MORPHOLOGICAL CHARACTERS, POLYTOMIES, AND HOMOPLASY INDEXES - RESPONSE

This commentary responds to Wrens and Hillis' critique of a simulation study I conducted that examined the ability of maximum parsimony to f ind the true tree when morphological characters were used in the phylo genetic analysis. It rebuts all the main criticisms of those authors, and in the process: 1) emphasizes significant differences between morp hological characters and other types of discrete character data; 2) ex plains why my study was necessary and why it still remains the only st udy to dale that has examined the unconfounded effects of seven import ant parameters on the accuracy of maximum parsimony with morphological characters; 3) recognizes and explains the significant and relevant d ifferences between the evolutionary trees that are generated by nature and the corresponding phylogenetic trees that are generated by humans (and cladistic computer programs); 4) properly interprets and stresse s the importance of and the necessity for polytomies in phylogenetic t rees; 5) examines properties of several indices of homoplasy, includin g the consistency index and a new index that I considered for use in m y simulation study, the restricted homoplasy index; 6) exposes the log ical flaws in some statistical methods that are used for assigning sup port or confidence to nodes or branches of a phylogenetic tree, and 7) points out several instances where Wiens and Hillis attribute to me s tatements I never made. I stand by every point I made in the original paper.