Are crop yields limited by pollinators? Proper assessments using pollinator gradients require measurements of flower density and yield potential

Abstract It is widely documented that many crops depend on animal—and primarily insect—pollination, but the degree to which pollinators limit yield in comparison with other factors is poorly understood. Recent studies conclude that pollinator visitation rates limit yields of many crops, based upon positive correlations between these two variables. However, these studies typically suffer from incomplete data on two key variables that affect both pollinator recruitment and yield: flower density per area and yield potential per flower, both driven by maternal resource availability. Here we review the literature on animal‐pollinated crops showing that (i) yields can be positively associated with flower density, and (ii) pollinator density can also be positively correlated with flower density; the third positive association observed between pollinator density and yield may result from the two first correlations, without causal relationship. Likewise, positive associations observed between (i) the amount of maternal resources per flower and yield potential per flower, and (ii) between the amount of maternal resources per flower and pollinator density per flower can both explain the positive correlation observed between yield per flower and pollinator density per flower. We conclude with an illustration reusing data from the literature to show that, without incorporating data on flower density and yield potential per flower, measuring yields and gradients of pollinator density per area alone can overestimate the degree to which pollinator visitation rates directly limit yield. Read the free Plain Language Summary for this article on the Journal blog.