Double Mutations Drive Multiple Resistances to Herbicides in Greek Rigid Ryegrass (Lolium rigidum Gaudin)

Abstract

Based on the complaints of malt barley growers about the insufficient control of rigid ryegrass (Lolium rigidum Gaudin) after applying the ACCase inhibitor pinoxaden, a survey was conducted during the early spring growing season of 2019/20; 20 barley fields located in Thessaloniki and 20 fields in Serres were marked with poor weed control levels. Before the barley harvest, representative weed seeds were collected from all 40 fields. After performing seed germination tests, fourteen populations (six from Thessaloniki and eight from Serres) with the highest seed germination ability were selected for further study. The whole-plant dose–response assays conducted in 2019–2020 indicated that most of the populations were multi-resistant to ACCase and ALS inhibitors. The estimated GR₅₀ values (the herbicide dose required to reduce the fresh weight of treated plants by 50%) for pinoxaden and mesosulfuron-methyl + iodosulfuron-methyl-sodium ranged from 1.15 to 52.41 g ai ha⁻¹ and 4.75 to 31.25 g ai ha⁻¹, respectively. Furthermore, the sequencing of acccase gene fragments from plants that survived pinoxaden application revealed that 11 out of 14 plant populations had a double accase point mutation at Ile1781 and Ile2041 codons. In addition, the sequencing of als gene fragments from the plants that survived mesosulfuron-methyl + iodosulfuron-methyl-sodium application revealed that 11 out of 14 plant populations had a point mutation at the Pro197 codon and 2 of them had a second als mutation at the Trp574 codon. These findings indicate that L. rigidum populations are multi-resistant to ACCase and ALS inhibitors, with individuals exhibiting either double accase or double als mutations.

1. Introduction

Malt barley (Hordeum vulgare subsp. distichum L.) is an important and promising crop in Greece due to its potential for growth in a contract farming system [1], which is attractive to producers as they can dispose of the entire quantity of their product at a pre-arranged price [2]. However, to ensure the sustainability and viability of this crop, farmers should produce high yields with excellent quality characteristics, including a grain protein content between 9.5% and 11.5% and more than 90% of the harvested grains must be larger than 2.5 mm [3]. In order to reach these goals, farmers rely on the effective management of weeds, as their presence causes significant reductions in barley yield and quality due to competition for nutrients, water, light, and space [4,5,6].

Rigid ryegrass (Lolium rigidum Gaudin) and sterile oat (Avena sterilis L.) are the most important grass weeds of malt barley grown in Greece [7,8,9]. L. rigidum is a diploid (2n = 14) and obligated, cross-pollinated, annual grass species [10] with a gametophytically self-incompatibility system [11]. It combines many biological characteristics such as prolific seed production, seed viability, pollen movement, a high degree of genetic variability, and high phenotypic plasticity, enabling it to adapt to adverse environmental conditions with rapidly evolved resistance to most herbicides in use [9,12]. It is highly distributed and abundant in both winter cereals and other crops [13,14]. It competes with barley mainly for nutrients, causing great yield losses ranging from 10% to 83% depending on the malt barley variety and L. rigidum density [15,16,17,18].

The control of this weed in winter cereals relies almost exclusively on post-emergence acetyl-coenzyme A carboxylase (ACCase) and acetolactate synthase (ALS) inhibitors [7,8,9,10]. However, the low efficacy of these herbicides has been reported in all Mediterranean countries during the last decade, which could be attributed to their intensive use, resulting in the evolution of many multi-resistant populations [11,12,16]. The difficulty in chemically controlling Lolium species is attributed to their capacity to rapidly evolve various herbicide-resistant mechanisms, such as seed viability, obligated cross-pollination, genetic variability, and high phenotypic plasticity [12]. In addition, due to their diploidy [10,11], they are capable of hybridizing with each other, resulting in individuals with multi-resistant alleles [19]; this reduces the option of using alternative herbicides for the integrated management of these weeds [20].

The resistance of L. rigidum involves target-site (TSR) and non-target-site (NTSR) mechanisms. TSR mechanisms include altering the target enzyme caused by point mutations or the over-expression of a specific gene encoding the target protein [21,22]. On the other hand, the decreased absorption and translocation of a plant’s herbicide, or even an enhancement of herbicide degradation via plant metabolism, is the outcome of NTSR mechanisms [22]. In some cases, TSR and NTSR have been reported to coexist either in the same individuals or in different plants of the same population [23]. TSR resistance in ACCase herbicides occurs due to a single point mutation in the carboxyl transferase (CT) domain of the ACCase enzyme. In particular, the substitutions of Ile 1781 to Leu, Val, or Thr, Trp 1999 to Cys, Leu, or Ser, Trp 2027 to Cys, Ile 2041 to Asn or Val, Asp 2078 to Gly, Cys 2088 to Arg, and Gly 2096 to Ser or Ala are mainly responsible for the resistance to ACCase inhibitors [24]. In addition, TSR resistance to ALS inhibitors occurs due to point mutations at the ALS codons of Ala 122, Pro 197, Ala 205, Asp 376, Arg 377, Trp 574, Ser 653, and Gly 654 [25].

Recently, a significant number of malt barley growers in Northern Greece have complained about the unsatisfactory control of L. rigidum, mainly after the application of pinoxaden (ACCase inhibitor) over the last three years; pinoxaden is the only effective post-emergence herbicide registered for the control of grass species grown in malt barley. Taking this information into consideration, this research aimed to provide data for the possible evolution of resistance in L. rigidum to ACCase inhibitor pinoxaden and possibly multi-resistance to ALS inhibitors. Therefore, the aim of this study was (1) to identify the most troublesome malt barley fields infested with L. rigidum, (2) to investigate whether the unsatisfactory control of 14 putative resistance populations collected from fields grown with malt barley in Northern Greece is due to resistance to ACCase pinoxaden and/or ALS inhibitors mesosulfuron-methyl + iodosulfuron-methyl, and (3) elucidate the possible presence of a point mutation in the als or/and accase genes.

2. Materials and Methods

2.1. Plant Material

This study was conducted in two winter cereal monoculture regions of Northern Greece, where the ALS-inhibiting herbicides chlorsulfuron, mesosulfuron-methyl + iodosulfuron-methyl-sodium, pyroxsulam, and the ACCase inhibitors diclofop-methyl, clodinafop-propargyl, fenoxaprop-p-ethyl, and pinoxaden had been continuously used for more than three decades to control grass weeds. This study was initiated after personal communication with malt barley growers in the Thessaloniki (40°37′45.3684″ N 22°56′50.6832″ E) and Serres (41°0.5′4.80″ N 23°32′35.39″ E) regions of Northern Greece, who complained about the unsatisfactory control of L. rigidum populations in recent years after the post-emergence application of the ACCase-inhibiting herbicide pinoxaden, which is the only graminicide registered for use in malt barley.

Based on the information collected from the farmers, a roadside survey [26,27] was conducted during the late spring growing season of 2019/20 to help identify 40 barley fields heavily infested with L. rigidum (20 per region) possibly due to the reduced efficacy of post-emergence-applied pinoxaden. Then, before the barley harvest, a representative bulked sample of seeds was collected by hand from 60 to 70 plants grown in different patches within each field. The pooled seeds from each field were considered different populations. It is worth noting that some of the collected seeds were mature, but immature seeds were also sampled in most populations. The collected seeds were then transferred to the Laboratory of Agronomy of the Agricultural University of Athens, where they were air-dried, threshed, placed in paper bags, and stored at room temperature (18–25 °C) for use in subsequent experiments.

Seeds from the 40 field populations were subjected to germination tests. Therefore, 50 seeds from each population were placed on wet filter paper in Petri dishes. There were three replication Petri dishes for each population, and all dishes were transferred to a temperature-controlled chamber at 25 °C for 5 days. Then, the number of germinated seeds in each Petri dish was counted, and the results were expressed as a percentage (%).

A separate ANOVA was performed for the seed germinability data obtained from each region, using a 20 (field populations) × 3 (replication Petri dishes), completely randomized design. Fisher’s protected LSD (least significant difference) test was used to compare the differences between means (p < 0.05) [28]. Based on the comparison of seed germinability means, 6 (AS3, AS6, AS9, AS10, AS16, AS19) out of 20 populations from Thessaloniki and 8 (N1, Ν4, N6, N11, N13, N15, N19, N20) out of the 20 populations originating from Serres were found to have the highest germination ability. In addition to the higher seed germination ability, 14 L. rigidum populations were selected for further study based on the following: the fields’ history of herbicide use (reported by malt barley growers after personal communication), poor control of L. rigidum in recent years after the post-emergence application of pinoxaden, and high rate of field infestation as recorded during the roadside survey.

2.2. Whole-Plant Dose–Response Assays for the 14 Putative Resistance Populations

A pot experiment was carried out on the farm of the Agricultural University of Athens in 2021 to evaluate the 14 collected L. rigidum populations for their possible resistance to the ACCase inhibitor pinoxaden (Axial 60 EC, Syngenta Hellas S.A., Anthousa, Greece) and/or the evolution of ALS-multi-resistance to mesosulfuron-methyl + iodosulfuron-methyl-sodium (Atlantis WG, Bayer Crop Science Hellas, Leverkusen, Germany). The experiment was carried out in plastic pots (1 L) filled with clay loam soil (35.9% sand, 29.8% clay, 34.3% silt, and 7.24 pH). In each pot, 6 L. rigidum pre-germinated seeds were sown at approximately 2 cm depth. When seedlings reached the three-leaf stage, they were carefully reduced to three per pot. Herbicides were applied when all plants were at the 22–24 Zadoks [29] growth stage (2–4 tillers). Both herbicides were applied at half, two, four, and sixteen times the recommended field label rate for pinoxaden (9, 18, 36, 72, 108 g ai ha⁻¹) and mesosulfuron-methyl + iodosulfuron-methyl (22.5, 45, 90, 180, 270 g ai ha⁻¹). Treatments were applied using a pressurized CO₂ AZO sprayer equipped with flat-fan nozzles spaced 50 cm apart on a 3 m boom calibrated to deliver 300 L ha⁻¹ of spray volume at 280 kPa pressure. After the herbicide applications, the plants were placed under field conditions (outdoor) and watered regularly until the final assessment (6 weeks after herbicide treatments). The obtained data indicated that the population AS19 could be effectively controlled using the recommended rate of both herbicides; thus, it was considered a susceptible population and used in two identical experiments conducted to evaluate the efficacy of the x/16, x/8, x/4, and x (recommended) rates of pinoxaden (1.1, 4.5, 9, 18 g ai ha⁻¹) and mesosulfuron-methyl + iodosulfuron-methyl-sodium (2.81, 11.25, 22.5, 45 g ai ha⁻¹). The establishment of the experiments and the herbicide applications were similar to the ones described above.

The experiment was performed twice using a completely randomized design with three replicated pots per herbicide treatment. Untreated control pots were also included. Six weeks after the herbicide treatments, the L. rigidum control was assessed by measuring the fresh weight (aboveground biomass) of all surviving herbicide-treated plants in each pot expressed as a percentage of the fresh weight of the untreated control plants. A combined ANOVA over the two experiments was performed to evaluate the resistance level of the 14 populations, using a 2 (experiments) × 14 (populations) × 2 (herbicides) × 5 (herbicide rates) × 3 (replication pots) split-plot approach, where the populations were the main plots and rate of two herbicides by five herbicides were the sub-plots. In addition, a combined ANOVA was performed over the two experiments to evaluate the susceptibility of the AS19 L. rigidum population using a 2 (experiments × 2 (herbicides) × 4 (herbicide rates) × 3 (replication pots) approach. Bartlett’s test determined the homogeneity of variances [28], which indicated that the departure of normality was not significant, thus allowing the data to be pooled and analyzed over the two experiments [28]. Fisher’s protected LSD (least significant difference) test was used to compare the differences between means at the significance level of p < 0.05. The ANOVA for all experiments was performed using the Statgraphics Centurion (version XVI) software package (Statpoint Technologies, Inc., Warrenton, VA, USA).

To determine the resistance level of the populations to pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium, the amount of herbicide required for 50% fresh weight reduction was estimated. This was calculated by analyzing the regression of L. rigidum fresh weight reduction (% of untreated control) data in the survived plants against pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium rates using a four-parameter log-logistics model (1) [30] as follows:

y = c + {[d − c]/1 + exp[b(log(x) − log(GR₅₀))]}

(1)

where y represents fresh weight as a percentage of the untreated control; x is the herbicide rate (where g ai ha⁻¹ + 0.1 was added because x is equal to 0); d and c represent the upper and lower values of y; and b is the slope of the curve near GR₅₀, which is the rate required to halve the fresh weight relative to d. The level of population resistance for each herbicide was determined by the resistance ratio (R/S) value, calculated as the respective GR₅₀ of the R population divided by the respective value of the S population (AS19).

2.3. Amplification and Sequencing of Als and Accase Gene Fragments

The accase and als gene fragments, encompassing potential ACCase (Ile1781, Ile 2041) or ALS (Pro197, Trp574) mutation sites, are known to confer resistance to ACCase- and/or ALS-inhibiting herbicides, respectively; they were amplified and sequenced in plant samples taken from three surviving individuals of each population six weeks after the application of the recommended (few samples) or two times the recommended field label rate of pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium herbicides used in whole-plant dose–response assays. All samples were stored at 20 °C for DNA extraction. Genomic DNA was extracted from 200 mg of ground leaf tissue using the cetyl trimethylammonium bromide (CTAB) method according to the protocol of Doyle and Doyle [31].

The amplification of the accase fragment containing the codon Ile-1781 (1000 bp) was conducted using the forward primer 5′-GCTCAATGACATTGGTATGGTAGCCTGG-3′ and the reverse primer 5′-TCCAGTTAGCAAGGATGAACAGAGG-3′ [11]. The PCR consisted of 0.5 μM of each forward and reverse primer; 0.2 mM dNTPs; 2 μL of the supplied 10× KAPA Taq PCR buffer; 1 μL genomic DNA; and 0.5 units of a standard KAPA Taq polymerase in 20 μL mixture. The amplifications are described above except for 20 s annealing at 61 °C and 1 min elongation at 72 °C. A final elongation step was performed at 72 °C for 1 min. Regarding the amplification of the accase gene fragment containing the codon Ile-2041 (500 bp), this was achieved using the forward primer 5′-AGGACAGCCTGATTCCCATGAG-3′ and the reverse 5′-GCTTCTATGCTCTTCTGAATGG-3 [11]. The PCR consisted of 0.5 μM of each forward and reverse primer, 0.2 mM deoxyribonucleotide triphosphate (dNTPs); 2 μL of the supplied 10× KAPA Taq PCR buffer; 1 μL genomic DNA; and 0.5 units of standard KAPA Taq polymerase in 20 μL volume. Amplification was performed as follows: 95 °C for 3 min, 35 cycles at 95 °C for 30 sec, 51 °C for 20 sec, and 72 °C for 30 s, and a final elongation step at 72 °C for 5 min.

The amplification of the als gene fragment containing the Pro-197 codon (400 bp) was conducted using the forward 5′-GCCACCAACCTCGTCTCC-3′ and the reverse 5′-CCACCGCCAACATARAGAAT-3′ primers [11]. The polymerase chain reaction (PCR) consisted of 0.5 μM of each forward and reverse primer; 0.2 mM deoxyribonucleotide triphosphate (dNTPs); 2 μL of the supplied 10× KAPA Taq PCR buffer; 2 μL genomic DNA; and 0.5 units of a standard KAPA Taq polymerase in 20 μL mixture. Amplifications were carried out in a SureCycler 8800 thermocycler (Agilent Technologies, Santa Clara, CA, USA) using the following cycles: DNA denaturation for 3 min at 95 °C and 35 cycles of 30 s denaturation at 95 °C; 30 s annealing at 61 °C; and 1 min elongation at 72 °C. A final elongation step was performed at 72 °C for 5 min. Regarding the amplification of the als gene fragment containing the codon Trp-574 (1396 bp), this was conducted using the forward primer 5′-ACTCCATCCCCATGGTGGC-3′ and the reverse primer 5′-ATAGGCAGCACATGCTCCTG-3′ [11]. The PCR consisted of 0.5 μM of each forward and reverse primer; 0.2 mM deoxyribonucleotide triphosphate (dNTPs); 2 μL of the supplied 10× KAPA Taq PCR buffer; 2 μL genomic DNA; and 0.5 units of standard KAPA Taq polymerase at a 20 μL volume. The PCR was carried out in a SureCycler 8800 thermocycler (Agilent Technologies, Santa Clara, CA, USA) using the following cycles: 95 °C for 3 min, 35 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 120 s, with a final extension step at 72 °C for 8 min.

All PCR products were separated in 1.5% agarose gel and purified according to the protocol outlined in the NucleoSpin^® Extract II kit (MACHEREY NAGEL GmbH & Co. KG. Postfach 10 13 52. D-52313 Düren, Germany). The purified products were sent immediately for sequencing to the CeMia Center in Larissa. Each PCR product was sequenced once using the forward primer, except for Trp574, which was sequenced with both forward and reverse primers due to its size. The sequences were aligned using the CLUSTAL W algorithm and the MEGA version X software [32].

3. Results

3.1. Lolium rigidum Seed Germination Ability of 40 Populations

The germination ability of the collected seeds from the Thessaloniki region indicated significant differences between the populations (F = 43.73, p = 0.01). Specifically, 57%, 60%, 62%, 66%, 42%, and 59% of the collected seeds were germinated from AS3, AS6, AS9, AS16, AS19, and AS20 plant populations, respectively (Figure 1). In addition, the respective germination rates of the selected seeds from Serres were significantly affected by the studied populations (F = 2.44, p = 0.01) at 70%, 86%, 86%, 46%, 60%, 50%, 72%, and 46% for the N1, N4, N6, N11, N13, N15, N19, and N20 populations, respectively.

3.2. Whole-Plant Dose–Response Assays

The ANOVA of the whole-plant dose–response experiments indicated that fresh weight reductions (FWRs) in the 14 populations were significantly affected by population, herbicide, herbicide rates, and their interaction. Therefore, to determine the fresh weight reduction in each population using the recommended rate of either pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium, the estimated GR₅₀ and R/S values were presented and the fresh weight reduction means were compared using the LSD test at a p < 0.05 level of significance (

Table 1. The fresh weight reduction (FWR, % of untreated control) using the recommended rate of pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium, GR₅₀ and R/S values of the 14 L. rigidum populations.

Population	Pinoxaden			Mesosuluron-Methyl + Iodosulfuron-Methyl-Sodium
	FWR 1	GR50 2	R/S 3	FWR 1	GR50 2	R/S 3
AS3 5	70 bcdf 4	-	-	54 bcde 4	17.13	3.6
AS6	61 defg	-	-	69 ab	7.2	1.51
AS9	62 cdefg	42.40	37.24	60 bcde	17.68	3.72
AS10	44 g	44.12	38.7	41 de	29.04	6.11
AS16	76 abcd	-	-	63 bcd	11.26	2.37
AS19	93 a	1.15	-	91 a	4.75	-
N1	71 bcde	-	-	66 b	17.28	3.63
N4	72 bcde	-	-	59 bcde	13.31	2.8
N6	81 ab	-	-	71 ab	9.12	1.92
N11	52 fg	37.30	32.71	91 a	-	-
N13	77 abcd	-	-	55 bcde	13.37	2.8
N15	49 g	52.41	45.97	37 e	24.26	5.09
N19	79 abc	-	-	64 bc	9.66	2.03
N20	57 egf	33.51	29.39	42 cde	25.96	5.32

LSD_0.05 = 6 to compare the fresh weight reduction in the population by means of the recommended rate of each herbicide at p = 0.05. ¹ The fresh weight reduction in each population via the recommended rate of each herbicide. ² The herbicide rate (g ai ha⁻¹) that causes 50% fresh weight reduction in each population. ³ R/S = GR₅₀ (R population)/GR₅₀ (S, AS6). ⁴ Values within the same herbicide column followed by the same letter are not different using Student’s multiple range test at a = 0.05. ⁵ Samples of populations AS and N were collected from Thessaloniki and Serres, respectively.

). In general, the fresh weight of most populations, averaging over five rates for each herbicide, was reduced more for pinoxaden than mesosulfuron-methyl + iodosulfuron-methyl-sodium. Specifically, the fresh weight of seven populations was reduced by 80–92% using pinoxaden, whereas the respective reduction in the seven other populations ranged from 61% to 76%. On the other hand, mesosulfuron-methyl + iodosulfuron-methyl-sodium exhibited a reduced fresh weight in four populations by 75–96% when averaged over five rates; the respective reduction in the remaining ten populations ranged from 57% to 72%. Regarding the populations using the recommended rate of each herbicide, pinoxaden reduced the fresh weight of AS19 and Ν6 populations by 93% and 81%, respectively, and the reduction in the remaining five AS and seven N populations ranged from 44% to 62% and 70% to 79% (Table 1). However, mesosulfuron-methyl + iodosulfuron-methyl-sodium reduced the fresh weight of AS19 and Ν11 populations by 91% and 91%, respectively, and the reductions in the remaining two groups of six populations ranged from 37% to 59% and 60% to 71% (Table 1). The GR₅₀ values of pinoxaden were estimated only for the S and five R populations as the fresh weight reduction in the other eight populations using the lowest rate of pinoxaden (half of the recommended rate) was greater than 50%. This was not the case for mesosulfuron-methyl + iodosulfuron-methyl-sodium, where GR₅₀ values were estimated for 13 populations. Based on the estimated GR₅₀ and R/S values, pinoxaden showed higher GR₅₀ (33.51–52.41 g ai ha⁻¹) and R/S (29.39–45.97) values than the respective GR₅₀ (7.2–29.04 g ai ha⁻¹) and R/S (1.51–6.11) values for mesosulfuron-methyl + iodosulfuron-methyl-sodium (Table 1).

3.3. Amplification and Sequencing of Accace and Als Gene Fragments

The amplification and sequencing of the accase gene fragment from plants surviving the recommended use of pinoxaden or two times the recommended application rate revealed that 11 out of 14 plant populations had either Ile1781 or Ile2041 or double accase point mutations at Ile1781 and Ile2041 codons (Figure 2,

Table 2. ACCase and ALS mutations detected in resistant Lolium rigidum plants from 14 populations grown in malt barley fields of Northern Greece.

Population	ACCase Mutations			ALS Mutations
	Ile-1781	Ile-2041	Mutant Plants/ Analyzed Plants	Pro-197	Trp-574	Mutant Plants/ Analyzed Plants
AS3	Leu(3)	Asn(3)	3/3	Pro(3)	Trp(3)	0/3
AS6	Leu(3)	Asn(1), Ile(2)	3/3	Pro(2), Thr(1)	Trp(3)	1/3
AS9	Leu(3)	Ile(3)	3/3	Pro(3)	Trp(3)	0/3
AS10	Leu (3)	Asn(1), Ile(2)	3/3	Thr(1), Ser(2)	Leu(3)	3/3
AS16	Leu(1) or Val(1) Leu(1), Ile(1)	Ile(3)	2/3	Gln(1), Ala(2)	Trp(3)	3/3
AS19	Ile(3)	Asn(1), Ile(2)	1/3	Ser(1), (2)Pro	Trp(3)	1/3
N1	Leu(3)	Asn(1), Ile(2)	3/3	Ser(3)	Trp(3)	3/3
N4	Leu(2), Ile(1)	Asn(1), Ile(2)	2/3	Ser or Thr, Pro(2)	Trp(3)	1/3
N6	Leu(3)	Asn(1), Ile(1) Val(1)	3/3	Pro(3)	Trp(3)	0/3
N11	Leu(3)	Asn(3)	3/3	Ser(3)	Trp(3)	3/3
N13	Leu(1), Ile(1) Val or Leu	Asn(2), Val(1)	3/3	Ser(3)	Trp(3)	3/3
N15	Leu(2), Ile(1)	Ile(3)	2/3	Ser(2), Pro(1)	Trp(3)	2/3
N19	Leu(3)	Asn(1), Ile(2)	3/3	Thr(2), Leu(1)	Leu(3)	3/3
N20	Leu(3)	Asn(1), Ile(2)	3/3	Thr(1), Pro(2)	Trp(3)	1/3

Amino acid substitution detected in two codons of accase or als genes of analyzed plants. Number of plants carrying the mutation are in brackets.

). In particular, 35 out of the 42 sequenced plants from 14 L. rigidum populations had a point mutation (MTA, YTA, WTA, TTA, CTA, or RTA) at the first base of the codon Ile(ATA)-1781 that resulted in the substitution of Ile-1781 by Leu compared to the coding chloroplastic ACCase sequence of the L. rigidum reference (DQ184647.1 isolate WLR96-S) (Table 2). Based on the sequence chromatograms, most point mutations appear in a heterozygous state, validated by the presence of double peaks (Figure 2). However, one plant each from the AS16 and N13 populations with an RTA mutation showed the substitution of Ile1781 by Leu or Val; three plants from the AS19 population and one plant each from the AS16, N4, N13, and N15 populations were homozygous for the wild-type allele (ATA) at position Ile-1781. Regarding the Ile(ATT)-2041 codon, 16 out of the 42 sequenced plants from 11 populations had a point mutation at the second base of the codon (AWT) that resulted in the substitution of Ile-2041 by Asn (Table 2); 1 plant from the N6 and N13 populations had a point mutation at the first base of the codon (RTT) that resulted in the substitution of Ile-2041 by Val. Most of them showed a heterozygous state, as revealed by the presence of double peaks in the sequence chromatograms (Figure 2).

The amplification and sequencing of the als gene fragment from plants that survived the mesosulfuron-methyl + iodosulfuron-methyl-sodium used at the recommended or two times the application rate revealed a point mutation at the Pro197 codon in 11 out of 14 populations; 2 of them also had a second als mutation at the Trp574 codon (Figure 2, Table 2). More specifically, 22 out of the 42 sequenced plants of the 14 L. rigidum populations had a heterozygous point mutation at the first base of the Pro(CCG)-197 codon (TCG, MCG, SCG, YCG, WCG), as shown by the presence of double peaks (Figure 2). This resulted in the substitution of Pro-197 by Ser (15 plants), Thr (6 plants), and Ala (2 plants) compared to the coding chloroplastic ALS sequence of the L. rigidum reference (GenBank accession number EF411170) (Table 2). However, one plant from the AS16 and one plant from the N19 populations had a point mutation at the second base (CAG, CTG) of the Pro(CCG)-197 codon that resulted in the substitution of Pro-197 by Gln or Leu. Conversely, 18 sequenced plants of eight populations were homozygous for the wild-type allele (CCG) at position Pro-197I. Regarding the Trp574 codon (TCG), three sequenced plants in each of the AS10 and N19 populations had a point mutation in the heterozygous state in addition to the Pro-197 mutation at the second base of the codon (TKG), resulting in the substitution of Trp-574 by Leu (Table 2).

4. Discussion

The estimated significant seed germination differences between the populations and regions could be explained by the differences in maturity (immature embryo), dormancy (presence or absence of phytohormones, germination inhibitors, germination enhancers-inducers), genetic background, pleiotropic effect associated with the herbicide resistance of collected seeds, environmental conditions, and agricultural practices applied. This means that a possible synergism or antagonism between these factors could have a different effect on seed germination ability, emergence under field conditions, fitness, and, consequently, the evolution of weed presence within barley crops. These differences in seed germination ability under controlled conditions may also result from variations in seed banks, soil temperature, diurnal temperature variation, soil moisture, light, the concentration of nitrates in the soil, soil pH, and gaseous environments in the soil [12]. As reported earlier, the selection of the 14 L. rigidum populations for further study was based both on their higher seed germination ability and the herbicide use history in the fields studied, including the unsatisfactory control of L. rigidum populations in recent years after the post-emergence application of pinoxaden, and the high field infestation rates with L. rigidum, as recorded during the roadside survey conducted.

The various numbers of individual plants from the 14 L. rigidum populations that survived the application of ACCase-inhibiting pinoxaden and ALS-inhibiting mesosulfuron-methyl + iodosulfuron-methyl-sodium clearly indicate that these populations have evolved different levels of resistance to these herbicides. Specifically, the AS19 and Ν6 populations were considered susceptible to pinoxaden as its recommended rate reduced their fresh weight by 93% and 81%, respectively, and the reduction in the remaining populations ranged from 44% to 79%. The AS19 and Ν11 populations were also considered susceptible to mesosulfuron-methyl + iodosulfuron-methyl-sodium because their recommended rate reduced their fresh weight by 91% and 91%, respectively, and the reductions in the remaining populations ranged from 37% to 71%. These results indicate that 11 out of the 14 populations studied were multi-resistant to pinoxaden and mesosulfuron-methyl + iodosulfuron-methyl-sodium; one population (AS19) was susceptible to pinoxaden and mesosulfuron-methyl + iodosulfuron-methyl-sodium, while Ν6 and Ν11 were resistant to mesosulfuron-methyl + iodosulfuron-methyl-sodium and pinoxaden, respectively. The higher GR₅₀ and R/S values estimated for pinoxaden than mesosulfuron-methyl + iodosulfuron-methyl-sodium could be attributed to the greater selection pressure of pinoxaden due to its continuous use as the only registered grass herbicide for post-emergence application in malt barley fields. The few surviving plants (7–19%) of the considered susceptible populations after the application of either herbicide’s recommended rates provide evidence of the possible beginnings of weed resistance evolution. These findings could be attributed to the fact that the populations studied in this work were selected due to their suspected resistance to pinoxaden, which was not confirmed for the AS19 and Ν6 populations, suggesting that pinoxaden was applied at lower rates and/or at an inappropriate growth stage of L. rigidum.

The multiple resistance results obtained from the whole-plant dose–response assays of this study agree with other findings [8,9,12,13,33,34,35]. The slightly different GR₅₀ and R/S values for pinoxaden and mesosulfuron-methyl + iodosulfuron-methyl-sodium compared to those reported by others [14] can be explained by considering the different levels and mechanisms of herbicide resistance, environmental conditions, and procedures used for the experiments.

The amino acid substitution of Ile-1781 by Leu in 35 plants or by Val in 2 out of the 42 sequenced plants that survived pinoxaden treatment in 14 L. rigidum populations confirms the results obtained from the whole-plant dose–response assays and strongly supports the evidence of resistance to this herbicide. In addition, the amino acid substitution of Ile-2041 by Asn in 16 out of 42 sequenced plants of 11 populations, along with 1 plant from the N6 and N13 populations that had the substitution of Ile-2041 by Val, clearly indicated that the resistance of these plants to pinoxaden is based on the presence of a double mutation in the accase gene. The amino acid substitution of Pro-197 by Ser (15 plants), Thr (6 plants), Ala (2 plants), Gln (1 plant), and Leu (1 plant) in 22 out of the 42 sequenced plants that survived after mesosulfuron-methyl + iodosulfuron-methyl-sodium application in 14 L. rigidum populations also confirms the results obtained from the whole-plant dose–response assays and strongly supports the evidence of weed resistance to these herbicides. Also, the fact that three plants each of the AS10 and N19 populations had the substitution of Trp-574 by Leu clearly shows that the resistance of these populations to mesosulfuron-methyl + iodosulfuron-methyl-sodium results from the presence of a double mutation in the als gene. It is worth mentioning that most ACCase- or ALS-resistant plants showed a great heterogeneity within the populations and among regions, which is in agreement with the results reported by others [16]. The fact that 7 and 18 surviving plants after the application of recommended rates of pinoxaden or mesosulfuron-methyl + iodosulfuron-methyl-sodium did not reveal any point mutation at Ile-1781 or Pro-197 in the whole-plant dose–response assays could be attributed to different and non-studied mutations in the accase and als genes or to NTSR mechanisms. Regarding NTSR, mechanisms have been identified, such as reduced herbicide absorption due to alterations in leaf characteristics (high epicuticular wax density), vacuolar sequestration that keeps the herbicide away from the target site of action, altered translocation (altered membrane transporter activity and foliar hypersensitivity), and enhanced herbicide metabolism [20,34,36]. Enhanced metabolic resistance is considered the most common mechanism of NTSR in grass weeds, which coexists in some cases with target-site resistance to herbicides and can confer cross-resistance to herbicides with the same mode of action and/or multiple resistance to herbicides with different modes of action [20,21]. This type of resistance has already been confirmed in L. rigidum populations from Morocco and Tunisia, which were found to be cross- or multi-resistant to ALS- and ACCase-inhibiting herbicides due to coexisting TSR and NTSR mechanisms of action [37]. In addition, Scarabel et al. [14] reported that the mechanisms underlying resistance are rather complex and diverse among Lolium spp. populations from Italy, Denmark, and Greece, demonstrating how the coevolution of both target-site resistance and metabolic-based herbicide resistance is a common feature in Denmark and Italy. Furthermore, metabolic resistance has also been reported by many researchers, especially after the application of the ALS-inhibiting herbicides chlorsulfuron [38,39,40] and iodosulfuron [41].

The above-identified mutations in one or two accase and als codons of L. rigidum plants strongly support the evidence of widespread multi-resistance to ACCase and ALS herbicides and are in agreement with the results reported by others [9]; such studies found plants of L. rigidum populations originating from Northern Greece with coexisting point mutations in the accase (Ile-2041 substitution by Asn or Thr) and als (Pro-197 substitution by Leu, Glu, Ser, Ala, Thr, or Gln) genes. Others studying Lolium spp. populations collected in Denmark, Greece, Italy [14], Morocco, and Tunisia [37] also identified different ACCase (Leu1781, Cys2027, Asn2041, Val2041, Gly2078, Arg2088, Ala2096) and ALS-mutant alleles (Gly122, Ala197, Gln197, Leu197, Ser197, Thr197; Val205, Asn376, Glu376, Leu574) endowing multi-resistance to ACCase (clodinafop-propargyl and pinoxaden) and ALS (mesosulfuron-methyl + iodosulfuron-methyl-sodium) inhibitors. In addition, Tavassoli et al. [42] found four L. rigidum populations in Iran with multi-resistance to the ACCase inhibitors clodinafop-propargyl and haloxyfop-methyl as well as the ALS inhibitor mesosulfuron-methyl + iodosulfuron-methyl-sodium; however, they were susceptible to sethoxydim, pinoxaden, and isoproturon+ diflufenican. The resistance in these populations resulted from the Ile-1781-Leu and Pro-197-Ser substitutions in accase- and als-encoding genes, respectively. The Ile-2041-Asn and Ile-2041-Val mutations were also reported to occur in L. rigidum individuals [43,44], and the Pro-197 substitutions by Ala, Gln, Leu, or Ser were found in resistant L. rigidum populations originating from Northern Greece [8]. Furthermore, the substitution of Pro197 by His or Leu has been reported in L. perenne populations originating from Northern Greece [35]. In addition, a study on 19 populations originating from Tunisia and Morocco indicated that 13 populations were resistant to ALS-inhibiting herbicides due to point mutations in positions Pro-197-Thr, Pro-197-Ser, Pro-197-Leu, Pro-197-Gln, and Trp-574-Leu; resistance was also detected in 18 populations to ACCase-inhibiting herbicides as a result of a point mutation in positions Asp-2078-Val, Trp-2027-Cys, Ile-1781-Leu, Gly-2096-Ala, and Ile-2041-Asn of the enzymes conferring TSR Ile2041 [37]. Furthermore, Vázquez-García et al. [41] reported that the accumulation of target-site mutations in L. rigidum and L. multiflorum populations collected from barley and wheat fields in Chile confer multi-target-site resistance to ALS-, ACCase-, and EPSPS-inhibiting herbicides. Specifically, the sequencing of target-site genes revealed that their resistance resulted from the Pro-106-Ser/Ala (EPSPS), Ile-2041-AsnCCAsp-2078-Gly (ACCase), and Pro-197-Ser/Gln + Trp-574-Leu (ALS)Trp mutations.

The detected ACCase alleles (1781-Leu/2041-Asn and 1781-Leu/2041-Val) and ALS alleles (Ala197, Leu197, Ser197, Thr197, Leu574) in the populations studied agree with the results of Scarabel et al. [14,45], who detected different mutant ACCase alleles (Leu1781, Cys2027, Asn2041, Val2041, Gly2078, Arg2088, Ala2096) and ALS alleles (Gly122, Ala197, Gln197, Leu197, Ser197, Thr197, Val205, Asn376, Glu376, Leu574) endowing multi-resistance in Greek and Italian populations [14], as well as seven different mutant ACCase alleles (1781-Leu, 1999-Leu, 2041-Asn, 2041-Val, 2078-Gly, 2088-Arg and 2096-Ala) and 13 combinations with two types of mutations in the pinoxaden-resistant plants of Lolium spp. [45]. These findings could be attributed to the fact that Lolium is an obligatory cross-pollinated species, and pollen-mediated gene flow can occur over a distance of at least 3 km [46]. The recorded dominance of 1781-Leu and 2041-Asn in pinoxaden-resistant plants in our work is also in agreement with the results reported by others [45]; 1781-Leu was found to be the dominant ACCase mutant allele in five Lolium populations grown in Italian durum wheat fields, conferring resistance to pinoxaden, clodinafop, haloxyfop, sethoxydim and clethodim, and the 2041-Asn allele was found to confer dominant resistance to clodinafop-propargyl and haloxyfop, no resistance to sethoxydim and clethodim, and moderate resistance to pinoxaden. In addition, the 2041-Val allele was associated with resistance to haloxyfop, partial resistance to clodinafop-propargyl, no resistance to sethoxydim, and a likely moderate resistance to pinoxaden. The two detected types of alleles (197-Ser/574-Leu, 197-Thr/574-Leu, 197-Leu/574-Leu) in the same mesosulfuron-methyl + iodosulfuron-methyl-sodium-resistant plant in this work agree with the results reported by Lu et al. [47], who found double als gene mutations (Pro197-Ser plus Trp574-Leu) in shepherd’s-purse (Capsella bursa-pastoris) plants conferring a high level of cross-resistance to the ALS-inhibiting herbicides mesosulfuron-methyl, tribenuron-methyl, bensulfuron-methyl, and penoxsulam.

The fact that the control of this weed in winter cereals relies almost exclusively on post-emergence ACCase and ALS inhibitors [7,8,9,10], along with their low-determined efficacy against most of the populations studied due to the evolution of multi-resistance, increases the complexity of managing resistant L. rigidum in cereal fields. It is worth mentioning that among the alternative chemical options, the soil-applied herbicides chlorotoluron + diflufenican, flufenacet + diflufenican, flufenacet + diflufenican + metribuzin, and prosulfocarb were found to be promising tools for the effective management of susceptible and ACCase/ALS-resistant L. rigidum populations grown in Greece [48].

5. Conclusions

The whole-plant dose–response assays and sequencing of the accase and als genes confirmed the occurrence of L. rigidum populations in malt barley fields with multiple target sites and mediated resistance to ACCase (pinoxaden) and ALS (mesosulfuron-methyl + iodosulfuron-methy-sodium) inhibitors. Differences were highlighted in the patterns and levels of resistance among populations and between regions. Therefore, these findings justify the complaints made by farmers about the insufficient control of L. rigidum when using pinoxaden and highlight the widespread evolution of multi-resistance among L. rigidum infesting winter cereal monoculture fields in Northern Greece. Based on these findings, measures should be taken to reduce the further use of these herbicides as they were not effective against most L. rigidum populations present in the winter cereal fields in the two studied regions.