Optimized Water Management Strategies: Evaluating Limited-Irrigation Effects on Spring Wheat Productivity and Grain Nutritional Composition in Arid Agroecosystems

Abstract

The Hetao Plain Irrigation District of Inner Mongolia faces critical agricultural sustainability challenges due to its arid climate, exacerbated by tightening Yellow River water allocations and pervasive water inefficiencies in the current wheat cultivation practices. This study addresses water scarcity by evaluating the impact of regulated deficit irrigation strategies on spring wheat production, with the dual objectives of enhancing water conservation and optimizing yield–quality synergies. Through a two-year field experiment (2020~2021), four irrigation regimes were implemented: rain-fed control (W0), single irrigation at the tillering–jointing stage (W1), dual irrigation at the tillering–jointing and heading–flowering stages (W2), and triple irrigation incorporating the grain-filling stage (W3). A comprehensive analysis revealed that an incremental irrigation frequency progressively enhanced plant morphological traits (height, upper three-leaf area), population dynamics (leaf area index, dry matter accumulation), and physiological performance (flag leaf SPAD, net photosynthetic rate), all peaking under the W2 and W3 treatments. While yield components and total water consumption exhibited linear increases with irrigation inputs, grain yield demonstrated a parabolic response, reaching maxima under W2 (29.3% increase over W0) and W3 (29.1%), whereas water use efficiency (WUE) displayed a distinct inverse trend, with W2 achieving the optimal balance (4.6% reduction vs. W0). The grain quality parameters exhibited divergent responses: the starch content increased proportionally with irrigation, while protein-associated indices (wet gluten, sedimentation value) and dough rheological properties (stability time, extensibility) peaked under W2. Notably, protein content and its subcomponents followed a unimodal pattern, with the W0, W1, and W2 treatments surpassing W3 by 3.4, 11.6, and 11.3%, respectively. Strong correlations emerged between protein composition and processing quality, while regression modeling identified an optimal water consumption threshold (3250~3500 m³ ha⁻¹) that concurrently maximized grain yield, protein output, and WUE. The W2 regime achieved the synchronization of water conservation, yield preservation, and quality enhancement through strategic irrigation timing during critical growth phases. These findings establish a scientifically validated framework for sustainable, intensive wheat production in arid irrigation districts, resolving the tripartite challenge of water scarcity mitigation, food security assurance, and processing quality optimization through precision water management.

1. Introduction

The Hetao Plain Irrigation District, a vital production area for medium- and strong-gluten high-quality wheat in northern China, relies fundamentally on Yellow River irrigation to sustain its spring wheat cultivation system, yet faces critical sustainability challenges stemming from suboptimal water management practices [1]. The current field operations paradoxically combine excessive irrigation volumes with inefficient irrigation scheduling, resulting in both substantial water wastage and constrained yield–quality improvements. These anthropogenic inefficiencies, exacerbated by persistent aridity and stringent water allocation policies for the Yellow River, have created a self-reinforcing cycle of agricultural water scarcity that threatens the long-term viability of wheat production systems. This pressing context necessitates the urgent optimization of irrigation protocols to reconcile competing demands for water conservation, yield preservation, and quality enhancement.

Extensive research demonstrates that irrigation management critically determines wheat productivity and grain quality through its regulation of photosynthetic partitioning and nitrogen metabolism [2,3]. Regarding yield, the soil moisture status governs physiological processes governing photoassimilate accumulation and translocation, with optimized water supply enhancing drought resilience while promoting biomass production and canopy architecture development for yield stabilization [4,5,6,7,8,9,10], through yield response plateaus beyond critical irrigation thresholds [11]. Concurrently, water availability modulates nitrogen assimilation dynamics, influencing grain protein composition and starch/protein ratios that ultimately define flour functionality [12,13,14,15]. Notably, controlled post-anthesis water stress enhances the remobilization efficiency of vegetative reserves and stimulates albumin/globulin synthesis [16,17,18], whereas supra-optimal irrigation degrades key quality parameters through nutrient dilution effects. These findings collectively underscore the necessity for precision irrigation strategies that navigate the narrow optimum between water conservation, yield maximization, and quality preservation.

Water-efficient agricultural research prioritizes the strategic exploitation of crop compensatory mechanisms under controlled water stress, aiming to optimize photosynthetic architectures and root system efficiency while maintaining yield stability through improved yield component synergies. Extensive studies confirmed that regulated deficit irrigation strategies can maintain wheat productivity while enhancing water use efficiency (WUE) through optimized irrigation frequency and volume control [19,20,21], though current methodological advancements predominantly focus on winter wheat systems [22,23,24]. In the Hetao Irrigation District, a regionally adapted spring wheat cultivation system has been developed through longitudinal research, combining autumn soil moisture preservation with two growth-stage-specific irrigations (750~1050 m³ ha⁻¹) during the tillering–jointing and heading–flowering phases. This protocol achieves concurrent yield (6.75 t ha⁻¹) and WUE (1.75 kg m⁻³) optimization [25,26,27,28], yet critical knowledge gaps persist regarding its impacts on grain processing quality parameters and the underlying physiological mechanisms governing water-stress-induced quality modulation—a crucial oversight, given the district’s specialization in high-value gluten wheat production.

Therefore, this study employs a two-year field experiment to systematically evaluates the impacts of limited irrigation treatment on spring wheat yield formation, WUE, and grain quality parameters. Through a mechanistic analysis of irrigation–plant interactions, we aim to quantify the relationships between irrigation amount and wheat performance metrics (grain yield, protein yield, and WUE), while establishing an optimized irrigation protocol that synergistically achieves water conservation, premium grain quality, and yield preservation in the Hetao Irrigation District, thereby providing actionable insights for developing precision water management protocols for semi-arid wheat systems.

2. Materials and Methods

2.1. Experiment Site

Field experiments were implemented during consecutive growing seasons (2020~2021) at two representative sites in Wuyuan County, Bayannur City, Inner Mongolia, i.e., Xingongzhong (41°08′ N, 108°01′ E; elevation 1095 m) and Longxingchang (41°07′ N, 108°31′ E; elevation 1110 m), characterized by a temperate continental climate with mean annual temperatures of 6.1 °C, frost-free periods of 117~136 days, and pronounced aridity indices (precipitation of 170.0 mm and evaporation of 2067.8 mm). Precipitation amount and temperature are shown in Figure 1. The experimental soils, classified as alluvial loams with homogeneous texture profiles, exhibited baseline physicochemical properties in the 0~20 cm plow layer as detailed in

Table 1. Initial soil properties along the root zone profile of the 0~20 cm soil layer.

Year	Site	Organic Matter (g kg−1)	Alkali N (mg kg−1)	Alkali P (mg kg−1)	Alkali K (mg kg−1)	pH
2020	Xingongzhong	19.63 ± 1.12	53.74 ± 3.65	25.86 ± 1.48	147.98 ± 6.84	7.67 ± 0.33
2021	Longxingchang	17.65 ± 0.88	57.45 ± 4.42	26.83 ± 1.99	152.42 ± 8.36	7.32 ± 0.21

. Growing-season precipitation during the trial years registered 60.02 and 90.19 mm in 2020 and 2021, respectively.

2.2. Experiment Design

The study employed the spring wheat cultivar Yongliang 4 under four regulated deficit irrigation treatments: W0 (rainfed control), W1 (single irrigation at the tillering–jointing stage), W2 (dual irrigation at the tillering–jointing and heading–flowering stages), and W3 (triple irrigation involving the grain-filling stage), with uniform irrigation applications of 900 m³ ha⁻¹ per event. Irrigation water was pumped from adjacent canals using electric pumps and delivered to individual plots through polyethylene pipelines, with inline flow meters installed to precisely regulate the irrigation amount and ensure volumetric precision across treatments. A randomized complete block design with three replicates yielded 12 experimental plots (10 × 10 m), separated by 1.0 m buffer zones and 0.5 m reinforced ridges to prevent hydrological interference. Mechanized drilling achieved the sowing density of 375 kg ha⁻¹, with 15 cm row spacing, coupled with basal fertilization (150 kg ha⁻¹ urea with 46% N and 300 kg ha⁻¹ diammonium phosphate with 18% N, 46% P₂O₅). Conventional agronomic practices were maintained throughout the trial periods, with sowing dates on 12 March 2020 and 15 March 2021, reaching physiological maturity on 17 July 2020 and 15 July 2021, respectively.

2.3. Sampling and Measurements

2.3.1. Flowering Stage Agronomic Trait

Agronomic trait quantification during the flowering stage involved triplicate 1 m linear sampling rows per plot, from which 20 representative plants were systematically selected for the measurement of main stem height (cm), basal diameter (cm), and total green leaf area (cm²), as well as the calculation of the top-three leaf area (cm²) and leaf area index (LAI), followed by drying of all plants for biomass determination after oven desiccation at 85 °C to constant biomass (kg ha⁻¹). The chlorophyll content (SPAD) was measured in situ on flag leaves using a SPAD-502 chlorophyll meter (Konica Minolta, Tokyo, Japan), while flag leaf photosynthetic capacity (Pn, μmol m⁻² s⁻¹) was quantified via a CIRAS-3 portable photosynthesis system (PP Systems, Amesbury, MA, USA). Canopy light interception dynamics were characterized using a SunScan-SS1 canopy analysis system (Delta-T Devices, Cambridge, UK), with photosynthetically active radiation (PAR) measurements recorded at the upper, middle, and lower canopy strata to compute population transmittance coefficients (%).

2.3.2. Grain Yield and Its Components

At physiological maturity, 1 m² areas were sampled in triplicate per plot for yield determination (kg ha⁻¹), mechanical threshing, and gravimetric assessment to determine the grain yield normalized to a 13% moisture content. Concurrently, representative 50 cm row segments were analyzed for spike density (10⁴ ha⁻¹) quantification, followed by manual dissection of 30 randomly selected spikes to determine the number of grains per spike, complemented by 1000-grain weight (g) measurements derived from three 500-seed subsamples.

2.3.3. Grain Processing Quality and Protein Component Content

Grain processing quality assessment involved the systematic processing of 200 g air-dried samples per treatment through an Infratec™ 1241 Near-Infrared Grain Analyzer (FOSS, Hillerød, Denmark) for comprehensive quality profiling. Quality parameters including starch content (%), wet gluten percentage (%), sedimentation value (mL), test weight (g L⁻¹), flour yield (%), water absorption (%), dough stability time (min), dough development time (min), extensibility area (cm²), dough malleability (mm), maximum resistance to extension (E.U.), protein content of the grains (%), were determined. Grain protein yield (kg ha⁻¹) was calculated by the product of compositional and productivity metrics (protein content × grain yield).

Protein component fractionation (%) was conducted following He’s sequential extraction protocol [29], commencing with 0.5 g of sample and performing the differential solubility-based partitioning of albumin, globulin, glutenin, and gluten. Albumin extraction involved mechanical homogenization in deionized water followed by three successive extractions (30 min shaking, 15 min centrifugation), with the pooled supernatants retained for nitrogen quantification. Globulin, glutenin, and gluten were extracted using 10% NaCl, 70% ethanol, and 0.2% NaOH, respectively, following identical extraction cycles as for albumin. The nitrogen content in the fractionated extracts was determined via automated Kjeldahl analysis (K1100 analyzer, Haineng Instrument Co., Weifang, China).

2.3.4. Total Water Consumption and Water Use Efficiency

Soil moisture monitoring involved stratified sampling (0~100 cm depth, 20 cm increments) using stainless-steel augers across all treatment plots during the pre-sowing and post-harvest phases. Core samples from each soil horizon were homogenized, placed in pre-weighed aluminum crucibles, and oven-dried at 105 °C to constant mass for gravimetric water content determination. The following formulas were used to calculate the soil water storage (SWS), total water consumption (TWC), and water use efficiency (WUE):

Soil water storage (m³ ha⁻¹) = W × D × H

(1)

where W indicates the soil water content (%); D indicates the soil bulk density (g cm⁻³); and H indicates the soil layer thickness (cm);

Total water consumption (m³ ha⁻¹) = P + I + ΔSWS

(2)

where P and I indicate the precipitation and irrigation amounts (m³ ha⁻¹); and ΔSWS indicates the difference in soil water storage between before sowing and after harvest (m³ ha⁻¹); and

Water use efficiency (kg m⁻³) = Y/TWC

(3)

where Y indicates the yield (kg ha⁻¹); and TWC indicates the total water consumption (m³ ha⁻¹).

2.4. Data Analysis

All data were analyzed by ANOVA to detect differences (least significant difference, LSD) between the treatment means at the 5% probability level. Statistical analysis and Spearman’s rank correlation tests were performed using SAS 9.0 software (SAS, San Antonio, TX, USA). Origin 2021 software (Origin LAB, Northampton, USA) was used for graphing.

3. Results

3.1. Wheat Agronomic Traits at the Flowering Stage

The irrigation treatments exerted significant differential effects (p < 0.05) on wheat flowering-stage agronomic performance, with trait optimization following threshold-dependent nonlinear responses peaking under the W2/W3 regimes (

Table 2. Agronomic traits of wheat at flowering stage under different irrigation treatments.

Years	Treatment	PH	SD	TLA	SPAD	Pn	LAI	Biomass	LT
2020	W0	65.7 b	0.30 a	61.3 b	43.76 c	19.4 b	2.91 c	8043.9 b	21.61 a
	W1	74.0 a	0.31 a	65.6 b	48.28 b	24.8 a	4.55 ab	9240.9 a	20.11 ab
	W2	79.3 a	0.33 a	74.7 a	51.96 ab	26.7 a	4.69 a	9697.8 a	18.09 c
	W3	78.7 a	0.34 a	75.4 a	53.71 a	26.3 a	4.36 b	9612.2 a	18.33 bc
2021	W0	67.4 b	0.30 a	60.4 b	40.26 c	20.6 c	2.84 c	8436.7 b	20.43 a
	W1	76.2 a	0.32 a	67.9 a	49.21 b	23.5 b	4.42 b	9171.5 a	19.27 a
	W2	80.3 a	0.34 a	74.2 a	52.86 a	25.9 a	4.71 a	9536.7 a	16.12 b
	W3	81.5 a	0.34 a	73.7 a	48.25 b	26.2 a	4.72 a	9498.7 a	15.87 b
r1		0.979 **	0.970 **	0.968 **	0.868 **	0.961 **	0.887 **	0.957 **	−0.916 **
r2		0.899 **	0.812 *	0.863 **	0.893 **	0.910 **	0.918 **	0.932 **	−0.763 *

Values followed by different lowercase letters correspond to significant differences at p < 0.05. r1, r2 are the correlation coefficients between grain yield, protein yield, and each agronomic trait. * and ** mean that the correlation coefficient reached the significant level at p < 0.05 and p < 0.01. PH: plant height (cm); SD: stem diameter (cm); TLA: top-three leaves area (cm²); Pn: photosynthetic rate (μmol m⁻² s⁻¹); LAI: leaf area index; Biomass: wheat dry matter accumulation (kg ha⁻¹); LT: light transmission rate (%).

). Progressive irrigation intensification induced either monotonic improvements or unimodal patterns in phenotypic and physiological parameters, notably enhancing plant height, photosynthetic capacity (flag leaf SPAD; photosynthetic rate), and population structure (LAI; biomass) across both trial years. Contrastingly, canopy light transmittance exhibited inverse response dynamics under W2/W3 versus W0 (p < 0.05). Inter-treatment stability was observed in stem diameter and top-three leaf area. A multivariate correlation analysis revealed strong inverse correlations between productivity metrics (grain yield/protein yield) and canopy transmittance (r = −0.76 to −0.92, p < 0.05), contrasting with the positive associations among other traits (r = 0.81~0.98, p < 0.05). These findings collectively demonstrate that precision irrigation scheduling during critical phenophases enhanced photosynthetic source strength while optimizing source–sink relationships, creating synergistic conditions for concurrent yield–quality enhancement through improved light interception efficiency and assimilate accumulation.

3.2. Grain Yield, Total Water Consumption, and Water Use Efficiency

The irrigation strategies demonstrated significant (p < 0.05) threshold-dependent responses in yield architecture and water productivity (

Table 3. Wheat yield and its components, water consumption and water use efficiency, under different irrigation treatments.

Year	Treatment	Grain Yield (kg ha−1)	Spike (104 ha−1)	Grains per Spike	1000-Grain Weight (g)	Total Water Consumption (m3 ha−1)	Water Use Efficiency (kg m−3)
2020	W0	5044.2 c	625.5 c	28.9 b	36.9 b	2296.5 c	2.20 a
	W1	5957.0 b	714.5 b	31.4 a	37.7 b	2914.5 b	2.04 a
	W2	6579.9 a	727.5 b	28.2 b	40.1 a	3163.5 b	2.07 a
	W3	6704.1 a	772.5 a	27.2 b	40.8 a	3829.5 a	1.75 b
2021	W0	5352.3 c	604.5 c	30.4 a	37.2 b	2530.4 c	2.12 a
	W1	6009.2 b	669.0 b	31.6 a	39.5 ab	3115.4 b	1.92 a
	W2	6860.7 a	715.5 a	32.1 a	40.9 ab	3349.4 b	2.05 a
	W3	6722.6 a	721.5 a	31.3 a	41.3 a	3926.4 a	1.71 b

Values followed by different lowercase letters correspond to significant differences at p < 0.05.

). In 2020, incremental irrigation linearly enhanced grain yield, spike density, and 1000-grain weight, with improvements of 18.1~32.9%, 14.2~23.5%, and 2.2~10.6% vs. the values recorded under W0, respectively, while grains per spike exhibited unimodal variation and peaked under W1, with an 8.7% increase. Contrastingly, 2021 showed parabolic yield and grains per spike gains, with enhancements of 12.3~28.2% and 3.0~5.6% (peak under W2), and diminishing returns under W3. Spike density and 1000-grain weigh, characterized by linear increments, plateaued at 10.7~19.4% and 6.2~11.0%, respectively. Critical trade-offs between water use efficiency and consumption were identified: while water consumption increased by 23.1~66.7% under the irrigation treatments, and water use efficiency peaked under W2 (2.06 kg m⁻³), exceeding the value under W3 by 19.1% (p < 0.05), through optimized yield/water consumption ratios. These results demonstrated that irrigation intensification beyond the tillering–jointing and heading–flowering stages triggered non-proportional yield gains against escalating water inputs, establishing W2 as the optimal solution for concurrent yield maximization and WUE preservation in water-limited systems.

3.3. Grain Processing Quality

The irrigation strategies induced significant quality differentiation in wheat grains, with the starch content demonstrating linear increases of 5.0~12.2% (p < 0.05) across intensified irrigation treatments, while the protein-associated parameters exhibited unimodal optimization patterns peaking under the W1 or the W2 treatment (

Table 4. Wheat grain processing quality under different irrigation treatments.

Treatment	STA (%)	WGC (%)	SV (mL)	TW (g L−1)	FE (%)	WA (%)	DST (min)	DDT (min)	EA (cm2)	DM (mm)	MR (E.U.)
2020
W0	54.81 c	32.7 ab	42.2 bc	836.5 a	74.1 a	60.9 a	8.8 b	4.5 a	97.2 d	166.0 c	406.8 c
W1	58.04 b	38.8 a	48.1 a	838.4 a	75.2 a	62.5 a	10.5 a	4.7 a	101.6 c	175.0 b	425.3 b
W2	60.26 ab	35.5 a	45.3 ab	845.3 a	74.0 a	64.2 a	11.1 a	5.1 a	117.3 a	186.0 a	468.1 a
W3	62.43 a	30.6 b	39.3 c	840.1 a	73.0 a	62.6 a	10.6 a	4.8 a	108.3 b	174.3 b	455.1 a
2021
W0	57.01 c	32.0 bc	39.4 b	815.3 a	73.2 a	61.3 a	7.1 b	4.4 a	101.5 b	163.5 b	433.2 b
W1	60.02 b	37.5 a	48.3 a	819.2 a	73.1 a	62.7 a	9.8 a	4.5 a	112.3 a	171.2 a	456.8 a
W2	63.30 a	34.7 ab	46.4 a	824.6 a	74.3 a	63.7 a	10.4 a	4.9 a	115.6 a	175.4 a	461.2 a
W3	64.76 a	30.0 c	37.9 b	822.7 a	74.4 a	64.6 a	9.1 a	4.7 a	114.8 a	172.3 a	459.4 a
r	0.938 **	−0.093	0.023	0.162	0.086	0.887	0.671	−0.129	0.737 *	0.717 *	0.571

Values followed by different lowercase letters correspond to significant differences at p < 0.05. r is the correlation coefficient between grain yield and each grain processing quality indicator. * and ** mean that the correlation coefficient reached the significant level at p < 0.05 and p < 0.01. STA: starch content (%); WGC: wet gluten content (%); SV: sedimentation value (mL); TW: test weight (g L⁻¹); FE: flour extraction (%); WA: water absorption (%); DST: dough stability time (min); DDT: dough development time (min); EA: extensibility area (cm²); DM: dough malleability (mm); MR: maximum resistance (E.U.).

). Specifically, W3 led to depressed wet gluten content and sedimentation values, with 18.3~21.5% and 13.2~18.3% (p < 0.05) reductions compared to the values obtained under W1 and W2, contrasting with the W0-associated inferior starch levels (5.0~12.2% deficit) and dough rheological properties, including dough stability time, extensibility, dough malleability, and maximum resistance, which decreased by 16.2~31.7%, 4.3~17.1%, 4.5~10.8%, and 4.3~13.1% (p < 0.05), respectively, compared to the values determined under the irrigation treatments. Notably, the flour functionality parameters (test weight, flour extraction, water absorption, and dough development time) remained irrigation-insensitive, while grain yield correlated positively with starch content, extensibility, and dough malleability (p < 0.05) but showed no association with other quality indices. These results established W2 as the irrigation optimum, achieving simultaneous starch content elevation (9.9% superior to that under W0) and gluten quality preservation (13.2% superior to that under W3), thereby resolving the yield–quality trade-off through precise water management during the reproductive stages.

3.4. Grain Protein and Its Components’ Content

The irrigation regimes induced threshold-dependent dynamics in protein metabolism, with both protein content and yield following unimodal response patterns peaking under W2 (

Table 5. Content of wheat grain protein and its components under different irrigation treatments.

Year	Treatment	Protein Content (%)	Protein Yield (kg ha−1)	Albumin (%)	Globulin (%)	Gliadin (%)	Glutenin (%)	Glutenin /Gliadin
2020	W0	13.92 b	702.2 c	1.95 ab	1.24 b	3.19 b	3.85 b	1.21 b
	W1	15.01 a	894.1 b	2.13 a	1.43 a	3.42 a	4.36 a	1.27 b
	W2	15.02 a	988.3 a	2.07 a	1.45 a	3.28 ab	4.52 a	1.38 a
	W3	13.64 b	914.4 b	1.85 b	1.16 b	3.05 c	3.89 b	1.28 b
2021	W0	13.84 ab	740.8 c	1.90 a	1.20 b	3.02 a	3.68 b	1.22 c
	W1	14.95 a	898.4 b	2.04 a	1.40 a	3.25 a	4.25 a	1.31 b
	W2	14.87 a	1020.2 a	1.99 a	1.35 a	3.16 a	4.41 a	1.40 a
	W3	13.21 b	888.1 b	1.69 b	1.13 b	2.99 a	3.31 c	1.11 d
r1		0.081	0.904 **	−0.198	0.053	−0.131	0.187	0.343
r2		/	0.498	0.933 **	0.981 **	0.851 **	0.954 **	0.812 *

Values followed by different lowercase letters correspond to significant differences at p < 0.05. r1 and r2 are the correlation coefficients between grain yield, protein content, and protein components. * and ** mean that the correlation coefficient reached the significant level at p < 0.05 and p < 0.01.

). The protein content under W3 demonstrated 9.1~11.6% and 9.2~11.2% reductions (p < 0.05) relative to the values under W1 and W2, respectively, while W0 exhibited the lowest protein yield, with 17.5~28.9% deficits (p < 0.05) across the irrigation treatments. Fractionation revealed irrigation-driven protein compositional shifts, exhibited unimodal optimization patterns peaking under the W1 or the W2 treatment. Specifically, W3 showed significant depressed (p < 0.05) albumin, globulin, gliadin, and glutenin contents compared to W1 and W2, while no significant difference was detected compared with the W0 treatment. A multivariate analysis confirmed strong yield–protein yield correlations (p < 0.01), while protein content correlated positively with glutelin proportion and albumin–globulin complexation but showed no direct linkage to protein yield parameters.

3.5. Relationship Between Water Consumption and Grain Yield, Protein Yield, and WUE

The total seasonal water consumption demonstrated significant quadratic relationships with grain yield, protein yield, and water use efficiency (WUE), exhibiting distinct optimization thresholds across three irrigation regimes (Figure 2). Regression modeling identified peak grain yield (6781 kg ha⁻¹) at 3903 m³ ha⁻¹ water input, while maximum protein yield (1022.9 kg ha⁻¹) and WUE (2.16 kg m⁻³) occurred at 3467 and 2058.8 m³ ha⁻¹, respectively. Further analysis showed that below 3250 m³ ha⁻¹, steep increases in grain yield and protein yield occurred with marginal WUE declines, representative of the high yield-efficiency optimization zone. Between 3250 and 3500 m³ ha⁻¹, grain yield showed continues gains alongside peak protein yield stabilization and controlled WUE reduction, establishing this range as the optimal zone for yield quality–water synergy. Beyond 3500 m³ ha⁻¹, diminishing returns prevailed, with grain yield increasing only slightly, contrasting sharp protein yield and WUE deteriorations, indicating this value as corresponding to the increased production and low-quality and -efficiency zone. Crucially, the 3250~3500 m³ ha⁻¹ bandwidth achieved simultaneous maximum attainable yield and peak WUE while maintaining optimal protein yield, thereby resolving the tripartite trade-off through precision hydrological management.

4. Discussion

4.1. Optimizing Water-Limited Irrigation to Enhance Grain Yield and WUE

Water availability critically governs wheat productivity through the stage-specific modulation of yield components, with grain number per spike demonstrating the greatest sensitivity to hydrological stress, followed by spike density and grain weight—parameters that can be strategically optimized through precision irrigation. The jointing–flowering phase emerged as the water stress phenological window for wheat, where precision irrigation timing and volume critically determine root system architecture, canopy structure optimization, and yield component synchronization. Strategic water application during this period optimizes the sink/source ratio, achieving yield enhancement, and maximizes the harvest index and water use efficiency improvements [30,31]. Controlled post-anthesis water stress during grain filling accelerates leaf senescence while triggering compensatory photoassimilate redistribution mechanisms—enhancing the remobilization efficiency of vegetative reserves and the stem contribution to grain yield compared to full irrigation regimes, thereby offsetting reduced post-anthesis carbon assimilation through optimized sink–source dynamics, as demonstrated in semi-arid cultivation systems [32].

The two-year dataset revealed significant trade-offs in the water-restricted treatments, where W0 and W1 improved water use efficiency by 24.9 and 14.5% compared to W3, but at the expense of yield reduction by 22.6 and 10.9%, including spike density (17.7 and 7.4% reduction) and 1000-grain weight (9.7 and 6.0% decline), which confirmed that single or absent irrigation fails to sustain the development required for wheat plant growth, with hydrological stress impairing tiller survival and grain filling efficiency. These deficits stemmed from compromised canopy functionality, manifested through reductions in photosynthetic architecture (leaf area index; canopy light transmittance) and decreases in flag leaf physiological performance (SPAD; Pn), ultimately curtailing biomass accumulation. The W2 irrigation protocol demonstrated superior water efficiency, enhancing water use efficiency by 19.1% relative to W3, while maintaining equivalent yields through minimal reductions in spike density (3.4%) and 1000-grain weight (1.3%). This irrigation strategy (W2) synchronized early-stage population architecture development with late-stage assimilate partitioning efficiency, substantially reducing the water inputs while amplifying grain filling precision through enhanced stem reserve mobilization [25,26,27,28]. Crucially, this optimized strategy achieved maximal yield potential with water input reduction, establishing an equilibrium between productivity and water efficiency that synergistically enhanced both metrics—a critical advancement for sustainable intensification paradigms in water-limited agricultural systems.

4.2. Water-Limited Irrigation Improves Grain Quality

Precision water management serves as a critical determinant of wheat end-use quality, with irrigation intensity directly modulating grain biochemical composition through nitrogen–carbon distribution [33]. A growing body of evidence demonstrates that supra-optimal irrigation compromises gluten matrix formation through protein dilution effects, reducing grain protein content and weakening dough rheological properties, such as extensibility and stability time compared to optimized water regimes [34,35]. Our results establish an irrigation quality optimization threshold, with the W1 and W2 regimes enhancing grain processing through higher protein content and dough stability compared to the W0 and W3 treatments, demonstrating that appropriate irrigation enhances grain processing quality, whereas excessive water inputs trigger quality degradation through dilution effects. This quality modulation stems from irrigation-induced nitrogen–carbon competition dynamics, where supra-optimal irrigation reduces soil nitrogen availability, diminishing root uptake efficiency [36], ultimately causing a reduction in grain protein content. Concurrently, precision irrigation elevates starch synthesis through enhanced photoassimilate partitioning, driving starch accumulation, which further exacerbates protein dilution through mass-balance effects—a dual mechanism evidenced by strong negative correlations between starch content and protein fractions. In addition, excessive irrigation disrupts grain protein accumulation through sink competition and impaired nitrogen remobilization. Prolonged soil moisture extends vegetative organ viability, creating sustained competition for photoassimilates during grain filling. Vigorous stem/leaf growth preferentially allocates carbohydrates and nitrogen away from reproductive sinks [37], while delayed senescence traps nutrients in vegetative tissues through suppressed protease activity [38] and cytokinin-mediated inhibition of nutrient transfer signals [39].

Protein, as the terminal product of nitrogen assimilation pathways, constitutes the primary determinant of wheat nutritional value and flour functionality with its compositional profile, particularly the glutenin/gliadin ratio directly governing dough rheological properties and end-use quality. The soil moisture status has been demonstrated as a critical determinant of wheat grain protein composition, with both water excess and deficit adversely affecting the accumulation and proportional distribution of protein fractions [40]. Specifically, Zhao et al. [41] revealed that limiting the irrigation frequency during reproductive growth stages effectively enhances the globulin, gliadin, and total protein content compared with conventional practices. Furthermore, Xu et al. [16] reported that strategic irrigation at the jointing and heading stages promotes the biosynthesis of storage proteins and facilitates gliadin polymer formation in developing kernels. These findings collectively highlight the importance of stage-specific water management in optimizing both the quantity and the quality of wheat grain proteins.

4.3. Synergistic Improvement of Wheat Yield and Quality and Water Efficiency

Crop productivity and grain quality emerge as physiologically antagonistic traits governed by photoassimilate accumulating and partitioning, where yield-exceeding thresholds trigger a significant negative correlation between grain yield and functional quality markers such as protein content in cereal crops—a metabolic trade-off arising from preferential carbon allocation to starch biosynthesis over nitrogen assimilation processes [42]. Conventional wisdom posits that agricultural systems face inherent trade-offs between achieving elevated concentrations of nutritionally valuable components and maximizing yield potential, where agroecological conditions and management strategies multifactorially regulate yield, quality, and water efficiency. Emerging evidence reveals threshold-dependent responses where optimal irrigation scheduling alone can enhance water use efficiency while maintaining maximal yield capacity. Specifically, precision irrigation and stage-specific fertilization can synchronize nitrogen translocation efficiency with photoassimilate partitioning, achieving a synergistic yield–quality improvement. This nonlinear yield–protein relationship, governed by sink–source coordination during reproductive phases, provides a physiological foundation for transcending trade-offs through water-smart agronomy that elevates both productivity and functional quality while reducing hydrological inputs. This study validated the effectiveness of dual-phase irrigation during the tillering–jointing stage and heading-flowering stage (W2) in achieving yield stability with water conservation compared to conventional practices, while simultaneously demonstrating the benefits of the W2 treatment in enhancing wheat grain quality, including the content of protein and its components, wet gluten content, and sedimentation value.

5. Conclusions

The study established critical water thresholds for spring wheat optimization, demonstrating that progressive irrigation intensification enhanced flowering-stage agronomic traits and grain and protein yields until reaching performance plateaus under the W2 (dual irrigation at the tillering–jointing and heading–flowering stages) regime. Both grain yield and protein yield exhibited unimodal responses, peaking at 3250~3500 m³ ha⁻¹ of water consumption—an optimization window achieving maximal yield potential while maintaining peak water use efficiency and protein yield retention. The W2 irrigation strategy reconciled these traditionally competing objectives through phenology-aligned water allocation during tillering–flowering to establish the photosynthetic capacity, and the following reproductive-phase irrigation enhanced assimilate remobilization, ultimately reducing the total water inputs compared to conventional practices, while sustaining yield stability and elevating quality parameters. These findings validate W2 as a scalable precision irrigation protocol that provides actionable solutions for sustainable intensification in the Hetao Irrigation District.