Do plant‐eating insect lineages pass through phases of host‐use generalism during speciation and host switching? Phylogenetic evidence

Abstract

The Oscillation Hypothesis posits that plant-eating insect diversity is generated by cycles of diet breadth expansion and contraction. Although at any given time most plant-eating insect species are host specialists, host-use evolution and speciation tend to entail a phase of generalism. The main evidence for this comes from comparative phylogenetic studies, but with mixed support. Here, I review and add to this evidence. I show that some of the original work that inspired the Oscillation Hypothesis is flawed in a way that leads to spurious inferences about trends in the evolution of diet diversity. And I present a new analysis which fails to support its predictions about patterns of species diversity. On the other hand, some of the published work that claims to reject the Oscillation Hypothesis may actually provide some of the strongest support for it, and I present new analyses which support its prediction that host-use generalism facilitates host-use evolution. In summary, the Oscillation Hypothesis successfully predicts some phylogenetic patterns but not others. Generalism appears to facilitate host-use evolution, but it does not appear to be inevitably chased by host-use specialization and speciation.

Bayesian inference of ancestral host-parasite interactions under a phylogenetic model of host repertoire evolution

“…Although these methods are designed for biogeographic inference, a similar approach is clearly suitable for more realistic modeling of host-parasite coevolution dynamics, where colonization and loss of hosts (instead of discrete areas) is modeled as a continuous-time Markov process (e.g. Hardy 2017). In biogeography, the colonization of a new area or the disappearance from a previously occupied area is modeled as a binary trait: the species is either present or absent in the area.…”

Colonizations drive host shifts, diversification of preferences and expansion of herbivore diet breadth

“…One possible mechanism to generate increase in diet breadth after colonizations would be that colonization events were followed by host shifts, and additional hosts were added to the diet during evolutionary transitions from traditional to novel hosts, as suggested by Hardy 9 . A model of parasite evolution in the context of a fitness landscape with heterogeneous hosts does, indeed, generate this scenario 12 .…”

“…Hardy 9 asks “does experimental adaptation of a plant-eating insect population to a novel host result in host-use generalism, and improve the odds of evolving additional new host associations?” Braga et al 12 use an experiment “in silico” to answer this question in the affirmative. Here we answer it “in vivo,” applying a combination of observation and experiment to a single butterfly species, Edith’s checkerspot (Euphydryas editha) , and showing that habitat colonizations were followed by diversification of individual host preferences and increases of population-level diet breadth.…”

Do major host shifts spark diversification in butterflies?

“…All analyses were performed in R (R Core Team, ). First, we used Dispersal Extinction Cladogenesis (DEC) models to reconstruct the phylogenetic history of the use of host orders and families (as in Hardy, ). Although DEC models were initially developed to estimate ancestral geographic ranges, they are well‐suited for the estimation of ancestral states for any multi‐state discrete character such as host use (Hardy, ).…”