CODING POLYMORPHIC DATA - EXAMPLES FROM ALLOZYMES AND ONTOGENY

Limitations for using polymorphic characters in phylogenetic analysis include the lack of computer algorithms that allow input of complex branched or reticulate character state trees and the inability to use polymorphisms as data in tree building and character optimization. The step matrix option in PAUP version 3.0 and MacClade versions 2.1 and 3.0 provides a solution to these problems, allowing all homology information among character states to be utilized while retaining a biologically realistic definition of characters. For genetic data, this means that the locus can be coded as the character and allelic combinations can be coded as the character states. For ontogenetic data, this means that the entire ontogenetic sequence can be coded as the character and combinations of observed conditions within ontogenies can be coded as the character states. The unity of the locus or character ontogeny is preserved, and homology information from shared electrophoretic or ontogenetic conditions is not lost. Attention has not been called to the loss of homology information that accompanies coding the locus or an entire ontogeny as the character, possibly because of the lack of a coding option through which these data could be retained. The step matrix option allows any and all assumptions of transformation (order) among character states while retaining the unity of the locus or a whole ontogeny as the character and can be utilized for coding these and other polymorphic characters for phylogenetic analysis.