Oceanic island ecosystems have been recognised as natural laboratories for studying evolution owing to their discrete geographical nature and diversity of species and habitats (Brown and Lomolino Reference Brown and Lomolino2000, Lomolino Reference Lomolino2000, Emerson Reference Emerson2002). Species richness within islands can be the result of a number of factors, including the diversification of a founding population into an array of species differentially adapted to diverse environmental niches, multiple successful colonisations of islands from neighbouring islands or continents, the diversification of a founding population by vicariance, and speciation through bottleneck and founder flush events (Templeton Reference Templeton1980, Carson and Templeton Reference Carson and Templeton1984, Clegg et al. Reference Clegg, Degnan, Kikkawa, Moritz, Estoup and Owens2002, Maley and Winker Reference Maley and Winker2010). Phylogenetic data can shed light on the processes that have led to species diversity on islands and the understanding of their origins (Emerson Reference Emerson2002).
The Falkland Islands (Malvinas) are located at 52°S in the South Atlantic, approximately 450 km from Tierra del Fuego and 600 km east of mainland Argentina (Figure 1). They lie close to the edge of the Patagonian continental shelf, and are influenced by the Falkland Current that derives from the westward wind drift around the Southern Ocean (McDowall Reference McDowall2005). Geological evidence suggests that the Falkland Islands (Malvinas) were historically adjacent to south-eastern Africa and reached their present position and orientation by the late Mesozoic approximately 130 Mya (Storey et al. Reference Storey, Curtis, Ferris, Hunter and Livermore1999, Trewin et al. Reference Trewin, MacDonald and Thomas2002). There are strong similarities between the biota of the Falkland Islands (Malvinas) and southern South America, including numerous island endemic species or subspecies. The avifauna comprises c.36 land and freshwater species (Woods Reference Woods1988, Strange Reference Strange1992). There is only one recognised Falkland endemic bird species, the Falkland Steamer Duck Tachyeres brachypterus (Weller Reference Weller1972). In several instances, the bird fauna is characterised by two or more species per genus, but in all cases the insular species of such genera are also present in Patagonia. It is believed that the Patagonian biota reached the Falkland Islands (Malvinas) either across a former land connection, in periods of lowered sea levels during Pleistocene glaciations, or across an ocean gap thorough dispersal from Patagonia (McDowall Reference McDowall2005).
The sheldgeese Neochen and Chloephaga comprise six species of waterfowl endemic to South America. With the exception of the Orinoco Goose Neochen jubata, they are distributed in the southern part of the continent along the Andes and Patagonia. Four of the five species also occur on the Falkland Islands (Malvinas) (Figure 1).
Neochen is monotypic, inhabiting rivers and wetlands in the tropical rainforests and savannas of northern South America (Johnsgard Reference Johnsgard1978). The Andean Goose Chloephaga melanoptera is resident to the Andes above 3,000 m elevation from Peru to the latitude of Mendoza in Argentina and Nuble in Chile; no subspecies are recognised (Johnsgard Reference Johnsgard1978). Breeding distribution of the Ashy-headed Goose Chloephaga poliocephala is strongly associated with the Nothofagus forest of Andean Patagonia from 37°S to Tierra del Fuego (Kear Reference Kear2005). The species, with no recognised subspecies, is more common in Chile than anywhere else in its range and is rare in the Falkland Islands (Malvinas) (Johnsgard Reference Johnsgard1978). The Upland Goose Chloephaga picta has two recognised subspecies; C. p. picta inhabits the mainland, and C. p. leucoptera is a non-migratory resident of the Falkland Islands (Malvinas) (Johnsgard Reference Johnsgard1978, Carboneras Reference Carboneras, del Hoyo, Elliott and Sargatal1992). The Kelp Goose Chloephaga hybrida also has a resident population in the Falkland Islands (Malvinas), C. h. malvinarum, and a partially migratory population, C. h. hybrida, on the coast of Patagonia in Chile and Argentina, including Tierra del Fuego (Carboneras Reference Carboneras, del Hoyo, Elliott and Sargatal1992). Finally, the Ruddy-headed Goose Chloephaga rubidiceps also occurs in two distinct populations, a migratory population that breeds in Tierra del Fuego and nearby in mainland Chile and Argentina, and a non-migratory population in the Falkland Islands (Malvinas) (Carboneras Reference Carboneras, del Hoyo, Elliott and Sargatal1992; Figure 1). No subspecies of C. rubidiceps are currently recognised.
Sheldgeese populations in the Falkland Islands (Malvinas) seem to be stable whereas continental populations have declined (Wetlands International 2006). In mainland southern South America, three of the species breed in southern Patagonia and Tierra del Fuego and migrate northwards during the winter to central Argentina and Chile: C. picta, C. poliocephala and C. rubidiceps. In Argentina, they were declared an agricultural pest in 1964 (Rumboll Reference Rumboll1975), when population control programmes were implemented. Control methods included mass destruction of nests and eggs in the breeding grounds and uncontrolled hunting in the wintering grounds (Pergolani de Costa Reference Pergolani de Costa1955, Blanco and de la Balze Reference Blanco, de la Balze, Boere, Galbraith and Stroud2006). In 1947, 250,000 eggs were destroyed (Delacour Reference Delacour1954), and 150,000 eggs per year in the 1970s (Weller Reference Weller1975) in Tierra del Fuego alone. These eradication techniques decreased the sheldgeese populations in Argentina, especially C. rubidiceps (Rumboll Reference Rumboll1975). In northern Tierra del Fuego, Crawshay (Reference Crawshay1907) recorded sheldgeese as ‘countless’ during the breeding season, and C. rubidiceps as the most abundant species (Scott Reference Scott1954). However, current population estimates of C. rubidiceps are 768–1,178 individuals in the breeding grounds of Chile and Argentina (Madsen et al. Reference Madsen, Matus, Blank, Benegas, Mateazzi and Blanco2003, Blanco and de la Balze Reference Blanco, de la Balze, Boere, Galbraith and Stroud2006). Population declines > 50% were also recorded in censuses of wintering areas in southern Buenos Aires province (Blanco et al. Reference Blanco, Matus, de la Balze, Blank, MacLean and Zalba2009). For the three species of sheldgeese wintering in Buenos Aires more than 35,000 individuals were counted in 1976, 16,547 in 1984, and 9,278–13,618 individuals during 2007–2009 (Petracci Reference Petracci, Ibáñez, Scorolli, Cozzani and Blanco2008, Reference Petracci, Ibáñez, Scorolli, Faillá and Blanco2009, Reference Petracci, Ibáñez, Baigún, Hollmann and MacLean2010), suggesting that sheldgeese have declined by approximately two-thirds.
Therefore, it is necessary to establish whether the continental populations of sheldgeese represent a significant unit for conservation, either evolutionarily or for management. Evolutionarily significant units (ESU) have been defined as reciprocally monophyletic for mitochondrial DNA (mtDNA) alleles and showing significant divergence of alleles at nuclear loci, whereas management unit (MUs) have been recognised as populations with significant divergence of allele frequencies at nuclear or mitochondrial loci, regardless of the phylogenetic distinctiveness of the alleles (Moritz Reference Moritz1994). The focus of the MU is on contemporary population structuring and short-term monitoring rather than historical factors. However, consensus about the applicability of the concepts of ESU and MU is lacking (Crandall et al. Reference Crandall, Bininda-Emonds, Mace and Wayne2000, Green Reference Green2005). Fraser and Bernatchez (Reference Fraser and Bernatchez2001) proposed a unifying concept for defining conservation units, and suggested that the best available biological information be used when exercising such decisions on a case-by-case basis. Despite the controversy, we aimed at defining areas of importance for sheldgeese conservation.
Motivated by the critical decrease of continental populations of sheldgeese, and the taxonomic uncertainty of continental versus island populations, we used mtDNA to compare and quantify divergence between populations of two species of particular interest C. rubidiceps and C. p. picta/leucoptera inhabiting the Falkland Islands (Malvinas) and continental South America (southern Chile and Argentina). By incorporating multiple samples for each Chloephaga species and Neochen jubata, we also generated a phylogeny for all species of sheldgeese and provided a likely biogeographic scenario for the split of the insular and mainland sheldgeese populations.
Sampling, PCR, and DNA sequencing
We collected blood or tissue samples from all six species of sheldgeese, including three insular populations from the Falkland Islands (Malvinas). These included Neochen jubata (n = 1), C. melanoptera (n = 10) from Peru, Bolivia, and Argentina; C. poliocephala from Argentina (n = 4); C. rubidiceps from Chile (n = 9) and the Falkland Islands (Malvinas) (n = 6); C. p. picta from Argentina (n = 15) and C. p. leucoptera (n = 28) from the Falkland Islands (Malvinas), and C. hybrida malvinarum (n = 3) from the Falkland Islands (Malvinas) (Figure 1). Five representative dabbling ducks Anas crecca crecca, A. c. carolinensis, A. acuta, A. americana and A. clypeata were selected as outgroups. Anas dabbling ducks are appropriate outgroups as shown by varied datasets and analyses (Donne-Goussé et al. Reference Donne-Goussé, Laudet and Hänni2002, Gonzalez et al. Reference Gonzalez, Düttmann and Wink2009, Bulgarella et al. Reference Bulgarella, Sorenson, Peters, Wilson and McCracken2010).
DNA sequences were obtained using modified protocols for DNA extraction, PCR, and sequencing (McCracken and Sorenson Reference McCracken and Sorenson2005). We amplified and sequenced 636 bp corresponding to the 5’-end of the mtDNA control region from each of the 81 sampled specimens. We initially used the overlapping primer pairs L78–H774 (Sorenson and Fleischer Reference Sorenson and Fleischer1996, Sorenson et al. Reference Sorenson, Ast, Dimcheff, Yuri and Mindell1999), but after initial inspection of the sequences and preliminary data analyses, we determined that all DNA extracts obtained from blood yielded co-amplified or purely nuclear mtDNA (numts; e.g. Sorenson and Quinn Reference Sorenson and Quinn1998), putative copies of the mtDNA sequence (GenBank accession numbers KC109081–KC109104). We therefore redesigned a 5’ forward primer L108-Chloephaga (CATATATTTATATGCCCCATACAATTTAGA) to selectively amplify the mtDNA sequences we obtained from muscle tissue extracts. This 30-bp primer is mostly identical to the mtDNA template at all sites but mismatches the nuclear sequence at the last position on the 3’-end (the nuclear copy possesses a G at the last site whereas the mtDNA possesses an A at this site). Additionally, Neochen and C. melanoptera each posses a deletion of the A nucleotide at position 24 (relative to the primer), but since we amplified these two species from muscle we were able to obtain clean mtDNA sequences using the L78 primer. The DNA extracted from blood samples from C. p. leucoptera inhabiting the Falkland Islands (Malvinas) were determined to have CG at the first and second position of the L108 primer, respectively, whereas the mtDNA template possessed CA in all six species. Thus, we were able to preferentially amplify and sequence the mtDNA copy, as well as the nuclear copy from blood samples using allele-specific priming (Bottema et al. Reference Bottema, Sarkar, Cassady, Li, Dutton and Sommer1993, Peters et al. Reference Peters, McCracken, Zhuravlev, Lu, Wilson, Johnson and Omland2005). Sequences from opposite strands were reconciled and verified for accuracy using Sequencher v.4.7 (Gene Codes, Ann Arbor, Michigan). Sequences and specimen voucher information, including geo-referenced localities, are archived in GenBank (accession numbers KC109005–KC109080, HM063476–HM063480).
The control region sequences for Neochen and Chloephaga were aligned by eye. However, when we included the five Anas dabbling ducks species as outgroup taxa, the sequences varied in length due to insertions and deletions of nucleotides. We therefore aligned control region sequences for Chloephaga, Neochen, and Anas using the Geneious v.5.4 (Drummond et al. Reference Drummond, Ashton, Buxton, Cheung and Cooper2011) alignment algorithm with the default parameter settings.
We first constructed a simple UPGMA tree by using the HKY genetic distance method as implemented in Geneious v.5.4, with a resampling of 1,000 bootstrap replicates. Secondly, we used AIC (Akaike Reference Akaike, Petrov and Csaki1973) as implemented in Modeltest 3.7 (Posada and Crandall Reference Posada and Crandall1998) to determine the model of sequence evolution that best fit the data. We then conducted a maximum likelihood analysis as implemented in the PhyML (Guindon and Gascuel Reference Guindon and Gascuel2003) plug-in for Geneious v5.4. Statistical support for clades was evaluated by 1,000 replicate nonparametric bootstrapping (Felsenstein Reference Felsenstein1985). Clade probabilities were obtained from the posterior distribution using MrBayes v.3.2.0 (Ronquist and Huelsenbeck Reference Ronquist and Huelsenbeck2005). Bayesian analyses were replicated twice, each with four Markov chains of 2.5 million generations. Trees were sampled every 1,000 generations, of which the first 0.5 million generations were discarded as burn-in.
Furthermore, to examine population differentiation of C. rubidiceps and C. p. picta/leucoptera between the Falkland Islands (Malvinas) and continental South America, we calculated ΘST for each population of sheldgeese using the Tamura and Nei (Reference Tamura and Nei1993) nucleotide substitution model in Arlequin v3.5 (Excoffier and Lischer Reference Excoffier and Lischer2010), with a significance level of 0.05. To determine the existence (or not) of shared haplotypes, we drew an allelic network calculated using the median-joining algorithm in NETWORK v.4.1 (Bandelt et al. Reference Bandelt, Forster and Röhl1999).
Population-level samples of sheldgeese mitochondrial DNA have not been previously published, due in part to the fact that DNA extracts obtained from blood frequently yield co-amplified or purely nuclear mtDNA (numts) copies of the mtDNA sequence in these species. Figure S1 in the online supplementary material shows the most parsimonious tree constructed with sequences comprising a mix of putative numts (extracted from blood, shown in bold) and mtDNA copies. The numts were an average of 3.21% (2.68–5.52%) divergent from the nearest mtDNA copy.
The UPGMA tree grouped C. melanoptera and N. jubata as sister taxa, and sister to all other Chloephaga species. Chloephaga rubidiceps from the Falkland Islands (Malvinas) formed a reciprocally monophyletic clade with C. rubidiceps from Chile, with an average uncorrected sequence divergence of 1.03% (0.94–1.26%). Chloephaga rubidiceps was sister to C. poliocephala. This clade was in turn sister to C. hybrida plus C. picta (Figure 2).
No haplotypes were shared between C. rubidiceps populations from the Falkland Islands (Malvinas) and Chile or between C. p. picta from Argentina and C. p. leucoptera from the Falkland Islands (Malvinas) (Figure 3, Figure S2). Both clades of C. picta were reciprocally monophyletic with an uncorrected sequence divergence of 0.59% (0.31–0.79%), defined by three transitions at positions 85, 109, and 317, respectively. Pairwise ΘST for the two C. picta populations was ΘST = 0.478 (P < 0.0001). The two C. rubidiceps populations had a ΘST = 0.596 (P < 0.0001), and the clades were differentiated by six character changes: five transitions at positions 91, 108, 114, 136, and 166, and one transversion at position 607.
The maximum likelihood tree agrees with the UPGMA tree (Figure 3). Chloephaga hybrida was placed as sister taxon to C. picta with low bootstrap (57.4%) but high posterior probability (1.00) support. Two reciprocally monophyletic clades of C. melanoptera were recovered with an uncorrected sequence divergence of 2.33% (1.88–2.67%), one including three specimens from north-western Argentina and one from Bolivia (Oruro), and the other one with three specimens from Peru (Ancash, Huancavelica, and Moquequa) and three from Bolivia (La Paz). To our knowledge, this is the first time that a possible differentiation between northern and southern populations of C. melanoptera is acknowledged, warranting further studies.
Phylogenetic relationships of the sheldgeese
Previous phylogenetic studies of the sheldgeese were based on behaviour and morphology (Delacour and Mayr Reference Delacour and Mayr1945, Johnsgard Reference Johnsgard1961, Livezey Reference Livezey1997). The shelducks of the genus Tadorna and the sheldgeese that included the genera Cyanochen, Alopochen, Neochen and Chloephaga historically were grouped in the monophyletic Tribe Tadornini. Based on morphology, Livezey (Reference Livezey1997) confirmed the monophyly of this group. Later, a molecular phylogeny based on mitochondrial DNA included the Tadornini within the Anatinae (Donne-Goussé et al. Reference Donne-Goussé, Laudet and Hänni2002), and the result has been since confirmed with different datasets and analyses (Gonzalez et al. Reference Gonzalez, Düttmann and Wink2009, Bulgarella et al. Reference Bulgarella, Sorenson, Peters, Wilson and McCracken2010); the Tadornini are closer to the true ducks not the geese (Anserinae). The sheldgeese comprise the Neotropical genera Neochen and Chloephaga; the shelducks include the genera Tadorna and Alopochen, whereas Cyanochen is more closely related to Hartlaub’s Duck Pteronetta hartlaubi (Gonzalez et al. Reference Gonzalez, Düttmann and Wink2009, Bulgarella et al. Reference Bulgarella, Sorenson, Peters, Wilson and McCracken2010). Recently, McCracken et al. (Reference McCracken, Barger and Sorenson2010) presented the first molecular phylogeny for the sheldgeese that included all six species. However, this study focused on haemoglobin genes and did not include mtDNA or population-level sampling for each of the Chloephaga species. Here, we present a fully resolved mtDNA phylogeny for Chloephaga and Neochen. Chloephaga picta and C. hybrida were found to be sister taxa, as were C. rubidiceps and C. poliocephala, and these two groups are sister to the clade formed by C. melanoptera and N. jubata. The mainland and insular subspecies of C. picta are reciprocally monophyletic as are the two populations of C. rubidiceps. Our results are mostly in agreement with those of McCracken et al. (Reference McCracken, Barger and Sorenson2010). However, they did not find C. picta and C. hybrida to be sister taxa, as these two species did not form a clade in their analyses (see Figure 2 in McCracken et al. Reference McCracken, Barger and Sorenson2010).
In this study, mainland and insular populations of C. rubidiceps did not share haplotypes, being as divergent from each other as they are from C. poliocephala. This suggests either the lack of gene flow between the islands and the continent or female-biased philopatry or both. Waterfowl are unusual among birds in that females show greater philopatry to natal and breeding areas than do males (Rohwer and Anderson Reference Rohwer, Anderson and Johnston1988), and sex-biased dispersal has frequently been invoked as a possible explanation for low levels of mtDNA haplotype sharing among geographic regions (Sonsthagen et al. Reference Sonsthagen, Talbot, Lanctot, Scribner and McCracken2009, Bulgarella et al. Reference Bulgarella, Peters, Kopuchian, Valqui, Wilson and McCracken2012, Peters et al. Reference Peters, Bolender and Pearce2012). Furthermore, animal mtDNA evolves rapidly, usually reaching reciprocal monophyly within species and is considered to be not recombining. In contrast, nuclear genes have a longer coalescence time (approximately four times that of mtDNA). Consequently, lineage sorting will be complete for mtDNA for some vicariant events for which it will not yet have occurred for nuclear markers. In such cases, mtDNA is more informative than nuclear DNA and therefore considered a ‘leading’ indicator (Zink and Barrowclough Reference Zink and Barrowclough2008). Multilocus genetic data may be particularly informative about dispersal patterns in sheldgeese because little data on demographics and movements are currently available. Nevertheless, a well-resolved mitochondrial DNA gene tree is a starting point for any phylogeographical investigation and well-supported geographical patterns of mtDNA differentiation can be sufficient to provide a strong inference of evolutionary history (Zink and Barrowclough Reference Zink and Barrowclough2008).
Avian diversification on the Falkland Islands (Malvinas)
McDowall (Reference McDowall2005) concluded that the biota of the Falkland Islands (Malvinas) are almost exclusively Patagonian and must have reached the Islands either across a former land connection or by dispersal across an ocean gap during the Pleistocene. Ponce et al. (Reference Ponce, Rabassa, Coronato and Borromei2011) demonstrated that during the last glacial maximum (LGM, 24,000 bp) the sea level was approximately 120–140 m below present sea level. At the time, the East and West Falkland Islands (Malvinas) formed a single island four times larger in area than the present day archipelago. This ancient island was separated from the rest of the continent by a strait less than 220 km wide (Ponce et al. Reference Ponce, Rabassa, Coronato and Borromei2011) giving opportunity for dispersal and colonisation.
Over 200 bird species are recorded in the Falkland Islands (Malvinas) (Woods and Woods Reference Woods and Woods2006), the majority of which are vagrants that do not breed on the islands. Weller (Reference Weller1972) counted 11 regular species of breeding waterfowl. Nine passerine taxa regularly breed there (Woods and Woods Reference Woods and Woods2006, Campagna et al. Reference Campagna, St Clair, Lougheed, Woods, Imberti and Tubaro2012). Six of the 11 recorded penguin species breed on the islands, making it one of the richest penguin faunas in the world (McDowall Reference McDowall2005). Seven species of raptors, four dotterels and snipes, and semi-aquatic birds such as herons complete the breeding resident bird list. Recently, there has been a resurgence of interest in the phylogeography of the avifauna of the Falkland Islands (Malvinas) (McCracken and Wilson Reference McCracken and Wilson2011, Fulton et al. Reference Fulton, Letts and Shapiro2012, Campagna et al. Reference Campagna, St Clair, Lougheed, Woods, Imberti and Tubaro2012).
Using multilocus genetic data, McCracken and Wilson (Reference McCracken and Wilson2011) reported differentiation between insular and mainland populations of Speckled Teal Anas flavirostris, a widespread migratory duck species in southern South America. They found the island population to be significantly differentiated from populations in Argentina in their mtDNA, with four private alleles in the Falkland Islands (Malvinas) population, and at three nuclear loci, with restricted gene flow from the continent; the insular population of Speckled Teal likely comprising a distinct demographic unit. Next, Fulton et al. (Reference Fulton, Letts and Shapiro2012) studied speciation in steamer ducks Tachyeres spp. finding that the two species of steamer ducks of the Falkland Islands (Malvinas) are indistinguishable from each other based on a single-locus mtDNA, and that this clade is genetically distinct from the three species in continental South America. They placed the divergence between continental and insular steamer ducks between 2.2 and 0.6 Mya consistent with the great Patagonian glaciation (GPG; Rabassa et al. Reference Rabassa, Coronato and Martínez2011) when extensive ice sheets covered Patagonia. Finally, Campagna et al. (Reference Campagna, St Clair, Lougheed, Woods, Imberti and Tubaro2012) studied the divergence between nine mainland and insular passerine populations. They found marked genetic differentiation between the conspecific Troglodytes aedon and T. cobbi, with T. cobbi being the oldest Falkland Islands (Malvinas) land bird that split from continental T. aedon during the GPG of the Pleistocene.
Although we have not dated the divergence between the Falkland Islands (Malvinas) and mainland populations of Chloephaga, several scenarios are possible. First, Chloephaga might have colonised the islands from the continent during the GPG when Patagonia was heavily glaciated and the birds could have found refuge in the unglaciated parts of the islands. Alternatively, the colonisation of the islands might date back to the earlier Pleistocene (or even the Pliocene) with birds dispersing to ice-free areas either in Northern Argentina/Chile or within the Falkland Islands (Malvinas) themselves during the glaciation/deglaciation cycles of the Pleistocene.
Indeed, multilocus phylogeographic analyses are necessary to infer demographic history parameters such as effective population sizes, direction and timing of colonisation, and gene flow but this study is an important first step towards the understanding of the sheldgeese.
The Ruddy-headed Goose, C. rubidiceps, has been removed from the threatened species list due to the stability of the insular Falkland Islands (Malvinas) population (BirdLife International 2012). However, the current mainland population of C. rubidiceps is estimated to comprise < 1,000 individuals and is decreasing (Blanco et al. Reference Blanco, Matus, de la Balze, Blank, MacLean and Zalba2009); therefore, it meets IUCN criterion C (fewer than 2,500 mature individuals in the population; IUCN 2001) and should be classified as ‘Endangered’. Unfortunately, we did not include individuals from the wintering population of C. rubidiceps in southern Buenos Aires, Argentina, because the extremely low numbers of individuals made it difficult to obtain samples. It would be useful to collect genetic samples from this population to quantify how much it differs from the Falkland Islands (Malvinas) population. Based on our findings, we suggest that the continental population of Ruddy-headed Goose be considered an evolutionarily significant unit (ESU; sensu Moritz Reference Moritz1994), distinct from the Falkland Islands (Malvinas) population.
Upland Geese populations from the mainland show significant genetic, morphological and ecological divergence from the subspecies in the Falkland Islands (Malvinas). While the C. p. leucoptera population is stable, the continental C. p. picta population appears to have decreased in recent years and its status should be considered carefully until more data or estimations of the current population sizes are undertaken. Moreover, for both species, we showed the reciprocal monophyly of the insular and continental populations and the implicit lack of female-mediated gene flow, supporting inferences that these populations have diverged and are geographically isolated units deserving different conservation status than they currently have.
The supplementary materials for this article can be found at journals.cambridge.org/bci
We thank the following people and agencies for their assistance: Dirección de Fauna Silvestre, Secretaría de Ambiente y Desarrollo Sustentable de la República Argentina, Dr. Pablo Tubaro from the Museo Argentino de Ciencias Naturales “Bernardino Rivadavia”, Dirección de Fauna y Flora Silvestre del Chubut, Dirección de Fauna de Santa Cruz, Falkland Island Government, and Alan Eagle and Sonia Felton from Fitzroy Farm. Two anonymous reviewers provided helpful comments on the manuscript. ASDG is a CONICET Research Fellow and CK is a Postdoctoral Researcher of IDRC (iBOL, International Barcode of Life Project). Funding was provided by the Institute of Arctic Biology at the University of Alaska Fairbanks, a grant from Frank M. Chapman Fund at the American Museum of Natural History to MB, and NSF grants (DEB-0444748,IOS-0949439) to KGM.