Response of native bees to local farm management and landscape features in a perennial agroecosystem

Background/Question/Methods

The expansion of modern agriculture and human activity has led to the alteration and fragmentation of the landscape.  The resulting loss of habitat has profound effects on the ability of organisms to persist and the provisioning of ecosystem services.  Pollination is a particularly important ecosystem service for both natural and agricultural systems.  Bees, the most important pollinators, are especially sensitive to changes in the landscape and exposure to agricultural chemicals.  Therefore, management decisions made at the local and landscape scale by farmers can have important consequences for bees.  In order to ensure their persistence and enhance the conservation of native bees in agroecosystems, it is important to understand how they respond to both local and landscape features.  The objective of this study was to determine to what extent native bees in a perennial cropping system respond to local farm management practices and landscape features.  To address our objective we pan trapped bees at 42 commercially managed cranberry bogs in central Wisconsin.  Prior to site selection we used a GIS to classify the landscape in central Wisconsin using the Cropland Data Layer and chose sites to encompass a gradient of surrounding landscape from highly agricultural to highly wooded. 

Results/Conclusions

We found that native bee abundance was positively correlated with percent woodland (R² = 0.20, p < 0.01) in the surrounding 1km.  Species richness was also positively correlated with percent woodland (R² = 0.18, p < 0.01) in the surrounding 1km.  Neither bee abundance nor species richness were significantly correlated with local use of agricultural chemicals based on a score derived from chemical toxicity, application timing, and number of applications.  These results show that native bees respond in number and diversity to features in the surrounding landscape.  The lack of correlation with agricultural chemical use suggests either that growers are using bee-friendly practices (e.g., selecting targeted chemicals versus broad spectrum, spraying at night when pollinators are not active) or that simplifying local management to a single toxicity score is insufficient.  These findings suggest that management decisions that impact the landscape composition surrounding pollinator-dependent crops have the potential to cause declines in the pollinator community and therefore pollination services within the crop fields.