Although wound-induced responses in plants are widespread, neither the
ecological nor the evolutionary significance of phytochemical inducti
on is clear. Several studies have shown, for example, that induced res
ponses can act against both plant pathogens and herbivores simultaneou
sly. We present the first evidence that phytochemical induction can in
hibit a pathogen of the herbivore responsible for the defoliation. In
1990, we generated leaf damage by enclosing gypsy moth larvae on branc
hes of red oak trees. We then inoculated a second cohort of larvae wit
h a nuclear polyhedrosis virus (LdNPV) on foliage from the damaged bra
nches. Larvae were less susceptible to virus consumed on foliage from
branches with increasing levels of defoliation, and with higher concen
trations of gallotannin. Defoliation itself was not related to any of
our chemistry measures. Field sampling supported the results of our ex
periments: death from virus among feral larvae collected from unmanipu
lated trees was also negatively correlated with defoliation. In 1991,
defoliation and gallotannin were again found to inhibit the virus. In
addition, gallotannin concentrations were found to be positively corre
lated with defoliation the previous year. Compared with previous resul
ts that demonstrated a deleterious effect of induction on gypsy moth p
upal weight and fecundity, the inhibition of the virus should confer a
n advantage to the gypsy moth. Since leaf damage levels increase as gy
psy moth density increases, and since leaf damage inhibits the gypsy m
oth virus, there is the potential for positive feedback in the system.
If phytochemical induction in red oak can inhibit an animal pathogen
such as LdNPV, it suggests to us that induction in red oak is a genera
lized response to tissue damage rather than an adaptive defense agains
t herbivores.