Overcoming soil dust impact with paraquat and bentazon tank mixture on redroot pigweed (Amaranthus Retroflexus L.) control at different pH levels

Abstract

Background:  Dust, one of the destructive and growing atmospheric phenomena, often antagonizes the performance of herbicides. Under such a scenario, using herbicide tank mixtures may improve herbicide effectiveness.

Objective:  This study (2023/2024) investigated how different soil dust and pH agents affect the herbicidal activity and the tank mixture of paraquat and bentazon using the additive dose model (ADM) as a reference.

Methods:  Three separate greenhouse experiments were conducted on Amaranthus retroflexus L. using five doses of paraquat and bentazon. A three-parameter log-logistic model was used to obtain ED₅₀ and ED₉₀ values - a dose causing a 50% reduction in fresh weight. The binary mixture of herbicides were applied in various ratios as 0:100, 25:75, 50:50, 75:25, and 100:0 to obtain their effects in R and Excel® softwares.

Results:  The results indicated that dust reduced the effectiveness of both paraquat and bentazon herbicides in controlling redroot pigweed. The negative impact of dust on herbicide efficacy, measured by ED₅₀, increased as pH rose from 5 to 9. Specifically, soil dust increased the ED₅₀ of paraquat by 27–59% and bentazon by 29–57% across the pH range tested. Isobolographic analysis confirmed that paraquat and bentazon exhibited a mixture effect in controlling redroot pigweed. The combination of paraquat and bentazon showed strong synergistic effects at pH 5 and 7, indicated by λ-values of 3.34 and 2.45, respectively. At pH 9, the herbicide mixture’s effect shifted to additive, with a slight tendency toward antagonism (λ<1).

Conclusion:  These results suggest that tank mixing herbicides is an effective strategy to overcome the negative impact of dust, improving weed control and preventing weed flora shifts.

Keywords:
Dust Effect; Isobolographic Analysis; Herbicide Mixture; Reduced Dosage

1. Introduction

Dust pollution has become a growing critical issue in the world in recent years. Human activities, including changes in vegetation cover and land use, are driving forces behind both climate change and widespread environmental degradation, which in turn are major contributors to the emergence of air pollutants. Iran is exposed to numerous dust sources due to its proximity to the deserts of neighboring countries, primarily Iraq, Syria, and Saudi Arabia. The soil dust phenomenon can cause serious negative impacts on agricultural activities, threatening food security (Babaei et al., 2023). Both the quality and quantity of agricultural products are significantly impaired by the unwanted growth of weeds in agricultural fields. Crop yield reductions due to weeds are estimated to range from 10% to over 70% (Chitband et al., 2021). On the other hand, to achieve a 60% increase in the human food supply by 2050, chemical control is often favored for its cost- and time-effectiveness and as a powerful weed management approach in modern agricultural production (Prostko, 2018). However, effective weed control to maintain increased efficiency requires the timely and appropriate application of herbicides under suitable environmental conditions.

Dust storms have emerged as a significant environmental challenge in many regions, and their adverse effects on agricultural productivity are well-documented. Dust is primarily composed of soil dust particles (over 70%), along with organic and inorganic materials, and heavy metals such as Cu, Zn, Cr, Ni, Co, Pb, Cd, and Fe. These particles are typically less than 100 microns in diameter (Babaei et al., 2023). Factors potentially affecting herbicide absorption by dust and soil particles include these compounds, the pH, and the cation exchange capacity of the spray solution. Herbicide adsorption to soil is driven by the electrostatic attraction between soil colloid surfaces and herbicide molecules (Zhou et al., 2006). Paraquat is a cationic herbicide are strongly adsorbed to dust and soil particles via electrostatic interactions, a process that is considered almost irreversible. Hence, these interactions between dust particles and paraquat significantly reduce the herbicides’ effectiveness against weeds. Consequently, the reduced paraquat uptake by weeds diminishes the overall availability of the herbicide at the site of action (Rytwo, Tropp, 2001). Such can be attributed to the adsorption of positively charged paraquat molecules to negatively charged soil particles (mainly clays). Several previous studies have indicated that herbicide effectiveness against weeds can be greatly reduced when soil is present in spray water (Simarmata et al., 2017) or when soil dust settles on the weeds themselves (Rytwo, Tropp, 2001). Likewise, the effectiveness of various systemic herbicides– including glyphosate (Simarmata et al., 2017; Zhou et al., 2006), furamsulfuron (Nosratti et al., 2016), mesosulfuron, iodosulfuron and diflufenican (Shahbazi et al., 2015), endothal (Poovey, Skogerboe, 2004), floridon and imazethapyr (Babaei et al., 2023), and nicosulfuron (Nosratti et al., 2016)- is reduced by soil contamination of spray water or weed surfaces.

On the other hand, the adsorption at the solid/liquid interface is influenced by environmental factors, such as the pH of the spray solution and the ionic strength of the compound. Spray mixture pH is a critical factor affecting herbicide performance. Acidic or alkaline spray mixture pH can enhance or diminish herbicide absorption, transport, and efficacy by affecting the solubility, hydrolysis, dissociation, or chemical decomposition of the herbicide molecule. For weak-acid herbicides like bentazon, clethodim, sethoxydim, glyphosate, glufosinate, and 2,4-D, uptake through the leaf cuticle is favored by acidic rather than alkaline spray water conditions. Because they are protonated and non-ionic, these molecules can readily diffuse across the lipophilic plant cuticle. Spray mixture pH has a contrasting effect on weak-acid herbicide uptake. Increasing pH enhances solubility when low pH limits uptake. However, at pH levels above the herbicide’s pKa, the herbicide becomes ionic, hindering its passage through plant barriers (Matocha et al., 2006; Dan et al., 2009). Devkota and Johnson (2016) reported a 10–12% reduction in glufosinate efficacy on palmer amaranth (Amaranthus palmeri S. Watson) and giant ragweed (Ambrosia trifida L.) when spray water pH was increased from 4 to 9. To combat the negative impact of dust on herbicide effectiveness under different spray water pH, optimizing herbicide mixtures is crucial; this can also lead to reduced herbicide use.

Tank mixing herbicides to control weeds is an advanced strategy that enhances herbicide efficiency while reducing crop damage. Combining herbicides also optimizes farm management, broadens the weed control spectrum, and helps manage herbicide resistance. Studies indicate that herbicide interactions may occur physically or chemically within the spray solution, biologically within the plant, or during absorption and transport to the site of action, resulting in synergism (increased activity than expected), antagonism (decreased activity than expected), and additivity (no change in activity than expected) (Ritz et al., 2021).

Predicting the joint action of mixtures is not easy and requires a reference model to characterize the type of interaction. Herbicide interactions are predicted using two different reference models (Chitband et al., 2024; 2025). The Multiplicative Survival Models (MSM), assuming independent modes of action, predict herbicide interactions by considering that plants affected by one herbicide are no longer susceptible to the second. This approach is well-suited for qualitative and binary data, such as live/dead assessments. The Additive Dose Model (ADM) predicts herbicide interactions by assuming that one herbicide behaves as a diluted solution of the other. The combined effect of the mixture is then determined based on the potency of each herbicide, as quantified by its ED₅₀ value. This model is typically used to analyze quantitative data, including dry weight, relative growth, height, and seed weight (Streibig, Jensen, 2000; Teymourinia et al., 2023). Research has shown that the ADM model predicts the effect of tank mixing herbicides better than the MSM model. For example, Wehtje et al. (1992) reported synergistic effects on smallflower morningglory (Jacquemontia tamnifolia (L.) Griseb.) when bentazon, a photosystem II inhibitor, was applied in mixture with paraquat, a photosystem I electron diverter.

Although existing publications on the effect of dust on herbicide efficacy, no studies have addressed how to manage redroot pigweed (Amaranthus retroflexus L.) under dusty conditions using tank mixtures with varying pH levels. Also, redroot pigweed poses a significant threat in regions with frequent dust storms due to its widespread contamination of arable and wastelands. Therefore, tank mixing herbicides provides a novel and advanced approach for controlling invasive weeds like redroot pigweed under dusty conditions. Hence, the goals of this study were to: i) To evaluate the efficacy of three binary mixtures of paraquat and bentazon herbicides across a range of pH levels for reducing dust impact on redroot pigweed control, using the ADM Model as a reference; (ii) and to investigate and compare these binary mixtures using two different approaches to test for deviations from the ADM model.

2. Materials and Methods

2.1 Plants materials and growth conditions

Redroot pigweed seeds were collected from the research farm of the Faculty of Agriculture at Lorestan University. Before sowing, seed dormancy in redroot pigweed was broken by scarifying the seeds with 98% sulfuric acid (H₂SO₄) for two minutes, followed by washing according to the procedure described by Andersen (1968). The results showed a high germination rate, with up to 96% of the seeds germinating. From May to August in both 2023 and 2024, three dose-response experiments were conducted in the greenhouse at the Faculty of Agriculture, Lorestan University, Iran (32°3’ N, 48°21’ E; 1117 m altitude). Dose-response and isobole curves were constructed using data representing the average response over two years. The greenhouse was equipped with an intelligent meteorological system to regulate environmental conditions. The system maintained a photoperiod of approximately 16:8 h (light: dark), and the temperature ranged from 23–26 °C during the day and 17–19 °C at night, with relative humidity levels of 46% during the day and 67% at night. Upon seedling emergence, the plants were thinned to a density of five seedlings per pot at the cotyledon stage. Subsequently, each pot received 30 mL of a water-soluble N-P-K fertilizer solution (20:20:20).

2.2 Chemicals and experimental design

Redroot pigweed plants were exposed to five doses of commercial formulations of paraquat (Gramoxone, 200 g a.i. L⁻¹, Partonar, Iran, dose range 9.38, 18.75, 37.5, 75, and 150 g a.i. ha⁻¹) in combination with bentazon (Bazageran, 480 g a.i. L⁻¹, Ariashimi, Iran, dose range 22.5, 45, 90, 180, and 360 g a.i. ha⁻¹) in a dose-response experiment under three different pH levels (5, 7, and 9). All herbicides were applied as single treatments or in binary mixtures with fixed ratios. The ratios of the herbicides in the binary mixtures were selected to represent a range of combinations: 100:0, 75:25, 50:50, 25:75, and 0:100. Carrier water pH was adjusted to 5, 7, or 9 using acidic or alkaline solutions. Sulfuric acid (98% liquid; Merck, Germany) at 0.26 mL and potassium hydroxide (90% solid flakes; Unid, Korea) at 0.39 g were each dissolved in 1 L of deionized water to prepare solutions with pH values of 5, 7, and 9, respectively. A pH meter was used to measure the pH following the addition of each solution. The experiments were conducted using a completely randomized factorial design with four replications of five doses for each herbicide mixture. Ten untreated controls were also included across three experiments.

2.3 Soil dust preparation

To prepare the dust, soil samples were crushed, air-dried, and then sequentially sieving through 25, 50, 100, and 200 mesh screens. The resulting particles, with a maximum size of 75 microns, were collected and used to simulate dust exposure by sprinkling them onto the pots within a dust chamber. The results of the physicochemical analysis of the soil used for dust generation are summarized in Table 1. Dust pollution was applied using a simulator. The dust chamber, measuring 45 cm long, 45 cm wide, and 120 cm high, was constructed of 5 mm thick transparent plastic. Four 50-mesh sieves (14 cm diameter openings) were interconnected with wire and attached to the chamber wall via a pull mechanism. A spring (45 cm long, 2 cm diameter) was mounted 3 cm above the sieves on the chamber wall. Prepared soil is poured into the sieves through a door located on the top of the chamber. A rope connected to the center of the spring and extending through the door is pulled to vibrate the spring, which in turn vibrates the sieves. A thin rayon fabric at the base of each sieve ensures uniform distribution of the soil dust onto the redroot pigweed foliage. After dusting, the dust within the chamber was allowed to settle on the aerial parts of redroot pigweed for 15 minutes. The plants were then exposed to the herbicide (Zhou et al., 2006; Babaei et al., 2023). At the four-to-six-leaf stage, binary mixtures were applied using an overhead trolley sprayer. The sprayer was equipped with a flat-fan nozzle (8002 Tee-jet; Spraying Systems Company, Wheaton, IL, USA) and delivered 200 L ha⁻¹ at 300 kPa. The physicochemical properties of the water used for foliar spraying of the experimental treatments are shown in Table 2. Four weeks after treatment, redroot pigweed plants were harvested, and their fresh weight were recorded.

Table 1
Properties of the analyzed soil dust samples in both experimental years.

Sampling depth (cm)	Soil texture	Clay (%)	Sand (%)	Silt (%)	Electrical Conductivity (ds m−1)	Organic matter (g kg−1)	Nitrogen (mg kg−1)	Phosphorus (mg kg−1)	Potassium (mg kg−1)	pH
0-30	Silty-clay	35	13	39	0.59	1.01	0.13	7.3	301	6.99

Table 2
Properties of the water used in the analysis for spraying experimental treatments.

Water sample	pH	Na (%)	Ca (%)	Mg (%)	SO4 (%)	Cl (%)	HCO3 (%)
Sprayer water	6.92	0.51	0.78	0.98	3.93	2.3	2.89

2.4 Statistical analysis

An F-test, based on residual sums of squares from the models, was used to compare the nested regression models and assess whether adding predictors significantly improved model fit. A Box-Cox transform-both-sides approach was implemented to address non-homogeneity of variance. The goodness of fit was then evaluated using graphical analysis of residuals and an F-test for lack of fit. R statistical software version 3.5.1 (R Core Team, 2021) with the drc add-on package (Ritz et al., 2021) was used for data analysis. ADM isobolograms were also created using Microsoft Excel 2019. Visual assessments of herbicide efficacy on redroot pigweed were conducted at 14 and 28 days after treatment, using the EWRC scale ranging from 0% (no control) to 100% (complete control) (Chitband et al., 2021; Teymourinia et al., 2023).

2.5 Dose-response analysis

A three-parameter log-logistic dose-response model was fitted to the fresh weight data of redroot pigweed in response to different mixture ratios (Streibig, Jensen, 2000):

(1) $U=\frac{D}{\left[1+{\left(\frac{Z}{ED50}\right)}^B\right]}$

Where U represents the fresh weight at dose Z, D is the upper limit of the fresh weight at zero dose, ED₅₀ (Effective Dose) is the dose required to reduce weed fresh weight by 50%, and B is a parameter proportional to the slope of the curve near ED₅₀.

2.6 Joint action analysis

The mixing ratio of the two herbicides at each ratio tested was calculated at predefined response levels (ED₅₀ and ED₉₀) using the ADM model, according to the formula below:

(2) ${\partial }_1=r\cdot {\partial }_2=d_1+r\cdot d_2$

Where ∂₁ and ∂₂ are the ED_x values of the pure herbicides, and d₁ and d₂ are the doses of those herbicides in the mixture, the resulting (d₁ + rd₂) of the mixture, calculated at ED_x, is expressed in units of ∂₁.

Based on the above assumptions, the relative potency (R) between herbicides ∂₁ and ∂₂ was calculated using the following relationship:

(3) $R={\partial }_1/{\partial }_2$

The relative potency (R) represents the difference in biological exchange rate between two herbicides when applied individually. Three isobole models were used to evaluate the effects of the binary herbicide mixtures. The first model, known as the concentration-addition model, is a simple linear model defined as follows (Streibig, Jensen, 2000; Chitband et al., 2024):

(4) $\frac{d_1}{{\partial }_1}+\frac{d_2}{{\partial }_2}=1$

The second isobole model, the Hewlett model, is more complex than the concentration-addition model, incorporating an additional parameter, lambda (λ), which indicates the deviation of the isobole (convexity or concavity) and is calculated as follows:

(5) ${\left(\frac{d_1}{{\partial }_1}\right)}^{1/\lambda }+{\left(\frac{d_2}{{\partial }_2}\right)}^{1/\lambda }=1$

The λ-value represents the interaction parameter between the herbicides. When λ = 1, the herbicides interact additively, following the ADM. Values of λ less than 1 indicate antagonism, while values greater than 1 indicate synergism. In addition to the previous models, the Voelund isobole model has been proposed, incorporating two eta parameters (η₁ and η₂) to describe the isobole curvature. This model exhibits better convergence than the Hewlett model and can describe asymmetric isoboles as explained below:

(6) ${\left(\frac{d_1}{{\partial }_1}\right)}^{n_1}{\left(\frac{d_1}{{\partial }_1}+\frac{d_2}{{\partial }_2}\right)}^{1-n_1}+{\left(\frac{d_2}{{\partial }_2}\right)}^{n_2}{\left(\frac{d_1}{{\partial }_1}+\frac{d_2}{{\partial }_2}\right)}^{1-n_2}=1$

The η-value indicates asymmetric isoboles. In this model, values of η greater than one indicate antagonistic effects, while values of η less than one indicate synergistic effects.

According to ADM, the ED₅₀ of a mixture (ED_50mix) can be calculated, as well as other ED_x values (e.g., ED_90mix), based on the ED₅₀ and ED₉₀ values of the individual compounds and their proportions in the mixture (Streibig, Jensen, 2000; Chitband et al., 2025).

(7) $E{D}_{50or90mix}=E{D}^{50or90A}/\left(\alpha +\left(1-\alpha \right)RP\right)$

where ED_{50 or 90A} are the ED₅₀ and ED₉₀ doses of the herbicide 1, α is the ratio of herbicide 1 in the mixture, and RP is the relative potency, similar to Eq. (3).

In Excel, the isobole representing the ADM can be constructed by defining the ED₅₀ and ED₉₀ doses of each herbicide in the mixture as coordinates on the x and y axes and drawing a straight line between them. This visual representation allows for a comparison of the observed ED₅₀ and ED₉₀ values with the expected isobole shape under the ADM. The position of data points relative to the ADM isobole revealed the nature of the herbicide interaction. Points above the line suggested an antagonistic effect (less than predicted by ADM), while points below the line indicated a synergistic effect (greater than predicted by ADM). Furthermore, we investigated whether the ED₅₀ and ED₉₀ values predicted by Equation 8 for the herbicide mixture were consistent with the experimentally derived ED₅₀ and ED₉₀ values from Equations 3 and 5, specifically checking if the predicted values fell within the 95% confidence intervals of the experimental values. Significant deviations from the ADM isobole were classified as synergism when the observed ED₅₀ and ED₉₀ values were lower than the values predicted by the ADM, and as antagonism when the observed values were higher. To quantify the relative amount of a chemical in a mixture and to assess the degree of interaction, the sum of toxic units (ΣTU) is calculated. The ΣTU at the 50:50% effect ratio (ΣTU_50:50) is particularly useful for measuring the deviation from additivity (ADM) in symmetric isoboles. The formulas for calculating ΣTU_50:50 differ depending on the isobole model. For the Hewlett isobole model, the ΣTU_50:50 is calculated as 2^1-λ (Ritz et al., 2021; Chitband et al., 2024; 2025).

3. Results

3.1 Individual analysis

The individual impacts of paraquat and bentazon herbicides on redroot pigweed control were evaluated under different spray tank pH conditions. A goodness-of-fit test indicated that the three-parameter logistic function provided the most accurate representation of the dose-response data under all experimental conditions. Therefore, this function was chosen to fit the observed dose-response relationships in the data. Furthermore, a lack-of-fit test applied to the fresh weight data of redroot pigweed at the 5% level using the three-parameter logarithmic model for paraquat and bentazon was not significant (Table 3), indicating a good fit of the model to the data.

Table 3
The curve slope, upper limit, and ED₅₀ and ED₉₀ values of the three-parameter log-logistic model for paraquat and bentazon herbicides in pure application on the fresh weight of Amaranthus retroflexus under application of different pHs and soil dust 0 and 20 kg.ha⁻¹ in four to six leaves stage in 28 days after treatment (DAT).

Soil Dust (Kg.ha−1)	Herbicide (g.a.i.ha−1)	Ba	Db	Herbicide effective dose (g.a.i.ha−1)		Lack of fit test (5%)
				ED50c	ED90
pH = 5
Zero	Paraquat	1.36 (0.44)**	24.63 (6.89)*	203.16 (17.14)*	630.07 (75.09)*	0.19 (NS)d
	Bentazon	1.46 (0.48)*	24.50 (6.93)*	150.15 (12.22)**	593.03 (61.18)*	0.13 (NS)
Twenty	Paraquat	1.81 (0.84)**	24.71 (7.81)**	278.10 (40.12)*	904.16 (311.27)**	0.17 (NS)
	Bentazon	1.92 (0.94)**	24.88 (7.89)**	211.08 (36.05)**	879.52 (284.01)*	0.29 (NS)
pH = 7
Zero	Paraquat	1.26 (0.41)**	24.96 (7.48)**	179.36 (29.94)**	748.06 (189.90)*	0.59 (NS)
	Bentazon	1.36 (0.47)**	24.59 (7.43)*	158.75 (22.28)**	832.56 (216.66)**	0.35 (NS)5
Twenty	Paraquat	1.61 (0.64)*	24.99 (7.74)**	322.70 (53.72)*	1194.50 (410.57)**	0.97 (NS)
	Bentazon	1.31 (0.44)**	24.77 (6.99)**	269.82 (44.35)**	1179.03 (460.21)*	0.89 (NS)
pH = 9
Zero	Paraquat	1.63 (0.62)*	24.66 (7.11)*	206.16 (33.22)*	796.25 (198.12)*	0.39 (NS)
	Bentazon	1.59 (0.57)*	24.81 (7.59)**	186.32 (30.21)**	889.79 (263.13)**	0.34 (NS)
Twenty	Paraquat	1.63 (0.62)*	24.96 (7.79)**	503.13 (53.71)*	1412.54 (613.27)*	0.91 (NS)
	Bentazon	1.49 (0.47)**	24.55 (7.41)*	436.09 (48.39)**	1389.18 (595.83)**	0.83 (NS)

• All parameter values are presented with standard errors (SE) in parentheses. ^aSlope around the ED₅₀. ^bUpper limit, when herbicide doses are zero. ^cEffective dose: The dose causes 50% decrease in response (weed fresh wieght). ^dNS, *and **: Non-significant at the 5% level, Significant at 5% and 1% levels of probability, respectively.

The effects of paraquat and bentazon, applied alone or in combination with dust at varying pH levels, are summarized in Table 3 and Figure 1. Based on the results, the efficacy of paraquat and bentazon were reduced by the presence of soil dust on redroot pigweed, so the ED₅₀ values for paraquat and bentazon at pH = 5 increased from 203.16 and 150.15 g a.i. ha⁻¹ to 278.10 and 211.08 g a.i. ha⁻¹, respectively, in the presence of soil dust compared to conditions without dust. To reduce redroot pigweed biomass by 50% under conditions without dust at pH 7 and 9, 179.36 to 206.16 and 158.75 to 186.32 g a.i. ha⁻¹ for paraquat and bentazon, respectively, were required. Similar increases in ED₅₀ and ED₉₀ values for paraquat and bentazon, due to the presence of dust, were also observed at pH = 7 and 9. The antagonistic effect of soil dust on paraquat and bentazon efficacy, as measured by ED₅₀ and ED₉₀, was significantly greater at pH 9 than at pH = 7. Hence, to achieve a 50% reduction in redroot pigweed biomass in the presence of soil dust, higher dosages of both paraquat and bentazon were necessary at pH = 9 than at pH = 7. Specifically, dosages of 322.70 to 503.13 g a.i. ha⁻¹ of paraquat and 269.82 to 436.09 g a.i. ha⁻¹ of bentazon were required at pH = 9 (Table 3). The greatest reduction in both herbicides’ efficacy under dust conditions was observed at pH = 9, significantly different from other pHs. Soil dust conditions increased the ED₅₀ values of paraquat and bentazon at pH = 5 by 26.95% and 28.87%, respectively. The increased values were even more pronounced at higher pH levels, rising by 44.41% and 41.26% at pH = 7, and 59.02% and 57.27% at pH = 9, indicating reduced efficacy of paraquat and bentazon in soil dusty conditions compared to soil dust-free conditions.

Figure 1
Response of biomass in redroot pigweed affected by different doses of paraquat and bentazon in the presence of 20 kg.ha⁻¹ soil dust and pH agents (5, 7, and 9) in the carrier water. A three-parameter log-logistic model was used for data fitting.

3.2 Joint action analysis

Isobole analyses were conducted to evaluate the interaction of three combinations of herbicides. Table 4 lists the calculated parameters, including ED₅₀, λ-values, and the sum of toxic units (ΣTU) at the 50:50% effect ratio, along with related statistics. Figures 2, 3, and 4 illustrate the isobolograms of these combinations, examined at three different pH levels. Additionally, six different mixture ratios of paraquat and bentazon on redroot pigweed were evaluated in Excel.

Figure 2
Isobolograms of mixtures of paraquat (Gramoxone) + bentazon (Bazagaran) at pH = 5 of the carrier water on redroot pigweed control with two software programs approaches for the first experiment; presented by three parameters with logistic logarithm in R software (a-c). The points represent the ED₅₀ and their corresponding standard error at 95% confidence intervals. The straight dot-dashed line is the ADM isobole; the curved solid line is the best-fitting isobole model, and the estimated ADM line based on all mixtures when there is a significant interaction. By Excel (d); at the ED₅₀ (•) and ED₉₀ (♦) response levels. Bars indicate 95% confidence intervals for the estimated ED₅₀ and ED₉₀ doses. The doses have been scaled so that the doses of the herbicides applied separately are 1.0

Figure 3
Isobolograms of mixtures of paraquat (Gramoxone) + bentazon (Bazagaran) at pH = 7 of the carrier water on redroot pigweed control with two software programs approaches for the second experiment; presented by three parameters with logistic logarithm in R software (a-c). The points represent the ED₅₀ and their corresponding standard error at 95% confidence intervals. The straight dot-dashed line is the ADM isobole; the curved solid line is the best-fitting isobole model, and the estimated ADM line based on all mixtures when there is a significant interaction. By Excel (d); at the ED₅₀ (•) and ED₉₀ (♦) response levels. Bars indicate 95% confidence intervals for the estimated ED₅₀ and ED₉₀ doses. The doses have been scaled so that the doses of the herbicides applied separately are 1.0

Figure 4
Isobolograms of mixtures of paraquat (Gramoxone) + bentazon (Bazagaran) at pH = 9 of the carrier water on redroot pigweed control with two software programs approaches for the third experiment; presented by three parameters with logistic logarithm in R software (a-c). The points represent the ED₅₀ and their corresponding standard error at 95% confidence intervals. The straight dot-dashed line is the ADM isobole; the curved solid line is the best-fitting isobole model, and the estimated ADM line based on all mixtures when there is a significant interaction. By Excel (d); at the ED₅₀ (•) and ED₉₀ (♦) response levels. Bars indicate 95% confidence intervals for the estimated ED₅₀ and ED₉₀ doses. The doses have been scaled so that the doses of the herbicides applied separately are 1.0

3.2.1 Paraquat:Bentazon at pH = 5

The experiment investigating the effect of mixing paraquat and bentazon on redroot pigweed at pH = 5 (the first experiment) revealed no significant difference between four-parameter and three-parameter logistic logarithm models, as indicated by a non-significant F-test (p = 0.16). Therefore, the simpler three-parameter model was used to fit the data. Furthermore, the lack-of-fit test for the three-parameter logistic model was not significant (p = 0.45) (Table 4), indicating a good fit to the data. On the other hand, the construction of isoboles was justified by the non-significant F-test (p = 0.09), which indicated no significant difference between models with independent and identical upper and lower limits. Based on the primary hypothesis of no synergistic or antagonistic effects in the mixture, the F-test between the freely model and the concentration addition (CA) model was non-significant (p = 0.92), indicating a good fit of the CA model to the data. Next, the Hewlett model was fitted to the paraquat and bentazon mixture data. The significant F-test (p = 0.05) between the CA and Hewlett models provides evidence that the mixture’s interaction could be better described by the Hewlett model. Furthermore, the Hewlett model’s λ-value (λ = 3.34) supports the presence of synergistic effects between paraquat and bentazon at pH = 5 (Table 4). Turning to the shape of the isobole for paraquat + bentazon at pH = 5 in the control of redroot pigweed exhibited complete inward curvature (concave), indicating synergistic effects (Figure 2). Finally, the Voelund model was fitted to the paraquat + bentazon mixture data at pH = 5. However, the F-test between the Hewlett and Voelund models was not significant (p = 0.39), indicating that the Voelund model did not provide a good fit. This conclusion is supported by the figure generated from the Voelund model fitting (data not shown). Therefore, it can be concluded that the Hewlett model best describes the interaction between paraquat and bentazon at pH = 5. The sum of toxic units (ΣTU) for a 50:50%-mixture was calculated as 0.20, indicating that approximately 20% of the chemicals are needed to achieve the same 50% effect compared to the prediction of ADM. In Excel, all dose mixtures of paraquat:bentazon at pH = 5 at the ED₅₀ and ED₉₀ response levels were all located inside the isobole, suggesting completely and potent synergistic interactions (Figure 2).

Table 4
Summary estimates of statistical parameters of the three-parameter log-logistic model for three separate experiments (the binary herbicide combinations tested; paraquat:bentazon under application of different pHs; 5, 7, and 9) on Amaranthus retroflexus. Hewlett isobole model for the λ-values is given with standard errors four weeks after treatment. The sum of toxic units (∑TU) of the 50:50 effect ratio is given under the ADM model as a predefined response of the mixtures.

Mixtures	Ratios	Slope (b)	Upper limit (d)a	ED50	λ	∑TU50:50	Lack of fit test (5%)	Relative potency (RP)d
pH = 5 (Expt I)
Paraquat:Bentazon	100:0	2.48 (0.48)**	24.23 (2.89)b**	52.29 (3.40)**
	75:25	2.17 (0.38)*	24.94 (2.46)*	31.25 (2.95)*
	50:50	2.03 (0.23)**	24.04 (2.15)**	13.08 (3.33)*	3.34 (0.73)	0.20	0.45 (NS)c	0.22
	25:75	2.11 (0.21)*	24.43 (2.34)*	31.39 (2.83)**
	0:100	2.45 (0.44)**	24.03 (2.35)**	237.93 (27.92)*
pH = 7 (Expt II)
Paraquat:Bentazon	100:0	1.69 (0.89)**	24.24 (6.24)*	54.55 (4.46)**
	75:25	1.30 (0.85)*	24.57 (6.48)**	48.59 (4.09)**
	50:50	1.36 (0.27)**	24.22 (5.62)*	30.92 (4.87)**	2.45 (0.27)	0.37	0.47	0.19
	25:75	1.65 (0.30)**	24.62 (6.33)**	42.27 (5.76)*
	0:100	1.09 (0.19)**	24.98 (6.12)*	286.98 (53.07)**
pH = 9 (Expt III)
Paraquat:Bentazon	100:0	2.24 (0.29)**	24.51 (6.53)*	691.88 (65.64)**
	75:25	2.33 (0.36)*	24.15 (6.13)**	578.14 (39.05)*
	50:50	2.64 (0.65)**	24.78 (6.79)*	473.99 (17.27)**	0.87 (0.16)	1.09	0.05	2.18
	25:75	2.83 (0.86)*	24.33 (6.39)**	432.53 (26.51)*
	0:100	3.05 (0.93)**	24.72 (6.76)**	318.04 (19.93)**

• In all experiments, the test for lack of fit was not significant, indicating that the logistic model was able to describe the data better than an ordinary ANOVA. ^aThe upper limit when the dose is zero. ^bStandard errors are in parentheses. ^cNS, * and **: Non-significant at the 5% level, Significant at 5% and 1% levels of probability, respectively. ^dThe relative potency is for the two herbicides applied alone.

3.2.2 Paraquat:Bentazon at pH = 7

For a binary mixture of paraquat:bentazon at pH = 7 (the second experiment), the data were analyzed using a three-parameter logarithmic logistic model because the F-test between four- and three-parameter models was non-significant (p = 0.33) and the lack-of-fit test was also non-significant (p = 0.47) (Table 4). The F-test between the freely and the CA models was non-significant (p = 0.43), but the difference between the CA and Hewlett models was significant. Furthermore, the F-test provided no evidence to reject the Hewlett model in favor of the Voelund model (p = 0.55), suggesting that the Hewlett model adequately describes the effects of mixtures. Turning to the isobologram, all mixture ratios were significantly lower than the ADM isobole (Figure 3). Similar to the previous results, the isoboles obtained for the paraquat + bentazon mixture at pH = 7 displayed a concave curvature, further confirming the synergistic interaction between these two herbicides in controlling redroot pigweed. The isobole-derived λ value of 2.45 (calculated in R software) for the paraquat:bentazon mixture at pH = 7 indicates synergistic interactions with the ADM model, as shown in Table 4.

The sum of toxic units (ΣTU) of 0.37 was calculated for the 50:50 paraquat:bentazon mixture at pH = 7, demonstrating a deviation from the ADM. Only 37% of the amount of herbicide predicted by the ADM was needed to reduce the response by 50%. In Excel, all three mixture ratios at the ED₅₀ response level were placed below the isobole, indicating that the ED₅₀ was lower than predicted by the ADM and confirming a synergistic interaction. At the 90% effect level, one observation was significantly located under the ADM and exhibited synergistic behavior, while the remaining two observations were below the ADM line; there was no statistically significant evidence to suggest that they deviated from the ADM and followed the ADM in the 95% confidence interval (Figure 3).

3.2.3 Paraquat:Bentazon at pH = 9

For the third combination (paraquat:bentazon at pH = 9), the three-parameter logarithmic logistic model was also chosen due to the non-significant F-test for the three-parameter logarithmic logistic model and the lack-of-fit test (p > 0.05) (Table 4). The F-test showed no significant difference (p = 0.07) between the freely and the CA models for paraquat:bentazon mixtures at pH = 9. Although the CA and Hewlett models differed significantly (p = 0.04), the F-test revealed no significant difference (p = 0.16) between the Hewlett model and the more complex Voelund model. Therefore, the Hewlett model is likely the most appropriate for describing the effects of the paraquat:bentazon mixtures at pH = 9. The ED₅₀, λ-estimates, ΣTU_50:50, and related statistics for this combination are presented in Table 4. Unlike the previous two mixtures, the paraquat:bentazon mixture at pH = 9 exhibited additive or weakly antagonistic effects according to the ADM model, as indicated by a λ-value below 1 (λ = 0.87), calculated using R software. Also, turning to the isobologram, the isobole curves for the paraquat:bentazon mixture at pH = 9, controlling redroot pigweed, exhibit an outward (convex) curvature, demonstrating an antagonistic interaction (Figure 4). Therefore, using ADM as a reference model, the Hewlett model provided a better fit to the data than the CA model, as shown by the significant F-test mentioned earlier. However, because its λ-value was close to one, the combination of paraquat + bentazon in controlling redroot pigweed is best described as weakly antagonistic to additive. The ΣTU_50:50 value for the paraquat:bentazon mixture at pH = 9 was highest (1.09) (Table 4), and its mixture ratios were well-distributed along the isobole compared to the other two mixtures. In Excel, at the 50% response level, two of the mixture ratios were placed on the ADM line and did not deviate significantly from the assumption of additivity of doses, and one observation was situated above the isobole, indicating lower responses than predicted by ADM, i.e., showing antagonistic effects. At the 90% response level, all observations generated ED₉₀’s significantly higher than predicted according to ADM; that is, all mixture ratios were located above the isobole, indicating antagonism (Figure 4).

3.3 Visual observations analysis

Paraquat, as a photosynthesis inhibitor, rapidly induced damage symptoms. Treated plants exhibited wilting, rapid chlorosis, and necrosis along the leaf margins. Subsequent symptoms progressed to leaf desiccation and scorching, eventually leading to browning and complete desiccation. In dusty conditions, paraquat alone, applied at 75 and 150 g a.i.ha⁻¹, controlled 76-78.33%, 73.33-76.67%, and 66.67–70.33% of redroot pigweed at pH 5, 7, and 9, respectively. In the same conditions, bentazon alone (180 and 360 g a.i.ha⁻¹) provided more effective control of redroot pigweed than paraquat, with control ranging from 81.67–86.67%, 78.33–83.33%, and 73.33–80% across pH levels of 5, 7, and 9. Bentazon caused similar damage symptoms to paraquat due to its similar mode of action. In dusty conditions, paraquat generally provided inferior control of redroot pigweed compared to bentazon when applied alone. However, given the widespread prevalence of this weed in different areas of Iran and the potential for resistance development with repeated bentazon applications, it is advisable to use bentazon in tank mixtures with other herbicides to reduce the risk of resistance. This is particularly important in soil dusty conditions where weed spraying is necessary (Table 5).

Table 5
The visual observations injury (EWRC) of Amaranthus retroflexus L. 14 and 28 days after paraquat:bentazon application of pH 5, 7, and 9 at four-to six-true leaf stage.

	Individual herbicidal dose	pH = 5		pH = 7		pH = 9
		14 Dayb	28 Day	14 Day	28 Day	14 Day	28 Day
Mixture ratio	g a.i. ha−1a	Control Percentage		Control Percentage		Control Percentage
Control	-	0	0	0	0	0	0
100:0	9.38:0	5.67e	23.33d	3.33d	16.67d	3.67d	10d
"	18.75:0	15.33d	45c	8d	35c	8.67d	26.67c
"	37.5:0	39.33c	68.33b	25.33c	53.33b	18.67c	46.67b
"	75:0	53b	76a	46.33b	73.33a	38.67b	66.67a
"	150:0	65.33a	78.33a	58.67a	76.67a	45.33a	70.33a
75:25	7.03:5.63	10.33d	38.33d	6.67d	18.33e	4.33c	11.67d
"	14.06:11.25	35.33c	63.33c	13d	38.33d	8.33c	31.67c
"	28.13:22.5	51b	80b	30.33c	58.33c	23.67b	51.67b
"	56.25:45	65.33a	93.33ab	51.33b	76.67b	43a	70a
"	112.5:90	73a	100a	63.33a	88.33a	48.67a	76.67a
50:50	4.69:11.25	15.33c	45d	13d	40d	8.33c	16.67d
"	9.38:22.5	51b	80c	36.67c	65c	13c	36.67c
"	18.75:45	63.67a	91.67b	58.67b	85b	28.67b	56.67b
"	37.5:90	69.33a	98.33a	67ab	95a	48a	75a
"	75:180	71a	100a	73a	100a	53.67a	81.67a
25:75	2.34:16.88	5.33d	33.33e	7.33d	26.67d	5d	21.67d
"	4.69:33.75	30.33c	58.33d	21.33c	46.67c	13.67c	41.67c
"	9.38:67.5	48.67b	76.67c	38.67b	66.67b	33.67	61.67b
"	18.75:135	63.67a	91.67b	58a	85a	55a	80a
"	37.5:270	69.33a	98.33a	67.33a	93.33a	58.67a	86.67a
0:100	0:22.5	10.67c	41.67d	8.67d	38.33d	5.33e	35.d
"	0:45	43.67b	71.67c	38c	65c	28.67d	56.67c
"	0:90	58.67a	81.67ab	53.67b	78.33ab	46.67c	73.33ab
"	0:180	63.67a	86.67b	61.33ab	81.67b	56.33b	75b
"	0:360	69.67a	86.67a	6533a	83.33a	61.670a	80a

• ^agram active ingredient per hectare. ^bdays after application.

All three different mixture ratios of 75:25, 50:50, and 25:75 for paraquat:bentazon were better at pH 5 and 7 than at pH 9 on redroot pigweed. So that application of 101.25 to 202.5 g a.i.ha⁻¹ (56.25:45 and 112.5:90 g a.i.ha⁻¹ of paraquat:bentazon) resulted in 93.33 to 100% control, respectively, at a 75:25 mixture ratio at pH 5. At the same mixing ratio, control was lower at pH 7 (76.67–83.33%) and pH 9 (70–76.67%) (Table 5). At a 50:50 mixture ratio of paraquat:bentazon at pH 5, 7, and 9, applying 127.5 g a.i. ha⁻¹ (37.5:90 g a.i. ha⁻¹) caused 98.33, 95, and 75% control of redroot pigweed, respectively, while 100, 100, and 81.67% of redroot pigweed control were acquired using 255 g a.i. ha⁻¹ (75:180 g a.i. ha⁻¹) and of paraquat:bentazon at pH 5, 7, and 9, respectively, indicating that this mixing ratio was more effective than the others tested (Table 5). Although the 25:75 mixing ratio achieved a level of control over redroot pigweed, it was not as effective as the control obtained with the other ratios tested. Also, achieving significant control of redroot pigweed with the 25:75 ratio required higher doses of bentazon. According to the visual evaluation results, application of 153.72 g a.i.ha⁻¹ (18.75:135 g a.i.ha⁻¹) and 307.5 g a.i.ha⁻¹ (37.5:270 g a.i.ha⁻¹) of paraquat:bentazon at pH 5 and 7 resulted in 91.67–98.33 and 85–93.33% control of redroot pigweed. While the lower control of 80–86.67% redroot pigweed occurred by the same doses of application of paraquat:bentazon at pH 9 (Table 5).

4. Discussion

4.1 Individual analysis

The reduced effectiveness of paraquat applied to redroot pigweed after soil dust is caused by the herbicide strongly adsorbing to soil particles on the leaves, preventing its absorption, and thus reducing the amount available to the site of action. Paraquat and diquat, positively charged bipyridylium herbicides, are strongly attracted to and bound by negatively charged soil dust particles, a process that results in limited displacement through ion exchange. Previous research showed that soil dust particles settled on plants can decrease the activity of paraquat (Simarmata et al., 2017) and diquat (Rytwo, Tropp, 2001), which was consistent with our results. Aliverdi and Ahmadvand (2020) reported that paraquat activity under muddy rain containing 1, 2, 4, and 8 kg of soil per hectare reduced the biomass of winter wild oat by 1.2-, 1.6-, 2.5-, and 7-fold compared to the control. Babaei et al. (2023) found that the presence of soil dust significantly reduced the efficacy of bentazon in controlling common purslane, leading to an increase in the herbicide’s EDs parameters.

On the other hand, the pH of the spray solution is a critical water quality factor that can significantly affect herbicide performance. The effectiveness of weak acid herbicides is reduced when the water sprayer’s pH is not within the optimal range (pH < 7 or > 7). At pH levels above or below 7, chemical interactions decrease herbicide solubility, thus reducing their availability for plant uptake. Researchers have extensively investigated the mechanisms by which lowering spray solution pH enhances the efficacy of weak acid herbicides. Examples include studies on bentazon (Liu, 2004), glyphosate (Molin, Hirase, 2004), clopyralid and picloram (Palma et al., 2015), and 2,4-D (Devkota, Johnson, 2019). This improved efficacy is often attributed to “ion trapping,” where a decrease in pH increases the permeability of the herbicide through the plant cuticle. Specifically, weak acid herbicides exist as negatively charged, hydrophilic molecules under alkaline conditions. However, under acidic conditions, these molecules gain a proton (H⁺) and become uncharged and more hydrophobic (Liu, 2002; 2004). This effect is demonstrated by bentazon’s water solubility, which ranges from 490 mg L⁻¹ at pH 3 to 570 mg L⁻¹ at pH 7 (Food and Agriculture Organization of United Nations, 1999). By adding citric acid to lower the spray carrier pH, bentazon molecules become uncharged, facilitating their passage through the cuticle and cell membrane. Once inside the cell, where H⁺ ions are actively pumped out, the bentazon molecule loses a proton, reverting to its charged, hydrophilic form. This charged form is then “trapped” within the cell or phloem, leading to more efficient transport and herbicidal action (Dan et al., 2009). For example, Sterling and Lownds (1992) found that picloram absorption into broom snakeweed (Gutierrezia sarothrae (Pursh) Britt. & Rusby) peaked at pH 4 and decreased at higher pH levels. Liu (2004) demonstrated that bentazon acid uptake into bean (Vicia faba) leaves at pH 5, 7, and 9 was 53%, 30%, and 21%, respectively, when applied at a very low concentration (0.002%). The pH-dependent uptake of bentazon supports the weak acid hypothesis of herbicide absorption, which posits that absorption is greatest at pH values near the compound’s pKa. Bentazon uptake was highest at pH 5, close to its pKa of 3.30, and decreased as the solution pH became more alkaline (pH 7 or 9) (Table 4). These results suggest that bentazon primarily diffuses into the leaf and then across the membrane in its undissociated acid form. This is because a higher concentration of undissociated bentazon exists at acidic pH levels compared to neutral pH. Consistent with previous studies on other weak acids (Matocha et al., 2006), the cuticle is likely more permeable to undissociated bentazon than to its dissociated and polar form. Petroff (2000) suggests that acidic water conditions are most suitable for post-emergence application of weak acid herbicides, such as glyphosate or bentazon. This is because these herbicides are less likely to dissociate when the concentration of H+ ions is high. Conversely, alkaline conditions promote herbicide dissociation into simpler units, which are either poorly absorbed by weeds or completely ineffective. According to Taheri et al. (2024), reducing spray water pH from 8 to 5 improved sethoxydim’s effectiveness against wild barley (Hordeum spontaneum K. Koch), as evidenced by a significant decrease in the ED₅₀ parameter from 136.64 to 113.35 g a.i.ha⁻¹. Liu (2002) and Devkota and Johnson (2016) linked the enhanced efficacy observed at acidic spray water pH to the prevalence of undissociated herbicide molecules, allowing for easier diffusion across the leaf cuticle. Therefore, herbicide mixtures offer a new advanced strategy for sustainable redroot pigweed control in dusty soil conditions, mitigating herbicide resistance and reducing herbicide use and environmental impact.

4.2 Joint action analysis

Isobolographic analysis is a statistical technique for assessing the effect of mixing two chemicals. An isobologram itself represents a series of cross-sections taken from a dose-response surface, with each section corresponding to a different ratio of the two chemicals. The correlation between ED₅₀ values and isobolographic analysis in this experiment suggests that isobolographic analysis, when used in conjunction with dose-response studies, provides a more accurate assessment of herbicide tank-mix efficacy.

This paper’s results indicate that the isobolographic analyses of binary herbicide mixtures were consistent across the two different statistical software programs used (R versus Excel). Based on the relationship observed between ED₅₀ values and isobolographic analyses, the first and second experiments demonstrated that a tank mixture of paraquat and bentazon effectively alleviates the negative impact of soil dust accumulation on the weed. Therefore, optimizing the mixing ratio(s) of these two herbicides can effectively mitigate soil dust-related weed control issues. Generally, synergistic interactions between the herbicides under pH 5 and 7 indicated that they could be effectively combined; i.e., the mixed application of these two herbicides exhibited a greater effect than their individual applications. Mixing them also enables reduced herbicide applications, provides broader and complete control of diverse weed species simultaneously, slows the evolution of herbicide resistance, reduces operational costs, and facilitates improved farm management practices. The lambda values (3.34 compared to 2.45) suggest that the herbicide mixture’s efficacy against redroot pigweed was greater in the first experiment than in the second. Prostko (2018) and Eason et al. (2020) reported that combining a photosystem II inhibitor (bentazon) with a photosystem I electron diverter (paraquat) resulted in synergistic effects on morning glory (Ipomoea purpurea (L.) Roth.). They also reported that adding bentazon to the spray solution reduced foliar damage caused by paraquat in peanut fields (Eason et al., 2020; Daramola et al., 2024). Similarly, Daramola et al. (2024) demonstrated that a paraquat/bentazon tank mixture controlled a broader spectrum of broadleaf weeds, including bristly starbur (Acanthospermum hispidum DC.), purple nutsedge (Cyperus rotundus L.), common cocklebur (Xanthium strumarium L.), coffee senna (Senna occidentalis (L.) Link), prickly sida (Sida spinosa L.), and J. tamnifolia (L.) Griseb. This tank mixture also reduced peanut foliar injury compared to paraquat applied alone. Although these herbicides target related biochemical pathways, the interaction ranged from additive to synergistic depending on the herbicide dose. It seems that the addition of bentazon to paraquat spray entirely mitigates the negative effects of soil dust on the foliage of redroot pigweed. This might be attributed to the positively charged paraquat molecules in the spray mixture binding to negatively charged soil dust particles on the redroot pigweed’s aerial surfaces (Rytwo, Tropp, 2001). Once paraquat binds to dust particles, the bentazon molecules remain free, allowing them to penetrate the weed leaf cuticle and control redroot pigweed. Rytwo, Tropp (2001) demonstrated that diquat molecules rapidly bind to dust particles on weed shoots, with complete binding occurring in approximately 80 seconds after application. Therefore, in the paraquat/bentazon mixture, paraquat molecules are sacrificed due to binding to soil dust particles; bentazon molecules have a good opportunity to freely penetrate (without binding to dust particles) into the leaf cuticle, ultimately increasing its effectiveness. On the other hand, deviations from the optimal spray solution pH can compromise herbicide performance by diminishing active ingredient solubility or promoting its rapid degradation to inactive metabolites, thereby affecting absorption and translocation. Generally, plant tissues absorb weak-acid herbicides better at lower spray solution pH because more molecules exist in their uncharged (undissociated) form. Our findings from both experiments indicated that lower pH significantly alters the herbicide interaction, with the paraquat/bentazon mixture exhibiting strong synergy at pH 5 for suppressing redroot pigweed (Tables 4 and 5).

The third experiment revealed that the paraquat/bentazon mixture at pH 9 exhibited an additive to a slightly antagonistic effect on redroot pigweed control. On the other hand, mixing these two herbicides resulted in an effect similar to that of each herbicide applied individually. Although the mixture displayed a mild inhibitory effect (λ = 0.87), their similar mechanisms of action theoretically support their use in combination. Combining these two herbicides offers several advantages, including simultaneous control of multiple weed types, time savings, and reduced labor and equipment costs (e.g., fewer spray applications and reduced equipment movement). As noted earlier, bentazon, a weak acid herbicide, displays a solubility that is a function of both the spray solution pH and the pKa of the herbicide molecule in the solution. However, when solubility is not a limiting factor, increasing the pH above bentazon’s pKa (3.30) causes the weak-acid herbicide to become ionic. This ionic form experiences reduced penetration through the lipophilic cuticle, the negatively charged cell wall, and the membrane. This is partly due to increased adhesion of herbicide molecules to soil dust particles at higher pH (Liu, 2002; 2004). Consistent with Ruiz and Ortiz’s (2005) findings, glyphosate’s efficacy on broadleaf signalgrass (Brachiaria extensa L.) was diminished when applied with alkaline spray water pH compared to acidic pH. Dan et al. (2009) confirmed this trend, observing similar results for glyphosate’s effectiveness on palisade grass (Brachiaria brizantha L.). Alkaline spray water pH reduces herbicide efficacy, according to Dan et al. (2009) and Matocha et al. (2006), due to increased dissociation of herbicide molecules. In this ionized state (negative charge), the herbicide molecule readily forms a cation-herbicide complex, which, according to Dan et al. (2009), reduces plant accumulation and overall effectiveness. Therefore, the reduced efficacy of bentazon at alkaline spray solution pH can be attributed to ionization of the weak-acid herbicide. This ionization, combined with cation complex formation and the presence of dust particles in the spray solution, hinders bentazon’s ability to effectively control redroot pigweed. Nalewaja et al. (1994) found in Wisconsin that sethoxydim’s effectiveness was unaffected in acidic spray solutions (pH 3.5) but decreased at alkaline pH values (above 7).

5. Conclusion

According to our results, soil dust negatively impacted the performance of both paraquat and bentazon when applied individually. Utilizing herbicide mixtures based on the ADM is considered an effective and efficient approach to optimizing herbicide application, reducing consumption, and fostering sustainable agricultural practices. The overall outcome of isobolographic analysis for binary mixtures was consistent between the two statistical software programs, R and Excel. Under soil dust conditions, the paraquat + bentazon mixture displayed variable efficacy against redroot pigweed depending on the pH level. The interaction was synergistic at pH 5 and 7, but as the pH rose to 9, the interaction shifted from additive to a weakly antagonistic effect. Across all three paraquat + bentazon mixtures tested under dust conditions and at pH levels of 5, 7, and 9, increasing the paraquat:bentazon ratio led to a steeper dose-response curve, indicating a faster and stronger effect on redroot pigweed. Therefore, higher bentazon ratios are recommended in these mixtures. Given the similar mechanisms and modes of action of paraquat and bentazon, their combination demonstrated synergistic activity at acidic to neutral pH (5 and 7). This synergy enables lower herbicide application rates, leading to both economic advantages and reduced environmental impact. Due to redroot pigweed’s widespread presence in diverse agricultural and non-agricultural settings, and the requirement for effective spraying even with dust, the paraquat + bentazon combination provides a valuable broad-spectrum solution for controlling a range of weeds, including this troublesome species. Applying these herbicide mixtures, even in dusty environments, is recommended as a strategy to decrease overall herbicide consumption, minimize the number of spraying events, and ultimately reduce the potential for herbicide resistance to emerge.

Acknowledgements

This work was performed in partial fulfillment of requirements for the MSc. degree (Project No. 71119552) in Weed Science at Lorestan University that was financed by a grant from the College of Agriculture.

• Funding
  No applicable.

Data availability

Data will be made available on request.

References

• Andersen RN. Germination and establishment of weeds for experimental purposes: a Weed Science Society of America handbook. Urbana: Weed Science Society of America; 1968.
• Aliverdi A, Ahmadvand G. Tank-mix adjuvants to reduce the adverse effect of muddy rain on the activity of paraquat against winter wild oat. Crop Prot. 2020;128:105013. Available from: https://doi.org/10.1016/j.cropro.2019.105013
  » https://doi.org/10.1016/j.cropro.2019.105013
• Babaei R, Chitband AA, Aliverdi A. [The evaluation effect of dust on the pure and mixed efficiency of bentazone, imazethapyr and pyridate with PCGate on control of common purslane (Portulaca oleracea L.)]. Appl Res Field Crops. 2023;36(2):103-29. Persian. Available from: https://doi.org/10.22092/aj.2024.364810.1669
  » https://doi.org/10.22092/aj.2024.364810.1669
• Chitband AA, Noghondar MN, Sarabi V. Yield of sweet corn varieties and response to sulfonylurea and mix herbicides. Adv Weed Sci. 2021;39:1-11. Available from: https://doi.org/10.51694/AdvWeedSci/2021;39:00018
  » https://doi.org/10.51694/AdvWeedSci/2021;39:00018
• Chitband AA, Rastgoo M, Asadi GA. Joint action of malathion with herbicides against Alhagi pseudalhagi (Bieb.) Desv. Adv Weed Sci. 2024;42:1-11. Available from: https://doi.org/10.51694/AdvWeedS-ci/2024;42:00021
  » https://doi.org/10.51694/AdvWeedS-ci/2024;42:00021
• Chitband AA, Sarabi V, Aliverdi A. Joint action of common herbicides on the control of Alhagi pseudalhagi (Bieb.) Desv.: A comparison study. Weed Res. 2025;65(1). Available from: https://doi.org/10.1111/wre.12679
  » https://doi.org/10.1111/wre.12679
• Dan HA, Dan LG, Barroso AL, Souza CH. Effect of pH of spraying at drying of the Braquiaria brizanta with glyphosate herbicide. Global Sci Technol. 2009;2:1-6.
• Daramola OS, Iboyi JE, MacDonald GE, Kanissery RG, Tillman BL, Singh H, Devkota P. A systematic review of chemical weed management in peanut (Arachis hypogaea) in the United States: challenges and opportunities. Weed Sci. 2024;72(1):5-29. Available from: https://doi.org/10.1017/wsc.2023.71
  » https://doi.org/10.1017/wsc.2023.71
• Devkota P, Johnson WG. Glufosinate efficacy as influenced by carrier water pH, hardness, foliar fertilizer, and ammonium sulfate. Weed Technol. 2016;30(4):848-59. Available from: https://doi.org/10.1614/WT-D-16-00053.1
  » https://doi.org/10.1614/WT-D-16-00053.1
• Devkota P, Johnson WG. Influence of carrier water pH, foliar fertilizer, and ammonium sulfate on 2,4-D and 2,4-D plus glyphosate efficacy. Weed Technol. 2019;33(4):562-8. Available from: https://doi.org/10.1017/wet.2019.31
  » https://doi.org/10.1017/wet.2019.31
• Eason KM, Grey TL, Tubbs RS, Prostko EP, Li X. Peanut and weed response to postemergence herbicide tank-mixtures including paraquat and inorganic liquid nutrients. Peanut Sci. 2020;47:94-102. Available from: https://doi.org/10.3146/PS20-12.1
  » https://doi.org/10.3146/PS20-12.1
• Food and Agriculture Organization of United Nations - FAO. FAO specifications and evaluations for plant protection products: Bentazone. Rome: Food and Agriculture Organization of United Nations; 1999. Available from: http://www.fao.org/waicent/faoinfo/agricult/agp/pdf
  » http://www.fao.org/waicent/faoinfo/agricult/agp/pdf
• Liu ZQ. Bentazone uptake into plant foliage as influenced by surfactants and carrier pH. Aust J Agric Res. 2004;55(9):967-71. Available from: https://doi.org/10.1071/AR04046
  » https://doi.org/10.1071/AR04046
• Liu ZQ. Lower formulation pH does not enhance bentazone up-take into plant foliage. N Z Plant Prot. 2002;55:163-7. Available from: https://doi.org/10.30843/nzpp.2002.55.3887
  » https://doi.org/10.30843/nzpp.2002.55.3887
• Matocha MA, Krutz LJ, Senseman SA, Koger CH, Reddy KN, Palmer EW. Spray carrier pH effect on absorption and translocation of triflox-ysulfuron in Palmer amaranth (Amaranthus palmeri) and Texasweed (Caperonia palustris). Weed Sci. 2006;54(6):969-73. Available from: https://doi.org/10.1614/WS-06-029.1
  » https://doi.org/10.1614/WS-06-029.1
• Molin WT, Hirase K. Comparison of commercial glyphosate formulations for control of prickly side, purple nutsedge, and sicklepod. Weed Biol Manag. 2004;4(3):136-41. Available from: https://doi.org/10.1111/j.1445-6664.2004.00130.x
  » https://doi.org/10.1111/j.1445-6664.2004.00130.x
• Nalewaja JD, Matysiak R, Szelezniak E. Sethoxydim response to spray carrier chemical properties and environment. Weed Technol. 1994;8(3):591-7. Available from: https://doi.org/10.1017/S0890037X00039749
  » https://doi.org/10.1017/S0890037X00039749
• Nosratti I, Saeidi M, Barbastegan H, Jalali-Honarmand S, Ghobadi M. [Effect of airborne particles on herbicides efficiency for control of corn (Zea mays) weeds in Kermanshah region]. Res Crop Ecosyst. 2016;3(1-2):55-66. Persian.
• Palma G, Demanet R, Jorquera M, Mora ML, Briceπo G, Violante A. Effect of pH on sorption kinetic process of acidic herbicides in a volcanic soil. J Soil Sci Plant Nutr. 2015;15(3):549-60. Available from: https://doi.org/10.4067/S0718-95162015005000023
  » https://doi.org/10.4067/S0718-95162015005000023
• Petroff R. Water quality and pesticide performance. Bozeman: Montana State University Extension Service; 2000[access Aug 11, 2010]. Available from: https://extension.msu.montana.edu/
  » https://extension.msu.montana.edu/
• Poovey AG, Skogerboe JG. Using diquat in combination with endothall under turbid water conditions to control Hydrilla. Vicksburg: US Army Engineer Research and Development Center; 2004. Report No.: ERDC/TN APCRP-CC-02.
• Prostko EP. Peanut weed control. In: University of Georgia Cooperative Extension, editor. Georgia pest management handbook. Athens: University of Georgia Cooperative Extension; 2018. p. 215-34.
• R Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2021[access May 18, 2021]. Available from: https://www.R-project.org/
  » https://www.R-project.org/
• Ritz C, Streibig JC, Kniss A. How to use statistics to claim antagonism and synergism from binary mixture experiments. Pest Manag Sci. 2021;77(9):3890-9. Available from: https://doi.org/10.1002/ps.6348
  » https://doi.org/10.1002/ps.6348
• Ruiz IG, Ortiz MS. Influence of water pH on the effectiveness of various herbicides used in sugarcane. Fitosanidad. 2005;9:37-40.
• Rytwo G, Tropp D. Improved efficiency of a divalent herbicide in the presence of clay by addition of monovalent organocations. Appl Clay Sci. 2001;18(5-6):327-33. Available from: https://doi.org/10.1016/S0169-1317(01)00035-7
  » https://doi.org/10.1016/S0169-1317(01)00035-7
• Shahbazi T, Saeedi M, Nosrati I, Jalai-Honarmand S. [Evaluation of the effect of airborne particles on herbicides efficiency on weed control in wheat (Tritucum aestivum)]. Res Crop Ecosyst. 2015;2:63-72. Persian.
• Simarmata M, Taufik M, Peranginangin ZZA. Efficacy of paraquat and glyphosate applied in water solvents from different sources to control weeds in oil palm plantation. ARPN J Agric Biol Sci. 2017;12:58-64.
• Sterling TM, Lownds NK. 1992. Picloram absorption by broom snake-weed (Gutierrezia sarothrae) leaf tissue. Weed Sci. 1992;40(3):390-394. Available from: https://doi.org/10.1017/S0043174500051791
  » https://doi.org/10.1017/S0043174500051791
• Streibig JC, Jensen JE. 2000. Actions of herbicides in mixtures. In: Cobb AH, Kirkwood RC, editors. Herbicides and their mechanisms of action. Boca Raton: CRC Press; 2000. p. 153-80
• Taheri Sh, Aliverdi A, Ahmadvand G. The effect of pH and light on the efficacy of spray solution stored of haloxyfop-r-methyl, fluazifop-p-butyl, and sethoxydim against wild barley (Hordeum spontaneum K. Koch). J Iranian Plant Prot Res. 2024;37(4):425-39. Available from: https://doi.org/10.22067/jpp.2023.80667.1128
  » https://doi.org/10.22067/jpp.2023.80667.1128
• Teymourinia M, Chitband AA, Rezaee Gh, Khayrandish S. The joint action of glyphosate, clethodim, and imazethapyr to control cogongrass (Imperata cylinderica L. Beauv) in the margin of the irrigation canals (a case study using two different approaches). Crop Prot. 2023;174. Available from: https://doi.org/10.1016/j.cropro.2023.106413
  » https://doi.org/10.1016/j.cropro.2023.106413
• Wehtje G, Wilcut JW, McGuire JA. Influence of bentazon on the phytotoxicity of paraquat to peanuts (Arachis hypogaea) and associated weeds. Weed Sci. 1992;40(1):90-5. Available from: https://doi.org/10.1017/S0043174500057015
  » https://doi.org/10.1017/S0043174500057015
• Zhou J, Tao B, Messersmith CG. Soil dust reduces glyphosate efficacy. Weed Sci. 2006;54(6):1132-6. Available from: https://doi.org/10.1614/WS-06-107R.1
  » https://doi.org/10.1614/WS-06-107R.1

Edited by

• Editor in Chief:
  Carol Ann Mallory-Smith

• Associate Editor:
  Victor Ribeiro

Publication Dates

• Publication in this collection
  24 Aug 2026
• Date of issue
  2026

History

• Received
  27 July 2025
• Accepted
  29 May 2026