The local population ideally is composed of an aggregation of organisms sharing in a common gene pool and exchanging genes freely (Wright, 1931, 1943a, 1949). Such freedom, however, rarely is achieved in nature, for the dispersal of genes is highly leptokurtic (Dobzhansky and Wright, 1943, 1947; Bateman, 1947a, 1950; Griffiths, 1950; Colwell, 1951). Within a local population of outbreeding organisms, this form of gene dispersal leads to the formation of a series of overlapping panmictic units referred to as neighborhoods (Wright, 1940, 1943a, 1946). Knowledge of the neighborhood structure of plant and animal populations is meager. In animals, neighborhood studies have been conducted only on Drosophila (Dobzhansky and Wright, 1941, 1943, 1947), Cepaea (Lamotte, 1951) and Sceloporus (Kerster, 1964). In plants neighborhood investigations have dealt almost exclusively with wind-pollinated trees including Fraxinus, Pseudotsuga, Populus, Ulmus, Picea, Cedrus (J. Wright, 1953) and Pinus (Bannister, 1965); a notable exception is the work on the entomophilous, annual herb, Linanthus parryae (Wright, 1943b). To the authors' knowledge, the neighborhood structure of populations of perennial herbs has never been determined. The work discussed below was undertaken to fill this void. The purpose of this paper is to consider intrapopulation gene dispersal and the type of neighborhood engendered by the dispersal pattern in the self-incompatible, perennial herb, Phlox pilosa L. Phlox pilosa is one of the most highly diversified species in the genus, and has evolved a series of subspecies and races fitting its ecologically diversified area (Wherry, 1930, 1955; Levin and Smith, 1965). The species is a conspicuous element of the spring flora of mesic prairies and open woodlands throughout much' of the eastern United States. Colonies of P. pilosa contain from tens to thousands of individuals, and often display marked exomorphic variation. Pollination is performed by an array of insects, the Lepidoptera being especially suited for the task (Grant and Grant, 1965). Seed-set in natural populations is about 40%, i.e., the triovulate capsule yields an average of 1.2 seeds (Levin and Kerster, 1967a). Seeds are ripened during the early summer and scattered by the force of the dehiscing capsule; there are no special mechanisms for longdistance dispersal. Neighborhood determinations were made on three colonies in Cook County, Illinois. The colonies inhabited undisturbed sand prairies, and each was composed of several thousand plants which occupied an area in excess of 1000 M2. Colony designation (I, II, III) is assigned on the basis of density, colony I being the most dense. Data on pollen and seed dispersal and density were collected from each colony, and synthesized for neighborhood calculations.
Read full abstract