REVIEW OF METHODOLOGICAL PROBLEMS OF COMPUTER CLADISTICS EXEMPLIFIED WITH A CASE-STUDY ON ISOPOD PHYLOGENY (CRUSTACEA, ISOPODA)

Using a computerized phylogenetic analysis of the Isopoda (Crustacea: Peracarida) as source of typical errors and misunderstandings, problem s that may occur in computer cladistics are reviewed. It is concluded that in addition to the errors that are possible in a conventional Hen nigian analysis some specific methodological problems exist in compute r cladistics. It is recommended that the OTU be replaced by the ground pattern concept. Tree statistics are not useful for comparing differen t competing hypotheses. Arguments ought to concentrate on the hypothet ico-deductive steps of the analysis, i.e. on character analysis. The u se of computers does not add objectivity to character analysis. Single outgroup taxa should not be used in assessing the character states of ingroups.Concerning isopod phylogeny, it is argued here that the tail fan of the Isopoda can probably be derived from the eumalacostracan g round pattern and did not evolve de novo within the Isopoda.