The Beringian Coevolution Project: holistic collections of mammals and associated parasites reveal novel perspectives on evolutionary and environmental change in the North

As a land bridge that repeatedly linked Eurasia to North America, Beringia mediated the flow of diversity between Arctic continents. During the Quaternary, the flow was predominantly west to east due to climatological gradients that developed in the late Pliocene and the relative size and timing of major dispersal barriers across the region (Hopkins et al. 1982; Waltari et al. 2007; Hoberg et al. 2012a). The general eastward trend (Rausch 1994), however, contrasts with some histories revealed by BCP analyses that reflect an important North American influence on Asian faunas (e.g., Demboski and Cook 2003; Hope et al. 2012). For example, data from cold-adapted species such as pikas (Galbreath and Hoberg 2012), arctic ground squirrels (Galbreath et al. 2011; McLean et al. 2016b), American marten now resident in interior Alaska (Small et al. 2003), jumping mice (Malaney and Cook 2013), and shrews (Hope et al. 2012) show evidence for distinct episodes of northward expansion from temperate North American latitudes into eastern Beringia or even farther into Asia. New evidence for multiple distinct transberingian colonization events by helminths associated with pikas as well as the presence of parasite lineages in Asia that may have North American ancestry (Toman and Galbreath, unpublished data), indicate a more important role for North American faunas (hosts and parasites) as a source for westward colonization than previously recognized. Further, phylodemographic studies are revealing complex episodes of geographic contraction and range expansion out of Asia (Fig. 5). For example, separate lineages of both northern red-backed voles (Myodes rutilus) and brown lemmings (Lemmus trimucronatus/sibiricus) appear to have undergone range retraction and subsequent recolonization across western Beringia (Fedorov et al. 1999; Kohli et al. 2015). Each of these snapshots of population history yields insight into the role and timing of processes shaping communities, such as the consequences of abiotic forces (e.g., climate change) for species range dynamics and population divergence (Hope et al. 2014). Still, issues such as the causes of the Late Pleistocene pause in eastward colonization of North America (Tamm et al. 2007; Meiri et al. 2014) or the influence of a possible mesic “buckle” in Beringia (Guthrie 2001) will be more clearly defined in the near future.

In addition to mediating intercontinental faunal exchange, climatic oscillations at northern high latitudes created new opportunities for population- and species-level contact. New specimen-derived genetic data from these areas have contributed to our general understanding of diversification and the permeability of species boundaries in the face of gene flow. In three species of Urocitellus ground squirrels, two sequential instances of ancient gene flow were identified, each resulting in mitochondrial capture (McLean et al. 2016b). Unlike many other examples of high-latitude hybridization, these taxa (U. parryii, U. richardsonii, and U. elegans) are now largely allopatric, demonstrating that historic genetic exchange among high-latitude mammals cannot be predicted based on current distributions alone. This and other instances of hidden gene flow within Beringian faunas are best understood when placed in the context of continental-scale analyses, an approach that BCP specimen archives are poised to contribute to significantly.

Investigation of distinct lineages of ermine (Mustela erminea) utilizing BCP specimens led to novel insights about the origin of endemic, at-risk island populations (Colella et al., unpublished data) in the Alexander Archipelago. Additionally, next-generation DNA sequencing uncovered highly divergent brown bear (Ursus arctos) populations (Lindqvist et al. 2010; Miller et al. 2012) within the same archipelago. Admiralty Island, in particular, supports distinctive lineages of both ermine and brown bears. These discoveries highlight shared demographic histories across topographically complex regions of Beringia and the role of glaciation events in driving diversity.

As an expansive region that remained relatively ice-free during Pleistocene glacial periods, Beringia served as a refuge for populations of northern species through repeated climatic oscillations. It also provided a source for post-glacial recolonization of high latitudes. However, questions have persisted regarding the existence of additional smaller refugia in the North and their relative contributions to diversity and post-glacial colonization (Fedorov and Stenseth 2002; Fleming and Cook 2002; Shafer et al. 2010). For example, divergent genetic lineages endemic to Beringia were found (Fedorov et al. 1999; Fedorov and Stenseth 2002; Brunhoff et al. 2003; Fedorov et al. 2003) in the collared lemming (Dicrostonyx), brown lemming (Lemmus), and tundra vole (Microtus oeconomus), a finding that suggests that these species did not disperse from Beringia after the last ice age and implies limited importance of Beringia as a source of post-glacial colonization.

Phylogeographic analyses based on broad BCP sampling are revealing concordant zones of genetic differentiation (Fig. 5) that help interpret the impact of the boundaries of the Beringian refugium, such as the Upper Kolyma region of Siberia in the west (e.g., Galbreath and Cook 2004; Kohli et al. 2015; Haukisalmi et al. 2016) and near the Yukon–Alaska border in the east (Dawson et al. 2014). In addition, the influence of smaller peripheral refugia in the Canadian Arctic (Fedorov and Stenseth 2002; Fedorov et al. 2003; Cook et al. 2016b), southeast Alaska (Cook et al. 2001; Sawyer and Cook 2016), and in the periglacial zone south of the Laurentide and Cordilleran ice sheets (Fedorov et al. 2003) is being explored. Early attempts to unravel the evolutionary history of arctic ground squirrels (Urocitellus parryii), one of the quintessential high-latitude mammals, relied on a view of Beringia as a vast steppe with few natural barriers to gene flow (e.g., Nadler and Hoffmann 1977). However, our work to test those ideas in U. parryii elucidated range-wide phylogeographic structure and a dynamic history of diversification spanning the Bering Land Bridge (Eddingsaas et al. 2004; Galbreath et al. 2011; McLean et al. 2016b), suggesting that late Pleistocene environmental changes within central Beringia supported significant regional differentiation in this species. Such insights establish a template for testing hypotheses of fine-scale phylogeographic structure in other northern taxa.

Five groups of cyclophyllidean tapeworms represented by the genera Arostrilepis, Anoplocephaloides, and Paranoplocephala among cricetids, Versteria among small to medium mustelids and rodents, and Staphylocystoides in shrews further serve as critical exemplars demonstrating the importance of Beringia as a center of diversification. Based on patchy geographic and host sampling, Arostrilepis was historically thought to be represented by a single geographically widespread and hypervariable species with a considerable host range across the Holarctic (Schiller 1952; Ryzhikov et al. 1978). By contrast, recent morphological and molecular investigations of Arostrilepis specimens collected primarily through the BCP led to the formal description (or redescription) of 13 species spanning the Holarctic and centered on Beringia (e.g., Makarikov et al. 2013). Further, the discovery of at least 15 unnamed genetic lineages appears consistent with a yet broader cryptic assemblage of species among cricetid, geomyid, heteromyid, and dipodid rodents extending into Boreal and Temperate zones in the Nearctic (Haas et al., unpublished data). A complex history involving recurrent expansion events across Beringia (intercontinental) and also from multiple refugia within Eurasia or North America (intracontinental) has driven geographic and host colonization and was the primary determinant of diversification through considerable faunal mixing over time (e.g., Hoberg et al. 2012a).

Similarly, Paranoplocephala and Anoplocephaloides continue to represent complicated and taxonomically problematic assemblages of species with broad geographic ranges in the Holarctic (Haukisalmi and Henttonen 2003; Haukisalmi 2009; Haukisalmi et al. 2009b, 2014, 2016). With important contributions from BCP collections, tapeworms that are predominately associated with northern arvicolines and previously assigned to the Paranoplocephala omphalodes complex are now recognized as at least 14 distinct genera based on integrated morphological and molecular data (Haukisalmi et al. 2014). Phylogenies for these cestodes suggest that episodes of host colonization by parasite lineages contributed to diversification while they created novel host–parasite interactions due to past episodes of range expansion from multiple refugia (e.g., Wickström et al. 2003; Makarikov et al. 2013; Haukisalmi et al. 2016).

Parasite diversity, in combination with phylogenetic and population data, is critical to revealing otherwise cryptic histories for mammalian hosts. For example, Rauschoides arctica (formerly Paranoplocephala) found in Dicrostonyx lemmings contains subclades that are either geographically widespread or restricted to islands of the Arctic archipelago (Wickström et al. 2003), a pattern consistent with isolation of multiple populations of hosts and parasites north and south of the continental glaciers during glacial maxima. Post-glacial expansion from south to north tracked deglaciation (Lessa et al. 2003) and brought previously isolated parasite populations into contact. Further, the boundaries for Nearctic and Eurasian host populations were not always at the Bering Strait, as indicated by Dicrostonyx of Neartic origin on Wrangel Island (Fedorov and Stenseth 2002), while R. arctica in these collared lemmings is represented by two discrete clades of Palearctic origin. Among lemmings, this may be another example of complex lineage persistence through host colonization prior to the local extinction of the original host group (Hoberg and Brooks 2008), in this case Eurasian Dicrostonyx. Complex signatures of geographic colonization are also indicated in Anoplocephaloides in lemmings (Haukisalmi et al. 2016) and in the nematode Soboliphyme baturini in marten (Koehler et al. 2009a; Hoberg et al. 2012b).

In its role as a high-latitude refugium, Beringia has long been recognized as stimulating the evolution of endemic northern faunas (Sher 1984, 1999). However, as population histories and species distributions become more fully resolved across Beringia, a more nuanced picture is emerging that links climatic oscillations, opening and closing of dispersal barriers, and geographic range fluctuations as major drivers of diversification in the region (Cook et al. 2004; Galbreath et al. 2011; Hoberg et al. 2012a; Hoberg and Brooks 2013). This new concept of Beringia as both a facilitator of intercontinental dispersal and a venue for in situ diversification significantly extends our view of land bridges and their importance in biotic diversification, including in expansion into the Americas by ancient humans (Hoffecker et al. 2014). This complex history is evident especially in some northern rodents (e.g., Fedorov et al. 2003) as well as in the tremendous diversity of their helminths, the full extent of which is only now becoming apparent through investigation of extensive BCP collections (e.g., Wickström et al. 2003, 2005; Haukisalmi et al. 2006, 2007, 2014; Lee et al. 2016).

Beringia has had a strong influence not only on Eurasian and Nearctic faunas but also on faunal dynamics in the Neotropical region (e.g., proboscideans: Shoshani and Tassy 1996). Our understanding of these biogeographic linkages is improving due to increased availability and geographic and taxonomic coverage of mammal and parasite specimens (e.g., Conroy et al. 2001; Hoberg 2005; Pozio et al. 2009; Esteva et al. 2010; Hoberg et al. 2012a; Korhonen et al. 2016). For example, assemblages of Trichinella nematodes that circulate through carnivory or scavenging by various mammals (e.g., carnivores, suids, and rodents) provide insights into colonization processes from Eurasia through the Nearctic and extending into the tropics (e.g., Pozio et al. 2009; Korhonen et al. 2016). The clade of “encapsulated” Trichinella initially diversified in central Eurasia with subsequent history dominated by sequential events of geographic and host colonization that account for the distribution of nine recognized species (Zarlenga et al. 2006). A single species, T. patagoniensis, is endemic to southern South America among small felids that radiated near 8 MYBP (Johnson et al. 2006). Distribution and phylogeny are consistent with geographic expansion from Eurasia with small felids, across a proto-Beringia during the late Miocene, through the Nearctic, and across an emergent or nascent Panamanian Isthmus near 10 MYBP prior to the Great American Interchange that ensued about 3–5 MYBP (Bacon et al. 2015).

Subsequently, the temporal duration and spatial history for radiation of an assemblage of Trichinella at high latitudes appear constrained by events in the late Pliocene and through the Quaternary. Climatological forcing and ecological perturbation served as determinants of recurrent episodes for expansion, colonization, and the origins of extensive faunal mosaics (e.g., Hoberg et al. 2012a; Korhonen et al. 2016). Contact zones and a broad geographic boundary demonstrated between T. nativa and the currently unnamed T-6 genotype among wolverines in northwestern Canada may reflect the most recent post-glacial expansion events following the last glacial maximum about 11 KYBP (e.g., Dunams-Morel et al. 2012).

Global climate warming is having a disproportionate influence on Arctic ecosystems (IPCC 2013) and the consequences of this change are yet to be fully realized. BCP collections are helping us to understand the implications of climate change for species distributions and community diversity (e.g., Baltensperger and Huettmann 2015; Hope et al. 2015) and to monitor fundamental biological attributes such as life histories, life cycles, and parasite transmission dynamics (e.g., Kutz et al. 2005; Jenkins et al. 2006; Laaksonen et al. 2015). For example, a series of discoveries based on field collections radically altered our understanding of the ungulate lungworm fauna in the Beringian region and broadly across the Holarctic (e.g., Hoberg et al. 1995, 2008, 2017; Kutz et al. 2001b, 2007; Jenkins et al. 2005). Links between lungworm demography and climate were elucidated through the development of DNA sequence-based diagnostic methods that probed extensive archival samples from across the North American Arctic (e.g., Kutz et al. 2012; Hoberg et al. 2013; Verocai et al. 2014). Application of these methods demonstrated recent range expansion onto Victoria Island from the mainland of the central Canadian Arctic by the lungworms Umingmakstrongylus pallikuukensis in muskoxen and Varestrongylus eleguneniensis primarily in barren-ground caribou (Kutz et al. 2013; Hoberg and Brooks 2015). Global warming that led to a climatological tipping point contributed to this expansion as development and transmission of Umingmakstrongylus shifted from a 2- to a 1-year cycle (Kutz et al. 2005). Accelerated development by larvae in gastropod intermediate hosts resulted in increasing abundance and infection pressure for parasites in muskoxen. Changing infection dynamics, in conjunction with a shifting balance in the distribution of permissive environments in the low Arctic islands, facilitated northward expansion since 2000 (Kutz et al. 2013; Hoberg and Brooks 2015). Lungworms provided the first examples of climate-related changes in development and distribution for any macroparasite in a large mammal (Kutz et al. 2005).

The generality of climate as an important determinant for host–parasite assemblages emerges from insights derived from specimens. Climate change influences population establishment and subsequent range shifts, structuring diversity across both temporally shallow (e.g., as demonstrated on Victoria Island) and deep time (e.g., episodic events linked to climate oscillations throughout the Quaternary). On the periphery of expanding geographic ranges, ecological perturbation drives the opportunity for colonization due to a shifting balance in the range of permissive environments suitable for parasite development and opportunities for expansion and establishment that are directly limited by capacities for transmission, resilience, and persistence in an array of host species (Agosta et al. 2010; Hoberg and Brooks 2015; Laaksonen et al. 2015). Further, these signify the Beringian nexus as a fundamental exemplar of biotic assembly as addressed in the Stockholm Paradigm (e.g., Hoberg and Brooks 2008, 2013, 2015; Hoberg et al. 2012a; Araujo et al. 2015).

The global climate is changing with an expected irreversible state shift that may already be in play (Barnosky et al. 2012). Northern high latitudes are a special focus because the biota is currently experiencing heavy impacts from three directions: competition from invasive and nonnative species, habitat conversion, and climate warming. Both intentional and accidental introductions of nonnative species have already led to significant changes in northern environments. BCP collecting efforts identified introductions of Peromyscus maniculatus in the Copper River Basin (MacDonald et al. 2009) and red squirrels (Tamiasciurus hudsonicus) to several islands in southeast Alaska (MacDonald and Cook 2007) as well as a series of new distributional records of mammals (e.g., Cook et al. 2016b). We also provided further documentation of introduced populations of Nearctic Neovison and Ondatra in the Palearctic. These and other anthropogenic introductions have considerable implications for host and parasite diversity and should be carefully monitored (e.g., Hoberg et al. 2015). BCP collections provided the critical sampling required to test questions related to how introduced populations expand with considerable impacts on native, often naïve organisms.

Of particular concern are populations isolated on islands, especially those of the Aleutian Islands, Alexander Archipelago, and Canadian Arctic Archipelago. Fur “ranchers” introduced fur-bearers such as arctic and red fox, marten, and mink and their prey, especially arctic ground squirrels, to more than 450 islands in Alaska (Bailey 1993) in the 20th century. These introductions had devastating impacts on the native fauna, including nesting waterfowl such as geese (Bailey 1993). Molecular genetic analyses of BCP samples of 17 arctic ground squirrel populations (including 12 North Pacific islands) throughout southwestern Alaska (Cook et al. 2010) provided a framework for disentangling recent human-mediated introductions from older endemic populations, with the study largely corroborating the written record of early explorers. Such specimen-based projects allow federal managers to prioritize eradication efforts while also protecting endemic populations.

Conversely, some species that show deep phylogeographic structure in northwestern North America occur naturally on the western islands of the Alexander Archipelago (e.g., Dawson et al. 2014). Some of these distinctive lineages of mammals and associated parasites represent paleoendemics, meaning that their DNA signatures reflect long-term persistence and divergence in glacial refugia along the North Pacific Coast (Koehler et al. 2009a; Hoberg et al. 2012b; Sawyer and Cook 2016), and these signatures are consistent with the Coastal Refugium Hypothesis. Phylogeographic studies along the North Pacific Coast, combined with insight from fossils recovered from the extensive karst systems (Heaton and Grady 2003), suggest that a suite of terrestrial species persisted nearby in coastal ice-free refugia during glacial advances. Identification of these endemic taxa in the region brings them to the attention of natural resource managers, increasing the likelihood of their persistence (Dawson et al. 2007).

In the early 1900s, the Alaska Game Commission introduced American marten (Martes americana) from the mainland to select islands of Southeast Alaska’s Alexander Archipelago to strengthen local economies under the assumption that no marten species previously inhabited these islands. Genetic (Stone et al. 2002; Small et al. 2003; Dawson et al. 2017) and parasitological (Koehler et al. 2007; Hoberg et al. 2012b) investigations stemming from BCP collections now suggest that several islands in the Alexander Archipelago harbor a different species, the endemic Pacific marten (Martes caurina), and that some of their populations may have been negatively impacted by competition and hybridization with introduced M. americana. Salvaged marten carcasses collected annually over the last two decades, if bolstered by continued specimen collecting, will offer rare opportunities to explore the dynamics of these human-mediated hybrid zones. Because translocated, endemic, and hybrid populations occur across a series of islands, the Alexander Archipelago provides a powerful series of independent experiments that can illuminate outcomes of species introductions and other perturbations (e.g., industrial logging) on these insular mesocarnivore populations.

Another mustelid example from high latitudes hinges on a series of studies of ermine across the Holarctic that revealed four independent phylogenetic lineages with disparate demographic and evolutionary histories (Fleming and Cook 2002; Dawson et al. 2014; Colella et al., unpublished data). Two ermine lineages are widely distributed across the Arctic and meet in south coastal Alaska. Fine-scale BCP sampling within that region and including the Alexander Archipelago shows that a third distinctive lineage is endemic to just a few islands (Prince of Wales Archipelago and Haida Gwaii). This island lineage roughly corresponds to the subspecies Mustela erminea haidarum, and is federally protected in Canada; however, in Alaska, these ermine are not protected within their restricted range. Given the history of intensive habitat conversion that these coastal regions have experienced, combined with change underway due to warming climate (Hennon et al. 2012), informed wildlife management depends on a solid understanding of both the distribution and the biogeographic history of species because an unusually high percentage of taxa have been identified as endemic along the North Pacific Coast (Cook and MacDonald 2001). This foundation can only be provided by spatially extensive and site intensive archives like those of BCP (Cook et al. 2006; Cook and MacDonald 2013).