Introduction
The evolution of herbicide-resistant weeds has been highlighted as a major challenge for agriculture (Baucom Reference Baucom2019). Since the first case of herbicide-resistant rigid ryegrass (Lolium rigidum Gaudin) reported in 1982 (Heap and Knight Reference Heap and Knight1982), Lolium species have evolved resistance to active compounds with more than 10 sites of action (Heap Reference Heap2019).
Lolium spp. are a major weed problem in winter cereal crops from five continents, and they leads to wheat (Triticum aestivum L.) yield losses of up to 55% (Bararpour et al. Reference Bararpour, Korres, Burgos, Hale and Tseng2018; De Prado et al. Reference De Prado, Osuna, Heredia and De Prado2005; Lemerle et al. Reference Lemerle, Verbeek and Combes1995; Mahmood et al. Reference Mahmood, Mathiassen, Kristensen and Kudsk2016; Sabet Zangeneh et al. Reference Sabet Zangeneh, Mohammaddust Chamanabad, Zand, Alcántara-de la Cruz, Travlos, De Prado and Alebrahim2018; Scursoni et al. Reference Scursoni, Palmano, De Notta and Delfino2012; Smit et al. Reference Smit, Smit and de Villiers1999; Zhang et al. Reference Zhang, Wu, Xu, Gao and Zhang2017). In Argentina, L. perenne and Italian ryegrass [Lolium perenne (L.) ssp. multiflorum (Lam.) Husnot] occur in 40% of wheat and barley (Hordeum vulgare L.) crops from the Pampas (Istilart and Yanniccari Reference Istilart and Yanniccari2012; Scursoni et al. Reference Scursoni, Gigón, Martin, Vigna, Leguizamón, Istilart and López2014). Chemical weed management has been based on glyphosate treatments during fallow periods before winter crop planting and acetyl-CoA carboxylase (ACCase)- or acetohydroxyacid synthase–inhibiting herbicides applied POST in wheat or barley crops.
In response to that management regime, cases of glyphosate-resistant L. perenne have been detected over large areas where glyphosate exerts a strong selection pressure on weeds (Yanniccari et al. Reference Yanniccari, Istilart, Giménez and Castro2012). ACCase-inhibiting herbicides have been used to control glyphosate-resistant plants; specifically, clethodim or haloxifop have been applied early preplant and pinoxaden or clodinafop later on wheat crops (Yanniccari et al. Reference Yanniccari, Istilart, Giménez and Castro2012). However, the evolution of ACCase herbicide–resistant Lolium spp. has been demonstrated in several countries (De Prado et al. Reference De Prado, Osuna, Heredia and De Prado2005; Kaundun Reference Kaundun, Bailly, Dale, Hutchings and McIndoe2013a; Yu et al. Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007), and putative ACCase-resistant populations of Lolium spp. have been detected in Argentina (Gigón and Yanniccari Reference Gigón and Yanniccari2018).
The biochemical basis of ACCase-herbicide resistance could mainly be either through enhanced rates of herbicide metabolism or molecular mutations in the ACCase gene, endowing target site–based herbicide resistance (Powles and Yu Reference Powles and Yu2010; Yu et al. Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007). In the first case, herbicides are metabolized to products with reduced phytotoxicity by cytochrome P450 monooxygenases (Yu and Powles Reference Yu and Powles2014). Since the first case of resistance to diclofop-methyl in L. rigidum, endowed by enhanced detoxification, several Lolium spp. populations have shown resistance to a large number of herbicides with several different modes of action (Preston Reference Preston2004; Preston et al. Reference Preston, Tardif, Christopher and Powles1996; Vila-Aiub et al. Reference Vila-Aiub, Neve and Powles2005).
Three classes of commercially available herbicides, aryloxyphenoxy propionates (FOPs), cyclohexanediones (DIMs), and phenylpyrazolines (DEN), target the carboxyltransferase (CT) activity of ACCase; however, 14 spontaneous mutations in the CT domain of this enzyme have been shown to confer herbicide resistance: Ile-1781-Leu, Ile-1781-Val, Ile-1781-Arg, Ile-1781-Thr, Trp-1999-Cys, Trp-1999-Leu, Trp-1999-Ser, Trp-2027-Cys, Ile-2041-Asn, Ile-2041-Val, Asp-2078-Gly, Cys-2088-Arg, Gly-2096-Ala, and Gly-2096-Ser (amino acid residues correspond to the native blackgrass [Alopecurus myosuroides Huds.] ACCase sequence) (Jang et al. Reference Jang, Marjanovic and Gornicki2013; Kaundun et al. Reference Kaundun, Bailly, Dale, Hutchings and McIndoe2013a, Reference Kaundun, Hutchings, Dale and McIndoe2013b; Yu et al. Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007, Reference Yu, Kim and Tong2010). Target site–based herbicide resistance can also be associated with major mutations underlying adaptive phenotypes (Baucom, Reference Baucom2019). In general terms, Asp-2078-Gly and Cys-2088-Arg mutations confer resistance to FOPs, DIMs, and DEN (Kukorelli et al. Reference Kukorelli, Reisinger and Pinke2013). The remaining mutations confer different levels of resistance to these herbicides, depending on specific amino acid changes, the number of resistant alleles, weed species, plant growth stage, and recommended field rates for the herbicides (Kaundun Reference Kaundun2014).
While a resistance gene increases its frequency depending on the herbicides used, operational and biological factors (Powles and Yu Reference Powles and Yu2010) and knowledge of the mechanism of resistance can support an appropriate management strategy (Evans et al. Reference Evans, Tranel, Hager, Schutte, Wu, Chatham and Davis2016). That might mean an adequate rotation of herbicides with different sites of action (Scarabel et al. Reference Scarabel, Varotto and Sattin2007) or that are detoxified by different metabolic pathways, for example, cytochrome P450 monooxygenases and glutathione S-transferase (Busi Reference Busi2018); the correct mixture of specific herbicides (Beckie and Reboud Reference Beckie and Reboud2009); the possibility of exploiting fitness costs(Keshtkar et al. Reference Keshtkar, Abdolshahi, Sasanfar, Zand, Beffa, Dayan and Kudsk2019); and an assessment of the capacity for resistance spread within and among populations (Mithila and Godar Reference Mithila and Godar2013).
As indicated earlier, because of the importance of ACCase-inhibiting herbicides in controlling glyphosate-resistant Lolium spp. plants during both fallow and POST periods, a putative pinoxaden-resistant L. perenne population threatens to undermine the efficacy of these herbicides. As such, this study evaluated the sensitivity of a putative pinoxaden-resistant L. perenne population from Argentina to ACCase-inhibiting herbicides. In addition, the cause of the poor plant control achieved by pinoxaden in this population was determined.
Materials and Methods
Experimental Site and Plant Material
Experiments were carried out in Chacra Experimental Integrada Barrow (38.32°S, 60.23°W) from 2017 to 2019. Each experiment was repeated twice.
In December 2016, seeds were sampled from a putative pinoxaden-resistant L. perenne population that had arisen in a wheat field (38.32°S, 60.47°W), where the crop rotation involved wheat–soybean [Glycine max (L.) Merr.] and barley–soybean. In this system, weed control had been based on recurring applications of glyphosate and clethodim or haloxyfop during fallow; pinoxaden, dicamba, 2,4-D, and metsulfuron in wheat and barley crops; and glyphosate in soybean crop. In addition to the putative pinoxaden-resistant population, seeds were sampled from a neighboring population susceptible to ACCase-inhibiting herbicides established as weeds in a wheat field (38.43°S, 60.97°W) (without application of ACCase-inhibiting herbicides in the last 5 yr). Thirty plants of each population were chosen at random from wheat fields and were stored under laboratory conditions at 20 to 25 C before being used in the current study.
Dose–Response Assays
Herbicide sensitivities of progeny obtained from both populations were compared using the assessments outlined in the following sections.
Germination Test
Seed germination percentage and plumule growth were evaluated under different pinoxaden concentrations following the methodology described by Murray et al. (Reference Murray, Friesen, Beaulieu and Morrison1996) with modifications. Fifty seeds were placed in petri dishes containing filter paper and 10 ml of the following: pinoxaden (Axial®, pinoxaden [50 g ai L−1] plus cloquintocet-mexyl [12.5 g L−1]; Syngenta Agro S.A., Av. Del Libertador 1855, Vicente López, Argentina) media (each pinoxaden concentration included cloquintocet-mexyl in a 4:1 ratio, respectively): 0, 0.01, 0.1, 1, 10, 100, and 1,000 µM. For each treatment, there were four replicate petri dishes in a completely randomized design. The seeds were incubated in a growth chamber with 75 mmol m−2 s−1 of photosynthetically active radiation in a 12-h light/12-h darkness regime and temperatures of 25 and 20 C for day and night, respectively. After 7 d, the number of germinated seeds for which the radicle protruded from the seed coat ≥ 0.5 mm was recorded. In addition, plumule length was measured from the point of attachment to the seed to the tip of the coleoptile.
Plant Survival and Seed Production
In August 2017 and 2018, seeds of each population were sown in 2-L pots filled with soil. Seedlings on emerging were thinned to four to five per pot. The plants were grown in a greenhouse at 21 C (average temperature) in a completely randomized design and were irrigated at least every 2 d. Pinoxaden, clethodim, or quizalofop-P-tefuryl was applied to plants with 1 to 2 tillers using a laboratory belt sprayer calibrated to deliver 200 l ha−1. The herbicides were applied at the following doses: pinoxaden (Axial®, pinoxaden [50 g ai L−1] plus cloquintocet-mexyl [12.5 g L−1]; Syngenta Agro S.A.) at 0, 12.5, 25, 50, 100, and 200 g ai ha−1 (recommended dose: 30 to 40 g ai ha−1); clethodim (Select®, clethodim 240 g ai L−1; Bayer S.A., Ricardo Gutierrez 3652, Munro, Argentina) at 0, 32.5, 65, 130, 260, and 520 g ai ha−1 (recommended dose: 110 to 280 g ai ha−1); and quizalofop (Rango® GR, quizalofop-P-tefuryl 120 g ai L−1; Arysta Lifescience Argentina S.A., Enciso 1463, Tigre, Argentina) at 0, 21.8, 43.7, 87.5, 175, and 350 g ai ha−1 (recommended dose: 35 to 112 g ai ha−1). Following the manufacturers’ recommendations, 0.2% v/v ethoxylated alcohol was used as surfactant for clethodim and quizalofop spraying. For each treatment, there were five replicates in a completely randomized design, wherein each pot was a sampling unit.
At 45 d post herbicide application, plant survival was evaluated by recording the number of dead and live plants. Plants with severe visual injury (chlorosis of newly emerged leaves, general browning, and severe growth reduction) were recorded as “dead” plants, while “surviving” plants showed no differences in their growth or color of their leaves compared with control plants. These data were used to calculate the percentage of surviving plants per sampling unit (the number of surviving plants was divided by the total number of plants per pot and multiplied by 100).
At the end of the plant life cycle, all plants were harvested and manually threshed. After seeds were separated from the residual inflorescences, the number of seeds produced per plant was determined using a seed counter (Pfeuffer GMBH, Kitzingen, Germany).
The seeds obtained were stored at room temperature for 3 mo (time required to terminate seed dormancy). The viability of the seeds obtained from surviving plants was evaluated by incubating seeds in petri dishes containing filter paper and 10 ml of water and determining the germination percentage. One hundred seeds per treatment were evaluated. This test was carried out in a growth chamber under the conditions previously described.
Peroxidase Activity under Pinoxaden Treatment
Plants obtained from seeds of the susceptible and putative pinoxaden-resistant populations were grown in a greenhouse following the methodology and using the materials described earlier. When plants had developed 1 to 2 tillers, four individuals from each population were treated with 0 or 50 g ha−1 pinoxaden. The last expanded leaf of each plant was sampled at 10 d after application of the herbicide. Immediately, 0.3 g fresh weight of these samples were ground in nitrogen liquid and placed in 1.5-ml tubes containing 1 ml 50 mM potassium phosphate buffer (pH 6.5) and 10% (v/v) glycerol at 4 C. The tubes were then centrifuged at 400 × g for 5 min at 4 C. Peroxidase activity was determined for an aliquot of each supernatant, based on the oxidation of pyrogallol to purpuro-galline (ε = 2.47 mM−1 cm−1) (Puntarulo et al. Reference Puntarulo, Sánchez and Boveris1988). The reaction mixtures consisted of 50 mM potassium phosphate buffer (pH 6.5), 45 mM pyrogallol, and 8 mM H2O2 at 25 C. Absorbance was measured on a spectrophotometer at 430 nm, and peroxidase activity was expressed in terms of the amount of purpuro-galline produced (mmol min−1 g−1 fresh weight of leaf).
Test of the Inhibition of Pinoxaden Detoxification through Cytochrome P450 Monooxygenases
The inhibition of pinoxaden detoxification had been performed using malathion as a cytochrome P450 monooxygenase inhibitor (Keith et al. Reference Keith, Lehnhoff, Burns, Menalled and Dyer2015; Matzrafi et al. Reference Matzrafi, Gadri, Frenkel, Rubin and Peleg2014). Based on this, the interaction between pinoxaden and malathion was evaluated on plumule growth. The susceptible and putative pinoxaden-resistant populations were tested in this assessment; however, another pinoxaden-resistant L. perenne population was also included in the evaluation as a positive control, as this was known to undergo detoxification through cytochrome P450 monooxygenases (Yanniccari et al. Reference Yanniccari, Gigón, Istilart and Castro2018).
Three hundred seeds of each population were incubated in petri dishes containing a wet filter paper in a growth chamber, with 75 mmol m−2 s−1 of photosynthetically active radiation, in a 12-h light/12-h dark regime, at temperatures of 25 and 20 C for day and night, respectively. After 48 h, germinated seeds with a radicle length of ≥ 0.2 mm were transferred to 10-ml glass test tubes (4 seeds per tube) containing cotton and 1 ml of one of the following four treatments: 1 µM pinoxaden, 10 ppm malathion, 1 µM pinoxaden plus 10 ppm malathion, and deionized water (control). Ten tubes were used as replicates for each population and treatment. After incubation for 5 d in a growth chamber under the conditions described above, the plumule length was measured from the point of attachment to the seed to the tip of the coleoptile.
Partial Sequencing of ACCase
Total DNA was extracted from leaf tissue of six plants of the susceptible and resistant populations by the cetyltrimethylammonium bromide protocol of Doyle and Doyle (Reference Doyle and Doyle1990). DNA yield and quality were evaluated spectrophotometrically. The DNA was used as a template to amplify two regions of the ACCase gene sequence, using primer pairs, ACCase A (F 5′-GGCTCAGCTATGTTCCTGCT-3′ and R 5′-CAAGCCTACCCATGCATTCT-3′) and ACCase B (F 5′-GGCTCAGCTATGTTCCTGCT-3′ and R 5′-CAAGCCTACCCATGCATTCT-3′), according to Matzrafi et al. (Reference Matzrafi, Gadri, Frenkel, Rubin and Peleg2014). The amplified sequences encompassed the positions of all the known mutations described earlier. Polymerase chain reaction (PCR) products were purified by ethanol precipitation at −20 C. The samples were then centrifuged at 32,000 × g for 20 min at 4 C. The resulting pellets were washed three times with 70% ethanol and then resuspended in ultrapure water. Nucleotide sequences were obtained for the PCR fragments in these DNA preparations. The sequence data obtained were cleaned, aligned, translated, and compared at the Ile-1781, Trp-1999, Trp-2027, Ile-2041, Asp-2078, Cys-2088, and Gly-2096 codons.
Statistical Analysis
Dose–response data were used to build dose–response curves with a nonlinear log-logistic regression model:
$$y = D/\left( {{\rm{1 }} + {\rm{ }}{{\left( {x/{{\rm{I}}_{{\rm{5}}0}}} \right)}^b}} \right)$$where y represents the response at herbicide dose x, D is the upper asymptote, I50 is the herbicide dose required to achieve 50% of the maximum response, and b is the slope of the line at I50. To assess the accuracy of the model, F-tests for model significance, coefficient of determination (R2), and residual variance analysis were calculated. Models obtained were compared with the extra sum-of-squares F-test using GraphPad Prism® v. 6.01 (GraphPad Software, San Diego, CA, USA). Based on this analysis, population data from replicated experiments were pooled when P > 0.05 and model parameters were recalculated using the combined evidence. Finally, the I50 values from susceptible and resistant populations were compared with the F-test (P < 0.05) (GraphPad Software). Resistance indices (RI) were calculated as the ratio of I50 of the pinoxaden-resistant population compared with the susceptible population.
Data from peroxidase activity assays and tests for inhibition of pinoxaden detoxification were analyzed by multifactorial analysis of variance (ANOVA). Data from replicated experiments were pooled when no significant differences between data sets were detected (P > 0.05). In all cases, residual plots indicated that variances were normally distributed and homogenous. When ANOVA indicated significant effects, means were compared using Fisher’s protected least significant difference test (P < 0.05).
Results and Discussion
Pinoxaden Resistance
At a pinoxaden dose of 50 g ai ha−1 (25% higher than the recommended dose of 30 to 40 g ha−1; Syngenta Agro S.A.) approximately 80% of plants from the putative resistant population survived, while only 7% of plants remained alive in the susceptible population (Figure 1A). However, only surviving plants from the pinoxaden-resistant population produced seeds at this dose (Figure 1B). When 200 g ai ha−1 of pinoxaden (the maximum dose evaluated) was applied, pinoxaden-resistant plants produced around 15% of the seeds obtained from untreated control plants (Figure 1B).
Effect of pinoxaden doses on plant survival (A) and seed production (B) in a resistant and a susceptible Lolium perenne population. Black and white symbols represent mean values from experiments conducted during 2017 and 2018, respectively, and bars indicate ±1 standard error of the mean. The predicted responses are shown by lines according to the adjusted models (y = D/(1 + (x/I50)b).
Regression models of plant survival and seed production for both populations were compared, and I50 parameters differed significantly between susceptible and pinoxaden-resistant populations. RIs for plant survival and seed production were 5.1 and 2.8, respectively (Table 1).
Variables measured in dose–response assays and parameter estimates from the nonlinear log-logistic regression model (y = D/(1 + (x/I50)b) of resistant and susceptible Lolium perenne populations treated with pinoxaden (data from two replicated experiments).a
a Asterisks indicate significant differences between herbicide doses required to achieve 50% of the maximum response calculated on susceptible and resistant populations. Resistance indices (RI) are shown.
When peroxidase activity was evaluated, ANOVA results showed a significant interaction between population (susceptible or resistant) and treatment (with or without pinoxaden) (P = 0.04), but differences between replicated experiments were not statistically significant (P > 0.05). After pinoxaden treatment, the susceptible plants showed indications of oxidative stress, with an increase in peroxidase activity of about 70% compared with controls not treated with herbicide (Figure 2). ACCase-inhibiting herbicides induce oxidative stress in grasses, resulting in enhanced lipid peroxidation and membrane damage (Lukatkin et al. Reference Lukatkin, Gar’kova, Bochkarjova, Nushtaeva and Teixeira da Silva2013). Nevertheless, the putative pinoxaden-resistant plants showed no significant difference in peroxidase activity between the untreated control and those treated with pinoxaden (Figure 2). The putative pinoxaden-resistant plants also did not exhibit stress symptoms associated with the herbicide application and completed their life cycles, producing viable progeny, showing an average percentage germination of 77% (SD ±5%). As this result included seed from surviving plants of the susceptible population treated with pinoxaden, it illustrated the capacity of this species for perpetuation.
Peroxidase activity of susceptible and resistant Lolium perenne plants. Purpuro-galline produced by peroxidase at 10 d post application of pinoxaden compared with untreated control plants. Columns represent mean values, and vertical bars indicate the standard error of the mean (pooled data from replicated experiments). Asterisk indicates significant differences (P < 0.05) between a treatment and the control in the same population.
The results supported an inherited ability of plants from the putative resistant population to survive and reproduce after a normally lethal dose of pinoxaden. This population had a 5-fold lower sensitivity to pinoxaden than the susceptible population (Table 1). Interestingly, this level of resistance, relative to the susceptible population, seemed to change through the life cycle when RI values were compared. When the effects of pinoxaden on germination and plumule growth were evaluated, the resistant population showed RI values of 27.0 and 29.0, respectively (Table 1).
The low pinoxaden sensitivity of the resistant population was expressed from germination (Figure 3A and B). ACCase-inhibiting herbicides primarily cause the inhibition of fatty-acid biosynthesis, triggering reduction in growth (Cob and Reade Reference Cob, Reade, Cob and Reade2010; Kukorelli et al. Reference Kukorelli, Reisinger and Pinke2013). Fatty-acid synthesis is required for the early stages of cell growth, playing an important role during the processes of germination and seedling growth (Ghosh and Sen-Mandi Reference Ghosh and Sen-Mandi2018; Sasaki and Nagano Reference Sasaki and Nagano2004). The germination and plumule growth percentages for the susceptible population were reduced by half at pinoxaden concentrations of 0.6 and 0.3 µM, respectively, in comparison to controls without herbicide treatment (I50; Table 1). However, those concentrations of pinoxaden did not affect the germination or plumule growth of the pinoxaden-resistant population (Figure 3A and B). These seedlings could adequately carry out fatty-acid biosynthesis at pinoxaden treatments ≤ 1 µM (Figure 3A and B).
Effect of pinoxaden concentrations on germination (A) and plumule growth (B) in a resistant and a susceptible Lolium perenne population. Black and white symbols represent mean values from replicated experiments, and bars indicate ±1 standard error of the mean. The predicted responses are shown by lines according to the adjusted models (y = D/(1 + (x/I50)b).
As was discussed in a previous report, the contrasting behavior of susceptible and resistant plants on the germination test could not be explained by differential transport of the herbicide (Yanniccari et al. Reference Yanniccari, Istilart, Giménez and Castro2012). Imbibing embryos with pinoxaden ensured that all tissues were exposed to the herbicide, therefore, eliminating the consequences of any differential transport between populations.
Mechanism of Resistance
Frequently, the development of resistance to ACCase-inhibiting herbicides has been associated with elevated levels of cytochrome P450 monooxygenase activity, suggesting a mechanism of detoxification (Powles and Yu Reference Powles and Yu2010; Siminszky Reference Siminszky2006). In the current work, malathion was used as an inhibitor of cytochrome P450. Only the non–target site resistance (NTSR) population with its known pinoxaden detoxification responded to the inhibitor, neutralizing its mechanism of resistance. ANOVA results showed no differences between replicated experiments (P > 0.05), but a significant interaction between population (susceptible, resistant, and NTSR) and treatment (malathion, pinoxaden, pinoxaden + malathion, and control without herbicide) was detected (P < 0.001). The pinoxaden treatment did not significantly inhibit plumule growth of the NTSR population; however, the combination of pinoxaden + malathion reduced growth to the same level as the susceptible population treated with the herbicide (Figure 4). By contrast, no significant differences were detected in plumule growth of the pinoxaden-resistant population treated with pinoxaden, pinoxaden + malathion, and the control without herbicide (Figure 4). As a result, there was no evidence of pinoxaden detoxification mediated by cytochrome P450 monooxygenase in this population.
Effect of cytochrome P450 inhibition. Plumule growth of the susceptible and resistant Lolium perenne populations compared with an L. perenne population with cytochrome P450–mediated pinoxaden resistance (non–target site resistance [NTSR] as positive control), treated with 1 µM pinoxaden, 10 ppm malathion, 1 µM pinoxaden + 10 ppm malathion, and distilled water (control). Columns represent mean values, and vertical bars indicate the standard error of the mean (pooled data from replicated experiments). Asterisks indicate significant differences (P < 0.05) between a treatment and the control in the same population.
When potential target-site mechanisms of resistance were explored, a single-point mutation from adenine to guanine was detected, indicating a Asp-2078-Gly substitution in ACCase as the source of the phenotype in the pinoxaden-resistant population (Figure 5). It is known that a Asp-2078-Gly mutation confers resistance to pinoxaden and/or other ACCase-inhibiting herbicides in A. myosuroides, wild oat (Avena fatua L.), sterile oat (Avena sterilis L.), goosegrass [Eleusine indica (L.) Gaertn.], L. perenne ssp. multiflorum, L. rigidum, littleseed canarygrass (Phalaris minor Retz.), and hood canarygrass (Phalaris paradoxa L.) (Beckie et al. Reference Beckie, Warwick and Sauder2012; Cruz-Hipolito et al. Reference Cruz-Hipolito, Osuna, Domínguez-Valenzuela, Espinoza and De Prado2011, Reference Cruz-Hipolito, Fernandez, Alcantara, Gherekhloo, Osuna and De Prado2015; Délye Reference Délye2005; Gherekhloo et al. Reference Gherekhloo, Osuna and De Prado2012; Hochberg et al. Reference Hochberg, Sibony and Rubin2009; Liu et al. Reference Liu, Harrison, Chalupska, Gornicki, O’Donnell, Adkins, Haselkorn and Williams2007; Osuna et al. Reference Osuna, Goulart, Vidal, Kalsing, Ruiz Santaella and De Prado2012; Petit et al. Reference Petit, Bay, Pernin and Délye2010; Yu et al. Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007). Although the D2078 residue is not part of either herbicide-binding site, a Asp-2078-Gly mutation would introduce large conformational changes in the binding site for the ACCase inhibitors, contributing to a low affinity for several herbicides (Jang et al. Reference Jang, Marjanovic and Gornicki2013; Yu et al. Reference Yu, Kim and Tong2010).
Sequence of the ACCase gene obtained from the pinoxaden-resistant (R) and pinoxaden-susceptible (S) plants and the conceptual translation of the amino acid sequence. The resistance-conferring codon is show in the box. Numbers refer to amino acid positions of full-length ACCase in Alopecurus myosuroides.
Cross-resistance
The sensitivity of the populations to clethodim and quizalofop was also examined. While the recommended dose of clethodim ranged from 110 to 280 g ai ha−1 (Senseman Reference Senseman2007), there was 85% and 63% plant survival in the resistant population at 130 and 260 g ha−1, respectively (Figure 6A). Moreover, around 40% of the plants survived at 520 g ha−1, the highest dose of clethodim assayed (Figure 6A). At this dose, seed production was 45% that of control plants (without herbicide; Figure 6B). The seeds produced showed an average percentage of germination of 81% (SD ±7%), including all plants resistant to clethodim, indicating their capacity for perpetuation.
Dose response of a resistant and a susceptible population of Lolium perenne treated with clethodim, for plant survival (A) and seed production (B), and with quizalofop, for plant survival (C) and seed production (D). Black and white symbols represent mean values from experiments conducted during 2017 and 2018, respectively, and bars indicate ±1 standard error of the mean. The predicted responses are shown by lines according to the adjusted models (y = D/(1 + (x/I50)b).
The I50 parameters calculated for plant survival and seed production were significantly different between the resistant and susceptible populations (Table 2). In the resistant population, a 5.4-fold greater dose of clethodim was necessary to match the level of plant death achieved in the susceptible population (RI; Table 2). The RI for seed production was 3.3 when comparing the effects of clethodim on the resistant and susceptible populations (Table 2). The Asp-2078-Gly mutation endowed a significant level of resistance to clethodim at the field rate of application, consistent with the findings of Yu et al. (Reference Yu, Collavo, Zheng, Owen, Sattin and Powles2007). In addition, the RI values for plant survival and seed production on treatment with pinoxaden (5.1 and 2.8, respectively; Table 1) were similar to the RIs obtained for clethodim (5.4 and 3.3, respectively; Table 2). Therefore, the resistance conferred by the Asp-2078-Gly mutation resulted in the same level of insensitivity to both pinoxaden and clethodim.
Results of dose–response assays using clethodim and quizalofop showing variables measured and parameter estimates from the nonlinear log-logistic regression model (y = D/(1 + (x/I50)b) of resistant and susceptible Lolium perenne populations (data from two replicated experiments).a
a Asterisks indicate significant differences between herbicide doses required to achieve 50% of the maximum response calculated on susceptible and resistant populations. Resistance indices (RI) are shown.
Quizalofop application is recommended at doses of 35 to 112 g ha−1, with the lowest dose sprayed on soybean crops and the highest dose applied to non-crop areas (Senseman Reference Senseman2007). However, at 43.7 and 87.5 g ha−1, around 80% of the plants from the resistant population survived in the current study (Figure 6C), and they produced more than 50 seeds per plant (Figure 6D). At the highest dose sprayed (350 g ai ha−1), plant survival was around 40% and every plant produced an average of 41 seeds (Figure 6D). The seeds produced showed an average percentage germination of 71% (SD ±6%) for all surviving plants treated with quizalofop. For plant survival and seed production, the resistant populations showed the highest RI values after quizalofop application compared with pinoxaden and clethodim treatments (Table 2). The Asp-2078-Gly mutation confers resistance to all three classes of herbicide. The effect of this mutation on the level of sensitivity of ACCase to several herbicides representing the three chemical classes (but not including clethodim) was studied by Jang et al. (Reference Jang, Marjanovic and Gornicki2013), and they found that it provoked higher levels of resistance to quizalofop than pinoxaden.
The level of resistance to different ACCase-inhibiting herbicides depends on the specific mutation, weed species, plant growth stage, and recommended field rates of herbicides. In the case of Asp-2078-Gly and Cys-2088-Arg, both confer broad resistance to all ACCase-inhibiting herbicides currently tested (Kaundun Reference Kaundun2014). However, the advantages provided to the plant by Asp-2078-Gly seem not be the general rule, as Vila-Aiub et al. (Reference Vila-Aiub, Yu, Han and Powles2015) found that this mutation was associated with constraints in ACCase activity. These constraints included significant reductions in the specific activity and the maximal reaction velocity of this enzyme, which may lead to a shortage of lipids available for rapid growth, correlated with detrimental effects in the growth of L. rigidum plants homozygous for this mutation. Similarly, A. myosuroides plants have displayed a reduction of around 40% in biomass and seed production associated with the Asp-2078-Gly mutation in the homozygous state (Menchari et al. Reference Menchari, Chauvel, Darmency and Délye2008). Nevertheless, despite the fitness cost, Asp-2078-Gly has been the most frequently found ACCase mutation, being detected in nine of 54 L. multiflorum populations sampled from the United Kingdom (Alarcón-Reverte et al. Reference Alarcón-Reverte, Shanley, Kaundun, Karp and Moss2013). The frequency of this mutation appeared to be governed by the herbicide selection pressure applied by FOPs and DIMs (Kaundun Reference Kaundun2014). In Argentina, the selection pressure associated with the use of clethodim and haloxifop in preplant applications and pinoxaden on wheat and barley crops would also have favored the resistance mediated by the Asp-2078-Gly mutation in the L. perenne population studied.
The inherited ability of plants from the resistant population to survive and reproduce after a normally lethal dose of ACCase-inhibiting herbicides was demonstrated. Pinoxaden-, clethodim-, or quizalofop-treated L. perenne plants completed their life cycle, producing viable seeds.
The level of resistance to these ACCase inhibitors was assessed by dose–response evaluation of plant survival or seed production, and RIs were higher for plant survival than seed production. While growth or survival comparisons between susceptible and resistant plants gave an approximation about the relative sensitivity to an herbicide, considering these results, viable seed production should also be a variable taken into account in evaluating herbicide resistance.
The resistant L. perenne population would evolve under a high use of ACCase inhibitors in order to control glyphosate resistant plants. In this context, nonchemical methods of weed management (seedbank control, crop rotation, and competitive crops, among others) (Jabran and Chauhan Reference Jabran, Chauhan, Jabran and Chauhan2018) and use of herbicide mixtures (Diggle et al. Reference Diggle, Neve and Smith2003) and herbicide rotations, taking into account the possible mechanisms of resistance (target-site resistance or detoxification, mainly), should be implemented to delay the evolution of herbicide resistance and prevent Asp-2078-Gly gene flow.
Acknowledgments
This research was partially supported by Consejo Nacional de Investigaciones Científicas y Técnicas and Instituto Nacional de Tecnología Agropecuaria (2019-PE-E4-I086-001). We sincerely thank four anonymous reviewers for their insightful comments that substantially improved the article. No conflicts of interest have been declared.
