Plant-herbivore coevolution and plant speciation.

More than five decades ago, Ehrlich and Raven proposed a revolutionary idea-that the evolution of novel plant defense could spur adaptive radiation in plants. Despite motivating much work on plant-herbivore coevolution and defense theory, Ehrlich and Raven never proposed a mechanism for their "escape and radiate" model. Recent intriguing mechanisms proposed by Marquis et al. include sympatric divergence, pleiotropic effects of plant defense traits on reproductive isolation, and strong postzygotic isolation, but these may not be general features of herbivore-mediated speciation. An alternate view is that herbivores may impose strong divergent selection on defenses in allopatric plant populations, with plant-herbivore coevolution driving local adaptation resulting in plant speciation. Building on these ideas, we propose three scenarios that consider the role of herbivores in plant speciation. These include (1) vicariance, subsequent coevolution within populations and adaptive divergence between geographically isolated populations, (2) colonization of a new habitat lacking effective herbivores followed by loss of defense and then re-evolution and coevolution of defense in response to novel herbivores, and (3) evolution of a new defense followed by range expansion, vicariance, and coevolution. We discuss the general role of coevolution in plant speciation and consider outstanding issues related to understanding: (1) the mechanisms behind cospeciation of plants and insects, (2) geographic variation in defense phenotypes, (3) how defensive traits and geography map on plant phylogenies, and (4) the role of herbivores in driving character displacement in defense phenotypes of related species in sympatry.