UNDERSTANDING THE EVOLUTION OF BASAL MOSASAURINAE USING A SPECIMEN-LEVEL PHYLOGENY: PITFALLS, POLYTOMIES, AND HIGH LEVELS OF VARIATION WITHIN “CLIDASTES”

I examine the relationships among and variation within species of Clidastes to better understand the early evolution of Mosasaurinae, a clade of marine squamates that are known globally from Cretaceous strata. Despite high diversity and an abundance of specimens, a number of phylogenetic and taxonomic issues within the clade are unresolved. One of those is the paraphyly of the genus Clidastes, a basal grade of mosasaurine with an abundant fossil record in North America and three recognized species. To better understand the evolution and systematics of Clidastes, I scored discrete morphological characters for over 50 specimens of Clidastes, as well as other species of mosasaurines. I then performed specimen-level phylogenetic analyses, using both traditional character scorings from previous studies and character states that more precisely describe the anatomy present on a given specimen. Those analyses produced a polytomy within Mosasaurinae, with little resolution even amongst mosasaurine taxa with derived character states. Traditionally, only a few characters of the frontal and quadrate have been used to diagnose species of Clidastes. Examination of a large sample of specimens demonstrates that there are more characters that vary within Clidastes than there are autapomorphies that unite individual species. I performed additional phylogenetic analyses after successively removing variable characters and incomplete specimens. This did little to improve phylogenetic resolution. One possible conclusion is that there were at least three highly variable species of basal mosasaurines that cannot be united under a monophyletic genus and have few autapomorphies that differentiate them. Another possible conclusion is that there was a single, highly variable mosasaurine lineage in North America from which the fossil record is sampling. If this is the case, individual species within that lineage may be difficult to distinguish. The use of specimens as operational taxonomic units in phylogenetic analyses has provided some insight into the taxonomy of other vertebrate groups. However, current methods test for one mode of speciation: cladogenesis. If budding speciation or anagenesis were biological realities for a given clade, a specimen-level analysis likely will not result in a dichotomously branching tree.