Flower preference and visit duration of European honey bees (Apis mellifera) on watermelon (Citrullus lanatus)

In light of Colony Collapse Disorder, the pollinator efficiency of honey bees on various crops is of vital interest to agriculture. A common method of approximating pollinator efficiency on watermelon involves quantifying the amount of pollen bees deposit on virgin pistillate flowers in a single, initial visit. Our goal was to determine if estimates of visit duration and pollen deposition approximated by initial visits are comparable to subsequent visits, and whether time of day affects these measurements. Triploid (seedless) pistillate watermelon flowers were bagged prior to anthesis then excised from the plant and presented for visitation by bees during two separate time periods, early and late morning, for both one- and four-visit treatments. Initial visits in the late morning were found to be about three times longer than those in the early morning, with initial visits lasting longer than those of subsequent visits for both time periods. Despite shorter visits, pollen deposition was twice as high during the early morning. For both time periods, each subsequent visit deposited only a fraction of the amount of pollen deposited during the initial visit. Our results indicate that pollen deposition is not inherently correlated with visit duration, perhaps because the amount of pollen available in the system is dissipated by pollinators as the day progresses. Most importantly, our results suggest that studies evaluating pollinators on the basis of initial visits to virgin flowers may be overestimating per-visit contributions to pollination. Time of day effects should likely be taken into account as well.