1. Introduction
Legumes are the second most economically important crop in worldwide agriculture after cereals and are grown for both forage and grain [
1]. In the past, they were understated in European cropping systems because of cereals and non-legume oilseed dominance. However, the great call for high-protein materials for livestock feed, the need for a reduction in European dependence on imported protein, and the opportunities to use legumes in new foods brought legumes back to the forefront of public debate in Europe [
2].
It is currently acknowledged that grain legumes can be a popular choice in farming systems since they can contribute to the nutritional security and resilience of agricultural ecosystems [
3]. As a valuable source of vitamins, minerals, protein, and dietary fibers, they play a vital role in human and animal diets [
4,
5,
6]. Moreover, grain legumes provide ecosystem services through their biological nitrogen fixation capacity, improving physical soil properties and enriching soil with N [
7,
8]. Thus, they contribute to the reduction in overdependence on inorganic nitrogen fertilizers. Due to their cultivation mostly in marginal environments, they can also survive in problematic soils and resist abiotic stresses [
9].
The underlined significance of grain legumes led the European Union (EU) to undertake various support measures [
2]. Nowadays, grain legumes are cultivated in an area of about 81 million ha and produce more than 92 million tons universally [
10]. Among grain legumes, peas (
Pisum sativum L.), faba beans (
Vicia faba L.), vetch (
Vicia sativa L.), and white lupine (
Lupinus albus L.) are considered important crops for several reasons. Peas are the most widely cultivated grain legume in Europe [
11]. They are mainly grown for their green pods and consumed as fresh vegetables and cooked green seeds. Faba beans are the fourth most significant cool-season grain legume and are broadly preferred as a green vegetable, while in several countries they are used as feed. Moreover, the production of faba beans has increased yearly by 2% during the past three decades. Concerning the vetch crop, another cool-season member of the Leguminosae family, it is mainly used as animal feed and is also studied for its seeds as a promising new source of starch and a sustainable source of food for humans [
12]. White lupine, a multifunctional legume crop, is cultivated for a broad range of uses, from forage for livestock and food for humans to medicinal and pharmaceutical uses [
13].
Even though legume production has been rising globally [
14], grain legumes are hardly ever chosen by farmers for cultivation as compared to other crops [
2]. The main reason for this fact is that grain legume yields are more variable against biotic and abiotic stresses than those of cereals, presenting yield fluctuations due to their indeterminate growth habit and the relative absence of investment in breeding for stress-resistant cultivars [
2]. These yield losses would be partially attributed to the lack of farming practices capable of withstanding biotic stresses such as weed infestation.
Grain legumes present a poor competitive ability against grasses and broadleaf weeds [
15]. Due to the relatively slow establishment after sowing and the insufficient soil surface coverage shaped by small tendrils and leaflets, grain legumes are unable to effectively suppress weed growth [
15]. Thus, weeds pose a serious threat to the sustainability of grain legumes since they strongly compete with the crop for water, sunlight, and nutrients [
16].
Given that weed control management is a major challenge for grain legume production, several cultural, mechanical, and chemical practices have long been studied. Increased crop density improved weed control in vining peas [
17] and faba beans [
18], which significantly raised the seeding cost of the crop. For faba beans, the combination of row spacing and mechanical control led to a similar level of weed control in comparison to herbicide application, while for field peas, some herbicides could be used alongside these other methods [
16]. Another notable strategy for optimizing weeding in grain legumes is the use of competitive cultivars against weeds, especially in farming systems where herbicides are avoided [
2].
In most grain legume cropping systems, chemical control is mostly preferred by farmers since herbicides are a more reliable, effective, and profitable method of controlling different weed species [
19]. In grain legumes, it is well documented that the most important type of herbicide is the pre-emergence (PRE) one [
20]. However, PRE herbicides are significantly dependent on rainfall soon after application, and therefore, in semi-arid conditions (i.e., Mediterranean conditions), frequently inconsistent or partial weed control is documented [
20]. The post-emergence selective (POST) herbicides are mostly used to control broadleaf weeds; only a few are registered in Europe for a small range of weed species, officially registered for a limited number of crops; they are the following: aclonifen, bentazone, and imazamox [
21].
While the benefits of grain legumes for the nutritional security and resilience of agricultural ecosystems are underlined, the best cultivar practices, including weed management, are urgently needed to support high yields and enhanced profitability of the crops. To the best of our knowledge, the design of a weed management program for these crops is still in its infancy, as little information exists, and there is also a lack of registered herbicides for these crops. In this context, the need for research aiming to identify potential herbicides becomes even more imperative in environments affected by large fluctuations in temperature and precipitation. Climatic factors can affect herbicide efficacy and crop safety [
20]. Therefore, environments such as the Mediterranean are of high scientific interest for the study of herbicide efficacy and selectivity in grain legume crops.
In this framework, there is a great need to identify herbicides (single or ready-mixed ones) that would provide broadleaf weed control with adequate crop safety. Therefore, this study aimed to (a) measure the efficacy of seven PRE and two POST herbicides against broadleaf weeds and (b) assess the selectivity of these herbicides on the four grain legumes (i.e., peas, faba bean, vetch, and white lupine). In this study, the effects of herbicide treatments on weeds and legumes were studied under semi-arid Mediterranean conditions during two growing seasons.
4. Materials and Methods
4.1. Experimental Site
Field experiments were conducted in Central Greece (latitude 39°18′25.28″ N, longitude 22°7′30.19″ E, altitude 131 m above sea level) for two consecutive growing seasons: November–June 2019–2020 and November–June 2020–2021. The soil was sandy loam (sand 36%, clay 24%, and loam 40%) with a pH value of 7.7, organic matter 1.26%, and CaCO
3 10.9%. Weather data were collected daily from the meteorological station located near the experimental area. Rainfall and average, minimum, and maximum air temperatures were reported as monthly mean data for the growing season (
Figure 2). The weather conditions during the growing seasons GS1 and GS2 were quite different. The average air temperatures during the growing periods did not differ greatly. The total rainfall was higher from November 2019 to July 2020 than in the second growing season. However, the total rainfall of both growing seasons remained lower than the historical mean annual rainfall at the site of 310 mm. These climatic conditions are typical of the Mediterranean basin, where winters are cold and rainy.
4.2. Plant Materials and Experimental Design
Grain legumes, pea (
Pisum sativum L.), faba bean (
Vicia faba L.), vetch (
Vicia sativa L.), and lupine (
Lupinus albus L.), which are widely grown in Greece, were selected for the experiments. The specific characteristics of the four commercial cultivars are as follows: cv. Olympos is an early pea cultivar (Institute of Industrial and Fodder Plants, Larissa, Greece), cv. Tanagra is a mid-late faba bean cultivar (Institute of Industrial and Fodder Plants, Larissa, Greece), cv. Evinos is an early vetch cultivar (Agroland SA Company, Karditsa, Greece), and cv. Multitalia is an early lupine cultivar (Agroland SA Company, Karditsa, Greece). A randomized complete block (RCB) design with three replications was arranged for each species. The experiments included ten treatments: untreated weedy check (control), application of seven different pre-emergence (PRE) and two post-emergence (POST) herbicides, and herbicide mixtures. The weedy check (control) plots received no weed management treatment. The description and doses of the herbicides tested are summarized in
Table 7.
All herbicides were sprayed once at the recommended doses with a local-made precision sprayer utilizing an air-pressure system, equipped with flat-fan nozzles (XR TeeJet®, TeeJet® Technologies, Glendale Heights, IL, USA), calibrated to deliver 350 L ha−1 at 300 kPa. The spraying boom was approximately 50 cm above the plant canopy. The spraying environmental conditions were ideal (no wind, no rain after the application) for the experiment. Pre-emergence herbicides were applied in the period after sowing and before the emergence of grain legume crops. Post-emergence herbicides were applied when most of the weeds were at the 3–4-leaf stage. Water was applied to the untreated control in the same manner as in the other treatments. The total size of each plot was 4 m in length and 1 m in width, and each plot consisted of 4 rows spaced 0.25 m apart.
4.3. Crop Management
Seedbed preparation was conducted by plowing at a depth of 20 cm and followed by shallow tillage. Seeds of commercial cultivars were sown separately at a depth of 3 cm on 25 November 2019 and 28 November 2020 to achieve the recommended seed rate of 160 kg ha−1 for faba bean and lupine and 112 kg ha−1 for pea and vetch. No basal fertilization or disease and pest control practices were followed. Plants were rainfed without any supplemental irrigation during the growing periods. The four grain legume crops reached seed maturity after the same period for the two consecutive growing seasons and were harvested on 22 June 2020 and 15 June 2021.
4.4. Sampling and Measurements
The experiment aimed to determine the effect of herbicide treatments on overall, specific weed control and legume plant growth and grain yield. The herbicide efficacy and plant phytotoxicity evaluations were performed 4 weeks after each treatment (WAT) and were assessed by visual rating. A scale of 0 to 100 was used for the herbicide efficacy, where 0 depicted no weed control and 100 depicted the total control of weeds. Based on the visual estimations, three broad control scales were developed as follows: high efficacy (>70% control), moderate efficacy (55–69% control), and low efficacy (<54% control). Scaling was based on the overall presence/growth of the weed species in the whole plot. Regarding plant injury level, it was also evaluated visually on a scale of 0 to 100, where 0 represented no injury of grain legumes and 100 represented complete plant necrosis [
43]. More specifically, the classification of crop damage was performed according to the European Weed Research Council (EWRC) standard method as follows: damage or less necrosis (1–3.5%), inconsistent less damage (3.5–7%), average and more consistent damage on crop (7–12.5%), average and consistent damage on crop (12.5–20%), heavy damage on crop (20–30) [
44].
Data on the weed population and dry matter were recorded 62 days after sowing (DAS), when most of the weeds had already emerged. Weeds were sampled from three 0.5 m−2 quadrants placed on the diagonal of each experimental plot. Aboveground weed biomass was collected, identified, and counted, and then it was oven-dried at 70 °C for 48 h (Binder FD023, binder GmbH, Tuttlingen, Germany) to a constant weight to allow the recording of weed dry weight (Kern FCF 30K-3, KERN & Sohn GmbH, Balingen, Germany). The values obtained (means of three quadrants) were converted per unit area for biomass (g m−2).
At harvest time, five plants were randomly selected from the middle rows of each plot, hand-harvested, and threshed using a laboratory thresher (LD350, Wintersteiger, Ried im Innkreis, Austria) in order to assess yield. Seed yield was measured at 13% seed moisture content. After the pods were removed, the aboveground stems and leaves of each harvested plant were cut above the soil, dried at 70 °C for 48 h (Binder FD023, Binder GmbH, Tuttlingen, Germany), and then weighed (Kern FCF 30K-3, KERN & Sohn GmbH, Balingen, Germany) to determine the total dry weight of each plant.
4.5. Statistical Analysis
All measured and derived data were subjected to analysis of variance (ANOVA), using the statistical software package Statgraphics Centurion XVI Version (Statgraphics Technologies, Inc., The Plains, VA, USA). Before analysis, the aboveground biomass weight was transformed as a square root to homogenize the variance. The significance of differences among treatments was estimated using Fisher’s least significant difference (LSD) test where probabilities are equal to or less than 0.05 (a = 5%).
5. Conclusions
Based on the results of the present work, we observed that specific weeds presented a negative effect due to weed–crop competition on the growth and yields of four grain legumes (i.e., vetch, pea, faba bean, and lupine) since plant dry biomass and yields were reduced in the weedy check plots. All the herbicide treatments (i.e., seven PRE and two POST herbicide treatments) decreased the weed presence of the major weed species compared with the untreated plots. Among the dominant weeds, the most easy-to-control weed species were V. arvensis and Sonchus spp., whereas the most difficult-to-control weeds were C. album, S. arvensis, and S. marianum. The best PRE herbicide was flumioxazin, which had the greatest control over major weeds and resulted in the lowest total weed biomass, followed by benfluralin and the mixture of terbuthylazine plus pendimethalin (the latter ones had only limited control of S. arvensis). The best POST herbicide was imazamox, with only limited control of S. arvensis. The herbicides tested caused low to medium levels of phytotoxicity mainly in vetch and secondarily in peas with negligible phytotoxicity in faba bean and lupine. Among the grain legumes, vetch was the most competitive against weeds, whereas peas were the most susceptible. Concerning all weed management treatments, benfluralin resulted in the highest grain yields for all four grain legume crops for both growing seasons. Further studies are needed to document the sustainable use of herbicides in grain legumes, with an emphasis on integration with non-chemical methods to enhance crop profitability and sustainability.