BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofitpublishers, academic institutions, research libraries, and research funders in the common goal of maximizing access tocritical research.

Fluctuating Breeding of Arctic Terns (Sterna paradisaea) inArctic and High-Arctic Colonies in GreenlandAuthor(s): Carsten Egevang and Morten FrederiksenSource: Waterbirds, 34(1):107-111. 2011.Published By: The Waterbird SocietyDOI: http://dx.doi.org/10.1675/063.034.0114URL: http://www.bioone.org/doi/full/10.1675/063.034.0114

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in thebiological, ecological, and environmental sciences. BioOne provides a sustainableonline platform for over 170 journals and books published by nonprofit societies,associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated contentindicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should bedirected to the individual publisher as copyright holder.

107

Fluctuating Breeding of Arctic Terns (Sterna paradisaea )in Arctic and High-arctic Colonies in Greenland

CARSTEN EGEVANG1,* AND MORTEN FREDERIKSEN2

1Greenland Institute of Natural Resources, Postbox 570, 3900 Nuuk, Greenland

2National Environmental Research Institute, Aarhus University, Department of Arctic Environment,Frederiksborgvej 399, P.O. Box 358, DK-4000 Roskilde, Denmark

*Corresponding author; E-mail: cep@dmu.dk

Abstract.—Arctic Terns show large variations in colony attendance between breeding seasons, making compa-rable counts of colony size complex and population status estimates difficult. Here, three areas in two regions ofGreenland were surveyed in consecutive breeding seasons between 2002 and 2009. Between-year variation in pop-ulation size in the small and mid-sized Arctic Tern colonies in West Greenland was considerable (mean CV of indi-vidual colonies 117.5%, CV of total 49.6%). In the largest colony in Greenland, Kitsissunnguit, overall colony sizeshowed minor variations (CV 14.6%), but variation in numbers was pronounced at the sub-colony level (mean CV47.4%). When combining the surveyed colonies in West Greenland, the total population size varied little (CV 6.7%)and less than expected if colonies fluctuated independently (P = 0.023), indicating that local movements betweenthe colonies took place and annual variation was linked to local rather than to large-scale phenomena. In NortheastGreenland, complete breeding failure, likely caused by Arctic Fox, was recorded in two out of four seasons. FutureGreenland Arctic Tern colony monitoring should either 1) survey multiple colonies within the same season cover-ing adjoining colonies over a larger area, or 2) survey one large, representative colony in multiple years in order totrack potential changes in population size. Received 15 March 2010, accepted 19 May 2010.

Key words.—Arctic, Arctic Tern, breeding, Greenland, seabirds, Sterna paradisaea, survey methods.

Waterbirds 34(1): 107-111, 2011

Seabirds as a group are generally charac-terized by long life expectancy with a smallannual investment in reproductive outcome,often breeding in large colonies with a highdegree of site fidelity (Coulson 2002; We-imerskirch 2002). Colonial breeding at pre-dictable sites makes seabirds ideal for moni-toring and survey schemes, monitoring notonly the status of seabird populations, but al-so as indicators for the state of the marineenvironment (Piatt et al. 2007; Parsons et al.2008).

The Arctic Tern (Sterna paradisaea) usual-ly breeds in land predator-free environ-ments, such as islands and skerries, and isparticularly effective at defending the colonyagainst avian predators. The annual repro-ductive outcome of Arctic Terns fluctuateswidely, with breeding failure over large areasreported (Avery et al. 1992; Boertmann et al.1996). Further, colony desertion within thebreeding season and high chick mortalityfrom starvation (Monaghan et al. 1989; Mon-aghan et al. 1992; Suddaby and Ratcliffe1997) seem to occur on a more regular basisthan in most other seabirds.

Although site fidelity at a regional levelis high in nesting Arctic Terns, dispersal toneighboring breeding colonies occurs fre-quently (Devlin et al. 2008; Møller et al.2006; Brindley et al. 1999; Ratcliffe 2004)resulting in large variation in colony atten-dance between seasons at some Arctic Terncolonies. A combination of the above-mentioned factors makes it particularlydifficult to study population size, popula-tion status and trends in Arctic Terns (Rat-cliffe 2004).

In Greenland, the Arctic Tern is a wide-spread breeder with its core distribution inWest Greenland between 68° and 74°N(Boertmann et al. 1996). Although no sys-tematic monitoring schedule of ArcticTern colonies has been initiated in Green-land, opportunistic counts during the lastfive to six decades indicate a decline in thepopulation (Salomonsen 1950; Egevang etal. 2004; Burnham et al. 2005). Due to thelack of systematic counts, fluctuations incolony size are poorly documented inGreenland, although years with non-breeding and breeding failure have been

108 WATERBIRDS

noted by several authors (Salomonsen1950; Boertmann et al. 1996; Levermannand Tøttrup 2007; GSCD 2009).

In this paper we present the results ofArctic Tern colony surveys from four to fiveconsecutive seasons conducted in bothsmall-, mid- and large-sized colonies in twodifferent regions of Greenland in order topresent management recommendations forfuture Arctic tern monitoring.

METHODS

Arctic Tern surveys were conducted in three studyareas located in two different regions and climate zonesof Greenland. Kitsissunnguit (Grønne Ejland) is locat-ed in the southern part of Disko Bay, West Greenland(68.85°N, 52.00°W).

The site consists of four major islands (size: 0.44-3.55km2) and many islets and skerries (Fig. 1A). Kitsissun-nguit holds the largest Arctic Tern colony in Greenlandand probably one of the largest in the world (Cramp1985; Hatch 2002). The archipelago furthermore holdslarge diversity (by arctic standards) and high densitiesof other breeding waterbirds, likely a result of breedingassociation between Arctic Terns and other waterbirds(Egevang et al. 2004; Jørgensen et al. 2007; Egevang andBoertmann 2008).

The “Akunnaaq area” is located in the southern partof Disko Bay, West Greenland (approximately 20 kmsouthwest of Kitsissunnguit) between the town of Aa-siaat and the settlement of Akunnaaq (68.72-68.80°N,52.20-52.73°W). The study area consists of numerousskerries and small islands (Fig. 1A), of which many holdbreeding Arctic Terns. Fourteen colonies on islets be-tween 0.002 and 0.2 km2 in size (Table 1, Fig 1A) werecounted annually between 2002 and 2005.

Sand Island (Sandøen) is located within the high-arctic zone of Northeast Greenland, at the mouth ofYoung Sound (74.26°N, 20.16°W, Fig. 1B). The small is-land (0.22 km2) is highly influenced by sea ice andbreak-up of the ice-covered surrounding fjord occurs inearly July, but drift ice in varying amounts may bepresent throughout the breeding season.

Both Kitsissunnguit and the Akunnaaq area are situ-ated within the core distribution area for Arctic Tern inGreenland, and neighboring Arctic Tern colonies (notincluded in either of the study sites) are few kilometresaway. In contrast, Sand Island is more isolated and thenearest tern colony is more than 50 km away.

The high number of breeding birds combined withthe fact that terns at Kitsissunnguit breed scattered overa large consecutive area called for special census meth-ods. Line transects (Buckland et al. 2001) recording thenest of Arctic Terns were applied on the major islandsbetween 18 June and 3 July in 2002-2006. North-southoriented lines 250 m apart were used in the field withdata subsequently analyzed using the software Distancever. 5.0 (Thomas et al. 2006). Counts were always con-ducted by two persons: an observer and a navigator.

Direct counts were applied at the Akunnaaq areaand at Sand Island. We used flush counts to estimatecolony size, and the number of birds present at thecolony was transformed into breeding pairs by divid-ing by a factor of 1.5 (Bullock and Gomersall 1981).The colonies were located on small islands that couldbe surveyed from the sea, and counts were performedwith a boat as platform. The counts in the Akunnaaqarea were conducted within a narrow time window(17-19 June), which corresponds to mid- incubation(Table 1).

Counts at Sand Island were conducted between 20and 26 July (mid-late incubation). The island was divid-ed into segments that were counted (flush counts) sep-arately by walking through the colony on foot.

We used permutation tests to evaluate whether re-gional annual total nest counts were less variable thanexpected if colonies fluctuated independently, whichwould indicate between-colony movements. Withineach colony, annual counts were permuted randomly9,999 times, simulated regional annual totals and theirvariance (or equivalently CV) were calculated, and aone-tailed P value determined as the proportion of sim-ulated totals having lower variance than the observedregional annual totals.

RESULTS

The overall population size at Kitsissun-nguit (all islands combined) showed littlevariation (CV = 13.9%) between years witha population size between 15,400-21,800pairs in 2002-2006 (Table 2). However, atindividual islands a more pronounced vari-ation (Table 2) was recorded with differ-ences of almost a factor three in breedingpopulation (e.g. Niaqornaq 2002 vs. 2003).The breeding density at Kitsissunnguit (allyears and all islands combined) averaged4,823 pairs/km2 (±1,483). At a smaller

Figure 1. Study areas. Inset map in upper left cornershows the location of A Kitsissunnguit and the Akun-naaq Area in Disko Bay and B of Sand Island in North-east Greenland. The map scale (lower right corner)applies to both A and B.

ARCTIC TERN COLONIES IN GREENLAND 109

scale (comparing between the islands),breeding densities varied between 7,213and 629 pairs/km2.

The 14 colonies in the Akunnaaq areashowed considerable differences in popula-tion size between years, with coefficient ofvariation ranging from 76.6% to 188.4% (Ta-ble 1), whereas CV of the colonies combinedwas lower at 49.6%. All of the colonies expe-rienced years without breeding terns (mostof them in 2003), and the largest absolutevariation in colony size was found in the larg-est colony (68013), ranging from 0 to 2,000individuals (Table 1).

The permutation test showed no evi-dence of lower regional variation withineach of the two regions (Kitsissunnguit P =0.20, Akunnaaq P = 0.90), but when the two

regions were pooled to form an overall DiskoBay total, variance and thus CV were lowerthan expected if individual colonies fluctuat-ed independently (P = 0.023).

At Sand Island, variation in breedingnumbers was more pronounced. Both2006 and 2009 were years with total breed-ing failures, whereas the breeding popula-tion in 2007 and 2008 was estimated at be-tween 700 and 1,000 pairs, based on countsvarying between 1,150 and 1,500 individu-als. In the two years with breeding failure,birds returned to the colony at the start ofthe season, but egg-laying was either post-poned or laid eggs depredated, untilbreeding was eventually abandoned. In2007/2008 the breeding density equalled2,574-3,676 pairs per km2 on Sand Island.

Table 1. Colony size (individuals) at 14 Arctic Tern colonies in the southern part of Disko Bay in 2002-2005. Colonycode refers to the Greenland Seabird Colony Database (GSCD 2009).

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 85.468142 10 30 0 0 141.468155 600 0 600 1,050 76.668166 30 0 0 12 135.068167 1,000 0 520 330 90.368174 80 0 40 200 108.068175 0 0 0 0 —68188 30 0 0 650 188.468189 55 0 260 0 156.968190 300 0 400 50 103.068191 200 0 140 400 89.768192 100 0 1,000 300 128.868193 300 0 0 100 141.468194 0 1,200 430 950 83.2

Total 4,705 1,230 4,040 5,842 49.6

Table 2. Arctic Tern population size (pairs) at Kitsissunnguit (Grønne Ejland) in 2002-2006. Population size was es-timated using line transects. Data were analyzed using software Distance 5.0 (Thomas et al. 2006) applying a halfnormal cosine detection function with a 3% right truncation.

2002 2003 2004 2005 2006CV 2002-2005

(%)CV 2002-2006

(%)

Angissat 0 0 0 0 0 — —Innarsuatsiaaq 0 0 776 6001 8001 68.7 93.0Basis Ø 11,525 17,626 15,901 12,112 17,060 15.8 19.1Niaqornaq 3,159 1,245 1,984 2,346 2,5001 27.7 31.3Saattuarsuit 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.

110 WATERBIRDS

DISCUSSION

The information obtained from compar-ing surveys in this study shows a high degreeof variation in population size between years.The highest coefficient of variation wasfound at colony level, whereas lower varia-tion was found for the two study areas, Kitsis-sunnguit and the Akunnaaq Area (Table 1and 2). When combining the two study ar-eas, which include more than 80% (GSCD2009) of the Arctic Tern population in DiskoBay, to address annual population size varia-tions at a large scale, a CV of only 6.7% wasfound with the grand total ranging between20,000 and 23,000 pairs (Table 3), a level ofvariation that was smaller than expected ifcolonies fluctuated independently. Thesefindings led us to conclude that year-to-yearvariation in the Arctic Tern population inDisko Bay is not governed by large-scale phe-nomena. Instead, the switching of coloniesbetween seasons is more likely driven bymore local phenomena, such as predation ordisturbance.

Variation was also found among coloniesin the distribution and density of the breed-ing birds. The highest and most pronouncedvariation was found among the smaller andmidsized colonies in the Akunnaaq area.Here colonies with up to 2,000 breeding in-dividuals one year could be followed by yearswithout or with very few pairs breeding.

At the high-arctic Sand Island, the shortsummer leaves only a brief window of oppor-tunity for completing the breeding cycle.The four years of data included in this studyindicate that complete breeding failuretakes place at a regular basis (two out of fourseasons). When the birds arrive to breed be-tween late June and early July, sea ice stillcovers the surroundings of the island con-necting it with the mainland. From here,

Arctic Foxes are able to access the island andcause considerable nest predation. The pres-ence of a predator on the island in egg-layingperiod seemed to make the birds postponetheir egg laying, as suggested by Levermannand Tøttrup (2007), or more likely postponethe decision of investing in a replacementclutch after having the first clutch depredat-ed early in the season.

The fluctuating occurrence of ArcticTerns between breeding seasons stressed bythis study clearly emphasizes the need for aspecialized monitoring program for this spe-cies. Monitoring of Arctic Terns in single col-onies conducted in single years does not pro-vide a reliable estimate of the population sta-tus, as Arctic Terns show a lesser degree ofsite fidelity compared with most other Arcticseabird species. The use of counts of breed-ing birds or nests in fixed study plots for laterextrapolation is furthermore questionable,as within-colony movements of the locationof nest along with variation in breeding den-sity between years seems to occur at a regularbasis.

Future monitoring of Arctic Tern popula-tions in Greenland is recommended to applyone of the following two approaches (or ide-ally a combination of both): 1) Survey ofmultiple colonies within the same seasoncovering adjoining colonies over a larger ar-ea, or 2) survey of one large, representativecolony in multiple years. However, since lo-gistics in Greenland are both expensive andtime-consuming, the first approach seemsmost cost effective and likely to produce asuitable estimate for managing purposes.

In terms of the survey method, this willprobably be determined by the size, struc-ture and breeding density of the colony.Smaller and mid-sized colonies up to 2,000birds breeding on skerries and islets that canbe surveyed from the sea are best counted us-

Table 3. Grand total (from Tables 1 and 2) and coefficient of variation (CV) from Arctic colony censuses in DiskoBay 2002-2005.

2002 2003 2004 2005 CV (%)

Kitsissunnguit 15,354 19,541 19,191 16,478 11.6Akunnaaq area 4,705 1,230 4,040 5,842 49.6

Grand Total 20,059 20,771 23,231 22,320 6.7

ARCTIC TERN COLONIES IN GREENLAND 111

ing a boat as platform. Large colonies withhigher number of birds are recommendedto be surveyed by counts in subdivisions ofthe colony (preferably delimited by easilyrecognizable, natural boundaries in thefield). Large colonies with high breedingnumbers and nests evenly distributed at lowor medium densities should be surveyed us-ing line transects or a similar approach.

ACKNOWLEDGMENTS

The fieldwork was financed by the Danish EnergyAgency (the climate support program to the Arctic) andthe Commission for Scientific Research in Greenland(KVUG). Special thanks go to those who assisted in thefield: D. Boertmann and M. Bjerrum (National Environ-mental Research Institute, Denmark), A. Tøttrup, M.Willemoes and P. S. Jørgensen (University of Copen-hagen), O. S. Kristensen, L. Witting and R. Lauridsen(Greenland Institute of Natural Resources), B. Peters-en, Q. and N. Ratcliffe and S. Benn (Royal Society forProtection of Birds).

LITERATURE CITED

Avery, M. I., D. Suddaby, P. M. Ellis and I. M. W. Sim.1992. Exceptionally low body-weights of Arctic TernsSterna paradisaea on Shetland. Ibis 134: 87-88.

Boertmann, D., A. Mosbech, K. Falk and K. Kampp.1996. Seabird colonies in western Greenland (60°-79°30’ N lat.). Technical Report 170, National Envi-ronmental Research Institute, Roskilde, Denmark.

Brindley E., G. Mudge, N. Dymond, C. Lodge, B. Rib-bands, D. Steele, P. Ellis, E. Meek, D. Suddaby and N.Ratcliffe. 1999. The status of Arctic Terns Sternaparadisaea at Shetland and Orkney in 1994. AtlanticSeabirds 1: 135-143.

Buckland, S. T., D. R. Anderson, K. P. Burnham, J. L.Laake, D. L. Borchers and L. Thomas. 2001. Intro-duction to Distance Sampling - Estimating abun-dance of biological populations. Oxford UniversityPress, Oxford, U.K.

Bullock, I. D. and C. H. Gomersall. 1981. The breedingpopulation of terns in Orkney and Shetland in 1980.Bird Study 28: 187-200.

Burnham, W., K. K. Burnham and T. J. Cade. 2005. Pastand present assessments of bird life in UummannaqDistrict, West Greenland. Dansk Ornitologisk Foren-ings Tidsskrift 99: 196-208.

Coulson, J. C. 2002. Colonial breeding in seabirds. Pag-es 87-113 in Biology of Marine Birds (E. A. Schreiberand J. Burger, Eds.). CRC Press, Boca Raton, Florida.

Cramp, S., Ed. 1985. Terns to woodpeckers. Birds of theWestern Palearctic. Oxford University Press, Oxford,U.K.

Devlin, C. M., A. W. Diamond, S. W. Kress, C. S. Hall andL. Welch. 2008. Breeding Dispersal and Survival of

Arctic Terns (Sterna paradisaea) Nesting in the Gulfof Maine. Auk 125: 850-858.

Egevang, C. and D. Boertmann. 2008. Ross’s Gulls (Rho-dostethia rosea) breeding in Greenland: A Review,with Special Emphasis on Records from 1979 to2007. Arctic 61: 322-328.

Egevang, C., K. Kampp and D. Boertmann. 2004. Thebreeding association of Red Phalaropes (Phalaropusfulicarius) with Arctic Terns (Sterna paradisaea): Re-sponse to a redistribution of terns in a major Green-land colony. Waterbirds 27: 406-410.

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. In TheBirds of North America (A. Poole and F. Gill, Eds.).The Birds of North America Inc., No. 707, Philadel-phia, Pennsylvania.

Jørgensen, P. S., M. W. Kristensen and C. Egevang. 2007.Red Phalarope Phalaropus fulicarius and Red-neckedPhalarope Phalaropus lobatus behavioural responseto Arctic Tern Sterna paradisaea colonial alarms. Dan-sk Oritologisk Forenings Tidsskrift 101: 73-78.

Levermann, N. and A. P. Tøttrup. 2007. Predator effectand behavioral patterns in Arctic Terns (Sternaparadisaea) and Sabine’s Gulls (Xema sabini) duringa failed breeding year. Waterbirds 30: 417-420.

Møller, A. P., E. Flensted-Jensen and W. Mardal. 2006.Dispersal and climate change: a case study of theArctic Tern Sterna paradisaea. doi:10.1111/j.1365-2486.2006.01216.x. Global Change Biology 12: 2005-2013.

Monaghan, P., J. D. Uttley, M. D. Burns, C. Thaine andJ. Blackwood. 1989. The relationship between foodsupply, reproductive effort and breeding success inArctic Terns Sterna paradisaea. Journal of AnimalEcology 58: 261-274.

Monaghan, P., J. D. Uttley and M. D. Burns. 1992. Effectof changes in food availability on reproductive effortin Arctic Terns Sterna paradisaea. Ardea 80: 71-81.

Parsons, I., P. I. Mitchell, A. Butler, N. Ratcliffe, M. Fre-deriksen, S. Foster and J. B. Reid. 2008. Seabirds asindicators of the marine environment. ICES Journalof Marine Science 65: 1520-1526.

Piatt, J. F., W. J. Sydeman and F. Wiese. 2007. Seabirds asindicators of marine ecosystems. Marine EcologyProgress Series 352: 199-204.

Ratcliffe, N. 2004. Arctic Tern Sterna paradisaea. Pages328-338 in Seabird Populations of Britain and Ire-land (P. I. Mitchell, S. F. Newton, N. Ratcliffe and T.E. Dunn, Eds.). T & A D Poyser, London, U.K.

Salomonsen, F. 1950. The Birds of Greenland. Munks-gaard, Copenhagen, Denmark.

Suddaby, D. and N. Ratcliffe. 1997. The effects of fluctu-ating food availability on breeding Arctic Terns (Ster-na paradisaea). Auk 114: 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. Dis-tance. University of St. Andrews, U.K.

Weimerskirch, H. 2002. Seabird demography and its re-lationship with the marine environment. Pages 115-135 in Biology of Marine Birds (E. A. Schreiber andJ. Burger, Eds.). CRC Press, New York, New York.