Compared with other environmental indicators, few U.S. studies have quantified the relationships between production of biomass crops and biodiversity. Here, we developed a modeling framework, Bio-EST (Bioenergy-biodiversity Estimation), to assess the change in avian species richness associated with change in land management to grow biomass crops. Bio-EST considers effects on birds from changes associated with growing dedicated energy crops, including perennial grasses, annual crops such as sorghum, and short-rotation woody crops (SRWCs). We collected spatially-referenced presence records for selected species from the Biodiversity Information Serving Our Nation database. We used boosted-regression tree methods to develop species distribution models (SDMs) that included land-use/land-cover (LULC) classes, bioclimatic variables and elevation as predictors. The outputs of the SDMs were local estimates of the probability of occupancy by a species within 1-km raster pixels for the conterminous U.S.. Summing these estimates produces an index of bird richness for each grid cell. We modeled the probability of occurrence in a biomass crop replacing a non-biomass crop based on the response ratio of a species for the two LULC classes. We separately report results for three groups of bird species: (i) 12 predominantly grassland bird species in perennial grasses, and (ii) 16 predominantly forest birds in SRWCs, and (iii) 24 generalist birds in SRWCs.
Results/Conclusions
Overall, the performance of SDMs was excellent as indicated by the Kappa statistics varying from 0.42 to 1.0 (all p-value <0.0001) across all bird species. Overall changes in suitable habitat (i.e., ranges) of avian species were small (for the majority (>98%) of areas, bird species showed no change in occupancy under the BC1-2040 landscape compared to the 2014 landscape). Forest generalists tended to benefit from changes, whereas forest and grassland specialists had larger ranges in the 2014 landscape. The analysis suggests that grassland birds currently in agricultural landscapes dominated by corn and other rowcrops will respond positively to switchgrass. In addition, field studies are needed to quantify responses of the native bird community to growing miscanthus in the United States. Geographic patterns in grassland species reflect responses to conversion of agricultural lands to BC1 2040 future switchgrass and energy sorghum. Patterns for forest projected decreases appear to be concentrated in the middle of the country. The analysis presented here can also be extended to represent other wildlife taxa once enough comparisons of wildlife performance (e.g., density, reproductive success) in multiple food crop and biomass crop habitats have been made.