Links between prey assemblages and poison frog toxins: A landscape ecology approach to assess how biotic interactions affect species phenotypes.

Ecological studies of species pairs showed that biotic interactions promote phenotypic change and eco-evolutionary feedbacks. However, it is unclear how phenotypes respond to synergistic interactions with multiple taxa. We investigate whether interactions with multiple prey species explain spatially structured variation in the skin toxins of the neotropical poison frog Oophaga pumilio. Specifically, we assess how dissimilarity (i.e., beta diversity) of alkaloid-bearing arthropod prey assemblages (68 ant species) and evolutionary divergence between frog populations (from a neutral genetic marker) contribute to frog poison dissimilarity (toxin profiles composed of 230 different lipophilic alkaloids sampled from 934 frogs at 46 sites). We find that models that incorporate spatial turnover in the composition of ant assemblages explain part of the frog alkaloid variation, and we infer unique alkaloid combinations across the range of O. pumilio. Moreover, we find that alkaloid variation increases weakly with the evolutionary divergence between frog populations. Our results pose two hypotheses: First, the distribution of only a few prey species may explain most of the geographic variation in poison frog alkaloids; second, different codistributed prey species may be redundant alkaloid sources. The analytical framework proposed here can be extended to other multitrophic systems, coevolutionary mosaics, microbial assemblages, and ecosystem services.