Introduction
Weed control in lentil crops is a challenge because lentils are an inherently weak competitor with weeds, they have low vigor in early growth, and few postemergence herbicides are available to control broadleaf weeds that grow among the crop (McDonald et al. Reference McDonald, Hollaway and McMurray2007). Over the last several decades, the dicotyledonous annual weeds common sowthistle (Sonchus oleraceus L.) and prickly lettuce (Lactuca serriola L.) have become more common in southern Australia because they are favored by no-till crop seeding practices (Weaver and Downs Reference Weaver and Downs2003; Widderick et al. Reference Widderick, Sindel and Walker1999). These weeds have proven to be especially problematic in lentil crops and are common in areas where lentil is intensively grown. Due to demand for alternative plant-based protein sources and market price signals, lentil is an increasingly popular break crop in regions of Australia where winter annual grains are grown. However, weed management is a barrier to the sustainability of lentil production as an economically viable component in these crop-rotation systems.
S. oleraceus and L. serriola produce a milky sap and can cause mechanical problems at lentil harvest, they increase moisture levels in harvested grain and they also affect grain quality by staining the lentil grains (Amor Reference Amor1986; Widderick Reference Widderick and Preston2019). Their seed is readily dispersed by wind (Hutchinson et al. Reference Hutchinson, Colosi and Lewin1984; Weaver and Downs Reference Weaver and Downs2003), and they are taller than the lentil crop, which further facilitates their dispersal. In addition, the weeds are prolific seed producers, especially in the absence of competition (Amor Reference Amor1986; Hutchinson et al. Reference Hutchinson, Colosi and Lewin1984). Furthermore, both species have evolved widespread resistance to the sulfonylurea (Lu et al. Reference Lu, Baker and Preston2007) and imidazolinone (IMI) chemical families of the acetolactate synthase (ALS)-inhibiting herbicides in Australia (Merriam et al. Reference Merriam, Boutsalis, Malone, Gill and Preston2018). Cases of resistance to 2,4-D and glyphosate have also been reported in Australia in both species (Heap Reference Heap2022), although these are less common.
Due to their short stature and low early vigor, lentil plants are weakly competitive and have a long critical period of weed control before canopy closure is achieved (Fedoruk et al. Reference Fedoruk, Johnson and Shirtliffe2011; McDonald et al. Reference McDonald, Hollaway and McMurray2007). Furthermore, lentils have little natural tolerance to herbicides and very few postemergence options are available for controlling weeds that grow among them (McDonald et al. Reference McDonald, Hollaway and McMurray2007). Weed management in lentil focuses heavily on selecting fields with low weed burden and controlling weeds prior to crop seeding and emergence (GRDC 2018). However, weeds such as S. oleraceus and L. serriola are known to germinate year-round if conditions are favorable (Chadha et al. Reference Chadha, Florentine, Chauhan, Long, Jayasundera, Javaid and Turville2019; Chauhan et al. Reference Chauhan, Gill and Preston2006), and they can readily colonize fields from adjacent roadsides and uncropped areas due to their mobility (Hutchinson et al. Reference Hutchinson, Colosi and Lewin1984; Lu et al. Reference Lu, Baker and Preston2007).
The first herbicide-tolerant lentil variety in Australia was commercially released in 2011 and had tolerance to the IMI herbicides that allowed growers to use the herbicides in-crop for broadleaf weed control (Bruce et al. Reference Bruce, Roberts, Gutsche and Day2019; Pulse Breeding Australia 2011; Rodda et al. Reference Rodda, Rosewarne, Sounness and Longson2016). These varieties are now widely grown in Australia. However, the ALS-inhibiting herbicides have a high propensity for evolving resistance (Tranel and Wright Reference Tranel and Wright2002). This is frequently realized by the availability of IMI-resistant varieties of other crops such as barley, oat, and wheat, leading to the overuse of these herbicides. Currently in Australia, more than 20 different grass and broadleaf weed species that grow with winter annual crops have developed resistance to ALS-inhibiting herbicides (Heap Reference Heap2022).
Throughout southern Australia, 78% of S. oleraceus populations have become resistant to sulfonylurea herbicides and 68% have become resistant to the IMIs (Merriam et al. Reference Merriam, Boutsalis, Malone, Gill and Preston2018). Resistance frequencies in L. serriola have previously been reported as 66% and 82% in 1999 and 2004, respectively (Lu et al. Reference Lu, Baker and Preston2007), and have likely increased since. This highlights the need for alternative strategies, including new herbicide-tolerance lentil traits, to effectively control these weeds and reduce the selection pressure for resistance to ALS-inhibitor herbicides. New lines of metribuzin-tolerant lentil were discovered through mutagenesis and have recently been released (McMurray et al. Reference McMurray, Preston, Vandenberg, Mao, Bett and Paull2019a, Reference McMurray, Preston, Vandenberg, Mao, Bett and Paull2019b, Reference McMurray, Preston, Vandenberg, Mao, Oldach, Meier and Paull2019c). Novel varieties of lentil being developed through this pipeline could be a valuable tool in improving weed control options for lentil growers.
Metribuzin inhibits photosystem II (PSII), and although resistance to it has been reported in Australia, cases are currently limited to only three weed species in winter annual cropping systems, and no resistance has been found in S. oleraceus or L. serriola (Heap Reference Heap2022). Resistance to PSII-inhibiting herbicides is primarily conferred by a target-site mutation and associated with a significant fitness penalty, although cases of metabolic resistance have also been reported (Gronwald Reference Gronwald, DePrado, Jorri and GarciaTorres1997). Herbicide tolerance in metribuzin-tolerant lentil is also conferred by a target-site mutation and accompanied by a fitness penalty. This fitness penalty will continue to be a focus for breeding programs (McMurray et al. Reference McMurray, Preston, Vandenberg, Mao, Bett and Paull2019a). However, the fitness penalty, if present in weeds, can be beneficial for their management, because it would result in selection for susceptible biotypes in the absence of herbicide selection pressure (Keshtkar et al. Reference Keshtkar, Abdolshahi, Sasanfar, Zand, Beffa, Dayan and Kudsk2019). Therefore, provided it is used in moderation as part of a diverse rotation of herbicides, the metribuzin-tolerant crop technology has a benefit over ALS-inhibitor resistant technology for which little to no fitness penalty exists in resistant weeds (Tranel and Wright Reference Tranel and Wright2002).
Metribuzin is registered for control of S. oleraceus and L. serriola in Australia in lentil crops, but at the preemergence stage only (Genfarm 2019). Rates range from 135 to 285 g ai ha−1, depending on soil texture, and the label also specifies a minimum sowing depth and warns of several situations that can result in herbicide damage to the crop. Other PSII-inhibiting herbicides, such as terbuthylazine, are also registered in lentil at the preemergence stage and can provide suppression of these weeds; however, crop damage can occur (Sipcam 2011). Both herbicide labels warn that significant rain shortly after application can wash the herbicide into the crop seed zone, causing injury. Heavy rains occur regularly in the winter-dominant crop production areas of southern Australia. Therefore, the use of these herbicides comes with a degree of risk.
A lentil variety with high levels of tolerance to metribuzin and allows the use of metribuzin at higher rates would significantly help improve control of broadleaf weeds such as S. oleraceus and L. serriola for lentil growers, and it would reduce the risk of crop damage. A novel metribuzin-tolerant lentil germplasm line, M043, was developed by the South Australian Research and Development Institute.
The aims of this study were to 1) identify the efficacy of metribuzin-based herbicide treatments for controlling ALS inhibitor-resistant S. oleraceus and L. serriola in a PSII-inhibiting herbicide-tolerant lentil variety, and 2) examine the crop safety of these herbicide treatments on the herbicide-tolerant lentil.
Materials and Methods
Experimental Design and Sampling
Three experimental sites were established at two locations in South Australia: Maitland in 2018 (34.37°S, 137.69°E) and 2019 (34.46°S, 137.71°E), and Roseworthy in 2019 (34.54°S, 138.70°E). Both locations have a Mediterranean-type climate with hot, dry summers and cool, wet winters. Both locations are dominated by annual winter crop production. The experimental plots were sown at a seed rate of 50 kg ha−1 for a target crop density of 120 plants m−2 with a six-row plot seeder, with a knife-point and press wheel system that prevents furrow wall collapsing back onto the crop row. The experiments were laid out in a randomized complete block design with three replications. Plots were 10 m long and contained six crop rows spaced 22.5 cm apart. Monoammonium phosphate fertilizer (Incitec Pivot Ltd, Melbourne, Vic, Australia) was applied at 80 kg ha−1 banded below the seed. Gravimetric soil moisture (Shukla et al. Reference Shukla, Panchal, Mishra, Patel, Srivastava, Patel and Shukla2014) at a soil depth of 2 to 10 cm was low at sowing time (15.5% at Maitland in 2018 and 14.3% at Maitland in 2019). Soil moisture was not measured at the Roseworthy site. Weeds were controlled prior to sowing at all experimental sites with an application of 900 g ai ha−1 propyzamide and 1,080 g ae ha−1 glyphosate, with the addition of 12 g ai ha−1 carfentrazone-ethyl at the Maitland sites (Table 1). The experiment areas were rolled shortly after crop emergence, and fungicides and insecticides were applied throughout the season as required. A postemergence application of 240 g ai ha−1 clethodim (Table 2) was applied across all experimental sites in July to control grass weeds.
Table 1. Herbicide used in the experiments.
a Applied to the entire experimental areas for site preparation.
b Applied postemergence to the entire experimental areas tocontrol of grass weeds that were not the focus of the experiment.
Pre-plant incorporated (PPI) herbicide treatments were applied immediately before sowing, post-sowing preemergence (PSPE) treatments were completed within the week following sowing, and postemergence treatments were applied at the 5-node crop growth stage (Table 2). At Roseworthy, all herbicides were applied using a hand-held CO2-pressurized hand boom equipped with TT110015 nozzles (TeeJet Australasia, Charlton, Vic, Australia) operating at 150 kPa and 5 km h−1, while at Maitland herbicides were applied using a tractor-mounted shrouded sprayer equipped with Minidrift 015 nozzles (Hardi, Cavan, SA, Australia) operating at 200 kPa and 5.8 km h−1. At both sites, nozzles were spaced 50 cm apart, 50 cm above the crop, and total spray volume output was 100 L ha−1. Herbicide treatments were designed around two rates of metribuzin, applied PSPE or postemergence, and with or without a supplemental postemergence application of diflufenican. This treatment list was supplemented with two rates of postemergence terbuthylazine and a low rate of metribuzin applied PSPE + diflufenican applied postemergence for comparison purposes (Tables 1 and 3).
Table 2. Site characteristics and key experimental management dates of the three study site-years.
a Abbreviations: GSR, growing season rainfall (in temperate Australia, growing season rainfall is that received between April and October); IBS, POST, postemergence; PSPE, post sowing–preemeergence.
b Soil type based on the Australian Soil Classification system described by Isbell (Reference Isbell2002).
c Experimental year and average GSR are presented in millimeters (mm).
Table 3. Densities of S. oleraceus and L. serriola assessed at 8 wk following application of final postemergence herbicide treatments at Maitland in 2018 and 2019, and Roseworthy in 2019.a
a Abbreviations: PPI, pre-plant incorporated; PSPE, post-sowing preemergence; POST, postemergence.
b Means within a column followed by the same letter are not significantly different at P < 0.05.
To supplement the background levels of the target weeds at Roseworthy, seeds of S. oleraceus and L. serriola from locally representative populations were evenly spread across the plots at a density of 120 seeds m−2 and 25 seeds m−2, respectively. The Maitland experimental sites had sufficient background populations of these weeds. Densities of S. oleraceus and L. serriola were assessed just prior to 8 wk postemergence herbicide applications from three randomly selected spots within a plot using a 50-cm by 50-cm quadrant.
Due to the potential for crop damage from some of the herbicides used in the study, lentil plants were also visually assessed for any signs of chlorosis, necrosis, and stunting during each visit to the experimental sites. Plant chlorosis score (1 = no chlorosis, 9 = plant death), necrosis score (1 = no necrosis, 9 = plant death), and stunting score (1 = no stunting, 9 = plant death) were used to assess the herbicide damage. The herbicide damage ratings were completed at 4 wk after postemergence herbicides were applied. Lentil seed yield was determined by harvesting plots with a small-plot harvester. The collected seed sample was cleaned to remove weed seeds and chaff material and then weighed to determine yield.
Due to the strong relationship between the activity of herbicides used in the experiment and rainfall, daily rain totals for each experimental site were sourced from the nearest Bureau of Meteorology (BOM) weather station and overlaid with herbicide application dates to aid in interpreting the results (Figure 1). Data for Maitland were obtained from BOM weather station 022008, 2 km from the 2018 experimental site, and 5 km from the 2019 site. Data for the Roseworthy site were obtained from BOM weather station 023122, located 4 km from the experimental site (AGBM 2019).
Figure 1. Daily rainfall totals and important experimental management dates during the growing season, April (A) to October (O), at each of the site-years of the experiment: A) Maitland 2018, B) Maitland 2019, C) Roseworthy 2019. Herbicide treatments are indicated by red asterisks (pre-plant incorporated (PPI) applications), blue asterisks (post-sowing pre-emergence (PSPE) applications) and green asterisks (POST (POST) applications). Rainfall data are from the nearest available weather station: Maitland for the 2018 and 2019 Maitland sites (2 km and 5 km from the experimental sites, respectively) and Roseworthy AWS for the 2019 Roseworthy site (4 km from the experimental site) (Australian Government Bureau of Meteorology, 2019).
Data Analysis
Data were analyzed using the R statistics package (R Core Team 2022) along with the packages agricolae (de Mendiburu Reference de Mendiburu2020) and ASReml (Gilmour et al. Reference Gilmour, Gogel, Cullis and Thompson2009). Two-way ANOVA revealed that site differences were present, so herbicide treatment data were analyzed separately for each site. A square root transformation was used for weed density data before analysis to normalize the distribution of residuals. Herbicide treatment effects on weed density were found to be significant using ANOVA, and multiple comparison was applied based on square root–transformed data using Tukey’s HSD. Plant count data are presented as the untransformed values to aid easy identification of population size.
Results and Discussion
Weed Control
Significant herbicide treatment effects that reduced weed populations were observed for both L. serriola and S. oleraceus in all site-years (Table 3). The density of both weed species was less after all herbicide treatments at Maitland in 2018 (P < 0.05). However, some treatments at Maitland in 2019 failed to reduce weed populations to any significant degree, including terbuthylazine applied at either rate and metribuzin applied PSPE at the low rate to control S. oleraceus and postemergence at the low rate to control L. serriola. At Roseworthy in 2019, reductions in L. serriola populations were significantly reduced only when terbuthylazine was applied at the low rate, metribuzin applied PSPE at the low rate followed by diflufenican, metribuzin applied postemergence at the high rate, and metribuzin applied postemergence at the high rate followed by diflufenican. Also at Roseworthy in 2019, six treatments resulted in a significant reduction in S. oleraceus populations. These included 1) terbuthylazine applied at the high rate, 2) metribuzin applied PSPE at the high rate, 3) and 4) metribuzin applied PSPE at both rates when followed by diflufenican, 5) metribuzin applied postemergence at the high rate, and 6) metribuzin applied postemergence at the high rate with diflufenican (Table 3).
The herbicide treatments that resulted in the lowest densities of both weeds across all sites included metribuzin (540 g ai ha−1) applied postemergence (with and without diflufenican), PSPE (with and without diflufenican), and metribuzin (270 g ai ha−1) applied to PSPE (with diflufenican). The treatments that were least effective included the terbuthylazine treatments (particularly the lower rate), metribuzin (270 g ai ha−1) applied PSPE (without diflufenican), and metribuzin (270 g ai ha−1) applied postemergence (with and without diflufenican).
S. oleraceus density was higher on average than L. serriola at Maitland in 2018 and Roseworthy in 2019, whereas at Maitland in 2019, L. serriola density was higher. Although the weed seedbank was artificially supplemented at the Roseworthy site, it still had a lower weed density than both Maitland sites. Although the combined weed density at Maitland in 2019 was similar to that of Maitland in 2018, despite 42.6 mm less rain in the 2019 growing season, the dominant species was different.
With the exception of S. oleraceus at Maitland in 2018 and L. serriola at Maitland in 2019, initial weed densities in the study were low. This reduced our ability to differentiate between herbicide treatments because differences were small, such as at Roseworthy in 2019. However, although crop yield impacts occur only when the density of these weed species is very high (Manalil et al. Reference Manalil, Ali and Chauhan2020), other issues can arise even when weed density is low. For example, weed plants that grow along with a weakly competitive lentil crop can produce large amounts of seed (Amor Reference Amor1986; Hutchinson et al. Reference Hutchinson, Colosi and Lewin1984). Additionally, large S. oleraceus plants present at harvest can reduce grain quality by increasing moisture levels at harvest and stain the grain (Amor Reference Amor1986; Widderick Reference Widderick and Preston2019).
Effect of Metribuzin Rate
Metribuzin provided better weed control at the higher rate (540 g ai ha−1) than at the lower rate (270 g ai ha−1), especially control of S. oleraceus (Table 3). When metribuzin was applied PSPE without a follow-up application of diflufenican, the higher rate of metribuzin provided significantly better control of S. oleraceus at Maitland in 2019. At Maitland in 2018 there was a significant difference in S. oleraceus populations between plots that received high and low rates of metribuzin at PSPE and plots that received postemergence applications of metribuzin alone. There were no differences between rates of metribuzin when diflufenican was applied as a supplement.
Effect of Application Timing and Supplementation with Diflufenican
Applying metribuzin PSPE versus postemergence showed no significant difference in weed control of either species in any of the trials (Table 3). However, PSPE applications of metribuzin at both the 270 and 540 g ai ha−1 rates followed by an application of diflufenican achieved greater control of L. serriola at Maitland in 2019 compared with postemergence applications (Table 3), indicating there may be an advantage of applying metribuzin earlier under some conditions. This increased control of weeds with the PSPE treatments with a supplemental application of diflufenican versus a postemergence application could be due to splitting the herbicide applications over two periods of exposure, thereby diminishing the seed emergence of L. serriola at several points over the growing season. Although the metribuzin label does not specifically list using it to control L. serriola, it notes that to achieve optimal control, weeds should not be larger than the 3-leaf stage (Genfarm 2019). When metribuzin plus diflufenican is applied later, some weeds may be too large to control.
Adding diflufenican to metribuzin (270 g ai ha−1) when applied PSPE provided better control of S. oleraceus at Maitland in 2019 (Table 3). Metribuzin applied at 540 g ai ha−1 PSPE provided complete control of S. oleraceus populations at all three trial sites. There was no increase in control efficacy of either S. oleraceus and L. serriola when metribuzin was applied postemergence with diflufenican as a supplement (Table 3).
Comparison with Terbuthylazine Treatments
An increased rate of terbuthylazine (1,050 g ai ha−1 vs. 750 g ai ha−1) did not provide better control of either weed species in any site year. Terbuthylazine treatments performed similarly to the most effective metribuzin treatments on L. serriola in all site-years, and on S. oleraceus at both sites in 2019 (Table 3). However, at Maitland in 2018 all the high-rate metribuzin treatments provided significantly better control of S. oleraceus than the lower rate of terbuthylazine, and all treatments but one (metribuzin 540 g ai ha−1 applied postemergence without diflufenican) outperformed the high rate of terbuthylazine.
Precipitation at the Sites
Metribuzin and terbuthylazine are primarily absorbed by plant roots, and therefore, their efficacy is reliant on adequate soil moisture levels that ensure mobility and root uptake (Black Reference Black1984; Congreve and Cameron Reference Congreve and Cameron2023; Genfarm 2019; Sipcam 2011). The registered use patterns of these herbicides in Australia recommends that both herbicides be applied to moist soils, and best results are achieved when followed by 6 to 12 mm of rain within 2 wk of metribuzin application, and 20 to 30 mm within 2 to 3 wk of terbuthylazine application (Genfarm 2019; Sipcam 2011). However, heavy rain following application can wash the herbicide into the crop seed zone, causing crop damage and thereby reducing its effectiveness. Diflufenican applied postemergence is primarily absorbed through the leaves of target weeds, so it must be applied before canopy closure, ideally between 4 and 6 wk after sowing (Bayer Crop Science, 2002). All three products function best when weeds are actively growing and unaffected by moisture stress.
In 2018, the Maitland site received an above-average amount of growing season rain that was distributed more evenly throughout the growing season than at the two sites in 2019 (Figure 1). The crop was sown following 30 mm of rain in the previous week, providing adequate soil moisture. However, only 14 mm of rain fell in the 3 wk following terbuthylazine application. Metribuzin was applied PSPE about a week after the PPI treatments, when no rain fell in the previous 10 d, and was followed by about 12 mm in the following two weeks. Metribuzin and diflufenican were also applied postemergence under suitable conditions shortly after a rain, followed by sufficient rain. This site had the longest interval between sowing and postemergence treatments at just under 8 wk, but these were still applied prior to canopy closure.
In 2019, Maitland experienced a low-growing season rain characterized by one significant event (>30 mm), several medium events (>10 mm), and many smaller events throughout the season (Figure 1B). Approximately 25 mm of rain fell in the week before sowing and PPI treatments, and more than 30 mm of rain fell in the 3 wk following, providing optimal conditions for terbuthylazine to take effect. PSPE treatments of metribuzin were applied a week later and were also followed by sufficient rain. Metribuzin and diflufenican were applied postemergence about a week after the heavy rain of the season, and nearly 30 mm of rain fell in the following 2 wk, also providing good conditions for herbicide activity. The diflufenican treatments were applied 6 wk after sowing, within the optimal window for herbicide effectiveness.
The 2019 Roseworthy site had little growing-season rain, and it was primarily concentrated in two heavy rains, with the rest consisting of small, sporadic rains (Figure 1C). Only 8 mm fell in the week before sowing, meaning that less surface soil moisture was present at this site compared with the other sites. A heavy rain of 33 mm fell 8 d after sowing. Postemergence treatments were applied approximately 6 wk after sowing and under more favorable conditions with sufficient rain in the weeks before and after.
The patterns of rain at each site-year (Figure 1) may have affected the behavior of individual herbicides and contributed to some of the differences in weed control between the sites. In particular, herbicides such as metribuzin and terbuthtylazine (Group 5 herbicides as categorized by the Herbicide-Resistance Action Committee and Weed Science Society of America), which are absorbed via soil uptake (Black Reference Black1984), will likely exhibit greater efficacy with better soil moisture following a rain, leading to higher uptake of the herbicide by weed roots. For example, the higher amounts of rain at Maitland in 2018 would improve the activity of these herbicides.
Crop Safety and Yield
The herbicide treatments used in this study had no effect on crop establishment or yield in any of the site-year trials. The M043 lentil line used in this study has a 33-fold greater tolerance of the field rate of metribuzin than the conventional parent line, PBA Flash (McMurray et al. Reference McMurray, Preston, Vandenberg, Mao, Bett and Paull2019a), with 50% reduction in dry weight (GR50) values greater than the field-use rate of metribuzin (McMurray et al. Reference McMurray, Preston, Vandenberg, Mao, Bett and Paull2019a). Some bleaching of lentil leaves was observed after diflufenican was applied, which lasted for up to 4 wk; however, this did not affect lentil yield (Table 4). Due to effective weed control achieved by the herbicide treatments, there was no effect on crop yield between the untread control and any of the herbicide treatments (data not presented).
Table 4. Lentil crop establishment in all 3 site years and yield in 2018 at Maitland and 2019 at Roseworthy.a,b
a Means within a column followed by the same letter are not significantly different at P < 0.05.
b Yield data were not recorded at Maitland in 2019 due to significant yield loss through pod drop prior to harvest caused by extreme winds.
The experimental plots were sown at a seed rate of 50 kg ha−1 for a target crop density of 120 plants m−2. Crop establishment was higher at Maitland in both years (124 plants m−2 in 2018 and 120 plants m−2 in 2019) than at Roseworthy (77 plants m−2 in 2018; Table 4). Grain yield was measured at the 2018 Maitland and 2019 Roseworthy sites only, and was significantly higher at Maitland in 2018 (1,410 kg ha−1) than at Roseworthy in 2019 (660 kg ha−1). Poor crop establishment at Roseworthy could be related to low soil moisture at sowing. The poor establishment and the lower amount of growing-season rain at Roseworthy (209 mm at Roseworthy in 2019 compared with 326 mm at Maitland in 2018) likely contributed to the lower yield. All three sites experienced lower than average growing-season rain, but the shortfall was greater at Roseworthy (25%) than at Maitland in 2018 (17%).
Yield at Roseworthy was lower than the regional average of 1,300 kg ha−1 for lentil in that year (Lewis Reference Lewis2020). Yield at the Maitland site was equal to the regional average of 1,400 kg ha−1 (Lewis Reference Lewis2019). However, experiments on the Yorke Peninsula have previously demonstrated a yield penalty for this lentil cultivar resulting from the herbicide tolerance mutation (Roberts et al. Reference Roberts, Mao, Aggarwal, Day, Michelmore, Sutton, Mould and Gutsche2020). Improving yield and water use efficiency of this line will be a focus of breeding programs to prepare it for commercial release.
These results demonstrate the crop safety of the herbicide treatments on the M043 line in the field. In 2019, both sites received significant rain of more than 30 mm within weeks after application of PPI and PSPE herbicides, which is a risk factor for herbicide injury to the crop (Genfarm 2019; Sipcam 2011), yet no crop damage was observed. Currently, growers are using terbuthylazine for weed control in lentil crops, but crop damage can occur under certain environmental conditions. The introduction of the M043 metribuzin-tolerant line would allow growers to use metribuzin at higher and more effective rates with lentil crops.
The novel metribuzin-tolerant line has promise for weed control. Both PSPE and postemergence application timings, including a postemergence application of diflufenican, were effective. The addition of diflufenican provided some additional control of S. oleraceus when combined with the lower rate of metribuzin applied PSPE at one site. The metribuzin treatments were found to provide a superior level of control than the currently registered rates of terbuthylazine in conventional lentil. Because improved weed control is the primary driver for development of this breeding line, further research prior to its commercial release should focus on other aspects of weed management, such as control of other broadleaf weeds that are known to be problematic in lentil, and tolerance to other PSII-inhibiting herbicides.
Practical Implications
ALS-inhibitor resistance in S. oleraceus and L. serriola is extremely common throughout the areas where the field experiments were conducted (Lu et al. Reference Lu, Baker and Preston2007; Merriam et al. Reference Merriam, Boutsalis, Malone, Gill and Preston2018), and was confirmed in the local population of both species at the Roseworthy experimental site (Merriam et al. Reference Merriam, Malone, Gill and Preston2021) and the S. oleraceus population at the Maitland 2018 site (50% to 55% survival after exposure to Intervix 500 ml ha−1 and imazapic 200 g ha−1, respectively). The Maitland 2019 experimental site was selected following failure of Intervix to control both species the previous year. While this study shows imidazolinone-resistant biotypes are effectively controlled by the high rates of metribuzin in a tolerant lentil crop, the development of biotypes with multiple-herbicide resistance is a risk that will need to be managed. Multiple resistance to the herbicides that inhibit ALS and PSII is reported in several weed species, including annual bluegrass (Poa annua; Singh et al. Reference Singh, dos Reis, Reynolds, Elmore and Bagavathiannan2021) and Palmer amaranth (Amaranthus palmeri) (Chaudhari et al. Reference Chaudhari, Varanasi, Nakka, Bhowmik, Thompson, Peterson, Currie and Jugulam2020) in the United States, redroot amaranthus (Amaranthus retroflexus) (Francischini et al. Reference Francischini, Constantin, Oliveira, Takano and Mendes2019) and hairy beggarticks (Bidens pilosa) (Takano et al. Reference Takano, de Oliveira, Constantin, Braz, Franchini and Burgos2016) in Brazil, and Wimmera ryegrass (Lolium rigidum) in Australia (Ma et al. Reference Ma, Lu, Han, Yu and Powles2020).
Responsible use of herbicide-tolerant crop technology would see these new lentil cultivars included as part of a diverse crop rotation along with cereals, oilseeds, and other pulses, and including conventional lentil varieties where appropriate to ensure a diverse herbicide rotation. This will give growers better integrated weed management, cropping rotation, and herbicide resistance management options (GIA 2025). Other triazine-tolerant crops such as triazine-tolerant canola, should be used only sparingly in these rotations to reduce exposure to the PSII-inhibiting herbicides. Regular monitoring of crops for herbicide failures and suspected resistance can help to catch problems early, and seed-set reduction of any plants surviving to crop maturity can prevent problems carrying over into the following season (Boutsalis et al. Reference Boutsalis, Gill and Preston2016). While the fitness penalty inherent in target-site resistance to the PSII-inhibiting herbicides can be exploited to manage resistant weed populations (Keshtkar et al. Reference Keshtkar, Abdolshahi, Sasanfar, Zand, Beffa, Dayan and Kudsk2019), this relies on the removal of the herbicide selection pressure for several years to allow time for the susceptible biotype to outcompete the resistant biotype. This, however, may not be viable within cropping systems, and an alternate approach would be to reduce weed densities to very low levels, as identified in the present study, so that the chances of resistance evolution to any herbicide treatment, including the PSII inhibitors, are substantially reduced.
Acknowledgments
We thank Larn McMurray, Dili Mao, Jeff Paul and Klaus Oldach for their contributions. This research was made possible by the significant contributions of growers through both trial cooperation and the support of the Grains Research and Development Corporation. We also thank David Brunton, Ben Fleet, and Jerome Martin of the University of Adelaide Weed Science Research Group; and Greg Walkley, John Nairn, Phil Rundle, Patrick Thomas, and Jacob Nicolai of the South Australian Research and Development Institute’s Clare Research Centre for their technical assistance.
Funding
This research was supported by an Australian Government Research Training Program Scholarship, a Grains Research and Development Corporation (GRDC) Grains Research Scholarship (project UOA1801-003RSX), and by a strategic joint investment from GRDC and SARDI via project DAS 00168BA. The metribuzin-tolerant M043 lentil germplasm line was developed through GRDC-SARDI-University of Adelaide funded projects DAS 00107 and DAS00131.
Competing Interests
The authors declare they have no conflicts of interest.
Open access
