ao Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), pelo suporte financeiro durante o período de desenvolvimento da dissertação; por fim, de todo coração e com toda a sinceridade, agradeço meus familiares, amigos, professores, à Universidade de São Paulo, à Universidade Federal de Uberlândia e todos que contribuíram direta ou indiretamente para que esse trabalho fosse desenvolvido.A todos vocês, muito obrigada.Antes o voo da ave, que passa e não deixa rastro, Que a passagem do animal, que fica lembrada no chão.A ave passa e esquece, e assim deve ser.O animal, onde já não está e por isso de nada serve, Mostra que já esteve, o que não serve para nada.A recordação é uma traição à Natureza, Porque a Natureza de ontem não é Natureza.O que foi não é nada, e lembrar não é ver. Extrafloral nectaries (EFNs) are glands that produce nectar not related to pollination and may be present in both vegetative and reproductive plant parts. These structures attract a great diversity of arthropods, mainly ants, which protect plants from herbivorous attack. In general, ant-plant interactions mediated by EFNs are positive, however, in some cases they may have negative effects, interfering directly in plant's fitness. From this perspective, our study aimed to evaluates whether the attraction of ants by an EFNs bearing plant, effectively protects the vegetable against herbivorous attack, but indirectly has anegative impact on pollination. We used as a model, Palicourea rigida (Rubiaceae), a common shrub in the Cerrado, that presents pericarpial nectaries (PNs -EFNs located in fruits). The main hypothesis of the work was that ants benefits P. rigida, reducing the action of foliar herbivores, but at the same time negatively interfere on its reproductive success by inhibiting the action of the main pollinators. The field work was developed between November 2016 and February 2017, in the ecological cerrado reserve of the Clube Caça e Pesca Itororó de Uberlândia, in Uberlândia city, Minas Gerais, Brazil. The results corroborated our main hypothesis, demonstrating that ants in fact protect the plant reducing the action of foliar herbivores and can, depending on the ant species, negatively interfere in fruit set, since they inhibit the action of the main pollinators -the hummingbirds.These results show that P. rigida has a complex system of interactions, involving ant-plantpollinator. This is the first study to evaluate the indirect costs of ant-plant mutualism on the fruit set of a plant pollinated by birds.