In most mutualisms, hosts interact with multiple mutualists. Competition among these mutualists is commonly observed, and can have distinct effects on the persistence of the interaction. It is also common for hosts to interact with multiple exploiters of mutualism, but interactions among these exploiters has received almost no attention. In pollination mutualisms, interactions between nectar robbers can be both competitive (since they share a limited pool of nectar) and facilitative (when “secondary” robbers feed through holes pierced by “primary” robbers). We investigated experimentally whether facilitation between two exploiters of a pollination mutualism (a primary and a secondary nectar robber) is strong enough to overcome the cost of competition for nectar. We observed foraging bouts of individual Bombus bifarius, a secondary-robbing bumble bee, on Linaria vulgaris flowers in a flight cage. Plants were manipulated to simulate the actions of Bombus occidentalis, a primary-robbing bumble bee that visits the same plants. Individual stalks were manipulated in a crossed design to have either low or high primary robbing rates and either rewarding or unrewarding primary-robbed flowers. We calculated net energetic benefit per bout using observed flower handling times and known energetic benefit acquisition and metabolic rates. Additionally, we calculated the strength of the interaction using Relative Interaction Intensity (RII) to determine whether treatment conditions produced a more competitive or facilitative interaction outcome for B. bifarius.
The secondary nectar-robber B. bifarius acquired a significantly higher net energetic benefit from patches with low primary robbing as opposed to patches with high primary robbing, and from patches with rewarding as opposed to unrewarding primary robbed flowers. There was no significant interaction between the main effects. Hence, B. bifarius performs better in the absence of B. occidentalis, suggesting that competition between primary and secondary nectar robbers is strong enough to mask the facilitative effect primary robbing has on secondary robbing. Calculation of RII supports this finding, showing significantly negative values for all treatment conditions relative to the control. Our results highlight an important but commonly overlooked intraguild interaction that has the potential to affect all parties in a multispecies mutualism and, ultimately, interaction persistence.