Pollinator species richness is correlated to flowering plant richness: A meta-analysis

Background/Question/Methods

Bees are important pollinators, critical for the continued survival of plants in both natural and agricultural ecosystems. Yet there is growing concern that pollinators, especially bees, are declining globally. An understanding of the mechanisms responsible for bee community composition is crucial for their conservation in natural systems and their restoration in human-dominated urban and agricultural landscapes. As bees are completely dependent upon flowering plants for their food resources, pollen and nectar, it has often been assumed that flowering plant communities structure bee communities. For lack of experiments that could determine directionality, we pose a simpler question about the association between bees and their floral resources. A number of observational surveys of natural, semi-natural, and human-impacted ecosystems have been undertaken that measure the diversity of plant communities and of the pollinator or flower visitor communities. We tested the hypothesis that flowering plant species richness positively correlates with bee species richness using a meta-analysis approach. This approach allows us to determine the direction and strength of the relationship between bees and flowering plants, as well as the generality of the pattern across ecosystems. In order to find eligible articles we searched ISI Web of Knowledge^SM using relevant topic search terms. Richness was the only diversity indicator available in sufficient abundance to be used in the meta-analysis. Using Fisher’s F-ratio or the r² value, Pearson’s correlation coefficient, r, was calculated and used as effect size.

Results/Conclusions

Fifteen effect sizes were used in the final meta-analysis. The result of the meta-analysis was a strongly positive mean correlation between flowering plant species richness and pollinator (primarily bee) species richness (r = 0.573 ± 0.09 S.E., p < 0.0001). Although it is impossible to determine causation from this meta-analysis, it is likely that flowering plant richness contributes to bee species richness through: 1) provisioning bees with food resources; 2) the correlation between floral richness and floral abundance at the community level; and 3) temporal consistency of floral resources due to phenological complementarity. However, the causal mechanisms resulting in this correlation in bee species richness and flowering plant species richness need to be explored further through experiments that manipulate either floral diversity or bee diversity or both. More surveys in underrepresented ecosystems would also provide important evidence for the existence of a general relationship between pollinators and floral resources, or the lack thereof.