimatuutut allaatigisaq manna imeqqutaallanik (Sterna paradisaea) 2002- 

2008-mi misissuinerit inernerinik saqqummiussivoq. Allaatigisamut 

atatillugu paasissutissat Kalaallit Nunaanni piffi nni marluusuni: Kitsissunnguit, 

Qeqertarsuup Tunuani Kitaata issittortaani aamma Sandøenimi 

Tunup avannaani issittorsuarmi, misissuiartornikkut katersorneqarsimapput. 

Kalaallit Nunaanni imeqqutaalaq pillugu 2002 sioqqullugu ilisimasanut 

tunngaviusut annikitsuinnaasimapput, pissuseqatigiit nunatsinni uumasuusut 

kinguaassiornikkut pissusaat arlalitsigut nassuiarneqarsimanatik. 

Allaatigisaq manna imeqqutaalaq pillugu paasisanik nutaanik nunap iluani 

nunallu tamat akornanni nallersuussinnaasunik saqqummiussivoq. 

Siullertut pingaarnertullu imeqqutaallap ingerlaartarnera pillugu misissuineq 

(Manus I) – uumasuni tamani ingerlaarnerit nalunaarsorneqarsimasut 

tannersaat – misissuineruvoq nunat tamat akornanni soqutiginaateqartussaq. 

Misissuinerup nassuiarpaa qanoq imeqqutaallat Tunup 

avannaaniit Islandimiillu, Sikuiuitsumi kujallermiittumut Weddelip Imartaanut, 

utimullu, ingerlaartarnersut. Ingerlaarnerup takissusaa immini 

(agguaqatigiissillugu 71.000 km) soqutiginarpoq, misissuinerulli aamma 

nassuiarpaa, qanoq imeqqutaallat ingerlaarnermi nalaanni imartat immikkut 

pinngorarfi ulluartut pinngitsoorsinnaanngikkaat. Kujammut ingerlaarnerminni 

timmissat qaammatingajammik (agguaqatigiissillugu 

ulluni 25-ni) sivisussusilimmik ingerlaarnertik unitsittarpaat Atlantikullu 

avannaata qiterpasissuani, kujammut ingerlaqqinnginnerminni, uninngaartarlutik. 

Ækvatorip qanittuani (~10º N) ingerlaarnerup aqqutaani 

avissaarfeqartarpoq: Nalunaaqutserneqarsimasuni 11-usuni arfi neq 

marluk Afrikap sineriaa atuarlugu ingerlaqqipput, sisamallu Atlantiku 

ikaarlugu Amerikkap kujalliup sineriaa atuarlugu ingerlaqqiffi galugu. 

Avannamut ukiivimmiit piaqqisarfi mmut ingerlaarneq sukkangaatsiartumik 

ingerlavoq (agguaqatigiissillugu ullormut 520 km) ingerlaarfi llu iluaqutaasunik 

sammivilinnik anoreqarfi usoq atuarsimallugu. 

Allaaserisap massuma aamma imarivaa imeqqutaallap mannilioqqissinnaaneranik 

– piaqqinerit/manniliornerit siulliit iluatsinngippata mannilioqqissinnaaneq 

- kisitsisinngorlugu missingersuinerit siullersaat 

(Manus II aamma III). Naak allaatigisani imeqqutaallat ajornasaaratik 

mannilioqqittarnerarlugit oqaatiginiarneqartaraluartut, misissuinerup 

massuma, aappariit piaqqisut misileraatigalugit uukapaatinneqartut 

kinguaassiornikkut qisuariarnerannik nakkutiginninnerit siullersaraat. 

Paasivarput, timmissat affaasa missaat (53,3 %) ivanerup naggataata tungaani 

manniliaat peerneqaraangata mannilioqqittartut. Tupaallannartumik 

paasinarsivoq, piaqqat mannilioqqinnerniit tukersimasut alliartornerat 

umaannarsinnaanerallu piaqqanit sapaatit akunnerini sisamani 

siusinnerusukkut allisarsimasuninngarnit allaanerunngitsut, naak, neriniarnikkut 

piniagaasalu angissusiisigut allaanerussutit ilisarineqarsinnaasimagaluartut 

(Manus III).Nunap iluani soqutiginaateqarnertut allaatigisamut 

atatillugu imeqqutaallat Kalaallit Nunaanni neriniartagaannik 

misissuinerit siulliit suliarineqarput. Naak aalisakkat uumasuaqqallu 

ikerinnarsiortut assigiinngitsut arlallit Kitaani, Qeqertarsuup Tunuani 

piaqqanut apuunneqartartunut ilaagaluartut, paasinarsivoq ammassak 

(Mallotus villosus) piaqqani tamanik angissuseqartuni nerisassat pingaarnersarigaat 

(Manus III). 

13

Figure 3. Map of study areas 

where fi eld work in the thesis 

were conducted. Inset map in upper 

left corner shows the location 

of A) Kitsissunnguit in Disko Bay 

and B) of Sand Island in Northeast 

Greenland. The map scale 

(lower right corner) applies to 

both A and B 

14 

Allaatigisami aamma sammineqarput imeqqutaallat piaqqiortartut amerlassutsimikkut 

ilaatigut allanngorartarnerisa, misissuiffi gineqarnerat 

(Manus IV). Naak imeqqutaalaq piaqqisarfi mminut amerlanertigut aalajaattaraluartoq, 

misissuinitta takutippaat, Qeqertarsuup Tunuani piaqqisarfi 

nni minnerni akunnattunilu timmissat piaqqisut piaqqinermiit piaqqinermut 

malunnaatilimmik allanngorartut. Allanngorarnerit tamakku 

piffi mmi annikitsumi pisunut attuumassuteqarnissaat ilimanaateqarput, 

assersuutigalugu kiisortunit akornusersorneqarneq, piffi mmilu annertunerusumi 

pisut nassuiaataasinnaanerat ilimanaateqannginnerulluni. 

Imeqqutaalaqarfi nnut misissuiartortarnerni sivisoqisuni, imeqqutaallat 

timmissallu naluusillit piaqqisut allat imminnut iluaqusersoqatigiittarneri 

nalunaarsorneqarsinnaasimapput. Naluumasortut Kajuaqqallu 

(Phalaropus fulucarius aamma P. lobatus) piaqqisut Kitsissunnguani misissuiffi 

ginerisigut (Manus V), imeqqutaallat kalerrisaarinerinut qisuariaatit 

ersarissut ilisarineqarsinnaasimapput. Inernerit taakkua qularnaallisippaat, 

naluumasortukkut Kitsissunnguani siammarsimaffi mmikkut 

allannguutaasa imeqqutaallat qeqertani sumiinneri malittarigaat, soorlu 

Egevang et al. 2004-mi siunnersuutigineqarsimasoq. 

Allaatigisap aamma imarai inernerit Kalaallit Nunaanni oqartussanut 

saaffi ginnittut. Kitaani Tunullu avannaani imeqqutaalaqarfi nni pingaarnerpaani 

piaqqisut amerlassusaannik missingersuinerit saniatigut, siunissami 

piujuartitsilluni imeqqutaallanik manissarluni atuisinnaanermik 

inassuteqaatit saqqummiunneqarput (Manus II), Kalaallillu Nunaanni 

imeqqutaalaqarfi it nakkutigineqarsinnaanerannut tigussaasunik siunnersuuteqartoqarluni 

(Manus IV).

1 Synopsis 

1.1 Introduction 

There are several compelling reasons to study the migration and breeding 

biology of the Arctic tern in Greenland. From a global perspective, the 

Arctic tern is the very epitome of long-distance migration in birds, and 

no other animal species connects the two Polar Regions, as the Arctic tern 

does. Thus, the study of Arctic tern migration is not only an examination 

of migrating animals at the edge of their physical performance, it is also an 

exploration of a species moving through a vast proportion of the world’s 

marine areas in a single calendar year, experiencing a rapidly changing 

environment with altered foraging opportunities on both the breeding 

and wintering grounds. 

At a local scale, no other bird species in Greenland encapsulates the feeling 

of summer in the Arctic. The cry of the Arctic tern is for many in Greenland 

the ultimate sign of the end of a long winter and the onset of summer, 

the return of the sun and warmer temperatures. Although tern eggs 

are small and may not contribute much as a source of protein, the habit of 

harvesting Arctic tern eggs in summer has a long tradition in Greenland, 

and is considered one of the highlights of the summer season. There is no 

doubt that, to the people of Greenland, the Arctic tern has a special status 

amongst birds. 

From a more scientifi c point of view, the Arctic tern is interesting to study 

due to its obvious appearance on the landscape and the placement of 

its nest directly on the ground. This makes the species relatively easy to 

study, compared with many other seabird species, where the nest is concealed 

in a rocky crevice or burrow, or on rocky ledges of high steep cliffs. 

Furthermore, the Arctic tern is a surface feeder and only capable of foraging 

in the top half meter of the water column. Theoretically, this twodimensional 

feeding niche makes the species more vulnerable to changes 

in prey availability, and a direct response in breeding performance can be 

detected within a breeding season. 

From an environmental management perspective, the Arctic tern is interesting 

because of its function as a “biodiversity generator”. Smaller species, 

such as shorebirds, apparently benefi t from the agonistic behaviour 

of the Arctic tern, and increased breeding densities of other species are 

often found within the boundaries of an Arctic tern colony (Nguyen et al. 

2006, Egevang et al. 2008, Egevang et al. 2004, Egevang et al. 2006). 

The Greenland Institute of Natural Resources generates specifi c knowledge 

on the Greenland marine and terrestrial fauna to provide the Greenland 

Government with management advice on the sustainable use of living 

resources. In Greenland, there is a long traditional of bird harvesting, 

and research on seabirds has mainly focused on the two most harvested 

species, Brünnich’s Guillemot (Uria lomvia) and Common Eider (Somateria 

mollissima), with little attention directed towards other harvested species, 

such as the Arctic tern. As monitoring programmes for guillemots and eiders 

developed over the last decade, opportunities to expanding research 

activities of GINR (to species like little auks, black-legged kittiwakes, and 

Arctic terns) arose. 

15

16 

1.2 Objectives 

Aim of the thesis: The initial aims of and motivations for my research 

were to illuminate a number of issues regarding the biology of the Arctic 

tern in Greenland, particularly from a management perspective. Basic 

questions, like “What is the current status of the Arctic tern population 

in Greenland ?”, “What is the key prey species for Arctic terns in Greenland 

?” and “Do Arctic terns relay in the Arctic ?” were unanswered. Later 

in the process, technological advances rapidly reduced the mass of geolocators, 

and allowed the Arctic tern to be targeted as a study subject, and 

the focus of my research broadened to also include this aspect. 

Aim of the synopsis: The aim of this synopsis is to present background 

information on the biology of the Arctic tern and highlight how the results 

of this work have contributed to an improved understanding of this 

species, either globally or specifi cally in Greenland. In this context, the 

synopsis also includes results generated during the course of my research, 

but not presented in the manuscripts – especially those relating to issues 

specifi c to Greenland. 

1.3 Distribution 

The Arctic tern has a circumpolar distribution between the Boreal and 

High Arctic zones. The breeding range in Greenland is large (more than 

350,000 km 2 ), and the species is found breeding in a variety of habitats. 

In Greenland, Arctic terns are found breeding throughout the country. In 

West Greenland, the Arctic tern shows a patchy distribution with large 

gaps evident. The core breeding areas are in Disko Bay and in Upernavik 

District (Boertmann et al. 1996, Egevang and Boertmann 2003). In East 

Greenland, the colonies are fewer, more scattered, and smaller than in 

West Greenland. Typically, colonies in Greenland are found at small, isolated 

islets along the coast, and range from a few tens of pairs to a few 

hundred pairs. Large colonies of more than 2,000 pairs are rare, and even 

solitary breeding has been recorded in Greenland (Boertmann 1994). 

1.4 Population status in Greenland 

The total world population of Arctic terns is estimated at 1 to 2 million 

pairs (Lloyd et al. 1991). Global estimates of the Arctic tern population are 

rough, however, and some large populations (e.g. in Russia and Alaska 

are diffi cult to estimate (Wetlands International 2006). 

Prior to 2002, there was little effort to monitor the Arctic tern population 

in Greenland. Counts were opportunistic and only performed in single 

years. The total Arctic tern population in Greenland is estimated to include 

at least 65,000 pairs (Egevang and Boertmann 2003), but limited data 

is available. However, three studies provide information that can be used 

to address population status: Salomonsen (1950) visited Kitsissunnguit, 

the largest Arctic tern colony in Greenland, for a short (1 ½ day) period in 

1949, and estimated the population to be 100,000 pairs. This high number 

of breeding birds has never been encountered again in censuses conduct-

ed between 1980 and 2009, where estimates range from 0 to 25,000 pairs 

(Kampp 1980, Frich 1997, Hjarsen 2000, Egevang et al. 2004, Manus IV). 

Burnham et al. (2005) revisited Arctic tern colonies censured in Uummannaq 

District by Berthelsen (1921) in the period 1905 to 1920, to fi nd that 

population numbers had declined to less than one-third between counts. 

Egevang and Boertmann (2003) compared Arctic tern colony counts in 

the Greenland Seabird Colony Database (GSCD) between 1921 and 2002, 

and concluded that a long-term decline was present, but the population 

seemed to stabilize in more recent counts, between 1986 and 2002. 

Although the above studies provide only circumstantial evidence of a decline, 

it seems reasonable to conclude that the Arctic tern population in 

Greenland has undergone a decline since the middle of the last century. 

However, the magnitude of this decline is diffi cult to quantify and attempts 

to estimate the historical population size should be treated with caution. 

The Arctic tern is categorized as “near threatened” in the Greenland Red 

List (Boertmann 2007). 

1.5 Population status at Kitsissunnguit 

The Arctic tern population at Kitsissunnguit has been followed with special 

interest by both managers in Greenland and international, non-government 

environmental organisations. The archipelago is home to one of 

the largest Arctic tern colonies in the world, and has been designated both 

as a Ramsar site, a special protected area for breeding birds in national 

legislation, and an Important Bird Area (IBA). Despite these designations, 

there was little effort to monitor the Arctic tern population in a systematic 

way prior to 2002. The sparse historical information on population size 

available from Kitsissunnguit indicated a crash in the population, with 

numbers going from 100,000 pairs in 1949 (Salomonsen 1950) to no breeding 

birds in 2000 (Hjarsen 2000), and a great concern arose within government 

and non-governmental institutions. 

The Greenland Institute of National Resources introduced a modifi ed distance 

sampling and line transect method to estimate population size at 

Kitsissunnguit in 1996 (Frich 1997), and annual counts were performed 

between 2002 and 2006 (Manus IV). This series of consecutive population 

estimates showed that the Arctic terns at Kitsissunnguit had not suffered 

a complete population crash, and found breeding numbers to be relatively 

stable between 15,000 and 22,000 pairs (Manus IV). Although these estimates 

are not in the same magnitude as the former estimate by Salomonsen 

in the mid 20th century, it is within the same magnitude as the 

estimate of roughly 25,000 pairs by Kampp et al. in 1980. 

1.6 Fluctuating breeding 

The Arctic tern is often referred to as a species with fl uctuating colony 

attendance, large variation in breeding numbers, and years of skipped 

breeding (Manus IV). Breeding site fi delity at a regional level is high in 

Arctic terns, but dispersal to neighbouring breeding colonies occurs frequently 

(Devlin et al. 2008, Møller et al. 2006, Brindley et al. 1999, Ratcliffe 

17

18 

2004). Although poorly understood and documented in the literature, it 

is important to address whether variations in colony size may be caused 

by locally occurring phenomena, such as predation and disturbance, or if 

these variations are linked to large scale events, such as food availability 

or climatic variations. In order to make sense of counts at Arctic tern colonies 

where variation occurs, it is important to understand the scale (from 

within-colony to regional scale) at which variations in colony attendance 

occur. A particular combination of factors makes it diffi cult to address 

population size, status, and trends in Arctic terns (Ratcliffe 2004). This is 

most pronounced in the Arctic zone where Arctic tern breeding sites may 

be diffi cult to access and colonies can be widely distributed. 

In addition to estimates of population size from the two study sites, Kitsissunnguit 

(2002-2006) and Sand Island (2006-2009), it was possible to conduct 

annual counts in 14 Arctic tern colonies in the southern part of Disko 

Bay (the Akunnaaq area) in the period 2002-2005 (Manus IV). Considerable 

fl uctuation was found between years in these small and mid-sized 

colonies (mean CV of individual colonies equalled 117.5 %), whereas the 

colonies at Kitsissunnguit only showed minor annual variation (47.4 % at 

sub-colony level, 14.6 % at overall level). When combined, however, the 

total population size of the Disko Bay colonies (more that 80 % of the Artic 

tern colonies in Disko Bay), varied little (CV 6.7 %), indicating local 

movements between the colonies and that annual variation was linked to 

locally occurring phenomena. 

Variation at high-arctic Sand Island was even more pronounced, with complete 

breeding failures recorded in two out of four seasons (Manus IV). 

1.7 Plausible cause of decline in population size 

Arctic tern populations in Greenland are affected by numerous factors regulating 

population size and the observed long-term decline is likely to be 

caused by more than one factor. Arctic foxes (Alopex lagopus) are known to 

cause devastating predation when they are able to access a colony (Bianki 

and Isaksen 2000, Manus IV), and multiple avian predators are known to 

depredate egg, chicks, and, to a lesser degree, adult Arctic terns (Cramp 

1985, Hatch 2002). Although these predators have always been present in 

Greenland, local Inuit claim that the number of predators in Greenland 

has increased, causing terns to decline. Although the relationship between 

food availability and reproductive performance in Arctic terns in Greenland 

is poorly documented, the study on chick diet and chick survival at 

Kitsissunnguit (Table 2, Manus III) indicates that productivity is low compared 

with other studies abroad (Appendix I in Hatch 2002). Food shortage 

or inaccessibility of suitable prey items can occur in Greenland waters 

and may affect the ability of Arctic terns to reproduce in some years. 

In Greenland, harvesting of Arctic tern eggs for consumption has probably 

taken place as long as humans have inhabited the country (Manus II). 

The impact of this harvesting was initially constrained by a small human 

population size and limited forms of transportation. The human population 

in Greenland has more than quadrupled since the 1950’s (Fægteborg 

2000), however, and the number and size of motorized boats has increased 

rapidly, allowing greater mobility, and likely increasing pressure on the 

Arctic tern population.

The extent of the annual egg harvest in Greenland remains undocumented, 

although a few references may provide indications of its magnitude: 

Salomonsen (1950) estimated the annual harvest at Kitsissunnguit to be 

as high as 100,000 eggs, while Frich (1997) estimated that 150-200 persons 

visited the archipelago between 6 and 25 June 1996, and harvested between 

3,000 and 6,000 eggs. Since 2002, the level of (illegal) harvesting has 

declined at Kitsissunnguit (pers. obs.). 

Conservation concerns for the Arctic tern population in Greenland led to 

an altered managing of the resource: from an open season for egg harvest 

until 1 July before 2001, a ban on Arctic tern harvesting was introduced in 

2002. Egg harvesting in Greenland is very much a cultural activity, where 

several generations participate in gathering eggs during summer. The ban 

on Arctic tern egg harvesting has been met by opposition in Greenland 

where some even consider it a violation of basic human rights. Since 2002, 

there has been a considerable pressure on national politicians to re-open 

the Arctic tern egg harvest season in Greenland. 

A harvest of Arctic tern eggs is also carried out in Iceland, and, apparently, 

this is conducted without an adverse effect on the population. The egg 

harvest in Iceland differs from the Greenland harvest in several ways: in 

Iceland the season closes on 15 June, the harvest is only conducted by private 

landowners, and a high degree of predator control (foxes, gulls, and 

ravens) takes place (Aevar Petersen pers. com.). 

The outcome of a model dealing with egg harvesting at Kitsissunnguit 

(Manus II) indicates that a fairly high number of eggs could be removed 

from the breeding population each year as long as egg harvesting is only 

conducted early in season. The model predicted 7 June as a date for harvest 

closure – the date where 92% of the maximum number of eggs would 

be available to harvesters, but the impact on productivity would be minor. 

The highest number of eggs could be harvested on 16 June, whereas harvest 

until 22 June would negatively affect population productivity. Based 

on the results of the model, high effort egg harvesting until 1 July (as was 

the case prior to 2002), is very likely to have placed an unsustainable pressure 

on the population, and could very well explain the observed decline. 

1.8 Renesting in Arctic terns 

In order to address the impact of egg harvesting on the Arctic tern population 

in Greenland, basic parameters of breeding biology are required 

for the species. Although the Arctic tern is often mentioned as a species 

that willingly produces a replacement clutch if the fi rst clutch is lost (e.g. 

depredated or fl ooded) this is in fact very poorly described in the literature. 

The only scientifi c reference to my knowledge relating to relaying 

in Arctic terns is a paper by Bianki (1967) in Russian, quoted by Cramp 

(1985) and Hatch (2002), showing that relaying only occurs if eggs are lost 

during the fi rst 10 days of the incubation period. Cramp (1985) also states 

(without references) that relaying mainly takes place at lower latitudes, 

but is “unlikely in the high Arctic”. Population dynamic parameters (such 

as 1) what proportion of the population will relay, 2) at what stage in the 

incubation period relaying will occur, and 3) growth and survival rates of 

chicks originating from replacement clutches), remain largely unknown 

in Arctic terns. 

19

20 

Renesting in Arctic terns was an important component in the fi eldwork at 

Kitsissunnguit (Manus II and III). By experimentally removing eggs and 

monitoring the response of the breeding birds, the fi rst estimates of relaying 

rate, and chick growth and survival rates were obtained for Arctic terns 

(Manus III). However, this challenge was not without large logistical diffi 

culties. The laying date in Arctic terns breeding in Disko Bay varied (see 

section “Variation in breeding phenology”) by as much as three weeks, 

making the onset of fi eldwork diffi cult to plan. If breeding terns were to 

be followed from early incubation, and egg removal experiments allowing 

the birds to mate, egg formation to take place, an incubation period of 22 

days, and chicks were to be followed until fl edging, the fi eld season had 

to be of two to three months in duration. In practice, the demands of such 

a long fi eld period proved diffi cult to meet, and not all aspects of renesting 

in Arctic terns could be covered within the framework of this research. 

Despite the fact that eggs were removed in the latter half of incubation, 

and past the 10 day period suggested by Bianki (1967), at least 16 out of 30 

breeding pairs (53.3%) at Kitsissunnguit produced a replacement clutch 

(Manus III). The clutch size (1.6) at these nests was signifi cantly lower 

than in control nests (2.1), as was the total Internal Egg Volume (IEV). 

None of the replacement eggs were relayed in the original nest cup, but 

were instead placed in a new nest at an average distance of 23.4 meters 

from the original nest cup. 

The results from Kitsissunnguit verify the general perception that Arctic 

terns do produce a replacement clutch – even if the eggs are removed 

relatively late in the breeding season. Although the long term impact of 

relaying, such as the post-fl edging survival of chicks, and the extra energy 

expenditure required of adult females, are not accounted for in this study, 

it does indicate that the birds are cable of compensating for the loss of 

eggs to some extent. The fi ndings in Manus III, where chicks from replacement 

clutches did not differ in their growth and survival (at least to eleven 

days), further support this conclusion. 

1.9 Variation in breeding phenology 

The fi eldwork at Kitsissunnguit and at Sand Island also revealed that 

breeding phenology may vary considerably between years, with considerable 

variation in egg laying between breeding seasons, as also seen in 

Arctic terns in Shetland (Suddaby and Ratcliffe 1997). At Kitsissunnguit, 

the majority of eggs hatched in the fi rst week of July (indicating that egg 

laying took place in mid-June), but in some years early breeding takes 

place – as was the case in 2004, where eggs hatched three weeks earlier 

in mid June (Egevang et al. 2005). At Sand Island, hatching took place approximately 

two weeks earlier in 2008 than in 2007, with a broader distribution 

of hatching dates observed (Egevang et al. 2008, Egevang and 

Stenhouse 2007). 

The onset of egg laying is likely correlated with sea ice conditions in Arctic 

regions (Hatch 2002) and this may explain the observed variation in this 

study. The variation of up to three weeks difference in egg laying may 

have implications for population monitoring at Arctic tern colonies. If 

pair formation is not completed, fl ush count estimates will underestimate 

population numbers. In line transect estimates (Manus IV), both the count

efore (not all pairs will have laid their eggs) and after (eggs will have 

hatched and chicks not recorded by the observer) the median incubation 

period will produce underestimates. Before conducting counts in a given 

year it is recommended that the stage of incubation is addressed – preferably 

by estimating egg age (Egevang et al. 2005). 

1.10 Clutch size 

From the study in Disko Bay and in Young Sund (Tab. 1) information on 

average clutch size was obtained. The distribution of one-, two-, and threeegg 

clutches differed signifi cantly (χ 2 = 126.02, df = 8, p = < 0.001) between 

seasons. At high-arctic Sand Island, no 3-egg clutches were observed in 

the study plots (nor in random observations) during the fi eld seasons in 

2007 and 2008, but a difference (χ 2 = 5.99, df = 1, p = < 0.014) in the proportion 

of one- and two-egg clutches was identifi ed. The overall clutch size 

at Sand Island (1.51, SD = 0.50, n = 169) was lower than at Kitsissunnguit 

(1.84, SD = 0.50, n = 1752), and the distribution of one-, two-, and three-egg 

clutches differed signifi cantly (χ 2 = 66.45, df = 2, p = < 0.001). 

From the fi ndings in this study from Greenland, it seems reasonable to 

conclude that investment in a 3-egg clutch only occurs on a regular basis 

at lower latitudes in the Arctic zone and rarely in the High Arctic. This 

hypothesis supported by a summary of clutch sizes found in Arctic terns 

across North America (Hatch 2002, Appendix 1), where 3-egg clutches are 

absent from locations north of 70° N. 

Table 1. Average clutch size (± Standard Deviation) and sample size (n = no. of nests) and 

the distribution of one-, two-, and three-egg clutches in Arctic tern nests from Kitsissunnguit, 

western Greenland, 2002-2006, and Sand Island, northeast Greenland, 2007-2008. 

Location Year Clutch size (SD) Clutch size % (n) 

1.11 Hatching and fl edging success 

1 2 3 

Kitsissunnguit 2002 1.80 (0.50) n=292 24.3 (71) 72.2 (208) 4.5 (13) 





Sand Island 2007 1.44 (0.50) n=109 56.0 (61) 44.0 (48) 0 

Sand Island 2008 1.65 (0.50) n=60 35.0 (21) 65.0 (39) 0 

Hatching success (Tab. 2) at Kitsissunnguit did not differ between years 

(χ 2 = 0.32, df = 3, p = 0.956), but hatching success at Sand Island in 2008 

was signifi cantly lower (χ 2 = 4.04, df = 1, p = 0.045) than Kitsissunnguit (all 

years combined). Fledging success in 2002, 2005, and 2006 (2003 data excluded 

due to short fi eld season) at Kitsissunnguit did not differ between 

years (χ 2 = 0.19, df = 2, p = 0.911), whereas fl edging success turned out to be 

signifi cantly higher at Sand Island (χ 2 = 5.06, df = 1, p = 0.0259). 

The highest proportion (14.2 %) of 3-egg clutches was found in 2006 –the 

same year the highest average chick survival was encountered. Further- 

21

Figure 4. Growth of Arctic tern 

chicks in 2005 at Kitsissunnguit, 

Disko Bay, West Greenland. The 

fi gure shows average mass and 

SE bars in chicks (15.2 daily 

measurements on average) from 

hatching to fl edgling. In the linear 

growth period (4-14 days) the 

equation describing daily mass 

increase is inserted. 

22 

Table 2. Hatching success (proportion of eggs hatched), fl edging success (proportion of 

chicks fl edged) and productivity (number of fl edglings per nest) from Kitsissunnguit, western 

Greenland, 2002-2006, and Sand Island, northeast Greenland, 2007-2008. 

Location Year Hatching success a Fledging success a Productivity (SD) n 

Kitsissunnguit 2002 0.90 (n=65) 0.37 (n=59) 0.56 (0.50) n=39 

Kitsissunnguit 2003 0.93 (n=54) 0.58b (n=50) 0.58b (0.319) n=24 

Kitsissunnguit 2004 * * * 



Sand Island 2007 * * * 

Sand Island 2008 0.84 (n=85) 0.69 (n=71) 1.04 (0.771) n=48 

aProductivity and Hatching/Fledging success is not necessary based on measurements in the same eggs/nests. 

bIn 2003 chicks could only be followed to the age of 7-15 days old. Values of fl edging success and productivity 

may not be representative. 

* = no data on chick production in 2004 and 2007. 

more, the year with the second highest recorded chick production (2005) 

was also the season with the second highest proportion of 3-egg clutches. 

Although the few data points in the study may not be suffi cient to test if 

Arctic terns are able to adjust their clutch size to food availability in the 

current breeding season (Suddaby and Ratcliffe 1997, Kilpi et al. 1992), the 

results suggest that this may in fact the case. 

1.12 Growth rate and chick survival 

During the course of fi eldwork, estimates of hatching success, chick 

growth and chick survival were obtained. Arctic tern chicks becomes agile 

and mobile after less than two days and will immediately start running 

away from the nest site if a land predator (humans included) approaches 

(Bianki and Isaksen 2000). In order to obtain consistent daily measurements 

of chick growth and survival, a small enclosure made from chicken 

wire was set up around Arctic tern nests at the study site. This method has 

previously been applied (Monaghan et al. 1989 a, b) in Arctic tern studies, 

and the birds adapt well to the enclosure after a short habituation 

period. Nests were found during the incubation period and the nest area 

was enclosed prior to hatching. As land predators were absent from the 

study sites during the chick rearing period, and the nest area could easy be 

approached from the air, it is unlikely that the enclosures induced biased 

estimates of the level of predation. 

Mass (g) 

120 

100 

80 

60 

40 

20 

0 

0 5 10 15 20 25 30 

Chick age (days) 

y = 5.558x – 3.6022 

R 2 = 0.9568

Table 3. Table of average growth daily rates (mass and wing) in Arctic tern chicks aged 4-14 

days at Kitsissunnguit, West Greenland (see Fig. 4). The number of daily measurements decreases 

over the season. Average daily measurements are presented in parenthesis. 

Year MASS WING 

2002 5.32 (n=15.5) 7.61 (n=15.5) 

2003 5.87 (n=25) 8.05 (n=19) 

2005 5.56 (n=15.2) 6.83 (n=15.6) 

2006 6.30 (n=39.8) 7.66 (n=34.4) 

Arctic tern chick growth rates were measured as the daily increase in mass 

and wing length. As growth is known to follow a sigmoidal pattern, with 

little increase in the fi rst 3 days and levelling off after age 15 days (Fig. 

4), the linear period between 4 and 14 days was used to compare groups 

and seasons (Manus II and III, Tab. 3). This period was also used in terns 

at Shetland (Suddaby and Ratcliffe 1997) to address differences in growth 

rates between seasons. In Shetland, chick growth varied with year and 

ranged between 6.9 and 7.8 g per day – the same order of magnitude as 

recorded at Kitsissunnguit (Tab. 3). 

When obtaining measures for chick survival (Tab. 2), the enclosures were 

an essential tool in keeping track of the chicks from hatching to fl edging. 

As the chicks grow older they become very mobile and diffi cult to fi nd 

and information on their fate (alive or dead) becomes almost impossible 

to achieve without enclosures. 

1.13 Feeding 

The Arctic tern is an opportunistic, plunge-diving and surface-dipping 

feeder, with a diet comprising mainly small fi sh and large zooplankton 

(Cramps 1985, Hatch 2002). The species shows a high diversity of food 

items throughout its breeding distribution and during migration. The key 

prey species of the Arctic tern varies with its distribution, but fi sh is reported 

to be important in the diet throughout the breeding range (Hatch 

2002). There are no systematic studies on the foraging range of the Arctic 

tern available, but “most” feeding is reported to take place within a 10 km 

(Boecker 1967) or 20 km (Pearson 1968) radius of the colony. 

The literature on the diet of Arctic Terns in Greenland is very limited 

(Salomonsen (1950) mentioned zooplankton as important), which makes 

the results of the fi eldwork at Kitsissunnguit and Sand Island relevant 

in a general context. At Kitsissunnguit, capelin (Mallotus villosus) was the 

single most important prey item over the study period (2002-2006). More 

detailed studies (Manus III) revealed that capelin accounted for up to 

99.9 % of the energy intake of Arctic tern chicks in periods. The study at 

Kitsissunnguit also revealed variation in diet composition between seasons 

(Manus III and unpubl. data). In some years, other prey items (e.g. 

Stichaeidae sp. in 2005 and 2006, Wolffi sh (Anarhichas spp.) in 2006) may 

become important for shorter periods during chick rearing. Even discards 

from shrimp vessels appear in the chick diet when the commercial catch is 

sorted close to colony islands. Although these occasional infl uxes of alternative 

prey items appear in the diet, the importance of capelin is unquestionable 

as the key prey species of Arctic terns in Disko Bay, and probably 

23

Figure 5. Arctic tern chick fed with polar cod at Sand Island. 

24 

across most of West Greenland, where capelin constitutes the link between 

zooplankton and higher trophic levels (Carscadden et al. 2002; Friis-Rødel 

and Kanneworff 2002). 

At high-arctic Sand Island, opportunistic observations of prey items carried 

by adult birds during 2007 and 2008 were conducted, along with systematic 

observations of chick diet in 2008 (Egevang and Stenhouse 2007, 

Egevang et al. 2008). Polar cod (both larvae and juvenile fi sh) appeared 

to be the most important prey species in the chick diet, with crustaceans 

(especially Thysanoessa ssp.) as secondary prey items (Egevang et al. 2008). 

1.14 Breeding association 

Photo: Carsten Egevang. 

The Arctic tern shows a strong antagonistic, anti-predator behaviour towards 

potential predators within the colony boundaries. Both land mammals 

and avian predators are fi ercely attacked from the air (Hatch 2002). 

Shorebirds such as purple sandpiper (Calidris maritima), semipalmated 

plover (Charadrius semipalmatus), and red phalarope (Phalaropus fulicarius) 

are known to breed in higher densities and with reduced nest predation 

when found within or close to an Arctic tern colony (Nguyen et al. 2006, 

Smith et al. 2007, Summer and Nicoll 2004). In Greenland, Sabine’s gulls 

(Xema sabini) breed almost exclusively with Arctic terns (Levermann and 

Tøttrup 2007), with 21 out of 24 colonies found in association with tern 

colonies. Ross’s gull (Rhodostethia rosea) has only been found breeding ten 

times in Greenland, and nine of these breeding records have been in association 

with breeding Arctic terns (Egevang and Boertmann 2008). At Kitsissunnguit, 

there is a strong breeding association between Arctic terns and 

red-necked phalaropes (Phalaropus lobatus) and red phalarope (Egevang et al.

2006, Egevang et al. 2004). Analysis of the historical and present distribution 

of breeding terns and phalaropes at Kitsissunnguit showed that phalarope 

breeding densities are higher on islands where terns breed. In the case of the 

red phalarope, the breeding association is particularly strong and breeding 

only occurred where breeding terns are found (Egevang et al. 2004). Breeding 

association and higher breeding densities are also documented for mallards 

(Anas platyrhynchos) and red-breasted mergansers (Mergus serrator), 

while long-tailed ducks (Clangula hyemalis) are exclusively found breeding 

at islands withterns (Egevang et al. 2006). 

At Kitsissunnguit, where high numbers of Arctic terns breed, the response 

to predators within the boundaries of the colony is very strong. 

The presence of an avian predator results in mobbing behaviour of hundreds, 

sometimes thousands, of terns with intense vocalisation. In order 

to document the breeding association between phalaropes and terns, we 

conducted a study (Manus V) on phalarope behavioural response to tern 

alarm calls. Potential avian predators over-fl ying the colony and both land 

and avian predators at the outskirts of the colony do not trigger a response 

from the terns. As soon as the predator enters an area with breeding terns, 

alarm calls and mobbing is initiated by the terns. These alarm calls vary, 

not only in duration, but also in the character of the call (Hatch 2002). 

Alarm calls in the study were divided into three groups originating from: 

1) falcons or gulls, 2) human disturbance, or 3) dreads (false alarms). By 

recording behavioural response from both red-necked and red phalaropes 

to tern alarm calls, we were able to document a strong relationship between 

the species. Phalaropes did not respond differently to alarms 

caused by a predator compared to false alarms, but extended tern alarms 

caused stronger phalarope response than short ones. In conclusion, phalaropes 

benefi t from both a higher predator detection rate and aggressive 

defence in the tern nesting area. This supports the hypothesis (Egevang et 

al. 2004, Egevang et al. 2006) of a decline in phalarope numbers and shift 

in phalarope distribution at Kitsissunnguit, governed by the distribution 

of breeding Arctic terns (Manus V). 

1.15 Migration 

The Arctic tern is known to perform the longest annual migration in the 

animal kingdom. The distance alone, from the breeding colonies at high 

latitudes of the northern hemisphere to the winter quarters at high latitudes 

in the southern hemisphere, makes the Arctic tern a candidate for 

the longest annual roundtrip (Berthold 2001, Newton 2008). There were 

several early attempts to interpret the migration of the species (Kullenberg 

1946, Storr 1958), but the fi rst comprehensive review was presented by 

Salomonsen (1967). Based on ringing recoveries and observations from 

the southern seas, Salomonsen suggested that the southbound migration 

follows the coast of West Europe and later West Africa, with the birds arriving 

in the waters off South Africa in November. Salomonsen suggested 

the main wintering areas to be off the southern tip of Africa and South 

America, and that Arctic terns might circumnavigate the Antarctic continent 

in early spring before initiating their northbound migration. This 

hypothesis was later supported by radar and visual observations of terns 

migrating from the Bellingshausen Sea to the Weddell Sea (Gudmundsson 

et al.1992). It has been suggested that Arctic terns complete the southbound 

migration in a step-wise fashion, moving between rich feeding 

25

Figure 6. An Arctic tern equipped 

with a geo-locator in 2007 returns 

to the breeding site at Sand Island 

in Northeast Greenland the 

following year, after having completed 

a round-trip migration of 

more than 70,000 km. 

26 

grounds (Alerstam 1985). Crossing over land areas and the Greenland 

inland ice at high altitude was suggested to take place in the migration 

of the Arctic tern in the mid-1980s using radar observations (Alerstam et 

al. 1986). There was little evidence of the northbound migration of the 

Arctic tern in spring available for Salomonsen, but a ringing recovery of a 

Greenland ringed bird from Columbia in the highland of South America, 

led him to speculate that at least some birds may end up in the Pacifi c 

Ocean and cross over land to enter the Atlantic Ocean. However, the major 

northbound migration was believed to be initiated in March in the Atlantic 

Ocean and occur more rapidly and over a wider front compared to 

autumn (Kampp 2001, Bourne and Casement 1996). 

Several authors (e.g. Berthold 2001, Newton 2008) have attempted to estimate 

the total distance travelled during the annual migration of the Arctic 

tern. As the fl ight path was largely unknown, these estimates vary, but 

typically a fi gure of 40,000 km was quoted as the annual trip from breeding 

grounds to the wintering grounds, and back. 

As advances in technology make devices for mapping animal migration 

smaller and lighter, the chance of documenting the long migration of the 

Arctic tern crept closer. The ideal tool for this would be a satellite transmitter 

with a full year of battery capacity. However, at present, satellite transmitters 

are still too heavy for birds with a mass under 250 gram to carry on 

migration. Instead, the less accurate, but much lighter geolocators (miniature 

archival light loggers), have proven to be an effective tool (Richards et 

al. 2004). By recording and storing ambient light intensity, the geolocators 

reveal information on sunrise and sunset. When these data are combined 

with time recordings, two daily geographical positions can be calculated 

and migration routes can be mapped. The disadvantage of geolocators is 

that they cannot transmit data, and the tagged bird has to be caught again 

at the end of the study period – a logistical challenge that normally requires 

that the bird nest at the same location two years in a row. 

Although often quoted in both scientifi c and popular literature as the 

longest migration performed by any bird in the world, a study (Shaffer et 

al. 2006) questioned the Arctic tern’s status as the longest migrant. Using 

geolocators, Shaffer et al. (2006) were able to document that sooty shearwaters 

(Puffi nus griseus) make an annual round trip from New Zealand to 

the waters in the northern Pacifi c Ocean and back, of 64,037 (± 9,779) km 

on average.

Figure 7. Simplifi ed fi gure showing 

migration patterns of the 

Arctic tern, along with approximate 

position of the 

birds each month, from 

the breeding sites in 

Greenland and Iceland 

to the winter grounds 

at Antarctica. After 

initiating the southbound 

migration 

(yellow line) the 

birds paused their 

migration in the 

central part of 

the North Atlantic 

(small circle) for 

almost a month 

before they continue 

towards the 

wintering sites at 

Antarctica (large 

circle). In spring, the 

northbound migration 

(white line) is conducted 

more than twice as fast 

in a gigantic “S” shaped 

pattern through the Atlantic 

Ocean. Areas particularly 

rich in biological productivity 

are indicated by yellow 

and green colours. 

At Sand Island, we conducted a study (Manus I) which recorded a full 

year of migration in Arctic terns. 

Prior to deploying geolocators on Arctic terns, however, a small-scale 

dummy test (Egevang 2006) was conducted on terns breeding at Kitsissunnguit 

in 2006. In order to observe how the birds would respond to objects 

attached to the leg, eleven arctic terns were equipped with dummies 

with the approximate mass and physical dimensions as loggers scheduled 

to be manufactured in 2007. Although some negative behavioural 

response (“leg-lifting”) to the loggers were observed the days following 

the deployment, the overall conclusion was that very little or no effect 

could be detected in the chick rearing performed by the tagged adult bird. 

On the basis of these fi ndings, it was decided to go ahead with a full scale 

study deploying geolocators on Arctic terns at Sand Island in 2007. 

The study on Arctic tern migration revealed the longest migration ever 

recorded in any animal, with a total annual average distance of 70,900 km 

(range 59,500-81,600 km) travelled. The study also identifi ed a previously 

unknown stop-over site in the central part of the North Atlantic Ocean 

(Fig. 5), where the birds spent on average 24.6 (±6 days) before continuing 

south. This at-sea hot-spot is located between cold and warm water 

masses with high eddy variability. Satellite imagery analysis of Chlorophyll 

a. distribution revealed a high biological productivity in this particularly 

area whereas areas to the south were generally of low productiv- 

ity. It seems likely that Arctic terns use this 

area to forage and “fuel up” 

1 June to 

before conducting 

1 August 

their 

1 September 

1 November 

1 December to 1 April 

1 May 

1 November 

27

28 

migration into tropical marine regions with low productivity and limited 

food availability. It is likely that this Arctic tern stop-over site may serve 

as a hot-spot area for other migrating seabirds, e.g. long-tailed skua (Stercorarius 

longicaudus) has been observed in high densities in this area during 

ship-based surveys (D. Boertmann pers. com). 

The migration study also revealed a divide in the southbound migration 

of the Arctic tern. Just south (~10° N) of the Cape Verde Island, seven individuals 

migrated south along the coast of West Africa (as expected from 

ringing recoveries), while four individuals crossed the Atlantic Ocean 

to migrate south along the coast of South America. Furthermore, three 

individuals drifted east along the southern Polar Front and entered the 

Indian Ocean. Although divides in migratory routes are known in other 

avian species breeding at a single location, for example ospreys (Pandion 

haliaetus) breeding in Sweden (Hake et al. 2001), the study on Arctic tern 

migration illustrates goal orientation at a global scale. Although the terns 

travelled through vastly different areas, they all wintered in a relatively 

restricted geographical area in the Weddell Sea. 

The Arctic tern migration study also highlighted differences in the pace 

of southbound versus northbound migrations. While the migration south 

to the winter quarters was conducted over a period of three months (average 

93 days and an average speed of 330 km per day), the northbound 

migration back to the breeding sites was much faster. The terns covered 

the 25,700 km from the winter site north to 60º N in an average of 40 days. 

This long leg of migration was conducted with average daily distances of 

520 km – with some individuals fl ying up to 670 km per day. Unequal migration 

speeds of autumn versus spring are known from numerous bird 

species (Newton 2008) but, in this study, we were able to correlate the fast 

northbound migration with the prevailing global wind systems. 

Most studies on bird migration are performed on land birds and much of 

our understanding of the mechanism behind migration is based on land 

bird studies. Seabirds are poorly represented in text books on migration 

(e.g. Newton 2008, Berthold 2001), even though it is amongst seabirds that 

the most spectacular migration patterns are found. Even in modern seabird 

text books seabird migration may be described “as most seabirds migrate 

exclusively at sea, they have an opportunity to rest or feed whenever 

they feel like it: not an option for land birds crossing water or deserts” 

(Gaston 2004). Our study show that the Arctic tern migration in many 

ways resembles that of land birds with distinct wintering areas and stopover 

sites along the migration route, and a high level of synchrony in phenology 

amongst individuals.

2 Conclusion 

In conclusion, the studies performed at Kitsissunnguit and Sand Island 

elevated the level of knowledge on the Arctic tern in Greenland considerably. 

From the series of population size estimates in Disko Bay we learned 

that the breeding population at the largest colony in Greenland, Kitsissunnguit, 

had not crashed as suggested, but instead numbers were stable 

over a fi ve year period at 18,500 + 3,500 pairs. Although considerably 

lower than population size estimates from fi ve or six decades ago, it seems 

reasonable to conclude that the decline in the Arctic tern population at 

Kitsissunnguit happened prior to 1980, which may also be the case for the 

other tern colonies in Greenland. The example from Kitsissunnguit highlights 

the need for a systematic monitoring plan for Arctic terns, if natural 

resource managers in Greenland are to address the status of birds like the 

Arctic tern. However, the fl uctuating breeding of the species requires a 

specially designed monitoring where many colonies are monitored within 

the same breeding season, or alternatively that large colonies, like Kitsissunnguit 

are monitored on a regular basis, preferably in consecutive 

years. The decline in the Greenland Arctic tern population has been linked 

to an egg harvest of a magnitude beyond what was believed to be a sustainable 

level. The output of the harvest model (Manus II) indicate that 

this is very likely to have been the case – egg harvesting beyond mid-June 

is likely to affect the population severely and should be avoided in future 

management. 

The studies included in this thesis also produced the fi rst estimates of the 

key prey species of the Arctic tern during the breeding season. In the central 

parts of West Greenland the capelin proved to be of very high importance. 

Over the last decade the idea of initiating an industrial fi shery for 

capelin in Greenland has been proposed on several occasions. The identifi 

cation of the diet of the Arctic tern is simply confi rmation of yet another 

Arctic bird, fi sh, or mammal species that depends heavily on capelin, 

making the consequences of opening up a large scale commercial fi shery 

uncertain but nonetheless alarming. 

The study at both Kitsissunnguit and Sand Island verifi ed the strong connection 

between Arctic terns and a number of other avian species that 

benefi t from the protection that terns offer from intruders within the perimeters 

of the colony. In identifying important natural sites with a high 

diversity of breeding bird species, one could consider breeding sites of 

Arctic terns. 

29

30 

3 Future research 

The study on the long-distance migration of the Arctic tern (Manus I) revealed 

a number of new facts not previously known about the impressive 

annual migration and verifi ed existing knowledge obtained from ringing 

recoveries. This study also raises some new questions that would be interesting 

to pursue in future studies: 1) at the breeding areas, the Arctic 

tern is known to be on the wing most of the time. When it rests, it sits 

on the ground or alternatively on an iceberg. Unlike most other seabirds, 

the Arctic tern rarely rests on the water and only in very calm weather. 

The geolocators used in the migration study also record dry or wet condition 

during the period of deployment, which can be used to indicate 

the individual’s behaviour during migration (Guilford et al. 2009). From 

a both physiological and energetic perspective it could be interesting to 

investigate whether Arctic terns manage the long stretches of migration 

between the polar areas without long periods of resting on the water; 2) 

The migration study identifi ed a divide into two different southbound 

routes. This divide could be coincidental, where wind systems drive the 

birds to take a western or eastern route. This would mean that the same 

individual may migrate south by either of the two different routes each 

year. An alternative explanation would be that the route may be genetically 

coded or learned in the individual and the same bird will choose the 

same route each year. The geolocators have the capacity to log for two 

years and are cable of storing data for up to ten years. In theory, loggers 

applied in 2007 and not retrieved in 2008 (potentially up to 40 loggers) 

should contain data on two round trips which could help shed light on 

which of the two hypotheses is most plausible. An attempt to retrieve loggers 

at Sand Island was made in 2009, but was unsuccessful due to a year 

of breeding failure. Hopefully, an attempt in 2010 will prove to have more 

success; 3) The migration study identifi ed an area in the central part of the 

Atlantic Ocean to be of high importance for the terns on their southbound 

migration. It seems likely that this area may be important for other surface 

feeding seabirds as an at-sea foraging hot-spot, and it could be interesting 

to expand the number of study species in a migration study using geolocators. 

Such a study could prove to be an effi cient tool to identify at-sea 

areas of high biodiversity otherwise diffi cult or cost-prohibitive to obtain. 

Although the study on relaying in Arctic terns (Manus III) did produce 

estimates of the species’ ability to produce a replacement clutch when the 

fi rst clutch was removed, additional detailed information is needed in order 

to address the impact of harvesting and relaying at the population 

level. However, large logistical diffi culties are associated with a study of 

this kind. Relaying Arctic terns may easily move within or even well out 

of the colony to place their new nest. The chances of detection of the new 

nest site, and establishing an accurate estimate of what proportion of the 

population will relay, decreases with distance from the original nest site. 

Unless devices, such as radio transmitters, that are able to track the location 

of the bird in a study are used, it will be diffi cult to obtain accurate 

estimates of relaying in Arctic terns.

4 References 

Alerstam, T. (1985). “Strategies of migratory fl ight, illustrated by arctic and common 

terns, Sterna paradisaea and Sterna hirundo”. Contrib. Mar. Sci. (Suppl.) (27): 580–603. 

Alerstam, T., C. Hjort, G. Högsted, P. E. Jönsson, J. Karlsson and B. Larsson (1986). 

“Spring migration of birds across the Greenland Inlandice”. Meddelser om Grønland, 

Bioscience 21: 38. 

Bertelsen, A. (1921). “Fuglene i Umánaq distrikt”. Meddelser om Grønland 62(2): 139-214. 

Berthold, P. (2001). Bird migration: a general survey Bird migration: a general survey. 

Oxford University Press, ed. 2. 

Bianki, V. V. (1967). “Waders, gulls and alcids of Kandalaksha Bay (In Russian)”. Kandalaksha 

State Reserve 6, 364 pages. 

Bianki, V. V. and K. Isaksen (2000). Arctic tern Sterna paradisaea. The status of marine 

birds breeding in the Barents Sea region. A.-N. T., B. V., S. H.et al. Tromsø, Norsk 

Polarinstitutt Rapportserie. 113: 131-136. 

Boecker, M. (1967). “Vergleichende Untersuchungen zur Nahrungsund Nistökologie 

der Flugseeschwalbe und der Kiistenseeschwalbe”. Bonn Zool. Beitr. 18: 15-126. 

Boertmann, D. (1994). “An annotated checklist to the birds of Greenland”. Meddelelser 

om Grønland, Bioscience 38: 1-63. 

Boertmann, D. (2007). ”Grønlands Rødliste”. Danmarks Miljøundersøgelser, Afdeling 

for Arktisk Miljø, Aarhus Universitet. 

Boertmann, D., A. Mosbech, K. Falk and K. Kampp (1996). Seabird colonies in western 

Greenland (60º - 79º 30’ N lat.). Roskilde, National Environmental Research 

Institute, Denmark: 1-148. 

Bourne, W. R. P. and M. B. Casement (1996). “The migrations of the Arctic Tern”. Bull. 

Br. Ornithol. Club 116: 117–123. 

Brindley E., G., N. Mudge, C. Dymond, B. Lodge, D. Ribbands, P. Steele, E. Ellis, D. 

Meek and S. a. N. Ratcliffe. ( 1999). “The status of Arctic Terns Sterna paradisaea at 

Shetland and Orkney in 1994”. Atlantic Seabirds 1: 135-143. 

Burnham, W., K. K. Burnham and T. J. Cade (2005). “Past and present assessments of 

bird life in Uummannaq District, West Greenland”. Dansk Ornitologisk Forenings 

Tidsskrift 99: 196-208. 

Carscadden, J. E., W. A. Montevecchi, G. K. Davoren and B. S. Nakashima (2002). 

“Trophic relationships along capelin (Mallotus villosus) and seabirds in a changing 

ecosystem”. ICES Journal of Marine Science 59: 1027-1033. 

Cramp, S., Ed. (1985). “Terns to woodpeckers”. Birds of the Western Palearctic. Oxford, 

Oxford University Press. 

Devlin, C. M., A. W. Diamond, S. W. Kress, C. S. Hall and L. Welch (2008). Breeding 

Dispersal and Survival of Arctic Terns (Sterna paradisaea) Nesting in the Gulf of 

Maine. 125: 850-858. 

Egevang, C. 2006. Geo-locator dummy study on Arctic terns 2006. http://issuu.com/ 

egevang/docs/geo-locator_dummy_study_06_1. 

Egevang, C. and D. Boertmann (2003). Havternen i Grønland - status og undersøgelser 

2002. Roskilde, National Environmental research Institute: 72. 

Egevang, C. and D. Boertmann (2008). “Ross’s Gulls (Rhodostethia rosea) Breeding in 

Greenland: 

A Review, with Special Emphasis on Records from 1979 to 2007”. Arctic 61(3): 322-328. 

Egevang, C., D. Boertmann and O. S. Kristensen (2005). Monitering af havternebestanden 

på Kitsissunnguit (Grønne Ejland) og den sydlige del af Disko Bugt, 

2002-2004. Nuuk, Grønlands Naturinstitut: 41. 

31

32 

Egevang, C., K. Kampp and D. Boertmann (2004). “The Breeding Association of Red 

Phalaropes (Phalaropus fulicarius) with Arctic Terns (Sterna paradisaea): Response to 

a Redistribution of Terns in a Major Greenland Colony”. Waterbirds 27(4): 406-410. 

Egevang, C., K. Kampp and D. Boertmann (2006). ”Declines in breeding waterbirds 

following a redistribution of Arctic Terns Sterna paradisaea in West Greenland”. 

Waterbirds around the World. C. A. G. D. A. S. G. C. Boere. The stationery Offi ce, 

Edinburgh, UK: 154. 

Egevang, C. and I. J. Stenhouse (2007). “Field report from Sand Island, Northeast 

Greenland, 2007”. Greenland Institute of Natural Resources. 

Egevang, C., I. J. Stenhouse, R. A. Phillips, A. Petersen, J. W. Fox and J. R. D. Silk 

(2010). “Tracking of Arctic terns Sterna paradisaea reveals longest animal migration”. 

Proceedings of National Academy of Science of the United States of America, 

107: 2078-2081. 

Egevang, C., I. J. Stenhouse, L. M. Rasmussen, M. Willemoes and F. Ugarte (2008). 

“Field report from Sand Island, Northeast Greenland, 2008”. Greenland Institute 

of Natural Resources. 

Fægteborg, M. (2000). ”Grønland idag – en introduktion år 2000”. Copenhagen, Arctic 

Information. 

Frich, A. S. (1997). ”Fuglelivet og dets udnyttelse på Grønne Ejland i Vestgrønland, 

juni 1996. Nuuk, Greenland”. Pinngortitaleriffi k, Greenland Institute of Natural 

Resources, Technical report 1, 19 pages. 

Friis-Rødel, E. and P. Kanneworff (2002). “A review of Capelin (Mallotus villosus) in 

Greenland waters”. ICES Journal of Marine Science 59: 890-896. 

Gaston, A. j. (2004). Seabirds: A Natural History. New Haven, Connecticut, Yale University 

Press. 

GSCD (2009). The Greenland Seabird Colony Database. http://www.dmu.dk/Greenland/Olie+og+Miljoe/Havfuglekolonier/ 

accessed October 2009. 

Gudmundsson, G., T. Alerstam and B. Larsson (1992). “Radar observations of northbound 

migration of the Arctic tern, Sterna paradisaea, at the Antarctic Peninsula”. 

Antarctic Science 4(2): 163-170. 

Guilford, T., J. Meade, J. Willis, R. A. Phillips, D. Boyle, S. Roberts, M. Collett, R. Freeman 

and C. M. Perrins (2009). “Migration and stopover in a small pelagic seabird, 

the Manx shearwater Puffi nus puffi nus: insights from machine learning”. Proceedings 

of The Royal Society, Biological Science 276: 1215-1223. 

Hake, M., N. Kjellén and T. Alerstam (2001). “Satellite tracking of Swedish Ospreys 

Pandion haliaetus: autumn migration routes and orientation”. Journal of Avian 

Biology 32: 47–56. 

Hatch, J. (2002). “Arctic Tern Sterna paradisaea”. The Birds of North America. 707: 1-40. 

Hjarsen, T. (2000). Diskobugten - Besigtigelses- og informationstur 20 - 26 juni 2000. 

Nuuk, Grønlands Hjemmestyre - Direktoratet for Miljø og Natur. 

Kampp, K. (1980). “Ornitologiske undersøgelser i Disko Bugt 1980”. Unpublished 

fi eld report. 

Kampp, K. (2001). “Seabird observations from the south and central Atlantic Ocean, 

Antarctica to 30°N, March–April 1998 and 2000”. Atlantic Seabirds 3: 1–14. 

Kilpi, M., K. Lindström and Wuorinen (1992). “A change in clutch size in Arctic Tern 

Sterna paradisaea in nothern Baltic”. Ornis Fennica 69: 88-91. 

Kullenberg, B. (1946). “Über verbreitung und wanderungen von vier Sterna-arten. [On 

the distribution and migrations of four Sterna species.] “ Arkiv Zoology 38: 1–80. 

Levermann, N. and A. P. Tøttrup (2007). “Predator Effect and Behavioral Patterns in 

Arctic Terns (Sterna paradisaea) and Sabine’s Gulls (Xema sabini) During a Failed 

Breeding Year”. Waterbirds 30(3): 417-420.

Lloyd, C., Tasker, M.L. and Partridge, K. (1991). The status of seabirds in Britain and 

Ireland. T. & A.D. 

Wetlands International (2006). Waterbird Population Estimates – Fourth Edition. Wetlands 

International, Wageningen, The Netherlands. 

Møller, A. P., E. Flensted-Jensen and W. Mardal (2006). “Rapidly advancing laying 

date in a seabird and the changing advantage of early reproduction”. Journal of 

Animal Ecology 75(3): 657-665. 

Monaghan, P., J. D. Uttley, M. D. Burns, C. Thaine and J. Blackwood (1989). “The relationship 

between food supply, reproductive effort and breeding success in Arctic 

Terns Sterna paradisaea”. Journal of Animal Ecology 58: 261-274. 

Monaghan, P., J. D. Uttley and J. D. Okill (1989). “Terns and sandells: seabirds as indicators 

of changes in marine fi sh populations”. Journal of Fish Biology 35 (supplement 

A): 339-340. 

Newton, I. (2008). The migration ecology of birds. London, Burlington, San Diego, 

Academic Press, Elsevier. 

Nguyen, L. P., K. F. Abraham and E. Nol (2006). “Infl uence of Arctic Terns on Survival 

of Artifi cial and Natural Semipalmated Plover Nests”. Waterbirds 20(1): 100-104. 

Pearson, T. H. (1968). “The Feeding Biology of Seabird species Breeding on the Farne 

Islands, Northumberland”. Journal of Animal Ecology 37: 521-552. 

Phillips, R. A., J. R. D. Silk, J. P. Croxall, V. Afanasyev and D. R. Briggs (2004). “Accuracy 

of geolocation estimates for fl ying seabirds”. Mar Ecol Prog Ser (266): 265–272. 

Ratcliffe, N. (2004). “Arctic Tern Sterna paradisaea”. Seabird Populations of Britain and 

Ireland. P. I. Mitchell, S. F. Newton, N. Ratcliffe and T. E. Dunn. London, T & A D 

Poyser: 328-338. 

Salomonsen, F. (1950). Grønlands Fugle. The Birds of Greenland. Copenhagen, 

Munksgaard. 

Salomonsen, F. (1967). “Migratory Movements of the Arctic Tern (Sterna paradisaea 

Pontoppidan) in the Southern Ocean”. Det Kongelige Danske Videnskabernes 

Selskab 24(1): 1-42. 

Shaffer, S. A., Y. Tremblay, H. Weimerskirch, D. Scott, D. R. Thompson, P. M. Sagar, 

H. Moller, G. A. Taylor, D. G. Foley, B. A. Block and D. P. Costa (2006). “Migratory 

Shearwaters Integrate Oceanic Resources across the Pacifi c Ocean in an Endless 

Summer”. Proceedings of the National Academy of Sciences of the United States 

of America 103(34): 12799-12802. 

Smith, P. A., G. H. Gilchrist and J. N. M. Smith (2007). “Effect of nest habitat, food and 

parental behaviour on shorebird nest success”. The Condor 109(1): 15-31. 

Storr, G. M. (1958). “Migration routes of the Arctic Tern”. Emu 1: 59–62. 

Suddaby, D. and N. Ratcliffe (1997). “The Effects of Fluctuating Food Availability on 

Breeding Arctic Terns (Sterna paradisaea)”. Auk 114(3): 524-530. 

Summers, R. W. and M. Nicoll (2004). “Geographical variation in the breeding biology 

of the Purple Sandpiper Calidris maritima”. IBIS 146: 303–313. 

33

Manus I 

TRACKING OF ARCTIC TERNS 

STERNA PARADISAEA REVEALS 

LONGEST ANIMAL MIGRATION 

Carsten Egevang a,b , Iain J. Stenhouse c , Richard A. Phillips d , 

Aevar Petersen e , James W. Fox d and Janet R. D. Silk d 

a Greenland Institute of Natural Resources 

DK-3900 Nuuk Greenland 

b National Environmental Research Institute 

Department of Arctic Environment, Aarhus University 

DK-4000 Roskilde, Denmark 

c Poland, ME 04274 

d British Antarctic Survey 

Natural Environment Research Council, High Cross 

Cambridge CB3 0ET, United Kingdom 

e Icelandic Institute of Natural History 

125 Reykjavik, Iceland 

Proceedings of the National Academy of Sciences of the United States. Vol. 107: 2078-2081

Tracking of Arctic terns Sterna paradisaea reveals 

longest animal migration 

Carsten Egevang a,b,1 , Iain J. Stenhouse c , Richard A. Phillips d , Aevar Petersen e , James W. Fox d , and Janet R. D. Silk d 

a Greenland Institute of Natural Resources, DK-3900 Nuuk Greenland; b National Environmental Research Institute, Department of Arctic Environment, Aarhus 

University, DK-4000 Roskilde, Denmark; c Poland, ME 04274; d British Antarctic Survey, Natural Environment Research Council, High Cross, Cambridge CB3 0ET, 

United Kingdom; and e Icelandic Institute of Natural History, 125 Reykjavik, Iceland 

Edited by Colleen Cassady St. Clair, University of Alberta, Edmonton, AB, Canada, and accepted by the Editorial Board December 11, 2009 (received for review 

August 20, 2009) 

The study of long-distance migration provides insights into the 

habits and performance of organisms at the limit of their physical 

abilities. The Arctic tern Sterna paradisaea is the epitome of such 

behavior; despite its small size (400 g) (3–7), neglecting smaller species 

which, arguably, have even greater migratory capabilities and are 

potentially more sensitive indicators of the status of marine 

ecosystems (8, 9). These include a species long regarded as the 

classic exponent of long-distance migration in vertebrates, the 

Arctic tern (Sterna paradisaea). 

The Arctic tern exhibits a circumpolar breeding distribution at 

high latitudes of the northern hemisphere and winters at high 

latitudes in the southern hemisphere (10, 11). The species nests on 

the ground, and is an opportunistic plunge-diving and surfacedipping 

feeder, with a diet comprising mainly small fish and large 

zooplankton (12, 13). There were several early attempts to infer its 

movements (14, 15), but the first comprehensive review (16) was 

presented by Salomonsen in 1967. He suggested that the main 

wintering areas were most likely to be off the southern tip of Africa 

and South America, and that Arctic terns might circumnavigate 

the Antarctic continent in early spring before initiating their 

northbound migration. This appeared to be supported by subsequent 

radar and visual observations of terns migrating from the 

Bellingshausen Sea to the Weddell Sea (17). It has been suggested 

that Arctic terns complete the southbound migration in a step-wise 

fashion, moving between rich feeding grounds, and may use high 

altitude migration flights to cross large expanses of land, such as 

the Greenland ice sheet (18, 19). Until now, there was little 

information on the northward return to the colony, but it was 

36 

generally believed to start in March and occur more rapidly and 

over a wider front (20, 21). Several authors (10, 11) have attempted 

to estimate the total distance traveled during this impressive 

annual migration, typically quoting a figure of 40,000 km. 

Although many (but not all) of these suppositions have subsequently 

proved correct, they were based on limited banding 

recoveries and at-sea observations, and hence inevitably reflected a 

range of potential biases (including, respectively, spatial and temporal 

variability in recovery effort, and lack of information on 

provenance, age, and status of individuals). Nor could they provide 

information on variation among individuals and populations in the 

phenology of migration, flight paths, or preferred wintering habitats. 

With the aim of providing much greater detail on individual movement 

patterns, we developed miniature (1.4-g) archival light loggers 

(geolocators) similar to those used in a recent study of passerines 

(22), for deployment on Arctic terns at two Arctic colonies. 

Results and Discussion 

With an accuracy of ∼185 km in flying seabirds (23), the geolocators 

were used to document the migration routes, stopover 

sites, and wintering areas of Arctic terns from colonies in high- 

Arctic Greenland (n = 10) and Arctic Iceland (n = 1). At the end 

of the breeding season, tagged birds traveled southwest to a 

stopover region of deep water in the eastern portion of the 

Newfoundland Basin and the western slope of the mid-North 

Atlantic Ridge between 41–53° N and 27–41° W, in which they 

remained for an average ± SD of 24.6 ± 6 days (Fig. 1). This 

previously unknown oceanic hotspot for terns was located at the 

junction between cold, highly productive northern water and 

warmer, less-productive southern water, and was characterized by 

high Chl. a concentrations and high eddy variability (24, 25). 

Between 5 and 22 September 2007, all 11 birds continued their 

migration southeast toward the West African coast. South of the 

Cape Verde Islands (∼10° N), however, migration routes diverged: 

seven birds continued to fly south parallel to the African coast, 

whereas four others crossed the Atlantic to follow the east coast of 

Brazil. Birds in both groups ceased their directed southbound 

transits at ∼38–40° S, and shifted to a pattern of predominantly 

east–west movements. Three birds then traveled east into the 

Indian Ocean (one as far as 106° E). All birds subsequently moved 

south, spending the austral summer (December–March) south of 

58° S and between 0 and 61° W in the Atlantic sector of the 

Southern Ocean. This region, which includes the Weddell Sea, is 

particularly productive, and supports higher densities of a key prey 

Author contributions: C.E., R.A.P., A.P., and J.W.F. designed research; C.E., I.J.S., and A.P. 

performed research; C.E., I.J.S., R.A.P., and J.R.D.S. analyzed data; and C.E., I.J.S., R.A.P., 

A.P., J.W.F., and J.R.D.S. wrote the paper. 

The authors declare no conflict of interest. 

This article is a PNAS Direct Submission. C.C.S. is a guest editor invited by the 

Editorial Board. 

Freely available online through the PNAS open access option. 

1 

To whom correspondence should be addressed. E-mail: cep@dmu.dk. 

www.pnas.org/cgi/doi/10.1073/pnas.0909493107 PNAS Early Edition | 1of4 

ECOLOGY

Fig. 1. Interpolated geolocation tracks of 11 Arctic terns tracked from breeding colonies in Greenland (n = 10 birds) and Iceland (n = 1 bird). Green = autumn 

(postbreeding) migration (August–November), red = winter range (December–March), and yellow = spring (return) migration (April–May). Two southbound 

migration routes were adopted in the South Atlantic, either (A) West African coast (n = 7 birds) or (B) Brazilian coast. Dotted lines link locations during the 

equinoxes. 

for many seabirds, Antarctic krill (Euphausia superba), than elsewhere 

in the Southern Ocean (26). 

All birds began the return migration to breeding colonies in 

early–mid April, always traveling over deep water at considerable 

distance from continental shelf margins, and taking a sigmoidal 

route, counterclockwise around the South Atlantic and clockwise 

around the North Atlantic gyres. Upon return in late May to the 

same general stopover region in the North Atlantic used during 

the southbound migration, two individuals transited rapidly west 

through the area, three remained there for 3–6 days before 

continuing to the north, and the remaining five birds appeared 

not to pause or deviate from their general course. Thus, there 

was little indication that the area was particularly important for 

northbound terns. Although the Arctic tern from Iceland began 

its outbound and return migration earlier than the Greenlandbreeding 

birds (Table 1), there was no obvious difference in 

migration routes between this individual and some of the terns 

from Greenland. Thus, the emerging pattern for transequatorial 

migrants is of a high degree of mixing of birds from different 

breeding populations in wintering areas, as evident in this study 

and previous ones (5, 7, 27). This contrasts with the dominant 

strategies of southern hemisphere albatrosses, which, despite a 

similar capacity for long-distance migration and hence overlap, 

nevertheless tend to show a high degree of between-population 

segregation in nonbreeding distributions (28). 

The Greenlandic birds exhibited clear synchrony in timing of 

migration, all reaching the North Atlantic stopover site, 

departing the wintering area and crossing the Equator within a 

few days of each other (Table 1), but there was no indication that 

they traveled together in the same flocks. Similarly, at-sea 

observations suggest that flock sizes of migrating terns are typically 

very small (

migration from Antarctica to south Greenland (c. 60° N), an 

average distance of 24,270 km (range 20,070–27,790 km) covered 

in only 40 days (range 36–46 days), with average travel distances 

of 520 km·day −1 (range 390–670 km·day −1 ). As birds seem to 

have been exploiting favorable winds, travel (ground) speed will 

be higher than air speed. Nevertheless these flight speeds are of 

the same order as the maximum range speed (9.7 ms −1 ) calculated 

from aerodynamics theory, and mean groundspeeds 

observed using radar (11.3 ms −1 ) (17), which would correspond 

to 838 and 976 km·day −1 , respectively, if birds were to fly nonstop 

(which is, of course, unlikely). Overall, the northbound migration 

took less than half the time (40 vs. 93 days), despite being three 

quarters the length (25,700 vs. 34,600 km) of the southbound 

journey. Indeed, the average annual distance traveled, from 

departing the breeding site in August to return in late May/early 

June (i.e., excluding movements within the breeding season) was 

70,900 km (range 59,500–81,600 km). This is the longest roundtrip 

animal migration ever recorded electronically (3, 5, 7). As 

our estimates of travel distance, and hence speed, were affected 

by the fractal dimension of the flight path (i.e., the step length for 

the distance measurements) (30), and by the considerable errors 

inherent in geolocation (23), they are not directly comparable 

with previous studies. Nevertheless, the tracked birds were considered 

to travel nearly twice the total distance generally cited 

for the annual Arctic tern migration. As Arctic terns can live for 

more than 30 years (31), the total distance traveled in a lifetime 

may exceed 2.4 million km, equivalent to approximately three 

return journeys to the Moon. 

Our study demonstrated heterogeneity in migration routes within 

a population, yet clear migratory connectivity between populations. 

We confirmed that the main wintering region was the marginal ice 

zone around Antarctica, which agrees with at-sea observations (ref. 

18 and references therein). Given that the highest densities of 

Arctic terns in the Southern Ocean have been observed in the 

Weddell Sea (17), and that a very high (likely >50%) proportion of 

the worlds Arctic terns breed in Greenland and Iceland (13), the 

wintering areas used by our tracked birds may well be typical of the 

majority of the North Atlantic population. Thus, the long-term 

changes in duration and extent of winter sea ice, and declines in 

abundance of Antarctic krill in this region (26, 32) should be viewed 

with considerable concern. 

Material and Methods 

We fitted miniature archival light loggers (Mk14 geolocators, mass 1.4 g; 

British Antarctic Survey), attached to plastic leg rings (mass of logger, ring, 

tape, and cable tie: 2.0 g, ∼1.9% of adult body mass), to 50 breeding Arctic 

terns in July 2007 at Sand Island (74° 43′ N, 20° 27′ W), Young Sound, 

Northeast Greenland. An additional 20 geolocators were deployed on Arctic 

terns at Flatey Island (65° 22′N; 22° 55′W), Breiðafjörður, Iceland, in June 

2007. The following season, a minimum of 21 birds equipped with loggers 

(>42%) were observed at Sand Island, but only 10 loggers (20%) were 

retrieved. At Flatey Island, four birds were resighted (20%), and one (5%) 

logger was retrieved. All recaptured birds were in good physical condition 

1. Block BA, et al. (2005) Electronic tagging and population structure of Atlantic bluefin 

tuna. Nature 434:1121–1127. 

38 

Table 2. Summary of distances traveled during migration stages of Arctic terns (Greenland and 

Iceland birds combined, n = 11) 

Migration segment Distance traveled 

Total distance traveled on migration 70,900 km (59,500–81,600 km) 

Distance traveled on southbound migration 34,600 km (28,800–38,500 km) 

Distance traveled per day on southbound migration 330 km·day −1 (280–390 km·day −1 ) 

Distance traveled on northbound migration 25,700 km (21,400–34,900 km) 

Distance traveled per day on northbound migration 520 km·day −1 (390–670 km·day −1 ) 

Distance traveled within winter site 10,900 km (2,700–21,600 km) 

and no significant difference (t = −1.57, P = 0.133, df = 18, n = 10) could be 

detected in their body mass between the 2 years (mean ± SD in 2007 and 

2008: 106.0 g ± 6.3 and 110.3 ± 6.0, respectively). The average clutch size of 

breeding pairs in 2008 at Sand Island equipped with loggers was 1.3 (± 0.43, 

n = 8), whereas clutch size in control birds the same year was 1.7 (±0.48, n = 

60). Although poor in power of analysis because of low sample size in the 

logger group, a one-sided Fisher’s exact test indicated no statistical difference 

(P = 0.038) between the two groups. Not all birds observed at colonies 

in 2008 could be recaptured; some birds were clearly breeding, but the nest 

could not be located or the targeted bird would not enter the trap; others 

were likely to be nonbreeders or failed breeders, which are very difficult to 

capture. In one case (Iceland), the nest of an equipped bird was abandoned 

after being trampled by domestic sheep before the bird could be trapped. 

Although fidelity at the regional level is high in nesting Arctic terns (33), 

dispersal to neighboring breeding colonies occurs frequently (33–36) and is 

likely driven by food availability, presence of predators, disturbance, or 

changing climatic conditions. The only study to apply modern capture– 

mark–recapture techniques to adult Arctic terns showed a low reencounter 

probability, ranging from 0.12 to 0.74, depending on colony and year (33). 

Ten of the 11 loggers retrieved were downloaded successfully, providing a 

full year of migration data (July 2007 to July 2008). Six months of migration 

data were extracted from the remaining logger. Light data were processed 

following the approach of Phillips et al. (23). Times of sunrise and sunset were 

calculated from light records and converted to location estimates using 

TransEdit and BirdTracker (British Antarctic Survey) using thresholds of 10, 

an angle of elevation of −4.7°, and applying the compensation for movement. 

Locations were unavailable at periods of the year when birds were at 

very high latitudes and experiencing 24 h daylight. In addition, only longitudes 

were available around equinoxes, when day length is similar 

throughout the world. Overall, after omitting periods with light level 

interference and periods around equinoxes, the filtered data sets contained 

between 166 and 242 days of locations for each individual (mean = 207.2 

days, SD = 20.58, n = 10). The data set contained two daily positions, after 

fitting points to a smooth line, travel distances were calculated in ArcMap 

(ESRI) using great circle distances (Table 2). A straight-line path was assumed 

between the preceding and succeeding valid fix around periods when 

locations were unavailable (discussed above). The only exceptions were 

during the autumn migration around the west African coast, where it was 

assumed that birds did not cross land and the projected route was altered to 

follow the coast, and for movement north and south to the breeding colonies 

in Greenland and Iceland from 60° N. As positions obtained from 

geolocators exhibit a relatively low accuracy, we used a conservative 

approach when calculating distances and speed, and hence values presented 

here should be regarded as minimum estimates. Great-circle distances were 

calculated using the mean of the two daily positions, interpolating between 

adjacent valid locations across periods when data were unavailable because 

of proximity to the equinox or light level interference. Stopover sites were 

identified following the approach of Guilford et al. (6); where latitudinal 

movements were

3. Croxall JP, Silk JRD, Phillips RA, Afanasyev V, Briggs DR (2005) Global circumnavigations: 

Tracking year-round ranges of non-breeding albatrosses. Science 307:249–250. 

4. Weimerskirch H, Wilson RP (2000) Oceanic respite for wandering albatrosses. Nature 

406:955–956. 

5. González-Solís J, Croxall JP, Oro D, Ruiz X (2007) Trans-equatorial migration and 

mixing in the wintering areas of a pelagic seabird. Front Ecol Environ 5:297–301. 

6. Guilford T, et al. (2009) Migration and stopover in a small pelagic seabird, the Manx 

shearwater Puffinus puffinus: Insights from machine learning. Proc R Soc B Biol Sci 

276:1215–1223. 

7. Shaffer SA, et al. (2006) Migratory shearwaters integrate oceanic resources across the 

Pacific Ocean in an endless summer. Proc Natl Acad Sci USA 103:12799–12802. 

8. Einoder LD (2009) A review of the use of seabirds as indicators in fisheries and 

ecosystem management. Fish Res 95:6–13. 

9. Furness RW, Tasker ML (2000) Seabird-fishery interactions: Quantifying the sensitivity 

of seabirds to reductions in sandeel abundance, and identification of key areas for 

sensitive seabirds in the North Sea. Mar Ecol Prog Ser 202:253–264. 

10. Berthold P (2001) Bird Migration: A General Survey (Oxford University Press, USA), 

2nd Ed. 

11. Newton I (2008) The Migration Ecology of Birds (Academic Press, Elsevier, London). 

12. Hatch H (2002) (Sterna paradisaea). The Birds of North America, No. 707 (Arctic Tern) 

(Philadelphia). 

13. Cramp S, ed (1985) Terns to Woodpeckers. Birds of the Western Palearctic (Oxford 

University Press), Vol IV. 

14. Kullenberg B (1946) Über verbreitung und wanderungen von vier Sterna-arten. Arkiv 

Zool. 38:1–80. 

15. Storr GM (1958) Migration routes of the Arctic tern. Emu 1:59–62. 

16. Salomonsen F (1967) Migratory movements of the arctic tern (Sterna paradisaea 

Pontoppidan) in the Southern Ocean. Biol Medd Dan Vid Selsk 24:1–42. 

17. Gudmundsson GA, Alerstam T, Larsson B (1992) Radar observations of northbound 

migration of the Arctic tern, Sterna paradisaea, at the Antarctic Peninsula. Antarct Sci 

4:163–170. 

18. Alerstam T (1985) Strategies of migratory flight, illustrated by arctic and common 

terns, Sterna paradisaea and Sterna hirundo. Contrib Mar Sci 27 (Suppl):580–603. 

19. Alerstam T, et al. (1986) Spring migration of birds across the Greenland Inlandice. 

Meddr Grønland Biosci 21:1–38. 

20. Kampp K (2001) Seabird observations from the south and central Atlantic Ocean, 

Antarctica to 30°N, March–April 1998 and 2000. Atlantic Seabirds 3:1–14. 

21. Bourne WRP, Casement MB (1996) The migrations of the Arctic tern. Bull Br Ornithol 

Club 116:117–123. 

22. Stutchbury BJM, et al. (2009) Tracking long-distance songbird migration by using 

geolocators. Science 323:896. 

23. Phillips RA, Silk JRD, Croxall JP, Afanasyev V, Briggs DR (2004) Accuracy of geolocation 

estimates for flying seabirds. Mar Ecol Prog Ser 266:265–272. 

24. National Aeronautics and Space Administration. NASA Earth Observations. Available 

at http://neo.sci.gsfc.nasa.gov, accessed February 2009. 

25. National Aeronautics and Space Administration. MODIS satellite data, NASA SeaWiFS 

Project. Available at http://oceancolor.gsfc.nasa.gov/, accessed February 2009. 

26. Atkinson A, Siegel V, Pakhomov E, Rothery P (2004) Long-term decline in krill stock 

and increase in salps within the Southern Ocean. Nature 432:100–103. 

27. Felicísimo ÁM, Muñoz J, González-Solis J (2008) Ocean surface winds drive dynamics 

of transoceanic aerial movements. PLoS One 3:e2928. 

28. Phillips RA, Croxall JP, Silk JRD, Briggs DR (2008) Foraging ecology of albatrosses and 

petrels from South Georgia: Two decades of insights from tracking technologies. 

Aquat Conserv Mar Freshwat Ecosyst 17:S6–S21. 

29. Elkins N (1988) Weather and Bird Behaviour (Poyser, Calton). 

30. Turchin P (1998) Quantitative Analysis of Movement: Measuring and Modeling 

Population Redistribution in Animals and Plants (Sinauer Associates, Sunderland, 

MA). 

31. Hatch JJ (1974) Longevity record for the Arctic tern. Bird-Banding 45:269–270. 

32. Rau GH, Ainley DG, Bengtson JL, Torres JJ, Hopkins TL (1992) 15 N/ 14 N and 13 C/ 12 Cin 

Weddell Sea birds, seals and fish: Implications for diet and trophic structure. Mar Ecol 

Prog Ser 84:1–8. 

33. Devlin CM, Diamond AW, Kress SW, Hall CS, Welch L (2008) Breeding dispersal and 

survival of Arctic terns (Sterna paradisaea) nesting in the Gulf of Maine. Auk 125: 

850–858. 

34. Møller AP, Flensted-Jensen E, Mardal W (2006) Dispersal and climate change: A case 

study of the Arctic tern Sterna paradisaea. Glob Change Biol 12:2005–2013. 

35. Brindley EG, at al (1999) The status of Arctic terns Sterna paradisaea at Shetland and 

Orkney in 1994. Atlantic Seabirds 1:135–143. 

36. Ratcliffe N (2004) Arctic tern Sterna paradisaea. Seabird Populations of Britain 

and Ireland, eds Mitchell PI, Newton SF, Ratcliffe N, Dunn TE (Poyser, London), pp 

328–338. 

4of4 | www.pnas.org/cgi/doi/10.1073/pnas.0909493107 Egevang et al. 

39

Manus II 

SUSTAINABLE REGULATION OF ARCTIC 

TERN EGG HARVESTING IN GREENLAND 

Carsten Egevang 1 , Norman Ratcliff e 2 and Morten Frederiksen 3 

1 Greenland Institute of Natural Resources, Postbox 570, DK-3900 Nuuk, Greenland 

E-mail: cep@dmu.dk 

2 British Antarctic Survey, High Cross, Cambridge, CB3 0ET, United Kingdom 

3 National Environmental Research Institute, Department of Arctic Environment, 

Aarhus University, Frederiksborgvej 399, DK-4000 Roskilde, Denmark 

Manuscript

42 

Sustainable regulation of Arctic tern egg 

harvesting in Greenland 

Carsten Egevang 1 , Norman Ratcliff e 2 and Morten Frederiksen 3 

1Greenland Institute of Natural Resources, Postbox 570, DK-3900 Nuuk, Greenland, 



3 National Environmental Research Institute, Department of Arctic Environment, 

Aarhus University, Frederiksborgvej 399, DK-4000 Roskilde, Denmark 

Abstract 

The Arctic tern population in West Greenland is likely to have undergone 

a severe decline since mid 20 th century. This decline has been suggested 

to be a consequence of an unsustainable egg harvest conducted by local 

Inuit. In this study we model breeding parameters obtained over a fi ve 

year period at the largest Arctic tern colony in Greenland, Kitsissunnguit 

to predict the impact of harvest on the local population. The model incorporates 

estimates of nest and chick survival to fl edging, clutch size in 

both “normal” and replacement nests to predict how harvest size and the 

impact on reproduction varies with harvest effort, duration and start date. 

Model output indicates the least impact on the Arctic tern population if 

harvest is conducted early in the season. Effort (man hours spent harvesting) 

governs the total number of eggs removed from the population 

whereas duration of the harvest only has a minor impact. A Impact- Harvest 

Ratio helped us determine a time period where egg were available 

to harvesters but where the harvest would only have minor impact as the 

birds were allowed to relay. Based on model output, we recommend that 

an Arctic tern egg harvest at Kitsissunnguit with the least impact should 

be conducted early in the breeding season preferably with 1) a stopping 

date before 10 June or 2) a harvest carried out during one day. An opening 

of egg harvest at Kitsissunnguit should be followed by monitoring 

of population size and fl edging success at both affected and unaffected 

islands in the archipelago. In the context of the model output, the former 

closing date (1 July) in the Greenland Arctic tern harvest has probably 

caused high impact on reproduction and it seems plausible that this may 

have caused a population decline. 

Key words: Seabirds, Arctic Tern, Sterna paradisaea, Arctic, Greenland, egg 

harvest, model, sustainable use, regulation.

Introduction 

Harvesting of seabird eggs, chicks and adults is widespread across most 

parts of the world. Seabirds are harvested for subsistence, income and 

recreation, which can cause additive mortality and population declines 

(Moller 2006, Nettleship et al. 1994). Seabird life histories are characterized 

by high adult survival, low annual reproductive rates and delayed maturation 

(Croxall et al. 1984, Furness and Monaghan 1987). This makes seabirds 

as a group particularly vulnerable to overexploitation. Furthermore, 

seabirds often aggregate at high densities in predictable colony locations 

which make harvesting of a large proportion of the population easier than 

would be the case in dispersed or mobile species (Coulson 2002). The remoteness 

of many seabird colonies also means that monitoring, and enforcement 

of regulations, is diffi cult to conduct. 

In the past, the collection of seabird products, such as eggs, chicks and 

adults for food and feathers for insulation was fundamental to the survival 

of indigenous human communities. Today, reliance on seabirds as a resource 

has lessened, yet harvesting is still valued by indigenous people as 

part of their cultural heritage and ethnic identity. This has led to confl icts 

between conservationists and indigenous people, but their aspirations are 

not mutually exclusive provided harvests are managed in a sustainable 

manner. Achieving this can be a major challenge owing to compensatory 

demographic mechanisms that can buffer populations against harvesting 

mortality. Designing sustainable harvesting regulations therefore requires 

estimation of key demographic parameters and simulation modelling of 

various management options. 

In modern Greenland, seabird harvesting is still important with approximately 

150,000 birds, mainly Brünnich’s guillemots (Uria lomvia) and common 

eider (Somateria mollissima) shot annually during winter (Merkel and 

Barry 2008). Before new harvest regulations were introduced in 2002, the 

seabird harvest level caused concern and strong criticism from conservationists 

and non-government environmental organisations (Hansen 2002). 

The Arctic tern (Sterna paradisaea) is a wide-spread breeder in Greenland 

with a patchy distribution. It is usually found breeding on skerries, inlets 

and islands in colonies varying from few pairs to tens of thousands pairs 

(Boertmann 2004). Harvesting of tern eggs has probably taken place as 

long as humans have inhabited the country, and is even today considered 

an important activity during summer. Although historical censuses 

of Greenland Arctic tern colonies are scarce, there is evidence that the 

population has suffered a decline since mid 20 th century (Burnham et al. 

2005, Egevang et al. 2004, Salomonsen 1950). This decline has been associated 

with an increase in egg harvesting, as both human population size 

and mobility through better boats with larger engines, have increased in 

the same period (Hansen 2002). The concern for the Greenland Arctic tern 

population led to a ban on Arctic tern egg harvesting from 2002 and onwards. 

Before 2002, egg harvesting was allowed with a closing date at July 

1 st .The new regulations have led to protests from local people who view 

this as an infringement of their cultural heritage, and so there is pressure 

for the ban to be lifted. Conservationists are sympathetic to these pleas but 

maintain that any future harvesting should be conducted in a sustainable 

manner. The challenge for conservation biologists was to identify the suite 

of management prescriptions that will allow this. 

43

44 

There is little written documentation of the Greenland Arctic tern harvest. 

Salomonsen (1950) estimated the annual number of eggs harvested to be 

as high as 100,000, while Frich (1997) estimated that 150-200 persons visited 

the archipelago between 6 and 25 June 2006, and harvested between 

3,000 and 6,000 eggs with a harvesting rate of 20-30 eggs per day per person. 

The fi rst year after the ban on egg harvesting, 42 persons were recorded 

in illegal egg harvesting activities between 18 and 24 June 2002. 

On 27 June 2002 two families were caught with a total of 398 boiled Arctic 

tern eggs – collected in less than 24 hours. Since 2002, the level of illegal 

harvesting has declined at Kitsissunnguit (pers. obs.). 

In this study we combine breeding parameters on Arctic terns obtained 

from Greenland in a model to predict whether a sustainable egg harvest 

can be practiced in Greenland. The goal of this study is to see if a balance 

could be found between confl icting interests of egg harvest by indigenous 

people and preserving wildlife as applied in similar studies, e.g. sooty 

terns (Sterna fuscata) and Glaucous-winged gulls (Larus glaucescens) (Feare 

1976; Feare 1976; Zador et al. 2006). The aim is to produce recommendations 

for managers on how enforceable regulations can secure that egg 

harvesting is still possible for local Inuit with minimum impact on the 

Arctic tern population. 

Methods and material 

Study site 

The study was conducted between 2002 and 2006 at Kitsissunnguit 

(Grønne Ejland), the largest Arctic tern colony in Greenland. The archipelago 

contains four major islands and a small number of skerries and islets, 

and is located in the southern part of Disko Bay, West Greenland (68˚50’N, 

52˚00’W). Arctic terns breed at relatively low densities throughout the interior 

parts of the islands on habitat comprising dwarf scrub heath. 

Data collection 

Study plots were defi ned with the tern colony, and all nests laid within 

these through the season were located by intensive searching. The clutch 

size of each nest was recorded and the egg length, breadth and mass was 

recorded using Vernier callipers and electronic scale with 0.1 g accuracy. 

Laying date were determined by backtracking hatching dates by 22 days 

(Cramp 1985, Hatch 2002). Nests were enclosed within a 5-8 meter diameter, 

25 cm high fence of chicken wire, with shelter for the chick being 

provided in the form of piled rocks, peat or driftwood. Nests were then 

checked every 1-2 days to determine their fate. Once the chicks hatched 

they were fi tted with a metal ring and had their wing and weight measured 

on every visit. In the small number of instances where chicks were 

missing from their enclosure, watches were stationed on vantage points 

overlooking the adult’s territory. Any chicks seen walking around nearby 

or being fed by adults were recaptured, and returned to the enclosure if 

their ring number indicated they belonged there. 

Renesting 

Estimates of the replacement period in Arctic terns were obtained by experimentally 

removing clutches of 30 breeding pairs within a study plot

during 2006. Before removing the eggs in the nest, one of the adult birds 

in the pair was caught and dyed with picric acid. After removal, picric 

marked birds in the study plot and surrounding area were observed from 

a blind each day. When these birds were observed incubating their nests 

were located and treated as described above. 

Population size 

The high number of breeding birds combined with the fact that terns at 

Kitsissunnguit breed scattered at low densities over extensive areas meant 

that complete colony counts usually employed for tern censuses were not 

applicable. Instead, densities of Arctic tern nests were sampled along a 

line transect (Buckland et al. 2001). North-south oriented lines 250-m apart 

were used in the fi eld with data subsequent analysed using the software 

Distance version 5.0 (Thomas et al. 2006). Transects were always conducted 

by two people: an observer and a navigator. The observer would walk 

along a transect line at an equal pace, searching for nests and measuring 

the distance between the nest and the transect line with a measuring tape. 

The navigator would walk approximately 15 meters behind the observer 

and give directions to keep them on the track line. The line transect counts 

were conducted between 18 June and 3 July in 2002-2006 at a time in the 

breeding cycle when the majority of the colony had stopped laying eggs 

but prior to the eggs hatching. 

Harvest effi ciency 

When egg harvesting takes place at Kitsissunnguit, participants walk side 

by side searching the ground for Arctic tern nests. When a nest is found, 

all eggs in the clutch are collected (own observations). The Inuit way of 

searching for eggs in the colony resembles the method used under line 

transect population estimation. As input in the harvest model, an estimate 

of harvest effi ciency (eggs encountered per time unit) could be obtained 

from the line transect. 

Parameter estimation 

Nest survival and chick survival were estimated separately using the 

Stephens method described in Rotella et al. (2004), implemented in SAS. 

This uses a non-linear mixed model with a binomial error distribution 

and logit link to estimate daily survival probability. We fi tted a number of 

models containing pertinent covariates and conducted model selection by 

comparing AIC values. The most parsimonious model was that with the 

lowest AIC value, unless a competing model with fewer parameters differed 

from this by less than two AIC units. In the case of nest survival the 

covariates were calendar date and whether the nest was a fi rst or a relay 

clutch. In the case of chicks the covariates were calendar date, chick age, 

hatch order and whether the brood was from a fi rst attempt or a relay. 

The likelihood of a clutch containing two eggs was modelled using a generalised 

linear model with a logit link and binomial error distribution, 

with two eggs clutches adopting a value of one and single egg clutches 

a value of zero. The covariate fi tted to the model was whether the clutch 

was a fi rst or relay nesting attempt. The rate of egg inviability (defi ned as 

the likelihood of an egg that fails to hatch at the end of the incubation period 

owing to infertility, early embryo death or exhaustion during hatching) 

was estimated using a similar model, in which the response variable was 

45

46 

the number of inviable eggs in the clutch and the binomial denominator 

was the total clutch size. Covariates fi tted to the model were calendar date 

and whether the eggs were in a fi rst or relay clutch. Model selection in 

both cases was conducted using likelihood ratio tests. Both these models 

were implemented in SAS. 

Incubation and fl edging periods and their ranges were taken from the literature 

(Cramp 1985). SDs were estimated by generating random normal 

distributions from the published mean and iteratively varied SD values until 

the maximum and minimum values matched the published range in values. 

The mean replacement period and its SD were estimated from the intervals 

between clutch removal and relaying in the experimental study plot. 

The start date (average date upon which individuals laid their fi rst clutch) 

and the stop date (the average date after which individuals would not 

initiate further relay clutches) and their SD were estimated by iteratively 

varying these values in a simulation model of the laying season (see below 

for details). The resulting estimated frequency distribution of nest 

initiation dates were compared with those observed in the experimental 

study plot. The values that minimised the sum of the squared deviations 

between observed and expected values were those used in further simulation 

models (Green 1988, Ratcliffe et al. 2005). 

Simulation model 

A simulation models that allowed for re-nesting (Beintema and Müskens 

1987; Green 1988; Green et al. 1997; Ratcliffe et al. 2005) was used to estimate 

productivity and number of nesting attempts made by Arctic terns, 

and the numbers of eggs that were harvested by humans, under a range 

of harvest management options. Options involved variations in the date 

harvesting started, the number of days the harvest remained open and 

the amount of harvesting effort (spread equally over the open period) in 

terms of man hours. We also only allowed one of the islands in the archipelago, 

Basisø, to be open to harvesting, although the model has the fl exibility 

to include variation in the area open for harvest. 

Breeding pairs were randomly allocated to a management zone by generating 

a random probability: if this was less than the proportion of pairs occupying 

Basisø then the pair was at risk of harvesting, otherwise they bred 

on another island where only natural mortality occurred. Breeding pairs 

were also randomly allocated incubation, fl edging and replacement periods 

and a start and stop date, all of which were drawn randomly from a 

normal distribution with the mean and standard deviation appropriate to 

each parameter. Clutch size was two unless a randomly generated probability 

exceeded the estimated proportion of two egg clutches, in which 

case it was one. The laying period was calculated as the clutch size minus 

1 day, on the basis of one egg being laid per day and incubation started the 

day the last egg was laid. 

During each day of the laying and incubation periods, the clutch was 

subjected to a likelihood of natural failure by testing whether a random 

probability exceeded the date-specifi c daily nest survival rate until the incubation 

period elapsed or the nest failed. On those days when harvesting 

occurred, nests on Basisø were further subjected to a likelihood of being 

harvested in the same way. The daily likelihood of harvest-induced failure 

was calculated by multiplying the effort on that day by the probability of

a nest being encountered per man hour. A tally of the number of eggs harvested 

from the population during the season was kept. 

Pairs that lost eggs during the laying period were allowed to continue to 

lay eggs until clutch completion. Those that failed during incubation were 

allowed to relay if the date of failure plus the replacement period was 

earlier than the stop date, otherwise they would give up breeding for the 

season and simulations for that pair ceased. A tally of the number of nesting 

attempts made by the population during the season was kept. 

If the nest survived to the end of the incubation period, a random likelihood 

was generated for each egg in the clutch: if this was lower than the 

likelihood of egg inviability the egg failed to hatch, otherwise it produced 

a chick. Each chick that hatched was subjected to an age- and hatching order-specifi 

c daily mortality rate until its fl edging period elapsed or it died. 

In the unlikely event of an A-chick dying before its sibling, the B-chick 

would thereafter be exposed to the lower mortality rates experienced by 

A-chicks. A tally of the number of chicks that fl edged from the population 

during the season was kept. 

This procedure was repeated for each of 300 pairs in each season simulated. 

This number of pairs was suffi ciently large to avoid demographic 

stochasticity, but small enough to allow reasonably rapid processing 

times. The number of chicks fl edged and the number of nesting attempts 

were each divided by the number of pairs to estimate productivity and 

nesting attempts per pair. These values were estimated 999 times for each 

harvest management option, and the mean and SD (equivalent to SE) of 

these bootstrapped replicates were calculated. Numbers of eggs harvested 

were treated similarly, but values were scaled up according to the actual 

number of pairs in the harvested area before the average and SD were 

calculated. A bespoke program written in Microsoft Visual Basic 6·0 was 

used to perform the simulations. 

Population model 

In order to evaluate the potential impact of egg harvesting on population 

growth rate, we constructed a simple matrix model in ULM (Legendre 

and Clobert 1995). Matrix models provide a simple and fl exible tool for 

exploring how variation in demographic parameters affects population 

growth (Caswell 2001). Little demographic information is available for the 

Arctic tern, and the model parameters were therefore mainly based on 

best guesses. The model had 3 age classes; adult survival was set to 0.9 

(Devlin et al. 2008), age of recruitment to 3 years (Hatch 2002), and fi rst- 

second- and third-year survival to 0.5, 0.8 and 0.85 respectively. 

Parameter estimates are given ± 1 Standard Error or with 95% confi dence 

intervals when noted. Parameters of the nest survival and chick survival 

models are given on the logit scale. 

47

Figure 1. Daily survival rates in 

relation to age and hatch order for 

a breeding attempt initiated at the 

average day of fi rst clutch initiation 

(day 34). Age is expressed in 

days to or from hatching (where 

day zero is the hatching date). 

The black line represents the survival 

rate of eggs and the A-chick, 

and the grey line the survival rate 

of the B-chick. 

48 

Results 

Model input 

The average clutch size at Kitsissunnguit in 2002-2006 equalled 1.84 

(± 0.012, n = 1752) expressed as a 0.921 (95% CI 0.850-0.960) probability of a 

two egg clutch in fi rst nesting attempt. Clutch size (1.6 eggs, ± 0.13, n = 16) 

in replacement nests was signifi cantly lower (χ2 = 16.8, p = 0.0002, df = 2) 

than non-manipulated nests (2.04 ± 0.026, n = 367) in 2006 and expressed 

as a probability of 0.654 (95 % CI 0.483-0.792) in the model. Daily nest survival 

(egg harvest excluded) varied over the course of the breeding season, 

and daily chick survival varied with chick age (slope 0.067 ± 0.021, 

p = 0.0013) and with hatching order (Fig. 1). Egg inviablity probability was 

estimated to 0.035 (95 % CI 0.020-0.058). 

Daily survival rate 

1.02 

1.00 

0.98 

0.96 

0.94 

0.92 

0.90 

0.88 

The incubation period was set to 22 days and the period from hatching to 

fl edging to 27 days (Cramp 1985, Hatch 2002) whereas the replacement 

period (period between a nest is harvested and a replacement nest produced) 

was set to 9.5 days (± 0.23, n=16). 

The population estimates for Kitsissunnguit based on transects varied between 

15,354 and 21,760 pairs (Egevang and Frederiksen 2010) with an 

average of 18,465 (± 1,023) pairs used in the simulation model. The land 

area (Basisø) exposed to harvesting equals 38.0 % (3.21 vs. total of 8.43 

km 2 ) of the total land area whereas 80.4 % (14,845 ± 1,136) of the total Kitsissunnguit 

population numbers (average 2002-2006) is found at Basisø. 

The line transect also produced estimates of nest encounter probability, 

which can be used as an expression of harvest effi ciency as search for nests 

in the two situations resemble each other. In 2004, 555 minutes of effective 

searching over 6,360 meters of line resulted in 415 found eggs originating 

from 234 nests. These results were used to calculate a probability that an 

individual nest was encountered person hour -1 : 234 [nests found] / (9.25 

[hours search time] × 18,465 [nests, total] ) = 0.001370 

Model output 

0.86 

–30 –25 –20 –15 –10 –5 0 5 10 15 20 25 30 

Nest age (days from hatching) 

Model output consisted of predicted harvest size (number of eggs harvested) 

and productivity (number of chicks fl edging per nest) under various 

harvest scenarios (Figs 2 and 3, A and B). An Harvest-Impact Ratio (HIR) 

is presented in Figure 2 C and Figure 3 C, and was calculated as: 

(HIR = Nh/(P0-Ph)),

Figure 2. A) Harvest (number of 

eggs), B) productivity (fl edglings 

produced per nest) and C) HIR 

(Harvest-Impact Ratio – see text) 

of simulated Arctic terns egg 

harvest plotted against harvest 

start (days after May 1 st ) over the 

breeding season with constant 

duration of harvest season (2 

days) and an effort of 25, 150, 

300 and 500 man hours. 

where Nh is the number of eggs harvested, P0 is productivity when no 

eggs are harvested (unaffected productivity value equals 0.9 chick per 

nest) and Ph is productivity under harvest. 

The variation of harvesting and breeding parameters is most marked in 

relation to season (Fig. 2 and 3). Harvests increased through the fi rst half 

of the season owing to the cumulative increase in the proportion of birds 

that have laid eggs, but then decline later in the season owing to hatching. 

The impact of harvesting on productivity increases through time as losses 

are too late in the season to be replaced. This is evident from the relationship 

between timing of harvests and number of nesting attempts per pair. 

Increasing effort clearly results in greater harvests, and a linear increase 

in impact when controlling for season. Duration has relatively little effect, 

with only a subtle decrease in impact during the middle of the season owing 

to eggs having the opportunity to hatch before being harvested at a 

given level of effort. 

Eggs harvested 

Productivity 

HIR 

16000 

14000 

12000 

10000 

8000 

6000 

4000 

2000 

0 

0.95 

0.90 

0.85 

0.80 

0.75 

0.70 

0.65 

0.60 

A 

B 

C 

25 

150 

300 

500 

10 20 30 40 50 60 70 80 

Days after 1 May 

49

Figure 3. A) Harvest (number of 

eggs), B) productivity (fl edglings 

produced per nest) and C) HIR 

(Harvest-Impact Ratio – see text) 

of simulated Arctic terns egg 

harvest plotted against harvest 

start (days after May 1 st ) over the 

breeding season with constant 

effort (250 man hours) and a duration 

of harvest season of one, 

three, seven or ten days. 

50 

Eggs harvested 

Productivity 

HIR 

9000 

8000 

7000 

6000 

5000 

4000 

3000 

2000 

1000 

0 

0.95 

0.90 

0.85 

0.80 

0.75 

0.70 

0.65 

0.60 

A 

B 

C 

10 20 30 40 50 60 70 80 

Days after 1 May 

At constant effort (Fig. 3C) a peak in HIR occurred between 4 and 8 June defi 

ning the day where harvest can be carried out with minimum effect on the 

overall productivity relative to harvest size. Harvest was high during this 

period (Fig. 3A), although the peak in harvest occurred a bit later (9-21 June). 

The model can help identifying periods in the breeding season when egg 

harvesting is possible/optimal (harvesters will encounter a signifi cant 

number of egg during a visit in the colony) but where the number of eggs 

removed will have a minimum impact at population level. The model predicts 

that the highest number of eggs can be harvested at day 47 (from 1 

May = 16 June) whereas the greatest reduction in productivity (due to harvesting) 

will appear at day 53 (22 June). The Harvest-Impact Ratio (HIR) 

predicts the highest value to occur at day 38 (7 June, i.e. the day with least 

impact on productivity relative to harvest size). Although this is nine days 

away from the peak harvesting date, harvest size on day 38 are on average 

92.2% (± 0.0048) of that on day 47. 

1 

3 

7 

10

At constant duration (Fig. 2) the overall pattern repeats. Highest values of 

HIR (Fig. 2 C) are found around 4 to 8 June (day 35-39) although low effort 

(25 man hours) peaked somewhat later. Highest number of eggs harvested 

(Fig. 2A) also took place in the period around 9-21 June. 

With a breeding success of 0.9 chicks per nest, the matrix model projected 

the population to grow at 4% per year. Reducing breeding success to 0.6 

chicks per nest, similar to the impact of the most intensive harvest scenario 

simulated (Fig. 2B), led to a 4% decline in growth rate so that the 

population was stable. Given the limited data available on Arctic tern demography, 

these values should be taken as indicative. Varying the input 

values for the model led to somewhat different projected growth rates, but 

for all realistic values a decline in breeding success from 0.9 to 0.6 caused 

a 3-4% decline in population growth rate. 

Discussion 

The harvest model applied in the study provides an effective tool to evaluate 

the effects of egg harvesting at Kitsissunnguit. Although a useful tool 

to design future management initiatives, the predictions derived from the 

model must interpreted with caution as these are only as robust as the data 

and parameter estimates upon they are based (review in Bassinger 1998). 

As in other studies (Zador et al. 2006, Feare 1976) exploring modelling of a 

sustainable egg harvest, our main fi ndings suggest that harvesting should 

be conducted as early in the season as possible allowing the birds to relay 

and minimize the impact at population level. The clutch size in replacement 

nests were lower, as found in several Larid species (Golet et al. 2004, 

Wood et al. 2009), and delayed fl edging dates signifi cantly. Despite this, 

Arctic terns at Kitsissunnguit were cable to compensate for the lost of the 

fi rst clutch early in the season by relaying and maintaining a production 

of similar magnitude as non-harvested nests. Although, early breeding 

individuals in bird populations are reported to be more productive that 

late ones (Lack 1966, Perrins 1970), studies has shown that late breeding in 

seabirds may be possible without a decline in productivity (de Forest and 

Gaston 1996, Hipfner et al. 1999, Harris and Wanless 2004) or even benefi - 

cial in some cases (Brown and Morris 1996). 

The aim of this study was to produce management advice on how a sustainable 

egg harvest could be carried out in Greenland. The overall goal is 

to ensure that egg harvesters will be rewarded when visiting the colony, 

while the impact at population level is kept at a minimum allowing the 

birds to relay and produce a signifi cant number of fl edglings. In practice, 

there are three management scenarios which could help insure this: 

1) introduction of a closing date – a fi xed date during the breeding season 

where harvesting will stop, 2) a bag limit – a fi xed number of eggs allowed 

to be harvested over the season or 3) restrictions in effort – a fi xed amount 

of harvesting man hours allowed during the season. 

The harvest model predicts the largest potential harvest to take place six 

days earlier (22 June vs. 16 June) than the date where the highest impact 

was identifi ed, but the date with highest Harvest-Impact Ratio falls nine 

days earlier (7 June) with the numbers of eggs available for harvesters in 

the colony being almost (92%) as high as in the peak period. A simple pop- 

51

52 

ulation model furthermore predicts that high effort harvesting can result 

in a 4% per annum decline in growth rate of the Arctic tern population. 

A closing date (scenario 1) in egg harvesting around 7 June would ensure 

that locals visiting the colony would collect high numbers of eggs. At the 

same time, the proportion of eggs removed from the population would 

only have a minor impact, as most birds would be able to produce a replacement 

clutch. Moreover, early season harvests will be more palatable 

to people as the eggs have not yet developed, although some older Inuit 

prefer the ripe taste of eggs containing partially developed embryos. 

The model helps to predict a theoretical optimal closing date by comparing 

potential harvest versus productivity, but the actual impact on productivity 

would obviously depend on harvest effort. Although the overall 

goal of sustainability in egg harvest also could be reached by management 

scenario 2 or 3, the logistic challenges of enforcing bag numbers or man 

hours spent egg harvesting seems an overwhelming task in rural Greenland. 

The risk of under- or non-reporting of the number of eggs harvested 

is high, and can result in inverse density dependence in a declining 

population owing to a fi xed harvest representing an increased proportion 

impact. Monitoring accumulated egg numbers harvested in the current 

breeding season would furthermore introduce a high level of bureaucracy 

into the management system. 

Another option would be to regulate effort by opening up for harvesting 

on a single day (a “harvest day”) and otherwise keep the colony closed for 

the public. Under the supervision of the hunting offi cer or another local 

designated authority harvest could be allowed to take place at Basisø. This 

would represent an enforceable scenario with well-defi ned dates where 

access to the island was allowed and no public access otherwise. More importantly, 

an effort restricted scenario is positively density dependent, so 

if the population density declines so does the harvest. Long-term changes 

in the population size or years with low breeding numbers will thus result 

in a low proportion of harvested eggs and limited negative impact at 

population level. 

A disadvantage of the introduction of a closing date is that the breeding 

phenology of Arctic terns breeding in Arctic regions may vary considerably 

between years, potentially correlated with sea ice break-up (Hatch 

2002). At Kitsissunnguit the majority of the eggs hatch in the fi rst week of 

July (indicating that egg laying took place in mid-June), but in some years 

early breeding takes place – as the case in 2004 where eggs hatched three 

weeks earlier in mid June. This means that in some years a fi xed closing 

date will be “off synch” with egg laying, resulting in years with either a 

higher impact on breeding productivity than predicted (“early” years) or 

in “late years” with reduced or absent egg harvest before the closing date. 

Ideally, the stage (laying date) of the breeding cycle should be addressed 

early in a current year before a closing date or a harvesting period is determined. 

However, the amount of time acquired to obtain on-site estimates 

of laying date, validate data and communicate an closing date/harvest 

date to the scatted settlements in the Disko Bay, mean that this would not 

be realistic within the time frame available. 

In conclusion, based on the fi ndings in this study future management of 

the Arctic tern population at Kitsissunnguit could include a limited egg 

harvest. This should be restricted to only one island in the archipelago

(Basisø) and include either a) a closing date between 7-10 June or b) introduction 

of a harvest day in the same period. It is however important that 

an opening of egg harvest at Kitsissunnguit is followed by a monitoring 

programme to assess the impact of the harvest to allow fi ne tuning of the 

harvest model. We recommend monitoring productivity and population 

size of birds in closed and open areas of the archipelago so that the model 

input parameters can be adjusted. It would furthermore be especially 

valuable to study fl edging success of replacement clutches in situations 

where a large proportion of the colony relays. 

The output of the harvest model can furthermore be used to address the 

cause of the observed historical decline in Arctic tern population size. The 

fi nding of an optimal harvest period in early June and a severe impact 

on productivity predicted during late June puts the pre-2002 closing date 

of 1 July into perspective. Combining this with the outcome of the simple 

population model with high effort harvest of a 4 % decline in annual 

growth rate (equals a 62.5 % decline over 25 years if the population would 

otherwise be stable), it seems plausible that a high effort historical harvest 

conducted late in the breeding season could have accounted for the observed 

decline in the Greenland Arctic tern. 


The fi eldwork was fi nancially supported by the Danish Energy Agency (the 

climate support program to the Arctic) and the Commission for Scientifi c 

Research in Greenland (KVUG). Special thanks go to the people who assisted 

in the fi eld over the years: David Boertmann (National Environmental 

Research Institute, Denmark), Anders Tøttrup, Mikkel Willemoes and Peter 

S. Jørgensen (University of Copenhagen), Ole S. Kristensen, Lars Witting 

and Rasmus Lauridsen (GNIR), Bjarne Petersen, Qasigiannguit and Stuart 

Benn (Royal Society for Protection of Birds). Also thanks to Finn Steffens 

(Qeqertarsuaq) and Elektriker-it (Aasiaat) for logistic support. 

Literature 

Barrett, R.T. 1989. The effect of egg harvesting on the growth of chicks and breeding 

success of the Shag Phalacrocorax aristotelis and the Kittiwake Rissa tridactyla on 

Bleiksøy, North Norway. Ornis Fennica 66: 117-122. 

Bassinger, S.R. 1998. On the use of demographic models of population viability in endangered 

species management. Journal of Wildlife Management 62: 821-841. 

Beintema, A.J. and G.J.D.M. Müskens 1987. Nesting success of birds breeding in 

Dutch agricultural grasslands. Journal of Applied Ecology, 24, 743–758. 

Boertmann, D. 1994. A annotated checklist to the birds of Greenland. Meddelelser om 

Grønland, Bioscience 38: 1-63. 

Brown, K.M. and R.D. Morris 1996. From Tragedy to Triumph: Renesting in Ring- 

Billed Gulls. The Auk 113(1): 23-31. 

Buckland, S.T., D.R. Anderson, K.P. Burnham, J.L. Laake, D.L. Borchers and L. Thomas 

2001. Introduction to Distance Sampling – Estimating abundance of biological 

populations. Oxford, Oxford University Press. 

53

54 

Burnham, W., K.K. Burnham and T.J. Cade (2005). Past and present assessments of 

bird life in Uummannaq District, West Greenland. Dansk Ornitologisk Forenings 


Caswell, H. 2001. Matrix population models. Construction, analysis and interpretation. 

– Sinauer. 

Coulson, J.C. 2002. Colonial Breeding in Seabirds p. 87-114. In: Schreiber, E.A. and 

Burger, J. (Eds.) Biology of Marine Birds 2002, CRC Marine Biology Series. 

Cramp, S., Ed. 1985. Terns to woodpeckers. Birds of the Western Palearctic. Oxford, 


Croxall, J.P., P.G.H. Evans and R.W. Schreiber 1984. Status and Conservation of the 

World’s seabirds. International Council of Bird Preservation. Technical Publication 

no. 2. 

Devlin, C.M., A.W. Diamond, S.W. Kress, C.S. Hall and L. Welch 2008. Breeding 


Maine. The Auk 125, 850-858. 

Egevang, C. and M. Frederiksen. 2010. Fluctuating breeding of Arctic Terns (Sterna 

paradisaea) in Arctic and High-arctic colonies in Greenland. Manuscript submitted 

WATERBIRDS. 

Egevang, C., K. Kampp and D. Boertmann 2004. The Breeding Association of Red 


a Redistribution of Terns in a Major Greenland Colony. Waterbirds 27(4): 406-410. 

Feare, C.J. 1976. The breeding of the Sooty Tern in the Seychelles and the effect of experimental 

removal of its eggs. J. Zool., (Lond.) 179: 317-360. 

Feare, C.J. 1976. The exploitation of sooty tern eggs in the Seychelles. Biological Conservation(10): 

169-182. 

de Forest, L.N. and A.J. Gaston 1996. The effect of age on timing of breeding and reproductive 

success in thick-billed murre. Ecology 77: 1501-1511. 

Frich, A.S. 1997. Fuglelivet og dets udnyttelse på Grønne Ejland i Vestgrønland [Birds 

and exploitation of birds on Grønne Ejland (Kitsissunnguit)]. Greenland Institute 

of Natural Resources, Nuuk, Greenland. Technical report no. 19. 

Furness, R.W. and Monaghan, P. 1987. Regulation of seabird populations. Seabird 

Ecology, pp. 35-52. Blackie, Glasgow. 

Golet, G.H., J.A. Schmutz, D.B. Irons and J.A. Estes 2004. Determinants of reproductive 

costs in the long-lived black-legged kittiwake: A multiyear experiment. 74: 353-372. 

Green, R.E. 1988. The effects of environmental factors on the timing and success of breeding 

of common snipe (Aves: Scolopacidae). Journal of Applied Ecology, 25, 79-93. 

Green, R.E., Tyler, G.A., Stowe, T.J. and Newton, A.V. 1997 A simulation model of the 

effect of mowing of agricultural grassland on breeding success of the corncrake 

(Crex crex). Journal of Zoology, London, 243, 81–115. 

Hansen, K. 2002. A Farewell to Greenland’s Wildlife, Nowlan, Denmark. 

Hipfner, J.M. 1997. The effect of parental quality and timing of breeding on the 

growth of nestling Thick-billed Murres. Condor 99: 353-360. 

Lack, D. 1966. Population studies in Birds. Oxford University Press, Oxford. 

Legendre, S. and Clobert, J. 1995. ULM, a software for conservation and evolutionary 

biologists. Journal of Applied Statistics 22: 817-834. 

Merkel, F. and Barry, T. (eds.) 2008. Seabird harvest in the Arctic. CAFF International 

Secretariat, Circumpolar Seabird Group (CBird), CAFF Technical Report No. 16. 

Moller, H. 2006. Are current harvest of seabirds sustainable? Acta Zoologica Sinica 52: 

649-652.

Nettleship, D. N., J. Burger and M. Gochfeld 1994. Seabirds on Islands: Threats, Case 

Studies and Action Plans. Birdlife Conservation Series no. 1. 

Perrins, C. M. 1970. The timing of bird’s breeding seasons. Ibis 112: 245-255. 

Ratcliffe, N., S.A. Schmitt and M. Whiffi n 2005. Sink or swim? Viability of a black-tailed 

godwit population in relation to fl ooding. Journal of Applied Ecology 82: 834-843. 

Rotella, J.J., Dinsmore, S.J., Shaffer, T.L., 2004. Modeling nest-survival data, a comparison 

of recently developed methods that can be implemented in MARK and SAS. 

Anim. Biodiv. and Cons. 27: 187–204. 

Salomonsen, F. 1950. Grønlands Fugle. The Birds of Greenland. Copenhagen, 

Munksgaard. 

Thomas, L., J.L. Laake, S. Strindberg, F.F.C. Marques, S.T. Buckland, D.L. Borchers, 

D.R. Anderson, K.P. Burnham, S.L. Hedley and J.H. Pollard 2006. Distance. University 

of St. Andrews, UK. Waterbirds. Vol. 22 (1): 68-79. 

Wood, P.J., M.D. Hudson and C.P. Doncaster 2009. Impact of egg harvesting on breeding 

success of black-headed gulls, Larus ridibundus. Acta Oecologica-International 

Journal of Ecology 35(1): 83-93. 

Zador, S.G., J. . Piatt and A.E. Punt 2006. Balancing predation and egg harvest in a colonial 

seabird: A simulation model. Ecological Modelling 195(3-4): 318-326. 

55

Manus III 

FOOD-RELATED CONSEQUENCES OF 

RELAYING IN THE ARCTIC TERN: 

CONTRASTING PREY AVAILABILITY WITHIN 

A PROLONGED BREEDING SEASON 

Carsten Egevang 1 , Mikkel Willemoes Kristensen 2 , 

Norman Ratcliff e 3 and Morten Frederiksen 4 

1 Greenland Institute of Natural Resources 

Postbox 570, DK-3900 Nuuk, Greenland 

E-mail: egevang@natur.gl 

2 Department of Population Ecology 

Institute of Biology, University of Copenhagen, 

Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark 

3 British Antarctic Survey, High Cross 

Cambridge, CB3 0ET, United Kingdom 

4 National Environmental Research Institute 

Department of Arctic Environment, Aarhus University 

Frederiksborgvej 399, DK-4000 Roskilde, Denmark 

Manuscript submitted IBIS

Page 1 of 26 Ibis Submitted Manuscript 

1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

58 

Food-related consequences of relaying in the Arctic Tern: 

Contrasting prey availability within a prolonged breeding season 

Carsten Egevang 1 , Mikkel Willemoes Kristensen 2 , Norman Ratcliffe 3 and Morten 

Frederiksen 4 

IBIS Review Copy 

Key words: Arctic Tern, Sterna paradisaea, seabird, replacement clutch, Arctic, diet, 

Greenland, Capelin, Mallotus villosus, egg harvesting, intraspecific kleptoparasitism. 

1 Greenland Institute of Natural Resources, Postbox 570, DK-3900 Nuuk, Greenland, E-mail: 

egevang@natur.gl 

2 

Department of Population Ecology, Institute of Biology, University of Copenhagen, 

Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark 


4 National Environmental Research Institute, Department of Arctic Environment, Aarhus 

University, Frederiksborgvej 399, DK-4000 Roskilde, Denmark

Abstract 

Egg harvesting may affect seabird population dynamics at several levels. The rate of 

renesting is one of the most important single factors influencing annual reproductive 

success in harvested populations. However, those birds that do renest may face 


additional costs since late season declines in prey availability may result in elevated 

chick mortality owing to starvation compared to birds hatching their first clutch. Such 

costs are likely to be most pronounced at higher latitudes where windows of opportunity 

for successful breeding are at their narrowest. 

We experimentally removed entire clutches from 30 Arctic Tern nests. Subsequently we 

monitored chick diet, provisioning rate, growth and survival in both the replacement 

nests and a control group. 

Ibis Submitted Manuscript 

The experimental egg removal resulted in 16 replacement nests (53.3 %) located at an 

average distance of 23.4 m (+ 7.65) from the original nest site. The replacement 

clutches were four weeks (+ 3 days) delayed in the breeding cycle compared to non- 

manipulated breeding pairs and clutch size in replacement nests (1.6 eggs, + 0.13, n = 

16) was significantly lower than in the control group (2.1 eggs, ± 0.08, n = 46). 

This late breeding season for replacement clutches resulted in a different foraging 

situation compared to controls. Feeding in control nests was characterised by long 

feeding trips and a low feeding rate, large Capelin as the principal prey item and a high 

incidence of kleptoparasitism. Feeding of chicks from replacement nests was 

characterised by short feeding trips/high feeding rate, fish larvae as principal prey and 

Page 2 of 26 

59


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

60 

an absence of kleptoparasitism. In both groups, Capelin was the single most important 

prey species. 

Despite the altered food availability, the energetic provisioning rate, chick growth and 

chick survival did not differ between the two groups. The parent birds in replacement 

nests switched to low quality prey items and compensated for this by significantly 

increasing feeding rate. 


Introduction 

The ability to produce a replacement clutch may be an important adaptation to increase 

an individual’s annual reproductive success. This is particularly the case in avian 

species that are prone to the loss of offspring due to stochastic events such as flooding 

or to a high risk of nest predation (Brown and Morris 1996; Wendeln et al. 2000). 

However, the production of an additional clutch of eggs requires a costly investment in 

terms of energy, and only females of “high quality” (older, experienced with high 

fitness) in a population may be able to allocate the needed energy into a replacement 

clutch (Hegyi and Sasvari 1998; Hipfner et al. 1999). Furthermore, the delay in the 

breeding cycle may result in altered, sub-optimal conditions for chick rearing compared 

with chick rearing earlier in the season owing to deteriorating food availability (Hipfner 

2001; Perrins 1970). 

In Greenland the Arctic Tern (Sterna paradisaea) has a widespread, patchy distribution 

with the largest colonies found in central West Greenland (Boertmann 1994). The 

Arctic Tern in Greenland has traditionally been subject to intensive exploitation by local

Inuit in terms of egg harvesting (Salomonsen 1950), which most likely has caused a 

general decline in the population over the last decades (Boertmann et al. 1996; Egevang 

et al. 2004). Although egg harvesting has never been economically important in 

Greenland, the harvesting of tern eggs has always been a popular recreational activity, 

often carried out by several generations in family groups. Egg harvesting would usually 

take place during early incubation, but especially the older Inuit generation favour the 


ripe taste of tern eggs in late development state, and harvesting could continue until late 

incubation. Arctic Tern egg harvesting was prohibited in 2002 in Greenland and now 

only occurs illegally at far lower levels (Egevang et al. 2004). Although Arctic Terns 

may produce a replacement clutch to compensate for the harvested eggs, it is unlikely 

that all affected individuals will invest in a second clutch that season. The likelihood of 

relaying may also be related to the stage of the incubating period of which egg loss 

occurs, as found by Feare (1976) in Sooty Terns (Onychoprion fuscatus). Furthermore, 

egg harvesting may create a mismatch between food demand and food availability due 

to the delay in breeding season. 


Surface-feeding seabirds are only capable of feeding in a 2-dimensional space whereas 

diving species exploit a 3-dimensional feeding area. The general perception is that this 

makes surface feeders particularly prone to breeding failure during years of poor food 

availability, a dependency that is particularly pronounced when birds rely on few or a 

single prey species (Regehr and Rodway 1999, Furness and Tasker 2000). 

Reproductive success of Arctic Terns is known to vary greatly with the food 

availability, which has been documented on several occasions in the Boreal zone where 

Lesser Sandeel (Ammodytes marinus) is the key food species (Monaghan et al. 1989; 

Monaghan et al. 1992; Suddaby and Ratcliffe 1997). Although literature on the diet of 


61


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

62 

Arctic Terns in Greenland is very limited (Salomonsen 1950 mentioned zooplankton as 

important), it is likely that Capelin (Mallotus villosus) and invertebrates are the main 

components of the diet. In southern and central West Greenland, Capelin is regarded as 

the key seabird prey species (Carscadden et al. 2002; Friis-Rødel and Kanneworff 

2002), but the link between seabird breeding success and Capelin abundance has not yet 

been quantified. 


In this paper, we study the effect of simulated harvesting of Arctic Tern eggs on a) 

chick growth and reproductive success, and b) chick diet and foraging strategies within 

the species’ core distribution area in West Greenland. The study furthermore presents 

the first estimate of renesting rates in Arctic Terns along with the first detailed 

description of Arctic Tern chick diet in West Greenland. 

Study site 

Materials and Methods 

Fieldwork was conducted in the period 16 June to 30 July 2006 at the archipelago 

Kitsissunnguit (Grønne Ejland) in the southern part of Disko Bay (68˚50’N, 52˚00’W), 

West Greenland (Figure 1). Kitsissunnguit holds the largest Arctic Tern breeding site in 

Greenland with approximately 20,000 pairs, with 17,000 of the pairs breeding at Basis 

Ø – the location of the fieldwork conducted in 2006. The terns breed scattered over 

most of the island, usually with at least 2-4 meters between nests, and only few areas 

have no nests. The total area with breeding terns was estimated to be 3.12 km 2 in 2006, 

with a low overall breeding density of 5,400 nests per km 2 (0.005 pairs per m 2 ).

Arctic Terns arrive in this region during the second half of May, and egg laying takes 

place in early June. The first fledglings are usually seen around mid July, and colony 

exodus happens at the end of July. In early August only few birds are found in the 

colony. Although hatching is well synchronized within large sections of the archipelago, 

a tendency towards later hatching is found when moving from the central part of the 

islands towards the outskirts of the colony. This tendency is even more pronounced 


(with up to two week delay in hatching date) when moving W from Basis Ø towards the 

more exposed skerries, Saattuarsuit, in the archipelago (Pers. obs. Kitsissunnguit 2002- 

2006). 

Study plots 


Two study plots were established within the first week of the field season: a control plot 

(46 nests; size app. 13,000 m 2 ) and a plot where clutches were experimentally removed, 

designated the “relay plot” (30 nests; size app. 8,200 m 2 ) hereafter. Well-drained, flat 

heather sections at the edge of large continuous areas of breeding Arctic Terns 

comprised the habitat in both study plots, with a distance to the shoreline of 

approximately 200 and 100 m, respectively. Daily measurements of hatching stage, 

chick mass (to the nearest 0.1 g) and wing length (to the nearest 1mm) were conducted 

in both plots in order to quantify chick age, growth rate and survival. The chicks in both 

study plots were kept in enclosures (chicken wire, mesh size 25 mm, height 25 cm, 

diameter app. 5 m), which ensured consistent measurements up to the age of fledging. 

Chicks in the control plot were monitored with daily measurements to the day of 

fledging while the last day of measurements in the relay plot was 30 July, when A- 

chicks had an age of 9-11 days (day of hatching = day 0). 


63


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

64 

In the relay plot one parent at each nest was caught using Kieler-traps, and biometric 

measurements, including bill length, were taken. The adult bird was dyed with picric 

acid on the right or left cheek and upper rump and fitted with a metal ring and coloured 

darvic ring before it was released. To ensure that no nests were abandoned due to the 

handling of the bird, the marked adult bird in each of 30 nests was observed incubating 

before the eggs were removed. On 19 June, the eggs of the nests in the relay plot were 


collected and the reproductive behaviour of the adult birds closely monitored in the 4-28 

days post removal. The combination of marks used ensured that individual 

identification could be made from a distance of up to 30 meters using a 20-60x spotting 

scope. The search for replacement nests was conducted within the relay plot and in a 

100m strip around the plot. 

Feeding observations 

Observations of food provisioning by the adult birds to the enclosed chicks were 

conducted in both the control plot and the relay nests from a hide positioned 10-25 

meters from the nests. Variables recorded during the feeding observations included: 

time of the prey delivery, which chick was fed when the brood comprised two chicks, 

prey identity (to the highest taxonomic resolution possible), prey size (in relation to the 

bill of the adult bird) and the outcome of the prey delivery (successful or 

kleptoparasitised by other adult birds). In the relay plot, the provisioning (marked or 

unmarked) adult was also recorded. In order to minimise bias originating from 

individual prey preferences in a parent bird, the hide was relocated on a regular basis to 

avoid long periods of observations on the same nests. The relocation of the hide was

always done at least 20 hours prior to feeding observations to allow birds to habituate to 

its presence. Feeding observations was conducted by three observers doing 3-hour 

sessions, where feeding observations during and 10 minutes after changing shifts were 

omitted from data analysis. Total duration of observation periods varied between three 

and twelve hours, involving 3-4 nests at a time. Feeding observations were conducted 

between 08:30 and 22:00 local time (GMT minus 3 hours) on clear days. The weather 


during chick rearing in both the control and relay group was generally good and did not 

include severe weather systems preventing the birds from foraging. Furthermore, on 

days with strong wind, heavy rain or thick fog, feeding observations were not 

conducted, to avoid bias from reduced foraging performance in the adult bird (Hatch 

2002). 

Data analysis 

The data on feeding observations from both study plots were subdivided into three age- 

classes according to chick age: 0-3 days, 4-7 days and 8-11 days. When comparing 

chick diet between the two groups, feeding observations were pooled into “nest hours” 

(i.e. four hours of observation on three nests equals twelve nest hours). When 

comparing feeding rates between the two groups, observations were pooled into 5-hour 

periods. 

The first hatched chick (the A-chick) is superior in competition with siblings for food 

brought to the nest due to larger size and higher mobility (Langham 1972). In the 

current analysis, chick rank is included as a prediction when comparing diet and growth 

rates in control versus relay chicks. 



65


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

66 

Conversions of food item size from bill length to mm were done using the average 

culmen length of the adults caught and measured in the relay plot. Food items were 

scored in relation to the total bill length in the field, so in order to achieve an estimate of 

prey item size, a factor of 0.25 was added to culmen length. Conversion factors from 

length to weight and calorific content were taken from Weslawski et al. (1994), Finley 

et al. (1990), Bradstreet (1980), Harris and Hislop (1978) and Wigley et al. (2003). Wet 


weights of both fish and fish larvae were converted to dry weight by dividing by 4.35 

(Payne et al. 1999). 

In the comparison of means with equal variances and normal error distribution, t-tests 

were used. Data not normally distributed were normalized with an arcsinh- 

transformation. To test for difference in variances, F-tests were used, and differences in 

means were tested using Welch’s approximate t (Zar 1999). All tests were two-tailed, 

and = 0.05 was taken as the threshold for significance. Estimates are presented as 

means with one standard error throughout the text. Daily chick survival probability was 

estimated using known-fate models in MARK (White and Burnham 1999), and 

between-group differences were evaluated using likelihood ratio tests. These models 

allow for right-censoring, i.e. the fact that chicks in the experimental nests were only 

followed until they were 9-11 days old. 

Calculations of area size were performed in GIS software MapInfo 8.5. The Internal 

Egg Volume (IEV) of the Arctic Tern egg was estimated as IEV = 0.00048 LB 2 , where 

egg length (L) and breadth (B) were measured with electronic callipers. 

Results

Relaying 

Of the 30 pairs whose clutches were removed, 16 pairs (53.3 %) were observed to 

produce a replacement clutch. All embryos in the collected eggs in the relay plot were in 

the last half of their development stage (estimated 2-9 days from hatching), with 

expected hatching dates between 21 and 28 June. Eggs in the control plot hatched 

between 20 June and 3 July (median hatching date 23 June). The first replacement nests 


were observed in the plot 28 June, and replacement eggs hatched between 19 July and 

22 July. A backwards calculation using a 22 day incubation period (Hatch 2002, Cramp 

1985) showed that replacement eggs were laid between 27 and 30 June – eight to eleven 

days post egg collection. Hence, the removal of eggs at the end of incubation delayed 

the nesting season of experimental nests by approximately four weeks (28 + 3 days) 

compared to the original hatching date. 


The mean clutch size (Table 3) of the control plot was 2.1 (+ 0.08, n = 46), similar to 

the relay plot where clutch size was 2.1 eggs (+ 0.07, n = 30) before egg removal. The 

mean clutch size (1.6 eggs, + 0.13, n = 16) in replacement nests was significantly lower 

( 2 = 16.8, p = 0.0002, df = 2) than in control nests and nests with experimentally 

removed eggs (pooled data). Total clutch volume (expressed as Internal Egg Volume) 

equalled 35.4 ml (+ 8.68, n = 42) on average per nest in the control plot, whereas IEV in 

relay nests (27.8 ml, + 8.76, n = 11) was significantly lower (t = 2.53, df = 51, p = 

0.014). However, IEV of A-eggs in relay nests (17.34 ml, n = 11) were not different (F 

= 0.58, P = 0.45) from A-eggs in the control plot (17.69 ml, n = 55). B-eggs in relay 

nests (16.373 ml, n = 7) were near-significant smaller (F = 3.56 P = 0.06) compared 


67


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

68 

with B-eggs from the control group (17.24 ml, n = 51), consistent with the smaller 

clutch size. 

Hatching rate in the relay plot (84.0%, + 37.4, n = 25) was lower, but not significantly 

different ( 2 = 0.6, p = 0.433, df = 1), than hatching in the control plot (91.8%, + 27.7, n 

= 97). The lower hatching rate in the relay plot was caused by higher nest abandonment 

with two nests (both with two eggs) out of 16 nests (12.5%) abandoned before hatching. 


In the control plot only a single case (2.2%) of nest abandonment was observed. 

There were no instances where the same nest scrapes were reused, and replacement 

nests were located between 6.0 and 100.6 m (mean 23.4 + 7.65) from the original nest. 

Chick production and growth rate 

It is very rare that B-chicks (or C-chicks) survive to the age of fledging in Disko Bay 

(pers. obs. Kitsissunnguit 2002-2005), but the 2006-season had a “high” proportion of 

B-chicks that survived to fledgling. In the control plot, where the fate of chicks from 2- 

egg clutches was known in 37 nests, five of the pairs (13.5%) fledged two chicks. This 

was not different ( 2 = 0.15, df = 1, p = 0.7) from the relay plot where one (11.1%) of 

nine two-egg clutches produced two chicks beyond the age of 11 days. 

Daily chick survival varied with age, and this variation could parsimoniously be 

modelled with four age groups: 0-1 days, 2-5 days, 6-10 days, and 11 days and over. 

These age groups were based on post-hoc examination of estimates from a model with 

full daily variation in survival and thus to some extent arbitrary, but the 

main conclusions in the following were unchanged when using other models for age 

variation (constancy, full daily variation, quadratic trend). Survival differed between A,

B, and C chicks ( 2 = 20.5, df = 2, P < 0.0001), but there was no difference between 

control and experimental chicks (2 = 0.033, df = 1, P = 0.85). Estimated daily survival 

was very high from 0-1 days (0.996, 0.986, and 0.964 for A, B and C chicks, 

respectively), low from 2-5 days (0.947, 0.836, and 0.658), higher from 6-10 days 

(0.982, 0.939, and 0.852), and very high after 11 days (0.998, 0.992, and 0.979). 

Estimated overall survival to fledging was 0.691 (+ 0.066) for A chicks, 0.293 (+ 0.073) 


for B chicks and 0.050 (+ 0.081) for C chicks. 

The daily growth rate (Table 3), from day four to eleven of the control group (6.39 g + 

0.62) did not differ significantly (t = 0.01; df = 58, p = 0.992) from that of chicks 

hatched from experimental clutches (6.61 g + 0.82). Similarly, growth rate of wing 

length did not differ between the two groups (control 7.66 mm + 1.55 vs. relay 7.66 mm 

+ 0.67; test: t = 0.31; df = 58, p = 0.761). 

Diet 


A total of nine different prey species were identified from 1237 prey deliveries to the A- 

chick (summarized in Table 1) in the control and the relay plot combined, originating 

from 203 and 155 nest hours of observation, respectively (Table 2). Unidentified prey 

items comprised approximately 7 % of the total number of observed feeds. The majority 

(>90%) of these feeds were unidentified fish larvae. The most numerous prey species 

was Capelin, accounting for 20.0-77.6 % of prey numbers in the three age groups and 

between 97.1 and 99.9 % of the corresponding energy (fish and fish larvae combined) in 

the control group. In the relay group only one feed with a juvenile Capelin (i.e. older 


69


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

70 

than 0-group but not spawning) was observed. Still, Capelin larvae accounted for 

between 47.7 and 73.7 % by numbers and between 37.1 and 51.4 % of the energy 

brought to replacement chicks. The size of Capelin larvae (0-group) averaged 29 mm 

estimated from an average observed prey size of 0.74 (+ 0.005, n = 627) bill length 

when applying the average bill size (culmen = 31.6 mm, n = 26, + 0.255) found in the 

adult birds at Kitsissunnguit. 


The only other post-larval fish observed in the chick diet was two successful deliveries 

with Polar Cod (Boreogadus saida) to replacement chicks (Table 1). Two other fish 

larvae were recorded in the diet: Stichaeidae sp. and Wolffish (Anarhichas spp.) with 

362 and 91 successful deliveries, respectively and only found in the diet of replacement 

chicks. Other prey species identified in the diet was Gammarus spp. (46 deliveries), 

shrimp Pandalus ssp. (2 deliveries) – from trawler discard and butterfly (Lepidoptera 

spp.) larvae (2 feeds). These did not, however, contribute significantly to energy 

provision within the three age groups (


in the mean amount of energy brought to the chicks (t = 1.26, df = 30, P = 0.219). The 

oldest chicks showed the same pattern, with the number of feeds being significantly 

higher in replacement chicks than in the control chicks (Welch’s approximate t = 2.42, 

df = 7, P = 0.046), while the amount of energy was not (t = 1.01, df = 24, P = 0.111). 

Kleptoparasitism was only observed in the control plot, where 30 of 81 Capelin feeds 

(37.0 %) were stolen by other adult birds before a successful delivery to the chick in the 


nest. Significant differences in kleptoparasitism between experimental and control plot 

could be detected in the two oldest age groups (4-7 days: 2 = 146.67, df = 1; 8-11 days: 

2 = 39.64, df =1). The only prey species targeted by kleptoparasites was large Capelin. 

Stolen fish (10.28 cm + 3.54, n = 29) were significantly larger (t = 2.677, df = 78, P = 

0.007) than Capelin fed to the chick (8.82 cm + 2.11, n = 51). 

Discussion 

The study at Kitsissunnguit reveals a striking difference in diet composition between 

non-manipulated nests and nests with simulated egg harvesting within the same 

breeding season. The predominant prey during the first chick-rearing period, juvenile 

Capelins, was virtually absent with only one delivery observed in the relay group versus 

81 (51 successful) Capelin deliveries in the control group. During the four-week delay 

in the breeding cycle, it is likely that a shift in available food items in the upper layers 

of the water column have taken place. The parent birds switched to an alternative prey 

and hence altered the duration of the feeding trips from long to short trips. In the two 

older age groups (4-11 days), this shift resulted in 10-15 times higher item provisioning 

rate in the relay group, however Capelin fish larvae contain less than 2% of the energy 


71


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

72 

of a 10 cm Capelin and so the total amount of energy brought to the relay chicks did not 

differ from that of the control group. 

Surprisingly, despite the altered food availability, survival and growth of chicks hatched 

from replacement eggs were similar to those in the control group up to the age of 11 

days at least. Thus, the adult birds were able to compensate for the low quality diet by 

increasing the feeding rate significantly without adverse effects on chick growth or 


survival. However, the fate of the replacement chicks could only be monitored to the 

age of nine to eleven days, whereas the control group could be followed to the day of 

fledging. Chick mortality due to starvation usually occurs within the first three to eight 

days for chicks hatched from first nesting attempts, and we assumed that survival to day 

11 was equivalent to fledging. However, starvation of near-fledged tern chicks has been 

documented in Shetland (Monaghan et al. 1989), probably owing to vertical migration 

of prey fish, and this could have affected terns chicks hatched from replacement nests in 

Kitsissunnguit after the study ceased. 

Another behavioural feature associated with feeding that differed between the two 

groups of chicks was the absence of intraspecific kleptoparasitism in the relay group. 

Steele and Hockey (1995) found that prey representing long handling times was most 

likely to be stolen by conspecific individuals. Other investigators (Hopkins and Wiley 

1972; Ratcliffe et al. 1997; Stienen et al. 2001; Dies and Dies 2005) found that larger 

fish stand a higher risk of kleptoparasitism. This was also the situation at Kitsissunnguit, 

where only larger fish and not fish larvae were kleptoparasitised, and stolen fish on 

average were 16.5% larger than those fed to chicks. Kleptoparasitism may also 

influence chick feeding in other ways than the actual chance of losing the fish. Handling

time when delivering a fish was very high due to other adults pursuing the feeder, hence 

reducing provisioning rate even if the prey was delivered to the chick. 

Potential predators on Arctic Tern eggs and chicks, such as Glaucous Gull (Larus 

hyperboreus), Great Black-backed Gull (Larus marinus), Common Raven (Corvus 

corax) and Arctic Skua (Stercorarius parasiticus) were observed close to or sometimes 

within the borders of the colony. The presence of avian predators along the shoreline 


did not cause a response from the many breeding birds, but when the predator moved 

into the air space above areas with nests, large numbers (usually hundreds) of terns 

would join in an effective, mutual mobbing of the intruder. There were no incidents of 

egg predation within the study plots (n = 121 eggs) and amongst the hatched eggs (n = 

108), 52 chicks survived to fledging (or presumed fledging (n = 13) in the relay plot), 

51 chicks died from starvation and five chicks “disappeared” from the enclosures and 

could not be found again – it is likely that these five chicks were predated. No chick 

predation was observed in the relay plot. Predation from Peregrine Falcons (Falco 

peregrinus) was regularly observed on fledglings and adult birds. Overall, the predation 

level at Kitsissunnguit must be considered low compared to other studies (Hatch 2002, 

Cramp 1985). 


A weakness of the study at Kitsissunnguit is that we could not obtain information on 

survival beyond day 11 in replacement chicks. However, the high survival rate of chicks 

older that 11 days within the control group in the same breeding season, combined with 

low observed predation rate and an equal growth rate amongst replacement chicks and 

control chicks, makes it very likely that these chicks would in fact survive to fledging. 

Another weakness is the estimate of the proportion of relaying breeding pairs. Although 


73


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

74 

considerate observation time was spent in the relay plot, at the surroundings and in the 

colony in general, we can not completely rule out the possibility that more replacement 

nests may have occurred. The estimate in this study, where approximately half of the 

birds relayed, should be regarded as a low-end estimate. This estimate should 

furthermore be seen in a context of variation of favourable feeding conditions between 

years that may affect the ability to relay, and more importantly, that removal of eggs 


earlier in the season may have induced a higher degree of relaying, as found by Feare 

(1976) in Sooty Terns. 

When summarizing the results of this study regarding the potential effects of egg 

harvesting, the equal growth rate and chick survival in the two groups combined with 

the absence of kleptoparasitism in the relay group may indicate only a minor negative 

effect of harvesting. However, only 53.3 % of breeding pairs produced a replacement 

clutch, which in itself reduced the reproductive outcome of the current breeding season 

substantially. Also, clutch size was lower in the breeding pairs that did produce a 

replacement clutch. Furthermore, the long-term effects on the adult birds of the 

markedly higher feeding rates found in replacement nests may be an important aspect at 

the individual level. 


We would like to thank Rasmus Lauridsen, Greenland Institute of Natural Resources 

and Peter Søgaard Jørgensen, University of Copenhagen for assistance with data 

collection, and Finn Steffens, captain of M/S “Maja S” of Qeqertarsuaq for logistical 

support. The fieldwork at Kitsissunnguit in 2006 was financially supported by 

Greenland Institute of Natural Resources and the Commission for Scientific Research in

Greenland (KVUG). Permits required to conduct fieldwork, handle birds and collect 

eggs were provided by the Greenland Home Rule, Directorate of Environment and 

Nature. 

References 


Boertmann, D. (1994). An annotated checklist to the birds of Greenland. Meddelelser 

om Grønland, Bioscience 38: 1-63. 

Boertmann, D., A. Mosbech, K. Falk and K. Kampp (1996). Seabird colonies in western 

Greenland (60˚-79˚30´ N. lat.). National Environmental research Institute, Denmark. 

148 pp – NERI Technical Report No. 170. 

Bradstreet, M. S. W. (1980). Thick-billed murres and black guillemots in the Barrow 

Strait area, N. W. T., during spring: diets and food availability along ice edges. 

Canadian Journal of Zoology 58, pp 2120-2140. 

Brown, K. M. and R. D. Morris (1996). From tragedy to triumph: Renesting in ring- 

billed gulls. Auk 113(1): 23-31. 

Carscadden, J. E., W. A. Montevecchi, G. K. Davoren and B. S. Nakashima (2002). 

Trophic relationships along capelin (Mallotus villosus) and seabirds in a changing 

ecosystem. ICES Journal of Marine Science 59: 1027-1033. 

Cramp, S. Eds. (1985). Terns to woodpeckers. Birds of the Western Palearctic. Vol 4. 

Oxford, Oxford University Press. 



75


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

76 

Dies, J. I. and B. Dies (2005). Kleptoparasitism and Host Responses in a Sandwich Tern 

Colony of Eastern Spain. Waterbirds: 167-171. 

Egevang, C., K. Kampp and D. Boertmann (2004). The Breeding Association of Red 




Feare, C. J. (1976). The breeding of the Sooty Tern in the Seychelles and the effect of 

experiemental removal of its eggs. J. Zool., (Lond.) 179: 317-360. 

Finley, K. J., M. S. W. Bradstreet and G. W. Miller (1990). Summer feeding ecology of 

Harp Seals (Phoca groenlandica) in relation to Arctic Cod (Boreogadus saida) in the 

Canadian high Arctic. Polar Biology, 10, pp 609-618. 

Friis-Rødel, E. and P. Kanneworff (2002). A review of Capelin (Mallotus villosus) in 

Greenland waters. ICES Journal of Marine Science 59: 890-896. 

Furness, R. W. and Tasker, M. L. (2000) Seabird-fishery interactions: quantifying the 

sensitivity of seabirds to reductions in the sandeel abundance, and identification of key 

areas for sensitive seabirds in the North Sea. Marine Ecology Progress Series, 202: 253- 

264. 

Harris, M. P. and J. R. G. Hislop (1978). The food of young Puffins Fratercula arctica. 

Journal of zoology, London, 185, pp 213-236. 

Hatch, J. J. (2002). Arctic Tern (Sterna paradisaea). In The birds of North America, no. 

707 (A. Poole and F. Gill, eds.). The Birds of North America, Inc., Philadelphia, PA.

Hegyi, Z. and L. Sasvari (1998). Parental Condition and Breeding Effort in Waders. The 

Journal of Animal Ecology 67(1): 41-53. 

Hipfner, J. M. (2001). Fitness-related consequences of relaying in an arctic seabird: 

survival of offspring to recruitment age. The Auk 118 (4), pp 1076-1080. 

Hipfner, J. M., A. J. Gaston, D. L. Martin and I. L. Jones (1999). Seasonal declines in 


replacement egg-laying in long-lived seabirds: costs of late breeding or variation in 

female quality? Journal of Animal Ecology 68: 988-998. 

Hopkins, C. D. and R. H. Wiley (1972). Food parasitism and competition in two terns. 

Auk 89: 583-594. 


Langham, N. P. E. (1972). Chick survival in terns (Sterna spp.) with special reference to 

the common tern. Journal of Animal Ecology 41: 385-395. 

Monaghan, P., J. D. Uttley and M. D. Burns (1992). Effect of changes in food 

availability on reproductive effort in arctic terns Sterna paradisaea. Ardea 80(1): 71-81. 

Monaghan, P., J. D. Uttley, M. D. Burns, C. Thaine and J. Blackwood (1989). The 

relationship between food supply, reproductive effort and breeding success in Arctic 

Terns Sterna paradisaea. Journal of Animal Ecology 58: 261-274. 

Payne, S. A., B. A. Johnson and R. S. Otto (1999). Proximate composition of some 

north-eastern Pacific forage fish species. Fisheries Oceanography 8:3, pp 159-177. 

Perrins, C. M. (1970). “The timing of birds' breeding seasons. Ibis 112:242-255 


77


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

78 

Ratcliffe, N., D. Richardson, R. Lidstone Scott, P. J. Bond, C. Westlake and S. Stennett 

(1997). Host Selection, Attack Rates and Success Rates for Black-Headed Gull 

Kleptoparasitism of Terns. Colonial Waterbirds 20(2): 227-234. 

Regehr, H. M. and M. S. Rodway (1999). Seabird Breeding Performance during Two 

Years of Delayed Capelin Arrival in the Northwest Atlantic: A Multi-Species 

Comparison. Waterbirds 22(1): 60-67. 


Salomonsen, F. (1950). Grønlands Fugle. The Birds of Greenland. Copenhagen, 

Munksgaard. 

Steele, W. K. and P. A. R. Hockey (1995). factors influencing rate and success of 

intraspecific kleptoparasitism among kelp gulls (Larus dominicanus). The auk 112(8): 

847-859. 

Stienen, E. W. M., A. Brenninkmeijer and C. E. Geschiere (2001). “Living with gulls: 

the consequences for Sandwich Terns of breeding in association with Black-headed 

Gulls. Waterbirds 24:68-82. 

Suddaby, D. and N. Ratcliffe (1997). The Effects of Fluctuating Food Availability on 

Breeding Arctic Terns (Sterna paradisaea). Auk 114(3): 524-530. 

Uttley, J., P. Monaghan, & S. White (1989). Differential effects of reduced sandeel 

availability on two sympatrically breeding species of tern. Ornis Scandinavica 20: 273- 

277.

Wendeln, H., P. H. Becker and J. Gonzáles-Solís (2000). Parental care of replacement 

clutches in common terns (Sterna hirundo). Behaviour Ecology Sociobiology 47: 382- 

392. 

Weslawski, J. M., L. Stempniewicz and K. Galaktionov (1994). Summer diet of 

seabirds from the Frans Josef Land archipelago, Russian Arctic. Polar research 13(2), 

pp 173-181. 


White, G.C. and K.P. Burnham (1999). Program MARK: survival estimation from 

populations of marked animals. Bird Study 46 (suppl.), 120-139. 

Wigley, S. E., H. M. McBride and N. J. McHugh (2003). Lenght-weight relationships 

for 74 fish species collected during NEFSC research vessel bottom trawl surveys, 1992- 

99. NOAA Technical Memorandum NMFS-NE-171. U.S. Department of Commerce. 

Zar, J. H. 1999. Biostatistical Analysis. 4th edition, Prentice Hall, Englewood, New 

Jersey. 



79


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

80 

Table 1. The number of successful prey deliveries (n = 1237), prey group and 

corresponding energy (% of total energy within the age group) fed to Arctic tern Achicks 

at Kitsissunnguit 2006. Observations are subdivided into three age groups for 

both control chicks (203 nest hours) and replacement chicks (155 nest hours). The prey 

group “Fish larvae” covers three “0-group” fish species whereas “Others” covers both 

marine (Gammarus spp., Pandalus ssp.) and terrestrial (Lepidoptera spp., Arachnids) 

species (see Results for details). 


Control Relay 

Chick age Species/group Number (%) Energy (%) Number (%) Energy (%) 

Capelin 4 (7.3) 84 1 (1.5) 22 

0-3 days Fish larvae 7 (12.7) 11 65 (98.5) 78 

Others 44 (80.0) 5 0 0 

Capelin 24 (28.2) 99 0 0 

4-7 days Fish larvae 15 (17.7) 1 648 (99.7) 100 

Others 46 (54.1) 0 2 (0.3) 0 

Capelin 22 (44.9) 98 0 0 

8-11 days 

Arctic cod 

Fish larvae 

0 

16 (32.7) 

0 

2 

2 (0.6) 

329 (99.4) 

13 

87 

Others 11 (22.5) 0 0 0

Table 2. Feeding rate (numbers and energy), prey size and kleptoparasitism in control 

and replacement nests of Arctic Tern, Kitsissunnguit 2006 divided into age groups 

(days). Numbers (n) of five-hour periods are presented under each age group and values 

are listed as mean (+ SE) except for kleptoparasitism (frequency and proportion 

presented). The number of 5-hour periods in feeding rate observations are indicated by 

“n =” while “n =” in prey size indicate the number of prey items recorded. P-value from 

statistic tests (t-test and Chi square for kleptoparasitism) is shown for each group. 

Mean feeding rate 

(feedings/5 hours) 

Mean feeding rate (kj/5 

hours) 

Mean prey size (dw in 

mg) 

Frequency of 

kleptoparasitism 

(proportion stolen) 



Age group Control 

Experimental 

nests 

P 

value 

0-3 (n = 4/10) 13.8 (6.2) 6.6 (1.3) 0.333 

4-7 (n = 19/13) 4.5 (0.9) 49.2 (12.0) 0.003 

8-11 (n = 18/8) 2.7 (0.5) 41.0 (15.8) 0.046 

0-3 (n = 4/10) 8.8 (3.7) 10.0 (3.0) 0.816 

4-7 (n = 19/13) 27.2 (7.9) 32.1 (7.4) 0.219 

8-11 (n = 18/8) 21.3 (5.3) 37.1 (9.0) 0.111 

0-3 (n = 55/66) 26.3 (11.4) 61.3 (14.9) 0.077 

4-7 (n = 85/651) 238.6 (51.0) 26.5 (1.2) < 0.001 

8-11 (n = 49/331) 306.8 (70.8) 37.1 (4.9) < 0.001 

0-3 2%(1/56) 0 0.928 

4-7 19% (20/105) 0 < 0.001 

8-11 18% (11/60) 0 < 0.001 


81


1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

11 

12 

13 

14 

15 

16 

17 

18 

19 

20 

21 

22 

23 

24 

25 

26 

27 

28 

29 

30 

31 

32 

33 

34 

35 

36 

37 

38 

39 

40 

41 

42 

43 

44 

45 

46 

47 

48 

49 

50 

51 

52 

53 

54 

55 

56 

57 

58 

59 

60 

82 

Table 3. Summary of the findings in replacement nests and control nests in Arctic terns 

at Kitsissunnguit 2006. 

Control N Relay n 


Clutch size 2.1 + 0.08 1 76 1.6 + 0.13 16 

Internal Egg Volume (ml) per nest 35.4 + 1.34 1 42 27.8 + 2.64 11 

Nests with 3 eggs (%) 7 (15.2%) 1 46 0 (0%) 16 

Hatching success 90.6% 96 84.0% 25 

Reproductive outcome (chicks per nest) 0.85 ± 0.088 46 0.88 + 0.125 2 16 

Daily wing growth rate (age 4-11 days) 7.66 mm + 1.55 35 3 7.66 mm + 0.67 13 4 

Daily mass increase (age 4-11 days) 6.39 g + 0.62 35 3 6.61 g + 0.82 13 4 

1 2 

: Collected eggs from the relay plot included. : Survival in relay chicks only measured until chick age 9- 

11 days. 3 : Number of daily measurements varied between 38 and 28 chicks, 4 : Number of daily 

measurements varied between 15 and 9 chicks



Figure 1: Map of the archipelago Kitsissunnguit with field work location, Basis Ø. Inset 

map in the right upper corner indicates location of Disko Bay. 


83

Manus IV 

FLUCTUATING BREEDING OF ARCTIC TERNS 

(STERNA PARADISAEA) IN ARCTIC AND 

HIGH-ARCTIC COLONIES IN GREENLAND 

Carsten Egevang 1 and Morten Frederiksen 2 

1 Greenland Institute of Natural Resources 

Postbox 570, 3900 Nuuk, Greenland 


2 National Environmental Research Institute 

Aarhus University, Department of Arctic Environment 

Frederiksborgvej 399, PO Box 358, DK-4000 Roskilde, Denmark 

Manuscript submitted WATERBIRDS

86 

Fluctuating breeding of Arctic Terns 

(Sterna paradisaea) in Arctic and 

High-arctic colonies in Greenland 

CARSTEN EGEVANG 1 AND MORTEN FREDERIKSEN 2 

1Greenland Institute of Natural Resources, Postbox 570, 3900 Nuuk, Greenland 


2 National Environmental Research Institute, Aarhus University, Department of Arctic 

Environment, Frederiksborgvej 399, PO Box 358, DK-4000 Roskilde, Denmark 

Abstract 

The Arctic Tern is regarded as a species showing large variations in colony 

attendance between breeding seasons. This makes comparable counts of 

colony size complicated, and reliable population status estimates are diffi 

cult to obtain. 

In this study, three different areas in two different regions with breeding 

Arctic Terns were surveyed in consecutive breeding seasons. Both small, 

mid and large sized colonies were surveyed over a four to fi ve year period 

in Arctic and High-arctic Greenland. Colony size was estimated with simple 

direct counts or line transects to estimate nest density. 

A considerable between-year variation in population size was found in 

the small and mid sized Arctic Tern colonies in Disko Bay, West Greenland 

(mean CV of individual colonies 117.5 %, CV of total 49.6 %). In the largest 

colony in Greenland, Kitsissunnguit, overall colony size only showed 

minor variations (CV 14.6 %), but variation in breeding density was pronounced 

at sub-colony level (mean CV 47.4 %). When combining the surveyed 

colonies in Disko Bay, the total population size varied only little (CV 

6.7 %). This amount of variation was smaller than expected if colonies fl uctuated 

independently (P = 0.023), indicating that local movements between 

the colonies took place and that annual variation was linked to local phenomena 

rather than to large-scale phenomena. At High-arctic Sand Island 

in Northeast Greenland, Arctic Terns returned to the colony at the start of 

each season, but complete breeding failure, likely caused by the presence of 

Arctic Fox in the colony, was recorded in two out of four seasons. 

Based on the results of this study, we recommend future Greenland Arctic 

Tern colony surveys to 1) survey multiple colonies within the same season 

covering adjoining colonies over a larger area, or 2) survey one large, representative 

colony in multiple years. 

Keywords: Arctic Tern, seabirds, survey methods, breeding, Sterna paradisaea, 

Arctic, Greenland. 

Running head: Arctic Tern colonies in Greenland

To obtain sustainable use of natural resources or to identify a need for 

conservation actions, wildlife managers around the world depend on reliable 

population status estimates. Whether ecosystem-based management 

or conservation issues of important natural sites are the focus, the assessment 

of population size development form counts is a key tool. 

Seabirds as a group are generally characterised by long life expectancy 

with a relatively small annual investment in reproductive outcome, often 

breeding in large colonies with a high degree of site fi delity (Coulson 2002; 

Weimerskirch 2002). Colonial breeding at predictable sites makes seabirds 

ideal for monitoring and survey schemes, monitoring not only the status 

of seabird populations, but also as indicators for the state of the marine 

environment (Piatt et al. 2007; Parsons et al. 2008). 

Arctic Terns perform the longest known annual roundtrip migration in 

any animal (Egevang et al. 2010) and are surface-feeders that plunge-dive 

or surface-dip for small fi sh or invertebrates (Cramp 1985; Bianki and Isaksen 

2000; Hatch 2002). The Arctic Tern (Sterna paradisaea) usually breeds 

in land predator-free environments, such as islands and skerries, and is 

particularly good at defending the colony against avian predators. Arctic 

Tern breeding ground requirements are simple and the nest, with one, two 

or more rarely three eggs, is placed on the ground in a large variety of 

substrates and habitats. The annual reproductive outcome of Arctic Terns 

fl uctuates widely, with breeding failure over large areas reported (Avery et 

al. 1992; Boertmann et al. 1996). Furthermore, colony desertion within the 

breeding season and high chick mortality from starvation (Monaghan et 

al. 1989; Monaghan et al. 1992; Suddaby and Ratcliffe 1997) seem to occur 

on a more regular basis than in most other seabirds. 

Although site fi delity at a regional level is high in nesting Arctic Terns, 

dispersal to neighbouring breeding colonies occurs frequently (Devlin et 

al. 2008, Møller et al. 2006, Brindley et al. 1999, Ratcliffe 2004). This means 

that large variation in colony attendance between seasons is observed at 

some Arctic Tern colonies. Although poorly understood, it is important to 

address whether these variations may be caused by locally occurring phenomena, 

such as predation and disturbance, or whether they are linked to 

large-scale events such as food shortage linked to e.g. climatic variations. 

When interpreting counts of Arctic Tern colonies where variation occurs, 

it is furthermore important to address the scale (from within-colony to 

regional scale) at which the variations in colony attendance occur. 

A combination of these factors makes it particularly diffi cult to address 

population size, population status and trends in Arctic Terns (Ratcliffe 

2004). This is specifi cally the case in the Arctic zone where breeding sites 

may be diffi cult to access and colonies can be far apart. 

In Greenland, the Arctic Tern is a widespread breeder with its core distribution 

in West Greenland between 68º and 74º N (Boertmann et al. 

1996; Egevang and Boertmann 2003). Although no systematic monitoring 

schedule of Arctic Tern colonies has been initiated in Greenland, opportunistic 

counts during the last fi ve to six decades indicate a decline 

in the population (Salomonsen 1950; Egevang et al. 2004; Burnham et al. 

2005). This decline is assumed to be linked to non-sustainable egg harvesting, 

which has been a popular recreative activity carried out by local 

family groups. Arctic Tern egg harvesting was offi cially banned in 2002, 

and although some illegal harvesting still takes place in the Disko Bay 

87

88 

area, the magnitude of tern egg harvesting seems to have decreased markedly 

(own observations). Due to the lack of systematic counts, fl uctuations 

in colony size are poorly documented in Greenland, although years with 

non-breeding and breeding failure have been noted by several authors 

(Salomonsen 1950, Boertmann et al. 1996, Levermann and Tøttrup 2007, 

GSCD 2009). Although breeding failure (1992 – the summer following the 

Mount Pinatubo eruption) over a large scale has been documented on one 

occasion in Greenland (Boertmann et al. 1996), the scale and reasons for 

fl uctuating breeding are largely undocumented. 

In this paper we present the results of Arctic Tern colony surveys from 

four to fi ve consecutive seasons conducted in both small, mid and large 

sized colonies in two different regions of Greenland. In conclusion, a 

number of recommendations for future monitoring of Arctic Tern colonies 

in Greenland are listed. 

METHODS 

Study area 

Arctic Tern surveys were conducted in three study areas located in two 

different regions and climate zones of Greenland. 

1. Kitsissunnguit (Grønne Ejland) is located in the southern part of Disko 

Bay, West Greenland (68.85º N, 52.00º W). The site consists of four major 

islands (size: 0.44-3.55 km 2 ) and many islets and skerries (Fig. 1A). Kitsissunnguit 

holds the largest Arctic Tern colony in Greenland and probably 

one of the largest in the World (Cramp 1985, Bianki and Isaksen 

2000, Hatch 2002). The archipelago furthermore holds large diversity 

(by Arctic standards) and high densities of other breeding waterbirds, 

likely a result of breeding association between Arctic Terns and other 

waterbirds (Egevang et al. 2004; Jørgensen et al. 2007; Egevang and Boertmann 

2008). The coast of Kitsissunnguit is mostly rocky, whereas the 

interior is made up of large consecutive areas covered by dwarf scrub 

heath with Crowberry (Empetrum nigrum) and mosses (Bryophyta) as 

the dominant species. 

2. The “Akunnaaq area” is located in the southern part of Disko Bay, 

West Greenland (approximately 20 km southwest of Kitsissunnguit) 

between the town Aasiaat and the settlement of Akunnaaq (68.72- 

68.80º N, 52.20-52.73º W). The study area consists of numerous skerries 

and small islands (Fig. 1A), of which many hold breeding Arctic Terns. 

Fourteen colonies on islets between 0.002 and 0.2 km 2 in size (Tab. 1, 

Fig 1A) were counted annually between 2002 and 2005. The colonies 

were all low-lying with rocky coasts and dwarf bush heath found in the 

central parts of islets. 

3. Sand Island (Sandøen) is located within the high-arctic zone of Northeast 

Greenland, at the mouth of Young Sound (74.26º N, 20.16º W, Fig. 

1B). The small island (0.22 km 2 ) peaks at only three meters above sea 

level, with a habitat consisting of mostly sand and gravel with vegetation 

being extremely sparse. The surroundings of the island are highly 

infl uenced by sea ice. Break-up of the ice-covered fjord occurs in early 

July, but drift ice in varying amounts may be present throughout the 

breeding season.

Figure 1. Map of study areas. Inset map in upper left corner shows the location of A) Kitsissunnguit and the Akunnaaq Area in 

Disko Bay and B) of Sand Island in Northeast Greenland. The map scale (lower right corner) applies to both A and B. 

Table 1. Colony size (individuals) at 14 Arctic Tern colonies in the southern part of Disko Bay in 2002-2005. Colony code refers 

to the Greenland Seabird Colony Database (GSCD 2009). Asterisk (*) indicates “dog islands” – sites where sledge dogs are located 

through the summer. 

Colony code 2002 (18 June) 2003 (17 June) 2004 (19 June) 2005 (18 June) CV 2002 to 2005 (%) 

68013 2,000 0 650 1,800 85.4 

68142 10 30 0 0 141.4 

68155 600 0 600 1,050 76.6 

68166 30 0 0* 12 135.0 

68167 1,000 0 520 330 90.3 

68174 80 0 40 200 108.0 

68175 0* 0 0 0* – 

68188 30 0 0 650 188.4 

68189 55 0 260 0 156.9 

68190 300 0 400 50 103.0 

68191 200 0 140 400 89.7 

68192 100 0 1,000 300 128.8 

68193 300 0 0 100 141.4 

68194 0 1,200 430 950 83.2 

Total 4,705 1,230 4,040 5,842 49.6 

89

90 

Both Kitsissunnguit and the Akunnaaq area are situated within the core 

distribution area for Arctic Tern in Greenland, and neighbouring Arctic 

Tern colonies (not included in neither of the study sites) are few kilometres 

away. In contrast, Sand Island is more isolated and the nearest tern 

colony is more than 50 km away. 

Predators at all three study areas included Arctic Fox (Alopex lagopus), Glaucous 

Gull (Larus hyperboreus), Great Black-backed Gull (L. marinus), Common 

Raven (Corvus corax), Peregrine Falcon (Falco peregrinus), Gyrfalcon 

(Falco rusticolus) and occasionally Lesser Black-backed Gull (L. fuscus), Parasitic 

Jaeger (Stercorarius parasiticus) and Long-tailed Jaeger (S. longicaudus). 

Methods 

Due to differences in colony area and population size at the study sites 

two different methods were applied when estimating population size. 

1. The high number of breeding birds combined with the fact that terns 

at Kitsissunnguit breed scattered over a large consecutive area called 

for special census methods. Line transects (Buckland et al. 2001) recording 

the nest of Arctic Terns were applied on the major islands. Northsouth 

oriented lines 250 m apart were used in the fi eld with data subsequently 

analysed using the software Distance ver. 5.0 (Thomas et al. 

2006). Counts were always conducted by two persons: an observer and 

a navigator. The observer would walk along a transect line at an equal 

pace, searching for nests and measuring the distance between the nest 

and the transect line with a measuring tape. The navigator would walk 

approximately 15 meters behind the observer and using a GPS directing 

the observer to stay on the line. The line transect counts were conducted 

between 18 June and 3 July in 2002 -2006 at a time in the breeding 

cycle when the majority of the colony had stopped laying eggs but 

prior to the start of hatching. 

2. Direct counts were applied at the Akunnaaq area and at Sand Island. We 

used fl ush counts to estimate colony size, and the number of birds present 

at the colony was transformed into breeding pairs by dividing with a 

factor of 1.5 (Bullock and Gomersall 1981). The colonies were located on 

small islands that could be surveyed from the sea, and counts were performed 

with a boat as platform. The counts in the Akunnaaq area were 

conducted within a narrow time window (17-19 June), which correspond 

to mid incubation (Table 1). Roosting terns (i.e. larger fl ocks aggregated at 

the shoreline) were ignored in the population size estimate. Islands with 

sledge dogs (“dog islands”) were recorded on the same occasion. 

Counts at Sand Island were conducted between 20 and 26 July (mid-late 

incubation). The island was divided into segments that were counted 

(fl ush counts) separately by walking through the colony on foot. 

We used permutation tests to evaluate whether regional annual total nest 

counts were less variable than expected if colonies fl uctuated independently, 

which would indicate between-colony movements. Within each 

colony, annual counts were permuted randomly 9,999 times, simulated 

regional annual totals and their variance (or equivalently CV) were calculated, 

and a one-tailed P value determined as the proportion of simulated 

totals having lower variance than the observed regional annual totals.

RESULTS 

The overall population size at Kitsissunnguit (all islands combined) 

showed little variation (CV=13.9 %) between ye ars with a population size 

between 15,400–21,800 pairs in 2002-2006 (Table 2, details in Appendix I). 

However, at individual islands a more pronounced variation (Tab. 2) was 

recorded with differences of almost a factor three in breeding population 

(e.g. Niaqornaq 2002 vs. 2003). Furthermore, the island Innarsuatsiaaq 

showed two years without terns followed by two years with low densities 

of breeding terns. The breeding density at Kitsissunnguit (all years and all 

islands combined) averaged 4,823 pairs/km 2 (+1,483). At a smaller scale 

(comparing between the islands), breeding densities varied between 7,213 

and 629 pairs/km 2 (Appendix I). 

The fourteen colonies in the Akunnaaq area showed considerable differences 

in population size between years, with coeffi cient of variation 

ranging from 76.6 % to 188.4 % (Table 1), whereas CV of the colonies combined 

was lower at 49.6 %. Only one site (68175) did not include breeding 

terns in any of the surveyed years, whereas the other thirteen colonies 

had breeding terns in at least two of surveyed years. All of the colonies 

experienced years without breeding terns (most of them in 2003), and the 

largest absolute variation in colony size was found in the largest colony 

(68013), ranging from 0 to 2,000 individuals (Tab. 1). When ignoring years 

without breeding terns (0-years) the average difference in maximum and 

minimum individuals in the colonies were 372 (SD = 387) with two colonies 

varying by as much as a factor 10 (68192) or 20 (68188) between years. 

The permutation test showed no evidence of lower regional variation 

within each of the two regions (Kitsissunnguit P = 0.20, Akunnaaq P = 0.90), 

but when the two regions were pooled to form an overall Disko Bay total, 

variance and thus CV were lower than expected if individual colonies 

fl uctuated independently (P = 0.023). 

At Sand Island, variation in breeding numbers was more pronounced. 

Both 2006 and 2009 were years with total breeding failures, whereas the 

breeding population in 2007 and 2008 was estimated at between 700 and 

1000 pairs, based on counts varying between 1150 and 1500 individuals. 

In the two years with breeding failure, birds returned to the colony at the 

start of the season, but egg-laying was either postponed or laid eggs depredated, 

until breeding was eventually abandoned. 

Table 2. Arctic Tern population size (pairs) at Kitsissunnguit (Grønne Ejland) in 2002-2006. Population size was estimated using 

line transects. Details on estimates are found in Appendix I. 

2002 2003 2004 2005 2006 CV 2002-2005 (%) CV 2002-2006 (%) 

Angissat 0 0 0 0 0 – – 

Innarsuatsiaaq 0 0 776 6001 8001 68.7 93.0 

Basis Ø 11,525 17,626 15,901 12,112 17,060 15.8 19.1 

Niaqornaq 3,159 1,245 1,984 2,346 2,5001 27.7 31.3 

Saattuarsuit 6701 6701 5301 1,4201 1,4001 46.9 46.3 

Total 15,354 19,541 19,191 16,478 21,760 11.6 13.9 

1Estimate from direct counts 

91

92 

Although the breeding numbers in 2007 and 2008 were similar, the distribution 

of tern nests differed between the two seasons. In 2007 birds bred 

scattered over most of the island, whereas in 2008 the majority (estimated 

80%) of the terns bred at the elevated central part of the island (Egevang 

& Stenhouse 2008, Egevang et al. 2009). In 2007/2008 the breeding density 

equalled 2,574-3,676 pairs per km 2 on Sand Island. 

DISCUSSION 

The information obtained from comparing surveys in this study shows 

that a high degree of variation in population size between years. The highest 

coeffi cient of variation was found at colony level, whereas lower variation 

was found for the two study areas, Kitsissunnguit and the Akunnaaq 

Area (Tab. 1 and 2). When combining the two study areas, which include 

more than 80 % (GSCD 2009) of the Arctic Tern population in Disko Bay, 

to address annual population size variations at a large scale, a CV of only 

6.7% was found with the grand total ranging between 20,000 and 23,000 

pairs (Tab. 3), a level of variation that was smaller than expected if colonies 

fl uctuated independently. This led us to conclude that year-to-year 

variation in the Arctic Tern population in Disko Bay is not governed by 

large scale phenomena. Instead, the switching of colonies between seasons 

is more likely driven by more locally occurring phenomena, such as 

predation or disturbance. However, in 2003 it seems likely that mesoscale 

movements from the southern colonies (Akunnaaq area) to Kitsissunnguit 

may have taken place. 

Variation was also found among colonies in the distribution and density 

of the breeding birds. The highest and most pronounced variation was 

found amongst the smaller and midsized colonies in the Akunnaaq area. 

Here colonies with up to 2,000 breeding individuals one year could be 

followed by years without or with very few pairs breeding. However, 

tern colonies are located densely in the southern part of Disko Bay, and it 

is likely that the local terns may shift between suitable nesting grounds. 

Even in 2003, which seemed to be a poor overall breeding season in the 

Akunnaaq area, a site (Pullat, code 68148) less than 10 km outside the 

Akunnaaq study area held 3,000 breeding Arctic Terns. Although direct 

evidence that the same individuals move around between colonies from 

year to year in the (southern) Disko Bay area is diffi cult to obtain, the 

circumstantial evidence of this study indicates that this is in fact the case. 

At the high-arctic Sand Island, the short summer leaves only a brief window 

of opportunity for completing the breeding cycle. The four years of 

data included in this study indicate that complete breeding failure takes 

place at a regular basis (two out of four seasons). When the birds arrive to 

breed between late June and early July, sea ice still covers the surroundings 

of the island connecting it with the mainland. From here, Arctic Foxes are 

able to access the island and cause considerable nest predation. The time 

for ice break-up and isolation of the island from land predators may vary 

between years, but usually takes place in early July (ZERO 1997-2009). 

The presence of a predator on the island in egg–laying period seemed to 

make the birds postpone their egg laying, as suggested by Levermann 

and Tøttrup (2007), or more likely to postpone the decision of investing in 

a replacement clutch after having the fi rst clutch depredated early in the 

season.

Table 3. Grand total (from table 1 and 2) and coeffi cient of variation (CV) from Arctic colony 

censuses in Disko Bay 2002-2005. 

The fl uctuating occurrence of Arctic Terns between breeding seasons 

stressed by this study clearly emphasises the need for a specialized monitoring 

programme for this species. Monitoring of Arctic Terns in single 

colonies conducted in single years does not provide a reliable estimate of 

the population status, as Arctic Terns show a lesser degree of site fi delity 

compared with most other Arctic seabird species. The use of counts 

of breeding birds or nests in fi xed study plots for later extrapolation is 

furthermore questionable, as within-colony movements of the location of 

nest along with variation in breeding density between years seems to occur 

at a regular basis. 

Future monitoring of Arctic Tern populations in Greenland is recommended 

to apply one of the following two approaches (or ideally a combination 

of both): 1) Survey of multiple colonies within the same season covering 

adjoining colonies over a larger area, or 2) survey of one large, representative 

colony in multiple years. However, since logistics in Greenland are 

both expensive and time consuming, the fi rst approach seems most cost 

effective and likely to produce a suitable estimate for managing purposes. 

In terms of the survey method, this will probably be determined by the 

size, structure and breeding density of the colony. Smaller and mid-sized 

colonies up to 2,000 birds breeding on skerries and islets that can be 

surveyed from the sea are best counted using a boat as platform. Large 

colonies with higher number of birds are recommended to be surveyed 

by counts in subdivisions of the colony (preferably delimited by easily 

recognisable, natural boundaries in the fi eld). Large colonies with high 

breeding numbers and nests evenly distributed at low or medium densities 

should be surveyed using line transects or a similar approach. 

The period for conducting surveys should be timed to coincide with late 

incubation. As timing of egg laying may vary considerably between seasons 

and locations, it is important to address the state of the breeding 

cycle before conducting the counts. This is particularly the case for line 

transects, where recording of tern nests need to be conducted when the 

majority of the birds have started incubation but before eggs starts hatching 

and chicks can move away from the nest area. 

ACKNOWLEDGEMENTS 

2002 2003 2004 2005 CV (%) 

Kitsissunnguit 15,354 19,541 19,191 16,478 11.6 

Akunnaaq area 4,705 1,230 4,040 5,842 49.6 

Grand Total 20,059 20,771 23,231 22,320 6.7 

The fi eldwork was fi nancially supported by the Danish Energy Agency 

(the climate support program to the Arctic) and the Commission for Scientifi 

c Research in Greenland (KVUG). Special thanks go to the people who 

assisted in the fi eld over the years: David Boertmann and Morten Bjerrum 

(National Environmental Research Institute, Denmark), Anders Tøttrup, 

Mikkel Willemoes and Peter S. Jørgensen (University of Copenhagen), 

93

94 

Ole S. Kristensen, Lars Witting and Rasmus Lauridsen (GNIR), Bjarne Petersen, 

Qasigiannguit and Norman Ratcliffe and Stuart Benn (Royal Society 

for Protection of Birds). Also thanks to Finn Steffens (Qeqertarsuaq) 

and Elektriker-it (Aasiaat) for logistic support. 

LITERATURE CITED 

Avery, M.I., D. Suddaby, P.M. Ellis and I.M. W. Sim. 1992. Exceptionally low bodyweights 

of Arctic Terns Sterna paradisaea on Shetland. Ibis 134: 87-88. 

Bianki, V.V. and K. Isaksen. 2000. Arctic Tern Sterna paradisaea. In The status of marine 

birds breeding in the Barents Sea region. Anker-Nilssen. T. et al. Tromsø, Norsk 

Polarinstitutt Rapportserie. 113: 131-136. 

Boertmann, D., A. Mosbech, K. Falk and K. Kampp. 1996. Seabird colonies in western 

Greenland (60 º– 79º 30’ N lat.). National Environmental Research Institute, Denmark. 

Technical Report 170, NERI, Roskilde. 

Brindley E., G. Mudge, N. Dymond, C. Lodge, B. Ribbands, D. Steele, P. Ellis, E. 

Meek, D. Suddaby and N. Ratcliffe. 1999. The status of Arctic Terns Sterna paradisaea 

at Shetland and Orkney in 1994. Atlantic Seabirds 1, 135-143. 

Buckland, S.T., D.R. Anderson, K.P. Burnham, J.L. Laake, D.L. Borchers and L. Thomas. 

2001. Introduction to Distance Sampling - Estimating abundance of biological 

populations. Oxford, Oxford University Press. 

Bullock, I.D. and C.H. Gomersall. 1981. The breeding population of terns in Orkney 

and Shetland in 1980. Bird Study 28: 187-200. 

Burnham, W., K.K. Burnham and T.J. Cade. 2005. Past and present assessments of 

bird life in Uummannaq District, West Greenland. Dansk Ornitologisk Forenings 


Coulson, J.C. 2002. Colonial breeding in seabirds Biology of Marine Birds. E.A. 

Schreiber and J. Burger. New York, CRC Press: 87-113. 

Cramp, S., Ed. 1985. Terns to woodpeckers. Birds of the Western Palearctic. Oxford, 


Devlin, C.M., A.W. Diamond, S.W. Kress, C.S. Hall and L. Welch. 2008. Breeding 


Maine. The Auk 125, 850-858. 

Egevang, C. and D. Boertmann. 2003. Havternen i Grønland – Status og undersøgelse 

2002. National Environmental Research Institute, Denmark. Technical Report 438, 

NERI, Roskilde. 

Egevang, C. and D. Boertmann. 2008. Ross’s Gulls (Rhodostethia rosea) Breeding in 

Greenland: A Review, with Special Emphasis on Records from 1979 to 2007. Arctic 

61(3): 322-328. 

Egevang, C., K. Kampp and D. Boertmann. 2004. The Breeding Association of Red 



Egevang, C. and I.J. Stenhouse. 2008. Mapping long-distance migration in two Arctic 

seabird species p. 74-76. In: Klitgaard, A.B. and Rasch, M. (eds.) 2008. Zackenberg 

Ecological Research Operations, 13 th Annual Report, 2007. – Copenhagen, Danish 

Polar Center, Danish Agency for Science, Technology and Innovation, Ministry of 

Science, Technology and Innovation.

Egevang, C., I.J. Stenhouse, L.M. Rasmussen, M.W. Kristensen and F. Ugarte 2009. 

Breeding and foraging ecology of seabirds on Sandøen 2008. In: Jensen, L.M. & 

Rasch, M. (eds.) 2009. Zackenberg Ecological Research Operations, 14 th Annual 

Report, 2008. National Environmental Research Institute, Aarhus University, Denmark. 

116 pp. 

Egevang, C., I.J. Stenhouse, R.A. Phillips, A. Petersen, J.W. Fox & J.R.D. Silk 2010. 

Tracking of Arctic terns Sterna paradisaea reveals longest animal migration. – Proceedings 

of the National Academy of Sciences of the United States 107: 2078-2081. 

GSCD. 2009. The Greenland Seabird Colony Database. http://www.dmu.dk/ 

Greenland/Olie+og+Miljoe/Havfuglekolonier/ accessed October 2009. 

Hatch, J. 2002. Arctic Tern Sterna paradisaea. The Birds of North America 707. Philadelphia, 

USA. 40 pp. 

Jørgensen, P.S., M.W. Kristensen and C. Egevang. 2007. Red Phalarope Phalaropus 

fulicarius and Red-necked Phalarope Phalaropus lobatus behavioural response 

to Arctic Tern Sterna paradisaea colonial alarms. Dansk Oritologisk Forenings 


Levermann, N. and A.P. Tøttrup. 2007. Predator Effect and Behavioral Patterns in 

Arctic Terns (Sterna paradisaea) and Sabine’s Gulls (Xema sabini) During a Failed 

Breeding Year. Waterbirds 30(3): 417-420. 

Møller, A.P., E. Flensted-Jensen and W. Mardal. 2006. Dispersal and climate change: a 

case study of the Arctic Tern Sterna paradisaea. 

doi:10.1111/j.1365-2486.2006.01216.x. Global Change Biology 12(10): 2005-2013. 

Monaghan, P., J.D. Uttley and M.D. Burns. 1992. Effect of changes in food availability 

on reproductive effort in Arctic Terns Sterna paradisaea. Ardea 80(1): 71-81. 

Monaghan, P., J.D. Uttley, M.D. Burns, C. Thaine and J. Blackwood. 1989. The relationship 

between food supply, reproductive effort and breeding success in Arctic 

Terns Sterna paradisaea. Journal of Animal Ecology 58: 261-274. 

Parsons, I. Mitchell, A. Butler, N. Ratcliffe, M. Frederiksen, S. Foster and J.B. Reid. 

2008. Seabirds as indicators of the marine environment. ICES Journal of Marine 

Science 65: 1520–1526. 

Piatt, J.F., W.J. Sydeman and F. Wiese. 2007. Seabirds as indicators of marine ecosystems. 

Marine Ecology Progress Series 352: 199–204. 

Ratcliffe, N. 2004. Arctic Tern Sterna paradisaea. Seabird Populations of Britain and 

Ireland. P.I. Mitchell, S.F. Newton, N. Ratcliffe and T.E. Dunn. London, T & A D 

Poyser: 328-338. 

Salomonsen, F. 1950. The Birds of Greenland. Copenhagen, Munksgaard. 

Suddaby, D. and N. Ratcliffe. 1997. The Effects of Fluctuating Food Availability on 

Breeding Arctic Terns (Sterna paradisaea). Auk 114(3): 524-530. 

Thomas, L., J.L. Laake, S. Strindberg, F.F.C. Marques, S.T. Buckland, D.L. Borchers, 

D.R. Anderson, K.P. Burnham, S.L. Hedley and J.H. Pollard. 2006. Distance. UK, 

University of St. Andrews, UK. 

Weimerskirch, H. 2002. Seabird demography and its relationship with the marine environment. 

Biology of Marine Birds. E.A. Schreiber and J. Burger. New York, CRC 

Press: 115-135. 

ZERO 1997-2009. Zackenberg Ecological Research Operations annual reports in the 

period 1997-2009. 

95

Pooled data 

96 

Appendix I 

Detailed results from line transect counts conducted on Kitsissunnguit, 

2002 to 2006. Data on search effort from the islands in the archipelago are 

joint (pooled data) while point estimates of density (pairs per km 2 ) with 

standard error (SE), confi dence variance (CV) and confi dence interval (CI) 

are given for each of the individual islands. Data were analysed using 

software Distance 5.0 (Thomas et al. 2006) applying a half normal cosine 

detection function with a 3% right truncation. 

2002 2003 2004 2005 2006 

ESW1 (SE) 1.9–(0.12) 2.3–(0.14) 3.0–(0.11) 3.3–(0.18) 2.6–(0.15) 

f (0) 2 (SE) 5.23–(0.319) 4.37–(0.270) 3.28–(0.126) 3.03–(0.169) 3.75–(0.213) 

Effort3 L(n) 8.6–(11) 9.4–(12) 17.8–(26) 8.9–(11) 7.7–(8) 

Observations (nests) 139 228 433 232 227 

Basis Ø West (0.90 km 2 ) 

Density (SE) 4199–(1253.1) 6587–(1690.9) 4049–(2195.8) 3600–(1233.7) 5746–(2097.9) 

Population size (SE) 3785–(1129.4) 5938–(1524.1) 3650–(1979.0) 3244–(1111.6) 5177–(1890.0) 

CV (%) 29.84 25.67 54.22 34.27 36.51 

95% CI 177–81101 2791-12635 732-18198 1158-9084 1731-15487 

Basis Ø East (2.22 km 2 ) 

Density (SE) 3483–(518.8) 5261–(651.1) 5514–(964.4) 3989–(322.4) 5345–(853.7) 

Population size (SE) 7740–(1152.8) 11688–(1446.6) 12251–(2142.7) 8868–(716.7) 11883–(1897.7) 

CV (%) 14.89 12.38 17.49 8.08 15.97 

95% CI 5368-11160 8561-15958 8264-18163 7422-10596 7627-18513 

Niaqornaq (0.45 km 2 ) 

Density (SE) 7040–(2795.0) 2665–(1119.7) 4420–(1466.2) 5213–(1502.1) 

Population size (SE) 3159–(1254.1) 1245–(523.1) 1984–(658.0) 2346–(503.3) 

CV (%) 39.70 42.01 33.17 28.81 

95% CI 965-10342 356-4357 721-5461 562-5425 

Innersuatsiaaq (1.23 km 2 ) 

Density (SE) 629–(340.8) 

Population size (SE) 776–(420.1) 

CV (%) 54.14 

95% CI 242-2494 

1Effective strip width (m) 

2 -3 Probability density function (10 ) 

3Effort = length (L) of lines (in km) and numbers (n) of lines

Manus V 

RED PHALAROPE PHALAROPUS FULICARIUS 

AND RED-NECKED PHALAROPE 

PHALAROPUS LOBATUS 

BEHAVIOURAL RESPONSE TO ARCTIC TERN 

STERNA PARADISAEA COLONIAL ALARMS 

Peter Søgaard Jørgensen 1 , Mikkel Willemoes Kristensen 2 and Carsten Egevang 3 

1 Hovmålvej 80 st. -15 

DK-2300 Copenhagen S, Denmark 

(psjoergensen@hotmail.com) 

2 Langøgade 10, 1. tv. 

DK-2100 Copenhagen Ø, Denmark 

3 Greenland Institute of Natural Resources, 

P.O. Box 570, 3900 Nuuk, Greenland 

Dansk Ornitologisk Forenings Tidsskrift 101 (3): 73-78, 2007

98 

Red Phalarope Phalaropus fulicarius and Red-necked 

Phalarope Phalaropus lobatus behavioural response to 

Arctic Tern Sterna paradisaea colonial alarms 

PETER SØGAARD JØRGENSEN, MIKKEL WILLEMOES KRISTENSEN and CARSTEN EGEVANG 

(Med et dansk resumé: Thorshanes og Odinshanes respons på alarmer i en havternekoloni) 

Abstract The breeding association between Arctic Terns and Red and Red-necked Phalaropes has been mentioned 

several times in the ornithological literature, though mostly anecdotally. The present study, conducted 

at the Kitsissunnguit archipelago in western Greenland, investigates phalarope behavioural response to Arctic 

Tern colonial alarms, specifically the relationship between type and duration of tern alarms and phalarope 

response type. The majority of tern alarms elicited a visible response from the phalaropes. No difference in 

phalarope response type was detected between true alarms (tern responses to the presence of gulls, falcons 

or humans) and dreads (false alarms). Tern alarms to which the phalaropes reacted were of longer duration than 

alarms that caused no change in phalarope behaviour, whereas no relationship was found between tern alarm 

duration and phalarope response type (passive vs active). In conclusion, phalaropes respond strongly to tern alarms, 

but do not differentiate between predator-related alarms and dreads. These findings support the view that phalaropes, 

where possible, take advantage from Arctic Tern colonies, and that this relationship – fulfilling the criteria of 

both the information parasitism and the defence parasitism hypotheses – explains the high densities of phalaropes 

at Kitsissunguit. 

Introduction 

Beneficial breeding associations are well known 

in multi-species colonies of birds. Many examples 

are known among colonial waterbirds (e.g., Alberico 

et al. 1991). Typically, such relationships form 

between a semi-aggressive, bold species (often a 

larid) and a more timid species of waterbird, where 

the latter benefits from the strong colony defence 

Dansk Orn. Foren. Tidsskr. 101 (2007): 73-78 

of the bold species. In most of these cases, however, 

the details of the relationship are unknown and 

the mechanism through which the timid species 

benefits from the breeding association has not 

been documented. 

In order to obtain a higher individual survival 

and reproductive output, a timid species can benefit 

from breeding associations with more aggressive 

73

74 Phalarope response to tern alarms 

species if the bold species provides a higher predator 

detection rate and/or predator detection at a greater 

distance, giving the timid species more time to react 

(the information parasitism hypothesis, Nuechterlein 

1981); or if the bold species aggressively defends 

the surroundings of its nest and thereby indirectly 

provides protection to the timid species (defence 

parasitism hypothesis, Dyrcz et al. 1981). 

Arctic Terns Sterna paradisaea have been 

known to associate with high breeding densities 

of several species of shorebirds, including Red-necked 

Phalarope Phalaropus lobatus and Red Phalarope 

Phalaropus fulicarius (Løvenskiold 1964, Hildén 

& Voulanto 1972, Cramp & Simmons 1983, 

Egevang et al. 2004), Ruddy Turnstone Arenaria 

interpres (Brearey & Hildén 1985) and Purple 

Sandpiper Calidris maritima (Summers & Nicoll 

2004). 

The present study focuses on the association between 

Arctic Terns and phalaropes in a major Arctic 

Tern colony in western Greenland. Using relatively 

simple observations of phalarope response to tern 

alarms and dreads (false alarms), we investigate 

whether phalaropes use information parasitism to 

benefit from the terns, and which behavioural 

responses they express. 

Study area and methods 

Fieldwork was conducted at the Kitsissunnguit 

(Grønne Ejland) archipelago in the southern part 

of Disko Bay, West Greenland (68˚50'N, 52˚00'W) 

in the period 2-28 July, 2006. Kitsissunnguit is 

the largest Arctic Tern colony in Greenland with 

16000-22000 estimated breeding pairs in 2002- 

2006 (Egevang et al. 2004; this study). By Arctic 

standards, the site holds a rich avifauna with 

25 breeding species recorded, some of them rare 

in West Greenland (e.g., Red Phalarope, Long-tailed 

Skua Stercorarius longicaudus, Ross's Gull 

Rhodostethia rosea, cf. Kampp & Kristensen 

1980, Kampp 1982, Frich 1997). The major avian 

predators are Glaucous Gull Larus hyperboreus, 

Great Black-backed Gull Larus marinus, Gyrfalcon 

Falco rusticolus, Peregrine Falcon Falco peregrinus 

and Raven Corvus corax. Arctic fox 

Alopex lagopus also occur at the archipelago 

where it occasionally breeds, and although Arctic 

Terns and foxes may co-exist at Kitsissunnguit 

for a few years (Kampp 1982), the terns probably 

avoid the easternmost island in the archipelago 

because foxes have been established there for 

decades (Egevang et al. 2004). The coasts of the 

islands are generally rocky, whereas the inland is 

covered by dwarf scrub heath with willow Salix 

spp., dwarf birch Betula nana and crowberry 

Empetrum nigrum as the dominant plant species. 

There are many moist areas, ponds and small 

lakes with sedge Carex spp. and cotton grass 

Eriophorum spp. 

The current study was conducted at Basisø, a 

fox-free island of approximately 3.5 km 2 . In 2006 

there were 45 pairs of Red-necked Phalaropes and 

7 of Red Phalaropes on the island, which numbers 

may underestimate the actual number of nests 

because females may be polyandrous (Cramp & 

Simmons 1983). 

Observations of phalarope and tern behaviour 

were conducted from the edge of two small ponds. 

All phalarope individuals present at the ponds 

during an observation period were kept under observation. 

The three main behaviours performed 

by phalaropes outside tern alarm periods were 

feeding, resting and preening. Observation periods 

varied from one to three hours, in total covering 30 

hours, and were evenly distributed over the study 

period. Initially the phalaropes preferred a pond of 

c. 2500 m 2 , but after the first week the behaviour 

of the birds changed and, apparently as a result of 

the fledging of the chicks, they began to frequent 

a new pond adjacent to the first. Both sites offered 

a good view of the water surface and the brinks. 

Apart from the phalaropes and a high number of 

Arctic Terns, the breeding species in the surrounding 

area were Mallard Anas platyrhynchos, Longtailed 

Duck Clangula hyemalis, and Red-breasted 

Merganser Mergus serrator. 

Both phalarope species were included in the 

study but no distinction between them were made, 

due to their very similar behaviour and response to 

aerial predators. The dataset mainly includes local 

breeders but, as the season advanced, these became 

mixed with migrants and with local juveniles. 

The following behavioural patterns were recorded: 

Tern alarms. The duration of tern alarms, defined 

as mass up-flight by at least half of the terns in 

the study area, was recorded. The type of alarm 

was categorized as a) originating from disturbance 

from gull or falcon, b) originating from human 

disturbance, or c) dreads (false alarms). 

Phalarope response to tern alarms. During tern 

alarms the behavioural response of the phalaropes 

was recorded. Six different types of response were 

identified: 1) no response (no change in behaviour), 

2) tucking to water surface (Tucking), 3) stopping 

ongoing behaviour and/or looking up (Stop/looking 

up), 4) take-off from pond (Upflight), and 5) 

99

100 

swimming out on pond surface or 6) to the brink 

(Swim surface/brink). Phalarope response during 

each tern alarm sometimes included more than one 

of these types, but only the predominant one was 

included in the statistical analysis. 

Response types 2) and 3) were classified as 

passive, types 4), 5) and 6) as active. 

The frequency distributions of phalarope responses 

were statistically tested using c 2 -tests. 

Mann-Whitney U-tests were used to test the 

difference in duration of tern alarms grouped on 

basis of phalarope response. 

Results 

The terns responded strongly to predators within 

the boundaries of the colony, and the presence 

of aerial predators resulted in fierce mobbing 

behaviour involving hundreds, on a few occasions 

thousands of birds. The type and intensity of the 

vocalisation of the terns varied somewhat, depending 

on the predator species, so that at least three 

types of alarm calls (falcon, gull, human) could be 

distinguished in the field. The mean duration of 

gull/falcon alarms was 30.3 s (n=17), which is longer 

than the mean duration of dreads (23.0 s, n=44) 

(Mann-Whitney U-test, U=239.5, P


Discussion 

The study on phalarope behavioural response to 

Arctic Tern alarms at Kitsissunnguit reveals a close 

association between the species. The phalaropes 

reacted instantly to tern alarms in 84% of the recorded 

alarms, indicating a one-way transfer of 

information from terns to phalaropes, as implied 

by the information parasitism hypothesis. At large 

Arctic Tern colonies (such as Kitsissunnguit) a 

predator will be detected at a considerable distances, 

and the intensity of the alarm will increase as the 

predator travels through the colony. This provides 

terns and other birds in the colony with an early 

warning system, passing on information on direction, 

distance and species of the approaching predator. 

This early warning gives the breeding birds the 

time necessary to perform anti-predator behaviour 

and warn offspring to seek cover. 

Spontaneously occurring dreads without an 

obvious cause are common in tern colonies. The 

reason is not known; among the proposed causes 

is that it may contribute to the maintainance of 

pair bonds (Cullen 1960), and irrespective of 

the underlying cause the dreads may serve as an 

indicator of the frequency of predator visits to the 

colony (Meehan & Nisbet 2002). In the present 

study more than two thirds of the recorded alarms 

were dreads. The fact that phalaropes did not 

show any difference in behavioural response to 

dreads versus alarms involving predators indicates 

that the phalaropes react to tern alarms before they 

have visually identified a potential predator. 

Tern alarm intensity and duration depends on 

the predator species, but also on how close the 

predator is to a given part of the colony and of the 

behaviour exhibited by the predator. For example 

will low hunting flights result in much more 

intense alarms from the terns than straight transits 

above the colony. These factors may also contribute 

to explaining why phalaropes apparently do not 

differentiate between real tern alarms and dreads, 

even though predatory alarms on average last 

longer than dreads, and phalaropes respond more 

frequently to longer alarms. 

The dominance of passive responses to active 

ones may serve to lower depredation on eggs or 

chicks, since it allows the phalarope to stay in its 

territory, whereas the upflight response means that 

it has to leave, even in the case of a false alarm 

(Alberico et al. 1991). 

The mean frequency of tern upflights (2.6 hour -1 ) 

is comparable to the mean rate of 2 hour -1 quoted 

by Cramp (1985); it may, however, vary through 

the breeding period (Cramp l.c.). The frequency 

of gull/falcon alarms at Kitsissunnguit (0.26 

hour -1 ) is lower than the about 4 hour -1 – caused 

by Herring Gull alone – found during the hatching 

and chick rearing stage in a small (505 pairs) mixed 

tern colony in Canada (Whittam & Leonard 1999). 

This may indicate that predator numbers are 

relatively low at Kitsissunnguit. The many terns 

in the colony form an effective defence towards 

intruders such as Ravens, gulls and skuas. However, 

falcons are able to outfly the terns and are less 

affected by their mobbing (Egevang et al. 2005). 

Egevang et al. (2004) documents that a marked 

decline in phalarope numbers took place after 

Arctic Terns abandoned the easternmost island in 

the Kitsissunnguit archipelago, Angissat. In addition, 

they discuss whether the unusually high breeding 

density of Red-necked Phalarope – and the presence 

of the rare Red Phalarope – at Kitsissunnguit is 

a direct result of the presence of a major tern colony, 

i.e., if the distribution of terns determines the 

distribution of phalaropes, or if both terns and 

phalaropes simply prefer a fox-free environment 

so that the presence or absence of foxes determines 

the distribution of both terns and phalaropes. The 

results of the present study verify the strong 

behavioural response of phalaropes to tern alarms, 

showing that the relationship between the species 

fulfils the criteria of both the information parasitism 

and the defence parasitism hypothesis, and thereby 

strengthen the first hypothesis, without being able 

to reject the second. 


We wish to express our gratitude to Rasmus Lauridsen, 

Greenland Institute of Natural Resources, for excellent 

company in the field, and to Finn Steffens, captain of m/v 

Maja S of Qeqertarsuaq, for logistic support. The fieldwork 

in 2006 was financed by the Greenland Institute 

of Natural Resources and the Commission for Scientific 

Research in Greenland (KVUG). 

Resumé 

Thorshanes og Odinshanes respons på alarmer i en 

havternekoloni 

Den positive ynglerelation mellem svømmesnepperne 

Thorshane Phalaropus fulicarius og Odinshane P. lobatus 

på den ene side, og Havternen Sterna paradisaea på den 

anden, er velkendt, men er oftest behandlet rent anekdotisk 

i den ornitologiske litteratur. Fra den 2. til 28. juli 2006 

undersøgte vi de to svømme sneppers adfærdsrespons på 

kollektive havternealarmer på øgruppen Kitsissunnguit 

(Grønne Ejland) i Disko Bugt i Grønland. Undersøgelsen 

blev foretaget ved relativt simple observationer af typen 

101

102 

og længden af havternealarmerne og typen af svømmesneppernes 

respons herpå. 

Hyppigheden af havternealarmer på hvilke svømmesnepperne 

responderede var signifikant højere end alarmer 

uden respons. Der blev ikke fundet nogen forskel 

mellem falske ternealarmer og ternealarmer forårsaget 

af måger, falke eller mennesker hvad angår hyppighed 

og fordeling af manglende respons, passiv respons, og 

aktiv respons. Alarmer af den sidstnævnte type var dog af 

signifikant længere varighed end falske alarmer. Ternealarmer 

med svømmesneppe-respons varede i gennemsnit 

længere end alarmer uden respons, mens der ikke 

var nogen tilsvarende signifikant forskel på længden af 

ternealarmer med henholdsvis aktiv og passiv svømmesneppe-respons. 

Resultaterne viser, at svømmesnepperne reagerer markant 

på ternealarmer, og at de i deres respons ikke skelner 

mellem rovdyrsrelaterede alarmer og falske alarmer. 

Resultaterne understøtter formodningen om, at den 

høje yngletæthed af Odinshane – og tilstedeværelsen af 

Thorshane – på Kitissunnguit skyldes at ynglende svømmesnepper 

tiltrækkes af havternekolonier. 

Phalarope response to tern alarms 77 

Phalaropes breeding in tern colonies use tern alarms (upflights and calls) as warning signals indicating the presence 

of predators, but are unable to distinguish between true and false alarms. Photo: Mikkel Willemoes Kristensen. 

References 

Alberico, J.A.R., J.M. Reed & L.W. Oring 1991: Nesting 

near a Common Tern colony increases and decreases 

Spotted Sandpiper nest predation. – Auk 108: 904- 

910. 

Brearey, D. & O. Hildén 1985: Nesting and egg-predation 

by Turnstones Arenaria interpres in larid colonies. 

– Ornis Scand. 16: 283-292. 

Cramp, S. & K.E.L. Simmons (eds) 1983: The birds of 

the western Palearctic. Vol. 3. – Oxford University 

Press. 

Cramp, S. (ed.) 1985: The birds of the western Palearctic. 

Vol. 4. – Oxford University Press. 

Cullen, J.M. 1960: The aerial display of the Arctic Tern 

and other species. – Ardea 48: 1-37. 

Dyrcz, A., J. Witkowski & J. Okulewicz 1981: Nesting of 

timid waders in the vicinity of bold ones as an antipredator 

adaptation. – Ibis 123: 542-545. 

Egevang, C., K. Kampp & D. Boertmann 2004: The 

breeding association of Red Phalaropes with Arctic 

Terns: response to a redistribution of terns in a major 

Greenland colony. – Waterbirds 27: 406-410. 

Egevang, C., D. Boertmann & O. Stenderup Kristensen 

2005: Monitering af havternebestanden på Kitsis-


sunnguit (Grønne Ejland) og i den sydlige del af Disko 

Bugt, 2002-2004. – Techn. rep. 62, Pinngortitaleriffik/ 

Grønlands Naturinstitut. 

Hildén, O. & S. Voulanto 1972: Breeding biology of the 

red-necked phalarope Phalaropus lobatus in Finland. 

– Ornis Fenn. 49: 57-85. 

Kampp, K. 1982: Notes on the Long-tailed Skua Stercorarius 

longicaudus in West Greenland. – Dansk Orn. 

Foren. Tidsskr. 76: 129-135. 

Kampp, K. & R.M. Kristensen 1980: Ross's Gull Rhodostethia 

rosea breeding in Disko Bay, West Greenland, 

1979. – Dansk Orn. Foren. Tidsskr. 74: 129-135. 

Løvenskiold, H.L. 1964: Avifauna Svalbardensis. – Norsk 

Polarinstitut, Oslo. 

Meehan, T.D. & I.C.T. Nisbet 2002: Nest attentiveness 

in common terns threatened by a model predator. 

– Waterbirds 25: 278-284. 

Nuechterlein, G.L. 1981: Information parasitism in 

mixed colonies of Western Grebes and Forster's Terns. 

– Anim. Behav. 29: 985-989. 

Summers, R.W. & M. Nicoll 2004: Geographical variation 

in the breeding biology of the Purple Sandpiper 

Calidris maritima. – Ibis 146: 303-313. 

Whittam, R.M. & M.L. Leonard 1999: Predation and 

breeding success in Roseate Terns. – Can. J. Zool 77: 

851-856. 

Accepted 2 June 2007 

Peter Søgaard Jørgensen (psjoergensen@hotmail.com) 

Hovmålvej 80 st. - 15, 

DK-2300 Copenhagen S, Denmark. 

Mikkel Willemoes Kristensen 

Langøgade 10, 1. tv. 

DK-2100 Copenhagen Ø, Denmark 


Greenland Institute of Natural Resources, 

P.O. Box 570, 3900 Nuuk, Greenland 

103



This thesis presents novel fi ndings for the Arctic tern in 

Greenland. 

Included is a study on Arctic tern migration– the longest 

annual migration ever recorded in any animal. The study 

documented how Greenland and Iceland breeding terns 

conduct the roundtrip migration to the Weddell Sea in 

Antarctica and back. Although the sheer distance (71,000 

km on average) travelled by the birds is interesting, the 

study furthermore showed how the birds depend on highproductive 

at-sea areas and global wind systems during 

their massive migration. 

Furthermore included is the fi rst quantifi ed estimate of the 

capacity to produce a replacement clutch in Arctic terns. 

We found that approximately half of the aff ected birds 

would produce a replacement clutch when the eggs were 

removed late in the incubation period. 

At a level of more national interest, the study produced the 

fi rst estimates of the key prey species of the Arctic tern in 

Greenland. Although zooplankton and various fi sh species 

were present in the chick diet of terns breeding in Disko 

Bay, Capelin was the single most important prey species 

found in all age groups of chicks. 

The thesis also includes a study on the fl uctuating breeding 

found in Arctic terns. Breeding birds showed a considerable 

variation in colony size between years in the small and mid 

sized colonies of Disko Bay. These variations were likely to 

be linked to local phenomena, such as disturbance from 

predators, rather than to large-scale occurring phenomena. 

Included is also a study that documents the breeding association 

between Arctic terns and other breeding waterbirds. 

At Kitsissunnguit a close behavioural response to tern 

alarms could be identifi ed. These fi ndings imply that the altered 

distributions of waterbirds observed at Kitsissunnguit 

were governed by the distribution of breeding Arctic terns. 

Included in the thesis are furthermore results with an appeal 

to the Greenland management agencies. Along with 

estimates of the Arctic tern population size at the two most 

important Arctic tern colonies in West Greenland and East 

Greenland, the study produced recommendations on how a 

potential future sustainable egg harvest could be carried out, 

and on how to monitor Arctic tern colonies in Greenland. 

ISBN: 87-91214-46-7 

ISBN (DMU): 978-87-7073-166-9

Migration and breeding biology of Arctic terns in Greenland

Create successful ePaper yourself

Delete template?

Save as template?