Effect of nitrogen and water treatment on leaf chemistry in horsenettle (Solanum carolinense), and relationship to herbivory by flea beetles (Epitrix spp.) and tobacco hornworm (Manduca sexta).

We studied the interaction between plants (horsenettle; Solanum carolinense) and herbivorous insects (flea beetles; Epitrix spp., and tobacco hornworm; Manduca sexta) by focusing on three questions: (1) Does variation in nitrogen availability affect leaf chemistry as predicted by the carbon-nutrient balance (CNB) hypothesis? (2) Does variation in plant treatment and leaf chemistry affect insect feeding? (3) Is there an interaction between the insect herbivores that is mediated by variation in leaf chemistry? For three successive years (1998-2001), we grew a set of clones of 10 maternal plants under two nitrogen treatments and two water treatments. For each plant in the summer of 2000, we assayed herbivory by hornworms in both indoor (detached leaf) and outdoor (attached leaf) assays, as well as ambient flea beetle damage. Estimates of leaf material consumed were made via analysis of digitized leaf images. We also assayed leaves for total protein, phenolic, and glycoalkaloid content, and for trypsin inhibitor, polyphenol oxidase, and peroxidase activity. Despite strong effects of nitrogen treatment on growth and reproduction, only total protein responded as predicted by CNB. Leaf phenolic levels were increased by nitrogen treatment, polyphenol oxidase activity was decreased, and other leaf parameters were unaffected. Neither hornworm nor flea beetle herbivory could be related to plant treatment or genotype or to variation in any of the six leaf chemical parameters. A negative relationship between flea beetle and hornworm herbivory was found, but was not apparently mediated by any of the measured leaf chemicals. Because leaf resistance was maintained in low nitrogen plants at the apparent expense of growth and reproduction, our results support the concept of a fitness cost of defense, as predicted by the optimal defense hypothesis.