Introduction

Agriculture and animal husbandry lay the foundation of national stability by ensuring food security. China's annual demand for high-quality forage reaches 120 million tons, and more than 2 million tons of high-quality forage such as alfalfa and oat grass are imported annually. The annual gap is more than 50 million tons, and the market demand is huge1. The irrigation area along the Yellow River in Inner Mongolia is not only the main grain producing area, but also an important forage base. Forage oats belong to gramineae, with high nutritional quality, salt tolerance, drought resistance, cold tolerance, wide adaptability and high yield2,3. In order to alleviate the shortage of forage grass, the oat planting pattern in this area was changed from one season a year to two seasons a year. The system change has led to the unclear growth process and water consumption law of oats, the contradiction between supply and demand of regional water resources has intensified, the utilization efficiency of farmland irrigation water has declined, and the bottleneck of water resources has threatened the sustainable development of agriculture and animal husbandry4,5. Therefore, exploring the growth and development and water consumption process of drip-irrigated oats in the Yellow River irrigation area of Inner Mongolia is the core of scientific planning and management of irrigation systems and the key link to achieve water conservation6. With the iteration of agricultural water-saving technology4,5,6,7,8 Hydrological models9,10 have shown irreplaceable value in the study of crop water consumption and the impact of water resources management11.

Extensive experimental and numerical studies have been conducted to understand water consumption in irrigation regions12,13,14. However, experimental approaches are often costly, time-intensive, and constrained by site-specific outcomes. Models such as AquaCrop15,16,17 HYDRUS18,19,20,21 Soil and Water Assessment Tool (SWAT)22 and Soil-Water-Atmosphere-Plant (SWAP)23 are effective in analyzing soil water migration, transformation, and consumption processes within these systems. For instance, Liang et al.24 employed the HYDRUS-1D model to simulate water and salt migration in wasteland soils, demonstrating that intense evaporation drives this migration, thereby validating the model's accuracy in simulating vertical water and salt movement. Ruimin et al.22 utilized the SWAT-DynamicParam model to effectively characterize time-varying parameters, emphasizing the importance of capturing dynamic hydrological processes. Similarly, Yuan et al.23 applied the SWAP model to simulate the water and salt balance in sunflower cultivation in the Hetao region, highlighting the frequent interactions between soil water and groundwater. Their findings demonstrated that sunflowers can efficiently utilize groundwater to meet their growth demands. Post-harvest irrigation to manage soil salinity is essential for ensuring the normal growth of sunflowers in subsequent years.

While hydrological models have been successfully applied to simulate comprehensive water consumption processes, crop-specific parameters—such as canopy cover and root depth—are often oversimplified or treated as static, failing to capture their continuous and dynamic changes25. To address these limitations, researchers have developed coupled models that integrate hydrological and crop growth processes for more accurate simulations of water consumption in irrigation regions. For example, Hao et al.26 developed a distributed model based on the one-dimensional agricultural hydrological HYDRUS-EPIC framework, which analyzed soil water, salt, and crop growth under existing irrigation practices. Their findings highlighted high soil salinity as a major constraint on crop yields. Wang et al.27 created a distributed water transformation model that couples irrigation and drainage processes with soil water movement and crop growth, enabling a detailed quantitative characterization of the supply-consumption-drainage dynamics within irrigation systems. Similarly, Wang et al.28 employed a HYDRUS-1D and EPIC coupling model to simulate soil water and salt dynamics as well as crop growth for winter wheat at the Fengqiu National Key Agricultural Ecological Experimental Station. Their study demonstrated that salt stress reduced evapotranspiration, primarily by diminishing crop transpiration, and that saline-alkali stress adversely impacted grain yield.

Although these coupled models have improved the understanding of water consumption processes in agricultural systems, their added complexity can present challenges. To address these issues, the AquaCrop model was developed as a water-driven crop hydrological model that simulates crop productivity under both soil moisture and rain-fed conditions using meteorological data29. WANG et al.30 AquaCrop was used to analyze the yield change and soil water balance of summer maize in the piedmont of North China Plain, and the growth and water consumption process of maize were characterized. Zhu et al.31 utilized AquaCrop to study the effects of soil water and salinity on winter wheat growth under saline-fresh irrigation schemes, providing optimized irrigation strategies. Their results validated AquaCrop's capability to effectively simulate soil water and salinity dynamics, biomass, and grain yield for winter wheat. M et al.32 further demonstrated AquaCrop's utility in predicting the impacts of soil, water, nutrients, and climate on biomass, yield, and water use efficiency across various crops. While elucidating the water consumption process in the agricultural environment, they emphasized its effectiveness in simulating rice growth and yield under different water conditions. Although these advances have been made, there are few studies on the water consumption process of forage oats under double cropping patterns and shallow drip irrigation techniques in arid and semi-arid areas, which fills the gap of existing research, and also verifies the applicability of AquaCrop under double cropping patterns and shallow drip irrigation techniques in this area.

In this study, the AquaCrop model was used to simulate and verify the water consumption process of forage oats in arid and semi-arid areas under the double cropping mode and shallow buried drip irrigation technology. By calibrating and validating the model, the growth and water consumption processes of forage oats were simulated, and the water balance and irrigation water consumption processes were evaluated. The study solved the challenges brought by the practice of shifting from single-season planting to double-season planting and limited irrigation, and provided important insights for improving the irrigation water use efficiency of the forage production system and optimizing the allocation of agricultural water resources in the Yellow River irrigation area of Inner Mongolia.

Materials and methods

Overview of the study area

The Yellow River Basin in Inner Mongolia, located at the northernmost section of the Yellow River (37°35′–41°50′N, 106°10′–112°50′E), has a total conventional water resource availability of 5.145 billion m³, with per capita water resources estimated at 856 m³, leading to a rigid water deficit of 1.816 billion m³. The cultivated land area spans 46.18 million acre, accounting for 27.31% of the total cultivated land in the region, with approximately 9 million acre classified as saline-alkali land, representing about 23% of the cultivated area. Annual precipitation ranges from 250 to 300 mm, with primary crops including corn, wheat, oats, alfalfa, and sunflowers. The study area is situated in the Tumd Left Banner of Hohhot (40°38′–40°44′N, 111°16′–111°24′E), characterized by gentle terrain and a continental monsoon climate. The total area of the study site is 2,560 acre, of which oats occupy 472 acre, with an average annual precipitation of 318.9 mm. Most precipitation occurs during the summer months from June to August, with 257.8 mm (80.91% of annual rainfall) falling during the oat growth period (April 6 to October 8 or October 14). The average temperature during this period is 18.93 °C, with a maximum daily temperature of 28.7 °C occurring in June and July, and a minimum daily temperature of 2.6 °C recorded in April. The average annual evaporation is 1,870.3 mm, with 2,876.5 h of sunshine and total annual solar radiation of 133.82 kcal.cm−2. The frost-free period is relatively short, lasting about 133 days. The soil in the top 0-40 cm is primarily sandy loam, while soil below 40 cm is predominantly sandy. Oats are irrigated using a subsurface drip irrigation method. The location of the test site and groundwater observation well is shown in Fig. 1. (The satellite image is from https://www.91weitu.com/default.htm).

Fig. 1
figure 1

Test site and groundwater observation well location.

Full size image

Experimental design and data acquisition

Field experiments were conducted in the demonstration area of Tumd Left Banner, Hohhot, Inner Mongolia, from April to October in 2022 and 2023. The planting area of oats is 40 hm2 and the irrigation method is shallow drip irrigation. The first crop variety used was Baylor 2, followed by Meida for the second crop. The subsurface drip irrigation system employed had a belt spacing of 45 cm, with oat row spacing of 12 cm and plant spacing ranging from 3 cm to 5 cm. Each drip irrigation belt managed an average of four oat rows, with a subsurface installation depth of 8 cm. For AquaCrop model management, field conditions indicated no runoff, a weed coverage rate of 5%, and no plastic film mulching. The experimental data collected in 2022 and 2023 were used as the simulated values of AquaCrop model, and the soil water content, canopy coverage and yield of the model were calibrated and verified by trial and error method. Five-point sampling method was used for sampling detection, and each point was sampled three times. The soil water content, canopy coverage and other indicators were monitored once a week, and the yield was measured at harvest. Soil water content was measured by drying method, and crop height was measured by tape method. The yield was evaluated by collecting 1 m2 samples from each plot, drying at 105 °C for 30 min, then drying to constant weight at 75 °C, and then multiplying the dry weight by the planting density to calculate the yield. Canopy coverage was determined by leaf area index, and the extinction coefficient was 0.7136,37. Before the experiment, the soil texture of each monitoring point was determined by Wilkes method soil particle grading (Table 1). The depths of sampling layers were 0–10 cm, 10–20 cm, 20–40 cm, 40–60 cm, 60–80 cm, 80–100 cm and 100–120 cm, respectively. A groundwater level observation well was established in the test field and monitored once a week. Meteorological data include rainfall, radiation, temperature, wind speed, humidity, etc. The data is from the automatic weather station in the experimental area (HOBO-U30, Onset Computer Corp., Bourne, MA, USA). The growth process of oats is shown in Fig. 2.

Fig. 2
figure 2

Sketch of oat growth process.

Full size image

(1) Soil physical properties in the study area.

Soil samples were collected from seven layers in the experimental site. The sampling depths were 0–10 cm, 10–20 cm, 20–40 cm, 40–60 cm, 60–80 cm, 80–100 cm, and 100–120 cm respectively. Three samples were taken from each layer. The undisturbed soil samples were soaked and saturated by Wilcox method35 and then placed on the air dried soil. The gravity water in the soil samples was discharged under the action of the suction of the air dried soil. After a period of time, the soil field capacity, saturated water content and soil bulk density could be determined, and the soil particle size gradation at each point was determined by dry particle size analyzer. The sand, silt, and clay content for each soil layer at each sampling point is summarized in Table 1.

Table 1 Soil physical characteristics of 0 ~ 120 cm.
Full size table

(2) Crop growth period.

Oat forage is cultivated in two seasons annually. In 2022, the first season was planted on April 6 and harvested on July 9. The second season was planted on July 22 and harvested on October 8. For 2023, the first crop was planted on April 5 and harvested on July 8, followed by the second crop, which was planted on July 25 and harvested on October 14.The statistics of oat growth period are shown in Table 2.

Table 2 The crop growth period table of two crops of Oats in 2022 and 2023.
Full size table

(3) Irrigation practice.

According to the local actual situation, the irrigation times, irrigation time and irrigation amount in 2022 and 2023 were counted. When the soil water content reached 50% of the total soil available water, irrigation was carried out. In 2022, the first crop only received three irrigations, with irrigation quotas of 25 mm, 30 mm and 30 mm, respectively, and the total irrigation quota was 85 mm. In 2022, the second crop will be irrigated 6 times, each irrigation quota is 30 mm, and a total of 180 mm will be irrigated. In 2023, the first crop was irrigated seven times, and the irrigation quota was 30, 30, 40, 40, 40, 40 mm, a total of 260 mm. In 2023, the second crop also received seven irrigations, with quotas of 15, 30, 30, 30 and 30 mm, respectively, for a total of 195 mm. Therefore, the total irrigation amount of the two crops was 230 mm in 2022 and 455 mm in 2023.The irrigation schedule for 2022 and 2023 is shown in Table 3.

Table 3 Irrigation schedule in 2022 and 2023.
Full size table

(4) Groundwater depth.

The test site primarily contains deep groundwater, with an average annual depth ranging from 25 to 29 m. The average groundwater depth during the growth periods in 2022 was 27.9 and 27.6 m, while in 2023 it was 25.7 and 27.5 m. The months of June, July, and AuEquation 1gust represent peak irrigation periods, during which groundwater levels exhibit significant fluctuations. After crop harvest, water consumption decreases, leading to a subsequent rise in groundwater levels. The data indicate that, with the implementation of water-saving measures in recent years, groundwater depth has gradually increased. Given that the groundwater depth exceeds 3 m, its impact on this study is minimal36. The data come from national monitoring wells. The change of groundwater depth is shown in Fig. 3.

Fig. 3
figure 3

Groundwater level changes in 2022 and 2023.

Full size image

Aquacrop model simulation

(1) The basic principle of AquaCrop model.

The AquaCrop model is a universal crop growth simulation tool developed by the Food and Agriculture Organization (FAO) to reflect the response of crop biomass and yield to water availability. It evaluates food security and assesses the environmental and management impacts on crop production. The model strikes an effective balance between simplicity, accuracy, and stability, utilizing fewer parameters, and is widely employed in designing deficit irrigation strategies. The theoretical foundation of the model is based on the yield-water response relationship expressed as37, according to Eq. 1.

$$\:\left(1-\frac{\text{Y}}{{\text{Y}}_{\text{x}}}\right)\text{=}{\text{K}}_{\text{y}}\left(1-\frac{\text{E}\text{T}}{{\text{E}\text{T}}_{\text{x}}}\right)\text{}\text{}\text{}$$
(1)

Yx and Y are the maximum and actual yield, kg × hm−2; Ky the proportionality factor between relative yield decline and relative reduction in evapotranspiration; ETx and ET the maximum and actual evapotranspiration. Through further refinement, the equations for calculating crop transpiration, aboveground biomass, and yield are as follows38, according to Eqs. 2–4.

$$\:{{\rm{T}}_{\rm{r}}}{\rm{ = }}{{\rm{K}}_{\rm{s}}}{\rm{ \times }}\left( {{{\rm{K}}_{{\rm{CTr,x}}}}{\rm{ \times C}}{{\rm{C}}^{\rm{*}}}} \right){\rm{ \times E}}{{\rm{T}}_o}$$
(2)
$$\:{\rm{B = 100 \times }}{{\rm{K}}_{{\rm{sb}}}}{\rm{ \times W}}{{\rm{P}}^{\rm{*}}}{\rm{ \times }}\sum \: \frac{{{{\rm{T}}_{\rm{r}}}}}{{{\rm{E}}{{\rm{T}}_{\bf{o}}}}}$$
(3)
$$Y = B \times HI$$
(4)

Tr represents crop transpiration, mm; Ks represents soil water stress coefficient, %; KCTr, x represents all the different factors that distinguish the actual crop from the reference crop, %; CC\(\:\text{*}\) represents the calibrated canopy coverage, %; B denotes biomass, t × hm−2; Ksb represents the low temperature stress coefficient, %; WP\(\:\text{*}\) denotes standardized water production efficiency, kg × m−2; HI represents crop harvest index, %; Y represents the final crop yield, t × hm−2.

REW (Readily Evaporable Water) is derived from the field capacity (FC) of the soil surface layer and the soil water content of the permanent wilting point (PWP) (5) :

$$\:0\le\:\text{R}\text{E}\text{W}=10\left(\text{F}\text{C}-\frac{\text{P}\text{W}\text{P}}{2}\right){\text{Z}}_{\text{e},\text{s}\text{u}\text{r}\text{f}}\le\:$$
(5)

FC volume water content at field capacity vol%; PWP volume water content at permanent wilting point vol%; Ze, surf thickness of the evaporating soil surface layer in direct contact with the atmosphere 0.040 m.

Total Available Soil Water (TAW) is the amount of water that can be contained in the fine soil portion of the considered soil volume between field capacity and permanent wilting point (6):

$$\:\text{T}\text{A}\text{W}=1000\left({{\uptheta\:}}_{\text{F}\text{C}}-{{\uptheta\:}}_{\text{W}\text{P}}\right)\text{Z}\left(1-\frac{{\text{V}\text{o}\text{l}\text{\%}}_{\text{g}\text{r}\text{a}\text{v}\text{e}\text{l}}}{100}\right)$$
(6)

TAW total available soil water in the considered soil volume, mm;\(\:{{\uptheta\:}}_{\text{F}\text{C}}\) volumetric water content at field capacity, m3 × m−3; \(\:{{\uptheta\:}}_{\text{W}\text{P}}\) water content at permanent wilting point, m3 × m−3; Z depth of the considered soil volume, m; Vol%gravel volume percentage of the gravel fraction in the soil volume.

(2) Model simulation unit division.

The simulated profile depth of the oat field was set to 120 cm. According to the measured soil texture (Table 1), the profile was divided into 7 layers, and observation points were set at 10,20,40,60,80,100 and 120 cm, respectively. The simulation timeline for the oat field is as follows: April 6 to July 9 for the first crop of 2022, July 22 to October 8 for the second crop, April 5 to July 8 for the first crop of 2023, and July 25 to October 14 for the second crop. The simulation durations for these periods are 95 days, 79, 95and 82 days, respectively. The initial time step for the model is set to one day.

(3) Weather variables.

The meteorological data required for the study were from an automatic weather station located in the experimental area (HOBO-U30, Onset Computer Corp., Bourne, MA, USA). Figure 4 (2022a and 2023b) illustrate the temperature and rainfall statistics for 2022 and 2023. The annual rainfall for 2022 was 316.9 mm, while in 2023 it was 320.81 mm. Temperature ranges during these years varied from −20.9 to 37 °C in 2022 and −27 to 34.41 °C in 2023. The average temperatures for 2022 and 2023 were −14.9 to 28.4 °C and −21.5 to 28.4 °C, respectively. During the growth periods (April 6 to October 8 in 2022 and April 5 to October 14 in 2023), rainfall totaled 298.8 and 194.9 mm, 94.29 and 60.75% of the annual rainfall, respectively. The average temperatures during these growth periods were 19.19 °C for 2022 and 18.66 °C for 2023. The daily average maximum temperatures were 29.4 and 28.0 °C, observed in June and July, while the daily average minimum temperatures were 3.7 and 2.6 °C, recorded in April. Crop evapotranspiration (ETo) was calculated using the Penman-Monteith equation.

Fig. 4
figure 4

Chart of meteorological changes in 2022 and 2023.

Full size image

(4) Crop parameters.

Crop parameter files are mainly composed of crop development, ET, crop production and soil water stress, soil salinity stress, air temperature stress and so on. The crop growth parameters such as Maximum canopy coverage (CCx), number of plants per hectare, Initial canopy coverage (CCo) and Maximum effective root depth (Zmax) were obtained by field observation. The base temperature and upper temperature limit are recommended for C3 crops by the model. Reference Harvest Index (HI0), Canopy growth coefficient (CGC), Canopy attenuation coefficient (CDC), Standard water productivity (WP*), Crop coefficient when canopy is complete but prior to senescence (KcTr, x) and Water stress parameters (Pexp, upper, Pexp, lower, Psto, upper, Psen, upper) were determined based on the reference parameters provided by the model, and the’trial and error method’31 was used to correct them.

Water Productivity(WP).

In this paper, Water Productivity (WP) were calculated by using oat yield in 2022 and 202343.

The water productivity (WP) of each irrigation system was calculated using Eq. 7.

$$\:\text{W}\text{P}=\frac{\text{Y}}{\text{E}\text{T}}$$
(7)

WP is water productivity kg × m−3; Y is yield kg × da−1; and ET is Water requirement during growth period mm.

Model calibration verification.

In this study, soil moisture content, leaf area index, and yield data from 2022 were employed to calibrate the model parameters, while data from 2023 were used to verify the simulation accuracy. The model calibration and validation processes utilized metrics such as root mean square error (RMSE), model performance index (EF), coefficient of determination (R240 The flow chart of calibration and verification is shown in Fig. 5 according to Eqs. 8–10.

$$\:\text{RMSE=}\sqrt{\frac{{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{M}}_{\text{i}}\text{-}{\text{S}}_{\text{i}}\right)}^{\text{2}}}{\text{n}}}\text{}\text{}$$
(8)
$$\:\text{EF=}\frac{{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{M}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{M}}\right)}^{\text{2}}\text{-}{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{M}}_{\text{i}}\text{-}{\text{S}}_{\text{i}}\right)}^{\text{2}}}{{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{M}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{M}}\right)}^{\text{2}}}\text{}\text{}\text{}$$
(9)
$$\:{\text{R}}^{\text{2}}\text{=}{\left[\frac{{\sum\:}_{\text{i}\text{=1}}^{\text{n}}\left({\text{M}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{M}}\right)\left({\text{S}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{S}}\right)}{{\left[{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{M}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{M}}\right)}^{\text{2}}\right]}^{\text{0.5}}{\left[{\sum\:}_{\text{i}\text{=1}}^{\text{n}}{\left({\text{S}}_{\text{i}}\text{-}\stackrel{\text{-}}{\text{S}}\right)}^{\text{2}}\right]}^{\text{0.5}}}\right]}^{\text{2}}\text{}\text{}\text{}\text{}\text{}\text{}\text{}$$
(10)

n is the number of samples; Mi and Si are the measured values and simulated values of the parameters, respectively; M is the measured average value; RMSE is used to describe the error of the model estimation, and the smaller the value, the better; EF and d are used to describe the relative error of the model, and the value is 0~1. The closer the simulated value is to 1, the smaller the deviation between the simulated and measured values, indicating a higher degree of accuracy in the model simulation.

Fig. 5
figure 5

Calibration, verification flow chart.

Full size image

Results

Model calibration

The model was calibrated using measured soil moisture data from 2022. The simulated soil moisture values aligned closely with the observed values, effectively reflecting the dynamic changes in soil moisture. As shown in Table 4, the calibration accuracy for soil water content across the two oat seasons was RMSE ≤ 4, with 0.7 ≤ EF ≤ 0.9 and R2 ≥ 0.85. These findings confirm that the simulated soil moisture values were consistent with the measured data. Research demonstrates that the AquaCrop model accurately captures the fundamental changes in soil moisture dynamics. 

The model calibration also included canopy coverage and yield data from 2022. The results indicated that the simulated values for both canopy coverage and yield were in good agreement with the measured data, effectively representing the dynamic changes in crop indices. According to Table 5, the calibration accuracy for canopy coverage (CCw) and yield for oats across the two seasons yielded RMSE ≤ 9, 0.85 ≤ EF ≤ 0.95,0.95 ≥ R2 ≥ 0.85.

In this experiment, the measured data of field experiment in 2022 was used to calibrate the model parameters, and the measured data of field experiment in 2023 was used to verify the model. In this study, two brands of Oats were selected, of which Baylor 2 Oats had a longer growth period of 90 ~ 100 days and a higher yield of 3500 ~ 6000 kg·hm⁻². Because forage Oats are planted twice a year, the remaining planting period is short, so the second crop is selected by meida. The growth period is generally 75–85 days, and the yield is relatively stable at 4500–5500 kg·hm⁻². Due to different varieties and different genes, the growth period and physiological performance are different, so the crop parameters are also relatively changed. Most of the parameter values are similar, but there are also different parameter values (such as canopy growth coefficient, maximum effective root depth, etc.) that affect the results. The main crop parameters of aquacrop model41,42 are shown in Tables 6 and Table 7.

Table 4 Evaluation of accuracy indicators for soil water content..
Full size table
Table 5 Evaluation of accuracy indicators for CCw and yield of Oats.
Full size table
Table 6 AquaCrop model first crop Oats (Baylor 2) parameter list.
Full size table
Table 7 AquaCrop model second crop Oats (Meida) parameter list.
Full size table

Model validation

The model was validated using the measured soil moisture, canopy coverage, and yield data from 2023, with all parameters set to the values obtained during model calibration. The verification results are illustrated in Fig. 6. The evaluation indices for soil moisture are as follows: RMSE was 3.17, EF was 0.76, and R2 was 0.88 for the first season; for the second season, RMSE was 2.47, EF was 0.83, and R2 was 0.85. The evaluation indices for canopy coverage were: RMSE of 9.54, EF of 0.85, and R2 of 0.92 for the first season; for the second season, RMSE was 9.03, EF was 0.93, and R2 was 0.94. For yield, the evaluation indices were as follows: RMSE of 0.30, EF of 0.90, and R2 of 0.94 for the first season; and for the second season, RMSE was 0.28, EF was 0.94, and R2 was 0.91. As shown in Fig. 5, the accuracy of the parameters meets the requirements and closely aligns with the calibration results. These findings indicate that the AquaCrop model effectively captures the fundamental dynamics of crop growth and development.

Fig. 6
figure 6

Verification diagram of soil moisture content, canopy cover, and yield for the year 2023. Note: Blue dots denote the scatter plot comparing simulated versus observed values, while the red line represents the fitted regression line.

Full size image

Soil moisture dynamic analysis process

The dynamic changes of soil moisture in oats were analyzed by using the simulation results of AquaCrop model in 2022 and 2023. The detailed soil moisture changes are shown in Fig. 7. In 2022, the soil moisture dynamics during the first crop growth period showed that the irrigation amount was small and the rainfall was large, accounting for 78.3% of the annual precipitation. On the 40 th day of growth, the maximum daily rainfall reached 76.9 mm, and the soil water content reached the peak of 38.5, 36.2, 33.2, 28.7, 26.4% and 23.9% in 10, 20, 40, 60, 80 and 120 cm soil layers, respectively. Due to frequent rainfall, surface soil moisture exhibited considerable fluctuations, while deeper soil layers retained moisture more effectively, resulting in slower consumption and minimal fluctuation. During the growth period of the second crop, rainfall was sparse, totaling only 19.7 mm throughout the entire growth period. Consequently, fluctuations in surface soil moisture were minimal, and the moisture content in deeper layers remained relatively stable. Notably, soil moisture decreased rapidly from the 40th to the 60th day during the second crop growth period.

The changes in soil moisture for oats in 2023 are presented in Fig. 8. According to local irrigation observations, the test site was irrigated seven times in 2023, with the last irrigation occurring on July 10. Throughout the growth period in 2023, rainfall accounted for 60.75% of the total annual precipitation, with the first and second stubbles contributing 22.44 and 38.31%, respectively. Consequently, the soil moisture of the second stubble exhibited greater variability compared to the first stubble. The average soil water content for each layer in the second crop was 0.24, 0.25, 0.22, and 0.21 cm3˖cm−3, respectively. Before and after the first irrigation, the increases in soil moisture content for the 10, 20, 40, 60, 80 and 100 cm layers were 87.82, 89.58, 69.47, 47.85, 31.87 and 17.56%, respectively. In 2022, changes in soil moisture for oats were primarily influenced by rainfall, while in 2023, irrigation had a more significant impact, resulting in notable differences in soil moisture dynamics between the two years.

Fig. 7
figure 7

The change process of soil moisture content in the experimental site in 2022.

Full size image
Fig. 8
figure 8

The change process of soil moisture content in the experimental site in 2023.

Full size image

The change process of crop physiological index

(1) Canopy coverage effect.

Figure 9 presents the canopy coverage index for oats in 2023. The canopy coverage index exhibited significant variation corresponding to the growth stages of the crops, with growth patterns in both years following a similar trend of gradual increase, culminating in a single peak. In 2022, the first crop experienced a rapid increase in canopy coverage from the 15th to the 45th day. However, due to low soil moisture from the 45th to the 71 st day, the increase slowed. Following irrigation on the 71 st day, the growth rate of the canopy coverage index accelerated once more. The second crop in 2022 also grew rapidly from the 30th to the 49th day, but then growth decelerated after the 49th day, ultimately achieving a canopy coverage index of 81% by the end of the growth period. In 2023, the growth curve for the first crop's canopy coverage index was relatively smooth, with fewer fluctuations. The second crop experienced a rapid increase in the canopy coverage index from the 20th to the 45th day, followed by a slowdown in growth thereafter. Notably, when the canopy coverage index increased rapidly, there was a significant decrease in soil moisture content. Conversely, when the canopy coverage index increased slowly, the decrease in soil moisture was less pronounced, suggesting that a more rapid increase in the canopy coverage index correlates with greater water demand.

The maximum canopy coverage index for the two crops in 2022 was 85 and 79%, respectively. This difference can be attributed to the concentrated rainfall during the growth period of the first crop, resulting in a higher canopy coverage index compared to the second crop. In 2023, the maximum canopy coverage index for the two crops was 81 and 82%, respectively, indicating minimal difference due to similar rainfall and irrigation quotas for both crops.

Fig. 9
figure 9

Comparison of simulated and measured values of canopy coverage index of experimental land in 2022 and 2023.

Full size image

(2) Characterization of yield change of oats in two stubbles per year.

Oat yields increased progressively with plant growth. In 2022, the rainfall was predominantly concentrated during the growth period of the first crop, which, combined with a longer growth duration, resulted in a yield of 808 kg·hm2 higher than that of the second crop. In 2023, rainfall was more evenly distributed during the growth period of the second crop, leading to a yield increase of 129 kg·hm2 compared to the first crop. The average yield growth rate throughout the entire growth period was approximately 50.89 to 60.54 kg·hm2 per day. Notably, the growth rate for the second crop in both 2022 and 2023 was faster than that of the first crop, with rates ranging from 0.71 to 9.64 kg·hm2 respectively. The evaluation table of canopy coverage and oat yield accuracy index is shown in Table 8.

Table 8 Canopy coverage and yield index of forage Oats in 2022 and 2023.
Full size table

Water balance analysis

Based on the principle of water balance, the water balance of soil moisture during the growth period was estimated by using the data of 2022 and 2023. The results are presented in Tables 9 and 10. In 2022, the evapotranspiration for the first and second crops of oats was 309 and 214.7 mm, respectively, with total drainage recorded at 10.6 and 8.9 mm. In 2023, evapotranspiration for the first and second crops was 315.4 and 304.2 mm, respectively, with total drainage at 1.2 and 2.6 mm. Overall water consumption during the growth period accounted for 93.39 to 95.69% of the total introduced water (rainfall plus irrigation), while drainage accounted for 0.36 to 5.40% of the introduced water. In 2022, soil evaporation for the first and second stubbles constituted 27.14 and 25.30% of total evapotranspiration, respectively, while plant transpiration represented 72.86 and 74.70%, respectively. In 2023, soil evaporation for the first and second crops accounted for 18.58 and 23.54% of total evapotranspiration, corresponding to 17.65 to 22.52% of the introduced water. Plant transpiration accounted for 81.42 and 76.46% of total evapotranspiration, comprising 73.17 to 77.34% of the introduced water. Because the canopy coverage of the second crops in 2022 and 2023 was greater than that of the first crops, transpiration for the second crops was higher. Total evapotranspiration for both crops in 2023 exceeded that of both crops in 2022 by 95.5 mm. The amount of drainage during the growth periods in 2022 and 2023 ranged from 1.2 to 10.6 mm, representing 0.36 to 5.40% of the total introduced water. Notably, the total drainage for both crops in 2023 was 15.7 mm less than that for both crops in 2022.

Table 9 Water balance of soil profile of different crops in 2022.
Full size table
Table 10 Water balance of soil profile of different crops in 2023.
Full size table

Analysis of water productivity

Based on the simulation results of oat yield in 2022 and 2023, the relationship between crop yield and water productivity (WP) was analyzed by using WP. The calculation results are summarized in Tables 11 and 12. The WP for the first and second crops of oats in 2022 were were 1.61and 2.01 kg × m−3, respectively, while in 2023, these values were 1.53 and 1.63 kg × m−3, respectively. The WP for the first and second stubbles decreased by 0.08 and 0.38 kg × m−3, respectively. This decrease may be attributed to higher evapotranspiration in 2023 compared to 2022, resulting in a slight reduction in unit water output.

Table 11 Water productivity of oats in 2022.
Full size table
Table 12 Water productivity of oats in 2023.
Full size table

Discussion

In this study, field experiments and numerical simulations were conducted in the demonstration area of Tumed Left Banner, Hohhot, Inner Mongolia, from 2022 to 2023. A comprehensive database encompassing meteorological, soil, crop, and irrigation factors was established, and the AquaCrop model was utilized for simulation research. The model underwent calibration and validation using data from 2022 to 2023, including soil water content, canopy coverage, and crop yield. After calibration and validation, the simulated values generated by the AquaCrop model exhibited good agreement with the observed values, effectively simulating the growth and water consumption processes of forage oat under the double-season planting mode.

Zhou Qilong et al.43 reported that the hay yield of oat Beile 2 in Tibet reached 9,082 kg·hm−2, while the yield for oat Meida was approximately 6,900 kg·hm−2. In contrast, this study observed hay yields for oat Beile 2 ranging from 4,835 to 5,132 kg·hm−2, and for oat Meida, yields varied between 432 and 4,964 kg·hm−2. The lower yields in this study can be attributed to the two-season planting model and the shorter growth period of forage oats. Additionally, environmental conditions and regional differences in oat cultivation contribute to the observed yield reductions. Ma et al.44 noted that with a planting row spacing of 15 cm, the hay yield was 4,300 kg·hm−2, which is slightly lower than the yields observed in this study. This difference may stem from the narrower planting row spacing of 12 cm and variations in planting density in this research. Despite the lower yields, Ma's study reported a water productivity of 2.55 kg·m−2, which is higher than that observed in this research. Shah et al.45 found that the fresh weight of oats was between 12 411.6 kg · hm − 2 and 16 572.85 kg·hm⁻². The experimental area of Shah has a humid subtropical climate, characterized by an average annual rainfall of 790 mm, which may have caused waterlogging and eventually led to lower yields than in this study. Cheng et al.46 examined the impact of climate change on forage oat yields in northern Shanxi and concluded that climatic conditions were the primary factors influencing yields, with soil parameters having minimal significance. The climatic conditions and yield (fresh weight: 14,006–17,786 kg·hm⁻²) of the yield area in this study were consistent with the results of this research. Zhao et al.47 cultivated oats in a rainproof pond, reporting hay yields between 4,395.55 and 12,960.64 kg·hm−2, significantly higher than those recorded in this study. This underscores the influence of climatic conditions on oat growth and development.

The study identified water consumption for oats during the two seasons of 2022 and 2023 as follows: 309.2, 147.7, 315.4, and 304.2 mm, respectively. The average water consumption rates were calculated at 3.25, 2.72, 3.32, and 3.71 mm.d−1. Xu Bing et al.48 conducted field experiments in Lhasa, determining that the water demand for oats throughout the entire growth period was 570 mm, with water consumption varying between 464 and 570 mm under different conditions. Utilizing the Hydrus model, Wang et al.49 found that total water consumption for oats ranged from 404 to 571 mm. The discrepancies in water consumption observed in this study can be attributed to significant regional variations and different planting methodologies, such as single versus double growing seasons. Wang et al.50 reported findings from the ecotone between agriculture and animal husbandry at the northern foot of the Yinshan Mountain in Wuchuan County, Inner Mongolia, where water consumption of oats aligned closely with this study's results. However, fluctuations in soil moisture from 0 cm to 40 cm in this study were greater than those reported elsewhere, particularly between the 40 cm and 60 cm depths. The water productivity (WP) for oats in Ma et al.51 study ranged from 3.41 to 3.86 kg·m−3, exceeding the WP observed in this research by 1.4 to 2.33 kg·m−3, possibly due to the application of nitrogen fertilizer, which enhanced water productivity.

This study provides valuable insights into the growth and water use patterns of drip-irrigated oats cultivated in two seasons per year. The findings serve as a reference for enhancing irrigation water use efficiency in forage production systems and optimizing the allocation of agricultural water resources in irrigated regions.

Conclusions

This study utilized the AquaCrop model to simulate the growth and water consumption processes of oats, identifying the patterns of water use throughout the entire growth cycle across two seasonal plantings. The key conclusions are summarized as follows:

The AquaCrop model has a high fitting accuracy between the simulated value and the observed data, indicating that it has high reliability in simulating the growth and water consumption process of oats. The dynamic change trend of canopy coverage of oats in the two years was similar, which showed the stable simulation ability of the model to this index in different years. Although the yield of the first crop was higher than that of the second crop in both years, the change of water productivity (WP) was significantly different. The WP of the first crop in 2022 and 2023 was lower than that of the second crop, which may be related to the water use efficiency of different seasons. The total water consumption during the growth cycle ranged from 214.7 to 315.4 mm, accounting for 93.39 % to 95.69 %of the total water input (rainfall plus irrigation), indicating a high dependence of oat growth on water. The drainage loss is the smallest, accounting for only 0.36 % to 5.40 % of the total water delivery, indicating that there is room for optimization in water management. In each growth stage, the soil water content decreased the fastest and the water consumption was the largest at the jointing stage, which emphasized the importance of water management at this stage.

This study explored the growth process and water consumption process of drip-irrigated oats in two seasons in one year. The experimental scale was limited to the small-scale water consumption process of farmland, and there was a lack of comparison of variables such as different drip irrigation belt spacing, buried depth and irrigation quota. According to the needs of relevant policies, this experiment will reset the comparison of different variables in the future, and combine the conditions on the regional scale to quantitatively evaluate the growth process and water consumption process of drip-irrigated oats in the irrigation area along the Yellow River in Inner Mongolia in one year and two seasons, so as to provide an effective basis for the implementation of relevant policies.