The Phylogeny of Euryapsid Reptiles
The Analysis of CT Scan Data
Phylogenetic Simulation Studies
My paleontological research is committed to improving our picture of the tree of vertebrate evolution using the methods of phylogenetic systematics. Since the first application of these methods to vertebrate evolution during the nineteen seventies a coherent, encompassing, and scientifically falsifiable hypothesis of vertebrate evolution has taken shape. Nevertheless, it is confined by limitations in the scope of available morphological data and methodological difficulties attending highly transformed taxa. Controversies over alternate methodological assumptions, moreover, render many details of this picture somewhat ambiguous. My research seeks to optimize our understanding of vertebrate evolutionary history by directly assaulting these limitations. To this end, I have analyzed enigmatic taxa omitted from earlier phylogenetic studies, have exploited innovative sources of morphological data sets, and have identified optimal methodological strategies through empirical tests of numerous options.
My dissertation research addressed a persistent gap in systematists' understanding of vertebrate evolution, the position of the dominant marine reptiles of the Mesozoic Era, ichthyosaurs and sauropterygians. As secondarily aquatic descendants of terrestrial forebears, these groups pose special challenges to systematists, in that the hierarchy of anatomical innovations used to reconstruct phylogeny, acquired in this case during evolution on land, tends to be "erased" by adaptation to aquatic life.
Successful completion of this project required the identification of derived characters shared by these marine taxa, and their close terrestrial relatives despite the former's drastic transformations. This was done using an exhaustive set of osteological morphological characters taken from all regions of the body, and from a wide taxonomic range of marine taxa and potential close relatives: 58 ingroup taxa and 478 characters. The result illuminated the close relationship of ichthyosaurs and sauropterygians to one another, and their position as basal members of Archosauromorpha. New features of the phylogeny of related non-marine reptiles, including the close relationship of rhynchosaurs and choristoderes, were reciprocally illuminated by the combination of marine and non-marine taxa in a single analysis. Based on the revision of the general phylogenetic framework of Archosauromorpha provided by my dissertation research, I am currently moving forward to address in greater detail specific issues arising within it, including the internal phylogeny of the ichthyosaurs and of the placodonts, derived durophagous sauropterygians.
A phylogenetic result is said to be an hypothesis of phylogeny. The hypothesis yielded by my dissertation research was tested for stratigraphic congruence, the degree to which the branching order of the phylogenetic tree matches the stratigraphic sequence of the appearance of taxa. Although generally congruent, several cases were noted in which the recovered phylogeny suggested the presence of "ghost lineages," persistent evolving lineages with no trace in the fossil record. The concentration of many character changes on these lineages indicates prolonged evolution, suggesting that the ghost lineage is real, not the result of an error in the analysis. Such identification of ghost lineages effectively highlights pieces of evolutionary history that field paleontologists have failed to sample. The actual recovery of fossil representatives of these lineages is a high priority for my future field work.
Return to top:
My second major research emphasis involves the employment and interpretation of high-resolution CT scan data. Confidence in the results of phylogenetic analysis increases with the range and depth of character information included. I have, therefore, augmented data obtained from visual inspection of fossils with information on their internal structure, gained non-invasively by CT scan, whenever possible. My efforts fortuitously coincided with a major effort by other researchers at the University of Texas at Austin to probe the utility of high-resolution industrial CT scanners for natural sciences. These machines differ from medical CT scanners by achieving scan resolutions one to two orders of magnitude greater and by generating data in exportable, rather than proprietary file formats. Currently, U. T. possesses the only industrial scanner in the country dedicated to the study of natural science specimens. My association with this effort has enabled me to participate in developing scanning algorithms specifically for paleontological specimens, in developing post-hoc computer algorithms for enhancing and interpreting the scanned images, and in the multi-media publication of CT data sets. In my future research, I will use my connection with the U. T. CT scanning facility to obtain, interpret, and publish high-resolution CT scan data of unique or enigmatic fossil specimens.
Return to top
My third major interest is the empirically testing of the assumptions that underlie phylogenetic analysis. The reconstruction of phylogeny by cladistic parsimony analysis necessarily involves the acceptance of assumptions such as the parsimony criterion that holds that the most parsimonious hypothesis is the best. Indeed, results often vary depending on the researcher's choices among numerous alternate assumptions of character polarization and evolution. Because the true tree of evolution is not known with certainty, the merits of these options are usually debated only in a theoretical manner.
In contrast, I have explored them by digitally generating known artificial phylogenies, complete with artificial character data sets, analyzing them using alternative standard methods of phylogeny reconstruction, and assessing the degree of difference between the known true phylogeny and the reconstructed one. Artificial phylogenies, albeit unnatural, can be generated according to known assumptions of evolution, allowing the consequences of the use of the wrong assumptions in phylogeny reconstruction to be explicitly evaluated. Also, they can be generated in sufficient numbers to facilitate statistical characterization of results. To date, I have reported on the results of investigations of the consequences of using multiple primitive character states, and upon the effect of the presence and position of taxa with anomolously high rates of evolution. A study of the consequences of alternate assumptions of character state ordering is currently underway. I expect to address several other similar issues in the coming years.
The reconstruction of phylogeny ramifies into a wide range of disciplines that address the history of evolving organisms and ecological communities, including biomechanics and biostratigraphy. The ultimate aim of my research is to provide a firm foundation for all such endeavors by expanding and improving our basic knowledge of vertebrate phylogenetic history.
Return to top