Format: ORAL

Amanda V. da Silva, Marjorie G. Weber, Anselmo Nogueira, Thais N. Vasconcelos, Avery Russell, Tamiris D. Lima4, Lilian F. Melo5, Laura C. Leal6

1 Center of Natural and Human Sciences, Federal University of ABC, São Bernardo do Campo, Brazil 2 Department of Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, USA 3 Department of Biology, Missouri State University, Springfield, USA 4 Department of Botany, São Paulo State University, Botucatu, Brazil 5 Department of Ecology, Conservation and Biodiversity, Federal University of Uberlândia, Uberlândia, Brazil 6 Department of Ecology and Evolutionary Biology, Federal University of São Paulo, São Paulo, Brazil.

Mutualistic interactions involve varying degrees of conflicts, including the interference between different types of mutualism. One common interference occurs between protective ants and pollinators in plants bearing extrafloral nectaries (EFNs). While patrolling the plant, ants can reduce herbivory but also repel pollinators, especially bees. Since this conflict can affect plant reproductive performance, it might be selectively disadvantageous for plants to have both traits associated with ant defense and specialized bee pollination. Although conflicts in mutualisms have been extensively studied on the ecological scale, the evolutionary consequences of traits mediating interactions have not been considered. We hypothesize that there is a negative evolutionary correlation between traits mediating plant protection by ants and exclusive pollination by bees as in buzz pollinated plants. For that, we used a phylogenetic comparative approach, mapping all angiosperms reported to bear EFN associated with ant attendance and poricidal anthers associated with buzz pollination in an angiosperm tree comprising ca. 8,000 genera. We found that EFNs occur in 755 genera while poricidal anthers occurs in 544. However, only 45 plant genera have both traits. Moreover, we found a negative evolutionary correlation between EFNs and poricidal anthers, supporting our hypothesis that the interference of one mutualism in another can affect the evolutionary trajectory of the traits driving these interactions. This result means plants protected by ants are less likely to be involved in buzz pollination, or vice-versa, throughout angiosperm diversification. Altogether, our results evidenced that protective and pollination mutualisms may exert a reciprocal selective force that reduces the likelihood of the simultaneous occurrence of both ants and bee partners across angiosperm evolution. This emphasizes that we studies focused on the evolution of mutualistic traits should consider the joint effect of simultaneous mutualistic interactions instead of evaluating the traits related to different mutualisms separately