A diverse array of sublethal plant secondary compounds are commonly found i
n the foliage of temperate deciduous trees. These traits are thought to def
end a plant in two principal ways, either directly by reducing insect ovipo
sition, feeding, reducing insect oviposition, feeding, or biomass gain, or
indirectly, through digestive inhibition. Such inhibition is hypothesized t
o slow the rate of herbivore development, thereby increasing their suscepti
bility to natural enemies (the slow-growth-high-mortality hypothesis). To c
larify the defensive role of these compounds, field experiments were conduc
ted to examine the relationships among oak leaf quality, herbivore family,
and three herbivore performance measures: survivorship, development time, a
nd pupal mass, for a bivoltine leaf-tying caterpillar, Psilocorsis quercice
lla (Lepidoptera: Oecophoridae). Two experiments, one for each generation o
f the insect, were conducted to examine the effects of intraspecific variat
ion in leaf chemistry of its host, white oak trees (Quercus alba). In each
experiment, full-sib neonate larvae were placed in experimental leaf ties o
n high- versus low-quality trees and allowed to feed for 2 weeks under fiel
d conditions. To determine the effect of the third trophic level, a portion
of each family in each leaf-quality treatment was bagged to prevent attack
from natural enemies. This treatment also allowed us to test a prediction
of the slow-growth-high-mortality hypothesis, i.e., that development time,
as measured for full sibs in the bagged treatment, should be positively cor
related with mortality of their full sibs exposed to natural enemies. Low l
eaf quality significantly reduced survivorship of the caterpillars in the f
irst generation but not the second. The third trophic level decreased survi
vorship in both generations. Larval development time was not affected by le
af quality in either generation, but varied significantly among insect fami
lies in both generations. In turn, larvae from slower-developing families d
id not suffer increased predation and parasitism, as predicted by the slow-
growth-high mortality hypothesis. In contrast to develop ment time, pupal m
ass showed a greater response to intraspecific variation in leaf quality, a
lthough the effect was only significant in generation 1. Concentrations of
both total phenolics and hydrolyzable tannins in Q. alba foliage appear to
be important negative predictors of pupal mass in P. quercicella. In marked
contrast to development time, no main family effect was found for pupal ma
ss in either experiment; however, significant familyxenvironment interactio
ns were found for the effects of the bagging treatment (generation 1) and t
he leaf-quality treatment (generation 2). Overall, the first trophic level
had a greater influence on pupal mass (a fecundity correlate), while larval
development time was determined more by the insect's family (genotype+mate
rnal environment). The third trophic level was a consistently strong source
of mortality in both experiments, but as a whole did not respond to famili
al differences in development time. Thus, from the perspective of P. querci
cella, plant quality appears to serve as a defense more through its direct
effect on herbivore survivorship and fecundity than through an indirect eff
ect on predation via changes in development time.