Cross-scale influences on jack pine budworm abundance during an outbreak: Local impacts by host trees and landscape-scale synchronicity of parasitoid attacks

A jack pine budworm (Choristoneura pinus) outbreak in northwestern Wisconsin was monitored from 2004 to 2006 employing an array of hierarchically nested sampling plots to explore the factors influencing changes in late-larval abundance and rates of parasitoid attack at multiple spatial scales. Each year, late-instar larval counts were documented on 192 individual branch tips divided among four sites with two transects at each site and three canopy type plots (closed, edge and open) in each transect. Larvae were collected and then reared in the lab to assess parasitism rates. Change in larval abundance from 2004 to 2005 differed significantly by site (hierarchical ANOVA, p<0.0096); and change from 2005 to 2006 differed by both site and plot canopy type (hierarchical ANOVA, p<0.0078 and p<0.0003). Parasitism rates showed no significant differences at any level in any year but increased synchronously across the landscape from an average rate of 11% in 2004 to 26% in 2006 regardless of larval abundance, defoliation levels or occurrence in edge plots. Changes in the larval budworm population were largely influenced by the physiological response of individual trees to defoliation at the plot-level scale. However, peak population numbers are expected to be reduced in 2007 by the action of parasitoid wasps and flies which are operating synchronously in many sites across the landscape. The extensive spatial scale of parasitoid response to elevated budworm population levels may account for the delay in significant mortality from parasitoid attack and the persistence of the outbreak phase for more than three years.