Tolerating castration by hiding flowers in plain sight

A key challenge in the study of inter-specific cooperation, or mutualisms, is to understand the mechanisms that prevent cheating. However, many mutualisms are still invaded by parasites, and we propose that plant ‘tolerance strategies’, which allow an individual to re-grow and/or reproduce after damage, might have evolved to recover some of the fitness lost to parasitism. We focus on the ant plant Cordia nodosa and its protecting ant symbionts Allomerus octoarticulatus and Azteca spp. Allomerus is a castration parasite, destroying host flowers. However, some flowers are produced on new branch shoots, where they escape castration. Here, we ask whether C. nodosa tolerates castration by directing floral growth to new shoots. We find that C. nodosa produces four times more floral buds on the new shoots of Allomerus-inhabited plants, compared to Azteca-inhabited plants. Additionally, on Allomerus-inhabited plants, the production of floral buds is three times greater on new than on mature shoots, and more floral buds are grown on plants with new shoots than on those without. We conclude therefore that C. nodosa tolerates castration by re-allocating floral resources to new shoots. We also test whether tolerance exploits behavioural differences between the brood-tending ‘nurse’ ant worker caste, which is active on mature shoots, and the new shoot-protecting ‘patroller’ caste. Behavioural assay experiments reveal that nurse ants are strongly attracted to extracts of flowers, whereas patrollers are not, suggesting that nurses are the primary castrating caste. Thus, producing flowers on new shoots avoids the castrating caste. We finish by proposing that tolerance strategies can promote the evolution and maintenance of mutualism.