Morphology, Fossils, Divergence Timing, and the Phylogenetic Relationships of Gavialis

Although morphological data have historically favored a basal position for the Indian gharial (Gavialis gangeticus) within Crocodylia and a Mesozoic divergence between Gavialis and all other crocodylians, several recent molecular data sets have argued for a sister-group relationship between Gavialis and the Indonesian false gharial (Tomistoma schlegelii) and a divergence between them no earlier than the Late Tertiary. Fossils were added to a matrix of 164 discrete morphological characters and subjected to parsimony analysis. When morphology was analyzed alone, Gavialis was the sister taxon of all other extant crocodylians whether or not fossil ingroup taxa were included, and a sister-group relationship between Gavialis and Tomistoma was significantly less parsimonious. In combination with published sequence and restriction site fragment data, Gavialis was the sister taxon of all other living crocodylians, but the position of Tomistoma depended on the inclusion of fossil ingroup taxa; with or without fossils, preferred morphological and molecular topologies were not significantly different. Fossils closer to Gavialis than to Tomistoma can be recognized in the Late Cretaceous, and fossil relatives of Tomistoma are known from the basal Eocene, strongly indicating a divergence long before the Late Tertiary. Comparison of minimum divergence time from the fossil record with different measures of molecular distance indicates evolutionary rate heterogeneity within Crocodylia. Fossils strongly contradict a post-Oligocene divergence between Gavialis and any other living crocodylian, but the phylogenetic placement of Gavialis is best viewed as unresolved.