Wild bees are important pollinators playing a key role in crop cultivation and wild plant reproduction. Like other insects, wild bees face declines due to pollution, land use change, agricultural intensification and climate change.
This study proposes a framework that combines advantages of species distribution modelling and ecosystem service mapping to develop a high-resolution model that predicts spatial patterns of pollination ecosystem services potential from wild bees at the local scale.
In the first step, researchers used GBIF-mediated occurrences combined with five bioclimatic variables and seven land use/cover classes to model the potential distribution of 25 important Andrena bee species at the global scale.
In the study area of Bornhöved Lake District in Schleswig-Holstein, Germany, the authors then applied habitat filters and foraging ranges of each species to refine the output of the species distribution models at a local scale, producing a map of relative pollination potential.
By deploying traps in rapeseed fields of the study area, the authors were able to test the predictive powers of their model, finding correlation between predicted pollination service potential and observed Andrena richness and abundance.