Optimizing irrigation scheduling in apple orchards: soil moisture thresholds balance yield, fruit quality, and water use efficiency

Abstract

Water scarcity necessitates efficient irrigation strategies in high-value crops like apple. While regulated deficit irrigation (RDI) is known to influence apple yield and quality, the optimal soil moisture threshold for automated systems in region-specific contexts remains poorly defined. This two-year study (2024–2025) evaluated the effects of four precise, soil moisture-based irrigation thresholds on the yield, fruit quality, and water productivity (WP) of ‘Super Chief’ apples in the northwestern Himalayas. An automated irrigation system using a Human-Machine Interface (HMI) and calibrated soil moisture sensors maintained Volumetric Water Content (VWC) at four thresholds: 100% (I100, 45% VWC), 80% (I80, 36% VWC), 60% (I60, 27% VWC), and 40% (I40, 18% VWC) of field capacity (θfc). Full irrigation (I100) maximized yield (70.80 t ha⁻¹) and fruit size (165.60 g). Severe deficit (I40) significantly enhanced fruit firmness (9.56 N) and soluble solids content (13.95 °Brix) but at the cost of a 20% yield reduction. Crucially, the I80 treatment achieved a near-optimal yield (66.24 t ha⁻¹) while using 20% less irrigation water than I100, resulting in a 31% higher water productivity (WP). Multi-criteria TOPSIS analysis objectively ranked I80 as the best compromise strategy. For ‘Super Chief’ apples in the Himalayan region, maintaining soil moisture at 80% θfc via sensor-based automation is the superior strategy, effectively balancing marketable yield, enhanced fruit quality, and water conservation.

Introduction

Global freshwater resources are under unprecedented strain, with agriculture accounting for over 70% of withdrawals¹. In this context, improving water productivity (WP) is critical for sustainable food production, particularly for high-value perennial horticultural crops like apple (Malus domestica Borkh.) that are often grown in water-scarce regions². India is the world’s fifth-largest apple producer, with the northwestern Himalayan state of Jammu and Kashmir contributing over 77% of national production, making efficient water management in this region an economic and environmental imperative³.

Conventional irrigation scheduling in apple orchards, often based on fixed timetables or visual assessment, typically leads to over-irrigation, promoting water loss and nutrient leaching, or under-irrigation, inducing plant stress⁴. The adoption of precision irrigation, which applies water based on actual plant or soil moisture status, is a promising solution. Soil moisture sensor-based irrigation provides a direct measure of plant-available water, enabling precise application that matches crop requirements⁵.

Regulated Deficit Irrigation (RDI), which involves applying water below full crop evapotranspiration during specific, drought-tolerant phenological stages, has been extensively studied in fruit trees⁶. In apples, RDI can improve fruit quality traits such as firmness, soluble solids content, and red coloration by moderating vegetative growth and altering carbohydrate partitioning⁷. However, the yield response to water deficit is complex and can be detrimental if the stress is too severe or ill-timed, highlighting a critical trade-off⁸.

Previous research has demonstrated that RDI effects are highly cultivar-, rootstock-, and climate-specific. Studies on traditional cultivars such as ‘Gala’ and ‘Fuji’ under Mediterranean or semi-arid conditions have identified optimal deficit levels ranging from 50% to 75% of full irrigation, often applied during post-harvest or fruit expansion stages^9,10. Similarly, work on ‘Golden Delicious’ and pear have shown that maintaining soil moisture at 75–80% of field capacity can balance yield and quality^11,12.

Despite substantial advances in regulated deficit irrigation (RDI), a critical gap remains between experimental research and its practical implementation in automated irrigation systems. Most sensor-based systems operate using a conventional “refill-to-field-capacity” approach, restoring soil moisture to field capacity whenever predefined thresholds are reached, rather than maintaining a controlled and sustained deficit required for effective RDI. Moreover, optimum soil moisture thresholds are highly dependent on cultivar, rootstock, soil type, and local climatic conditions, limiting the direct transferability of recommendations developed for other production regions. Consequently, precise sensor-based volumetric water content (VWC) thresholds for automated irrigation of ‘Super Chief’ apple under the temperate Himalayan conditions remain unavailable, highlighting the need for region-specific optimization.

Therefore, this study addresses these critical translational and contextual gaps by introducing and validating a novel irrigation management paradigm: sustained deficit maintenance via automation. Our core innovations are threefold:

1. Paradigm Shift in Automation Logic: We move beyond conventional “refill” systems by deploying an automation system explicitly programmed to maintain soil moisture at a constant, sub-field-capacity threshold, thereby implementing a true, sustained RDI strategy.

2. Cultivar- and Region-Specific Precision: We establish and validate exact, sensor-based Volumetric Water Content (VWC) set points that optimize the trade-off between yield, fruit quality, and water productivity for ‘Super Chief’ apples under Himalayan conditions, a combination previously unstudied.

3. Holistic, Decision-Ready Output: We employ Multi-Criteria Decision Analysis (TOPSIS) to objectively identify the single best threshold, providing a definitive, actionable recommendation for growers and system designers.

The objectives of this two-year study were:

1. To evaluate the effects of four precise, automatically-maintained soil moisture thresholds (40%, 60%, 80%, and 100% of field capacity) on the yield, fruit quality, and water productivity of ‘Super Chief’ apples.

2. To identify the optimal, singular soil moisture threshold that best balances high marketable yield, superior fruit quality, and enhanced water productivity.

3. To demonstrate the efficacy of a cost-effective HMI-based automation system for implementing this sustained deficit maintenance strategy.

Materials and methods

Site description, plant material, and experimental design

A two-year field study was conducted during the 2024 and 2025 growing seasons (March to September) at the experimental farm of SKUAST-K, Shalimar, Jammu & Kashmir, India (34°05’N, 74°50’E; 1605 m above sea level). The experiment utilized mature (planted in 2013) ‘Super Chief Sandidge’ apple trees grafted on M.9 T337 dwarfing rootstock, trained in a high-density vertical axis system. The soil was classified as a silty clay loam; its detailed physicochemical properties are presented in Table 1.

Table 1.

Physicochemical properties of the experimental orchard soil.

Property	Value	Unit
Particle Size Distribution
Sand	19.6	%
Silt	48.0	%
Clay	32.4	%
Textural Class	Silty Clay Loam	–
Water Retention
Field Capacity (θfc)	45	%
Permanent Wilting Point	13	%
Physical Properties
Bulk Density	1.25	g cm⁻³
Particle Density	2.46	g cm⁻³
Chemical Property
Electrical Conductivity	0.47	dS m⁻¹

The experimental layout followed a randomized complete block design (RCBD) with four irrigation treatments and five replications. The four treatments (I100, I80, I60, I40) were randomly assigned within each block to minimize spatial variability. Each experimental plot (replicate) consisted of six trees arranged in a row, with a 1.0 m spacing between trees and 3.5 m between rows.

To avoid edge effects and ensure measurements reflected treatment impacts without interference from adjacent plots, data were collected exclusively from the two central trees within each replicate plot (i.e., two trees per replication, across all five replications per treatment).

Weather conditions

The regional climate is temperate, with warm summers. Total seasonal rainfall during the 2024 and 2025 growing seasons was 677.2 mm and 475.5 mm, respectively. Pan evaporation was higher, recorded at 777.2 mm in 2024 and 828.2 mm in 2025, indicating a consistent soil water deficit that necessitated irrigation. Daily meteorological data (maximum and minimum temperature, relative humidity, rainfall, and pan evaporation) were recorded from an on-site agrometeorological station (Fig. 1a and b).

Fig. 1.

(a) Daily meteorological data during experimental period (2024). (b) Daily meteorological data during experimental period (2025).

Irrigation treatments and automation system

The four VWC thresholds (40%, 60%, 80%, and 100% of θfc) were selected to represent a gradient from severe deficit to full irrigation, based on prior research in apple RDI (e.g^4,13, and the specific water retention characteristics of the silty clay loam soil at the experimental site (Table 1). These thresholds allowed for the evaluation of both physiological and agronomic responses across a realistic range of soil moisture availability for apple trees in the region.

• I100: Irrigation triggered when VWC < 100% θfc (45% VWC).

• I80: Irrigation triggered when VWC < 80% θfc (36% VWC).

• I60: Irrigation triggered when VWC < 60% θfc (27% VWC).

• I40: Irrigation triggered when VWC < 40% θfc (18% VWC).

A drip irrigation system with pressure-compensating emitters (discharge: 4 L h⁻¹) delivered water to each treatment plot. Irrigation frequency was determined by the automated system based on real-time soil moisture data, ranging from daily to every 3–4 days during high evaporative demand, and less frequently during cooler or rainy periods (up to 7–10 days). This variability ensured that soil moisture was consistently maintained near the target VWC thresholds without over-irrigation.

A custom, cost-effective automation system was deployed, consisting of:

Controller:

A Kinco Green Series HMI panel housed in an IP67 enclosure, programmed to execute control logic without a separate PLC.

Sensors: Calibrated capacitive soil moisture sensors (Model: TEROS 10, Meter Group) were installed at 45 cm depth (primary root zone) in each plot. Calibration was performed using gravimetric soil moisture measurements before study initiation.

Actuation:

2-way solenoid valves (AEROTAC, Model 2 W-250-25) controlled water flow from overhead HDPE storage tanks.

Irrigation Logic:

IF (Average VWC < Set Threshold) THEN OPEN solenoid valve; ELSE CLOSE valve. Irrigation continued until soil moisture reached 3–5% above the threshold, ensuring uniform moisture distribution.

Water was delivered via a drip irrigation system with pressure-compensating emitters (discharge: 4 L h⁻¹). Irrigation frequency varied by treatment and weather conditions, ranging from daily to every 3–4 days during peak evapotranspiration periods. The total seasonal irrigation water applied for each treatment is detailed in Table 2.

Table 2.

Seasonal irrigation water applied per treatment during growing seasons 2024 and 2025.

Treatment	Irrigation Threshold (% Ɵfc)	Total Seasonal Water applied (mm) 2024	Total Seasonal Water applied (mm) 2025
I100	100%	840.60	895.7
I80	80%	672.48	716.6
I60	60%	504.36	537.4
I40	40%	336.24	358.3

Soil moisture sensor calibration and management of spatial variability

The reliability of soil moisture data is paramount for precision irrigation. To ensure accuracy and address the inherent spatial variability of soil moisture under drip irrigation, a rigorous calibration and sensor deployment protocol was implemented.

Sensor placement justification and protocol

Capacitive soil moisture sensors (TEROS 10, Meter Group) were used for continuous monitoring. The sensor depth was set at 45 cm based on the known root architecture of mature ‘Super Chief’ trees on M.9 T337 dwarfing rootstock, whose functional, water-absorbing roots are predominantly concentrated in the 20–60 cm soil layer in high-density plantings.

To ensure that the single sensor per plot provided a representative measure of root-zone moisture status, it was installed equidistant between the two central sample trees, approximately 30 cm from the base of each trunk. This placement locates the sensor within the area of highest fine-root density and directly within the wetted bulb created by the drip emitter, ensuring that the recorded VWC reflects the moisture environment most relevant for tree water uptake. This approach is a standard and validated practice in precision irrigation research for drip-irrigated tree crops.

Sensor calibration protocol

Sensors were calibrated in situ prior to the start of each growing season to account for site-specific soil properties (silty clay loam). The calibration procedure followed a destructive gravimetric method:

1. Sensor Installation: Sensors were installed at the target depth (45 cm) in representative locations outside the experimental plots but within the same orchard block.

2. Controlled Wetting & Drying: The soil around each sensor was saturated and allowed to drain naturally, creating a range of moisture conditions.

3. Paired Sampling: At multiple points during the drying cycle, the sensor’s raw output (dielectric permittivity, ε) was recorded. Immediately following each reading, a soil core was extracted using a standard auger from the immediate vicinity (within a 10 cm radius) of the sensor at depths of 40–50 cm.

4. Gravimetric Analysis: The fresh weight of each soil sample was recorded, then samples were oven-dried at 105 °C for 24 h to obtain dry weight. Volumetric Water Content (VWC, %) was calculated as: [(Fresh Weight - Dry Weight)/Dry Weight] × Bulk Density.

5. Regression Model: A linear regression model was established between the sensor’s raw output (ε) and the gravimetrically-determined VWC for each sensor. The average calibration equation across all sensors was VWC = 0.132 × ε – 3.71, with an average coefficient of determination (R²) exceeding 0.92. The Mean Absolute Error (MAE) was 1.2% VWC across the calibration range of 10% to 50% VWC. The resulting equation was then uploaded to the data logger for real-time conversion of raw sensor output to VWC.

Strategy to manage spatial variability

Spatial variability in soil moisture, particularly under point-source drip irrigation, is a critical challenge. The following measures were taken to ensure sensor data were representative and irrigation was applied uniformly:

1. Strategic Sensor Placement: As described in Sect. 2.4.1, sensors were placed within the primary wetted zone to standardize the measurement point relative to the water source and tree position.

2. System Design to Minimize Variability: The irrigation system was designed to promote uniformity: (a) Pressure-compensating emitters (4 L h⁻¹) were used to ensure equal discharge across the field regardless of line pressure changes; (b) Regular system maintenance checks were performed to prevent clogging; (c) Irrigation run times were sufficiently long to ensure lateral water movement and achieve the target “3–5% above threshold” moisture level, promoting a more uniform wetting front.

3. Validation of Uniformity: To confirm that the single sensor reading was representative of the plot’s root zone condition, periodic spot checks were conducted during the first season. Using a handheld portable probe (similar technology), VWC was measured at 2–3 additional points within the same plot’s wetted zone. The readings showed a mean coefficient of variation of less than 15% from the installed sensor’s reading, confirming acceptable representativeness.

4. Data Aggregation for Control: The irrigation control logic used a moving average of the most recent 6 sensor readings (over 6 h) to trigger irrigation, buffering against momentary, localized fluctuations and ensuring irrigation decisions were based on a stable trend in root zone moisture.

Continuous monitoring and logging

Following calibration, sensors were installed in all experimental plots as described. Volumetric water content was logged at 60-minute intervals throughout both growing seasons via the central HMI system.

Crop management practices

Fertilization followed regional recommendations for high-density apple orchards. Urea (46% N) was applied at a total rate of 200 kg N ha⁻¹ year⁻¹, split into three applications: at bud break (40%), fruit set (40%), and post-harvest (20%). NPS fertilizer (18% N, 46% P₂O₅, 0% K₂O) was applied at 150 kg ha⁻¹ during the dormant season (late winter). All fertilizers were broadcast evenly within the tree rows and lightly incorporated. Pest and disease management followed integrated pest management protocols, applied uniformly across all treatments.

Data collection

Yield components

At commercial harvest, total yield per tree was recorded and converted to tonnes per hectare (t ha⁻¹). A random sample of 20 fruits per replication was used to determine average fruit weight (g), diameter (mm), and length (mm).

Fruit quality

Fruit firmness (N)

Measured on two pared sides using a handheld penetrometer (Guss, Model: FT-327)

Total soluble solids (TSS, °Brix)

Determined with a digital refractometer (Atago, Model: PAL-1)

Titratable acidity (TA, %)

Assessed by titrating apple juice with 0.1 N NaOH and expressed as percent malic acid.

Water productivity (WP) calculation

WP was calculated using only irrigation water applied, excluding rainfall, to isolate treatment effects:

WP (kg m⁻³) = Fruit Yield (kg ha⁻¹)/Total Irrigation Water Applied (m³ ha⁻¹).

Effective rainfall was not included in WP calculations to maintain clarity in irrigation treatment comparisons.

TOPSIS multi-criteria decision analysis

The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) was employed to comprehensively evaluate and rank irrigation treatments based on five beneficial criteria: yield, fruit weight, firmness, TSS, and WP. Equal weighting was adopted to avoid subjective prioritization among agronomic, quality, and water-efficiency criteria, ensuring an unbiased, holistic assessment. The analysis followed these steps:

1. Normalization of the decision matrix.

2. Determination of ideal best (A+) and ideal worst (A−) solutions.

3. Calculation of Euclidean distances from A+ (D+) and A− (D−).

4. Computation of relative closeness (Ci) as:

   

Where  and represent the Euclidean distances from the ideal best and ideal worst solutions, respectively.

Higher Ci values indicate better overall performance.

Statistical analysis

Data were analyzed using a two-way analysis of variance (ANOVA), with irrigation treatment and year considered fixed factors. The experimental unit consisted of one replicate plot (six trees), with measurements obtained from the two central trees of each plot. Mean separation was performed using Duncan’s Multiple Range Test (DMRT) at p ≤ 0.05. Statistical analyses were conducted using IBM SPSS Statistics version 24.0.

Results

Statistical overview and treatment consistency

A two-way ANOVA revealed that irrigation treatment had a highly significant effect (p < 0.001p < 0.001) on all measured agronomic and qualitative parameters (Table 3). In contrast, neither Year nor the Year × Treatment interaction reached significance (p > 0.05p > 0.05) for any variable, confirming that treatment responses were consistent across the two growing seasons despite contrasting rainfall patterns (677.2 mm in 2024 vs. 475.5 mm in 2025). This temporal stability validates the robustness of the findings and permits the presentation of pooled data.

Table 3.

Two-way ANOVA results (p-values) for the effects of Year, Treatment, and their Interaction.

Parameter	Year (Y)	Treatment (T)	Y × T Interaction
Yield (t ha⁻¹)	0.312	< 0.001	0.450
Fruit Weight (g)	0.285	< 0.001	0.520
Fruit Diameter (mm)	0.401	< 0.001	0.485
Fruit Firmness (N)	0.210	< 0.001	0.630
TSS (°Brix)	0.380	< 0.001	0.550
Titratable Acidity (%)	0.295	< 0.001	0.510
Water Productivity (kg m⁻³)	0.335	< 0.001	0.490

p-values indicate significant effects (p < 0.05p < 0.05).

Yield and fruit size: identifying the inflection point for sustainable deficit

Irrigation treatment significantly affected fruit weight, fruit diameter, and yield (Table 4). Full irrigation (I100) produced the highest fruit weight (165.6 g), fruit diameter (81.3 mm), and yield (70.80 t ha⁻¹), whereas all deficit irrigation treatments resulted in progressive reductions in these parameters. The I80 treatment maintained a yield of 66.24 t ha⁻¹, representing approximately 94% of the maximum yield obtained under full irrigation while using 20% less irrigation water than I100.

Table 4.

Impact of irrigation treatment on yield and fruit physical characteristics (pooled data for 2024 and 2025).

Treatment	Fruit Weight (g)	Fruit Diameter (mm)	Yield (t ha⁻¹)
I100	165.6 ± 5.2a	81.3 ± 2.1a	70.80 ± 2.3a
I80	150.2 ± 4.8b	78.5 ± 1.9b	66.24 ± 2.1b
I60	132.4 ± 4.1c	74.2 ± 1.7c	60.12 ± 1.9c
I40	122.1 ± 3.8d	70.8 ± 1.5d	56.64 ± 1.8d

*Values are mean ± SD (n = 5). Different letters within columns indicate significant differences (Duncan’s Multiple Range Test, p ≤ 0.05p ≤ 0.05).*.

Soil moisture dynamics across treatments

Soil moisture monitoring confirmed that the automated system successfully maintained distinct moisture regimes (Table 5). I100 maintained VWC near field capacity (42–45%), while I40 operated at 40% of field capacity (17–19% VWC). I80 consistently maintained soil moisture at 34–37% VWC. The vertical soil moisture profile showed that deficits were most pronounced in the upper 30 cm, with deeper layers (40–60 cm) showing less variation between treatments (Fig. 2).

Table 5.

Seasonal average VWC, standard deviation, and irrigation frequency across treatments (pooled 2024–2025).

Treatment	Mean VWC (%)	SD (±%)	Avg. Irrigation Frequency (days interval)
I100	43.2	1.8	1–2
I80	35.1	1.5	2–4
I60	26.4	1.2	4–6
I40	17.9	1.1	6–10

Fig. 2.

Mean seasonal volumetric water content (VWC) maintained under the four automated irrigation treatments during the study period. Bars represent mean values, and error bars indicate ± standard deviation (SD) of repeated sensor measurements.

Fruit quality and water productivity: the synergy of the moderate deficit regime

Fruit firmness, total soluble solids (TSS), titratable acidity, and water productivity were significantly influenced by irrigation treatment (Table 6). Fruit firmness and TSS increased progressively with increasing irrigation deficit, reaching maximum values under I40 (9.56 N and 13.95 °Brix, respectively). Water productivity also increased significantly as irrigation volume decreased, with I40 recording the highest value (13.51 kg m⁻³), followed by I60 (8.43 kg m⁻³) and I80 (8.05 kg m⁻³). The I80 treatment combined improved fruit quality with substantially higher water productivity than full irrigation while maintaining comparatively high yield.

Table 6.

Effect of irrigation treatment on fruit quality parameters and water productivity (pooled data for 2024 and 2025).

Treatment	Firmness (N)	TSS (°Brix)	TA (%)	WP (kg m⁻³)
I100	7.35 ± 0.30d	12.75 ± 0.40c	0.180 ± 0.010c	6.16 ± 0.20d
I80	8.50 ± 0.33c	12.90 ± 0.43c	0.190 ± 0.010bc	8.05 ± 0.30c
I60	8.74 ± 0.32b	13.72 ± 0.41b	0.220 ± 0.011b	8.43 ± 0.30b
I40	9.56 ± 0.34a	13.95 ± 0.50a	0.240 ± 0.012a	13.51 ± 0.50a

*Values are mean ± SD (n = 5). Different letters within columns indicate significant differences (Duncan’s Multiple Range Test, p ≤ 0.05p ≤ 0.05).*.

Integrated assessment via multi-criteria decision analysis: validating the novel strategy

TOPSIS analysis ranked irrigation treatments according to their overall performance based on yield, fruit weight, firmness, TSS, and water productivity (Table 6). The I80 treatment achieved the highest relative closeness coefficient (C_i = 0.745), followed by I100 (0.572), I60 (0.431), and I40 (0.288), indicating the highest overall ranking among the evaluated irrigation strategies (see Table 7).

Table 7.

TOPSIS analysis results including Euclidean distances and relative closeness scores.

Treatment	D⁺ (Distance to Ideal Best)	D⁻ (Distance to Ideal Worst)	Ci (Relative Closeness)	Rank
I80	0.124	0.362	0.745	1
I100	0.267	0.356	0.572	2
I60	0.398	0.301	0.431	3
I40	0.512	0.207	0.288	4

I80 achieved the highest TOPSIS score (0.745), with the smallest distance to the ideal best (D⁺ = 0.124) and largest distance from the ideal worst (D⁻ = 0.362), significantly outperforming both full irrigation (I100: 0.572) and more severe deficits (I60: 0.431; I40: 0.288). This mathematically confirms that I80 represents the optimal compromise, effectively balancing the competing objectives of production, quality, and sustainability.

Discussion

Collectively, the present findings indicate that the moderate deficit irrigation treatment (I80) provided the most favorable balance between productivity, fruit quality, and water productivity under the temperate Himalayan conditions. While full irrigation (I100) achieved the highest fruit yield, increasing irrigation deficit progressively improved fruit quality attributes and water productivity. However, severe deficit irrigation (I40), despite maximizing water productivity and fruit firmness, resulted in a substantial reduction in yield. These findings demonstrate that maintaining soil moisture at approximately 80% of field capacity (36% VWC) represents an effective compromise between maximizing orchard productivity and improving resource-use efficiency under automated irrigation.

Re-interpreting deficit irrigation: from periodic stress to sustained optimization

Our results challenge the conventional application of regulated deficit irrigation (RDI) as a practice of imposing periodic, sometimes severe, water stress during specific phenological stages. Instead, they support an emerging paradigm of sustained moisture optimization, where automation maintains soil moisture within a productive deficit zone. For ‘Super Chief’ under Himalayan conditions, this zone corresponded to approximately 80% of field capacity (36% VWC). This approach continuously favored fruit quality and water-use efficiency while avoiding the severe stress associated with larger yield reductions under I40. Furthermore, the non-significant Year × Treatment interaction indicates that this automated moisture-maintenance strategy produced consistent responses across contrasting growing seasons, suggesting its suitability as a stable irrigation management approach.

Yield preservation under moderate deficit: physiological mechanisms and the deficit-irrigation trade-off

The observed decline in yield with increasing water deficit is not only reasonable but is the foundational trade-off that defines regulated deficit irrigation (RDI). The core objective of RDI is rarely to increase absolute yield, but to optimize the yield-per-unit-water-applied (water productivity) and enhance fruit quality, often while accepting a modest, controlled reduction in total output. Our finding that full irrigation (I100) maximized yield validates the experimental design by confirming that the control treatment was not under-watered.

The key advance of this study lies in precisely quantifying the magnitude of this trade-off for ‘Super Chief’ in the Himalayan region. The I80 threshold incurred only a 6.4% yield penalty in exchange for a 31% gain in water productivity and significant improvement in fruit firmness. A plausible explanation for this favorable balance is that maintaining soil moisture at approximately 80% of field capacity (36% VWC) may have avoided severe plant water stress while inducing only mild physiological responses associated with moderate deficit irrigation. Previous studies have suggested that moderate water deficits can maintain leaf water status above critical thresholds, while stimulating abscisic acid (ABA)-mediated stomatal regulation and moderating vegetative growth, thereby improving assimilate partitioning towards fruit development and enhancing water-use efficiency¹⁴. Although these physiological mechanisms were not directly measured in the present study, they provide a reasonable explanation for the observed balance between yield retention and improved water productivity under the I80 treatment.

Furthermore, the M.9 T337 dwarfing rootstock, with its shallow, fibrous root system concentrated in the top 0–60 cm, is exceptionally responsive to the consistent moisture environment provided by our drip-based, sustained-deficit strategy. This rootstock-specific adaptation helps explain the remarkable yield preservation under I80, as the irrigation system directly targets the zone of highest root activity—precisely where our soil-moisture sensors were placed (45 cm depth, within the wetted bulb).

Fruit quality enhancement: beyond solute concentration to active metabolic regulation

The significant improvement in fruit firmness and total soluble solids (TSS) under deficit irrigation, particularly under the I80 treatment, suggests that moderate water deficit may induce physiological and biochemical adjustments beyond simple concentration effects resulting from reduced fruit water content. A plausible explanation, though not directly tested in this study, is that moderate water deficits may influence carbohydrate metabolism and modify the expression or activity of cell wall-associated enzymes, which have been reported to contribute to fruit firmness and sugar accumulation in other fruit species^14,15. However, as these enzymatic and molecular responses were not directly measured in the present investigation, they should be regarded as hypothetical mechanisms warranting future research rather than demonstrated processes.

The improved fruit firmness (8.50 N) and TSS (12.90 °Brix) observed under the I80 treatment are consistent with these previously reported responses to moderate deficit irrigation, while simultaneously maintaining high yield and water productivity. In addition, increased fruit firmness may contribute to improved postharvest handling and storage potential, although postharvest performance was beyond the scope of the present investigation and warrants future study.

Water productivity and the economics of efficient irrigation

The 31% improvement in water productivity (WP) under the I80 treatment highlights the hydrological efficiency of maintaining a moderate, sustained soil moisture deficit through automated irrigation. Compared with full irrigation, the I80 treatment reduced seasonal irrigation water application by 20% while maintaining approximately 94% of the maximum yield, indicating a more efficient utilization of irrigation water. From a practical perspective, the lower irrigation requirement is expected to reduce pumping time, electricity consumption, and overall irrigation costs, which is particularly beneficial for Himalayan apple orchards where water availability and energy resources are often constrained. Furthermore, the automated irrigation system minimizes manual intervention and simplifies irrigation scheduling, thereby improving management efficiency. The TOPSIS analysis integrates these agronomic, quality, and water-use efficiency attributes, confirming I80 as the most balanced irrigation strategy for sustainable orchard management.

Comparative analysis and global context

Placing our findings in a global context highlights their specificity and relevance. A comparative analysis shows that the 31% WP improvement at I80 exceeds the 10–25% gains typically reported in many apple RDI studies in semi-arid regions (e.g¹⁶,. Concurrently, the yield preservation (> 93%) is at the high end of the spectrum reported for deficit irrigation in apples (cf. Atay et al.⁹,. This favorable outcome underscores the efficiency of the sustained maintenance approach compared to strategies employing more severe or variable deficits.

Crucially, this comparison reinforces our central thesis: optimal irrigation thresholds are not universal. Our recommended 80% θfc (36% VWC) is the calibrated optimum for the *’Super Chief’/M.9/Himalayan silty clay loam* system. It differs from recommendations for ‘Gala’ in Mediterranean climates or ‘Fuji’ in arid zones, thereby filling a critical knowledge gap and providing the missing, locally-validated dataset necessary for true precision agriculture in this major Indian apple-producing region.

Implications for long-term orchard health and sustainability

While this two-year study provides robust evidence for the immediate benefits of maintaining soil moisture at 80% of field capacity, the long-term implications of sustained deficit irrigation on orchard health and productivity warrant careful consideration. Perennial fruit trees may exhibit cumulative physiological responses to repeated water stress that are not captured in short-term trials.

Potential long-term effects could include altered carbon allocation patterns affecting reserve storage, gradual changes in root system architecture, and influences on biennial bearing tendencies. However, the moderate nature of the I80 deficit—imposing only a mild, consistent reduction in soil moisture—is designed to minimize such risks. By avoiding the severe stress imposed in the I40 treatment, which incurred a 20% yield loss, the I80 strategy likely poses a lower risk of inducing long-term decline in tree vigor or productive capacity.

Importantly, our automated, sensor-based approach ensures that trees experience a stable, predictable moisture environment, which may be less disruptive than traditional irrigation scheduling that subjects trees to cycles of saturation and dry-down. This stability could be beneficial for long-term root health and function.

Limitations and future research directions

While this study provides robust evidence supporting the performance of the I80 irrigation threshold under temperate Himalayan conditions, several limitations should be acknowledged. First, the experiment was conducted using a single cultivar (‘Super Chief’) grafted onto M.9 rootstock; therefore, the observed responses may differ among other commercially important cultivar-rootstock combinations. Second, the study evaluated a sustained deficit irrigation strategy throughout the growing season rather than stage-specific deficit irrigation, which may offer additional opportunities for optimizing water use and fruit quality. Third, although the discussion proposes plausible physiological mechanisms underlying the observed responses, these processes (e.g., hormonal regulation, carbohydrate partitioning, and cell wall metabolism) were not directly measured and should therefore be regarded as hypotheses rather than experimentally verified mechanisms. Finally, the conclusions are based on two consecutive growing seasons at a single experimental location, and longer-term, multi-location validation would further strengthen the applicability of the proposed irrigation threshold.

Future research should therefore investigate: (1) the performance of the 80% field-capacity threshold across diverse apple cultivar-rootstock combinations and production environments; (2) the integration of stage-specific deficit irrigation with automated sensor-based scheduling; (3) the physiological and molecular mechanisms governing fruit quality responses through measurements of plant water relations, hormone dynamics, carbohydrate metabolism, and cell wall-modifying enzymes; and (4) comprehensive economic analyses quantifying water, energy, and labor savings together with long-term orchard productivity. Such investigations would further refine precision irrigation recommendations and facilitate broader adoption of automated irrigation systems in commercial apple production.

Conclusion

This two-year study demonstrated that maintaining soil moisture at approximately 80% of field capacity (36% VWC) using a sensor-based automated irrigation system provided the most favorable balance between yield, fruit quality, and water productivity for ‘Super Chief’ apple under the northwestern Himalayan conditions evaluated. Compared with full irrigation, the I80 treatment retained approximately 94% of maximum yield while reducing seasonal irrigation water application by 20%, improving water productivity by 31%, and increasing fruit firmness by approximately 16%. Multi-criteria decision analysis (TOPSIS) further identified I80 as the highest-ranked treatment among the irrigation strategies evaluated.

These findings indicate that maintaining soil moisture near 80% of field capacity is a promising irrigation strategy for ‘Super Chief’ apple orchards under temperate Himalayan conditions. However, because this study was conducted using a single cultivar-rootstock combination over two growing seasons, further validation across additional cultivars, rootstocks, environments, and longer-term experiments is warranted before broader generalization.

Overall, the results provide practical evidence supporting the use of sensor-based automated irrigation to improve water-use efficiency while maintaining commercially acceptable productivity under conditions of increasing water scarcity.

Author contributions

Nowsheen Nazir: Conceptualization, Methodology, Investigation, Formal analysisMeinaz Nissar: Formal analysis, Data curation, VisualizationKrishna Nemali: Sensor calibration, Data collection, Validation.Abdul Raouf Malik: VisualizationAbdul Malik: Data interpretation, Funding acquisition, Project administrationRifat Bhat: Data interpretationAmit Kumar: ValidationShafat Ahmad Banday: SoftwareNikhat J Siddiqi: Experimental designCharu Bisht: Data curation, VisualizationShowkat Rasool: Sensor calibration, Data collection, Validation.Immad A. Shah: Formal analysis, Data curationAb Waheed Wani: Data interpretationTashooq Ahmad Bhat: Formal analysis, Data curation, Visualization, Writing – original draft.

Funding

This work was supported by the Science and Engineering Research Board, Department of Science and Technology (SERB-DST), India. The authors are grateful to Ongoing Research Funding Program (ORF-2026-376), King Saud University, Riyadh, Saudi Arabia.

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.