Effects of Wildfires and Liming of Pine-oak-heath Communities in the Linville Gorge Wilderness, Western North Carolina

Linville Gorge Wilderness (LGW) is a Class I area in the southern Appalachian Mountains, western North Carolina. Over the last 150 years, LGW has been subject to several wildfires, varying in intensity and extent (Newell and Peet 1995). In November 2000, a wildfire burned 4000 ha in the wilderness; the fire ranged in severity across the northern portion of the wilderness from low severity in coves to high severity along ridges and bluffs (Reilly and others 2006). In May 2007, the Pinnacle wildfire burned ca. 2000 ha in the southern portion of the wilderness. A large portion of the Pinnacle fire overlapped the area previously burned in 2000, resulting in much of the central section of LGW being burned twice in less than 7 years. In addition, dolomitic lime was aerially applied at a rate of 1120 kg ha-1 on the most severely burned area. We hypothesized that liming the most severely burned area would accelerate the restoration of acidic, nutrient depleted soils by adding basic cations, balancing soil pH, reducing soil and soil solution Al, and subsequently increase ecosystem productivity. We sampled vegetation (composition, biomass, and foliar nutrients) and soil nutrients in five treatment areas in LGW over a 2-year period following the most recent wildfire. The treatment areas were: severely burned twice (2000 and 2007) plus dolomitic lime application (2xSBL); moderately burned twice plus lime (2xMBL); severely burned twice, no lime (2xSB); moderately burned once (2000), no lime (1xMB); and an unburned and unlimed reference area (REF). All wildfire burned sites experienced overstory mortality (>300 stems ha-1). There were no live overstory trees on 2xSBL, and 2xSB had a larger number of dead trees than all other sites. The large number of dead pines on all sites, including the reference, was due in large part to a southern pine beetle (Dendroctonus frontalis Zimm.) outbreak in 1999-2001. The three areas impacted by the recent wildfire had little to no live pine in the overstory. Overstory biomass was lower on the severely burned areas (2xSBL and 2xSB) than the other treatments, with no significant change between 2008 and 2009 (Fig. 1). By 2009, understory density was higher on 2xSBL than the other treatments (Fig. 2a), with higher numbers of tree species (Fig. 2b). We found no differences in shrub density among burned treatments, and only 2xSB had higher shrub density than REF (Fig. 2c). As expected, ericaceous (heath) species sprouted after fire, and their density increased between 2008 and 2009 for all recently burned treatments. We found no differences in foliar nitrogen (N) concentration among treatments for either evergreen or deciduous species (Table 1). Evergreen foliar calcium (Ca) was greater on 2xMBL and REF than the severely burned sites (2xSBL and 2xSB); whereas deciduous foliar Ca and phosphorus (P) were greater on 2xMBL than 2xSBL, 1xMB and REF (Table 1). With few exceptions, we found significant date, treatment, date x treatment interaction effects for soil exchangeable cations for both shallow (0-10 cm