The interactions between plants and their insect herbivores are one of the main generators of biological diversity. A fundamental process generating this outstanding diversity is diet expansion to novel host plants. During the last four decades scientists accumulated evidence showing that co-adaptation between plants and herbivores is a major process assembling plant-herbivore interactions. However, rescent research suggests that adaptation is not always a prerequisite to generate novel plant-herbivore interactions. Novel associations between plants and insect herbivores may be assembled by ecological fitting - an ecological process whereby herbivores colonize novel host plants as a result of the suites of preadapted traits that they carry at the time of colonization. A widespread assumption concerning the architecture of insect herbivore genotypes is the "Jack of all trades master of none" principle. This principle proposes that there is a trade-off in genotype performances between host plants. The main prediction of this principle is that genotype performance will be negatively correlated among hosts. Genotypes displaying high performance on a given host will perform poorly on other hosts. This constraint of adaptation to multiple host plants implies that diet specialization will be selected over generalization. Contrary to these theoretical expectations, in most cases, genotypes that perform well in one host will also perform well in other host plants. Positive correlations in cross-host performance represent ecological and evolutionary dynamics opposite to the "Jack of all trades" principle. In this scenario genotypes with high performance on one host plant also have high performance on other plants, promoting generalization. The predictions of the current theory on the assemblage of novel plant-herbivore interactions focus on the fact that most insect herbivores are specialists. However, to fully understand the processes underlying the assembly of novel plant-insect interactions, it is necessary to study diet expansions in both specialist and generalist insect herbivores. This dissertation was performed at La Selva Biological Station, a tropical rain forest in Costa Rica, Central America. We studied a group of neotropical herbivores, the "rolled-leaf beetles" (Cephaloleia, Chrysomelidae: Cassidinae) and their host plants, neotropical plants in the order Zingiberales. Cephaloleia beetles have evolved with neotropical Zingiberales for the last 40-60 MY. Four paleotropical and one South American members of the Zingiberales have been introduced to La Selva during the last decade. After these introductions, currently seven Cephaloleia beetles are expanding their diets to exotic Zingiberales. These incipient diet expansions represent an opportunity to understand: 1) the relative roles of adaptation vs ecological fitting on the demography and colonization success of novel hosts in generalist and specialist herbivores and 2) whether genotypic performance across original and novel hosts are negatively correlated, as predicted by the "Jack of all trades" principle, or genotype performances across original and novel hosts are positively correlated. For most of the experiments included in this dissertation, I focused on the performance of larvae and adults of two generalist (Cephaloleia belti, C. dilaticollis) and two specialist beetles (Cephaloleia dorsalis, C. placida) reared in the laboratory on native or exotic Zingiberales. Generalist and specialist species display similar responses when changing their diets to novel hosts. Larvae preferred and performed better in the original than in the novel hosts. Adults usually displayed the opposite pattern, i.e. higher preference for and longevity on the exotic than on the novel hosts. In most novel interactions between Cephaloleia beetles and exotic Zingiberales, larval performance required adaptation, but adult performance was pre-adapted to the novel hosts. Therefore, both adaptation and ecological fitting are playing a role during diet expansions to novel hosts. Vital rates estimated through experimental demography show that population growth is reduced on novel host plants for both generalist and specialist Cephaloleia. Although in some cases population growth on the novel hosts is negative, suggesting the potential outcome of extinction after colonization or source-sink dynamics, several beetle species displayed positive population growth in the novel host plants. Positive instantaneous population growth rates in novel hosts supports diet expansions without substantial initial evolutionary change through ecological fitting. In quantitative genetics experiments testing for cross-host genetic correlations in performance between the original and novel host plants, we did not find evidence for negative genetic correlations, as predicted by the "Jack of all trades" principle. Most genetic correlations in performance between original and novel hosts were either not significant or they were positive. These results represent very different ecological and evolutionary dynamics than those predicted by the "Jack of all trades" principle. In this case, genotypes with high performance on original hosts also displayed high performance on novel hosts, promoting generalization. In conclusion, interactions between Cephaloleia beetles and plants from the order Zingiberales are labile. In some cases diet expansions may occur without substantial evolutionary change. In addition the genetic architecture of genotypes promotes generalization during diet expansions to novel hosts.

No description available.