HERBICIDES TO AID ESTABLISHMENT OF FOURWING SALTBUSH Joseph L. Petersen Darrell N. Ueckert Matthew W. Wagner

ABSTRACT of this study was to identify herbicides effective for selec­ tive control of competing vegetation in fourwing saltbush Failures in attempts to establish fourwing saltbush plantings. (Atriplex canescens) are often attributed to interference from weeds. Field experiments were conducted in western Texas to evaluate preemergence and postemergence appli­ MATERIALS AND METHODS cations of selected herbicides for we~d control in fourwing The study was conducted on an Angelo clay loam saltbush plantings. Sprays ofmetolachlor at 2.0 kg a.i.lha (fine, mixed, thermic Torrertic Calciustolls) at the Texas and alachlor at 2.5 kg/ha appeared promising as preemer­ A&M University Agricultural Research and Extension gence treatments. Acifluorfen at 0.56 kglha and clopyralid Center, 8 km northwest of San Angelo in the southern at 0.28 kg/ha effectively controlled broadleafed weeds, and Rolling Plains resource area. Soil pH was 7.8 and organic fluazifop-P at 0.21 kg/ha effectively controlled grasses matter content was 1.2 percent.. Elevation is 580 m and while causing no or only slight injury to fourwing saltbush mean annual precipitation is 47 em. The most predictable plants that were 1 month to 2 years old. Postemergence and abundant periods of rainfall are April-June and herbicides appear more practical than preemergence herbi­ September-October. cides in arid and semiarid areas where fourwing saltbush is utilized. Preemergence Herbicide Experiments INTRODUCTION Selected preemergence herbicide treatments were evaluated in three experiments. Seedbeds were disked Fourwing saltbush (Atriplex canescens) is used exten­ twice and packed before planting. About 800 dewinged sively for rangeland seeding because of its wide range fourwing saltbush seeds (4 g) from a commercial source of adaptation to edaphic and climatic conditions and its (harvested in western Texas) were hand planted 1.3 em potential to produce nutritious forage for livestock and deep in each of two 120-cm rows in 2- by 2-m plots sepa­ wildlife, particularly during winter (Plummer and others rated by 1.2-m borders. Germination of the seed on moist 1966; Petersen and others 1987). However, success in blotter paper in a controlled environment was 22 percent. establishing the shrub by seeding or transplanting Herbicide treatments included trifluralin [2,6-dinitro­ seedlings has been highly variable. Failures have been N ,N-dipropyl-4-(trifluoromethyl)benzenamine] at 0.8 kg attributed to poor seedbed preparation, poor seed quality, a.i .lha, oryzalin [4-( di propyl amino)-3,5-dini trobenzene­ inadequate soil moisture, grazing, and excessive competi­ sulfonamide] at 1.1 kg/ha, ethalfluralin [N-ethyl-N-(2- tion from associated plants (Springfield 1970; Nord and methyl-2-propenyl)-2, 6-dinitro-4-(trifl uoromethyl)­ others 1971; Aldon 1972; Petersen and others 1986). benzenamine] at 1.9 kg/ha, metolachlor [2-chloro-N-(2- Plant competition is most often the primary factor ethyl-6-methyl phenyl)-N-(2-methoxy-1-methylethyl)­ limiting establishment offourwing saltbush plantings acetamide] at 2.0 kg/ha, and alachlor [2-chloro-N-(2,6- (Giunta and others 1975; VanEpps and McKell 1977, diethylphenyl)-N-(methoxymethyl)acetamide] at 1983; Geist and Edgerton 1984; Petersen and others 2.8 kg/ha. 1986). Selective weed control practices are often needed Treatments were applied in separate experiments on to enhance and hasten shrub establishment and return April27, 1987, September 15, 1987, and May 13, 1988. the investment in seed, seedbed preparation, fencing, Other treatments included hand-weeding and untreated and deferment. Various weed control strategies have checks. Herbicides were applied in water at a total vol­ increased the survival and growth of tree plantings in ume of 1.9 Uplot using hand-held sprayers. Trifluralin semiarid regions (Felker and others 1986). The objective was incorporated 3 to 5 em deep with garden rakes imme­ diately after application, just prior to planting fourwing saltbush seed. All other herbicides were applied immedi­ ately after seeding, without incorporation. The experi­ Paper presented at the Symposium on Cheatgrass Invasion, Shrub Die­ Off, and Other Aspects of Shrub Biology and Management, Las Vegas, NV, ments were arranged as randomized complete blocks with April 5-7, 1989. four replications. Soil water contents were maintained Joseph L. Petersen and Darrell N. Ueckert are Research Associate and Range Ecologist, respectively, Texas Agricultural Experiment Station, at field capacity by frequent irrigation with garden sprin­ 7887 N. Hwy. 87, San Angelo, TX 76901. Matthew W. Wagner is Biologist klers to assure seed germination, emergence, and estab­ II, Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, lishment offourwing saltbush. The study sites were TX 78744. fenced to eliminate rodents, lagomorphs, and livestock.

305 This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. Phytotoxicity of the herbicides to fourwing saltbush was Phytotoxicity of the herbicides to fourwing saltbush quantified by counting the live seedlings 15 and 90 days plants was estimated 15 and 90 DAT by visually ranking after treatment (DAT) and by measuring seedling heights percent necrosis in each plot (0 percent = no visible injury, 90 DAT. Canopy cover (percent) of grasses and broad­ 100 percent = no live tissue visible). Efficacy of the treat­ leafedweeds was visually estimated in five 0.1-m2 ran­ ments for weed control was estimated 90 DAT by visually domly located quadrats in each plot 90 DAT. Grasses estimating weed canopy cover from five 0.1-m2 randomly and broadleafed weeds were then harvested at ground located quadrats in each plot. Grasses and broadleafed level and dried to a constant weight at 50 °C for estimates weeds in the quadrats were then harvested at ground of standing crop. level and dried to a constant weight at 50 oc for estimates of standing crop. The data were subjected to analyses of variance. Trans­ Postemergence Herbicide formations of percentage data did not affect data interpre­ Experiments tation, so actual values are presented. Results from the Field experiments were established to evaluate the April 1987 and May 1988 experiments with preemergence effects of selected postemergence herbicides on fourwing herbicides were similar, thus the data were pooled. Means were separated by Duncan's multiple range test (P 0.05) saltbush seedlings. Plots were 3 by 3m and separated = where appropriate. by 1.2-m borders. Plots were rototilled and packed just prior to planting seed and transplanting seedlings. Four age classes offourwing saltbush plants occurred in each RESUL~PREEMERGENCE plot when treatments were applied. These included: HERBICIDES (1) 1-month-old seedlings, hand planted at 8 g/2.4-m row, 1.3 em deep, about 6 weeks prior to treatment. The Preemergence applications of metolachlor and alachlor number of seedlings present at time of treatment ranged had little effect on emergence or growth of fourwing from 12 to 80/plot; (2) 4-month-old seedlings (seven/plot), saltbush seedlings. Shrub seedling densities in plots transplanted on 38-cm centers 1 month prior to treat­ treated with metolachlor or alachlor were similar to ment; (3) 9-month-old seedlings (seven/plot), transplanted those in hand-weeded plots after 90 days (table 1). on 38-cm centers 1 month prior to treatment; and ( 4) 2- However, shrub seedling densities were significantly year-old, stem-cut seedlings (five/plot), transplanted on lower (P = 0.05) in plots treated with trifluralin, oryzalin, 61-cm centers 6 months before treatment. One row of and ethalfluralin than in hand-weeded plots after 90 each plant age class was included in each plot. days. Seedling emergence and establishment were much Seeds used for the first three age classes were pur­ lower following autumn seeding compared to spring seed­ chased from a commercial source harvested in western ing, but the responses to autumn herbicide applications Texas. The stem-cut seedlings were from mature plants were generally similar to those observed for spring of an accession from Texon, TX, growing in a nursery at treatments. the Research Center. The plots were frequently irrigated The preemergence herbicides did not affect saltbush as discussed earlier. seedling heights compared to heights of seedlings in hand­ Herbicide treatments applied postemergence to the weeded plots (table 1). The high mortality of saltbush weeds June 16, 1988 included acifluorfen {5-[2-chloro-4- seedlings during the first 3 months after planting in well­ (trifluoromethyl)phenoxy]-2-nitrobenzoic acid} at 0.56 kg prepared seedbeds with hand weeding and frequent irri­ a.i .lha, metsulfuron {2-[[[[( 4-methoxy-6-methyl-1,3,5- gation indicated that poor stand establishment in these triazin-2yl)amino]carbonyl]amino]sulfonyl]benzoic experiments was partly related to poor seed quality, low acid} at 0.04 kg/ha, sulfometuron {2-[[[[(4,6-dimethyl-2- inherent vigor of seedlings, or both. pyrimidinyl)amino]carbonyl]amino]sulfonyl]benzoic acid} at 0.05 kglha, fluazifop-P {(R)-2-[4-[[5-(trifluoromethyl)-2- pyridinyl]oxy]phenoxy]propanoic acid} at 0.21 kg/ha, and Table 1-Fourwing saltbush seedling densities (No./2.4-m row) 30 clopyralid (3,6-dichloro-2-pyridinecarboxylic acid) at 0.28 and 90 days (d) and heights (mm) 90 days after spring or kg/ha. Herbicide treatments applied on October 4, 1988 autumn applications of selected preemergence herbicides were the same as above except that metsulfuron and sul­ near San Angelo, TX 1 fometuron were applied' at 0.02 kg/ha based on results Average seedling from the first experiment. Broadleafed and grass weeds densities Average seedling were actively growing and varied phenologically from the 2 Treatment S~ring Autumn height at 90 d two-leaf stage to flowering at time of herbicide application Herbicide Rate 30d 90d 30d 90d S~ring Autumn in both experiments. kglha --- -No./2.4-m row------mm ----- Herbicides were applied as broadcast sprays in water Trifluralin 0.8 20b 7b 2c 1c 44b 4 at a total volume of 1.9 Uplot with hand-held sprayers. Oryzalin 1.1 38ab 12b 7bc 2c 74ab 17 A nonionic surfactant (trimethylnonylpolyethoxyethanol) Ethalfluralin 1.9 33b 12b 9bc 2c 34b 13 was included at 0.13 percent (v/v) with acifluorfen and Metolachlor 2.0 49ab 19ab 17b Sbc 64ab 13 at 0.25 percent (v/v) with all other herbicides. Other Alachlor 2.8 39ab 17ab 12b Sbc 66ab 14 treatments in both experiments included hand weeding Hand weeded 66a 29a 14b 7ab 71ab 18 and untreated checks. The experiments were arranged None 40ab 12b 27a 11a 96a 19 as randomized complete blocks with three replications. 1Means within a column followed by similar lower case letters or without Plots were fenced to exclude herbivores. letters are not significantly different (P= 0.05). 2Average of two experiments.

306 Weeds in the experimental area included common RESUL~POSTEMERGENCE purslane (Portulaca oleracea), Dakota verbena (Verbena bipinnatifida), spear globemallow (Sphaeralcea hastu­ HERBICIDES lata), common devilsclaws (Martynia louisianica), western One-Month-Old Seedlings-Mid-June applications of ragweed (Ambrosia psilostachya), kleingrass (Panicum metsulfuron at 0.04 kg/ha and sulfometuron at 0.05 kg/ha coloratum), and common oats (Avena sativa). All herbi­ killed 1-month-old seedlings offourwing saltbush within cides applied in the spring experiments reduced weedy 90 DAT (table 3). Visual estimates of injury caused by the plant cover compared to that on untreated plots (table 2). other treatments were similar to those of untreated and Canopy cover of weeds was significantly less on alachlor­ hand-weeded plots at 90 DAT, although acifluorfen and treated plots compared to plots treated with ethalfluralin, clopyralid appeared to cause slight injury to 1-month-old but there were no other differences among the herbicide seedlings. treatments. All herbicides except ethalfluralin reduced Application rates of metsulfuron and sulfometuron were the standing crop of grasses compared to untreated plots reduced to 0.02 kglha in the second experiment (October but differences in broadleafed weed standing crop among 1988), but visual estimates of injury to 1-month-old seed­ the treatments were not significant (table 2) due to high lings at 90 DAT were still high (100 and 83 percent for within-treatment variation caused by the "patchy" dis­ metsulfuron and sulfometuron, respectively) (table 4). tribution of weeds. Canopy cover and standing crop of Initial injury caused by acifluorfen applied at 0.56 kg/ha annual weeds were considerably less in the autumn ex­ was moderate ( 43 percent), but seedlings had generally periment compared to the spring experiment (table 2). recovered after 90 days. Clopyralid applied at 0.28 kg/ha Weed canopy cover and standing crop did not differ sig­ in October caused moderate injury ( 40 percent), whereas nificantly with treatment in the autumn experiment. fluazifop-P applied at 0.21 kglha caused no apparent phy­ Most of the preemergence herbicides initially sup­ totoxicity to 1-month-old seedlings (table 4). pressed weed establishment and growth, but rapid weed emergence and growth began in most plots after about Four· and 9-Month-Old Seedlings-Four- and 60 days in all experiments. Rates of application used 9-month-old fourwing saltbush plants appeared vigorous were those recommended for medium-textured soils. when herbicides were applied. Injury to 4-month-old Weed emergence and growth after 60 days probably re­ seedlings was near 100 percent 90 days after June appli­ flect the short residual of these preemergence herbicides. cations ofmetsulfuron and sulfometuron applied at 0.04 Higher rates of application do not appear feasible because and 0.05 kg/ha, respectively (table 3). Phytotoxicity of the tendency of most of the herbicides to reduce the symptoms resulting from June applications of the other establishment, growth, or both, of saltbush seedlings herbicides were slight to moderate after 90 days. Nine­ (table 1). month-old seedlings were also killed by metsulfuron. Sulfometuron was less toxic than metsulfuron, as evi­ denced by decreased phytotoxicity as seedling age increased. Clopyralid sprays applied in June caused moderate injury to 9-month-old seedlings (table 3). Table 2-Canopy cover (percent} and standing crop 2 Four- and 9-month-old seedlings were more tolerant of (g/m } of competing vegetation 90 days after spring or autumn applications of selected the lower rates (0.02 kglha) ofmetsulfuron and sulfome­ preemergence herbicides to plantings of turon applied in October (table 4), but injury by the lower fourwing saltbush near San Angelo, TX1 rate ofmetsulfuron was still unacceptable (63 percent). The 4- and 9-month-old shrub seedlings were generally Treatment Canopy Weed standing cro~ uninjured or only slightly injured by sprays of acifluorfen, Herbicide Rate cover Broadleaf Grasses fluazifop-P, or clopyralid applied in October. kg!ha Percent ------glmZ ------Two-Year-Old Seedlings-Metsulfuron applied in Spring planting2 mid June at 0.04 kglha caused almost total (94 percent) foliar necrosis to 2-year-old fourwing saltbush plants Trifluralin 26bc 0.8 90 1Gb (table 3). Evidence of initial phytotoxicity caused by sul­ Oryzalin 1.1 25bc 206 42b fometuron at 0.05 kglha (25 percent necrosis) was not Ethalfluralin 1_.9 28b 124 76ab Metolachlor 2.0 22bc 84 8b present after 90 days. There was no apparent injury Alachlor 2.8 15c 38 2b to 2-year-old fourwing saltbush 90 days after sprays of None 58a 114 158a acifluorfen, fluazifop-P, clopyralid, or sulfometuron were applied in June. Mature fourwing saltbush plants have Autumn planting been reported to be tolerant to sprays of clopyralid applied Trifluralin 0.8 10 2 10 at rates of 0.56 to 1.12 kglha (Jacoby and others 1981). Oryzalin 1.1 40 2 118 Metsulfuron applied at 0.02 kglha in October caused Ethalfluralin 1.9 14 2 40 moderate to severe foliar necrosis (54 percent) of the Metolachlor 2.0 12 2 24 2-year-old plants (table 4). October applications of the Alachlor 2.8 4 2 10 None 31 12 60 other herbicides caused little or no injury. 1Means within a column and planting season followed by similar lower case letters or without letters are not significantly different (P = 0.05). 2Average of two experiments.

307 2 Table 3-Necrosis (percent) of four age classes of fourwing salt- Table 5-Canopy cover (percent) and standing crop (g/m ) bush plants 15 and 90 days after spring applications of competing vegetation 90 days after spring or of selected postemergence herbicides near San Angelo, autumn applications of postemergence herbicides TX1 to fourwing saltbush plantings near San Angelo, TX1 Treatment Plant age class Herbicide Rate 1 month 4 months 9 months 2 years Treatment Canopy Weed standing cro~ kglha ------Percent necrosis ------Herbicide Rate cover Broadleaf Grasses kglha Percent ------glrrfl ----- 15 days after treatment Spring planting Acifluorfen 0.56 23b 14b 29cde Be Metsulfuron .04 B7a B1a 79a 43a Acifluorfen 0.56 29abc 74bc 48bc Fluazifop-P .21 5b 29b 13de 12bc Metsulfuron .04 22bc 42bc 56 be Clopyralid .2B 27b 21b 35bcd 3c Fluazifop-P .21 34ab 20Bab 6c Sulfometuron .05 7Ba 91a 59ab 25b Clopyralid .28 40a 10c 190a Hand weeded 7b 12b Be 2c Sulfometuron .05 16c 1Bc 142ab None Ob 19b 44bc 1c None 39ab 262a 48bc 90 days after treatment Autumn planting Acifluorfen 0.56 17b 17b 27cd Ob Acifluorfen 0.56 21 16 80 Metsulfuron .04 100a 100a 100a 94a Metsulfuron .02 15 12 38 Fluazifop-P .21 Ob 25b Bd 7b Fluazifop-P .21 13 52 2 Clopyralid .28 25b 29b 44bc Ob Clopyralid .28 30 32 116 Sulfometuron .05 100a 94a 63b 1b Sulfometuron .02 B 26 14 Hand weeded Ob 10b 7d Ob None 40 56 156 None Ob 15b 41bc Ob 1Means within a column and planting date followed by similar 1Means within a column and evaluation date followed by similar lower lower case letters or without letters are not significantly different case letters are not significantly different (P= 0.05). (P=0.05).

Table 4-Necrosis (percent) of four age classes of fourwing salt­ The dominant weeds during the June 1988 experiment bush plants 15 and 90 days after autumn applications were prostrate euphorbia (Euphorbia prostrata), Dakota of selected postemergence herbicides near San Angelo, verbena, spear globemallow, queensdelight (Stillingia 1 TX sylvatica), browntop panicum (Panicum fasciculatum), Treatment Plant age class stinkgrass (Eragrostis cilianensis), and kleingrass. Only Herbicide Rate 1 month 4 months 9 months 2 years sulfometuron applied at 0.05 kg/ha in June reduced weed kglha ------Percent necrosis ------canopies compared to those on untreated plots (table 5). All herbicides except fluazifop-P reduced broadleafed 15 days after treatment weed standing crop compared to untreated plots. Acifluorfen 0.56 43b 15ab 13abc 16a Fluazifop-P applied at 0.21 kg/ha significantly reduced Metsulfuron .02 7Ba 22a 25a 12ab the standing crop of grasses compared to clopyralid or Fluazifop-P .21 2cd 5b 3c Oc sulfometuron. Clopyralid .2B 13c 6b Bbc 1c Weeds present during the October experiment, al­ Sulfometuron .02 53b 15ab 17ab Bb though not abundant, were similar to the spring experi­ Hand weeded Od 5b 1c Oc ment with the addition of common purslane, Texas filaree None 3cd 4b 1c Oc (Erodium texanum), western ragweed, and common oats. 90 days after treatment Weed canopies and standing crop were low for most herbi­ Acifluorfen 0.56 7d 3c 3c 6b cide-treated plots, but the means were not significantly Metsulfuron .02 100a 51 a 63a 54 a different (table 5). Fluazifop-P .21 O.d 9bc 4c 4b Clopyralid .2B 40c 9bc 15bc 22b DISCUSSION Sulfometur:on .02 B3b 31ab 24b 10b Hand weeded Od 2c 2c 3b Herbicides have been an effective tool for vegetation None Od 9bc 7c 4b manipulation to improve rangelands for many years, 1Means within a column and evaluation date followed by similar lower but to our knowledge there have been no previous case letters are not significantly different (P = 0.05). attempts to use herbicides to enhance establishment of forage shrub plantings. Metolachlor and alachlor

308 applied at planting appeared promising as preemergence Geist, J. M.; Edgerton, P. J. 1984. Performance tests of treatments, but weed control was temporary because of fourwing saltbush transplants in eastern Oregon. In: the short half-life of these herbicides. Applications of Tiedemann, Arthur R.; McArthur, E. Durant; Stutz, clopyralid or acifluorfen after weed emergence satisfacto­ Howard C.; Stevens, Richard; Johnson, Kendall L., rily controlled broadleafed weeds, and postemergence ap­ compilers. Proceedings: symposium on the biology plications offluazifop-P controlled grasses with no or only of Atriplex and related chenopods. Gen. Tech. Rep. slight phytotoxicity to fourwing saltbush seedlings and INT-172. Ogden, UT: U.S. Department of Agriculture, established plants. Forest Service, Intermountain Forest and Range Single-herbicide treatments are unlikely to provide Experiment Station: 244-250. acceptable suppression of both grasses and broadleafed Giunta, B. C.; Christensen, D. R.; Monsen, S. B. 1975. weeds in shrub plantings. Additional research is needed Interseeding shrubs in cheatgrass with a browse to evaluate herbicide combinations and sequential appli­ seeder-scalper. Journal of Range Management. 28(5): cations of different herbicides. The integrated use of cul­ 398-402. tivation and selective herbicides might have application Jacoby, P. W.; Meadors, C. H.; Foster, M.A. 1981. Control in some shrub establishment efforts. ofhoney mesquite (Prosopisjuliflora var. glandulosa) Postemergence herbicide treatments appear more with 3,6-dichloropicolinic acid. Weed Science. 29(4): practical than preemergence treatments because of the 376-378. uncertainty of precipitation (and hence the uncertainty Nord, E. C.; Hartless, P. F.; Nettleton, W. D. 1971. of weed problems) in most areas where fourwing saltbush Effects of several factors on saltbush establishment is utilized for range improvement. Herbicidal control of in California. Journal of Range Management. 24(3): undesirable plants may not be a viable practice where 216-223. fourwing saltbush is planted in mixtures with grasses, Petersen, J. L.; Ueckert, D. N.; Potter, R. L. 1986. desirable forbs, or other shrub species. Cultural practices for establishing fourwing saltbush The high level of seedling mortality observed under within perennial grass stands. Journal of Range Man­ the ideal conditions of hand weeding and irrigation in agement. 39(5): 460-463. this study suggests that some seeding failures that have Petersen, J. L.; Ueckert, D. N.; Potter, R. L.; Huston, J. E. been attributed to interference from associated vegetation 1987. Ecotypic variation in selected fourwing saltbush may have been a result of poor seed quality or low vigor populations in western Texas. Journal of Range Man­ offourwing saltbush seedlings. Additional research on agement. 40(4): 361-366. increasing fourwing saltbush seed quality and seedling Plummer, A. P.; Monsen, S. B.; Christensen, D. R. 1966. vigor appears warranted. Fourwing saltbush: a shrub for future game ranges. Pub. No. 66-4. Salt Lake City, UT: Utah State Depart­ ment of Fish and Game. REFERENCES Springfield, H. W. 1970. Germination and establishment Aldon, E. F. 1972. Critical soil moisture levels for field of fourwing saltbush in the Southwest. Res. Pap. planting fourwing saltbush. Journal of Range Manage­ RM-55. Fort Collins, CO: U.S. Department of Agri­ ment. 25(4): 311-312. culture, Forest Service, Rocky Mountain Forest and Blauer, A. C.; Plummer, A. P.; McArthur, E. D.; Range Experiment Station. 48 p. Stevens, R.; Giunta, B. C. 1976. Characteristics VanEpps, G. A.; McKell, C. M. 1977. Shrubs plus grass and hybridization of important Intermountain shrubs. for livestock forage: a possibility. Utah Science. 38: II. Chenopod family. Res. Pap. INT-177. Ogden, UT: 75-78. U.S. Department of Agriculture, Forest Service, Inter­ VanEpps, G. A.; McKell, C. M. 1983. Effect of weedy mountain Forest and Range Experiment Station. 42 p. annuals on the survival and growth of transplants Felker, P.; Smith, D.; Wiesmann, C. 1986. Influence of under arid conditions. Journal of Range Management. mechanical and chemical weed control on growth and 36(3): 366-369. survival of tree plantings in semiarid regions. Forest Ecology and Management. 16: 259-267.

309