DETERMINING THE MOST ROBUST METHOD OF MEASURING DISAGGREGATED FOSSIL MATERIAL: THE COLONIAL CONUNDRUM

Determining the most robust method of counting fossils is necessary to achieve consistency and accuracy in paleoecological research. Modern studies often measure biomass of taxa directly. However, due to taphonomic processes, proxies for biomass must be used when examining paleocommunities. Often, abundance counts are used to tally disaggregated material, but this method leads to problems equating solitary brachiopods and mollusks with colonial (bryozoans, corals) and easily disarticulated taxa (crinoids). Here, we evaluate the method of equating 1 cm of bryozoan, crinoid, and colonial coral (BrCrCo) to one individual brachiopod or mollusk as a means of equally including all taxa in abundance analyses.

Thirty-three 3 L bulk sediment samples were collected from the Pennsylvanian Finis Shale of Texas and disaggregated. Fossils were identified to genus and counted. Two different taxon-sample matrices were created; the first included the total number of each solitary taxon, with BrCrCo counted as one if present or zero if absent (P/A) for a total of 4911 individuals from 73 genera. The second included the total number of each solitary taxon, with BrCrCo included by adding 1 individual per 1 cm of length (maximum dimension) for a total of 5371 individuals. Using these two methods, three categorizations (all taxa, three dominant brachiopods removed, and only BrCrCo) were produced for a total of six data matrices. Polar Ordinations were run for each of the six matrices using Sorenson and relative Sorenson distance measurements.

The abundance of all taxa and the abundance of brachiopods and mollusks with P/A of BrCrCo ordinations show the same three stratigraphically- and dominant-brachiopod-driven clusters. With the three dominant taxa removed from these two matrices, ordinations reveal the same three clusters, but they are less defined. Using BrCrCo only matrices, there is a clear change in pattern and grouping between the length and P/A ordinations. These comparisons demonstrate: 1) when dealing with brachiopod or mollusk dominated samples, the method of including BrCrCo, or even whether they are included, is not as important; and 2) when dealing with taxonomically even or bryozoan- and crinoid-dominated communities, the method of tallying these taxa can have a major effect on the resulting community signal.