Predicting Post-Fire Tree Survival for Restoring Oak Ecosystems

Reducing tree encroachment is challenging for restoring light-requiring habitats including prairies, savannas, and woodlands. Such ecosystems frequently support high biodiversity and are of keen interest for restoration (Haney et al. 2008). Periodic fires often limited tree recruitment during the evolutionary history of these ecosystems, maintaining open conditions (Hutchinson et al. 2012). Prescribed fire is a restoration strategy for reducing tree encroachment that can be cost-effective (compared to mechanical cutting) and have other advantages, such as avoiding soil disturbance from mechanized equipment. However, the effectiveness of prescribed fire hinges on its ability to selectively remove undesired trees. Knowledge of the largest trees that fires can top-kill could aid restoration planning for identifying sites amenable to fire management, versus sites where trees are likely too large for fire to be effective.

In eastern North America, one of the most extensive restoration goals is reestablishing open-structured oak ecosystems, including a continuum of habitat from oak savanna to open forests maintained through fire (Dey and Hartman 2005). During the last century, non-oak tree species such as Acer rubrum (red maple) have formed dense understories below oak canopies due to the cessation of fires ignited by Native Americans and early Euro-American settlers, the fragmentation of landscape fuel continuity, fire suppression, and the alteration of fuel types. Small stems of encroaching non-oak species are generally intolerant to fire, but their resistance increases with size (Hruska and Ebinger 1995). Moreover, in open sites during fire-free periods, high densities of oaks can

Figure 1.

Examples of burn sites in forest (left) and oak savanna (right) in Oak Openings Preserve, northwestern Ohio, USA. The center foreground around the measuring tape in the left photo shows top killed and resprouting Acer rubrum trees that were < 13 cm in diameter at 1.4 m (DBH). The larger tree in the left foreground is a 32 cm DBH Quercus velutina. The photo on the right shows top killed and resprouting small Q. velutina (< 3 cm DBH), while Q. velutina ranging from 4–9 cm DBH (center and left-center of the photo) had live canopies. Photos by S.R. Abella, August 2018, five months after April 2018 prescribed fires.

attain sizes resistant to most fires, potentially undermining restoration goals of maintaining open structure (Cole et al. 1992). While there is a small, slowly growing body of research modeling post-fire survival based on tree size in eastern hardwoods (e.g., Loomis 1973, Regelbrugge and Smith 1994, Dey and Hartman 2005, Keyser et al. 2018), predicting the sizes of trees killable by restoration burns across the diversity of species, sites, and fire severities in this biome remains challenging.

Our objective was to develop models using stem diameter and trunk scorch to predict tree survival following restoration burns. We conducted the study in the 1497-ha Oak Openings Preserve, managed by the Metroparks of the Toledo Area, within the 40,000-ha Oak Openings region in northwestern Ohio, USA. Climate is temperate with warm, humid summers and cold winters. Summer precipitation (May through August), measured 10 km from the preserve, averaged 34 cm/year from 1955 through 2018 (Toledo Express Airport weather station; National Oceanic and Atmospheric Administration, Asheville, North Carolina). During 1817–1832 land surveys, the region including the preserve supported a fire-maintained mixture of prairie, oak savanna, and oak woodland (Brewer and Vankat 2004). Much of the region, now long without fire, supports closed-canopy forest with overstory Quercus velutina (black oak) and Quercus alba (white oak) and dense understories of non-oak trees (e.g., A. rubrum, Sassafras albidum [sassafras]). Non-oak trees were present but rare during the pre-Euro-American settlement land surveys. Based on the 1817–1832 reference conditions and a goal of stimulating groundlayer plant diversity, the ecological restoration objective in this study was to reestablish open-structured oak ecosystems with ≤ 50% canopy cover, reduce the density of non-oak trees, and reinstate fire as a process maintaining oak ecosystems.

In August 2018, we sampled four sites...