Abstract
Biotic interactions can promote phenotypic change, yet we have a limited understanding of how phenotypes respond to concomitant interactions with many species. We introduce a framework to investigate how biotic interactions contribute to spatially structured phenotypes and apply it to the drivers of chemical defense variation among populations of the poison frog Oophaga pumilio (Dendrobatidae). Specifically, we assess how beta-diversity of alkaloid-bearing arthropod prey assemblages (based on projected distributions of toxic ant species) and evolutionary divergence among O. pumilio populations (based on a neutral genetic marker) contribute to poison composition variation over the range of these frogs. Under Generalized Dissimilarity Modeling, ant assemblage turnover predicted alkaloid turnover and geographic regions harboring unique toxin combinations in O. pumilio. Evolutionary relatedness between frog populations had a weak effect on poison composition variation based on a Multiple Matrix Regression approach. By leading to spatially structured phenotypes, biotic interactions promote eco-evolutionary feedbacks and functional diversity in ecological communities. The analytical framework proposed here can be extended to other multi-trophic systems, coevolutionary mosaics, microbial assemblages, and ecosystem services.
