Determining Primitive Character States for Phylogenetic Reconstruction

Publisher Website
Google Scholar
Abstract
The reconstruction of evolutionary trees by either conventional or cladistic methods depends upon the determination of evolutionary directionality (or polarity) of character states. Many criteria have been used to determine primitive states of characters, but these sometimes have duplicated the same concept and sometimes have not been based on evolutionary reasoning. Some apply only to specific taxa. We recognize two broad categories of criteria of primitiveness: first-level criteria, in which nothing is known about any primitive condition; and second-level criteria, in which at least one primitive condition already is known. We recognize nine criteria, the first six being first-level and the last three being second-level criteria: (1) fossil evidence; (2) common is primitive; (3) co-occurrence of primitive states; (4) earliest ontogenetic state; (5) minor abnormalities of organogenesis; (6) vestigial organs; (7) association; (8) correlation; and (9) group trends. The many specific criteria of primitiveness (e.g., hypogyny preceding epigyny in the angiosperms) are regarded as tendency statements of restricted utility. The only absolute criterion of primitiveness is that the oldest character state is primitive, which can never be determined with certainty because we lack the true phylogeny. The criteria, therefore, can only help approximate real phylogenies. Some criteria are more generally useful than others, but their efficacy depends to some extent on the nature of characters and character states within individual taxa.