© 1981 by The Society for Integrative and Comparative Biology
Functional-Adaptive Analysis in Evolutionary Classification1
Department of Biological Sciences, Columbia University New York, New York 10027
SYNOPSIS. Higher level systematic study can be divided into two parts, namely: (a) the formulation of classificatory and of phylogenetic hypotheses about groups and their testing against taxonomic properties of characters; and (b) the formulation of hypotheses about taxonomic properties of characters and their testing against empirical observations of these features. Thesecond step—character analysis—is most important because it constitutes the actualobjective testing within systematic study and provides the basis for one's conviction in the validity of a particular classification and/or phylogeny. Yet insufficient attention has been given to the methods used in the character analysis phase of systematics.
Classifications and phylogenies are historical narrative explanations of the past evolutionary history of organisms. As such they must be based on and tested against the underlying theory of organic evolution. Classifications are not used to test the validity of mechanisms of evolutionary change; rather the reverse is true. For the construction of a classification or of a phylogeny, it matters absolutely how the organisms evolved—the classification and phylogeny depend completely on the accepted mechanisms of evolutionary change, not just on the hypothesis that organic evolution had occurred. Thus, the methods used in the character analysis phase of systematic study must be based on accepted mechanisms of evolutionary change. If the evolutionary change is adaptive, then functional and adaptive analyses form an integral part of the character analysis.
Several commonly used methods of character analysis, most importantly that of outgroup comparison, are invalid because they are based on simple circularity, etc. The valid methods available for testing taxonomic properties, e.g., homology, apomorphy, etc., of characters are poortests in that they have low resolving power, i.e., poor ability to distinguish between correctand incorrect answers to individual tests. Additional methods must be used to establish a degree of confidence in particular answers of valid tests of low resolving power. A set of valid methods are outlined for testing character hypotheses about homology, apomorphy—plesiomorphy and synapomorphy, and for establishing the degree of confidence in the results of individual tests. Functional-adaptive analyses are a necessary and essential part of the valid methodsused to test and/or establish the degree of confidence in individual results for all hypotheses about taxonomic properties of characters.