Influence of grazing history on the community structure of grasshoppers of a mixed-grass prairie

A study was conducted to determine if grazing history affected community composition of plants and grasshoppers by comparing communities in an undisturbed grassland that has not been grazed by livestock for approximately 60 yr (Badlands National Park, South Dakota) with communities on adjacent grasslands that are managed for grazing (Buffalo Gap National Grasslands). Grasshoppers were sampled with sweep nets in August 1986 at 29 sites within the Badlands National Park and Buffalo Gap National Grasslands. Concurrently, plant communities were sampled at the sites to determine species richness of grasses and percentage of coverage of all grass species, forbs, sedges, and bare ground. Detrended correspondence analysis and Spearman correlation analysis of insect abundance and incidence data were used to assess spatial changes in grasshopper community structure and to relate vegetation variables to grasshopper community composition. Results indicated that grasshopper communities changed along habitat gradients defined by grazing history and coverage and species richness of grasses. Sites within the undisturbed grassland habitat had similar community compositions of grasshoppers and were dominated by mixed forb- and grass-feeding species. Grazed sites supported higher populations of obligate grass-feeding grasshoppers. Differences in grasshopper communities between the grazed and undisturbed sites were, in part, due to differences in plant community composition; grazed sites were dominated by short grasses, whereas ungrazed sites were composed mainly of tallgrass species. In general, the mixed forb- and grass-feeding species of grasshoppers increased along vegetation gradients defined by increasing coverage of Agropyron smithii Rydb., Bromus spp., Poa spp., Stipa viridula Trin., and Carex spp., and decreasing coverage of bare ground and Bouteloua gracilis (H.B.K.) Lag. ex Steud. Communities dominated by obligate grass-feeding species were associated with high coverage of bare ground and B. gracilis. Grasshopper species richness was greatest at sites with high richness of grasses and total coverage of grasses and sedges