Introduction

Cultivated land, as a special natural resource, is associated with national food security, benign economic development, social harmony, and stability1,2,3. In long-term agricultural practices, parent materials, climate, topography, biology, time (natural factors) and anthropic activities (human-induced factors) are considered the six major factors affecting cultivated land soil fertility4. Cultivated land soil fertility was also strongly influenced by regional social factors5. Anthropogenic activities are a major factor influencing pedogenesis, particularly in managed landscapes4. Globally, significant variations exist in geographical conditions, land utilization patterns, and soil characteristics across different regions6,7. A study in northwestern Chong-qing, China8 analyzed the chemical properties of the soil under the influence of anthropogenic activities (tillage perturbation and engineering perturbation) and found that anthropic activities changed the soil fertility level. In a representative region of the Horqin Sandy Land, when comparing an uncultivated native grassland control with some areas that had been continuously cultivated for 60 years, significant changes in soil properties were observed9. And after 5 years of reclamation, the soil bulk density increased significantly and then decreased steadily, organic carbon decreased by 60%, while total nitrogen, available phosphorus, and available potassium increased. In the North China Plain, compared with flood irrigation, drip irrigation led to a notable increase in soil organic matter content within the depth range of 10–20 cm10. In subtropical India, when compared to conventional irrigation, drip irrigation significantly enhanced the soil available nitrogen content, with increments ranging from 6 to 8%, demonstrating a positive correlation between drip irrigation and soil nutrient retention11.

Xinjiang is a typical continental arid climate zone. The Tianshan Mountains in the central part of Xinjiang divide Xinjiang into two major parts, which are northern Xinjiang and southern Xinjiang. The Tianshan Mountains exhibit pronounced climatic disparities between their northern and southern regions, characterized by distinct precipitation and temperature patterns12. Specifically, northern Xinjiang receives significantly higher mean annual precipitation compared to its southern counterpart, while southern Xinjiang maintains consistently higher mean annual temperatures than the northern region13. As a result of these climatic differences, the irrigation patterns in the north and south of Xinjiang are different. Water-saving irrigation methods developed at different paces in northern and southern Xinjiang due to climatic constraints14. Before 2010, drip irrigation in northern Xinjiang developed rapidly. Due to the slow technological development and lack subsidy programs, drip irrigation in southern Xinjiang started to develop after 201015. The proportion of water-saving irrigation in northern Xinjiang has been reported to be higher in northern Xinjiang14. While large-scale drip irrigation in northern Xinjiang is exclusively implemented during the crop growing season (April-September), southern Xinjiang employs a dual irrigation regime combining drip irrigation during the growing season with supplementary flood irrigation during winter and spring non-growing periods. Drip irrigation applied 480 mm annually in northern Xinjiang, whereas a combination of drip and flood irrigation in southern Xinjiang totalled 900 mm, following local farming practices.

Different irrigation patterns impacted the changes of soil physical and chemical properties16. Drip irrigation affected soil infiltration, effectively retained water and nutrients, improved irrigation water efficiency, and minimized chemical leaching from the root zone17. Flood irrigation has the limitations of large amount of irrigation, poor uniformity18, significantly regulated by environmental factors, difficulty in control19, and evaporation of the field20. Consequently, long-term flood irrigation can elevate groundwater levels over time21, increase the risk of secondary salinization of soil and affect cultivated land soil fertility22. A recent study conducted in the midstream of the Heihe River Basin of northwestern China, revealed that the irrigation amount reaches 1096 mm under flood irrigation, deep percolation account for 67%, potentially triggering secondary soil salinization in subsurface layers and groundwater contamination23. Proper irrigation volumes can prevent excessive nutrient losses and maintain soil fertility24. Inappropriate irrigation practices can reduce soil organic matter content25. Uncontrolled flood irrigation can lead to soil nitrogen leaching losses26 and soil salinization27, which eventually lead to low soil fertility. Abbasi et al.28 observed a significant increase in nitrogen leaching with elevated irrigation levels, with the 100% of the standard crop evapotranspiration treatment exhibiting the highest leaching risk. Drip irrigation can maintain soil moisture29, increase the soil temperature, and improve crop yield in extremely arid areas30. Research findings in the Hetao Irrigation District, China31 have established that due to its superior performance in carbon assimilation and light utilization efficiency, drip irrigation achieves an average yield increase of approximately 20.5% compared to flood irrigation. Compared with the rainfed conditions, under the drip irrigated conditions, the soil organic matter was significantly increased by 8.06%32. Especially in arid areas, drip irrigation played a major role in all the agronomic practices for enhancing soil quality and fertility33.

Significant regional variations manifest across distinct regions, and diverse irrigation practices exert a further influence on the physicochemical properties of soil. Owing to the long-term disparities in irrigation patterns, notable variations in soil quality have emerged between northern and southern Xinjiang. However, the pattern in which soil quality changes in northern and southern Xinjiang under different irrigation patterns remains unclear. Therefore, this study presents a comparative analysis of the long-term soil quality dynamics between northern and southern Xinjiang under varying irrigation patterns, grounded in a 35-year soil dataset. Based on the preliminary analysis, field sampling and validation were carried out in 2022 to verify the analytical results. Try to elucidate the evolutionary patterns of soil fertility of in the northern and southern Xinjiang, and propose actionable strategies for sustainable soil fertility management in other arid and semi-arid agricultural regions.

Materials and methods

Comparison of the study area

In the middle of the northern and southern sides of the Tianshan Mountains, there are rivers such as the Manas River in the Shihezi Irrigation District and the Aksu River in the Alar Irrigation District. The two geomorphic units share some similarities: (1) both are under the unified management model of the Xinjiang Production and Construction Corps; (2) they are respectively located on the northern and southern sides of the Tianshan Mountains, deriving irrigation water primarily from glacial meltwater; (3) their perpendicular distances to the Tianshan are similar. However, blocked by the Tianshan Mountains, they exhibit differences in climate, rainfall, and irrigation regimes. The Shihezi Irrigation District is characterized by long-term application of drip irrigation (DI) in Northern Xinjiang (NX), while the Alar Irrigation District is characterized by long-term application of drip irrigation during the growth season and flood irrigation during the non-growing season (DF) in Southern Xinjiang (SX). Thus, we selected these areas for a comparative analysis.

Shihezi Irrigation District (84° 58′ ~ 86° 24′ E,43° 26′ ~ 45° 20′ N) is located in the center of the northern foothills of Tianshan Mountain and the southern Junggar Basin, in Xinjiang, China (Fig. 1). The region has a temperate continental climate34, with an annual average sunshine time of about 2318–2732 h. The annual average temperature is 7.5–8.2 °C. The average annual rainfall is 180–270 mm, and the annual potential evaporation ranges from 1000 to 1942 mm35. The frost-free period of the crop is short (148–187 d). The Shihezi Irrigation District has three main soil types, namely gray desert soil, tidal soil, and meadow soil. The main crops in the irrigated area of Shihezi are wheat, maize and cotton, while the average amount of fertilizer applied was 7.115 × 104 kg/hm2. The irrigation period is generally from May to September, and the average irrigation times are 7–8 times in the Shihezi Irrigation District (Table 1)36.

Fig. 1
figure 1

Location of the study area. (a) Map of China; (b) Geographical location of Xinjiang; (c) (d) Location of the Shihezi Irrigation District and Alar Irrigation District. Note: the map was created using ArcGIS software (version10.8.1 http://www.esri.com).

Full size image
Table 1 Irrigation patterns, months and times of Shihezi Irrigation District and Alar Irrigation District.
Full size table

Alar Irrigation District (79° 22′ ~ 81° 53′ E, 40° 20′ ~ 41° 47′ N) is located in the southern foot of the Tianshan Mountains and the northern edge of the Taklimakan Basin, in Xinjiang China (Fig. 1). The region has a typical continental desert climate, with an annual average sunshine time of about 2654–3025 h37 and annual average temperature is 7.5–11.5 °C. The average annual rainfall is 19.8–81.4 mm, and the annual potential evaporation ranges from 1870 to 2660 mm38. Alar Irrigation District has mainly soil types such as tide soil, irrigation meadow soil, brown desert soil, salt soil and wind sand soil. The main crops in the irrigated area of Alar are mainly cotton, wheat and maize, while the average amount of fertilizer applied was 8.1045 × 104 kg/hm2. The irrigation period was divided into drip irrigation in growth season and flood irrigation in non-growing season, in Alar Irrigation District. The drip irrigation period generally lasted from April to September during the growing season, and the average irrigation times was 7–9 times. Non-reproductive irrigation was generally from December to March, and the average irrigation times was 1–2 times (Table 1)39.

All four sampling rounds followed the regional indicator criteria—considering terrain, soil, primary crops, and administrative boundaries. Proportional sampling was used each time (with varying points), and all regions received standard fertilization.

The distinct irrigation patterns and environmental conditions in these two regions make them suitable for comparing the long-term impacts of different irrigation regimes on soil quality.

Data sources and processing

The data employed in this study are sourced from records of typical soil fertility documented during the second national soil survey of China, investigation data regarding the cultivated land soil fertility of Xinjiang Production and Construction Corps, and the data of the third national land survey. Moreover, field sampling and validation were carried out in 2022 to corroborate the analytical results.

  1. 1.

    The typical soil fertility recorded during China’s second national soil survey. The data for cultivated land soil fertility were derived from "Xinjiang Soil" and "Xinjiang Soil Seed Annals" in 1985. The data set was generated based on the 1:1,000,000 scale soil map and 344 soil profiles of the second national soil survey of Xinjiang. In this study, the soil fertility of surface soils (those within a depth of 0–20 cm) was selected as the research object. This data set does not contain information on micronutrients;

  2. 2.

    The investigation data regarding the cultivated land soil fertility of Xinjiang Production and Construction Corps. Soil samples were collected after the main crop harvest in either October or November from 2005 to 2009. The research focuses on the surface soils (0–20 cm), with 15–20 points were collected from each mixed sample that covered an area of 6–13 hm2;

  3. 3.

    The data of the third national land survey were utilized. Soil samples were collected after the main crop harvest in either October or November in 2020. The research focuses on the surface soils (0–20 cm), with 15–20 sampling points collected from each mixed sample that covered an area of 6–13 hm2. There are 535 sample points in Shihezi Irrigation District and 333 sample points in Alar Irrigation District.

  4. 4.

    The data of field sampling and validation in 2022 were obtained. Considering various factors such as topography, land use patterns, and crop planting types in the irrigation area, 24 typical representative sampling points were selected. Among them, 14 were along the banks of the river basin of the Manas River irrigation area and 10 were in the Aksu River irrigation area. And, using a soil drilling rig (Geoprobe, AMS, United States), a total of 235 soil samples were obtained from May to July 2022 at different soil depths (0–30 cm, 30–60 cm, 60–100 cm, 100–200 cm, 200–300 cm, 300–400 cm, 400–500 cm), in order to avoid the short-term effects of surface layer caused by factors such as irrigation or fertilization. The collected samples were promptly stored in sealed plastic bags and carefully transported back to the laboratory for further analysis. This research employed soil depth data collected from the 0–100 cm.

Analysis methods

The data of determination and analysis

In accordance with the national standards for soil analysis methods (GB/T 42,363–2023), the physical and chemical properties of the soil samples were systematically measured. Specifically, the soil bulk density was determined using the ring knife drying method. For the measurement of soil pH, a soil pH meter (PB-10, Sartorius, Göttingen, Germany) was employed. Soil organic matter (SOM) content was estimated through the external heating method using potassium dichromate32. The Kjeldahl method was utilized to quantify the soil total nitrogen (TN). To assess the available phosphorus (AP), the sodium bicarbonate extraction-molybdenum-antimony colorimetric method was applied. Additionally, the available potassium (AK) was measured by means of flame photometry.

The integrated soil fertility assessment

By reference to the Cultivated Land quality grade (GB/T 33,469–2016), classification of type regions and fertility of cultivated land in China (NY/T 309–1996), five indexes of SOM, TN, pH, AP and AK were selected to grade the soil nutrient indexes of Xinjiang (Table 2). In this study, a modified Nemero Index was used to assess the integrated soil fertility separately in NX and SX40.

Table 2 Soil physical and chemical properties classification standards.
Full size table

The equations for calculating the nutrient fertility coefficient (IFIi) are:

$$IFIi = \left\{ {\begin{array}{*{20}l} {X/Xa} & {X \le Xa} \\ {1 + (X - Xa)/(Xb - Xa)} & {Xa \le X \le Xb} \\ {2 + (X - Xb)/(Xc - Xb)} & {Xb \le X \le Xc} \\ 3 & {X \ge Xc} \\ \end{array} } \right\}$$
(1)

The integrated soil fertility index (IFI) was calculated from the obtained nutrient fertility coefficient:

$$IFI = \sqrt {\frac{{Ave(IFIi)^{2} + Min(IFIi)^{2} }}{2}} \times \frac{(n - 1)}{n}$$
(2)

where X is the soil nutrient index value; Xa, Xb, and Xc are the thresholds of the index grade thresholds; Ave (IFIi) and Min (IFIi) represent the average and minimum values of fertility coefficients for all nutrient indicators across distinct sampling sites, respectively; N represents the number of indicators involved in the calculation, and in this study, n = 5. The IFI reference value is between 1 and 3 by this algorithm.

Statistical analysis

All the statistical analyses were performed with SPSS software 17.0. The effects of irrigation patterns on the soil fertility of cultivated land were investigated using Analysis of Variance (ANOVA). Significant differences of soil nutrient were determined with the LSD test at the 0.05 levels of probability. Correlation analysis and path analysis were performed using Pearson’s correlation coefficients and multiple regression analysis, respectively. All graphs were prepared using Excel, Arcgis10.8.1 and Origin 2022.

Results and analysis

Analysis on changes of soil physical and chemical properties from 1985 through 2020

Change in the soil bulk density and pH from 1985 through 2020

During the period from 1985 to 2009, the soil bulk density in SX remained relatively stable, maintained at 1.35 g∙cm−3, while the soil bulk density in NX slightly increased, no significant differences were observed (Fig. 2a). From 2009 to 2020, the soil bulk density in NX remained relatively stable, while the soil bulk density in SX slightly increased. No significant differences were observed among the treatments (p > 0.05). Overall, the soil bulk density in NX was consistently lower than in SX.

Fig. 2
figure 2

Plot of soil bulk density and soil pH from1985 through 2020 in NX and SX. (a) represents the change of soil bulk density content, (b) represents the change of pH.

Full size image

During the study period, the pH in NX and SX showed a slight decrease (Fig. 2b). In 2020, the pH in NX was lower than SX. No significant differences were observed among the treatments.

Change in soil organic matter and available nutrients from 1985 through 2020

Irrigation patterns significantly influenced SOM content in both 2009 and 2020. From 1985 to 2020, the SOM content of NX showed an overall upward trend. The SOM content in NX significantly increased (p < 0.05), rising from 1.8 g∙kg−1 in 1985 to 14.75 g∙kg−1 in 2020, with an annual average growth rate of 9.16% (Fig. 3a). After 2009, the increase in the SOM content of NX was less significant compared to that in the previous 25 years. In comparison with NX, the SOM content in SX exhibited a relatively smaller increase, rising from 2.46 g∙kg−1 in 1985 to 10.52 g∙kg−1 in 2009, with an annual average growth rate 6.24%. The SOM content of SX showed a slight downward trend and no statistically significant difference was found, after 2009. Overall, the SOM content in NX was higher than that in SX.

Fig. 3
figure 3

Changes of soil organic matter and available nutrients from 1985 through 2020 in NX and SX. (a) represents the change of SOM content, (b) represents the change of TN content, (c) represents the change of AP content, (d) represents the change of AK content.

Full size image

Different irrigation patterns had a great influence on TN in 2009 and 2020. Over the period from 1985 to 2009, TN content in SX had an increase trend, which was significantly higher than that in NX. The TN content in SX increased from 0.1 g∙kg⁻1 to 0.74 g∙kg⁻1, with an annual average growth rate of 8.70% (Fig. 3b). In comparison with SX, the growth of TN content in NX was relatively slow before 2009, and the annual average growth rate decreased by 1.34%. However, after 2009, the TN content of NX showed a continuous increase, while the TN content in SX showed a downward trend and was significantly lower than that in NX, reaching only 0.58 g∙kg−1 in 2020. In general, DI excelled in increasing the content of TN, which was more conducive to enhancing soil fertility.

Different irrigation patterns influenced the AP content to varying degrees. There was statistically significant difference in 2009. From 1985 to 2009, in SX, the AP content exhibited an increasing trend and was significantly higher than that in NX. It increased from 7.82 mg∙kg−1 to 26.54 mg∙kg−1, with an annual average growth rate of 5.22% (Fig. 3c). After 2009, the AP content in SX slightly declined and no statistically significant difference was found. However, owing to the elevated phosphorus fertilizer input in fertigation, the growth of AP in NX accelerated after 2009, increasing from 15.42 mg∙kg−1 to 31.00 mg∙kg−1, with an annual average growth rate of 6.55%, which was significantly higher than that in SX. In terms of the long-term effect, AP content in NX was superior to SX.

Irrigation patterns significantly affected the content of AK in NX and SX (p < 0.05). We found that the AK content in NX presented an overall increasing trend, increasing from 236.97 mg∙kg−1 to 350.26 mg∙kg−1, with annual average growth rate at 1.12% (Fig. 3d). Due to desalination irrigation, the most significant decline of the AK content in SX was observed between 1985 and 2009, decreasing form 186.65 mg∙kg−1 to 142.21 mg∙kg−1, with an annual average decrease of 1.13%. The AK content of SX remained relatively stable and slowly increased between 2009 and 2020. The AK content of NX was consistently higher than that of SX.

Change in beneficial soil micronutrients status from 2009 through 2020

Irrigation patterns affected the content of available copper in NX and SX. Specifically, the available copper content in SX was higher than NX, and both exhibited a slight increase (Table 3). However, the change trends of available boron in NX and SX was consistent and the available boron content in NX and SX slightly increased over time.

Table 3 Descriptive statistics of soil micronutrients from 2009 through 2020 in NX and SX.
Full size table

Different irrigation patterns had a great influence on available iron in 2009 and 2020. The available iron content in NX and SX both demonstrated an increasing trend over time. The available iron content in SX showed an increasing tendency between 2009 and 2020, increasing from 9.42 mg∙kg−1 to 30.73 mg∙kg−1, with an annual average growth rate of 11.34%. In comparison with SX, the available iron in NX exhibited a relatively smaller increase, with an annual average growth rate of 6.79%. In general, DF excelled in increasing available iron.

There was a significant difference in 2009 in available manganese between different patterns. The available manganese content of SX showed an upward trend, with an annual average growth rate of 8.46%. Compared to SX, the available manganese content in NX demonstrated a decreasing trend, with an annual average decrease of 3.55%. DF may superior to drip irrigation on increasing the content of available manganese in SX. The available zinc content in NX and SX exhibited a decreasing trend, with an annual average decline rate of 3.74% and 4.48% respectively. This phenomenon could potentially be attributed that the AP content in NX and SX had relatively increased (Fig. 3), while zinc has been fixed, leading to a decrease in its availability.

Contribution of soil physical and chemical properties to the integrated soil fertility from 1985 through 2020

The integrated soil fertility index (IFI) for different years was analyzed using a specific algorithm (Table 4). In NX, the IFI showed an upward trend over time. Conversely, in SX, the IFI displayed a downward trend after 2009.

Table 4 Differences in IFI in NX and SX in different years.
Full size table

The results of the Pearson correlation coefficient showed that in NX, the IFI had a significant positive correlation with SOM, TN, AP and AK, and a significant negative correlation with pH (Table 5, Fig. 4). The path analysis showed that in NX, the IFI had a direct and positive effects with SOM, TN, and AK. The SOM content in NX had the greatest impact on soil fertility, followed by TN. Conversely, the IFI had a significant positive correlation with SOM, TN, pH and AP, and a significant negative correlation with AK (Table 5, Fig. 4). The path analysis showed that in SX, the IFI had a direct and positive effect with TN and AP. The TN content in SX had the greatest impact on soil fertility, followed by AP.

Table 5 Correlation and path analysis between IFI and soil properties in NX and SX from 1985 to 2020.
Full size table
Fig. 4
figure 4

Correlation heat map of IFI in NX and SX and each nutrient index from 1985 to 2020.

Full size image

Analysis on changes of soil physical and chemical properties and IFI in 2022

The results of soil physical and chemical properties and IFI in 2022 were similar to the analysis of 1985–2020. The average values of SOM, AK content, pH and IFI of different soil depths (0–100 cm) in NX were higher than those of SX (Fig. 5). In NX, the average values of SOM, AK content and pH were 10.24%, 76.86% and 5.24% higher than those of SX, respectively. The content of AP in NX was 28.52% lower than SX. The average values of IFI in NX were 7.6% higher than those of SX. In general, the content of soil nutrients in NX was higher than that in SX. Moreover, the results of field-sampling-based IFI further verified the higher soil fertility in NX compared to SX.

Fig. 5
figure 5

Distribution characteristics of soil nutrients and IFI in NX and SX in 2022.

Full size image

Discussion

Drip irrigation led to lower soil pH and decreased compaction, resulting in looser soil structure in northern Xinjiang

A study from 1980 to 2020 showed that the drought trends in southern Xinjiang were more pronounced than in northern Xinjiang41, nearly half (49.6%) of the irrigated area in southern Xinjiang was salty farmland42. Therefore, flood irrigation was mostly used in southern Xinjiang in winter and spring to alleviate soil salinization. While in northern Xinjiang, the drought degree was mild, and with the support of government policies and subsidy programs15, drip irrigation technology was developed rapidly. Drip irrigation drips water directly into the soil near the root-zone of crops, forming an ellipsoid or spherical wet body in the root-zone of crops43. High-frequency drip irrigation can effectively moderate soil salinity by leaching salt from the root zone, which was partly conducive to decreasing the pH. In this study, the pH in NX slight decreased, from 8.52 in 1985 to 8.24 in 2020, and the soil bulk density slightly increased from 1.25 g∙cm−3 to 1.31 g∙cm−3, which were consistently lower than SX. Soil pH is a key factor determining nutrient availability and also influences microbial activity44. Soils rich in nutrients typically support higher microbial diversity and activity, which can improve soil health and fertility45. Li et al.46 found that after 12a, under long-term mulched drip irrigation, the soil salt content of cotton fields tended to stabilize, the average salt content in the root zone was less than 2.49 g∙kg−1, and the annual average desalination rate was 24.93% in root-zone of crops. Wang et al.47 reported that deep (40–120 cm) soil salinity decreased significantly, below 21–74% of the initial profile in drip irrigation under freshwater film (300–700 mm of the irrigation water volume). Wang et al.48 has shown that prolonged mulched drip irrigation application reduced salinity levels of soil, the desalination rate stabilized at 80–90% beyond eight years of mulched drip irrigation application. The large amount of water used in irrigation will cause excessive soil moisture, which in turn will lead to the increase of soil bulk density49. The results of the previous studies are similar to those of this study. But, Hu et al.50 reported under the condition of mulch drip irrigation, the non-growth winter irrigation salt was comprehensively used, the salt content in the root zone of the soil did not accumulate significantly. Under high frequency of drip irrigation, soil salt is basically in the state of leaching, avoiding the phenomenon of return salt and salt accumulation brought about by traditional irrigation, so, improved soil quality. Wang et al.51 found that with drip irrigation, soil salinity in the cultivation layer consistently decreased over the years, which improved soil quality, thereby facilitating proper crop emergence and growth. Therefore, DI improved soil structure by maintaining optimal moisture, which reduced compaction, the key reasons are: (1) Optimal water application. The optimal irrigation amount reduces the soil salt content and water-salt leaching, resulting in a decrease in pH. It is also the main reason why the soil bulk density in northern Xinjiang is lower than that in southern Xinjiang. (2) Changes in the law of water and salt migration. Under drip irrigation, groundwater level drops and the salt formed by diving evaporation decreases, thus reducing pH52. These findings are crucial for understanding the long-term effects of drip irrigation on soil quality and can provide valuable insights for sustainable agricultural practices in northern Xinjiang.

The content of soil macronutrients was relatively higher under drip irrigation in northern Xinjiang

The content of soil macronutrients is affected by factors such as surface litter input, mineralization process and so on53. Compared with the north of the Tianshan Mountains, the south of the Tianshan Mountains had lower surface vegetation coverage54. Under the influence of drip irrigation, only the soil areas in the root-zone of crops are wetted. The soil mineralization is strong under freeze–thaw cycle conditions55, which is conducive to increasing the soil nutrient content. The results from 1985 to 2020 showed that the SOM, TN, AP and AK content in NX increased, with an annual average growth rate of 7.39%, 7.09%, 4.71% and 1.12%, respectively. The results of field sampling in 2022 presented that the average values of SOM, AK content and pH in NX were 10.24%, 76.86% and 5.24% higher than those of SX, respectively, which validated our analysis of 1985 to 2020. The SOM content in NX had the greatest impact on soil fertility (Table 5). In the Central Anatolia region of Turkey, Demir56 found that among the 22 soil quality parameters, soil SOM had the highest weight of 0.431, and could be considered as the first restriction limiting the working potential of soils, which consistent with our research results. We observed a negative correlation between AK and IFI in SX (Table 5), which may be attributed to soil salinization and poor drainage under flood irrigation practices. Soil is a vastly intricate system where quality depends on combined chemical, physical, and biological processes57. Tan et al.58 reported that the irrigation level of 15 mm and fertilization level of 93.00 kg·hm-2 resulted in favorable soil physicochemical properties and bacterial community diversity, maintaining soil quality at an optimal level. Li et al.59 found that in extremely arid areas, the implementation of long-term drip irrigation and the addition of organic fertilizer reached the highest content seven years later, with a significant increase of 44.19%, TN, AP and AK content were significantly higher than conventional irrigation. Song et al.60 reported that in Northeast China, the total organic carbon content of controlled irrigation was higher than that of conventional flooded irrigation at same N fertilizer application amount. With the total organic carbon content in surface layers (0–20 cm) increasing by 0.22–1.35% and the total organic carbon in deep layers (40–60 cm) increasing by 0.19–1.31%. In this study, DI increased the content of soil macro-nutrients. The key reasons are as follows: (1) Groundwater lowering. Long-term drip irrigation reduces the groundwater level, which changes the structure of soil microflora, accelerates the decomposition of organic matter, promotes the formation of soil humus, and is conducive to the transformation of soil mineral elements, so the soil macronutrients keep increasing61. (2) Salt leaching rate increase. In long-term drip irrigation, water and salt shower decreases, soil pH decreases, the effectiveness of calcium in the root area plays a greater role, promoting soil macronutrients stability and increasing62.

However, the impact of irrigation patterns on soil micronutrients is a different story. Drip irrigation can effectively promote the increase of soil macronutrients in NX, but the effect on soil micronutrients is not obvious; it can even lead to a decline, such as in the case of available manganese and available zinc. DF had obvious effect on increasing the available iron content in SX, and the specific reasons need to be further studied. Future research could establish a long-term and dynamic soil quality monitoring system to continuously track changes in specific indicators (Soil macronutrients: total phosphorus available nitrogen et al.; Soil micronutrients: available molybdenum et al.) under different irrigation practices, thereby providing more comprehensive data support for precision agriculture and sustainable land utilization. In general, there was a significant effect of drip irrigation on promoting soil fertility, which can be confirmed by the IFI in NX, in this study.

Conclusions

Compared to SX, from 1985 to 2020, the SOM, TN, AP and AK content of NX had an increase trend. The application of zinc fertilizer should be emphasized in both NX and SX. In 2022, the results of field sampling validated our analysis from 1985 to 2020. The IFI confirmed that soil fertility in NX was higher than SX. With regard to sustaining and enhancing long-term soil fertility, northern Xinjiang was superior to southern Xinjiang in SOM, TN, AP, AK, pH and soil bulk density. However, this study has several limitations: (1) The comparison of soil quality was only conducted between two irrigation districts, which may lack representativeness; (2) The analysis relied on historical data and limited field sampling points, resulting in insufficient spatial coverage and sampling depth. Future studies should incorporate data from additional regions to provide a more comprehensive understanding of regional soil quality differences of Xinjiang. These findings suggest that expanding DI practices could be a viable strategy to enhance soil fertility in arid and semi-arid agricultural systems. In conclusion, NX has demonstrated superior soil fertility and nutrient retention compared to SX over the past thirty-five years under the influence of irrigation patterns, drip irrigation may represent a more sustainable long-term agricultural strategy than traditional flood irrigation in this region.