THE ACCURACY OF THE MAXIMUM PARSIMONY METHOD FOR PHYLOGENY RECONSTRUCTION WITH MORPHOLOGICAL CHARACTERS

The accuracy of maximum parsimony for phylogeny reconstruction using m orphological data has never been adequately tested. In this study 128, 000 known phylogenies of eight species and their associated discrete u nordered character state data sets were generated by computer simulati on. In order to mimic morphological data, values of the following seve n simulation parameters were varied: 1) number of characters: 2) maxim um number of character states per character: 3) number of character st ate changes per speciation event, and 4) the probabilities of parallel isms, reversals, phyletic changes, and extinctions. Character state da ta for the known simulated phylogenies were input into the program Hen nig86, which found all equally parsimonious trees. The accuracy of max imum parsimony was judged primarily by the percent of time the strict consensus tree was found to be identical to the true tree. Parallelism s, reversals, phyletic changes, and extinctions all significantly decr ease maximum parsimony's ability to find the true phylogenetic tree. I n addition, maximum parsimony is less accurate when there are fewer ch aracters, more character states per character, and more character stat e changes per speciation event. For the parameter values considered, m aximum parsimony did not do particularly well at finding the true phyl ogenetic tree. For 24.2% of the simulations, the chance that maximum p arsimony found the true tree was less 25%; for 60.2% of the simulation s, it was less than 50%; and for 85.2% of the simulations, it was less than 75%. In order to determine the relevance of these results to rea l data sets, values of the consistency index from the simulated phylog enies were compared to values reported in published plant morphologica l studies. For simulated data sets having values of the consistency in dex in the range reported in the literature, maximum parsimony found t he true phylogenetic tree only from 0.7% to 27.8% of the time (mean = 10.3%). I advise caution, therefore, in using phylogenies reconstructe d from morphological data to make inferences about biological processe s or patterns that rely on knowledge of the true phylogeny. Additional ly, I raise the question whether the great emphasis in systematics on phylogeny reconstruction using maximum parsimony and other cladistic m ethods is misplaced.