Reproduction of most flowering plants depends on pollination by animals. Plant-pollinator interactions can be influenced by the diversity of the resident plant community, but recent studies have highlighted the potential importance of soil-borne microbial communities in affecting pollinator behavior. It is currently unknown to what extent these changes in pollinator behavior result in a change in individual plant fecundity. In a tall grass prairie system, we tested whether plant-soil microbial interactions mediate plant-pollinator interactions, and thus indirectly affect plant fecundity, and whether the diversity of resident plant community alters these relationships. To untangle direct and indirect effects of above- and belowground interactions that influence plant fecundity, we performed pollination manipulations and pollinator observations in experimental plant communities that varied in the number of plant species (1, multiple species) and in the composition of soil microbial communities (with, without live soil microbes). To determine how these factors affected fecundity, we conducted a germination experiment and measured germination rate as well as proportion of seeds germinated.
Results/Conclusions
We found that soil microbes negatively affected the number of seeds produced by both manually and naturally pollinated flowers in Rudbeckia hirta when individuals were grown in monocultures. However, manually pollinated flowers in polycultures produced equal numbers of seeds when grown in either soil condition, while naturally pollinated flowers produced significantly fewer seeds in live soil compared to sterile soil. Pollinator visitation did not significantly differ between soil treatments in polyculture, but was significantly reduced in monocultures with live soil microbes. These results suggest that the negative effects of soil microbes on individual fecundity in polyculture are not caused by a change in pollinator behavior. Rather, the lower fecundity of naturally pollinated flowers compared to manually pollinated flowers in polyculture could indicate an increase in pollen limitation in these conditions, leading to greater susceptibility when receiving heterospecific pollen. We did not find a difference between soil or diversity treatments in the proportion of seeds germinated. Our results indicate that soil microbes can lead to changes in pollinator behavior, but these changes do not have straightforward effects on fecundity.