Indirect Interactions Between Temporally Separated Herbivores Mediated by the Host Plant

I tested the hypothesis that early season herbivory by leaf—chewing insects affects distribution, densities, and survivorship of late—feeding, leaf—mining insects; early chewing may physically and chemically alter quality of leaf miners on the shared host plant, Quercus emoryi. Proportions of intact leaves and leaves damaged by native leaf chewers on six control and six experimental trees were determined over two growing seasons. I manually damaged °50% of leaves on experimental trees to increase the total fraction of damaged leaves to °75%. Leaf—miner densities, distribution, survivorship, and mortality were monitored on control and experimental trees and in intact and damaged leaves within trees. Leaf miners occurred more frequently than expected by chance on intact than damaged leaves for both control and experimental trees. Leaf—miner densities did not differ between control and experimental trees. This result suggests ovipositing leaf miners did not discriminate between damaged and undamaged trees, but selected leaves within trees. More leaf miners survived intact than damaged leaves in both growing seasons. Survivorship was less in damaged leaves because of significantly increased rates of parasitism on these leaves. However, death from other causes (including bacterial and fungal death) was significantly less for leaf miners on damaged leaves. This positive effect did not compensate for overall lower survivorship of leaf miners in damaged leaves due to increased parasitism. Survivorship and mortality did not differ between control and experiment trees when damaged and intact leaves were pooled within treatments. However, in 1982—1983, survival was lower and parasitism greater for leaf miners in control—damaged leaves than experimental—damaged leaves. Survival and parasitism were not different between control—intact and experimental—intact leaves. These result suggest that (1) the effect of early season herbivory on leaf miners was inversely density dependent and, (2) occurrence of leaf miners on intact leaves may be caused by avoidance of damaged leaves rather than preference for intact leaves. Early season herbivory caused localized changes in phytochemistry within trees, and these chemical alterations were consistent with observed distributions and survivorship of leaf miners. Both early season herbivory and experimental damage resulted in higher levels of condensed tannins and lower protein content in damaged leaves within trees, but herbivory had no effect on between—tree chemistry differences. Leaf miner distributions correspond to these localized chemical changes: leaf miners avoided damaged leaves within trees but showed no between—tree preferences. Parasitism of miners on damaged leaves was higher, possibly because parasitoids used physical and chemical changes as cues to locate leaf—miner hosts, or because exposure of leaf miners to parasitoids was prolonged. However, when many leaves were damaged (experimental trees), the negative effect of parasites was mitigated because physical and chemical cues associated with damaged leaves were less effective. Lower leaf—miner mortality from other causes may be related to the bactericidal of fungicidal properties of increased tannins in damaged leaves. This study demonstrates that temporally separated guilds can interact subtly at low levels of herbivory through changes in the host plant. Current theories of within— and between—guild organization of phytophagous insects may need to be re—evaluated if such interactions are common.