1. Introduction

Rice is the staple food of more than three billion people and is essential for the subsistence of communities around the world. It serves low-income families, who depend on it as a primary source of energy and nutrients, to those with greater economic power, where it appears as an ingredient in elaborate and traditional dishes. Its relevance goes beyond nutrition, playing a fundamental role in the culture, economy and food security of countless nations. It is the second most consumed cereal in the world, behind only corn, and stands out for its adaptability to different climates and cultivation systems. In Brazil, rice has a special place in the population's daily diet, with a significant total production of 9.2 million tons in the 2022/23 session (Companhia Nacional de Abastecimento, 2023). This performance not only reinforces the importance of rice for the country's agricultural economy, but also for national food security, ensuring the food supply for millions of homes.

Among the factors that interfere with crop yield, competition with weeds can cause significant yield losses, while weeds also serve as hosts for pests and diseases (Agostinetto et al., 2022). The main weeds present in flooded rice in Rio Grande do Sul (RS) are weedy rice (Oryza sativa), barnyardgrass (Echinochioa spp.) and sedges (Cyperus spp.) (Fruet et al., 2020; Silva et al., 2020). Yield reductions in flooded rice can reach 600 kg ha^-1 due to competition with barnyardgrass (Richter et al., 2019). Furthermore, the competition for nutrients between barnyardgrass and flooded rice in the vegetative phase is greater for nitrogen (N), and in the reproductive phase for potassium (K), calcium (Ca) and magnesium (Mg), with each barnyardgrass plant m^-2 capable of reducing rice yield by 1.13% (Ulguim et al., 2020).

Weedy rice stands out as a problematic weed species in flooded rice due to its taxonomic similarity to rice, making it difficult to control with selective herbicides. It constitutes a serious problem, with only a few chemical control alternatives before and after crop emergence, which can reduce grain yield and quality. In more drastic situations, fields are abandoned due to high infestations of the weed (Silva et al., 2021). To promote the selective control of weedy rice, genotypes have been commercialized with tolerance to herbicides that are naturally non-selective to the Oryza genus such as imidazolinones (inhibition of acetolactate synthase - ALS) and aryloxphenoxy-propionates (inhibition of acetyl coenzyme-A carboxylase - ACCase).

The Clearfield (CL) technology was launched in Brazil in 2003, through rice cultivars with a gene for tolerance to herbicides from the imidazolinone (IMI) chemical family of Group 2 herbicides (Avila et al., 2021). Due to inadequate use and minimal implementation of stewardship recommendations, the evolution of herbicide-resistant biotypes of weedy rice was inevitable. The monoculture of the CL system for more than six consecutive years, the use of herbicides with the same mechanism of action and the use of rates greater than 1.5x above the label rate, increased the selection pressure for weed resistance, reducing the efficacy of the technology in rice production areas (Ulguim et al., 2019). Five IMI-resistant weed species are currently documented in flooded rice crops in southern Brazil, including weedy rice and barnyardgrass (Heap, 2024).

An alternative for controlling IMI-resistant weedy rice and barnyardgrass is the technology of rice genotypes with tolerance to herbicides of ACCase inhibitors. In Brazil, this technology was commercially launched in the 2022/2023 growing season, with tolerance to propaquizafop and quizalofop herbicides in the Max-Ace® and Provisia®, respectively (RiceTec, 2024; BASF Corporation, 2024). Both were obtained through an induced mutation, in which Max-Ace® presents a change from a serine to a glycine at position 2096 (mutation G2096S) in ACCase makes the enzyme tolerant/resistant to some herbicides, including propaquizafop (Hinga et al., 2016). However, these herbicides have a spectrum limited to plants of the Poaceae family only (Hinga et al., 2016).

There is still a need for chemical control alternatives for eudicotyledonous and sedge weed species in rice production. The tank mixing of herbicides is a common practice, used by 97% of producers in Brazil (Gazziero, 2015), and practiced by 65% of farmers in rice fields (Fruet et al., 2020). This practice, in addition to increasing the control spectrum, also reduces application costs and optimizes land management (Bianchi et al., 2020). Therefore, understanding the effect of ACCase inhibitor herbicides in tank mix with other herbicides is necessary to optimize weed management in flooded rice tolerant to ACCase inhibitor herbicides.

The interaction of quizalofop when mixed with florpyrauxifen-benzyl is additive for the control of barnyardgrass, being a tank mix option to broaden the spectrum of weed control (Sanders et al., 2021). However, there is little information on the tank mix of herbicides with propaquizafop, used in Max-Ace® technology in Brazil. Therefore, the objective of this study was to evaluate the interaction of propaquizafop in tank mix with herbicides used in rice tolerant to ACCase inhibitors to control weedy rice and barnyardgrass.

2. Material and Methods

Two experiments were conducted, in the 2022/2023 season, in two sites: Dom Pedrito (30°58'54"S 54°40'39"W) and Formigueiro (30°0'32"S 53°29'54"W), located in Rio Grande do Sul, Brazil. The design used was randomized blocks, with four replications, in 3x5m plots. The soil in both sites is classified as Alfisols (United States Department of Agriculture, 2024). The state's climate, according to the Köppen classification, is characterized as subtropical (Cfa), with hot summers and no defined dry season (Alvares et al., 2013).

Rice was planted on October 5, 2022 in Dom Pedrito and on November 10, 2022 in Formigueiro, using the early maturity hybrid XP739MA (Max-Ace® technology), at 45 kg ha^-1. Rice was fertilized based on research recommendations. In Dom Pedrito 20 kg ha^-1 of N, 104 kg ha^-1 of P₂O₅ were applied at sowing. Topdressing 120 kg ha^-1 of K₂O and 135 kg ha^-1 of N, divided into two applications, the first in the phenological stages in V₃/V₄ and the second in R₀ (Counce et al., 2000). In Formigueiro 18 kg ha^-1 of N, 70 kg ha^-1 of P₂O₅ and 70 kg ha^-1 of K₂O was applied at sowing. Topdressing 108 kg ha^-1 of N, divided into two applications, in V₃/V₄ e R₀.

In both locations, 25 treatments were tested (Table 1). The herbicides were applied before flood irrigation, when the rice was at the V₃/V₄ stage. The treatments were arranged in a 3x8+1 factorial design, with factor A being the rates of propaquizafop (0, 125 and 180 g ha^-1) and factor B was herbicides in tank mixes. The herbicides added to propaquizafop were: carfentrazone, propanil, quinclorac, penoxsulam, florpyrauxifen-benzyl, bentazon, saflufenacil, and no additional herbicide (Table 1). The nontreated check had no propaquizafop and no other herbicide. The additional treatment consisted of two applications of propaquizafop alone at 125 g ha^-1, as specified on the herbicide label (Adama Brasil S/A, 2024). The first propaquizafop application was at the V₃/V₄ stage of rice before flooding, and the second was 15 days after the first.

The herbicides were applied using a backpack sprayer, pressurized with CO₂ and a handheld spray boom with four 110.015 flat-fan nozzles spaced 0.5 m apart. The equipment was calibrated to deliver 150 L ha^-1. Flood irrigation was initiated two days after herbicide, and the water level remained stable after eight days. At the time of application, Dom Pedrito had weedy rice and barnyardgrass infestations of 11 and 8 plants m^-2, respectively, at the 5-8 leaf stage. In Formigueiro, barnyardgrass infestation was at 224 plants m^-2 (6-8 leaves), while weedy rice was at 56 plants m^-2 (4-6 leaves).

Rice injury and weed control was evaluated visually by two personnel together, based on a rating scale of 0 to 100 where 0 = absence of symptoms and 100 = death. Evaluations were done at 10, 20 and 30 days after application (DAA), for both locations. Rice was harvested on February 28 and March 14, 2023 in Dom Pedrito and Formigueiro, respectively.

Rice was harvested from 6 m² area. Samples were threshed, their weight corrected to 13% moisture content, and yields converted to kg ha^-1. In Dom Pedrito, the yield for the untreated check and the zero rate of propaquizafop was estimated visually and based on treatments with single herbicides (without propaquizafop), as these plots could not be harvested due to severe weed infestation. Additionally, the number of panicles m^-2, grains per panicle, and 1,000-grain weight were evaluated. Panicles were counted directly, grains per panicle were averaged from 10 randomly collected panicles per plot, and 1,000-grain weight was determined using five replicates of 100 grains, corrected to 13% moisture content.

The variables were analyzed with R software (R Core Team, 2019). Using the ExpDes.pt package (Healy, 1956), subjected to analysis of variance, homogeneity of residual variances (Barlett) e normality of errors (Shapiro-Wilk). When the factor effect was significant, means were compared using the Scott-Knott test (p<0.05). For the additional treatment, a Dunnett test was performed, comparing the tank mix with it. Weed control data were used to test for herbicide interactions following the Colby (1967) methodology, which consists of evaluating the effect of herbicides alone and in tank mix. Equation 1 gives the expected control value of the tank mix, obtained from the control of each herbicide alone, and compared with the observed value of the tank mix.

where: E = expected value of the herbicide tank mix in each combination; X and Y = control. If the observed control is greater than expected, the interaction is synergistic, if the observed control is equal to the sum of the two alone herbicides, the interaction is additive, and if the observed control is less than expected, the interaction is antagonistic (Colby, 1967). The interactions were compared by Student's t-test (p≤0.05).

3. Results and discussion

The data were adjusted to the requirements of homogeneity (Barlett) and normality (Shapiro-Willk) tests. The interaction effect between propaquizafop rate and tank mix partner, was significant at Formigueiro and Dom Pedrito for the number of panicles m^-2, weight of 1,000-grain and yield, and number of grains panicle^-1 in Formigueiro but not in Dom Pedrito. Rice injury was similar across herbicide treatments, not exceeding 15% (data not shown).

For the weedy rice control in Dom Pedrito, the activity of herbicides in mixtures were mainly additive (Table 2). Overall and across the three evaluation periods, 64% of the comparisons between the observed and estimated control using the Colby test showed additive interactions, while 36% were antagonistic, and none were synergistic. At 30 DAA there was antagonism of the mixtures of propaquizafop with propanil and carfentrazone at a rate of 125 g ha^-1, and with bentazon and propanil at a rate of 180 g ha^-1 (Table 2). In Formigueiro, in the weedy rice control, there was antagonism only with the tank mix of propaquizafop with propanil in all evaluation periods for both rates (Table 3).

Similar results were found in other studies published in the literature, where tank mix of herbicides such as propanil, penoxsulam and quinclorac with quizalofop resulted in antagonism (Lancaster et al., 2019). The tank mix of quizalofop with propanil and quinclorac was antagonistic of barnyardgrass and weedy rice control, resulting in control 7 to 20% lower than the expected values calculated from the results of the herbicides applied alone (Lancaster et al., 2019).

For barnyardgrass control in Dom Pedrito and Formigueiro 26% of the interactions were considered antagonistic (Tables 4 and 5). Thus, around 16% and 14% antagonistic interactions for Dom Pedrito and Formigueiro, respectively, resulting from the tank mix at the highest rate of propaquizafop (180 g ha^-1). The result is justified by propaquizafop alone had high control and expected value for Colby's test is higher than that of the 125 g ha^-1 rate (Tables 4 and 5). Furthermore, for barnyardgrass control at 30 DAA there was antagonism only of propaquizafop mixed with propanil, in the two rates of the first herbicide (Tables 4 and 5), as was observed for weedy rice (Tables 2 and 3).

The tank mix of propaquizafop and propanil should not be recommended for controlling weedy rice, possibly because they have different modes of action on the plant. The antagonism can be explained due to systemic translocation for propaquizafop and effect of contact for propanil. Tank mix of cyhalofop-butyl and propanil for control of barnyardgrass were antagonistic, because propanil reduced the rate of metabolism of cyhalofop-butyl by decreasing the apoplastic esterase enzyme that converts cyhalofop-butyl to cyhalofop-acid (Ottis et al., 2005).

For both sites, the tank mix of propaquizafop with penoxsulam was antagonistic for barnyardgrass control only in the evaluation at 20 DAA (Tables 4 and 5). For the barnyardgrass control in both sites, there was antagonism of the tank mix of propaquizafop with florpyrauxifen and quinclorac in general until the evaluation at 20 DAA (Tables 4 and 5). As observed for weedy rice, this result may be associated with the mode of action of the quinclorac and florpyrauxifen herbicides.

The antagonistic interaction of propaquizafop with florpyrauxifen is expected because this herbicide belongs to the mode of action of auxin mimics, with reports in the literature of antagonism of these tank mix (Barbieri et al., 2022). An example is tank mix of clethodim and 2,4-D that resulted in an antagonistic effect for ryegrass control, but an increase in the rate of clethodim compensated for the antagonism with 2,4-D (Polito et al., 2021). However, as seen previously, the same was not observed for weedy rice control for tank mix of propaquizafop with florpyrauxifen, since even increasing the rate of propaquizafop the antagonism was not overcome.

The number of panicles m^-2 in Dom Pedrito, the highest number was obtained with the tank mix of propaquizafop at a rate of 125 g ha^-1 with carfentrazone (Figure 1). At the highest rate (180 g ha^-1), the tank mix with carfentrazone, bentazon and quinclorac were superior to the other treatments (Figure 1). These mixtures had control above 96% at 30 DAA for weedy rice and barnyardgrass (Tables 2 and 4). For the variable 1000-grain weight the tank mix with propanil at a rate of 125 g ha^-1 resulted in the lowest values (Figure 2). For the highest rate of propaquizafop, no difference was observed between treatments. These results corroborate the control data, where the observed antagonism of the tank mix of propaquizafop with quinclorac and florpyrauxifen for barnyardgrass (Tables 4 and 5) and with propanil for weedy rice (Tables 2 and 3) reduced the yield components (Figures 1 and 2).

Grain yield in Dom Pedrito apresented in a significant interaction between the tank mix and the rates of propaquizafop (Figure 3). Propaquizafop at 125 g ha^-1, the tank mix with bentazon, quinclorac and propanil resulted in lower yield than the other treatments. At the highest rate of propaquizafop, there was no statistical difference between the treatments (Figure 3). Herbicide treatments without a tank mix partner to control barnyardgrass and weedy rice resulted in low yields similar to, or lesser than, those of antagonistic treatments, since these herbicides have a control spectrum mainly in eudicotyledonous plants.

It is important to highlight that knowledge of antagonistic tank mix for the of weedy rice control is extremely important in the context of flooded rice production (Tables 2 and 3). The herbicide propaquizafop provides effective of weedy rice control, making it an important tool for managing this species, given the challenges of selective control (Avila et al., 2021). This difficulty lies in the fact that the weed and crop belong to the same species, whose crop tolerance to propaquizafop occurs through induced mutation events (Hinga et al., 2016). Therefore, failures in weedy rice control in flooded rice may favor the occurrence of gene flow from cross-pollination of tolerant cultivars and the weed (Menezes et al., 2009). This fact can result in rapid evolution of herbicide resistance in commercial grain production areas, as observed for CL rice (Goulart et al., 2014). Therefore, such antagonism results may imply loss of crop yield, with an impact on rice yield components.

The tank mix of propaquizafop (125 g ha^-1) with bentazon produced the highest number of panicles m^-2 in Formigueiro (Figure 4). At higher propaquizafop rates, only the mix with propanil led to fewer panicles m^-2. The number of grains panicle^-1 was lower than the untreated check when propaquizafop (125 g ha^-1) was combined with quinclorac, florpyrauxifen, propanil and penoxsulam (Figure 5). For propaquizafop at 180 g ha^-1, only the tank mix with propanil was lower than the untreated check. The 1000-grain weight in Formigueiro did not differ between the treatments with of propaquizafop at 125 g ha^-1 (Figure 6).

The tank mix of quizalofop with penoxsulam resulted in antagonism for barnyardgrass control, showing an 8% reduction in control when compared to quizalofop alone (Lancaster et al., 2019). Then, the sequential application of propaquizafop as labeled recommended is important to control possible failures and new weed plant emergence flows, preserving the technology. In the present study, sequential application was only used in the additional treatment, in order to evaluate the differences in the other treatments by mixing the herbicides only, since sequential application would result in the control of all plots.

The tank mix of propanil and propaquizafop in Formigueiro resulted in lower grain yield than other treatments including untreated check (Figure 7). For this tank mix at the highest rate of propaquizafop the yield reduces 14.6% more than 125 g ha^-1 rate (Figure 7). It is worth noting that yield values close to zero when applying the alone herbicides are due to the fact that competition with barnyardgrass and weedy rice.

The Dunnet test was used to compare the treatments with the additional treatment for the yield variable (Figure 8 and 9). The additional treatment consists of the sequential application of propaquizafop, following the recommendations for the use of Max-Ace® technology (Adama Brasil, 2024).

Means followed by the same letter do not differ from each other according to the Scott-Knott test (p≤0.05). Uppercase letters compare herbicides within the same dose of propaquizafop; lowercase letters compare propaquizafop doses for a given herbicide

In Dom Pedrito, the treatments with bentazon, saflufenacil, propanil, and carfentrazone alone had lower yields compared to the additional treatment, which obtained 3,233 kg ha^-1 of yield (Figure 8). Additionally, the tank mix with bentazon, quinclorac, and propanil with the lower rate of propaquizafop resulted in lower yields than the additional treatment, although there was no statistical difference due to the overlap of confidence intervals (Figure 8). The results suggest that antagonism promoted by some tank mixes can be minimized by alone application sequential of propaquizafop. The tank mixes with florpyrauxifen at the two rates of propaquizafop had higher yield than additional factor, mainly due to broad spectrum of control and additive interaction (Figure 8).

An additive response was observed between the tank mix of quizalofop and quinclorac, carfentrazone or thiobencarb, with yields did not differ from the treatment with quizalofop alone (Rustom et al., 2019), as observed in the present study. Furthermore, the application of quizalofop with propanil or 3 days after application of ACCase may result in reductions in rice yield (Rustom et al., 2020), as collected for both sites (Table 3 and 4).

It should be noted that the average yield of the treatments that stood out most in Dom Pedrito was around 6000 kg ha^-1, below the average for Rio Grande do Sul of 8790 kg ha^-1 (Companhia Nacional de Abastecimento, 2023). This may be a result, among other factors, of the high temperatures faced by the crop at the time of flowering, which cause sterility of the spikelets (Yoshida, 1981).

In Formigueiro, the Dunnet test showed that the herbicides applied alone let to lower yields than the additional treatment, which yield 7,878 kg ha^-1, except for quinclorac and propaquizafop (Figure 9). Tank mix had yields similar or higher than the additional treatment, except for propaquizafop plus propanil (Figure 9).

The application of propaquizafop with broad-leaf spectrum herbicides proved to be a viable option for producers, expanding the control spectrum and allowing only one application, thus reducing costs. However, care should be taken as to which herbicides to use, since propanil is an antagonistic mixture and should be avoided. Furthermore, further studies could be conducted to evaluate other herbicides, especially auxin mimics.

4. Conclusions

Tank mix of propaquizafop with other herbicides used in flooded rice for broadleaf weed control can be used to increase the spectrum of action of the herbicides. Mixtures with carfentrazone, bentazon, penoxsulam, florpyrauxifen, quinclorac, and saflufenacil result in additive interaction and are an alternative for use in rice with propaquizafop tolerance. Tank mix of propaquizafop with propanil are antagonistic, reducing control of barnyardgrass and weedy rice and grain yield of flooded rice, and should not be used.

Acknowledgements

The authors would like to thank UruAgro Assessoria Agropecuária for its assistance during the conduct of the experiments, and the producers Thiago Bolson and Celso

Ricardo Mario for providing the area for conducting the experiments.

References

Edited by

Publication Dates

History