Abstract:
This study is concerned with the reconstruction and interpretation of evolutionary events as evinced by the fossil record of the shallow marine gastropod Amalda. The collection of Amalda fossils is extensive both in time and space, and offers an historic record that is as complete as one could reasonably expect. This collection, which consisted of many thousands of individual specimens, all of which had good locality and stratigraphic documentation, represents over a century of collecting effort from every Tertiary locality in New Zealand. The data set was good; the task was of historical reconstruction and interpretation. Reconstruction is concerned with recreating the unique historical details of a particular phylogeny and the context in which those evens occurred. Reconstruction is not something that can be 'read from the rocks', but has to be extracted from the data by appropriate methods of analysis. Interpretation refers to the processes that are hypothesised to explain the details of the reconstruction, and strictly speaking go beyond palaeontological data. The data have to be consistent with the interpretations, but many interpretations may be possible for a particular set of data. The decision about which interpretation is to be preferred has to be made by considering evidence from living organisms. Chapters 1&2 (both submitted to Systematic Zoology) are primarily concerned with interpretation. They set out the theoretical perspective which the study aims to illustrate. Chapter 1 outlines the theoretical basis and methodological protocol of Organotaxism an organismic theory of the punctuated equilibria pattern seen in the fossil record. This "way of seeing" the fossil record accepts the reconstruction of fossil evidence favoured by punctuated equilibria, namely that species appear suddenly and remain unchanged for their duration, but does not accept the interpretation that the gaps in the fossil record are artifacts of the imperfection of the fossil record due to intermediate populations being of small size and restricted geographic range. It does not accept that intermediate populations ever existed. That is, the gaps are seen as real. The process through which form transformation occurs is not seen as a populational phenomenon, but rather as a general developmental response by individuals to conditions which break their homeostatic balance. Chapter 2 is concerned with species. Are species just a taxonomic category representing some degree of morphological (used in its widest sense) differentiation, that have no meaning beyond representing lineage (populational) continuity, or do they define themselves? The question is central to this thesis because Organotaxism argues that if species remain unchanged through time (in the sense that all fossil variability lies within statistically defined limits), then form change and speciation are linked. Species and their identification in the fossil record are important because it is not correct to interpret form change independently of speciation events. This has important implications for the interpretation of such phenomena as morphological trends. It is also important because if species are more than abstractions then any theory of evolution must adequately account for their occurrence.
Chapter 3 (accepted for publication, New Zealand Journal of Zoology) details the results of an allozymic study of four modem species of Amalda. This study confirms the specific identity of the modern representatives of these lineages that first appeared in the Pliocene. The specific distinctness is confirmed by criteria that are independent of shell characters. Fossil members of these lineages, which are identified by shell shape and characters, are also assumed to be biologically discrete. That is, these biological species can be traced back through time. The allozymic data is also analysed cladistically. This analysis confirms that the phylogeny implied by genetic data is congruent with that implied by morphological data. For a group with rather simple morphology this increases confidence in phylogenies based purely on taxonomic criteria. Chapter 4 (submitted to Paleobiology) examines the morphological variation through time of three of the biological species discussed in Chapter 3. Shell shape of the modem species is described using canonical variate analysis and the shell shapes of fossil specimens are shown to fall within the range defined by the modem species. The variation of fossil means around the modem means cannot be interpreted biologically in this case, because the effects of bias introduced through post-death sorting and fossilisation cannot be quantified. Chapter 5 (submitted to Alcheringa) describes a cladistic analysis of New Zealand modem and Tertiary Amalda. The analysis was performed and the phylogeny obtained was tested against that implied by stratigraphy. It is concluded that cladistic methodology has the potential to reconstruct the actual phylogeny of a group without recourse to any fossil evidence, provided that transformation series linking states of multistate characters are known. It is suggested that this evidence has the potential to be directly obtained from developmental experimentation, at least in the case of molluscs. A proposed revision of the taxonomy of this group is suggested. The generic classification is phylogenetic, but for reasons outlined in the chapter the subgeneric classification is based on phenetic similarity. Chapter 6 (to be submitted to Nature) is a summary of the evolutionary history of this group. It is concerned with methods of historical reconstruction. It is argued that history can only be reconstructed by methods which are ahistorical, that is by methods which work on general principles that operate independently of time. Croizat's panbiogeographic method for the analysis of biotic distributions, Hennig's method for the reconstruction of phylogenies, and the developmental understanding of form transformation, are all methods based on general principles that can be used to reconstruct and interpret any history. They are not dependent on the unique details of particular cases; indeed the particulars become intelligible only by reference to these generalities. The validity of the reconstruction of an evolutionary history depends on how well it synthesises independent evidence from many diverse fields rather than on the criterion of falsifiability. There is no complete and unequivocal record of what happened which would allow for falsification; the challenge is to develop methods to reconstruct history with increasing degrees of confidence. Chapter 7 is a final discussion concerning the philosophical basis from which this work was conceived. Chapter 3 reports the results of an electrophoretic analysis of four species of the molluscan genus Amalda. The results of this analysis confirm that four taxonomically defined species are also biologically distinct, that their often considerable morphological variability is a within-species phenomenon, and that a previously undescribed form of A. depressa, which is genetically and morphologically distinct, occurs in the Bay of Islands. The genetic data were analysed cladistically and the resulting phylogenetic classification supports that based on morphology.
In Chapter 4, ten measurements, taken from each of 700 shells of four biologically distinct species, were used to describe phenotype in multidimensional space. Canonical discriminant analysis was performed on the data and a set of allocatory rules was derived. These allocatory rules were then applied to 644 fossil specimens of three of these biological species. Fossil individuals are shown to occupy the appropriate phenotypic space as defined by their modern descendants. The variation of fossil sample means about the modern means is illustrated. A discussion of the biological significance of these differences in means is framed in terms of the various sources of bias in estimating sample statistics, and it is concluded that for this example it is not possible to resolve the question of whether these differences are due to biological or physical processes. A phylogenetic analysis of Recent and fossil species of the genus Amalda was undertaken in Chapter 5. The resulting cladogram is compared with the phylogeny implied by stratigraphic evidence. The two phylogenies are in broad agreement, and both suggest that the subgenus Alocospira should be elevated to generic rank. The status of the subgenus Gracilispira is uncertain on both cladistic and stratigraphic grounds. Of the two remaining subgenera, only Baryspira is unequivocally monophyletic. Gemaspira and Baryspira are sister taxa in the cladistic analysis. However, stratigraphic evidence strongly suggests that Gemaspira is a paraphyletic group. The implied-sister-group relationship is interpreted as resulting from coding procedures which assumed no transformation series between states of a multistate character. Clearly, stratigraphic evidence implies that the synapomorphies linking Baryspira spp. are transformations from (rather than being independent of) the synapomorphies linking Gemaspira spp. This underlines the importance of deducing the correct transformation series for multistate characters in any attempt to reconstruct phylogenies. In this case the transformation series implied are counter-intuitive. Finally a pragmatic rather than a strictly phylogenetic revision of the taxonomy is proposed and included as Appendix A. In chapter 6 the panbiogeographic method of Leon Croizat is discussed. Track analysis is used in the reconstruction of the context in which evolutionary processes operate, evolutionary history being seen as resulting from an interaction between process and context. A reconstruction of the evolutionary history of the gastropod genus Amalda is used as an example to illustrate this view. Finally, the problems associated with historical reconstruction are discussed, and it is argued that the methods employed in reconstruction should be a historical.