James S. Crampton, GNS Science (New Zealand)
Simon Hills, Massey University (New Zealand)
Mark Fenwick, Te Papa Tongarewa (New Zealand)
Mary Morgan-Richards, Massey University (New Zealand)
Bruce Marshall, Te Papa Tongarewa (New Zealand)
Alan Beu, GNS Science (New Zealand)
Austin Hendy, Yale University (United States)
The definition of the species concept and the operational recognition of species are highly contentious problems; they are topics upon which "more paper has been consumed... than any other in evolutionary and systematic biology" (Wiley, 1978, Systematic Zoology, 27:17). These problems are acute when considering species in the fossil record, which may, or may not, be equivalent to species living today. The issue bears upon patterns of diversification, extinction and evolution as inferred from the fossil record: can these historical patterns be interpreted in terms of biological processes observed in living communities of plants and animals?
To explore this question, we have obtained molecular phylogenetic and morphological data from two very different groups of living molluscs from New Zealand: marine volutid gastropods and freshwater unionid bivalves. The volutid dataset is based on 11 nominal species living at subtidal to outer shelf depths. Approximately 7500 base pairs of mitochondrial DNA have been sequenced from these species and phylogenies were reconstructed using maximum likelihood and Bayesian methods. To study morphological variation within and between volutid taxa, we used two approaches: Geometric morphometric description of five landmarks and six semi-landmarks around the shell, and Fourier shape analysis of the aperture.
The unionid dataset is based on three nominal species found in lakes and rivers throughout New Zealand. The molecular analysis is based on a 500 base pairs fragment of mitochondrial DNA with an independent test using nuclear DNA. Morphological variation within the unionids has been described using Fourier shape analysis of the lateral projection of the shell.
Comparisons of these molecular and morphometric data indicate that, in general, morphological data are able to discriminate molecular species with some success, even in morphologically "challenging" groups such as the unionids. This encouraging result lends confidence to some taxic paleobiological analyses of diversity, extinction and evolution based on fossil morphospecies. In contrast, however, our results suggest that reconstruction of phylogenies based solely on morphology is more difficult, especially in groups that lack abundant, discrete, hard-part morphological characters. This problem may be addressed, in part, by incorporation of stratophenetic interpretations and constraints in phylogenetic analyses of fossil groups, at least in well sampled taxa such as New Zealand Neogene molluscs that have average per-stage sampling probabilities of between 40% and 80%.