Abstract
In the summer of 1980, gypsy moth populations were monitored in 13 sparsely infested forests in southwestern Quebec; counts of living and dead larvae and pupae were made on 1,870 trees representing 28 deciduous and one coniferous species. Contrary to our null expectations, the proportionate numbers of pupae compared with larvae on the various host species were not equal. Hosts more preferred by larvae were less preferred by pupae, and vice versa. For example, pupae were disproportionately abundant on host species like Acer pensylvanicum L., Carya ovata (Mill.) K. Koch, and Juglans cinerea L. which are not generally favored larval hosts. Conversely, favored larval hosts like Quercus rubra L. and Ostrya virginiana (Mill.) K. Koch carried lower numbers of pupae than expected from the numbers of larvae feeding on them. Such differential utilization of host trees by larvae versus pupae, which can arise either from host-dependent differences in larval mortality or from late instar migration between hosts, may contribute to maintaining the broad polyphagy of gypsy moth larvae.