Effective pollination is a complex phenomenon determined by the outcome of the interaction between pollen transfer and a plants’ pollinator dependency, yet most studies investigate pollinator effectiveness without consideration of plant mating system differences.

We investigated pollinator effectiveness in three types of blueberry that differed in their degree of pollinator dependency as measured by plant mating system: two self-compatible highbush cultivars and one partially self-incompatible rabbiteye cultivar. We quantified pollinator effectiveness as a function of the fruit set and fruit weight resulting from single and multiple floral visits (2 – 15 visits), in comparison with estimates of fruit set and fruit weight resulting from experimental pollination treatments (open-pollination, cross-pollination and self-pollination).

Single-visit effectiveness of fruit set was similar across pollinator taxa but considerably higher in both self-compatible cultivars. The probability of fruit set in all three blueberry types improved in response to an increasing number of visits, but this relationship was steeper in self-compatible cultivars: >90% probability of fruit set was achieved in three to five visits. In the self-incompatible rabbiteye cultivar, 58% fruit set was achieved with 15 visits. Multiple visits improved fruit weight by 27% – 48% in self-compatible cultivars, but not there was no relationship in rabbiteye. Pollination deficits in fruit set and fruit weight due to self-pollination were most pronounced in rabbiteye. Synthesis and applications: Improved understanding of cultivar-level mating system differences in plants will inform pollination planning and management in agroecosystems. Self-compatible (highbush) cultivars require less floral visitation to maximise fruit production. Therefore, these cultivars may be best suited to landscapes in which pollinator abundance is low, such as intensive and/or simple landscapes. In contrast, self-incompatible (rabbiteye) cultivars may benefit from the implementation of mixed-cultivar crop row plantings to facilitate cross-pollination through the movement by bees between cultivars.

R script and datasets used to estimate floral visitor abundance, single and multiple visit pollinator effectiveness and pollination dependence in three types of blueberry: northern highbush (Vaccinium corymbosum), rabbiteye (V. virgatum), and southern highbush (V. corymbosum hybrid).