SPECIES CONCEPTS - REPLY

Ernst Mayr has recently analysed the flora of Concord, Massachusetts, and concluded that over 90 % of the native species of plants are well described by his ''biological species concept'', which defines species as interbreeding communities that are reproductively isolated from one another. However, many of the species he considers to be good ''biological species'' either show little or no outcrossing or else regularly exchange genes with other species. Mayr has assessed the species of Concord, not by the two reproductive properties in his definition, but by the properties in the traditional morpho-geographical species concept, that is, by the existence of discontinuities in the natural interpopulational variation of morphological or other markers. The flaws Mayr attributes to morphological species concepts - that taxonomists assess variation in individuals without reference to the underlying population-genetic variation, that taxonomic species are based on overall phenetic distance rather than phenetic discontinuities, and that they grossly overestimate the frequency of natural hybridization in plants - are not part of any modem species concept, and his refutations of critics of his concepts do not adequately address the problems that exist in his definition. He is quite correct in concluding that the nature of species is similar in animals and plants, since uniparental reproduction and interspecific gene exchange are also widespread in animals. To the extent that Mayr is defending the classical morpho-geographical approach to systematics, he is quite right, as most taxonomists have long recognized. The controversy over Mayr's biological species concept is due to his linking the morpho-geographical approach to an inaccurate definition.