Decades of research have shown that phytochemicals play a key role in structuring interactions between plants and their insect herbivores. However, it is still unclear why plants produce such an incredible diversity of different compounds—a single leaf may contain hundreds or even thousands of different phytochemicals with no known role in primary metabolism. We conducted an experiment to test three hypotheses that may explain this diversity: 1) combinations of compounds work synergistically and provide more effective defense than any individual compound alone, 2) different compounds are required to defend against diverse enemies, or 3) plants produce mostly inactive compounds but biosynthetic pathways are maintained to increase the probability of effective novel defenses (screening hypothesis). These hypotheses lead to distinct predictions about the effects of plant secondary metabolites on insect performance, but we know of no studies that have experimentally tested these predictions. We reared four species of Lepidopteran herbivores on artificial diets in which we manipulated three key aspects of phytochemical diversity—compound richness, compound evenness, and functional diversity. We focused on phenolic metabolites due to their diversity and ubiquity in the plant kingdom.
Results/Conclusions
Unexpectedly, we found no negative effects of increasing richness, evenness, or structural diversity of phenolic mixtures on insect performance. In fact, in the few cases where we detected a significant effect of richness on performance, insects performed better on more diverse diet mixtures. However, most individual compounds (13 of 14 compounds tested) did impact insect performance in some way. Different compounds variably affected different species of insect herbivores and different metrics of performance (survival, pupal weights, and development time). Thus, our results support the hypothesis that phytochemical diversity in plants is best explained by diverse interactions in the complex biotic environments in which plants have evolved. Our experimental approach sheds new mechanistic understanding that can complement field-based studies to further disentangle patterns of phytochemical diversity across plant species and environments.