Effects of Water Application Frequency and Water Use Efficiency Under Deficit Irrigation on Maize Yield in Xinjiang

Abstract

Water conservation is critical for global maize production, particularly in arid regions where water scarcity, exacerbated by climate change, threatens conventional irrigation sustainability. Optimizing irrigation strategies to reconcile water productivity and yield remains a key scientific challenge in water-limited agriculture. This four-year study (2018–2021) evaluated integrated irrigation management that combined frequency and volume adjustments. A field experiment compared three strategies: high-frequency limited irrigation (HL: 2400 m³·hm⁻²), low-frequency conventional irrigation (LC: 2400 m³·hm⁻²), and high-frequency conventional irrigation (HC: 4800 m³·hm⁻²). The four-year mean yield showed that HL (10,793.78 kg·hm⁻²) had a non-significant 18.2% numerical advantage over LC (9129.11 kg·hm⁻², p > 0.05). The WUE for HL reached 3.63 kg·m⁻³, representing an 18.6% numerical increase compared to LC (3.06 kg·m⁻³; p > 0.05). Physiological parameters (plant height + 2.6%, leaf area + 9.9%, SPAD + 1.5%) showed marginal improvements in HL, yet lacked both statistical significance (p > 0.05) and strong yield correlation. Multi-year analyses confirmed no statistically distinguishable differences between strategies (p > 0.05), demonstrating that irrigation frequency adjustments alone cannot reliably enhance drought resilience. These findings caution against advocating for HL as a superior practice, given the statistical equivalence between HL and LC despite water savings, and the non-significant yield gap between HL and HC. Future research must establish causality through models integrating real-time soil–crop–climate feedback prior to recommending altered irrigation regimes.

1. Introduction

Water is a fundamental resource in agricultural production, directly influencing crop growth and development [1]. However, the global disparity in water distribution, exacerbated by escalating scarcity, poses critical threats to agricultural sustainability [2]. Xinjiang, a typical arid and semi-arid region in Northwest China, faces acute water shortages [3,4]. Water scarcity combined with inefficient irrigation practices has led to insufficient water supply for cultivated lands, increasing the risk of agricultural abandonment [5]. Water scarcity, inefficient irrigation practices, and the risk of farmland abandonment in Xinjiang not only undermine regional agricultural productivity but also jeopardize China’s food security [6]. Globally, agricultural land in arid and semi-arid regions is under water stress [7], underscoring the imperative to optimize irrigation strategies for sustainable agriculture.

To address these challenges, drip irrigation technology has emerged as a pivotal solution for water resource management in arid agricultural systems. As a globally adopted water-saving method, drip irrigation has demonstrated effectiveness in arid regions worldwide [8,9], including having well-documented implementations in Xinjiang [10]. This system delivers water precisely to crop root zones, minimizing evaporation and leaching losses, thereby improving water use efficiency (WUE) [11,12,13]. Nevertheless, Xinjiang’s extreme water scarcity strains crop water requirements, necessitating innovative irrigation strategies. Given the critical role of water management under scarcity, selecting maize—a crop sensitive to water stress—provides an ideal model for determining optimal irrigation techniques [14].

Maize (Zea mays L.), a globally vital cereal crop with pronounced water sensitivity [15], is an ideal model for studying water management in arid agriculture. In China, maize occupies the largest cultivated area and achieves the highest grain yield [16], while in Xinjiang, it constitutes a cornerstone of the local agriculture. Maize’s water demand exhibits marked temporal heterogeneity: 40% of its total water requirement occurs during the jointing stage, with 35% consumed during grain filling [17], as evidenced by studies in arid regions (e.g., the U.S. Midwest [18], North China Plain [19], and Sub-Saharan Africa [20]). Water deficit directly impairs maize physiology, manifesting as reduced stomatal conductance and photosynthetic rates under stress conditions [21], thereby depressing yields. Although plastic film-mulched drip irrigation is widely adopted in Xinjiang, farmers’ reliance on suboptimal irrigation schedules compromises the system’s efficiency and long-term sustainability [22]. Under water-limited conditions, irrigation frequency becomes a critical determinant of both WUE and crop performance [23]. High-frequency irrigation maintains stable soil moisture levels, facilitating root water uptake [24], whereas low-frequency irrigation induces moisture fluctuations detrimental to crop health [25]. The conventional irrigation quota of 4800 m³·hm⁻² yr⁻¹ fails to meet crop water demands, resulting in partial field abandonment. Thus, deficit irrigation emerges as a viable strategy to reduce water consumption while maintaining acceptable productivity levels.

This study hypothesizes that high-frequency limited irrigation (HL) maintains root-zone soil moisture stability, thereby improving water use efficiency (WUE) without compromising yield compared to low-frequency conventional irrigation (LC). To test this hypothesis, three irrigation strategies are evaluated: high-frequency conventional (HC), high-frequency limited (HL), and low-frequency conventional (LC). The objectives are to (1) quantify the effects of irrigation frequency and volume on maize growth and yield and (2) identify the optimal strategy for balancing water savings and agricultural productivity. The findings aim to provide a scalable framework for drip irrigation optimization in global arid regions.

2. Materials and Methods

2.1. Experiment Site

The research took place between 2018 and 2021 at an experimental station in Shihezi, Xinjiang, China (86°09′ E, 45°38′ N), located in the western suburbs of Shihezi City in northwestern China. The site, which is at an altitude of 452.8 m above sea level, features a semi-arid climate characterized by an average yearly temperature of 22.46 °C and an annual evaporation reaching 1942 mm. The annual precipitation during the breeding season was 68 mm in 2018, 93 mm in 2019, 42.5 mm in 2020, and 42 mm in 2021 (Figure 1). The groundwater table fluctuates seasonally between 2 and 3 m below the surface. The experimental site contains gray desert soil (USDA Soil Taxonomy) with the following physicochemical properties measured at a 0–20 cm depth using stainless-steel auger sampling: organic matter was measured to be 16.79 g·kg⁻¹, total nitrogen was measured at 1.44 g·kg⁻¹, available phosphorus stood at 26.52 mg·kg⁻¹, and available potassium at 415.98 mg·kg⁻¹. The pH level was recorded as 8.19, with a bulk density of 1.56 g·cm⁻³, a saturated water content of 32.01%, and a soil saturation of 60–80%. Additionally, the key physicochemical parameters of the irrigation water are detailed in

Table 1. Physicochemical properties of irrigation water.

Index	Total Hardness (mg/L) (CaCO3)	Mineralization Degree (mg/L)	(NH4-N) (mg/L)	Permanganate Index (mg/L)	SO42− (mg/L)	CL− (mg/L)	Phenol	EC (mS/m)
Content	155	367	<0.05	13.53	92.09	25.53	<0.002	1

.

2.2. Experimental Design

The experiment utilized plastic film-mulched drip irrigation with water treatments designed based on local water availability, comprising three treatments: (i) high-frequency conventional irrigation (HC, 4800 m³·hm⁻², 100% quota), (ii) high-frequency limited irrigation (HL, 2400 m³·hm⁻², 50% quota), and (iii) low-frequency conventional irrigation (LC, 2400 m³·hm⁻², 50% quota). We arranged the treatments in a randomized complete block design with three replications, resulting in nine experimental plots (20 m × 5.5 m each). Adjacent plots were separated by 2.2 m buffer zones to minimize edge effects (Figure 2a). A combined seeder performed three synchronized operations: (1) laying drip tapes (inner diameter: 16 mm; wall thickness: 0.18 mm; emitter spacing: 300 mm; Xinjiang Drip Irrigation, Sprinkler Irrigation and Water Pipe Water-saving Equipment Co., Ltd.), (2) applying plastic film mulch, and (3) sowing maize seeds. The irrigation system operated at 0.1–0.15 MPa, with each emitter delivering 2.0 L·h⁻¹. Drip tapes were spaced 110 cm apart, and each plot had independent water meters and fertilizer tanks for precise irrigation and fertigation control. Maize was planted in alternating wide–narrow rows (80 cm wide, 30 cm narrow), achieving a target density of 1.26 × 10⁵ plants hm⁻² through a 14.4 cm within-row spacing (Figure 2b).

2.3. Material

The experimental maize variety was ZD985 (green maize), provided by Beijing Denong Seed Technology Co., Ltd. (Beijing, China). All plant experiments and field studies adhered to the relevant institutional regulations, as well as to national (Ministry of Agriculture and Rural Affairs of the People’s Republic of China) and international guidelines for field research (CIMMYT). The fertilizers included the following: urea (N content ≥ 46.4%, granular; Xinjiang Xinlianxin Co., Ltd., Xinjiang, China), monoammonium phosphate (N ≥ 12%, P₂O₅ ≥ 61%, powder; Guizhou Kai Phosphorus Group Co., Ltd., Guiyang, China), and potassium sulfate (Xinjiang Lop Nur Potassium Salt Co., Ltd., Xinjiang, China). The irrigation and fertilization quotas for each growth stage are detailed in

Table 2. Irrigation and fertilizer application during maize growth stages.

Treatment/Period		Seedling Stage	Jointing Stage	Bell- Mouth Stage	Heading Stage	Flowering Stage	Silking Stage	Grain Formation Stage	Milk- Ripe Stage	Maturity Stage	Total
HC	Irrigation quantity (kg3·hm−2)	164.0	600.0	600.0	600.0	600.0	600.0	600.0	564.0	472.0	4800.0
	Urea (kg·hm−2)	0.0	81.8	81.8	90.9	81.8	81.8	72.7	54.5	0.0	545.3
	Monoammonium phosphate (kg·hm−2)	36.4	36.4	45.5	45.5	45.5	27.3	18.2	18.2	0.0	273.0
	Potassium sulfate (kg·hm−2)	0.0	18.2	27.3	27.3	36.4	22.7	18.2	13.6	0.0	163.7
HL	Irrigation quantity (kg3·hm−2)	164.0	300.0	300.0	300.0	300.0	300.0	300.0	282.0	154.0	2400.0
	Urea (kg·hm−2)	0.0	81.8	81.8	90.9	81.8	81.8	72.7	54.5	0.0	545.3
	Monoammonium phosphate (kg·hm−2)	36.4	36.4	45.5	45.5	45.5	27.3	18.2	18.2	0.0	273.0
	Potassium sulfate (kg·hm−2)	0.0	18.2	27.3	27.3	36.4	22.7	18.2	13.6	0.0	163.7
LC	Irrigation quantity (kg3·hm−2)	164.0	600.00		600.0		600.0		436.0		2400.0
	Urea (kg·hm−2)	0.0	163.6		172.7		154.5		54.5		545.3
	Monoammonium phosphate (kg·hm−2)	36.4	81.9		91		45.5		18.2		273.0
	Potassium sulfate (kg·hm−2)	0.0	45.5		63.7		40.9		13.6		163.7

. The schedules for drip irrigation, sowing, harvesting, and sampling are summarized in

Table 3. Maize sowing, sampling, and harvest timelines across four growth stages (2018–2021).

Years	Sowing Date	Flowering Stage	Maturity Stage	Harvest Date	Total Days
2018	28 Apr	18 Jul	25 Aug	27 Sep	152
2019	30 Apr	14 Jul	22 Aug	22 Sep	145
2020	26 Apr	15 Jul	2 Sep	1 Oct	158
2021	7 May	19 Jul	27 Aug	24 Sep	140

Total Days: Duration from sowing to harvest, calculated based on the Gregorian calendar.

.

2.4. Sampling and Measurements

2.4.1. Stand Growth Index

To assess the impact of irrigation treatments on maize growth, several growth parameters were measured at the different stages of the growing season. These included the following:

Plant height: Measured from the base to the apical meristem using a tape meter [26].

Leaf area: Determined by destructive sampling. The length and width of fully expanded leaves were recorded, and the leaf area was calculated as (length × width) × 0.75 [27].

Chlorophyll content (SPAD measurement): The measurement of SPAD values, which served as an indicator of leaf chlorophyll content, was conducted using a SPAD-502 m from Konica Minolta Tokyo, Japan at the flowering and maturity stages. For each plot, 10 representative fully expanded leaves were selected.

Dry matter biomass: Aboveground biomass (leaves, stems, and reproductive organs) was harvested at the flowering and maturity stages. The samples were dried in an oven at 105 °C for 30 min and then at 75 °C until they reached a constant weight.

Dry matter translocation (kg·hm⁻²) was calculated as the difference in stem and leaf dry matter between the flowering and maturity stages:

$$\begin{array}{l}\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}(\mathrm{k}\mathrm{g}·\mathrm{h}{\mathrm{m}}^{-2})=\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f} \mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{a}\mathrm{t} \mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}-\\ \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f} \mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{a}\mathrm{t} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{i}\mathrm{t}\mathrm{y} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}\end{array}$$

(1)

Dry matter transfer efficiency (%) was determined by dividing the translocation value by the stem and leaf dry matter at the flowering stage, then multiplying by 100:

$$\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{f}\mathrm{e}\mathrm{r} \mathrm{e}\mathrm{f}\mathrm{f}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}(\%)={\displaystyle \frac{\mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}{\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m} \mathrm{a}\mathrm{n}\mathrm{d} \mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f} \mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{a}\mathrm{t} \mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}}}\times 100$$

(2)

2.4.2. Maize Yields and Yield Components

At maize maturity, 20 randomly selected plants from each plot were harvested. Ears were manually threshed, and the grains were dried to a 14% moisture content. Total kernel weight and 1000-kernel weight were recorded. Grain yield per hectare was converted based on the plot area.

The yield (kg·hm⁻²) was calculated using the formula [28]:

$$\mathrm{Y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}(\mathrm{k}\mathrm{g}·\mathrm{h}{\mathrm{m}}^{-2})={\displaystyle \frac{20\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}(\mathrm{g})}{20\mathrm{p}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\mathrm{s}}}\times {\displaystyle \frac{126000}{1000}}\times {\displaystyle \frac{1-\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n} \mathrm{m}\mathrm{o}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}(\%)}{1-14\%}}$$

(3)

The harvest index (%) was derived as follows [29]:

$$\mathrm{H}\mathrm{a}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{t} \mathrm{i}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{x}(\%)={\displaystyle \frac{\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}}{\mathrm{a}\mathrm{b}\mathrm{o}\mathrm{v}\mathrm{e}\mathrm{g}\mathrm{r}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{d} \mathrm{b}\mathrm{i}\mathrm{o}\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}}}\times 100$$

(4)

For the contribution of dry matter translocation to the grain (%), it was calculated by [30]:

$$\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{b}\mathrm{u}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{o}\mathrm{f} \mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{t}\mathrm{o} \mathrm{g}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}(\%)={\displaystyle \frac{\mathrm{d}\mathrm{r}\mathrm{y} \mathrm{m}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{n}\mathrm{s}\mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}}{\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n} \mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}}}\times 100$$

(5)

2.4.3. Water Use Efficiency (WUE) Calculation

The $\mathrm{E}{\mathrm{T}}_{\mathrm{C}}$ was calculated by means of the soil water balance [31]:

$$\mathrm{E}\mathrm{T}\mathrm{c}=\mathrm{I}+\mathrm{P}+{\mathrm{C}}_{\mathrm{r}}-{\mathrm{R}}_{\mathrm{f}}-{\mathrm{D}}_{\mathrm{p}}\pm \mathrm{\Delta }\mathrm{S}$$

(6)

where $\mathrm{I}$ denotes irrigation (mm); $\mathrm{P}$ represents precipitation (mm); ${\mathrm{C}}_{\mathrm{r}}$ indicates capillary rise (mm); ${\mathrm{R}}_{\mathrm{f}}$ refers to runoff (mm); ${\mathrm{D}}_{\mathrm{p}}$ signifies deep percolation (mm); and $\mathrm{\Delta }\mathrm{S}$ represents the variation in soil water storage (mm). Based on the field conditions, we made the following assumptions:

${\mathrm{C}}_{\mathrm{r}}=0$: The groundwater table was located 2 m beneath the soil surface.

${\mathrm{R}}_{\mathrm{f}}\mathrm{\approx }0$: Flat terrain minimized runoff.

${\mathrm{D}}_{\mathrm{p}}\mathrm{\approx }0$: The soil water content below 90 cm in depth never reached field capacity.

$\mathrm{\Delta }\mathrm{S}={\mathrm{S}}_{\mathrm{f}}-{\mathrm{S}}_{\mathrm{i}}$, representing alterations in the soil water storage within the profile, where ${\mathrm{S}}_{\mathrm{i}}$ denotes the soil water storage at the time of sowing and ${\mathrm{S}}_{\mathrm{f}}$ signifies the soil water storage at harvest.$\mathrm{\Delta }\mathrm{S}$ suggest that the soil water storage during sowing is comparable to that at harvest, and thus can be disregarded.

Water use efficiency (WUE) and irrigation water use efficiency (IWUE) were computed as follows [32,33]:

$$\mathrm{W}\mathrm{U}\mathrm{E}={\displaystyle \frac{\mathrm{Y}}{\mathrm{E}{\mathrm{T}}_{\mathrm{c}}}}$$

(7)

$$\mathrm{I}\mathrm{W}\mathrm{U}\mathrm{E}={\displaystyle \frac{\mathrm{Y}}{\mathrm{I}}}$$

(8)

where the grain yield ($\mathrm{Y}$) is measured, $\mathrm{E}{\mathrm{T}}_{\mathrm{c}}$ evapotranspiration represents the total for the entire growing season (mm), and $\mathrm{I}$ the irrigation amount refers to the maize growth period (mm).

2.5. Statistical Analysis

Data from all experiments were assessed using Microsoft Excel 2010 and SPSS Statistics 27.0. The assumption of normality and the homogeneity of variances were tested using the Shapiro–Wilk and Levene’s tests, respectively. Treatment comparisons were conducted using one-way ANOVA, followed by LSD post hoc tests with a significance level set at p < 0.05. The nonparametric Kruskal–Wallis H (multiple groups) or Mann–Whitney U (two groups) tests were used for non-normal data, while Welch’s test addressed unequal variances. Graphical representations of the results were generated using Origin 2021 software.

3. Results

3.1. Effects of Irrigation Frequency and Amount on Maize Growth Parameters

The impact of irrigation frequency and amount on maize growth parameters demonstrated numerical variations. Although the differences in plant height, leaf area, and SPAD values (chlorophyll content) between the two low-water (total irrigation water volume) treatments were not statistically significant, high-frequency limited irrigation (HL) showed numerically higher growth metrics with no statistical significance compared to low-frequency conventional irrigation (LC). Specifically, HL was associated with a 2.6% numerically greater plant height (244.09 ± 14.34 vs. 237.84 ± 19.08 cm), a 9.9% numerically larger leaf area (371.32 ± 32.42 vs. 337.73 ± 22.11 cm²), and a 1.5% numerically higher SPAD value (49.40 ± 0.56 vs. 48.65 ± 0.40). However, these differences were not statistically significant (Figure 3a–c). At flowering and maturity, the plant height values under HL and LC shared no statistical difference despite numerical variation (Figure 3a), indicating comparable growth despite numerical differences.

Dry matter accumulation in maize during the flowering and maturity stages varied across irrigation treatments and between irrigation treatments. Aboveground biomass accumulation was higher under high-frequency limited irrigation (HL) compared to low-frequency conventional irrigation (LC). At maturity, the total biomass under HL reached 22,310.58 ± 884.10 kg·hm⁻², representing a 6.83% increase compared to LC (20,884.84 ± 645.47 kg·hm⁻²). Dry matter accumulation in all plant organs: leaf biomass under HL was significantly higher (+14.7%), while stem (+2.7%) and reproductive organ (+9.8%) accumulations showed non-significant gains (Figure 4a,b). HL treatment increased the translocation efficiency of dry matter from the vegetative organs to grains by 12.2% compared to LC, resulting in an 18.6% higher grain yield (Table 5).

Note: Different lowercase letters indicate significant differences (p < 0.05) between treatments for each indicator. As the dry matter transfer efficiency data did not meet the normality assumption, a non-parametric test was employed; the LSD method was used for the rest of the data.

3.2. Yield and Its Components

Deficit irrigation and irrigation frequency influenced the maize yield and its components (

Table 4. Yield and components of maize under different irrigation treatments.

Year	Treatment	Ear Diameter (mm)	Kernel Number Per Row	Row Number Per Ear	1000-Kernel Weight (g)	Yield(kg·hm−2)
2018	HC	46.85 ± 0.91 a	32.45 ± 1.82 a	14.30 ± 0.48 a	330.43 ± 14.66 a	15,041.43 ± 1022.43 a
	HL	44.70 ± 0.62 b	31.85 ± 0.82 a	13.45 ± 1.03 ab	310.58 ± 8.08 b	11,929.02 ± 320.29 b
	LC	42.70 ± 1.15 b	30.65 ± 2.07 b	13.15 ± 0.25 a	302.93 ± 9.85 b	11,047.74 ± 398.14 b
2019	HC	43.55 ± 0.74 a	33.05 ± 1.06 a	16.11 ± 0.36 a	329.25 ± 9.46 a	13,164.46 ± 1506.54 a
	HL	42.95 ± 0.53 a	32.35 ± 1.68 a	14.75 ± 0.19 b	306.75 ± 19.6 ab	9465.83 ± 752.97 b
	LC	42.70 ± 0.53 a	31.95 ± 0.77 a	13.95 ± 0.20 c	300.00 ± 15.10 b	8677.52 ± 252.19 b
2020	HC	43.50 ± 0.87 a	31.40 ± 2.12 a	14.80 ± 1.26 a	333.75 ± 5.56 a	15,775.93 ± 2027.60 a
	HL	42.60 ± 0.53 a	30.95 ± 1.24 a	14.20 ± 0.23 a	329.66 ± 14.57 ab	10,972.99 ± 279.96 b
	LC	42.35 ± 1.02 a	29.95 ± 1.64 a	13.70 ± 0.50 a	309.60 ± 17.56 b	8068.67 ± 514.82 c
2021	HC	43.53 ± 0.76 a	31.40 ± 0.85 a	14.80 ± 0.35 a	340.00 ± 3.56 a	16,569.64 ± 301.29 a
	HL	42.18 ± 1.21 a	28.58 ± 1.39 b	13.48 ± 0.36 b	317.00 ± 21.59 ab	10,805.78 ± 333.15 b
	LC	40.73 ± 0.77 b	27.00 ± 1.67 b	13.20 ± 0.41 b	302.50 ± 3.42 b	8722.51 ± 547.40 c
Mean	HC	44.35 ± 1.66 a	32.08 ± 0.82 a	15.00 ± 0.77 a	333.36 ± 4.82 a	15,137.86 ± 1456.10 a
	HL	43.19 ± 1.11 a	30.93 ± 1.67 a	13.97 ± 0.63 b	316.00 ± 10.05 b	10,793.78 ± 1013.98 b
	LC	42.12 ± 0.94 a	29.89 ± 2.10 a	13.50 ± 0.39 b	303.76 ± 4.10 c	9129.11 ± 1313.38 b

Note: Different lowercase letters denote significant differences (p < 0.05). Non-parametric tests were applied to the 2018 and 2021 average yields and ear diameter due to non-normality/unequal variances; Welch’s test was applied for the 2020 rows per ear, yield, and 1000-kernel weight; and LSD was applied for the remaining data.

). Yield data showed that high-frequency limited irrigation (HL) was numerically higher than low-frequency conventional irrigation (LC) by 18.2% (10,793.78 ± 1013.98 kg·hm⁻² vs. 9129.11 ± 1313.38 kg·hm⁻²). In terms of yield components, HL treatment exhibited better numerical values across several parameters. The numerical values of ear length, number of kernels per row, and thousand-kernel weight were all higher under HL than under LC, contributing to the increased yield. But these differences were not statistically significant, as indicated by the same letter annotations in Table 4.

3.3. Harvest Index

The effects of irrigation volume and frequency on maize biomass translocation and related indices showed numerical variations (p < 0.05, LSD test for significant parameters). As irrigation volume and frequency decreased, dry matter translocation and associated metrics tended to decline. Specifically, the HL treatment exhibited a numerically 23.0% higher dry matter translocation than LC, though not statistically significant (

Table 5. Effect on harvest index of maize under different irrigation treatments.

Treatment	Dry Matter at Flowering Stage (kg·hm−2)	Dry Matter at Maturity (kg·hm−2)	Dry Matter Translocation (kg·hm−2)	Dry Matter Transport Efficiency (%)	Grain Contribution (%)	Harvest Index (%)
HC	21,215.07 ± 511.32 a	26,341.57 ± 3011.91 a	7825.96 ± 948.32 a	65.00 ± 12.42 a	53.52 ± 11.82 a	58.13 ± 10.00 a
HL	16,513.99 ± 1071.80 b	24,164.32 ± 2368.45 ab	5467.53 ± 1247.47 b	48.31 ± 7.52 b	50.99 ± 12.45 a	48.46 ± 5.45 ab
LC	14,720.43 ± 823.66 c	21,869.60 ± 2948.25 b	4444.22 ± 1222.41 b	43.07 ± 12.12 b	50.59 ± 18.11 a	43.78 ± 6.78 b

Note: Different lowercase letters indicate significant differences (p < 0.05) between treatments for each indicator. As the dry matter transfer efficiency data did not meet the normality assumption, a non-parametric test was employed; the LSD method was used for the rest of the data.

). In terms of dry matter translocation efficiency, HL showed a 12.2% numerical increase compared to LC, though not statistically significant (Table 5). Furthermore, the harvest index in the HL treatment declined less than in LC: HL had a 16.6% reduction, versus 24.7% for LC, resulting in an 8.1% smaller decline in HL, though this was not statistically significant (Table 5).

3.4. Water Use Efficiency (WUE)

Irrigation water use efficiency (IWUE) was significantly higher in high-frequency limited irrigation (HL) compared to low-frequency conventional irrigation (LC) (

Table 6. Effect on water use efficiency of maize under different irrigation treatments.

Year	Treatment	Irrigation Amount in Maize Growth Period (m3·hm−2)	Yield (kg·hm−2)	WUE (kg·m−3)	IWUE (kg·m−3)	PUE (kg·m−3)
2018	HC	4800	15,041.43 ± 1022.43 a	2.74 ± 0.17 c	3.13 ± 0.21 c	22.12 ± 1.50 a
	HL	2400	11,929.02 ± 320.29 b	3.87 ± 0.10 a	4.97 ± 0.13 a	17.54 ± 0.47 b
	LC	2400	11,047.74 ± 398.14 b	3.59 ± 0.13 b	4.60 ± 0.17 b	16.25 ± 0.59 b
2019	HC	4800	13,164.46 ± 1506.54 a	2.29 ± 0.27 b	2.73 ± 0.33 b	14.08 ± 1.68 a
	HL	2400	9465.83 ± 752.97 b	2.86 ± 0.24 a	3.97 ± 0.33 a	10.24 ± 0.86 b
	LC	2400	8677.52 ± 252.19 b	2.61 ± 0.08 ab	3.62 ± 0.11 a	9.33 ± 0.27 b
2020	HC	4800	15,775.93 ± 2027.60 a	3.04 ± 0.39 b	3.29 ± 0.42 b	39.94 ± 5.13 a
	HL	2400	10,972.99 ± 279.6 b	3.93 ± 0.10 a	4.57 ± 0.12 a	27.78 ± 0.71 b
	LC	2400	8068.67 ± 514.82 c	2.89 ± 0.18 b	3.36 ± 0.21 b	20.43 ± 1.30 c
2021	HC	4800	16,569.64 ± 301.29 a	3.20 ± 0.06 b	3.45 ± 0.06 c	43.60 ± 0.79 a
	HL	2400	10,805.78 ± 333.15 b	3.89 ± 0.12 a	4.50 ± 0.14 a	28.44 ± 0.88 b
	LC	2400	8722.51 ± 547.40 c	3.14 ± 0.20 b	3.63 ± 0.23 b	22.95 ± 1.44 c
Mean	HC	4800	15,137.89 ± 1456.10 a	2.82 ± 0.40 a	3.15 ± 0.31 c	29.93 ± 14.13 a
	HL	2400	10,793.78 ± 1013.98 b	3.63 ± 0.52 a	4.50 ± 0.42 a	20.99 ± 8.74 a
	LC	2400	9129.11 ± 1313.38 b	3.06 ± 0.42 a	3.80 ± 0.55 b	17.24 ± 5.95 a
P	Year		***	***	***	***
	Treatment		***	***	***	***
	Year × Treatment		***	***	***	***

Note: Different lowercase letters denote significant differences (p < 0.05) between treatments for each indicator. Due to the homogeneity of variances, Welch’s test was applied to WUE, IWUE, and PUE in 2020, as well as to the average PUE; the average WUE, failing the normality test, was analyzed using a non-parametric method. The LSD method was used for the remaining data. Here, *** denotes p < 0.001.

). The WUE for HL reached 3.63 kg·m⁻³, representing an 18.6% numerical increase compared to LC (3.06 kg·m⁻³, p > 0.05, Kruskal–Wallis H test). The irrigation water use efficiency (IWUE) was calculated as the yield per unit of irrigation water applied. The HL treatment exhibited an 18.4% enhancement in IWUE relative to LC (p < 0.05, LSD test). Regarding precipitation use efficiency (PUE), the HL treatment showed a 21.8% numerical improvement over LC in PUE values (20.99 ± 8.74 kg·m⁻³ vs. 17.24 ± 5.95 kg·m⁻³, p > 0.05, Welch test). Year and irrigation treatment significantly influenced yield, water use efficiency (WUE), irrigation water use efficiency (IWUE), and precipitation use efficiency (PUE), with a significant interaction observed between the two factors.

3.5. Key Growth Parameters Driving Maize Yield

The experimental data demonstrated that maize yield correlated strongly with total biomass, ear diameter, and thousand-kernel weight (Figure 5), with statistical significance denoted by asterisks (no significance p > 0.05, * 0.01 ≤ p ≤ 0.05, ** p ≤ 0.01, and *** p ≤ 0.001). Specifically, yield showed a significant positive correlation with ear diameter (R² = 0.66, p = 0.019) and total biomass dry matter (R² = 0.73, p = 0.008), and demonstrated a highly significant positive correlation with thousand-kernel weight (R² = 0.86, p < 0.001). Thousand-kernel weight exhibited the strongest predictive power (86% explained variance), followed by total biomass (73%) and ear diameter (66%). These quantitative correlations highlighted that thousand-kernel weight was the dominant driver of yield variation, while ear diameter and total biomass provided complementary explanatory value.

4. Discussion

This study revealed the resource optimization effect of high-frequency limited irrigation (HL, irrigation water volume 2400 m³·hm⁻²) in maize production in arid areas. Compared to low-frequency conventional irrigation (LC, irrigation water volume 2400 m³·hm⁻²), the HL treatment achieved multiple benefits while reducing the irrigation water volume by 50%: (1) the water use efficiency (WUE) was increased by 18.6% compared to LC (p > 0.05), showing an efficiency optimization trend, which was consistent with the threshold irrigation theory proposed by Du et al. [34]; (2) the harvest index increased by 10.7%, indicating that limited water was preferentially allocated to the grains. This characteristic of “reducing irrigation water volume by half while improving efficiency” provides a new path for balancing water-saving goals and yield stability in arid areas.

4.1. Physiological Mechanisms of Water Regulation Efficiency

HL’s effectiveness originated from precision water management during the reproductive phases (35–75 days post-sowing). By applying frequent, small-volume irrigation events (60 mm per event), HL maintained soil saturation within the 60–80% range in the 0–40 cm layer. This approach effectively mitigated the drought-induced suppression of carbon assimilation. Specifically, a 9.9% expansion in leaf area value (Figure 3b) and a 1.5% increase in SPAD values (Figure 3c) were observed. While statistically non-significant, the coordinated improvements in leaf area (+9.9%) and SPAD (+1.5%) aligned with the reported delayed leaf senescence under a stable water supply [35]. Given that 90–95% of maize grain dry matter originated from reproductive-stage photosynthates [36], it was hypothesized that HL enhanced the assimilate translocation to grains by stabilizing the functionality of photosynthetic organs. This hypothesis was supported by phenotypic evidence: a 3.8% increase in ear diameter under HL (Table 4) suggested enhanced sink capacity through an expanded grain spatial arrangement, mirroring the source–sink coordination mechanism identified in drought-resistant hybrids [15].

4.2. Climate-Adaptive Optimization

The successful implementation of HL was facilitated through climate-adaptive optimization. Specifically, 65% of seasonal precipitation was concentrated during the low-water-demand seedling stage (Figure 1). This allowed HL to reduce early-stage irrigation by 15%, prioritizing water conservation during the moisture-sensitive reproductive phase. This contrasted with the “stage-specific water allocation” strategy, which increased WUE by 22% through concentrated jointing-stage irrigation in the North China Plain [19]. Under Xinjiang’s extreme heat (28 °C monthly average during reproduction), HL’s high-frequency irrigation counteracted the rapid root-zone water depletion caused by high temperatures, thereby extending the applicability of Tan et al.’s [37] irrigation quota–efficiency quadratic relationship theory.

4.3. Unresolved Issues and Implementation Challenges

Although HL achieved a 12% increase in irrigation water use efficiency (IWUE) (Table 6), indicating improved shallow water utilization, the underlying root system responses still need to be verified. While Wu et al. [38] documented water fluctuation-induced increases in shallow root length density, this contrasted with Zhang et al.’s [31] recommendation of a 60 cm optimal root depth in gray desert soils. In situ, root imaging should be employed to elucidate the interactions between vertical water distribution and root architecture. Practically, HL erects a sustainable irrigation framework for groundwater-scarce regions like Xinjiang but faces implementation hurdles. Smallholder farmers typically lack access to high-frequency irrigation technologies (e.g., drip systems) and the associated energy supplies. Policy interventions, including infrastructure subsidies and cultivar-specific adaptation trials, are pivotal to catalyzing HL adoption. Future research should prioritize the following: (1) root phenomics to illuminate water uptake mechanisms; (2) multi-climate zone trials to authenticate strategy robustness; and (3) the development of dynamic irrigation models integrating frequency, volume, and climatic variables.

5. Conclusions

This study shows that high-frequency limited irrigation (HL) numerically improves water use efficiency (WUE) and has a yield advantage over low-frequency conventional irrigation (LC) under the same irrigation volume. Though physiological parameters improved marginally in HL, there were no statistically significant differences between the strategies. HL can help to save water and maintain maize yield in water-limited conditions, offering a potential adaptive approach for arid regions. However, future research with models integrating real-time feedback is needed before recommending HL.