Introduction
Prior to the colonization and expansion of European settlers in North America, grasslands covered approximately 15% of the continent, including the western shortgrass prairie, tallgrass prairies in the Great Plains and Upper Midwest, and the mixed-grass prairie in between the two regions (Anderson Reference Anderson2012; Arkansas Natural Heritage Commission 2016). North American prairies have been shrinking due to the conversion of land to agricultural production, overgrazing of native prairie, development of cities and towns, and climate change. Just more than half of the shortgrass prairie land (54%), 11% of the tallgrass prairie, and a quarter (24%) of the mixed-grass prairie remain nationwide (Masters et al. Reference Masters, Nissen, Gaussoin, Beran and Stougaard1996; Wilsey et al. Reference Wilsey, Grand, Wu, Michel, Grogan-Brown and Trusty2019).
Integrated roadside vegetation management (IRVM) has become increasingly popular since the 1980s to combat the steep decline in North American prairies, conserve intact grasslands near roadsides, and establish prairies along rights-of-way (Brandt et al. Reference Brandt, Henderson, Uthe and Urice2015; Harper-Lore Reference Harper-Lore1998). IRVM includes mowing, herbicide use, grazing, burning, replanting, and cultivating desirable vegetation with native plants to establish and maintain safe and functional roadsides that require less maintenance (Brandt et al. Reference Brandt, Henderson, Uthe and Urice2015; Eck and McGeeh Reference Eck and McGeeh2008). IRVM also includes planting diverse stands of native vegetation along roads to prevent erosion by including deep and expansive-rooted native plants, especially warm-season grasses, to stabilize sediment and reduce stormwater runoff (Brandt et al. Reference Brandt, Henderson, Uthe and Urice2015; Nemec Reference Nemec2023). These practices connect fragmented remnant land along roadsides to create native pollinator refuges, provide habitat and food for birds and insects, reduce soil erosion, improve water quality, and allow mobility between areas for animals and seed dispersion for plant species (Angelella et al. Reference Angelella, Stange, Scoggins and O’Rourke2019; Armstrong et al. Reference Armstrong, Christians, Erickson, Hopwood, Horning, Kramer, Landis, Moore, Remley, Riley, Riley, Roberts, Skinner, Steinfeld, Stella, Teuscher, White and Wilkinson2017; Foti Reference Foti1978; USDA-NRCS 2011).
Prairie establishment can be disrupted by competitive weed species crowding out desirable native species (Martin et al. Reference Martin, Moomaw and Vogel1982; Masters et al. Reference Masters, Nissen, Gaussoin, Beran and Stougaard1996). Preventing weed interference and allowing native species to grow uninhibited is integral to the success of the establishment phase of prairie restoration (Armstrong et al. Reference Armstrong, Christians, Erickson, Hopwood, Horning, Kramer, Landis, Moore, Remley, Riley, Riley, Roberts, Skinner, Steinfeld, Stella, Teuscher, White and Wilkinson2017; Wenzel et al. Reference Wenzel, Shaw, Hedtke, Luniewski, Mohring, Norris, Powell and Strojny2012). Therefore, removing existing vegetation prior to planting native grassland species is vital, but interventions are often labor intensive and highly target species-specific (Brandt et al. Reference Brandt, Henderson, Uthe and Urice2015; Masters et al. Reference Masters, Nissen, Gaussoin, Beran and Stougaard1996; Randall Reference Randall1996). Current recommendations are to generally avoid spraying herbicides on new prairie sites after native grasses and forbs have been planted because of a lack of selective weed control or uncertainty in the desirable species response to a given herbicide. Herbicides can kill or damage the seedlings before the plants are established and create space for invasive plants (Brandt et al. Reference Brandt, Henderson, Uthe and Urice2015; Durling and Leif Reference Durling and Leif2014; USDA-NRCS 2018).
Previous research has demonstrated that some herbicides applied preemergence, such as aminopyralid, clopyralid, or picloram, cause no injury or reduction in canopy cover of big bluestem, sideoats grama, or switchgrass (Lym et al. Reference Lym, Becker, Moechnig, Halstvedt and Peterson2017). Postemergence applications of quinclorac and sulfosulfuron caused ephemeral reductions in switchgrass cover that were not distinguishable from nontreated checks at 8 wk after treatment (Curran et al. Reference Curran, Ryan, Myers and Adler2011). Switchgrass coverage was reduced and sideoats grama coverage increased in response to postemergence applications of metsulfuron and chlorsulfuron (Lair and Redente Reference Lair and Redente2004). Postemergence applications of clopyralid, dithiopyr, metsulfuron, MSMA, and quinclorac individually did not injure buffalograss or reduce buffalograss coverage (Fry and Upham Reference Fry and Upham1994). While native grass species have exhibited varying levels of tolerance to some common roadside herbicides, many of the native forbs are sensitive to those herbicides (Durling and Leif Reference Durling and Leif2014). More specifically, postemergence applications of mesotrione, clopyralid, halosulfuron, and imazaquin caused injury to black-eyed Susan (Henry et al. Reference Henry, Tucker and McCurdy2023). If postemergence herbicides can be applied to newly emerged native plant species without apparent plant injury or adverse effects on growth, selective weed control may be possible when establishing native species or mixtures of native species. Herbicides serve as a tool in IRVM for prairie establishment and conservation on rights-of-way, so assessments of native plant species tolerance to common roadside herbicides are necessary.
The aim of this study was to identify several Great Plains native grass and forb species that may exhibit tolerance at the seedling stage to exposure to four commonly used postemergence herbicides. The present study assessed plant injury response and growth responses of 11 container-grown native grass and forb species treated with four postemergence herbicides used on roadsides and Conservation Reserve Program (CRP) land: clopyralid, florpyrauxifen-benzyl, metsulfuron, and quinclorac.
Materials and Methods
Native grass and forbs species were selected based on a series of preliminary germination trials conducted in May through August 2022 (data not shown). Multiple seeding densities and depths were tested for each species, using Ray Leach Cone-tainers™ (Stueve and Sons, Inc., Tangent, OR) (3.8 cm diam by 21.6 cm length) filled with PRO-MIX M growing medium with mycorrhizae (Premier Tech, Quakertown, PA). Species that failed to produce suitable plants in ≥30% of Cone-tainers were excluded from herbicide screens due to limited and unreliable plant material for assessments. Based on germination and emergence in the preliminary trials, five grass species and nine forb species were included in the greenhouse trial (Table 1). However, reduced germination rates of Indiangrass [Sorghastrum nutans (L.) Nash] and prairie blazing star (Liatris pycnostachya Michx.) necessitated their removal from the trial prior to randomization and herbicide treatment application. Additionally, common milkweed (Asclepias syriaca) was initially included in the first run of the trial but was excluded from the trial following herbicide application due to high mortality in nontreated plants (33%) at the first data collection at 7 d after treatment (DAT). Thus, a total of 11 species were included for assessment in this study (Table 1). Four commonly used roadside and CRP-approved herbicides were selected for this trial: quinclorac, florpyrauxifen-benzyl, clopyralid, and metsulfuron (Table 2). Three of the herbicides are synthetic auxins in Herbicide Resistance Action Committee (HRAC) Group 4: quinclorac, florpyrauxifen-benzyl, and clopyralid (Heap Reference Heap2022). Metsulfuron is a HRAC Group 2 herbicide that inhibits the production of acetolactate synthase (Heap Reference Heap2022). A nontreated control was included for each species.
Scientific and common names of native prairie species included in greenhouse trials conducted in 2022 and 2023.a,b
a All seeds sourced from Roundstone Native Seed, LLC 9764 Raider Hollow Road, Upton, KY 42784.
b Greenhouse trials were conducted in Fayetteville, AR, at the Milo J Shult Research and Extension Center
c Species removed from trial prior to completion of experiment.
Herbicide treatments included in greenhouse trials conducted in 2022 and 2023.a
a Greenhouse trials were carried out in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
b Rates for herbicides are g ae ha−1for Clopyralid and g ai ha−1 for metsulfuron, florpyrauxifen-benzyl, and quinclorac.
c Manufacturer locations and websites: BASF, Research Triangle Park, NC (https://agriculture.basf.us); Corteva Agriscience, Indianapolis, IN (https://www.corteva.us/); Nufarm, Alsip, IL (https://nufarm.com/).
d Nonionic surfactant at 5 mL L−1 was added to the treatment.
e Nonionic surfactant at 3 mL L−1 was added to the treatment.
f Methylated seed oil at 5 mL L−1 was added to the treatment.
g Methylated seed oil at 10 mL L−1 was added to the treatment.
Plant materials were prepared by direct-seeding native species into Cone-tainers filled with PRO-MIX M growing medium with mycorrhizae and maintained in a greenhouse located at the Milo J. Shult Research and Experiment Center in Fayetteville, AR (36.10°N, 94.17°W). Cone-tainer cells were sown with sufficient seed of each species to produce at least one viable plant per Cone-tainer. Excess seedlings in Cone-tainers were thinned within 72 h after emergence. For each replication, each species included 14 Cone-tainers arranged in 2 × 7 rows on a planting rack to receive herbicide treatments. Sowing dates were staggered to allow plants from each species to reach similar heights (10.2 cm) at the time of treatment. Herbicides were applied in a compressed air-powered spray chamber, and it was necessary to avoid having excessively tall plants of one species that could disrupt proper herbicide coverage on adjacent plants during application (Table 3). Greenhouse temperature was set to 24 C, and Cone-tainers were checked daily and watered regularly to ensure the substrate maintained medium levels of moisture.
Planting dates of native prairie species included in greenhouse trial in Run 1 (2022) and Run 2 (2023).a
a Greenhouse trials were carried out in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
b Plants were removed from the trial due to insufficient germination
c Due to poor germination in initial sowing, additional seeds were sown into new Cone-tainers and used.
For herbicide application, plants were moved from the evaluation greenhouse, and 14 Cone-tainers of each species were arranged on trays for each replication of herbicide treatment with a minimum of 7.6 cm between species within the spray chamber. Herbicide treatments were applied on November 15, 2022 (Run 1) and April 5, 2023 (Run 2), using a compressed air-powered spray chamber calibrated to deliver 187 L ha−1 at 1.6 kph and fitted with two flat-fan TeeJet 1100067 nozzle tips (TeeJet Technologies, Glendale Heights, IL) placed 50 cm apart and at a height 50 cm above native plant canopies. Postemergence treatments included clopyralid (689 g ae ha−1), metsulfuron (4.18 g ai ha−1), florpyrauxifen-benzyl (38.4 g ai ha−1), and quinclorac (418 g ai ha−1), plus appropriate adjuvants (Table 2).
Treated plants were returned to the observation greenhouse immediately following herbicide treatment and remained arranged by herbicide application for 48 h. Plants were then arranged in a randomized complete block design with five replications, including all treatment combinations: 12 species by five herbicide treatments in Run 1 and 11 species by five herbicide treatments in Run 2. Visual assessments of plant injury were conducted on a scale of 0% to 100% (0% representing no injury and 100% representing plant death) relative to a nontreated check of each species at 7, 14, 21, and 28 DAT. Plant injury symptoms manifested in different ways depending on species and herbicide chemistry. Injury symptoms were rated as a composite score of all visible injury, which included epinasty, leaf-cupping, vein discoloration, buggy whipping, lodging, and stunting. Aboveground biomass was recorded following a destructive harvest at 28 DAT. Termination at 28 DAT was necessary due to plant growth beyond the capacity of Cone-tainers. All plants were cut at the media line, bulked into paper bags, and dried in an industrial drying oven (Power-O-Matic-60; Blue M, New Columbia, PA) for a minimum of 72 h at 70 C prior to weighing as an experimental unit.
The greenhouse trial was conducted as a two-factor factorial with herbicide and plant species as main effects. The experiment was arranged in a randomized, complete block design with five replications, repeated over two runs. Analysis of variance was conducted using the GLIMMIX procedure with SAS software (v. 9.4; SAS Institute, Cary, NC) with the main effects of herbicide, plant species, and experimental run as well as their respective interaction effects, treated as fixed effects. Replication nested within the greenhouse run was treated as a random effect. Where interactions of experimental run and main effects were found to be significant (P < 0.05), the SLICE statement in GLIMMIX was used to compare means from each treatment and experimental run combination. Least square means were generated using the LINES statement, and means were compared using Tukey’s honestly significant difference at a significance level of 0.05. While the experimental design accommodates comparisons of response variables across all levels of treatment combinations (i.e., herbicide×plant species), response variables were distinctly measured for many species (e.g., herbicide injury), making statistical comparisons across species inappropriate. Furthermore, analyzing plant injury data across all species was problematic because extreme injury levels adversely affected outputs of the pooled data. Instead, data were analyzed separately by species with ANOVA tests conducted on the main effects of herbicide, experimental run, and their interaction. If a significant herbicide×run interaction was observed, means are presented separately for each run.
Results and Discussion
Plant Injury
Big bluestem exhibited low levels or no injury at any evaluation date. Peak injury of 8% occurred in response to metsulfuron 14 DAT (P = 0.0009; Table 4). This level of injury is below the acceptable herbicide injury thresholds established for turfgrasses (Boeri et al. Reference Boeri, Unruh, Kenworthy, Trenholm and Rios2021; Peppers and Askew Reference Peppers and Askew2023) and would not be noticeable along a roadside. Metsulfuron is considered appropriate to apply postemergence on CRP land and grass pastures, so the low levels of injury in the greenhouse conditions are of little concern (Shaner et al. Reference Shaner, Jachetta, Senseman, Burke, Hanson, Jugulam, Tan, Reynolds, Strek, McAllister, Green, Glenn, Turner and Pawlak2014). Furthermore, big bluestem injury levels in response to metsulfuron were ≤5% for the remainder of the trial and were not different from the other herbicide treatments. The findings for big bluestem are consistent with those of previous studies showing minimal big bluestem injury in response to the selected herbicides, including preemergence applications of clopyralid and postemergence applications of metsulfuron (Lym et al. Reference Lym, Becker, Moechnig, Halstvedt and Peterson2017; Peters et al. Reference Peters, Moomaw and Martin1989).
Visual assessment of big bluestem plant injury treated with four postemergence herbicides.a– d
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
The other three grasses, switchgrass, buffalograss, and sideoats grama, were all injured by florpyrauxifen-benzyl (Tables 5–7). Florpyrauxifen-benzyl treatment resulted in buggy whipping of switchgrass and lesions at the base of the plants, and caused 40% injury by the end of the trial (P < 0.0001; Figure 1). Similarly, buffalograss injury from florpyrauxifen-benzyl increased for several weeks in both runs ranging 14% to 62% in the first run and 24% to 41% in the second (P < 0.0001). Buffalograss exhibited unacceptable injury at 28 DAT with 49% injury and 28% injury in the first and second runs, respectively (P < 0.0001). Buffalograss injury manifested as browning, curling, and a thinner appearance of grass blades (Figure 2). During the first run, sideoats grama injury in response to florpyrauxifen-benzyl was 11% and 10% at 21 DAT (P = 0.0005) and 28 DAT (P = 0.0021), respectively. However, there was no significant difference in visual injury ratings between the herbicides 21 or 28 DAT in the second run (P > 0.05). Clopyralid and quinclorac did not injure buffalograss or switchgrass in previous seedling and establishment studies (Curran et al. Reference Curran, Ryan, Myers and Adler2011; Fry and Upham Reference Fry and Upham1994; Lym et al. Reference Lym, Becker, Moechnig, Halstvedt and Peterson2017). In this trial, metsulfuron caused negligible injury to the grass species, which is consistent with previous research showing that buffalograss was unaffected by postemergence applications of metsulfuron (Fry and Upham Reference Fry and Upham1994). However, metsulfuron has been shown to increase sideoats grama coverage while decreasing switchgrass coverage in field trials (Lair and Redente Reference Lair and Redente2004). Although florpyrauxifen-benzyl was not previously tested on the native grass species in this trial, the herbicide is active on several nonrelated Poaceae species (Miller and Norsworthy Reference Miller and Norsworthy2018).
Visual assessment of switchgrass plant injury treated with four postemergence herbicides.a –d
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
e Data are combined over 2022 and 2023 for 7 and 28 DAT due to no significant (P > 0.05) herbicide×run interaction at those time points.
Visual assessment of buffalograss and Illinois bundleflower plant injury treated with four postemergence herbicides.a–d
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0 to 100% scale, with 0 indicating a healthy plant and 100% indicating total plant death.
e Data are combined over 2022 and 2023 for 7 and 14 DAT due to no significant (p > 0.05) herbicide×run interaction at those time points.
Switchgrass buggy whipping and lesion growth following florpyrauxifen-benzyl herbicide application in 2022 a greenhouse trial in Fayetteville, AR.
Buffalograss symptoms 28 d after postemergence herbicide treatment in a 2022 greenhouse trial in Fayetteville, AR. From left to right: nontreated, clopyralid, florpyrauxifen-benzyl, metsulfuron, and quinclorac treated.
Butterfly milkweed treated with florpyrauxifen-benzyl exhibited more injury than with other herbicide treatments 7 DAT (8% and 13%; P < 0.001) and 14 DAT (4%; P = 0.0170; Table 8). In 2022, butterfly milkweed injury did not differ in response to herbicide treatments 21 or 28 DAT (P > 0.05). However, in 2023, butterfly milkweed injury in response to florpyrauxifen-benzyl was 12% and 10% at 21 and 28 DAT, respectively (P < 0.001). Other herbicides caused no substantial injury across rating timings, suggesting that butterfly milkweed may have considerable tolerance to clopyralid, metsulfuron, and quinclorac (Figure 3). Butterfly milkweed is a larval food for monarch butterfly (Danaus plexippus) and other butterfly species, making them frequent inclusions as perennial forbs in prairie restoration and wetland rehabilitation sites (Stevens Reference Stevens2006). However, milkweed species are toxic to humans and livestock, therefore, efforts have been made to eradicate milkweed plants from range and pasture areas (Stevens Reference Stevens2006). In a 2014 herbicide screen, butterfly milkweed was found to not be tolerant to clopyralid, but was tolerant to halosulfuron, a sulfonylurea similar to metsulfuron (Durling and Leif Reference Durling and Leif2014). Some systemic herbicides such as glyphosate are effective in reducing butterfly milkweed stands because they impact the plant’s rhizome; however, other systemic herbicides, including 2,4-D, 2,4,5-T, and dicamba, have been ineffective in reducing stands (Bhowmik Reference Bhowmik1982).
Visual assessment of butterfly milkweed and Mexican hat plant injury treated with four postemeregence herbicides.a–c
a Abbreviation: POST, postemergence herbicide; DAT, days after treatment.
Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
b Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s Honest Significant Difference Test.
c Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
d Data are combined over 2022 and 2023 for 14 DAT due to no significant (P > 0.05) herbicide×run interaction at that time point.
Butterfly milkweed symptoms 28 d after postemergence herbicide treatment in a 2023 greenhouse trial in Fayetteville, AR. From left to right: nontreated, clopyralid, florpyrauxifen-benzyl, metsulfuron, and quinclorac treated.
Black-eyed Susan injury was high across all treatments and runs, with all herbicides causing unacceptable injury (≥52%) at 28 DAT in both runs (P < 0.0001; Table 9). Clopyralid and florpyrauxifen-benzyl caused highest injury levels at 7 and 14 DAT (P < 0.001). Clopyralid caused 93% injury at 28 DAT in 2023 and caused higher injury than metsulfuron and quinclorac at 21 and 28 DAT; however, injury was similar between plants treated with florpyrauxifen-benzyl (81% to 84%) and metsulfuron (79% to 81%) at 28 DAT (P < 0.0001). High levels of injury caused by these herbicides are not unexpected, given previous studies by Henry et al. (Reference Henry, Tucker and McCurdy2023), which found that clopyralid and halosulfuron were injurious to black-eyed Susan. Clopyralid is known to control weedy plants belonging to the Asteraceae family (Anonymous, 2022); therefore, it is unsurprising that clopyralid causes injury to black-eyed Susan, also a member of the Asteraceae family. Reever Morghan et al. (Reference Reever Morghan, Leger and Rice2003) found that applications of clopyralid in bunchgrass prairies significantly reduced the overall presence of native and nonnative Asteraceae species and Asteraceae plant cover.
Visual assessment of ashy sunflower and black-eyed Susan plant injury treated with four postemergence herbicides. a–d
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
e Data are combined over 2022 and 2023 for 7 and 14 DAT due to no significant (P > 0.05) herbicide×run interaction at those time points.
Desert false indigo injury was severe (≥75%) in response to all herbicides for at least one rating timing, and observed injury ratings increased from 7 DAT to 28 DAT for all herbicide treatments (Table 10). Throughout the trial, desert false indigo plants treated with metsulfuron exhibited significantly lower injury symptoms than plants treated with the other herbicides (P ≤ 0.0003). At 28 DAT, metsulfuron-treated desert false indigo exhibited 75% injury in 2022 and 28% injury in 2023, a substantial disparity. Desert false indigo plants were larger at the time of herbicide application in Run 1 than in Run 2, despite having similar numbers of true leaves; this could potentially explain some variability in injury between runs. Taller desert false indigo plants in Run 1 were etiolated and more stressed than in Run 2, which may make them more susceptible to herbicide injury. Desert false indigo injury levels are unacceptable for new seedlings in both runs, indicating a postemergence application of metsulfuron would not be suitable for seedlings of this species. Herbicide-treated desert false indigo plants were smaller than nontreated checks, with leaflets turning yellow to brown, and then full leaves withering and dying.
Visual assessment of desert false indigo plant injury treated with four postemergence herbicides. a –d
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were a composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0% to 100% scale, with 0% indicating a healthy plant and 100% indicating total plant death.
e Data are combined over 2022 and 2023 for 7 DAT due to no significant (P >0.05) herbicide×run interaction at that time point.
Illinois bundleflower was generally injured more by clopyralid (26% to 80% in Run 1; 65% to 97% in Run 2) and florpyrauxifen-benzyl (45% to 93% Run 1; 75% to 99% Run 2) than metsulfuron (2% to 69% Run 1; 10% to 83% Run 2) or quinclorac (27% to 77% Run 1; 35% to 67% Run 2; Table 6; P < 0.0001). Initially, the Illinois bundleflower plants exhibited auxin-type herbicide symptoms that included dysregulated growth, with stems curling and twisting. At 14 DAT, the treated plants began to defoliate with entire leaves drying out despite moisture being maintained, and the plants began dropping their leaflets. The plants continued to defoliate, and stems began to dry out before the end of the trial.
Mexican hat plant was generally injured more by clopyralid (52% to 96% Run 1; 69% to 99% Run 2) and florpyrauxifen-benzyl (51% to 95% Run 1; 74% to 97% Run 2) than by metsulfuron (0% to 43% Run 1; 0% to 45% Run 2) or quinclorac (21% to 45% Run 1; 43% to 85% Run 2; Table 8). Mexican hat plants were larger at the time of herbicide application in Run 1 than in Run 2, potentially explaining some variability between runs. Injury symptoms from the synthetic auxins clopyralid, florpyrauxifen-benzyl, and quinclorac were similar in form but differed in degree (Figure 4). All three herbicides caused leaf curling, stunting, and chlorosis. Injury symptoms in response to metsulfuron were similar to symptoms caused by halosulfuron, another sulfonylurea, which caused stunting and chlorosis to Mexican hat plants (Wiese et al. Reference Wiese, Keren and Menalled2011).
Mexican hat plant symptoms 28 d after postemergence herbicide treatment in a 2022 greenhouse trial in Fayetteville, AR. From left to right: nontreated, clopyralid, florpyrauxifen-benzyl, metsulfuron, and quinclorac treated.
Ashy sunflower plants were taller at the time of treatment during the second run than the first, however, they still had similar trends in herbicide injury between the two runs (Table 9). Plants treated with clopyralid or florpyrauxifen-benzyl exhibited higher levels of injury than those treated with quinclorac or metsulfuron throughout the study (P < 0.0001). By 28 DAT, while metsulfuron-treated ashy sunflowers exhibited only 15% or 25% injury (Run 1 vs. Run 2), plants treated with quinclorac reached 63% to 69% injury, while those treated with clopyralid or florpyrauxifen-benzyl exhibited between 89% and 96% injury (P < 0.0001). All herbicides damaged the apical meristem of the ashy sunflowers, with the growing point becoming bleached or necrotic and the top pair of true leaves curling down and away from the top of the plant. Ashy sunflowers treated with clopyralid and florpyrauxifen-benzyl exhibited the most prominent stem bending, with the tops entirely bending over, and showing more chlorosis and necrosis than the other herbicide treatments.
In the first run, at 7 DAT, purple coneflower treated with florpyrauxifen-benzyl had the highest injury (28%), followed by plants treated with clopyralid (18%), then quinclorac (1%) and metsulfuron (0%; P < 0.0001; Table 11). However, at 7 DAT in the second run, purple coneflowers sprayed with either clopyralid or florpyrauxifen-benzyl exhibited similar increased levels of injury (35% and 44%, respectively) compared with plants treated with metsulfuron (1%) or quinclorac (3%; P < 0.0001). This trend continued at 14 DAT with clopyralid or florpyrauxifen-benzyl causing more injury than other herbicide treatments (P < 0.0001). At 21 and 28 DAT, only purple coneflowers treated with quinclorac exhibited less injury relative to the other three herbicide treatments (27% at 21 DAT and 37% at 28 DAT; P < 0.0001). Purple coneflowers treated with clopyralid manifested injury symptoms as petioles curling, while plants treated with metsulfuron exhibited red veins on the underside of the leaf, and all four herbicide treatments led to generally smaller true leaves than those on the nontreated purple coneflowers.
Visual assessment of purple coneflower plant injury treated with four postemergence herbicides. a–e
a Abbreviation: DAT, days after treatment.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Visual ratings were composite score of 14 individual plants for each replication. Injury ratings were assessed visually on a 0 to 100% scale with 0 indicating a healthy plant and 100% indicating total plant death.
e Data are combined over 2022 and 2023 for 14, 21, and 28 DAT due to no significant (P > 0.05) herbicide×run interaction at those time points.
Given the low maintenance areas of roadsides and higher standards for turfgrasses, injury levels of ≥25% were considered unacceptable. The native grasses in this trial did not exhibit ≥10% injury in response to clopyralid, metsulfuron or quinclorac, indicating that these herbicides may be suitable for postemergent use in the field at the selected rates. Florpyrauxifen-benzyl caused unacceptable injury to switchgrass and buffalograss, but not to big bluestem and sideoats grama. Thus, native grasses had a species-specific response to florpyrauxifen-benzyl, while clopyralid, metsulfuron, and quinclorac did not cause injury that exceeded 25% and were selective for the grass species in this study.
Florpyrauxifen-benzyl and clopyralid typically caused greater injury to the forb species than metsulfuron or quinclorac. Florpyrauxifen-benzyl and clopyralid caused high injury (≥79%) to ashy sunflower, black-eyed Susan, Illinois bundleflower, and purple coneflower. Although metsulfuron and quinclorac caused less damage the desirable plants, they still caused unacceptable levels of injury (≥25%) to all forb species except for butterfly milkweed. Thus, these selected herbicides do not seem safe to use on new mixed-species plantings of native forbs.
Biomass
The biomass of one grass and four forb species had a significant herbicide×run interaction, so their masses are presented separately by year (Table 12). The other three grass and three forb species did not have a significant herbicide×run interaction for biomass and are presented with runs combined (Table 13). There were no significant differences between herbicides and how they affected the dried biomass of big bluestem, desert false indigo, Mexican hat plant, or sideoats grama when analyzed individually by species (P > 0.05; Table 13).
Dried plant weights of five greenhouse-grown native grass and forb species treated with four postemergence herbicides. a–d
a Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
b Means were derived from the total aboveground biomass of all plants in each experimental unit (14) of each replication.
c Means within a column followed by the same lowercase letter are not significantly different at α = 0.05 according to Tukey’s honestly significant difference test.
d Weights were combined for the entire experimental unit per replication.
Dried plant weights of six greenhouse-grown native grass and forb species treated with four postemergence herbicides. a,b
a Means were derived from the total aboveground biomass of all plants in each experimental unit (14) of each replication.
b Greenhouse trials were conducted in 2022 and 2023 in Fayetteville, AR, at the Milo J Shult Research and Extension Center.
Nontreated purple coneflower and Illinois bundleflower generally produced more biomass than those treated with a herbicide. In the first run of the trial, nontreated purple coneflower plants produced more dried biomass (3.6 g) than herbicide-treated plants (2.4 to 2.7 g; P < 0.0001; Table 13). In the second run of the trial, all plants produced less biomass than in the first run. Despite this, the trend of nontreated purple coneflower producing the most biomass (1.0 g) was mostly maintained whereas the smallest biomass (P = 0.0392; Table 12) was produced when quinclorac (0.6 g), metsulfuron (0.5 g) and clopyralid (0.5 g) were used. However, in the second run, nontreated purple coneflower biomass was not different from that of plants treated with florpyrauxifen-benzyl (0.8 g; P = 0.0392).
Nontreated Illinois bundleflower produced the most biomass in both runs (3.7 g and 2.1 g in runs 1 and 2, respectively; P < 0.0001; Table 12). In the first run, mean biomasses were similar for all herbicide-treated Illinois bundleflowers (1.8 to 2.1 g; P < 0.0001). In the second run, quinclorac-treated Illinois bundleflower produced more biomass (1.3 g) than plants treated with clopyralid (0.6 g), florpyrauxifen-benzyl (0.5 g), or metsulfuron (0.6 g; P < 0.0001).
Although there was a disparity in ashy sunflower mass between the two runs, with the plants in the second run being more massive in general than in the first run, the effect of the herbicides on the plants’ dried biomass trend was similar (Table 12). In both the first and second runs of the trial, ashy sunflower produced less biomass when treated with clopyralid (2.1 g Run 1; 8.2 g Run 2) or florpyrauxifen-benzyl (2.7 g Run 1; 8.3 g Run 2) than plants that were not treated (4.9 g Run 1; 15.4 g Run 2) or treated with metsulfuron (5.0 g Run 1, P = 0.0011; 14.4 g Run 2, P < 0.0001). In Run 1, the weight of ashy sunflower plants treated with quinclorac were not statistically different from those treated with florpyrauxifen-benzyl (3.9 g; P = 0.0011). However, in the second run, quinclorac-treated ashy sunflowers were more massive (10.5 g) than those treated with florpyrauxifen-benzyl or clopyralid (P < 0.0001).
Buffalograss plants that were not treated with a herbicide (3.1 g) or were treated with metsulfuron (2.9 g) had more dried biomass than plants treated with florpyrauxifen-benzyl (2.0 g; P < 0.0030; Table 13). This coincides with the increased buffalograss injury caused by the use of florpyrauxifen-benzyl.
Black-eyed Susan biomass did not significantly differ between herbicide treatments in the first run (P = 0.3548; Table 12). However, in the second run, the nontreated plants produced 9.0 g of biomass and those treated with quinclorac produced 8.2 g, which is more than all other treatments (P < 0.0001). The smallest amounts of Black-eyed Susan biomass in the second run occurred when clopyralid (5.3 g), florpyrauxifen-benzyl- (4.9 g), and metsulfuron (5.4 g) were applied (P < 0.0001).
In the second run, the effect of herbicide on switchgrass biomass was insignificant (P = 0.6642; Table 12). However, in the first run, switchgrass treated with florpyrauxifen-benzyl (3.0 g) or metsulfuron (3.6 g) produced less biomass than nontreated checks (4.4 g) or plants treated with clopyralid (4.6 g) or quinclorac (4.9 g; P < 0.0001).
Butterfly milkweed responded differently from the other plant species in terms of biomass (Table 13). Of all the treatments, the nontreated butterfly milkweed produced the least biomass (1.6 g; P = 0.0375), although it was not statistically different from the biomass of plants treated with florpyrauxifen-benzyl (1.6 g), metsulfuron (1.7 g), or quinclorac (1.7 g). Butterfly milkweed treated with clopyralid produced more biomass (2.0 g) than the nontreated check, indicating a potential for herbicide tolerance.
Although clopyralid-treated butterfly milkweed plants produced more biomass than the nontreated plants, in general, the herbicides caused reductions in forb biomass. Ashy sunflower, black-eyed Susan, buffalograss, Illinois bundleflower, and purple coneflower all treated with clopyralid and florpyrauxifen-benzyl produced less biomass than their nontreated counterparts. This coincides with the increased forb injury caused by clopyralid and florpyrauxifen-benzyl in this trial. Metsulfuron and quinclorac also caused biomass reductions to most of the forb species, excluding ashy sunflower (both herbicides in Run 1 and metsulfuron in Run 2), black-eyed Susan (quinclorac in Run 2), desert false indigo, and Mexican hat plant. The least reduction in biomass occurred in the grass species, including big bluestem and side oats grama, likely due to herbicide treatment, with no statistically significant differences. Buffalograss biomass was reduced by florpyrauxifen-benzyl, and switchgrass biomass was reduced by both florpyrauxifen-benzyl and metsulfuron use in the first run, but no differences were observed in the second run.
Practical Implications
This greenhouse herbicide screening found that postemergence herbicides such as clopyralid, metsulfuron, and quinclorac may be appropriate for use on native warm-season grass plantings without causing significant injury or reductions in biomass. However, these herbicides may be inappropriate for use on native forb or mixed, first-year plantings of native grass and forb species because of the forbs’ sensitivity to all herbicides investigated in this trial, with undesirable levels of injury. Due to the injury levels to switchgrass, buffalograss, and most of the native forbs, the use of florpyrauxifen-benzyl while establishing native grass and forb plantings does not appear to be a suitable option. Additional trials in the field would be beneficial to determine whether factors such as weed interference, soil system, temperature, or rainfall affect the resilience of the plants to these herbicides.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wet.2025.34
Acknowledgments
We thank Drew Kirkpatrick, Kayla Knepp, Sarah Pascal, Aleksy Cybulski, Trevor McClain, Ali Ablao, and Jackson Ball for their help with these studies.
Funding
This work was supported in by the Arkansas Department of Transportation and by U.S. Department of Agriculture–National Institute of Food and Agriculture, Hatch project 1024455.
Competing interests
The authors declare they have no competing interests.
Open access
