Robertson.indd 85Robertson et al. 2001: Polar Research 20(1), 85–94 Clutch, egg and body size variation among common eiders breeding in Hudson Bay, Canada Gregory J. Robertson, Austin Reed & H. Grant Gilchrist The Hudson Bay common eider (Somateria molissima sedentaria) is a unique subspecies of eider that remains within the confines of Hudson Bay throughout the year. We compared clutch, egg and body size varia- tion among populations of common eiders breeding in eastern and west- ern Hudson Bay. Clutch size did not differ substantially among these populations. All eiders in Hudson Bay laid larger clutches than other sub- species in eastern North America. As Hudson Bay common eiders do not undergo extensive migrations, they may have more energy reserves avail- able to them for egg production. Eiders nesting in eastern Hudson Bay laid larger eggs than eiders nesting in western Hudson Bay. Further, eiders in eastern Hudson Bay tended to be structurally larger, but had smaller bill processes. These differences may have a genetic basis. Smaller egg size and body size may arise in western Hudson Bay from mixing with the smaller borealis subspecies nesting to the north. Further work to resolve genetic affinities, determine levels of male and female dispersal, and examine variation in reproductive ecology are needed to resolve the sources of these differences. G. J. Robertson, Canadian Wildlife Service, 6 Bruce Street, Mount Pearl, Newfoundland A1N 4T3, Canada; A. Reed, Canadian Wildlife Service, 1141 Route de l’Église, Ste-Foy, Québec G1V 4H5, Canada; H. G. Gilchrist, Canadian Wildlife Service, Suite 301, 5204 50th Avenue, Yellowknife, Northwest Territories X1A 1E2, Canada. Many species of birds show variation in morpho- logical, behavioural and/or life history characters across their ranges. This variation can be clinal or discrete (Rockwell & Barrowclough 1987), and populations showing discrete differences may be designated as subspecies. Alternatively, differ- ences among populations may reflect differential phenotypic expression of traits in different envi- ronments. Documenting variation among popula- tions has important implications for evolutionary, systematic and biogeographical studies and for conservation efforts (Avise & Hamrick 1996). Among waterfowl, ducks tend to show less intraspecific variation than geese in morpholog- ical and life history traits across their range. Common eiders (Somateria mollissima) are one of the most variable ducks and have seven dis- tinct subspecies (Palmer 1976; Cramp & Sim- mons 1977), which show morphological and life history trait differences. In this regard common eiders are similar to geese, which also tend to show considerable variation among populations and subspecies (Dunn & MacInnes 1987). Clutch, egg and body size are all important life history/morphological traits. In particular, clutch size is an important component of reproductive output (Winkler & Walters 1983). Variation in clutch size among populations may reflect differ- ent selective regimes, environmental conditions, and/or age structure (Rohwer 1992). Factors gov- erning egg size variation are less clear (Williams 1994). Variation in egg size among populations 86 Clutch, egg and body size variation among common eiders may result from selection for different sized eggs in different environments, from genetic drift, dif- ferent environmental conditions, and/or differ- ences in age structure. Body size is heritable (Boag & van Noordwijk 1987; van Noordwijk et al. 1988), although environmental conditions, specifically food availability during growth, can influence final adult body size (Cooch et al. 1991; Leafloor et al. 1998). Therefore, differences in body size among populations may reflect different selec- tive pressures, population isolation and genetic drift, or different environmental conditions. In this study we investigated patterns of clutch, egg and body size variation among populations of the Hudson Bay subspecies of the common eider (Somateria mollissima sedentaria). As a subspe- cies, Hudson Bay common eiders are poorly defined and not well studied (Snyder 1941; Free- man 1970; Mendall 1980, 1986). The entire popu- lation of Hudson Bay common eiders is assumed to spend the whole year within the icy confines of Hudson Bay (Snyder 1941; Freeman 1970), unlike other subspecies of northern breeding eiders, which migrate to southern waters. In this paper, we compare clutch, egg and body size among populations of common eiders at three spatial scales: 1) among populations of common eiders breeding on the east coast of Hudson Bay, 2) between common eiders breeding in eastern and western Hudson Bay, and 3) between the Hudson Bay subspecies and the other two subspecies of common eiders breeding in eastern North America. Methods Fieldwork for this study was conducted in two general areas within Hudson Bay. The first site was a large area along the coast of south-eastern Hudson Bay and the Belcher Islands (Figs. 1 and 2). Data on egg size and clutch size in this region were collected during two intensive surveys of nesting eiders, one from 1985 to 1988, the other in 1997. Nine nesting regions were visited in 1985-88, while only four regions in the Belcher Island archipelago were surveyed in 1997. Clutch size data for these sites were extracted from Flem- ing & McDonald (1987), Nakashima & Murray (1988), McDonald & Fleming (1990) and Robert- son & Gilchrist (1998). This population was esti- mated at 83 000 breeding pairs in the mid-1980s (Nakashima & Murray 1988), but has declined to an estimated 20 000 pairs by 1997 (Robertson & Gilchrist 1998). The other study area was at La Pérouse Bay, on the west coast of Hudson Bay, 30 km east of Churchill, Manitoba (Fig. 1). Data from this site were collected as part of an intensive study on the breeding biology of common eiders in 1991–93. The colony consists of approximately 400 female eiders nesting on the small freshwater islands of Fig. 1. Location and place names of common eider study areas in Hudson Bay. 87Robertson et al. 2001: Polar Research 20(1), 85–94 the Mast River Delta (Schmutz et al. 1983; Rob- ertson 1995a). Egg and clutch size data for this site have been presented in Robertson & Cooke (1993) and Robertson (1995b); the data presented in this paper have been reanalysed to keep the analyses between the two study areas consistent. Because the colony at the Mast River is small, it is likely that clutch size and egg size for a few females were measured in more than one year. This potential problem will artificially inflate sample sizes, but will not bias estimates of means. Information for other subspecies (borealis and dresseri) was extracted from the literature and unpublished data provided by researchers. Only clutches that were complete and being incubated were included in the analyses. Egg volume (cm3) was calculated from measurements of egg length and maximum egg breadth using the equation found in Guild (1974) and Robertson & Cooke (1993): vol. = 0.000515*length*breadth2. Two analysis methods were used, recognizing that the size of eggs within a clutch are not inde- pendent samples. For data collected in eastern Hudson Bay in 1985-88 and the western Hudson Bay data, a mean egg size for the clutch was calculated and used in all analyses. During the 1997 survey in eastern Hudson Bay only one egg was measured from each clutch to reduce time spent on nesting islands. The egg that was meas- ured was chosen at random (the egg closest to the north-east corner of the nest). Both methods esti- mate mean egg size without bias and effectively control for the non-independence of the size of eggs within a clutch. The only difference is that by selecting one egg at random, both intra- and inter-clutch variance are estimated, whereas using mean egg size, only inter-clutch egg size vari- ance is estimated (i.e. the variance will be overes- timated with the random eggs, making statistical tests more conservative). Body size measurements were collected by the senior author at all locations. Measurements, taken with vernier calipers (± 0.1 mm), were col- lected from eiders trapped for ringing (western Hudson Bay) or from birds shot by local Inuit field assistants (eastern Hudson Bay). Birds col- lected in eastern Hudson Bay where shot on the breeding grounds (in some cases directly off the nests) and so were representative breeders. Head length, culmen 1 (culmen midline) and culmen 2 (total bill length), nostril extension, and total tarsus were taken according to Mendall (1980, 1986) and Dzubin & Cooch (1992). Results Clutch size There were some differences in the clutch sizes among regions in eastern Hudson Bay (Table 1). Eiders at the most northerly and southerly sites Fig 2. Location and place names of common eider study areas in south-eastern Hudson Bay. 88 Clutch, egg and body size variation among common eiders laid larger clutches. Clutch size from sites in the middle of the study area did not differ (4.0 - 4.4). Clutch sizes laid in western Hudson Bay were in the middle of the range of clutch sizes laid in east- ern Hudson Bay (Table 1). Egg size Within eastern Hudson Bay, there were significant differences among mean egg volumes measured at the different locations in 1985-86 (ANOVA, F= 2.55, df = 7, 199, p = 0.016). Egg volumes were larger in Long Island Sound than all other regions (Table 2); the sample size for this region was very small and may not be representative. No differ- ences in mean egg volume were detected among the four regions of eastern Hudson Bay studied in 1997 (F = 1.43, df = 3, 224, p = 0.24). Mean egg volumes for clutches from western Hudson Bay (Table 2) were significantly smaller than clutches laid in eastern Hudson Bay (pooled mean 104.2 cm3 ± 7.2, n = 435) (t-test with une- qual variances, t = 20.8, df = 754.7, p = 0.0001). Body size In general, female common eiders in eastern Hudson Bay had longer tarsi and head lengths, and shorter nostril extensions than females in western Hudson Bay (Table 3). Similarly, males from eastern Hudson Bay had longer tarsi and head lengths and shorter nostril extensions and culmen 2 (total bill length) than males from west- ern Hudson Bay (Table 3). Differences among subspecies Clutch sizes were largest among sedentaria females. Some populations of dresseri laid simi- lar sized clutches (Table 4), but some laid smaller ones. To the north, borealis females tended to lay smaller clutches. Egg sizes were smaller in bore- alis than in sedentaria in eastern Hudson Bay, but not western Hudson Bay (Table 4). Only limited information is available for dresseri, but this sub- species lays similar or larger sized eggs than sed- entaria (Table 4). Location Latitude Mean clutch size Eastern Hudson Bay Western Hudson Bay 1985-88a 1997b 1991-93c Koktac River 59°15’ 4.8 ± 1.3 B (612) La Pérouse Bay, MB 58°45’ 4.3 ± 1.0 CD (323) Sleeper Islands 57°30’ 4.3 ± 1.4 CD 4.4 ± 1.1 C (2557) (412) Nastapoka Islands 57°00’ 4.1 ± 1.4 DE (925) N. Belcher Islands 56°45’ 4.0 ± 1.3 E 4.0 ± 1.1 E (1140) (206) Salikuit Islands 56°15’ 4.2 ± 1.4 CDE (649) S. Churchill Sound 56°00’ 4.1 ± 1.4 DE 4.4 ± 1.2 C (876) (275) S. Flaherty Islands 55°45’ 4.0 ± 1.1 DE 4.4 ± 1.0 C (1980) (419) Long Island Sound 54°45’ 5.6 ± 1.0 A (113) Table 1. Mean clutch size (± 1 s.d.) for Hudson Bay common eiders (Somateria mollissima sedentaria) across their range (see Figs. 1 and 2 for locations). Letters represent means not significantly different from one another (Student-Neuman-Keuls post-hoc means test, ANOVA, F = 34.4, df = 12, 10165, p = 0.0001). a Data from Fleming & McDonald (1987), Nakashima & Murray (1988), McDonald & Fleming (1990). b Data from Robertson & Gilchrist (1998). c Data from Robertson (1995b). 89Robertson et al. 2001: Polar Research 20(1), 85–94 Discussion Clutch size Comparisons of clutch sizes across studies are confounded by many factors that can influence measurements of clutch size. The timing of when clutch size is surveyed is important as partial clutches may be counted if the survey is too early. Further, eiders, along with many other bird spe- cies, show significant annual variation in clutch sizes laid by individuals and populations (Coul- son 1999). A component of this annual variation is correlated with laying date (Rohwer 1992). To minimize these sources of variation, at all sites in this study clutch size was measured during early to mid-incubation. Although it is difficult to con- trol for annual differences, we were fortunate in most years of this study because nesting was usu- ally early and conditions were generally favour- able (Nakashima & Murray 1988; Robertson 1995b; Robertson & Gilchrist 1998). Predation can greatly impact measures of clutch size, as many eggs are taken during the egg-laying stage (Robertson & Cooke 1993), especially by avian predators. Differences in predation rates may be reflected in differences in clutch sizes among populations and among years. In spite of the fact that different sites were examined in different years, within Hudson Bay there was no clear pattern of clutch size variation among populations of common eiders. Clutch sizes consistently averaged between 4 and 4.5 eggs. We are not sure why clutch sizes laid in Long Island Sound were so large in 1985 at 5.6 eggs. Intraspecific nest parasitism may have resulted in abnormally large clutches being incu- bated at this site (Robertson et al. 1992; Bjørn & Erikstad 1994), and sample sizes were not large. In general, sedentaria laid slightly larger clutches then dresseri to the south and larger clutches than borealis to the north. Other water- fowl species show a similar trend: clutch size tends to decrease in the northern portions of their range (Dunn & MacInnes 1987; Rohwer 1992). Individuals and populations that lay later in the season tend to lay smaller clutches (Rohwer 1992). However, the relative timing of egg laying cannot explain the variation in clutch size laid between these subspecies of common eiders. In good to normal years, eiders nesting near the Belcher Islands do not begin egg-laying until mid- to late June (Nakashima & Murray 1988; Robert- son & Gilchrist 1998), similar to borealis eiders Location Latitude Mean egg volume of clutch (cm3) Eastern Hudson Bay Western Hudson Bay 1985-86 1997 1991-93 Koktac River 59°15’ 103.4 ± 5.8 B (72) La Pérouse Bay, MB 58°45’ 94.5 ± 5.5 C (323) Sleeper Islands 57°30’ 104.3 ± 4.7 B 103.2 ± 8.1 B (14) (61) N. Nastapoka Islands 57°15’ 102.2 ± 7.0 B (41) S. Nastapoka Islands 56°45’ 105.9 ± 5.7 B (10) N. Belcher Islands 56°45’ 103.0 ± 7.1 B 105.4 ± 7.0 B (20) (38) Salikuit Islands 56°15’ 105.5 ± 8.1 B (16) S. Churchill Sound 56°00’ 104.5 ± 7.5 B (74). S. Flaherty Islands 55°45’ 102.7 ± 5.2 B 105.9 ± 7.9 B (25) (55) Long Island Sound 54°45’ 110.9 ± 8.0 A (9) Table 2. Mean egg volumes (± 1 s.d.) of Hudson Bay common eiders (Somateria mollissima sedentaria) across their range (see Figs. 1 and 2 for locations). Letters represent means not significantly different from one another (Student-Neuman-Keuls post-hoc means test). 90 Clutch, egg and body size variation among common eiders in good nesting years far to the north (Cooch 1965; Prach et al. 1986; HGG unpubl. data). This is months later than dresseri nesting in New Eng- land and Maritime Canada, which begins in April for many populations. Therefore, common eiders nesting in Hudson Bay tend to lay large clutches in spite of their late timing of breeding. Common eiders rely exclusively on nutrients gathered and stored before laying for egg pro- duction and incubation (Korschgen 1977; Parker & Holm 1990). Eiders nesting in Hudson Bay may have more nutrient reserves during egg pro- duction because they do not undertake an exten- sive migration to the breeding grounds (Freeman 1970; Guild 1974). Migrating individuals can utilize considerable nutrient reserves (Berthold 1993). Most populations of dresseri do not make extensive migrations either, and clutch sizes tend to be larger than borealis populations, which migrate from south-west Greenland, Newfound- land and the Gulf of St. Lawrence to the eastern Canadian Arctic (Palmer 1976). Populations of common eiders nesting in Europe (S. m. mollis- sima) that do not undergo extensive migrations also show relatively large clutch sizes (means ranging from 4.2 to 5.0 eggs; Scotland [Milne 1974], the Netherlands [Swennen 1983], Sweden [Götmark & Åhlund 1988], Finland [Hario & Selin 1988], Norway [Erikstad et al. 1993]). Simi- lar to North America, borealis breeding on remote islands north of Europe lay much smaller clutches (3.0 to 3.2, Spitsbergen, Ahlén & Andersson [1970]; Bjørn & Erikstad [1994]). These popula- tions undergo substantial migrations, wintering off the south coast of Greenland, eastern North America or the Barents Sea (Dement’ev & Glad- kov 1952). These comparisons must be made cau- tiously, however, as other factors, such as high incidents of nest parasitism in Hudson Bay among common eiders, may artificially inflate measures of clutch size, or high levels of partial clutch dep- redation in borealis may lower clutch size. Egg size Eiders nesting in western Hudson Bay laid smaller eggs. Interestingly, there is no correlation between egg size and female body size or condition (Lau- rila & Hario 1988; Swennen & van der Meer 1992), so differences in body condition or size would not explain egg size differences among populations. One environmental component has been identified that explains some variation in egg size: Robertson (1995b) showed that eiders laid smaller eggs when temperatures were cold during egg-laying. However, the average mean egg size difference between a very warm year and a very cold year was only 3% (Robertson 1995b). The mean difference between east and west coast Hudson Bay common eiders was three times larger at 9%. Nesting conditions were gen- erally favourable in most years included in this study, so different annual environmental condi- tions were not likely to be a major source of variation in egg size between sites. Egg size is highly repeatable in common eiders (Laurila 1988; Swennen & van der Meer 1992) and likely to be highly heritable (Rohwer 1992). Therefore, we suggest that egg size differences among pop- ulations may reflect mostly genetic differences, although some portion of the variation (up to 3%) could be explained by the fact that eiders nesting on the east coast of Hudson Bay enjoy warmer temperatures at nesting. Female eiders show natal philopatry to their breeding grounds (Wakeley & Mendall 1976; Swennen 1990), so breeding populations of eiders are expected to show some genetic isolation (Anderson et al. 1992; Avise et al. 1992). Indeed, Tiedemann et al. (1999) found high levels of dif- ferentiation between colonies in maternally inher- ited mtDNA sequences. However, male common eiders in Europe disperse widely (Swennen 1990). Genetic isolation among breeding populations of Eastern Western Hudson Baya Hudson Bay t p Females (n) 30 79 Culmen 1 51.6 ± 2.6 52.0 ± 3.0 0.59 0.5552 Culmen 2 71.4 ± 2.8 71.0 ± 2.4 0.75 0.4527 Nostril extension 34.0 ± 2.0 35.9 ± 1.6 5.16 0.0001 Head length 121.1 ± 2.3 118.6 ± 2.6 4.60 0.0001 Tarsus length 66.4 ± 2.0 62.8 ± 2.0 8.38 0.0001 Males (n) 10 9 Culmen 1 55.8 ± 2.9 58.1 ± 3.5 1.59 0.1305 Culmen 2 79.7 ± 3.3 82.8 ± 2.3 2.29 0.0353 Nostril extension 40.5 ± 2.7 45.3 ± 2.2 4.20 0.0006 Head length 128.8 ± 2.6 125.9 ± 2.5 2.35 0.0327 Tarsus length 68.6 ± 1.3 66.2 ± 2.0 3.20 0.0053 a Only 28 female and 8 male head measurements available for eastern Hudson Bay. Table 3. Morphometrics (mm, ± 1 s.d.) of common eiders in eastern and western Hudson Bay. Culmen 1 (Dzubin & Cooch 1992) corresponds to Mendall’s (1980, 1986) culmen midline and culmen 2 corresponds to total bill length. Tarsus length is total tarsus (Dzubin & Cooch 1992). 91Robertson et al. 2001: Polar Research 20(1), 85–94 common eiders in Hudson Bay could be main- tained if male dispersal is restricted. Suitable wintering habitat in the form of open water leads and polynyas is mostly restricted to south-eastern Hudson Bay, where large numbers of sedentaria common eiders are known to winter (Freeman 1970; Gilchrist & Robertson 2000). It is assumed that these eiders breed in eastern Hudson Bay (Freeman 1970). Although they may winter in the Belcher Islands, common eiders from west- ern Hudson Bay are suspected to winter off the landfast ice edge of the western coast, or in a large recurring polynya in Roes Welcome Sound (Abra- ham & Finney 1986). Potentially, eastern and western Hudson Bay common eiders are isolated from each other during winter, the time when pair bonds and, therefore, gene flow occurs. Even if there is extensive nuclear gene flow among eiders breeding in Hudson Bay due to male dispersal, genetic isolation could be main- tained if additive genetic variance for egg size is maintained on mitochondrial genes or the het- erogametic chromosome. Williams et al. (1994) suggested that there may be a genetic correlation between egg size and physiological and/or met- abolic processes. If the genetic variance in egg size and covariance of egg size with other met- abolic traits are occurring within the mitochon- drial DNA of females, then egg size will be solely maternally inherited. In this case, male dispersal will not result in exchange of genes influencing egg size, and genetic population structure for egg size would be solely influenced by female disper- sal patterns (Avise et al. 1992). Body size As with egg size, eiders in eastern Hudson Bay Subspecies Location Mean egg volume (cm3) Mean clutch size Source borealis Devon Island 94.6 ± 7.7 3.3 ± 1.0 Prach et al. (1986) (158) (544) Southampton Island 90.2 ± 6.6 3.6 ± 1.2 H. G. Gilchrist, unpubl data (262) (262) Cape Dorset 3.4 ± 1.0 Cooch (1965) (1598) Ungava Bay 94.7 ± 6.2 3.6 ± 1.1 Chapdelaine et al. (1986) (33) (2387) sedentaria La Pérouse Bay 94.5 ± 5.5 4.3 ± 1.0 this study (323) (323) La Pérouse Bay 96.1 4.1 ± 1.2 Guild (1974) (298) (175) Eastern Hudson Bay 104.2 ± 7.2 4.2 ± 1.3 this study (435) (9855) S. Belcher Islands 116 (97 - 139) 4.5 ± 1.4 Freeman (1970) (84) (290) dresseri St. Lawrence Estuary 112.8 ± 7.4 3.7 ± 1.4 Milne & Reed (1974) (72) (1521) Gulf of St.Lawrence 4.3 ± 1.2 Guignion (1968) (315) Gulf of St.Lawrence 4.0 ± 1.0 Lewis (1939) (1131) New Brunswick 3.6 ± 1.2 Paynter (1951) (134) New Brunswick 116.4 ± 10.2 4.4 (2-7) Gross (1938) (45) (100) New Brunswick 106.3 ± 7.2 4.0 ± 1.1 K. Mawhinney, unpubl. data (642) (1921) Maine 107.4 ± 8.2 3.9 ± 1.7 K. Mawhinney, unpubl. data (736) (395) Table 4. Volumes of eggs and clutch sizes laid by common eiders in eastern North America. Mean ± 1 s.d. (or range) with sample sizes in parentheses below. 92 Clutch, egg and body size variation among common eiders were structurally larger. Body size, similar to egg size, is highly heritable (Boag & van Noordwijk 1987), suggesting that there some isolation among eider populations breeding in Hudson Bay might be possible. Head length and tarsus tend to be better indices of overall structural size than bill measurements (Rising & Somers 1989) and these were larger in eastern Hudson Bay. Mixing with the smaller borealis race to the north may explain the smaller body size (and possibly egg size) of eiders nesting in western Hudson Bay. In devel- oping his key, Mendall (1980) tentatively identi- fied some individuals collected from north-west Hudson Bay as borealis–sedentaria intergrades. A north—south cline along western Hudson Bay, with sedentaria to the south and borealis to the north is plausible. Further, that cline may extend south-east, with the largest eiders occurring at the south-east corner of Hudson Bay. Inuit hunt- ers from the Belcher Islands mention that large brown eiders are found in the southern portion of the archipelago and smaller grey eiders are found at the northern portions of the Belchers Islands (Nakashima & Murray 1988). This cline may be maintained by male disper- sal. Genes for body size are known to be pater- nally inherited in waterfowl (Larsson & Forslund 1992; Cooke et al. 1995), and clines are usually maintained by limited dispersal. Movements of male eiders along the coast of Hudson Bay during winter are possible but probably restricted, to a degree, by ice and food availability. Alternatively, the cline could be maintained by mate choice, if mates choose partners from their natal colony more often (but not exclusively) than expected at random (Tiedemann et al. 1999). Notwithstand- ing potential genetic differences, body size differ- ences between eastern and western Hudson Bay may be due to different environmental conditions, especially during duckling growth (Leafloor et al. 1998). The core of the breeding range of eiders in Hudson Bay is the south-eastern portion of the bay. Therefore the western portion of Hudson Bay may represent marginal habitat for eiders and may be responsible for their slower growth and subsequent smaller adult body size (and possibly egg size). Unlike head and tarsus length, bill measure- ments tend to be larger in western Hudson Bay. These differences are not as easily explained by differences in conditions during duckling growth. Bill measurements are used in discriminating between the different subspecies of common eiders (Mendall 1980, 1986). In general, dresseri individuals have longer culmen 2 (total bill length) and nostril extension measurements, and larger frontal lobes than sedentaria or borealis. Interest- ingly, individuals, especially males, from western Hudson Bay had bill measurements like dresseri birds. In fact, using the keys provided in Mendall (1986), the average male from western Hudson Bay was classified as a dresseri male. The key was based on small samples sizes for male seden- taria (n = 11) and may be insufficient to discrim- inate the subspecific status of eiders in Hudson Bay (Mendall 1986). Overall, common eiders breeding in Hudson Bay are large compared to other subspecies in eastern North America (Freeman 1970). Only v-nigra, which breeds and winters in the north Pacific, is larger (Palmer 1976). Common eiders in Hudson Bay may have been under selective pressure to become larger. Larger birds are better able to maintain homeothermy and they can sur- vive longer during periods of stress (Calder 1974; Dunn & MacInnes 1987). During the cold, short days of winter at these latitudes, larger body size may be advantageous for common eiders that do not migrate to southern waters. Conclusion Clutch sizes were relatively large and similar among populations of common eiders breeding in eastern and western Hudson Bay. As any differ- ences in environmental conditions would be most likely be seen as differences in clutch size, differ- ences in egg and body size among these popula- tions suggest that there is some genetic population structuring of common eiders breeding in Hudson Bay. The population structure seen in eiders may be due to female natal philopatry and/or a lack of dispersal by males because of restricted win- tering habitat in the ice. Clearly, more work is needed to resolve the genetic relationships among eiders in Hudson Bay and studies are required to determine actual movements and breeding and wintering ground affinities of these populations. Eider populations in this region have declined dramatically in recent years (Robertson & Gil- christ 1998) and an understanding of the genetic structure, affinities and movements of these pop- ulations is certainly needed to guide management recommendations in the future. 93Robertson et al. 2001: Polar Research 20(1), 85–94 Acknowledgements.—Kim Mawhinney kindly provided us with unpublished data. J.-Å. Nilsson provided helpful com- ments on an earlier draft. We thank the many individuals who helped us with the data collection. Maureen Kay prepared the figures. Funding and logistical support was provided by the Sanikiluaq Hunters and Trappers Association, the Cana- dian Wildlife Service (CWS), the Nunavut Wildlife Research Trust, and the Environmental Studies Revolving Funds for the work in eastern Hudson Bay. GJR was supported by the Nat- ural Sciences and Engineering Research Council of Canada (NSERC), the Northern Scientific Training Program (Depart- ment of Indian Affairs and Northern Development, DIAND), the Frank M. Chapman Memorial Fund and the Association of Canadian Universities for Northern Studies while at La Pérouse Bay. The La Pérouse Bay Tundra Biology Station is generously funded by NSERC, DIAND, CWS, the Wildlife Management Institute, the Central and Mississippi Flyway Councils, Ducks Unlimited, the Arctic Goose Joint Venture, the Manitoba Department of Renewable Resources and the City University of New York. References Abraham, K. F. & Finney, G. H. 1986: Eiders of the eastern Canadian Arctic. In A. Reed (ed.): Eider ducks in Canada. Can. Wildl. Serv. Rep. Ser. 47, 55–73. Ahlén, I. & Andersson, Å. 1970: Breeding ecology of an eider population on Spitsbergen. Ornis Scand. 1, 83–106. Anderson, M. D., Rhymer, J. M. & Rohwer, F. C. 1992: Phi- lopatry, dispersal, and the genetic structure of waterfowl populations. In B. D. J. Batt et al. (eds.): Ecology and man- agement of breeding waterfowl. Pp. 365–395. Minneapolis: University of Minnesota Press. Avise, J. C., Alisauskas, R. T., Nelson, W. S. & Ankney, C. D. 1992: Matriarchal population genetic structure in an avian species with female natal philopatry. Evolution 46, 1084–1096. Avise, J. C. & Hamrick, J. L. (eds.) 1996: Conservation genet- ics. London: Chapman and Hall. Berthold, P. 1993: Bird migration: a general survey. Oxford: Oxford University Press. Bjørn, T. H. & Erikstad, K. E. 1994: Patterns of intraspecific nest parasitism in the high Arctic common eider (Somate- ria mollissima borealis). Can J. Zool. 72, 1027–1034. Boag, P. T. & van Noordwijk, A .J. 1987: Quantitative genet- ics. In F. Cooke & P. A. Buckley (eds.): Avian genetics. Pp. 45–78. London: Academic Press. Calder, W. A., III 1974: Consequences of body size for avian energetics. In R. A. Paynter (ed.): Avian energetics. Pp. 86–144. Cambridge, MA: Nuttall Ornithological Club. Chapdelaine, G., Bourget, A., Kemp, W. B., Nakashima, D. J. & Murray, D. J. 1986: Population d’eider à duvet près des côtes du Québec septentrional. (The common eider popula- tion of the northern Québec coast.) In A. Reed (ed.): Eider ducks in Canada. Can. Wildl. Serv. Rep. Ser. 47, 39–50. Cooch, F. G. 1965: The breeding biology and management of the northern eider (Somateria mollissima borealis) in the Cape Dorset area, Northwest Territories. Canadian Wild- life Service Wildl. Manage. Bull. Ser. 2, No. 10. Cooch, E. G., Lank, D. B., Rockwell, R. F. & Cooke, F. 1991: Long-term decline in body size in a snow goose population: evidence of environmental degradation? J. Anim. Ecol. 60, 483–496. Cooke, F., Rockwell, R. F. & Lank, D. B. 1995: The snow geese of La Pérouse Bay: a study in natural selection. Oxford: Oxford University Press. Coulson, J. C. 1999: Variation in clutch size of the common eider: a study based on 41 breeding seasons on Coquet Island, Northumberland, England. Waterbirds 22, 225–238. Cramp, S. & Simmons, K. E. L. 1977: The birds of the western palearctic. Vol. 1. Oxford: Oxford University Press. Dement’ev, G. P. & Gladkov, N. A. 1952: Birds of the Soviet Union. Vol. IV. (Translated from Russian by the Israel Pro- gram for Scientific Translations, Jerusalem, 1967.) Dunn, E. H. & MacInnes, C. D. 1987: Geographic variation in clutch size and body size of Canada geese. J. Field Orni- thol. 58, 355–371. Dzubin, A. & Cooch, E. G. 1992: Measurements of geese: general field measurements. Sacremento: California Water- fowl Association. Erikstad, K. E., Bustnes, J. O. & Moum, T. 1993: Clutch-size determination in precocial birds: a study of the common eider. Auk 110, 623–628. Fleming, B. & McDonald, M. 1987: A nest census and the economic potential of the Hudson Bay eider in the southern Belcher Islands, N.W.T: a report prepared for the Depart- ment of Indian and Northern Affairs, GNWT region. Project no. 176057. Sanikiluaq, Canada: Municipality of Saniki- lauq. Freeman, M. M. R. 1970: Observations on the seasonal behav- iour of the Hudson Bay eider (Somateria mollissima seden- taria). Can. Field-Nat. 84, 145–153. Gilchrist, H. G. & Robertson, G. J. 2000: Observations of marine birds and mammals wintering at polynyas and ice edges in the Belcher Islands, Nunavut, Canada. Arctic 53, 61–68. Götmark, F. & Åhlund, M. 1988: Nest predation and nest site selection among eiders Somateria mollissima: the influence of gulls. Ibis 130, 111–123. Gross, A. O. 1938: The eider ducks of Kent’s Island. Auk 55, 387–400. Guignion, D. 1968: Clutch size and incubation period of the American eider (Somateria mollissima dresseri) on Bran- dypot Island. Nat. Can. 95, 1145–1152. Guild, B.L. 1974: The breeding biology of the Hudson Bay eider at La Pérouse Bay, Manitoba. M.Sc. thesis. Wright State University, Ohio. Hario, M. & Selin, K. 1988: Thirty-year trends in an eider population: timing of breeding, clutch size, and nest site preferences. Finn. Game Res. 45, 3–10. Korschgen, C. E. 1977: Breeding stress of female eiders in Maine. J. Wildl. Manage. 41, 360–373. Larsson, K. & Forslund, P. 1992: Genetic and social inherit- ance of body and egg size in the barnacle goose (Branta leucopsis). Evolution 46, 235–244. Laurila, T. 1988: The role of genetic factors on clutch size and egg dimensions in the common eider Somateria mollissima. Finnish Game Res. 45, 11–18. Laurila, T. & Hario, M. 1988: Environmental and genetic fac- tors influencing clutch size, egg volume, date of laying and female weight in the common eider Somateria mollissima. Finn. Game Res. 45, 19–30. Leafloor, J. O., Ankney, C. D. & Rusch, D. H. 1998: Envi- ronmental effects on body size of Canada geese. Auk 115, 94 Clutch, egg and body size variation among common eiders 26–33. Lewis, H. F. 1939: Size of sets of eggs of the American eider. J. Wildl. Manage. 3, 70–73. McDonald, M. & Fleming, B. 1990: Development of a com- munity-based eider down industry in Sanikiluaq: resource management and business strategies: a report prepared for the Canada-Northwest Territories economic develop- ment agreement. Project no. 561 510. Sanikiluaq, Canada: Municipality of Sanikilauq. Mendall, H. L. 1980: Intergradation of eastern American common eiders. Can. Field-Nat. 94, 286–292. Mendall, H. L. 1986: Identification of eastern races of the common eider. In A. Reed (ed.): Eider ducks in Canada. Can. Wildl. Serv. Rep. Ser. 47, 82–88. Milne, H. 1974: Breeding numbers and reproductive rate of eiders at the Sands of Forvie National Nature Reserve, Scotland. Ibis 116, 135–154. Milne, H. & Reed, A. 1974: Annual production of fledged young from the eider colonies of the St. Lawrence estuary. Can. Field-Nat. 88, 163–169. Nakashima, D. J. & Murray, D. J. 1988: The common eider (Somateria mollissima sedentaria) of eastern Hudson Bay: a survey of nest colonies and Inuit ecological knowledge. Environmental Studies Revolving Funds Rep. 102. van Noordwijk, A. J., van Balen, J. H. & Scharloo, W. 1988: Heritability of body size in a natural population of the great tit (Parus major) and its relation to age and environmental conditions during growth. Genet. Res. Camb. 51, 149–162. Palmer, R. S. 1976: Handbook of North American birds. Vol. 3. New Haven: Yale University Press. Parker, H. & Holm, H. 1990: Patterns of nutrient and energy expenditures in female common eiders nesting in the high Arctic. Auk 107, 660–668. Paynter, R. A., Jr. 1951: Clutch-size and egg mortality of Kent Island eiders. Ecology 32, 497–507. Prach, R. W., Smith, A. R. & Dzubin, A. 1986: Nesting of the common eider near Hell Gate—Cardigan Strait polynya, 1980-1981. In A. Reed (ed.): Eider ducks in Canada. Can. Wildl. Serv. Rep. Ser. 47, 127–135. Rising, J. D. & Somers, K. M. 1989: The measurement of overall body size in birds. Auk 106, 666–674. Robertson, G. J. 1995a: Factors affecting nest site selection and nesting success in the common eider Somateria mollis- sima. Ibis 137, 109–115. Robertson, G. J. 1995b: Annual variation in common eider egg size: effects of temperature, clutch size, laying date, and laying sequence. Can. J. Zool. 73, 1579–1587. Robertson, G. J. & Cooke, F. 1993: Intraclutch egg-size vari- ation and hatching success in the common eider. Can. J. Zool. 71, 544–549. Robertson, G. J. & Gilchrist, H. G. 1998: Evidence of popula- tion declines among common eiders breeding in the Belcher Islands, Northwest Territories. Arctic 51, 378–385. Robertson, G. J., Watson, M. D. & Cooke, F. 1992: Frequency, timing and costs of intraspecific nest parasitism in the common eider. Condor 94, 871–879. Rockwell, R. F. & Barrowclough, G. 1987: Gene flow and the genetic structure of populations. In F. Cooke & P. A. Buck- ley (eds.): Avian genetics. Pp. 226–255. London: Academic Press. Rohwer, F. C. 1992: The evolution of reproductive parameters in waterfowl. In B. D. J. Batt et al. (eds.): Ecology and man- agement of breeding waterfowl. Pp. 486–539. Minneapolis: University of Minnesota Press. Schmutz, J. M., Robertson, R. J. & Cooke, F. 1983: Colonial nesting of the Hudson Bay eider duck. Can. J. Zool. 61, 2424–2433. Snyder, L. L. 1941: On the Hudson Bay eider. Occas. Pap. R. Ont. Mus. Zool. 6, 1–7. Swennen, C. 1983: Reproductive output of eiders Somateria mollissima on the southern border of its breeding range. Ardea 71, 245–254. Swennen, C. 1990: Dispersal and migratory movements of eiders Somateria mollissima breeding in the Netherlands. Ornis Scand. 21, 17–27. Swennen, C. & van der Meer, J. 1992: Variation in egg size of common eiders. Ardea 80, 363–373. Tiedemann, R., von Kistowski, K. G. & Noer, H. 1999: On sex-specific dispersal and mating tactics in the common eider Somateria mollissima as inferred from the genetic structure of breeding colonies. Behaviour 136, 1145–1155. Wakeley, J. S. & Mendall, H. L. 1976: Migrational homing and survival of adult female eiders nesting in Maine. J. Wildl. Manage. 40, 15–21. Williams, T. D. 1994: Intraspecific variation in egg size and composition among birds: effects on offspring fitness. Biol. Rev. Camb. Philos. Soc. 68, 35–59. Williams, T. D, Lank, D. B., Cooke, F. & Rockwell, R. F. 1994: Fitness consequences of egg-size variation in the lesser snow goose. Oecologia 96, 331–338. Winkler, D. W. & Walters, J. R. 1983: The determination of clutch size in precocial birds. Curr. Ornithol. 1, 33–68.