Statistics and Systematics

See allHide authors and affiliations

Science  28 Apr 2000:
Vol. 288, Issue 5466, pp. 577
DOI: 10.1126/science.288.5466.577f

Biologists since Linnaeus have routinely used the species as the fundamental unit of classification. Despite the many recent advances of phylogenetic and systematic methodology, the vast majority of species are delimited on the basis of a very small number of preserved specimens that may not be representative of the true range of variation. Obtaining the ideal sample sizes or conducting detailed morphological and molecular analyses on more than a tiny proportion of the world's species is impractical by any yardstick. Can statistical methods help?

Wiens and Servedio present a new statistical model that allows systematists using traditional character-based methods to determine whether they have sampled enough characters (usually morphological) or individuals to diagnose species differences with confidence. The model promises hope for more standardized approaches to traditional taxonomic methods, still used for the bulk of systematic work.—AMS

Proc. R. Soc. Lond. B267, 631 (2000).

Navigate This Article