Content uploaded by Richard B Lanctot
Author content
All content in this area was uploaded by Richard B Lanctot on Sep 19, 2017
Content may be subject to copyright.
HIERARCHICAL SPATIAL GENETIC STRUCTURE OF COMMON EIDERS
(Somateria molliSSima) BREEDING ALONG A MIGRATORY CORRIDOR
R.
—La documentación de la discordancia genética espacial entre poblaciones reproductivas de especies de aves que
crían en el Ártico es importante debido a que los cambios antropogénicos están alterando las conexiones ambientales a escalas micro
y macrogeográficas. Utilizando marcadores moleculares y datos de captura, marcado y recaptura (CMR), estimamos los niveles de
subdivisión poblacional en Somateria mollissima v-nigrum considerando las áreas de reproducción ubicadas en islas de barrera en el
oeste del Mar de Beaufort, Alaska. Las poblaciones estuvieron estructuradas a una escala microgeográfica. Las comparaciones regionales
entre poblaciones que crían en grupos de islas separados por km (Bahía de Mikkelsen y Laguna Simpson) mostraron estructura en
loci microsatélites (FST = ., P < .), un intrón nuclear (FST = ., P = .) y ADN mitocondrial (ΦST = ., P < .). Los datos
de CMR (n = ) no indicaron que existiera dispersión de las hembras entre grupos de islas. La concordancia entre los datos genéticos
y de CMR indica que las hembras que se reproducen en el oeste del Mar de Beaufort son fuertemente filopátricas a grupos de islas
pero no a una isla en particular. A pesar de la aparentemente alta fidelidad a los sitios de las hembras, modelos de flujo génico basados
en coalescencia sugieren que existe disperdsión asimétrica al oeste entre grupos de islas, lo que probablemente esté mediado por el
hecho de que las hembras de la Bahía de Mikkelsen se detienen temprano durante la migración de otoño en la Laguna Simpson para
reproducirse. Alternativamente, las hembras que arriban tarde podrían estar predispuestas a anidar en la Laguna Simpson debido a la
mayor disponibilidad y distribución más amplia del hábitat de anidación. Nuestros resultados indican que, a escalas microgeográficas a
lo largo de corredores de migración establecidos, pueden existir discontinuidades genéticas mediadas por la filopatría de las hembras.
—
744
—
e Auk 126(4):744−754, 2009
e American Ornithologists’ Union, 2009.
Printed in USA.
e Auk, Vol. , Number , page s −. ISSN -, electronic ISSN -. b y e American Ornit hologists’ Un ion. All rights re served. Plea se direct
all requests for permission to photocopy or reproduce article content th rough the University of Cali fornia Press’s Rights and Permissions website, http://www.ucpressjournals.
com/reprintInfo.asp. DOI: ./ auk. .
Estructura Genética Jerárquica Espacial de Poblaciones Reproductivas de Somateria mollissima a lo largo de un
Corredor Migratorio
Sa r a h a. So n S t h a g e n ,1,6 Sa n d y L. ta L b o t ,2 ri c h a r d b. La n c t o t ,3
Kim t. Sc r i b n e r ,4 a n d Ke v i n g. mcCr a c K e n 1, 5
1Institute of Arctic Biology and Department of Biology and Wildlife, University of Alaska Fairbanks, Fairbanks, Alaska 99775, USA;
2Alaska Science Center, U.S. Geological Survey, 4210 University Drive, Anchorage, Alaska 99508, USA;
3U.S. Fish and Wildlife Service, Migratory Bird Management, 1011 E. Tudor Road, MS 201, Anchorage, Alaska 99503, USA;
4Department of Fisheries and Wildlife, 13 Natural Resources Building, Michigan State University, East Lansing, Michigan 48824, USA; and
5University of Alaska Museum, 907 Yukon Drive, Fairbanks, Alaska 99775, USA
6Present address: Alaska Science Center, U.S. Geological Survey, 4210 University Drive, Anchorage, Alaska 99508, USA.
E-mail: ssonsthagen@usgs.gov
A .—Documentation of spatial genetic discordance among breeding populations of Arctic-nesting avian species is important,
because anthropogenic change is altering environmental linkages at micro- and macrogeographic scales. We estimated levels of population
subdivision within Pacific Common Eiders (Somateria mollissima v-nigrum) breeding on barrier islands in the western Beaufort Sea,
Alaska, using molecular markers and capture–mark–recapture (CMR) data. Common Eider populations were genetically structured
on a microgeographic scale. Regional comparisons between populations breeding on island groups separated by km (Mikkelsen Bay
and Simpson Lagoon) revealed structuring at microsatellite loci (FST = ., P < .), a nuclear intron (FST = ., P = .), and
mitochondrial DNA (ΦST = . , P < .). e CMR data (n = ) did not indicate fema le dispersal between island groups. Concordance
between genetic and CMR data indicates that females breeding in the western Beaufort Sea are strongly philopatric to island groups
rather than to a particular island. Despite the apparent high site fidelity of females, coalescence-based models of gene flow suggest that
asymmetrical western dispersal occurs between island groups and is likely mediated by Mikkelsen Bay females stopping early on spring
migration at Simpson Lagoon to breed. Alternatively, late-arriving females may be predisposed to nest in Simpson Lagoon because of
the greater availability and wider distribution of nesting habitat. Our results indicate that genetic discontinuities, mediated by female
philopatry, can exist at microgeographic scales along established migratory corridors. Rec eived O ctob er , accepte d May .
Key words: barrier islands, Beaufort Sea, Common Eider, dispersal, gene flow, microsatellites, mtDNA, philopatry, Somateria mollissima,
spatial genetic structure.
02_Sonsthagen_08-224.indd 744 9/2/09 2:25:26 PM
oc t o b e r 2009
—
co m m o n ei d e r Po P u L a t i o n St r u c t u r e
—
745
T which avian populations are genetically struc-
tured at microgeographic scales is greatly influenced by disper-
sal. For many species, detecting dispersal is difficult, especially for
mobile organisms that may travel long distances prior to and be-
tween breeding attempts. Birds are of particular interest, because
most species that breed in Arctic or temperate regions migrate
to other areas during the non-breeding season and, thus, show
less geographic structure than other vertebrate groups (Avise
). Lack of population structure has been attributed to the en-
vironmental variability of Arctic and temperate regions (Scribner
et al. , Pearce et al. ), which increases dispersal and mi-
gratory behavior in birds and can homogenize gene frequencies
(Winker et al. ). Conversely, many birds exhibit high natal-
and breeding-site fidelity, which is expected to restrict gene flow
among neighboring populations (Avise ), leading to popula-
tion subdivision in the face of countervailing dispersal.
Differences in the degree of philopatry also may exist between
males and females. Females of most waterfowl species typically
exhibit greater natal and breeding philopatry than males (Rohwer
and Anderson ). Males and females typically pair on the win-
tering grounds, and males accompany females back to the females’
natal area to breed. Because ducks from different breeding areas
frequently share a common wintering ground, males may disperse
long distances. Additionally, males may not mate with the same
female each year and, thus, individual males may breed in distant
locations from year to year (Anderson et al. ). erefore, male
dispersal behavior is expected to cause genetic mixing among in-
dividuals from multiple breeding areas. Sex-biased dispersal is
expected to result in distinct patterns of spatial variance when
genetic markers with different modes of inheritance are used to
characterize individuals from different breeding populations.
Common Eiders (Somateria mollissima) have a circumpolar
distribution and inhabit coastal regions throughout the Holarctic
(Goudie et al. ). Pacific Common Eiders (S. m. v- ni gr um ) breed
on the barrier islands of western Canada and Alaska (Goudie et al.
) and have experienced a marked population decline (%)
since the mid-s (Suydam et al. ). Reasons for the decline
are unknown, but recent surveys indicate that the population may
have stabilized (R. Suydam pers. comm.). However, these birds are
long lived and have a low reproductive rate, which may limit popu-
lation growth (Goudie et al. ).
Satellite-telemetry studies indicate that adult female Com-
mon Eiders that nest on islands in the Beaufort Sea may intermix
with other Common Eider populations during migration to win-
tering grounds in the Bering Sea between Alaska and the Chu-
kotka Peninsula, Russia (Petersen and Flint , L. Dickson pers.
comm.). Females originating from Alaskan and Canadian breed-
ing areas likely pair with males from other breeding populations
during winter, thus mediating gene flow through male-biased dis-
persal, even if levels of female dispersal are low. Such gene flow
would be expected to occur despite the observation that all trans-
mitter-equipped females returned to their breeding areas in the
western Beaufort Sea the following summer (Petersen and Flint
, L. Dickson pers. comm.). us, differences between male
and female dispersal should be evident in contrasting levels of
subdivision in maternally and biparentally inherited markers.
Data describing population genetic structure are available
for European Common Eiders (S. m. mollissima) that breed in the
Baltic Sea (colonies –, km apart; Tiedemann et al. )
and Pacific Common Eiders that breed on the Yukon-Kuskok-
wim Delta, Alaska (colonies – km apart; Sonsthagen et al.
). High levels of structure in maternally inherited mitochon-
drial DNA (mtDNA) were observed among colonies in the Bal-
tic Sea and Yukon-Kuskokwim Delta (ΦST = .−., FST =
.−., P < .). Significant, but lower, levels of structure
were detected among Baltic Sea colonies at microsatellite loci
(FST = .−., P < .). Both Tiedemann et al. () and
Sonsthagen et al. () attributed high levels of mtDNA spatial
structure to high rates of female natal philopatry. Tiedemann et
al. () attributed proportionally lower structure at microsatel-
lite loci to nonrandom mating by males on the wintering grounds
(i.e., males mate with females from the same locality more often
than expected).
Information on the degree of spatial population structure in
Pacific Common Eiders in the western Beaufort Sea is of particu-
lar interest to management and industry agencies because of the
proximity of nesting areas to petrochemical projects (Minerals
Management Service ). Philopatry to barrier-island nesting
sites that are close to industrial infrastructure (Reed , Swen-
nen ) makes this population particularly susceptible to hu-
man disturbance and climate change. In addition, increases in
avian and mammalian predators near human developments, in-
cl udin g pet roc hemi ca l ins ta ll ati on s, m ay a dve rsel y aff ec t ne st su c-
cess and duckling survival (Eberhardt et al. , Johnson ,
R. H. Day unpubl. data). Population risks may be exacerbated if
genetically distinct populations occur in proximity to existing
or proposed developments. Evaluation of population structure of
Common Eiders breeding in the western Beaufort Sea thus pro-
vides a means to assess potential effects of oil and gas explora-
tion, and of changes in Arctic ecosystems resulting from climate
change, on populations of this species.
We estimated levels of spatial population structure in Pacific
Common Eiders breeding on barrier islands within two island
groups of the western Beaufort Sea using microsatellite geno-
types and sequence information from the mtDNA control region
and two nuclear introns, coupled with capture–mark–recapture
(CMR) data. We hypothesized that the nuclear markers (micro-
satellites and intron sequences) would show little population ge-
netic structure, because these markers are biparentally inherited
and Common Eiders breeding on these islands share a common
wintering ground with Common Eiders from several other breed-
ing areas. Over time, male dispersal among breeding populations
could contribute to high levels of gene flow throughout the nu-
clear genome. However, we predicted that population structure
would be observed in maternally inherited mtDNA because of the
high degree of female natal and breeding philopatry in Common
Eiders.
Me t h o d s
Sample collection.—Blood or feather samples from breeding fe-
male Common Eiders and eggs from nests were collected during
mark–recapture and monitoring efforts on the barrier islands in
the Beaufort Sea, Alaska, between June and July of –.
Samples were collected from two island groups consisting of
islands (Fig. ). e western group, hereafter called “Simpson
02_Sonsthagen_08-224.indd 745 9/2/09 2:25:27 PM
746
—
So n S t h a g e n e t a L .
—
au K , vo L .
126
Lagoon,” consists of five islands: Stump, “Wannabe,” Egg, Long,
and Spy islands (Fig. A). e eastern group, hereafter called
“Mikkelsen Bay,” consists of seven islands: “Camp,” Point om-
son, Mary Saches, North Star, Duchess, Alaska, and Challenge is-
lands (Fig. B). Distances between islands within each of the two
groups ranged from . to . km, and distances between islands
located in Simpson Lagoon and Mikkelsen Bay ranged from .
to . km. Samples later identified as having identical genotypes
and mother–offspring groups were not included in analyses (see
below).
Females (n = ) were captured on nests using a dip net dur-
i ng i ni ti al ne st s ea rc he s a nd u si ng a bo w n et w he n ne st s w ere re vi s-
ited during late incubation (Sayler ). Blood was collected from
brachial or jugular veins and placed in blood lysis buffer (Long-
mire et al. ). Feather samples (n = ) were collected from
nest bowls of unsampled females and stored in silica gel desiccant
at room temperature. Tissues from eggs (n = from five clutches)
were collected opportunistically from abandoned or depredated
nests to verify that sequences were mitochondrial (see below). Em-
bryos were placed in tissue preservation buffer (. M urea, . M
NaCl, mM EDTA, .% N-lauroyl-sarcosine, and mM tris-
HCl [pH .]; S. Talbot unpubl. data). Genomic DNAs were ex-
tracted using either a “salting out” protocol described by Medrano
et al. (), with modifications described in Sonsthagen et al.
(), or a Qiagen DNeasy Tissue Kit (Qiagen, Valencia, Califor-
nia). Genomic DNA extractions were quantified using fluorom-
etry and diluted to ng L− working solutions.
Microsatellite genotyping.—Primers used for microsatellite
genotyping of Common Eiders (n = ) were obtained via cross-
species screening of microsatellite primers developed for other
waterfowl. We screened individuals at microsatellite loci
reported to be variable in other waterfowl species and selected
microsatellite loci found to be polymorphic: Aph, Aph,
Aph, Aph (Maak et al. ); Bca, Bca, and Hhi (Bu-
chholz et al. ); Cm (Maak et al. ); Sfi (Libants et
al. unpubl. data); and Smo, Smo, Smo, Smo, and Smo
(Paulus and Tiedemann ). Microsatellites were amplified us-
ing the polymerase chain reaction (PCR), and products were elec-
trophoresed following protocols described in Sonsthagen et al.
() for tailed primers (Aph, Aph, Aph, Aph, Cm,
Fig . 1. Location of barrier islands in (A) Simpson Lagoon (western group) and (B) Mikkelsen Bay (eastern group) within the western Beaufort Sea. The
number of Pacific Common Eiders sampled on each island is in parentheses; the first value is the number of samples (blood and feather) genotyped at
14 microsatellite loci, and the second value is the number of samples (blood) sequenced for mtDNA and two nuclear introns. “Wannabe” and “Camp”
islands are designations used by the authors and are not official names of islands.
02_Sonsthagen_08-224.indd 746 9/2/09 2:25:30 PM
oc t o b e r 2009
—
co m m o n ei d e r Po P u L a t i o n St r u c t u r e
—
747
Using only adult breeding females, we calculated allelic rich-
ne ss , in bre ed in g co effic ien t (FIS) , ex pe cte d a nd o bse rv ed h eter ozy -
gosities, Hardy-Weinberg equilibrium (HWE), and linkage
disequilibrium (LD) for each nuclear intron and microsatellite lo-
cus in FSTAT, version .. (Goudet ; see Acknowledgments).
Mitochondrial DNA control region and LMNA and GAPDH se-
quences were tested for selective neutrality and historical fluc-
tuations in population demography, using Fu’s Fs (Fu ) and
Tajima’s D (Tajima ) in ARLEQUIN, version . (Schneider et
al. ). Critical significance values of % require a P value <.
for Fu’s Fs (Fu ). Unrooted phylogenetic trees for each gene
we re c onst ru cte d i n NET WOR K, v ersi on . .. (se e Ac kno wled g-
ments), using the reduced median network (Bandelt et al. ),
to illustrate possible reticulations in the gene trees resulting from
homoplasy or recombination.
Estimation of population subdivision.—e degrees of popu-
lation subdivision among islands and between island groups were
assessed by calculating overall and pairwise FST, RST, and ΦST for
microsatellite and sequence data in ARLEQUIN, adjusting for
multiple comparisons using Bonferroni corrections (α = .).
Pairwise ΦST was calculated using a Tamura-Nei nucleotide sub-
stitution model with an invariant site parameter (Tamura and Nei
), as determined using MODELTEST, version . (Posada
and Crandall ), and Akaike’s information criterion (Akaike
). Because the upper possible FST value for a set of microsatel-
lite loci is usually <. (Hedrick ), we used RECODEDATA,
version . (Meirmans ), to calculate the uppermost limit of
FST for a given data set. Hierarchical analysis of molecular variance
(AMOVA) was performed using ARLEQUIN to determine the
magnitude of spatial variance in haplotypic and allelic frequencies
among populations within and among island groups. In addition,
a nested AMOVA was performed to assess the partitioning of ge-
netic variation within individuals and among individuals within
populations. An isolation-by-distance analysis was performed in
IBD, version . (Bohonak ), with microsatellite data and
nuclear intron data, to determine whether more geographically
distant population pairs are also more genetically differentiated.
IBD tests the statistical significance of the relationship between
genetic and geographic distance using a Mantel test and calculates
slope and intercept from reduced major axis regressions following
Sokal and Rohlf (), with confidence limits.
Estimation of gene flow among populations.—We used MI-
GRATE, version .. (Beerli and Felsenstein , ; see
Acknowledgments), to calculate the number of migrants per gen-
eration (Nem) for microsatellite and nuclear intron data and num-
ber of female migrants per generation (Nfm) for mtDNA between
the two island groups. Individuals breeding in Mikkelsen Bay and
Simpson Lagoon, respectively, were pooled and treated as two
separate populations. Full models, θ (Ne; composite measure of
effective population size and mutation rate), and all pairwise mi-
gra tio n p ara me te rs ( M), were estimated individually from the data
and compared with restricted island models for which parameters
θ and M were symmetrical among populations.
MIGRATE was run using maximum-likelihood search pa-
rameters: short chains (, used trees out of , sampled),
five long chains (, used trees out of , sampled), and
five adaptively heated chains (start temperatures: , ., , , and
; swapping interval = ). Full models were run three times to
Smo, Smo, Smo, Smo, and Smo) and Pearce et al. ()
for direct-labeled primers (Bca, Bca, Hhi, and Sfi). For
quality-control purposes, % of the samples were randomly se-
lected, re-amplified, and genotyped in duplicate.
Mi to cho nd ria l D NA a nd nuc le ar i nt ron se que nc ing .—We a m-
plified and sequenced a -base-pair (bp) portion of the control-
region domain I and II (Baker and Marshall ) using primer
pairs (L and H) and protocols described in Sonsthagen et
al. (). Sequences from opposite strands were assembled using
SEQUENCHER, version .. (Gene Codes, Ann Arbor, Michi-
gan). Because of the existence of nuclear pseudogenes in avian spe-
cies (Sorenson and Fleischer ), we verified that the amplified
sequences were mtDNA control region by comparing sequences
from heart and blood samples from five putative mother–offspring
groups (e.g., Pearce et al. ). Individuals that contained dou-
ble peaks within mtDNA electropherograms were resequenced.
If co-amplified peaks were still detected, presumably because of
nuclear pseudogenes present in this species (Tiedemann and von
Kistowski ) or heteroplasmy, those individuals were removed
(~%). Sequences were deposited in GenBank (accession numbers
EU–EU and GQ–GQ).
Six nuclear introns were screened for variability in Com-
mon Eiders (Sonsthagen et al. ), and two polymorphic in-
trons were selected for sequencing, lamin A (LMNA; bp) and
glyceraldehyde--phosphate dehydrogenase (GAPDH; –
bp) (McCracken and Sorenson ). Introns were amplified us-
ing PCR and sequencing protocols described by Sonsthagen et
al. (). Sequences that contained double peaks of approxi-
mately equal peak height, indicating the presence of two al-
leles, were coded with International Union of Pure and Applied
Chemistry degeneracy codes and treated as polymorphisms.
Gaps were coded as a fifth character state. Sequences were de-
posited in GenBank (accession numbers GQ–GQ
and GQ–GQ).
Estimation of genetic diversity.—To determine whether the
same individual was sampled across multiple years (between
feather and blood or feather and feather samples), probabilities of
identity for a randomly mating population (PID) and among sib-
lings (PIDsib) were calculated in GIMLET, version .. (Valière
), using genotypes from the microsatellite loci.
Data from islands with low sample sizes were pooled on the
basis of geographic proximity (i.e., females captured within km
of each other on neighboring islands). Samples from Challenge
Island were pooled with those from Alaska Island, samples from
Mary Saches Island were pooled with those from North Star Is-
land, and samples from “Wannabe Island” were pooled with those
from Egg Island (Fig. ).
Allelic phases for LMNA and GAPDH were inferred from
diploid sequence data using PHASE, version . (Stephens et al.
), which uses a Bayesian approach to reconstruct haplotypes
from population genotypic data and allows for recombination
and the decay of linkage disequilibrium (LD) with distance. e
PHASE analysis (parameters: , iterations with , burn-in
iterations) was repeated three times to ensure consistency across
runs. A four-gamete test was performed in DNASP, version .
(Rozas et al. ), to estimate the minimum number of recom-
bination events and identify inferred break points within nuclear
intron sequences.
02_Sonsthagen_08-224.indd 747 9/2/09 2:25:31 PM
748
—
So n S t h a g e n e t a L .
—
au K , vo L .
126
ensure the convergence of parameter estimates. Restricted mod-
els were run once. Competing models were evaluated for good-
ness-of-fit given the data using a log-likelihood ratio test and a
chi-square distribution with the degrees of freedom equal to the
difference in the number of parameters estimated in the two mod-
els (Beerli and Felsenstein ).
Re s u l t s
Genetic Diversity
Microsatellites.—e number of alleles per locus at the micro-
satellite loci ranged from to , with an average of . alleles
per locus. Allelic richness for each population ranged from .
to .. e observed heterozygosity for each population ranged
from .% to .%, with an overall value of .%. e inbreeding
coefficients (FIS) ranged from −. to . across all islands,
with an overall value of ., and did not differ significantly from
zero (P > .).
Nuclear introns.—Twenty-five LMNA alleles were recon-
structed from individuals using PHASE (Fig. A). Sixty
individuals (%) were homozygous at all variable sites, and in-
dividuals (%) were heterozygous at one site. Pair probabilities of
reconstructed haplotypes for individuals that were heterozygous
for more than one site ranged from . to . (n = ), except for
two individuals with haplotype probabilities of . and .. e
background recombination rate (ρ) was ., with factors exceed-
ing ρ ranging from . to . among variable sites. A mini-
mum of five recombination events was identified in DNASP, with
break points occurring between sites and , and ,
and , and , and and .
Twenty-two GAPDH alleles were reconstructed from
individuals (Fig. B). Six individuals (%) were homozygous at
all variable sites, and one individual (%) was heterozygous at
one site. Pair probabilities of all other reconstructed haplotypes
ranged from . to . (n = ) and from . to . (n = ),
which may be attributable to potentially high levels of recom-
bination occurring within this marker (.–. factors ex-
ceeding ρ = ., between variable sites) and autapomorphies
(single novel polymorphisms occurring on one allele in one in-
dividual). At least two recombination events were identified in
DNASP, with break points occurring between sites and
and sites and .
Haplotype (h) and nucleotide (π) diversity ranges were .–
. and .–., respectively, for LMNA, and .–.
and .–., respectively, for GAPDH (Table ). Observed
and expected heterozygosity for LMNA were .% and .%,
respectively, which significantly deviated from HWE (P = .).
Observed and expected heterozygosity for GAPDH was .% and
.%, respectively, which also significantly deviated from HWE
(P < .). Significantly negative values for Fu’s Fs (P < .) were
observed for North Star and Mary Saches, Duchess, and Long
(Table ) islands, indicating an excess of rare alleles, which may
result from population expansion.
Mitochondrial DNA.—Eleven mtDNA control-region haplo-
types were identified from individuals (Fig. C) defined by
variable sites. Haplotype and nucleotide diversity were high for
most populations, with values ranging from . to . and
from . to ., respectively (Table ). Spy Island possessed
a single haplotype. Other islands were represented by two to six
haplotypes. Tajima’s D and Fu’s Fs were not significant (Table ).
Fig . 2. Unrooted parsimony tree illustrating relationships of (A) 25 lamin A (LMNA) alleles, (B) 22 glyceraldehyde-3-phosphate dehydrogenase (GAP-
DH) alleles, and (C) 11 mtDNA control-region haplotypes. The 95% probability set of parsimony trees is illustrated with bold branches, with the size
of the circle node corresponding to the frequency of each allele. Dashed lines indicate alternative branching patterns and possible reticulations. Small
black circles indicate intermediate ancestral alleles that were not sampled.
02_Sonsthagen_08-224.indd 748 9/2/09 2:25:33 PM
oc t o b e r 2009
—
co m m o n ei d e r Po P u L a t i o n St r u c t u r e
—
749
Population Structure
Microsatellites.—After removing mother–offspring groups and
identical genotypes (n = ) at microsatellite loci, the overall FST
(., P < .) was significant. e upper limit of FST calculated
with RECODEDATA was .. erefore, the overall FST of .
accounts for .% of the maximum possible level of genetic struc-
ture. We did not observe a significant level of differentiation using
RST (., P > .). Significant differences in allelic frequencies
were observed within and among island groups on the basis of FST
(.–.; Table ). No significant comparisons were observed
among adjacent islands. e regional comparison between Mik-
kelsen Bay and Simpson Lagoon was also significant (FST = .;
Table ). e significant FST values we observed, albeit low, are
noteworthy given the geographic proximity of the islands (.–
km apart; Wright ). Hierarchical analyses of molecular
variance uncovered low but significant variance within islands,
among islands within a group, and within individuals (Table ).
We found no evidence of isolation-by-distance correlations
between genetic and geographic distances (r = ., P = .).
Nuclear introns.—Significant differences in the spatial distri-
bution of allelic frequencies were observed for LMNA among is-
lands (overall ΦST = ., P = .), between Mikkelsen Bay and
Simpson Lagoon (ΦST = .; Table ), and among islands within
these waters (ΦST = .−.; Table ). AMOVA analyses also
detected significant variance among islands within island groups,
within islands, among individuals within islands, and within in-
dividuals (Table ). is pattern appears to be driven by a single
C/T nucleotide polymorphism at one of polymorphic positions
in LMNA: site (FST = . ± . between Mikkelsen Bay
and Simpson Lagoon). Pairwise FST comparisons among islands
ranged from . to . (Table ). However, site was mono-
morphic for all islands in Mikkelsen Bay.
No population structure was observed for GAPDH (overall
ΦST = −., P = .; Table ). As with the microsatellite data,
we detected no significant correlations between genetic and geo-
graphic distances for LMNA and GAPDH combined (r = . , P =
.) or analyzed separately (LMNA r = −. , P = .; GAPDH
r = ., P = .).
Mitochondrial DNA.—AMOVA analysis detected signifi-
cant variance within islands and among islands within each is-
land group (Table ). Mean inter-island variance in haplotypic
frequency was low but significant (ΦST = ., P = .). Regional
comparison between Mikkelsen Bay and Simpson Lagoon was
also significant (ΦST = ., P < .; Table ). High levels of ge-
netic discordance were observed between Duchess Island, located
in Mikkelsen Bay, and all four islands located in Simpson Lagoon
(ΦST = .−.; Table ).
Estimates of Gene Flow
Asymmetrical dispersal was observed between Mikkelsen Bay and
Simpson Lagoon across all three marker types, and the full model
(all parameters allowed to vary independently) was found to have
significantly higher likelihoods than the restricted island model
(equal interpopulation migration rate and equal θ across popu-
lations; Table ). e biases in the variances and the means indi-
cate that, on average over generations, gene flow has been greater
from Mikkelsen Bay to Simpson Lagoon (east to west) than vice
tab L e 1. Estimates (± SD) of nucleotide (π) and haplotype (h) diversity for lamin A (LMNA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and mtDNA control region for Pacific
Common Eiders sampled at islands in the western Beaufort Sea (Fig. 1). Fu’s Fs and Tajima’s D values in bold indicate significance (Fs, P < 0.02; D, P < 0.05).
Spy Long
“Wannabe”
and Egg Stump
Challenge
and Alaska Duchess
Mary Saches
and North Star Pt. Thomson “Camp”
LMNA
n915 13 24 314 5 6 7
h0.876 ± 0.045 0.915 ± 0.027 0. 818 ± 0.052 0.849 ± 0.036 0.600 ± 0 .215 0.884 ± 0.034 0.844 ± 0.103 0.864 ± 0.072 0.901 ± 0.0 47
π0.008 ± 0.005 0.009 ± 0.006 0.007 ± 0.00 4 0.008 ± 0.005 0.005 ± 0.0 04 0.007 ± 0.0 05 0.005 ± 0.004 0.006 ± 0.00 4 0.009 ± 0.00 6
Fu’s Fs −1.896 −4.943 −3.033 −3.499 0.381 −4.704 −2.690 −2.00 4 −1.503
Tajima’s D0.070 −0.119 − 0.126 − 0.16 9 0.338 0.226 −0.682 − 0.212 − 0.147
GAPDH
n613 13 18 311 4 5 6
h0.954 ± 0.047 0.886 ± 0.036 0.874 ± 0.032 0.90 0 ± 0. 024 0.933 ± 0.12 2 0.909 ± 0.037 0.927 ± 0.0 84 0.911 ± 0.077 0.924 ± 0.058
π0.007 ± 0.005 0.007 ± 0.004 0.006 ± 0.0 04 0.006 ± 0.004 0.008 ± 0.00 6 0.006 ± 0.0 04 0.007 ± 0.005 0.009 ± 0.0 06 0.006 ± 0.0 04
Fu’s Fs −2.162 −0.996 −0.773 −2.850 −1.466 −3.602 −0.930 − 0.907 −1.5 84
Tajima’s D−0.964 −0.306 −0.836 −0.853 −0.631 −0.989 −0.856 −0.810 − 0.323
MtDNA
n713 13 19 311 5 6 6
h0.000 ± 0.000 0.654 ± 0.106 0.526 ± 0 .152 0.579 ± 0. 114 0.667 ± 0.314 0.891 ± 0.06 3 0.400 ± 0.237 0.800 ± 0 .172 0.333 ± 0.215
π0.000 ± 0.000 0.003 ± 0.002 0.003 ± 0.0 02 0.0 03 ± 0.0 02 0.0 01 ± 0.001 0.009 ± 0.005 0.001 ± 0.001 0.007 ± 0.00 4 0.002 ± 0.002
Fu’s Fs — 0.399 0.868 0 . 1 8 3 0.201 0.543 0 . 0 9 0 0 . 5 6 7 2.1 39
Tajima’s D0.000 −1.249 − 0.167 −0.869 0.000 0.591 − 0 .817 − 0.516 −1. 295
02_Sonsthagen_08-224.indd 749 9/2/09 2:25:34 PM
750
—
So n S t h a g e n e t a L .
—
au K , vo L .
126
versa (Table ). Nem and θ (Ne) values calculated in MIGRATE
from microsatellite genotypes, mtDNA, and nuclear intron se-
quence data ranged from . to . migrants per generation
from Simpson Lagoon to Mikkelsen Bay, with θ ranging from
. to ., and from . to . migrants per generation
from Mikkelsen Bay to Simpson Lagoon, with θ ranging from
. to . (Table ).
Female Site Fidelity
Using GIMLET, we calculated an overall PID of . × − for a
population composed of randomly mating individuals and . ×
− for siblings using genotypes collected from microsatellite
loci. ese denominator values are much larger than the popula-
tion breeding on islands in the western Beaufort Sea (~ nests
found on the islands; Noel et al. ), which gave us confidence
that identical genotypes for samples taken in different years were
from the same individual.
roughout the course of the four-year study, females
were detected breeding in multiple years (based on observations
of banded individuals and genetic techniques). Most (n = of ;
%) nested on the same islands, whereas (%) moved to a new
island within the same island group. Inter-nest distances between
breeding attempts ranged from . to . km on the basis of ob-
servations using band CMR data (J. Reed unpubl. data) and from
. to . km on the basis of a comparison of genetic samples. We
found no evidence of female dispersal between Mikkelsen Bay and
Simpson Lagoon island groups. Females that dispersed to a differ-
ent nest site between years generally moved to an adjacent island
within the same island group to breed ( of , or %). However,
three females breeding in Mikkelsen Bay moved from islands in
the ocean to islands near the coast (Alaska Island to Pt. omson
Is la nd , . km an d . k m; D uc hes s I sl an d t o “C am p I sla nd ,” . km ;
Fig. ). One female breeding in Simpson Lagoon dispersed three
islands east of its original nest site (Long Island to Stump Island,
. km).
di s c u s s i o n
Population genetic structure.—Population subdivision was ob-
served in three types of genetic markers in the Pacific Common
Eiders sampled in the western Beaufort Sea. Higher levels of spa-
tial structure were observed at maternally inherited mtDNA than
at biparentally inherited nuclear introns and microsatellite loci for
inter-island comparisons, which is consistent with our prediction
tab L e 2. Pairwise inter-island FST calculated from 14 microsatellite loci (above diagonal) and pairwise inter-island φST calculated from mtDNA control region (below diagonal) for Pacific
Common Eiders breeding in the western Beaufort Sea, Alaska. Significant comparisons (α = 0.05) are in bold.
Islands Spy Long Egg “Wannab e” Stump Challenge Alaska Duchess North Star Mary Sache s Pt. Thomson “Camp”
Spy —0.009 0.010 0.000 0.008 0.012 0.007 0.000 0.016 0.007 0.007 0.008
Long 0.029 —0.006 0.001 0.005 0.016 0.010 0.001 0.014 0.002 0.001 0.0 07
Egg 0.002 − 0.003 0.000 0.002 0.003 0.008 0. 013 0.000 0.003
“Wannabe” 0.068 0.037 — — 0.001 0.008 0.003 −0.007 0. 011 0.006 − 0.010 −0.004
Stump 0.008 −0.023 −0 .010 NA —0.007 0.007 0.001 0.003 0.013 −0.002 0.000
Challenge —0.002 0.006 0. 013 0 .013 −0.001 0.0 07
Alaska 0.300 − 0.19 9 −0.002 NA −0.141 NA —0.002 0.001 0.0 06 −0.003 0.004
Duchess 0.271 0.230 0.135 NA 0.183 NA 0.116 —0.0 07 0.002 −0.004 0.002
North Star — 0 .017 0.004 0.002
Mar y Sach es 0.073 −0.095 0.040 NA −0.066 NA −0.299 0.201 NA —−0.002 0.020
Pt. Thomson 0 .127 0.026 −0.082 NA −0.024 NA −0.174 − 0. 024 NA 0.032 —−0.005
“Camp” 0.028 0.000 −0.082 NA −0.083 NA − 0.058 0.0 97 NA −0.018 −0.085 —
tab L e 3. Hierarchical nested analyses of molecular variance (AMOVA) of
allelic and haplotypic frequencies (LMNA = lamin A, GAPDH = gly cera l-
dehyde-3-phosphate dehydrogenase, and µsats = microsatellite loci) for
Pacific Common Eiders sampled from islands within Mikkelsen Bay and
Simpson Lagoon. Significant comparisons (P < 0.05) are in bold.
Sources of variation µsats LMNA GAPDH MtDNA
Among groups (FCT)0.001 −0.003 −0.000 0.028
Among islands within groups (FSC)0.004 0.025 −0.019 0.055
Within populations (FST)0.004 0.022 − 0.019 0.082
Among individuals within islands (FIS )0.014 0.682 0.073 —
Within individuals (FIT)0.019 0.678 0.051 —
02_Sonsthagen_08-224.indd 750 9/2/09 2:25:35 PM
oc t o b e r 2009
—
co m m o n ei d e r Po P u L a t i o n St r u c t u r e
—
751
and known patterns of male-biased dispersal (Swennen ).
We observed higher differentiation among colonies separated by
similar geographic distance than that reported in European Com-
mon Eiders in the Baltic Sea (Tiedemann et al. ) and Pacific
Common Eiders in the Yukon-Kuskokwim Delta (Sonsthagen et
al. ). Tiedemann et al. () proposed that differences in
migration phenology among geographic regions coupled with a
selective advantage for early pair formation were the main mecha-
nisms promoting genetic subdivision among populations in the
Baltic Sea. ese mechanisms may not provide a satisfactory ex-
planation in our study area, because differences in migratory phe-
nology do not appear to occur among island groups, and satellite
telemetry data indicate that there is no difference in the start of
autumn migration among Common Eiders breeding in northern
and western Alaska (~, km apart; Petersen and Flint ).
Lack of differences in migration phenology may explain, in part,
the lower levels of differentiation observed at nuclear markers, be-
cause Common Eiders from different island groups likely admix
on the wintering grounds.
We observed higher levels of population structure at mtDNA.
Population subdivision was observed between Duchess Island, lo-
cated in Mikkelsen Bay, and all islands located in Simpson La-
goon. Although we did not detect significant pairwise comparisons
a mon g a ll i sl an ds in M i kke ls en B ay and Si mp son La goo n, w e b eli ev e
that the structuring observed is noteworthy given that Duchess Is-
land is the only island that contains a colony of breeding Common
Eiders in Mikkelsen Bay. Nests on the other islands in Mikkelsen
Bay were scattered thinly, with relatively few nests per island. e
presence of a colony on Duchess Island is likely driving the signifi-
cant pairwise comparisons observed at this marker, given that when
Duchess Island was removed from the analysis the overall ΦST was
no longer significant (mtDNA ΦST = −., P = .). Common Ei-
ders are typically colonial nesters (Goudie et al. ), and the low-
density nesters in Mikkelsen Bay could be “overflow” from Duchess
Island, though demographic data are needed to support this hy-
pothesis. Although there also are colonies on three islands in Simp-
son Lagoon (Egg, Long, and Stump islands), these colonies occur
on islands that are adjacent to each other (within several hundred
meters) and, thus, unlikely to become genetically isolated because
birds may disperse among islands. Although we do not have natal-
dispersal data for this population, we have breeding-dispersal dis-
tances from recaptured individuals. Given that no breeding females
were obser ved d ispersi ng between Mikke lsen Bay and Sim pson L a-
goon, we hypothesize that females breeding in the western Beaufort
Sea are strongly philopatric to island groups rather than to a partic-
ular island. is differs from observations of S. m. d res seri breeding
tab L e 4. Pairwise inter-island φST values calculated for lamin A (complete sequence below diagonal and site 116 above diagonal) for Pacific Common
Eiders breeding on islands in the western Beaufort Sea, Alaska. Significant comparisons (α = 0.05) are in bold.
Simpson Lagoon Mikkelsen Bay
Islands Spy Long
“Wannabe”
and Egg Stump
Alaska and
Challenge Duchess
Mar y Sach es
and North Star Pt. Thomson “Camp”
Spy —−0.086 0.284 0.352 0.053 0.330 0.151 0.182 0.208
Long 0.022 —0.169 0. 217 − 0.006 0.205 0.078 0.102 0.121
“Wannabe” and Egg 0.106 0.025 —−0.024 −0.091 −0.007 −0.050 −0.040 −0.032
Stump 0.092 0.022 − 0. 012 —−0.092 − 0.012 −0.051 − 0.042 −0.035
Alaska and Challenge 0.148 0.0 24 −0.042 0.003 — — — — —
Duchess 0.053 0.007 − 0.002 − 0.007 0 .024 — — — —
Mary Saches and North Star 0.033 0.0 03 0.003 −0.025 0.119 −0.029 — — —
Pt. Thomson 0.089 0.024 0.059 0.037 0 .173 0.009 0.034 — —
“Camp” 0.052 0.026 0.027 0.074 0.088 0.001 −0.0 07 −0.034 —
tab L e 5. Comparison of alternative models of gene flow between Pacific Common Eiders in Mikkelsen Bay and Simpson Lagoon. Full-model migration
matrix (allowing all parameters to vary independently) and restricted-model (symmetrical gene flow) migration rates calculated from 14 microsatellite
loci, nuclear introns lamin A and glyceraldehyde-3-phosphate dehydrogenase, and mtDNA control region, were evaluated for significance using a log
likelihood ratio test (df = 4). Ninety-five percent confidence intervals are in parentheses.
Simpson Lagoon to Mikkelsen Bay Mikkelsen Bay to Simpson Lagoon
Marker Hypothesis Ln(L) PNfm or Nem θNfm or Nem θ
Microsatellites Full −8,782.1 <0.001 18.8 0.683 2 7.1 0.635
( 17. 8 − 2 0 . 3 ) (0.650− 0.717) (25.4−29.6) (0.612−0.659)
Restricted −8,888.0 78.3 2.247 78.3 2.247
MtDNA Full 1.9 <0.001 5.1 0.001 24.4 0.006
(0.9 −28.1) (0.000−0.002) (2.4−95.9) (0.005− 0.015)
Restricted −12.0 12.3 0.003 12. 3 0.003
Nuclear introns Full −401.8 <0.0 01 24.2 0.003 34.2 0 .010
(18.6 −31.5) (0.003−0.004) (26.7−43.7) (0.009−0.011)
Restricted −468.9 22.3 0.006 22.3 0.006
02_Sonsthagen_08-224.indd 751 9/2/09 2:25:36 PM
752
—
So n S t h a g e n e t a L .
—
au K , vo L .
126
in Maine (Wakeley and Mendall ). Distances among islands
in Maine are similar to those observed in our study (.–. km
apart); however, % of females returned to their previous breed-
ing island, and only % dispersed to neighboring islands, compared
with the % observed in our study (Wakeley and Mendall ).
Over many generations, females dispersing among neighboring is-
lands would have a homogenizing effect within island groups, while
maintaining population subdivision between island groups.
Behavioral responses to a stochastic Arctic environment
may play a role in the inferred degree of breeding philopatry we
observed between populations in Maine and the Beaufort Sea.
Common Eider nests in the western Beaufort Sea are associated
with driftwood (Goudie et al. , Johnson ), and changes
in driftwood distribution and abundance will affect where Com-
mon Eiders nest. Storms dramatically modify the shape and to-
pography of these barrier islands, thus changing where available
habitat is located annually (Noel et al. ). Finally, Common Ei-
ders breeding on the Beaufort Sea postpone nesting attempts un-
til the island is surrounded by open water, reducing predation risk
(Schamel ). Islands located in the same vicinity may not be
surrounded by water at the same time (S. Sonsthagen pers. obs.).
erefore, in years when ice break-up is late, Common Eiders may
initiate nesting on the first “suitable” island, regardless of where
they nested in previous years or where they hatched, because of
the presumed selective advantage of nesting early (Milne ).
e population structure observed in the present study also
was likely influenced by differences in the rates of lineage sort-
ing between the nuclear and mitochondrial genomes (Avise ).
Mitochondrial DNA has a lower effective population size than nu-
clear DNA. us, genetic drift has a larger effect on mtDNA than
on nuclear DNA (Avise ), which translates into higher esti-
mates of population subdivision (FST). e effects of lineage sort-
ing and sex-biased differences in philopatry on spatial genetic
subdivision are not mutually exclusive, and both may play a role in
the degree of population structure observed.
Gene flow.—We do not completely understand the factors
that influence the degree of migratory and homing behavior in
this species. Common Eiders appear to move the minimum dis-
tance to wintering areas (Petersen and Flint ), and the degree
of movement is likely environmentally induced (Swennen ).
is may explain, in part, the directionality of gene flow observed.
Microsatellite, nuclear intron, and mtDNA loci all indicate sig-
nificant asymmetrical gene flow, such that, on average, more in-
dividuals dispersed from Mikkelsen Bay to Simpson Lagoon (i.e.,
east to west). Because islands become ice-free about two weeks
earlier in Simpson Lagoon than in Mikkelsen Bay (Schamel ),
female Common Eiders from Mikkelsen Bay may benefit from
short-stopping migration and initiating nests at Simpson Lagoon,
enabling them to hatch broods sooner. Should these females suc-
cessfully hatch young, they may be more likely to nest in Simpson
Lagoon in succeeding years (Milne ). us, earlier nest initia-
tion and previous nest success may be factors that influence females
that hatched in Mikkelsen Bay to breed in Simpson Lagoon and to
return there in successive years. e relatively earlier ice break-up in
Simps on Lag oon may have d rive n the w estward bias in di spersal.
Alternatively, Common Eiders may be dispersing west be-
cause of the more abundant and broader distribution of available
nesting habitat in Simpson Lagoon (based on number of nests
found in each island group between and [Noel et al.
] and driftwood distribution [R. Lanctot and S. Sonsthagen
pers. obs.]). is differential distribution of nesting habitat may
especially affect first-time breeding females from Mikkelsen Bay
that tend to arrive later from the wintering grounds (Johnson et al.
), because females arriving earlier on the breeding grounds
may have already secured suitable nest sites.
Comparison with other waterfowl.—e fine-scaled spatial
genetic structuring we observed in Pacific Common Eiders breed-
ing on island groups km apart in the western Beaufort Sea is
noteworthy, especially when compared with other Arctic-nesting
waterfowl. King Eiders (S. spectabilis) sampled from Russia,
Alaska, and Canada exhibited evidence of high levels of dispersal
among western and eastern Arctic populations using mtDNA and
six nuclear microsatellite loci (Pearce et al. ). Among Har-
lequin Duck (Histrionicus histrionicus) populations that breed in
Alaska, no evidence of genetic discordance among sampled sites
was found in analyses of four autosomal microsatellite loci, two
Z-specific microsatellite loci, and mtDNA (Lanctot et al. ).
Lack of structure was attributed to recent range expansion. Stell-
er’s Eiders (Polysticta stelleri) that breed in Alaska and Russia ex-
hibit low population subdivision at nuclear markers (Pearce et
al. ). However, estimates based on mtDNA were not signif-
icant. Scribner et al. () documented high levels of differen-
tiation in mtDNA among sampled sites in Spectacled Eiders (S.
fisheri), but they did not detect any differences in allelic frequen-
cies at autosomal or Z-linked microsatellite loci. Canada Geese
(Branta canadensis) exhibit high levels of genetic differentiation
among sampled sites at autosomal and Z-linked microsatellite loci
and mtDNA (Scribner et al. ). Several recent studies of dab-
bling ducks (Anas spp.) also have shown a range of differentiation
at continental scales (Kulikova et al. , Peters et al. , Mc-
Cracken et al. ). Although these studies all documented sig-
nificant differences in gene frequencies among sampled sites, they
were conducted at much larger spatial scales than our study.
Conclusion.—Oil and gas development has occurred and is
being planned for barrier islands where Common Eiders currently
nest in the Beaufort Sea. We observed females dispersing ≤. km
between years to nest. us, females may exhibit some plasticity
in choosing their nesting sites; however, it is unclear how much
plasticity occurs at the population level. e presence of micro-
geographic genetic structure between the island groups for all
marker types assayed in the present study indicates that this plas-
ticity may be limited. Furthermore, it is unclear how much genetic
variation might be lost should these barrier-island populations be
extirpated. Additional genetic analyses are needed to quantify the
genetic diversity and levels of population structure present across
the Pacific Common Eider’s entire breeding range.
Ac k n o w l e d g M e n t s
Funding was provided by Mineral Management Service (-
--CA-) through the Coastal Marine Institute, Univer-
sity of Alaska Fairbanks, U.S. Geological Survey, Alaska EPSCoR
Graduate Fellowship (NSF EPS-), University of Alaska
Foundation Angus Gavin Migratory Bird Research Fund, and
BP Exploration (Alaska), Inc. We thank all the U.S. Geological
Survey researchers and biologists that worked on the Beaufort
02_Sonsthagen_08-224.indd 752 9/2/09 2:25:37 PM
oc t o b e r 2009
—
co m m o n ei d e r Po P u L a t i o n St r u c t u r e
—
753
Sea Common Eider project, especially P. Flint, J. C. Franson, D.
LaCroix, and J. Reed, as well as J. Gust and G. K. Sage, who pro-
vided laboratory assistance. J. Gleason, C. Monnett, J. Pearce, M.
Petersen, and J. Gust, U.S. Geological Survey, and three anony-
mous reviewers provided valuable comments on earlier drafts
of the manuscript. MIGRATE software and documentation are
available from P. Beerli at popgen.scs.fsu.edu/migrate.down-
load.html. NETWORK is available at www.fluxus-engineering.
com. FSTAT is available from J. Goudet at www.unil.ch/popgen/
softwares/fstat.htm.
li t e R A t u R e c i t e d
A, H.
. A new look at the statistical model identification.
IEEE Transactions on Automatic Control :–.
A, M. G., J. M. R, F. C. R
. . Philo-
patry, dispersal, and the genetic structure of waterfowl popula-
tions. Pages – in Ecology and Management of Breeding
Wat erfowl (B. D. J. Batt, A. D. Afton, M. G. Anderson, C. D.
Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, Eds.). Uni-
versity of Minnesota Press, Minneapolis.
A, J. C.
. ree fundamental contributions of molecular
genetics to avian ecolog y and evolution. Ibi s :–.
A, J. C.
. Molecular Markers, Natural History, and Evolu-
tion, nd ed. Sinauer Associates, Sunderland, Massachusetts.
B, A. J., H. D. M
. . Mitochondrial control
region sequences as tools for understanding evolution. Pages –
in Avian Molecular Evolution and Systematics (
D.
Mindell,
Ed.). Ac ademic Pres s, Sa n Diego, C alifor nia.
B, H.-J., P. F, B. C. S, M. B. R
.
. Mitochondrial portraits of human populations using
median networks. Genetics :–.
B, P., J. F
. . Maximum-likelihood estima-
tion of migration rates and effective population numbers in two
populations using a coalescent approach. Genetics :–.
B, P., J. F
. . Maximum likelihood estima-
tion of a migration matrix and effective population sizes in n sub-
populations by using a coalescent approach. Proceedings of the
National Academy of Sciences USA :–.
B, A. J.
. IBD (isolation by distance): A program for
analyses of isolation by distance. Journal of Heredity :–.
B, W. G., J. M. P, B. J. P, K. T. S-
. . Dinucleotide repeat polymorphisms in waterfowl
(family Anatidae): Characterization of a sex-linked (Z-specific)
and autos omal loci. Animal Ge netics :–.
E, L. E., W. C. H, J. L. B, R. A. G,
E. E. H
. . Arctic fox home range characteris-
tics in an oil-development area. Journal of Wildlife Management
:–.
F , Y. -X
. . Statistical tests of neutrality of mutations against pop-
ulation growth, hitchhiking and background selection. Genetics
:–.
G, J.
. FSTAT (version .) : A co mput er p rogra m to cal cu-
late F-statistics. Journal of Heredity :–.
G, M. L., G. J. R, A. R
. . Common
Eider (Somateria mollissima). In e Birds of North America,
no. (
A.
Poole and
F.
Gill, Eds.). Birds of North America,
Philadelphia.
H, P. W.
. A standardized genetic differentiation mea-
sure. Evolution :–.
J, D. H., J. D. N, M. D. S
. . Popu-
lation dynamics of breeding waterfowl. Pages – in Ecol-
ogy and Management of Breeding Waterfowl (
B. D. J.
Batt,
A.
D.
Afton
, M. G.
Anderson,
C. D.
An kn ey,
D. H.
Johnson,
J. A.
Kadlec
,
and
G. L.
Krapu, Eds.). University of Minnesota Press,
Minneapolis.
J, S. R.
. Pacific Eider. Pages – in e Natural
History of an Arctic Oil Field: Development and the Biota (
J. C.
Truet t
and
S. R .
Johnson, Ed s.). Academic Press, San Diego, Ca li-
fornia.
K, I. V., S. V. D, D. D. G, R. J. H,
S. R, M. D. S, K. W, Y. N. Z,
K. G. MC
. . Phylogeography of the Mal-
lard (Anas platyrhynchos): Hybridization, dispersal, and lin-
eage sorting contribute to complex geographic structure. Auk
:–.
L, R., B. G, K. S, S. T, B. P,
D. E, D. Z
. . Harlequin Duck recov-
ery from the Exxon Valdez oil spill: A population genetics per-
spective. Auk :–.
L, J. L., A. K. L, N. C. B, J. M. B,
L. M. C, M. D. J, L. J. M, J. M, R. L. R-
, F. A. R
, . . Isola tion and mol ecu lar char-
acterization of a highly polymorphic centromeric tandem repeat
in the family Fa lconidae. Genomics :–.
M, S., K. N, G. L, R. G-
. . First seven microsatellites developed for the Peking
Duc k (Anas platyrhynchos). A nimal Geneti cs :.
M, S., K. W, S. W, K. N
. .
Isolation and characterization of microsatellites in the Peking
Duck (Anas platyrhynchos) and their application in other water-
fowl sp ecies. Molecu lar E colog y Notes :–.
MC, K. G., M. B, K. P. J, M. K. K-
, J. T, T. H. V, R. E. W, J. L. P
.
. Gene flow in the face of countervailing selection: Adapta-
tion to high-altitude hypoxia in the bA hemoglobin subunit of
Yellow-billed Pintails in the Andes. Molecular Biology and Evo-
lution :–.
MC, K. G., M. D. S
. . Is homoplasy or
lineage sorting the source of incongruent mtDNA and nuclear
gene trees in the stiff-tailed ducks (Nomonyx–Oxyura)? System-
atic Biolo gy : –.
M, J. F., E. A, L. S
. . DNA extraction
from nucleated red blood cells. Biotechniques :.
M, P. G.
. Using the AMOVA framework to estimate a
st and ard iz ed ge net ic d iffe rent iat ion mea sur e. Ev olut ion : –
.
M, H.
. Breeding numbers and reproductive rate of eiders
at the Sands of Forvie National Nature Reserve, Scotland. Ibis
:–.
M M S
. . Alaska Annual Stud-
ies Plan Final FY . U.S. Department of the Interior, Miner-
al s Ma nageme nt Se rvice , Al aska Oute r Conti nenta l She lf Re gion,
Anchorage, Alaska.
N, L. E., S. R. J, G. M. O’D, M. K.
B
. . Common Eider (Somateria mollissima) nest
02_Sonsthagen_08-224.indd 753 9/2/09 2:25:38 PM
754
—
So n S t h a g e n e t a L .
—
au K , vo L .
126
cover and depredation on central Alaskan Beaufort Sea barrier
islands. Arctic :–.
P, K. B., R. T
. . Ten polymorphic auto-
somal microsatellite loci for the eider duck Somateria mollissima
and their cross-species applicability among waterfowl species
(Anatidae). Molecular Ecology Notes :–.
P, J. M., S. L. T, M. R. P, J. R. R
.
. Limited genetic differentiation among breeding, molting, and
wintering groups of threatened Steller’s Eider: e role of historic
and contempo rar y fac tors. Co nser vat ion Ge netics :–.
P, J. M., S. L. T, B. J. P, M. R. P, K. T.
S, D. L. D, A. M
. . Lack of spa-
tial genetic structure among nesting and wintering King Eiders.
Condor :–.
P, J. L., Y. N. Z, I. F, E. M. H,
K. E. O
. . Multilocus phylogeography of a Hol-
arctic duck: Colonization of North America from Eurasia by Gad-
wa ll (Anas strepera). Evolution :–.
P, M. R., P. L. F
. . Population structure of
Pac ific Com mon E ider s breed ing i n Al aska. C ondor :–.
P, D., K. A. C
. . MODE LTES T: Tes tin g the
model of DNA substitution. Bioinformatics :–.
R, A .
. Mi grat ion, homing , and mor tal ity of breeding fema le
eiders Somateria mollissima dresseri of the St. Lawrence estuary,
Quebec. Ornis Scandinavica :–.
R, F. C., M. G. A
. . Female-biased philo-
patry, monogamy, and the timing of pair formation in migratory
waterfowl. Pages – in Current Ornithology, vol. (
D. M.
Power, Ed.). Plenum Press, New York.
R, J., J. C. S-DB, X. M, R.
R
. . DnaSP, DNA polymorphism analyses by the coales-
cent and other methods. Bioinformatics :–.
S, J. W.
. A bow-net trap for ducks. Journal of Wildlife
Management :–.
S, D.
. Breeding of the Common Eider (Somateria mol-
lissima) on the Beaufort Sea Coast of Alaska. Condor :–.
S, S., D. R, L. E
. . ARLE-
QUIN, version .: A software for population genetic data anal-
ysis. Genetics and Biometry Laboratory, University of Geneva,
Geneva, Switzerland.
S, K. T., M. R. P, R. L. F, S. L. T, J. M.
P , R. K . C
. . Sex-biased gene flow in Spec-
tacled Eiders (Anatidae): Inferences from molecular markers with
contrasting modes of inheritance. Evolution :–.
S, K. T., S. L. T, J. M. P, B. J. P, K. S.
B, D. V. D
. . Phylogeography of Can-
ada Geese (Branta canadensis) in western North America. Auk
:–.
S, R. R., F. J. R
. . Biometry, nd ed. W. H. Free-
man, New York.
S, S. A., S. L. T, K. G. MC
. .
Genetic characterization of Common Eiders breeding in the
Yukon-Kuskokwim Delta, Alaska. Condor :–.
S, S. A., S. L. T, C. M. W
. . Gene
flow and genetic characterization of Northern Goshawks breed-
ing in Utah. Condor :–.
S, M. D., R. C. F
. . Multiple inde-
pendent transpositions of mitochondrial DNA control region
sequences to the nucleus. Proceedings of the National Academy
of Sciences USA :–.
S, M., N. S, P. D
. . A new statisti-
cal method for haplotype reconstruction from population data.
American Journal of Human Genetics :–.
S, R. S., D. L. D, J. B. F, L. T. Q-
. . Population declines of King and Common eiders of
the Be aufor t Sea . Cond or : –.
S, C.
. Dispersal and migratory movements of eiders
Somateria mollissima breeding in e Netherlands. Ornis Scan-
dinavica :–.
T , F.
. e effect of change in population size on DNA
polymorphism. Genetics :–.
T, K., M. N
. . Estimation of the number of nucle-
otide substitutions in the control region of mitochondrial DNA
in humans and chimpanzees. Molecular Biology and Evolution
:–.
T, R., K. G. K
. . Novel primers
for the mitochondrial control region and its homologous nuclear
pseudogene in the eider duck Somateria mollissima. Animal
Genetics :.
T, R., K. G. K, H. N
. . On
sex-specific dispersal and mating tactics in the Common Eider
Somateria mollissima as inferred from the genetic structure of
breeding colonies. Behaviour :–.
V , N .
. GIMLET: A computer program for analysing
genetic individual identification data. Molecular Ecology Notes
:–.
W, J. S., H. L. M
. . Mig rati ona l hom ing a nd
survival of adult female eiders in Maine. Journal of Wildlife Man-
agement :–.
W, K., G. R. G, M. J. B
. . Genetic differ-
entiation among populations of a migratory songbird: Limnothly-
pis swainsonii. Journal of Avian Biology :–.
W , S .
. e genetical structure of populations. Annals of
Eugenics :–.
As soci ate E ditor: D . B. McDo nald
02_Sonsthagen_08-224.indd 754 9/2/09 2:25:38 PM