Spatial and Temporal Variations of Soil pH in Farmland in Xinjiang, China over the Past Decade

Abstract

Soil pH is crucial for the quality of the farmland and crop growth. The objective of this study is to analyze the spatial and temporal variations of farmland soil pH in Xinjiang (XJ), and to provide a scientific basis for soil improvement and agricultural production optimization. Based on soil pH data from XJ farmland in two periods, 2008~2010 and 2019~2021, geostatistical methods and kriging spatial interpolation techniques were employed to analyze the spatiotemporal changes in soil pH and to investigate the relationship between relevant influencing factors and pH over more than a decade. The results indicate that the spatiotemporal distribution of soil pH in XJ farmlands is uneven. Between 2019 and 2021, the average soil pH was 8.11, generally being on the higher side, with a coefficient of variation (CV) of 3.85%, indicating minimal spatial variability. In the farmland soil of Southern Xinjiang (S-XJ), the average pH value was 8.17, significantly higher than that of Northern Xinjiang (N-XJ), which was 8.10, demonstrating a spatial pattern of “higher in the south, lower in the north”. Over the past decade, soil pH in XJ has significantly increased from 8.11 to 8.13, with a 0.05 unit increase in the north and a 0.01 unit increase in the south (p < 0.05). Regionally, Altay saw the largest increase of 0.22 units, while Bortala Mongol Autonomous Prefecture (Bortala) experienced the most significant decrease, dropping by 0.59 units. Furthermore, this study found that factors such as topography, nutrients, and irrigation methods all have certain influences on the spatial distribution of soil pH in XJ farmland, while variations in climate factors and fertilization levels may affect its long-term temporal changes. These research findings will provide new insights for adjusting and updating agricultural management measures related to soil pH regulation in XJ.

1. Introduction

The total area of farmland in China is approximately 1.411 × 10⁶ km², yet the per capita farmland area is less than 40% of the global average [1]. Moreover, the total area of farmland has been on a long-term decline [2], leading to tight farmland resources and reduced soil fertility. As China’s economy rapidly develops, the issue of declining farmland quality is becoming increasingly prominent. Soil pH is critical as it not only affects crops’ absorption of soil nutrients [3], but also influences the valence, migration, and transformation of metal elements in the soil [4], which are essential for crop growth and soil quality. A high soil pH can adversely affect the activity and community structure of soil microorganisms, inhibit crop root growth, and lead to issues such as soil compaction and reduced crop yields [5,6]. Conversely, a low soil pH decreases soil biodiversity and accelerates soil degradation [7]. Under natural conditions, changes in soil pH typically occur over extended periods. However, with the intensification of human activities, the natural evolution process of soil pH has accelerated, posing significant threats to China’s food security and the sustainable development of agriculture.

The traditional soil pH monitoring method is primarily based on locating points to obtain measured data for monitoring [8]. This method can achieve accurate monitoring results, but it is time-consuming and labor-intensive, making it difficult to conduct over large areas and extended time periods. In recent years, with the development of modern agriculture, the application of “3S” technology in the field of agriculture has become increasingly widespread. The use of geostatistics combined with “3S” technology to study the spatial and temporal variability of soil’s physical and chemical properties and their driving factors has gradually become a hot research topic [9,10,11,12,13,14]. Wu et al. [10] used various sampling methods and spatial interpolation techniques to select the optimal prediction method for the spatiotemporal variation of pH under different conditions; Yang et al. [11] analyzed the characteristics of pH changes in agricultural soils in China over the past 30 years based on geostatistics, revealing that the vast majority of soils in China have shown a tendency toward acidification over the past 30 years; Liu et al. [12] found that the pH of farmland soil in the Poyang Lake watershed significantly increased by 0.1 units from 2012 to 2018, and the pH changes were greatly influenced by rainfall; Li et al. [13] and Han et al. [14] both analyzed the spatial and temporal variation characteristics of farmland soil pH at the county level, utilizing a combination of geostatistics and GIS technology.

Xinjiang (XJ) is a significant agricultural production base in the western part of China and is also one of the main distribution areas for saline–alkali farmland in the country. Saline–alkali farmland covers more than one-third of the total farmland area in the region, and soil salinization is a serious issue [15]. With its vast territory and complex topography, XJ urgently needs to understand the spatiotemporal variability of soil pH in farmland, amidst the extensive use of chemical fertilizers and the ongoing innovations in farming and irrigation methods. Currently, research on soil pH in XJ farmland is predominantly focused on small-scale investigations, lacking broader-scale studies [16]. This study focuses on XJ as the research area, and utilizes geostatistical methods to investigate the spatiotemporal variations of farmland soil pH in XJ and their relationship with related influencing factors based on soil testing and formula fertilization data from 2008 to 2010 and from 2019 to 2021. Specifically, the objectives of this study are (1) to examine the spatial distribution and temporal changes in soil pH in XJ farmland during these two periods, and (2) to explore the impacts of different factors on soil pH in XJ farmland.

2. Materials and Methods

2.1. Overview of the Study Area

XJ (73°45′ E~96°39′ E, 34°34′ N~49°17′ N) is located in the heart of the Eurasian continent and on the northwest border of China (Figure 1). With a total area of about 1.66 × 10⁶ km², it is the largest provincial-level administrative region in China, comprising one-sixth of the country’s land area. The region has a complex terrain, with alternating mountains and basins surrounded by high mountains, forming a unique geographic structure known as ‘three mountains sandwiching two basins’. The Tianshan Mountains traverse central XJ, dividing the region into Northern Xinjiang (N-XJ) and Southern Xinjiang (S-XJ). N-XJ, situated north of the Tianshan Mountains, includes Altay, Bortala Mongol Autonomous Prefecture (Bortala), Changji Hui Autonomous Prefecture (Changji), Urumqi, Karamay, Ili Kazakh Autonomous Prefecture (Ili), Tacheng, Barkol Kazakh Autonomous County of Hami, and Yiwu County of Hami (collectively referred to as the northern part of Hami, abbreviated as ‘Hami _N-XJ’). S-XJ, south of the Tianshan Mountains, comprises Aksu, Turpan, Bayingolin Mongol Autonomous Prefecture (Bayingolin), Hotan, Kashgar, Kizilsu Kyrgyz Autonomous Prefecture (Kizilsu), and Yizhou District of Hami (the southern part of Hami, abbreviated as ‘Hami _S-XJ’). XJ predominantly has a temperate continental arid climate, with an average of 2836 h of sunshine per year [17], ensuring ample sunlight. The region’s average annual temperature ranges from −4 to 9 °C [18], while evaporation ranges between 1000 and 4000 mm, significantly exceeding other regions in China at the same latitude. Annual precipitation is relatively low, around 150 mm, with highly uneven spatial and temporal distribution. N-XJ receives about 150~200 mm of annual rainfall, while S-XJ receives less than 100 mm [19]. XJ, with its vast territory and abundant land resources, has a wide variety of soil types. In the basins, desert soils, saline soils, and cumulated irrigated soils formed by land reclamation, as well as fluvo-aquic soils, are predominantly found. In the mountainous regions, there are mostly Calcium layer soils, meadow soils, and Calcic gray-cinnamon soils [20]. Under the combined influence of soil parent material, climatic conditions, and farming management, among other natural and human factors, XJ has become one of the most severely affected areas of soil salinization in China.

2.2. Data Sources

The data on farmland soil pH, irrigation methods, and soil nutrients content (SOM, AN, AP, and AK) from 2008 to 2010 and from 2019 to 2021 involved in this study were obtained from the soil testing and fertilizer formulation records for the corresponding years, provided by the Soil and Fertilizer Station of the Xinjiang Uygur Autonomous Region. The sampling period was during the autumn of the corresponding years, after the crop harvest (from mid-September to November), which minimized the impact of crop growth and fertilization factors. The soil nutrients content was classified into levels 1 to 6 according to the Second National Soil Census Standard (

Table 1. Soil nutrient content grading standard of the second national soil census.

Soil Nutrients	Grade
	1	2	3	4	5	6
SOM (g/kg)	≥40	30~40	20~30	10~20	6~10	<6
AN (mg/kg)	≥150	120~150	90~120	60~90	30~60	<30
AP (mg/kg)	≥40	20~40	10~20	5~10	3~5	<3
AK (mg/kg)	≥200	150~200	100~150	50~100	30~50	<30

Note: SOM represents soil organic matter, AN, AP and AK respectively represent soil available N, available P and available K.

). After outlier removal, there were a total of 23,362 data points from 2008 to 2010, and 22,320 data points from 2019 to 2021 (Figure 2). Fertilizer application data (2008, 2012 and 2021) were sourced from the official website of the Bureau of Statistics of the XJ Uygur Autonomous Region (https://tjj.xinjiang.gov.cn/ (accessed on 25 May 2024)). Elevation data corresponding to the years of sampling are derived from NASA DEM data products for the Shuttle Radar Topography Mission (https://lpdaac.usgs.gov/products/nasadem_hgtv001/ (accessed on 25 May 2024)). Land use data (2010 and 2020) are from the Center for Resource and Environmental Science and Data, Chinese Academy of Sciences (https://www.resdc.cn/ (accessed on 25 May 2024)). Total annual precipitation (Pre) data from 2008 to 2021 are sourced from the CHIRPS precipitation dataset, while Total annual evapotranspiration (ET) and average annual land surface temperature (LST) data from 2008 to 2021 are obtained from MODIS products. All the aforementioned meteorological data were downloaded via the Google Earth Engine (https://code.earthengine.google.com/ (accessed on 25 May 2024)) platform and were resampled to 30 m × 30 m.

2.3. Data Analysis

Using Excel 2021 and SPSS 26.0 software, the soil pH data were subjected to normality test, conventional statistical analysis, ANOVA, and t-tests, and the soil pH data were transformed so that they approximately obeyed a normal distribution; the optimal semivariogram model was fitted to the soil pH of the XJ farmland using GS+ 9.0 software; spatial autocorrelation analysis was performed based on the global Moran’s I index using ArcGIS 10.8 software; spatial interpolation analysis and mapping was carried out using the spatial autocorrelation analysis based on the global Moran’s I index using ArcGIS 10.8 software; and the spatial interpolation analysis and mapping were performed using the ordinary kriging interpolation method of the Geostatistical Analyst module.

2.4. Methods

2.4.1. Semivariogram

The semivariogram is a fundamental concept in geostatistics. In this study, we employ the semivariogram, grounded in the theory of regionalized variables, utilizing the variogram as the principal analytical tool. Our investigation focuses on the spatial variability of farmland soil pH, which exhibits significant spatial correlation and dependency. Additionally, we explore the underlying interaction mechanisms that contribute to this variability. The formula for the semivariogram is as follows [21,22]:

$$\gamma \left(h\right)=\frac{1}{2N\left(h\right)}{\displaystyle \sum _{i=1}^{N\left(h\right)}\left[Z\left(x_i\right)-Z\left(x_i+h\right)\right]}$$

(1)

where $\gamma \left(h\right)$ is the value of the semivariogram corresponding to all points at spatial distance $h$; $h$ is the step size indicating the spatial interval distance of the sample points; $N\left(h\right)$ is the logarithm of points with distance h; and $Z\left(xi\right)$ and $Z\left(xi+h\right)$ are the values of the variable $Z$ at spatial positions $x_i$ and $\left(x_i+h\right)$, respectively [23,24].

2.4.2. Spatial Autocorrelation

Spatial autocorrelation is generally used to characterize the heterogeneity and spatial clustering of variables, which can be classified into global spatial autocorrelation and local spatial autocorrelation according to different focuses. At present, the spatial autocorrelation analysis of variables is mainly determined based on Moran’s index (Moran’s I), and in this study, the spatial autocorrelation of soil pH in XJ farmland was analyzed based on Moran’s I. The global Moran’s I was calculated using the following formula [25,26]:

$$I=\frac{n{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{\omega }_{ij}\left(x_i-\overline{x}\right)}}\left(x_j-\overline{x}\right)}{\left({\displaystyle \sum _{i=1}^n{\displaystyle \sum _{i=1}^n{\omega }_{ij}}}\right){\displaystyle \sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}}$$

(2)

where $n$ is the number of spatial data; $x_i$ and $x_j$ are the attribute values of the spatial elements in zones $i$ and $j$, respectively; $\overline{x}$ is the average value of all the spatial data; and ${\omega }_{ij}$ is the element of the spatial weight matrix, which is generally a symmetric matrix, and ${\omega }_{ij}$ = 0 [25].

In the analysis of global Moran’s I, the significance of spatial autocorrelation among spatial elements is commonly assessed using the standardized statistic $Z\left(i\right)$. The general formula for this calculation is presented below:

$$Z\left(I\right)=\frac{I-E\left(I\right)}{\sqrt{Var\left(I\right)}}$$

(3)

$$E\left(I\right)=\frac{-1}{n-1}$$

(4)

where $Var\left(I\right)$ is the theoretical variance of Moran’s I; and $E\left(I\right)$ is the theoretical expected value of Moran’s I. At the 0.05 level, when $\left|Z\right|$ > 1.96, it indicates that there is spatial autocorrelation for this spatial element [26].

3. Results

3.1. Descriptive Statistical Analysis of Farmland Soil pH

As indicated in

Table 2. Descriptive statistics of farmland soil pH in XJ from 2008 to 2010 and from 2019 to 2021.

Area	Cities	2008~2010			2019~2021
		Sample Size	Mean	CV (%)	Sample Size	Mean	CV (%)
N-XJ	Altay	1695	7.84 ± 0.37l	4.74	1329	8.06 ± 0.37fg	4.52
	Bortala	951	8.38 ± 0.26b	3.11	792	7.79 ± 0.29h	3.78
	Changji	2784	7.97 ± 0.31k	3.85	3067	8.01 ± 0.33g	4.07
	Urumqi	303	8.02 ± 0.22j	2.76	369	8.05 ± 0.27fg	3.40
	Karamay	/	/	/	185	8.25 ± 0.32ab	3.86
	Ili	2917	8.13 ± 0.32efg	3.90	2988	8.13 ± 0.26cde	3.20
	Tacheng	1795	8.06 ± 0.25hi	3.10	3493	8.22 ± 0.23b	2.85
	Hami N-XJ	463	8.10 ± 0.28gh	3.46	198	8.10 ± 0.25ef	3.21
	Total N-XJ	10908	8.05 ± 0.34ij	4.18	12421	8.10 ± 0.31ef	3.82
S-XJ	Aksu	2199	8.14 ± 0.32ef	3.92	3223	8.17 ± 0.31c	3.83
	Turpan	287	8.46 ± 0.35a	4.14	320	8.17 ± 0.27cd	3.27
	Bayingolin	2395	8.11 ± 0.35fg	4.29	1633	8.14 ± 0.35cde	3.27
	Hotan	1788	8.30 ± 0.26c	3.16	992	8.28 ± 0.27a	3.30
	Kashgar	5058	8.10 ± 0.22fg	2.73	3213	8.16 ± 0.30cd	3.67
	Kizilsu	362	8.25 ± 0.30d	3.57	265	8.11 ± 0.36de	4.38
	Hami S-XJ	365	8.33 ± 0.30c	3.55	253	8.01 ± 0.26g	3.25
	Total S-XJ	12454	8.16 ± 0.29e	3.61	9899	8.17 ± 0.31c	3.84
XJ	Total	23362	8.11 ± 0.32fg	3.94	22320	8.13 ± 0.31cde	3.85

Note: CV is the coefficient of variation of the soil pH data of farmland, which is used to measure the degree of dispersion of the data; different letters for data in the same column indicate significant differences (p < 0.05).

and

Table 3. Results of the t-test for the soil pH change.

Area	Cities	$∆\mathbf{p}\mathbf{H}$	Significance (2-Tailed)
N-XJ	Altay	0.22	0.000 **
	Bortala	−0.59	0.000 **
	Changji	0.04	0.000 **
	Urumqi	0.03	0.137
	Karamay	/	/
	Ili	0.00	0.818
	Tacheng	0.16	0.000 **
	Hami N-XJ	0.00	0.951
	Total N-XJ	0.05	0.000 **
S-XJ	Aksu	0.03	0.000 **
	Turpan	−0.29	0.000 **
	Bayingolin	0.03	0.011 *
	Hotan	−0.02	0.041 *
	Kashgar	0.06	0.000 **
	Kizilsu	−0.14	0.000 **
	Hami S-XJ	−0.32	0.000 **
	Total S-XJ	0.01	0.011 *
XJ	Total	0.02	0.000 **

Note: * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).

, from 2008 to 2010, the average pH value of farmland soil in XJ was 8.11, showing an overall alkaline nature. The average pH value of farmland soil in N-XJ was 8.05, and in S-XJ, it was 8.16. The pH of the farmland soil in S-XJ was significantly higher than that in N-XJ (p < 0.05), demonstrating a spatial distribution pattern of “higher in the south and lower in the north”. Among the regions in XJ, the farmland soil in Turpan had the highest pH, reaching 8.46, which was significantly higher than that in other regions (p < 0.05). In contrast, the farmland soil in Altay had the lowest pH, at 7.84, which was significantly lower than that in other regions (p < 0.05).

From 2019 to 2021, the average pH value of farmland soil in XJ was 8.13, showing a slight increase from 2008 to 2010. The average pH value of farmland soil in N-XJ was 8.10, while in S-XJ, it was 8.17. Overall, the soil pH in S-XJ remains significantly higher than that in N-XJ, with both regions experiencing a certain degree of increase (p < 0.05). Among all regions in XJ, the farmland soil in Hotan had the highest pH, reaching 8.28, which was significantly higher than that in all other cities except Karamay (p < 0.05). In contrast, the farmland soil in Bortala had the lowest pH, at 7.79, which was significantly lower than that in other regions (p < 0.05).

The CV of soil pH in farmland of XJ cities were lower than 10% in both observation periods. It belongs to a very low degree of variability (CV < 25%) [27]. In the past decade, the CV of farmland soil pH in Xinjiang has slightly decreased, indicating that the degree of variability of farmland soil pH is decreasing. It is noteworthy that the CV in N-XJ decreased by 0.32 percentage points, whereas the CV in S-XJ increased by 0.23 percentage points, which highlights the regional differences in soil pH stability.

Over the past decade, farmland soil pH in XJ has significantly increased by 0.02 units (p < 0.05). Specifically, soil pH in S-XJ has significantly increased by 0.01 units (p < 0.05), while in N-XJ, it has significantly increased by 0.05 units (p < 0.05). Among all regions (excluding Karamay), the changes in soil pH in the farmland of various regions in XJ were significant (p < 0.05), except for Urumqi, Ili, and Hami _N-XJ. Altay had the most significant increase, with the pH rising by 0.22 units. In contrast, Bortala in the western part of N-XJ experienced the greatest decrease, with the soil pH dropping by 0.59 units.

3.2. Analysis of the Spatiotemporal Distribution Characteristics of Farmland Soil pH

3.2.1. Spatial Autocorrelation Analysis

The results of the spatial autocorrelation analysis (

Table 4. Spatial autocorrelation analysis of soil pH in different time periods.

Periods	Moran’s I	$\mathit{E}\left(\mathit{I}\right)$	Variance	$\mathit{Z}\left(\mathit{I}\right)$	p
2008~2010	0.568	−4.30 × 10−5	1.34 × 10−4	49.138	<0.001
2019~2021	0.152	−4.50 × 10−5	9.70 × 10−5	15.422	<0.001

) show that the $Z\left(I\right)$ for farmland soil pH in XJ was 49.138 from 2008 to 2010 and was 15.422 from 2019 to 2021, both exceeding 1.96 with corresponding probabilities p of less than 0.05. This indicates the strong spatial autocorrelation of soil pH in both periods, suggesting that spatial interpolation methods can be utilized to study its distribution. Moreover, Moran’s I for both periods was greater than zero, indicating a positive spatial correlation for farmland soil pH during these times. Over the past decade, Moran’s I for XJ farmland soil pH has significantly decreased, suggesting a weakening in spatial autocorrelation and an increase in the influence of random factors, which may be related to intensified human activities.

3.2.2. Semivariogram Analysis

The semivariogram analysis results (

Table 5. Semivariogram models and their parameters for farmland soil pH in Xinjiang.

Periods	Area	Optimal Model	Nugget Variance (C0)	Sill (C0 + C)	Structural Variance [C0/(C0 + C)], %	Range (A0) km	R²	RSS
2008~ 2010	XJ	Exponential	0.0260	0.0849	30.62	26.70	0.93	8.50 × 10−5
	N-XJ	Exponential	0.0305	0.1060	28.77	24.00	0.94	1.65 × 10−4
	S-XJ	Exponential	0.0211	0.0711	29.68	22.30	0.87	2.33 × 10−4
2019~ 2021	XJ	Exponential	0.0367	0.0878	41.80	23.60	0.85	4.16 × 10−5
	N-XJ	Exponential	0.0361	0.0896	40.29	21.20	0.82	2.90 × 10−5
	S-XJ	Gaussian	0.0277	0.0858	32.28	19.30	0.87	5.01 × 10−6

Note: R² indicates the degree of explanation of the dependent variable by the independent variables in the model; and the residual sum of squares (RSS) indicates the sum of the squared differences between the predicted and actual values in the regression model, which is used to measure the goodness of fit of the model to the data. Generally speaking, the higher the R² and the lower the RSS of a model, the better the model fits.

) indicate that from 2008 to 2010, the optimal semivariogram model for farmland soil pH across XJ, N-XJ, and S-XJ, and from 2019 to 2021 in XJ and N-XJ, was the exponential model, while from 2019 to 2021 in S-XJ, it was the Gaussian model. Model residuals were near zero, and the coefficient of determination (R²) exceeded 0.8, indicating a robust fit. The nugget variance (C₀) quantifies the variability attributed to random factors, such as experimental errors or variations below the sampling resolution. Nugget values for XJ, N-XJ, and S-XJ during the periods 2008~2010 and 2019~2021 were 0.0260, 0.0305, 0.0211, 0.0367, 0.0361, and 0.0277, respectively, suggesting that variations at the sampling scale are influenced by errors in sample collection, sampling extent, and variable factors such as fertilization and cultivation. The structural variance [C₀/(C₀ + C)] represents the proportion of spatial heterogeneity caused by random factors. The structural variance for XJ farmland soil pH during the two periods was 30.62% and 41.80%, respectively, both within the 25% to 75% range, indicating a moderate degree of spatial variability [28]. This suggests that the spatial variability of soil pH in XJ farmlands is primarily driven by structural factors such as parent material, elevation, and soil type. However, random factors such as cultivation and fertilization also significantly influence the spatial variability of soil pH. The range (A₀) is the sample distance at which the semivariogram reaches the sill, beyond which there is typically no spatial correlation. Over the past decade, the range for XJ farmland soil pH decreased from 26.70 km to 23.6 km, with reductions of 2.8 km in N-XJ and 3.0 km in S-XJ. This indicates that the impact of human activities and other random factors on farmland soil pH is increasing.

3.2.3. Analysis of the Spatiotemporal Distribution of Farmland Soil pH

In order to more accurately and intuitively reflect the spatial distribution pattern of farmland soil pH in XJ, this paper interpolates farmland soil pH in XJ from 2008 to 2010 and from 2019 to 2021 based on semivariogram function simulation results using ordinary kriging interpolation (Figure 3).

Analysis reveals that the pH of most farmland soils in XJ is concentrated between 7.5 and 8.5, with S-XJ generally exhibiting higher pH levels than N-XJ, demonstrating a “higher in the south, lower in the north” trend. This pattern may be associated with the climatic, geographical conditions, and cultivation management practices of S-XJ and N-XJ. From 2019 to 2021, farmland soil pH in N-XJ exhibited a spatial distribution pattern of “higher in the southwest and lower in the northeast”. Central and eastern N-XJ, including Changji, Urumqi, and Hami _N-XJ, as well as the northern part of the Altay region, had relatively lower soil pH levels. In contrast, the southwestern parts, including Karamay, Tacheng, Ili, and the southwestern part of the Altay, generally showed higher pH levels. However, the Bortala in the western N-XJ was an exception, recording the lowest soil pH in N-XJ during 2019~2021. In S-XJ, most farmland soil pH values were above 8.0, indicating generally higher soil pH levels. Lower pH values were sporadically found in Hami _S-XJ, the western parts of Aksu, parts of Bayingol bordering the Aksu, and areas of Kashi bordering the Kizilsu. Notably, Hami _S-XJ had the lowest soil pH in S-XJ during this period.

From 2008 to 2010, the distribution of farmland soil pH in N-XJ was patchy. Lower pH levels were observed in the northern Altay, Karamay, southern Tacheng, Changji, Urumqi, and Hami _N-XJ. In contrast, higher soil pH was recorded in the southwestern regions such as the Bortala, Ili, and the western parts of Tacheng. In S-XJ, the spatial distribution pattern of soil pH during this period largely mirrored that observed from 2019 to 2021, with the exception of Hami S-XJ, which exhibited notably higher pH levels, contrasting with its lower levels in the later period.

The variation range of soil pH in farmland > 0.1 was defined as ‘Rise’, the variation range <−0.1 was defined as ‘Decrease’, and the variation range between [−0.1,0.1] was defined as ‘Stability’ [12]. Further analysis of the changes in soil pH in farmland in XJ over the past decade was performed. According to Figure 4, over the past decade, the farmland in N-XJ with soil pH defined as ‘Rise’ has been the largest, accounting for 38.26% of the total farmland, mainly in Altay, Tacheng, and Changji in N-XJ, and Aksu and Kashgar in S-XJ. The areas of farmland with soil pH defined as ‘Decrease’ and ‘Stability’ were basically the same, accounting for 30.14% and 31.60% of the total cropland area. Among them, the farmland with reduced soil pH was mainly distributed in Bortala in N-XJ, Turpan and Hami _S-XJ in S-XJ, and to some extent in the rest of the region; the farmland with stable soil pH was scattered throughout the territory. The soil pH of most of the farmland in N-XJ has risen to varying degrees, and its area accounts for 49.36% of the total farmland in N-XJ; while the soil pH of only 27.51% of the farmland in the S-XJ has risen in the last decade or so, and 35.04% of the farmland has experienced varying degrees of decline in soil pH. Overall, the trend in soil alkalization of farmland in N-XJ has intensified over the past decade or so, and although the trend in soil alkalization of farmland in the S-XJ has been curbed to a certain extent, the pH of farmland is still high and on an upward trend.

3.3. Analysis of Factors Affecting Farmland Soil pH

3.3.1. Topographic Factors

XJ possesses unique geographical features, with the Altai Mountains, Tianshan Mountains, and Kunlun Mountains located in the north, central, and southern regions, respectively, with relatively high elevations. In contrast, the Junggar Basin and Tarim Basin, enveloped by these major mountain ranges, feature lower terrain. Consequently, it becomes imperative to investigate the spatial distribution of farmland soil pH across varying topographic conditions to inform agricultural practices and soil management strategies in the XJ region. As depicted in

Table 6. Changes in pH of farmland soils under different elevation conditions.

Elevation (m)	Sample Size	Min	Max	Mean	CV (%)
≤0	322	7.57	8.90	8.34 ± 0.34a	4.04
0–1000	21,522	7.26	9.00	8.10 ± 0.33d	3.17
1000–2000	22,040	7.26	9.00	8.13 ± 0.30cd	3.17
2000–3000	860	7.30	8.90	8.14 ± 0.26bc	3.21
>3000	617	7.30	8.90	8.16 ± 0.26b	3.17

Different letters for data in the same column indicate significant differences (p < 0.05).

, a discernible trend emerges, wherein farmland soil pH demonstrates an increasing pattern with elevation above sea level (>0 m). Notably, a statistically significant disparity is observed between elevations exceeding 3000 m and those below or equal to 2000 m. However, within each adjacent elevation gradient, the observed differences were found to be statistically insignificant (p < 0.05). It is noteworthy that soil pH attains its peak at elevations below or equal to 0 m, registering at 8.34, significantly surpassing pH values recorded at elevations above 0 m (p < 0.05). The CV of soil pH in XJ farmland under different elevation gradients is less than 10%, which indicates a weak spatial variability.

3.3.2. Soil Nutrients Factors

Soil nutrients are important indicators of farmland productivity and farmland quality. A large number of studies have proven that there is a significant relationship between soil nutrients and pH, and the decomposition of nutrients in the soil may cause soil acidification and change soil pH [29,30,31]. In this study, after screening farmland SOM, AN, AP, and AK contents and farmland soil pH for Pearson’s correlation analysis (

Table 7. Effect and correlation of soil nutrient content on soil pH.

Grade	SOM		AN		AP		AK
	Mean pH	CV (%)	Mean pH	CV (%)	Mean pH	CV (%)	Mean pH	CV (%)
6	8.23 ± 0.36a	4.17	8.23 ± 0.34a	4.17	8.20 ± 0.32a	3.91	8.28 ± 0.44a	5.31
5	8.14 ± 0.35b	3.96	8.18 ± 0.32b	3.96	8.19 ± 0.35ab	4.25	8.26 ± 0.33a	3.96
4	8.13 ± 0.30bc	3.69	8.11 ± 0.30c	3.69	8.18 ± 0.31ab	3.73	8.21 ± 0.31ab	3.82
3	8.11 ± 0.29c	3.61	8.09 ± 0.29d	3.61	8.15 ± 0.30bc	3.73	8.15 ± 0.31bc	3.79
2	8.06 ± 0.35d	3.58	8.04 ± 0.29e	3.58	8.10 ± 0.32cd	3.95	8.13 ± 0.30bc	3.73
1	8.05 ± 0.28d	3.62	8.02 ± 0.29e	3.62	8.08 ± 0.31d	3.86	8.08 ± 0.31c	3.85
Pearson’s correlation coefficient	−0.108 **		−0.197 **		−0.126 **		−0.248 **

Different letters for data in the same column indicate significant differences (p < 0.05). ** Correlation is significant at the 0.01 level (2-tailed).

), it was found that farmland soil pH and the contents of SOM, AN, AP, and AK showed a highly significant negative correlation (p < 0.01). Farmland soil pH decreased with the increase in the contents of SOM, AN, AP, and AK. Among them, the pH of farmland soil corresponding to a SOM content of grade 2 or above was significantly lower than that of other grades (p < 0.05), and the pH of farmland soil corresponding to a SOM content of grade 6 was significantly higher than that of other grades (p < 0.05). When the AN content is above grade 2, the corresponding farmland soil pH is significantly lower than other grades (p < 0.05). When the AN content is below grade 6, the corresponding farmland soil pH is significantly higher than other grades (p < 0.05). The differences in AP and AK were not significant at all adjacent grades. The coefficients of variation of farmland soil pH under different conditions of SOM, AN, AP, and AK content were less than 10%, which indicates a weak spatial variation.

3.3.3. Irrigation Methods

Different irrigation methods vary in terms of water consumption, infiltration depth, and drainage, which can significantly influence the content and distribution of soil acidogenic ions, ultimately affecting soil pH [32]. Therefore, it is important to explore the differences in soil salinity under different irrigation methods. Data from

Table 8. Farmland soil pH under different irrigation methods.

Irrigation Methods	Sample Size	Min	Max	Mean	CV (%)
Drip and flood irrigation	161	7.72	8.98	8.39 ± 0.18a	2.18
Flood irrigation	9229	7.30	9.00	8.20 ± 0.28bc	3.41
Drip irrigation	6602	7.27	9.00	8.15 ± 0.32cd	3.88
Border irrigation	1749	7.36	8.98	8.13 ± 0.23cd	2.89
Furrow irrigation	5575	7.29	8.98	8.13 ± 0.35cd	4.37
Flood and border irrigation	3364	7.30	8.90	8.09 ± 0.30d	3.77
No irrigation	517	7.28	8.99	7.98 ± 0.32e	4.01
Sprinkler irrigation	24	7.40	8.98	7.97 ± 0.40e	4.98

Different letters for data in the same column indicate significant differences (p < 0.05).

show that the soil pH values in farmland vary according to the irrigation method used, ranked from highest to lowest as follows: drip and flood irrigation > flood irrigation > drip irrigation > border irrigation > furrow irrigation > flood and border irrigation > no irrigation > sprinkler irrigation. Specifically, fields with combined drip and flood irrigation exhibit the highest soil pH, at 8.39, which is significantly greater than that seen with other methods (p < 0.05). The lowest pH values, 7.97 and 7.98, occur under sprinkler and no irrigation conditions, respectively, both significantly lower than other methods (p < 0.05). Differences in soil pH among the other irrigation methods are not statistically significant. Furthermore, the CV for the pH under each irrigation condition is less than 10%, indicating a low level of spatial variability.

4. Discussion

4.1. Spatiotemporal Variability Characteristics of Farmland Soil pH in Xinjiang

Changes in soil pH can directly impact the physical, chemical, and biological processes of soil ecosystems, and are key factors limiting the effectiveness and mobility of soil nutrients and pollutants such as heavy metals. A comprehensive analysis of the spatiotemporal variability of soil pH helps in assessing the quality of farmland and the effectiveness of soil nutrients [33,34]. This study found that the overall soil pH in farmland in XJ is relatively high. Between the periods of 2008~2010 and 2019~2021, the soil pH exhibited a “high in the south, low in the north” spatial distribution pattern, which aligns with China’s soil development pattern of “acidic near the coast and alkaline inland” [35]. Over the past decade, the overall soil pH in XJ has shown an increasing trend, with a rise of 0.05 units in N-XJ, and of 0.01 units in S-XJ. Although the increase is more pronounced in N-XJ, the soil pH in S-XJ has consistently been higher than that in the north, likely due to the presence of the Tianshan Mountains. The Tianshan Mountains, stretching across central XJ, block the southward movement of warm and moist air currents, leading to distinct climatic differences between N-XJ and S-XJ [36]. The relatively humid climate of N-XJ, with more rainfall, can leach minerals and basic ions from the soil, increasing the surface soil H+ content and effectively limiting the rise in soil pH, whereas the climate in S-XJ is dry, with sparse rainfall and high evaporation, making the soil more prone to alkalization [19,37]. Additionally, human activities are also important factors affecting the spatiotemporal variability of farmland soil pH. Numerous studies have indicated that applying an appropriate amount of biochar can effectively inhibit soil alkalinization [38,39,40]. Among these, research by Jing et al. [40] has confirmed that biochar, which contains abundant AP, can partially substitute for mineral phosphorus fertilizers, promote plant growth, and reduce the need for phosphorus fertilizer application in agriculture. Additionally, measures such as increasing the application of organic fertilizers [41], introducing and cultivating vegetation that can improve soil pH (e.g., Suaeda salsa (L.) Pall.) [42], and using alkalinity ameliorants can all effectively suppress soil alkalinization.

Among the various regions in XJ, Bortala has experienced the greatest decrease in farmland soil pH over the past decade, with a drop of 0.59 units. This significant decrease is likely due to abundant local rainfall and increased fertilization. Studies suggest that, as a result of increased fertilization, soil nutrient content, especially SOM, has been rising in Bortala [43]. Additionally, the high rainfall in the area intensifies soil eluviation, which contributes to the notable decrease in soil pH. In contrast, the Altay has seen the largest increase in farmland soil pH, rising by 0.22 units. The soil in this area predominantly consists of chestnut calcic soils, which are characterized by slow nutrient transformation and weak eluviation. This leads to the accumulation of base ions in the soil surface layer, causing an increase in the soil pH of the region.

4.2. Impact of Different Factors on Farmland Soil pH

Influence of topographic factors on farmland soil pH. Topographic factors can affect soil pH by altering the redistribution of soil water, nutrients, and heat [44]. Our study finds that above sea level (>0 m), farmland soil pH increases with elevation, consistent with findings by Zhao et al. [45] and Zuo et al. [46]. This may be due to decreasing temperatures and thinner soil layers at higher altitudes, which slow organic matter decomposition and increase the direct influence of parent rock weathering, thereby gradually increasing pH levels [47]. Conversely, at altitudes ≤ 0 m, farmland soil pH is highest and is significantly greater than at altitudes > 0 m. This is likely because areas in XJ below sea level are mainly concentrated in the Turpan Basin, where high temperatures, high evaporation rates, and low rainfall accelerate rock weathering, releasing more base ions [45]. These ions are less likely to leach downward, resulting in a higher soil pH in these farmlands.

Influence of soil nutrient content on farmland soil pH. Our study found that there is an extremely significant negative correlation between the pH of farmland soil in XJ and soil nutrients (SOM, AN, AP, and AK). The soil pH decreases as the content of SOM, AN, AP, and AK increases. This may be due to the release of large amounts of acid during the decomposition of soil nutrients. For example, when SOM decomposes, it generates a large amount of organic acids, thus causing the soil pH to gradually decrease [30]. In soils with higher levels of AN, the nitrification process is also stronger [31] and the release of H+ during this process leads to a decrease in soil pH. The results of studies by Dai et al. [47] also support this conclusion.

Influence of irrigation methods on farmland soil pH. In XJ, where arid conditions prevail and water resources are scarce, most farmland relies on artificial irrigation to meet the water needs of crops. Different irrigation methods can cause the movement of salt ions in the soil, leading to changes in soil pH. This study found that under different irrigation methods, the pH of farmland soil ranged from highest to lowest as follows: drip and flood irrigation > flood irrigation > drip irrigation > border irrigation > furrow irrigation > flood and border irrigation > no irrigation > sprinkler irrigation. Traditional flood irrigation uses a large amount of water, while the groundwater table in XJ is generally shallow and the groundwater mineralization is high [48]; thus, the evaporation of residual water after full irrigation tends to cause salts and alkaline-based ions to accumulate in the soil surface layer, resulting in soil salinization problems. One significant advantage of sprinkler irrigation is its ability to prevent the accumulation of salts and alkaline residues that might occur with drip or flood irrigation. By atomizing water into the air, sprinkler irrigation ensures that water droplets are distributed more evenly across the soil surface, reducing the likelihood of salt and other dissolved substances accumulating [49]. Consequently, soil pH tends to be lower under sprinkler irrigation conditions. Drip irrigation is a technique that has been extensively promoted in XJ in recent years, known for its strategy of “less irrigation, more frequency”. Due to the region’s arid climate and intense evaporation, prolonged use of this method can lead to secondary salinization of the soil, thereby increasing soil pH [50]. However, some studies have found that with the extension of the duration of drip irrigation, the alkalinity of the soil in the study area will gradually decrease [51].

Influence of other factors on farmland soil pH. Numerous studies have shown that arid climatic conditions with high temperatures, high evapotranspiration, and low rainfall are important factors in imaging soil pH changes [52,53]. Temperature primarily affects the rate of rock weathering, while evaporation and precipitation affect the movement of base ions in the soil, which, in turn, affects the physicochemical properties of the soil, causing changes in soil pH [54]. In this study, Pearson’s correlation coefficient was utilized to examine the relationship between farmland soil pH and XJ annual ET, Pre, and LST (

Table 9. Pearson’s correlation coefficient of farmland soil pH with climatic factors.

	ET	Pre	LST
pH	0.082 **	−0.201 **	0.121 **

Note: ET indicates total annual evapotranspiration; Pre indicates total annual precipitation; LST indicates annual mean surface temperature. ** Correlation is significant at the 0.01 level (2-tailed).

). It was found that soil pH in XJ farmlands has a highly significant negative correlation with Pre, and a highly significant positive correlation with ET and LST. Due to the barrier effect of the Tianshan Mountains, the climatic characteristics of N-XJ and S-XJ are markedly different. Studies indicate that compared to N-XJ, S-XJ experiences higher temperatures and evaporation rates, while its precipitation is significantly lower, with annual evaporation reaching 7 to 20 times that of annual precipitation [37,55]. The high temperatures, high evaporation rates, and low precipitation levels are significant factors contributing to the elevated soil pH in XJ. Increased temperatures enhance weathering, which raises the base saturation [46,54]. When precipitation is less than evaporation, base cations accumulate at the soil surface, leading to an increase in soil pH. Compared to N-XJ, S-XJ has higher temperatures, lower precipitation, and greater evaporation rates, which exacerbates the salinization issue. Consequently, the soil pH of farmlands in XJ during both periods shows a pattern of “higher in the south and lower in the north”.

Comparing two periods, there has been a slight increase in the soil pH of farmlands in XJ, with a rise of 0.05 units in N-XJ, and of 0.01 units in S-XJ. Soil pH is significantly correlated with Pre, ET, and LST. Utilizing the Google Earth Engine (GEE) remote sensing big data platform, this study accessed the trends in the annual average LST, Pre, and ET for XJ over the past decade (Figure 5). Analysis shows that, over more than ten years, both the annual average LST and ET in XJ have been on an upward trend, while Pre has shown a downward trend. These trends are critical drivers of the observed rise in soil pH in the farmlands of XJ.

With the advancement of agricultural modernization, in addition to the natural factors previously mentioned, human activities are increasingly impacting the physicochemical properties of soil. One of the key contributors to changes in soil acidity and alkalinity is the application of chemical fertilizers [56]. Analysis of

Table 10. Fertilizer application in different regions and periods in Xinjiang.

Periods	Regions	Fertilizer Application (× 107 kg)	N Fertilizer Application (× 107 kg)	P Fertilizer Application (× 107 kg)	K Fertilizer Application (× 107 kg)
2008	N-XJ	51.10	23.58	11.94	2.57
	S-XJ	61.62	31.34	19.93	3.02
	XJ	112.72	54.93	31.87	5.59
2012	N-XJ	66.27	30.09	15.51	3.84
	S-XJ	82.90	39.22	26.59	5.06
	XJ	149.17	69.31	42.10	8.90
2020	N-XJ	65.74	27.41	16.49	6.49
	S-XJ	105.27	43.59	32.22	7.86
	XJ	171.01	71.00	48.71	14.35

shows that from 2008 to 2020, the usage of chemical fertilizers in XJ has consistently risen. In N-XJ, the amount of fertilizer used initially increased and then decreased, with a relatively small range of change; in contrast, in S-XJ, the usage significantly increased, reaching 1.7 times the 2008 level by 2020. The indiscriminate use of chemical fertilizers, especially ammonium nitrogen fertilizers, while neglecting the application of organic fertilizers, is a major factor causing changes in soil pH. The ammonium ions added to the soil release a large amount of H+ through nitrification reactions [57]. The greater and more substantial increase in fertilizer usage in S-XJ might explain why the rise in soil pH there has been relatively slower over the past decade. Since 2005, XJ has actively responded to national initiatives by comprehensively implementing soil testing and formulated fertilization techniques. This approach is beneficial for scientifically adjusting the ratio of chemical to organic fertilizers, which is vital for controlling soil pH and improving the quality of farmland.

This study examined the spatiotemporal changes in farmland soil pH in XJ and their associated factors. The insights gleaned provide valuable directions for updating local agricultural management strategies and refining fertilization practices. These findings hold particular importance for addressing soil salinity and alkalinity issues in XJ, as well as for tailoring fertilization plans to specific site conditions. Additionally, this study lays a foundational framework for further investigation into the mechanisms and influencing factors behind soil pH variations. Nevertheless, constrained by limitations in data availability and other practical factors, the analysis is confined to a relatively short timeframe, thus not capturing the long-term trends in soil pH in XJ. To conduct a more scientifically rigorous assessment of the spatiotemporal variability of soil pH in XJ, we are dedicated to acquiring longer-term data and undertaking more comprehensive research efforts.

5. Conclusions

(1) The average soil pH value of farmland in Xinjiang is 8.11. In S-XJ, it is 8.17, which is significantly higher than the 8.10 in N-XJ, showing a spatial distribution pattern of “higher in the south and lower in the north”. The spatial variability of farmland soil pH is low, classified as weak variability (CV < 10%). Spatial autocorrelation analysis indicates that the soil pH of farmlands exhibits strong spatial correlations in both periods. semivariogram analysis shows that the soil pH of XJ farmlands during both periods is influenced by both structural and anthropogenic factors, with a greater influence from structural factors.

(2) Over the past decade, the soil pH in XJ farmland has significantly increased by 0.02 units, with a significant increase of 0.05 units in N-XJ, and of 0.01 unit in S-XJ, showing a general trend towards alkalization. Spatial autocorrelation analysis reveals that Moran’s I value for soil pH in XJ has significantly declined during this period, indicating a reduction in spatial autocorrelation. Additionally, semivariogram analysis verifies that the A₀ values for both N-XJ and S-XJ, as well as the entire region, have decreased to varying degrees, suggesting that the impact of random factors is growing.

(3) The spatiotemporal changes in soil pH across XJ farmlands are influenced by both natural and human factors, including topography, soil nutrient content, irrigation practices, climate, and fertilization. The analysis indicates that (a) Soil pH in XJ farmlands increases with altitude when the altitude is greater than 0, and it is highest at altitudes ≤ 0, significantly higher than at altitudes > 0; (b) there is a highly significant negative correlation between soil pH and soil nutrient content (SOM, AN, AP, and AK) in XJ farmlands, with pH decreasing as nutrient levels increase; (c) depending on the irrigation method, the soil pH in XJ farmlands varies, with the following order from highest to lowest: drip and flood irrigation > flood irrigation > drip irrigation > border irrigation > furrow irrigation > flood and border irrigation > no irrigation > sprinkler irrigation; and (d) climate change and fertilization management practices are also important factors influencing the spatiotemporal variability of soil pH in XJ farmlands.