Competition and Coexistence in a North Carolina Grassland: II. The Effects of the Experimental Removal of Species

Abstract

(1) Species were removed singly (by hand) or in groups (by herbicide or soil sterilization) in a mown field. The vegetation of treated and control quadrats was assessed before and after the removals by measuring the cover of each species. (2) In all cases total cover returned to its original value within a year, although the species composition remained different. (3) The effects of the removals of single species upon the abundance of the other species were small, accounting for an average of only 7% of the variance. Fourteen of the seventy-two pairwise effects tested were significant. Interactions between the different pairs of species did not differ greatly in magnitude. The results of this experiment do not support the hypothesis that this community is divided into well-defined groups of competing species (guilds). A previous study provided evidence for the division of the community into temporal guilds. But within each temporal guild the species are characterized by diffuse competition. (4) Interactions between pairs of species were in general non-reciprocal. (5) Some interactions were found among three or more species that were not predictable from interactions of pairs of the constituent species. Two species with vigorous vegetative spread pre-empted space from competitors, and there were several significant negative responses to species removals. (6) No divisions of the community along taxonomic or morphological lines were found. However, the temporal guilds correspond to the distribution of the C3 and C4 photosynthetic pathway in the grass species. (7) The spatial patterning of the community showed no relationship to the responses to removals.