Frequency of herbicide resistance in wild oats (Avena spp.), brome grass (Bromus spp.) and barley grass (Hordeum spp.) as determined by random surveys across south-eastern Australia

Introduction

Grass weeds are a major problem in grain cropping systems across south-eastern Australia. Annual ryegrass (Lolium rigidum Gaud.) is the major species of concern; however, other species such as wild oats (Avena spp.), brome grass (Bromus spp.) and barley grass (Hordeum spp.) are of high importance (Llewellyn et al. 2016). Changes in management practices for crop production combined with the development of non-herbicidal control systems used to control herbicide-resistant annual ryegrass (Walsh and Powles 2007; Llewellyn et al. 2016) may increase the importance these other grass weed species, if they are either more suited to the cropping systems or less suited for control under new weed management systems.

A survey of weed species remaining in fields ranked wild oats, brome grass and barley grass the third-, fifth- and ninth-most important weeds, respectively, across south-eastern Australia with regard to crop yield and revenue losses (Llewellyn et al. 2016). In parts of this area they are considered even more important; for example, in the Mallee region of South Australia (SA) and Victoria, brome grass is considered the most important weed of cereal crops and second-most important in canola and pulses (Llewellyn et al. 2016).

In 1993, a survey of southern New South Wales (NSW) found wild oats to be the second-most common weed of cereal crops, present in 72% of fields, behind only capeweed (Arctotheca calendula) and ahead of annual ryegrass (Lemerle et al. 1996). A subsequent survey of a portion of this area in 2010 recorded wild oats in 63% of fields, behind only annual ryegrass in prevalence (Broster et al. 2012a). Barley grass was recorded in 25% of the fields in 1993 and 17% in 2010, whereas in both surveys, brome grass was recorded in ~10% of the fields (Lemerle et al. 1996; Broster et al. 2012a).

Although annual ryegrass is the main herbicide-resistant weed in Australia, herbicide resistance has also been reported in other species of grass weeds. Resistance was first reported in wild oats (Avena fatua L.) in 1985 (Piper 1990) and Avena sterilis subsp. ludoviciana in 1989 (Mansooji et al. 1992); in two species of barley grass (Hordeum glaucum Steud. and H. leporinum Link) in 1982 and 1987, respectively (Powles 1986; Tucker and Powles 1991); and in two species of brome grass (Bromus diandrus Roth and B. rigidus Roth) in 1998 (Boutsalis and Preston 2006).

Many of the herbicides used for annual ryegrass control are also effective on these other grass species; therefore, these weed species have been subject to selection for resistance to these herbicides even if they were not the major targeted species in the fields. Previous surveys looking at herbicide resistance in wild oats have found resistance to several herbicides, albeit at lower levels than for annual ryegrass collected from the same regions (Owen and Powles 2009; Broster et al. 2011a, 2012b, 2013; Owen and Powles 2016). Similarly, the results for brome grass and barley grass samples collected in surveys show either very low levels of resistance (Owen et al. 2015) or no resistance (Broster et al. 2010).

The aim of this study was to test for herbicide resistance in wild oats, brome grass and barley grass samples and where present determine its distribution across south-eastern Australia. The samples collected were tested with a range of herbicides used for the control of these species in winter cropping in south-eastern Australia.

Materials and methods

Results

Samples collected and crop distribution

Over the period 2013–17, 996 of the 1760 fields visited contained one or more of wild oats, brome grass and barley grass. In total, 663 wild oats, 366 brome grass and 262 barley grass samples were collected from these fields (Table 2). Of these 996 fields, 741 contained only one of these species (458 wild oats, 169 brome grass, 114 barley grass), 215 contained two species (107 wild oats and brome grass, 58 wild oats and barley grass, 50 brome grass and barley grass), and 40 contained all three species (Fig. 1).

Table 2.Number of wild oats, brome grass and barley grass samples collected by region and year.

State/region	Year surveyed	No. of fields visited	No. of samples collected
			Wild oats	Brome grass	Barley grass
New South Wales
Slopes	2013	199	100	39	43
Eastern	2014	75	35	11	13
Western	2015	164	96	30	50
Plains	2016	152	71	9	7
Southern	2017	162	108	23	10
Tasmania	2014	75	17	12	6
South Australia
Mid North	2013	150	47	45	19
Eyre Peninsula	2014	167	31	52	56
Mallee	2017	164	11	56	39
South East	2017	74	20	0	0
Victoria
Southern	2014	116	26	22	0
Wimmera–Mallee	2015	140	42	50	14
North East	2016	122	59	17	5
Total		1760	663	366	262

Fig. 1.

Sites at which wild oats, brome grass and barley grass samples were collected.

Discussion

For each of the grass species, resistance to at least one herbicide was identified in each of the states where samples were collected. The extent of resistance in these three grass species is lower than observed for annual ryegrass in Australia not only for the equivalent surveying period and region (Broster et al. 2022), but also in previous annual ryegrass surveys (Broster et al. 2011b, 2012b, 2013; Boutsalis et al. 2012; Owen et al. 2014). However, more than 20% of samples of wild oats collected were resistant to the ACCase-inhibiting herbicide clodinafop-propargyl and brome grass to the ALS-inhibiting herbicide mesosulfuron-methyl.

In the case of wild oats, there was considerable variation among regions sampled, with 32% of samples resistant to clodinafop-propargyl in NSW, but only 8% in Victoria and 4% in SA (Table 3). This higher frequency of resistance in NSW is also higher than reported by Owen and Powles (2016), who found only 7% of samples from Western Australia (WA) resistant to diclofop-methyl, another ACCase-inhibiting herbicide. These frequencies are lower than those reported from Canada, where Beckie et al. (2013) found 44% of wild oats samples from the Canadian Prairies to be resistant to one or more fop herbicides. However, Travlos et al. (2011) found that 28% of wild oats samples from wheat fields in Greece were resistant to clodinafop-propargyl. By contrast, a low level of resistance to fop herbicides (9%) was identified in the Canterbury region of New Zealand (Buddenhagen et al. 2021).

Previous surveys have shown differences in the extent of resistance between the fop herbicides. Owen and Powles (2009) reported that the number of samples resistant to diclofop-methyl (16%) was higher than to clodinafop-propargyl (3%) in WA fields. Travlos et al. (2011) reported similar findings, with resistance to clodinafop-propargyl lower than to either fenoxaprop-P-ethyl (41%) or diclofop-methyl (64%). Beckie et al. (2013) screened wild oats to three fop herbicides (fenoxaprop-P-ethyl, clodinafop-propargyl and quizalofop-ethyl) but did not report results for the different herbicides.

The frequency of resistance to the fop herbicides in NSW Slopes and Plains has increased over the 6-year period since previously surveyed. Previously, 30% of samples from NSW Slopes (2007) (Broster et al. 2011a) and 19% from the NSW Plains (2010) (Broster et al. 2013) were resistant to diclofop-methyl compared with 49% of samples from the NSW Slopes resistant to clodinafop-propargyl in 2013 (P < 0.0001) and 31% from the NSW Plains in 2016 (P = 0.009) (Table 3).

Resistance to the ALS-inhibiting herbicides was lower than to ACCase-inhibiting herbicides in wild oats (Table 5). However, resistance was more common in SA (12% of samples) and Victoria (10% of samples) than in NSW (5% of samples; Table 3). These differences between states likely reflect different crop rotations and herbicides used. Owen and Powles (2016) identified one sample with resistance to sulfonylurea herbicides in 118 samples of wild oats tested from WA. Beckie et al. (2013), on the other hand, identified resistance to ALS-inhibiting herbicides in 12% of samples collected from the Canadian Prairies.

In contrast to wild oats, resistance to ACCase-inhibiting herbicides was less common in barley grass and brome grass (Table 4). Only three samples of brome grass were resistant to quizalofop-ethyl, all from Victoria Wimmera–Mallee (Table 5), and five samples of barley grass, all from SA (Table 6). There was no difference in the frequency of ACCase-inhibitor resistance between brome grass and barley grass (Table 4). Previous surveys in NSW and Tasmania identified no resistance to the ACCase-inhibiting herbicides in these two weed species (Broster et al. 2010, 2012b; Owen et al. 2015). Boutsalis et al. (2014) reported resistance to ACCase-inhibiting herbicides in brome grass from the Wimmera–Mallee at a similar frequency to the present survey and in 14% of samples from South East SA. A smaller survey of barley grass in the Eyre Peninsula and Mid North of SA identified resistance (>20% survival) to quizalofop-ethyl in 6 of the 90 barley grass samples screened (Shergill et al. 2015), similar to the frequency of resistance identified in this survey (Table 6). In a survey of brome grass and barley grass in WA, 1% of brome grass samples had survivors when treated with fluazifop-butyl and clethodim; however, all of the barley grass samples were completely controlled by ACCase-inhibiting herbicides (Owen et al. 2015).

Over 40% of samples from all three regions in SA where brome grass was collected showed resistance to sulfonylurea herbicides (Table 5) with resistance also occurring in >10% of the samples from Western NSW, Tasmania and the Wimmera–Mallee in Victoria. For barley grass, all regions except for the Victorian North East had samples resistant to mesosulfuron-methyl (Table 6), with two of the three regions in which it was collected in SA having >40% of samples resistant (Table 6). In both brome grass and barley grass, there was more resistance to sulfonylurea herbicides than to ACCase inhibitors, and there was more resistance to sulfonylurea herbicides than occurred in wild oats (Table 4). Previous surveys of brome grass (2008–12) in SA and Victoria reported that 2–45% of samples collected were resistant to mesosulfuron-methyl (Boutsalis et al. 2014). There was less resistance found to mesosulfuron-methyl in Victoria and more in SA in this survey. Previous surveys in Tasmania (Broster et al. 2012b) and SA (Shergill et al. 2015) identified no resistance to ALS-inhibiting herbicides in barley grass. However, Owen et al. (2015) reported 8% of barley grass samples collected in WA were resistant to sulfonylurea herbicides. There has been a rapid increase in resistance to ALS-inhibiting herbicides in barley grass in south-eastern Australia in recent years.

Across the surveyed regions, annual ryegrass was the most commonly collected grass weed species, collected from 81% of fields, and in no region was it collected from <60% of fields (Broster et al. 2022). By comparison, wild oats was collected from 38% of fields, brome grass 21% and barley grass 15%. There were differences among regions in the incidence of the different weed species, likely resulting from different environments and soil types present. Wild oats was most common in NSW, collected from 55% of fields, and less common in SA and Victoria, collected from 20% and 34% of fields, respectively (Table 2). On the other hand, brome grass was most common in SA, collected from 28% of fields, compared with only 15% of fields in NSW (Table 2).

Resistance in wild oats was more common in NSW than in the other states (Table 3), and likewise, resistance in brome grass was more common in SA than in NSW (Table 5). This suggests that resistance is being selected more often where there are higher populations of these grass weeds. This agrees with studies from Canada and Greece. In Canada, wild oats was the most common weed species, found in 67% of paddocks, with 44% of these resistant to a fop herbicide, whereas in Greece, wild oats was collected from ~20% of fields and the resistance level was lower at 28% of samples (Travlos et al. 2011; Beckie et al. 2013). If the initial frequency of resistance to herbicides is similar among the weed species, then lower populations make it less likely that resistance will evolve (Preston and Powles 2002; Kreiner et al. 2018).

This group of surveys of three major grass weed species across south-eastern Australia shows that these species are found across this region and that resistance occurs in all species. The extent of resistance differs among species, herbicides and regions, reflecting the importance of the weed in a region and/or the herbicides used for weed control in that region as has been shown in annual ryegrass (Broster et al. 2019). Despite the high levels of resistance in some regions, there are still herbicides that control all populations, allowing alternatives for control.

Data availability

The data that support this study are available in the article and upon request to the corresponding author.

Conflicts of interest

Christopher Preston is an Associate Editor of Crop & Pasture Science, but was blinded from the peer-review process for this paper. All other authors declare that they have no conflicts of interest.

Declaration of funding

This research was funded by Grains Research and Development Corporation (GRDC) grant number UCS00020. The GRDC had no involvement in study design, collection, analysis or interpretation of data, or in writing the manuscript.