The gypsy moth (Lymantria dispar) was once a major defoliator in the eastern USA but has been quiescent for nearly thirty years. A surprisingly severe gypsy moth outbreak began in 2015 across southern New England, resulting in dramatic oak (Quercus spp.) mortality. We leveraged the natural experimental gradient of defoliation frequency (0-3 years of defoliation) to address questions about the causes of tree mortality:
1) How does mortality vary by defoliation frequency, species, and tree size?
2) Does carbon starvation cause tree mortality?
We established 10 permanent 0.12 ha plots in mixed oak forests in central Massachusetts that were defoliated 0, 1, 2, or 3 years. Defoliation history was inferred from Landsat imagery and verified in the field. We recorded tree species, condition (live or dead), crown dieback (percent dead branches), defoliation (percent leaves eaten), and size (diameter at breast height). In February 2019 and 2020, we collected stem and root samples from oak trees with varying defoliation frequency (0-3 years) and severity (percent defoliation in 2018 and 2019) from these plots (n=41 trees) and a study of road-side trees (n=47 trees) and then extracted labile carbohydrate concentrations (bulk sugars and starch) in the lab.
Results/Conclusions
Mortality increased with increasing defoliation frequency. Average plot-level biomass loss to mortality was 12% for plots defoliated 1 year, 25% for plots defoliated 2 years, and 62% for plots defoliated 3 years. Although gypsy moth is considered a generalist, nearly all trees that died were oaks. Despite past work indicating that smaller-diameter trees were more susceptible to mortality, we found that oaks of all sizes were equally likely to die. Forests in the region are generally older now than they were during past outbreaks, which may result in different susceptibility patterns by size.
Defoliation in 2019 was minimal. In the aftermath of the outbreak, we found that oaks either died or survived with minor (<25%) crown dieback; few of the surviving oaks exhibited major (25-99%) crown dieback. Therefore, we expect that the productivity of surviving trees will rebound within a few years, and that carbohydrate levels of surviving trees will have recovered in 2020, but that carbohydrate levels of trees that died will be below a critical threshold. All cores for carbohydrate analysis have been collected and pre-processed for lab analysis, which will be completed in Spring 2020. Our results will improve the mechanistic understanding of plant mortality under insect outbreaks.