摘要:Feral horses are widespread in the western United States, with the majority of feral horse herds found in the Great Basin. There is a federal mandate to manage these herds in order to maintain “ecological balance”; however, understanding of the specific effects of feral horse grazing on rangeland plant communities in this region is incomplete. To address this research gap, we utilized long-term grazing exclosures and fenceline contrasts to evaluate the impacts of feral horses on several plant community variables (diversity, richness, dominance, and biomass) and species composition. Because the effects of grazing can vary with site precipitation and productivity, we selected 5 sites from 4 different rangeland types (Great Basin Desert, Colorado Plateau, Rocky Mountain grassland, and mixed grass prairie) that spanned a mean annual precipitation gradient of 229 to 413 mm. Our results did not reveal a significant effect of feral horse grazing on plant community composition, species richness, diversity, evenness, or dominance. In contrast, total aboveground herbaceous biomass and grass biomass were significantly reduced with feral horse grazing, but these effects did not vary with mean annual precipitation. Our results suggest that, at least at the sites we studied, feral horses have affected the plant community by reducing herbaceous biomass but have not caused plant community shifts. Additional multisite studies, preferably with standardized exclosures and larger sample sizes, would increase our understanding of feral horse grazing effects and inform management of feral horse herds in the western United States.Al oeste de los Estados Unidos se han dispersado caballos ferales, localizándose, en su gran mayoría, en la Gran Cuenca (Great Basin). Existe una orden federal para administrar estas manadas a fin de mantener el “equilibrio ecológico”; sin embargo, las consecuencias específicas del pastoreo de caballos ferales en los pastizales de esta región de EE.UU. no se conocen por completo. Para abordar esta falta de conocimiento comparamos el uso de zonas exclusivas de pastoreo a largo plazo y áreas de cercado contrastantes, con el fin de evaluar los efectos de los caballos ferales en diversas variables de las comunidades vegetales (diversidad, riqueza, dominio y biomasa) y en su composición. Debido a que los efectos del pastoreo pueden variar según la lluvia y la productividad del lugar, seleccionamos cinco sitios de cuatro tipos diferentes de pastizales (desierto de la Gran Cuenca, meseta de Colorado, praderas mixtas y pastizales de las Montañas Rocosas) que comprendieron un gradiente de precipitación media anual