Optimizing Wheat Planting Density by Adjusting Population Structure and Stabilizing Stem Strength to Achieve High and Stable Yields

Abstract

Increasing wheat (Triticum aestivum L.) planting density is the most effective production management method for increasing yields; however, excessive crop populations under high planting densities may experience elevated risk of stem lodging. We conducted this study to assess the relationship between reduced lodging and increased yield, investigate the effects of planting density on wheat population structure, stem strength, and material transport, and provide a basis for rationale planting densities. The experiments were carried out using a split-plot design with three replicates. The main plots contained two wheat varieties: Bainong 5819 (BN5819) and Bainong 4199 (BN4199), and the sub-plots contained four planting density treatments: 90 × 10⁴ plants/ha (D1), 180 × 10⁴ plants/ha (D2), 270 × 10⁴ plants/ha (D3), and 360 × 10⁴ plants/ha (D4). A two-year field trial was conducted in 2021–2023. The relationships between population structure characteristics, changes in stem strength, activation, and retransport of stem material after anthesis, and achievement of high and stable yields were investigated at the different planting densities. When the planting density of wheat increased from D1 to D4 treatment, the activity of fructan hydrolase was significantly increased. Compared with D1 treatment, the highest activity of fructan hydrolase was increased by 457.47 μg/h/g under D4 treatment. At the same time, the increase of density also increased the contribution rate of dry matter accumulation (CDMA) to grain after anthesis increased, with the highest increase in CDMA at 33.67%, which significantly reduced stem strength. Correlation analysis revealed a significant negative association between CDMA and stem strength. Specifically, CDMA levels were significantly lower with the D3 treatment than the D4 treatment, while stem strength remained higher after anthesis as an adaptive response to mitigate lodging risk. Stem storage compounds can promote grain filling and a weight increase in inferior grains. The number of spikes per unit area increased significantly with increasing planting density, but the number of grains per spike and 1000-grain weight decreased significantly. In two years, the number of spikes in D3 treatment increased by a maximum of 211.67 × 10⁴ ha⁻¹ and 99.17 × 10⁴ ha⁻¹, respectively, compared to D1 and D2 treatments. The number of grains per spike was significantly higher than that of D4 treatment, the highest being 3.68 grains. Therefore, in the North China Plain with suitable water, fertilizer, and temperature, the sowing density of 270 × 10⁴ plants/ha established population structure, significantly reduced CDMA, maintained post-anthesis stem strength, enhanced resilience of stems against post-anthesis lodging, and resulted in high yields by stabilizing the number of grains per spike and increasing the number of wheat spikes.

1. Introduction

Wheat (Triticum aestivum L.) is one of the most widely cultivated cereal crops in the world. There has been a decline in land devoted to cultivation as a result of urbanization and global population growth [1]. Therefore, increasing the per-unit yield is the main way to improve grain productivity. Under the current production levels, increasing the population size is still the key to increasing output [2]. At present, the research on wheat yield increase is carried out from many angles, such as tillage and gene mining [3,4]. However, adjusting the planting density is one of the most effective ways to increase wheat yield per unit area in field management [5,6]. Most users prefer to use high seeding volume to achieve high yield. However, the unreasonable sowing density not only cannot increase the yield excessively, but also seriously affects the stem strength of wheat after anthesis stage [7]. Therefore, optimizing wheat planting density is imperative for enhancing the population structure and regulating stem quality, thereby ensuring the attainment of high and stable wheat yields without significant lodging.

The planting density of wheat significantly affects stem quality [8]. Wheat stem lodging primarily occurs between the second node and the base of the stem [9]. Higher planting density has been shown to significantly reduce stem diameter and wall thickness, leading to a decline in stem strength [6,10]. High planting density intensifies interplant competition for diverse resources, compromises stem strength, and significantly increases the risk of wheat lodging [5,11]. Previous studies have shown that, within reasonable planting densities, increasing planting density can lead to absorption of more light energy, increasing canopy photosynthesis and total biomass accumulation and ultimately resulting in a high yield of wheat [12,13]. However, with an over-dense wheat population, plants are shaded, the canopy structure and stem quality deteriorate, and the risk of plant lodging (displacement from stems from their vertical orientation) increases. These factors pose a serious threat to yield stability [14,15]. The annual yield reduction caused by lodging ranges from 10% to 57% [16], significantly impeding the enhancement of grain productivity.

Moreover, increasing seeding density can promote translocation of assimilates from vegetative organs to grains before anthesis [17]. Previous studies by our research group found that the contribution rate of dry matter accumulation before anthesis is significantly negatively correlated with stem strength. Stem assimilation transport is conducive to grain filling but increases the risk of stem lodging [9]. Huang et al. [18] found that the more assimilates in the stem are transported to the grain after anthesis, the greater the impact on stem quality and reduction in lodging resistance of the stem. At the same time, for wheat planted at a low density, stem microstructure, stem fullness, and lignin content were improved, significantly enhancing stem strength [6]. Although wheat planting density can regulate stem and grain growth and development, the determination of wheat planting density in different regions is also significantly influenced by local climate and soil type. Weather conditions and soil quality are seen as key environmental factors that directly affect crop production [19,20]. Changes in temperature may have a negative impact on crop productivity [21].

Spike number per unit area, grain number per spike, and 1000-grain weight are the most important factors affecting wheat yield [22]. Different wheat seeding densities can affect yield by adjusting yield components [23]. Under low planting density conditions, the limited interplant competition for water and nutrient resources leads to an extended tillering duration in wheat plants, increasing tiller number and survival rate and contributing to a higher yield per plant [24,25]. At the same time, low planting density can also increase the number of spikes produced by each plant and the weight of a single spike, so that wheat has a higher 1000-grain weight and number of grains per spike [23,26]. However, these benefits are not sufficient to compensate for the yield loss resulting from a lower spike number per unit area [27]. Some studies have suggested that high-density planting of crops can increase competition among plants, accelerate leaf senescence, and induce more grain abortions because of a reduction in spike assimilate accumulation during the filling process [28,29]. The main way to reduce wheat grain weight with a high planting density is to significantly increase the number of spikes per unit area, intensify competition among individual plants, and reduce the number of grains per spike and grain weight [22]. However, Yang et al. [30] found that increasing planting density could adjust the number of tillers, increasing the number of high-quality tillers and thereby increasing the yield. Higher planting densities result in higher spike numbers and yields than lower planting densities [11]. High planting density can also increase the dry matter contribution of pre-anthesis vegetative organs to grain filling and reduce the photosynthetic contribution of post-anthesis flag leaves to grain filling; the contribution of pre-anthesis vegetative organ remobilization to grain filling was 22.7% [31]. However, when utilizing assimilates stored in the stem for grain filling, it is imperative to ensure superior stem quality to effectively mitigate lodging risks during later stages of filling [32,33].

Previous studies have primarily focused on the impact of planting density on stalk lodging or examined the response of wheat yield, dry matter accumulation, and reactivation to planting density. However, few studies have assessed how different planting densities affect yield by adjusting the yield components and effects of stem material transport on stem strength and grain filling of superior and inferior grains under different planting densities. We hypothesized that planting density could improve stem strength and grain yield by adjusting wheat population structure. Therefore, the objectives of this study were as follows: (i) to investigate the impact of different planting densities on post-anthesis material transport and wheat stem strength; (ii) to quantify the contribution of stem storage assimilates to superior and inferior grain filling; and (iii) in terms of yield components, to explore how different planting densities lead to high and stable yields by regulating population structure.

2. Materials and Methods

2.1. Experiment Site

The field trial was conducted in an experimental field in Xinxiang County (35.17° N, 113.90° E), Henan Province, China, from 2021 to 2023. The soil composition in the experimental field was loam, in which the proportion of sand, powder, and clay is 40%–40%–20%.

Table 1. The average concentration of essential nutrients in the top 20 cm soil layer prior to sowing in the experimental area.

Year	Organic Matter (g/kg)	Available N (mg/kg)	Available P (mg/kg)	Available K (mg/kg)	Soil PH
2021–2022	17.01	106.33	28.23	157.33	7.12
2022–2023	16.58	104.28	30.3	164.63	7.02

presents the levels of organic matter, total nitrogen, available phosphorus, available potassium, and soil PH in the 0–20 cm soil layer of the test field prior to winter wheat sowing in both experimental years. Organic matter, total nitrogen, available phosphorus, and available potassium in soil layers were extracted by alkaline hydrolysis, Kjeldahl method, NaHCO₃ leaching, and flame photometry, respectively. The average temperature and precipitation of each wheat growth stage in the test field during 2021–2023 are shown in Figure 1.

2.2. Experimental Design

The experiments were carried out using a split-plot design with three replicates. The main plots consisted of two nationally approved wheat varieties: Bainong5819 (BN5819) and Bainong4199 (BN4199). Among them, BN5819 belonged to the medium and large spikes variety, BN4199 belonged to the multi-spike variety. According to the user in the field, 360 × 10⁴ plants/ha high sowing amount is frequently used as the highest seed. In order to reduce the amount of seeds used in production, four seed densities were set in the sub-plots, which were 90 × 10⁴ plants/ha (D1), 180 × 10⁴ plants/ha (D2), 270 × 10⁴ plants/ha (D3), and 360 × 10⁴ plants/ha (D4). The same planting row spacing (21.6 cm) was used for all four density treatments, and the difference in planting density was determined by the number of seeds planted in each row.

The spacing between wheat planting rows in each plot was set at 21.6 cm, while the plot area measured 30 m × 3.4 m with three repetitions. A compound fertilizer with a N:P₂O₅:K₂O ratio of 15:15:15 was applied at a rate of 705.9 kg/ha before sowing, whereas urea was applied at a rate of 215.7 kg/ha during the jointing stage. The two-year trial was sown yearly on 7 October 2021 and 5 October 2022 and harvested on 4 June 2022 and 1 June 2023.

2.3. Sampling and Determination

2.3.1. Mechanical Strength of the Second Internode of the Stem Base

Stem strength was quantified with a digital-display push-pull meter (HP-30) (Cambridge Scientific Labs, Watertown, MA, USA) using the vertical downward pressure method. Ten wheat plants exhibiting uniform growth at the anthesis stage were selected and measured every seven days. Each test was repeated three times, and the whole growth stage was measured six times. At the second segment of the stem, with the leaf sheathing removed, a 5 cm-wide slot was created, and vertical stem strength was quantified using a tension meter equipped with a V-shaped probe.

2.3.2. Stem Storage Assimilate Transport

Samples were taken at anthesis stage and maturity stage. Twenty wheat plants with uniform growth were taken at anthesis stage, and twenty wheat plants with uniform growth and their seeds were taken at maturity stage. The samples were dried to a constant weight at 70 °C, and the dry matter mass was determined. The contribution rate of dry matter accumulation to grains (CDMA) after anthesis was calculated as follows [34]:

$$\mathrm{CDMA}\%=\left({\displaystyle \frac{{\mathrm{G}}_1-{\mathrm{G}}_2}{\mathrm{GY}}}\right)\times 100$$

(1)

where G₁ (kg/ha) represents the dry weight of the whole plant at maturity. G₂ (kg/ha) represents the whole plant dry weight at anthesis stage, and GY (kg/ha) represents dried grain yield after harvest.

During the wheat jointing stage, ten plants exhibiting uniform growth were selected for analysis of assimilates. The second base leaf was treated according to the method described by Zhang et al. [35], using CO₂ containing a ¹³C isotope with an abundance of 99% as a labeling agent. Another set of ten plants was used as a control. The labeled plant grains were separated according to superior and inferior grains at the wheat maturation stage, followed by measurement of ¹³C and δ¹³C accumulation in each component after sample drying and grinding. The formula used was as follows:

$$F_i={\displaystyle \frac{({\mathsf{\delta }}^{13}\mathrm{C}+1000)\times R_{PBD}}{({\mathsf{\delta }}^{13}\mathrm{C}+1000)\times R_{PBD}+1}}\times 100$$

(2)

$${{}^{13}C}_i=C_i{\displaystyle \frac{\left(F_i-F_{nl}\right)}{100}}\times 1000$$

(3)

where $F_i$ is ¹³C abundance (%), R_PBD is the standard ratio of carbon isotopes (0.0112372), ¹³$C_i$ is the quantity of ¹³C, $C_i$ is the carbon content (g) of superior or inferior grains, and nl represents ¹³C abundance in the control group.

2.3.3. Fructan Hydrolase Activity

Ten wheat plants exhibiting uniform growth and simultaneous flowering were selected during the flowering stage. Fresh samples from the second stem at the base were rapidly frozen in liquid nitrogen, ground, and mixed thoroughly. A 0.2 g tissue sample was weighed and added to 1 mL of pre-cooled 95% ethanol in an ice bath, then placed at 4 °C for a duration of 10 min. The sample was centrifuged for 5 min, followed by discarding the supernatant and retaining the precipitate. This process was repeated once, and 1 mL of the pre-cooled extract was added to the precipitate to ensure thorough swirling and mixing. This was placed at 4 °C for 10 min before centrifuging again at 4 °C for another 10 min. The supernatant was retained, and the precipitate was discarded. Light absorption was measured at 540 nm using an enzyme label [36].

2.3.4. Grain Filling Rate

During the wheat anthesis stage, 200 marked main spikes with synchronized flowering and growth on the same day were selected. Samples were collected at 5-day intervals from 5 d after anthesis until grain filling was completed. Ten spikes were sampled each time, then killed in the oven at 105 °C, then weighed after drying at 75 °C, and finally converted into 1000-grain weight to calculate the filling rate of the superior and inferior grains.

Filling rate (g/d) = (C_n+5 − C_n)/5

(4)

where C represents the dry weight of 1000 grains at the time of sampling, and n represents the number of days after anthesis (i.e., 0 d, 5 d, 10 d, 15 d, 20 d, 25 d, or 30 d).

2.3.5. Yield Components and Grain Yield

An investigation was conducted by determining, in three replicates, the number of spikes on wheat plants in two rows per meter in each plot at the milk-ripe stage. The number from each replicate was subsequently converted into the number of spikes per unit area. Twenty-five wheat plants were randomly selected from each plot, and the number of grains per spike was determined. In the maturing stage of wheat, eight rows per meter of threshing were harvested from each experimental plot. When the grain water content was 12.5%, the yield was calculated by weighing the threshing and converting this into an amount per hectare. Whole grains of naturally air-dried wheat were randomly selected (five replicates) to obtain the weight of 1000 grains.

Ten wheat plants exhibiting uniform growth were selected from each plot at the milk-ripe stage, and the spikes of these ten plants were harvested to investigate characteristics of the superior and inferior grains within each spike. Similarly, at the milk-ripe stage, wheat plants of 30 cm with double rows and even potential were selected in triplicate to investigate the superior and inferior grains of wheat spikes. The number of sterile spikelets per spike was determined based on 50 randomly selected spikes from each plot at the milk-ripe stage.

2.3.6. Data Analysis

Microsoft Excel 2007 (Microsoft Corporation, Microsoft Way, Redmond, WA, USA) and Origin 2021 (Northampton, MA, USA) were used to process and analyze tabular data while generating charts. All data were expressed as the mean ± standard deviation. Significant differences among means were separated by the LSD test at the p < 0.05 probability level, and it was reflected by different letters in the figures. The statistical analysis software SPSS (version 13.0, SPSS Inc., Chicago, IL, USA) was used to conduct analysis of variance (ANOVA) and comprehensive data analyses.

3. Results

3.1. Basal Stem Strength

Following the completion of flowering, a gradual decline in the strength of the basal secondary stem was observed in wheat (Figure 2), and this trend remained consistent across planting densities for both varieties. The stem strength in the D1 treatment was significantly greater than that in the other treatments. Stem strength exhibited a maximum decline of 3.08 N after two years of D3 treatment, whereas the maximum decline in stem strength following D4 treatment was 3.73 N. The stem strength in the D3 treatment was significantly greater than that in the D4 treatment.

3.2. Reactivation and Transport of Stem Dry Matter

As shown in Figure 3, in 2021–2023, the post-anthesis CDMA to grain increased with an increase in planting density. The D1 treatment showed the lowest CDMA, which was significantly lower than that of the other treatments. In 2022–2023, under the D4 treatment, the BN5819 variety showed the highest CDMA (72.18%), and the CDMA ranged 52.18–72.18% with planting density. From 2021 to 2023, the CDMA ranged 66.97–72.18% and 61.99–68.12% in the D4 and D3 treatments, respectively. CDMA change in the D3 treatment was significantly less than that in the D4 treatment. ¹³C isotope labeling of stems revealed that the contribution of stem transport substances to the filling of superior and inferior grains differed (Figure 4). Quantitative analysis showed that the contribution of stem transport substances to the filling of inferior grains was greater than that of superior grains. With the increase of planting density, the contribution of stem transport substances to both strong and weak grains increased. During the two years, the carbon accumulation of superior and inferior grains treated with D4 and D3 treatments was significantly different. Compared with D3 treatment, the accumulation of ¹³C in superior and inferior grains under D4 treatment increased by 61.54% and 46.88%, respectively. However, there was no significant difference between D3 treatment and D2 and D1 treatments in the contribution of assimilates to inferior grain filling.

3.3. Stem Fructan Hydrolase

The activity of fructan hydrolase in stems was significantly influenced by planting density (Figure 5). The activity of fructan hydrolase in stems increased gradually with increasing planting density. Enzyme activity in the BN4199 variety was higher than that in the BN5819 variety under the same planting density, and trends were consistent over the two years. There was no significant difference in enzyme activity between the D3 and D4 treatments; however, the enzyme activity of the D3 treatment was significantly higher than that of the D1 and D2 treatments. The enzyme activity of D3 treatment increased by a maximum of 42.41% and 18.30% compared to D1 and D2 treatments.

3.4. Grain Filling Rate

According to Figure 6 and Figure 7, in the years 2021–2022 and 2022–2023, both strong and weak grain filling decreased with the increase in planting density in the same period after flowering of the two varieties. The filling rates of the superior and inferior grains under D1 treatment were significantly higher than those under the other treatments, and the filling rates of the superior and inferior grains under the D3 treatment were significantly higher than those under D4 treatment. At the same planting density, the grain-filling rates of the two varieties of superior and inferior grains showed a trend of increase followed by decrease. For both varieties in both years, the maximum filling rate of superior grains appeared 20 days after anthesis, while the maximum filling rate of inferior grains appeared 15 days after anthesis.

3.5. Wheat Spike and Grain Number

The number of sterile spikelets per spike was significantly affected by planting density (

Table 2. The number of sterile spikelets and the number of superior and inferior grains in the spikes of wheat at milk-ripe stage under different planting densities. Means within a column followed by different letters are significantly different (p < 0.05) according to LSD’s multiple range test. **, significantly different at p < 0.01; *, significantly different at p < 0.05; NS, not significant.

Variety	Planting Density	Sterile Spikelet (Number/Spike)	Inferior Grains (Number/Spike)	Superior Grains (Number/Spike)	Inferior Grains (Number/per Area)	Superior Grains (Number/per Area)
2021–2022
BN5819	D1	2.10 d	10.83 a	22.88 d	670.50 a	2417.00 d
	D2	2.95 c	8.00 b	26.63 c	533.00 b	2680.00 c
	D3	3.79 b	6.50 c	29.88 b	526.00 b	4481.00 a
	D4	4.41 a	4.33 d	35.25 a	506.00 b	3214.50 b
BN4199	D1	2.10 c	12.50 a	26.13 d	731.00 a	1715.00 d
	D2	3.54 b	11.17 b	28.00 c	618.00 b	2599.00 c
	D3	4.26 a	8.33 c	29.75 b	437.00 c	4251.50 a
	D4	4.49 a	6.17 d	32.38 a	336.00 d	3085.00 b
F-Value	Variety level (V)	8.26 **	188.48 **	3.37 NS	1.14 NS	13.26 *
	Planting density (P)	115.85 **	236.76 **	188.94 **	166.65 **	1316.09 **
	V × P	2.26 NS	3.78 *	20.24 **	41.56 **	27.76 **
2022–2023
BN5819	D1	2.74 d	10.33 a	23.38 d	1176.00 a	2509.00 d
	D2	4.44 c	8.33 b	25.38 c	760.50 b	3097.00 c
	D3	4.97 b	7.33 c	27.50 b	449.50 c	4106.50 a
	D4	6.41 a	3.67 d	30.63 a	418.00 c	3380.50 b
BN4199	D1	2.23 d	12.67 a	25.38 d	1152.00 a	3094.50 d
	D2	3.03 c	10.67 b	27.13 c	1106.50 a	3440.50 c
	D3	4.46 b	8.33 c	30.63 b	817.00 b	4095.00 a
	D4	5.90 a	7.33 d	33.63 a	680.00 c	3763.00 b
F-Value	Variety level (V)	40.28 **	490.00 **	615.31 **	48.60 **	376.48 **
	Planting density (P)	262.68 **	142.65 **	479.81 **	184.44 **	129.91 **
	V × P	5.50 **	6.53 **	5.06 **	18.80 **	6.74 **

). The number of sterile spikelets increased with increasing planting density during both wheat growth stages, and both varieties exhibited similar performance. During the two-year period, plants in the D4 treatment exhibited the highest number of sterile spikelets, whereas those in the D3 treatment showed a significantly lower number than that in the D4 treatment, with a maximum decrease of 24.41%, but significantly higher number than plants in either the D1 or D2 treatment. The interaction of year, variety, and planting density had a very significant relationship with the number of sterile spikelets per spike. The number of superior and inferior grains per plant decreased with increasing planting density, as shown in Table 2. In both wheat growth stages, the number of inferior grains per plant under D1 treatment was the highest, and the number of inferior grains per plant under D3 treatment was significantly higher than that under D4 treatment, with a maximum increase of 99.73%. The number of superior and inferior grains in the wheat population was significantly influenced by planting density. The number of inferior grains in the wheat population decreased with increasing planting density; however, the numbers of superior and inferior grains varied under different planting densities. Specifically, the number of superior grains in the D3 treatment was significantly higher than that in the other treatments; the number of inferior grains in the D3 treatment was also significantly higher than that in the other treatments. ANOVA showed that sowing density has a significant influence on the number of sterile spikelets, kernel number per plant, and kernel number per unit area in 2021–2023. The number of sterile spikelets, kernel number per plant, and kernel number per unit area in 2022–2023 were significantly affected by different varieties. However, their interaction had a significant effect on the panicle traits in 2022–2023.

3.6. Production Factors and Yield

Different planting densities had significant effects on both yield components and yields (

Table 3. Yield and component factors influenced by various planting densities. Means within a column followed by different letters are significantly different (p < 0.05) according to LSD’s multiple range test. **, significantly different at p < 0.01; *, significantly different at p < 0.05; NS, not significant.

Variety	Planting Density	Spike Number (×104 ha−1)	Grains per Spike	1000-Grain Weight (g)	Grain Yield (kg ha−1)
2021–2022
BN5819	D1	435.83 c	43.77 a	48.92 a	8704.39 b
	D2	548.33 b	40.14 b	46.23 b	9021.78 b
	D3	647.50 a	38.25 c	43.40 c	9529.03 a
	D4	670.83 a	36.32 d	43.03 de	9094.25 b
BN4199	D1	521.67 d	43.85 a	45.20 a	8731.93 c
	D2	664.17 c	38.90 b	43.78 b	9347.87 b
	D3	710.00 b	38.53 b	42.47 c	9909.47 a
	D4	755.00 a	36.90 c	41.53 c	9799.11 a
F-value	Variety level (V)	108.59 **	0.14 NS	565.31 **	50.06 *
	Planting density (P)	225.32 **	175.96 **	278.33 **	21.77 **
	V × P	2.52 NS	3.01 **	22.04 **	2.31 NS
2022–2023
BN5819	D1	517.50 d	44.27 a	42.75 a	7952.94 c
	D2	626.67 c	40.32 b	42.07 ab	8550.77 b
	D3	706.67 b	38.92 c	41.38 b	9438.45 a
	D4	749.17 a	36.13 d	40.93 b	9221.06 a
BN4199	D1	552.50 d	43.70 a	41.64 a	8460.19 c
	D2	681.67 c	39.92 b	40.38 b	8876.86 bc
	D3	763.33 b	37.98 c	39.74 bc	9325.41 a
	D4	820.00 a	34.30 d	38.70 c	9279.76 ab
F-value	Variety level (V)	28.53 *	3.71 NS	24.95 *	18.17 NS
	Planting density (P)	174.63 **	214.75 **	6.33 **	23.73 **
	V × P	0.80 NS	1.64 *	0.32 NS	1.56 NS

). The number of spikes per unit area increased proportionally with planting density. BN4199 exhibited a significantly higher number of spikes per unit area than BN5819 at the same planting density. Additionally, plants in the D3 treatment showed a significantly lower number of spikes per unit area than those in the D4 treatment, but this number was still significantly higher than those in either the D1 or D2 treatment. The number of grains per spike decreased with increasing planting density, and the number of grains per spike in the D3 treatment was significantly higher than that in the D4 treatment. The 1000-grain weight of wheat decreased with increasing planting density. The 1000-grain weight in the D1 treatment was significantly higher than that in the other treatments, whereas the 1000-grain weight in the D3 treatment was not significantly different from that of plants in the D4 treatment. As shown in Table 3, wheat yield first increased and then decreased with increasing planting density. In 2021–2022, under the D3 treatment, the BN4199 variety exhibited the highest grain yield, reaching 9909.47 kg/ha. The grain yield from the D1 treatment was significantly lower than that from the other treatments, and the grain yield from the D3 treatment was higher than that from the D4 treatment. However, there was no significant difference between the D3 and D4 treatments. According to the ANOVA, sowing density had a significant influence on grain yield and production factors in 2021–2023. Different varieties had significant effects on spike number and 1000-grain weight. Their interaction significantly affected the number of grains per spike in two years.

3.7. Correlation Analysis

Based on the correlation analysis of CDMA, 1000-grain weight, fructan hydrolase activity, basal stem strength, number of spikes per unit area, and grain yield (Figure 8), a significant negative correlation was observed between CDMA and stem strength, whereas a significant positive correlation was observed between CDMA and fructan hydrolase activity (correlation coefficient = 0.70). Surprisingly, there was a significant negative correlation between CDMA and 1000-grain weight, with a correlation coefficient as high as –0.72. The results also confirmed a significant positive correlation (correlation coefficient = 0.64) between the number of spikes per unit area and grain yield.

4. Discussion

Planting density regulates the retransport of post-anthesis stem assimilate activation during intensive agricultural production [37]. Suitable crop populations can coordinate stem assimilate transport and reduce the risk of stem lodging [38]. Arduini et al. [39] studied the activation of dry matter accumulation in wheat according to sowing rate and found that the reactivation of dry matter from vegetative growth to grains after flowering could be increased under high sowing rate conditions. However, another study showed that an increase in planting density led to negative effects, including a decrease in mechanical strength and number of vascular bundles in stems, thus reducing the lodging resistance of stems [40]. In the present study, a higher planting density increased the transport of dry matter to grain filling after anthesis, which was conducive to continuation of the grain filling process after anthesis (Figure 3). Correlation analysis showed that CDMA was significantly negatively correlated with stem strength, with a correlation coefficient of –0.69 (Figure 8). The excessive transfer of dry matter from stem to grain reduces stem strength. Compared with D3 treatment, the maximum stem strength decreased by 1.37 N in D4 treatment, which increased the risk of stem lodging after anthesis (Figure 2). Therefore, in wheat cultivation, it is necessary to consider the contradiction between stem material transport and the decline in stem strength after transport and enhancement of remobilization and activation ability of dry matter to grain filling to ensure that stems can cope with post-anthesis lodging.

Studies have shown that non-structural assimilates in stems exist mainly in the form of fructans [41]. Wang et al. [42] found that high-temperature treatment before wheat anthesis improved the activity of fructan hydrolase and promoted the transport of stem storage assimilates to grains under high-temperature conditions after anthesis. During grain filling, there is a decrease in the fructan content and increase in fructose content in stems, possibly due to the enhancement of fructan hydrolase [43]. The results of this study indicate that an increase in wheat planting density can increase the activity of fructan hydrolase in stems (Figure 5). The variation of enzyme activity ranged from 685.92 to 1448.99 μg/h/g. Compared with D1 treatment, the enzyme activity of D3 treatment was increased by a maximum of 42.41%. Thus, the reactivation ability of stem storage assimilates was enhanced. Further, CMDA was significantly and positively correlated with fructan hydrolase activity (Figure 8), indicating that increased wheat planting density was conducive to enhanced fructan hydrolase activity in stems, thereby improving stem-to-grain transport of storage compounds during post-anthesis grain filling (Figure 3). Zhang et al. [44] believed that optimizing the reactivation and transport capacity of stem storage assimilates to grains and promoting grain filling to the maximum extent could improve grain yield. This study showed that a greater CDMA increased grain weight and improved grain yield. However, while CDMA had little effect on yield improvement under good production conditions, stem strength decreased significantly because of the transport of stored substances in the stem after anthesis, which increased the risk of stem lodging and potential food loss (Figure 2 and Figure 3 and Table 3). Therefore, on the basis of increasing or not decreasing wheat yield, the activity of fructan hydrolase in stems at anthesis stage was decreased by adjusting planting density, thereby reducing CDMA and maintaining higher post-anthesis stem strength, thus reducing stem lodging risk and potential grain yield loss.

Planting density can affect grain yield by altering the number of spikes per unit area, number of grains per spike, and 1000-grain weight of wheat [30]. The optimal planting density can determine the relationship between crop populations and individuals to maximize the use of natural resources and maximize yield per unit area [45]. Some studies have shown that a higher plant density increases the wheat population, increases intraspecific competition, intensifies competition for resources between individual plants, and significantly reduces the ability of wheat plants to acquire assimilates [46,47]. In this experiment, with an increase in planting density, the number of spikes per unit area increased significantly, and ultimately, the grain yield increased. However, the number of grains per spike and 1000-grain weight decreased significantly (Table 3). Correlation analysis showed that grain yield was significantly positively correlated with the number of spikes per unit area (Figure 8). The difference in spike number per hectare between the D3 and D4 treatments of BN4199 in the second year was 56.67 × 10⁴ (Table 3). Therefore, the increase in wheat grain yield mainly depends on the increase in spike number per unit area. The reason why increasing the seeding rate can improve yield may be that a significant increase in spike number per unit area offsets a decrease in the number of grains per spike and 1000-grain weight (Figure 9 and Table 3). Although studies have indicated that spike numbers have little potential to increase wheat yield under ultra-high-yield conditions, it is necessary to ensure sufficient spike numbers to achieve a high yield [48].

The increase of seeding amount could increase the number of tillers, and partly compensate for the decrease of the number of fertile spikes per plant [49]. Under high-density conditions, the number of grains per spike is inhibited mainly by abortion of the top and bottom spikelets, thereby reducing wheat yield [22]. Our results are partially consistent with those of these previous studies. In this study, the number of sterile spikelets per plant increased under the condition of high planting density. At the same time, the decrease of inferior grain per plant was 5.34–6.66, and the decrease of superior grain per plant was 6.25–12.37 (Table 2). The above results indicate that increasing planting density mainly regulates the number of grains by reducing the number of superior grains per plant, ultimately reducing the number of grains per spike and the yield of wheat (Table 2 and Table 3). Relevant studies have shown that under high planting densities, the decrease in winter wheat yield is mainly related to a lower number of grains per spike and smaller diameter of grains [50]. Slafer et al. [51] and Mondal et al. [52] contended that an increase in wheat yield was mainly achieved through an increase in the number of grains per square meter rather than through an increase in the average grain weight. We observed that, while a high planting density led to a decrease in the number of inferior grains per unit area and per plant, it significantly increased the number of superior grains in the population (Table 2). The number of superior grains per unit area of D3 treatment increased by a maximum of 147.90% compared with D1 treatment. In two years, the 1000-grain weight was significantly reduced after D3 treatment. However, the population structure of wheat was coordinated under this density, and the number of grains per spike was stabilized while the number of spikes was increased. This planting density is an important way to achieve high yield in wheat (Figure 9, Table 2 and Table 3).

Grain filling plays a crucial role in determining grain weight, which is an important component of wheat yield. High-density planting can impede grain filling and subsequently reduce grain weight [22]. Feng et al. [53] posited that during grain formation, inferior grains contribute more to the increase in wheat yield than superior grains. The temporal and spatial variability of inferior grains is greater, rendering them more susceptible to environmental influences than superior grains [54]. The filling rate of superior grains was significantly higher than that of inferior grains at the same density (Figure 6 and Figure 7). With increasing planting density, both superior and inferior grain-filling rates decreased. Compared with the filling rate of superior grains, the filling rate of inferior grains appeared five days earlier, resulting in a lower filling rate and shorter filling time, which were mainly reflected in smaller particle sizes and lower grain weights (Figure 6 and Figure 7 and Table 3).

The accumulation and transfer of dry matter in the vegetative organs after anthesis is an important source of yield improvement [55]. In harsh natural environments, increasing wheat planting density can compensate for yield reductions by enhancing the contribution of pre-anthesis assimilates to grains [11]. During the rapid wheat filling period, stem dry matter decreases, and at the same time, transport of assimilates can promote grain filling [56]. Correlation analysis revealed a significant negative association between CDMA and 1000-grain weight. This indicates that post-anthesis stem storage assimilate transport did not increase 1000-grain weight; the contribution of stem storage assimilates to grain filling was small as a proportion of total grout assimilates, and the effect of these assimilates on grain yield was not significant in the process of grain filling (Figure 8). The grain-filling process remains unaffected by CDMA during years of favorable production. A previous report indicated that non-structural carbohydrates stored in stems not only contributed to grain filling but also played a crucial role in enhancing reservoir strength during the initial stage of grain filling [57]. In this experiment, a higher planting density was found to increase the contribution rate of post-anthesis stem assimilation to grains (Figure 3), which promoted grain filling and increased grain weight, mainly affecting the weight of inferior grains (Figure 4, Figure 6 and Figure 7). By quantifying the contribution of stem assimilate transport to the filling of superior and inferior grains, we determined that the contribution of stem assimilate transport to the filling of inferior grains exceeded that of superior grains. Reactivation of stem assimilates primarily facilitated yield formation in inferior grains, enhancing their filling rate, which consequently increased their weight (Figure 4).

5. Conclusions

Optimization of planting density enables the regulation of wheat population structure and stem characteristics. The results of our two-year field experiment showed that increasing wheat planting density to an appropriate level can better balance the relationship between strong stems and high yields. The activation and retransport of post-anthesis storage compounds can be reduced by reducing the activity of stem enzymes to maintain the strength of stems post-anthesis, which can effectively limit the occurrence of post-anthesis lodging. Simultaneously, the transfer of stem storage assimilates primarily facilitates the filling of inferior grains, thereby augmenting their weight. In this study, a planting density of 270 × 10⁴ plants/ha was shown to maintain stem strength after anthesis and stabilize the number of grains per spike by increasing the number of spikes, which compensated for the decrease in 1000-grain weight and led to a high wheat yield. The cultivation strategy that achieved a high and stable yield was planting at a density of 270 × 10⁴ plants/ha. However, the impact of planting density on the quality of superior and inferior grains remains uncertain, necessitating further investigations into the regulatory effects of planting density on grain quality for multiple varieties and across multiple years.