Date of Award


Document Type


Degree Name

Master of Science (MS)



First Advisor

J. Hartman


Species are the basic units of life relative to taxonomy, and a practical definition of species is imperative to paleontology in order to promote reproducible, unambiguous criteria for accurately interpreting past species relationships. The need for a well-defined rubric for identifying species is evident in the freshwater mussel assemblage (Family Unionidae) in the Hell Creek Formation. The Hell Creek Formation unionid species were first described by Whitfield in 1903 and 1907. His descriptions were often based on differences of a single trait. In 197 6, Russell reevaluated the original species descriptions, I but still based the classification of species on only a few qualitative morphologic differences. These previous species diagnoses did not incorporate phenetics (based on morphologic similarity) or cladistics (based on evolutionary relationships). As a result, the assignment of morphologically similar specimens to previously defined species is sometimes problematic. Additionally, without more knowledge, certain species may not clearly be considered distinct.

The goal of this study is to reevaluate four "sister" species sets in order to assess their validity as species using an analytical approach instead of relying on the traditional qualitative approach. Species diagnosis on the basis of a single trait or very few traits, as employed in the previous species descriptions, is not likely adequate to support the classification as distinct species or to show the relationships among species. To test the validity of species in the Hell Creek Formation, an extensive database of quantitative and qualitative character traits was constructed to be applied to various quantitative models. The morphological distinction of each species was ascertained by using discriminant and cluster analyses. The species distinction based on evolutionary relationships was also tested using cladistic methods.

Comparing morphologic differences without quantitative analysis, all eight species display significant morphologic differences, implying that they are discrete species. Discriminant analysis distinguished all eight species on the basis of quantitative traits; however, many sister species were indistinguishable solely on the basis of qualitative traits. Cluster analysis displayed apparent structure in the dendrogram that loosely applied to eight distinct species, but in some cases, there was substantial intermixing of specimens otherwise assigned to the recognized species observed in the clusters. Cladistics displayed minimal resolution in the phylogenetic relationships between species and was not able to distinguish between sisters species on the basis of the criteria given. The results herein suggest that each species is distinct on the basis of their morphology as shown by species comparison with and without quantitative models. Interpretation of the evolutionary relationships among the species shows that each sister species set is closely related, but within the sister sets, the species are phylogenetically indistinguishable.

Bingle (24176 kB)

Included in

Geology Commons