A PHYLOGENETIC HISTORY OF THE ORDOVICIAN DIPLOGRAPTACEA: ASSESSING THE CONGRUENCE OF STRATIGRAPHIC AND CLADISTIC DATA IN GRAPTOLITE EVOLUTIONARY PATTERNS

Cladograms are built to answer questions about the evolutionary relationships among species. They also predict the order in which taxa evolve and hence the order of appearance for fossils in the paleontologic record. When the phylogenetic order of appearance conflicts with the actual stratigraphic record, paleontologists must choose between the stratigraphic data and the phylogenetic hypothesis based on morphologic data. We examined the inconsistencies between stratigraphic and cladistic data in the phylogenetic history of the Diplograptacea. In order to study the evolutionary relationships within the Diplograptacea we conducted a cladistic analysis on 87 Ordovician species using PAUP 4.0. The strict consensus of 24 most parsimonious trees contained two sister group relationships among sub-clades that conflicted strongly with the stratigraphic record of diplograptids and contradicted current ideas regarding graptolite evolutionary history. For example, the Climacograptinae was projected as the sister group to the normalograptids, two groups that most graptolite workers think have independent ancestries. We then used MacClade to manually rearranged the tree based on Mitchell's (1987) work on astogenetic conservatism in graptoloids and to better fit the stratigraphic record. We evaluated the congruence of the stratigraphic and morphologic data in the two trees using stratocladistics and Norell's (1993) method of ghost lineage analysis. Stratigraphic debt was calculated using the Australasian stages. The strict consensus tree had a total tree length of 979.8 steps of which 174 were attributable to the coded stratigraphic character. The manually rearranged tree had a total length of 982.9 steps with 157 attributable to stratigraphic debt. Thus, although our favored tree saved 17 stratigraphic steps, the stratocladistic method does not produce a shorter tree. We also examined the apportionment of stratigraphic debt between ghost lineages (range extension) and ghost taxa (putative ancestors). The strict consensus tree contained 23 ghost taxa that accounted for 37% of the stratigraphic misfit. Conversely, our rearranged tree contained 25 ghost taxa that accounted for 42% of the stratigraphic debt. These results indicate that stratigraphy alone does not justify rearranging the tree.