Lack of pollinators limits fruit production in commercial blueberry

Background/Question/Methods

While most agricultural pollination is provided by a single domesticated species, the honeybee, recent research suggests that a diversity of native pollinators make important contributions to both the mean level and the stability of pollination services.  We studied the pollination of commercial blueberry in New Jersey, USA to answer the following questions. What are the relative contributions of native bees and honeybees to blueberry pollination? Do native pollinators provide more stable pollination services over time than honeybees? Does pollination by native bees or honeybees better predict fruit weight? Are blueberry flowers receiving sufficient pollination at current bee visitation rates to achieve maximum fruit weight? We measured flower visitation rate by native bees and honeybees on 9 separate days over three years at 16 farms, and also measured the number of pollen grains each type of bee deposits per flower visit. We calculated the mean and the inverse coefficient of variation (mean/standard deviation) of pollination services provided by honeybees and native bees (in aggregate) over time at each site. We also measured the extent of pollination limitation for two varieties of blueberry over two years by comparing berry weights from flowers either receiving ambient pollination or saturated with pollen by hand. 

Results/Conclusions

Overall, honeybees provided 86% of pollination function, while the remaining 14% was from native bees. We found that pollination by honeybees was more stable than pollination by the native bee community (honeybee mean/st dev =1.90±0.57, native bees =1.55±0.64), although the difference was not significant (t=1.63, p=0.11).  This is contrary to the prediction of the diversity-stability hypothesis and likely results from the greater natural variability of wild, as opposed to human-subsidized, species. Average berry weight was positively correlated with honeybee flower visitation rate in three of the four year-by-variety combinations studied (2011: Bluecrop R²=0.44, p=0.02, Duke R²=0.41, p=0.05; 2012: Bluecrop R²=0.30, p=0.05, Duke R²=-0.15, p=0.65). In contrast, berry weight was not significantly associated with native bee visitation rate in any of the four year-by-variety combinations. In 2011, hand-pollinated berries were significantly heavier than open pollinated standards in both Duke (t=3.69, p=0.0003) and in Bluecrop (t=4.46, p<0.0001). However, in 2012 there was no significant difference in either cultivar (Duke: t=1.92, p=0.06; Bluecrop: t=1.33, p=0.19). Overall, our study shows that pollination can be a limiting factor in commercial fruit production even though all farms in our study provided managed honeybees at the recommended densities.