AbstractPollination syndromes have been widely used to predict the groups of pollinators of plants. Their predictability has been demonstrated for plant species pollinated by bees, birds, and bats. However, the importance of other aspects on the predictability of pollination syndromes still remain unknown. For instance, plant life cycle has not been considered in evaluations of the pollination syndrome predictability. The goal of this study was to evaluate the predictability of pollination syndromes in nine species with contrasting life cycles. Pollination syndromes were inferred based on floral traits of each plant species, whereas pollinator effectiveness of each group of floral visitors was established through pollen load analysis. Predictability of pollination syndromes was tested through the association between floral traits and the abundance of the most‐effective pollinators. Bees, flies, birds, beetles, and butterflies were the primary and secondary pollinators inferred from floral traits of each species. Although six out of the nine plant species studied were visited by effective pollinators; only four species were effectively pollinated by the groups of primary or secondary pollinators predicted from floral traits. However,Dahlia coccineaandTigridia pavoniawere the only species in which the predicted pollinator group was both effective and showed an association with floral traits. Pollination syndromes were not predictive of pollinators for any of the annual and perennial plant species. Although pollination syndromes are useful to understand some aspects of the reproductive biology of plant species, several other factors might have a role on their predictability.
Read full abstract