Evaluation of interactions between honeybees and alternative managed pollinators: a meta-analysis of their effect on crop productivity

Abstract

The productivity of approximately 75% of crops worldwide depends to some extent on insect pollination. However, while global agriculture is becoming more dependent on pollinators, wild populations of pollinators are declining. For this reason, hives of Apis mellifera (honeybees), the most widely used pollinator, are commonly placed in the fields; in recent years, alternative managed pollinators (AMPs) such as Bombus spp. or Osmia spp. have also been used. Thus, for evidence-based pollination management, we need to know whether the pollination service provided by AMPs can replace, complement or synergistically interact with that provided by honeybees. We asked: Does crop productivity differ between fields with honeybees and those with AMPs? Does productivity increase by incorporating AMPs in addition to managed honeybees? Do the effects of managed honeybees and AMPs interact? We performed a meta-analysis based on 28 studies on 20 crops. We estimated effect sizes (ln(R)) for crop productivity (fruit/seed set, fruit/seed quality and yield) from 73 comparisons between honeybees and an AMP, and 21 comparisons between honeybees alone and honeybees plus an AMP. Overall, we found no evidence of difference in crop productivity between honeybees and AMPs when managed separately. However, the productivity of crops pollinated by honeybees together with AMPs was 22% ± 6 (SE) higher than that of crops pollinated only by honeybees. Moreover, we found a weak evidence of a positive effect of beehive density on crop productivity when an AMP was added, suggesting a synergistic interaction between honeybees and AMPs. We conclude that, on average, honeybee performance is similar to that of AMPs, and that increasing the number of managed pollinator species can improve crop productivity in the short-term, particularly in systems with impoverished pollinator faunas. More generally, this review confirms the positive effect of pollinator diversity on pollination service, suggesting this can be partly recreated using a suite of managed pollinators.