Characteristics of soil salinity and water-salt transport in the vadose zone of salt-impacted regions with variable permeability

Abstract

Soil salinization poses a significant ecological challenge, emerging as a critical constraint to agricultural development in the arid and semi-arid regions of China, especially in southern Xinjiang. In particular, Yuepuhu County, situated in Kashgar, faces a distinctive issue. Impermeable thin clay layers within the vadose zone impede year-round leaching of salts, significantly impacting the growth of cotton. Through a combination of indoor testing, experiments, and statistical analyses, this study elucidated the varying permeability of soil layers at different depths and explored the forms and accumulation characteristics of soil salts in Yuepuhu County. It unveiled patterns of water and salt movement in soils with variable permeability layers, identifying key influencing factors. The research also proposed an irrigation regime suitable for cultivating vadose zone soils in the local context. The findings revealed a progression of increasing soil complexity and decreasing burial depth of clay layers from northwest to southeast, aligned with the direction of groundwater flow. With increasing depth, a noticeable reduction in soil saturated hydraulic conductivity was observed, indicating significant variability in permeability. Predominantly chloride-sulfate type saline soils in Yuepuhu County contained potassium (K⁺) and sodium (Na⁺) as the main cations in surface soils. Salinity strongly correlated with calcium (Ca²⁺) and magnesium (Mg²⁺). Chloride (Cl⁻), sulfate (SO₄²⁻), K⁺, Na⁺, and bicarbonate (HCO₃⁻) reflected the degree of soil salinization in Yuepuhu County. The clay interlayers in variable permeability zones significantly impeded water and salt movement in the vadose zone. Moving from west to east, thicker and shallower clay interlayers hindered downward water movement, increasing the difficulty of salt leaching. Additionally, the irrigation regime influenced water and salt movement in the vadose zone. Under the same soil structure, flood irrigation with a higher water flux resulted in more significant salt leaching, and lower total dissolved solids (TDS) in irrigation water were more favorable for effective salt leaching. Collectively, our findings provided a theoretical foundation for improving and managing local saline soils, as well as guiding the implementation of rational agricultural irrigation practices.

Introduction

Salinization of soil, recognized globally as a critical resource and soil issue, has garnered widespread attention from various sectors of society (Farifteh et al., 2006; Sheng et al., 2010; Wang et al., 2018; Qian et al., 2019). Soil salinization refers to the process where soluble salts in the soil rise with capillary water and accumulate on or near the surface after water evaporation (Hui et al., 2022). This phenomenon is primarily influenced by natural factors such as soil parent material, climate conditions, groundwater levels (Ma, 2023), and anthropogenic irrigation practices (Su et al., 2022), with direct and significant impacts on agricultural production (Fu et al., 2023). In recent years, due to factors such as irrational irrigation practices by humans (Hu et al., 2012; Yu et al., 2022), the extent of saline soils continues to increase annually (Hong et al., 2022; Zhuang et al., 2021). In China, approximately 1.0 × 10⁸ hectares of land are affected by varying degrees of salinization, accounting for 10.38% of the total global area of saline-alkali land (Lu et al., 2023), severely impeding the sustainable development of regional agriculture and the protection of the ecological environment.

The primary occurrence of soil salinization in China is in the arid and semi-arid regions of the northwest (Wang, 2021), with Xinjiang being the most severely affected region (Li et al., 2009; Yu et al., 2022). The main causes are the region's low precipitation and intense evaporation, coupled with the irrational drainage and irrigation systems leading to significant secondary salinization (Qian et al., 2019; Wang et al., 2008), greatly affecting agricultural development in arid areas. Additionally, salts in saline soils often exist in ionic forms, mainly including K⁺, Na⁺, Ca²⁺, Mg²⁺, Cl^–, SO₄^2–, CO₃^2–, and HCO₃^– (Shrestha, 2006). Fan et al. (2012) have pointed out that if the concentration of a specific salt in soluble soil salts is excessively high, the harm is greater than when multiple salts coexist. Research by Hu et al. (2018) and Qin et al. (2019) indicates that plants are highly susceptible to salt stress in high-salinity environments, leading to osmotic stress, ion toxicity, changes in membrane permeability, and disruptions in physiological metabolism. Numerous studies suggest that salt stress disrupting plant ion homeostasis is primarily due to high concentrations of ions such as Na⁺, Cl^–, and SO₄^2– in the soil (Hu et al., 2018; Qi et al., 2020; Nan et al., 2022). Therefore, when studying the harm of saline soils to agricultural production, attention should not only be paid to the total salt content but also to the ionic composition and spatial distribution characteristics of soil salts.

In the process of salt accumulation in saline soils, the spatial and temporal variations of soil moisture serve as a medium and carrier, while salt follows the principle of "salt comes with water, salt goes with water". The salinization status of soil in a region is intricately influenced by the migration of water and salts. Therefore, investigating the laws governing soil water and salt movement holds immense significance in unveiling the formation mechanisms and developmental patterns of regional saline soils (Barrett-Lennard, 2003; Jin et al., 2019). Such studies not only establish a theoretical foundation but also provide essential insights for enhancing water-salt regulation techniques and the reclamation of saline-alkali land (Zhou et al., 2018). Contemporary scholars employ various methodologies, such as field instrument monitoring and indoor soil column experiments, to delve into soil water and salt movement. For instance, Huang et al. (2020) have combined field monitoring with numerical simulation to delineate water and salt movement patterns in farmland soils located behind dams. This comprehensive approach allows them to identify internal and external factors contributing to salinization. Similarly, Dong et al. (2021), in their investigation involving soil column irrigation and evaporation experiments, have compared the water and salt movement and spatiotemporal distribution characteristics between sandwiched structures and homogeneous soils. Their findings unveil the specific inhibitory effects of the sandwiched structure on both evaporation and salt distribution. However, existing achievements often pertain to specific experimental conditions, and the regularities of water and salt movement under different irrigation methods, irrigation water quality, and soil textures remain undisclosed. Hence, further research is imperative to consider factors such as irrigation evaporation, soil struc-ture, and the impact of biological and chemical amendments. Such research should aim to elucidate water and salt movement patterns in saline soils under diverse soil structures and irrigation regimes, facilitating the optimization of effective water-salt regulation strategies.

In Yuepuhu County, situated in the heart of Kashgar in southern Xinjiang, salinization is influenced not only by soil parent material and climatic conditions (Abudureheman, 2011; Lv et al., 2014) but also by the presence of impermeable thin clay layers in the local vadose zone. These unique variable permeability soil layers result in ineffective year-round leaching, detrimentally impacting cotton growth and yields and leading to the abandonment of certain cotton fields. While previous studies on salinization in the Xinjiang region have primarily focused on large-scale regional analyses in the Kashgar River Basin (Wang et al., 2020; Li et al., 2022; Wang et al., 2022; Zeng et al., 2022), fine-scale research on salinization in Yuepuhu County is relatively scarce. Moreover, studies on the characteristics of salt content and water-salt movement patterns in areas with variable permeability soil layers are particularly limited. Hence, we, in the present research, selected the salinized area in Yuepuhu County as its focus, employing a combination of field surveys, sampling, laboratory analysis, and simulation experiments. The research utilized both traditional statistics and geostatistics to investigate the composition and distribution characteristics of soil salts, unveil water-salt movement patterns in the vadose zone, and propose an irrigation regime suitable for locally cultivated vadose zone soils. Collectively, the outcomes of this research served as a scientific basis for agricultural production, saline soil prevention and control, and reclamation efforts in Yuepuhu County.

Materials and methods

Study area

Yuepuhu County is situated at the southern foothills of the Tianshan Mountains, in the western Tarim Basin, at the lower reaches of the Gaizi River, within a desert oasis alluvial plain. The region is characterized by mountains on three sides, nestled deep inland, and experiences a typical warm temperate continental arid climate (Lin et al., 2018). Throughout the year, the area is dry with minimal rainfall, boasting an annual average sunshine duration of 2,795.6 h (Nueramina, 2020), an average annual precipitation of 52.8 mm, and an annual evaporation of 2,584 mm (Luo et al., 2014). Additionally, due to high salinity in the soil parent material and inappropriate agricultural irrigation practices, nearly 90% of the arable land in Yuepuhu County is affected by saline conditions. The average groundwater mineralization in the study area ranges from 2 to 3 g/L, with some localities exceeding 5 g/L.

Field investigations revealed that the groundwater flowed from west to east in the study area (Fig. 1), and the soil profile complexity increased along this direction, transitioning from a single-layer structure to double and triple layers. The soil properties of the layers at depths of 20–80 cm shifted gradually from loamy to fine sandy soil intermixed with clay and clayey mud, as depicted in Fig. 2 from on-site survey images. In the vertical direction, due to the absence of red clayey mud at 80 cm in the exposed K1 soil layer, a comparative analysis of the saturated permeability coefficients of the layered K2, K3, K4, K5, and K6 soil samples was conducted, with results shown in Fig. 3. On the soil vertical profile, as the soil structure transitioned from layered or stratified soil to clayey interlayers or clay layers in the deep soil, the saturated permeability coefficient decreased with increasing depth, indicating a distinct variation in permeability. The transition from layers with permeability classes K2 to K4 indicated a reduction in burial depth from west to east, aligning with the direction of groundwater flow. Specifically, the depth decreased from 65 cm to approximately 35 cm. This observation implied a gradual shallowing of variable permeability layers in the study area.

Fig. 1.

Study area and the distribution of sampling points

Fig. 2.

Images from field survey

Fig. 3.

Vertical distribution of soil saturated permeability coefficients

Sampling design and sample analysis

For this study, seven strategic sampling points were selected along both the groundwater flow direction (from west to east) and the vertical groundwater flow direction (approximately north–south), as illustrated in Fig. 1. The progression from west to east was denoted as K1 to K7, with K7 located in a reclaimed construction site. In line with requirements for indoor analysis and experimental design, fieldwork involved a sequential approach. Initially, the soil profile was measured, followed by the vertical collection of soil samples at specified depths. Sampling at each point was stratified based on soil vertical variations, covering samples for soil particle analysis, natural soil moisture content (ω), natural density (ρ), soil–water characteristics, and readily soluble salt content.

The soil particle composition was determined using the sieving method in accordance with the Standard Test Methods for Soil Engineering (GB/T 50123–2019). Natural gravimetric water content (ω) was determined using the oven-drying method, with soil samples placed in a constant-temperature oven at 105 °C for 12 h (Jin, 2015). The natural density (ρ) of the soil was measured using the ring-knife method (Fu et al., 2013). Characteristic soil–water samples were sent to the Institute of Soil and Water Conservation at Northwest A&F University for analysis, determining the soil's water characteristic curve through the centrifugation method to obtain the saturated permeability coefficient. Soil samples for readily soluble salt analysis were sent to the Xinjiang Rock and Soil Testing Center. The WET instrument was utilized to analyze the total content of readily soluble salts, and ion concentrations were determined according to the Standard Test Methods for Soil Engineering (GB/T 50123–2019).

Analytical methods

Using ArcGIS 10.8 software, the sampling point data were converted into spatial points with coordinates. The inverse distance weighting (IDW) interpolation method was employed to generate contour maps of soil pH, salinity, and salinization levels at different depths, reflecting the spatial distribution characteristics of soil salinity in the study area. SPSS 26.0 software was used to perform Pearson correlation analysis and principal component analysis (PCA) on the ionic components of soil salinity at different depths, quantitatively describing the characteristics of salinized soil and the distribution of salt ions in the study area.

Experimental design for indoor simulation

One-dimensional dispersion experiment

The experimental apparatus comprised a rectangular percolation tank with dimensions of 80 cm in length, 17 cm in width, and 36.5 cm in height (inner diameter). This tank was divided into three sections: the water supply zone (tracer injection area), percolation zone, and drainage zone. Two waterproof boards, each 1 cm thick, separated these sections, as illustrated in Fig. 4. A Mariotte bottle was employed for water supply, ensuring constant head control. Real-time conductivity data were monitored using a 24-channel data acquisition instrument.

Fig. 4.

Schematic representation of one-dimensional infiltration trench structure

When filling the percolation zone with soil, it was crucial to achieve equivalent filling based on the field's actual conditions and soil dry density. The dry density (ρ_d), determined by Eq. (1), is presented in Table 1. The filling thickness was 18 cm, with compaction every 3 cm. Three observation wells were evenly distributed horizontally in the seepage area as monitoring channels, located 10 cm, 34 cm, and 58 cm from the side wall of the water supply area. After filling, the system was saturated, maintaining an 8-cm height difference between upstream and downstream water heads during the stable percolation phase. Subsequently, 1 L of NaCl tracer solution with a concentration of 8 g/L was injected into the water supply zone, and conductivity was monitored at 5-min intervals. Further details of the dispersion experiment are outlined in Table 2.

$$\rho d=\frac{\rho }{1+\omega }$$ 1

where ρ_d represents the soil dry density (g/cm³), ρ denotes the soil natural density (g/cm³), and ω signifies natural gravimetric water content of soil (%).

Table 1.

Basic parameters of soil samples

Soil ID	Soil layer	Sampling depth (cm)	Natural density (g/cm3)	Natural gravimetric water content (%)	Dry density (g/cm3)
K1	Fine Sand	0–80	1.26	10.29	1.14
K2-1	Silty Clay with Fine Sand	0–40	1.35	12.73	1.20
K2-2	Fine Sand	40–80	1.55	15.29	1.34
K3-1	Silty Clay with Fine Sand	0–45	1.51	18.62	1.27
K3-2	Clay	45–80	1.90	27.39	1.49
K4-1	Silt	0–25	1.44	17.36	1.23
K4-2	Silty Clay with Fine Sand	25–45	1.68	18.76	1.41
K4-3	Red–Black Clay	45–80	1.83	25.12	1.46
K5-1	Silt	0–25	1.47	33.48	1.10
K5-2	Silty Clay with Fine Sand	25–45	1.32	20.39	1.10
K5-3	Red Clay	45–80	1.53	28.75	1.19
K6-1	Silt	0–25	1.73	14.25	1.51
K6-2	Silty Clay with Fine Sand	25–45	1.55	15.86	1.34
K6-3	Red Clay	45–80	1.72	31.15	1.31
K7	Reclaimed Soil for Construction	0–80	1.23	13.03	1.09

Table 2.

Experimental design for dispersion tests

Experiment ID	Soil ID	Fill thickness (cm)	Fill medium	Hydraulic head (cm)	Tracer concentration (g/L)
K1	K1	18	Fine sand	8	8
K2	K2-1	9	Silty clay with fine sand	8	8
	K2-2	9	Fine sand	8	8
K3	K3-1	9	Silty clay with fine sand	8	8
	K3-2	9	clay	8	8
K4	K4-1	6	Silt	8	8
	K4-2	6	Silty clay with fine sand	8	8
	K4-3	6	Red–black clay	8	8
K5	K5-1	6	Silt	8	8
	K5-2	6	Silty clay with fine sand	8	8
	K5-3	6	Red clay	8	8
K6	K6-1	6	Silt	8	8
	K6-2	6	Silty clay with fine sand	8	8
	K6-3	6	Red clay	8	8
	K7	18	Reclaimed soil for construction	8	8

Soil column irrigation experiment

The experimental setup comprised a soil column apparatus, an infrared lamp, an irrigation system, and a real-time monitoring system. The soil column apparatus boasted dimensions of 70 cm in height, 19 cm in base diameter, and a wall thickness of 1 cm. To simulate local evaporation conditions, a 150 kW/h infrared lamp was utilized. After determining the evaporation intensity of the infrared lamp, the bulb was positioned 20 cm above the soil column apparatus and exposed for 2 h each morning and afternoon. Irrigation was facilitated through a Mariotte bottle, and real-time monitoring of soil water-salt status was executed using a soil three-parameter sensor, measuring soil volumetric water content, conductivity, and temperature.

For the soil column apparatus filling, we adhered to equivalent filling based on field conditions determined by soil dry density. The filling thickness was maintained at 60 cm, with compaction applied at 5-cm intervals, as illustrated in the schematic diagram presented in Fig. 5. The soil three-parameter sensor was strategically embedded at depths of 0.1 m, 0.3 m, and 0.5 m within the soil column apparatus, automatically capturing data on soil moisture content, temperature, and conductivity every 20 min. Following the filling process, saturation and gravity drainage were meticulously executed, followed by a 5-days static closure of the bottom drainage hole before commencing irrigation.

Fig. 5.

Schematic diagram of soil column filling

To simulate diverse irrigation water qualities, NaCl solutions with varying total dissolved solids (TDS) values were configured. Emulating local irrigation practices, different irrigation methods were simulated by controlling the irrigation water flow rate, as delineated in the design details outlined in Table 3. Seven sets of experiments were conducted, simulating four irrigation cycles each, with drip and flood irrigation featuring cycle durations of 7 and 10 days, respectively. Each irrigation event involved applying 750 mL of water to ensure that soil moisture remained at optimal levels while allowing for the observation of soil salinity dynamics under different irrigation conditions.

Table 3.

Irrigation experiment design

Experiment ID	Soil ID	Fill Thickness (cm)	Dry Density (g/cm3)	Low TDS Water Irrigation			High TDS Water Irrigation			Irrigation Method
				Irrigation Water Quality (g/L)	First Irrigation Cycle	Second Irrigation Cycle	Irrigation Water Quality (g/L)	Third Irrigation Cycle	Fourth Irrigation Cycle
K1	K1	60	1.14	1	1st day	8th day	3	15th day	22nd day	Drip Irrigation
K2	K2-1	30	1.20	0.54	1st day	8th day	3	15th day	22nd day	Drip Irrigation
	K2-2	30	1.34	0.54			3
K3	K3-1	30	1.27	0.79	1st day	11th day	3	21st day	31s day	Surface Irrigation
	K3-2	30	1.49	0.79			3
K4	K4-1	20	1.23	1	1st day	8th day	3	15th day	22nd day	Drip Irrigation
	K4-2	20	1.41	1			3
	K4-3	20	1.46	1			3
K5	K5-1	20	1.10	1	1st day	8th day	3	15th day	22nd day	Drip Irrigation
	K5-2	20	1.10	1			3
	K5-3	20	1.19	1			3
K6	K6-1	20	1.51	1	1st day	8th day	3	15th day	22nd day	Drip Irrigation
	K6-2	20	1.34	1			3
	K6-3	20	1.31	1			3
K7	K7	60	1.09	1	1st day	8th day	3	15th day	22nd day	Drip Irrigation

Indoor simulation experiment data analysis

In the indoor experiment, we employed the conductivity conversion method to obtain solution concentration values at different time points. A series of NaCl solutions with varying concentrations were configured to establish the standard curve for electric conductivity versus concentration, as illustrated in Fig. 6, with the mathematical relationship defined in Eq. (2). According to Eq. (2), the conductivity data obtained from the dispersion experiment were converted into solution concentrations. Subsequently, Origin software was utilized for graphical analysis.

$$y=2079.3x+61.873,R^2=0.999$$ 2

where y represents electrical conductivity (μS/cm), x denotes solution concentration (g/L), and R² signifies the goodness of fit, with higher values indicating a closer proximity to 1 and thereby, an enhanced fitting precision.

Fig. 6.

Standard curves for electrical conductivity and solution concentration

After transforming the original dispersion data, the point-by-point parameter determination method was employed to obtain the hydrodynamic dispersion coefficient using Eq. (3), and the dispersivity was calculated using Eq. (4).

$$DL=\frac{\left(t_1-t_2\right)\times \left(x_{}^{_2^{}}-u{t}_1{t}_2\right)}{4t_1{t}_{2}ln\left(\frac{c_1{t}_1}{c_2{t}_2}\right)}$$ 3

where D_L represents the hydrodynamic dispersion coefficient (m²/s), t₁ and t₂ denote two moments in the dispersion experiment process (s), x is the distance from the observation well to the water supply end (m), u is the cross-sectional average flow velocity (m/s), and c₁ and c₂ are the concentrations corresponding to the moments t₁ and t₂ (g/L), respectively.

$$DL=a\times u$$ 4

where a denotes the dispersivity (m).

Utilizing Origin software, the data from dispersion and irrigation experiments were graphically presented. The trends in soil moisture content and conductivity under different irrigation conditions were analyzed, aiming to elucidate the water-salt transport patterns in the variable permeability layers of Yuepuhu County.

Results

Spatial distribution characteristics of soil salinity

Soil salinization levels

The spatial distribution results of soil salinization levels at different depths, generated using the IDW interpolation method in ArcGIS 10.8 software, are shown in Fig. 7. It can be seen that the degree of salinization in the study area gradually intensified along the groundwater flow direction (from northwest to southeast). The severely and extremely severely salinized areas were the most widespread, accounting for over 70% of the total area. The salt content of all soil layers increases to a certain extent along the groundwater flow direction. The salinity content decreased from the upper to the lower layers, with soil salinity primarily exhibiting surface accumulation. Additionally, the higher the overall ion content in the soil, the more pronounced the vertical differentiation of salinity.

Fig. 7.

Distribution of soil salinization levels a: 0–25 cm, b: 25–45 cm, c: 45–80 cm

The changes in soil salinity content were mainly related to the movement of water in the soil's vadose zone. In the study area, strong daytime sunlight and rising temperatures enhanced evaporation, causing upward movement of water and salt accumulation on the surface. The variable permeability soil layers in the soil affected the infiltration of water and the leaching of salts. Under the same local evaporation conditions, the downstream areas had more complex soil structures, thicker and shallower buried variable permeability soil layers, resulting in poorer leaching of salts after irrigation and increased soil salinity content.

Soil pH

Using the IDW interpolation method, the pH distribution results of soil at different depths are shown in Fig. 8. Overall, the soil pH in the study area ranged from 8.0 to 8.6, classifying it as alkaline soil. Horizontally, the pH distribution at different depths was similar to the distribution of salinity, with alkalinity increasing along the groundwater flow direction (from northwest to southeast). Vertically, soil pH decreased to varying degrees with increasing soil depth.

Fig. 8.

Soil pH distribution map a: 0–25 cm, b: 25–45 cm, c: 45–80 cm

Correlation between soil salinity and ionic composition

The results of soluble salt analysis are presented in Table 4. Combining the sampling depths of each soluble salt sample, it could be observed that the salinity in the shallow soil was generally the highest, followed by the middle layer, with the lowest in the third layer. This indicated a concentration of soil salts in the shallow part. In terms of ionic composition throughout the soil profile, sulfate ions (SO₄²⁻) were the highest among anions, followed by chloride ions (Cl⁻), and among cations, potassium ions (K⁺) and sodium ions (Na⁺) were the absolute dominant ions. After conducting Pearson correlation analysis using SPSS 26.0 for the total salt content and compositional ions in different soil depths, it was observed that the correlation between soil salinity and compositional ions varied from shallow to deep soil layers.

Table 4.

Results of soluble salt detection in soil

Soil ID	Soil Depth (cm)	Cl− (mg/kg)	HCO3− (%)	CO32−	SO42−	Ca2+	Mg2+	K+ + Na+	Total	Soluble Salt Content
K1	0–80	1424	201	0	1896	380	137	1627	5665	0.570
K2-1	0–40	2130	178	12	5664	2060	159	2017	12,220	1.225
K2-2	40–80	1612	206	12	2956	1031	123	1518	7458	0.749
K3-1	0–45	1624	195	12	5126	1625	265	1637	10,484	1.051
K3-2	45–80	578	250	0	1321	526	118	365	3158	0.319
K4-1	0–25	1979	211	12	9812	2047	523	3661	18,245	1.827
K4-2	25–45	2645	201	0	10,080	2000	824	3693	19,443	1.947
K4-3	45–80	1206	250	0	5123	1102	196	2274	10,151	1.018
K5-1	0–25	1566	198	12	3859	940	279	1795	8649	0.868
K5-2	25–45	812	219	0	1632	325	132	1030	4150	0.418
K5-3	45–80	622	256	0	912	311	68	605	2774	0.280
K6-1	0–25	2357	195	12	10,358	1521	1219	3353	19,015	1.904
K6-2	25–45	1256	253	0	7659	2165	523	1465	13,321	1.335
K6-3	45–80	936	260	0	2956	1024	125	948	6249	0.628
K7	0–80	155	253	0	431	80	30	341	1290	0.132

In the 0–25 cm soil layer, the soil salinity was significantly or highly significantly correlated with all ions except for HCO₃⁻. Specifically, the correlation coefficients with Cl⁻, SO₄²⁻, and K⁺ + Na⁺ were 0.884, 0.994, and 0.963, respectively, showing a highly significant positive correlation. The correlation coefficients with CO₃²⁻, Ca²⁺, and Mg²⁺ were 0.780, 0.840, and 0.803, respectively, indicating a significant positive correlation. This finding suggested that the salinity in the shallow soil was mainly related to the levels of Cl⁻, SO₄²⁻, CO₃²⁻, Ca²⁺, Mg2⁺, and K⁺ + Na⁺. The correlations with Cl⁻ and SO₄²⁻ were higher than those with other anions, indicating that chloride and sulfate salts were overwhelmingly dominant in the soil. For Cl⁻ and SO₄²⁻, the correlation coefficients with K⁺ + Na⁺ were higher than those with other cations, suggesting that potassium and sodium salts were predominant among chloride and sulfate salts in the shallow soil. Regarding Mg²⁺, the correlation coefficient with SO₄²⁻ was higher than that with other anions, indicating a higher content of MgSO₄ in the shallow soil. The correlation results are shown in Table 5.

Table 5.

Correlation matrix of soil salinity and ionic composition in the 0–25 cm soil layer

Parameter	Salinity	Cl−	HCO3−	CO32−	SO42−	Ca2+	Mg2+	K+ + Na+
Salinity	1.000	0.884**	− 0.545	0.780*	0.994**	0.840*	0.803*	0.963**
Cl−		1.000	− 0.866*	0.773*	0.831*	0.822*	0.642	0.845*
HCO3−			1.000	− 0.654	− 0.458	− 0.665	− 0.273	− 0.482
CO32−				1.000	0.759*	0.872*	0.49	0.653
SO42−					1.000	0.811*	0.831*	0.953**
Ca2+						1.000	0.404	0.732
Mg2+							1.000	0.771*
K+ + Na+								1.000

^**Significant at the 0.01 probability level; *Significant at the 0.05 probability level, and so forth

In the 25–45 cm soil layer, soil salinity was significantly or highly significantly correlated with all ions except for HCO₃⁻ and CO₃²⁻. Specifically, the correlation coefficients with Cl⁻, SO₄²⁻, Ca²⁺, Mg²⁺, and K⁺ + Na⁺ were 0.891, 0.990, 0.906, 0.895, and 0.898, respectively, showing a highly significant positive correlation. This finding indicated that the salinity in the middle layer was mainly related to the levels of Cl⁻, SO₄²⁻, Ca²⁺, Mg²⁺, and K⁺ + Na⁺. The correlations with Cl⁻ and SO₄²⁻ were higher than those with other anions, indicating that chloride and sulfate salts still overwhelmingly dominated in the 25–45 cm soil layer. For Cl⁻, the correlation coefficients with K⁺ + Na⁺ were higher than those with other cations, suggesting that potassium and sodium salts were predominant among chloride salts in this layer. For SO₄²⁻, the correlation coefficients with Mg²⁺ and Ca²⁺ were higher than those with other cations, indicating that sulfate salts formed in this layer were mainly composed of MgSO₄ and sparingly soluble CaSO₄. The correlation results are shown in Table 6.

Table 6.

Correlation matrix of soil salinity and ionic composition in the 25–45 cm soil layer

Parameter	Salinity	Cl−	HCO3−	CO32−	SO42−	Ca2+	Mg2+	K+ + Na+
Salinity	1.000	0.891**	− 0.363	0.202	0.990**	0.906**	0.895**	0.898**
Cl−		1.000	− 0.729	0.368	0.816*	0.757*	0.681	0.945**
HCO3−			1.000	− 0.654	− 0.233	− 0.304	− 0.033	− 0.558
CO32−				1.000	0.146	0.448	− 0.204	0.092
SO42−					1.000	0.915**	0.920**	0.840*
Ca2+						1.000	0.692	0.656
Mg2+							1.000	0.826*
K+ + Na+								1.000

In the 45–80 cm soil layer, the salt forms and accumulation characteristics were highly similar to those in the middle layer. The salinity in this layer was still highly related to the levels of Cl⁻, SO₄²⁻, Ca²⁺, Mg²⁺, and K⁺ + Na⁺, indicating that chloride and sulfate salts continued to be overwhelmingly dominant. Chloride salts were primarily composed of potassium and sodium salts, while sulfate salts were mainly represented by MgSO₄ and sparingly soluble CaSO₄. The correlation results are presented in Table 7.

Table 7.

Correlation matrix of soil salinity and ionic composition in the 45–80 cm soil layer

Parameter	Salinity	Cl−	HCO3−	CO32−	SO42−	Ca2+	Mg2+	K+ + Na+
Salinity	1.000	0.804*	− 0.294	0.318	0.976**	0.883**	0.910**	0.930**
Cl−		1.000	− 0.754	0.579	0.656	0.661	0.722	0.822*
HCO3−			1.000	− 0.594	− 0.091	− 0.074	− 0.249	− 0.47
CO32−				1.000	0.201	0.423	0.077	0.254
SO42−					1.000	0.894**	0.895**	0.869*
Ca2+						1.000	0.779*	0.654
Mg2+							1.000	0.839*
K+ + Na+								1.000

PCA of Soil Salt Ions

PCA was applied to construct independent composite variables from multiple salt variables without or with minimal loss of information. This quantitative approach was used to study the characteristics of saline soils and the distribution of salt ions in the study area, identifying representative dominant factors and accurately evaluating the salinization status. The number of principal components was determined based on the principle that the cumulative variance contribution rate should be greater than 85%.

The eigenvalues and contribution rates obtained from the PCA for the 0–25 cm soil layer are presented in Table 8, and the factor loadings and score coefficient matrix are shown in Table 9. By analyzing the variance contribution rates of the principal components, it could be observed that the eigenvalues for the first two principal components were greater than 1, while the others were less than 1. The cumulative contribution rate of the first two principal components reached 90.319%, preserving the majority of the original variable information, with only 9.681% information loss. This was sufficient to represent the main information represented by the original variable factors. According to the analysis results, the contribution rate of the first principal component was 77.129%, showing a strong positive correlation with soil salinity, Cl⁻, CO₃²⁻, SO₄²⁻, Ca²⁺, and K⁺ + Na⁺. This finding indicated that the first principal component was a comprehensive indicator reflecting the salinization status of the shallow soil, with a larger first principal component implying a more severe soil salinization. The contribution rate of the second principal component was 13.19%, showing a significant positive correlation with HCO₃⁻ (correlation coefficient = 0.616). HCO₃⁻ not only represents the composition of salt ions but also indicates soil alkalinity. Therefore, the second principal component reflected the alkalization characteristics of the shallow soil, with a larger second principal component indicating a more severe soil alkalization. Considering the significant correlation between various indicators and the first and second principal components, Cl⁻, SO₄²⁻, Ca²⁺, K⁺, and Na⁺ could be considered characteristic factors for the salinization status of the shallow soil.

Table 8.

Principal component eigenvalues and contribution rates in the 0–25 cm soil layer

Principal component	Eigenvalue	Contribution rate (%)	Cumulative contribution rate (%)
1	6.17	77.129	77.129
2	1.055	13.19	90.319
3	0.462	5.776	96.095
4	0.244	3.049	99.143
5	0.067	0.843	99.987
6	0.001	0.013	100
7	0	0	100
8	0	0	100

Table 9.

Principal component loading and score coefficient matrix in the 0–25 cm soil layer

Salt Variable	Factor loading matrix		Score coefficient matrix
	Principal component 1	Principal component 2	Principal component 1	Principal component 2
Salinity	0.979	0.182	0.159	0.172
Cl−	0.949	− 0.196	0.154	− 0.186
HCO3−	− 0.694	0.616	− 0.112	0.584
CO32−	0.854	− 0.266	0.138	− 0.252
SO42−	0.956	0.268	0.155	0.254
Ca2+	0.882	− 0.286	0.143	− 0.271
Mg2+	0.748	0.558	0.121	0.529
K+ + Na+	0.922	0.262	0.149	0.248

The results of the PCA for the 25–45 cm soil layer are presented in Tables 10 and 11. According to the variance contribution rates, the cumulative contribution rate of the first two principal components reached 90.159%, indicating that they could replace the original variable factors to express the main information represented by the factors.

Table 10.

Principal component eigenvalues and contribution rates in the 25–45 cm soil layer

Principal component	Eigenvalue	Contribution rate (%)	Cumulative contribution rate (%)
1	5.48	68.496	68.496
2	1.733	21.663	90.159
3	0.716	8.949	99.108
4	0.059	0.735	99.843
5	0.011	0.135	99.978
6	0.002	0.022	100
7	0	0	100
8	0	0	100

Table 11.

Principal component loading and score coefficient matrix in the 25–45 cm soil layer

Salt variable	Factor loading matrix		Score coefficient matrix
	Principal component 1	Principal component 2	Principal component 1	Principal component 2
Salinity	0.987	0.142	0.18	0.082
Cl−	0.95	− 0.21	0.173	− 0.121
HCO3−	− 0.51	0.768	− 0.093	0.443
CO32−	0.298	− 0.858	0.054	− 0.495
SO42−	0.954	0.243	0.174	0.14
Ca2+	0.884	− 0.028	0.161	− 0.016
Mg2+	0.836	0.529	0.152	0.305
K+ + Na+	0.93	0.061	0.17	0.035

The contribution rate of the first principal component of the soil in the 25–45 cm layer was 68.496%, with correlation coefficients with soil salinity, Cl⁻, SO₄²⁻, and K⁺ + Na⁺ all exceeding 0.9. This finding suggested that the first principal component remained a comprehensive indicator reflecting the soil salinization status in the middle layer. The contribution rate of the second principal component was 21.663%, exhibiting negative loadings with Cl⁻, CO₃²⁻, and Ca²⁺ and positive loadings with other indicators. Notably, it had the highest positive correlation with HCO₃⁻, followed by Mg²⁺. Considering the significant correlation between various indicators and the first and second principal components, Cl⁻, SO₄²⁻, Ca²⁺, K⁺, and Na⁺ could be considered characteristic factors for the salinization status of the middle layer soil.

The results of the PCA for the 45–80 cm soil layer were presented in Tables 12 and 13. Similarly, the first two principal components could replace the majority of the original indicators. It could be observed that the distribution characteristics of salt ions in the deep soil layer were highly similar to those in the shallow and middle layers. The first principal component continued to reflect the soil salinization status, while the second principal component reflected the alkalization status. Considering the significant correlation between various indicators as well as the first and second principal components, Cl⁻, SO₄²⁻, Ca²⁺, K⁺, Na⁺, and HCO₃⁻ could be considered characteristic factors for the salinization status of the deep soil layer.

Table 12.

Principal component eigenvalues and contribution rates in the 45–80 cm soil layer

Principal component	Eigenvalue	Contribution rate (%)	Cumulative contribution rate (%)
1	5.418	67.731	67.731
2	1.617	20.217	87.948
3	0.714	8.927	96.875
4	0.175	2.188	99.063
5	0.057	0.719	99.781
6	0.017	0.219	100
7	0	0	100
8	0	0	100

Table 13.

Principal component loading and score coefficient matrix in the 45–80 cm soil layer

Salt variable	Factor loading matrix		Score coefficient matrix
	Principal component 1	Principal component 2	Principal component 1	Principal component 2
Salinity	0.98	0.181	0.181	0.112
Cl−	0.899	− 0.374	0.166	− 0.231
HCO3−	− 0.46	0.815	− 0.085	0.504
CO32−	0.443	− 0.706	0.082	− 0.437
SO42−	0.917	0.369	0.169	0.228
Ca2+	0.861	0.234	0.159	0.145
Mg2+	0.896	0.3	0.165	0.185
K+ + Na+	0.926	0.04	0.171	0.025

Dispersion test results

During the dispersion process, the salt concentration of soil water in each observation hole increased with displacement, and the appearance of peaks exhibited a lag with decreasing peak amplitudes. Over time, after reaching the peak concentration, the salt concentration of soil water gradually approached the initial concentration but remained slightly higher than the initial concentration. Comparing the results of each experimental group, it was observed that the more complex the soil structure and the higher the clay content, the more pronounced the lag in peak concentration, as seen in K2, K3, K4, K5, and K6 (corresponding to b, c, d, e, and f in Fig. 9, respectively). The dispersion test duration in sandy soil was shorter than that in clayey soil, as observed in K3 (Figure c), indicating that clay particles had a significant retention and adsorption effect on solutes. Figures a–f in Fig. 9 represent the concentration trends for the K1–K7 groups.

Fig. 9.

Concentration trends for groups K1–K7 (a–g)

Based on the experimental data, we computed the corresponding dispersivity parameters using Eq. (3) and (4). The overall trend indicated that as the infiltration rate increased, both the dispersivity coefficient and dispersivity exhibited higher values, as shown in Table 14.

Table 14.

Permeability rate, dispersivity, and dispersivity coefficient results

Experiment number	Permeability rate (cm/d)	Dispersivity (cm)	Dispersivity coefficient (cm2/d)
K1	2.18	17.10	37.27
K2	1.51	37.10	56.02
K3	1.36	29.65	40.32
K4	1.17	29.07	34.01
K5	1.10	28.71	31.58
K6	1.05	42.11	44.22
K7	2.09	11.37	23.77

Irrigation experiment results

Soil moisture variation patterns

Throughout the four irrigation cycles, the soil moisture at different depths exhibited an initial increase followed by a subsequent decrease, ultimately reaching a stabilized state, as illustrated in Fig. 10. Homogeneous structure surface soils (K1, K7) displayed the most pronounced fluctuations in moisture content with irrigation cycles, reaching saturation after the first irrigation (Fig. 10a and g). As the depth increased, the range of moisture content variation decreased, showing a lag effect.

Fig. 10.

Soil Moisture variation trends at different depths for groups K1-K7 (a–g)

The surface soil of the dual-layer structure (K2, K3), consisting of fine sand with intercalated clay, and the underlying layers comprising loamy sand and loam, demonstrated varying permeabilities. The permeability of the clay-intercalated layer was inferior to that of loamy sand and loam, and loam exhibited lower permeability than loamy sand. This manifested in the moisture content peak in the subsoil, with K3 exhibiting the most delayed response and the overall moisture content variation being relatively small. The difference in irrigation methods between K2 (drip irrigation) and K3 (surface irrigation) resulted in varying peak times within each irrigation cycle. For drip irrigation (K2), significant moisture content variations occurred in the initial 5 days of each cycle, while for surface irrigation (K3), this was observed in the initial 3 days (Fig. 10b and c).

In the case of the multi-layer structure (K4, K5, and K6), the moisture content exhibited a stepwise continuous increase over the four cycles. In the first two irrigation cycles, the soil moisture of the surface and middle layers in K4, K5, and K6 did not reach their maximum peaks. After reaching these peaks, their values were consistently higher than the other four groups. Notably, in K4 and K5, the moisture content in the deep soil did not show a significant increase, highlighting that soils with a multi-layer structure not only had a slower infiltration rate but also possessed strong water retention capacities in clay particles (Fig. 10d, e, and f). Additionally, the moisture trend in the deep soil of K6 differed significantly from that of K4 and K5. Based on the particle analysis results (Table 15), we believed this was due to the different particle compositions in the deep soil of K6 compared to K4 and K5. According to the particle analysis, the deep soils of K4 and K5 did not contain medium sand, whereas the deep soil of K6 had a medium sand content of 0.5%. This difference in soil texture might significantly impact the water-salt migration mechanism: the finer particles in the deep layers of K4 and K5 resulted in slow infiltration of water and salt into the deep layers, leading to no significant increase in deep-layer water and salt during the first three cycles. In contrast, K6, with its trace amount of medium sand, allowed more efficient water and salt migration in the deep soil, resulting in distinct patterns of salt changes in the deep soil during each irrigation cycle.

Table 15.

Particle analysis results

Sample ID	Sampling depth (cm)	Percentage of particle composition (%)
		Coarse sand		Medium sand	Fine sand Mm	Very fine sand, silt and clay
		2 ~ 1	1.0 ~ 0.5	0.5 ~ 0.25	0.25 ~ 0.075	< 0.075
K1	0–80	0.0	2.1	14.6	23.9	59.4
K2-1	0–40	0.0	0.0	11.2	36.4	52.4
K2-2	40–80	0.0	0.0	6.7	30.8	62.5
K3-1	0–45	0.0	0.0	5.7	15.8	78.5
K3-2	45–80	0.0	0.0	1.6	12.8	85.6
K4-1	0–25	0.0	0.0	1.3	22.9	75.8
K4-2	25–45	0.0	0.0	9.6	21.9	68.5
K4-3	45–80	0.0	0.0	0.0	8.5	91.5
K5-1	0–25	0.0	0.0	3.5	16.7	79.8
K5-2	25–45	0.0	0.0	5.6	23.3	71.1
K5-3	45–80	0.0	0.0	0.0	18.4	81.6
K6-1	0–25	0.0	0.0	2.6	12.2	85.2
K6-2	25–45	0.0	0.0	5.2	15.0	79.8
K6-3	45–80	0.0	0.0	0.5	14.9	84.6
K7	0–80	0.0	2.5	15.1	29.9	52.5

Soil electrical conductivity dynamics

The electrical conductivity of probes at different depths also exhibited an initial increase followed by a subsequent decrease, ultimately stabilizing over the four cycles. Comparing the electrical conductivity variations between the seven groups irrigated with low TDS (first two cycles) and high TDS (last two cycles), it was observed that during high TDS irrigation, the increase in electrical conductivity was more substantial at all soil layers. Moreover, the lowest electrical conductivity values during high TDS irrigation exceeded the peaks observed in low TDS irrigation, indicating that irrigation water quality influenced the water-salt transport process, with low TDS irrigation water being more conducive to salt leaching. Additionally, in the case of drip irrigation (K2), the electrical conductivity reached its peak early in the irrigation cycle and remained elevated until near the end of the cycle. For surface irrigation (K3), after reaching the peak electrical conductivity, it leveled off within 2–3 days, suggesting that different irrigation methods significantly affected the water-salt transport process within the vadose zone (Fig. 11b and c).

Fig. 11.

Trends in soil electrical conductivity at different depths for groups K1-K7 (a–g)

Comparing the homogeneous structure of K1, K7, and the other four groups with a multi-layer configuration, it was observed that the electrical conductivity trends in the surface and deep layers of the homogeneous structure were more similar. This finding implied that as the soil structure became more uniform, the differences in salt content variation at different depths became smaller (Fig. 11a and g). In K4, K5, and K6, the electrical conductivity at 10-cm and 30-cm depths exhibited a stepwise continuous increase over the first two cycles, with a significantly higher rate of increase compared to the single-layer K1 and dual-layer K2 and K3 structures. This suggested that with a more complex soil layer structure, surface salt accumulation became more pronounced (Fig. 11d, e, and f).

Discussion

Salt composition and spatial distribution characteristics in saline areas

The research results indicated that chloride and sulfate dominated the soils of the saline areas in Yuepuhu County. Specifically, chloride was predominantly in the form of potassium and sodium salts, while sulfate, in the surface layer, was mainly in the form of potassium and sodium salts, and in the middle to deep layers, it was primarily composed of MgSO₄ and sparingly soluble CaSO₄. Overall, Cl⁻, SO₄²⁻, K⁺, and Na⁺ could be considered characteristic factors reflecting soil salinization, while HCO₃⁻ served as a characteristic factor reflecting soil alkalinization. Spatially, the soil salt content in Yuepuhu County generally exhibited an increasing trend along the direction of groundwater flow (from west to east). Additionally, comparing the easily soluble salt content in the 0–25 cm, 25–45 cm, and 45–80 cm layers, it was observed that the salt content increased to a certain extent along the direction of groundwater flow in each layer. Vertically, the soil salt content primarily showed surface accumulation, and the higher the soil ion content, the more pronounced the vertical differentiation in salt content.

The primary reason for these patterns was closely related to the spatial distribution of the variable permeability soil layers in Yuepuhu County. Firstly, spatially, Yuepuhu County was located in the alluvial plain area of the Gaizi River basin, downstream of the Gaizi River. The runoff gradually decreased from west to east, and the particle size of suspended matter became smaller. Simultaneously, the overall topography of the county was higher in the west and lower in the east, with the groundwater level in the eastern region being lower than that in the western region. Under the influence of sedimentary sorting and long-term westward irrigation, the eastern soil has a higher clay content compared to the western soil (Wang, 2015), leading to stronger adsorption properties for salt. Furthermore, the gradual increase in soil layer complexity and clay content from west to east, combined with other factors, results in the characteristic of higher salt content in the eastern part of the county. The northwest and central regions generally had soil salt content below 4 g/kg, while in the central and southeast regions, the salt content exceeded 10 g/kg. The vertical aggregation of salt content was attributed to the intense weathering of parent materials, causing soil salts to migrate with water flow. In the dry climate of Yuepuhu County, with sparse rainfall and intense evaporation conditions, salts tend to accumulate in the surface soil (Abudureheman, 2011). Additionally, the burial depth of the variable permeability soil layer gradually became shallower from west to east in Yuepuhu County. This characteristic resulted in the accumulation depth of soil salt becoming shallower from west to east, leading to a more pronounced vertical differentiation in soil salt content in the eastern region. The surface soil salt content in the northwest was approximately 1.6 times that of the deep soil (K2), while in the southeastern region, the difference increased to 3.1–3.3 times (K3, K5).

Numerous studies have consistently demonstrated that the structure, thickness, sequence, and distance from the surface of stratified heterogeneous soils significantly influence the transport of water and salts in the soil (Huang et al., 2013; Pillai et al., 2009). Chen et al. (2020) have emphasized that as the structure of heterogeneous soils becomes more complex, the internal impermeable layers become thicker and closer to the surface. This complexity leads to a more pronounced hindrance to the transport of water and salts in the soil, making salts more likely to accumulate in the surface layer. In the later stages of irrigation with saline or slightly saline water, salt deposition during evaporation becomes more likely due to continuous salt enrichment. Liu (2022) has also suggested that impermeable layers in heterogeneous soils can elevate the water table from the dry surface to the top of the impermeable layer. The low permeability of these layers results in a significantly lower water transmission rate in the vadose zone compared to atmospheric evaporation intensity, leading to an increased proportion of water loss from the dry surface. Consequently, water carries salt away, and salt accumulates at the top of the impermeable layer.

It was evident that the spatial distribution characteristics of the variable permeability soil layers in Yuepuhu County further impacted the overall spatial distribution of salt. This ultimately contributed to the gradual increase in salt content from west to east in Yuepuhu County, with a corresponding variation in the depth of salt accumulation from deep to shallow.

Water-salt transport patterns and influencing factors

Based on the experimental results of this study, the soil moisture and electrical conductivity at different depths with variable permeability soil layers showed a trend of initial increase, followed by a decrease, and eventually stabilizing in each irrigation cycle. This pattern indicated that after irrigation water with a certain TDS background entered the soil, the soil salt content initially increased. Subsequently, as the irrigation water moved downward, salt accumulated in the surface layer due to intense evaporation and capillary rise of water (Natthawit, 2016), resulting in a periodic process of irrigation leaching and salt accumulation through evaporation. After irrigation with high TDS water, the surface soil water TDS of columns K1–K7 increased by 0.37 to 5.08 times compared to low TDS water irrigation (Fig. 12).

Fig. 12.

Influence of different irrigation water qualities on surface soil water TDS

The repeated surface salt accumulation phenomenon is closely related to local climate, hydrogeological conditions, and irrigation practices (Ma, 2023), with the irrigation regime playing a significant role (Su et al., 2022). The impact of the irrigation regime on the water-salt transport process is considered in terms of irrigation water quality and method. In terms of water quality, low TDS irrigation water is more favorable for salt leaching, while high TDS irrigation water is more likely to induce soil salt accumulation (Ma et al., 2023). The impact of irrigation methods on vadose zone water-salt transport is mainly attributed to the amount of irrigation water. Salt in the soil follows the principle of "salt follows water". Drip irrigation inherently has a lower rate of downward water movement, coupled with intense water evaporation, resulting in a small irrigation water flux. Therefore, salt leaching in drip irrigation is not significant. However, drip irrigation, through precise control of water volume and flow, significantly reduces surface evaporation and underground leakage of water (Shu et al., 2022; Yang et al., 2023), thus improving water use efficiency.

On the other hand, if excessive water is applied during surface irrigation, it may cause the groundwater level to rise and exacerbate secondary salinization. Comparing columns with similar soil structures but different irrigation methods, such as K2 (drip irrigation) and K3 (surface irrigation), it could be observed that the kurtosis of the soil moisture content curves at each layer during drip irrigation was smaller than that during surface irrigation (Fig. 10b and c). This finding indicated that compared to surface irrigation, drip irrigation not only had a lower rate of water infiltration but also a lower rate of water loss due to evaporation and infiltration. Therefore, from the perspective of water resource utilization efficiency, drip irrigation was more suitable for local water-saving agriculture. However, it is essential to consider that drip irrigation has a lower water flux, and salt may accumulate in the surface or crop root zone.

Furthermore, soil structure and texture also influence the water and salt transport in the vadose zone. In the study area, salinization intensified from west to east. This was attributed to the transition from a single layer (K1) to a progressively heterogeneous double layer (K2 and K3) and eventually to a three-layer structure (K4, K5, and K6) in the soil profile from west to east. The depth of impermeable soil layers within the vadose zone decreased, and clay content increased in the soil towards the east. Consequently, in the eastern regions, locations where downward water movement was impeded were shallower, leading to a greater accumulation of salts in the surface layer. The variation in salt content between different soil depths also became more pronounced. This phenomenon underscored significant differences in salt retention mechanisms between heterogeneous stratified soil and homogeneous soil. The salt retention in agricultural soils with stratified layers is primarily influenced by the characteristics of poorly permeable soil layers (Zhao et al., 2015). Shallow positions or increased thickness of clay interlayers are less conducive to leaching and eluviation of salts (Li et al., 2004). Simultaneously, in terms of soil texture, clay particles exhibit lower permeability compared to sand particles, resulting in stronger salt retention and adsorption capabilities (Wang et al., 2014).

Conclusions

The results of this study indicated that the spatial distribution of variable permeability soil layers in Yuepuhu County was closely related to the direction of groundwater flow. From northwest to southeast (along the groundwater flow direction), the soil layer composition became increasingly complex, and the depth of clay layers decreased. Vertically, the saturated hydraulic conductivity of the soil decreased to varying degrees with increasing depth, exhibiting significant variable permeability.

Overall, the spatial distribution of salinity in Yuepuhu County was primarily influenced by the spatial distribution of variable permeability soil layers. Soil salinity exhibited a low-to-high gradient from northwest to southeast, with a notable accumulation in the surface layer.

In the salinized areas, chlorides and sulfates dominated, mainly in the form of potassium salts and sodium salts. According to PCA, Cl⁻, SO₄²⁻, K⁺, Na⁺, and HCO₃⁻ could be considered characteristic factors reflecting the salinization status of the soil in Yuepuhu County.

Soil texture and structure, irrigation methods, and irrigation water quality all impacted the water-salt migration process in the vadose zone. Under the same irrigation mode, the more complex the soil structure, the shallower the depth of variable permeability soil layers in the vadose zone, and the thicker the clay interlayers, the stronger the hindrance to water-salt migration in the vadose zone. With the same soil structure, higher water flux in the irrigation mode resulted in more pronounced salt leaching. Additionally, lower TDS values in irrigation water favored salt leaching, while higher TDS values were more likely to cause evaporation and salt return.

Author contributions

All authors contributed to the conception and design of the study. The trial preparation, data collection and analysis were carried out jointly by Ma, Pang, Liang and Tuerhong. The first draft of the manuscript was written by Ma, and Ge proposed revisions. All authors have commented on previous manuscript editions. All authors read and ap-proved the final manuscript.

Funding

This work was supported by the University Student Innovation and Practice Project (Grant numbers S202110755001).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.