Half a million species of herbivorous insects have been described. Most of them are diet specialists, using only a few plant species as hosts. Biologists suspect that their specificity is key to their diversity. But why do herbivorous insects tend to be diet specialists? In this review, we catalog a broad range of explanations. We review the evidence for each and suggest lines of research to obtain the evidence we lack. We then draw attention to a second major question, namely how changes in diet breadth affect the rest of a species’ biology. In particular, we know little about how changes in diet breadth feed back on genetic architecture, the population genetic environment, and other aspects of a species’ ecology. Knowing more about how generalists and specialists differ should go a long way toward sorting out potential explanations of specificity, and yield a deeper understanding of herbivorous insect diversity. Expected final online publication date for the Annual Review of Ecology, Evolution, and Systematics, Volume 51 is November 2, 2020. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

When herbivorous insects interact, they can increase or decrease each other's fitness. As it stands, we know little of what causes this variation. Classic competition theory predicts that competition will increase with niche overlap and population density. And classic hypotheses of herbivorous insect diversification predict that diet specialists will be superior competitors to generalists. Here, we test these predictions using phylogenetic meta‐analysis. We estimate the effects of diet breadth, population density and proxies of niche overlap: phylogenetic relatedness, physical proximity and feeding‐guild membership. As predicted, we find that competition between herbivorous insects increases with population density as well as phylogenetic and physical proximity. Contrary to predictions, competition tends to be stronger between than within feeding guilds and affects specialists as much as generalists. This is the first statistical evidence that niche overlap increases competition between herbivorous insects. However, niche overlap is not everything; complex feeding guild effects indicate important indirect interactions.

Across herbivorous insect clades, species richness and host-use diversity tend to positively covary. This could be because host-use divergence drives speciation, or because it raises the ecological limits on species richness. To evaluate these hypotheses, we performed phylogenetic path model analyses of the species diversity of Nearctic aphids. Here, we show that variation in the species richness of aphid clades is caused mainly by host-use divergence, whereas variation in speciation rates is caused more by divergence in non-host-related niche variables. Aphid speciation is affected by both the evolution of host and non-host-related niche components, but the former is largely caused by the latter. Thus, our analyses suggest that host-use divergence can both raise the ecological limits on species richness and drive speciation, although in the latter case, host-use divergence tends to be a step along the causal path leading from non-host-related niche evolution to speciation.

The Escape and Radiate Hypothesis posits that herbivorous insects and their host plants diversify through antagonistic coevolutionary adaptive radiation. For more than 50 years, it has inspired predictions about herbivorous insect macro-evolution, but only recently have the resources begun to fall into place for rigorous testing of those predictions. Here, with comparative phylogenetic analyses of nymphalid butterflies, we test two of these predictions: that major host switches tend to increase species diversification and that such increases will be proportional to the scope of ecological opportunity afforded by a particular novel host association. We find that by and large the effect of major host-use changes on butterfly diversity is the opposite of what was predicted; although it appears that the evolution of a few novel host associations can cause short-term bursts of speciation, in general, major changes in host use tend to be linked to significant long-term decreases in butterfly species richness. K E Y W O R D Scoevolution, evolutionary ecology, plant-insect interactions, species diversification | 3637