Evolution in a marine gastropod : rocks, clocks, DNA and diversity : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Evolutionary Biology

Abstract

Comprehensive integration of paleontological and molecular data remains a sought-after goal of evolutionary research. This thesis presents a dataset unlike any previously studied to document changes over time in the evolutionary history of the New Zealand marine mollusc genus Alcithoe. In order to study evolutionary relationships in the Alcithoe, DNA sequence of approximately 8Kb of mitochondrial DNA was generated using universal and newly developed PCR primers. The gene composition of the resulting sequences has been thoroughly analysed, using a novel splits-based approach, to gain a clear understanding of the underlying phylogenetic signals in the data. Refinement of the phylogeny was achieved by considering subsets of both the taxa and genes. Taking these analyses into account the combined a robust phylogeny for the Alcithoe is presented for use in subsequent analyses. The Alcithoe genus includes species that are exemplars of the problem of correctly identifying species by morphological traits, in both the living and extinct taxa. Taxonomic assignments were explored in a population level analysis of the highly morphologically variable species A. wilsonae. Analyses revealed that the various recognised forms of A. wilsonae are genetically indistinguishable and that the previously recognised species A. knoxi is a synonym of A. wilsonae. This result has significant implications for the interpretation of the paleontological data, as A. knoxi specimens are known from the Tongaporutuan stage (10.92 – 6.5 Ma) of the New Zealand geological timescale. Therefore, this finding also has implications on the assignment of calibration data in molecular clock analysis. To ensure accurate estimation of divergence times and rates of molecular evolution, extensive explorations of parameter space in molecular-clock analyses were carried out. These analyses identified the most appropriate models and calibration settings for Alcithoe the dataset. The fossil data used to calibrate this analysis is amongst the most robust applied to molecular clock analyses to date. Statistical sampling uncertainty derived from the paleontological data was included in the calculation of prior distributions. Divergence dates inferred for the extant Alcithoe are largely consistent with the fossil record. However, the root of the tree was consistently inferred to be younger then expected. Rates of evolution in the species of Alcithoe included in this analysis are broadly consistent. However, some small rate differences are observed in some branches, for example, Alcithoe fusus appears to have a faster rate then the rest of the genus. This rate increase is the likely cause of topological inconsistencies observed for four closely related taxa, including A. fusus, and indicates that slight rate differences can cause phylogenetic instability when small genetic distances are involved. Direct comparison of diversification rates between the molecular and paleontological data for the Alcithoe illustrated that modern Alcithoe species have origins that are around 13 millions years younger then the oldest known Alcithoe fossils. The suggestion that A. fusus is descended from a series of fossil Leporemax species is directly contradicted by the molecular tree. In light of the molecular evidence this result highlights the problem of morphological convergence in the interpretation of fossil Alcithoe species. Comparison of the molecular and paleontological datasets was difficult for absolute speciation and extinction rates, as errors inherent to each dataset led to disparate estimates. For example, the fossil record clearly fails to record most recent speciation events observed in the molecular phylogeny, but the molecular data cannot sufficiently account for the amount of extinction evident in the fossil record. It is clear that the assumption of a constant and equal probability of speciation and extinction for all lineages is violated in the Alcithoe. However, the general long-term trends estimated for both datasets are concordant, and demonstrate an increase in both speciation and extinction rates over the Cenozoic era. The research described in this thesis represents significant progress toward the goal of more thorough integration of molecular and paleontological in the study of evolution. I have shown that reconciliation of molecular and paleontological data is not only possible, but can substantially improve the resulting interpretation of evolution. This study is the broadest analysis of the evolution in a single genus using combined molecular and paleontological data that the author is currently aware of. It illustrates the advantage of having quality paleontological data to compare to emerging molecular data, and how the molecular data can further inform the paleontological data. Furthermore, it adds support to the shift in perspective from an adversarial to a complementary approach to the consideration of molecular and paleontological data. This thesis is a comprehensive first step in the synthesis of molecular and paleontological data in the study of evolution of the New Zealand mollusc fauna, and alludes to many promising avenues for future study.