UC Riverside

Before the 1960s, phylogenetic data sets consisted only of morphological data. Now, molecular data is the most common data type used to create phylogenetic trees for extant taxa. However, morphological data is still required to include extinct taxa in phylogenetic analyses because molecular data cannot be obtained from most fossils and museum specimens. The accuracy of fossil placement in phylogenetics has been questioned due to morphological data being more prone to convergence, which can confuse hypothesized evolutionary relationships. For example, in extant placental mammals, morphological and molecular data support several clades that are incompatible with one another. Additionally, molecular and morphological data disagree on the timeline of the radiation of the placental mammals. The fossil record and morphological phylogenies mainly support an origin of placental mammals after the Cretaceous-Paleogene (K-Pg) boundary while relaxed molecular clocks support an origin within the Cretaceous. In this dissertation, I expand our understanding of these issues in several ways: 1) exploring the issue of sample size in morphological data sets by using the largest morphological matrix available for placental mammals, 2) utilizing ancestral state reconstructions to determine if hypothetical predicted ancestors, as a proxy for undiscovered fossil taxa, increase the accuracy of morphological phylogenetics, 3) exploring if ordered morphological characters change hypothesized phylogenetic relationships, and 4) testing the hypothesis that supports a Paleogene origin of the placental mammals by adding a new fossil taxon into the largest morphological data set available for placental mammals. I concluded the following: 1) increased sample size did not alleviate the problems of homoplasy in this morphological matrix, 2) the use of hypothetical predicted ancestors resulted in a greater number of monophyletic pseudoextinct orders, but the interordinal positions of these groups were often inaccurately reconstructed, 3), ordering an appropriate subset of morphological characters did not greatly change hypothesized phylogenetic relationships for placental mammals when using a molecular constraint, but character ordering did have a greater effect in the absence of this constraint and 4) there is still no definitive morphological support for a Cretaceous origin of placental mammals.