We characterized the structure of condensed tannins from 16 woody plant species (seven genera, six families) and determined their effects on six herbivorous insect species (four genera, two families). There were major differences in tannin structure, even between congeneric plant species. Condensed tannins differed markedly in their antiherbivore activity, averaged over these herbivores, and the herbivores differed in their sensitivity, averaged over these tannins. Furthermore, the same tannin can have different effects on different herbivores, presumably because of interactions between tannin structure and gut physiology. Results challenge the view that tannins provide an evolutionarily stable plant defense because of their uniform chemical properties. Condensed tannin can sometimes impact herbivore fitness through effects on survival and growth, but the largest effects in 45 insect-tannin combinations were less than that of many other plant metabolites at lower doses. Even at high doses, condensed tannins frequently had no strong antiherbivore activity, even against insects with no evolutionary history of encountering the tannin (Ͻ10% reduction in growth rate in 24 of 45 experiments). Most condensed tannins apparently do not have broad-spectrum antiherbivore activity. We doubt that selective pressures from folivorous insects can be the main explanation for the diversion of so much carbon, in so many plant species, into the synthesis of condensed tannins.

The resource allocation hypothesis is based on the assumption that defenses are costly, but relatively few studies have quantified the reproductive price of induced defenses, which represent the best means of measuring such costs in isolation from the genotypic costs that confound research involving constitutive defenses. Jasmonic acid (JA) is a plant signal molecule involved in the defensive responses of plants. It induces many of the same chemicals that are associated with herbivore damage, and thus offers a means of inducing plants without the removal of leaf area, which incurs its own costs. In tomato plants, JA induced resistance to Manduca sexta and increased levels of two defensive enzymes, polyphenol oxidase and peroxidase. We measured the impact of JA-induced defenses in tomato, Lycopersicon esculentum (Solanaceae), on several variables associated with reproductive success: fruit number, fruit weight, ripening time, time of fruit-set, number of seeds per fruit, total seeds per plant, the relationship between fruit weight and seed number, and germination success. Plants were grown in a pest-free greenhouse and treated biweekly with solvent or with JA at either of two concentrations: 10 mM or 1 mM. The high concentration of JA led to fewer but larger fruits, longer ripening time, delayed fruit-set, fewer seeds per plant, and fewer seeds per unit of fruit weight. The reproductive impact of induction was reduced at the lower dose, but still significant; 1 mM JA resulted in delayed fruit-set and fewer seeds per unit of fruit weight, compared to control plants. Our research indicates that JA-induced defenses impose significant costs on tomato plants.

The gypsy moth, Lymantria dispar, and the northern tiger swallowtail, Papilio canadensis, overlap geographically as well as in their host ranges. Adult female swallowtails are incapable of distinguishing between damaged and undamaged leaves, and the opportunities for competition between these two species are numerous. We designed field and laboratory experiments to look for evidence of indirect competition between P. canadensis and L. dispar larvae. Swallowtail caterpillars were reared in the laboratory on leaves from gypsy-moth-defoliated and undefoliated trees to explore host-plant effects. We tested for pathogen-mediated interactions by rearing swallowtail larvae on both sterilized and unsterilized leaves from defoliated and undefoliated sources. In addition, we measured the effects of known gypsy moth pathogens, as well as gypsy moth body fluids, on the growth and survival of swallowtail larvae. Field experiments were designed to detect the presence of parasitoid-mediated competition, as well: we recorded parasitism of swallowtail caterpillars placed in the field either where there were no gypsy moth larvae present, or where we had artificially created dense gypsy moth populations. We found evidence that swallowtails were negatively affected by gypsy moths in several ways: defoliation by gypsy moths depressed swallowtail growth rate and survival, whether leaves were sterilized or not; sterilization significantly reduced the effect of defoliation, and gypsy moth body fluids proved lethal; and swallowtail caterpillars suffered significantly increased rates of parasitism when they were placed in the field near gypsy moth infestations.

No abstract