Distribution Characteristics and Relationship Between Soil Salinity and Soil Particle Size in Ebinur Lake Wetland, Xinjiang

Abstract

A comprehensive understanding of soil salinity characteristics and the vertical and spatial distribution of particle sizes in lakes and wetlands within arid zones, as well as elucidating their interrelationship, is crucial for effective wetland soil salinization management. In this study, the typical salinized wetland, the Ebinur Lake wetland, was selected as the research object. A total of 50 sampling points were established along the edge of Ebinur Lake, resulting in the collection of 200 soil samples from depths of 0–60 cm. The particle size distribution (PSD) of the soil samples was obtained by laser particle sizer, and the fractal dimension of the soil structure was deduced by applying fractal theory. The soluble salt content (TSS) and salt ions content were measured by laboratory physicochemical experiments. Finally, Pearson correlation and other methods were used to explore the relationship between soil salinity and soil particle size. The results showed the following: (1) Soil salinization in the study area was severe, and the accumulation of surface salts was obvious, with a mean value of 46,410 mg/kg. The spatial distribution of TSS was predominantly influenced by Cl⁻, SO₄²⁻, Na⁺ + K⁺, Mg²⁺, and Ca²⁺. (2) Across various soil depths, silt and sand were the primary constituents, with soil fractal dimensions (D_soil) ranging from 1.91 to 2.76, averaging 2.54, and a poor soil textural structure. The spatial distribution of D_soil closely mirrored that of TSS. (3) According to the correlation analysis results, as TSS increased, D_soil continued to rise, with an increasing content of clay, while the sand content decreased. Simultaneously, as the soil particles became finer, TSS and D_soil also increased, suggesting that sandy loam to silty soils in the study area were more prone to salt accumulation.

1. Introduction

Soil salinization represents a significant challenge to the sustainable ecological functionality of lake wetlands in arid and semi-arid regions [1,2]. The essence of soil salinization lies in the overaccumulation of soluble salts, such as Na⁺, K⁺, Cl⁻, and SO₄²⁻, in the soil [3]. Currently, salinization affects more than 1 billion hectares of land in more than 100 countries, mainly in Africa, Western Asia and Europe, and Western America [4]. In China, salinized soil covers approximately 3690 km², mainly concentrated in the northern arid zone, the western irrigated zone, and the eastern coastal zone, where the proportion of salinization significantly exceeds the global average.

Soil salinization has become a pressing issue attracting extensive attention from experts both domestically and internationally. Numerous scholars have investigated the causes, characteristics, distribution, and change trends of salinization in arid and semi-arid zones, such as Xinjiang, China [5], the Hetao Irrigation District in Inner Mongolia [6], the coastal area [7], and Europe [8], on the basis of experiments and practical investigations, using methods such as field samples and indoor experiments, and have achieved certain results. For example, Gebremeskel analyzed the planar spatial distribution characteristics of saline soils in Northern Ethiopia and identified discernible patterns in soil salt distribution, providing valuable direction for future research [9]. Additionally, researchers have used correlation analysis to explore the relationships between soil salinity and ions, deepening the understanding of salinity structures. Zheng et al. identified strong correlations between Cl⁻, Na⁺ + K⁺, SO₄²⁻, Ca²⁺, and Mg²⁺ across soil layers [10]. Li et al. observed that total salt concentration was negatively correlated with HCO₃⁻ but positively correlated with other ions [11]. Among the anions, SO₄²⁻ and Cl⁻ showed the strongest correlation, with Na⁺ being most closely associated with Cl⁻. At the same time, considerable research was conducted on the spatial heterogeneity of soil salinity. Soil properties such as salinity, water content, and salt ion composition exhibit pronounced spatial and temporal variability. Geostatistical methods, which integrate geographic coordinates with soil data, have proven indispensable for analyzing this heterogeneity. In the 1980s, Burgess et al. introduced the theory of regionalized variables and spatial interpolation methods into the study of spatial heterogeneity of soil salinity to quantify it and push the research forward [12]. Liu et al. used GIS and geostatistical methods to study the spatial variable distribution characteristics of soil features in the Yellow River Delta, and found that the salinity between adjacent soil layers had strong spatial correlation [13]. In terms of research content, domestic and international research on soil salinization has mainly focused on salinization caused by irrigated agriculture [14,15,16,17,18]. However, saline wetlands with unique ecological roles are often overlooked [19] and there are no accurate estimates of current or future salinization of freshwater wetlands in inland and coastal areas on a global scale, which is clearly a pervasive global problem and is expected to worsen over time [20]. In China, wetlands are widespread, with numerous studies focusing on the salinization characteristics of coastal wetlands, including mechanisms and technical methods [21,22]; however, the salinization characteristics of inland lake wetlands differ due to regional environmental variations [1]. In arid and semi-arid regions with low rainfall and high evapotranspiration, lake wetlands are particularly prone to salt accumulation, necessitating urgent attention and research. Wetland salinization is progressing far beyond natural salinity variations [23] and poses a major threat to the sustainability of lake wetland ecosystems. Therefore, addressing soil salinization in these ecosystems is particularly important in arid regions.

The key issue in addressing soil salinization lies in the improvement of soil structure [24]. Essentially, measures to ameliorate saline soils involve regulating soil water and salt movement, promoting the downward leaching of salts, and preventing the upward migration of salts due to evaporation [25,26]. Therefore, understanding the relationship between soil salinity and particle size is crucial. Soil particle size significantly influences the retention and transfer of water, nutrients, air, and heat in the soil [27], and it influences soil permeability, soil water and salt transport, nutrient conversion, and solute migration [28,29], thus affecting soil quality. Numerous factors influence soil salinity accumulation [30]. Soil texture is one of the spatial drivers of soil salinity, characterized by the relative proportions of fine and coarse mineral particles. The size and shape of soil particles affects pore morphology, and the interactions between gravity, cohesion, and capillary forces can affect the movement of water and salts [31]. In assessing soil texture, researchers have found the soil fractal dimension (D_soil) to be a useful tool for characterizing particle size distribution (PSD) and soil structure [32,33]. This dimension serves as a quantitative indicator for evaluating soil fertility and degradation [34], reflecting the distribution of coarse and fine particles and their correlation with soil fertility, hydrodynamic properties, and erosion resistance, which are essential for ecological restoration.

During the formation of saline soils, soil particle size plays the role of mediator and carrier, influencing the process of salt stress. Soil salinity is significantly affected by the clay content, with positive correlations between clay and silt fractions and negative correlations with sand [35,36]. Changes in particle size also affect salinity, as alterations in capillary thickness modify water and salt redistribution [37]. In addition, soil texture and its vertical spatial heterogeneity may also greatly affect water and solutes distribution in the soil profile [38]. Zettle indicated that the distribution of soil texture in the vertical profile also significantly affected soil water–salt transport [39]. In arid zone lake wetlands, driven by natural and human factors, the lake water levels decrease, leading to an increase in exposed areas. The salts in these exposed areas migrate horizontally and vertically, accumulating on the surface, which exacerbates the salinization of lake wetlands [40], whereas in the process of water and salt migration, soil particle size, as a critical medium of capillary components, plays a key role in the distribution of salts [41]. However, most studies on soil salinization are limited to the soil surface layer only [42,43,44], ignoring the characteristics of vertical distribution of salts, whereas a quantitative study of the vertical distribution characteristics of soil salts can accurately obtain the information of potential salinization in the region, which is of great significance in guiding the ecological restoration of wetlands.

Soil texture is the root cause of soil properties, which can vary in properties such as adhesion, water retention, permeability, and ion exchange due to the size of the constituent soil particles and the content of different sized soil particles [45]. Therefore, this study aimed to (i) analyze the distribution characteristics of soil salts/salt ions as well as soil particle size in the vertical profile of Ebinur Lake wetland; (ii) study the relationship between soil salinity and soil particle size in Ebinur Lake wetland; and (iii) analyze the influencing factors of soil salinity, aiming to provide theoretical support for the restoration of the wetland ecosystem of Ebinur Lake.

2. Materials and Methods

2.1. Study Area

Ebinur Lake Wetland National Reserve is located in the southwest of the Junggar Basin in Xinjiang Uygur Autonomous Region (82°33′–83°53′ E, 44°37′–45°15′ N). Ebinur Lake in the region is the largest saltwater lake in Xinjiang. The region has a typical temperate continental climate, with scarce precipitation (<150 mm/yr), strong evaporation (>2000 mm/yr), and an average annual temperature of about 7.8 °C, making it a typical wetland–desert ecosystem of the temperate arid zone [41]. At the same time, because the wetland is located in the Alashankou gale channel, the wind is always gusty, and its special geographic location makes the area the main source of sand and dust storms in the northern border [46]. Typical soil types in the region are mainly desert soil consisting of piedmont psephitic, salt, and gypsum [47]. The soils are coarse textured and dominated by sandy soils and sandy loams [48]. Due to the special climatic characteristics of the wetland, the low level of the terrain, the soil texture, and the shallow water table, the study area is characterized by severe salinization and the distribution of saline soils over a wide area. Under these environmental conditions, the vegetation in the study area is sparse and dominated by saline plants. Salt-tolerant plants dominated by reeds, pikes, tamarisks, and poplars formed according to the different salt content of the soil [49]. Overall, the high degree of soil salinization and ecological sensitivity [50] of the Ebinur Lake wetland provides a good research example to explore the relationship between salinization and soil particle size. And this study is of great value for the ecological restoration and promotion of sustainable development in the Ebibur Lake wetland.

2.2. Data Collecting and Processing

The research methods include selecting representative oases, deserts, and ecotones based on the ecological landscape characteristics of the Ebinur Lake wetland, and sampling along the edge of the wetland. As shown in Figure 1, GPS was used for positioning, and 50 sampling points were established. At each profile, four layers were sampled at different depths to reflect the vertical variation in soil salinity, 0–10, 10–20, 20–40, and 40–60 cm, with a total of 200 samples collected. Sampling was conducted in late August 2023. Three soil samples were taken from different soil depths at each sampling point and mixed to form the test sample for that layer. The research technical roadmap was outlined as follows (Figure 2).

2.2.1. Soil Particle Size Measurements

Measurements of the relevant components of soil samples were mainly carried out in laboratories. The soil samples were allowed to air-dry naturally to remove impurities such as grass roots and were ground and passed through a 2 mm sieve before being used for the measurement of soil salt ions and the determination of soil particle size. Soil particle size was measured using a Mastersizer 3000 laser particle size analyzer(Malvern Panalytical, Shanghai, China) to determine the volume fraction of soil particle sizes. Each soil sample was measured three times, and the arithmetic mean was calculated. Soil particle size classification followed the USDA particle size classification standards: clay (<0.002 mm), silt (0.002–0.05 mm), and sand (0.05–2 mm).

2.2.2. Soil Salinity Measurement

Salinity content can be expressed as total soluble salt (TSS) or as soil electrical conductivity (EC) of saturated extract or soil–water suspensions [3]. Whereas EC measurements do not identify the type of salt present in the sample, but only the total cumulative concentration of soluble salts present [51]. Therefore, TSS is used in this paper to represent soil salt content. In general, the choice of the soil–water ratio should be based on the specific purpose, and various ratios can be used, such as 1:1, 1:2, 1:5, and 1:10 for soil-to-water ratios [31,52,53,54], while 1:2 and 1:5 ratios are more common when testing large numbers of soil samples. Therefore, all things considered, soil samples were used in this study to prepare a soil–water 1:5 soil suspension. The concentrations of CO₃²⁻ and HCO₃⁻ were determined by dual indicator titration; Cl⁻ was measured using the AgNO₃ method; SO₄²⁻, Ca²⁺, and Mg²⁺ were determined using EDTA; and Na⁺ + K⁺ were measured using the subtraction method. TSS was calculated as the sum of cations and anions [55]. Meanwhile, in this study, saline soil was categorized with reference to the criteria of the Saline and Alkaline Land Specialized Committee of the Soil Society of China [56]. The standards are as follows: when Cl⁻/SO₄²⁻ ≥ 2, it is classified as a chloride type; when 1 ≤ Cl⁻/SO₄²⁻ < 2, it is classified as a sulfate–chloride type; when 0.2 ≤ Cl⁻/SO₄²⁻ < 1, it is classified as a chloride–sulfate type; and when Cl⁻/SO₄²⁻ < 0.2, it is classified as a sulfate type.

2.2.3. Soil Fractal Modeling

The fractal dimension of soil (D_soil) particle size can accurately characterize the distribution characteristics of soil particles and the uniformity of soil texture. Therefore, in this study, D_soil was calculated based on the single fractal theory model of soil particle volume derived [57]. The expression formula is as follows:

$$3-\mathrm{D}=\lg \left({\displaystyle \frac{V_{r<R_i}}{V_T}}\right)/\lg \left({\displaystyle \frac{R_i}{R_{max}}}\right)$$

(1)

In the formula, D is the fractal dimension; r is the soil particle size; V (r < R_i) is the cumulative volume of particles smaller than a certain size; V_T is the total volume of soil particles; and R_max is the upper limit for all particle sizes, numerically the maximum particle size. To calculate, first determine the slope of the least squares fitted line of $\lg \left({\displaystyle \frac{V_{r<R_i}}{V_T}}\right)$ and $\lg \left({\displaystyle \frac{R_i}{R_{max}}}\right)$, which equals 3–D, and then derive the fractal dimension D.

2.2.4. Inverse Distance Weighting

The inverse distance weighting (IDW) interpolation method is primarily based on the first law of geography, which determines the value of the interpolation point based on the inverse of the distance between the interpolation point and the sample points [58]. In other words, the farther the interpolation point is from the sample points, the smaller the influence, and vice versa. The main advantage of this method is its high computational efficiency and simple structure. The formula is as follows:

$$\widehat{z}\left(s_0\right)={\displaystyle \sum }_{i=1}^N\left({\lambda }_i\right)z\left(s_i\right)$$

(2)

$${\lambda }_i={\displaystyle \frac{1}{{\left(D_i\right)}^p}}{\left[{\displaystyle \sum }_{i=1}^N{\displaystyle \frac{1}{{\left(D_i\right)}^p}}\right]}^{-1}$$

(3)

In the formula, $\widehat{z}\left(s_0\right)$ represents the estimated attribute value at the point $s_0$, N is the sample size, ${\lambda }_i$ is the weight, $z\left(s_i\right)$ is the attribute value at the sample point $s_i$, D_i is the distance between the sample point at point i and the estimation point, and p is the power exponent, typically defined as 2.

2.2.5. Ordinary Kriging

The essence of the kriging method is to use the observed data of the regionalized variables at the sampling points to perform an unbiased estimation of the data at the interpolation point, with the most commonly used being the ordinary kriging (OK) method [59] where the data at an interpolation point are estimated as a linear combination of the observed values from the sample points. The general formula is as follows:

$$Z\left(x_0\right)={\displaystyle \sum }_{i=1}^N{\lambda }_{1}Z\left(x_i\right)$$

(4)

In the equation, $Z\left(x_0\right)$ represents the value of the predicted unknown point; $Z\left(x_i\right)$ represents the value of the known sample points surrounding the predicted unknown sample point; N is the number of known sample points; and λ is the weight of the i-th sample point.

2.2.6. Statistical and Geostatistical Analysis

Descriptive statistical analysis was conducted on salinity data and particle size data using SPSS 25.0, and the minimum, maximum, mean, standard deviation (SD), and coefficient of variation (CV) were determined for all datasets to analyze the vertical distribution characteristics of soil salinity and particle size. The coefficient of variation is an indicator used to measure the degree of data variation. Based on the CV values, variations are generally categorized as low (CV < 10%), medium (CV between 10% and 100%), and high (CV > 100%). Then, SPSS was used to perform Pearson correlation analysis, Origin 2021 was used to plot correlation heatmaps, and the relationship between soil salinity and particle size was investigated. Finally, redundancy analysis [60] was used to investigate the influencing factors of salinity.

3. Results

3.1. Characterization of Soil Salinity/Salt Ion Distribution

3.1.1. Vertical Distribution Characteristics of Total Soil Salinity

The statistical description of TSS at various soil depths within the study area is presented in

Table 1. Statistical characteristics of total soluble salt content/salt ions at different soil depths. Unit: mg/kg.

Soil Depth (cm)	Statistic	CO32−	HCO3−	Cl−	SO42−	Ca2+	Mg2+	Na+ + K+	TSS
0–10	Min	0	24	71	960	120	100	700	7805
	Max	340	900	77,810	40,220	4850	5190	51,350	143,790
	Mean	86	353	18,533	10,458	2182	1189	13,604	46,406
	SD	114	182	21,035	5349	1259	1382	12,929	35,877
	CV	132	51	114	51	58	116	95	77
10–20	Min	0	50	70	2300	100	80	810	4490
	Max	380	710	43,040	19,100	3830	5870	24,940	88,400
	Mean	60	330	6100	9500	1560	620	6310	24,500
	SD	100	160	8230	2590	1100	870	4930	15,090
	CV	163	47	135	27	71	139	78	62
20–40	Min	0	50	70	1820	100	120	30	3930
	Max	430	970	24,990	12,380	3190	1710	17,270	54,290
	Mean	70	330	4480	7730	1210	520	4910	19,250
	SD	110	170	5520	2760	1110	400	3850	11,080
	CV	155	50	123	36	91	78	79	58
40–60	Min	0	20	70	1340	110	30	320	3970
	Max	310	780	26,940	14,210	3230	1850	19,370	56,710
	Mean	70	300	4040	7980	1030	530	4940	18,880
	SD	100	130	5520	2540	1100	380	3640	10,450
	CV	137	44	137	32	107	72	74	55

. The overall salt content in the study area was relatively high, with variations in TSS at different soil depths. TSS decreases with increasing soil depth, with the maximum TSS value at the soil surface (0–10 cm), reaching 143,790 mg/kg (Figure 3). The surface layer salt accumulation in the study area was evident. The coefficient of variation (CV) of the surface layer was 77%, significantly higher than that of other soil layers. As shown in Table 1, the salinity data at different soil depths all fall within the medium variability range. The CV decreases with increasing soil depth, in the order of 0–10 > 10–20 > 20–40 > 40–60 cm.

The relationship of TSS at different soil depths was investigated based on Pearson’s correlation analysis, and as shown in

Table 2. Relationship of soil salinity between different depths based on Pearson correlation analysis.

Soil Depth (cm)	0–10	10–20	20–40	40–60
0–10	1
10–20	0.611 **	1
20–40	0.600 **	0.669 **	1
40–60	0.579 **	0.629 **	0.843 **	1

** Correlation is significant at 0.01 level (two-tailed).

, the TSS between different soil layers showed a significant correlation, with the correlation coefficients ranging from 0.579 to 0.843. The correlation coefficients between adjacent soil layers, i.e., 0–10 and 10–20, 10–20 and 20–40, and 20–40 and 40–60, were 0.611, 0.669, and 0.843, respectively. The correlations were all significant at the 0.01 level, which indicated that the soil salinity at a particular depth was determined by the salinity values of the neighboring soil layers. Therefore, this study quantitatively investigated the relationship of different soil depths and TSS by establishing multiple linear regression equations. The regression equations were shown as a, b, c, and d. Analysis of variance and significance testing of regression coefficients were conducted for the established regression equations; the results indicated significant correlation in the regression equations.

$$\begin{gathered} T{S}_{0-10}=0.854T{S}_{10-20}+0.693T{S}_{20-40}+0.592T{S}_{40-60}+0.976 \\ {R}^2=0.448 F=12.421 Sig=0.00 \end{gathered}$$

(5)

$$\begin{gathered} T{S}_{10-20}=0.131T{S}_{0-10}+0.487T{S}_{20-40}+0.214T{S}_{40-60}+5.017 \\ {R}^2=0.552 F=16.742 Sig=0.00 \end{gathered}$$

(6)

$$\begin{gathered} T{S}_{20-40}=0.030T{S}_{0-10}+0.138T{S}_{10-20}+0.709T{S}_{40-60}+1.082 \\ {R}^2=0.748 F=45.528 Sig=0.00 \end{gathered}$$

(7)

$$\begin{gathered} T{S}_{40-60}=0.025T{S}_{0-10}+0.059T{S}_{10-20}+0.692T{S}_{20-40}+2.933 \\ {R}^2=0.723 F=40.049 Sig=0.00 \end{gathered}$$

(8)

3.1.2. Characteristics of Vertical Distribution of Salt Ions and Major Saline Soil Types

As can be seen in Figure 4, in conjunction with the depth of soil sampling, the overall salt ion content was highest in the surface layer, second highest in the second layer, and lowest in the fourth layer. The anions were Cl⁻ > SO₄²⁻ > HCO₃⁻ > CO₃²⁻ and the cations were Na⁺ + K⁺ > Ca²⁺ > Mg²⁺ in the soil surface. This showed that the major anions in the study region were Cl⁻ and SO₄²⁻ and the major cations were dominated by Na⁺ + K⁺. The coefficients of variation in Cl⁻, CO₃²⁻, and Mg²⁺ among all the ions were more than 100%, which were highly variable ions, while the other ions were of medium variation.

Correlation analysis between soil salt ions can help to reveal the synergistic migration of salt ions in soil [61]; therefore, this study was conducted using a Pearson correlation analysis between TSS and salt ions using SPSS 25.0. From the surface to the deeper layers, the correlation between soil salinity and salt ions showed significant differences. As shown in Figure 5, TSS showed non-significant correlation with both HCO₃⁻ in 0–60 cm soil layer; non-significant correlation with CO₃²⁻ in 20–40,40–60, and significant correlation in this soil layer; and significant correlation with SO₄²⁻ in all but 0–10 cm soil layer. In addition to this, TSS showed significant correlation with all other ions, with the largest correlation coefficients with Cl⁻, Na⁺ + K⁺, all of which had correlation coefficients above 0.9. It indicates that soil salinity is mainly associated with Cl⁻, SO₄²⁻, CO₃²⁻, Ca²⁺, Mg²⁺, Na⁺ + K⁺. From the results of the correlation of salt ions with each other, Cl⁻ and Na⁺ + K⁺ showed highly significant correlation with correlation coefficients above 0.9 in all soil layers. In addition, in 0–10 cm CO₃²⁻ vs. Na⁺ + K⁺; Cl⁻ vs. Ca²⁺; in 10–20 cm CO₃²⁻ vs. Mg²⁺; Cl⁻ vs. Mg²⁺; in 20–40 cm CO₃²⁻ vs. Ca²⁺; SO₄^2– with Ca²⁺; SO₄²⁻ with Na⁺ + K⁺; and CO₃²⁻ with Ca²⁺; Cl⁻ with Mg²⁺; SO₄²⁻ with Na⁺ + K⁺ in 40–60 cm showed highly significant correlation. Overall, the correlations between TSS and Cl⁻ and Na⁺ + K⁺, as well as between Cl⁻ and Na⁺ + K⁺, were high, suggesting that Na⁺ + K⁺ and Cl⁻ have the strongest synergistic migration ability.

In order to further explore the main types of salts in each soil layer, in this study, we classified salinized soils according to the standards of the Saline Soil Specialized Committee of the Soil Society of China. According to the results in

Table 3. Classification of saline soil types and degree of salinization in Ebinur Lake wetland.

Soil Depth (cm)	2 ≤ Cl−/SO42− (%) Chloride	1 ≤ Cl−/SO42− < 2(%) Sulfate–Chloride	0.2 ≤ Cl−/SO42− < 221(%) Chloride–Sulfate	Cl−/SO42− < 20.2(%) Sulfate
0–10	36.00	12.00	24.00	28.00
10–20	4.00	18.00	38.00	40.00
20–40	4.00	12.00	44.00	40.00
40–60	2.00	16.00	40.00	42.00
Grading of soil salinization	Chloride	Sulfate–chloride	Chloride–sulfate	Sulfate
Non-saline Criteria Value	TSS0–10 < 1500	TSS0–10 < 2000	TSS0–10 < 2500	TSS0–10 < 3000
Light salinized Criteria Value	1500 < TSS0–10 < 3000	2000 < TSS0–10 < 3000	2500 < TSS0–10 < 4000	3000 < TSS0–10 < 6000
Moderate salinized Criteria Value	3000 < TSS0–10 < 5000	3000 < TSS0–10 < 6000	4000 < TSS0–10 < 7000 2%	6000 < TSS0–10 < 10,000 2%
Heavy salinized Criteria Value	5000 < TSS0–10 < 8000	6000 < TSS0–10 < 10,000	7000 < TSS0–10 < 12,000 4%	10,000 < TSS0–10 < 20,000 20%
Salinized soil Criteria	TSS0–10 ≥ 8000 36%	TSS0–10 ≥ 10,000 12%	TSS0–10 ≥ 12,000 20%	TSS0–10 ≥ 20,000 6%

The percentages in the table are ratios of the corresponding sampling site to the total sampling site TSS, and the subscripts 0–10 indicate the total soluble salt (mg/kg) at soil depths of 0–10 cm. And Bold font indicates table header.

, the degree of soil salinity in the study area was relatively high, with Cl⁻ content in the surface layer accounting for approximately 36% of the total ion content. However, the proportion significantly decreases at other soil depths. Therefore, the surface saline soil was mainly of the chloride type, while other soil layers predominantly exhibited chloride–sulfate and sulfate types. Moderately saline soils, heavily saline soils, and saline soils in the study area accounted for 4, 24, and 74%, respectively, indicating that the study area had a high degree of salinization, which was mainly dominated by chloride-type salinization, with good a solubility of chloride.

3.1.3. Characterization of Spatial Distribution of Soil Salinity

Geostatistical analyses play a critical role in decision-making processes related to environmental remediation, monitoring, and land management [62]. Various interpolation methods were employed to predict the spatial distribution of soil salinity, including IDW and OK [63]. However, different interpolation methods often yield varying results. To achieve a more accurate understanding of the spatial distribution characteristics of TSS and salt ions, this study compared two widely used interpolation methods for soil salinity. The objective is to identify the most suitable interpolation method for the study area. Salinity data were interpolated in ArcGIS 10.6 using IDW and OK, and 80% (40 samples) were used as the training set, and 20% (10 samples) were used as the testing set. The accuracy and reliability of spatial interpolation were evaluated by three metrics: R², RMSE, and MAE. As shown in

Table 4. Validation and comparison of 0–10 cm IDW and OK interpolation methods. Unit: 10³ mg/kg.

Variable	Method	R2	RMSE	MAE
TSS	OK	0.43	7.31	6.36
	IDW	0.56	6.71	5.31
HCO3−	OK	0.35	0.04	0.03
	IDW	0.34	0.03	0.03
Cl−	OK	0.40	3.99	3.43
	IDW	0.68	3.09	2.53
SO42−	OK	0.78	2.08	1.37
	IDW	0.80	1.31	0.88
Ca2+	OK	0.43	0.26	0.21
	IDW	0.59	0.24	0.21
Mg2+	OK	0.53	0.26	0.21
	IDW	0.59	0.24	0.20
Na+ + K+	OK	0.54	2.50	2.16
	IDW	0.62	1.45	1.84
Dsoil	OK	0.44	0.04	0.04
	IDW	0.50	0.05	0.04
Clay	OK	0.55	2.20	1.77
	IDW	0.71	2.10	1.68
Silt	OK	0.53	19.19	16.15
	IDW	0.53	18.85	15.67
Sand	OK	0.43	21.28	16.15
	IDW	0.53	20.81	17.03

, two spatial interpolation methods exhibited significant differences. Except for the HCO₃⁻, the R², RMSE, and MAE of other ions were superior to those of OK, indicating that IDW achieved better fitting performance for the test data than OK. Overall, the IDW interpolation method exhibited smaller errors, higher accuracy, and greater stability, making it more suitable for salinity interpolation in the Ebinur Lake wetland compared to OK. It can more effectively reveal the spatial distribution of soil salinity and soil particle size. Furthermore, based on interpolation methods employed by other researchers in the Ebinur Lake wetland [64], this study selected IDW as the spatial interpolation method.

In this paper, the spatial distribution maps of TSS and each ion from 0 to 10 cm were obtained by spatial interpolation using IDW. As shown in Figure 6, the spatial distribution characteristics of TSS and salt ions are more similar, and the overall distribution is more dispersed. The high value areas of TSS were mainly distributed in the periphery of Lake Ebey, such as the northwest and southwest sides of the lake, as well as the southeast side of the lake. The spatial distributions of TSS were roughly the same as those of Cl⁻, SO₄²⁻, and Na⁺ + K⁺; and the trend of spatial distribution was opposite to that of HCO₃⁻. And from the comparison between ions, it can be seen that the spatial distribution of Cl⁻ and Na⁺ + K⁺ is extremely similar; contrary to the trend of the spatial distribution of HCO₃⁻, the spatial distribution of SO₄²⁻ is more similar to that of Ca²⁺ and Mg²⁺.

3.2. Statistical Description and Spatial Distribution Characteristics of Soil Particle Size

Descriptive statistics were analyzed for PSD at different soil depths in the Ebinur Lake wetland, and the results are shown in

Table 5. Statistical characteristics of soil particle size in each layer.

Soil Depth (cm)	Minimum	Maximum	Average Value	Standard Deviation	Coefficient of Variation	Skewness	Kurtosis
Clay (%)
0–10	0.10	11.26	5.64	2.48	0.44	0.49	0.11
10–20	0.11	13.98	5.07	2.99	0.59	0.78	0.80
20–40	0.03	16.99	5.04	3.64	0.72	1.08	1.40
40–60	0.15	18.00	5.23	3.71	0.71	1.12	1.63
Silt (%)
0–10	2.92	77.24	45.32	19.18	0.42	−0.33	−0.73
10–20	2.52	78.68	39.47	20.96	0.53	0.02	−1.11
20–40	1.94	79.41	39.51	22.97	0.58	−0.03	−1.28
40–60	2.35	80.82	40.64	21.90	0.54	−0.09	−1.10
Sand (%)
0–10	11.50	96.19	49.04	21.30	0.43	0.27	−0.64
10–20	9.88	96.77	55.46	23.48	0.42	−0.06	−0.96
20–40	9.23	97.60	55.45	25.90	0.47	−0.01	−1.18
40–60	8.78	96.94	54.12	24.79	0.46	0.04	−1.03

. The soil particles at different depths in the study area were primarily composed of silt and sand, with the maximum volume fraction of sand reaching 97.6% and the minimum volume fraction of clay being 0.03%. At the soil surface, sand was the most common category (49.04%), followed by silt (45.32%) and clay (5.64%). At other soil depths, the silt content ranges from 39.47 to 40.64%, while the sand content ranges from 54.12 to 55.46%. From the perspective of variability, the variability in clay content was the most pronounced, followed by silt, with sand showing the least variability. A triangular classification diagram of the top 50 samples was created according to the USDA soil texture classification (Figure S1). The surface soil texture mainly consists of sandy soil, sandy loam, loamy sand, and silty loam. The soil textures at other depths generally include sandy soil, sandy loam, loamy sand, loam, silty loam, and silt loam (Table S1). The characteristics of these soil types include a high sand content, low clay and silt content, good permeability, and poor water retention [65].

In reality, the distribution of soil types is complex and intricate, making it difficult to determine the distribution characteristics and texture uniformity of soil particles in the study area. Therefore, this study used the measured particle size results and the IDW interpolation method to obtain the spatial distribution of surface layer PSD and D_soil in the Ebinur Lake wetland (Figure 7). Regarding surface layer PSD, the volume fraction of clay particles was generally relatively low. It was mainly distributed in the northwest corner of Ebinur Lake, the narrow strip on the southwest side of Ebinur Lake (mainly referring to the lower reaches of the Jing River riparian zone), and the northeast side (mainly with some high value points in the lower reaches of the Kuitun River). The spatial distribution trend of silt content was similar to that of clay particles. In contrast, the spatial distribution of the sand particles was exactly opposite to that of the clay particles, with lower volume fractions of sand in some areas on the northwest, southwest, and northeast sides of Ebinur Lake. Generally, the spatial distribution trend showed a higher sand content on the southeast side of the lake compared to the northwest side.

The D_soil at different soil depths in the study area ranged from 1.91 to 2.76 with a mean value of 2.54. The spatial distribution of D_soil was extremely similar to that of clay and silt. At the same time, the spatial distribution of D_soil at different soil depths remained generally consistent. Overall, D_soil exhibited an increase with increasing content of clay, silt; a decrease with decreasing sand content; and an increase with increasing content of TSS (Figure S2).

3.3. Correlation Analysis Between Soil Salts/Ions and Soil Particle Size

3.3.1. Correlation Between Soil Salts/Ions and Soil Particle Size

To accurately reflect the relationship between soil fractal dimension and soil salinization, this study first averaged the TSS and D_soil of all soil depths in the study area to analyze the spatial distribution characteristics of soil fractal dimension and salinity. Secondly, Pearson correlation analysis was performed on TSS, D_soil, and salt ions to analyze the relationship between PSD and soil salinization at soil depth from 0 to 60 cm. As shown in Figure 8a, the trend line of D_soil in the green line at the Ebinur Lake wetland was concave, showing a decreasing-then-increasing trend overall, with a general decreasing trend in the blue line. The trend line of TSS in the green line was also concave, showing a decreasing-then-increasing pattern, while the blue line showed a linear decreasing trend (Figure 8b). Overall, D_soil and TSS follow the same trend.

Table 6. Pearson’s correlation analysis of salts/salt ions and soil particle size in Ebinur Lake wetland.

	Dsoil		CLAY		SILT		SAND
0–10	R	Sig	R	Sig	R	Sig	R	Sig
TSS	0.420 **	0.002	0.582 **	0.000	0.440 **	0.001	−0.464 **	0.001
CO32−	0.260	0.068	0.327 *	0.020	0.295 *	0.038	−0.304 *	0.032
HCO3−	−0.073	0.617	−0.059	0.683	−0.076	0.598	0.076	0.602
Cl−	0.395 **	0.004	0.568 **	0.000	0.410 **	0.003	−0.435 **	0.002
SO42−	0.171	0.234	0.142	0.326	0.201	0.162	−0.197	0.170
Ca2+	0.292 *	0.040	0.174	0.227	0.288 *	0.043	−0.279 *	0.050
Mg2+	0.182	0.206	0.198	0.168	0.283 *	0.046	−0.278	0.051
Na+ + K+	0.402 **	0.004	0.590 **	0.000	0.413 **	0.003	−0.441 **	0.001
10–20
TSS	0.448 **	0.001	0.550 **	0.000	0.418 **	0.002	−0.444 **	0.001
CO32−	0.233	0.103	0.295 *	0.037	0.207	0.149	−0.222	0.121
HCO3−	0.015	0.920	−0.052	0.719	0.108	0.453	−0.090	0.533
Cl−	0.447 **	0.001	0.568 **	0.000	0.450 **	0.001	−0.474 **	0.001
SO42−	0.182	0.207	0.154	0.285	0.046	0.753	−0.060	0.677
Ca2+	0.357 *	0.011	0.316 *	0.025	0.280 *	0.049	−0.290 *	0.041
Mg2+	0.266	0.062	0.349 *	0.013	0.342 *	0.015	−0.350 *	0.013
Na+ + K+	0.398 **	0.004	0.519 **	0.000	0.376 **	0.007	−0.401 **	0.004
20–40
TSS	0.493 **	0.000	0.703 **	0.000	0.476 **	0.000	−0.521 **	0.000
CO32−	−0.083	0.565	−0.170	0.238	−0.205	0.154	0.205	0.152
HCO3−	−0.001	0.992	−0.119	0.410	0.069	0.632	−0.045	0.757
Cl−	0.485 **	0.000	0.728 **	0.000	0.534 **	0.000	−0.576 **	0.000
SO42−	0.274	0.054	0.315 *	0.026	0.120	0.407	−0.150	0.297
Ca2+	0.375 **	0.007	0.432 **	0.002	0.362 **	0.010	−0.382 **	0.006
Mg2+	0.265	0.063	0.267	0.060	0.318 *	0.024	−0.319 *	0.024
Na+ + K+	0.393 **	0.005	0.609 **	0.000	0.383 **	0.006	−0.425 **	0.002
40–60
TSS	0.386 **	0.006	0.607 **	0.000	0.421 **	0.002	−0.462 **	0.001
CO32−	−0.198	0.168	−0.322 *	0.023	−0.323 *	0.022	0.334 *	0.018
HCO3−	0.120	0.406	0.116	0.423	0.175	0.224	−0.172	0.233
Cl−	0.458 **	0.001	0.685 **	0.000	0.494 **	0.000	−0.539 **	0.000
SO42−	−0.020	0.891	0.053	0.715	−0.006	0.965	−0.002	0.988
Ca2+	0.401 **	0.004	0.421 **	0.002	0.425 **	0.002	−0.438 **	0.001
Mg2+	0.378 **	0.007	0.453 **	0.001	0.538 **	0.000	−0.543 **	0.000
Na+ + K+	0.265	0.062	0.496 **	0.000	0.280 *	0.049	−0.321 *	0.023

showed that TSS had a significant positive correlation with D_soil, clay, and silt, and a negative correlation with sand. The strength of the correlation was TSS-Clay > TSS-Silt > TSS-Sand. In terms of salt ions correlating with each other, CO₃²⁻, HCO₃⁻, and SO₄²⁻ were significantly uncorrelated with D_soil in all layers, whereas Cl⁻ was the opposite, showing a high correlation. Ca²⁺ and D_soil among cations showed a gradual increase in correlation with increasing soil depth, while Na⁺ + K⁺ changed in the opposite trend. The trend of the correlation between Mg²⁺ and D_soil was mainly “S” shaped. In summary, it showed that Cl⁻ and Na⁺ + K⁺ were the overwhelmingly dominant ions in the migration of ions in the bottom–up process. Clay showed non-significant negative correlation with HCO₃⁻ in all soil layers; negative correlation with CO₃²⁻ in the 20–40 cm layer; except for this, clay was positively correlated with all other ions. Silt was the same as clay. In sand, CO₃²⁻ and HCO₃⁻ showed a positive correlation with sand except in some soil layers, which was the opposite of the case for both particle sizes of clay/silt, and all the rest of the ions showed a negative correlation with sand.

3.3.2. Percentage of Total Salts/Salt Ions in Different Soil Textures

To further explore the relationship between TSS and PSD, soil texture types at different depths were classified according to the USDA soil texture classification standards, and the relationship between soil salinity and particle size distribution in different soil textures was examined. According to

Table 7. Statistical analysis of soil salinity and soil particle size under different soil texture types.

Soil Texture Type	Layers (cm)	Samples	Clay (%)	Silt (%)	Sand (%)	Dsoil	Average Salt Content (mg/kg)
Sandy soil; loamy sandy soil	0–10	14	3.61	26.01	70.39	2.49	35,730
	10–20	11	2.19	16.78	81.03	2.40	14,950
	20–40	17	1.80	15.27	82.93	2.36	13,140
	40–60	13	1.76	14.37	83.88	2.36	15,200
Sandy loam; loam	0–10	28	6.22	50.85	42.93	2.59	43,490
	10–20	19	4.49	32.24	63.28	2.53	26,770
	20–40	14	5.52	36.34	58.14	2.57	19,780
	40–60	17	4.88	36.27	58.85	2.55	16,630
Silt; silty loam	0–10	8	7.17	60.58	32.24	2.62	68,930
	10–20	20	7.07	57.25	35.68	2.61	27,640
	20–40	19	7.58	63.53	28.88	2.62	24,311
	40–60	20	7.78	60.98	31.25	2.62	23,071

, the number of samples for sandy soil and loamy sand, sandy loam and loam, and silty loam and silt in the study area were 55, 78, and 67, respectively, with the total proportion of sandy loam and silty loam and silt reaching 72.5%. The lowest TSS appears in sandy soil and loamy sand, at 13,140 mg/kg, while the highest value appears in silty loam and silt, at 68,930 mg/kg, with the overall TSS of sandy soil and loamy sand being lower than that of the other soil types.

Figure 9 shows the differences in the contents of TSS and salt ions in different soil textures. CO₃²⁻ was excluded from the analytical section below due to the lack of CO₃²⁻ in most of the samples, resulting in low CO₃²⁻ levels and large errors. Overall, TSS and salt ions were highest in the 0–10 cm, with most ions showing a decrease in content with increasing soil depth. TSS, Cl⁻, SO₄²⁻, Ca²⁺, Mg²⁺, Na⁺ + K⁺ content increased with finer soil texture in different soil textures, except for HCO₃⁻ ions. In 0–20 cm, TSS, Cl⁻, Mg²⁺, and Na⁺ + K⁺ were more variable, with reductions ranging from 53 to 75%, and in 20–60 cm, the variations leveled off, with variations ranging from 0.5 to 14.5%. From 0 to 60 cm, the reduction of HCO₃⁻, SO₄²⁻, and Ca²⁺ varied between 0.1 and 36%, with small variations. HCO₃⁻ and SO₄²⁻ content varied little in different soil textures, with an overall “S” pattern. In summary, the average TSS content of the soil gradually increased as the soil texture became finer and D_soil increased. It also showed that there were significant differences in the migration of different salt ions across the soil profile.

4. Discussion

4.1. Analysis of Distribution Characteristics of Salinity and Salt Ions

The accumulation of surface soil salinity in the study area was prominent, and there was an interrelationship between the salinity levels at different soil depths. This was consistent with the findings in the Pariette Draw wetland in the United States, where capillary water movement caused soil salinity accumulation in the surface layer of wetland soils [66]. In the Ebinur Lake wetlands, soil salinity is strongly influenced by regional climate and topography. Redundancy analysis results clearly indicated that topographic and climatic factors had a significant impact on salt accumulation (Figure S3, Table S2). And freshwater wetlands located in dryland environments are characterized by high evapotranspiration rates and frequent periods of desiccation [67]. High temperatures and strong evapotranspiration during seasonal peaks cause shallow groundwater to migrate continuously to the surface, carrying salt-rich water that evaporates at the surface [68]. With a reduction in water content, the evaporation rate gradually declines, and the temperature decreases with increasing soil depth. The evaporation front is unable to move downward due to groundwater replenishment, and deeper soils are less influenced by evaporation [69]. Analyzing the Cl⁻/SO₄²⁻ equivalent ratios revealed asynchronous migration rates of different salt ions in the soil profile, with Cl⁻, which was highly mobile during upward migration, moving quickly through the soil, while SO₄²⁻, having low solubility, migrated relatively slowly. Additionally, the correlation between Cl⁻ and Na⁺ + K⁺ was the strongest in the soil profile, indicating synergistic transport of these ions, likely due to differences in their solubility, mobility, and microtopographic distribution [70]. Zhang et al. also showed that the water–salt migration changes in Ebinur Lake were influenced by microgeomorphology, soil texture, and other factors [71]. Therefore, during the upward mobilization of labile salts in the study area, chloride upward epimerization was the strongest, followed by sulfate. Additionally, human activities have played an indirect role in promoting the accumulation of soil salinity [72]. Ainur indicated that salt content increased significantly from west to east in the Bortala River Basin, with the most severe salinization occurring in the Ebinur Lake wetland at the eastern end of the basin, suggesting irrigation’s impact on salt accumulation and migration [73]. Excessive agricultural irrigation has reduced downstream surface runoff, even causing streamflow cessation, leading to the shrinkage of the Ebinur Lake area [47]. Under intense wind erosion in Alashankou, the exposed dried lakebed becomes one of the primary sources of saline dust storms [40]. This secondary landscape results from hydrological imbalances, reduced saline lake areas, the evaporation of saline lake waters, and salt accumulation on dried lakebeds [74]. Similar issues of saline dust storms were observed in the Aral Sea, driven by declining water levels and drought [75].

4.2. Fractal Dimension of Particle Size Distribution

Fractal dimension is an important parameter commonly used to characterize PSD [76]. Previous studies have shown that finer-textured, nutrient-rich soils generally have fractal dimensions of 2.60 to 2.80, and coarser-grained, less structured soils have fractal dimensions of 1.83 to 2.6453 [77]. Furthermore, Liu et al. indicated that well-structured soils tend to have a fractal dimension of approximately 2.75 [78]. In this study, the fractal dimension ranges from 1.91 to 2.76, with an average value of 2.54. Most of the D_soil values in this study were less than 2.75, indicating poor soil structure and a high sand content in the study area. According to Table 7, the combined proportion of sandy loam, silty loam, and silt in the soil profile reached 72.5%, indicating dominance by sandy and loamy soils with coarse texture and poor structure, which are unfavorable for retaining soil moisture and nutrients, resulting in relatively low soil fertility. In the soil profile, D_soil showed highly significant correlations with both soil particle size (Figure S2), highly significant negative correlations with sand (p < 0.01), and significant positive correlations with silt and clay (p < 0.01), with correlations above 0.8. The results confirmed that the finer the particle size, the greater the soil fractal dimension. These correlations indicated the usefulness of fractal dimension to characterize the PSD of saline soils.

4.3. Analysis of Relationship Between Soil Salinity and Soil Particle Size

Numerous studies to date have demonstrated a significant relationship between soil particle size and soil moisture, nutrients, and soil organic matter [39,79]. Soil texture and its vertical spatial heterogeneity may greatly influence the distribution of water and solutes in the soil profile [38], which is one of the important factors affecting salt ion retention [80]. The results of this study showed that TSS was positively correlated with D_soil, clay, and silt and negatively correlated with sand. At the same time, there were differences in salinity among different soil textures, mainly in the form of an increasing trend in TSS and D_soil with the increase in the fine-grained fraction. This may be due to the strong evaporation from the surface in summer [81], and the fact that soil texture in the vertical profile is dominated by sandy loam, loam, silty loam, and silt, and that sandy loam to silty loam have moderate sized capillary porosity compared to sandy loam and clay, and that groundwater rises through the soil capillaries at a faster rate and at a higher altitude [82], making the soils more susceptible to salinization. Meantime, areas with a higher distribution of sand exhibit higher permeability due to larger particles and porosity, resulting in a weaker salt-holding and adsorption capacity as salts are more likely to migrate deeper with water [83]. There were differences in the strength of correlations between D_soil and the various salt ions at different soil depths, with D_soil being significantly positively correlated with Cl⁻ and Na⁺ + K⁺ in the soil profile, and cations generally having higher correlations with D_soil than anions. Furthermore, Chen et al. modified the soil particle size composition and texture by varying the content of shell sand, revealing that the leaching rate of salt ions differed with varying shell sand contents [84]. This indicated that the PSD of the soil profile might be a driving factor affecting the migration of different soluble cations and anions [85], resulting in an increased Cl⁻/SO₄²⁻ equivalent ratio from the bottom to the top of the vertical profile. Due to the seasonal nature of wetlands, NaCl was eventually lost through surface evaporation and accumulated on the soil surface. The impact of excessive soil salinity on wetland ecosystems is destructive and long-lasting. The relationship between soil salinity and soil particle size presented in the study area provides a basis for understanding the effect of soil particle size on soil salinity and acts as a reference point for wetlands with similar environments. In addition to the results of this study, Wang et al. indicated that controlling the lake area of Ebinur Lake had a positive effect on the ecological restoration of the wetland [47].

5. Conclusions

In this study, soil profile samples of 0–60 cm from Ebinur Lake wetland were analyzed by combining Pearson correlation, IDW spatial interpolation, and redundancy analysis to explore the relationship between soil particle size and soil salinity, and to reveal the characteristics of vertical distribution of soil salinity and its influencing factors. The results indicated that the Ebinur Lake wetland was severely salinized, with pronounced surface soil salinity accumulation. The dominant salinization type was chloride, with Cl⁻ and Na⁺ + K⁺ playing major roles in the salinization process. TSS were significantly and positively correlated with Dsoil. Salt content varied among soil textures, with more pronounced salt accumulation in finer soil particles, particularly in sandy loam and silt soils. This study demonstrated that PSD is closely related to soil salinity, while climate, topography, and geomorphology also significantly influenced soil salt accumulation in the Ebinur Lake wetland. The findings are expected to provide valuable insights into wetland soil management and ecological restoration.