Diadophis punctatus

Fontanella, Frank & Siddall, Mark E., 2010, Evaluating hypotheses on the origin and diversification of the ringneck snake Diadophis punctatus (Colubridae: Dipsadinae), Zoological Journal of the Linnean Society 158 (3), pp. 629-640 : 633-635

DIADOPHIS PUNCTATUS

Our analyses agree neither with the hypothesis of Blanchard (1942), who suggested that the ancestral area of D. punctatus was in central Mexico and that D. p. dugesii arose first, nor with the hypothesis that the lineages throughout the eastern USA are derived from a Mexican ancestor. Under a phylogenetic framework, Blanchard’s (1942) hypothesis implies that the southern Mexico populations should be the earliest diverging lineage for any phylogenetic analyses. Likewise, the Gulf Coast corridor hypothesis suggests the drop in sea level during glacial periods of the Miocene/Pliocene exposed the Gulf Coast shelf, allowing dispersal from southern Mexico into the south-eastern USA. Yet the topologies that we estimated indicate that the populations from Mexico are nested well within D. punctatus , in a derived position more closely related to the lineages occupying the western USA rather than basal to Diadophis as a whole or to any of the eastern lineages. The results of all the topological constraint tests indicate a significantly worse topology when the Mexican populations are forced to be the first lineage of the genus Diadophis or sister to the lineages occupying the eastern USA ( Table 1). The placement of Mexican populations nested within the tree suggests that the possession of two posterior temporal scales and moderate number of scale counts, characters considered by Blanchard (1942) to be primitive, are actually derived and apomorphic for the Mexican lineages. That said, the results of the constraint tests indicate that the precise placement of the Mexican lineage is not statistically well supported with respect to the western lineages. The central Mexico lineage can be shifted as sister to all of the lineages of the Western clade without a significant change in likelihood score. Although the likelihood value of this topology is not significantly worse, it presents a convoluted biogeographical pattern.

Regardless of optimality criterion, the origin of the genus Diadophis appears to be in south-eastern USA. We further tested this hypothesis by constraining lineages in the eastern USA to be most closely related to lineages in the western USA, each of which resulted in significantly worse topologies ( Table 1). These conclusions are supported by the ancestral area reconstructions, which unequivocally inferred a south-eastern origin for D. punctatus . Additionally, the oldest fossil of the genus Diadophis was found in southern Florida, lending additional support for a south-eastern origin ( Holman, 2000). Species origins in the south-eastern USA have been attributed to a combination of geological and zoogeographical factors ( Adams, 1902). The South-east is an ancient and highly diverse landscape over half a billion years in Values in parentheses represent standard deviation (SD) and the 95% confidence interval (CI) estimated using the uncorrelated lognormal Bayesian relaxed molecular clock in BEAST v. 1.4.4. Numbers for nodes follow Figure 3 .

age and composed of numerous physiographic provinces. Differentiation of species has also been fostered by the South-east’s stable geological history. This region was neither significantly disturbed during ice ages nor completely inundated by rising sea levels during interglacial periods. These upland areas have served as the primary ‘spawning sites’ for the evolution of new species ( Adams, 1902; Butler & Mayden, 2003).

The combined data inferred an initial divergence within D. punctatus during the late Miocene approximately 6.5 Mya with the separation of the Gulf Coast lineages from the remaining conterminous lineages ( Fig. 2 , Table 2). This time period was marked by a major cooling trend with the progression of glaciers across the northern and Pacific regions ( Dorf, 1959). The remaining major divergences occurred during the warming trend of the Pliocene, between 5–3 Mya, following the retreat of the Miocene glaciers. From the south-eastern USA, the major biogeographical pattern follows a south-east to north-east then westward directionality of historical migration, with the divergence of the Mid-Atlantic lineage followed by a split across the north-eastern USA separating the Appalachian clade of Fontanella et al. (2008). The new biogeographical scenario within the Appalachian clade suggests a southern Appalachian origin followed by the divergence of the North-east and Western Kentucky lineages respectively.

Across the south-west, the results suggest a historical migration from the central USA into Mexico and across the south-western deserts. The position within the phylogeny, the date estimate, and the ancestral area reconstructions for the south-western + Mexico clade combined suggest a relatively recent invasion into central Mexico. Typically, reptiles occupying the Western USA are confined to the xeric habitats of the Chihuahuan and Sonoran deserts in northern Mexico ( Riddle & Hafner, 2006) and do not extend into the grasslands of central Mexico or as far south as the Trans-Volcanic Belt. Interestingly, the divergence date estimate for the central Mexican lineage indicates a recent origin, approximately 34 Kya. Based on palynological records, the Pleistocene conditions of Mexico were very different from present. Northern Mexico had a more temperate forested climate with extensive lakes and wetland areas, whereas in central Mexico, glaciers expanded and there were extensive pine forests mixed with alpine grasslands ( Metcalf et al., 2000). Since that time, the reorganization of atmospheric circulation caused major changes to seasonality and precipitation throughout Mexico, leading to the drying out of northern Mexico and resulting in significant changes to species distributions ( Metcalf, 2006). Unfortunately, the sparse sampling throughout Mexico still limits our understanding of the southern extent of the North Texas and Great Basin lineages and the genetic barriers between these lineages and the one sampled Mexican lineage. Further sampling throughout these areas is needed to more accurately define the range limits and potential barriers to these lineages.

Figures

Figure 2. Simplified tree with the mean dates of origin (in millions of years) for each of the major lineages and haplogroups. Numbers below nodes refer to Table 2. Letters above nodes refer to the optimal reconstructions of ancestral areas; A, Coast Plain east of the Mississippi; B, North-eastern USA; C, west of the Mississippi River to the Rocky Mountains; D, west of the Rocky Mountains; E, Mexico.

Figure 3. Map showing the distribution of each Diadophis punctatus lineage diagnosed by mtDNA variation. Symbols correspond to Figure 1. Dark line corresponds to distribution of D. punctatus throughout the USA and Mexico.