Interspecific pollen transfer between two coflowering species was minimized by bumblebee fidelity and differential pollen placement on the bumblebee body

Aims When sympatric flowering plant species in a natural community share pollinators, study of plant–plant interactions via interspecific pollen transfer (iPT) is essential for understanding species coexistence. However, little is known about the extent of iPT between interactive species and its causes. Methods To explore how sympatric flowering plants sharing pollinators minimize deleterious effects of iPT, we investigated the pollination ecology of two endemic species, Salvia przewalskii and Delphinium yuanum, in an alpine meadow in the Hengduan mountains, southwest China. We quantified conspecific and interspecific visits by shared bumblebee pollinators, amounts of pollen placed on differ ent body sites of the pollinators and stigmatic pollen loads on openpollinated flowers. To examine whether iPT affects female fitness, we measured pollen germination and seed production in the two species in an artificial pollination experiment. Important Findings one bumblebee species, Bombus trifasciatus, was found to be the sole effective pollinator for the two coflowering species. Pollination experiments indicated that deposition of heterospecific pollen could significantly decrease seed set in both species. Experiments showed that S. przewalskii pollen could germinate well on stigmas of D. yuanum, inhibiting conspecific pollen germination in D. yuanum. However, seed set was not lower under open pollination than under cross-pollination within species, suggesting that no female fitness loss was caused by iPT. in foraging bouts with pollinator switches, switches from S. przewalskii to D. yuanum were relatively more frequent (8.27%) than the converse (1.72%). However, iPT from S. prze walskii to D. yuanum accounted for only 1.82% of total stigmatic pollen loads while the reverse iPT to S. przewalskii was 8.70%, indicating that more switches of bumblebees to D. yuanum did not result in higher iPT. by contrast, selection for reduced iPT to S. przewalskii would limit pollinator switches from D. yuanum. We found that a bumblebee generally carried pollen grains from both species but the two species differed in the position of pollen placement on the bumblebee’s body; S. przewalskii’ s pollen was concentrated on the dorsal thorax while D. yuanum’ s pollen was concentrated ventrally on the head. This differential pollen placement along with pollinator fidelity largely reduced iPT between the two species with a shared pollinator.