Field study.

Field research to evaluate select PRE herbicides labeled for crabgrass control in common bermudagrass turf managed under different mowing heights was conducted at the East Tennessee Research and Education Center–Plant Sciences Unit (ETREC) in Knoxville, TN (lat. 35.54° N, long. 83.57° W) during 2012 and 2013. This experiment was repeated in 2013 at North Carolina State University Turf Field Laboratory (NCSU) in Raleigh, NC (lat. 35.73° N, long. 78.67° W) and Pine Hills Golf Club (UGA) in Winder, GA (lat. 33.97° N, long. 83.69° W) to evaluate treatment response across a wider geographic region of the southeastern United States.

At all locations, research was conducted on established common bermudagrass. At NCSU and ETREC, plots (1 × 2.5 m and 1.5 × 2.4 m, respectively) were established in areas with a history of smooth crabgrass incidence, whereas large crabgrass was the predominant species at UGA (1.5 × 3-m plots). Soil type at NCSU was a Cecil sandy clay loam (fine, kaolinitic, thermic typic Kanhapludult), measuring 6.4 in soil pH and 1.9% in organic matter. Soil type at ETREC was a Sequatchie loam soil (fine-loamy, siliceous, semiactive, thermic humic Hapludult), measuring 6.2 in soil pH and 2.1% in organic matter. At the UGA location, soil was a Madison sandy clay loam (fine, kaolinitic, thermic typic Kanhapludult) with a pH of 5.1 and 0.7% organic matter. In both years, turf at ETREC received two fertilizer applications at 49 kg nitrogen (N)/ha in spring (March, April) from a complete fertilizer (21.0N–0.4P₂O₅–13.3K₂O; Harrell’s LLC, Charlotte, NC). Supplemental nutrition was applied similarly at NCSU with two applications (April, June) made at 49 kg N/ha from a complete fertilizer (25.0N–2.6P₂O₅–10.0K₂O; Harrell’s LLC). Fertilizer (16.0N–1.7P₂O₅–6.6K₂O; Griggs Brothers, Albion, ID) was applied at UGA at a rate of 49 kg N/ha in April and June. To prevent wilt, overhead irrigation was applied to supplement rainfall at each location. Additionally, plots at all locations received ≈0.5 cm of irrigation within 24 h of herbicide or fertilizer applications.

Treatments included a factorial combination of two mowing heights (1.5 or 3.8 cm), two application regimes (single or split applications with split treatments applied at initial application and 8 weeks later), and six PRE herbicides including: dithiopyr (Dimension; Dow AgroSciences, Indianapolis, IN), indaziflam (Specticle; Bayer Environmental Sciences, Research Triangle Park, NC), oxadiazon (Ronstar; Bayer Environmental Sciences), pendimethalin (Pendulum AquaCap; BASF Corporation, Research Triangle Park, NC), prodiamine (Barricade; Syngenta Professional Products, Greensboro, NC), and prodiamine + sulfentrazone (Echelon; FMC Corporation, Philadelphia, PA). Herbicide application rates and dates are presented in Tables 1 and 2. All herbicides except oxadiazon used water as the carrier and were applied with CO₂-powered backpack sprayers at all locations. Carrier volumes at ETREC, NSCU, and UGA were 280, 304, and 280 L·ha⁻¹, respectively, using flat-fan nozzles (XR8002; TeeJet Technologies, Wheaton, IL). A granular formulation of oxadiazon was applied at all locations with a shaker jar.

Table 1.

Preemergent herbicide treatments evaluated in Knoxville, TN, Raleigh, NC, and Winder, GA, during 2012 and 2013.^z

View Table

Table 2.

Calendar dates and environmental conditions when applying preemergent herbicides during 2012 and 2013 in Knoxville, TN (ETREC), Raleigh, NC (NCSU), and Winder, GA (UGA).

View Table

Plots were mowed at 1.5 or 3.8 cm at all locations beginning 4 weeks before herbicide treatment. Plots maintained at 1.5 cm were mowed three times per week, whereas those maintained at 3.8 cm were mowed two times per week. In all cases, clippings were returned. ETREC maintained 1.5-cm and 3.8-cm plots with reel (Jacobsen LF 3800; Jacobsen, A Textron Company, Charlotte, NC) and rotary (Gravely Zero Turn; An Ariens Company, Brillion, WI) mowers, respectively. Rotary mowers were used to maintain both 1.5-cm (Honda HRC 215; American Honda Motor Company, Duluth, GA) and 3.8-cm (John Deere 997 Zero-Turn; Deere and Company, Moline, IL) plots at NCSU. Plots at UGA were maintained at 1.5 cm with a reel mower (Toro Reelmaster 5500-D; The Toro Company, Bloomington, MN) and 3.8 cm with a rotary mower (Toro Groundsmaster 3320; The Toro Company). Nontreated plots were also included for each mowing height.

Crabgrass control was visually evaluated on a 0% (no control) to 100% (complete control) scale 12, 16, 20, and 24 weeks after initial treatment (WAIT) in 2012 and 2013. Crabgrass plant counts were also taken 24 WAIT in 2013 using a 1-m² grid (10 cm × 10 cm spacing) placed in the center of each plot. All count data are expressed as a percent reduction compared with the non-treated control at each respective mowing height.

The experimental design was a randomized complete block with three replications at ETREC and UGA and four replications at NCSU. Crabgrass control data were arcsine-transformed before being subjected to analysis of variance in SAS (SAS Version 9.1; SAS Institute, Cary, NC) using expected means squares described by McIntosh (1983) for multiple location experiments. Interpretations of arcsine- and non-transformed data were similar; therefore, non-transformed means are presented for clarity. Data from the 2012 and 2013 experiments were analyzed separately. Significant treatment-by-location interactions were detected in 2013 data collected at ETREC, NCSU, and UGA; therefore, data from all 2013 experiments were analyzed and presented separately. In both years, Fisher’s protected least significant difference (lsd) test was used to separate treatment means at the α = 0.05 level. Pearson’s correlation coefficients were also calculated to evaluate relationships between visual assessments of crabgrass control and plant counts.

Residue analysis.

Laboratory experiments were conducted in 2013 to determine the effect of mowing height on prodiamine and pendimethalin residue from field research plots at ETREC and NCSU. Prodiamine and pendimethalin were selected because they are commonly used PRE herbicides in established common bermudagrass areas. Furthermore, each are dinitroaniline herbicides with varying persistence. Prodiamine and pendimethalin have moderate to persistent and short to moderate half-lives, respectively, in soil (Senseman, 2007). Single soil cores (2.5 × 2.5 × 2.0 cm) were collected from the center of each plot using a soil probe (AMS Inc., America Falls, ID) at 0 and 24 WAIT. Initial samples (0 WAIT) were collected by removing the above-ground vegetation in a small sampling area within the plot, thereby treating the soil and allowing the herbicides to air-dry for 2 h before collecting the samples. All samples were wrapped in aluminum foil and stored immediately at –20 °C until analysis occurred.

Prodiamine and pendimethalin were extracted according to U.S. EPA Method Number 445276-01 with slight modifications as noted (U.S. Environmental Protection Agency, 1996). Ten grams of soil were placed in a 175-mL polypropylene conical bottom centrifuge bottle and 25 mL deionized water was added and hand-shaken for 15 s. Seventy-five milliliters of 2% acidic methanol was added and shaken on an orbital shaker for 1 h at 240 rpm. Samples were then centrifuged for 10 min at 4000 rpm and supernatant was decanted. Samples were filtered through a 0.45-μm syringe filter, vialed, and analyzed by an ultra-performance liquid chromatography–tandem mass spectrometer. Residues were quantified based on peak area measurements using appropriate calibration curves with means separated using Fisher’s protected lsd test at the α = 0.05 level.

Field study.

By 24 WAIT in 2012, averaged over application regime, all PRE herbicides excluding prodiamine and indaziflam (52.5 g·ha⁻¹) provided greater crabgrass control on plots maintained at 3.8 cm compared with 1.5 cm (Table 3). When applied to bermudagrass maintained at 3.8 cm, prodiamine controlled crabgrass 97%, similar to indaziflam (52.5 g·ha⁻¹), oxadiazon, and prodiamine + sulfentrazone (Table 3).

Table 3.

Mowing height-by-herbicide interactions in smooth crabgrass (Digitaria ischaemum) control data collected 24 weeks after treatment in Knoxville, TN (ETREC) in 2012.

View Table

Significant (P < 0.0001) location-by-treatment interactions were detected in crabgrass control and plant count data collected in 2013; therefore, data from the ETREC, NCSU, and UGA locations are presented separately. By 24 WAIT at ETREC, dithiopyr applications to bermudagrass maintained at 3.8 cm controlled crabgrass greater than dithiopyr applied to turf at 1.5 cm (Table 4). Although this is similar to findings from ETREC in 2012, this response was not observed at NCSU or UGA in 2013. At these locations, averaged over application regime, similar crabgrass control was observed with dithiopyr (62% to 78%) regardless of mowing height 24 WAIT (Table 4). Applications of indaziflam at 52.5 g·ha⁻¹ to 3.8 cm turf at ETREC controlled crabgrass less than those made to turf mowed at 1.5 cm in 2013. This response was also observed with indaziflam (35 g·ha⁻¹) at NCSU (Table 4). No differences in crabgrass control with indaziflam were detected between the 3.8- and 1.5-cm mowing heights at UGA in 2013; however, overall levels of crabgrass control with indaziflam at this location were low by 24 WAIT (20% to 41%). Similar to 2012, prodiamine (1680 g·ha⁻¹) provided excellent crabgrass control 24 WAIT (94% to 100%) regardless of location, mowing height, or application regime in 2013.

Table 4.

Mowing height-by-herbicide interactions in crabgrass (Digitaria spp.) control and plant count data collected 12 and 24 weeks after treatment (WAT) in Knoxville, TN (ETREC), Raleigh, NC (NCSU), and Winder, GA (UGA) in 2013.

View Table

Location-by-application regime interactions (P = 0.02) were also detected in 2013. At ETREC, single and split applications of indaziflam (35 g·ha⁻¹) controlled crabgrass similarly; however, the level of crabgrass control with these treatments was unacceptable (46% to 50%) by 24 WAIT (Table 5). Application regime affected crabgrass control with the 35 g·ha⁻¹ rate of indaziflam in 2013 at NCSU. By 24 WAIT, split applications at 17.5 g·ha⁻¹ controlled crabgrass greater than a single application at 35 g·ha⁻¹. Previous research indicates sequential applications of mitotis-inhibiting PRE herbicides consistently provided season-long crabgrass control (Johnson, 1977; Proctor et al., 2012). The opposite relationship was observed between single and split application regimes of indaziflam at 52.5 and 35 g·ha⁻¹ at 12 WAIT and 35 g·ha⁻¹ by 24 WAIT at UGA. These responses suggest that 35 g·ha⁻¹ of indaziflam may not be a sufficient rate for crabgrass control in bermudagrass turf in the southeastern United States for season-long control. Overall crabgrass control with single and split applications of indaziflam at 35 g·ha⁻¹ in this 2-year study only ranged from 21% to 77%, regardless of mowing height (Tables 3 and 4). With the exception of oxadiazon at UGA, significant differences were not detected between other PRE herbicides applied as a single or as split applications. Relationships among treatments using plant count data were similar to those detected by visual assessments of crabgrass control at each location. Plant count data were significantly correlated (r = 0.89, P < 0.0001) to visual assessments of crabgrass control 24 WAIT.

Table 5.

Application regime-by-herbicide interactions in crabgrass species (Digitaria spp.) control and plant count data collected 12 and 24 weeks after treatment (WAT) in Knoxville, TN (ETREC), Raleigh, NC (NCSU), and Winder, GA (UGA) in 2013.

View Table

Although mowing height had a pronounced effect on crabgrass control with several PRE herbicides in 2012, responses observed in 2013 were indistinct in comparison. This could be the result of variable environmental conditions over years (Table 2). During the data collection period in 2012 at ETREC, daily maximum air temperatures were 3 °C higher (29 °C) and rainfall accumulation (48.5 cm) was considerably lower than 30-year averages for this location (National Climatic Data Center, 2013). The opposite was observed in 2013 because rainfall accumulation during the data collection period (86.5 cm) exceeded the 30-year average by 28% (National Climatic Data Center, 2013). At the UGA site, rainfall accumulation during the data collection period (94.2 cm) in 2013 exceeded the 30-year average by 54% as well (National Climatic Data Center, 2013). Increased rainfall in 2013 may have facilitated greater crabgrass germination compared with 2012, therefore making reductions as a result of increased mowing height less apparent. Several researchers have found crabgrass germination to be maximal when water potential of soil is low (i.e., conditions of ample moisture) in the top 5 cm (King and Oliver, 1994; Masin et al., 2005). Increased germination may in part be related to the fact that crabgrass seed moisture content must increase to 38% to 50% before germination (King and Oliver, 1994).

Residue analysis.

Prodiamine residues in soil 24 WAIT varied as a result of mowing height (Table 6). At NCSU, prodiamine soil residues at the 1.5-cm mowing height were greater than those detected at 3.8 cm. A similar response was observed at ETREC; however, it was not statistically significant (P = 0.07). We surmise that increasing the bermudagrass canopy by mowing at a higher height of may limit prodiamine soil exposure at application. Although herbicide treatments were irrigated after application, it is likely that some was retained on the above-ground vegetation with greater amounts retained at the higher mowing height. In the case of prodiamine, if retained on the above-ground vegetation, it would be subject to photodegradation, perhaps decreasing its persistence (Senseman, 2007; Shaner, 2014). However, additional research is needed to confirm this theory. No difference in pendimethalin soil residue (1.1% or less of applied) was detected at NCSU or ETREC, regardless of mowing height. This may be attributable in part to pendimethalin having a shorter soil half-life than prodiamine (Senseman, 2007).

Table 6.

Mowing height-by-herbicide interactions prodiamine and pendimethalin soil residue data collected 24 weeks after treatment in Knoxville, TN (ETREC) and Raleigh, NC (NCSU) in 2013.^z

View Table

For financial reasons, the residue analysis was limited to two of the herbicides tested in field trials. However, responses indicate that differences in crabgrass control managed under different mowing heights may not be related to degradation of evaluated herbicides in soil. Rather, a higher bermudagrass canopy may reduce the amount of light that penetrates the bermudagrass canopy, thereby reducing crabgrass germination and subsequent growth. Reduced smooth crabgrass incidence in turf mowed at higher heights of cut has been reported previously in cool-season turfgrass (Debels et al., 2012; Dernoeden et al., 1998, 1993; Dunn et al., 1981; Hoyle et al., 2014; Jagschitz and Ebdon, 1985; Voigt et al., 2001). Debels et al. (2012) noted that although both mowing height and N application rate affect weed incidence in cool-season turfgrasses, increasing mowing height has a more dramatic effect on reducing weed abundance than augmenting fertilizer rates when clippings are returned after mowing.

Overall, we observed that increasing bermudagrass mowing height from 1.5 to 3.8 cm increased the efficacy of several PRE herbicides for crabgrass control during a single year of a 2-year study. During the second year of the study, abnormally high rainfall at two locations may have diminished this response. Improved crabgrass control efficacy was likely the result of reduced PAR at the soil surface and not differential breakdown of applied herbicides at the 1.5- and 3.8-cm mowing heights based on our findings. Future research should explore effects of increasing bermudagrass mowing height on PAR required for crabgrass germination and growth.