Study on Alfalfa Water Use Efficiency and Optimal Irrigation Strategy in Agro-Pastoral Ecotone, Northwestern China

Abstract

Agro-pastoral ecotone is an important livestock production area in the north of China, and alfalfa is the main pasture crop in this area. Aiming to address the issues of groundwater overexploitation in the area with water demand, we assessed the consumption pattern, irrigation scheduling, and water usage efficiency of alfalfa under subsurface drip irrigation. Alfalfa was used as the research object in this study. A DSSAT model was used to simulate the soil moisture, yield, and other alfalfa grow characteristics during a two-year in situ observation study and provide information on the best irrigation techniques and the water-use efficiency of alfalfa in the agro-pastoral ecotone of Northwestern China. The results showed that the ARE, nRMSE, and R² values of the alfalfa soil water content, leaf area index, and yield varied between 3.82% and 5.57%, 4.81% and 8.06%, and 0.86 and 0.93, respectively, the accuracy of the calibrated and validated parameters were acceptable, and the model could be applied to this study. The water consumption of alfalfa ranged from 395.6 mm to 421.8 mm during the whole year, and the critical water consumption period was the branching stage and the bud stage. During the branching stage and the bud stage, water consumption was 30–31% and 31–33% of the total water consumption, and the water consumption intensity averaged 2.97–3.04 mm/d and 4.23–4.97 mm/d. The variations of WUE and IWUE were 11.74–14.39 kg·m⁻³ and 7.12–9.31 kg·m⁻³. Irrigation increased the water productivity of rain-fed alfalfa by 49.48–64.70% and increased the yield of alfalfa by 17.87–34.72%. With the highest yield as the goal, the recommended irrigation volumes for normal and dry flow years were 200 mm and 240 mm; with the goal of the highest utilization of groundwater resources, the recommended irrigation volumes for normal and dry flow years were 160 mm and 192 mm. The results of this study are expected to provide scientific and technological support for the rational utilization of groundwater and the scientific improvement of alfalfa yields in the agro-pastoral ecotone of Northwestern China.

1. Introduction

The Yinshanbeilu region in northern China is a typical agro-pastoral ecotone, playing a vital role in cattle and sheep farming. It is also a key area for the production of artificial forage. Alfalfa (Medicago sativa), a perennial leguminous plant, is particularly valuable in this region due to its high nutritional content, impressive yield, and resilience. This forage is not only grazing-tolerant but can also be harvested multiple times over the years [1,2,3]. As a result, alfalfa serves as a crucial source of high-quality forage in the Yinshanbeilu region. However, this area faces severe water resource scarcity and significant land desertification. Irrigation relies entirely on groundwater, and after years of extraction, groundwater levels have been steadily declining, leading to an increasingly critical shortage of groundwater resources. This shortage has become a limiting factor in the development of agriculture and livestock farming in the agro-pastoral ecotone of the Yinshanbeilu region [4,5]. Additionally, due to the high water consumption characteristics of alfalfa [6,7], its production is inevitably constrained by local climate conditions and available irrigation water. Therefore, developing water-saving and yield-enhancing techniques for alfalfa cultivation has become an urgent issue that needs to be addressed in the Yinshanbeilu region. Ma et al. [8], a field experiment in Northwest China, determined that subsurface irrigation at 75% of crop evapotranspiration (ET_C) combined with a nitrogen application rate of 145–190 kg·ha⁻¹ optimizes the alfalfa yield, quality, water, and nitrogen use efficiency, as well as economic benefits. Tong et al. [9], based on a two-year field irrigation experiment, investigated the effects of different soil moisture gradients under drip irrigation on the crop coefficient (Kc) and water use efficiency (WUE). Ma et al. [10], in order to explore suitable irrigation methods for alfalfa cultivation in the arid and semi-arid regions of northwest China, designed three irrigation systems: rolling sprinkler irrigation, center pivot irrigation, and shallow-buried drip irrigation. The study concluded that shallow-buried drip irrigation was the most suitable method as it significantly improved soil moisture conditions, increased the alfalfa yield, and enhanced irrigation water use efficiency (IWUE). Zheng et al. [11], in order to determine the optimal irrigation amount and drip line depth for subsurface drip irrigation of alfalfa in arid regions, conducted a field experiment. Using the water balance equation and the FAO-56 crop coefficient method to estimate evapotranspiration, the study identified the optimal irrigation amount for the research area to be 22.5–30.0 mm, with a drip line depth of 20 cm. Wang et al. [12] conducted a split-plot experiment with two factors to study the optimal irrigation and fertilization rates for alfalfa seed production under drip irrigation in the Ningxia region. Many scholars have studied the water consumption, growth physiology, irrigation methods, and irrigation regimes of alfalfa, laying a solid foundation for alfalfa research [13,14,15]. However, most of these studies have been concentrated in agricultural and pastoral areas with flat terrain, long growing seasons, and high accumulated temperatures. In contrast, research on alfalfa in the agro-pastoral ecotone, particularly on sloping farmland with short growing seasons and low accumulated temperatures, remains relatively scarce. Moreover, many researchers have primarily focused on alfalfa yield and irrigation regimes, with less consideration given to the impacts of different hydrological years.

In summary, extensive research has been conducted on alfalfa water use efficiency (WUE) and irrigation regimes in the arid and semi-arid regions of northwestern China, including Inner Mongolia, Xinjiang, and Ningxia. However, studies focusing on the water consumption patterns, soil moisture dynamics, yield variations, and optimal irrigation predictions for alfalfa in the groundwater over-exploitation areas of the Yinshanbeilu region’s agro-pastoral ecotone remain limited. Most current research consists of in situ monitoring experiments, where optimal irrigation regimes are proposed by designing different irrigation and fertilization levels. There has been relatively little application of crop growth models, highlighting the need to develop models suited to local crop conditions and establish a comprehensive set of alfalfa genetic parameters, field management practices, and soil parameters. To address these issues, this study focused on alfalfa and conducted two years of in situ monitoring experiments. The DSSAT (Decision Support System for Agrotechnology Transfer) model was used to simulate the soil moisture, yield, and other parameters for alfalfa, revealing water use efficiency (WUE) and optimal irrigation strategies under varying hydrological years in groundwater-irrigated farmland within the agro-pastoral ecotone of the Yinshanbeilu.

2. Materials and Methods

2.1. Overview of the Experimental Area

The study area is located in Wuchuan County (41°08.344′ N, 111°17.580′ E), within the agro-pastoral ecotone of the Yinshanbeilu in Inner Mongolia. Wuchuan County has a total arable land area of 144,436.32 hm², of which 57,645.96 hm² is irrigated land (The Third National Land Survey). Agricultural water use relies on groundwater, and due to excessive extraction, the water table has been declining at a rate of 27.50 cm/a to 47.50 cm/a, with well depths increasing by 50 to 120 m over the past 20 years. The climate of the study area is classified as a temperate continental monsoon, with an average annual temperature of 2.5 °C. Annual precipitation ranges from 100 to 400 mm, while annual evaporation is between 2400 and 2800 mm. The frost-free period lasts 90 to 120 days. The soil in the area is classified as chestnut soil, with a bulk density ranging from 1.2 to 1.7 g/cm³ in the 0–100 cm soil layer. The cropping system follows a single cropping cycle per year, with major crops including potatoes, oats, and artificial forage (Figure 1).

2.2. Field Experiment Setup

In a typical groundwater-irrigated unit, one irrigation well was set up to cover a total irrigation area of 529 acres. The crop distribution included 228 mu of potatoes, 162 mu of oats, 73 mu of alfalfa (Medicago sativa), and 66 mu of sunflowers. This study focused on alfalfa to investigate its water use efficiency (WUE) and irrigation strategies. The alfalfa variety used was ‘Algonquin’. The alfalfa was planted with a row spacing of 30 cm, with a drip irrigation line shared between every two rows. The irrigation system used subsurface drip irrigation, with the drip lines buried at a depth of 15 cm. The monitoring period lasted from 1 May to 30 September, with samples taken on the 1st and 15th of each month and more frequent sampling during irrigation periods; the additional sampling date is the next day after each irrigation. Soil samples were taken to a depth of 100 cm, and alfalfa growth indicators such as plant height and leaf area were monitored. Leaf area index (LAI) data were collected through direct field measurements during key growth stages of alfalfa using a portable LAI meter (LAI-2200C, origin and manufacturer: LI-COR, Lincoln, Nebraska, USA) for all treatment plots. Additionally, the growth stages, irrigation timing, frequency, and irrigation quotas for alfalfa were recorded (Figure 2).

2.3. Observation Indicators

2.3.1. Soil Physical Properties of the Study Area

In each crop area, soil moisture monitoring equipment (9 devices are installed, with an interval of 100 m; origin and manufacturer: Oriental Intelligence Technology Co., Ltd. Zhejiang, China) was installed to track soil moisture dynamics. The sensors were buried at depths of 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 cm, with data collected every hour. Additionally, soil moisture content was measured using the drying method at each crop growth stage and before and after irrigation, and these data were used to calibrate the soil moisture monitoring equipment. Soil samples were collected from five layers (0–20 cm, 20–40 cm, 40–60 cm, 60–80 cm, and 80–100 cm), with three samples taken from each layer. The soil bulk density was measured using the ring knife method, field capacity was determined through field measurements, and saturated soil moisture was measured using the drying method. Soil pH was measured with a digital pH meter, and organic matter content was sent for laboratory analysis. These data allowed for the determination of basic soil physical parameters for each crop layer in the study area [11,12]. The study area had poor soil texture, with higher fertility in the shallow soil layers and lower fertility in the deeper layers. As the soil depth increased, the sand content rose, reducing the soil’s water retention capacity. Detailed soil physical parameters are shown in

Table 1. Basic physical and chemical parameters of alfalfa soil.

Soil Layer Depth	Bulk Density		Field Moisture Capacity		Saturated Moisture		Available P		Available K		Soil Organic Matter		pH
	(g/cm3)		(cm3/cm3)		(cm3/cm3)		(mg·kg−1)		(mg·kg−1)		(g·kg−1)
(cm)	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023
0–20	1.36	1.31	0.29	0.28	0.31	0.32	11.16	11.63	159.95	154.91	2.19	2.34	7.21	7.02
20–40	1.57	1.54	0.23	0.22	0.27	0.25	4.39	5.14	129.56	125.23	5.54	5.68	7.03	6.84
40–60	1.59	1.55	0.21	0.17	0.22	0.18	2.98	2.87	72.77	70.06	1.42	1.64	6.96	6.88
60–80	1.63	1.61	0.16	0.19	0.20	0.20	1.77	1.78	64.78	65.62	0.67	0.71	6.86	6.76
80–100	1.65	1.63	0.12	0.13	0.16	0.15	1.51	1.63	90.10	86.62	0.49	0.45	6.81	6.76

.

2.3.2. Meteorological Data Collection

Temperature and rainfall during the crop growing season are shown in Figure 3. Rainfall amounts during the growing seasons of 2022 and 2023 were 213.6 mm and 257.8 mm, respectively, with temperatures ranging from 0.4 °C to 29.8 °C in 2022 and 1.3 °C to 26.9 °C in 2023.

2.3.3. Crop Growth Stages

The growth period of alfalfa, from planting to harvesting, is divided into two harvest stages and four growth stages: greening–branching, branching–budding, budding–blooming, and blossom–harvest. The growth stages and duration for 2022 and 2023 are shown in

Table 2. Growth stages of different alfalfa.

Crop Varieties	Growth Period	2022		2023
		Start and End Date	Fertility Days	Start and End Date	Fertility Days
Alfalfa	Greening–branching	11/5~30/5	19 d	10/5~25/5	15 d
	Branching–budding	31/5~22/6	23 d	26/5~18/6	24 d
	Budding–blooming	23/6~10/7	17 d	19/6~5/7	16 d
	Blossom–harvest	11/7~24/7	13 d	6/7~21/7	15 d
	Greening–branching	25/7~10/8	16 d	22/7~4/8	14 d
	Branching–budding	11/8~28/8	17 d	5/8~24/8	19 d
	Budding–blooming	29/8~9/9	12 d	25/8~6/9	12 d
	Blossom–harvest	10/9~20/9	10 d	7/9~18/9	11 d
	Total	11/5~20/9	127 d	10/5~18/9	126 d

.

2.3.4. Crop Irrigation Regime

The irrigation quotas for alfalfa during the growing seasons are shown in

Table 3. Irrigation quotas for different alfalfa growth stages.

Crop Varieties	Growth Period	Irrigation Amount (mm)
		2022	2023
Alfalfa	Greening–branching	16	16
	Branching–budding	39	35
	Budding–blooming	19	16
	Blossom–harvest	18	15
	Greening–branching	19	17
	Branching–budding	38	33
	Budding–blooming	18	18
	Blossom–harvest	15	14.0
	Total	182	164

.

2.4. Research Methods

2.4.1. Introduction to the DSSAT Model

The Decision Support System for Agrotechnology Transfer (DSSAT) is one of the most widely used crop growth models worldwide. It includes various modules such as CROPGRO, CERES, SUBSTOR, and CANEGRO, which can simulate the growth of numerous crops. The DSSAT model simulates crop growth and development processes daily, responding to factors such as crop variety genetics, water and nitrogen stress, field management practices, and climatic conditions. It can be used to optimize crop cultivation plans, simulate the potential productivity under varying light, temperature, and precipitation conditions, and provide reasonable predictions and decisions for agricultural technology choices [16].

2.4.2. Soil Water Movement Equation

The formula used by the DSSAT model to calculate changes in soil water content [17] is as follows:

$$∆\mathrm{S}=\mathrm{P}+\mathrm{I}-\mathrm{EP}-\mathrm{ES}-\mathrm{R}-\mathrm{D}$$

(1)

where ∆S represents the change in soil water content; P is the rainfall; I is the irrigation amount; EP is the transpiration; ES is the soil evaporation; R is the surface runoff; D is the soil profile drainage.

The DSSAT model calculates water stress using the following formula [18]:

$${\mathrm{SWDF}}_1={\displaystyle \frac{{\mathrm{WS}}_{\mathrm{P}}}{\mathrm{RWUEP}\times {\mathrm{EP}}_{\mathrm{O}}}}$$

(2)

$${\mathrm{SWDF}}_2={\displaystyle \frac{{\mathrm{WS}}_{\mathrm{P}}}{{\mathrm{EP}}_{\mathrm{O}}}}$$

(3)

where SWDF₁ is the first water stress factor in the model; WS_P denotes potential root water uptake (mm); RWUEP is a species-specific characteristic parameter; EP₀ represents the crop water demand (mm); SWDF₂ is the second water stress factor in the crop model.

2.4.3. Crop Dry Matter Accumulation Equation

The DSSAT model’s equation for dry matter accumulation is [19]

$$∆\mathrm{TOT}=0.758\times (\mathrm{PARCE}\times {10}^6\times \mathrm{IPAR} - 0.004\times \mathrm{TOT})\times \mathrm{SWD}$$

(4)

where ∆TOT is the daily increment of crop dry matter (t/ha); PARCE is the photosynthetically active radiation conversion efficiency (g/MJ); IPAR is the intercepted photosynthetically active radiation (MJ/ha); TOT is the total dry matter (t/ha); SWDF is the water stress factor affecting dry matter accumulation.

The calculation formula for PARCE in the DSSAT model is

$$\mathrm{PARCE}={\mathrm{PARCE}}_{\max }\times \left\{1-\exp \left.\left[-0.008\times \left(\mathrm{T} - 8\right)\right]\right\}\right.$$

(5)

where PARCE_max is the maximum value for photosynthetically active radiation conversion efficiency (g/MJ), and T is the daily average temperature (°C).

2.4.4. Model Simulation Units

Using field experiment data, including crop growth stages, leaf area index, and yield, the DSSAT model was employed to simulate the growth, yield, and soil moisture dynamics of alfalfa. The simulation periods for 2022 and 2023 were set from 1 May to 13 September.

2.4.5. Meteorological Data

The meteorological data required for the DSSAT model include solar radiation (MJ·m⁻²), maximum temperature (°C), minimum temperature (°C), and rainfall (mm) (Figure 3). Field management data needed for the model include crop variety, sowing method, sowing date, sowing density and depth, irrigation quotas and timing, and fertilization amounts and timing. Solar radiation was derived from sunshine hours recorded by the meteorological station using the conversion formula [20]:

$${\mathrm{R}}_{\mathrm{s}}={\mathrm{R}}_{\max }\times \left({\mathrm{a}}_{\mathrm{s}}-{\mathrm{b}}_{\mathrm{s}}{\displaystyle \frac{\mathrm{n}}{\mathrm{N}}}\right)$$

(6)

where R_s is the total solar radiation (MJ/m²); a_s is the clear sky solar radiation constant; b_s is a proportionality factor, both of which are constants related to atmospheric conditions, usually taken as 0.25 and 0.5, respectively; n denotes sunshine duration (h); N is the maximum possible sunshine duration (h); R_max is the solar radiation on a clear day (MJ/m²).

$${\mathrm{R}}_{\max }=37.586\times \mathrm{d}\times \left({\mathrm{W}}_{\mathrm{s}}+\sin \varnothing \sin \mathrm{\delta }+\cos \varnothing \cos \mathrm{\delta }\sin {\mathrm{w}}_{\mathrm{s}}\right)$$

(7)

where R_max is the solar radiation on a clear day (MJ/m²), d is the distance between the earth and the sun, W_s is the sunset hour angle (°), δ is the declination of the equator (°), and $\varnothing$ is the latitude of the meteorological station (°).

$$\mathrm{N}={\displaystyle \frac{24}{\pi }}{\mathrm{W}}_{\mathrm{s}}$$

(8)

where N represents the maximum possible duration of daylight (h), and W_s is the sunset hour angle (°).

$$\mathrm{\delta }=0.4093\times \sin \left({\displaystyle \frac{2\pi }{365}}\right)\times \mathrm{J} - 1.045$$

(9)

where δ is the declination of the equator (°), and J is the Julian day, which is the day of the year, with 1 January equal to 1.

$${\mathrm{W}}_{\mathrm{s}}=\mathrm{arc}\cos \left(-\tan \varnothing \tan \mathrm{\delta }\right)$$

(10)

where W_s is the sunset hour angle (°), δ is the declination of the equator (°), and ∅ is the latitude of the meteorological station (°).

2.4.6. Soil Parameters

The soil parameters for the DSSAT model require both the physical and chemical properties of the soil. The soil profile parameters at different depths include the wilting point, field capacity, saturated water content, bulk density, soil organic carbon content, total nitrogen content, and pH value. The data are shown in Table 1.

2.4.7. Calibration and Validation

The model’s accuracy in this study was evaluated using the Absolute Relative Error (ARE) and the normalized Root Mean Square Error (nRMSE), which measure the relative difference between the DSSAT model’s simulated values and the observed values [21,22]. Generally, the closer the ARE is to 0, the higher the model’s accuracy. A nRMSE of less than 10% is considered excellent; 10–20% is considered good; 20–30% is acceptable; greater than 30% is poor. The coefficient of determination (R²) was used to assess the fit between the simulated and observed values, with an R² closer to 1 indicating a higher degree of fit. The formula for these indices is as follows:

$$\mathrm{ARE}={\displaystyle \frac{\left|{\mathrm{S}}_{\mathrm{i}}-{\mathrm{M}}_{\mathrm{i}}\right|}{{\mathrm{M}}_{\mathrm{i}}}}\times 100\%$$

(11)

$$\mathrm{RMSE}=\sqrt{{\displaystyle \frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{M}}_{\mathrm{i}}-{\mathrm{S}}_{\mathrm{i}}\right)}^2}{\mathrm{n}}}}$$

(12)

$$\mathrm{nRMSE}={\displaystyle \frac{\mathrm{RMSE}}{\mathrm{M}}}\times 100\%$$

(13)

$${\mathrm{R}}^2={\displaystyle \frac{{\left[\sum _{\mathrm{i}=1}^{\mathrm{n}}\left({\mathrm{M}}_{\mathrm{i}}-\mathrm{M}\right)\left({\mathrm{S}}_{\mathrm{i}}-\mathrm{S}\right)\right]}^2}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{M}}_{\mathrm{i}}-\mathrm{M}\right)}^2\sum _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{S}}_{\mathrm{i}}-\mathrm{S}\right)}^2}}$$

(14)

where M_i is the observed value, S_i is the simulated value, M is the average of the measured values, S is the average of the simulated values, and n is the number of samples.

2.5. Crop Irrigation Evaluation

The crop water requirement during the growth period can be calculated using the water balance equation. The calculation method is as follows [23]:

$$\mathrm{ET}=\mathrm{R}+\mathrm{I}-∆\mathrm{W}-\mathrm{P}$$

(15)

where: ET is the crop water requirement during the growth period (mm), R is the amount of rainfall (mm), I is the amount of irrigation water (mm), ∆W is the change in soil water storage before and after the crop growth period (mm), P is the groundwater recharge and deep percolation (mm). Since the groundwater table in the study area is deeper than 6 m, groundwater recharge can be neglected in this calculation.

This paper uses water use efficiency (WUE, kg/m³) and irrigation water use efficiency (IWUE, kg/m³) as two indices [24,25] to evaluate and reflect the relationship between crop yield and irrigation water use. The calculation formula is

$$\mathrm{WUE}={\displaystyle \frac{\mathrm{Y}}{\mathrm{ET}}}$$

(16)

where Y is the crop yield (kg·ha⁻¹), ET is the actual evapotranspiration during the growth stage (mm).

$$\mathrm{IWUE}={\displaystyle \frac{\mathrm{Y}-{\mathrm{Y}}_{\mathrm{a}}}{\mathrm{I}}}$$

(17)

where Y is the crop yield (kg/ha), Ya is the crop yield without irrigation (kg·ha⁻¹), and I is the amount of irrigation water supplied during the growth stage (mm).

3. Results and Analysis

3.1. Model Calibration and Validation

3.1.1. Model Calibration

This study utilized 2022 experimental data for calibration, using the Absolute Relative Error (ARE), normalized Root Mean Square Error (nRMSE), and coefficient of determination (R²). The ARE for the leaf area index (LAI) of alfalfa was 3.82%, with an nRMSE of 6.31% and an R² of 0.87. For the yield, the ARE was 4.27%, the nRMSE was 5.88%, and the R² was 0.89 (

Table 4. Evaluation of accuracy indexes for alfalfa leaf area and alfalfa yield.

Crop Type	Statistical Indicators	LAI/(cm2·cm−2)	Yield/(kg·ha−1)
Alfalfa	ARE/%	3.82	4.27
	nRMSE/%	6.31	5.88
	R2	0.87	0.89

). The precision indicators for soil moisture in alfalfa, including ARE, nRMSE, and R², were 4.38–5.57%, 6.22–8.06%, and 0.86–0.87, respectively (

Table 5. Evaluation of accuracy indexes for the soil water content.

Crop Type	Soil Depth	ARE/%	nRMSE/%	R2
Alfalfa	0–20	4.38	6.22	0.87
	20–40	4.94	5.87	0.87
	40–60	5.57	8.06	0.86

). The genetic parameter calibration for the crop is shown in

Table 6. Genetic parameters of alfalfa varieties.

Argument	Definition	Calibrated Value
CSDL	Critical short day duration (h)	10.5
PPSEN	Relative response slope to photoperiod (1/h)	0.2
EM-FL	Duration of light and heat from seedling emergence to first blossom appearance (d)	21.5
FL-SH	From the initial inflorescence blossoming to the first inflorescence fruit setting, light and heat conditions (d)	6.7
FL-SD	The light and heat time from the first inflorescence blooming to the first inflorescence grain production (d)	12.6
SD-PM	Photothermal duration from seed production to the first inflorescence’s physiological ripening (d)	33.5
FL-LF	The photothermal time between the flowering of the first inflorescence and the cessation of leaf expansion (d)	16
LFMAX	Maximum photosynthetic rate of leaves (mg CO2/m2·s−1)	2.5
SLAVR	Specific leaf area (cm2 /g)	290
SIZLF	Maximum blade size (cm2)	5

. The parameters met the required accuracy, indicating high precision in the model simulation.

3.1.2. Model Validation

After calibration using the 2022 data, the corrected genetic parameters for different crops (Table 4, Table 5 and Table 6) were input into their respective modules for validation using the 2023 measured data. The validation results are shown in Figure 4. For soil moisture, the evaluation indicators were an ARE of 4.23%, an nRMSE of 5.74%, and an R² of 0.912. For LAI, the ARE was 5.13%, the nRMSE was 4.81%, and the R² was 0.919. For the yield, the ARE was 4.49%, the nRMSE was 5.97%, and the R² was 0.930. The model’s validation results were consistent with the calibration results, indicating a high degree of accuracy in simulating the alfalfa’s soil moisture, LAI, and yield.

3.2. Dynamics of Soil Moisture, Leaf Area Index, and Yield of Alfalfa

(1)

Soil Moisture

In 2023, alfalfa was irrigated eight times, with relatively short intervals between irrigations. During the rapid growth stages (30–45 days and 92–105 days), soil moisture in the 0–20 cm layer decreased by approximately 50–52%, in the 20–40 cm layer by 46–48%, and in the 40–60 cm layer by 40–45%. During the mid-to-late growth stages (45–50 days and 105–117 days), soil moisture in the 0–20 cm layer decreased by approximately 43–46%, in the 20–40 cm layer by 40–44%, and in the 40–60 cm layer by 32–35%. The soil moisture reduction during the rapid growth stage was 7–10% greater than that during the mid-to-late growth stage (Figure 5).

(2)

Leaf Area Index

Throughout the growing season, alfalfa exhibited multiple LAI peaks. After cutting, the LAI increased rapidly, and soil moisture decreased sharply, indicating a high water demand. As alfalfa entered the flowering stage, the decline in soil moisture slowed, reflecting a cyclical pattern of water consumption and recovery (Figure 6).

(3)

Yield

As a perennial plant, alfalfa’s yield increased rapidly during the periods of 45–60 days and 105–120 days, corresponding to the downward trend in soil moisture. After each cutting, alfalfa quickly recovered, with the growth rate of the first harvest averaging 35.7 kg·ha⁻¹ per day and the second harvest averaging 64.2 kg·ha⁻¹ per day, resulting in an annual yield increase of 6529 kg·ha⁻¹ (Figure 7).

3.3. Alfalfa Water Consumption, Water Intensity, and Water Module Patterns

As shown in

Table 7. Water consumption, water consumption intensity and water consumption intensity during the growth period of alfalfa.

Crop Varieties	Growth Period	Water Consumption (mm)		Modulus Ratio Coefficient		Water Consumption Intensity (mm/d)
		2022	2023	2022	2023	2022	2023
Alfalfa	Greening–branching	78.98	90.74	19%	20%	2.15	2.91
	Branching–budding	124.71	140.65	30%	31%	2.97	3.04
	Budding–blooming	128.87	149.72	31%	33%	4.23	4.97
	Blossom–harvest	83.14	72.59	20%	16%	3.44	2.60
	Total	415.7	453.7	100%	100%	3.11	3.35

, alfalfa’s water requirement during the growing season in 2022 was 415.7 mm, while in 2023, it was 453.7 mm. The average water requirement for each growth stage accounted for the following percentages of total water consumption: 19–20% during the regrowth stage, 30–31% during the branching stage, 31–33% during the bud stage, and 16–20% during the flowering stage. The water intensity for alfalfa in 2022 and 2023 ranged from 3.11 to 3.35 mm/day (Table 7).

3.4. Soil Water Balance Estimation and Water Use Efficiency Evaluation

3.4.1. Soil Water Balance Estimation

In 2022 and 2023, the evapotranspiration for alfalfa was 415.7 mm and 453.7 mm, respectively, with total percolation amounts of 27.3 mm and 24.8 mm. Soil evaporation and alfalfa transpiration accounted for 67.88% and 32.12% of total evapotranspiration in 2022 and 57.46% and 42.54% in 2023, respectively (

Table 8. Water balance of 0–60 cm soil profile for alfalfa in 2022 and 2023.

Parameter	R/mm		I/mm		ΔW/mm		ET/mm		E/mm		T/mm		P/mm
Years	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023
Alfalfa	213.6	257.8	182	164	−47.4	−56.7	415.7	453.7	184.3	181.2	231.4	272.5	27.3	24.8

).

3.4.2. Alfalfa Water Use Efficiency Evaluation

Based on the simulation results for alfalfa yield in 2022 and 2023, water use efficiency (WUE) and irrigation water use efficiency (IWUE) were used to evaluate the relationship between yield and irrigation water. The results, shown in

Table 9. Water use efficiency and irrigation water use efficiency of alfalfa in 2022 and 2023.

Parameter	I/mm		ET/mm		Y/(kg·ha−1)		Ya/(kg·ha−1)		WUE/ (kg·m−3)		IWUE/ (kg·m−3)
Years	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023	2022	2023
Alfalfa	182	164	415.7	453.7	4879	6529	3185	5362	11.74	14.39	9.31	7.12

, indicate that WUE ranged from 11.74 to 14.39 kg·m⁻³, and IWUE ranged from 7.12 to 9.31 kg·m⁻³. Each millimeter of water consumption produced about 11.74 to 14.39 kg of alfalfa per hectare, while each millimeter of groundwater used produced about 7.12 to 9.31 kg of alfalfa. Groundwater irrigation increased alfalfa yields by 17.87% to 34.72% compared to rainfed alfalfa. There is potential for further improvement in groundwater irrigation water productivity by reducing groundwater usage while maintaining current yield levels through the adoption of deep groundwater-saving and controlled irrigation practices.

3.5. Optimal Irrigation Strategy for Alfalfa Under Different Hydrological Years

3.5.1. Classification of Different Hydrological Years

Based on a frequency analysis of rainfall data from Wuchuan from 1991 to 2023, the years were classified into wet, normal, and dry years. A wet year is defined when P ≤ 25%, a normal year when 25% < P < 85%, and a dry year when P ≥ 85%, using the following calculation formula:

$${\mathrm{P}}_{\mathrm{i}}={\displaystyle \frac{\mathrm{m}}{\mathrm{n}+1}}\times 100$$

(18)

where P_i is the rainfall frequency for a given year, m is the rank of the rainfall amount from highest to lowest, and n is the number of data years.

Using hydrological frequency calculation software, the rainfall data from Wuchuan (1991–2023) were processed with a Pearson Type III curve, and the rainfall frequency curve is shown in Figure 8. Wet years had rainfall amounts ranging from 419.4 mm to 485.2 mm, normal years from 259.4 mm to 414.4 mm, and dry years from 223.2 mm to 255.8 mm. Over the 33-year period, there were 8 wet years, 20 normal years, and 5 dry years. The wet years were 1992, 1998, 2003, 2008, 2013, 2014, 2016, and 2018; the normal years were 1991, 1993–1997, 2000, 2001, 2004–2005, 2009–2012, 2015, 2017, and 2019–2023; and the dry years were 1999, 2002, 2006, 2007, and 2022.

3.5.2. Rainfall Analysis During Alfalfa Growth Stages over Multiple Years

The average rainfall during the alfalfa growth stages over multiple years was 259.4 mm, with the peak occurring in 2004 at 414.0 mm and the lowest value in 1997 at 169.3 mm, a difference of 2.45 times. The significant variation in rainfall during the growth stages indicates that local rainfall distribution is uneven, with pronounced differences between wet and dry seasons. Based on the classification of hydrological years (1991–2023) and the analysis of rainfall during the growth stages, the rainfall values for wet, normal, and dry years were 346 mm, 248 mm, and 178 mm, respectively (Figure 9).

3.5.3. Simulation Scenario Setup for Alfalfa Irrigation Strategies in Different Hydrological Years

To achieve efficient farm irrigation management, irrigation volumes should be adjusted based on the water requirements of the crops and the precipitation in different hydrological years. According to the irrigation practices of local farmers and the “Inner Mongolia Water Use Quota” (Standard number: DB15/T 385-2020 [26]), irrigation quotas for alfalfa in normal and dry years were set at 180 mm and 240 mm, respectively. To adapt to varying water conditions, different irrigation trial combinations were formulated to ensure a sufficient water supply in the same typical hydrological years (normal and dry). For each typical hydrological year, three different irrigation volume scenarios were set (

Table 10. Simulation scenario design for different hydrological years.

Hydrological Year	Crop Varieties	Irrigation Scenario	Irrigation Water (mm)	Hydrological Year	Crop Varieties	Irrigation Scenario	Irrigation Water (mm)
normal year	Alfalfa	1	200	Dry Year	Alfalfa	1	240
		2	180			2	216
		3	160			3	192

). Scenario 1 represents the basic irrigation quota for typical crops, Scenario 2 reduces the basic irrigation quota by 10%, and Scenario 3 reduces it by 20%.

3.5.4. Optimal Irrigation Volume for Different Hydrological Years

Figure 10 shows the impact of different irrigation scenarios on crop yield. As shown in the figure, the yield of alfalfa is highest in Scenario 1 for both normal and dry years, followed by Scenario 2, with Scenario 3 yielding the lowest. The difference between Scenario 1 and Scenario 2 is minimal, while the gap between Scenario 1 and Scenario 3 is larger, indicating that the yield decreases as the irrigation volume decreases. In normal years, Scenario 1 yields 6325 kg·hm², which is approximately 3.29% and 7.64% higher than Scenarios 2 and 3, respectively. In dry years, Scenario 1 yields 6184 kg·hm², which is about 4.95% and 10.56% higher than Scenarios 2 and 3, respectively. The yields in normal years are higher than in dry years by 0.95%, 1.63%, and 1.97%, respectively, across the three scenarios.

Based on the water use efficiency (WUE) formula, the relationship between the crop yield and irrigation volume was analyzed for different hydrological years. The results (

Table 11. Water consumption and water use efficiency of alfalfa in normal years and dry years.

Crop Varieties	Hydrological Year	Simulation Scenario	I/mm	ET/mm	Y/ (kg·ha−1)	WUE/ (kg·m−3)
Alfalfa	normal year	1	200	448	6325	14.12
		2	180	428	6117	14.29
		3	160	408	5842	14.32
Alfalfa	Dry Year	1	240	418	6184	14.79
		2	216	394	5878	14.92
		3	192	370	5531	14.95

) show that in normal years, WUE decreases significantly, with Scenario 1 having a WUE value of 14.39, which is 0.17 and 0.20 higher than Scenarios 2 and 3, respectively. In dry years, WUE increases across the three scenarios, with Scenario 3 having the highest WUE of 14.95, which is 0.16 and 0.03 higher than Scenarios 1 and 2, respectively, showing a notable improvement in WUE.

Based on the irrigation schemes and water use efficiency (WUE) evaluations simulated by the DSSAT model, the appropriate irrigation volumes for typical crops under different hydrological conditions were recommended based on local production practices. For the maximum yield, 200 mm of irrigation water is recommended for normal years and 240 mm for dry years. To conserve groundwater resources, 160 mm is recommended for normal years and 192 mm for dry years to maintain crop yield and economic efficiency (Table 11).

4. Discussion

The agro-pastoral ecotone is a crucial production area for alfalfa in northern China. Optimizing the water use efficiency (WUE) and irrigation strategies for alfalfa in regions of groundwater over-extraction holds great significance for the efficient use of groundwater resources and deep water-saving controls in the agro-pastoral ecotone. This study conducted experiments on water use efficiency (WUE) and the yield of alfalfa with two harvests per year in Inner Mongolia in 2022 and 2023. The soil moisture content, leaf area index, and yield data were used to calibrate and validate the DSSAT-Forages-Alfalfa model. The results showed that the simulated values of the calibrated model closely matched the observed values, with trends consistent with actual alfalfa growth, confirming the model’s reliability in simulating alfalfa growth processes. Zhang et al. [27] indicated that the DSSAT model’s simulation of the impact of different irrigation schemes on the final alfalfa yield was highly accurate, with normalized root mean square errors (RMSE) below 0.1. Similarly, Hou et al. [28] demonstrated that the DSSAT model accurately simulated the alfalfa yield under different water and fertilizer conditions, achieving an R² of 0.90. The R² in this study was higher than that of Hou Chenli’s due to the exclusion of fertilizer effects on alfalfa growth in Hou’s research. Additionally, the findings of Kou et al. [29] on water consumption intensity during alfalfa growth stages—showing the highest water consumption during the bud stage, followed by the branching stage and the pre-branching stage—were consistent with this study. In this research, the total water consumption during alfalfa growth stages was between 415.7 and 453.7 mm, with a yield ranging from 4879 to 6529 kg·ha⁻¹ and water productivity of 11.74 to 14.39 kg·m⁻³. In comparison, the study by Cao et al. [30] conducted in Ordos, Inner Mongolia, found a total water consumption of 400 to 500 mm, a total yield of 7500 to 12,000 kg·ha⁻¹, and water productivity of 18.0 to 25.0 kg·m⁻³. Although the water consumption in their study was similar to the results of this study, the yield and water productivity were significantly higher. The study area in the agro-pastoral ecotone in the Yinshanbeilu experiences a lower accumulated temperature and shorter growing season, resulting in only two alfalfa harvests, with a total yield of 13% to 45% lower and water productivity of 6.26 to 10.61 kg·m⁻³ lower. This highlights the substantial impact of local environmental conditions on alfalfa growth. Jia et al. [31] found that a planting method using 60 cm-wide ridges with plastic mulch was the most suitable for large-scale alfalfa cultivation in semi-arid regions, achieving a water use efficiency (WUE) of 9.97 kg·m⁻³. The WUE observed in this study was lower than that of Yu’s research, likely because this study used flat planting methods and applied more irrigation and fertilizer. These findings suggest that environmental factors, the duration of growth, and planting methods all significantly affect alfalfa yield and water productivity.

Many scholars have used the DSSAT model to simulate and optimize irrigation schedules for crops [32]. In this study, the irrigation volumes for alfalfa were determined by considering the optimal soil moisture range for crop growth in the region, as well as the actual irrigation practices of local farmers, in accordance with the standards set by the “Inner Mongolia Regional Water Use Quota”. Simulations of different irrigation scenarios under varying hydrological years were conducted, revealing that crop yields responded differently to these scenarios. The results indicated that alfalfa yields during dry years were generally lower than during normal years, with the largest yield reduction observed under Scenario 3 in normal years, which contrasts with the findings of Shao et al. [33]. This discrepancy is primarily due to the region’s tendency for water logging during periods of heavy rainfall, which leads to reduced yields. Using crop water use efficiency (WUE) to estimate yields across the three scenarios, the study found that different crops displayed varying levels of adaptability to irrigation water volumes and water use efficiency (WUE) across different hydrological years. Reductions in irrigation water and rainfall had a significant impact on crop yields [34]. Li et al. [35] found that as irrigation volumes decreased, both alfalfa ETc (crop evapotranspiration) and the forage yield declined, while WUE and crude protein (CP) content increased. This study’s simulation results similarly showed that WUE increased as irrigation volumes decreased, aligning with the trends observed in previous studies [36]. Simulation predictions indicated that, with the objective of maximizing yield, the recommended irrigation volume in a normal year is 200 mm, while in a dry year, it is 240 mm. If the goal is to maximize groundwater resource efficiency, the recommended irrigation volumes are 160 mm for normal years and 192 mm for dry years.

This study used the DSSAT model, calibrated and validated with in situ alfalfa observation data from 2022 and 2023, to investigate the WUE and optimal irrigation strategies for alfalfa in groundwater-irrigated farmlands of the Yinshanbeilu agro-pastoral ecotone. The research aimed to address the rational use of groundwater in over-extraction zones and improve water resource utilization efficiency by proposing optimal irrigation strategies for different hydrological years. This provides a theoretical basis for deep water-saving controls on groundwater resources. However, this study did not consider the effects of fertilization on alfalfa growth. Future research will incorporate fertilization factors into the analysis.

5. Conclusions

This study utilized the DSSAT model, calibrated and validated with in situ alfalfa field observations from 2022 and 2023, to explore the water use efficiency (WUE) and optimal irrigation strategies for alfalfa in groundwater-irrigated farmlands of the Yinshanbeilu agro-pastoral ecotone. The main conclusions are as follows:

(1)

The accuracy of the parameters for soil water content, leaf area index (LAI), and yield, measured by ARE, nRMSE, and R², met the required standards. The ranges for these parameters were as follows: for the soil water content, ARE: 4.23–5.57%, nRMSE: 5.74–8.06%, R²: 0.86–0.91; for LAI, ARE: 3.82–5.13%, nRMSE: 4.81–6.31%, R²: 0.87–0.92; for the yield, ARE: 4.27–4.49%, nRMSE: 5.88–5.97%, R²: 0.89–0.93.

(2)

The water consumption of alfalfa was 415.7 mm to 453.7 mm, with infiltration volumes ranging from 24.8 mm to 27.3 mm. The branching stage and budding stage accounted for 30–31% and 31–33% of total water consumption, respectively. The water consumption intensity ranged from 2.97 to 3.04 mm/day during the branching stage to 4.23 to 4.97 mm/day during the budding stage. Water use efficiency (WUE) varied from 11.74 to 14.39 kg·m⁻³, and irrigation water use efficiency (IWUE) ranged from 7.12 to 9.31 kg·m⁻³. The efficiency of water use was relatively high, with irrigation contributing 49.48% to 64.70% of water productivity. Groundwater irrigation increased alfalfa yields by 17.87% and 34.72% compared to rain-fed conditions.

(3)

To achieve the maximum yield, the recommended irrigation volume for alfalfa is 200 mm during normal years and 240 mm during dry years. If maximizing groundwater resource utilization is the goal, the recommended irrigation volumes are 160 mm for normal years and 192 mm for dry years.