Plants produce an enormous diversity of secondary metabolites, but the evolutionary mechanisms that maintain this diversity are still unclear. The interaction diversity hypothesis suggests that complex chemical phenotypes are maintained because different metabolites benefit plants in different pairwise interactions with a diversity of other organisms. In this synthesis, we extend the interaction diversity hypothesis to consider that fruits, as potential hotspots of interactions with both antagonists and mutualists, are likely important incubators of phytochemical diversity. We provide a case study focused on the Neotropical shrub Piper reticulatum that demonstrates: 1) secondary metabolites in fruits have complex and cascading effects for shaping the outcome of both mutualistic and antagonistic fruit-frugivore interactions, and; 2) fruits can harbor substantially higher levels of phytochemical diversity than leaves, even though leaves have been the primary focus of plant chemical ecology research for decades. We then suggest a number of research priorities for integrating chemical ecology with fruit-frugivore interaction research and make specific, testable predictions for patterns that should emerge if fruit interaction diversity has helped shape phytochemical diversity. Testing these predictions in a range of systems will provide new insight into the mechanisms driving frugivory and seed dispersal and shape an improved, whole-plant perspective on plant chemical trait evolution.

Premise of the study: The specialized metabolites of plants are recognized as key chemical traits in mediating the ecology and evolution of sundry plant-biotic interactions, from pollination to seed predation. Intra- and interspecific patterns of specialized metabolite diversity have been studied extensively in leaves, but the diverse biotic interactions which contribute to specialized metabolite diversity encompass all plant organs. Focusing on two species of Psychotria shrubs, we investigate and compare patterns of specialized metabolite diversity in leaves and fruit with respect to each organ's diversity of biotic interactions. Methods: To evaluate associations between biotic interaction diversity and specialized metabolite diversity, we combine UPLC-MS metabolomic analysis of foliar and fruit specialized metabolites with existing surveys of leaf- and fruit-centered biotic interactions. We compare patterns of specialized metabolite richness and variance among vegetative and reproductive tissues, among plants, and between species. Key results: In our study system, leaves are involved in a greater number of host-specific biotic interactions than fruit, while fruit-centric interactions are more ecologically diverse. This was reflected in specialized metabolite richness - leaves contained more than fruit, while each contained over 200 organ-specific specialized metabolites. Within each species, leaf- and fruit specialized metabolite composition varied independently of one another across plants. Specialized metabolite composition exhibited stronger contrasts between organs than between species. Conclusions: As ecologically disparate plant organs with organ-specific specialized metabolite traits, leaves and fruit can each contribute to the tremendous overall diversity of plant specialized metabolites.