Integrating Cultivation Practices and Post-Emergence Herbicides for ALS-Resistant False Cleavers (Galium spurium L.) Management in Durum Wheat

Abstract

False cleavers (Galium spurium L.) is a broadleaf weed species that affects wheat productivity because of its strong competition for resources. It has developed resistance to acetolactate synthase (ALS) inhibitors, such as sulfonylureas and triazolopyrimidines, which are herbicides widely used in durum wheat. Integrated weed management programs can contribute to the control of this species and delay the evolution of herbicide resistance. Thus, a two-year field experiment was conducted to evaluate the effects of sowing time, variety, and herbicides on crop yield, density, and dry weight of a false cleavers population with resistance to ALS inhibitors. In both growing seasons, a split-split-plot design was used with three replicates. The sowing date was chosen as the main plot factor, durum wheat varieties as the subplot factor, and herbicides as the sub-subplot factor. The herbicide treatments were: (1) metsulfuron-methyl/bensulfuron-methyl (4/50 g a.i. ha⁻¹), (2) aminopyralid/florasulam (9.9/4.95 g a.i. ha⁻¹), (3) pyroxsulam and florasulam/2,4-D (18.75 + 4.725/225 g a.i. ha⁻¹), (4) 2,4-D/bromoxynil (633.15/601.2 g a.i. ha⁻¹), non-treated control, and hand-weeded control for the first season, while in the second season one more herbicide treatment (halauxifen-methyl/florasulam, 5.6/5.15 g a.i. ha⁻¹) was added. Herbicide application was performed on 10 March 2021 and 28 March 2022, when the crop was at the end of tillering and the beginning of stem elongation. The results showed that the density of false cleavers was not affected by the variety or sowing time. However, its dry weight was 17.3–23.4% higher in early sowing (16 November in 2020 and 8 November 2021) than in late sowing (24 December 2020 and 2 December 2021). Among the herbicides tested, 2,4-D/bromoxynil and halauxifen-methyl/florasulam effectively controlled false cleavers, showing greater efficacy in late sowing (>88%), which ultimately led to a higher yield. In conclusion, our two-year findings demonstrate that delayed sowing as part of an integrated weed management strategy can contribute to controlling resistant populations of false cleavers to ALS-inhibiting herbicides without affecting the quantity and quality of durum wheat yield in areas with a Mediterranean climate.

1. Introduction

Chemical weed control is the primary method applied and preferred by farmers for winter cereals. However, several cultural weed control methods are also applied to these crops and contribute to weed management [1,2,3,4]. Some cultural methods that can be combined with chemical control in winter cereals within the framework of integrated weed management programs include crop rotation, false seedbed, sowing date, selection of competitive varieties, and sowing density [5,6,7,8]. Sowing date is a cultural technique that can help control several grass and broadleaf weed species [9,10] such as black-grass (Alopecurus myosuroides Huds.), milk thistle (Silybum marianum (L.) Gaertn), and wild mustard (Sinapis arvensis L.). However, the impact of sowing time on weed management varies by region and depends on the weed species present in cereal fields [7,10,11]. In a recent study, Khasraw et al. [7] observed that late sowing of wheat on 30 December with a seeding rate of 160 kg ha⁻¹ reduced the weed dry biomass by up to 66.2% compared to early sowing on 20 November with the same seeding rate. Additionally, it is important to note that sowing date can affect the efficacy of certain herbicides against specific weed species [10], while some researchers have reported variations in herbicide use based on the sowing date [12]. However, in some cases, a delay in sowing date can reduce plant growth and seed yield in wheat cultivation [13,14]. Lindell et al. [14] found that late sowing (18 and November 30 in the 2019/2020 and 2020/2021 growing seasons, respectively) resulted in a 20.6–57.3% decrease in the seed yield of winter wheat compared to that in the optimal sowing date (17 and October 9 in 2019/2020 and 2020/2021, respectively).

Regarding the competitive ability of wheat, various studies have shown that different varieties exhibit differences in their competitive ability. In an experiment conducted at two sites in Canada, Gerard et al. [15] found that the yield losses due to wild oat (Avena fatua L.) competition differed in eight wheat varieties (e.g., Colombus, Laura, and Oslo). Some characteristics associated with the greater competitive ability of wheat varieties are as follows: higher plant height [3,16,17], greater tillering ability [16,18], greater early plant growth [19], higher rate of stem growth [20], higher leaf area [18], larger leaf width [21], greater root system [18], and winter hardiness [16]. In this context, in a field experiment conducted in Turkey, Mennan and Zandstra [22] observed that four wheat varieties (Bezostaja, Kate A-l, Momtchill, and Panda) presented different competitive abilities against catchweed bedstraw (Galium aparine L.), as the Bezostaja variety reduced the growth and seed production of this weed to a greater extent than the other varieties. Thus, the inclusion of varieties with superior competitive abilities in integrated weed management systems can contribute to effective weed management. However, it is important to note that Fang et al. [18] observed that the winter wheat landrace PL40 exhibited greater competitive ability than the CW135 variety, but the grain yield in the moderate water stress treatment was 28.5% and 24.3% lower than that in the well-watered treatment in PL40 and CW135, respectively, revealing that the yield reduction was lower in the less competitive genotype under moderate water stress conditions.

Additionally, it is important to point out that due to climate change, it is critical to adapt farming practices (e.g., sowing time and variety selection) based on climatic data to mitigate the effects of climate change on crop productivity [23,24]. The adaptation of cultivation practices (e.g., sowing time, varieties, and time of herbicide application) is also important for optimizing weed management because climate change (e.g., high temperatures and drought) affects weed interference with crops as well as herbicide efficacy [25,26,27,28].

Among the various weeds, false cleavers (Galium spurium L.) and catchweed bedstraw (G. aparine) are significant broadleaf species in wheat cultivation. Specifically, false cleavers present intense competition for resources throughout the crop growth cycle, resulting in a decrease in seed yield [22,29]. Additionally, the evolution of resistance to various herbicides that inhibit the ALS (acetolactate synthase) enzyme in both Galium species (G. spurium and G. aparine) has made its control even more difficult in various areas where wheat is grown [29,30,31]. For the above-mentioned reasons, the implementation of integrated management programs can contribute to controlling false cleavers in winter cereal crops as well as delaying the evolution of the herbicide resistance problem. In this context, the adoption of integrated management systems by farmers can contribute to the reduction of the chemical pesticides (e.g., herbicides) use by 50% by 2030 [32,33], a target set in the farm-to-fork strategy adopted by the European Commission in the framework of the European Green Deal to achieve an environmentally friendly food system in the European Union [34].

Thus, the main objective of this two-year experiment was to evaluate the effect of a combination of cultural techniques and chemical control on the infestation of false cleavers populations with resistance to ALS herbicides in durum wheat. Specifically, the effects of durum wheat varieties, sowing time, and herbicides on the density and biomass of this weed species were studied, and the impact of these factors on the quality and yield of the durum wheat crop was also evaluated.

2. Materials and Methods

2.1. Experimental Design and Main Cultural Practices

Field experiments were conducted in the Oropos region of eastern Attica (38°18′N, 23°45′E), Greece, during the 2020–2021 and 2021–2022 growing seasons. In both growing seasons, the experiments were arranged in a split-split plot design with three replicates. The sowing date was selected as the main plot factor, durum wheat varieties as the subplot factor, and herbicides as the sub-subplot factor. The main physicochemical properties of the soil in the experimental field were as follows: clay loam texture (clay: 38%, silt: 28%, and sand: 34%), organic matter (4.9%), pH (7.6), and electrical conductivity (1.41 mS cm⁻¹). The meteorological data from the weather station located on the farm are shown in Figure 1.

The mail plot factor (sowing date) included two levels: early and late sowing dates. In the first growing season, the early sowing was conducted on 16 November 2020, and the late sowing on 24 December 2020, while in the second growing season, the early sowing took place on 8 November 2021, and the late sowing on 2 December 2021. In both growing seasons, the durum wheat seed rate was set at 240 kg ha⁻¹. On the same dates, a false cleavers population resistant to ALS-inhibiting herbicides, such as mesosulfuron-methyl/iodosulfuron-methyl-sodium, tribenuron-methyl (chemical family: sulfonylureas), and florasulam (chemical family: triazolopyrimidines) [29] was also sown. A mixture (7 g) of seeds and plant residues of false cleavers containing approximately 2000 seeds was evenly dispersed in each plot and incorporated into the soil of each plot. False cleavers seeds were collected from a durum wheat field in Agios Georgios (Domokos region, Central Greece) in June 2020 and 2021. Regarding fertilization, an inorganic fertilizer (16-20-0; N-P₂O₅-K₂O, 300 kg ha⁻¹) was applied as basal fertilization prior to crop sowing and then incorporated into the soil, while the inorganic fertilizer calcium ammonium nitrate (26-0-0; N-P₂O₅-K₂O, 300 kg ha⁻¹) was applied as top-dressing at the tillering stage of durum wheat.

The subplot factor (varieties) included two levels: durum wheat varieties Levante and Simeto. Levante is a mid-early variety with white awns that is widely cultivated in Greece. This variety has a medium height, good tillering, high protein content, and good resistance to low temperatures, lodging, and some foliar diseases. Simeto is also a mid-early variety with black awns that is widely grown in Greece due to its good adaptability. Moreover, the latter variety has moderate tillering capacity and height, satisfactory lodging resistance and cold tolerance, excellent heat and drought tolerances, and satisfactory resistance to foliar diseases.

Regarding the sub-subplot factor, in the first growing season, the herbicide treatments were as follows: herbicide 1: metsulfuron-methyl/bensulfuron-methyl (Phyton, UPL Europe Ltd., Warrington, United Kingdom), herbicide 2: aminopyralid/florasulam (Lancelot 450 WG, Corteva Agriscience Hellas, Athens, Greece), herbicide 3: pyroxsulam (Senior 75 WG, Corteva Agriscience Hellas, Athens, Greece) and florasulam/2,4-D as 2-ethylhexyl ester (Titanas SE, Adama Hellas, Marousi, Greece), tank-mix, herbicide 4: 2,4-D as 2-ethylhexyl ester/bromoxynil as actonate/heptanoate ester (Brominal Nuevo, Bayer Hellas, Marousi, Greece), non-treated control, and hand-weeded control. Specifically, 72 plots (6 m², 2 m × 3 m) were established in the experimental area. In the second growing season, an additional herbicide was applied to the experimental field. Thus, the experimental treatments were as follows: herbicides 1 to 4, herbicide 5: halauxifen-methyl/florasulam (Quelex, Corteva Agriscience Hellas), non-treated control, and hand-weeded control. Specifically, 84 plots (6 m², 2 m × 3 m) were established for the second experiment. In both growing seasons, two hand weedings were performed in the hand-weeded control treatment: the first on the day of herbicide application and the second about two weeks later. The herbicides were applied using a precision sprayer [flat fan nozzles (flow rate: 0.73 L/min, 110°), spray volume 300 L per hectare, pressure: 2.5 atm] to ensure uniform spraying and complete coverage. Herbicide application was performed on 10 March 2021 and 28 March 2022, when the crop was at the end of tillering and the beginning of stem elongation (BBCH 30–31). In both experiments, herbicides were applied at their maximum recommended doses (

Table 1. Post-emergence herbicides and application doses used in the field experiments in both growing seasons (2020/2021 and 2021/2022).

Common Name	Trade Name	Dose (g a.i. ha−1)	Modes of Action
metsulfuron-methyl/bensulfuron-methyl 1	Phyton	4/50	Inhibition of acetolactate synthase
aminopyralid/florasulam 2	Lancelot 450 WG	9.9/4.95	Auxin mimics/Inhibition of acetolactate synthase
pyroxsulam + florasulam/2,4-D 3	Senior 75 WG + Titanas SE	18.75 + 4.725/225	Inhibition of acetolactate synthase + Inhibition of acetolactate synthase/Auxin mimics
2,4-D/bromoxynil 4	Brominal Nuevo	633.15/601.2	Auxin mimics/Inhibition of photosynthesis at PS II
halauxifen-methyl/florasulam 5	Quelex	5.6/5.15	Auxin mimics/Inhibition of acetolactate synthase

Abbreviations: ¹ metsul/bensulf, ² amin/floras, ³ pyroxs + 2,4-D/floras, ⁴ 2,4-D/bromox, and ⁵ halaux/floras, active ingredient (a.i.).

), while an adjuvant (alkylphenol alkoxylate 99% w/v, Kaytar SL, Elanco Hellas S.A., Athens, Greece) was added to herbicides 1, 2, 3, and 5 at a dose of 200 mL per 100 L of the spray solution.

2.2. Sampling and Measurements

The dry weight and the density of false cleavers and other weed species were measured at random central points within the sub-subplots using a 60 cm × 60 cm square wooden frame, and after drying in an oven at 60 °C for four days. The measurements of these parameters took place at 37 DAA (days after herbicide application) and 38 DAA in 2020–2021 and 2021–2022, respectively. The herbicide efficacy (%) was calculated based on the weed dry weight (WDW) using the following equation:

$$Efficacy (\%) =100 \times {\displaystyle \frac{ WDW in nontreated control-WDW in herbicide }{WDW in nontreated control }}$$

(1)

To determine the dry weight of the aboveground parts of the durum wheat crop, destructive sampling was performed. Plants were cut from an area of 0.36 m² at 37 Days After herbicide Application (DAA) and 38 DAA in 2020–2021 and 2021–2022, respectively. The samples were dried in an oven at 60 °C for four days. Plant height measurements were conducted on the same dates as the dry weight measurements, while spike length was measured before crop harvest by randomly selecting five plants from each plot. For the determination of photosynthesis rate (μmol CO₂ m⁻² s⁻¹), transpiration rate (mmol H₂O m⁻² s⁻¹), and stomatal conductance (mol m⁻² s⁻¹), a portable device, the LCi Leaf Chamber Analysis System (ADC, Bioscientific, Hoddesdon, UK), was used on cloudless days at 38–39 and 39–40 DAA in 2020–2021 and 2021–2022, respectively. These physiological parameters were recorded for the flag leaves of durum wheat plants.

The durum wheat crop was harvested by hand in mid-June from an area of 1 m². The 1000-seed weight was calculated by measuring three samples of 100 seeds, while protein content and wet and dry gluten content were determined at the Laboratory of Agronomy and Applied Crop Physiology of the University of Thessaly. The above-mentioned quality traits were measured using an NIR analyzer (model DA 7250 NIR analyzer, Perten Instruments, Hägersten, Sweden).

2.3. Statistical Analysis

To evaluate the effects of the three factors as well as their interactions, the data of all measured crop and weed parameters recorded in the two field experiments were submitted to analysis of variance (Three-Way ANOVA) according to a split-split plot design using the statistical package IBM SPSS version 24 (IBM Corp., Armonk, NY, USA). Duncan’s post hoc test at p < 0.05 was used for the comparison of means (herbicides and interaction effects between the factors). The Shapiro-Wilk test was used to check the data normality for all measured parameters. Finally, a Pearson correlation analysis (two-tailed test, n = 72 and n = 84 for the first and second growing seasons, respectively) was conducted to determine the relationships between the main parameters determined in the two experiments.

3. Results

3.1. Weed Flora

3.1.1. False Cleavers Density and Dry Weight

In both growing seasons, the data analysis for the density of false cleavers showed no interactions among the three factors, while there were no significant differences between varieties or between the sowing date treatments. In 2020–2021, the false cleavers density ranged from 0 to 31.27 plants per m² (

Table 2. Effects of variety, sowing time, and herbicides on false cleavers density (plants m⁻²) and biomass (kg ha⁻¹) in durum wheat. Varieties (Levante and Simeto), sowing time (early and late), herbicide treatments [non-treated control, hand-weeded control, metsulfuron-methyl/bensulfuron-methyl (metsul/bensul), aminopyralid/florasulam (amin/floras), pyroxsulam + florasulam/2,4-D (pyroxs + 2,4-D + floras), 2,4-D/bromoxynil (2,4-D/bromox), and halauxifen-methyl/florasulam (halaux/floras)].

Factors	False Cleavers Density (Plants m−2)		False Cleavers Dry Weight (kg ha−1)
	2020–2021	2021–2022	2020–2021	2021–2022
Variety
Levante	19.13 ± 13.44	17.35 ± 14.18	480.0 ± 335.9	492.2 ± 440.3
Simeto	17.89 ± 12.43	17.94 ± 14.76	471.1 ± 334.8	467.7 ± 409.7
Sowing date
Early	17.73 ± 11.77	17.08 ± 13.12	538.6 ± 353.3 a	525.3 ± 435.4 a
Late	19.29 ± 14.02	18.20 ± 15.69	412.5 ± 303.2 b	434.6 ± 410.2 b
Herbicides
Non-treated control	31.27 ± 6.18 a	35.32 ± 8.53 a	855.9 ± 159.4 a	1018.4 ± 146.4 a
Hand-weeded control	0 ± 0 d	0 ± 0 e	0 ± 0 e	0 ± 0 e
metsul/bensul	26.37 ± 8.40 a	29.32 ± 8.47 b	716.0 ± 135.2 b	865.1 ± 116.9 b
amin/floras	18.90 ± 5.86 b	19.85 ± 4.39 c	404.6 ± 92 c	434.2 ± 158.2 c
pyroxs + 2,4-D/floras	27.53 ± 7.72 a	29.09 ± 6.94 b	707.3 ± 144.1 b	886.7 ± 130.8 b
2,4-D/bromox	7.00 ± 3.87 c	5.77 ± 3.43 d	169.5 ± 95 d	93.4 ± 63.1 d
halaux/floras	-	4.16 ± 3.01 de	-	61.8 ± 46.4 de
F values
FVariety(V)	0.668	0.188	0.136	1.208
FSowing Date(S)	1.044	0.671	27.187 ***	16.586 ***
FHerbicides(H)	44.952 ***	61.906 ***	132.343 ***	222.664 ***
FV × S	0.376	0.021	1.675	0.549
FV × H	0.161	0.118	0.067	0.676
FS × H	1.026	1.120	1.427	1.323
FV × S × H	0.282	0.124	0.217	0.274

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters indicate statistically significant differences according to Duncan’s post hoc test (herbicide factor) and analysis of variance (variety and sowing date factors). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.001 (***).

). As for the herbicides and their impact on the density of false cleavers, the highest density was recorded in the non-treated control (31.27 plants per m²), followed by the pyroxsulam + florasulam/2,4-D treatments (27.53 plants per m²), metsulfuron-methyl/bensulfuron-methyl (26.37 plants per m²), and aminopyralid/florasulam (18.9 plants per m²).

Among the applied herbicides, the lowest density was found in the 2,4-D/bromoxynil treatment, which was 77.6% lower than that in the untreated control. In 2021–2022, among the herbicides, the highest false cleavers density was measured in the non-treated control (35.32 plants m⁻²), with statistically significant differences from all other treatments. Moreover, the herbicides 2,4-D/bromoxynil and halauxifen-methyl/florasulam had statistically lower values (4.16–5.77 plants m⁻²) compared to those in other herbicides (metsulfuron-methyl/bensulfuron-methyl, pyroxsulam + florasulam/2,4-D, and aminopyralid/florasulam).

Regarding the dry weight of false cleaver, durum wheat varieties did not affect this parameter in either growing season, whereas significant differences were found between the herbicide treatments and between late and early sowing. No interactions were observed among the three experimental factors. Regarding the sowing date factor, our results showed that early sowing had significantly higher values of dry weight for false cleavers (538.6 kg ha⁻¹ and 525.3 kg ha⁻¹ in 2020–2021 and 2021–2022, respectively) than late sowing (412.5 kg ha⁻¹ and 434.6 kg ha⁻¹ in 2020–2021 and 2021–2022, respectively). Among the herbicide treatments, the highest values of false cleavers dry weight (855.9 kg ha⁻¹) were recorded in the non-treated control, while the 2,4-D + bromoxynil treatment had the lowest value (169.5 kg ha⁻¹). Moreover, there were no differences between the effects of metsulfuron-methyl/bensulfuron-methyl and pyroxsulam + florasulam/2,4-D on this parameter, but these treatments significantly differed from the other treatments. In 2021–2022, the non-treated control had the statistically highest value of false cleavers dry weight (1018.4 kg ha⁻¹), followed by the herbicides pyroxsulam and florasulam/2,4-D (886.7 kg ha⁻¹) and metsulfuron-methyl/bensulfuron-methyl (865.1 kg ha⁻¹), which did not differ significantly from each other. The aminopyralid/florasulam treatment had the next lowest value for false cleavers dry weight. Moreover, the 2,4-D/bromoxynil and halauxifen-methyl/florasulam treatments recorded the lowest dry weight values among all the herbicides.

3.1.2. Total Weed Density and Dry Weight

Apart from false cleavers, the most significant weed species observed in the field were common poppy (Papaver rhoeas L.), common field-speedwell (Veronica persica L.), and chickweed (Stellaria media (L.) Vill.). In contrast, wild mustard (Sinapis arvensis L.), sow thistle (Sonchus oleraceus L.), common fumitory (Fumaria officinalis L.), and henbit dead-nettle (Lamium amplexicaule L.) were recorded at lower densities. In 2020–2021, the data analysis for total weed density showed no interactions among the three factors, and there were no significant differences between varieties or between the sowing date treatments (

Table 3. Effects of variety, sowing date, and herbicides on total weed density (plants m⁻²) in durum wheat. Treatments for each factor are explained in Table 2.

Factors	Total Weed Density (Plants m−2)
	2020–2021	2021–2022
Variety
Levante	24.89 ± 18.56	22.36 ± 18.79
Simeto	23.64 ± 15.72	25.06 ± 20.93
Sowing date
Early	25.12 ± 17.84	26.84 ± 21.42 a
Late	23.41 ± 16.51	20.58 ± 17.79 b
Herbicides
Non-treated control	48.30 ± 9.33 a	55.40 ± 14.04 a
Hand-weeded control	0 ± 0 e	0 ± 0 e
metsul/bensul	31.97 ± 7.01 b	35.09 ± 8.21 b
amin/floras	25.67 ± 6.07 c	26.32 ± 10.13 c
pyroxs + 2,4-D/floras	29.87 ± 9.58 bc	33.24 ± 7.09 b
2,4-D/bromox	9.80 ± 4.22 c	6.69 ± 4.33 d
halaux/floras	-	9.23 ± 5.06 d
F values
FVariety(V)	0.595	2.647
FSowing Date(S)	1.126	14.213 ***
FHerbicides(H)	75.265 ***	79.049 ***
FV × S	0.753	0.984
FV × H	2.275	0.354
FS × H	0.361	1.410
FV × S × H	0.547	0.551

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters indicate statistically significant differences according to Duncan’s post hoc test (herbicide factor) and analysis of variance (sowing date factor). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.001 (***).

). Moreover, among the herbicide treatments, the non-treated control had the highest value (48.3 plants m⁻²), while the 2,4-D + bromoxynil treatment had the lowest total weed density (9.8 weeds m⁻²) among the applied herbicides, showing a difference of 79.7% compared to the non-treated control. In 2021–2022, the total weed density ranged from 0 to 55.4 plants m⁻². The data analysis also showed no interactions among the three factors, and no significant differences between the varieties. In contrast, there were significant differences between sowing dates, with the highest values recorded for early sowing. Moreover, among the herbicide treatments, the highest value of this parameter was again measured in the non-treated control (55.4 plants m⁻²), with statistically significant differences from the other herbicide treatments. The herbicides pyroxsulam and florasulam/2,4-D, aminopyralid/florasulam, and metsulfuron-methyl/bensulfuron-methyl resulted in greater total weed density values than the halauxifen-methyl/florasulam and 2,4-D/bromoxynil treatments.

For the total weed dry weight, in both growing seasons, there was an interaction between the sowing date and the herbicides, while the variety factor did not affect this parameter (

Table 4. Effects of variety and sowing date on total weed dry weight (kg ha⁻¹) and herbicide efficacy (%) against false cleavers. Treatments for each factor are explained in Table 2.

Factors/Interactions		Total Weed Dry Weight (kg ha−1)		Efficacy (%)
		2020–2021	2021–2022	2020–2021	2021–2022
Variety
Levante		991.3 ± 861.7	806.7 ± 764.5	54.2 ± 34.2	62.1 ± 37.4
Simeto		915.0 ± 844.4	765.3 ± 727.2	52.5 ± 35.1	61.2 ± 38.2
Sowing date × Herbicides
Early	Non-treated control	2502.1 ± 298.2 a	2506.5 ± 323 a	-	-
	Hand-weeded control	0 ± 0 e	0 ± 0 h	100.0 ± 0 a	100.00 ± 0 a
	metsul/bensul	1461.1 ± 407.9 b	1350.8 ± 225.3 c	16.6 ± 7.5 e	15.8 ± 5.7 e
	amin/floras	1145.9 ± 397.1 bc	840.4 ± 167.6 e	52.9 ± 5.0 d	49.6 ± 11.5 d
	pyroxs + 2,4-D/floras	1245.9 ± 351.7 b	1071.9 ± 126.4 d	17.2 ± 5.9 e	12.7 ± 8.5 e
	2,4-D/bromox	227.1 ± 84.8 e	237.6 ± 89.3 g	73.3 ± 5.8 c	87.4 ± 5.5 b
	halaux/floras	-	498.9 ± 146.3 f	-	92.8 ± 4.4 ab
Late	Non-treated control	2340.5 ± 386.8 a	1900.5 ± 270.5 b	-	-
	Hand-weeded control	0 ± 0 e	0 ± 0 h	100.0 ± 0 a	100.0 ± 0 a
	metsul/bensul	904.1 ± 319.8 cd	1013.1 ± 251.9 de	15.8 ± 4.3 e	13.5 ± 8.7 e
	amin/floras	676.5 ± 177 d	475.5 ± 183.8 f	52.0 ± 12.0 d	66.0 ± 9.0 c
	pyroxs + 2,4-D/floras	776.0 ± 166.5 d	836.5 ± 151.9 e	17.6 ± 7.0 e	12.5 ± 9.7 e
	2,4-D/bromox	158.4 ± 77.1 e	50.2 ± 27.9 gh	88.3 ± 2.9 b	94.6 ± 3.5 ab
	halaux/floras	-	222 ± 178.6 gh	-	95.1 ± 4.6 ab
F Values
FVariety(V)		1.526	1.091	0.929	0.252
FSowing Date(S)		21.666 ***	52.441 ***	2.544	5.303 *
FHerbicides(H)		129.106 ***	207.599 ***	385.112 ***	372.314 ***
FV × S		2.847	0.246	0.431	0.413
FV × H		0.317	0.983	0.131	0.860
FS × H		2.489 *	3.116 *	3.220 *	2.796 *
FV × S × H		1.722	0.955	0.116	0.461

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters indicate statistically significant differences according to Duncan’s post hoc test (sowing date x herbicides interaction). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.05 (*) and p < 0.001 (***).

). In 2020–2021, the non-treated control in early (2502.1 kg ha⁻¹) and late (2340.5 kg ha⁻¹) sowing had the statistically highest dry weight values, although there were no statistically significant differences between the two sowing dates. The lowest total dry weed weight values were observed for the herbicide 2,4-D/bromoxynil (227.1 and 158.4 kg ha⁻¹ in early and late sowing, respectively), with no differences between the two sowing dates. For the other herbicides, there were no significant differences in total dry weight between early and late sowing, while their values were higher than those of the 2,4-D + bromoxynil treatment. In 2021–2022, the non-treated control in early sowing (2506.5 kg ha⁻¹) and late sowing (1900.5 kg ha⁻¹) had the statistically highest dry weight values, which differed significantly from each other. The lowest values of total weed dry weight were recorded in the herbicide treatments 2,4-D/bromoxynil in early sowing (237.6 kg ha⁻¹), halauxifen-methyl/florasulam (222 kg ha⁻¹) in late sowing, and 2,4-D/bromoxynil in late sowing (50.2 kg ha⁻¹), with no significant differences among them. For the other herbicides, there were no significant differences in total dry weight between the early and late sowing dates.

3.1.3. Herbicides Efficacy Against False Cleavers

In 2020–2021, the statistical analysis of the efficacy data showed an interaction between sowing date and herbicides, while the varieties did not affect the efficacy rate (Table 4). The herbicide 2,4-D/bromoxynil had the highest efficacy rate among all the herbicides. Additionally, this herbicide was more effective against false cleavers in late sowing (88.3%) than in early sowing (73.3%). Regarding the control of false cleavers by the other herbicides, the results showed that the aminopyralid + florasulam herbicide had moderate efficacy (52–52.9%) on both sowing dates, without statistically significant differences between them. The lowest efficacy rates (15.8–17.6%) were observed for the herbicides metsulfuron-methyl/bensulfuron-methyl and pyroxsulam plus florasulam/2,4-D on both sowing dates, with no statistically significant differences between them.

In 2021–2022, statistical analysis of the efficacy data showed an interaction between sowing date and herbicides, while the varieties did not affect the herbicide efficacy. The herbicides 2,4-D/bromoxynil and halauxifen-methyl/florasulam were more effective against false cleavers in late sowing (94.6% and 95.1%, respectively) than in early sowing (87.4% and 92.8%, respectively), although the differences were not statistically significant. Additionally, the herbicide aminopyralid/florasulam was more effective in late sowing (66.0%) than in early sowing (49.6%). The herbicide metsulfuron-methyl/bensulfuron-methyl and the mixture of herbicides pyroxsulam and florasulam/2,4-D showed the poorest efficacy against this weed species, both in early sowing (12.7–15.8%) and in late sowing (12.5–13.5%), without statistically significant differences between them.

3.2. Durum Wheat

3.2.1. Physiological Parameters

In 2020–2021, for the physiological parameters (photosynthesis rate, transpiration rate, and stomatal conductance), significant effects of herbicide treatments were observed, as well as interactions between variety and sowing date (

Table 5. Effects of variety, sowing date, and herbicides on photosynthesis rate (PR, μmol CO₂ m⁻² s⁻¹), transpiration rate (TR, mmol H₂O m⁻² s⁻¹), and stomatal conductance (SC, mol m⁻² s⁻¹) of durum wheat. Treatments for each factor are explained in Table 2.

Factors/ Interactions		PR	TR	SC	Factors	PR	TR	SC
		2020–2021				2021–2022
Variety × Sowing date					Variety
					Levante	9.89 ± 1.61 b	3.85 ± 0.44 b	0.16 ± 0.04 b
Levante	Early	13.81 ± 1.95 b	5.30 ± 0.79 b	0.34 ± 0.09 b	Simeto	10.69 ± 2.07 a	4.11 ± 0.74 a	0.21 ± 0.07 a
	Late	13.59 ± 1.93 b	5.28 ± 0.69 b	0.30 ± 0.06 b	Sowing date
Simeto	Early	14.32 ± 1.60 b	5.85± 0.95 b	0.38 ± 0.09 b	Early	10.36 ± 2.03	4.06 ± 0.61	0.19 ± 0.06
	Late	16.52 ± 2.71 a	6.57 ± 0.81 a	0.54 ± 0.20 a	Late	10.21 ± 1.74	3.90 ± 0.62	0.18 ± 0.06
Herbicides					Herbicides
Non-treated control		12.06 ± 1.10 b	4.76 ± 0.69 b	0.29 ± 0.07 b	Non-treated control	6.99 ± 0.96 c	3.06 ± 0.37 b	0.10 ± 0.01 d
Hand-weeded control		15.32 ± 2.16 a	6.15 ± 0.62 a	0.41 ± 0.12 a	Hand-weeded control	11.37 ± 1.19 a	4.29 ± 0.40 a	0.22 ± 0.07 a
metsul/bensul		15.16 ± 2.15 a	5.90 ± 0.87 a	0.39 ± 0.15 ab	metsul/bensul	10.05 ± 1.10 b	3.93 ± 0.41 a	0.17 ± 0.05 c
amin/floras		14.74 ± 2.41 a	5.63 ± 0.79 a	0.37 ± 0.11 ab	amin/floras	10.45 ± 1.11 ab	4.00 ± 0.47 a	0.18 ± 0.05 bc
pyroxs + 2,4-D/floras		14.54 ± 2.28 a	5.88 ± 1.21 a	0.45 ± 0.21 a	pyroxs + 2,4-D/floras	10.83 ± 1.61 ab	3.99 ± 0.53 a	0.19 ± 0.05 abc
2,4-D/bromox		15.56 ± 2.30 a	6.18 ± 0.86 a	0.43 ± 0.18 a	2,4-D/bromox	10.94 ± 1.04 ab	4.27 ± 0.55 a	0.21 ± 0.05 ab
-		-	-	-	halaux/floras	11.13 ± 1.85 ab	4.21 ± 0.64 a	0.20 ± 0.06 abc
F Values					F values
FVariety(V)		14.442 ***	31.783 ***	24.612 ***	FVariety(V)	6.668 *	5.242 *	27.597 ***
FSowing Date(S)		4.787 *	4.691 *	5.372 *	FSowing Date(S)	0.074	1.511	1.825
FHerbicides(H)		5.354 **	6.967 ***	2.825 *	FHerbicides(H)	16.611 ***	9.997 ***	10.054 ***
FV × S		7.125 *	5.270 *	12,259 **	FV × S	1.369	3.593	0.360
FV × H		0.665	0.899	1.165	FV × H	0.721	1.684	1.495
FS × H		0.153	0.795	0.300	FS × H	1.371	1.004	1.176
FV × S × H		0.307	0.814	0.445	FV × S × H	0.585	0.391	1.073

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters have statistically significant differences according to Duncan’s post hoc test (herbicide factor) and analysis of variance (sowing date and variety factors). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***).

). Among the herbicide treatments, the lowest photosynthetic rate (12.06 μmol CO₂ m² s⁻¹) and transpiration rate (4.76 mmol H₂O m² s⁻¹) values were recorded in the non-treated control, while there were significant differences between this treatment and all the other treatments. Moreover, the lowest stomatal conductance value (0.29 mol m² s⁻¹) was recorded for the non-treated control, without significant differences between this treatment and the metsulfuron-methyl/bensulfuron-methyl or aminopyralid/florasulam treatments. As mentioned above, interactions were observed for all parameters regarding the effects of variety and sowing date. Plants of the Simeto variety in late sowing had the highest values (photosynthesis rate: 16.52 μmol CO₂ m² s⁻¹, transpiration rate: 6.57 mmol H₂O m² s⁻¹, and stomatal conductance: 0.54 mol m² s⁻¹) with statistically significant differences from the other combinations of variety and sowing date factors (Levante variety in early sowing, Simeto variety in early sowing, and Levante variety in late sowing).

In 2021–2022, for the photosynthesis rate, transpiration rate, and stomatal conductance of flag leaves, no interaction among the factors was observed, although significant effects of variety and herbicide treatments were noted. In contrast, sowing date did not affect any of these physiological parameters. The photosynthesis rate in the Simeto variety (10.69 µmol CO₂ m² s⁻¹) was significantly higher than that in the Levante variety (9.89 µmol CO₂ m² s⁻¹). Similarly, the transpiration rate and stomatal conductance values were also significantly higher in the Simeto variety than in the Levante variety. Regarding the herbicide treatments for photosynthesis rate, the metsulfuron-methyl/bensulfuron-methyl (10.05 µmol CO₂ m² s⁻¹) and the non-treated control (6.99 µmol CO₂ m² s⁻¹) treatments had the lowest photosynthesis rates, with a statistically significant difference between them, while there were no significant differences between all the other herbicide treatments. For the transpiration rate, the hand-weeded control had the lowest value (3.06 mmol H₂O m² s⁻¹), while the other treatments did not differ significantly from each other, ranging from 3.93 to 4.29 mmol H₂O m² s⁻¹. For stomatal conductance, the hand-weeded control had the highest value (0.22 mol m² s⁻¹), followed by the herbicides 2,4-D/bromoxynil (0.21 mol m² s⁻¹), halauxifen-methyl/florasulam (0.20 mol m² s⁻¹), and pyroxsulam plus florasulam/2,4-D (0.19 mol m² s⁻¹), with no statistically significant differences among them.

3.2.2. Plant Growth Parameters

For the plant height of durum wheat, in both growing seasons, the data analysis showed an interaction between variety and sowing date (

Table 6. Effects of variety, sowing date, and herbicides on aboveground dry weight (kg ha⁻¹) and plant height (cm) of durum wheat. Treatments for each factor are explained in Table 2.

Factors	Aboveground Dry Weight (kg ha−1)		Factor/Interactions		Plant Height (cm)
	2020–2021	2021–2022			2020–2021	2021–2022
Variety			Variety × Sowing date
Levante	13,832.5 ± 3542.1	15,833.3 ± 2108.0 b
Simeto	14,369.4 ± 3673.6	17,881.0 ± 3110.7 a	Levante	Early	100.33 ± 4.97 a	94.80 ± 2.99 a
Sowing date				Late	85.11 ± 3.53 c	91.43 ± 2.34 b
Early	17,053.5 ± 2524.7 a	17,583.3 ± 3135.2 a	Simeto	Early	91.72 ± 3.72 b	85.54 ± 2.22 c
Late	11,148.3 ± 1371.7 b	16,131.0 ± 2314.5 b		Late	81.94 ± 3.04 d	86.48 ± 1.74 c
Herbicides			Herbicides
Non-treated control	12,625.3 ± 3457.7 b	14,222.2 ± 2101.5 b	Non-treated control		89.25 ± 7.45 ab	91.29 ± 4.87
Hand-weeded control	15,516.6 ± 3883.7 a	17,805.6 ± 2264.9 a	Hand-weeded control		91.00 ± 8.50 a	89.49 ± 4.26
metsul/bensul	13,409.0 ± 3169.7 b	16,861.1 ± 2120.0 a	metsul/bensul		89.33 ± 7.57 ab	89.83 ± 4.58
amin/floras	13,986.3 ± 3227.8 b	17,111.1 ± 2878.3 a	amin/floras		92.08 ± 9.24 a	89.50 ± 4.02
pyroxs + 2,4-D/floras	13,374.0 ± 3283.0 b	16,416.7 ± 3458.6 a	pyroxs + 2,4-D/floras		86.33 ± 6.72 b	87.95 ± 4.47
2,4-D/bromox	15,694.2 ± 4074.8 a	17,986.1 ± 2743.7 a	2,4-D/bromox		90.67 ± 8.97 a	89.02 ± 4.60
halaux/floras	-	17,597.2 ± 2827.1 a	halaux/floras		-	89.87 ± 4.70
F values			F values
FVariety(V)	1.552	15.003 ***	FVariety(V)		47.915 ***	180.015 ***
FSowing Date(S)	187.796 ***	7.548 **	FSowing Date(S)		215.885 ***	5.317 *
FHerbicides(H)	5.558 ***	3.382 **	FHerbicides(H)		3.673 **	2.056
FV × S	0.242	3.192	FV × S		10.239 ***	16.547 ***
FV × H	0.518	0.610	FV × H		1.053	0.373
FS × H	0.831	0.416	FS × H		0.885	0.502
FV × S × H	0.425	0.910	FV × S × H		0.513	0.412

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters are statistically significantly different according to Duncan’s post hoc test (herbicides and variety × sowing date interaction) and analysis of variance (variety and sowing date). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***).

). In 2020–2021, the plants of the Levante variety in early sowing had the greatest height (100.33 cm), followed by the plants of the Simeto variety in early sowing (91.72 cm). In 2021–2022, the Simeto variety at both sowing dates had statistically significantly lower values than the Levante variety. However, differences were observed between early and late sowing only in the Levante variety. Regarding the herbicide treatments, the shortest plant height was recorded in plots treated with pyroxsulam + florasulam/2,4-D (86.33 and 87.95 cm in 2020–2021 and 2021–2022, respectively). However, no statistically significant differences were observed among the herbicide treatments.

For the dry weight of the aboveground parts of durum wheat plants in 2020–2021, no statistically significant differences were observed regarding the variety, and no interaction was noted among the three factors. However, sowing date and herbicide treatment showed statistically significant differences. Early sowing had a significantly higher dry weight (17,053.5 kg ha⁻¹) than late sowing (11,148.3 kg ha⁻¹). Regarding the herbicide treatments, the herbicide 2,4-D/bromoxynil and the hand-weeded control had statistically significantly higher dry weight values (15,694.2 kg ha⁻¹ and 15,516.6 kg ha⁻¹, respectively) than the other treatments, but there were no statistically significant differences between them. In 2021–2022, no interaction was observed between the three factors, but there were statistically significant differences between the treatments of each factor. The Simeto variety (17,881 kg ha⁻¹) had a greater dry weight than Levante (15,833.3 kg ha⁻¹), while in early sowing, the dry weight was about 8% higher than that in late sowing. Regarding the herbicide treatments, there were significant differences between the non-treated control and all the other treatments, while the five herbicides did not differ significantly from each other.

3.2.3. Yield Components

Regarding the effects of sowing date, variety, and herbicides on 1000-seed weight, no interaction was observed among the factors (

Table 7. Effects of variety, sowing date, and herbicides on 1000-seed weight (g) and spike length (cm) of durum wheat. Treatments for each factor are explained in Table 2.

Factors	1000-Seed Weight (g)		Factor/Interactions		Spike Length (cm)
	2020–2021	2021–2022			2020–2021	2021–2022
Variety			Variety × Sowing date
Levante	42.62 ± 4.98 b	42.31 ± 2.60 b
Simeto	57.84 ± 4.97 a	56.17 ± 3.39 a	Levante	Early	9.90 ± 0.51 a	9.09 ± 0.61 a
Sowing date				Late	9.08 ± 0.51 b	8.27 ± 0.67 b
Early	53.01 ± 9.14 a	49.93 ± 8.02 a	Simeto	Early	7.43 ± 0.50 c	7.30 ± 0.53 c
Late	47.45 ± 8.32 b	48.55 ± 7.17 b		Late	7.11 ± 0.42 d	7.19 ± 0.50 c
Herbicides			Herbicides
Non-treated control	49.42 ± 9.22 b	46.45 ± 6.59 b	Non-treated control		8.53 ± 1.34	7.93 ± 1.21
Hand-weeded control	51.17 ± 9.89 ab	49.84 ± 6.96 a	Hand-weeded control		8.47 ± 1.29	7.76 ± 0.93
metsul/bensul	50.60 ± 9.94 ab	48.88 ± 8.31 a	metsul/bensul		8.61 ± 1.43	8.11 ± 0.66
amin/floras	53.52 ± 8.70 a	50.34 ± 8.89 a	amin/floras		8.14 ± 1.29	8.33 ± 0.98
pyroxs + 2,4-D/floras	47.63 ± 8.40 b	48.39 ± 7.74 ab	pyroxs + 2,4-D/floras		8.28 ± 1.33	7.75 ± 0.94
2,4-D/bromox	49.04 ± 9.29 b	50.45 ± 8.24 a	2,4-D/bromox		8.27 ± 1.03	8.01 ± 0.92
halaux/floras	-	50.31 ± 7.26 a	halaux/floras		-	7.87 ± 1.12
F values			F values
FVariety(V)	259.073 ***	565.28 ***	FVariety(V)		401.506 ***	134.526 ***
FSowing Date(S)	34.605 ***	5.631 *	FSowing Date(S)		26.108 ***	14.162 ***
FHerbicides(H)	3.078 *	3.593 **	FHerbicides(H)		1.749	1.578
FV × S	0.746	2.396	FV × S		5.082 *	8.048 **
FV × H	0.216	0.629	FV × H		0.769	0.958
FS × H	0.331	1.986	FS × H		2.186	1.681
FV × S × H	1.255	0.591	FV × S × H		0.366	0.620

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters indicate statistically significant differences according to Duncan’s post hoc test (herbicide factor and variety × sowing date interaction) and analysis of variance (variety and sowing date factors). The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***).

). However, each of the three factors individually had a statistically significant effect in both growing seasons (2020–2021 and 2021–2022). The Simeto variety had statistically significantly higher values (57.84 g and 56.17 g in 2020–2021 and 2021–2022, respectively) for this parameter than the Levante variety (42.62 g and 42.31 g in 2020–2021 and 2021–2022, respectively). In early sowing, the 1000-seed weight was 2.8–10.5% higher than that in late sowing. Among the herbicide treatments, the lowest value was observed in the non-treated control.

For the spike length parameter, an interaction was observed between the variety and sowing date, while there were no significant differences among the herbicide treatments in both growing seasons. Specifically, spike lengths ranged from 7.11 to 9.90 cm and from 7.19 to 9.09 cm in the first and second growing seasons, respectively. The spike length in the Levante variety was about 8.3–9% higher in early sowing compared to late sowing, while in the Simeto variety, the differences between the two levels of sowing date were smaller (1.5–4.3%).

3.2.4. Seed Yield

In the first growing season, no interaction was observed among the factors for seed yield (

Table 8. Effects of variety, sowing date, and herbicides on seed yield (kg ha⁻¹) of durum wheat. Treatments for each factor are explained in Table 2.

Seed Yield (kg ha−1)
Factors	2020–2021	Factors/Interactions		2021–2022
Variety		Variety × Sowing date
Levante	4659.1 ± 623.5 b
Simeto	5181.2 ± 630.5 a	Levante	Early	5243.1 ± 824.2 b
Sowing date			Late	5528.2 ± 780.7 ab
Early	5116.3 ± 644.5 a	Simeto	Early	5866.1 ± 1149.3 a
Late	4724.0 ± 657.2 b		Late	5401.6 ± 904.5 ab
Herbicides		Herbicides
Non-treated control	3901.1 ± 371.3 c	Non-treated control		4120.2 ± 401.6 c
Hand-weeded control	5332.6 ± 541.5 a	Hand-weeded control		5848.2 ± 679.0 ab
metsul/bensul	4850.5 ± 523.7 b	metsul/bensul		5390.9 ± 791.5 b
amin/floras	5015.7 ± 445.2 b	amin/floras		5335.4 ± 815.4 b
pyroxs + 2,4-D/floras	4971.2 ± 396.4 b	pyroxs + 2,4-D/floras		5469.2 ± 562.7 b
2,4-D/bromox	5450.0 ± 494.4 a	2,4-D/bromox		6092.4 ± 629.5 a
-	-	halaux/floras		6312.1 ± 807.6 a
F values		F values
FVariety(V)	40.12 ***	FVariety(V)		2.960
FSowing Date(S)	22.65 ***	FSowing Date(S)		0.386
FHerbicides(H)	29.548 ***	FHerbicides(H)		14.075 ***
FV × S	0.079	FV × S		6.752 *
FV × H	0.471	FV × H		1.277
FS × H	0.196	FS × H		1.063
FV × S × H	0.636	FV × S × H		0.401

Values are presented as the mean of three replicates ± standard deviation. Mean values followed by different letters have statistically significant differences according to Duncan’s post hoc test [herbicide factor (2020–2021 and 2021–2022) and interaction variety × sowing date (2021–2022) and analysis of variance (variety and sowing date factors: 2020–2021)]. The absence of letters indicates non-significant statistical differences. Asterisks indicate significance at p < 0.05 (*) and p < 0.001 (***).

). However, each of the three factors (variety, sowing date, and herbicides) individually had a statistically significant effect. The Simeto variety had statistically significantly higher values (5181.2 kg ha⁻¹) than the Levante variety (4659.1 kg ha⁻¹), while in early sowing, the seed yield was 7.67% higher than that in late sowing. Among the herbicides, 2,4-D + bromoxynil (5450 kg ha⁻¹) and the hand-weeded control (5332.6 kg ha⁻¹) treatments had the highest seed yield values compared with all the other herbicide treatments. Statistically lower yields were found in the herbicides aminopyralid + florasulam, pyroxsulam, florasulam + 2,4-D, and metsulfuron-methyl + bensulfuron-methyl, with no significant differences among them. Finally, the non-treated control had about 26.8% lower seed yield than the hand-weeded control treatment.

In the second growing season, an interaction between sowing date and variety was observed for seed yield. The highest yield was recorded in the Simeto variety with early sowing (5866.1 kg ha⁻¹), followed by Levante with late sowing (5528.2 kg ha⁻¹), Simeto with late sowing (5401.6 kg ha⁻¹), and Levante with early sowing (5243.1 kg ha⁻¹). Concerning the herbicide factor, it was observed that the herbicides halauxifen-methyl/florasulam, 2,4-D/bromoxynil, and the hand-weeded control had statistically higher yields, followed by the herbicide treatments pyroxsulam + florasulam/2,4-D, metsulfuron-methyl/bensulfuron-methyl, and aminopyralid/florasulam. The non-treated control had the lowest yield (4120.2 kg ha⁻¹), with statistically significant differences from all other treatments.

3.2.5. Seed Quality Characteristics

In both growing seasons, the seed quality characteristics did not show statistically significant differences with respect to variety, sowing time, or herbicide application (

Table 9. Effects of variety, sowing date, and herbicides on protein, dry gluten, and wet gluten content (%) in durum wheat seeds. Treatments for each factor are explained in Table 2.

Factors	Protein Content (%)		Dry Gluten (%)		Wet Gluten (%)
	2020–2021	2021–2022	2020–2021	2021–2022	2020–2021	2021–2022
Variety
Levante	13.68 ± 1.42	13.36 ± 0.91	11.12 ± 1.28	10.51 ± 0.77	31.48 ± 3.58	29.87 ± 2.38
Simeto	13.54 ± 1.29	12.94 ± 1.06	11.04 ± 0.99	10.29 ± 1.25	30.97 ± 2.91	29.09 ± 2.82
Sowing date
Early	13.72 ± 1.26	13.22 ± 1.05	10.99 ± 1.10	10.50 ± 1.13	31.26 ± 3.29	29.65 ± 2.90
Late	13.51 ± 1.43	13.09 ± 0.97	11.16 ± 1.18	10.30 ± 0.94	31.19 ± 3.25	29.32 ± 2.35
Herbicides
Non-treated control	13.90 ± 1.56	12.79 ± 1.16	11.29 ± 1.11	9.98 ± 0.87	31.92 ± 2.95	28.35 ± 2.91
Hand-weeded control	13.16 ± 1.26	13.05 ± 0.75	10.85 ± 1.16	10.65 ± 0.58	31.05 ± 4.03	30.48 ± 1.45
metsul/bensul	13.70 ± 1.49	13.27 ± 1.06	10.97 ± 1.41	10.53 ± 1.42	31.10 ± 3.09	29.80 ± 2.94
amin/floras	13.98 ± 1.44	13.65 ± 0.88	11.38 ± 0.86	11.04 ± 1.19	31.97 ± 2.90	30.91 ± 2.67
pyroxs + 2,4-D/floras	13.33 ± 1.70	13.16 ± 0.56	10.90 ± 1.45	10.25 ± 0.51	30.24 ± 3.90	28.86 ± 1.35
2,4-D/bromox	13.62 ± 1.02	13.54 ± 1.21	11.06 ± 0.81	10.64 ± 0.99	31.07 ± 2.80	30.09 ± 2.88
halaux/floras	-	12.61 ± 1.06	-	9.72 ± 1.02	-	27.87 ± 2.67
F values
FVariety(V)	0.210	3.727	0.095	0.905	0.508	1.882
FSowing Date(S)	0.450	0.374	0.394	0.815	0.008	0.331
FHerbicides(H)	0.716	1.670	0.443	2.215	0.529	2.249
FV × S	3.454	1.083	3.327	1.283	2.948	0.671
FV × H	1.663	0.530	2.005	0.886	1.176	0.628
FS × H	1.596	1.253	1.085	0.570	2.383	0.927
FV × S × H	0.828	0.155	0.697	0.179	1.631	0.177

Values are presented as the mean of three replicates ± standard deviation. The absence of letters indicates non-significant statistical differences.

). No interactions were observed between the experimental factors. Specifically, the two varieties had protein content ranging from 12.94% to 13.68%, dry gluten content from 10.29% to 11.12%, and wet gluten content from 29.09% to 31.48%. Moreover, in early sowing, a slightly higher protein content was recorded than in late sowing. Regarding herbicide treatments, the protein content ranged from 12.61% to 13.98%, dry gluten content from 9.72% to 11.38%, and wet gluten content from 27.87% to 31.97%. It is essential to note that high weed control in the hand-weeded, 2,4-D/bromoxynil, and halauxifen-methyl/florasulam treatments did not improve the quality characteristics.

4. Discussion

4.1. Effects of Variety, Sowing Time, and Herbicides on Weed Parameters

In recent years, several populations of false cleavers and other broadleaf [e.g., bastard cabbage (Rapistrum rugosum L. All.), catchweed bedstraw, corn poppy (Papaver rhoeas L.), kochia (Bassia scoparia (L.) A.J. Scott, wild radish (Raphanus raphanistrum L.)] or grass weed species [e.g., loose silky-bent (Apera spica-venti (L.) P. Beauv.), rigid ryegrass (Lolium rigidum Gaud), and wild oat (Avena fatua L.)] evolved resistance to herbicides of various chemical groups [e.g., shulfonylureas, triazolopyrimidines (acetolactate synthase inhibitors), phenylpyrazolines, aryloxyphenoxy-propionates (acetyl-CoA carboxylase (ACCase) inhibitors), and benzoic acid (auxin mimics)] that are applied in cereal crops, reducing their productivity and causing high economic losses [29,30,31,35,36,37,38,39]. Minimizing wheat crop production loss is crucial, as it directly impacts global food sustainability [40]. Therefore, developing integrated management programs for false cleavers and other weed species in winter cereal crops is essential. These programs should combine various cultivation methods (e.g., sowing date and cultivar choice) with chemical control to improve weed management and minimize crop yield losses due to weed competition.

In the field experiments conducted over two growing seasons, valuable data were recorded on the impact of variety, sowing time, and herbicides on the density and dry weight of false cleavers, as well as on the total weed density and dry biomass. The data analysis revealed no statistically significant differences between the two varieties (Levante and Simeto) in terms of the dry weight of false cleavers and the total weed dry biomass. However, it is important to highlight that there are some studies that report differences in competitiveness between wheat varieties. Mennan and Zandstra [22] found that the lowest values of catchweed bedstraw biomass were recorded in the Bezostaja variety of common wheat, showing that this variety presents a highly competitive ability against this Galium species. In another study, De Vita et al. [41] found that when durum wheat was sown at a row spacing of 25 cm, the lowest values of weed biomass were recorded in the Cappelli variety compared to the PR22D89 variety, which recorded a significantly lower height, revealing that this variety exhibits lower competitive ability against weeds. The height trait is linked to the ability of wheat plants to compete with weeds for light [42]. In our results, the lack of reduction in the dry weight of false cleavers and other weeds in the Levante variety, which exhibited the highest plant height, is attributed to its slower growth during the early growth stages compared to Simeto. The importance of greater early plant growth in wheat for improving its competitive ability against weeds has also been noted by Lemerle et al. [43], who reported that traits such as greater initial growth of plants and high tillering ability are associated with greater competitive ability in wheat. Additionally, Lazzaro et al. [44] observed differences among 160 common wheat accessions regarding the trait of early growth, as some accessions had higher aboveground biomass values before the stem elongation stage than others.

Regarding the effect of sowing time on weed parameters, our results showed no statistically significant differences between early and late sowing for the density of false cleavers. However, the highest total weed density was found in early sowing. This result shows that the density of other weed species observed in our experiments decreased in late sowing compared to early sowing, which aligns with the findings of Klauk and Petersen [45] and Karkanis et al. [10], who observed a decrease in the density of black-grass, milk thistle, and wild mustard in the delayed sowing compared to the early sowing in winter cereal crops. It is important to note that the delay in sowing that did not affect the false cleavers density is related to the base temperature required for the germination of its seeds. De Roo et al. [46] found that the base temperature for the germination of false cleavers seeds, which were collected at different locations in Canada, was 2 °C, while the optimum temperature for the germination of 50% of the seeds was 6.4 °C. In contrast, Liava et al. [47] observed no germination of milk thistle seeds at 5 °C in three populations (Mesopotamia, Palaionterveno, and Spata) collected from three sites in Greece. Therefore, for species like false cleavers, where the base temperature for seed germination is quite low, late sowing does not help reduce density. In contrast to the density parameter, sowing time affected the dry weight of false cleavers and the total weed dry weight, with the highest values noted for the early sowing date. Specifically, the dry weight of false cleavers was 17.3–23.4% higher on the early sowing date than on the late sowing date. However, to the best of our knowledge, limited data are available on the effects of sowing date on the growth of Galium weed species. In a previous study, Aziz et al. [48] observed that the dry biomass of catchweed bedstraw was reduced by up to 42.5% in late sowing (23 November) compared to early sowing (7 November).

In this study, we also found that the highest efficacy against false cleavers was recorded for the herbicides 2,4-D/bromoxynil (73.3% to 94.6%) and halauxifen-methyl/florasulam (92.8% to 95.2%), followed by aminopyralid/florasulam (52% to 66%). It is essential to note that the herbicides 2,4-D/bromoxynil (73.3% to 94.6%) and halauxifen-methyl/florasulam showed greater efficacy in the late-sowing plots in both growing seasons, while aminopyralid/florasulam provided higher efficacy against false cleavers in the late-sowing date only in 2021–2022. This improvement might be attributed to the smaller growth of false cleavers at the time of spraying. In line with our findings, Askew et al. [49] found that the herbicide halauxifen-methyl had 10% greater efficacy against the Canadian horseweed (Conyza canadensis L.) when applied to plants with shorter growth (average plant height 5 cm) compared to plants with greater growth (average plant height 15 cm), while Karkanis et al. [10] noted 3.4% to 7.6% greater efficacy of post-emergence herbicides florasulam/clopyralid and tribenuron-methyl against wild mustard and milk thistle in durum wheat crop when applied in the late planted plots compared to the early planted plots. In addition, the herbicide metsulfuron-methyl/bensulfuron-methyl and the mixture of herbicides pyroxsulam and florasulam/2,4-D showed low efficacy (12.5% to 17.6%) against this resistant population of false cleavers to ALS-inhibiting herbicides, which agrees with our previous study conducted at two locations [29]. However, for all herbicides applied in both growing seasons, higher total weed dry weight values were recorded in the early sowing plots than in the late sowing plots. A similar trend was observed for total weed dry weight in the non-treated control, suggesting that sowing date affected the growth of weed species other than false cleavers. In line with our findings, Khasraw et al. [7] noted statistically significant values of weed dry weight with the untreated control in early sowing (20 November) compared to delayed sowing carried out 15 days later, regardless of the seeding rate used for common wheat establishment.

4.2. Effects of Variety, Sowing Time, and Herbicides on Crop Parameters

The results of this study showed that the plant height of durum wheat was not affected by the herbicide factor, while an interaction was observed between variety and sowing time, with the highest values found in the Levante variety compared to the Simeto variety. The difference in plant height between the two varieties was greater in early sowing (8.6% and 9.8% in 2020–2021 and 2021–2022, respectively) than in late sowing (3.7% and 5.4% in 2020–2021 and 2021–2022, respectively). In another study, Khasraw et al. [7] observed that late sowing of common wheat on 30 December with a seeding rate of 160 kg ha⁻¹ in non-treated control reduced the plant height by up to 19.4% compared to early sowing on 20 November with the same seeding rate and herbicide treatment. Additionally, the lowest values of shoot dry weight were found in the non-treated control plots compared to the other herbicide treatments, as well as in early sowing compared to late sowing. The dry weight values were also 3.8 to 11.5% higher in the Simeto variety than in the Levante variety. False cleavers and other weed interference in the untreated control plots decreased dry weight of durum wheat by up to 20.1% compared to the hand-weeded control treatment, as a result of the reduction in photosynthetic rates of durum wheat plants. In both growing seasons, a significant negative correlation (

Table 10. Correlation analysis (Pearson correlation coefficient, r; two-tailed test; n = 72 and n = 84 in 2020–2021 and 2021–2022, respectively) between the main crop and weed parameters. Asterisks indicate significance at p = 0.05 (*), p = 0.01 (**), and p = 0.001 (***). Parameters: photosynthetic rate (PR), transpiration rate (TR), stomatal conductance (SC), plant height (PH), crop dry weight (CDW), 1000-seed weight (SW), seed yield (SY), protein content (PC), dry gluten content (DGC), wet gluten content (WGC), false cleavers dry weight (FCDW), and total weed dry weight (TWDW).

Parameters	PR	TR	SC	PH	CDW	SW	SY	PC	DGC	WGC	FCDW	TWDW
PR	1	0.686 ***	0.661 ***	−0.197	−0.024	0.194	0.495 ***	−0.087	−0.037	−0.046	−0.337 **	−0.445 ***
TR	0.786 ***	1	0.712 ***	−0.268 *	0.017	0.318 **	0.459 ***	−0.119	−0.064	−0.086	−0.349 **	−0.536 ***
SC	0.832 ***	0.815 ***	1	−0.312 **	−0.074	0.300 *	0.349 **	−0.151	−0.113	−0.168	−0.233 *	−0.361 **
PH	−0.248 *	−0.160	−0.423 ***	1	0.652 ***	−0.034	0.165	0.019	−0.098	0.038	0.021	0.060
CDW	0.467 ***	0.317 **	0.475 ***	−0.383 ***	1	0.296 *	0.433 ***	0.053	−0.055	−0.027	−0.133	−0.138
SW	0.307 **	0.320 **	0.498 ***	−0.772 ***	0.463 ***	1	0.435 ***	−0.018	−0.045	−0.013	−0.033	−0.030
SY	0.523 ***	0.473 ***	0.460 ***	−0.233 *	0.372 ***	0.292 **	1	−0.115	−0.108	−0.118	−0.508 ***	−0.633 ***
PC	0.095	0.059	0.018	0.098	0.032	−0.282 **	−0.101	1	0.944 ***	0.864 ***	0.093	0.150
DGC	0.057	−0.041	−0.022	0.034	0.091	−0.137	−0.187	0.810 ***	1	0.875 ***	0.023	0.087
WGC	0.128	0.082	0.068	0.065	0.074	−0.206	−0.149	0.914 ***	0.893 ***	1	−0.024	0.073
FCDW	−0.506 ***	−0.477 ***	−0.420 ***	0.096	−0.302 **	−0.164	−0.563 ***	0.018	−0.013	−0.064	1	0.852 ***
TWDW	−0.646 ***	−0.589 ***	−0.520 ***	0.169	−0.349 ***	−0.171	−0.620 ***	−0.072	−0.097	−0.128	0.881 ***	1

The italicized values show the r values for 2021–2022.

) was found between false cleavers dry weight and photosynthetic rate (r = −0.337, p = 0.01 and r = −0.506, p = 0.001 in 2020–2021 and 2021–2022, respectively) and between total weed dry weight and photosynthetic rate (r = −0.445, p = 0.001 and r = −0.646, p = 0.001 in 2020–2021 and 2021–2022, respectively). In line with the findings of this study, Karkanis et al. [10] noted that competition between milk thistle, wild mustard, and some other broadleaf weeds on durum wheat decreased the dry weight of this crop. In contrast, Mennan and Zandstra [22] observed that the dry weight of common wheat was not affected by catchweed bedstraw competition, regardless of the cultivated variety.

The 1000-seed weight ranged from 42.31 to 47.84 g and was up to 26.3% lower in the Levante variety than in Simeto, while the delay in sowing date resulted in lower values for this yield component. A similar trend was observed for spike length, as its lowest values were recorded in the delayed sowing. In this context, in a 3-year experiment carried out in arid areas of China, Wu et al. [50] found that delayed sowing of winter wheat for 20 days and without plastic mulching application decreased the 1000-seed weight and number of seeds per spike up to 10.4% and 8.1%, respectively, while in another study, Khasraw et al. [7] noted that early sowing of wheat on 20 November with a seeding rate of 160 kg ha⁻¹ increased the 1000-seed weight compared to delayed sowing for 15 days with the same seeding rate, regardless of weed control level. Additionally, the 1000-seed weight values in the non-treated control plots due to the absence of weed competition were 6,8% lower than those recorded in the weed-free control plots. However, spike length was not affected by false cleavers and other weed competition, as the results of this study showed no differences in this parameter between the non-treated control and all the other herbicide treatments. In contrast, milk thistle at densities of 5 plants m⁻² or greater caused a significant decrease in spike length in wheat cultivation [51].

The results of this study also showed that all experimental factors affect the seed yield in durum wheat. Weed competition in the non-treated control plots caused up to 34.7% lower seed yield compared to highly effective herbicides (halauxifen-methyl/florasulam and 2,4/bromoxynil). A significant negative correlation (r = −0.633, p = 0.001 and r = −0.620, p = 0.001 in 2020–2021 and 2021–2022, respectively) was observed between seed yield and total weed dry weight. Moreover, the yield gap between halauxifen-methyl/florasulam or 2,4/bromoxynil and all the other herbicides that provided low efficacy against false cleavers revealed that interference between durum wheat and false cleavers resulted in a significant decrease in seed yield in this crop. In this context, a significant negative correlation (r = −0.508, p = 0.001 and r = −0.563, p = 0.001 in 2020–2021 and 2021–2022, respectively) was observed between the seed yield and dry weight of false cleavers. In another study, the yield loss of common wheat due to catchweed bedstraw (G. aparine) competition ranged from 9.1% to 48.8%, and differences in the competitive ability of four varieties against this weed species were observed [22]. These results show that this broadleaf weed exhibits extensive competitive ability against wheat. In this context, Rehman et al. [51] noted that milk thistle at densities of 5 plants m⁻² or greater caused a significant decrease in seed yield in wheat cultivation. Additionally, our results showed that Simeto had greater productivity than Levante, except at late sowing in 2021–2022, and the highest seed yield values were observed in early sowing, with the exception of the Levante variety in 2021–2022. Other studies have also shown a significant decrease in the seed yield of wheat crops in late sowing compared to early sowing [7,14]. However, Karkanis et al. [10] noted that in the weed-infested control, the highest seed yield in durum wheat was observed in late sowing compared to early sowing due to lower weed infestation.

Regarding the quality traits of durum wheat, the results of this 2-year experiment showed that none of the factors affected the protein and dry or wet gluten content. It is essential to note that the interference of false cleavers and other weed species with durum wheat plants had no impact on quality traits (e.g., protein content). Similarly, in an experiment conducted in Italy, De Vita et al. [41] noted that weed interference in the untreated control did not affect the protein content of two durum wheat varieties (Cappelli and PR22D89). In contrast, Akhter et al. [52] observed that the rattail fescue (Vulpia myuros (L.) C.C. Gmel.) interference with winter wheat decreased the protein content in 2017–2018 only when the crop was sown early (10 October). These results show that weed species vary in their competitive ability and consequently, their effects on wheat quality traits. Additionally, in our study, there were no differences in protein content (12.94% to 13.68%) between the varieties Simeto and Levante. However, differences in protein content between varieties of durum wheat have been observed in other studies. For example, De Vita et al. [41] found that the protein content was higher in the Cappelli variety than in the PR22D89 variety, and Graziano et al. [53] recorded that the grain protein content was 9% higher in the Svevo variety than in Iride. These results show that genetic factors affect the protein content of durum wheat seeds. Moreover, a strong and positive correlation (r > 0.80, p = 0.001) was found between the protein content and dry or wet gluten content in both growing seasons. Similarly, Matzen et al. reported a high correlation coefficient (r = 0.90) between protein and gluten content in winter wheat seeds [54].

5. Conclusions

The field experiment conducted over two growing seasons provided valuable insights into the integrated management of false cleavers in durum wheat cultivation by combining cultivation practices with herbicide applications. Our results showed that the density of false cleavers was not affected by either variety or sowing time, while its dry weight was affected by sowing time, with higher values recorded for early sowing. Additionally, the total weed density and weight values were higher in early sowing than in late sowing, clearly revealing that sowing time can affect the density and growth of specific weed species.

Among the herbicides applied, 2,4-D/bromoxynil and halauxifen-methyl/florasulam effectively controlled false cleavers, with better results in late sowing than in early sowing. Additionally, competition from false cleavers and other weeds negatively affected the physiological parameters of the crop; however, it did not affect the quality characteristics of durum wheat seeds. No interaction was found between the three factors for durum wheat dry weight. The Simeto variety had the highest dry weight, while late sowing resulted in a lower dry weight compared with early sowing. Simeto had shorter spikes, with no differences between the sowing time treatments for this variety, while the Levante variety had the lowest 1000-seed weight, with early sowing resulting in higher values. Weed competition also affected this parameter, with the lowest values recorded in the non-treated control. Seed yield was influenced by all three of these factors. The highest yield values were recorded in the hand-weeded control, 2,4-D/bromoxynil, and halauxifen-methyl/florasulam (5332.6–5450 kg ha⁻¹ and 5848.2–6312.1 kg ha⁻¹ in 2020–2021 and 2021/2022, respectively), with no statistically significant differences among them.

Overall, the optimization of cultivation practices in the context of integrated weed management programs can lead to improved management of false cleavers and other weed species in durum wheat cultivation. Considering that climate change affects agroecosystems, more experiments need to be conducted in various locations with diverse climate conditions to collect more data that will help develop effective integrated management systems for various weed species.