Lungfish: Phylogeny and parsimony

Miles ('77) has produced the only comprehensive cladistic analysis of the Dipnoi. His phylogeny involves both definitely and uncertainly placed genera. An analysis of the data for the definitely placed genera, using a Wagner routine, showed that his phylogeny is similar to two of eight equally parsimonious cladograms. Analysis of the data for all the genera of Miles's phylogeny produced sixteen equally parsimonious cladograms. They have a wide range of topologies, none of which corresponds to Miles's phylogeny. It is shown that Miles employed some character weighting to obtain his phylogeny. A cladogram is presented based on Campbell and Barwick's ('83, '84) study of early dipnoan dentitions. The phylogenetic conclusions drawn from Campbell and Barwick's work are dependent on stratigraphic and functional data. The resulting cladogram differs from all the trees produced by the re‐analysis of Miles's complete phylogeny; it is less parsimonious and requires that the type of dentition consisting of a shagreen of small denticles that are shed during growth is an advanced dipnoan feature rather than a primitive one. Campbell and Barwick's ('83) phylogeny is favored since, for theoretical and practical reasons, parsimony is not a good criterion for choosing between the possible phylogenies for dipnoans. Campbell and Barwick's ('83) phylogenetic conclusions are based on a richer empirical base and are more consistent with observed functional trends. A new phylogeny of the Dipnoi is presented, with the basic structure being defined by the conclusions drawn from Campbell and Barwick's ('83) work, namely, that patterns of dentition characterized by expanding growth and fusion of denticles or of the development of organized tooth plates each arose only once in evolution. The addition of Miles's ('77) characters, using the principle of parsimony, gives the fine structure of the phylogeny.