Annual Weeds Suppression and Oat Forage Yield Responses to Crop Density Management in an Oat-Cultivated Grassland: A Case Study in Eastern China

Abstract

Although weeds can be inhibited by high planting densities, canopy shading, elemental balance and soil microbial recruitment are not yet adequately considered when measuring competitive effects on weed control. The effects of oat (Avena sativa) planting density (60 to 600 plants m⁻²) on the biomass and shoot element balance of oat and weeds were evaluated in a field experiment. The shift in the microbial community of the dominant weed species was examined in a pot experiment by growing the weed alone and in competition with 360 oat plants m⁻² (recommended planting density) under greenhouse conditions. Increasing oat planting density beyond 360 plants m⁻² did not improve oat forage yield or weed suppression. Compared to 60 plants m⁻², the biomass of broadleaf and grass weeds decreased by 1122% and 111%, respectively, at a density of 360 plants m⁻², while oat forage biomass increased by 60% and leaf area index by 24%. The improved canopy properties suppressed competing weeds through increased shading. Typically, the C:N and C:P ratios of shoots of Echinochloa crus-galli and Digitaria sanguinalis were higher than those of Portulaca oleracea and Chenopodium album. At high planting densities, E. crus-galli and D. sanguinalis exhibited high P contents and low N:P ratios, suggesting a limited supply of N nutrients for growth. Soil bacterial community assay showed that the composition of microbial communities of the two grass weeds were shaped by the presence of oat competition, which also considerably depleted several important functional microbes associated with nutrient cycling in the weeds’ rhizosphere. These results highlight that increased crop density significantly improves the crop competitive advantage over weeds through increased shading, reduced elemental balance, and beneficial microorganisms of weeds, thereby reducing the need for herbicides or physical weed control in oat cropping system.

1. Introduction

Agricultural weed infestation has been estimated to cause severe yield losses globally [1]. Weeds compete with crops for solar radiation, nutrients, water, and other resources, and this subsequently reduces crop productivity. Therefore, weeds greatly threaten food security [2]. Herbicide resistance in weeds has evolved as a result of the current conventional intense cropping system’s heavy reliance on herbicides, which has also resulted in environmental pollution [3]. However, the competitive strategies employed by crop plants have received little attention. Therefore, as supplements or substitutes for chemical weed control, innovative and ecologically friendly methods of weed management are desperately needed.

Herbicide control alternatives would include improvements to cultural/agronomic practices, like planting density, which could increase a crop’s capacity to compete with weeds [4]. Previous studies showed that agricultural researchers have used high planting densities as weed management strategies, including oat [5], spring wheat (Triticum aestivum) [6], and maize (Zea mays) [7]. Increased crop density may enhance the crop’s initial size-asymmetric competitive advantage over surrounding weeds, allowing the crop to obtain a disproportionately larger share of resources in the early growing season [4]. High density planting produces shading by limiting the amount of sunlight available to weed species; therefore, light becomes a limited resource for weeds growth [8]. Weed species attempt to escape shade stress by exhibiting a known shade avoidance syndrome (SAS), which is associated with elongation growth of stems or petioles to improve light capture. Growth elongation, however, aims to prevent shading at the expense of carbon resource assimilation, which has a detrimental effect on weed growth and fitness [9]. Moreover, weeds are suppressed before they can “catch up” with the crop’s initial size−asymmetric advantage due to increased planting density, which facilitates crop canopy closure and the collective shading of weeds. [10]. Therefore, crop growth would benefit at the expense of weeds due to the radiation-asymmetric advantage in the crop−weed competition [11]. As planting density increases, the radiation interception potential would plateau significantly, and may slowly rise above an optimal density. Hence, investigating the optimum planting density to maximize resource interception and canopy closure in order to improve forage yield and suppress weeds could aid in directing the design of crop cultivation in agroecosystems.

Besides radiation competition, a crop may require additional nutrients to attain high biomass at high planting densities, which in turn compete nutrient resource absorption of neighboring weeds. Nutrients that are necessary for vegetation are nitrogen (N), phosphorus (P), and carbon (C), and together they frequently constrain plant growth and metabolism [12]. The C:N:P stoichiometry emphasizes the elemental interaction and balance in ecological processes, which is used to investigate the feedback and relationships between aboveground and belowground components of the ecosystems [12]. From a stoichiometric perspective, variations in the nutrient contents and stoichiometry of plant tissues indicate how well weed species have adapted to and competed in agroecosystems [13]. In plant tissue, N and P exhibit a close stoichiometric scaling relationship when there is a functional coupling between them, representing the synchronized plant species-specific demand and limitation for N and P nutrient [14]. Plant N and P stoichiometry tend to be flexible to changes in neighbor identity and/or soil resource conditions, suggesting that plant stoichiometry could have a complexity ecological relation in the plant-soil system [15]. Previous studies have found that high N, P-requirement weed species are greatly affected by intense crop competition in nutrient element acquisition compared to low N, P−requirement weeds [16,17,18]. At a high density, interspecific competition led to higher resource limitations for high N, P-requirement weeds, contributing to the greater changes in the elemental ratios of the weeds [17,18]. Weed species with higher competitiveness in nutrient-limited environments can increase their biomass’s nutrition content and nutrient utilization efficiency without dramatic inhibition in their growth and/or reproduction [19]. However, the effects of interspecific competition on crop and weed stoichiometry that depends on species-specific flexibility remains unclear. Understanding how planting density affects plant stoichiometry is essential for diagnosing crop–weed nutrient competitive outcomes and directing weed control strategies in agroecosystems.

Furthermore, crop-weed competition for soil mineral nutrients affects soil microorganisms. Soil microbial communities’ activities and functions can be impacted by variations in soil nutrients (e.g., organic matter, N and P) [20]. The soil microorganisms are important drivers for regulating the nutrient absorption capacity of plants and the nutrient cycling in the soil, which is closely linked to above-ground plant growth [21]. Therefore, in crop-weed competition systems, the interactions between microbes and nutrients may result in changes in the quantity and taxa of microbes, which in turn may affect the microecological environment of the soil [22]. Plant roots selectively recruit rhizosphere microbiota through their root exudates, which are closely linked to specific plant species [23]. Plants respond to the surrounding plants through their growth habits; they also respond to physiological and biochemical approaches, e.g., by alterations in composition of root exudates [24]. These changes would affect the composition of microbial communities in the rhizosphere, potentially inhibiting a number of beneficial functions required by the weed species [25]. In field, the rhizosphere microbial community of weeds can be affected by a crop, and the changes in soil microbial community may be crucial for enhancing weed control. Several studies have investigated the changes in the microbial characteristics of soils caused by interspecific competition [26,27,28]. However, it is not well understood how the rhizosphere microbial community changes at the family level in the interspecific competition, and even less information is available about this in crop/weed competition specifically.

In this study, we investigate the influence of increased crop density on weed suppression in an oat cropping system in Eastern China. Throughout the world, oats are the most extensively cultivated annual cool-season cereal, serving as a crucial source of nutrition for livestock production [5]. Four coexisting weeds, including two grass weeds, i.e., Echinochloa crus-galli and Digitaria sanguinalis and two broadleaf weeds, i.e., Portulaca oleracea and Chenopodium album, were chosen because they are the most prevalent and problematic weeds in the oat field. The objectives of the current study were: (i) to evaluate the effects of planting density on oat forage yield and weeds growth, stoichiometry of shoots for C, N, and P, light environment, as well as soil chemical properties; (ii) to explore the possible effects of environmental factors on variation in weed suppression in oat fields; and (iii) to clarify the shifts in rhizosphere bacterial community composition of weed species with respect to neighbor oat competition in greenhouse conditions. We hypothesized that (1) increasing planting density increased shading and decreased elemental ratios and soil nutrient resources, thus affecting weed biomass; (2) the response of driving variables (e.g., light, soil nutrient, and shoot stoichiometry) to weed growth and nutrients depends on the weed species; (3) the interspecific competition with oat alters soil bacterial community structure and decreases rhizosphere beneficial bacteria associated with soil nutrient cycling of weeds. We aim to provide effective information for improving weed suppression by increasing oat competition.

2. Materials and Methods

2.1. Filed Experiment: Effects of Planting Density on Growth and the Elements Balance of Oat and Weeds

2.1.1. Experimental Site Description

This study was conducted at the experimental fields in Jiaozhou City, Shandong Province, located in the northeast of the North China Plain (36°26′25″ N, 120°4′48″ E; elevation, 1 m). According to a meteorological station (2006-2021) located within 100 m of the experimental site, mean annual temperature is 14 °C and mean annual rainfall is 686 mm. Mean annual frost-free period is 206 d, and mean annual sunshine duration is 2573 h. Monthly meteorological data (e.g., air temperature, rainfall, and humidity) during the 2021 growing season are presented in Table S1. In 2021, average air temperatures and precipitation were close to the average over the past 15 years. There were no major weather fluctuations; thus, weather conditions could not primarily influence the variation in growth and development of oats and weeds and subsequent weed suppression under field conditions. The soil is classified as brown earth. The soil profile’s upper 0–20 cm had a pH of 7.14, and organic C, total N, and total P contents were 10.77, 1.18, and 0.61 g kg⁻¹, respectively. Available P and available N were 6.32 and 36.31 mg kg⁻¹, respectively.

2.1.2. Experimental Design and Field Management

The study was conducted from 16 April to 2 July in 2021. Oat (A. sativa cv. Monika) was sown at 16 April in 2021. The recommended planting density for oat is 360 plants m⁻² in the region. In this study, six planting density treatments were applied to 18 m² (3 × 6 m) plots that were set up in four replications using a completely randomized block design. All plots were manually furrowed with a row spacing of 30 cm and a sowing depth of 3–4 cm. Six oat planting density treatments were used, as follows: (1) 60 plants m⁻² (O1), (2) 120 plants m⁻² (O2), (3) 240 plants m⁻² (O3), (4) 360 plants m⁻² (O4), (5) 480 plants m⁻² (O5), and (6) 600 plants m⁻² (O6). Next, 50 kg ha⁻¹ nitrogen fertilizer was applied as basal fertilizer before sowing. No irrigation was applied, and weeds were uncontrolled. The preceding crop at the experiment site was maize.

2.1.3. Field Sampling and Measurement

In the present study, we focused exclusively on the shoot stoichiometry of weeds and oats at the milky stage, which represents the crucial period that determines the aboveground biomass and quality of oats. On 2 July 2021, five surface soil cores (0 to 20 cm depth) were randomly collected using soil auger from each plot. A composite soil sample for every plot was created by combining the soil cores. The soil samples were allowed to air dry for seven days in a ventilated and dry environment to maintain a constant weight in the laboratory before analyzing soil chemical properties. The soil samples were sieved through 2 mm and 0.25 mm sieves for further measurement.

At the oat milky stage, plant samples, i.e., oat and weeds, were clipped at the soil surface from two 1 m² quadrats centered over the row at each plot, dried in an oven at 65 °C for a week, and then weighed. In the laboratory, all the dried shoots (leaves and stem) were ground to a 0.15 mm diameter with a grinder for subsequent chemical analysis.

The methods used to determine the C and N in the soil and plant shoot were as follows: subsamples (0.1 g of plant tissues and 0.5 g of soil) were weighed for C and N analyses. An elemental analyzer (Elementar, Vario El Cube, Langenselbold, Germany) was used to determine the C and N. Analyzer (AA3, Bran-Luebbe, Hamburg, Germany) was used to measure plant P and soil total P after they were digested in a mixture of trace H₂SO₄-H₂O₂. Sodium bicarbonate (0.5 mol L⁻¹) was used to extract soil-available phosphorus (AP), which was measured photometrically using a spectrophotometer (UV2700, Shimadzu, Kyoto, Japan) at 660 nm. Soil−available N is the sum of NO₃⁻-N and NH₄⁺-N. An autoanalyzer (AA3, Bran-Luebbe, Hamburg, Germany) was used to analyze the soil samples for NO₃⁻−N and NH₄⁺−N after they were shaken with a 2 mol L⁻¹ potassium chloride (KCl) solution [25].

2.1.4. Weed Community and Density

Weed community composition and density was assessed before oat harvest. Weed data were gathered prior to oat harvest by counting the total number of weeds and categorizing them by species using 1 m² squares for each plot.

2.1.5. Measurement of Photosynthetically Active Radiation and Leaf Area Index

A ceptometer (AccuPAR LP-80, Decagon Devices, Pullman, WA, USA) was applied to measure photosynthetically active radiation (PAR) under different row configurations on the harvest day between 11:00 h and 12:00 h. The ceptometer was positioned and oriented in accordance with the protocol suggested by Tang et al. [29], with one reading taken above the canopy and four below it from randomly chosen locations in each plot. The spectral irradiance under different treatment canopy was measured using a HR550 spectrometer (Hipoint Inc., Gaoxiong, Taiwan) accompanied by PAR measurement.

2.2. Greenhouse Experiment: Effects of Oat Interspecific Competition on Weeds Rhizosphere Microbial Communities

2.2.1. Weed Species Collection

According to weed infestation in field, the E. crus-galli and D. sanguinalis gained more biomass and higher flexibility C, N, and P-related stoichiometry than that of broadleaf weeds. In addition, increasing oat planting density above the 360 plants m⁻² could not significantly improve weed suppression and oat forage biomass accumulation. Thus, we focused on the effects of neighboring oat on bacterial communities of E. crus-galli and D. sanguinalis rhizosphere at recommended oat planting density (360 plants m⁻²) rather than all the planting densities. For the recommended planting density (360 plants m⁻²), oat seeds were counted and then used to adjust the amount of seeds from the other weed species, both in monoculture and in mixed culture.

2.2.2. Experimental Design

The experiments were carried out in a greenhouse at the Experimental Greenhouse of Qingdao Agricultural University in Qingdao during the spring-summer season of 2022. Soil was collected from the above-mentioned study area of the experimental site. Plastic pots (30 cm diameter and 30 cm height) were filled with a sterilized 4:1 mixture of field soil and autoclaved vermiculite. The soil mixture profile indicates that the following values are present: pH 5.80 (CaCl₂), TC 14.07 g kg⁻¹, TN 1.41 g kg⁻¹, TP 0.54 g kg⁻¹, AP 6.6 mg kg⁻¹, NO₃⁻−N 19.1 mg kg⁻¹, NH₄⁺-N 34.3 mg kg⁻¹.

2.2.3. Seed Planting and Cultivation

The experimental treatments comprised E. crus-galli (Ec), D. sanguinalis (Ds), and A. sativa (As) mono-culture, and E. crus-galli vs. D. sanguinalis (Ec_Ds) and A. sativa vs. E. crus-galli and D. sanguinalis (As_Ec_Ds) mix-culture. We used a complete randomized block design with five replicates.

For three monoculture treatments, we used six A. sativa, one E. crus-galli, and two D. sanguinalis plants of each species per pot, respectively. E. crus-galli vs. D. sanguinalis treatment consisted of one E. crus-galli plant and two D. sanguinalis plants per pot. A. sativa vs. E. crus-galli vs. D. sanguinalis treatment consisted of six A. sativa plants, one E. crus-galli plant, and two D. sanguinalis plants per pot. The number of weed species in a pot is equal to the corresponding weed density at 360 oat plants m⁻² in field. Seeds were sterilized and hand sown evenly over each container. The sowing was done in May 2022. The temperature was maintained at 20–25 °C, corresponding to the growing season conditions of the region. Plants were watered with normal tap water as needed and fertilized once with ammonium nitrate (equivalent of 50 kg N ha⁻¹ pot⁻¹) in 1000 mL of half-strength Hoagland’s solution before sowing.

2.2.4. Sample Collection

By harvesting the oats just when they reached the almost milky stage, plant growth traits were measured for both the monocultures and the mixtures. For molecular analysis, oat and weed species were pulled from the soil and gently shaken to remove superfluous soil, with root-associated soil collected by shaking the plant roots [25]. A total of 48 samples (16 treatment × 3 repetitions) was established. Soil samples were divided into two subsamples. One subsample was air-dried to analyze soil nutrient, and the other was stored in a −80 °C refrigerator to determine the soil microbial DNA extraction.

2.2.5. Soil DNA Extraction and PCR Amplification

Information about the microbial extraction and determination procedures can be found in our previous study [25]. Briefly, soil genomic DNA was extracted from each sample using a FastDNA^® SPIN Kit (MP Biomedicals, Santa Ana, CA, USA). The bacterial 16S rRNA gene’s V3–V4 regions were amplified using primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GACTACHVGGGTWTC TAAT-3′) to create sequencing libraries. The following conditions applied to the PCR amplification cycle: 3 min of initial denaturation at 95 °C, 27 cycles at 95 °C for 30 s, 30 s of annealing at 55 °C, 45 s of extension at 72 °C, 10 min of single extension at 72 °C, and 4 °C of termination.

2.2.6. Processing of MiSeq Sequencing Data

On an Illumina MiSeq PE300 platform (Illumina, San Diego, CA, USA), paired ends of purified amplicons were sequenced after being pooled in equimolar ratios, which were in accordance with protocols of the Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). QIIME version 1.3.0 was used to process the raw data and diversity indices, such as Shannon and Chao1 richness. Reads were truncated to achieve a quality score of less than 20 on average and a sliding window of more than 50 bp over three consecutive bases; those that were shorter than 300 bp were discarded. After that, high-quality sequences meeting a 97% percent identity threshold were clustered into functional taxonomic units using UPARSE version 7.1 [30].

2.3. Statistical Analyses

SPSS version 17.0 software (SPSS Inc., Chicago, IL, USA) was used to data analysis. One-way analysis of variance (ANOVA) was conducted to evaluate the effects of planting density on the soil and plant nutrients and nutrients ratio, and to assess the effects of interspecific competition on soil bacterial α-diversity indices (Shannon and Chao 1). The Kruskal–Wallis H test was used to test the relative abundance of bacterial family at OTU levels among treatments. Levene’s test was used to determine the variables’ homoscedasticity. Duncan’s new multiple range test on least significant difference (l.s.d.) was used to determine least significant range between means. The redundancy analysis (RDA) was applied to analyze the relationship between environmental variables (light and soil nutrient) and biomass in field experiment, and between soil nutrient content and soil bacterial communities in laboratory experiment using Canoco 5. A Mantel R² statistic and p values were used to test the significance level and define the significance of each variable (permutations = 999). Spearman correlation heatmap was used to evaluate correlations between soil environmental variables and shoot nutrients in four weeds and A. sativa in field experiment, and between bacterial communities and soil environmental variables in greenhouse experiment. The oat and weed soil microbial β-diversity were identified in separate analyses at the family level using principal coordinate analysis (PCoA) based on Bray–Curtis similarity matrices. Significance was tested by similarity analysis (ANOSIM) using the vegan package in R statistical software (version 3.6.3).

3. Results

3.1. Effect of Planting Density on Crop Performance

Plant density significantly affected oat growth traits (i.e., plant height, stem diameter, till number, and crop biomass) (p < 0.001) (

Table 1. Effect of crop density on plant height, stem diameter, till number, crop biomass and leaf area index of oat crop.

Treatment	Plant Height (cm)	Stem Diameter (mm)	Till Number (Till/Plant)	Crop Biomass (kg/ha)	Leaf Area Index
O1	68.4 ± 1.45 bc	6.10 ± 0.11 a	7.2 ± 0.47 a	4297.2 ± 209.7 d	3.26 ± 0.079 e
O2	67.3 ± 0.88 c	5.27 ± 0.12 b	6.2 ± 0.29 b	5318.2 ± 197.2 c	3.58 ± 0.030 d
O3	69.9 ± 0.41 bc	4.89 ± 0.08 cd	4.6 ± 0.22 c	6746.0 ± 155.6 b	3.89 ± 0.035 c
O4	69.2 ± 0.59 bc	5.11 ± 0.11 bc	3.3 ± 0.15 d	6885.8 ± 192.8 ab	4.03 ± 0.042 bc
O5	70.7 ± 0.72 b	4.72 ± 0.10 d	2.7 ± 0.21 de	7040.4 ± 234.4 ab	4.15 ± 0.058 b
O6	73.5 ± 0.96 a	3.93 ± 0.17 e	2.2 ± 0.20 e	7392.6 ± 185.0 a	4.36 ± 0.060 a

Note: O1, O2, O3, O4, O5, and O6, refer 60, 120, 240, 360, 480, and 600 plants m⁻², respectively. All values are shown as mean (S.E.) (n = 4). Within a column for plant height, stem diameter, till number, and crop biomass of each plant density, values followed by different letters are significantly different by least significant differences of Duncan’s new multiple range test.

and Table S2). There was an increase in plant height resulting from increased crop density, accompanied by the highest oat biomass. On contrast, stem diameter and till number were 55.2% and 227.3% higher at 60 plants m⁻² than those at 600 plants m⁻². Crop biomass and leaf area index (LAI) greatly increased with the increase in planting density from 60 to 600 plants m⁻², and the greatest crop biomass and LAI gained at 600 plants m⁻² (Table 1).

3.2. Effect of Planting Density on Light Environment

The light environment, including intercepted photosynthetically active radiation (IPAR) and red to far-red light ratio (R:FR ratio), varied among planting density treatments (Table S3). The IPAR increased with increasing planting density to the O6 compared to the O1 treatment (Figure 1A). Increased planting density significantly increased IPAR by between 0.8 and 11.5%. In contrast, there was a decrease in R:FR ratio resulting from increased crop density; R:FR ratio was 55.4% (p < 0.001) lower in O6 than in O1 treatment (Figure 1B).

3.3. Effect of Planting Density on Weed Biomass and Weed Density

In the present study, different plant density treatments had a significant impact on the weed biomass (Table S4). The results present that increasing planting density from O1 to O6 decreased the weed biomass accumulation in field, but weed biomass did not differ significantly at O4, O5, and O6 treatments (Figure 2A). Weed biomass varied in weed species (Table S5). On average, E. crus-galli and D. sanguinalis gained more biomass than that P. oleracea and C. album (Figure 2B). Increased planting density from O1 to O6 significantly (p < 0.001) decreased the proportion of grass weeds biomass by between 91.6% and 58.8%, but significantly (p < 0.001) increased the proportion of broadleaf weeds biomass by between 8.4% and 41.2%.

The weed density varied among planting density treatments (Table S4). The weed density was the lowest at O1 treatment, and there was a greater increase in weed density in higher planting density treatments (Figure 2C). Weed density varied in weed species (Table S5). On average, weed density of E. crus-galli and D. sanguinalis were less than these of P. oleracea and C. album (Figure 2D). Weed density of broadleaf increased with increasing planting density by 79.6% (p < 0.001) from the O6 treatment to the O1 treatment, whereas that of grass was 55.8% (p < 0.001) lower in the O6 than in the O1 treatment.

3.4. Effect of Planting Density on Soil C, N, and P Contents and Stoichiometry in the Oat Field

Different plant densities had a significant impact on soil properties and stoichiometry (Table S6). The soil P decreased with increasing planting density in the O6 compared to the O1 treatment (

Table 2. Effect of crop density on contents of soil total carbon (TC), total nitrogen (TN), total phosphorus (TP), available phosphorus (AP), and available nitrogen (AN), and ratio of soil C to soil N (soil C/N), ratio of soil C to soil P (soil C/P), ratio of soil N to soil P (soil N/P).

Treatment	Soil TC g kg−1	Soil TN g kg−1	Soil TP g kg−1	Soil AP mg kg−1	Soil AN mg kg−1	Soil C:N	Soil C:P	Soil N:P
O1	10.60 ± 0.04 ns	1.20 ± 0.01 ns	0.633 ± 0.012 a	6.14 ± 0.19 a	26.70 ± 0.22 a	8.85 ± 0.09 c	16.75 ± 0.27 c	1.89 ± 0.04 b
O2	11.16 ± 0.13 ns	1.23 ± 0.03 ns	0.627 ± 0.014 a	6.23 ± 0.21 a	25.59 ± 0.86 ab	9.11 ± 0.09 bc	17.85 ± 0.56 bc	1.96 ± 0.08 b
O3	11.29 ± 0.39 ns	1.26 ± 0.04 ns	0.618 ± 0.014 a	6.07 ± 0.14 a	24.41 ± 0.86 bc	8.96 ± 0.06 bc	18.26 ± 0.51 b	2.04 ± 0.05 b
O4	11.08 ± 0.31 ns	1.20 ± 0.03 ns	0.590 ± 0.029 ab	5.54 ± 0.13 ab	22.78 ± 0.69 cd	9.22 ± 0.09 ab	18.83 ± 0.46 b	2.04 ± 0.07 b
O5	10.52 ± 0.47 ns	1.15 ± 0.03 ns	0.551 ± 0.014 b	5.14 ± 0.19 b	21.60 ± 0.62 d	9.15 ± 0.13 b	19.06 ± 0.50 b	2.08 ± 0.04 b
O6	10.82 ± 0.17 ns	1.14 ± 0.02 ns	0.486 ± 0.008 c	4.90 ± 0.30 b	21.09 ± 0.58 d	9.48 ± 0.08 a	22.29 ± 0.49 a	2.35 ± 0.07 a

Note: O1, O2, O3, O4, O5, and O6, refer 60, 120, 240, 360, 480, and 600 plants m⁻², respectively. All values are shown as mean (S.E.) (n = 4). Within a column for soil TC, TN, TP, AP, AN, soil C/N, C/P and N/P of each plant density, values followed by different letters are significantly different by least significant differences of Duncan’s new multiple range test. ns means no significant between different treatments.

), and soil P was 23.3% (p < 0.001) lower in the O6 than in the O1 treatment. Increased planting density from the O1 to O6 treatment decreased soil AP and soil AN (Table 2). Soil AP and AN in the O6 treatment were decreased by 21.4 and 21.0% (p < 0.001), respectively, compared to O1. Generally, when planting density increased from O1 to O6, soil C:N, soil C:P, and soil N:P increased (Table 2). On average, soil C:N, soil C:P, and soil N:P, were significantly increased by 7.1, 33.1, and 24.3% (p < 0.001), respectively, at the O6 treatment compared to the O1 treatment. Soil N:P did not differ significantly from O1 to O5.

3.5. Effect of Planting Density on Shoot C, N, and P Contents and Stoichiometry in the Weeds and Oat

The shoot C, N, and P contents and their ratios were significantly (p < 0.001) affected by treatment, species, and their interaction (Table S7).

There was a significant (p < 0.01) effect of planting density on shoot C of P. oleracea. Shoot C of P. oleracea increased gradually with increasing planting density from O1 to O4, whereas the reverse was true for planting density from O4 to O6 (Figure 3A). Shoot C of D. sanguinalis and A. sativa were significantly (p < 0.05) higher than that in E. crus-galli, P. oleracea, and C. album shoot, with the lowest shoot C in P. oleracea (Figure 3A).

Increasing planting density significantly decreased in shoot N of P. oleracea, C. album, and A. sativa (p < 0.001). By contrast, shoot N in D. sanguinalis was significantly (p < 0.05) increased at O6 compared to O1. Increased shoot N was found in E. crus-galli at O3 and O6 compared to O1 (Figure 3B). P. oleracea had the highest (p < 0.05) shoot N, followed by C. album and D. sanguinalis, and then E. crus-galli and A. sativa (Figure 3B).

Planting density showed no significant influence on shoot P of P. oleracea and A. sativa, and increased shoot P of E. crus-galli and D. sanguinalis at an increasing planting density, but decreased shoot P of C. album (Figure 3C). The shoot P in weeds was significantly higher than those in A. sativa, with the highest shoot P in C. album (Figure 3C).

The increasing planting density (O3 and O6) significantly (p < 0.05) decreased the C:N ratio in E. crus-galli, and a lower (p < 0.05) C:N ratio was also measured in D. sanguinalis at O6 than O1. In contrast, the C:N ratio in shoots from P. oleracea, C. album, and A. sativa significantly (p < 0.05) increased from O3 to O6, O5 to O6 and O4 to O6, respectively, compared to O1 (Figure 3D). E. crus-galli and A. sativa had the highest C:N ratio, followed by D. sanguinalis, C. album and P. oleracea (Figure 3D).

There was a significant (p < 0.05) decrease in C:P ratio in shoots from D. sanguinalis and E. crus-galli from O3 to O6. In contrast, the C:P ratio in shoots A. sativa, C. album, and P. oleracea significantly (p < 0.05) increased from O3 to O6 and O4 to O5, respectively, compared to O1. The shoot C:P ratio was higher (p < 0.05) in A. sativa than other four weeds, but the lowest C:P ratio was shown in C. album (Figure 3E).

The N:P ratio in shoot of E. crus-galli, D. sanguinalis, P. oleracea, and A. sativa significantly (p < 0.05) decreased from O5 to O6, O5 to O6, O2 to O6, and O4 to O6, respectively, compared to O1. In contrast, N:P ratio in leaves of C. album increased gradually with increasing planting density from O1 to O3, whereas the reverse was true for planting density from O3 to O6 (Figure 3F). The N:P ratio was higher (p < 0.05) in A. sativa than the other four weeds, with the lowest shoot C:P ratio in C. album (Figure 3F).

3.6. Relationship of C, N, and P Contents and Ecological Stoichiometry in the Shoot–Soil System

For grass weeds, there was a significant negative correlation between shoot C and soil C:P and N:P for E. crus-galli (Figure 4A). Shoot N and P were significantly negatively correlated with soil P and AP for E. crus-galli and with soil N, P, AP, and AN for D. sanguinalis, but were positively correlated with soil C:N, C:P, and N:P for both grass weeds (Figure 4A,B). Shoot C:N was positively correlated with soil P for E. crus-galli and with soil P, AP, and AN for D. sanguinalis, but there was a significant negative correlation between shoot C:N and soil C:N, C:P, and N:P for D. sanguinalis (Figure 4A,B). There was a significant positive correlation between shoot N:P and soil P, AP, and AN for the two grass weeds, but negatively correlated with soil C:P and N:P for E. crus-galli and with soil C:N, C:P, and N:P for D. sanguinalis (Figure 4A,B).

For broadleaf weeds, shoot C only showed a significant negative correlation with soil AN for P. oleracea. There was a significant positive correlation between shoot N and soil P, AP, and AN for P. oleracea and between shoot N and soil N, P, AP, and AN for C. album, but a negative correlation between shoot N and soil C:N, C:P, and N:P for both grass weeds (Figure 4C,D). There was a significant positive correlation between shoot P and soil P, AP, and AN for C. album, but a negative correlation between shoot P and soil C:N, C:P, and N:P. for P. oleracea; shoot P only showed a significant positive correlation with soil C:N. Compared to shoot N, shoot C:N showed reversed correlations with soil nutrient and stoichiometries for both grass weeds. Shoot C:P showed a negative correlation with soil P, AP, and AN, and a positive correlation with soil C:N, C:P, and N:P for C. album (Figure 4C,D). There was a significant positive correlation between shoot N:P and soil P, and between AP and AN, but a negative correlation between shoot N:P and soil C:N, C:P, and N:P for P. oleracea (Figure 4C,D).

For A. sativa, there was a significant positive correlation between shoot C and soil C and N. Shoot N and N:P were positively correlated with soil N, P, AP, and AN, but were negatively correlated with soil C:N, C:P, and N:P (Figure 4E). Shoot P showed a positive correlation with soil C:N but a negative correlation with soil AP and AN. In contrast to shoot N, shoot C:N had the reversed correlations with soil nutrient (N, P, AP, and AN) and stoichiometries (C:N, C:P, and N:P). Shoot C:P showed a significant positive correlation with soil AN and AP, but was negatively correlated with soil C:N (Figure 4E).

3.7. Correlations between Environmental Variables (Light and Soil Traits) and Plant Nutrient and Biomass

Redundancy analysis (RDA) showed a combination of variables that explained 72.6%, 73.7%, 72.6%, 74.1%, and 75.3% of the variance in nutrient and biomass for E. crus-galli, D. sanguinalis, P. oleracea, C. album, and A. sativa, respectively (Figure 5A–E). The contribution to total variance’s relative importance reveals the IPAR and soil AN have surprisingly dominant roles in explaining the biomass and nutrient content variation for four weeds and A. sativa (

Table 3. The Mantel test (permutations = 999) among light environment, soil nutrient contents and stoichiometries, and plants (E. crus-galli; D. sanguinalis; P. oleracea, C. album, and A. sativa) nutrient contents and biomass. The significance level was tested by the R² and p-values.

	E. crus-galli		D. sanguinalis		P. oleracea		C. album		A. sativa
	R2	p Value	R2	p Value	R2	p Value	R2	p Value	R2	p Value
IPAR	0.159	0.006	0.096	0.008	0.093	0.008	0.128	0.002	0.092	0.012
Soil C	0.022	0.368	0.017	0.49	0.025	0.268	0.039	0.1	0.019	0.386
Soil N	0.056	0.016	0.041	0.04	0.002	0.984	0.013	0.578	0.119	0.006
Soil P	0.022	0.35	0.047	0.048	0.016	0.456	0.053	0.034	0.003	0.896
Soil C:N	0.067	0.012	0.068	0.036	0.006	0.842	0.025	0.254	0.053	0.046
Soil C:P	0.045	0.114	0.028	0.268	0.085	0.016	0.055	0.04	<0.001	0.996
Soil N:P	0.064	0.026	0.013	0.616	0.031	0.172	0.007	0.8	0.028	0.192
Soil AN	0.312	0.002	0.417	0.002	0.421	0.002	0.408	0.002	0.433	0.002
Soil AP	0.009	0.732	0.01	0.684	0.047	0.046	0.013	0.538	0.005	0.806

).

3.8. The Diversity and Compositions of Bacterial Communities in the Rhizosphere of Oat and Weeds

The Shannon diversity index of soil bacterial communities was lower in A. sativa monocropping treatment (As) than A. sativa in interspecific competition with E. crus-galli and D. sanguinalis (As_Ds_Ec) (p < 0.05) (Figure 6A,D). E. crus-galli in interspecific competition with D. sanguinalis (Ec_Ds) and with D. sanguinalis and A. sativa (Ec_Ds_As) significantly increase rhizosphere Shannon diversity index and Chao 1 index compared to E. crus-galli root microbiota in monocropping treatment (Ec) (Figure 6B,E). Similarly, the interspecies competition of D. sanguinalis with E. crus-galli and A. sativa (Ds_Ec_As) treatment showed a significantly increased Shannon diversity index, which was larger than monocropping (Ds) treatment and interspecific competition with E. crus-galli (Ds_Ec) (Figure 6C,F).

The principal components analysis (PCoA) and ANOSIM test indicated that, in the rhizosphere of two weeds, there were notable variations in the β diversity among the soil bacterial communities (p < 0.05) (Figure 7). The PCoA showed two extracted principal coordinates that explained 79.7%, 65.4%, and 68.0% of the total variation in bacteria communities for A. sativa, E. crus-galli, and D. sanguinalis, respectively (Figure 7A–C). For A. sativa, monoculture treatment (As) clustered together and did not separate from interspecific competition treatment (As_Ds_Ec) (Figure 7A). In addition, the monoculture treatment of E. crus-galli and D. sanguinalis (Ec and Ds, respectively) clustered together and were separate from the interspecific competition treatments (Ec_Ds, Ec_Ds_As and Ds_Ec, Ds_Ec_As, respectively) (p < 0.05) (Figure 7B,C). There was less of a distinction among the interspecific treatments.

The variations in the bacterial OTUs abundance at the family level across the various treatments, and the bacterial family was chosen on the basis of OTU abundance greater than 1% (Figure 8A–C). For the A. sativa bacterial communities, Bacillaceae, Gemmatimonadaceae, Gaiellales, Solibacteraceae, Comamonadaceae, Roseiflexaceae, Haliangiaceae, and Rhizobiaceae were significantly (p < 0.05) increased in As_Ds_Ec treatment compared to As treatment, whereas the reverse was true for Intrasporangiaceae and LWQ8 (Figure 8D). Correspondingly, the relative abundances of the dominant bacterial family Bacillaceae, Xanthobacteraceae, Streptomycetaceae, Geodermatophilaceae, and Planococcaceae in the E. crus-galli rhizosphere were greatly decreased by interspecific competition treatment (p < 0.05), whereas the reverse was true for Chthoniobacteraceae (Figure 8E). The relative abundances of the dominant bacterial family Micromonosporaceae, Paenibacillaceae, Geodermatophilaceae, and Clostridiaceae in the D. sanguinalis rhizosphere were greatly decreased by interspecific competition treatment (p < 0.05), whereas the reverse was true for Roseiflexaceae (Figure 8F).

3.9. Correlations between Soil Properties and Microbial Communities

The redundancy analysis demonstrated a combination of soil chemical properties explained 61.5% and 62.2% of the variance in the bacterial communities for E. crus-galli and D. sanguinalis, respectively (Figure 9A,B). The partial Mantel test showed that NO₃⁻-N and AP contents have a significant influence in the bacterial communities of E. crus-galli, and pH, NO₃⁻-N, and AP contents have a significant influence in the bacterial communities of D. sanguinalis (Tables S8 and S9).

Spearman correlation analyses showed that both soil AP and NO₃⁻-N showed a positive correlation with Bacillaceae, Streptomycetaceae, Paenibacillaceae, and Geodermatophilaceae for E. crus-galli and with Paenibacillaceae and Thermoanaerobaculaceae for D. sanguinalis (p < 0.05) (Figure 9C,D). Positive correlations were found between pH and the abundance of beneficial bacterial family, including Bacillaceae, Streptomycetaceae, Micromonosporaceae, and Geodermatophilaceae for D. sanguinalis (p < 0.05). In addition, pH showed a negative correlation with the abundances of Intrasporangiaceae and Vicinamibacteraceae for D. sanguinalis (p < 0.05).

4. Discussion

Cultivated oat crops are usually spontaneously surrounded by weed species, which are responsible for significant feed losses in production systems worldwide. This study showed the benefits of increasing oat planting density to improve the crop competition to manage weeds as an integral part of integrated weed management (IWM) strategies. Maintaining a dense oat plant establishment can favor the oats’ competitive capacity to increase shading and decrease elemental ratios and rhizosphere beneficial microorganisms of weeds, which are associated with successful weeds suppression. Such results, therefore, emphasize the lower competitive ability of weeds when grown in areas with high oat planting density through increased interspecific competition. Therefore, understanding the responses of weeds to heterospecific neighbors would help to effectively utilize an agronomic approach in IWM programs. Below, the implications of these findings are discussed.

4.1. Effect of Oat Planting Density on Weed Suppression and Oat Production

Effective weed suppression by a crop could be achieved at high crop density due to population crowding [31]. In the current study, we found that an increase in oat density could take advantage of size-asymmetric competition with weeds. At high density, a rapid crop canopy closure greatly decreased intercept photosynthetically active radiation (IPAR) and increased reflection of far-red (FR) light to neighbor weeds, resulting in a decreased weed cohorts’ growth and consequently competitiveness [32]. In specific, the increased growth traits (e.g., height and leaf area index) of oat plants could contribute to shade of weeds, which is a ‘one-sided’ interaction in which higher leaves shade lower leaves. Such competition advantage increases with plant density, which would reduce IPAR and increase reflection of FR perceived by the weeds, thereby benefiting the crop at the expense of the annual weeds [33,34]. Similar suppressive effects of increased plant density on weed control have been reported for problem weed species in wheat [4] and soybean (Glycine max) [35] cropping system. It is possible that increasing planting density may improve plant competitiveness against weeds through enhanced shading, which may inhibit weed growth. By their nature, weeds always require sufficient sunlight for photosynthesis, which is essential for their rapid growth and spread [36]. Furthermore, when weeds are shaded, they develop a shade avoidance syndrome, which inhibits their biomass accumulation by regulating their physiological metabolism and adaptive morphology at the expense of assimilated resources. [37]. Thus, increased oat planting density inhibited weed growth, indicating that dense crop planting promotes a quick canopy closure to effectively suppress weed growth.

The current results showed that weed biomass dramatically decreased with increasing oat density, which confirmed findings from a previous study by Weisberger et al. [5]. However, increased crop density may increase intraspecific competition among crop plants resulting in faster senescence for leaves at the bottom, which to some extent explains the adverse effects, especially on seed crops [38]. Li et al. [39] found increasing crop density reduced the seed yield and yield components of winter oilseed rape (Brassica napus) when crop density increased from 58.5 to 69 plants m⁻². It suggests that optimizing the crop density is also crucial from a nutrient quality standpoint. Currently, these annual weeds could be greatly suppressed, and there is no significant difference in forage yield and protein from 360 to 600 plants m⁻² (Figure S1). This finding is in agreement with Berti and Samarappuli [40], who found that the forage yield and nutritional value of alfalfa (Medicago sativa) were not affected by an increase in sowing rate. Therefore, it is necessary to integrate different weed management strategies to achieve complete weeds control without decreasing forage biomass in fodder crops and seed yield in seed crops.

The case study is based on a single season, and needs to be repeated in different spaces and at different times to test the reliability of our results. Climatic factors (e.g., temperature, precipitation) play a crucial role in influencing the outcome of weed infestation in fields, and the factors are indirectly related to attack on weeds and plants by insect pests and diseases [3]. It was observed that there was no oat or weed diseases or insect damage during the growing season.

4.2. Effect of Oat Planting Density on Nutritional Elements Balance and Stoichiometry in the Weeds

The competitive ability of a specific plant species is related to its intrinsic ability (e.g., stoichiometric flexibility) to alter its metabolism in response to interspecific competition [41]. The current results clearly showed that the intense nutrient competition between oats and weeds led to a reduction in nutrient availability for each weed species, and subsequently reduced their nutrient uptake at high densities. The competitive ability of weed species decreases when their C:N:P stoichiometry exceeds the threshold of stoichiometric homeostasis [42]. E. crus-galli and D. sanguinalis decreased their C:N, C:P, and N:P ratios in response to competition with increased oat planting density, whereas the reverse was true for P. oleracea and C. album. As proposed by de Matos et al. [17], the species-specific stoichiometry could reflect variation in plant morphology, photosynthesis, and nutrient acquisition capacity, as observed in Bidens pilosa, Ipomoea grandifolia, and Amaranthus viridis. Additionally, weed species have evolved systems for determining the nitrogen status of the root system and soil, thereby regulating reactions in their leaves that would affect photosynthesis [43]. In this study, P. oleracea and C. album growth were primarily limited by available nitrogen in competition with oat at high density, suggesting that increasing their C:N ratio could be considered a physiological adaptation to the lower N availability in the soil. In contrast, the decreased C:N ratio of E. crus-galli and D. sanguinalis tissues in response to increasing crop density is likely to be a genetic and physiological adaptation to competition [44]. Therefore, weed species could optimize energy use when grown in competition with high-density oat plants, through regulate metabolism, transport, and assimilation of C and N.

The increase in C:P of P. oleracea and C. album in competition with high crop density may be due to limitations in P uptake, as rapid growth and development of oat plants provided them an advantage over the weeds competing for P. The fast growth of oats allowed their roots to take up most of the soil volume, which hindered the weed roots’ ability to spread out across the same space [45]. As a result, depletion of the soil-available P and decreased root growth led to reduced P uptake by P. oleracea and C. album. However, the failure of E. crus-galli and D. sanguinalis suppression at low oat planting density conditions might be mainly related to their high nutrient uptake [46]. One possible explanation was that a decline in shoot P content in grass weeds was due to the dilution effect caused by the increase in biomass and the need for P supply for growth at low density pressure [47].

In the present study, the soil-available N and P was practically deceased by increased oat planting density. Plant growth can be stimulated by both N and P, and the availability of one element affects how well the other is absorbed and utilized. Competition with oats may have resulted in reduced N:P ratios in these weeds, and the N:P ratio is mainly regulated by adjusting N and P uptake [47]. E. crus-galli and D. sanguinalis had higher P content and lower N:P ratios at high density than low density, suggesting that N limitation existed in the soil as a result of N uptake by oats. At high density, these weeds would promote shoot growth elongation to improve light-harvesting function to avoid shade stress, which increases the requirement of P for the synthesis of P-rich rRNA for rapid growth [48,49]. That is, weeds adjusted the N:P ratio of their tissues, allowing organisms to exhibit advantageous adaption in heterogeneous environments [50]. Therefore, adaptive modifications may result in increased weed success, as was also observed for A. viridis and B. pilosa when they were grown in interspecific competition with maize [17,51]. Therefore, the data from this study suggest that E. crus-galli and D. sanguinalis likely have greater infestation potential than P. oleracea and C. album in oat-cultivated grassland.

4.3. Effect of Interspecific Competition on Bacterial Community Diversity, Composition and Structure

A comparison between the interspecific competition treatment and the monoculture treatment for two weed species showed that the bacterial community in the former had a higher richness estimator (Chao 1) and diversity index (Shannon). These variations might result from interactions between different species altering the soil microenvironment [22]. The oat–weeds competition presented an intense heterogeneous vegetation cover and rooting pattern, thereby affecting soil properties and root-associated microbial communities. Variations in the bacterial communities of weeds competing with oats were probably caused either by microbial dispersal from nearby oat plants to weed plants or by competitive displacement of groups usually associated with weeds [52]. Furthermore, the PCoA result showed marked variations in the structures of bacterial communities across the monoculture and mixed culture treatments. Previous research has demonstrated that interactions between heterospecific plants alter the soil bacterial community structure, as reported for maize compared to B. pilosa [53]. It is possible that the co-existence of heterospecific neighbors trigger complex biochemical reactions in plants, such as how changes in metabolic substances in root exudates impact root microbiota recruitment [54,55,56].

The microbial community structure variations imply that the abundance of certain soil microbial populations may be influenced by interspecies interactions. The dominant taxonomic groups identified in the soils assayed were Bacillaceae, Gaiellales, Gemmatimonadaceae, Sphingomonadaceae, Xanthobacteraceae, Streptomycetaceae, Micromonosporaceae, Micrococcaceae, and Paenibacillaceae, all depicted as common inhabitants of soil [21]. Our dataset shows that competition with oat leads to a considerable decline of the relative abundance occurred in the families Bacillaceae, Xanthobacteraceae, Streptomycetaceae, Geodermatophilaceae, and Planococcaceae of E. crus-galli. A lower relative abundance of Micromonosporaceae, Paenibacillaceae, and Geodermatophilaceae and a higher abundance of Roseiflexaceae in competition with oat than in D. sanguinalis grown in monoculture. It indicated that both plant species and interspecific competition can change the abundance of dominant bacteria, which is due to their adaptability to a new microenvironment [57,58]. Moreover, the beneficial species (i.e., Bacillaceae, Xanthobacteraceae, Streptomycetaceae, Paenibacillaceae, Planococcaceae, Micromonosporaceae) can favor plant growth and improve nutrient availability by stimulating plants to secrete auxin while simultaneously fixing nitrogen, solubilizing phosphate, and decomposing organic matter [59,60,61,62,63,64]. Apart from their ability to stimulate plant growth, Bacillaceae, Streptomycetaceae, and Geodermatophilaceae can prevent plants from soil-borne disease and abiotic stresses (e.g., drought) through the synthesis of substances akin to lipopeptides that make plants more resistant to external stresses [65,66]. These findings imply that neighboring A. sativa can competitively impair plant growth-promoting bacteria of E. crus-galli and D. sanguinalis, resulting in decreased rhizosphere competence on nutrient uptake, growth, and soil-borne pathogen suppression.

Notably, changes in soil chemical properties are strongly correlated with the composition of the soil microbial community [67]. In the present study, both the AP and NO₃⁻N were found to be important drivers of the variations in the soil bacterial communities of E. crus-galli, and AP, NO₃⁻N, and pH showed a significant correlation with the bacterial communities of D. sanguinalis. Soil AP and NO₃⁻-N showed a positive correlation with Bacillaceae, Streptomycetaceae, Paenibacillaceae, and Geodermatophilaceae, as previously observed by Li et al. [62], Zhou et al. [68], Ribeiro et al. [59], and Panhwar et al. [69]. This occurs because these plant-beneficial bacteria exhibit a significant potential in N fixation and phosphates solubilization in the soil [65,70]. However, competition with oat significantly decreased abundance of these beneficial bacteria, resulting in decreasing nutrient acquisition of E. crus-galli and D. sanguinalis. Another important factor influencing the composition of bacterial communities is the soil pH. This is particularly true given the significant impact of soil pH on bacterial communities in general, particularly the increase in abundance and diversity of bacteria with increasing soil pH [71,72], as observed in E. crus-galli. In addition, our results showed positive correlations between pH and the abundance of certain bacterial family, including Bacillaceae, Streptomycetaceae, Micromonosporaceae, and Geodermatophilaceae, which have been reported to have the ability to increase nitrogen-fixing and reduce denitrification [71,72]. Overall, the soil bacterial communities seemed to be responsive to changes in plant species and environmental conditions. Further research is necessary to determine the specific influencing factors in this experiment.

When compared to crops, weeds are generally more reliant on soil microbiota for growth and competitiveness [54]. According to the current study, competition with A. sativa significantly reduced the photosynthesis and biomass accumulation of the E. crus-galli and D. sanguinalis (Figure S2 and Table S10), possibly as a result of competitive interactions in decreased beneficial microbial inhabitants, attributed to a decrease in mineral nutrient uptake for weed growth [73,74]. Plants release enormous amounts of photosynthetic carbon from their roots through active and passive processes. These rhizodepositions might combat phytopathogens and attract beneficial microorganisms for the resource acquisition [75]. Because of the complexity of the relationships that exist between weeds, crops, and soil microorganisms in agroecosystems, future studies should concentrate on identifying and characterizing these important microbes and their roles in crop–weed competition.

5. Conclusions

Our findings demonstrated how crop density influenced weed suppression and oat forage yield, with implications for the development of ecological weed control in oat-cultivated grassland. Weed biomass was significantly suppressed by an increase in oat density by up to 360 plants m⁻², allowing oat crops to gain an advantage in terms of population proportion effects and asymmetric radiation competition over weeds, which inhibited weed growth. These weeds appeared to have different survival strategies at high crop densities: E. crus-galli and D. sanguinalis could absorb and use nutrients more efficiently while P. oleracea and C. album tend to transport and accumulate nutrient reserves in their tissues, suggesting the E. crus-galli and D. sanguinalis possibly have a greater nutritional competitive advantage than P. oleracea and C. album. Furthermore, in competition with oat, rhizosphere bacterial communities of E. crus-galli and D. sanguinalis exhibited a decline in unique taxa associated with beneficial functions in nutrient cycling and plant growth. This study enhances our understanding of how weed density affects the effectiveness of weed control through reduced radiation resource availability, elemental ratios, and beneficial microorganisms of weeds, thereby achieving optimal forage yield while considering crop competition potential. Further research on crop–weed interactions under various climate changes and agronomic practices is required to expand our understanding of how to suppress weeds and increase yields based on crop competitiveness.