Effect of Planting Patterns and Seeding Rate on Dryland Wheat Yield Formation and Water Use Efficiency on the Loess Plateau, China

Abstract

Dryland winter wheat (Triticum aestivum L.) production plays an extremely important role in the southeast of the Loess Plateau. Planting patterns have great influence on improving soil water storage and yield, and should be matched with different seeding rates. In order to assess the effect of different sowing methods on the drought resistance and stable yield of dryland wheat, a field experiment was conducted in Wenxi County Dryland Wheat Experimental Base in Shanxi Province, China. In the current study, the effects of three planting techniques (drilling sowing, furrow sowing, and film-mulched sowing) and four seeding rates (150, 225, 300, and 375 kg ha⁻¹) were examined on water storage, dry matter formation, yield, and water use efficiency (WUE). The results showed that furrow sowing (FS) and film-mulched sowing (FM) treatments increased soil water storage in the 0–300 cm soil layer at overwintering and jointing stages. In addition, FS and FM increased soil water consumption in the 0–300 cm soil layer from overwintering to maturity of wheat. Furthermore, FS and FM significantly increased the dry matter accumulation from the overwintering to the mature stage, promoted its accumulation in vegetative organs and translocation to grains after anthesis, viz., increased yield by 6.2% and 7.9%, and WUE by 4.6% and 5.3%, respectively, as compared with those of the drilling sowing (DS) treatments. Pearson’s correlation analysis showed that grain yield had a significantly positive correlation with soil water storage at overwintering and jointing. Moreover, grain yield was significantly positively correlated with soil water consumption in the 0–300 cm soil layer from jointing to maturity. Additionally, the seeding rate of 150 kg ha⁻¹ with FS could obtain higher WUE and grain yield. Therefore, it is strongly recommended that the seeding rate of 150 kg ha⁻¹ is used with FS to improve the grain yield and WUE of dryland agricultural systems in China.

1. Introduction

Wheat (Triticum aestivum L.) is one of the main cultivated crops in the world, and essential to ensure world food security. Rainfed wheat accounts for one-third of the wheat planting area in China. The Loess Plateau, a typical rainfed farming area in China, makes up approximately 40% of rainfed farming in China [1], which has the problem of variable and unstable yield due to the mismatching between the time of precipitation (July–September) and the timing of the water demand of the wheat (October–June). Thus, the scarcity of water is one of the main restriction factors for the yield of wheat in this area [2].

The appropriate planting pattern is an important cultivation measure which can promote the collection of rainwater in soil, root growth of crops, and the efficient use of water during the growth period. This affects the accumulation of dry matter and transportation of nutrients, which can ultimately impact the growth, development, and yield of wheat [3,4,5]. In recent years, most of the studies were focused on the effect of planting patterns of drilling sowing (DS), furrow sowing (FS) [6], wide-space sowing [7], film-mulched sowing (FM), and plastic film mulching [8,9]. DS is the main planting pattern in dryland due to its simple operation, providing a time- and labor-saving impact, but the yield fluctuates with the amount of rainfall. FM can effectively inhibit soil water evaporation, reduce ineffective water consumption, and increase upper soil moisture in the early growth stage and deep soil moisture in the late growth stage. It also increased soil temperature, which facilitates the high spike number, yield, and WUE in semi-arid areas of the world [10]. FS has the advantages of rain collection, drought resistance, frost prevention, wide and narrow row spacing, ventilation, and light transmission, which can improve the seeding quality and yield of dry land wheat. In previous studies, Dong et al. [11] showed that FS was beneficial to water storage and soil moisture conservation, increasing surface temperature, water stability, and ultimately on wheat yield. On the other hand, Zhao et al. [12] found that FS was beneficial in increasing the total water consumption of wheat during its growth period, thus improving the yield and WUE. At present, the plastic film-mulched sowing and FS are the leading technologies used in wheat production. However, plastic film mulching increases the cost, and may also cause environmental pollution [13,14].

In addition, the seeding rate also plays an important role in affecting the formation of wheat yield, which has a significant regulatory effect on the construction of wheat population structure, and on the coordinated development of material accumulation and yield factors [15,16]. The optimum seeding rate could ensure a reasonable population structure before winter, improve the lodging resistance ability in the late growth period, and increase the final yield through an increase in the number of panicles [17,18,19]. Tao et al. [20] showed that when the sowing density was 270 × 10⁴ plants ha⁻¹, the dry matter accumulation at each stage of the plant was increased, and the wheat yield and water use efficiency were significantly improved. Furthermore, several studies have shown that crop yield of winter wheat showed a quadratic curve relationship with seeding density, and it reached a maximum level at the optimum seeding rate [21,22], indicating that the seeding rate also affects grain yield. However, the optimal seeding rate, which was suitable for FS, was still a major concern.

To date, many studies have examined the effect of planting patterns on soil water storage of dryland wheat; however, there are few studies focused on the mechanism of coordinating the effect of planting patterns on soil water and the effect of adjusting seeding rate on soil water, yield, and efficiency. Therefore, the present study aimed to further analyze the differences of soil water separation, storage, utilization, plant dry matter transportation, and enhancement of winter wheat grain yield using different planting methods at different growth stages, revealing the mechanism of yield increase and efficiency, and matching the optimum seeding rate for wheat. At the same time, these findings could provide a theoretical basis and technical support for high, stable, and efficient winter wheat cultivation in the Chinese Loess Plateau.

2. Materials and Methods

2.1. Site Description

Field experiments were performed at the Agriculture Research Station of Shanxi Agriculture University in Wenxi (35°20′ N, 111°17′ E), Shanxi Province of China in 2017–2019. The site represents the typical semi-arid climate of Northeast Loess Plateau, which has an altitude of 450–700 m, with an average annual air temperature of 11–13 °C, and 447.8 mm annual precipitation averaged over the long term (from 2009 to 2020). Winter wheat is a staple crop for the people who live in this area, which is usually planted in early October and harvested in early June of the following year. The land is left fallow in summer from July to September. During the two observed years, the weather conditions differed notably from the eleven-year average. The precipitation during the growth stages of the two trial years, 2017–2018 and 2018–2019 were 6.5% and 27.8% lower than the average precipitation during 2009–2020, respectively. There was no precipitation during the wintering to jointing stage in 2017–2018 and no precipitation during the sowing to overwintering period in 2018–2019 (

Table 1. Precipitation at the experimental site (mm).

Year	Fallow Period	SS–WS	WS–JS	JS–AS	AS–MS	Total
2009–2020	259.9	70.2	26.1	35.8	55.8	447.8
2017–2018	199.3	152.4	0.0	49.8	16.1	417.6
2018–2019	254.5	0.0	8.9	36.7	23.3	323.4

SS–WS: Sowing stage to overwintering stage; WS–JS: overwintering stage to jointing stage; JS–AS: jointing stage to anthesis stage; AS–MS: anthesis stage to maturity. Fallow period: mid-June to early October; Total: total precipitation.

). Compared with the fallow period of 2018–2019, the precipitation in 2017–2018 was 55.2 mm less, 152.4 mm more for sowing to overwintering, and 8.9 mm less for overwintering to jointing. The precipitation before overwintering in 2018–2019 was 70.2 mm less than that the long term average, and the early growth of winter wheat was severely affected by drought (Table 1).

The soil type at the site was classified as calcareous cinnamon soil using Chinese soil taxonomy. The soil fertility level in the 0–20 cm soil layer before sowing was measured on 10 June 2017 and 11 June 2018 (

Table 2. Soil essential nutrient content of 0–20 cm layer at the experimental site in two years.

Year	Organic Matter (g/kg)	Available N (mg/kg)	Available Phosphorus (mg/kg)	Available Potassium (mg/kg)
2017–2018	9.17	34.53	20.29	98.29
2018–2019	10.66	41.92	19.82	96.34

).

2.2. Experimental Design and Treatments

The experiment used a split-plot design with three replications. Three different planting patterns were used as main plot treatments: (1) drill sowing (DS) as control; (2) furrow sowing (FS); and (3) film-mulched sowing (FM), and four seeding rates (150, 225, 300, 375 kg ha⁻¹) were randomized in the subplot. The DS technique was a traditional planting method with 20 cm row spacing, and 3–5 cm sowing depth. For FS, wall trench seeding was done using a saw tooth disc furrow opener, which accomplished all the necessary actions of row cleaning seeding and fertilizer placement in a single operation. The furrow depth was 7–8 cm, and the ridge height was 3–4 cm. The fertilizer was applied in the middle part of the bottom of the furrow, and the seeds were planted in the wet soil on both sides of the furrow at 3–4 cm from the bottom of the ditch. For FM, the ridge was raised, the film was covered, and the seeding was completed on the same day. The ridge (10 cm high, 40 cm wide) and furrows (20 cm wide) were alternated and mulched with transparent polyethylene film (0.08 cm thick and 1.2 m wide). The film was spread over the ridges and its edges were covered by soil where wheat was planted, and two rows of wheat were planted on the ridge side and furrow film. Rainwater runoff was collected by the slopes of ridges and stored under the furrows (Figure 1). Tall stubble residue from the previous season (20–30 cm) remained to reduce water evaporation and increase soil organic carbon for the next season. In the middle of July, deep ploughing was done to control weeds. Later in August, rotary tillage and land leveling were performed before planting. The sowing dates of wheat were 1 October 2017 and 2 October 2018. Before sowing, nitrogen, phosphorus, and potassium were applied at rates of 150, 150, and 75 kg ha⁻¹, respectively. The plastic cover in FM treatment was removed in the middle of May. Weeds and pests were well controlled by hand, and no irrigation was applied across the two experimental years.

2.3. Sampling and Measurement

The representative wheat cultivar Yunhan805 was used. Before plot preparation, a 3 m deep profile pit was excavated, and soil samples were taken from 0 to 300 cm depth in 20 cm increments using the cutting ring method described by Dama et al. [23]. Soil bulk density and soil water content (SWS, %) were measured for each soil layer. SWS was determined and expressed as soil water storage (mm), in which SWS was measured in 20 cm increments to a depth of 300 cm at the beginning of plastic mulching and in the middle of May. SWS were also determined at four plant developmental stages. In each plot, four random soil samples were taken between plant rows. For each soil sample, SWS was calculated as the ratio of the mass of water per mass of dry soil. The dry weight of the soil sample was obtained by placing fresh soil samples in an oven at 105 °C for 72 h.

Soil water storage (SWS) for a given layer was calculated as follows [24]:

W (mm) = GSW (%) × ρb × SD

where ρb (g cm⁻³) is the dry soil bulk density of a given soil layer and SD (mm) refers to soil depth.

2.4. Ear Number, Grain Yield, and Water Use Efficiency (WUE)

At maturity level, plants of 1 m² from each plot were randomly sampled from the inner rows to determine ear number. Grain yield was obtained by harvesting all plants in the plot.

Economic water use efficiency (WUE, kg ha⁻¹ mm⁻¹) was calculated using the following formula:

WUE = Y/ET

where Y is grain yield (kg ha⁻¹) and ET (mm) is evapotranspiration in the crop growing season. The experiment was conducted under rainfed conditions, and no irrigation was provided over the growing season. Evapotranspiration was the sum of precipitation (mm) in the growing season and consumption of soil water storage (mm) in 0–300 cm depth from sowing to plant maturity.

Ten winter wheat plants were randomly selected from each plot at flowering and maturity stages in the two trial years. The total dry matter of roots and shoots from these ten plants at each stage was found after oven drying at 105 °C for one hour and then at 70 °C until a stable weight was reached. The biomass is the total weight of the different parts [25]. The dry matter remobilization amount (RA) of pre-anthesis and its contribution to grain (RC) were calculated using the equation below [26]:

RA = total aboveground dry matter at anthesis without grain at maturity

$$RC=\frac{RA}{grain yield}\times 100\%$$

2.5. Statistical Analysis

The SAS package (SAS Institute Inc., Cary, NC, USA) was used for the analysis of variance (ANOVA) and correlation analysis as well. Least Significance Difference (LSD) tests at p = 0.05 were used to compare means.

3. Results

3.1. Effects of Planting Patterns and Seeding Rate on Soil Water Storage (SWS)

Change in soil water storage varied with growth stages, planting patterns, and seeding rate treatments during the two experimental years (Figure 2). SWS of 0–300 cm at overwintering was significantly increased by 25.1 and 20.1 mm, and at jointing by 19.7 and 13.7 mm in the FS and FM treatments compared to DS in the two experimental years, respectively. SWS of 0–300 cm at anthesis showed a different tendency, from 2017 to 2018, the FS and FM treatments increased by 12.3 and 19.9 mm. Whereas, from 2018 to 2019, the SWS decreased by 1.9 and 7.6 mm under FS and FM compared with DS. At the maturation stage, the SWS of 0–300 cm were reduced by 6.8 and 9.7 mm, which was significantly higher in FS than in FM for both experimental years. No significant differences were observed in 2017–2018 between FS and FM for R150, R225, and R300 (Figure 2).

Compared with DS, FS and FM considerably enhanced the SWC at various seeding rates. With the increase in seeding rate, the SWS of the 0–300 cm soil layer during winter dormancy decreased, and the SWS capacity of R150 in FS and FM were 13.5 and 14.4 mm higher than the R225 treatment on average over two experimental years, while there was no difference between R150 and R225 in 2018–2019 (Figure 2A). At jointing, the SWS capacity of R150 has a similar tendency as winter dormancy, which was 3.10 and 5.39 mm higher than the R225 treatment, respectively (Figure 2B). At maturity, the soil water storage in the 0–300 mm layers in FS and FM were 14.9 and 8.5 mm less at R150 than at R225 (Figure 2D). During the two growing seasons, the initial SWC at sowing did not show a statistically significant difference among different treatments. It can be seen that the seeding rate of R150 was more conducive to water storage in the early growth period under FS.

3.2. Effects of Planting Patterns and Seeding Rate on Soil Water Consumption (SWC)

In comparison to DS, soil water consumption under the FS and FM treatments from sowing to overwintering was significantly reduced by 27.1 and 30.6 mm (Figure 3A), whereas from jointing to anthesis, it increased by 10.7 and 11.3 mm (Figure 3C). From anthesis to maturity the average increases were 15.8 and 12.0 mm, respectively. Soil water consumption from anthesis to maturity under FS was 3.8 mm higher than FM (Figure 3D).

Soil water consumption of winter wheat showed an increasing tendency with increasing seeding rate during sowing to overwintering in all three planting patterns (Figure 3A). R150 in both FS and FM (both seasons) had a significantly lower soil water consumption than R300 and R375 from the sowing to overwintering stages, while no significant difference for R150 and R225 was observed between FS and FM (Figure 3A). From the jointing to anthesis and anthesis to maturation stages, soil water consumption of R150 in FS and FM was 11.6 and 5.7 mm, and in the anthesis to maturation stages it was 6.5 and 11.4 mm higher than R225 treatment on the average over two experimental years (Figure 3C,D). These results show that the optimal seeding rate for FS treatments could reduce soil water storage up to the jointing stage, leading to more soil water available for later growth stages, and the plants treated with the optimal seeding rate also consume more soil water from flowering to maturity.

3.3. Effects of Planting Patterns and Seeding Rate on Dry Matter Accumulation at Four Growth Stages

The planting patterns and seeding rate significantly affect the dry matter accumulation above ground. Two-way interactions of S × R significantly affected the dry matter accumulation of the overwintering, flowering, and mature periods. Y × R, Y × P, and Y × R × P -highly (p ≤ 0.01) or significantly (p ≤ 0.05) affected the amount of dry matter accumulation in each growth period (

Table 3. Significance analysis (F value) of interactive effects of year, planting patterns, and seeding rate on the dry matter accumulation at four growth stages over two years.

Parameter	WS	JS	AS	MS
Year (Y)	14.039	1.490	233.075 **	233.718 **
Planting patterns (P)	197.711 **	265.635 **	615.046 **	1028.671 **
Seeding rate (R)	11.586 **	4.665 **	1106.158 **	2288.279 **
Y × P	16.989 **	5.088 *	54.425 **	36.501 **
Y × R	199.660 **	156.056 **	49.778 **	39.826 **
P × R	6.137 **	1.524	18.913 **	57.483 **
Y × P × R	14.566 **	4.765 **	15.260 **	22.459 **

* and ** indicate significance at p ≤ 0.05 and p ≤ 0.01, respectively.

).

Compared with the DS method, the dry matter accumulation significantly increased at overwintering by 23.2% and 24.15%, jointing by 12.6% and 18.7%, flowering by 9.8% and 10.8%, and maturity by 9.9% and 12.2%, respectively. There was no significant difference in dry matter accumulation at the jointing stage (2018–2019) or flowering stage (2017–2018). As the seeding rate increased, the dry matter accumulation at the flowering and maturity period showed a downward trend under furrow sowing and film-mulched sowing. The dry matter accumulation at the maturity stage was the highest in the treatment of the R150 seeding rate, and dry matter accumulation at the maturity stage R150 in DS, FS, and FM was 403.1, 818.55, and 854.54 kg ha⁻¹ higher than R225 treatment on average, respectively (Figure 4).

3.4. Accumulation and Translocation of Dry Matter at Pre- and Post-Anthesis Period

The main effect of these three factors was significant (p ≤ 0.05) or highly significant (p ≤ 0.01), except for dry matter amount of translocation before anthesis for planting patterns and years. Two-way interactions of P × R and Y × R × S were significant (p ≤ 0.05) and so highly significant (p ≤ 0.01), it effected the amount of dry matter translocation before anthesis, after anthesis, and its contribution rate to grain (

Table 4. Significance analysis (F value) of interactive effects on the accumulation and translocation of dry matter at pre- and post-anthesis period in two years.

Parameter	DMABA		DMAAA
	TA (kg ha−1)	CG (%)	AA (kg ha−1)	CG (%)
Year (Y)	0.846	49.764 *	504.748 **	49.764 *
Planting patterns (P)	1.102	24.296 **	51.813 **	24.296 **
Seeding rate (R)	49.724 **	10.012 **	398.647 **	10.012 **
Y × P	2.253	0.674	1.690	0.674
Y × R	1.932	3.762 *	10.791 **	3.762 *
P × R	8.214 **	2.786 *	29.492 **	2.786 *
Y × P × R	5.550 **	6.588 **	5.574 **	6.588 **

DMABA: Dry matter assimilation before anthesis; DMAAA dry matter assimilation after anthesis; TA: Translocation accumulation; CG: Contribution ratio to grain; AA: Assimilation accumation. * and ** indicate significance at p ≤ 0.05 and p ≤ 0.01, respectively.

)).

The contribution rate from pre-anthesis dry matter translocation to grains under different treatments of three planting patterns and four different seeding rates was 30.65–39.84%, and the post-anthesis dry matter accumulation was 60.16–69.35% (Table S1). It can be seen that the grain yield mainly came from the accumulation of dry matter after anthesis.

On average over two experimental years, FS and FM treatments significantly increased the dry matter assimilation after anthesis (DMAAA) by 9.6% and 13.9%, and its contribution rate to grain by 2.0% and 3.2%, compared with the DS treatment (Table S1), respectively, but the difference was not significant. As the seeding rate increased, the dry matter accumulation after anthesis and its contribution rate to grain decreased and reached the maximum at the seeding rate of R150. Dry matter assimilation after anthesis (DMAAA) of R150 in DS, FS, and FM was 52.5, 270.7, and 389.5 kg ha⁻¹ higher than R225.

3.5. Yield and Yield Attributes and Water Use Efficiency (WUE)

The main effect of these three factors was significant (p ≤ 0.05) or highly significant (p ≤ 0.01) except on thousand kernels weight. The two-way interaction of Y and P significantly affected the spike number. The interaction of Y and R had highly significant effects on spike number, 1000-kernel weight, yield, and water use efficiency of wheat. The interaction of P and R also had highly significant effects on all the five indicators. Three-way interactions of Y × R × S were significant on water use efficiency (p ≤ 0.01), spike number, 1000-grain weight, and yield (

Table 5. Significance analysis (F value) of interactive effects of year, planting patterns, and seeing rate on the yield, its components, and water use efficiency in two years.

Parameter	Spike Number (104 ha−1)	Kernels per Spike	1000-Kernel Weigh (g)	Grain Yield (kg ha−1)	Water Use Efficiency (kg ha−1 mm−1)
Year (Y)	590.967 **	266.230 **	944.841 **	69.342 *	46.025 *
Planting patterns (P)	26.682 **	11.301 **	4.094	27.107 **	18.649 **
Seeding rate (R)	91.102 **	25.382 **	16.280 **	644.983 **	555.393 **
Y × P	3.176 **	2.556	0.460	2.544	0.488
Y × R	9.216 **	2.001	45.474 **	6.851 **	7.940 **
P × R	4.475 **	3.415 **	30.417 **	52.885 **	51.442 **
Y × P × R	5.284 **	2.022	7.814 **	2.700 *	1.759

* and ** indicate significance at p ≤ 0.05 and p ≤ 0.01, respectively.

).

On average over two experimental years, FS and FM treatments significantly increased the spike number by 6.0% and 10%, and kernels per spike 4.5% and 4.0%, respectively, (Table S2) compared with DS. The thousand kernel weight of FS treatments was lower than that of DS and higher than that of FM. Finally, the yield increased by 6.2% and 7.9%. The WUE was significantly increased by 4.6% and 5.3%, but the yield and WUE were not significantly different between FS and FM. As the seeding rate increased, the spike number, yield, and the WUE decreased, reaching the maximum of 150 kg ha⁻¹, and the average yield increased by 229.0 and 293.8 kg, respectively, compared with that of DS. The seeding rates of 150 kg ha⁻¹ with FM achieved the highest grain yield and WUE, which was significantly different from other treatments. It can be seen that deep tillage combined with furrow sowing in a wheat field during the fallow period can increase the yield mainly by increasing panicle number and grain number per panicle, and thus increasing yield significantly.

3.6. Correlation Analysis

Grain yield was significantly positively correlated with SWS_WJ, SWC from overwintering to maturity, DMABA, and DMAAA; it was negatively related to SWSsw in 2017–2019; and non-significant for SWS_WJ in 2018–2019. Spike number and kernels per spike were significantly positively correlated with the DMABA (

Table 6. Correlation coefficients between grain yield, soil water storage, dry matter accumulation and water consumption from sowing to overwintering (SS-WS, SW), overwintering to jointing (WS-JS, WJ), jointing to anthesis (JS-AS, JA) and anthesis to maturity (AS-MS, AM) of wheat.

Parameter	2017–2018				2018–2019
	Spike Number	Kernels per Spike	1000-Kernel Weigh	Yield	Spike Number	Kernels per Spike	1000-Kernel Weigh	Yield
SWS (WS)	0.879 **	0.910 **	−0.654 *	0.864 **	0.708 **	0.557 *	0.194	0.747 **
SWS (JS)	0.839 **	0.887 **	−0.583 *	0.852 **	0.612 *	0.756 **	0.024	0.873 **
SWC (SS–WS)	−0.869 **	−0.904 **	0.649 *	−0.853 **	−0.681 *	−0.542	−0.200	−0.720 **
SWC (WS–JS)	0.692 **	0.661 *	−0.638 *	0.598 *	0.554 *	−0.016	0.389	0.211
SWC (JS–AS)	0.793 **	0.889 **	−0.671 *	0.767 **	0.795 **	0.537	−0.104	0.755 **
SWC (AS–MS)	0.753 **	0.885 **	−0.531	0.766 **	0.667 *	0.738 **	−0.004	0.887 **
DMABA	0.894 **	0.965 **	−0.653 *	0.903 **	0.814 **	0.713 **	−0.084	0.970 **
DMAAA	0.750 **	0.625 *	−0.485	0.798 **	0.371	0.654 *	−0.0049	0.777 **

* and ** indicate significance at p ≤ 0.05 and p ≤ 0.01, respectively.

) and WS_WJ, whereas they were significantly negatively correlated with 1000-grain weight in 2017–2018. Thousand-grain weight wasnegatively correlated with water consumption at JA, in 2017–2018. The spike number and kernels per spike were significantly or extremely significantly and positively correlated with SWS_WJ and SWC_AM, as well as DMABA from 2018 to 2019. SWC_SW was significantly negatively correlated with spike number, whereas SWC_WJ and SWC_JA were positively correlated with spike number. There was a significant or highly significant positive correlation between DMAAA and kernels per spike in 2018–2019.

3.7. Contribution of Water to Yield

In comparison to DS, the contribution rate of water consumption to yield (CRWCR) under the FS and FM treatments from sowing to overwintering was significantly reduced by 7.7% and 6.4% (Figure 5A). However, the CRWCR from jointing to anthesis increased by 2.4% and 2.8%, and from anthesis to maturity it increased by 3.7% and 2.5%, respectively, in both of the experimental years. Moreover, the contribution rate of water consumption to yield from anthesis to maturity under FS was higher than FM (Figure 5D).

The CRWCR of winter wheat showed an increasing tendency with increasing seeding rate during sowing to overwintering. In both seasons, R150 in both FS and FM had significantly lower values than R300 and R375 from the sowing to overwintering stages. Whereas, from the jointing to anthesis and anthesis to maturity stages, the CRWCR of R150 in FS and FM was 2.6% and 2.0%, and anthesis to maturation stages 3.9% and 3.8% higher than the R300 treatment on average over two experimental years (Figure 5C,D). It can be seen from the above that furrow sowing can optimize the water consumption structure of wheat, and balance the contribution of water consumption to yield from jointing to flowering and from flowering to maturity, which is conducive to stable water supply after the jointing of wheat. More water is used from jointing to maturity, which is conducive to ear formation and grain formation of wheat.

4. Discussion

Natural precipitation is the sole source of water supply in the winter wheat area on the Loess Plateau. However, 60% of the precipitation occurs in the fallow season from July to September [27]. Therefore, improving the capacity of SWS and continuing to use water to the later period is crucial for stable and increased crop yield. Relevant studies have shown that suitable planting patterns are beneficial for ensuring soil moisture distribution to crops then coordinating the proper utilization of water in the growth period [12,28,29]. Our results showed that FS and FM significantly reduced the sowing to overwintering water consumption (27.1 and 30.6 mm) in compared with DS. On the other hand, FS and FM increased the jointing–maturing water consumption, which was higher in FM than FS (Figure 2 and Figure 3). In the growth period, FS and FM treatments could increase SWS at the overwintering (25.1 and 20.1 mm) and jointing (19.7 and 13.7 mm) stages, prolong the period of moisture accessibility, and increase the total water consumption (5.2 and 8.9 mm) (Figure 2). Thus, FS improved the capacity of SWS, and optimized the contribution of water to yield, which was consistent with the previous reports’ results that the topsoil of traditional flat sowing wheat fields was relatively loose, so water was prone to evaporate or infiltrate. Conversely, FS could increase the tillage degree of topsoil, slow down the rate of water seepage downward, and play a crucial role in water retention [30,31]. FS increased the 0–200 cm soil water storage by 8.9–19.9 mm during the whole growth period of winter wheat [11].

Optimized planting patterns can ensure that wheat achieves uniform emergence and seedling robustness. Previous studies have shown that planting patterns affected winter wheat population structure, dry matter accumulation and translocation, yield, and its composition [32,33]. A higher amount of dry matter accumulation of wheat is the basis for a high yield of wheat [34]; it is required for population structure, which also can promote photosynthesis and accelerate dry matter transport and accumulation. Wang et al. [35] showed that the dry matter accumulation of wheat increased with the increase in planting density but decreased when it exceeded a certain range. This study showed that the dry matter quality of wheat showed a decreasing trend at the maturity stage with an increased seeding rate. In comparison with DS, FS and FM increased the dry matter accumulation of vegetative organs at the maturity stage (Figure 4). The formation of grain yield of wheat mainly came from the transport of stored substances before flowering and the accumulation of photosynthetic products in functional leaves after flowering, which was closely related to the accumulation, distribution, and transport of dry matter [36]. This study results showed that the dry matter transfer after flowering greatly contributed to the grain yield, which is the main source of increasing grain yield [37,38]. Promoting the transport of dry matter stored in vegetative organs before anthesis to grains, enhancing the assimilation rate of dry matter after anthesis, and increasing the accumulation and distribution ratio of dry matter in grains are the important basis for a high yield in FM and FS. Under the conditions of this experiment, compared with the DS, FS and FM patterns increased the dry matter assimilation after anthesis and filled the grain, which was the main reason behind the high yield of wheat (Table S2).

Different planting patterns will change the physiological metabolic process by changing the population structure as well as affecting the overall growth and development of wheat [39,40]. This ultimately improves the efficiency of water and fertilizer, to increase the yield [20,41,42]. Previous studies have shown that plastic film mulching could increase the yield of wheat in arid areas by more than 30% [43,44,45]. The reason behind the increasing yield is that it significantly increases the ground temperature during the early development stage and compensates for the absence of accumulated warmth before winter, which raises overwintering survival rates and encourages growth during the regreening stage [46,47,48,49]. Furthermore, while increasing SWS by inhibiting evaporation and preserving soil moisture, it transforms ineffective evaporation of water into effective transpiration of plants, improves the WUE [43,45,50], and promotes plant growth and development. Compared with plastic mulch sowing, furrow sowing could also effectively reduce soil moisture evaporation consumption, which is conducive to water storage and soil moisture conservation, as well as conducive to ventilation and light penetration in the later period. It helps to improve the photosynthetic efficiency and ultimately enhance the crop yield and WUE [51]. Xue et al. [52] pointed out that FS mainly increased the yield of winter wheat by increasing the number of effective panicles. FS was conducive to ventilation and light penetration in the later stage, and improved photosynthetic efficiency, so that the grain was larger and more uniform. Mao et al. [53] also pointed out that FS can significantly improve the winter wheat leaf area index (LAI), flag leaf photosynthetic rate, stomata conductance, and transpiration rate, thereby increasing yield. The results of this experiment showed that compared with drilling sowing, FS and FM significantly increased the number of panicles by 6.0% and 10%, the number of grains per panicle by 4.5% and 4.0%, and the yield by 6.2% and 7.9%, respectively (Table S2). The FS and FM strategies also increase the number of effective panicles and the number of grains per panicle, which is consistent with the results of previously reported studies. The rational distribution of population structure is mainly due to the influence of canopy morphology and the internal microenvironment of crops. In addition, the appropriate row spacing can significantly reduce the ineffective evapotranspiration of water and increase transpiration to improve the WUE in dryland.

An optimum seeding rate can reasonably construct the wheat population structure, which is conducive to the coordinated development of the three elements of yield. There have been numerous studies on the effect of seeding rate on wheat grain yield, but the results are not consistent [25,54,55]. With the increase in the seeding rate, soil water consumption gradually increased, and more soil moisture was consumed before jointing under a higher seeding rate [56]. We found that the water consumption increased from sowing to the overwintering stage with the increase of seeding rate. After jointing, the maximum water consumption of winter wheat was observed at the seeding rate of 150 kg ha⁻¹ under FM and FS. Continuing to increase the seeding rate, the water consumption of wheat decreased. The main factor was that winter wheat planted in the soil moisture detection ditch had a more logical population structure, which improved its ability to use soil moisture in dense planting conditions and reduced competition for soil moisture between individuals during the middle and late growth stages. It may be seen that the whole growth period of winter wheat fields adapts to the technology of soil moisture exploration, FS, and water conservation, which is more conducive to the preservation of water in the early and middle stages for further utilization. It can provide sufficient water for the increase of panicle number and the formation of grain number per panicle, and it can resist the subsequent drought to ensure the 1000-grain weight. Thus, the effect of the seeding rate on effective panicle number and grain number per panicle was extremely significant, and the yield decreased with the increase in seeding rate. The yield was the maximum when the seeding rate was 150 kg ha⁻¹, while the seeding rate had no significant effect on 1000-grain weight (Table S2).

Due to the difference in soil moisture caused by the planting pattern, the seeding rate needs to be adjusted to achieve the purpose of higher yield. Under the conditions of this experiment, FS increased soil water consumption in the middle and late stages of growth, and significantly increased the number of panicles at the maturity stage. When the seeding rate was 150 kg ha⁻¹, it was more conducive to the formation of a reasonable population than other seeding rates to achieve a high yield. The technology of FS in winter wheat fields can improve the accumulation ability of natural precipitation, make efficient and reasonable use of soil moisture, and reduce soil water consumption in the early stage, but increase consumption in the middle and late stages of growth. This is conducive to the improvement of spike number and grain formation, as well as optimizes the yield composition, yield, WUE, and economic benefits. Eco-friendly, green, and high-yield production is the main purpose of modern agricultural production. In this study, although the yield of FM was the highest, the residual film left in the soil will cause changes in soil structure, and reduce water permeability, thereby hindering root development, which is not conducive to sustainable agricultural development [13,57]. Therefore, winter wheat fields can use the furrow planting pattern to maintain a certain level of yield, and economic efficiency will improve green and efficient production and cultivation techniques.

5. Conclusions

Compared with drilling sowing (DS), furrow sowing (FS) and film-mulched sowing (FM) increased the soil water storage (SWS), dry matter accumulation at maturity, and the spike number, thus increased the grain yield. This new planting pattern (FS) also maximized water use efficiency (WUE) and grain yield in compared to FM. The highest WUE was at the seeding rate of R150, in which the spike number, dry matter weight at maturity, and assimilation after anthesis was the highest. With the seeding rate of R150, the healthiest plants could be obtained and the number of panicles could be increased. Thus, the yield and WUE could be increased gradually, and could maximize the yield. Considering the economic benefits, ecological benefits, and the difficulty of field operation, FS was more suitable planting pattern to large areas of dryland farming than FM, as well as agricultural areas with similar environmental, soil, and available water conditions.