Abstract
Life cycle assessment (LCA) aims at providing standardized evaluations of processes involving resource use, human health, and environmental consequences. Currently, spatial dependencies are most often neglected, though they are essential for impact categories like biodiversity. The "Swiss Agricultural Life Cycle Assessment for Biodiversity (SALCA-BD)" evaluates the impact of agricultural field management on 11 indicator species groups. We tested if its performance can be improved by accounting for the spatial context of the individual fields. We used high-resolution bird/butterfly point observations in two agricultural regions in Switzerland and built linear mixed models to compare SALCA-BD scores to the observed species richness at the field/landscape scale. We calculated a set of landscape metrics, tested their relationship with the landscape-model prediction errors, and then added all significant metrics as additional predictors to the landscape models. Our results show that field-scale SALCA-BD scores were significantly related to the observed field-scale richness for both indicator groups. However, the performance decreased when aggregated to the landscape scale, with high variability between regions. Adding specific landscape metrics improved the landscape model for birds but not for butterflies. Integrating the spatial context to LCA biodiversity assessments could provide moderate benefits, while its usefulness depends on the conditions of the respective assessment.