Interactions between biotic and climatic disturbances are likely to amplify the aggregate effects of disturbance within forests. Mesic forests in the northeastern U.S. are experiencing biotic disturbance from a large set of introduced insects and pathogens concurrent with extreme climate events. One example of concurrent disturbance in northeastern forests is the current outbreak of gypsy moth (Lymantria dispar) in southern New England, which began in 2015 and led to widespread defoliation during a period of severe drought. Disturbance and stress during the growing season reduces primary productivity, however the magnitude and sensitivity of ecosystem responses within northeastern mesic forests are uncertain. Furthermore, for defoliation, the timing of autumn phenology might influence the ability of trees to recover from growing season disturbance, considering that in some forests a later tree leaf-off date could create opportunity to “catch up” on growth. In this study, we combined field-based measurements, satellite data, and ecosystem modeling to quantify the magnitude and uncertainty of the concurrent impacts of defoliation, drought, and phenology on the ecosystem carbon, water, and energy balance during the 2015-2017 set of disturbance events in the northeastern U.S.
Results/Conclusions
Through simulation experiments based on field measurements and satellite data, we investigated the relative contributions of individual disturbance mechanisms in addition to their aggregate effects on northeastern forests. We found that the magnitude and frequency of defoliation across years played a critical role in determining the ecosystem response to drought, with lower carbon uptake, lower evapotranspiration, and more sensible heat flux in more stressed forest areas. However, an extension of fall phenology, i.e., a much later leaf-off date (observed in 2017), did increase tree productivity and the probability of survival in subsequent years in areas with low to moderate disturbance. At the start of the growing season, we found that defoliation phenology with respect to tree carbon allocation phenology is a critical window for predicting the response of forests, which is in line with recent empirical studies. Model-data mismatch and uncertainty reduction approaches guided us in improving the modeled representation of phenology-disturbance interactions, and point to opportunities for future model-data integration of phenology data.