Density-Dependent Gene Dispersal in Liatris

Abstract

The effect of colony density on gene migration has been interpreted in the perennial herb, Liatris aspera. Colonies with 1, 3.2, 5, and 11 plants/square meter, respectively, were studied. The pollen component of gene migration was inferred from pollinator flight distances, 1,614 of which were analyzed. Bees were the primary pollinators. The seed component of gene migration was measured in the experimental garden, utilizing transplants as the seed source. Axial distances were determined for 2,309 wind-borne seeds. Means of pollinator flight distances displayed a close curvilinear relationship to colony density, an increase in density being accompanied by a decrease in flight distance. Flights are predominately to near neighbors, and there is strong positive correlation between the variance of flight distance and the variance of plant spacing. Thus we conclude that pollinator behavior and pollen-dispersal distance are density-dependent. Gene migration has been considered within the framework of Wright's neighborhood model. Neighborhood area is density-dependent, the relationship between density and area being an inverse one. The dependency of Na on density is somewhat obscured by the seed-dispersal variance, which is density-independent and which dominates the total gene-dispersal variance. Since gene migration is influenced by colony density, the potential for local differentiation is to some extent density-dependent, high-density arrays being most conducive to subdivision. Genetically effective density increased with colony density. However, when only the density-dependent component was considered, Ne remained relatively constant over the range of densities. This feature is traceable to the Ne equation, which includes a square function of spacing (pollen dispersal) and an inverse square-root function of spacing (density).