Irrigation Suitability and Interaction Between Surface Water and Groundwater Influenced by Agriculture Activities in an Arid Plain of Central Asia

Abstract

Agricultural activities and dry climatic conditions promote the evaporation and salinization of groundwater in arid areas. Long-term irrigation alters the groundwater circulation and environment in arid plains, as well as its hydraulic connection with surface water. A comprehensive assessment of groundwater irrigation suitability and its interaction with surface water is essential for water–ecology–agriculture security in arid areas. This study evaluates the irrigation water quality and groundwater–surface water interaction influenced by agricultural activities in a typical arid plain region using hydrochemical and stable isotopic data from 51 water samples. The results reveal that the area of cultivated land increases by 658.9 km² from 2000 to 2023, predominantly resulting from the conversion of bare land. Groundwater TDS (total dissolved solids) value exhibits significant spatial heterogeneity, ranging from 516 to 2684 mg/L. Cl⁻, SO₄²⁻, and Na⁺ are the dominant ions in groundwater, with a widespread distribution of brackish water. Groundwater δ¹⁸O values range from −9.4‰ to −5.4‰, with the mean value close to surface water. In total, 86% of the surface water samples are good and suitable for agricultural irrigation, while 60% of shallow groundwater samples are marginally suitable or unsuitable for irrigation at present. Groundwater hydrochemistry is largely controlled by intensive evaporation, water–rock interaction, and agricultural activities (e.g., cultivated land expansion, irrigation, groundwater exploitation, and fertilizers). Agricultural activities could cause shallow groundwater salinization, even confined water deterioration, with an intense and frequent exchange between groundwater and surface water. In order to sustainably manage groundwater and maintain ecosystem stability in arid plain regions, controlling cultivated land area and irrigation water amount, enhancing water utilization efficiency, limiting groundwater exploitation, and fully utilizing floodwater resources would be the viable ways. The findings will help to deepen the understanding of the groundwater quality evolution mechanism in arid irrigated regions and also provide a scientific basis for agricultural water management in the context of extreme climatic events and anthropogenic activities.

1. Introduction

Groundwater is a vital water resource in water-scarce arid areas, playing a crucial role in ensuring human survival, supporting irrigation and industry, and maintaining ecosystems [1,2]. Globally, the amount of agricultural irrigation water accounts for 70% of total water consumption [3]. However, as water scarcity becomes a pressing issue, sustainable groundwater utilization faces severe challenges in arid areas [4,5]. Agricultural production, urbanization development, and groundwater overexploitation have changed the content of groundwater hydrochemical components to a certain extent, affecting groundwater quality [6,7,8,9]. In agricultural systems, groundwater quality directly impacts crop yield, soil health, and the long-term sustainability of irrigation practices [2]. Contaminated irrigation water could lead to soil salinization, accumulation of heavy metals, and the transfer of pathogens to crops, ultimately affecting food safety and economic stability [10]. Agricultural activities and dry climate promote the evaporation and salinization of groundwater in arid areas [11,12]. Long-term irrigation alters groundwater circulation and environment in arid plains, as well as their hydraulic connection to surface water [13]. As the largest inland river basin in the arid region of China, the Tarim Basin faces severe water availability challenge, where groundwater serves as a primary water resource to support domestic use, agricultural irrigation, ecosystem functions, and sustainable socioeconomic development [14,15]. Agriculture plays a key role in local socioeconomic development, heavily dependent on irrigation from streamflow and pumping groundwater, thus leading to groundwater level decline and environmental degradation [16,17]. Therefore, a comprehensive assessment of groundwater quality and surface water–groundwater conversion mechanism in arid areas is critically important for safeguarding groundwater resources and indispensable for agricultural sustainability, particularly in the Tarim Basin [2].

Groundwater hydrochemistry is controlled by natural factors (e.g., climate and geology) and anthropogenic influences (e.g., agricultural irrigation, fertilizer use, and wastewater discharge) [4,18,19,20]. In arid areas, the primary factors shaping groundwater hydrochemistry would include rock weathering, mineral dissolution, salt accumulation, cation exchange, evaporation, agricultural irrigation, and fertilizers [6,21,22,23]. At present, Na% (soluble sodium percentage) and SAR (sodium adsorption ratio) indices are widely used to evaluate groundwater irrigation suitability and to assess the potential influence of agricultural irrigation water on crops and soil [24]. The Wilcox and USSL diagrams are well-established methods for assessing the suitability of agricultural irrigation water, providing critical insight for groundwater resource management [2,25]. Wang et al. [13] reported that the long-term development of irrigation agriculture (e.g., large-scale land reclamation) in arid regions could cause groundwater level decline and salinization by analyzing LULC (land use and land cover) and NDVI (normalized difference vegetation index) variations. However, a comprehensive understanding of groundwater quality and irrigation suitability remains insufficient in arid areas, particularly in the Tarim Basin.

Previous research has gradually emphasized the important role of hydraulic connection between groundwater and surface water in groundwater quality changes, especially in irrigated areas or riparian zones [13]. It has been found that in the irrigation areas with shallow DGL (depth to groundwater level), the pollution of surface water can affect the shallow groundwater quality through leakage, while the irrigation return flow would also affect the water quality of agricultural drainage channels [16]. However, the understanding of the surface water–groundwater interaction mechanism in arid irrigated areas is still insufficient [13,26]. Yin et al. [26] reported that groundwater depletion was approaching hydrological limits, although this process was greatly compensated by rainfall and streamflow in an oasis–desert system. The plain region of the Hotan River basin is a typical arid irrigated plain in the Tarim Basin of Central Asia, experiencing rapid agricultural development. The plain heavily relies on groundwater and faces significant challenges, such as low precipitation, limited streamflow, strong evaporation, soil and groundwater salinization, and excess fertilization [17,23]. This contributes to the degradation of groundwater quality. However, most existing research in the Tarim Basin has mainly focused on the spatiotemporal variations, sources, and driving factors of groundwater ions, lacking a comprehensive assessment of groundwater irrigation quality and agricultural impacts on groundwater–surface water interactions [14,23,27].

The arid plain region of the Hotan River basin in the Tarim Basin was selected as our study area to assess groundwater irrigation suitability and its interaction with surface water. In this study, we evaluate the irrigation water quality and groundwater–surface water interaction influenced by agricultural activities using hydrochemical and stable isotopic data in the Hotan River basin. The specific objectives of this study are (ⅰ) to analyze the variation characteristics of climate, vegetation, land use, and groundwater hydrochemistry; (ii) to assess groundwater irrigation suitability; (iii) to evaluate the impact of agricultural activities on groundwater quality; (iv) to clarify the interaction between groundwater and surface water; and (v) to put forward recommendations for groundwater resource management. The findings will help to deepen the understanding of the groundwater quality evolution mechanism in arid irrigated plain regions and provide a scientific basis for agricultural water resource management, groundwater pollution prevention, and ecological security maintenance in the context of extreme climatic events and anthropogenic activities.

2. Study Area

Our study area (the plain region of Hotan River) covers approximately 6.3 × 10⁴ km² (77.55–84.08° E, 36.21–38.12° N), located in an arid region of northwestern China, lying at the northern foot of Kunlun Mountains and on the southern edge of Tarim Basin (Figure 1b). The plain region consists of the city of Hotan and the Pishan, Moyu, Luopu, Hotan, Cele, Yutian, and Minfeng. The terrain descends from the south to north, with an altitude between 1202 and 3159 m above sea level. This area experiences a typical temperate continental arid climate, characterized by windy and dry conditions, with annual average air temperature and potential evaporation of 12.6 °C and 2159–3137 mm, respectively [17]. Precipitation is scarce, and exhibits a seasonal fluctuation, mainly concentrated from early April to late August, with an annual mean of 35.7 mm. Hotan River is the largest river of the Hotan River basin, with an annual streamflow of 45.60 × 10⁸ m³ during 1957 to 2023, which originates in the northern Kunlun Mountains [17].

The Quaternary strata vary gradually from piedmont proluvial sediment to alluvial sediment from the south to north, which influences the spatial distribution of aquifers [13]. The lithology of the Quaternary strata is mainly composed of sandy gravel, coarse sandstone, fine sandstone, and silty sandstone, with a loose structure, and porous aquifers [23]. The Quaternary strata consist of the confined aquifer group and the phreatic aquifer group. This study mainly focuses on the phreatic aquifer group (phreatic water and semi-confined water). The material composition of aquifers is predominantly silicate and evaporite. The observed DGL (depth to groundwater level) in the study area ranges from 0.22 to 49.99 m, with an average value of 10.97 m. The groundwater flow direction in the piedmont alluvial plain is from the south to north, primarily controlled by the topographic gradient. Groundwater in the plain is mainly recharged by precipitation, surface water, lateral Quaternary pore water, and bedrock fissure water from the mountainous area [28]. The intense evaporation and unique climatic conditions promote the mineral salts accumulation in the upper layer of soil, thus exacerbating groundwater salinization. Furthermore, the dominant land use types are bare land, grassland, and cultivated land. Agriculture is the cornerstone of regional economy, and is critical for local food security and economic stability, depending on irrigation system. Agricultural irrigation in the plain takes place during the spring, summer, and autumn, relying on surface water and groundwater extraction. The average field irrigation rate is 41.32 m³/acre in the study area (data from the Statistical Yearbook of Hotan Region, Xinjiang Uygur Autonomous Region, China). The dominant crops of the cultivation are wheat, corn, and cotton. The forest and fruit industry is critical for the local economy, and the dominant artificial forests include walnut, red date, grape, and pomegranate trees.

3. Materials and Methods

3.1. Data Collection

Groundwater and surface water samples are systematically acquired in August 2024 within the plain region of Hotan River. The coordinates of sampling locations are precisely recorded by a GPS device (Figure 1c). A total of 51 water samples is acquired from the wells, streams, and reservoirs, including 21 surface water samples and 30 groundwater samples. Groundwater samples are collected from the wells used for domestic water supply, agricultural irrigation, and ecological irrigation in the study area. An amount of 250 mL of water sample is filtered through a 0.45 μm filter in situ, and then collects in high-density polyethylene bottle. All water samples are stored in a 4 °C refrigerator until laboratory analysis.

The main testing parameters for water samples include pH, TDS (mg/L), EC (electrical conductivity, μS/cm), Ca²⁺ (mg/L), K⁺ (mg/L), Na⁺ (mg/L), Mg²⁺ (mg/L), SO₄²⁻ (mg/L), Cl⁻ (mg/L), HCO₃⁻ (mg/L), CO₃²⁻ (mg/L), δ²H (‰), and δ¹⁸O (‰). At the water sampling site, pH, TDS, EC, and water temperature of samples are directly measured in situ using a YSI ProPlus instrument. Na⁺, Ca²⁺, K⁺, Mg²⁺, Cl⁻ and SO₄²⁻ are measured via ion chromatograph (Dionex-320), while HCO₃⁻ and CO₃²⁻ are analyzed via the titration method. The precision of ion measurement is below 1%. Moreover, the accuracy and reliability of the ion measurement is evaluated by the ion charge balance error (ICBE) for all water samples using Equation (1) [29]:

$$ICBE\left(\%\right)={\displaystyle \frac{\sum cations-\sum anions}{\sum cations+\sum anions}}\times 100$$

(1)

The δ²H and δ¹⁸O of water samples are determined using an LGR DLT-100 water isotope analyzer. The stable isotopes results are expressed in the delta (δ) notation in per million (‰), and defined relative to VSMOW (Vienna Standard Mean Ocean Water). The equation is given as Equation (2) [30]:

$${\delta }_{sample}=\left({\displaystyle \frac{R_{sample}-R_{standard}}{R_{standard}}}\right)\times 1000$$

(2)

where R represents the isotopic ratio of water sample (²H/¹H and ¹⁸O/¹⁶O) and $R_{standard}$ represents the corresponding ratio in the standard sample. The measurement precision is ±0.1‰ for δ¹⁸O and ±0.8‰ for δ²H.

Additionally, the stable isotopic data of precipitation are downloaded from the GNIP (Global Network of Isotopes in Precipitation) of the IAEA (International Atomic Energy Agency). For the Hotan station in Xingjiang Province, China, the weighted means of δ¹⁸O and δ²H for atmospheric precipitation are −5.47‰ and −32.63‰.

The grid data for air temperature and precipitation in the study area are collected from the National Earth System Science Data Center, National Science & Technology Infrastructure of China (http://www.geodata.cn (accessed on 1 May 2025)), with a monthly temporal resolution and 1 km spatial resolution. The NDVI data for the Hotan River basin, spanning 2000 to 2023, are sourced from the MODIS NDVI product (MOD13Q1), which offers a 250 m resolution (available at https://www.earthdata.nasa.gov/data/catalog/lpcloud-mod13q1-061 (accessed on 1 January 2025)). The LULC data for the Hotan River basin from 2000 to 2023 are obtained from the China Land Cover Dataset (CLCD; https://zenodo.org/records/8176941 (accessed on 15 May 2025)), with a 30 m spatial resolution [31]. In this study, the spatial averages of these variables are used to characterize the entire basin, including basin-wide precipitation, air temperature, and the NDVI.

3.2. Irrigation Suitability Assessment Methods

The hydrochemical characteristics of surface water and groundwater could indicate its suitability for agricultural irrigation. In this study, two key indices are employed to evaluate the irrigation suitability of surface water and groundwater resources: soluble sodium percentage (Na%) and sodium adsorption ratio (SAR). The Na% and SAR are applied for alkalinity hazard assessment, while EC is used for salinity hazard assessment, in addition to the USSL and Wilcox classification. This method considers the potential impacts of sodium ions and salinity on soil structure and plant growth. The SAR represents the relative activity of sodium ions in irrigation water, which could exchange with soil [32]. High SAR and Na% levels in irrigation water could cause soil salinization, thus adversely influencing plant growth and crop yield. Irrigation water with a high Na% could increase soil salinization risk. Prolonged irrigation with a high Na% and SAR would compound soil salinization risk, gradually altering the chemical and physical properties of soil, thus leading to the reducing of agricultural production [33]. The equations are given as Equations (3) and (4) [32]:

$$SAR={\displaystyle \frac{{Na}^{+}}{\sqrt{\frac{{Ca}^{2+}+{Mg}^{2+}}{2}}}}$$

(3)

$$Na\%={\displaystyle \frac{{Na}^{+}}{{Ca}^{2+}+{Mg}^{2+}+{Na}^{+}+K^{+}}}\times 100\%$$

(4)

where the unit of all ions is expressed in meq/L. The criteria of classifying irrigation water suitability based on the SAR and EC is shown in

Table 1. Criteria of classifying irrigation water suitability based on SAR and EC [34].

Alkalinity Hazard (SAR)	Salinity Hazard (EC)	Irrigation Water Suitability
Low: <10	Low: <250	Excellent quality
Medium: 10–18	Medium: 250–750	Good quality
High: 18–26	High: 750–2250	Acceptable quality
Very high: >26	Very high: >2250	Unacceptable quality

.

3.3. Statistical Analyses

In this study, statistical analysis and plotting are performed employed the IBM SPSS 27, ArcGIS 10.6, and Origin 2021. The Piper, Gibbs, and ion mixing plots are used to reveal the hydrochemical type and evolution mechanisms of surface water and groundwater [23].

4. Results

4.1. Variations in Climate, Vegetation, and Land Use

Figure 2 shows the changes in the annual precipitation, air temperature, NDVI, and cultivated land area in the study area from 2000 to 2023. In general, the annual precipitation demonstrates large fluctuations, with no obvious change trend from 2000 to 2023 (Figure 2a). The annual average air temperature shows a slightly downward trend during this period, decreasing by 1 °C (from 13.4 °C in 2000 to 12.4 °C in 2023). However, the values of the NDVI and cultivated land area exhibit an upward trend during 2000 to 2023, increasing by 0.03 (from 0.11 in 2000 to 0.14 in 2023) and 658.9 km² (from 2954.8 km² in 2000 to 3613.7 km² in 2023), respectively. This indicates that the increasing vegetation in the plain of Hotan River is mainly attributed to the increasing cultivated land due to agriculture development (large-scale land reclamation), while it has a weak relationship with local precipitation and temperature. Especially after 2010, the cultivated land area has increased significantly, mainly converting from grassland and bare land (Figure 2d).

Figure 3 shows the LULC spatial patterns in the plain region of the Hotan River basin from 2000 to 2023. Cultivated land is mainly distributed in the middle of the plain region, grassland and forest primarily in the middle and southern margin, and bare land in the north, middle, and south (Figure 3a–c). Additionally, Figure 3d,e illustrate the area and percentage change in different land cover types in the study area over the period 2000 to 2023. As shown in Figure 3, the dominant land cover type is bare land, and its area coverage is more than 80% of the study area (84.83%, 83.31%, and 82.41% in 2000, 2010, and 2023, respectively). Grassland is second (10.46%, 11.21%, and 11.82% in 2000, 2010, and 2023, respectively), and cultivated land is third in dominance (4.66%, 5.41%, and 5.70% in 2000, 2010, and 2023, respectively). Generally, the area of grassland, artificial surfaces, and cultivated land gradually increases from 2000 to 2023, while bare land area gradually decreases (Figure 3d). Overall, the area coverage of different land cover types changes significantly from 2000 to 2010 in the study area, while it changes slightly from 2010 to 2023 (Figure 3e). From 2000 to 2023, the highest area percentage change occurred in artificial surfaces (increasing by 151.6% between 2000 and 2023), followed by cultivated land (increasing by 22.3% between 2000 and 2023). On the contrary, the area of bare land undergoes a persistent decrease by 2.9% between 2000 and 2023. It is obvious that the increasing area of cultivated land, grassland, and forest was transferred primarily from bare land within the study area from 2000 to 2023. The area of water bodies and forest undergoes a general increase by 5.0% and 27.3% between 2000 and 2010, respectively, while decreasing by 13.9% and 7.1% between 2010 and 2023, respectively (Figure 3e).

4.2. Characteristics of Major Dissolved Ions

The ICBE of all water samples in this study is calculated using Equation (1). The calculation results show that the ICBE of water samples is within the acceptable range (−5% < ICBE < 5%), indicating the suitability of hydrochemical data. Figure 4 shows the violin plot of hydrochemical characteristics for surface water and groundwater samples collected within the study area. The distribution of various ions for groundwater is evidently wider than that for surface water, an indication of the complex and remarkable spatial difference in hydrochemical characteristics in the plain region (Figure 4). Except for pH, the mean concentrations of all the ions in groundwater are much higher than those in surface water, with the mean TDS, Cl⁻, and SO₄²⁻ of 1329.2 mg/L, 323.3 mg/L, and 400.5 mg/L, respectively. The 25th percentiles of groundwater TDS are much higher than 1000 mg/L, implying the widespread distribution of brackish water [35]. While surface water in the study area is more diluted, with a low TDS (Cl⁻ = 93.2 mg/L, TDS = 428.7 mg/L). The pH values are 8.6 for surface water and 8.3 for groundwater, implying that the water in the study area was a slightly alkaline.

The Piper trilinear diagram is prepared to reveal the differences and similarities among samples [36]. Figure 5 exhibits the Piper plots for surface water and groundwater in the study area. Na⁺ concentration (4.0 meq/L) is the primary cation in surface water samples, followed by Ca²⁺ (3.2 meq/L). By contrast, SO₄²⁻ (2.9 meq/L) and Cl⁻ (2.6 meq/L) are the dominant anions in surface water. As for the groundwater in this study area, Cl⁻ (9.1 meq/L) and SO₄²⁻ (8.3 meq/L) prevail as the dominant anions, while Na⁺ (14.1 meq/L) prevail as the dominant cation. Furthermore, the groundwater samples from the Yutian, Moyu, Minfeng, Luopu, and Hotan counties have similar hydrochemical type (Cl-SO₄-Na), but were different from the city of Hotan and the Pishan and Cele counties (SO₄-Cl-Na). In addition, surface water samples from Cele and Yutian counties belong to the Cl-SO₄-Na-Ca hydrochemical type, while are HCO₃-SO₄-Ca-Na hydrochemical type for Hotan the city and Hotan County, HCO₃-Ca for Luopu County, SO₄-Cl-Na-Ca for Minfeng County, and SO₄-Cl-Ca-Na for Moyu and Pishan counties (Figure 5).

As shown in Figure 6a,b, TDS concentrations of surface water and groundwater samples show marked spatial heterogeneity in the plain region. Generally, TDS values of stream water and shallow groundwater increase along the flow paths from the midstream to downstream. Overall, the water samples have high TDS and poor water quality at the counties located in the western and eastern margins, while there is low TDS and good water quality in the middle counties. TDS value in surface water samples varies from 132 to 1536 mg/L (Figure 6a). The high TDS values of surface water are found in the Minfeng (943.0 mg/L), Cele (531.0 mg/L), and Pishan (402.0 mg/L) counties, while the low TDS values are found in Hotan city (132.0 mg/L) and the Hotan (148.0 mg/L), Moyu (194.0 mg/L), and Luopu (299.0 mg/L) counties. Furthermore, the TDS value in groundwater samples varies from 516 to 2684 mg/L, implying the widespread distribution of brackish water (Figure 6b). The high TDS of groundwater is found in the Pishan (1670.7 mg/L), Hotan city (1530.0 mg/L), Minfeng (1442.0 mg/L), Yutian (1406.7 mg/L) and Cele (1385.0 mg/L) counties, while the low values are found in the Moyu (1033.0 mg/L), Hotan (1230.0 mg/L) and Luopu (1238.5 mg/L) counties.

4.3. Characteristics of Stable Water Isotopes

Figure 7 exhibits the relationship between δ²H and δ¹⁸O for surface water and groundwater samples in the study area. Stable isotopic compositions of water samples show a large range from −69.3‰ to −26.5‰ for δ²H and −10.8‰ to −4.2‰ for δ¹⁸O in the plain region of the Hotan River basin. In general, the δ¹⁸O values of water samples differ noticeably among sampling sites (Figure 6c,d). Obviously, the mean values of isotopic compositions for groundwater are very close to those of surface water. In addition, the majority of water samples are located near the local meteoric water line in the Tarim Basin (LMWL: ${\delta }^2\mathrm{H}=8{\delta }^{18}O+15$) [37], while above the global meteoric water line (GMWL: ${\delta }^2\mathrm{H}=8{\delta }^{18}O+10$) [30]. Furthermore, the stable isotopic data define a regression line of ${\delta }^2\mathrm{H}=7.3{\delta }^{18}O+10.8$ (R² = 0.90, n = 21) from surface water samples and ${\delta }^2\mathrm{H}=7.7{\delta }^{18}O+12.2$ (R² = 0.92, n = 30) from groundwater samples (Figure 7). The mean values of surface water isotopes (δ¹⁸O = −7.9‰) differ noticeably from the atmospheric precipitation isotopes for the Hotan station (δ¹⁸O = −5.5‰) in Xinjiang Province, China. This indicates that surface water in the plain region is not primarily supplied by local precipitation.

Figure 6c,d exhibit the spatial patterns of water samples δ¹⁸O values in the plain region of the Hotan River basin. Overall, isotopic signatures of groundwater and surface water in the plain region showed marked spatial heterogeneity, and differ noticeably among various counties (Figure 6 and Figure 8). The δ¹⁸O values of the surface water samples extend from −10.8‰ to −4.2‰ (Figure 6c). The high δ¹⁸O values of surface water are found in the Cele (−6.2‰), Minfeng (−6.8‰), and Pishan (−6.8‰) counties, while the low values are found in Hotan city (−9.8‰) and the Luopu (−10.8‰) and Hotan (−10.7‰) counties. Additionally, the δ¹⁸O values of the groundwater samples vary from −9.4‰ to −5.4‰ (Figure 6d). The groundwater δ¹⁸O values are low in Hotan city (−8.7‰) and the Luopu (−9.3‰) and Minfeng (−8.8‰) counties, while the high values are found in the Cele (−6.3‰), Pishan (−6.9‰), and Yutian (−7.2‰) counties. Moreover, groundwater and surface water samples in Minfeng County are located below the LMWL (Figure 8).

4.4. Irrigation Water Quality of Groundwater

Overall, according to the USSL diagram (Figure 9a), 62% of the surface water samples have low alkalinity and salinity, and are distributed in the Cele, Yutian, Moyu, Pishan, and Luopu counties, with good water quality for irrigation (Table 1). Only 24% of the surface water samples have moderate alkalinity and salinity, with a moderate water quality, which were distributed in the Minfeng, Pishan, Yutian, and Cele counties. However, 14% of the surface water samples have high alkalinity and salinity, with bad water quality, which are distributed in Hotan city and the Minfeng and Hotan counties. Among the groundwater samples, 47% have moderate salinity and alkalinity, with a moderate water quality, mainly located in the Luopu, Pishan, Yutian, Moyu, and Cele counties. While 53% of the groundwater samples have high salinity and alkalinity, with bad water quality, mainly located in Hotan city and the Minfeng, Luopu, Hotan, Cele, Yutian, and Pishan counties.

To further assess the irrigation suitability of groundwater in the plain region, the Wilcox diagram is prepared (Figure 9b). As shown in Figure 9b, surface water samples are classified as “Excellent to Good”, “Good to Permissible,” and “Doubtful to Unsuitable”. The percentages of these categories are 62% for “Excellent to Good” (originated from Pishan, Moyu, Luopu, Cele, Yutian, and Minfeng counties), 14% for “Good to Permissible” (originated from Cele, Yutian, and Pishan counties), and 24% for “Doubtful to Unsuitable” (originated from Hotan city and Minfeng and Hotan counties). Furthermore, groundwater samples are classified into five categories, 3% for “Excellent to Good” (from Pishan County), 23% for “Good to Permissible” (from Yutian, Luopu, Moyu, and Cele counties), 13% for “Permissible to Doubtful” (from Minfeng, Yutian, Moyu, and Cele counties), 47% for “Doubtful to Unsuitable” (from Hotan city and Yutian, Luopu, Moyu, Pishan, Cele, and Hotan counties), and 13% for “Unsuitable” (from Yutian, Pishan, and Minfeng counties; Figure 9b).

5. Discussion

5.1. Irrigation Suitability Assessment

Groundwater is an important source of agricultural irrigation in the Hotan River basin, which is vital for ecosystem sustainability and agricultural production [24]. The Hotan River basin is a typical agricultural region in the Tarim Basin, and irrigation depends on surface water and groundwater [17]. In 2016, the groundwater exploitation amount in the Tarim Basin was about 48.33 × 10⁸ m³, and 89.1% of the total amount was used for agricultural irrigation (data from the Water Resources Bulletin of Xinjiang Uygur Autonomous Region, China). Therefore, irrigation suitability of surface water and groundwater in our study area is evaluated using the alkalinity hazard (SAR and Na%) and salinity hazard (EC) values (Figure 9).

Overall, surface water samples are predominantly distributed in the low alkalinity (S1) and relatively low salinity (C1 and C2), accounting for 62% of all surface water samples, while only 14% exhibit poor water quality (Figure 9). This indicates that surface water in majority of the plain region is good and appropriate for irrigation at present. Thus, it is inferred that crops and plants are less affected by soil alkalinity and salinity hazards from surface water irrigation, and sustained irrigation with stream water would not cause soil problems and not impact crop production [13,32,34]. However, most groundwater samples fall within the high salinity and alkalinity categories, accounting for 60% of all groundwater samples, indicating the water quality for “Doubtful to Unsuitable” and “Unsuitable.” These samples are distributed in most counties of the study area. Therefore, shallow groundwater (well depth < 100 m) in the plain region is marginally suitable or unsuitable for irrigation at present (Na% > 40% for most samples). Thus, plants and crops in the plain region are mostly affected by soil alkalinity and salinity hazards from groundwater irrigation, and long-term use of shallow groundwater would cause soil problems and affect crop production [32,38]. Thus, groundwater quality protection ought to be considered in arid plain regions in the context of extreme climatic events and anthropogenic activities [13].

5.2. Influence of Agricultural Activities on Groundwater Hydrochemistry

Water hydrochemistry in the arid plain region is largely controlled by natural processes (e.g., water–rock interactions) and anthropogenic impacts, especially agricultural activities [39]. The Gibbs diagram has been extensively applied to study the hydrochemical evolution mechanisms of water, which is used to evaluate the hydrochemical processes and reactions occurred along flow paths in this study [27]. Rock weathering is the main mechanism controlling the surface water hydrochemistry in our study area, while groundwater hydrochemistry is primarily governed by evaporation and rock weathering (Figure 5 and Figure 10a). Moreover, the plot of Mg²⁺/Ca²⁺ versus Mg²⁺/Na⁺ is used to evaluate the hydrochemical processes of soil salt leaching (affected by human activities) and evaporation [40]. In the study area, evaporation is the major force governing the groundwater hydrochemistry, while surface water samples are mainly governed by soil salt leaching (Na and Ca salts) and rock interaction (Figure 10b) [13].

Furthermore, the dissolved species and their relationships can reveal the solute origin [41]. The mixing diagrams of Na-normalized ratios could reveal the sources and determining factors of primary hydrochemical compositions in water [42]. The dissolution and precipitation of minerals are important and common processes impacting the solutes in surface water and groundwater within the arid plain region [23]. As shown in Figure 10c,d, majority of surface water and groundwater samples are located between the evaporite and silicate end-members, indicating the dominance of evaporate dissolution (e.g., halite) and silicate weathering. It may be attributed to the complex lithology in the study area [43]. That is, the weathering of silicates and the dissolution of evaporite minerals are the effective processes for releasing Na⁺ into groundwater and streamflow [42]. Soil salinization in the plain region of the Hotan River basin is serious, significantly affecting the hydrochemical compositions of phreatic water [17].

Agricultural activities have had a profound impact on groundwater in arid areas, especially since 2000 [15]. In general, agricultural activities have changed the recharge, flow, discharge, and hydrochemical processes of groundwater, which are the dominant driving force for the evolution of groundwater environment in arid plains [13]. The expansion of cultivated land area and arid climatic conditions exacerbate the evaporation and salinization of phreatic water in the arid plain region (

Table 2. Information related to land use types, climate condition, groundwater, and human activities in the plain region of the Hotan River basin during 2002 to 2020.

Parameters	2002	2010	2014	2020
Precipitation (mm)	29.0	17.5	42.5	35.5
Air temperature (°C)	12.6	12.7	12.6	11.9
Cropland area (km2)	2880	3433	3642	3770
Bare land area (km2)	53,399	52,825	51,937	52,678
Depth to phreatic water level (m)	/	15.1	17.1	18.6
Total groundwater extraction (108 m3)	2.6	3.7	4.6	6.0
Water-saving irrigation area (km)	/	107	386	/
Irrigation water amount (108 m3)	25.8	44.7	44.8	38.8

). After 2000, the cultivated land area in the study area has expanded significantly, and the amount of irrigation water has increased accordingly, resulting in groundwater overexploitation [12]. In the Tarim Basin, groundwater has gradually become the main source of agricultural irrigation due to the limited surface water resources [17]. In the study area, groundwater withdrawal increases significantly from 2.6 × 10⁸ m³ in 2002 to 6.0 × 10⁸ m³ in 2020, and the irrigation water amount increases from 25.8 × 10⁸ m³ in 2002 to 38.8 × 10⁸ m³ in 2020 (Table 2). However, the groundwater level in the study area decreases by 3.5 m from 2010 to 2020. Furthermore, irrigation and excessive extraction of deep groundwater have changed the groundwater flow field and recharge path in the plain region, aggravating the salt accumulation, and saltwater process of shallow groundwater [12]. Thus, it causes the pollution of deep groundwater and aggravates the interaction between salt and fresh water within the aquifers [27]. In the oasis area of the Hotan River basin, the land cover types are dominated by cultivated land and artificial forest, and the TDS value of shallow groundwater is large, causing the processes of evaporation, migration, and deposition (Figure 6) Thus it promotes the enrichment of salt into soil and aquifers [23]. Evaporites and saline soil are widespread in the oasis area of Tarim Basin, due to strong evaporation, irrational agricultural irrigation, and inefficient drainage [36]. The mean TDS of groundwater in the plain region of the Hotan River basin (TDS = 1329.2 mg/L) is lower than that in the Luntai Oasis (TDS = 9025.5 mg/L) of Tarim Basin [27]. In addition, water-saving irrigation is widely promoted since 2000, effectively reducing the evapotranspiration and irrigation return water, but also reducing the infiltration amount into deep aquifer (Table 2).

Moreover, agricultural activities in the arid plain region could also affect the Cl⁻ and NO₃⁻ concentration in soil and groundwater, such as organochlorine pesticides and fertilizers [44]. The nitrogen compounds in groundwater are usually from the application of nitrogenous fertilizers (e.g., urea) in agricultural irrigation districts [45]. Urea is widely applied in the plain region of Tarim Basin, which is an organic nitrogenous fertilizer [13]. In the Hotan River basin, the application amount of nitrogenous fertilizer was 0.210 tons per hectare in 2001, and 0.188 tons per hectare in 2020 (data from the Statistical Yearbook of Hotan Regional, Xinjiang Uygur Autonomous Region, China). However, we did not analyze the variation characteristics of pollutants (e.g., pesticides, fertilizers, and heavy metals) in soil and groundwater, and their impacts on groundwater quality in the phreatic and confined aquifers. Future research should focus on these aspects.

5.3. Interaction Between Groundwater and Surface Water

Long-term groundwater overexploitation caused by agricultural irrigation has led to the depletion of groundwater reserve and the decline of groundwater level, which has changed the flow path of groundwater and its interaction with surface water in the arid plain region, thus affecting the hydrochemical evolution of groundwater [12]. The interaction between surface water and groundwater is complex, with the pollutants from agricultural practices affecting both systems [13]. Thus, the hydraulic connectivity between surface water and groundwater facilitates the spread of various contaminants. The stable isotopic and TDS signals could indicate the recharge sources and flow paths of groundwater (Figure 6 and Figure 7). Generally, groundwater in the piedmont alluvial plain region flows from the south to north, and is primarily recharged by the infiltration of rainfall, stream water, and irrigation return flow, as well as the Quaternary pore water and bedrock fissure water from mountainous areas [24]. Meanwhile, the discharge pathway of groundwater is mainly the phreatic water evaporation and spring flow, and the exploitation from phreatic and confined aquifers for agricultural irrigation and industry [16].

During the development period of rapid agriculture (after 1990), there is an intense and frequent exchange between surface water and groundwater in our study area. In the arid plain region, agricultural irrigation relied on streamflow and groundwater, leading to the irrigation return flow as a main source of shallow groundwater [27]. Streamflow in the plain region is dominantly recharged by the rainfall, meltwater, and lateral groundwater from mountainous areas, and could also change from soil erosion. Large-scale water resource control projects (e.g., reservoirs and water diversion projects) have been constructed in the Hotan River basin, and more than 80% of streamflow is directly diverted to the fields for irrigation, occurring intense evaporation along the flow path. This directly changes the way of groundwater recharge, shifting from natural infiltration to primarily artificial recharge (channel and reservoir leakage), as well as reducing the recharge amount of groundwater from streamflow [12]. In addition, irrigation water is from the streamflow diversion and pumping groundwater in the arid plain region. The vertical infiltration of irrigation water could leach and dissolve the soluble mineral and soil salt, and thus transfer the salt downward into shallow groundwater, causing groundwater salinization [46]. Part of residual fertilizers and pesticides in the soil could also percolate into aquifers via the irrigation water infiltrating. That is, agricultural activities promote the migration of surface water contaminants into groundwater. Meanwhile, the DGL (depth to groundwater level) of approximately half of monitoring wells in the study area is less than 7 m. Thus, the phreatic water in the plain region of the Hotan River basin could undergo evaporation, then causing the increase in groundwater salinity [13,16].

Furthermore, agricultural activities could lead to shallow groundwater salinization in the arid plain region, even the deterioration of confined water environment, which threatens natural desert vegetation. Groundwater (especially confined water) is overexploited for agricultural irrigation during the crop growing season, altering the leaking direction between confined and phreatic aquifers due to the formation of groundwater drawdown funnels [14]. Thus, phreatic water would leak downward into confined water by the groundwater hydraulic gradient, so the salt in the phreatic aquifer would also migrate into the deep groundwater, causing the mixing between salt phreatic water and fresh confined water [16]. Obviously, in the plain region of Tarim Basin, there is a close hydraulic relationship (water and salt exchange) between groundwater and surface water under the influence of human activities, and the transformation is frequent. This is an important reason for the complicated groundwater circulation. The increased groundwater recharge from agricultural irrigation may not compensate for the decreased groundwater level by groundwater pumping, thus resulting in a reducing base flow into streams and wetlands [28].

5.4. Recommendations for Water Resource Management

Our results above imply that extensive land reclamation, driven by irrigation agricultural development, has caused several consequences for water resource management and ecosystem security in the study area [26]. (ⅰ) Agricultural water consumption in the plain of the Hotan River basin accounts for more than 90% of the total water consumption, and even more than 95% in some regions. Thus, agricultural water use occupies the industrial and ecological water use [17]. (ii) The irrigation patterns are extensive in the study area, and the promotion and maintenance of efficient water-saving irrigation (e.g., drip irrigation) are insufficient, leading to a limited actual water-saving effect [13]. (iii) The agricultural planting structure is not optimized, and the planting proportion of high water-consuming crops (e.g., wheat, corn, and cotton) is high, which does not match the local water resource conditions [15]. (iv) The dynamic monitoring and early warning system for water resources is not perfect, especially for groundwater [27]. (v) Climate warming accelerates the melting of glaciers, and the trend of glacier retreat in the Kunlun Mountains is obvious. Thus, the future supply of water resources faces uncertain risks [17].

In order to sustainably manage water resources and maintain ecosystem stability in arid regions, we propound some recommendations as following: (ⅰ) In the arid plain region, the area of cultivated land should be strictly controlled, and the total amount of agricultural irrigation water should be restricted [12]. (ii) Efficient water-saving irrigation (e.g., drip irrigation) should be widely promoted and maintained, and the planting structure should be optimized, reducing the area of low efficiency and high water-consumption crops. Thus, the efficiency of water resource utilization would be enhanced [17]. (iii) The allocation of water resources should be optimized, and the modernization of storage projects and canal systems should be accelerated to reduce leakage and ensure the ecological flow of rivers [15]. (iv) The monitoring system of surface water and groundwater should be improved, and the total amount of groundwater exploitation should be limited, thus avoiding further groundwater overexploitation. (v) Abundant floodwater in the piedmont should be introduced into the ground by natural infiltration, and the storage and regulation function of an underground reservoir should be given play fully [12]. This could greatly increase the recharge amount of groundwater and thus would improve the connectivity among groundwater, surface water, and ecosystems.

6. Conclusions

In this study, the irrigation water quality and groundwater–surface water interaction influenced by agricultural activities in a typical arid plain region of the Hotan River basin are evaluated using hydrochemical and stable isotopic data from 51 water samples. Based on the analysis of hydrochemical characteristics, groundwater irrigation quality is assessed, and the impact of agricultural activities on groundwater quality is evaluated. Subsequently, the interaction between groundwater and surface water is examined, and the recommendations for groundwater resource management are propounded. The following conclusions would be drawn:

(i) From 2000 to 2023, the annual average NDVI and cultivated land area increased with time, increasing by 0.03 and 658.9 km², respectively, which was predominantly resulting from the conversion of bare land. TDS values of stream water and groundwater exhibited marked spatial heterogeneity in the plain region, increasing along the flow paths. Groundwater TDS ranged from 516 to 2684 mg/L and was high in the western and eastern margins, implying the widespread distribution of brackish water. Cl⁻, SO₄²⁻, and Na⁺ were the dominant ions in groundwater, with a widespread distribution of brackish water. Groundwater δ¹⁸O value ranged from −9.4‰ to −5.4‰, with the mean value close to surface water. The regression line for groundwater was ${\delta }^2\mathrm{H}=7.7{\delta }^{18}O+12.2$ (R² = 0.92, n = 30).

(ii) In total, 86% of the surface water samples in the plain region are good and suitable for agricultural irrigation. However, 60% of shallow groundwater samples in the plain region are marginally suitable or unsuitable for irrigation at present. Long-term use of shallow groundwater could cause soil problems and affect crop production. Groundwater hydrochemistry in the arid plain region is largely controlled by intensive evaporation, water–rock interaction (rock weathering, dissolution of silicate, and evaporite minerals), and agricultural activities (cultivated land expansion, irrigation, groundwater exploitation, and fertilizers).

(iii) During a period of rapid agricultural development, agricultural activities have profoundly affected the recharge and discharge, flow path, and hydrochemical evolution of groundwater, demonstrating an intense and frequent interaction between groundwater and surface water in arid plain regions. Agricultural activities could lead to shallow groundwater salinization, even the deterioration of confined water environment. Groundwater overexploitation could alter the leaking direction between confined and phreatic aquifers, and phreatic water could leak downward into confined water, causing the mixing between salt phreatic water and fresh confined water.

(iv) Hence, sustainable groundwater management should be implemented to protect groundwater environment in arid plain regions. Controlling cultivated land area and agricultural irrigation water amount, enhancing water resource utilization efficiency, limiting groundwater exploitation, and fully utilizing floodwater resources would be effective methods to restore groundwater quality and maintain ecosystem stability in the context of extreme climatic events and anthropogenic activities.