Remobilization of initially deposited pollen grains has negligible impact on gene dispersal in bumble bee-pollinated Brassica napus

The transfer of pollen among flowers is an important determinant of mating patterns in plants, and considerable attention has been devoted to understanding animal‐mediated pollen transfer and its causal mechanisms. Conventionally, pollen carryover has been envisioned as a one‐step process, but a pollen grain's journey may not be so simple. Potentially, pollen deposited on a flower's stigma could be remobilized and transferred to the stigma of a second flower. Alternatively, pollen may be deposited on unreceptive floral surfaces, such as petals, by either a pollinator or the wind and then remobilized and transferred to the stigma during a pollinator visit.

We quantified the potential for secondary dispersal of pollen by bumble bees (Bombus terrestris L.) visiting flowers of Brassica napus L., chosen because it typifies a generalized animal‐pollinated flower and because of concern about pollen escaping from genetically modified varieties in agriculture.

We tested whether genetically marked pollen that was initially deposited by a bumble bee on a flower's stigma could later be transferred by another bee and thereby fertilize ovules on a different flower. In a second experiment, we tested whether marked pollen deposited on a flower's petals (as if by a pollinator or the wind) could later be transferred to the stigma by a bee and thereby fertilize the flower's ovules. In both experiments, the numbers of genetically marked progeny were recorded and a model was developed to quantify the relative effectiveness of direct vs secondary pollen dispersal.

Our analysis revealed that only small numbers of pollen grains achieved secondary dispersal by a pollinator, and that the probability of fertilization being achieved through secondary dispersal was minute.