New Jersey.

The Freehold, NJ, site was the same as described for 2018 experiments in terms of soil characteristics and turfgrass management regimens. In 2019, the sites were adjacent ‘Dauntless’ kentucky bluegrass and ‘Revenge GLX’ perennial ryegrass (Lolium perenne L.) stands with natural infestations of annual bluegrass. See Table 1 for more information on the annual bluegrass infestation, which exhibited perennial behavior at the kentucky bluegrass site with a creeping growth habit and cover increase during summer. At the perennial ryegrass site, annual bluegrass exhibited a more upright growth habit and cover decreased during the summer, suggesting weak perennial or annual behavior. The degree to which turfgrass management practices (i.e., irrigation, mowing, and fungicide use) or annual bluegrass genetics contributed to perennial annual bluegrass behavior is not known. An additional experiment was conducted on mature ‘Traverse II’ tall fescue to evaluate turfgrass tolerance. No annual bluegrass was present on this site. The location, soil, and management were the same as described previously for the 2018 perennial ryegrass and kentucky bluegrass experiments. Plots were sized and arranged in the same way as described for the 2018 experiment. Treatments were applied in the same manner as in the 2018 experiment on 25 Apr., 9 May, and 22 May 2019 to all three sites.

Indiana.

Research was conducted at the William H. Daniel Turfgrass Research and Diagnostic Center in West Lafayette, IN (lat. 40°26'31″N, long. 86°55'53″W), on a 10-year-old perennial ryegrass (cultivars unknown) sward mown weekly at 1.3 cm. Annual bluegrass cover decreased slightly from spring to summer but the site was heavily infested (Table 1). The soil was a Starks-Fincastle silt loam (fine-silty, mixed, mesic Aeric Ochraqualf) with a pH of 6.7. Fertilizer was applied at 37 kg·ha⁻¹ N on 5 June 2019 (Shaw’s 24N–0P–22K Fairways Grade; Knox Fertilizer Company, Knox, IN) but no other fertilizer was added during the experimental period. Irrigation was applied as needed to prevent drought stress. Plots were sized 1.5 × 1.5 m and arranged in a randomized complete block design with four replications. Treatments were applied with water carrier at 815 L⋅ha⁻¹ through a three-nozzle boom equipped with flat fan nozzles (XR8003VS, Teejet; Spraying Systems Co., Glendale Heights, IL) using a CO₂-pressurized (207 kPa) sprayer on 24 Apr., 10 May, and 27 May 2019.

Iowa.

Research was conducted at Coldwater Golf Links in Ames, IA (lat. 42°1′36″N, long. 93°37′12″W), on kentucky bluegrass (cultivars unknown) rough mown at 4 cm. The annual bluegrass at the Iowa site was <1 year old, as the site was free of annual bluegrass before a flood in Summer 2018. The flood killed some kentucky bluegrass and allowed annual bluegrass to establish in Fall 2018. Annual bluegrass cover declined slightly from spring to summer at this site, suggesting annual or weak perennial behavior (Table 1). The soil was a Clarion Loam (fine-loamy, mixed, superactive, mesic Typic Hapludoll containing 5.2% organic matter). The site was fertilized with 112 kg·ha⁻¹ N annually (Lesco 28N–0P–3K; SiteOne Landscape Supply, Roswell, GA) applied in three applications and supplemented with 2.5 cm irrigation per week. Plots were sized 1.5 × 3.0 m and arranged in a randomized complete block design with four replications. Treatments were applied with water carrier at 420 L·ha⁻¹ through flat fan nozzles (XR8003VS, Teejet; Spraying Systems Co.) hand-held CO₂-pressurized (280 kPa) sprayer on 24 May, 7 June, and 24 June.

Data collection.

In 2018, annual bluegrass control was estimated visually on a 0% (no injury or control) to 100% (complete necrosis and/or weed absence) scale relative to the nontreated control. Turfgrass (kentucky bluegrass, perennial ryegrass or tall fescue) injury was evaluated visually when it was apparent on a 0% (no injury) to 100% (complete necrosis) scale relative to the nontreated control. Turfgrass injury <20% is generally considered acceptable by turf practitioners. Annual bluegrass cover was evaluated visually on a 0% (no cover) to 100% (complete cover) scale on the day of initial herbicide application and at the conclusion of the experiment in early autumn. In 2019, annual bluegrass control was not evaluated, but annual bluegrass cover was transformed and expressed as percent reduction (or increase) in cover compared with the initial application date on a plot-by-plot basis. To supplement visual estimates, a grid intersect count was used to determine annual bluegrass cover at the final evaluation in both years at all sites. The presence or absence of annual bluegrass was determined under each intersect. In New Jersey, a 0.8 × 0.9-m grid with 90 intersects was placed twice in each plot for a total of 180 intersects in 2018 and a 0.8 × 0.8-m grid with 80 evenly spaced intersects for a total of 160 intersects in 2019. The grid was placed systematically, the 0.8-m grid width centered within the 0.9-m width of each plot; the grid was first placed one-half of the 3.0 m plot length and then moved to the other half for the second count. In Iowa, counts were conducted using a 0.8 × 0.8-m grid with 49 evenly spaced intersects placed twice in each plot; the grid placed in the same way as described previously for the New Jersey location. Count data were pooled within each plot and not treated as subsamples. In Indiana, a 1.0 × 1.0-m grid with 81 evenly spaced intersects was placed once in the center of each plot. Grid intersect counts were transformed to express annual bluegrass as percent cover. Grid counts and final visual ratings were conducted on 30 Sept. in Indiana at 22 WAIT, 18 Oct. in Iowa (21 WAIT), and 10 Oct. in New Jersey (23 WAIT) in 2019 and on 15 Aug. in 2018 (14 WAIT) at both New Jersey sites.

Statistical analysis.

Data from all experiments were analyzed as a single-factor randomized complete block design (P = 0.05). Data were not transformed based on residual analysis (Shapiro–Wilk statistic) in SAS (Statistical Analysis Software, Inc., Cary, NC). Analysis of variance was performed using the mixed-model procedure in SAS and Fisher’s Protected least significant difference (P = 0.05) was used to compare means (Saxton, 2010). Treatment interactions with location were significant (P < 0.05), so data from each location are presented separately. Treatment and location effects were fixed and block effects were considered random in the mixed-model analysis.

Annual bluegrass control.

Amicarbazone alone controlled annual bluegrass ≤20% at all rating dates at both sites (Table 4). According to grid intersect counts at 14 WAIT, amicarbazone-alone programs did not reduce annual bluegrass cover compared with the nontreated control. In previous research, similarly poor annual bluegrass control was reported from two springtime amicarbazone applications at 100 g⋅ha⁻¹ in Indiana (McCullough et al., 2010).

Table 4.

Annual bluegrass control and cover following postemergence herbicide programs in Freehold, NJ, during 2018. Programs in 2018 were initiated on 26 Apr. on ‘Baron’ kentucky bluegrass and ‘Regenerate’ tall fescue. Annual bluegrass control was evaluated visually on a 0% (no injury or control) to 100% (complete necrosis and/or weed absence) percent scale relative to the nontreated control.

View Table

All mesotrione + amicarbazone tank-mixtures controlled annual bluegrass control more than amicarbazone alone from 4 to 14 WAIT at both sites. Tank-mixtures containing mesotrione at 175 g·ha⁻¹ provided more control than at 110 and 140 g·ha⁻¹ at 4 WAIT. Tank-mixtures containing mesotrione at 110, 140, and 175 g·ha⁻¹ provided similar annual bluegrass control from 6 and 14 WAIT at both sites, suggesting that sequential applications muted effects of rate. The tank-mixture containing mesotrione at 280 g·ha⁻¹ provided less control than other tank-mixtures at 6, 9, and 14 WAIT at the tall fescue site but not the kentucky bluegrass site. This indicates three sequential applications provide more control than two, just as Skelton et al. (2012) found. Although a direct statistical comparison cannot be made, treatments generally provided more control at the kentucky bluegrass site. We attribute the efficacy difference to a more severe annual bluegrass infestation at the tall fescue site (68% annual bluegrass cover compared with 32% at the kentucky bluegrass site 0 WAIT) combined with the rhizomatous growth of kentucky bluegrass that competes better with annual bluegrass.

Annual bluegrass grid intersect data support visual estimations. At the kentucky bluegrass site, grid counts determined annual bluegrass cover was 2% to 6% for mesotrione + amicarbazone tank-mixtures and 33% to 42% for amicarbazone alone and the nontreated. At the tall fescue site, annual bluegrass cover according to grid counts was 35% to 36% for tank-mixtures containing mesotrione at 110, 140, and 175 g·ha⁻¹ compared with 50% to 58% for amicarbazone alone and the nontreated. Control provided by amicarbazone alone was negligible in both experiments and would not be considered commercially acceptable.

Turfgrass injury.

Injury was only assessed in the kentucky bluegrass trial, as tall fescue density was too low to accurately assess injury. Injury was greatest at 4 WAIT and will be discussed but is not presented in a table. Amicarbazone alone at 53 and 70 g·ha⁻¹ caused <10% injury at 4 WAIT. The 110 g·ha⁻¹ regimen caused 15% injury, slightly more than the 6% injury reported from a similar regimen evaluated on kentucky bluegrass by McCullough et al. (2010). All amicarbazone + mesotrione combinations caused more injury than amicarbazone (53 g·ha⁻¹) alone. Amicarbazone + mesotrione at 280 g·ha⁻¹ caused 34% injury, more than amicarbazone (53 g·ha⁻¹) + mesotrione at 110 and 140 g·ha⁻¹, which caused 21% and 23% injury, respectively.

Annual bluegrass control.

The efficacy of amicarbazone + mesotrione tank-mixtures with UAN or urea + AMS was usually not different from the herbicides alone and is not discussed (Table 5). At 6 WAIT, mesotrione + amicarbazone tank-mixtures reduced annual bluegrass cover 59% to 83% in Indiana and at both New Jersey sites. Mesotrione and amicarbazone alone reduced annual bluegrass cover 27% to 35% and were less effective than the tank-mixture. In Iowa, the efficacy of all treatments was 88% to 100% at 6 WAIT.

Table 5.

Annual bluegrass cover reduction following sequential postemergence herbicide applications in Freehold, NJ (NJ), West Lafayette, IN (IN), and Ames, IA (IA) during 2019. Initial applications were made on 24 Apr., 25 Apr., and 24 May in New Jersey, Indiana, and Iowa, respectively. Two sequential applications were made at 2-week intervals following the initial application. Experiments were conducted on kentucky bluegrass and perennial ryegrass sites in New Jersey, perennial ryegrass in Indiana, and kentucky bluegrass in Iowa.

View Table

At 12 WAIT, efficacy across sites was more variable. In Iowa, all treatments reduced annual bluegrass cover by ≥85% and there were generally no differences between treatments. In Indiana, only amicarbazone + mesotrione at 175 g·ha⁻¹ reduced annual bluegrass cover more than mesotrione alone; otherwise, there were no differences among treatments. Treatment differences were most evident in New Jersey. At the kentucky bluegrass site in New Jersey, amicarbazone + mesotrione at 140 or 175 g·ha⁻¹ reduced annual bluegrass cover by 73% and 90%, respectively, whereas mesotrione alone did not reduce cover compared with the nontreated. At the perennial ryegrass site in New Jersey, amicarbazone alone was less effective than two of the three mesotrione + amicarbazone tank-mixtures 12 WAIT. Mesotrione alone was as effective as the tank-mixtures at the perennial ryegrass site 12 WAIT.

At the final rating date (10 October, 23 WAIT) at both New Jersey sites, trends were similar to those observed 12 WAIT, except that the efficacy of mesotrione alone had improved relative to the tank-mixtures. Amicarbazone tank-mixed with mesotrione at 140 and 175, but not 110 g·ha⁻¹ reduced annual bluegrass cover more than mesotrione alone at the kentucky bluegrass site. At the New Jersey perennial ryegrass site, mesotrione alone was as effective as the two of the three tank-mixtures.

In Indiana, there were no differences among herbicide treatments at the final visual and grid count evaluation on 30 Sept. (22 WAIT). The severity of the annual bluegrass infestation and summer decline of annual bluegrass at the Indiana site (annual bluegrass cover was reduced by 46% in the nontreated control at 12 WAIT) may explain the lack of treatment differences. No fungicides were applied in Indiana, which likely hastened the decline of annual bluegrass during summer. There were few treatment differences in Iowa, although treatments tended to be more effective in Iowa than Indiana. We attribute this to annual bluegrass immaturity and higher air temperatures at and after herbicide application (Table 3). Amicarbazone activity against annual bluegrass (McCullough et al., 2010; Yu et al., 2013) and mesotrione efficacy in other species (Gonçalves et al., 2020; Johnson and Young, 2002) increases with air temperature.

Turfgrass injury.

No injury was observed in Iowa or Indiana (data not presented). Injury observed in New Jersey on kentucky bluegrass in 2019 was similar to what was observed in 2018. Mesotrione alone caused less injury than tank-mixtures of amicarbazone + mesotrione tank-mixtures at 4 and 6 WAIT on all three turfgrass species (Table 6). Injury from amicarbazone and mesotrione alone caused <9% injury, except at 4 WAIT on kentucky bluegrass where amicarbazone caused 14% injury. Injury from amicarbazone + mesotrione tank-mixtures ranged from 10% to 30% and exceeded, at times, the threshold of 20% injury, which is deemed as unacceptable by some turf practitioners. The addition of UAN increased injury to kentucky bluegrass, but not perennial ryegrass or tall fescue. No injury was observed by 8 WAIT (4 weeks after the last application) in any species (data not presented).

Table 6.

Turfgrass injury following three sequential postemergence herbicide applications in Freehold, NJ, during 2019. Herbicide programs were initiated on 25 Apr. and sequential applications were made on 9 May and 22 May 2019 to separate kentucky bluegrass, perennial ryegrass, and tall fescue sites. Injury was evaluated visually on a 0% (no injury) to 100% (complete necrosis) scale relative to the nontreated control.

View Table

This research demonstrates that sequential applications of mesotrione + amicarbazone substantially improves control compared with amicarbazone alone. Improved efficacy from this tank-mixture was reported previously by Elmore et al. (2013a) on juvenile annual bluegrass that was thought to behave as an annual in overseeded bermudagrass. Our research was conducted where annual bluegrass behaved as a perennial due to management and climate. Highlighting the difference in herbicide efficacy, Elmore et al. (2013a) reported one application of amicarbazone at 75 g⋅ha⁻¹ provided substantially more annual bluegrass control than three sequential applications at 53 g⋅ha⁻¹ in this research. Except in Iowa, annual bluegrass control provided by amicarbazone in this research was poor, comparable to previous research evaluating amicarbazone at similar rates for annual bluegrass control in cool-season turf (McCullough et al., 2010). Whether mesotrione + amicarbazone improves control compared with mesotrione alone was less conclusive. In 2019, the tank-mixtures provided more control than mesotrione alone at three of the four sites 6 WAIT, but at only one site by the conclusion of the experiments. Unfortunately, mesotrione was not included as a standard in the 2018 research where mesotrione + amicarbazone provided ≥90% control in kentucky bluegrass. In both 2018 and 2019 in New Jersey, the tank-mixture was most effective at the kentucky bluegrass site which had the most mature annual bluegrass infestation.

Inconsistent annual bluegrass control across locations reported in our research is common. Reicher et al. (2011) reported 71% to 88% annual bluegrass control from three sequential applications of mesotrione at 175 g·ha⁻¹ in 1 year and 25% to 52% control when the experiment was repeated. Skelton et al. (2012) reported similarly inconsistent annual bluegrass control across experiments. This inconsistent control may also be due to genetic diversity between populations of this polyploid species (Chen et al., 2015; Mengistu et al., 2000) or annual bluegrass maturity and growth habit, which is not typically reported.

This research indicates that mesotrione + amicarbazone tank-mixtures are often more effective than amicarbazone alone for postemergence annual bluegrass control in the springtime. This tank-mixture was not always more effective than mesotrione alone. Turfgrass managers who experience poor efficacy of amicarbazone or mesotrione alone should consider this tank-mixture if transient turfgrass injury can be tolerated. This research found a tank-mixture of amicarbazone (53 g·ha⁻¹) with mesotrione at 110 g·ha⁻¹ applied three times on a 2-week interval is optimal. Higher rates of mesotrione (140 and 175 g·ha⁻¹) increased turfgrass injury in some instances and did not reliably increase annual bluegrass control.

Given the common theme of incomplete annual bluegrass control provided by herbicide programs in this and other research, future work should investigate multiyear herbicide programs for annual bluegrass control in conjunction with management practices that promote turfgrass competition. Mesotrione + amicarbazone tank-mixtures could be applied in the spring and preceded by ethofumesate applications in the autumn. Research should be conducted on mature annual bluegrass and consider management practices including irrigation, fungicide, fertilizer, management, and seeding that promote or prevent summer annual bluegrass decline.