Delayed Tree Mortality After Prescribed Fires in Mixed Oak Forests in Northwestern Ohio

Abstract

Abstract Delayed tree mortality can contribute to variability in fire effects in forests, but its prevalence is not well understood in eastern North American oak forests where a management goal is using prescribed fire to shape forest density and composition. To assess potential delayed mortality after prescribed fires, we tracked the fates of 690 trees of four species in burned and 542 trees in unburned oak forests in northwestern Ohio, USA, and modeled survival using tree diameter and bole char. Delayed mortality, occurring 3–4 growing seasons after fire and in addition to initial mortality (1–2 growing seasons after fire), varied with species and tree diameter. Compared to initial mortality, delayed mortality resulted in eleven times more small-diameter (1–13 cm) red maple (Acer rubrum) dying after fire. White oak (Quercus alba), 1–25 cm in diameter, also incurred delayed mortality (five-times increase in dead trees). Neither sassafras (Sassafras albidum) nor black oak (Quercus velutina) displayed delayed mortality. Background tree mortality in unburned sites was minimal (0.4% per year across species). Logistic regression to model canopy survival selected only stem diameter for burned red maple trees, whereas both diameter and bole char related to survival in other species. Results suggest that (1) monitoring postfire tree mortality in oak forests should extend for at least four growing seasons to detect delayed mortality in some species, and that (2) single surface fires may eventually reduce encroaching red maple in oak forests more than initial postfire years indicate.