Spinks et al. (2010, Mol. Ecol. 19: 542–556; 2014, Mol. Ecol. 23: 2228–2241; 2016, Mol. Phylogenet. Evol. 103: 85–97) demonstrated deep phylogeographic divergence within the genus, corresponding to the previously recognized subspecies, and recommended species recognition for pallida.
Note on genus:
Work by Bickham et al. (1996, Herpetologica 52: 89–97), Burke et al. (1996, Herpetologica 52: 572–584), Lenk et al. (1999, Mol. Ecol. 8: 1911–1922), Holman and Fritz (2001, Zoolog. Abhand. Staat. Mus. für Tierkunde Dresden 51: 331–354), Feldman and Parham (2002, Mol. Phylogenet. Evol. 22: 388–398), Seidel (2002, Copeia 2002: 1118–1121), Stephens and Wiens (2003, Biol. J. Linn. Soc. 79: 577–610), Wiens et al. (2010, Biol. J. Linn. Soc. 99: 445–461), Fritz et al. (2011, Zootaxa 2791: 41–53), and Spinks et al. (2016, Mol. Phylogenet. Evol. 103: 85–97) provided ample evidence that the genus Clemmys as previously recognized (e.g., McDowell, 1964, Proc. Zool. Soc. Lond. 143: 239–279) was paraphyletic with respect to the sister genera Emys and Emydoidea, and also possibly Terrapene. Two taxonomic schemes reflecting these relationships are currently in contention. Both would place sister taxa insculpta and muhlenbergii in the genus Glyptemys and leave guttata in the monotypic genus Clemmys (both changes are recognized in this list). However, one scheme (e.g., Feldman and Parham, 2002, op cit.; Spinks and Shaffer, 2005, Mol. Ecol. 14: 2047–2064; Spinks et al. (2016, op cit.) would expand the definition of Emys to include marmorata (and pallida), blandingii, orbicularis (European) and trinacris (Sicilian). This would involve two taxonomic changes and eliminate the genus Emydoidea, which is monotypic as a living taxon, but polytypic if the fossil record is included (Holman, 2002, Michigan Academician 34: 393–394). The other scheme (Holman and Fritz, 2001, op cit.; Stephens and Wiens, 2003, op cit.; Wiens et al. 2010, op cit.; Fritz et al. 2011, op cit.) involves only one taxonomic change, placing marmorata (and pallida) in the now polytypic genus Actinemys, and retaining the polytypic genera Emydoidea (North America) and Emys (Eurasia). The contention originally hinged on the relative importance of eliminating monotypic genera versus maintaining taxonomic stability (fewer changes being preferable). The former is supported primarily by taxonomists who consider monotypic genera to be redundant names and hence of no value in providing phylogenetic information. Thus, although the former scheme requires more changes, it eliminates the genus Emydoidea (which is monotypic only if the fossil record is ignored: Holman, 2002, op. cit), although it retains the monotypic genus Clemmys. Many proponents of the latter scheme believe that monotypic genera are not taxonomically redundant but rather reflect evolutionary distinctiveness (see Mayr and Bock, 2002, J. Zool. Syst. Evol. Research 40: 169–194 for a general discussion of the values of taxonomic stability and recording anagenesis in classification schemes). An analysis by Angielczyk and Feldman (2013, Biol. J. Linn. Soc. 108: 727–755), based on 14 nuclear genes, found that Emys broadly defined is paraphyletic with respect to Clemmys, but more recently Spinks et al. (2016, op cit.) resolved a monophyletic Emys sensu lato based on 30 nuclear loci. Because of the value of current stability, the belief that monotypic genera do provide some phylogenetic information, the uncertainty concerning the monophyly of Emys sensu lato, and the increasing use of the three separate genera in the turtle literature, we here follow the second scheme, recognizing Actinemys, Emydoidea and Emys*, as recommended by Fritz et al. (2011, op cit.).
John B. Iverson (Chair), Peter A. Meylan, Michael E. Seidel, 2018-02-12