Effects of herbaceous and woody plant control on longleaf pine growth and understory plant cover

fine sandy loam (fine-loamy, siliceous, semiactive, thermic Typic Paleudult) with a slope of 1–10% (Kerr et al. 1980). The other complex (92° 38 W, 31° 8 N at 66 m above sea level) is comprised of Beauregard silt loam and Malbis fine sandy loam (fine-loamy, siliceous, subactive, thermic Plinthic Paleudult) with a slope of 1–5%. Before harvesting, Study 2 was a closed canopy, mature, loblolly pine (P. taeda L.)-hardwood forest. The understory vegetation was mostly hardwood trees, shrubs, and vines and scattered shade tolerant herbaceous plants. The study sites are within the humid, temperate, coastal plain and flatwoods province of the West Gulf Region of the southeastern United States (McNab and Avers 1994). The climate is subtropical. During the 12-year period, December had the lowest average mean temperature of 10.3° C and August had the highest average mean temperature of 28.2° C (National Climatic Data Center 2012). Annual precipitation averaged 1,463 mm with 1,059 mm during the growing season, which included the months of March through November. Both studies are on uplands suitable for restoring longleaf pine forests (Turner et al. 1999). Treatment Establishment In Study 1, the vegetation was rotary mowed and the large woody debris was hand cleared in June 1997. In Study 2, the mature loblolly pine-hardwood forest on both complexes was clearcut harvested in 1996, roller drum chopped, and prescribed fire was applied by October 1997. Primarily grasses dominated the plant community in Study 1, and trees and shrubs dominated the plant community in Study 2 for the next 6 years (Haywood 2005). On plots that were only prescribed burned (checks), 1st-year herbaceous plant mass was 2,058 kg/ha oven-dried weight at Study 1 and 1,055 kg/ha at Study 2. After 4 years, tree and shrub stocking was 18,031 stems/ha with an average total height of 0.06 m and crown width of 0.03 m at Study 1, whereas at Study 2, stocking was 29,270 stems/ha with an average total height of 0.6 m and crown width of 0.3 m. In 1997, four treatments were randomly assigned to the research plots in a randomized complete block design (Steel and Torrie 1980)—check, herbaceous plant control (HPC), woody plant control (WPC), and HPC WPC. In both studies, the 16 research plots (four blocks by four treatments) each measured 22 by 22 m (0.048 ha) and contained 12 rows of 12 seedlings arranged in a 1.83by 1.83-m spacing. The center 64 longleaf pine seedlings (eight rows of eight seedlings each) were the measurement plot. In Study 1, blocking was based on soils with two blocks established on each soil type. In Study 2, blocking was by complex (two blocks on each soil complex) and topographic location within each complex. Longleaf pine seeds from a standard Louisiana seed source were sowed in containers in May 1997. The 28-week-old seedlings were planted on both sites in November 1997 using a planting dibble with a tip of the correct size and shape for the 3.8-cm-wide and 14-cm-deep root plug. Two herbicides were used for HPC: sethoxydim (2-[1(ethoxyimino)butyl]-5-[2-(ethylthio)propyl]-3-hydroxy-2-cyclohexen-1-one) for controlling bluestem grasses and hexazinone (3cyclohexyl-6-[dimethylamino]-1-methyl-1,3,5-triazine-2,4[1H, 3H]-dione) for general herbaceous plant control. In April of 1998 and 1999, the two herbicides were applied in 0.9-m bands over the rows of unshielded longleaf pine seedlings at Study 1. Within the 0.9-m bands, the rate of sethoxydim was 0.37 kg active ingredient (ai)/ha, and for hexazinone, the rate was 1.12 kg ai/ha. At Study 2, only hexazinone was banded in April 1998 and 1999 because sethoxydim was not needed for bluestem grass control. In both studies, WPC was done with triclopyr ([(3,5,6trichloro-2-pyridinyl)oxy]acetic acid) at 0.0048 kg acid equivalent/ liter. The triclopyr was tank-mixed with surfactant and water and applied as a directed foliar spray to hardwood trees and shrubs in April 1998. In Study 2, the brush was retreated in June 1999, but Study 1 did not need retreating because an intense prescribed fire earlier in May 1999 top-killed most of the woody vegetation. Recovering brush was hand-felled in February 2001 at both studies. Prescribed fire was routinely applied in both studies as a normal management practice. Fire management personnel with the KNF first set backfires to secure the boundaries of each site or complex. Then, the fire crew would set striphead fires with drip torches or spot fires using a helicopter-mounted ignition system until the entire site or complex was burned. Prescribed fire was applied at Study 1 in May 1999 (18 months after planting), April 2001, May 2003, June 2005, June 2007, and May 2009. All six were intense fires, which are common in established grass rough (Haywood 2009, 2011). In Study 2, the first prescribed fire was delayed until June 2000 (31 months after planting) because of a lack of grass development and subsequent poor fuel bed conditions. A wildfire in January 2003 burned Blocks 3 and 4, but the longleaf pines survived because this species commonly endures high-fire intensities (Haywood 2009, 2011). Prescribed fire was applied to the other two blocks in May 2003. The next three fires were set in May 2005, June 2007, and May 2009. Sampling Longleaf pine tree total height and diameter at breast height (dbh) measurements were taken at ages 7 and 12 years. Heights were measured with a calibrated pole at age 7 and with a laser instrument (Criterion 400 Survey Laser, Laser Technology, Inc., Centennial, CO) at age 12. Tree dbh was measured with a diameter tape. Total height and dbh were used to calculate outside-bark bole volume with Baldwin and Saucier’s (1983) formulas. In September of the thirteenth growing season, percent cover of litter and understory vegetation was estimated as five different taxagrasses, forbs (which included grasslike-plants and ferns), trees, shrubs (which included blackberry [Rubus spp.]), and woody vines. The measurements were taken at five 1.83by 1.83-m squares whose corners were the original planting locations for the longleaf pine seedlings. A square was located in the middle of each plot and in the center of each quarter section of the plot. Data Analysis In each study, number of longleaf pine per ha, average total height, basal area, and volume per tree, and basal area and volume per ha were compared among the four treatments using a randomized complete block design model at 0.05 (SAS Institute, Inc., 1985). Analyses compared treatments at ages 7 and 12 years and the difference in growth and production over the 5-year period. Percent cover of litter in the thirteenth growing season was analyzed with the same model. However, percent understory plant cover was analyzed with an analysis of covariance model in which the covariate for tree and shrub cover was tree and shrub cover in the fourth growing season. For woody vines, the covariate was the number of vines per ha in the fourth growing season. For grass and forbs, the covariates SOUTH. J. APPL. FOR. 37(2) 2013 109 were percent cover in the third growing season for Study 1 and fourth growing season for Study 2. If there were significant differences among the four treatments, mean comparisons were made with Tukey’s Studentized Range Test at 0.05. Percentages were arcsine transformed before analysis to equalize variances (Steel and Torrie 1980). Results Longleaf Pine In Study 1, longleaf pine total height on HPC WPC plots was significantly greater than on checks and WPC plots after seven growing seasons (Table 1). The treatment combination also resulted in greater basal area and volume per tree than on checks. However, after 12 growing seasons, the treatment combination had greater pine total height compared only to checks. There were no treatment differences in basal area and volume per tree at age 12 years, and changes in total height, basal area, and volume per tree over the 5-year period were not significantly different among the four treatments. After 12 years, total height ranged from about 7 m on checks to 8 m on HPC WPC plots, basal area per tree ranged from 0.6 dm on checks to 0.7 dm on HPC WPC plots, and volume per tree ranged from 25 dm on checks to 34 dm on HPC WPC plots. In Study 2, longleaf pine total height, basal area, and volume per tree were significantly greater on HPC WPC plots than on checks and WPC plots, and total height on HPC plots was greater than on checks after seven growing seasons (Table 1). At age 12 years, total height was greater on HPC WPC plots than on checks, and the treatment combination had greater basal area and volume per tree than the HPC plots. Change in basal area per tree over the 5-year period was greater on checks than on HPC plots. After 12 years, total height ranged from about 11 m on checks to 12 m on HPC WPC plots, basal area per tree ranged from 1.2 dm on HPC plots to 1.4 dm on HPC WPC plots, and volume per tree ranged from 69 dm on HPC plots to 85 dm on HPC WPC plots. Longleaf pine stocking after 12 growing seasons was not significantly affected by postplant vegetation control in either study (Table 2). Survival ranged from 60% on checks to 69% on HPC WPC plots in Study 1 and 57% on checks to 71% on HPC plots in Study 2. Most mortality occurred in the first growing season (Haywood 2005), and survival decreased by only three percentage points in both studies from the sixth through twelfth growing seasons. In Study 1, pine basal area and volume per ha were significantly greater on HPC WPC plots than on checks at age 7 and 12 years (Table 2). Likewise, change in basal area and volume per ha over the 5-year period was greater on the treatment combination than on checks. The HPC treatment also resulted in greater basal area and volume per ha than on checks after 12 growing seasons, and change in volume per ha over the 5-year period was greater on HPC plots than on checks. After 12 years, basal area per ha ranged from about 11 m/ha on checks to 15 m/ha on HPC