Bee species diversity enhances productivity and stability in a perennial crop

Abstract

Wild bees provide important pollination services to agroecoystems, but the mechanisms which underlie their contribution to ecosystem functioning--and, therefore, their importance in maintaining and enhancing these services-remain unclear. We evaluated several mechanisms through which wild bees contribute to crop productivity, the stability of pollinator visitation, and the efficiency of individual pollinators in a highly bee-pollination dependent plant, highbush blueberry. We surveyed the bee community (through transect sampling and pan trapping) and measured pollination of both open- and singly-visited flowers. We found that the abundance of managed honey bees, Apis mellifera, and wild-bee richness were equally important in describing resulting open pollination. Wild-bee richness was a better predictor of pollination than wild-bee abundance. We also found evidence suggesting pollinator visitation (and subsequent pollination) are stabilized through the differential response of bee taxa to weather (i.e., response diversity). Variation in the individual visit efficiency of A. mellifera and the southeastern blueberry bee, Habropoda laboriosa, a wild specialist, was not associated with changes in the pollinator community. Our findings add to a growing literature that diverse pollinator communities provide more stable and productive ecosystem services.

Figure 1. Relationship of pollination to Apis abundance (a) and wild-bee richness (b) in highbush blueberry.

Each point represents the mean seed set per location visit. Trend lines are based on model estimates described in table 2 (using mean of year estimates) at different levels of Apis and wild-bee richness. For (a), logarithmic fits represent wild-bee richness of 0 (dotted line), 2 (dashed line), and 4 (solid line). For (b), linear fits represent Apis abundance of 2 (dotted line), 20 (dashed line), and 200 (solid line).

Figure 2. Forager abundance and pollination during ‘inclement’ and ‘optimal’ weather conditions in blueberry fields.

Mean estimates with SE bars shown for Apis abundance (a), wild-bee abundance (b), and open-pollinated seed set (c). Replicates, included in parentheses, are location visits (a, b) and number of berries (c).