Pollen limitation: How much is really about the pollinator community?

Background/Question/Methods

Plant reproduction is frequently limited by the deposition of adequate compatible pollen.  Lack of visits by pollinators is often invoked as an explanation, especially in the context of the need to conserve declining pollinator populations.  We review the evidence for a link between the diversity and abundance of floral visitors and pollen limitation, and also present the results of a pollen limitation experiment on 6 wildflower species at 6 sites and two years.  We measured seed production of control and pollen-supplemented inflorescences and interpret significantly reduced seed production in control inflorescences as pollen limitation.  We additionally collected floral visitors from flowers of all focal species (and other wildflowers) at all sites.  We used Simpson’s index weighted by the total number of insects caught as our measure of pollinator diversity, so highest values are for plant species with high numbers of visitors from different species.  Finally, we constructed plant-pollinator interaction webs and calculated interaction asymmetry, an indication of the degree to which plants rely on specific species of floral visitors.  We then asked how pollen limitation was related to weighted diversity and asymmetry using regression. 

Results/Conclusions

We studied 38 combinations of species/sites/years, but found significant pollen limitation (difference in seed set between open vs. supplemented inflorescences) in just 6 instances.  This is a much lower frequency of pollen limitation than has been shown in recent reviews.  Focal plant species were visited by 4 to 48 insect species, but network asymmetry analysis indicated little reliance of plants on specific pollinators.  An across-species regression analysis indicated no relationship between pollen limitation and the weighted diversity of insect visitors.  Within-species analysis revealed one instance of a significant relationship—but pollen limitation increased with the diversity and abundance of the visiting insect community, contrary to predictions that a diverse pollinator community should improve plant reproductive output.  A review of other studies attempting to link pollinator diversity to pollen limitation (with higher measured values for pollen limitation than in our study) similarly suggests the link is weak.  Although in some crop species there is a positive relationship between pollinator diversity and reproductive output, the importance of pollinators as an ecosystem service in natural communities remains surprisingly unclear.