Ecological divergence in the yellow-bellied kingsnake (Lampropeltis calligaster) at two North American biodiversity hotspots

Highlights

• The Lampropeltis calligaster complex is composed of three species occupying distinct ecological niches in North America.

• Speciation in the Eastern Nearctic is associated with ecological processes occurring near the Mississippi River.

• Divergence occurred in the late Pliocene or early Pleistocene and was likely ecological rather than allopatric.

• Two of the three species delimited are likely imperiled by urbanization and climate change.

Abstract

Several biogeographic barriers in the Eastern Nearctic appear to reduce gene flow among populations of many species in predictable ways, however these patterns used to infer process of divergence may be deceiving if alternative modes of diversification are not considered. By using a multilocus statistical phylogeographic approach to examine diversity within a North American snake, Lampropeltis calligaster, we find that mode and timing of speciation near the Mississippi River embayment and peninsular Florida, two main biodiversity hotspots in eastern North America, challenge previously held notions of strict vicariance as the causal factor behind patterns of divergence seen among taxa at these locations. We found three species inhabiting distinct ecological niches with divergences dating to the mid- and early-Pleistocene with subsequently stable or increasing effective population sizes, further supporting the idea that the Pleistocene was an important driver of diversification in North America. Our results lead to a revised hypothesis that ecological divergence has occurred in this group across environments associated with the Mississippi River and at the Florida peninsula. Importantly, in their western distributions, we show that species divergence is associated with the ecological transition from distinct forested habitats to grasslands, rather than the nearby Mississippi River, a barrier often implicated for many other organisms. Additionally, we stress the importance of examining each delimited lineage with respect to conservation, since ecological niche models suggest that by the end of the century changes in climate may negatively alter habitat suitability and, barring adaptation, substantially reduce the suitable range of two of the three species we identified.

Introduction

Geographic barriers to gene flow that isolate populations and generate species have been characterized using phylogeographic analyses (Avise, 2000), which play a strong role in understanding how species form with regard to timing, degree of isolation (allopatric or parapatric) and location, while also identifying previously undescribed or unrecognized diversity (Pyron and Burbrink, 2010). In areas such as the Nearctic, physical and ecological barriers have been examined across many taxa and appear to isolate populations, reducing gene flow across organisms in similar ways (Avise, 2000, Pyron and Burbrink, 2010). For wide-ranging taxa, one or more of these barriers is often found to influence population structure predictably. Causes for these repeated patterns observed among taxa may be deceiving in that they appear to have similar superficial effects but vary in key characteristics such as fine-scale location, timing or mode of divergence. Examining these key characteristics associated with biogeographic barriers enhances our understanding of regional processes related to speciation. Certainly, patterns may appear repeated among taxa, but underlying processes may vary greatly for each individual species.

In eastern North America, there are several major areas where terrestrial taxa show strong overlapping patterns of divergence. First, the Mississippi River region is particularly important for isolating populations and generating species (Pyron and Burbrink, 2010, Robison, 1986). During periods of climatic change in the Pleistocene, the volume of glacial meltwater and the river bed itself was subject to fluctuation in size and location (Royall et al., 1991, Smith, 1996). The current prevailing hypothesis is that the Mississippi River was instrumental in generating biodiversity in eastern North America (Burbrink et al., 2000, Fontanella et al., 2008, Gamble et al., 2008, Guiher and Burbrink, 2008, Howes et al., 2006, Jackson and Austin, 2010, Leaché and Reeder, 2002, Lemmon et al., 2007, Pyron and Burbrink, 2009, Walker and Avise, 1998). Congruently, a longitudinal ecological transition zone from the central prairies to the eastern Nearctic forests suggests the possibility that ecological divergence in these different habitats drove speciation. This transition zone along the upper portions of the river was affected by glacial events during the Pleistocene and may not represent a long term, stable barrier (Robison, 1986, Royall et al., 1991), therefore the processes that isolated populations on separate sides of the Mississippi River (Burbrink et al., 2000, Fontanella et al., 2008, Guiher and Burbrink, 2008, Pyron and Burbrink, 2009, Ruane et al., 2014) may be due to ecological transition and not historical vicariance associated with the river itself. Both processes may be at work and the shared phylogeographic structure observed in codistributed taxa may only be pseudocongruent, indicating that spatial distributions of lineages are broadly similar across the Mississippi River, but may differ in mode, timing or actual location of speciation (Pyron and Burbrink, 2010, Soltis et al., 2006).

Another major region important for isolating populations in the Eastern Nearctic is located where the Florida peninsula meets the continental US (Fontanella et al., 2008, Guiher and Burbrink, 2008). The patterns of divergence there are typically explained by refugial isolation on islands in northern and central peninsular Florida during periods of elevation in sea level during the Oligocene, Pliocene and early Pleistocene (James, 1961, Soltis et al., 2006). Furthermore, this area is home to multiple rivers that may have been responsible for limiting secondary contact (Soltis et al., 2006). Alternatively, the Florida peninsula features an area of fairly sharp ecological turnover where major climate zones and ecoregional transitions occur (Omernik, 1987). This ecological transition from the southern coastal plain to the Piedmont may be a causal factor driving regional ecological speciation, particularly in groups with key characteristics that are not well explained by traditional island or riverine vicariance.

Although there is disagreement about the role of the Pleistocene in generating modern diversity (Klicka and Zink, 1997), recent work has shown that changes in climate during the Pleistocene may have significantly driven diversification, especially within the common snakes of the genus Lampropeltis (Ruane et al., 2014). Estimating timing of diversification while including post-divergence demographics also helps determine how populations have responded to climatic cycles. Testing these hypotheses to assess the importance of the Pliocene and Pleistocene in shaping current biodiversity requires credible estimates of speciation times within taxa to accurately reflect evolutionary history.

To elucidate processes of diversification associated with these barriers at the Mississippi River region and in the Southeastern United States (SEUS), it is important to first examine organisms with distributions that cross these transition zones and respond to landscape features. Snakes are exceptional candidates for studies on the effect of ecology and landscape isolation; they are known to respond both to vicariant barriers and ecological transitions (Brandley et al., 2010, Burbrink et al., 2011). The yellow-bellied kingsnake, Lampropeltis calligaster (Harlan, 1827), is found across much of the southern United States and three subspecies are currently recognized, distinguished by color pattern and average number of infralabial scales (Blanchard, 1921, Price, 1987). This morphological variation also corresponds to areas on either side of ecological transitions associated with the Mississippi River, and near the Florida peninsula, offering an opportunity to test associated patterns and processes of speciation at these two biodiversity hotspots. The eastern subspecies of L. calligaster, the Mole Kingsnake, Lampropeltis c. rhombomaculata (Holbrook, 1836) is generally found east of the Mississippi River and on the Piedmont and eastern seaboard, while Lampropeltis c. calligaster (Harlan, 1827), the Prairie Kingsnake, is primarily found in the eastern Great and Central Plains from Texas, Oklahoma and Kansas east to Indiana (Fig. 1). The two subspecies meet east of the Mississippi River, where morphological evidence suggests ranges overlap and hybridization occurs (Cook, 1945). The eastern and western subspecies were both originally described as unique species (Harlan, 1827, Holbrook, 1836) but were reduced to subspecies by Cook (1945) based on intergradation and parapatric distribution in the eastern portion of the State of Mississippi. A third and much rarer form, Lampropeltis c. occipitolineata (Price, 1987), the South Florida Mole Kingsnake, is poorly represented in museum collections and is found in only a few counties in South Florida (Krysko et al., 2000, Price, 1987). Importantly, the former two taxa were considered distinct species occupying distinctly different niches (Wright and Wright, 1957) prior to being arbitrarily delimited to subspecies (Cook, 1945).

While statistical species delimitation is becoming common (Fujita et al., 2012), and it is clear that even in North America where surveys of species diversity have been conducted for more than 200 years, phylogeographic studies are still demonstrating that a substantial amount of biological diversity remains undiscovered in terms of uncovering numerous cryptic lineages and populations (Burbrink and Guiher, 2015, McKelvy et al., 2016, Myers et al., 2013a, Pyron and Burbrink, 2009, Ruane et al., 2014). However, the potential fate with regard to conservation of newly recognized species is rarely tested. An often-overlooked consequence of uncovering cryptic diversity is that newly erected taxa are subsets of the total range of the species complex from which they were delimited. Distinct scenarios for each lineage regarding the influence of climate change and urbanization are likely. Conservation efforts to facilitate the survival of these species must reflect the best available data (Malaney and Cook, 2013). Recent discoveries of cryptic lineages with confined ranges are not rare, and small range size is understood as a factor correlated with extinction (Bielby et al., 2008, Niemiller et al., 2013); if researchers and stakeholders are expected to incorporate newly uncovered taxa into management plans, available information, including niche trajectories modeled from available climate data, should be examined.

In this study, we examine the phylogeographic history of the L. calligaster complex to first determine if distinct lineages occur at known geographical barriers. We then test specific causes of lineage diversification from among several potential candidate hypotheses. To do this, we investigate divergence times to assess if diversification corresponds with major Pleistocene events. We use redundancy analysis and Mantel tests to determine if the pattern of divergence we observe at the Mississippi River region is best explained by a model of vicariance at the river or by divergence across an ecological gradient. To test mode of speciation, we explore the niche identity of each species in a biogeographic context via environmental niche modeling and assess if gene flow is asymmetric between delimitable species in adjacent niches. We then use historical demographic analyses to determine if populations from different species respond to climatic fluctuations in similar ways. We generate an estimate of future suitable climatic space across North America, and discuss conservation implications due to changes in potential range size for the rare occipitolineata form of L. calligaster. Our results lead to a better understanding of what causes diversification at major biogeographic transitions in eastern North America.