Comatulid Crinoids in the Fossil Record: Methods and Results for the Extremely Imperfect.

Abstract

Paleontologists' attempts to understand patterns of evolutionary change have always been influenced by biases in both the fossil record and collection/description. This dissertation has two basic goals: to better understand biases, and to better understand patterns of evolutionary change. Comatulid crinoids, the most abundant modern crinoids, have a depauperate described fossil record, with an order of magnitude lower generic diversity reported from any stage than that described from the modern ocean. Two methods are used to address potential bias in the described record: morphological analysis of comatulid centrodorsals to test the role of differential material for modern and paleontological taxonomic descriptions, and a capture-mark-recapture (CMR) method to understand regional and temporal difference in detection rate. To address patterns of morphological change, we examine the disparity of comatulid centrodorsals, both within species, and for the whole group through geologic time. Investigation of intraspecific variation of comatulid centrodorsals finds no evidence of lumping bias in fossil taxonomic descriptions. However, analysis using CMR found substantial differences in detection rate both regionally and temporally. Extremely low detection rates, especially in the non-European Cenozoic, mean that there are substantial biases that hide a significant increase in comatulid diversity over the Cenozoic. Collection efforts should be undertaken to better understand whether the bias is in the fossil record, or in efforts to collect and describe comatulid material from the Cenozoic. Previous efforts to describe the pattern of morphological change in crinoids have shown a pattern of rapid expansion into morphospace followed by stasis. Such a pattern can be explained by a decrease in rates of evolution, or as constraints preventing further dispersion into morphospace. These explanations have differing signatures on subclade disparity and homoplasy. A lowering rate of change scenario should mean homoplasy is rare and subclade disparity to be low, while the constraints scenario should result in homoplasy being more common and subclade disparity being high. We found evidence for common homoplasy and high subclade disparity for the comatulids. This supports constraints as the explanation for the pattern of early expansion into morphospace followed by stasis.