Rossman and Eberle cite biochemical data that suggests New World "Natrix" are more closely related to each other and to New World natricine genera than they are to Old World Natrix. Nonetheless, such results are possible if Old World Natrix is paraphyletic with respect to New World watersnakes. Looking at the data of Dowling et al., this scenario appears to be strongly supported. Old world watersnakes are basal to the New World watersnakes in Dowling et al.'s data. Comparing the definitions of the genus Natrix with that of Nerodia in Rossman and Eberle, they are identical with the exception of the number and types of chromosomes and hemipenial structure.
The differences in karyotypes can be the result of an evolutionary change in the derived lineage of the New World natricines (an autapomorphy in this lineage), therefore it does not refute a close relationship between New World and Old World watersnakes. The hemipenial structure of Nerodia is indeed disparate from Old World Natrix. However, Regina, which was once included as part of New World "Natrix", has a hemipenial morphology that is very similar to that of Old World Natrix (compare figs. 4D and 4 H of Rossman and Eberle), suggesting that the hemipenial structure of Nerodia is a derived condition. Regina is basal to Nerodia and Old World watersnakes are in turn basal to Regina. Therefore the similarities between the hemipenis of Regina and Natrix appear to be a shared ancestral character which has since been modified in the derived lineage Nerodia.
Old World Natrix and New World "Natrix" (which includes Regina, Colonophis and Seminatrix) appear to be part of a monophyletic group. The extensive morphological similarities between Old World and New World watersnakes therefore appear to be due to common ancestry rather than convergent evolution.
The species in Regina, Nerodia, Storeria, Colonophis and Seminatrix could have been retained in a paraphyletic Natrix with Old World species of Natrix. Without including Regina and Colonphis, which are basal to Nerodia, Nerodia plus Natrix would be polyphyletic. Therefore, unless there is a major revision of the watersnakes, for the sake of taxonomic stability, it is best to recognize Natrix, Regina, Nerodia and Clonophis as valid genera.
The evolutionary history of this group is pretty clear from the morphological and biochemical data:
Water snakes originated in the Old World. One species of Old World watersnake migrated to the New World (most probably from Europe across a land bridge connecting the eastern US and Europe) and this species is ancestral to Storeria, Regina, Nerodia, Clonophis and Seminatrix. Storeria, Regina, Clonophis and Seminatrix are basal among the New World Natricines and they resemble Old World Natrix morphologically. Nerodia is a more derived group and differ more from Old World Natrix. These species also are restricted to the eastern US in distribution. It is from Nerodia that the species of Thamnophis evolved. Thamnophis appears to be more suited to the drier conditions of the western US and the members of this genus are the only Natricines that have spread westward. Thamnophis sauritus is probably not a member of the genus Thamnophis, since it is basal to the clade that includes Nerodia, Thamnophis elegans and Thamnophis sirtalis. T. sauritus may only be convergently similar to the other species of Thamnophis.