Rhinoclemmys

Le, Minh & Mccord, William P., 2008, Phylogenetic relationships and biogeographical history of the genus Rhinoclemmys Fitzinger, 1835 and the monophyly of the turtle family Geoemydidae (Testudines: Testudinoidea), Zoological Journal of the Linnean Society 153 (4), pp. 751-767 : 761-763

BIOGEOGRAPHY OF RHINOCLEMMYS

Our biogeographical results do not support the hypothesis that Rhinoclemmys migrated to the New World from Europe or Africa. In addition, as the sister clade of Rhinoclemmys within the Geoemydidae has all basal lineages restricted to Asia and the fossils of the family putative ancestors, the Lindholmemydidae , are all Asiatic ( Sukhanov, 2000; Claude & Tong, 2004), the origin of geoemydids can safely be placed in Asia. Other palaeontological evidence corroborates the hypothesis that the ancestors of the group migrated over the Bering Land Bridge during the warmer period in the early Eocene as did the mammals ( Sukhanov, 2000; Beard, 2002; Bowen et al., 2002).

The Bering Strait separating Asia and northwestern America was formed about 100 Mya and remained open occasionally until the Pleistocene, but due to its northern latitude it is clear that animals only migrated over the Bridge during warm periods ( Sanmartin et al., 2001). According to Bowen et al. (2002) and Beard (2002), a short global warming period facilitated the dispersal of at least three mammal groups (uintatheres, rodents and hyaenodontids) through this route in the late Tiffanian (57 Mya), early Clarkforkian (56 Mya) and early Wasatchian (55 Mya).

The abrupt occurrence of geoemydid fossils (genus Echmatemys ) in North America (Wyoming and South Dakota) around 55 Mya (in the earliest Wasatchian, early Eocene) ( Hutchison, 1996) is congruent with this hypothesis. The monophyly of Rhinoclemmys suggests that this group only colonized the Americas once, and this colonization probably corresponds with the third wave of mammal invasion of the Americas. Other fossil records of Bridgeremys , a genus related to Rhinoclemmys , were also found in Wyoming in the middle Eocene between 46 and 49 Mya ( Hutchison, 2006).

The fossil record also indicates that the diversification of geoemydids occurred very early in their history with fossils found in the early Eocene in North America and Europe ( Godinot & de Broin, 2003; Claude & Tong, 2004). This implies that the family had very widespread distribution and that the current distribution may just be relict. Thus far, palaeontological evidence supports two separate migration routes from Asia. According to Godinot & de Broin (2003), fossil forms of Europe and North America in the early Eocene are completely different from each other even though they both seem to be related to the Asian forms (but see Hutchison, 1996). The same hypothesis has been proposed for the migrations of tortoises, family Testudinidae , to Europe and North America ( Le et al., 2006). This pattern shows that these two families had already diversified in Asia well before they migrated to Europe and the Americas, and that there was no exchange of turtle fauna between Europe and North America.

In the Americas, the ancestors of Rhinoclemmys may have dispersed to tropical regions in Central America during the cooling period of the Eocene, as did other groups of reptiles and amphibians, due in part to the formation of uplands in western North America and Mexico ( Savage, 2002). Nevertheless, living lineages of this genus only started to diversify in the late Oligocene ( Fig. 6 ). Thus, the emergence of the Sierra Madres of Mexico, the Nuclear Highlands – a combination of the highlands of Chiapas, Guatemala and Honduras – and the Panama land bridge in the Oligocene, Miocene and Pliocene, respectively, substantially influenced the biogeographical patterns of the local herpetofauna ( Savage, 1982, 2002). Mountain uplift in northern Mexico, i.e. Sierra Madre Oriental, might have isolated the R. areolata group (Lowland Atlantic) from R. rubida (Lowland Pacific) during the early Miocene. This vicariance event also had important impacts on other reptiles and amphibians, resulting in similar divergence of frogs in the Hyla microcephala complex and lizard species in the Enyaliosaurus group of the genus Ctenosaura ( Savage, 1982, 2002).

The emergence of the Nuclear Highlands in the middle Miocene may have caused the divergences of R. pulcherrima and R. annulata , and also isolated R. areolata , largely distributed in the Yucatán Peninsula, from the R. punctularia + R. funerea group. Interestingly, our estimate indicates that these two events took place almost simultaneously ( Fig. 6 , Table 3). The species of the R. areolata group show a clear progression rule, where younger species are found further south, from the Lowland Atlantic to Amazon South ( Fig. 7 ). As shown by the phylogenetic results and by the distribution of R. punctularia + R. funerea , R. annulata and R. nasuta , Rhinoclemmys invaded South America at least four times. Three of these invasions are likely to have taken place after the emergence of the Isthmus (see Fig. 6 ).

Due to the fact that R. nasuta is endemic to the Choco, there are two explanations for its current distribution. Rhinoclemmys nasuta invaded the Choco through the Panamanian Land Bridge as did other species in the Pliocene and its populations in the north subsequently went extinct ( Carr, 1991). However, this hypothesis conflicts with the fact that all other species representing other major clades within this genus, R. pulcherrima , R. areolata and R. rubida , still occur in the north. Alternatively, R. nasuta migrated to South America before the emergence of the Isthmus of Panama. Due to the general limited dispersal ability and limited tolerance to seawater in turtles, it might have used different means to reach South America rather than island hopping and waif dispersal as shown in many groups of mammals ( Marshall, 1979; Webb, 1985). It is very likely that this species inhabited the Choco in the early Miocene when this region was still a series of volcanic islands close to Central America ( Savage, 2002). The subsequent movement of the Choco block led to its collision with northern South America in the late Miocene ( Duque-Caro, 1990; Savage, 2002), probably bringing with it part of the Central American fauna. The close relationship between the Choco and Central America has also been reported in different bird groups ( Cracraft & Prum, 1988; Brumfield & Capparella, 1996; Bates, Hackett & Cracraft, 1998).

The diversification within the R. punctularia and funerea groups is likely to have been influenced by dispersals across the Isthmus of Panama, although the split between these two groups might have occurred before the closure of the land bridge ( Figs 6 , 7 ). In addition, the Pleistocene effect in South America might have had impacts on the distribution of R. diademata and R. punctularia . In particular, their distribution seems to fit well with the refugia hypothesis ( Haffer, 1969) with the former being restricted to the Maracaibo Basin and the later distributed in the lower Amazon Basin. The increased aridity in the lowlands during the Pleistocene may explain the gap between the distributions of these two species. Because both of them are semi-aquatic, this phenomenon can have a significant impact on constraining their ranges. In fact, Rhinoclemmys presumably had a much wider range because fossils have been found in Brazil hundreds of kilometres south of its current range, and on the Santa Elena Peninsula ( Carr, 1991).

Figures

Figure 6. Time calibration using the program BEAST. The error bar on each node represents the 95% confidence interval calculated by the program. The column on the right shows the time slice of Isthmian closure (3.5–2.5 Mya; Coates & Obando, 1996). Pli + Ple: Pliocene + Pleistocene.

Figure 7. Areas of endemism and the area cladogram illustrating the relationship among species of Rhinoclemmys. Sources of these areas of endemism are from Savage (2002) (Central America), Cracraft (1985), and Haffer (1985) as assigned to zoographical regions of Stotz, Fitzpatrick & Moskovits (1996) (South America): AL, Atlantic Lowland; AN, Amazon North; AS, Amazon South; CHO, Choco; NSA, Northern South America; NUH, Nuclear Highland; PH, Panamanian Herpetofauna; PL, Pacific Lowland.