Site Description

A greenhouse experiment was conducted in 2019 at the University of Nebraska Panhandle Research and Extension Center, located near Scottsbluff, NE (41.89°N, 103.68°W). The first and second runs of the experiment were planted on June 12 and August 8, respectively. Palmer amaranth and sugarbeet were grown separately in plastic square pots 9 cm wide and 8 cm deep (T.O. Plastics, Clearwater, MN), and were filled with Sungro^® potting mix (Sungro Horticulture, Agawam, MA). The potting mix contained no supplemental nutrients and was composed of 80% peat moss and 20% perlite. Greenhouse temperatures were maintained between 31 C (day) and 24 C (night). Supplemental lighting (P.L. Lighting Systems Inc., Beamsville, Ontario, Canada) was used to maintain a consistent photoperiod of 16 hours light and 8 hours dark. Plants were hand watered once daily throughout the duration of the experiments.

Treatment and Experimental Design

The experiment was laid out in a randomized complete block design with six and four replications in the first and second run, respectively. In the high-plains sugarbeet production region of Colorado and Nebraska, Palmer amaranth first begins to emerge approximately 1 wk before sugarbeet is planted (late April). Early-season weed control, (from planting to the two true-leaf stage of sugarbeet) is the most challenging time for control of Palmer amaranth. Therefore, Palmer amaranth and sugarbeet were planted at the same time in each run of the experiment to simulate a difficult weed-control scenario and subsequently were thinned to 1 plant pot^-1 within 2 d after emergence. Palmer amaranth seed was treated with 5 mL of 0.1 M KNO₃ solution to help ensure adequate germination (Buhler and Hoffman 1999). Herbicide treatments were applied with a Generation III Research sprayer (DeVries Manufacturing, Hollandale, MN) equipped with an 8002 EVS TeeJet^© nozzle (TeeJet Technologies, Wheaton, IL) calibrated to deliver the equivalent of 140 L ha^–1 spray solution.

Three Palmer amaranth accessions were evaluated in both runs of the experiment: a local field population from Scottsbluff, NE; a confirmed triazine-resistant population from south-central Nebraska (Jhala et al. 2014); and a field population from Hays, KS. The first run of the experiment included two sugarbeet varieties, Crystal^® W611NT GEM 100 and Beta^® BTS 60RR27 Pro 50 (both from Betaseed Inc., Shakopee, MN). In the second run of the experiment, Beta BTS 60RR27 Pro 50 was used again along with Hilleshog^® H7-1 (Hilleshog Seeds, Longmont, CO) and a noncommercial Hilleshog conventional variety.

The Betamix formulated premix of desmedipham (77.9 g ai L^–1) + phenmedipham (77.9 g ai L^–1) was applied at 274, 547, 1,090, and 2,190 g ai ha^-1, corresponding to each individual component being applied at 137, 274, 547, 1,090 g ai ha^-1. Desmedipham and phenmedipham were applied individually as the formulated products Betanex, and Spin-Aid, respectively, at 137, 274, 547, 1,090 g ai ha^-1, representing their equivalent individual rate in the Betamix product.

The second run of the experiment contained the same treatments as the first, with the additional rates of 137 and 4,380 g ai ha^-1 of Betamix and equivalent rates of desmedipham and phenmedipham of 68 and 2,190 g ai ha^-1. Herbicide treatments were mixed with 1.5 % vol/vol methylated seed oil. The maximum labeled field rate of Betamix (desmedipham 274 g ai ha^-1 + phenmedipham 274 g ai ha^-1) and Betanex (desmedipham 547 g ai ha^-1) for application to the cotyledon and two true-leaf stages of sugarbeet is 547 g ai ha^-1. Treatments of 547 g ai ha^-1 hereafter in this article are referred to as a field rate or 1× rate, depending on context.

Treatments were applied at two separate timings on the basis of the size of Palmer amaranth and growth stage of sugarbeet. At the first application, both Palmer amaranth and sugarbeet were at the cotyledon stage. Palmer amaranth was 2 cm tall and sugarbeet was 1.5 cm tall 4 d after emergence. At the second application, Palmer amaranth was 7 cm tall and sugarbeet was 5 cm at the two true-leaf stage 12 d after emergence. Palmer amaranth and sugarbeet were treated together on the same day with the same treatment mixtures. Plants were not watered until the day after herbicide application to ensure effective herbicide absorption.

Data Collection

Plant death was assessed as a binary response (alive or dead) 2 wk after treatment. Treated plants failing to regrow after treatment and lacking any green tissues were considered dead. After assessing plant death, aboveground biomass of Palmer amaranth and sugarbeet was harvested and oven dried for 72 h at 50 C before weighing.

Statistical Analysis

Data from each run of the experiment were analyzed separately. Analysis was done with R software (R Core Team, 2019) by nonlinear regression using the DRC package (Ritz et al. 2015). Sugarbeet varieties responded similarly to all herbicide treatments, as did Palmer amaranth accessions. Therefore, all sugarbeet varieties were combined, and all Palmer amaranth accessions were combined. Hereafter in this article, discussion of plant response to herbicide treatments refers to either sugarbeet or Palmer amaranth, but not individual varieties or accessions. Variety and accessions were combined using nonlinear mixed-effects regression with the medrc package (Gerhard and Ritz, 2016). An LL.2 model was used for survival response, with sugarbeet variety and Palmer amaranth accession added as random effects, as shown in Equation 1:

Biomass data were also combined using the medrc package (Gerhard and Ritz, 2016), and an LL.3 model was fit with variety and accession as random effects, as shown in Equation 2:

The selectivity index (SI) was calculated as described by Kniss and Streibig (2018) by dividing the estimated ED₁₀ of sugarbeet by the ED₉₀ of Palmer amaranth (Equation 3):

The SI was only calculated using the mortality data and not biomass data.

Table 1.

Parameter estimates and SEs of the two-parameter log-logistic model of survival of sugarbeet and Palmer amaranth treated at the cotyledon growth stage in a greenhouse experiment in 2019.

Table 2.

Parameter estimates (b, d, and e) and SEs of the three-parameter log-logistic model for sugarbeet and Palmer amaranth biomass treated with desmedipham and phenmedipham at the cotyledon stage in a greenhouse experiment in 2019.

Herbicide Dose-Response at Cotyledon Application Stage

Palmer amaranth response with desmedipham + phenmedipham resulted in an LD₉₀ that was lower than a full labeled rate (547 g ai ha^-1) in both runs of the experiment, indicating an effective level of control (Table 1). Phenmedipham and desmedipham treatments had LD₉₀ values greater than the 1× field rate in the first run of the experiment, indicating they did not provide effective control (Table 1). However, the LD₉₀ of both phenmedipham and desmedipham was reduced in the second run of the experiment, and they both provided effective Palmer amaranth control at the 1× equivalent field rate (Table 1).

Phenmedipham was inconsistent in Palmer amaranth control. Phenmedipham resulted in the highest LD₅₀ and provided almost 50% Palmer amaranth control at the 1× rate in the first run of the experiment. In the second run, phenmedipham provided effective Palmer amaranth control and resulted in a lower LD₅₀ and GR₅₀, compared with desmedipham + phenmedipham, and provided greater than 90% control at the 1× rate (Tables 1 and 2). Desmedipham resulted in the lowest LD₅₀ and GR₅₀ for Palmer amaranth in both runs of the experiment and provided the greatest control efficacy compared with other treatments.

At the cotyledon stage, death was not observed in sugarbeet at the highest rates of desmedipham or phenmedipham applied alone, corresponding to a 2× field rate, in the first run of the experiment. The additional increased 4× rate in the second run achieved sugarbeet death and allowed for an LD₅₀ to be calculated for all herbicide treatments in sugarbeet. Sugarbeet response to desmedipham + phenmedipham resulted in an LD₁₀ above a labeled rate, indicating crop safety (Table 1). Desmedipham caused the highest mortality rate in sugarbeet, as evidenced by the lowest LD₅₀ and GR₅₀, in the second run of the experiment (Tables 1 and 2). The LD₅₀ for desmedipham is 612 g ai ha^-1 (Table 1), relatively close to the field rate of 547 g ai ha^-1, indicating nearly 50% mortality, which is an unacceptable level of crop injury. Phenmedipham exhibited the greatest crop safety; it had the highest GR₅₀ of sugarbeet in the first run, and the highest LD₅₀ and a similar GR₅₀ to desmedipham + phenmedipham in the second run (Tables 1 and 2).

As a general trend, equivalent rates in the second run provided better Palmer amaranth control than in the first run for desmedipham and phenmedipham (Table 1). Although adding additional rates in the second run improved model fit, the increased sugarbeet and Palmer amaranth injury observed between experiments may be best explained by normal variation in herbicide response observed from repeating experiments.

Desmedipham + phenmedipham provided effective control of cotyledon-sized Palmer amaranth in both runs of the experiment at the 1× field rate (Figure 1). The individual desmedipham and phenmedipham treatments controlled cotyledon Palmer amaranth at the 1× rate in the second run of the experiment only (Figure 1). Desmedipham + phenmedipham provided greater consistency of Palmer amaranth control and resulted in a similar LD₅₀ for Palmer amaranth in both runs (Table 1).

Desmedipham caused an unacceptable level of sugarbeet injury at the 1× field rate, resulting in nearly 50% mortality (Figure 1). The estimated LD₁₀ for cotyledon sugarbeet treated with phenmedipham and desmedipham + phenmedipham is greater than the 1× field rate for both herbicides, indicating that a low mortality rate is expected from these treatments (Table 1). However, sugarbeet injury was greater in the second run, indicated by reduced LD₅₀ and LD₁₀ (Table 1). Desmedipham + phenmedipham showed the most potential for effective control of Palmer amaranth at the cotyledon stage; the combination provided consistent weed control and improved crop safety, compared to desmedipham and phenmedipham applied alone.

Table 3.

Parameter estimates and SEs of the two-parameter log-logistic model of survival of sugarbeet and Palmer amaranth treated at the two true-leaf stage and 7 cm height, respectively, in a greenhouse experiment in 2019.

Figure 1.

Survival of sugarbeet and Palmer amaranth treated at the cotyledon stage 2 wk after desmedipham and phenmedipham were applied alone or together. A two-parameter log-logistic model was used to determine the response. Horizontal error bars represent the SE of the e parameter (LD₅₀).

Herbicide Dose-Response at Two True-Leaf Application Stage

At the two true-leaf application timings, desmedipham and phenmedipham did not cause significant death in sugarbeet at the highest application rate in either run of the experiment (Table 3). Desmedipham + phenmedipham caused sugarbeet death, but only when applied at rates far exceeding labeled guidelines (Table 3). The estimated LD₅₀ and GR₅₀ were considerably lower in the first run compared to the second, which can be partially attributed to the additional higher rate in the second run improving model fit. The LD₅₀ of both sugarbeet and Palmer amaranth at the two true-leaf application was increased for all herbicides compared with the cotyledon application timing. This observation agrees with the conclusions of Dexter (1994) and Weinlaeder and Dexter (1972) that both sugarbeet and redroot pigweed gain tolerance to desmedipham and phenmedipham as plants increase in size.

Desmedipham resulted in the lowest LD₅₀ and GR₅₀ of Palmer amaranth (Tables 3 and 4). The higher efficacy of desmedipham to control 7-cm tall Palmer amaranth follows the pattern of greater efficacy observed with desmedipham when applied at the cotyledon stage. Desmedipham + phenmedipham resulted in a higher LD₅₀ for Palmer amaranth than desmedipham (Table 3). The 1× field rate of desmedipham + phenmedipham did not provide effective Palmer amaranth control in either run of the experiment; the highest level of control observed was near 50% (Figure 2). Phenmedipham did not cause Palmer amaranth death in the first run, and in the second run resulted in the highest estimated LD₅₀ (Table 3). Phenmedipham applied at four times the equivalent of a field rate only provided 50% Palmer amaranth control (Figure 2).

None of the herbicides provided effective control of 7-cm tall Palmer amaranth at equivalent labeled field rates. Desmedipham resulted in lower LD₅₀ and lower GR₅₀ values compared to phenmedipham, indicating that desmedipham provides greater efficacy in controlling Palmer amaranth, but desmedipham also caused greater sugarbeet injury compared with phenmedipham (Tables 3 and 4). Desmedipham contributed to more of the observed weed control activity compared to phenmedipham in the formulated Betamix product. The increased weed control efficacy of desmedipham compared to phenmedipham has been observed at the field level. Adbollahi and Ghadiri (2004) found the control of common lambsquarters and redroot pigweed improved when desmedipham + phenmedipham was applied, compared with phenmedipham applied alone. Hendrick et al. (1974) reported that redroot pigweed could metabolize phenmedipham at an accelerated rate compared to desmedipham, and a similar mechanism may be responsible for the increased efficacy of desmedipham compared to phenmedipham for control of Palmer amaranth.

Table 4.

Parameter estimates (b, d, and e) and SEs of the three-parameter log-logistic model for sugarbeet and Palmer amaranth biomass treated with desmedipham and phenmedipham at the two true-leaf sugarbeet growth stage in a dose-response experiment in 2019.

Figure 2.

Survival of sugarbeet at the two true-leaf stage and Palmer amaranth when 7 cm tall at 2 wk after desmedipham or phenmedipham were applied alone or together. A two-parameter log-logistic model was used to determine the response. Horizontal error bars represent the SE of the e parameter (LD₅₀).

Selectivity

A selectivity index greater than 2 indicates there is potential selectivity for a product to be used safely in a crop (Bartley 1993). At the cotyledon application stage, all three herbicide treatments achieved selectivity between Palmer amaranth and sugarbeet, meaning that the rate estimated to cause 90% mortality in Palmer amaranth was lower than the rate estimated to cause 10% mortality in sugarbeet. Achieving selectivity in a controlled dose-response experiment indicates there is potential for a field application of the herbicide at specific rates that will cause acceptable control of a weed species without excessive crop injury.

At the cotyledon application timing, desmedipham + phenmedipham had the highest calculated selectivity index (3.7 to 3.1), whereas desmedipham had the lowest (1.6) (Table 1). Palmer amaranth treated with desmedipham resulted in the lowest LD₉₀, indicating desmedipham can provide 90% Palmer amaranth control at a lower rate than phenmedipham or desmedipham + phenmedipham (Table 1). However, sugarbeet treated with desmedipham resulted in a lower LD₁₀ compared with other treatments, reducing the selectivity (Table 1). Phenmedipham provided less control of Palmer amaranth at equivalent rates to desmedipham but resulted in a greater selectivity index (2.5), due to reduced injury in sugarbeet, indicated by a higher LD₁₀ (Table 1).

At the two true-leaf application timing, selectivity could only be calculated for desmedipham + phenmedipham, because an LD₁₀ of sugarbeet could not be estimated for other treatments (Table 3). Selectivity of desmedipham + phenmedipham was not achieved in the first run of the experiment between Palmer amaranth and sugarbeet, and the resulting selectivity index was less than 1, which indicates the application rate required to achieve control of Palmer amaranth is greater than the rate causing 10% mortality in sugarbeet (Table 3). In the second run of the experiment, the selectivity index for desmedipham + phenmedipham is 2.2 (Table 3). Although a value greater than 2 demonstrates potential for use in the crop, an LD₁₀ in sugarbeet and an LD₉₀ in Palmer amaranth were achieved at a use rate of 1,082 g ai ha^-1, which is twice the maximum labeled rate.

Figure 3.

Palmer amaranth and sugarbeet, treated with desmedipham + phenmedipham at the cotyledon stage, 10 d after application. Plants are arranged starting with nontreated control at the left, moving to the right up to the 8× rate.

The selectivity of desmedipham + phenmedipham at the two true-leaf application was reduced compared with the cotyledon application (Tables 1 and 3). The tolerance of sugarbeet to desmedipham + phenmedipham was increased at the two true-leaf stage compared to the cotyledon stage, as evidenced by the increased LD₅₀ and GR₅₀. However, the tolerance of Palmer amaranth to desmedipham + phenmedipham also increased from the cotyledon stage and required a higher rate to reach 90% mortality. This increase in the rate required to control Palmer amaranth reduced the range between rates that would selectively control Palmer amaranth and not cause excessive injury in sugarbeet.

Practical Implications

The 1× rate of desmedipham + phenmedipham controlled Palmer amaranth and resulted in a low mortality rate in sugarbeet when applied at the cotyledon stage (Figure 1). However, death is not a complete indication of crop injury; growers applied desmedipham + phenmedipham knowing that substantial injury would occur. Figure 3 shows desmedipham + phenmedipham applied to cotyledon sugarbeet 10 d after treatment. Sugarbeet treated with the 1× labeled field rate had a high amount of foliar injury and severe stunting compared to the nontreated control. Sugarbeet treated with the 1× field rate were considered as surviving if they began regrowth; however, yield loss is expected from this level of crop injury. The estimated GR₅₀ for cotyledon sugarbeet treated with desmedipham + phenmedipham was near the 1× field rate in the first run and lower that the 1× field rate in the second run, revealing the high amount of growth reduction and crop injury caused (Table 3).

The potential for effective Palmer amaranth control at the two true-leaf application was not evident, because none of the herbicide treatments provided effective Palmer amaranth control without risking unacceptable rates of plant mortality. The 1× rate of desmedipham + phenmedipham resulted in 50% Palmer amaranth mortality, whereas the 1× desmedipham rate resulted in 75% Palmer amaranth mortality (Figure 2), neither of which is an acceptable level of control. The phenmedipham treatment provided the poorest control of Palmer amaranth, as indicated by the higher LD₅₀ and GR₅₀ values (Tables 3 and 4).

The level of sugarbeet injury sustained from the 1× desmedipham + phenmedipham treatment at the two true-leaf application shows improved crop tolerance, in comparison to the cotyledon stage (Figures 3 and 4). However, a labeled application at two true leaves does cause crop injury without providing an effective level of Palmer amaranth control. The maximum application rate of desmedipham and desmedipham + phenmedipham increased to 820 g ai ha^-1 when sugarbeet reached four true leaves; however, the possibility of control at this rate is not evident on the basis of the performance of equivalent applications to 7-cm tall Palmer amaranth. Furthermore, Palmer amaranth that emerged at the same time as sugarbeet would be taller than 7 cm when the four true-leaf stage is reached, further diminishing control.

The lack of control achieved on 7-cm tall Palmer amaranth reveals that the success of a rescue treatment with desmedipham, phenmedipham, or desmedipham + phenmedipham is not possible even when applying rates exceeding label recommendations. In a rescue situation, Palmer amaranth will be 10–20 cm tall when glyphosate failure is identified and a rescue treatment can be made. Anecdotally, growers in the Panhandle of Nebraska have attempted late-season rescue treatments with desmedipham + phenmedipham at rates exceeding 1,457 g ai ha^-1 and failed to control glyphosate-resistant Palmer amaranth. There is potential for desmedipham + phenmedipham to control cotyledon-sized Palmer amaranth. However, application timing would be highly critical to the success of control, and growers would have to accept the potential of a high level of crop injury. Desmedipham + phenmedipham may still be effective when used in a split application or microrate program in which single application rates were reduced to improve crop injury (Dale et al. 2006; Dexter 1994); however, returning to such intensive weed management is likely unpalatable after years of relying on glyphosate for broadleaf weed control in sugarbeet.

Figure 4.

Palmer amaranth and sugarbeet, treated with desmedipham + phenmedipham when they were 7cm tall and at the two true-leaf stage, respectively, at 10 d after application. Plants are arranged starting with nontreated control at the left, moving to the right up to the 8× rate.

Sugarbeet growers using desmedipham or phenmedipham in any type of POST program will have to accept the increased crop injury and herbicide cost compared to glyphosate (Kniss et al. 2004; Wilson 1999). Desmedipham and phenmedipham have limited utility for controlling Palmer amaranth in western Nebraska and will not provide control of Palmer amaranth as a rescue treatment. With desmedipham and phenmedipham no longer labeled for US sugarbeet, there is a critical need for alternative herbicides to control glyphosate-resistant weeds, especially early in the season.