Scaling up effects of measures mitigating pollinator loss from local- to landscape-level population responses

1. Declining pollinator populations have caused concern about consequences for food production, and have initiated an increasing number of initiatives that aim to mitigate pollinator loss through enhancement of floral resources. Studies evaluating effects of mitigation measures generally demonstrate positive responses of pollinators to floral resource enhancement. However, it remains unclear whether this represents landscape-level population effects or results from a spatial redistribution of individuals from otherwise unaffected populations. 2. Here we present a method for estimating landscape-level population effects using data from commonly used standardized pollinator transect surveys. The approach links local density responses of pollinators in both mitigation sites and surrounding landscape elements to the area these habitats occupy in mitigation landscapes as well as control landscapes to obtain landscape-level population estimates. 3. We demonstrate the method using data from a two-year study examining the effects of experimental wildflower enhancements on wild bumble bees and solitary bees in Dutch agricultural landscapes. The results show that conclusions based on local responses may differ significantly from those based on landscape-level responses. 4. Wildflower enhancements significantly enhanced landscape-level abundance of both bumble bees and solitary bees. Bumble bees showed a pronounced positive local density response in mitigation sites and the surrounding landscape that was in line with significant landscape-level increases in abundance. However, solitary bees showed no local response to mitigation sites, and the landscape-level increases in abundance only became apparent when the area of bee habitat was taken into account. 5. Incorporating the area of both newly created and pre-existing pollinator habitats into effect estimates accounts for density-dependent processes such as dilution, spill-over and local concentration of individuals. It therefore results in more reliable estimates of the response to mitigation measures of pollinators, as well as other mobile arthropod groups that are often being surveyed using transect surveys.