Effects of Application Rate, Timing, and Formulation of Glyphosate and Triclopyr on Control of Chinese Privet (Ligustrum sinense)l

Chinese privet is a nonnative shrub that has invaded mesic forests throughout the southeastern United States during the past century. Foliar sprays of glyphosate and triclopyr were tested in three factorial experiments that included wide ranges of application rate, timing, and formulation to refine methods for controlling Chinese privet. For spring (April) and fall (October and December) applications, percentage control of privet cover averaged 93 to 100% and 49 to 70% for glyphosate and triclopyr treatments, respectively, whereas for summer (June and August) applications, control averaged 67 to 69% and 14 to 26%, respectively (study 1). However, privet control was not influenced by variation in herbicide rates of 1.7, 3.4, 5.0, or 6.7 kg aeha compared with each of the five application timings. No differences were found in August comparisons of liquid vs. dry glyphosate products or water-soluble vs. oil-soluble triclopyr products for each of the four rates (study 2). In a comparison of low rates of glyphosate applied in August with or without trenching of plot perimeters to isolate privet clumps (study 3), control increased from 12 to 65% as rate increased from 0 to 0.8 kg aeha, suggesting that rate responses may occur at lower values than those tested in studies 1 and 2. Isolation of privet clumps by trenching did not have a statistically detectable effect on privet susceptibility to glyphosate. Low rates of glyphosate (1.7 kg ae/ha or possibly lower) will provide effective control of privet when applied in the spring or fall. Nomenclature: Glyphosate; triclopyr; Chinese privet, Ligustrum sinense Lour. Additional index words: Bottomland hardwoods, crown cover, invasive weeds, response surface analysis. INTRODUCTION spread and thrive under dense forest canopies. As an Chinese privet is a rapidly encroaching plant that continues to invade disturbed sites, fencerows, and bottomland and upland forests in the Southeast (Dirr 1998; Haragan 1996; Miller 2003). This shade-tolerant, perennial shrub or small tree grows to a height of 9 m and has multiple stems (Miller 2003). Its foliage is evergreen to semievergreen, becoming deciduous in cold climates (Dirr 1998). Once liberated from their fleshy fruit, privet seeds will germinate promptly without cold stratification (Burrows and Kohen 1986; Young and Young 1992). The spread of its seeds by birds and other animals and abundant production of root sprouts enable the species to invade new areas and form dense thickets (Dirr 1998; Miller and Miller 1999). Because of the species' shade tolerance and abundant regeneration, privet is able to ' Received for publication September 23, 2003. and in revised form May 18. 2004. Research Forester, USDA Forest Service, Pacific Northwest Research Station, 3625 93rd Avenue Southwest, Olympia, WA 98512-9193; Research Ecologist, USDA Forest Service, Southern Research Station, 520 Devall Drive, Auburn University, AL 36849. Corresponding author's E-mail: [email protected]. additional layer of understory vegetation, privet may be an important factor limiting hardwood regeneration, wildlife habitat, biodiversity, and recreational activities. Introduced from China in 1852 as a woody ornamental, Chinese privet has escaped and now dominates understories of mesic forests throughout the southeastern United States (Haragan 1996) and is moving into New England and the Midwest (USDA-NRCS 2003a). During the period of 1950 to 1980, Chinese privet distribution expanded at an exponential rate, and today it is present in over 40% of southeastern U.S. counties (USDANRCS 2003b). In a survey conducted by the USDA Forest Service, the Forest Inventory and Analysis Program estimated that Ligustrunz spp. occupied approximately 5% of forestland area along the eastern seaboard from Virginia to Florida (Rudis and Jacobs 2002). Chinese privet is ranked among the top 10 exotic pest plants of Georgia (Georgia Exotic Pest Plant Council 2003) and Mississippi (Matlack 2002). Herbicides are an important tool for controlling Ligustrum spp., although comprehensive comparisons of application rate and timing are not available in the pubHARRIKGTON AND MILLER: CONTROL OF CHINESE PRIVET lished literature. In primary screening work, Miller (1998) observed 89 to 90% control of Chinese privet after 1 yr with high rates of glyphosate, imazapyr, or metsulfuron applied as foliar sprays in August, whereas control averaged only 60% after triclopyr. James and Mortimer (1984) successfully controlled privet with cutstump applications of picloram plus 2,4-D or picloram plus triclopyr and with foliage applications of metsulfuron (spring or autumn) or glyphosate (spring only). Similarly, Little (1982) achieved control of 97% of privet plants by cut-stump application of picloram plus 2,4D. Mowatt (1981) found consistently high levels of control when privet was injected with triclopyr or hexazinone but variable control when injected with glyphosate or dicarnba. Of the herbicides tested, glyphosate and triclopyr have no soil activity at registered rates (WSSA 1994) and pose little risk to associated vegetation when applied to privet as a directed foliar application. Other herbicides, such as picloram, imazapyr, and metsulfuron, have soil-activated phytotoxic effects on many hardwood tree species and therefore have restrictions when used for privet control in bottomland forests. To identify optimum application rates and timings of herbicides for a given target species, controlled studies are needed in which these factors are varied systematically and plant responses are quantified with objective measurements (Borders and Shiver 1989; Knowe et al. 1995). Therefore, the objective of this research was to compare control of Chinese privet abundance and height 2 yr after various application rates, timings, and formulations of glyphosate and triclopyr. Because a herbicide dose applied to privet in a small plot might be subject to excessive dilution within the creeping root system, we conducted a separate study in which privet control after low rates of glyphosate was compared in the presence vs. absence of trenching to sever the root system from nearby plants. MATERIALS AND METHODS Study Site and Treatments. The research was conducted in the understory of a 1.2-ha bottomland hardwood stand located at the confluence of McNutts and Barber creeks in Oconee County near Athens, GA (lat 33'57'N. long 83'19'W). Soils are gravelly sandy loams of the Madison series (fine, kaolinitic, thermic Typic Kanhapludults) and gravelly loams of the Louisa series (loamy, micaceous, thermic, shallow Ruptic-Ultic Dystrudepts) (USDA-NRCS 2003c). The upper canopy of the forest included, in decreasing order of abundance, river birch (Betula nigra L.), green ash (Fraxinus penns~~lvanica Marsh.), boxelder (Acer negundo L.), red maple (Acer rubrurn L.), yellow-poplar (Liriodendron tct11pifern L.), American hornbeam (Carpinus carolirzinnn Walt.), water oak (Quercus nigra L.), and sweetgum (Liquidanzbar styracijua L.). In spring 1999, a deme stand of privet, 2 to 4 m in height, was cut to a 15-cm height by the Georgia Department of Transportation in preparation for a stump application of the triethylamine salt of triclopyr in water. However, the herbicide treat ment was delayed for several weeks, and no signs plant injury were visually detectable at study initiation (spring 2000). Triclopyr entry into the privet stumps probably was prevented by blockage of the xylem vessels, which can occur within 2 h after cutting the stern of a woody plant (Newton and Knight 1981). In April 2000, the study site was dominated by a uniform stand of 1-yr-old privet sprouts about 1 m in height. A total of 218 plots, each 3 by 6 m in dimension, were located in a contiguous grid. Three studies were initiated to compare privet control subsequent to a variety of treatment specifications (Table 1). Study 1 compared four application rates (kg aelha) and five timings of glyphosate and triclopyr. Study 2 compared two formulations and four rates of glyphosate and triclopyr applied in August 2000. Studies 1 and 2 had randomized complete block designs with four replications of each treatment. Blocks ran parallel to McNutts Creek and were assigned according to distance from the creek because flooding can limit privet growth (Brown and Pezeshki 2000). Four of the plots (one per block) were randomly assigned as nontreated checks. Using the remaining plots, study 3 compared three application rates of glyphosate applied in August 2000 with or without trenching of plot perimeters to a depth of 50 cm with a Ditch Witch3 to isolate privet clumps. Study 3 had a cornpletely randomized design with three replications of each treatment because plot locations did not conform to the blocked designs of studies 1 and 2. Plots for the three studies were randomly interspersed. To evaluate control resulting from a nonherbicide treatment, four plots were designated for manual uprooting of privet in June (one plot per block). The time required to manually uproot the privet on a given plot was recorded (min/m2). Seedlings and small clumps were uprooted by hand, whereas larger clumps were uprooted with a winch puller." Herbicide treatments for study 1 were applied on the following dates in 2000: April 20, June 19, August 23, ' Model 1230, walk-along trencher, Ditch W~tch, 4501 East Second, Edmond, OK 73034-7500 'Model 144, w~nch puller, Ben Meadowa Co , PO Box 5277. Janesv~lle, W1 53547-5277 48 Volume 19, Issue 1 (January-March) 2005