INFERENCE FROM BINARY COMPARATIVE DATA

A variety of null hypotheses have been used to test for associations w hich might be construed as evidence of biologically interesting relati onships between binary character states exhibited by taxa. These model s assume that particular regions of a phylogenetic tree are independen t with respect to their probabilities of character evolution, but they differ in the regions they specify. Early analyses specified terminal taxa (usually species); more recent developments have specified all o r a subset of branches, or infinitesimally short sections of branches. Yet the central problem for comparative biologists is that branches t hroughout a phylogeny are not independent with respect to their evolut ionary possibilities. Tests which assume that they are may indeed prov ide evidence for non-random association between characters according t o the particular model of randomness used, but do not necessarily prov ide the basis for rationally inferring the existence of a biologically interesting link between the characters. In particular, they suffer f rom pseudoreplication of lineage-specific factors. By way of contrast, we resurrect in this context a different model of randomness, the ran dom assignment of treatments, which we argue provides a rationally acc eptable basis for inference. States are assumed to be randomly assigne d amongst sister taxa exhibiting different states of both variables. T his model allows that the probabilities of character evolution vary th roughout the tree, but does not require that these be specified, nor a ssumptions to be made about how evolution occurs. We illustrate this a pproach with reference to controversial associations between (i) warni ng coloration and larval gregariousness in butterflies, for which we f ind some support, and (ii) hybrid fitness and heterogamety (Haldane's Rule), for which we find no support, in contrast to Ridley's method wh ich demonstrates the opposite Rule.