The frequency and severity of disturbances in forested ecosystems is expected to increase under climate change, potentially driving substantial increases in disturbance interactions. Interactions between disturbances can result in compounding effects on forest growth and mortality, but the influence of temporal patterns on disturbance interaction outcomes is not well understood. Southern New England has recently experienced an unprecedented tree mortality event related to severe drought in 2016 and widespread gypsy moth defoliation from 2016-2018. Our objective was to assess how defoliation severity and timing (relative to drought) influence stand growth and tree mortality outcomes. Based on remotely-sensed assessment of defoliation timing, we selected 27 study sites across eastern Connecticut and central Massachusetts from existing sampling locations with field-based assessments of gypsy moth defoliation, caterpillar populations, and egg mass counts. At a site level, remotely sensed imagery was used to characterize within-season patterns of defoliation, vegetation response, and mortality timing and extent. Across all study sites, increment cores were collected from all live and dead canopy trees, including preferred and non-preferred gypsy moth host species. Dendrochronological techniques were used to assess tree and stand level growth patterns and mortality timing across a range of defoliation patterns and initial drought severity.
Results/Conclusions
Based on remote sensing image time series analysis, existing study sites were classified into six classes describing the temporal pattern of gypsy moth defoliation – none (3 sites), 2017 only (13 sites), 2018 only (2 sites), 2016 and 2017 (2 sites), 2017 and 2018 (3 sites), and all 3 years (4 sites). Field observations indicated that 81% (22 of 27) of sites experienced tree mortality, with overall mortality rates ranging from 3-70% and oak mortality rates ranging from 11-90%. Both overall and oak-specific mortality rates varied significantly among defoliation timing categories based on analysis of variance (p = 0.024 of overall mortality and p = 0.002 for oak-specific mortality), with significantly greater mortality in sites undergoing multiple years of defoliation. Analysis of increment core data illustrates site and species-specific variation in growth and mortality of individual trees and at the stand level, and general patterns arising from these analyses will be detailed. Results of this study will further understanding of the effects of temporal patterns of interacting disturbances on tree mortality and stand productivity, and have the potential to improve models focused on predicting the effects of disturbance and disturbance interactions on ecosystem functioning and resilience.