Fire and Fire Surrogate Study -

  Southeastern Piedmont

EXPERIMENTAL DESIGN

All sites involved with the FFS study follow the core experimental design as laid out by national protocols. (For details about national protocols, see the proposal on the national FFS web site.) The components of this design include treatments, replication and plot size, and response variables. Any deviations from these protocols are noted and have been approved by the FFS Executive Committee and Science Management Integration Committee (SMIC).

One of four treatments, as defined by FFS protocols, was assigned to each treatment area within a block. (A fifth treatment incorporating herbicide and thinning will be implemented in 2001.) Treatments include:

The levels of thinning and prescribed burning are defined by FFS protocols to be sufficiently heavy so that if a wildfire occurred on a day with weather conditions at the 80^th percentile, 80 percent of the overstory trees would survive. Eightieth-percentile weather conditions during the wildfire season for the Piedmont of South Carolina (February through early April) would include a high temperature of 22^o C, low relative humidity of 34%, and peak 5-minute windspeed of 13 m/sec (NCDC daily observations for Greenville/ Spartanburg airport).

Thinning was conducted by contract in winter of 2000-2001 and was specified as a thinning from below.  Small, merchantable-sized trees and diseased or insect-infested trees were selected first.  Other trees were removed as necessary to provide a residual basal area of approximately 18 m²/ha.  This was a heavy thinning because treatment areas prior to treatment implementation had basal areas ranging from 28 to 37 m²/ha.  Residual slash will remain spread over the treatment area.  

Prescribed burning of each treatment area will be conducted at least 6 weeks after thinning to allow sufficient drying of slash.  Burn-only treatment areas were burned April 2001.  Thin and burn treatment areas will be burned in the spring of 2002.  Strip headfires and flanking fire will be ignited by ground crews.  Weather conditions and strip placement will be selected to produce flames from 2 to 3 m high.  Such fires will likely result in topkill of all understory plants and some trees in the suppressed and intermediate canopy classes.

National protocols indicate "each site … should take necessary measures to maximize the probability of keeping all 3 replications in tact". The minimum size for each treatment area is 14 ha (10-ha measurement area and a surrounding buffer equal to the height of one tree). Large, uniform stands of this size do not exist in the Piedmont. Therefore, rather than using location as the blocking factor, we used stand age. Block 1 consists of pulpwood-sized trees (dbh 15-25 cm); Block 2 has a combination of pulpwood- and sawtimber-sized trees (dbh >25 cm); and Block 3 is dominated by sawtimber-sized trees . Each block has 4 treatment areas associated with it, representing one of the 4 treatments. Treatment areas contain 40 grid points on a 50 by 50 m spacing and range from 14 ha to 27 ha in size. 

Within the treatment area, core response variables are measured for each of the following disciplines:

Ten 0.1 ha plots are located systematically within each treatment unit. Each plot measures 50 by 20 m and is divided into 10 by 10 m subplots. Five of these subplots are used for tree, sapling, and shrub measurements. Variables measured for trees include:

Tree age is measured on one tree randomly selected from each 5 cm dbh class.

Saplings are recorded by species, status, and dbh class. Status is indicated as live, topkilled or harvested. Dbh classes include <3 cm, 3-6 cm, and >6 cm. Shrubs are identified by species and given an estimate of coverage within each subplot.

Tree seedlings and herbaceous vegetation are measured on 1 m² quadrats located in the upper right corner and lower left corner of each 10 by 10 m subplot. Tree seedlings are identified by species and tallied by origin and height class. Origin is indicated as 1^st year, established, or sprout and height classes include <10 cm, 10-50 cm, to >50 cm. Shrubs (<1.4 m tall) and herbs are assigned a cover class (<1%, 1 to 10%, 11 to 25%, 26 to50%, 51 to 75%, >75%) and identified by origin as either germinant, established, or sprout.

Fuel and Fire Behavior

Ground fuels are measured by destructively sampling the forest floor at each grid point. Litter and duff samples are collected from an area of 1 ft² and depth for each layer is measured. To ensure the collection of all organic matter, the duff sample is collected past the soil surface. Each sample is then washed to remove the soil and rock portion. The litter and duff samples are then dried in an oven at 85 EC until a constant weight is reached. One-, 10-, 100-hour fuels, and other component (cones, bark, and other vegetation parts) are separated out of the individual samples. The separation process supplements the woody material inventory by determining the woody component incorporated in the forest floor.

Downed woody fuels are measured using three 50-ft fuel transects located 2 m from every grid point at a randomly selected direction. These transects were established with the outer 2 being 45 degrees apart. Measurements on the outer transects begin at the end farthest from the grid point while the middle transect starts closest to the grid point. Fuels in the 1- and 10-hour class are measured on the first 6 ft of the transect and 100-hour fuels are measured along the first 12 ft. Fuels in the 1000-hour class are measured along the total length of the transect. At 10, 25, and 40 ft the litter depth, duff depth, and fuel depth are recorded.

Fire behavior is recorded for each treatment area by noting flame length, rate of spread, smoldering spread, and fuel moisture. Temperature, relative humidity, and wind speed and direction are noted prior to burning.

Soils and Forest Floor

Mineral soil chemistry is determined by analyzing 12 samples from each 50 by 20 m vegetation plot for carbon, nitrogen, and macronutrients. Soil nutrient availability is measured by comparing mineralization and nitrification of recently collected samples with those that are incubated in the ground for 20-30 days. Bulk density sampling and penetrometer readings are used to determine soil physical properties.

Wildlife

The wildlife discipline consists of 3 components: small mammals, herpetofauna, and avifauna.

Small Mammals - Impacts of the applied treatments on small mammal populations is determined through a combination of live and dead trapping. Sherman, Tomahawk, and snap traps are placed at each grid point. Trapping lasts from 5-10 days on 2 treatment areas within the same repetition. At the end of the trapping period, the traps are moved to another 2 treatment areas of the same rep and this procedure is repeated until are treatment areas are complete. Traps are baited with peanut butter and checked once daily. Any animals captured in live traps are marked and released.

Herpetofauna – Herpetofauna diversity and abundance will be monitored using pitfall traps and modified pitfall traps. Pitfall traps are located between each grid point. Five-gallon buckets are buried in the ground so that the top is flush with the soil surface. Ten pitfall traps were modified by adding three 1-m long sections of 30-cm tall flashing radiating out from the buckets every 120 degrees. Drift fence/pitfall arrays, another modification of the pitfall traps, consist of 3 drift fences 10 m in length and 40 cm in height with buckets placed at the end of each drift fence as well as in the middle of the array. Two drift fence arrays are located in each treatment area. Trapping is conducted on a 10-day period with all traps in one block open at the same time. After all blocks are complete, all traps will remain open for one year. Animals captured in the pitfall traps are marked and released.

Avifauna – Assessment of bird diversity and abundance will be determined using 3-4 point counts per treatment unit. These points are located 200 m apart with detection radii of 50 m. Surveys are conducted for 2-5 minutes and performed 3 times for each point, rather than 6 as indicated by national protocols. Nest searching is conducted to determine nest productivity by monitoring nests and determining number of young successfully fledged. Two 1-hour focal observations of woodpeckers and other bark-gleaners are conducted for each treatment area. Time spent foraging, tree species, dbh, tree condition, and length of fire scar are noted. Woodpecker activity on snags and trees affected by bark beetles is recorded. Microhabitat vegetation sampling is also conducted at each nest location and each tree used for foraging.

Entomology

Incidence of Ips beetles and southern pine beetles (Dendroctonus frontalis, Zimmerman) are of particular importance in this region. Risk- and hazard-rating, beetle trapping and monitoring of resin flow and tree growth are conducted to determine the short-term and long-term impact of thinning and prescribed fire, alone and in combination. Risk-rating indicates the probability of southern pine beetle infestation per acre and hazard-rating identifies the potential of southern pine beetle infestation growth. Eight traps are located in each treatment area at a height of 40 ft. Traps are checked once a week from April until November. 

Pathology

The national protocol focuses on Ophiostomoid fungi as the key species of interest. We have expanded our fieldwork to include P. cinnamomi and H. annosum. Trees sampled for the above stated pathogens are restricted to pine species located within the vegetation plots. Three trees are sampled per vegetation plot: two trees with disease (pitch canker, littleleaf, butt rot, thin top) or incidence of southern pine beetle and one control. For each tree, 2 roots with a minimum diameter of 5 cm are exposed and three 5-mm cores are taken from the upper face of each root. One-mm cross sections of the root cores are placed on agar plates with and without cycloheximide and incubated for at least 10 days at 21-22 degrees C. Incidence of fungal colonies are inspected and identified.

Treatment Costs and Utilization Economics

Operational hours of equipment used for tree harvest are recorded by electronic dataloggers mounted on the machinery. Paper records are also kept for all work performed. Scale tickets are kept from harvest production and separated by treatment area.

Costs associated with burning are estimated based on an expert opinion methodology. Analysis is based on information provided by individuals knowledgeable about burning under local conditions on similar plots of lands with units sized and staffed for operational treatments.