Effects of Oak Woodland Restoration Treatments on Sapling Survival and Tree Recruitment of Oaks in an Upland Mesic Oak-dominated Forest

Oak-dominated (Quercus spp.) woodlands, defined here as mixed woody and diverse herbaceous plant communities with > 70% tree canopy containing few non-oaks (see also Taft 1997), comprise a small fraction of the area in North America occupied prior to fire suppression. Historically, frequent low intensity fires maintained oak-dominated forests, woodlands, and savannas across the Midwest and interior South of the United States (Anderson and Bowles 1999, Fralish et al. 1999, Heikens 1999, Brewer 2001, Van Lear 2004, Ruffner and Groninger 2006). Fire exclusion in the 20^th century enabled fire-sensitive hardwoods to colonize previously fire-maintained oak woodlands (Hart et al. 2008, Nowacki and Abrams 2008). These tree species produced a more closed canopy (> 90%) than the historically open, sparse canopies of oak woodlands and savannas (Bowles and McBride 1998), leading to widespread oak regeneration failure and losses of groundcover plant diversity (Abrams 1992, Bowles and McBride 1998). These diverse oak communities are rare ecosystems (Anderson and Bowles 1999) and are thought to occupy less than 0.02% of the area in North America they occupied before fire exclusion (Nuzzo 1986).

One challenging aspect of oak woodland management and restoration is identifying prescribed fire regimes that will effectively promote oak regeneration and maintain groundcover plant diversity. Effective methods for promoting oak regeneration involve opening the canopy sufficiently to foster growth of an existing oak seedling/sprout layer and then, when (or if) necessary, implementing fire in such a way as to reduce competition with non-oak saplings but not cause excessive oak sapling mortality (Loftis 1990, Kruger 1997, Brose et al. 1999, Albrecht and McCarthy 2006, Iverson et al. 2008, Cannon and Brewer 2013). It remains unclear, however, whether the fire regimes necessary to maximize oak regeneration are compatible with the maintenance of groundcover plant diversity in oak woodlands. To the extent that fire regimes that promote oak regeneration conflict with the maintenance of groundcover plant diversity, practitioners must either implement fire regimes that create a mosaic of oak regeneration patches interspersed with patches of diverse groundcover vegetation or they must identify fire regimes that forge a compromise between oak sapling survival and the maintenance of groundcover plant diversity.

This study examined the effects of frequent fires proven to promote groundcover plant diversity in a mesic oak-dominated system (Brewer and Menzel 2009, K.S. Spiegel and J.S. Brewer, University of Mississippi, unpubl. data) on the survival of established oak sapling recruits and on oak tree recruitment. I tested two hypotheses: 1) frequent (biennial) burning of an established sapling layer in persistent gaps reduces sapling densities, including those of upland oak species; and 2) frequent burning prevents recruitment of tree-sized individuals from saplings (oak and non-oak).

In 2003, I established an oak woodland restoration experiment at Strawberry Plains Audubon Center (SPAC) in the loess plains of north-central Mississippi, a region characterized by gently rolling hills with moderately fertile, mesic silt and sandy loams in the uplands and floodplains. The primary objectives of the restoration experiment (i.e., the reference model; sensu Clewell and Aronson 2013) were 1) to increase the abundance and diversity of open oak woodland groundcover vegetation and 2) to restore natural regeneration of upland oak species (see Brewer 2001 and Surrette et al. 2008 for evidence that upland communities in this region were historically open oak woodlands). The current study examines the latter objective.

Figure 1.

Diagram of the two sites at Strawberry Plains Audubon Center, MS, USA, showing the layout of sapling and tree plots and the treatment arrangement and description. Sapling and tree plots at Site 1 were established in 2003; first burned and thinned in 2004; subsequently burned in 2006, 2008, 2010, and 2012. Sapling plot at the edge of a control area was damaged in 2009 and abandoned in 2010. Sapling plots at Site 2 were established in 2007; first burned and thinned in 2008, subsequently burned in 2010 and 2012. Tree plots (not shown) at Site 2 were established in...