Co-flowering with congeners does not affect buzz-pollinator specialization and pollination performance in Rhexia mariana, but does affect floral trait variance.

premisePollinator-mediated plant-plant interactions may be negative (i.e., competition, reproductive interference) or positive (i.e., facilitation). Especially when co-flowering with close relatives (e.g., congeners), negative interactions through reproductive interference may be strong and result in floral trait divergence and increased pollination niche partitioning. However, when pollination services are limited, positive effects of pollinator sharing through floral trait similarity may outweigh the costs of reproductive interference. We therefore tested for evidence of negative or positive pollinator-mediated plant-plant interactions across a gradient of varying congeneric co-flowering contexts in the genus Rhexia (Melastomataceae).

methodsWe studied pollinator interactions, pollination performance and floral traits of Rhexia mariana across nine localities of varying cogeneric co-flowering contexts (up to seven Rhexia species co-flowering) in central Florida, USA.

resultsRegardless of co-flowering context, differential pollinator specialization was weak, with Bombus impatiens visiting all Rhexia species, removing pollen through buzz-pollination. Co-flowering context did not affect visitation rates or pollination performance, but the floral traits of R. mariana differed signficantly and were less variable in low compared to high co-flowering contexts.

conclusionsWe did not find support for either negative or positive effects of co-flowering on pollination performance in Rhexia mariana, indicating that co-flowering may instead have neutral effects. Negative effects of co-flowering with close relatives hence do not seem to be strong enough to drive specialization on distinct buzzing bee pollinators in Rhexia. Sampling across more localities, paired with experimental approaches (e.g., manipulating co-flowering density, assessing post-zygotic reproductive barriers) will be essential to clarify whether reproductive interference through co-flowering is indeed low.