Conspecific trees growing at high and low-elevations encounter different gr
owing conditions and may vary in their suitability as hosts for herbivorous
insects. Mountain tree populations may be more resistant to herbivory if l
ow temperatures constrain growth more than they constrain photosynthesis, r
esulting in increased secondary metabolism (temperature hypothesis). Altern
atively, mountain trees may be fertilized by atmospheric nitrogen depositio
n and become more palatable to insects (atmospheric deposition hypothesis).
We evaluated these two hypotheses by comparing high-and low-elevation tree
s with insect bioassays and analyses of foliar nitrogen and condensed tanni
n. Contrary to the temperature hypothesis, high-elevation foliage had highe
r leaf nitrogen (six of six tree species) and allowed higher growth rates o
f Lymantria dispar larvae (five of six tree species). The nitrogen depositi
on hypothesis was broadly supported by measurements from two mountains show
ing that high-elevation trees tended to have higher leaf nitrogen, lower le
af tannins, and support higher insect growth performance than conspecific t
rees from lower elevations. The deposition hypothesis was further supported
by fertilization studies showing that simulated atmospheric nitrogen depos
ition changed the foliar chemistry of valley trees to resemble that of high
-elevation trees. Predictions that the altitudinal gradient in foliar chemi
stry and host suitability should be steepest on mountains receiving more de
position were largely not supported, but interpretations are complicated by
lack of replication among mountains. In the northeastern United States, in
creased host suitability of high-elevation trees seems sufficient to influe
nce the population dynamics and community composition of herbivores. Atmosp
heric nitrogen deposition offers a promising hypothesis to explain and pred
ict some important spatial patterns in herbivory.