- Top of page
- Author Contributions
- Data Accessibility
- Conflict of Interest
Population isolation results from changes in the geographic range of a species. Range expansion may result in colonization of previously unoccupied habitat patches, while range contraction can isolate peripheral populations into “islands” separate from the main range. Peripheral isolates can act as natural laboratories for evolutionary processes because they may experience different ecological and evolutionary pressures than populations in the species’ main range. While individual populations within a larger metapopulation may differ mildly from one another, geographically isolated populations are released from the homogenizing effects of gene flow and may take unique evolutionary trajectories. Geographical isolates may preserve and accumulate these differences over time, resulting in replicate natural experiments on speciation (Key 1968; Themudo and Arntzen 2007).
The evolution and ecology of peripherally isolated populations may also be influenced by interactions with parapatric (or sympatric) relatives along contact zones, where hybridization or ecological competition may occur (Barton and Hewitt 1985; Sætre and Sæther 2010). A peripherally isolated population completely surrounded by populations of a close relative is known as an enclave (Arntzen 1978). The formation of enclaves may be facilitated by the presence of habitat mosaics, or differential rates of hybrid zone movement. Moving contact zones, in which one species expands its range at the expense of another (see review in Buggs 2007), may result in either replacement or assimilation of the species whose range retracts, depending on the frequency of hybridization and genetic introgression (Rhymer and Simberloff 1996; Mallet 2005). These represent special cases of range boundary dynamics.
Comparisons between peripheral and main range populations of a species can help elucidate the ecological, environmental, and population genetic processes that shape an organism's responses to life at its range boundary. Here, we make such a comparison between enclave and main range populations of Black-capped Chickadee (Poecile atricapillus) in the Appalachian region. We assess genetic variation in these populations and those of a congener, Carolina Chickadee (P. carolinensis), with which atricapillus is known to hybridize. Using a multilocus molecular survey, we searched for genetic evidence of hybridization and introgression in an atricapilllus enclave in the Great Smoky Mountains (GSM), the highest range in the southern Appalachians. We particularly wished to investigate whether geographically separate contact zones between the same taxa can result in fundamentally different levels of hybridization, based on local ecological differences, because this enclave appears to have evolved a unique reproductive isolating mechanism based on elevational movement (Tanner 1952; Tove 1980). This study also has significant conservation implications, as human impact on Appalachian ecosystems has been severe, in the form of habitat destruction, invasive species (Tingley et al. 2002; Simons et al. 2002), and climate change (Thomas and Lennon 1999; Inouye et al. 2000; Crick 2004).
Poecile atricapillus are small songbirds that are found throughout northern North America (Foote et al. 2010). Like many northern taxa, the range of atricapillus includes a southern salient through the Appalachian Mountains (Fig. 1). Other northern birds with Appalachian range extensions include Common Raven (Corvus corax), Red-breasted Nuthatch (Sitta canadensis), Black-throated Green Warbler (Setophaga virens), and Dark-eyed Junco (Junco hyemalis) to name just a few (Price et al. 1995).
Figure 1. Locations and cytochrome-b haplotype proportions of sampled populations. Pie charts located at the geographic location of each sample denote the proportion of each population with the most common mtDNA haplotypes (white for atricapillus and black carolinensis). Alternate haplotypes are represented by smaller sectors of the pie charts, and are not shared between species. Dark gray shading on the background map indicates Poecile atricapillus range and light gray shading represents P. carolinensis range. See Table 1 for population abbreviations.
Download figure to PowerPoint
In this region, the species’ continuous range extends as far south as southern West Virginia (WV) and southwestern Virginia (VA) in upper elevation forest characterized by northern tree species, while in the southern Appalachians atricapillus are only found in high elevation habitats that support Red Spruce (Picea rubens)/Fraser Fir (Abies fraseri)/Yellow Birch (Betula alleghaniensis) ecosystems. These “sky island” communities are widely scattered in the Blue Ridge province and have long been subject to anthropogenic habitat disturbance. With one exception, all atricapillus populations in this region have gone extinct within the last century (Lee 1999). The only remaining large population of atricapillus in the southern Appalachians is in the GSM National Park, on the border of Tennessee (TN) and North Carolina (NC).
Today, this sky island population is separated from the species’ contiguous range by nearly 200 km of marginal or unsuitable habitat, making it probable that the level of genetic exchange between it and main range populations is low. Due to its small geographic range, and restrictive habitat requirements, atricapillus is considered a species of concern in both TN and NC. Although protected by the national park, atricapillus has been identified as the southern Appalachian bird species most likely to become extirpated due to habitat destruction and the least likely to become reestablished in suitable but unoccupied habitat (Hunter et al. 1999).
Extensive hybridization between atricapillus and carolinensis occurs along the main range interface from New Jersey to Kansas (Brewer 1963; Rising 1968; Merritt 1978; Robbins et al. 1986; Bronson et al. 2003a; Curry 2005; Reudink et al. 2007; Olson et al. 2010) and in the northern Appalachians (Johnston 1971; Sattler and Braun 2000; Sattler et al. 2007). Southern Appalachian atricapillus are under threat of ecological replacement or genetic assimilation by the more southerly distributed, morphologically similar P. carolinensis. Each Appalachian sky island left vacant by extirpated atricapillus populations in the past 100 years has been colonized by P. carolinensis (Tanner 1952; Lee 1999).
In contrast to the main range contact zone between the species, hybridization between atricapillus and carolinensis has not been observed in the GSM (Tanner 1952; Tove 1980). Although the two forms occur together in winter flocks in this area, a well-documented gap in their elevational distributions forms before the breeding season (Tanner 1952; Tove 1980), and has been implicated as a reproductive isolating mechanism. This gap in breeding distribution develops during early April, when carolinensis begin nesting below 900 m and atricapillus move upslope to the remaining spruce/fir forests above 1150 m. This distance is equivalent to at least 1.6 km horizontally, depending on slope (Tanner 1952). After the breeding season, the gap disappears as atricapillus move back downslope. The ultimate reason this elevational gap occurs is unknown, but carolinensis can be found breeding at elevations over 1800 m on nearby mountains where atricapillus are absent (Tanner 1952; Simpson 1992), suggesting that the gap is mediated by interspecific interactions rather than a difference in breeding habitat preferences.
Tanner (1952) and Tove (1980) concluded that the GSM population of atricapillus did not hybridize with local carolinensis based on the lack of morphological and vocal admixture, respectively. However, due to their similar morphology and the fact that their vocalizations are learned, molecular methods are more sensitive for differentiating these species and identifying hybrids (Sattler and Braun 2000; Bronson et al. 2005; Sattler et al. 2007). In fact, although hybridization was not always suspected in advance, all populations studied near the main range contact zone of atricapillus with carolinensis have been found to be heavily introgressed at the molecular level (Robbins et al. 1986; Sawaya 1990; Sattler and Braun 2000; Bronson et al. 2003a; Curry 2005; Reudink et al. 2007; Sattler et al. 2007; Olson et al. 2010). Thus, if the conclusions of Tanner (1952) and Tove (1980) regarding absence of hybridization are correct, the GSM population of atricapillus is unique in its purity.
The main goal of the present study was thus to assess the efficacy of the elevational gap as a reproductive isolating mechanism by determining whether cryptic hybridization or introgression between GSM P. atricapillus and local P. carolinensis is evident in the multilocus genotype of the GSM population. Previous studies of hybridization and introgression between these species have used relatively low numbers of highly differentiated molecular markers (Braun and Robbins 1986; Sattler and Braun 2000). Such diagnostic markers facilitate identification of hybrids, but may often underestimate the extent of genome-wide introgression because they are under selection opposing it (Sattler and Braun 2000). In fact, the degree of differentiation may vary dramatically among loci and genomic regions, probably as a result of the homogenizing effects of gene flow on some regions, while genetic incompatibilities or adaptive processes promote divergence in others (Harr 2006; Via and West 2008; Yuri et al. 2009). In order to gain a genome-level perspective on hybridization and introgression, we surveyed both mitochondrial cytochrome-b sequences and a relatively large number of amplified fragment length polymorphism (AFLP) loci from the nuclear genome. AFLP loci have several advantages over other marker types: They are largely neutral, being randomly generated from the whole genome, they require no prior sequence knowledge, they have high reproducibility and hundreds of loci can readily be studied in order to provide an approximation of genome-wide variation at low cost (Bensch and Åkesson 2005).