Thesis (Ph. D.)--University of Chicago, Committee on Evolutionary Biology, June 2001. / Includes bibliographical references. Also available on the Internet.

Thesis (Ph. D.)--University of Chicago, Committee on Evolutionary Biology, August 2003. / Includes bibliographical references. Also available on the Internet.

Thesis (Ph.D.) -- University of Adelaide, Dept. of Geology, 1963. / Typewritten.

Jawed vertebrates (gnathostomes) account for over 99% of living vertebrate diversity, with origins that stretch back nearly half a billion years, and comprise two groups: Osteichthyes (fishes and land-dwelling vertebrates) and Chondrichthyes (sharks, rays and chimaeras). Osteichthyans are the dominant clade, with at least 60,000 species approximately evenly divided between two clades: the Actinopterygii and the Sarcopterygii. However, our understanding of early osteichthyan evolution is skewed in favour of sarcopterygians, leaving the origin of nearly half of all vertebrate diversity critically understudied. Furthermore, recent upheavals in the early gnathostome tree have destabilised relationships amongst fossil taxa and eroded our understanding of primitive anatomical conditions of key groups. Central to understanding early gnathostome evolution is the braincase, an anatomically complex structure that provides a wealth of morphological characters. However, braincases rarely fossilise, and their position inside the skull makes them difficult to attain. X-ray tomography allows a comprehensive description of the internal and external anatomy of fossils, including the braincase. This thesis sets out to target phylogenetically pivotal taxa and incorporate new anatomical data in building up a picture of character evolution in early jawed vertebrates. In particular, I target the gnathostome stem, describing a new taxon that helps bridge the morphological gap between placoderms and crown gnathostomes, allowing a more comprehensive understanding of both dermal and endoskeletal evolution. I also focus on early actinopterygians, describing the endoskeleton of the first members of the group in order to understand primitive anatomical conditions. I then investigate actinopterygian braincase anatomy in the context of a revised phylogenetic analysis, illuminating the early evolution of the actinopterygians. Finally, I present a synthetic review of braincase anatomy across the early gnathostome tree. These results provide a more accurate picture of braincase evolution across gnathostomes and actinopterygians, clarifying our understanding of their evolution while revealing new information about when key innovations arose in the brains of the very first ray-finned fishes.

Estuaries are dynamic on many spatial and temporal scales. Distinguishing effects of unpredictable events from cyclical patterns can be challenging but important to predict the influence of press and pulse drivers in the face of climate change. Diatom assemblages respond rapidly to changing environmental conditions and characterize change on multiple time scales. The goals of this research were to 1) characterize diatom assemblages in the Charlotte Harbor watershed, their relationships with water quality parameters, and how they change in response to climate; and 2) use assemblages in sediment cores to interpret past climate changes and tropical cyclone activity. Diatom assemblages had strong relationships with salinity and nutrient concentrations, and a quantitative tool was developed to reconstruct past values of these parameters. Assemblages were stable between the wet and dry seasons, and were more similar to each other than to assemblages found following a tropical cyclone. Diatom assemblages following the storm showed a decrease in dispersion among sites, a pattern that was consistent on different spatial scales but may depend on hydrological management regimes. Analysis of sediment cores from two southwest Florida estuaries showed that locally-developed diatom inference models can be applied with caution on regional scales. Large-scale climate changes were suggested by environmental reconstructions in both estuaries, but with slightly different temporal pacing. Estimates of salinity and nutrient concentrations suggested that major hydrological patterns changed at approximately 5.5 and 3 kyrs BP. A highly temporally-resolved sediment core from Charlotte Harbor provided evidence for past changes that correspond with known climate records. Diatom assemblages had significant relationships with the three-year average index values of the Atlantic Multidecadal Oscillation and the El Niño Southern Oscillation. Assemblages that predicted low salinity and high total phosphorus also had the lowest dispersion and corresponded with some major storms in the known record, which together may provide a proxy for evidence of severe storms in the paleoecological record.

North American black bears (U. americanus) and brown bears (U. arctos) can be difficult to distinguish in the fossil record due to similar dental and skeletal morphologies. Challenges identifying ursid material from Oregon Caves National Monument (ORCA) called for an accurate tool to distinguish the species. Ursid teeth have a high degree of variability and morphological features are not always diagnostic. This study utilized a large database of lower tooth lengths (p4, m1, m2, and m3) and ratios (p4/m1, m2/m1, m3/m1, p4/m3, m2/m3) in an attempt to differentiate U. americanus and U. arctos in North America. Further, this project examined how these linear measurements differ in response to ecoregion, latitude, and climate. Analysis of variance (ANOVA) found significant differences between U. americanus and U. arctos from across North America for every variable studied. Stepwise discriminant analyses (DA) found lengths separated species better than ratios with 99.1% correct classification versus 77.5% correct classification for ratios. When sexes were analyzed, ANOVA only found significant differences for lengths while DA found lengths and ratios could not accurately distinguish between sexes; only 72.1% of sexes were classified correctly while utilizing lengths and 61% for ratios. Seventeen previously identified fossil specimens from across North America, in addition to the ORCA specimen, demonstrated the utility of this study, confirming several identifications and rejecting others, proposing the need for new designations.

A detailed morphological description of a proboscidean manus from the Gray Fossil Site (GFS), Gray, Tennessee is provided. Manus elements from an American mastodon (Mammut americanum), a Britt’s shovel-tusker (Amebelodon britti), an undescribed small gomphothere species, and a Columbian mammoth (Mammuthus columbi) are used for comparisons. Linear measurements indicate proportional differences between the GFS mastodon and other proboscidean taxa ranging from the Hemphillian to Rancholabrean land mammal ages. Possible pathologies are also described. The purpose of this study is to determine how the GFS mastodon differs in manus morphology and locomotion from different proboscidean taxa, including size, environmental, evolutionary, and taxonomic factors. Morphological differences are more pronounced in carpals and metacarpals in regard to shape, size, and orientation of articular surfaces but are not statistically compared due to the small sample size.

Mastodons (family Mammutidae) are widespread and abundant in of Pleistocene faunas across North America, exhibiting considerable variation in morphology over their broad geographic and temporal range. Mandibular tusks are a notably variable feature among mastodons; these tusks vary in size and shape, and many mastodons lack them entirely, or possess only one. Patterns in mandibular tusk distribution could potentially indicate important selective controls on different populations. Various hypotheses have been proposed to explain the distribution of mastodon mandibular tusks, attributing their presence to geographic, temporal, and sexual variation, but there have been no analyses investigating whether there is statistical support for patterns of mandibular tusk distribution. We analyze a dataset of more than 100 mastodons from throughout North America. Mandibles were coded for the presence/absence of mandibular tusks. These data were used to test whether tusk presence was related to geographic or chronological patterns. The most statistically-significant differences were recorded between interglacial- and late glacial-interval mastodons. Interglacial mastodons had mean differences of 0.958 and 0.827 from last glacial maximum and post-last glacial maximum mastodons, respectively, with essentially 0 p-values for both. Though this study is preliminary and uses simple statistical tests, it shows that patterns do exist in the presence of mandibular tusks among mastodons, and that this variability warrants further investigation, particularly with regards to differences in paleoenvironment and paleoecology between interglacial and glacial mastodons that may be responsible for their apparent differences in mandibular tusk presence.

Within Pleistocene cave deposits, snake fossils tend to be relatively common and generally occur as isolated vertebrae. However, the specific, and sometimes generic, identification of isolated snake vertebrae is often hindered by the significant amount of intra- and interspecific variation along the precloacal vertebral column, a subject that has largely been neglected in many fossil snake identifications. Identifications are typically based on the vertebrae of disarticulated modern specimens with preference given to mid-trunk vertebrae. This study utilizes 2-D geometric morphometrics to determine the extent of intervertebral variation along the precloacal vertebral column of North American pit vipers of the genera Crotalus (rattlesnakes) and Agkistrodon (copperheads and moccasins), two closely related genera geographically sympatric in northeast Tennessee and much of the eastern United States with similar vertebral morphologies. The focus of this study is to determine the need for identifying the morphological regionalization of the precloacal vertebral column and determining the regional position of isolated vertebra prior to identification. Using one individual from each genus, every third vertebra was chosen and analyzed in anterior view using geometric morphometrics and relative warp analyses. A discriminant function analysis was then performed to distinguish between the two genera. Viperid fossils from Hickory Tree Cave in northeast Tennessee underwent the same geometric morphometric and discriminant function analyses as a means of identification. It is expected that the relative warp analyses will show some morphological regionalization of the precloacal vertebral column, but may prove to not be significant enough for use in the identification of isolated vertebrae. In that case, comparison to mid-trunk vertebrae for identification may prove accurate. This study also shows the need for more modern skeletal specimens in herpetological collections and the need for preserving the vertebral order of those specimens.

A century ago, two large American mastodon Mammut americanum specimens from coastal South Carolina were mounted for display at Amherst College. Their robust build, broad and pentalophodont third molars, and well-developed mandibular tusks were noted as unusual. Here, these specimens are redescribed; three additional mastodon mandibles from the same region of South Carolina are also described. Though collection information on these specimens is sparse, they appear to span >150 ka in the Middle to Late Pleistocene, and probably include the first report of Mammut from the earliest-Rancholabrean Ten Mile Hill Formation. Some “unusual” features — moderately to extremely broad third molars, longitudinally-wrinkled enamel, very robust mandibles, large mandibular tusks — are present in most or all specimens, and may characterize regional mastodons. Description of these mastodons provides a useful starting point for more-thorough evaluation of the species in the Southeastern United States.