Interaction diversity explains the maintenance of phytochemical diversity.

The production of complex mixtures of secondary metabolites is a ubiquitous feature of plants. Several evolutionary hypotheses seek to explain how phytochemical diversity is maintained, including the synergy hypothesis, the interaction diversity hypothesis, and the screening hypothesis. We experimentally tested a set of predictions derived from these hypotheses by manipulating the richness and structural diversity of phenolic metabolites in the diets of eight plant consumers. Across 3940 total bioassays, there was clear support for the interaction diversity hypothesis over the synergy or screening hypotheses. The number of consumers affected by a particular phenolic composition increased with increasing richness and structural diversity of compounds. Furthermore, the bioactivity of phenolics was consumer-specific. All compounds tested reduced the performance of at least one consumer, but no compounds affected all consumers. These results show how phytochemical diversity may be maintained in nature by a complex selective landscape exerted by diverse communities of plant consumers.