Multi-scale considerations for grassland butterfly conservation in agroecosystems

Global change threatens the persistence of multiple taxonomic groups, including butterflies. While conservation efforts for butterfly populations have increased, they are often hampered by a lack of true density estimates and a better understanding of ecological factors that influence density. Our objective was to enhance current grassland butterfly conservation efforts by using line-transect distance sampling to calculate density estimates and model the influence of landscape and local variables on butterfly density in the Northern Great Plains, USA. We calculated density for five obligate and ten facultative grassland species to produce one of the most extensive datasets for butterfly densities to date. In contrast to most previous research, we found that landscape variables influenced butterfly density more often than local variables. Specifically, the percent cover of perennial grasslands, crop lands, and wetlands appeared in 90% of species models, whereas common local variables-forb richness and invasive plant cover- appeared in 60% of best-ranked models. We expected the density of obligate butterfly species to decrease as invasive plants increased, but butterfly species' responses varied with larval diet, instead of habitat associations. Best-ranked models for Danaus plexippus and Speyeria idalia, two species of conservation concern with obligate host plants, did not include local host plant availability. Our results reiterate the importance of modeling species responses to variables across multiple scales and the potential benefits of conserving large tracts of grasslands. Although results emphasize the need for conservation at the landscape scale, managing for heterogeneous local scale variables will also help conserve grassland butterflies.