Distinct Animal-Generated Edge Effects in a Tallgrass Prairie Community

Natural disturbances often affect adjacent vegeta? tion, and in some cases the influence extends beyond the immediate area of the disturbance (Carson and Fickett 1990). In extreme cases, plant communities may develop waves of plant biomass, which are ini? tiated when certain plants receive substantially more of a critical resource than others. For example, trees on the edge of a forest may grow tall in response to the unimpeded access to sunlight. Their neighbors, suffering from reduced sunlight, would be short, thus providing trees somewhat farther into the forest with ad? ditional sunlight, and so forth, yielding what has been termed a competition-induced wave of biomass. In most studies to date, edges yielding biomass waves were initiated at natural breaks (e.g., forest edges; Spru? gel 1976) or by artificially structured plantings (e.g., experimental garden plots; Franco and Harper 1988). Under natural conditions, animals may produce such edges directly (consumption; Reichman and Smith 1985) or indirectly (trampling, burrowing; Gillham 1956, Lock 1972, Inouye et al. 1987, Andrew 1988, Reichman 1988, Cantor and Whitham 1989, Gibson 1989, Huntly 1991). Here we present evidence that fossorial herbivores (North American pocket gophers) generate distinct edge patterns in overlying plant communities, and that these edges may initiate competition-induced waves of plant biomass. Pocket gophers produce disturbances that generate numerous gaps and edges in vegetation. They excavate extensive tunnels (Reichman et al. 1982) and deposit the tailings on the surface as mounds of soil, or in abandoned burrows (Huntly and Inouye 1988). Burrows may underlie up to 7.5% of a field (at various depths centered in the root zone), and mounds may cover up to 20% of an area (Grant et al. 1980, Reichman et al. 1982, Huntly and Inouye 1988, Reichman and Jarvis 1989). Plant biomass directly over burrows and on mounds is significantly lower (30-50%) than in adjacent control plots (Reichman and Smith 1985, Reichman and Jarvis 1989). Vegetation is often completely missing on fresh mounds and returns slowly, preserving open areas in the vegetation for weeks or months (Grant et al. 1980, Grant and McBrayer 1981, Spencer et al. 1985, Hobbs and Hobbs 1987, Reichman 1988). To determine if pocket gopher excavation activity generates specific patterns of vegetation, we measured plant biomass in a transect of quadrats extending from directly over gopher burrows and mounds out to 50 cm from the disturbances. In addition, we analyzed soil from the same quadrats for total nitrogen and mea? sured phosphorus, extractable nitrogen, and water po? tentials of sample plants from the quadrats in an attempt to identify the cause of any emergent biomass patterns.