EFFECTS OF SELECTIVE SEED PREDATION BY RODENTS ON SHORTGRASS ESTABLISHMENT

Our experiment tested whether selective seed predation by rodents affects the establishment of blue grama (Bouteloua gracilis) and buffalo grass (Buchloe dactyloides) at a disturbed site in the Colorado shortgrass steppe. We hypothesized that foraging rodent granivores respond positively to large seed sizes and to high seed densities, and we predicted that intensive foraging would constrain grass establishment. Capture-recapture studies confirmed that heteromyid granivores, kangaroo rats (Dipodomys ordii) and hispid pocket mice (Perognathus hispidus), comprised 48% of the local rodent community. Seeds of blue grama and buffalo grass were shallowly planted at three densities in replicated plots within and outside rodent exclosures. The frequency of digging in seeded rows (seed predation) and the frequency of herbivory on seedlings were calculated. We assessed grass establishment using seedling frequency and harvested aboveground biomass. Seed foraging was related positively to seed size, but foraging responses to seed densities were not confirmed. Herbivory on seedlings was not intensive and was similar between the two grasses. Seedling frequency and harvested biomass numbers for the species with the larger seeds, buffalo grass, were reduced more in rodent-accessible plots than were values for blue grama. Rodent seed predation can constrain the establishment of large-seeded species, but results should be interpreted to reflect site-specific conditions and the characteristics of seeded species.