Where east meets west: Phylogeography of the high Arctic North American brant goose

Abstract

Genetic variation in Arctic species is often influenced by vicariance during the Pleistocene, as ice sheets fragmented the landscape and displaced populations to low- and high-latitude refugia. The formation of secondary contact or suture zones during periods of ice sheet retraction has important consequences on genetic diversity by facilitating genetic connectivity between formerly isolated populations. Brant geese (Branta bernicla) are a maritime migratory waterfowl (Anseriformes) species that almost exclusively uses coastal habitats. Within North America, brant geese are characterized by two phenotypically distinct subspecies that utilize disjunct breeding and wintering areas in the northern Pacific and Atlantic. In the Western High Arctic of Canada, brant geese consist of individuals with an intermediate phenotype that are rarely observed nesting outside this region. We examined the genetic structure of brant geese populations from each subspecies and areas consisting of intermediate phenotypes using mitochondrial DNA (mtDNA) control region sequence data and microsatellite loci. We found a strong east-west partition in both marker types consistent with refugial populations. Within subspecies, structure was also observed at mtDNA while microsatellite data suggested the presence of only two distinct genetic clusters. The Western High Arctic (WHA) appears to be a secondary contact zone for both Atlantic and Pacific lineages as mtDNA and nuclear genotypes were assigned to both subspecies, and admixed individuals were observed in this region. The mtDNA sequence data outside WHA suggests no or very restricted intermixing between Atlantic and Pacific wintering populations which is consistent with published banding and telemetry data. Our study indicates that, although brant geese in the WHA are not a genetically distinct lineage, this region may act as a reservoir of genetic diversity and may be an area of high conservation value given the potential of low reproductive output in this species.

Keywords: Branta bernicla; brant geese; contact zones; genetic structure; refugia.

FIGURE 1

Photographs of the three different plumage coloration types of brant within North America: Pacific Black (nigricans, left), Western High Arctic (nigricans, middle), and Atlantic (hrota, right). Photograph credit: Maynard Axelson.

FIGURE 2

Population subdivision of brant geese (Brant bernicla) based on STRUCTURE (top panel) with breeding distribution (color shading) and subspecies/ecological stocks designations indicated by different color shading (bottom panel). STRUCTURE plot shows average assignment probability based on microsatellite data for K = 2. See Table S3 for proportion of samples that have nigricans‐like genotypes (>75% assignment to dark gray cluster), hrota‐like (>75% to light gray cluster), and to uncertain (Admixed group, less than 75% assignment to either group). Color blocks in the top panel correspond to color circles (molting or nesting sample locations) or squares (staging or wintering sample locations) in the bottom panel. Location of brant breeding populations representing the three subspecies (nigricans, hrota, and bernicla) and four ecological stocks (Pacific Black [blue; YKD, NS, LB, LRD], Western High Arctic [green; PP, MEL], Eastern High Arctic [orange: BI, ICE, IRE], and Atlantic [yellow; BAF, SH]). Arrows indicate general migration pathways to staging and wintering areas. Collection location abbreviations are defined in Section 2 and Table 1.

FIGURE 3

Unrooted haplotype network illustrating relationships of 130 mtDNA control region haplotypes assayed (star contraction option reduced the 135 haplotypes to 100; see Table S1) from brant (Branta bernicla). The size of the circle corresponds to the frequency of each haplotype and small white circles indicate intermediate ancestral haplotypes that were not sampled. As branch lengths are not to scale, the number of marks across branches indicate the number of variable sites between haplotypes. No marks indicate only a single variable site between haplotypes. The dashed line separates haplotypes of B. b. nigricans and B. b. hrota.

FIGURE 4

PCA plot of first two principal components based on 12 microsatellite loci. For illustrative purposes, Western High Arctic samples were not shown in the (a) top panel but are shown in (b) bottom panel. Both panels represent the same analysis. Although there is some overlap, the x‐axis divides the majority of nigricans‐like (Pacific/Black) and hrota‐like (Atlantic/EHA) genotypes. The table in the bottom panel reports the proportion of samples from the Western High Arctic on either side of the y‐axis (positive or negative PC1 coordinate value).

FIGURE 5

Results of full model migration matrix when parameters were allowed to vary independently to estimate gene flow among brant populations in North America. Migration direction and rates are based on (a) mtDNA control region data and (b) 12 microsatellite loci. Arrow thickness is proportionate to estimated levels of gene flow (thicker arrows indicate higher relative gene flow). Solid lines indicate 95% confidence intervals (CI) did not overlap while dashed lines indicate 95% CI overlap but did not encompass either mean value. No lines indicate either symmetrical gene flow or no gene flow between populations. See Table S4 for values.