Pollinator species-specific contribution to reproduction in tropical trees

Background/Question/Methods

The evolutionary rationale for the existence of pollinator syndromes centers on the idea that different pollinator species differ in their ability to successfully pollinate flowers. Pollinators may differ not only quantitatively, in terms of fruit and seed set, but also qualitatively in terms of the genetic composition of the plant’s offspring. This is especially critical given that many plant species exhibit inbreeding depression and high fine-scale spatial genetic structuring in their populations. Given that nearly 5.8 million ha of tropical forest are deforested per year, there is great need to identify which pollinator species are most vulnerable to habitat change, and what the independent contributions of each pollinator species is for the maintenance of reproductive processes and genetic diversity for tropical trees. We conducted a single visit experiment in three populations of the understory tree Miconia affinis in Panama. In each population, we randomly chose 30 trees and bagged five randomly chosen inflorescences per tree. The bagged inflorescences were exposed to a single insect visitor, which was subsequently caught and identified. The fruits were allowed to develop to maturity and collected. We compared different pollinator species in terms of success in setting fruits, the number of viable seeds per fruit, and sire diversity as estimated by a microsatellite-based population genetic analysis. Specifically, a total of 404 seeds were screened at 9 polymorphic loci to determine sire identity and assess pollinator dispersal ability. 

Results/Conclusions

Twenty different bee species visited M. affinis’ flowers. These species differ widely in their size with the intertegular distance ranging from 0.91 to 7.72 mm. Our results indicate that pollinator species differed significantly in their individual contribution to plant reproduction. Flowers visited by large bees set more fruits. Large bees also produced higher proportions of viable seeds per fruit than smaller-bodied species. Finally, fruits produced by large bees exhibited greater sire diversity than those produced by small bees. Our results reveal important differences in terms of effectiveness within a pollinator group traditionally considered distinct and homogeneous (i.e., bees). Furthermore, we present the first analyses of pollinator effectiveness that incorporates both seed quantities and sire diversity. Given the prevalence of genetic structuring and inbreeding depression in plant populations, we posit that assessments of pollinator quality should consider not only seed set numbers, but also the pollinator’s ability to promote genetic diversity in plant’s offspring.