1. Introduction
The ability of perennial ryegrass (
Lolium perenne (L.) ssp.
perenne), Italian ryegrass (
L. multiflorum Lam. or
L. perenne (L.) ssp.
multiflorum (Lam.) Husnot), and rigid ryegrass (
L. rigidum L.) to rapidly adapt in agricultural and non-agricultural areas has gained particular interest from weed scientists [
1]. These species are consistently included among the most serious weeds of winter cereal and perennial crops, such as orchards, olive groves, vineyards, and alfalfa (
Medicago sativa L.) [
2,
3,
4]. According to their distribution, they are native to Europe, temperate Asia, and North Africa, but their presence has also spread over the past two centuries to southern parts of Africa, Australia, South America, New Zealand, and North America [
5,
6,
7].
Ryegrass species have a diploid (2
n = 2
x = 14) chromosome number and are characterized by self-incompatibility, rendering them obligate out-crossers. Due to the free cross-pollination occurring between
L. perenne and
L. multiflorum, numerous hybrids with intermediate characteristics have been naturally developed, which are characterized by high genetic variability, adaptability in agricultural landscapes, high phenotypic plasticity, and prolific seed production [
7,
8,
9].
The control of various ryegrass species in winter cereal crops relies mainly on herbicides inhibiting the acetolactate synthase (ALS) and acetyl-CoA carboxylase (ACCase) enzyme activity, applied either alone or in mixtures. Unfortunately, these two herbicide groups are associated with the highest risk of the rapid evolution of target-site resistance, which is most commonly due to a single point mutation in the
ALS or
ACCase gene that results in the substitution of an amino acid in the ALS or ACCase target enzyme [
10,
11,
12]. The herbicide resistance genes of ryegrass species spread easily via wind-mediated pollen movement due to their self-incompatible and obligate out-crossing traits [
13,
14]. In addition, mutant alleles conferring resistance to ryegrass are generally not linked and, therefore, allow different combinations of resistance mechanisms to accumulate independently [
11]. Consequently, multiple herbicide-resistant ryegrass plants and populations evolve very rapidly, exhibiting complex herbicide resistance patterns, due to both target site-based resistance (TSR) and non-target site resistance (NTSR) mechanisms [
15,
16,
17]. The mechanisms conferring TSR include an alteration of the herbicide target enzyme caused by point mutations within the coding gene or an overproduction of the target enzyme [
18,
19,
20]. On the other hand, NTSR refers to all other mechanisms evolved within a weed that result in reduced herbicide activity, such as reduced uptake or translocation, increased sequestration, or enhanced metabolism [
18,
21,
22]. Ultimately, the evolution of weeds with multiple herbicide resistance due to the accumulation of various resistance mechanisms reduces the options for herbicide rotation with different modes of action to manage these weeds [
11].
Target site-mediated resistance in
L. perenne species has already evolved to ALS inhibitors in France [
23] and in California [
24], whereas populations from Chile were found to have TSR to ACCase and EPSPS, and NTSR to the ALS inhibitor iodosulfuron [
11]. In addition, an
L. perenne population in Argentina evolved cross-resistance to pinoxaden, clethodim, and quizalofop due to the Asp-2078-Gly mutation [
25], whereas
Lolium ssp. (
L. perenne,
L. multiflorum, and their hybrids) populations from this country were found to have multiple resistance to pinoxaden and iodosulfuron+mesosulfuron, which was mainly due to cytochrome P450-mediated herbicide metabolism [
26]. Also, three
L. perenne populations in Texas were found to be multiple resistant to diclofop-methyl and mesosulfuron [
27], whereas an
L. perenne population from New Zealand was found to have cross-resistance to pinoxaden and quizalofop-p-ethyl due to both the target-site mechanism (isoleucine to valine replacement at position 2041 of the
ACCase gene) and cytochrome P450-based metabolism [
28]. Moreover, many
L. perenne populations in southwestern North Carolina evolved cross-resistance to diclofop and pinoxaden, five of which were multiple-resistant to diclofop, pinoxaden, and mesosulfuron, and two were cross-resistant to imazamox, mesosulfuron, and pyroxsulam [
29]. Finally,
L. perenne populations with multiple resistance to both ACCase and ALS inhibitors were reported in cereal crops grown in Germany and Denmark, whereas populations of this weed in Portugal, New Zealand, and Argentina were found to be resistant to the EPSPS inhibitor glyphosate [
10,
30].
Plant traits (e.g., seed germination, seed production, early vigor, tillering, growth rate, biomass production, and competitive ability) linked to herbicide-resistant alleles are key players in the evolution of adaptive alleles fixation, but the rate of this process depends on their opposing cost and benefit effects [
31]. More specifically, the benefits of the resistance alleles are clear and usually appear very fast, as they allow plants to survive and reproduce under herbicide selection, while the cost effects of the resistance alleles on the surviving resistant weed plants are unclear and usually appear slowly because of their reduced adaptive and establishing potential compared to crops and other weed plants. Regarding fitness/adaptation penalty associated with herbicide resistance alleles, Vila-Aiub et al. [
32] reported that these are evident among plant species, but their expression depends on particular herbicide resistance gene and allele, the genetic background, the dominance of the fitness cost, and the abiotic and biotic environmental conditions.
Lolium perenne is less abundant than L. rigidum in winter cereal crops in Greece. However, some cereal farmers in northwestern Macedonia, Greece, observed poor control of this weed after spraying their crops with ALS-inhibiting herbicides during the 2020 growing season. This information prompted the current study, with objectives to (1) investigate whether three L. perenne populations from the study area have evolved resistance to ALS-inhibiting herbicides, (2) evaluate the potential preemergence and postemergence herbicides as alternatives for the management of these populations, (3) elucidate the possible target site-mediated resistance mechanism, (4) assess the growth rate (aboveground biomass and tiller number) of an R and a reference S L. perenne population without competition, and (5) compare their competitive ability against wheat.
4. Discussion
The fact that the recommended two-fold and four-fold rates of chlorsulfuron reduced fresh weight of the R1, R2, and R3 populations by 0–4%, 71–84%, and 86–98%, respectively, indicates that the R1 population has evolved resistance to chlorsulfuron, while R2 and R3 are moderately resistant and susceptible to this herbicide, respectively. However, the excellent control of these populations with chlorotoluron + diflufenican or prosulfocarb suggests their use as alternative chemical options to manage this weed [
10]. Regarding the S population, the above herbicides showed good to excellent efficacy, even after the application of lower than their recommended field label rates. It is worth noting that, regardless of the high efficacy of the soil-applied chlorotoluron + diflufenican or prosulfocarb against this weed, their use in Greece is very limited because their efficacy against weeds and selectivity in winter cereals are dependent on weed species, crop cultivars, soil, and weather conditions.
The 38–67% and 32–48% fresh weight reduction in the R1 population by the recommended and two-fold rates of mesosulfuron + iodosulfuron and pyroxsulam supports the evidence that this population has evolved cross-resistance to these ALS-inhibiting herbicides, which makes their further use in some cases unsatisfactory for weed management. However, the fact that imazamox provided excellent control of this weed justifies its possible use in rotational Clearfield® crops (tolerant to imidazolinone herbicides). Also, the excellent control of the susceptible population with half of the recommended rate of the above herbicides indicates that some susceptible populations of this weed can be controlled at lower than their recommended rates.
The R1 population fresh weight was reduced by 99–100% with the recommended field label rate of the ACCase-inhibiting herbicides clodinafop-propargyl, diclofop-methyl, pinoxaden, and clethodim, which indicates that this population with cross-resistance to ALS-inhibiting herbicides did not develop multiple resistance to ACCase inhibitors. In addition, the fact that one-fourth, half, or the recommended rate of the same herbicides provided 90–100% control of the S population shows that some S populations of this weed can be controlled at lower than their recommended rates.
The R1
L. perenne population evolution of cross-resistance to ALS-inhibiting herbicides could be attributed to overreliance on the preemergence application of chlorsulfuron or the postemergence application of mesosulfuron + iodosulfuron or pyroxsulam for many consecutive years. The high frequency of the ALS herbicide-resistant individuals occurring naturally in weed populations along with the rapid spread of resistance in
L. perenne due to its obligate cross-pollination and self-incompatibility could account for the selection of cross-resistant populations by the repeated application of these herbicides [
40,
41]. Moreover, the fact that the control of this weed in Greece is mainly based on the ACCase and ALS inhibitors only, which are at high risk for herbicide-resistance evolution, the herbicide-resistant populations of this weed threaten the sustainability of cereal crop production, and this danger requires a well-thought approach to manage field-evolved resistance.
The lack of point mutations at codon Pro-197 of the sequenced
ALS gene in the S population confirms its susceptibility to ALS inhibitors found in the whole-plant response experiments. However, the two identified point mutations in the
ALS gene, causing amino acid substitutions at Pro-197 position in the ALS enzyme of the R1 plants, support the evidence of target site-based herbicide cross-resistance and confirm the plant resistance findings for this population. The Pro-197-His and Pro-197-Leu amino acid substitutions found in separate individuals of the R1 population, according to our knowledge, are reported for the first time in
L. perenne, although their presence is very common in other weed species [
10]. These two amino-acid substitutions conferred broad target-site cross-resistance to chlorsulfuron, mesosulfuron + iodosulfuron, pyroxsulam, and propoxycarbazone, which agree with results reported by Menegat et al. [
23], who found that plants of a French
L. perenne population with the ALS Asp-376-Glu genotype were cross-resistant against mesosulfuron + iodosulfuron, pyroxsulam and propoxycarbazone. However, Vázquez-García et al. [
11] found that a
L. perenne population from Chile was resistant to the ALS-iodosulfuron (due to non-target-site resistance mechanisms), and also multiple-resistant to the ACCase-diclofop-methyl (due to an Asp-2078-Gly point mutation) and the EPSPS-glyphosate (due to
EPSPS gene amplification resulting in high enzyme activity). Finally, plant dose–response assays conducted by Singh et al. [
27] indicated that three
L. perenne populations in Texas were multiple-resistant to the ACCase-diclofop-methyl and the ALS-mesosulfuron-methyl.
The cross-resistance of the R1 population to sulfonylurea (chlorsulfuron, mesosulfuron + iodosulfuron) and triazolopyrimidine (pyroxsulam), but not to imidazolinone (imazamox) herbicides, is in contrast with the results reported by Saari et al. [
22], who found that a
L. perenne population from California was cross-resistant (due to less sensitive ALS target site enzyme) to both sulfonylurea (chlorsulfuron, sulfometuron, triasulfuron) and imidazolinone (imazapyr) herbicides.
The extremely high susceptibility of the R1 population to the ACCase inhibitors clodinafop-propargyl, diclofop-methyl, and pinoxaden, applied at even half of their registered field label rate, justifies their current use for the effective control of this weed in winter cereals. However, since the
L. perenne population has evolved cross-resistance to ACCase inhibitors due to the Asp-2078-Gly substitution in the ACCase enzyme [
11], measures should be taken for the rotational and appropriate use of these ACCase inhibitors.
The similar aboveground biomass and tiller number increasing trends of the R1 and S populations in the absence of crop competition strongly suggest that the R1 population has similar fitness compared that of the S population. In addition, the similar aboveground biomass of the R1 and S populations grown in competition with wheat is in agreement with the results found in the growth rate experiment (with the absence of competition). The similar vegetative productivity trends of both R1 and S populations grown in the absence or presence of wheat competition could be attributed to their similar product biosynthesis, which may result from the similar catalytic activity of their ALS enzyme and/or similar substrate affinity, and/or similar feedback inhibition [
32,
41]. In general, the similarly calculated slopes for the R1 and S populations from the linear equations fitted on their aboveground biomass and tiller number grown in the absence or presence of wheat competition support the evidence for similar growth rates. In contrast to these results, Menegat et al. [
23] found that a French
L. perenne population, with cross-resistance to ALS inhibitors due to Asp-376-Glu substitution of the
ALS gene, had no significant impact on shoot biomass and tiller number compared to the wild-type population, although its root biomass and ALS enzyme activity were 68 % and 48 % lower than those of the S population, respectively. In addition, Zangeneh et al. [
42], studying one S and two R
L. rigidum populations with target-site resistance to ACCase inhibitors, found that the R-2041-Asn population exhibited better fitness than the S and R-1781-Leu populations. Based on these results, it could be concluded that fitness cost endowed by field-selected
ALS target-site resistance genes cannot be easily predicted, because it depends on weed species, the mutant allele allowing resistance, the genetic background, and the experimental and environmental conditions [
43].
The lack of proportional aboveground biomass increases with increasing density of the R1 and S populations supports the evidence of both inter- and intra-species competition. In addition, their similar increasing aboveground biomass trends with increasing density were confirmed by their similar estimated growth rates. Finally, the similar wheat biomass growth in competition with either the R1 or S population suggests a similar competitive ability for both weed populations against wheat.