Abstract:

Some plants benefit from self-fertilization for reproductive assurance, and thus might experience little effect of nectar robbing on fruit production. In facultative selfing plants which involve both outcrossing by pollinators and selfing when pollinators are scare or not available, nectar robbing is expected to have no influence on fruit set, but may affect pollinator-mediated fruit set. In order to test this prediction, we manipulated robbing (open to be robbed or excluding nectar robbers by caging flowers) in an annual facultative selfing plant Invarvillea sinensis var. sinensis and quantified fruit set, seed number and weight per fruit. Pollinator-mediated fruit set was evaluated on the basis of the closure of stigmatic lobes, which generally occurs after pollinator visiting. The height of flowers was measured to test whether nectar robbers have a visiting preference for specific floral traits. Bumble bees, workers of Bombus patagiatus, were the primary pollinators of I. sinensis and some of them were also nectar robbers. Robbers stole nectar from both open flowers and flower buds. Averaged frequency of nectar robbing was 20.24%, ranging from 0 to 51.43%. Experiments excluding nectar robbers showed that nectar robbing did not have significant effects on proportion of fruit set, seed number per fruit and seed mass per fruit. However, the proportion of stigmatic lobes closed was significantly higher in robbed flowers than in unrobbed flowers, suggesting that robbing influences pollinator-mediated fruit set. Height of robbed flowers was significantly higher than that of unrobbed flowers, suggesting that nectar robbers prefer robbing from larger and higher flowers. These results provide insight into the effects of nectar robbing on the reproduction of plants.

Key words: Invarvillea, facultative selfing, flower height, nectar robbing, pollinator-mediated fruit set, reproduction