Theoretical and Procedural Aspects of Numerical Phyletics

Abstract

Eades, D. C. (Dept. Zool., Univ. Illinois, Urbana), 1970. Theoretical and Procedural Aspects of Numerical Phyletics. Syst. Zool., 19:142-171.-There is a possible common ground between conventional and numerical taxonomy involving computer analysis of many characters. It would utilize evolutionary concepts whenever possible to modify an otherwise phenetic classification. New algorithms are described for character analysis and for cluster analysis. Many combinations of computer analysis were tested on four sets of taxonomic data. Results were judged by the effectiveness in showing greater similarity between individuals of the same group as defined by reliable evidence not included in the data sets. Statistically significant improvements were obtained by inclusion of computer procedures for adjusting correlated characters and for weighting characters. A variety of numerical procedures are discussed [Quantitative systematics; character analysis; phenetics; phyletics]. The split between conventional and numerical schools of taxonomic thought is too well known to require elaboration here. This polarization has been most troublesome for me because I have seen much virtue on both sides of the fence. Numerical taxonomy has a great deal to offer if viewed as an improved method for certain operations in conventional taxonomy, but not if viewed as a substitute for the whole of conventional taxonomy. My own prediction is that the future development of taxonomy will see the incorporation of the most useful taxometric methods into evolutionary taxonomy but with a rejection of the backward ideology underlying phenetics, Mayr (1965:94). With this in mind, I have attempted to think in a theoretical framework assembled from the best of both methods. The result is an approach that could be called numerical phyletics. The enthusiastic reception of a talk by Throckmorton (1966) convinced me that many taxonomists are thinking along similar lines, and the time is ripe for someone to assemble and present such philosophy. Rogers, Fleming, and Estabrook (1967) and Farris (1967) also provide important contributions in this di-