Research on the Fallow Compensation Mechanism for Groundwater Overexploitation in the Tarim River Basin Under Bidirectional Collaboration

Abstract

Exploring the differentiated fallow compensation (FC) standards in different regions is of great significance for formulating and improving the mechanism of fallow compensation and ensuring the sustainability of policies. The groundwater overexploitation area in the Tarim River Basin was selected as the research area; this study breaks through the perspective of a single subject and integrates the “opportunity cost” of the compensated subject and the “ecosystem service value” of the compensating subject into a unified analysis framework to obtain the fallow compensation standard, and the logistic model is used to analyze the influencing factors of farmers’ compensation method selection. The results are as follows: (1) The FC standards exhibit significant spatial heterogeneity. The range of FC standards in various counties is 5540.40 to 7770.53 CNY/hm² (769.50 to 1079.24 USD/hm²), which is generally lower than the current standard. (2) There are three main compensation methods chosen by farmers, ranked in descending order of selection ratio: monetary compensation (72.06%) > physical compensation (19.37%) > technical compensation (8.57%). (3) The factors influencing the choice of compensation method are quite complex. The dependency ratio is the main influencing factor in the choice of monetary compensation (β = 0.738); the evaluation of economic conditions has a significant negative correlation with the choice of physical compensation (β = −0.562), and nonfarm household income is the main influencing factor for choosing technical compensation (β = 0.747). This study provides a new perspective for determining FC standards and aims to provide a theoretical basis for local governments to improve their fallow policies.

1. Introduction

Cropland is an important cornerstone for ensuring food security [1]. It is estimated that the global population will exceed 9 billion by 2050 [2], and global food demand will also increase by nearly 70% [3]. However, cropland has been in a high-intensity and overloaded state for a long time in China [4], faced with many problems such as cropland pollution [5,6], soil erosion [7], desertification [8], groundwater overexploitation, and declines in biodiversity [9]. In this context, how to coordinate food security and cropland protection has become one of the main challenges facing the sustainable development of global agriculture today [10,11].

Currently, the international community has formed a multi-level practice system for protecting farmland, such as improving the quality of farmland through land leveling and optimizing irrigation and drainage systems, remediating degraded farmland via soil restoration technologies [12]. Other protection measures include formulating policies for the exploitation of wells to control groundwater overexploitation, ensuring the sustainability of aquifer conditions for long-term exploitation [13]. In addition, farmers are also guided to participate in protecting farmland through fallow compensation and the promotion of large-scale operations to improve resource utilization efficiency [14]. In arid and semi-arid regions, the above measures need to focus on the coordinated safety of “water soil food”, while fallow policy exhibits distinct advantages in ecologically sensitive areas such as those suffering from groundwater overexploitation through phased cessation of cultivation and ecological restoration measures. Its core value lies not only in restoring farmland fertility, but also in alleviating resource and environmental pressures and reserving ecological resilience for long-term sustainable development.

The main purpose of fallow is to promote land cultivation and regeneration, with the aim of restoring the fertility of damaged farmland, especially farmland with continuous ecological degradation, groundwater overexploitation, and heavy metal pollution [15]. Due to the significant achievements of fallow policy in improving soil quality and restoring ecological functions, extensive policy practices have been carried out both domestically and internationally. Long-term practices in countries such as the United States, Japan, and the European Union have shown that fallowing farmland with high ecological risks or severe pollution can effectively alleviate the degradation of water and soil resources caused by overloaded cultivation, and also contribute to the improvement of soil quality and the enhancement of farmland productivity [16,17,18,19,20]. In 2016, China’s “Thirteenth Five Year Plan Outline” clearly proposed to explore the system of farmland rotation and fallow based on regional characteristics, as well as promote pilot projects of farmland rotation and fallow, aiming to improve farmland quality, increase grain yield, control soil pollution, alleviate groundwater overexploitation, etc. The theoretical research on fallow in China is still in its infancy, mainly focusing on the summary and reference of foreign fallow experiences [21], the willingness and influencing factors of farmers to participate in fallow [22], and the scale and layout of fallow [23,24]. Although the current fallow policy has made certain contributions to achieving goals such as improving the physical and chemical properties of soil and reducing soil degradation risks [14], due to budget constraints and the necessity of long-term policy implementation, improving the long-term effectiveness of policy implementation has become a hot topic of concern in both the government and academia [25].

One of the key factors affecting the effectiveness of fallow policy is the formulation of fallow compensation (FC) standards. The formulation of reasonable FC standards should be based on natural, socio-economic, and other factors of different regions, and they should adhere to the principle of whoever benefits should be compensated, clarify the compensation subject and the compensated subject, implement differentiated compensation, explore multiple compensation models, and achieve coordinated development of social and economic development and farmland protection [26]. Therefore, scholars at home and abroad have conducted research on FC standards. At present, there are several methods for calculating FC standards, including the direct cost method, the opportunity cost method, the contingent valuation method, and the ecological compensation quantification method [27]. The direct cost method calculates compensation standards based on the losses suffered by participants or the construction costs invested by builders. The opportunity cost method uses the opportunity cost incurred by farmers due to fallow as the compensation standard [28]. For the contingent valuation method, also known as the willingness survey method, a survey questionnaire was used to determine the FC standards based on the willingness of farmers [29]. The quantification method of ecological compensation refers to directly measuring the value of fallow on the improvement of ecosystem service functions to indirectly reflect compensation standards [30]. Although the existing methods for calculating FC standards have their own focuses, there are still significant shortcomings in adaptability in complex scenarios of groundwater overexploitation in arid areas. Specifically, the direct cost method and opportunity cost method only consider the economic losses of farmers, but do not take into account the long-term ecological benefits of fallow. The contingent valuation method relies on the subjective willingness surveys of farmers, which may lead to distorted results due to information asymmetry or strategic biases. The quantification method of ecological compensation focuses on enhancing ecological value, but does not take into account the actual losses of farmers.

As the largest inland river basin in China, the Tarim River Basin (TRB) has experienced a significant increase in irrigation water usage due to the continuous increase in farmland area over the past 70 years. From 1949 to 2008 alone, the expansion of farmland area in the basin reached as high as 98.70 × 10⁴ hm². In 2020, the overloaded farmland area was 1.35 × 10⁴ hm². Agricultural irrigation water has always accounted for more than 90% of the total social and economic water use in the area. The Kashgar and Hotan regions within the watershed suffer from severe groundwater overexploitation, long-term continuous cropping leading to poor soil quality, intensified non-point source pollution in farmland, frequent disasters such as sandstorms, droughts, and floods, severe water scarcity, extremely fragile ecological environments, and land use intensity exceeding normal thresholds, posing a serious threat to farmland protection and regional ecological security [31,32,33]. In this context, this paper selects a typical area suffering from groundwater overexploitation in the Tarim River Basin as the study area. Firstly, a bidirectional collaborative framework for FC is constructed, breaking through the single subject perspective and incorporating the “opportunity cost” of the compensated subject and the “ecosystem service value” of the compensated subject into a unified analysis framework, achieving multidimensional consideration of FC standards. Finally, logistic regression function was used to analyze the influencing factors of FC methods for farmers. Our research aims to provide a theoretical basis and experience reference for food security and ecological security.

2. Materials and Methods

2.1. Study Area

The area suffering from groundwater overexploitation in the Tarim River Basin is located in the western part of the Tarim River Basin, mainly including seven counties in Kashgar and three counties in Hotan. The region has a typical temperate continental climate, with scarce rainfall of about 51.2 mm per year, an average annual temperature of 11.0 °C, and annual evaporation of up to 2123.7 mm [34]. Water resources rely on glacier meltwater for replenishment, forming a typical desert oasis ecosystem. Cotton, wheat, and other water-consuming crops are the main crops planted in the area, with agricultural water consumption accounting for over 90% of total water use. Agricultural expansion has led to a gradual decline in groundwater levels and the continuous expansion of overexploited areas, exacerbating the contradiction between water scarcity and agricultural water consumption. This has brought enormous pressure to the regional ecological environment [35], severely restricting the construction of regional farmland ecosystems and sustainable agricultural development [36]. The overview of the study area is shown in Figure 1.

2.2. Data Sources and Description

This study selects the overexploited groundwater area in the Tarim River Basin as the research sample, aiming to calculate the loss in opportunity cost caused by fallow from the perspective of farmers and explore the factors that affect the choice of FC methods. The data for this study comes from an in-depth field survey conducted by our team in June 2025 in the region, including 7 counties in the Kashgar region and 3 counties in the Hotan region. Given the differences in fallow area and local response to policies, a combination of random sampling and stratified random sampling was used to randomly select 2–3 townships in each county based on fallow land area and village concentration. Then, 1–4 villages were randomly selected from the selected townships for questionnaire surveys. The survey mainly focuses on the characteristics of farmers and family members, the input–output situation during the wheat-growing season, and the influencing factors on the choice of FC methods. The respondents in the research area are mainly middle-aged and elderly, with the highest proportion in the age group of 40–60 years old. Their highest education level is concentrated in junior high school, and the overall education level is relatively low. The highest proportion of family size is 3–4 people. A total of 300 questionnaires were distributed for this survey, with 275 valid questionnaires and an effective rate of 91.67%. To ensure the scientific validity of the research results, reliability and validity analyses on the questionnaire were conducted, among which, KMO = 0.783, indicating that the validity of the questionnaire is good. The Cronbach’s α coefficient is 0.825, indicating good reliability of the questionnaire data.

In addition, crop yield and planting area data are sourced from the Xinjiang Uygur Autonomous Region Statistical Yearbook 2024 and various county-level statistical bulletins in 2024 (https://tjj.xinjiang.gov.cn, accessed on 10 July 2025). The price data of agricultural products comes from the “Compilation of National Agricultural Product Cost Benefit Information” (https://www.stats.gov.cn, accessed on 10 July 2025). Administrative boundary data was acquired from Standard Map Service System (http://bzdt.ch.mnr.gov.cn, accessed on 23 June 2025) using the Chinese standard map [Approval No.: GS (2024)0650] without cartographic modifications.

2.3. Research Methods

2.3.1. FC Standard Calculation from the Perspective of Bidirectional Collaboration

When determining the FC standard, it is necessary to comprehensively consider factors such as the ecosystem service value and opportunity cost of fallow, and calculate an FC standard that is in line with the interests of all parties from a two-way coordination perspective. Due to the fact that the compensation standards calculated by the opportunity cost method are considered from the perspective of the compensated party, and the compensation standards calculated by the ecological compensation quantification method are considered from the perspective of the compensated party, this paper sets weight coefficients for the compensation standards of different parties and calculates the FC standards under bidirectional coordination [31]. The formula is as follows:

$$FC=mQC+nOC$$

(1)

In the formula: FC is the fallow compensation standard; QC refers to the fallow compensation standard, calculated based on the quantification method of ecological compensation; OC refers to the fallow compensation standard, calculated based on the opportunity cost method; m refers to the weight coefficient of the QC; and n refers to the weight coefficient of the OC.

2.3.2. Quantitative Method of Ecological Compensation

The ecological compensation quantification method is based on the organic coordination of multiple indicators such as ecological service value, ecological overload index, farmland ecological compensation correction coefficient, and farmland pressure index to calculate the ecosystem service value obtained by the compensation subject.

• Calculate the total ecological service value of farmland

Drawing on the research of Xie et al., the equivalent factor method was used to calculate the ecological service value of farmland in the region [37,38]. The formula is as follows:

$$A=5.91\times E_a\times S$$

(2)

In this formula, A represents the total ecological service value of farmland within the region, in billions of CNY; E_a is the unit equivalent factor value, CNY/hm²; and S is the total farmland area within the region, hm².

• Calculate the ecological overload index of farmland

The ecological overload index of farmland can be used to quantify the degree of imbalance between the ecological footprint and ecological carrying capacity of farmland and then calculate the amount of ecological compensation for farmland [39,40,41,42]. The calculation formula is as follows:

$$I={\displaystyle \frac{EC-EF}{EC}}$$

(3)

In this formula: I refers to the ecological overload index of farmland; EC refers to the ecological carrying capacity of farmland, hm²; and EF refers to the total ecological footprint of farmland, hm².

• Calculate the amount of ecological compensation for farmland

The correction coefficient for ecological compensation of farmland aims to solve the problem of policy failure due to excessively high theoretical compensation values or ecological degradation of farmland due to excessively low values. The specific formula is as follows:

$$R={\displaystyle \frac{1}{1+e^{-E_n}}}$$

(4)

$$E_n{=E}_a\times \theta +E_b\times \left(1-\theta \right)$$

(5)

In the formula, R represents the correction coefficient for ecological compensation of farmland; E_n, E_a, and E_b respectively represent the comprehensive Engel coefficient, urban Engel coefficient, and rural Engel coefficient; and θ represents the urbanization rate.

The amount of ecological compensation for farmland is a comprehensive evaluation of the ecological benefits of farmland and the actual situation of the compensation area. It is calculated through a multi-index model to ensure the rationality and feasibility of the compensation standard [43]. The specific formula is as follows:

$$A_{ec}=A\times I\times R$$

(6)

In this formula, A_ec represents the ecological compensation standard for farmland, in billions of CNY; A represents the total ecological service value of farmland, in billions of CNY; I represents the ecological overload coefficient of farmland; and R represents the correction coefficient for ecological compensation of farmland.

• Maximum scale calculation of fallow based on modified farmland pressure index

The farmland pressure index is a variable that reflects the relationship between the minimum per capita farmland area required to ensure food security in a certain region and the actual per capita farmland area; it reflects the degree of farmland resource tension in the study area [44]. Due to the heterogeneity of farmland quality in different regions, this paper draws on previous research results, and the standard coefficient of farmland productivity, $\phi$, was used to correct the farmland pressure index [45]. The formula is as follows:

$$K_s={\displaystyle \frac{\beta \times \left[G_a/\left(p\times q\times l\right)\right]}{S_a}}\times \phi$$

(7)

$$\phi ={\displaystyle \frac{p\times q}{p_n\times q_n}}$$

(8)

In the formula, β is the self-sufficiency rate of grain, in %; G_a is the per capita food demand, kg/person; p is the yield per unit of grain in the region, kg/hm²; q is the ratio of grain sowing area to total sowing area in the region, in %; l is the index of crop replanting, in %; and S_a is the actual per capita farmland area, hm²/person. p_n is the national yield per unit of grain, kg/hm² and q_n is the ratio of the national grain sowing area to the total sowing area, in %.

On the basis of correcting the farmland pressure index, the maximum fallow land scale that ensures the minimum required farmland area is calculated using the deviation degree between the farmland pressure index K and 1 [46]. The formula is as follows:

$$S_f=S\times \left(1-K\right)$$

(9)

In the formula, S_f represents the maximum area of farmland that can be fallow in the region, hm²; S is the total area of farmland in the region, hm²; and K is the pressure index of farmland.

• Fallow compensation standard

Due to the fact that the farmland in the research area is mainly irrigated land, the winter wheat land that previously relied mainly on groundwater extraction for irrigation will be fallow during ecological restoration. The calculation formula for the expected fallow compensation in the region is as follows:

$$A_{ec\_f}=A_{ec}\times {\displaystyle \frac{S_f}{S}}$$

(10)

In this formula, A_{ec_f} represents the fallow compensation standard, in CNY; A_ec represents the ecological compensation standard for farmland, in CNY; S_f represents the area of fallow within the region, hm²; and S represents the total farmland area, hm².

According to the quantification method of ecological compensation, the fallow compensation amount is evenly distributed among the farmland that implements fallow policy, and this is used as the fallow compensation standard. The formula is as follows:

$$A_{ec\_f\_a}={\displaystyle \frac{A_{ec\_f}}{S_f}}$$

(11)

In this formula, A_{ec_f_a} represents the average fallow compensation standard in the area, in CNY/hm².

2.3.3. Opportunity Cost Method

The opportunity cost method essentially uses the losses caused by winter fallow by farmers as compensation standards [28], and this method has been widely applied in determining ecological compensation standards [25]. In this paper, the opportunity cost method was used to represent the input level of regional land from the perspective of the compensated subject, based on the input characteristics of farmers’ land in the region. Six element variables, including irrigation, seeds, fertilizers, machinery, pesticides, and labor, were selected to characterize the input level of regional land. The calculation method for net income from crops is as follows:

$$M=P\times Y-X_1-X_2-X_3-X_4-X_5-X_6$$

(12)

In this formula, P is the selling price of wheat; Y is the wheat yield; X₁, X₂, X₃, X₄, X₅, and X₆ respectively represent the input costs of irrigation, seeds, fertilizers, machinery, pesticides, and labor.

The labor input (X₆) is estimated using the shadow wage method and calculated as follows:

$$X_6=W\times p=\alpha \times {\displaystyle \frac{I}{T}}\times p$$

(13)

Among them, W is the shadow of labor force (CNY/day); P represents the labor input per unit area (days/hm²); I is the total income of winter wheat (CNY); T is the total labor input (in days); and α is the elasticity coefficient of labor input.

The elasticity coefficient of labor input is calculated using the Cobb–Douglas production function [47]. The calculation is as follows:

$$\ln I=a+b\ln T+c\ln S+d\ln C+\sum \left(\gamma +\delta \right)+ϵ$$

(14)

where a represents a constant term, T; S and C represent the labor, land, and capital inputs for wheat, respectively; b, c, and d are the elastic coefficients of T, S, and C, respectively; γ, δ is the external factor and its corresponding coefficient; and ε is the random error term.

2.3.4. Logistic Model

There are three types of compensation preferences for the compensated party: monetary compensation, physical compensation, and technical compensation. The logistic model was used to analyze the influencing factors of farmers’ choice of compensation methods for fallow [48]. The specific formula is as follows:

$$\mathit{log}\left({\displaystyle \frac{P_i}{1-P_i}}\right)=a_0+a_1{x}_1+a_2{x}_2+\dots +a_n{x}_n+ϵ$$

(15)

In this formula, P_i represents the probability of selecting a certain compensation method for fallow; a₀ is a fixed intercept; x₁, x₂, …, x_n and a₁, a₂, ..., a_n respectively represent explanatory variables and coefficients; and ε is the random error term. In this study, a total of 11 explanatory variables were selected from five aspects: natural capital, material capital, human capital, social capital, and financial capital (

Table 1. Definition and description of the independent variables.

Criterion Layer	Variable Name	Description/Definition
Natural capital (N)	Per capita area of farmland (N1)	Ratio of total farmland area to total number of family members (hm2/person)
	Farmland quality (N2)	Proportion of high-quality farmland in households to total farmland area (%)
Human capital (H)	Education level (H1)	Illiterate = 0; Primary school = 0.25; junior middle school = 0.5; high school = 0.75; college or higher = 1
	Dependency ratio (H2)	Proportion of household labor force to total household population (%)
	Health status of labor force (H3)	Excellent = 1; Good = 0.75; Medium = 0.5; Poor = 0.25
Material capital (M)	Household asset ownership status (M1)	Truck = 1; sedan = 0.8; tractor = 0.6; agricultural tricycle = 0.4; motorcycle = 0.2
	Economic condition evaluation (M2)	Excellent = 1; Good = 0.75; Medium = 0.5; Poor = 0.25
Social capital(S)	Family members and neighborhood relationships (S1)	Excellent = 1; Good = 0.75; Medium = 0.5; Poor = 0.25
	Family members’ participation in training (S2)	Number of training sessions per year
Financial capital (F)	Nonfarm household income (F1)	Household nonfarm total income (CNY)
	Agricultural household income (F2)	Annual total income from family agriculture (CNY)

).

3. Results

3.1. Calculation of FC Standards

3.1.1. FC Standards Using the Ecological Compensation Quantitative Method

The total farmland service value of the Tarim River Basin (TRB) in 2023 is CNY 11.735 × 10⁹ (USD 1.630 × 10⁹), of which the farmland service value in the Hotan region (HR) is CNY 1.664 × 10⁹ (USD 0.231 × 10⁹), and in the Kashgar region (KR) is CNY 10.071 × 10⁹ (USD 1.399 × 10⁹), accounting for 14.18% and 85.82% of the total research area, respectively. The average farmland service value in HR is CNY 5.55 × 10⁸ (USD 0.771 × 10⁸), while in KR it is CNY 1.439 × 10⁹ (USD 0.200 × 10⁹). There is a significant difference in farmland service value between the two regions, with KR being significantly higher than HR.

The farmland overload index of the watershed is all greater than 0, with the farmland overload index concentrated between 0.57 and 0.81, indicating that the farmland ecology is in a surplus state, and the regional ecological pressure is low, showing a trend of higher ecological pressure in KR than in HR. Among them, the high values of farmland overload index are concentrated in Jiashi County, Bachu County, and Maigaiti County, all of which are higher than 0.8. The low values are mainly concentrated in Hotan County and Moyu County. Within the watershed, there is a consistent spatial distribution between the ecological footprint of farmland and the ecological carrying capacity of farmland. Generally, areas with larger ecological footprints of farmland also have higher ecological carrying capacities.

Based on the correction coefficient of farmland ecological compensation, the final calculated farmland compensation in the watershed is CNY 5.33 × 10⁹ (USD 0.740 × 10⁹). The farmland compensation amounts in HR and KR are CNY 0.624 × 10⁹ (USD 0.867 × 10⁸) and CNY 4.706 × 10⁹ (USD 0.654 × 10⁹), respectively (

Table 2. Indicators related to farmland compensation in TRB groundwater overexploitation in 2023.

	The Farmland Service Value (108 CNY/108 USD)	Farmland Overload Index (I)	Farmland Pressure Index (K)	Fallow Compensation (108 CNY/108 USD)
Hotan County	3.75/0.528	0.59	0.57	1.50/0.208
Luopu County	4.72/0.656	0.63	0.53	1.87/0.260
Moyu County	8.17/1.135	0.57	0.54	2.87/0.399
Shache County	23.00/3.194	0.72	0.48	10.11/1.404
Jiashi County	22.16/3.078	0.81	0.42	10.94/1.519
Bachu County	15.96/2.217	0.80	0.41	7.77/1.079
Zepu County	6.51/0.904	0.75	0.50	2.95/0.410
Shufu County	8.53/1.185	0.77	0.54	4.00/0.556
Yecheng County	13.52/1.878	0.72	0.55	5.91/0.821
Maigaiti County	11.03/1.532	0.80	0.47	5.38/0.747

). Based on the revised farmland pressure index, the maximum scale of fallow in the TRB was calculated. Under the premise of ensuring food security, the proportion of fallow in each county is concentrated between 43.17% and 59.08%. From a county-level perspective, the total fallow compensation is ranked from high to low as follows: Jiashi County, Shache County, Bachu County, Yecheng County, Maigaiti County, Shufu County, Zepu County, Moyu County, Luopu County, and Hotan County. There is a significant difference in fallow compensation among different counties.

3.1.2. FC Standards Using the Opportunity Cost Method

Perform least squares regression on 275 questionnaires using the Cobb–Douglas production function. The dependent variable is the total agricultural income, and the explanatory variables are the age of the farmer/decision maker, the education level of the decision maker, farmland area, quality of farmland, the number of days of labor input, and capital input (

Table 3. Estimation of the Cobb–Douglas production function in TRB.

Influencing Factors	Coefficient	Standard Error	Z Statistic	p Value
Age	0.0006	0.0232	0.4924	0.6343
Education level	−0.0443 **	0.0216	−2.2531	0.0424
Farmland area	0.7752	0.0756	10.3171	0.0000
Quality of farmland	0.5934 **	0.1183	4.3762	0.0000
Labor input	0.0321	0.0354	0.9205	0.4075
Capital input	0.2054	0.0621	4.3251	0.0502
R2 = 0.8237
Prob(F-statistic) = 0

Note: ** indicates statistical tests were at the 5% significant levels.

). The model R² = 0.8237 indicates a good fit. Among them, the elasticity coefficient of labor input is 0.0321, and based on this, the average shadow wage in TRB is 41.65 CNY/day (5.78 USD/day).

The average yield of wheat in the research area is 5441.14 kg/hm². Among the research area, Jiashi County has the highest yield, reaching 6265.66 kg/hm², while Moyu County has the lowest yield of only 4858.44 kg/hm². There is a significant positive correlation between the level of capital investment and output. The total capital investment in Jiashi County ranks first, followed by Shufu County; Zepu County, Shache County, and Maigaiti County, which have similar investment scales. There are significant differences in the structure of capital investment among different counties (Figure 2). The water conservancy infrastructure in the irrigation area of Jiashi County is weak, with an irrigation water utilization coefficient of only 0.49, which is significantly lower than the average value for southern Xinjiang (0.53). The water resource utilization efficiency is low, and the proportion of irrigation investment in this county reaches 13.67%, far higher than other counties in the TRB. Bachu County has the highest proportion of fertilizer input (35.27%) and absolute input in the TRB, and its fertilizer input intensity is strongly positively correlated with yield. The county’s yield is among the top in the TRB. The comprehensive mechanization rate of major crops in Hotan County is 77.42%, which is lower than the average level of the basin, resulting in the lowest mechanical input in the TRB. Correspondingly, the labor input has risen to the highest level. The differences in seed and pesticide inputs among the 10 counties are relatively small. Overall, the highest proportion of capital investment in each county in the TRB is fertilizer investment, followed by machinery investment. In terms of net income, the average net income in the study area is 7237.94 CNY/hm² (1005.27 USD/hm²), with the highest value in Jiashi County and the lowest in Moyu County, at 7700.40 CNY/hm² (1069.50 USD/hm²) and 5515.80 CNY/hm² (766.08 USD/hm²), respectively (Figure 3). The results indicate that the efficiency of fund allocation and structural optimization play a decisive role in improving agricultural production efficiency.

3.1.3. FC Standards Under Bidirectional Collaboration

Based on the quantification method of ecological compensation and the opportunity cost method, the weight coefficients of compensation standards calculated from different perspectives were set to 0.5, and the final FC standards in the TRB were obtained (

Table 4. FC standards in TRB in 2023.

County	FC Standards Using Ecological Compensation Quantitative Method (CNY/USD/hm2)	FC Standards Using Opportunity Cost Method (CNY/USD/hm2)	FC Standards Under Bidirectional Collaboration (CNY/USD/hm2)
Hotan County	5141.85/714.15	6048.00/840.00	5594.93/777.07
Luopu County	6086.10/845.29	6382.65/886.48	6234.38/865.89
Moyu County	5565.60/773.00	5515.80/766.08	5540.40/769.54
Shache County	6829.80/948.58	6526.35/906.44	6678.08/927.51
Jiashi County	7840.65/1088.98	7700.40/1069.50	7770.53/1079.24
Bachu County	6653.10/924.04	7085.25/984.06	6869.18/954.05
Zepu County	6963.30/967.13	6695.25/929.90	6829.28/948.51
Shufu County	7867.95/1092.77	7005.75/973.02	7436.85/1032.90
Yecheng County	7812.45/1085.06	6313.80/876.92	7063.13/980.99
Maigaiti County	6891.45/957.15	6259.80/869.42	6575.63/913.28

).

From a county-level perspective, the FC is ranked from high to low as follows: Jiashi County, Shufu County, Yecheng County, Zepu County, Bachu County, Shache County, Maigaiti County, Luopu County, Hotan County, and Moyu County. There is a significant difference in the amount of FC in each county, with compensation standards ranging from 5540.40 to 7770.53 CNY/hm² (769.50 to 1079.24 USD/hm²).

3.2. Factors Influencing the Selection of FC Methods

At present, the only compensation method for the overexploitation of groundwater in the TRB is monetary compensation. During the research process, it was found that in addition to monetary compensation, physical compensation and technical compensation are also acceptable compensation methods for farmers. The FC methods chosen by farmers mainly focus on monetary compensation, physical compensation, and technical compensation. Monetary compensation is the main compensation method chosen by farmers, accounting for 72.06%, followed by physical compensation, accounting for 19.37%, and technical compensation has the lowest proportion, at only 8.57%.

The influencing factors of farmers’ choice of FC methods through the logistic model are shown in

Table 5. Binary logistic regression analysis results.

Indicators	Monetary Compensation		Physical Compensation		Technical Compensation
	Coefficient	Standard Error	Coefficient	Standard Error	Coefficient	Standard Error
Per capita area of farmland	0.486 ***	0.149	0.365 ***	0.112	0.139	0.120
Farmland quality	0.124	0.093	0.113	0.372	0.086	0.086
Education level	0.151 *	0.107	−0.087	0.009	0.649 ***	0.189
Dependency ratio	0.738 ***	0.226	−0.153	0.101	0.172	0.134
Health status of labor force	−0.062	0.085	−0.203 **	0.089	0.034	0.075
Household asset ownership status	0.238	0.201	−0.473	0.312	−0.156	0.190
Economic condition evaluation	−0.297	0.158	−0.562 **	0.249	0.622	0.332
Family members and neighborhood relationships	1.337	0.841	1.238	0.863	0.961	0.712
Family members’ participation in training	0.131	0.092	0.093	0.081	0.596 **	0.174
Nonfarm household income	0.352 **	0.156	1.032	0.724	0.747 ***	0.256
Agricultural household income	0.336 *	0.182	0.489 ***	0.172	0.097	0.118
Nagelkerke R2	0.693		0.675		0.638
Sig	0.038		0.043		0.047

Note: ***, **, and * indicate statistical tests were at the 1%, 5%, and 10% significant levels.

. The results indicate that the preference for monetary compensation is significantly influenced by natural capital, human capital, and financial capital. The dependency ratio (β = 0.738, p < 0.01) showed a highly significant positive effect, and the per capita farmland area (β = 0.486, p < 0.01) was significantly positively correlated with nonfarm household income (β = 0.352, p < 0.05), indicating that households with abundant farmland resources or diversified income rely more on cash compensation to alleviate economic losses from fallow.

The choice of physical compensation is constrained by factors related to natural capital and human capital. Agricultural income (β = 0.489, p < 0.01) and per capita cropland area (β = 0.365, p < 0.01) are significantly positively driven, reflecting the urgent need for agricultural dominant households with large farmland to supplement production materials. The two indicators of economic condition evaluation (β = −0.562, p < 0.05) and the health status of labor force (β = −0.203, p < 0.05) have a significant negative effect on the selection of physical compensation. The worse the economic condition evaluation and the health status of labor force, the more likely farmers are to solve their survival problems, which will increase their demand for food to alleviate their concerns about survival.

The demand for technical compensation is highly correlated with human capital, social capital, and financial capital. Nonfarm household income (β = 0.747, p < 0.01), labor education level (β = 0.649, p < 0.01), and training participation (β = 0.596, p < 0.05) have a significant positive effect, highlighting the tendency of highly skilled farmers to achieve sustainable development through technological compensation.

4. Discussion

4.1. The Novelty of the Comprehensive Model in Determining FC Standards

The determination of FC standards is crucial for the implementation of compensation policies [49]. However, most existing research only considers single factors such as ecosystem service value or opportunity cost [50,51], or only considers a single entity, without comprehensively considering the input–output, ecosystem service supply, and consumption of different entities. In view of this, this paper has developed a more comprehensive method to determine the FC standards. Starting from the perspectives of the compensated subject and the compensating subject, the opportunity cost method is used to calculate the loss of fallow by the compensated subject, and the ecological compensation quantification method is used to calculate the ecological service value obtained by the compensating subject. This not only meets the requirements of the compensating subject to improve the ecological quality of farmland, but also satisfies the loss of income from fallow by the compensated subject. Taking into account the wishes of both parties, the FC standards under the synergistic effects are calculated to promote the sustainable implementation of fallow policies. Zhang et al. (2023) established an ecological compensation standard accounting model under different levels of fertilizer reduction based on WTA, and determined the compensation standard [52], while Zhang et al. (2017) estimated the ecological compensation standards for non-point source nitrogen pollution in farmland using WTA and opportunity cost methods [53]. In summary, it is reasonable to determine compensation standards based on multiple levels and factors. This paper provides a new perspective for determining the FC standards, which can provide scientific guidance for developing compensation plans in other fallow areas.

4.2. FC Standard and the Selection of Compensation Methods

Comparing the calculation results with the current fallow compensation for groundwater overexploitation in the TRB, it was found that the average compensation calculated in this study for the basin is 6659.24 CNY/hm² (924.89 USD/hm²), which is lower than the standard of 7500.00 CNY/hm² (1041.67 USD/hm²) that was determined in the “Implementation Plan for Crop Rotation and fallow Projects in 2023” issued by the Agriculture and Rural Department of Xinjiang Uygur Autonomous Region. In addition, the reference value determined by Wang et al. in Cang County, Hebei Province, is 5250.00 CNY/hm² (729.17 USD/hm²) [54], while the reference value obtained by Ti et al. in the survey of farmers in the North China Plain is 8781.00 CNY/hm² (1219.58 USD/hm²) [25]. The reason for this may be that the current standard adopts a method of single economic compensation, which includes implicit social stability costs such as insurance premiums to prevent land reclamation, while this paper focuses on explicit economic costs. The functional positioning and emphasis of the two are slightly different. The differences in the reference values calculated by different scholars may be due to the different conditions of the research area itself and the different methods used. The FC standards of various counties in the TRB are concentrated in the range of 5540.40 to 7770.53 CNY/hm² (769.50 to 1079.24 USD/hm²). The reason for the different FC standards in each county is due to the varying proportions of land capital investment, resulting in different opportunity costs [25]. As shown in Figure 2, fertilizer input is the main capital input for wheat and has a strong impact on net income. The main reason is that the soil fertility in the watershed is poor. Reasonable application of chemical fertilizers can increase yield, but the vicious problems of soil compaction and organic matter decline caused by excessive fertilization should be noted. Capital investment in farmland, wheat prices, regional ecological overload index, etc., all have an impact on FC standards. Therefore, FC standards should not be a one-size-fits-all approach, but should be tailored to local conditions to establish compensation standards. In addition, the loss of income from fallow is less than the current compensation standard, and the current standard only compensates based on the fallow area. Although this model is simpler and more efficient to implement, the widespread implementation of fallow policies with this compensation standard will undoubtedly bring a huge burden to government finances. Therefore, different FC standards should be set according to the actual situation in various regions to achieve the purpose of fallow more efficiently and promote the sustainable development of farmland.

The analysis of the selection of compensation methods for farmers shows that monetary compensation is the main intended compensation method, accounting for 72. 06%, followed by physical compensation (19.37%), and the lowest technical compensation (8.57%). This is consistent with the research results of Zhang et al. (2023) and Chen et al. (2018), where monetary compensation is the most acceptable form of compensation [48,52]. Among them, income level greatly affects the choice of compensation method selected by the compensated subject. The results of Yu et al. indicate that farmers with a higher proportion of agricultural income have a lower preference for technological compensation, while those with a higher proportion of nonfarm income have less dependence on land and are more inclined to obtain technological compensation in order to better achieve nonfarm income growth. This is roughly consistent with the results of this paper [55].

4.3. Policy Suggestions for Groundwater Overexploitation in the TRB

China has achieved certain results in the farmland protection compensation, but the legal system is still not sound. It is urgent to establish a complete set of farmland protection and compensation practices for the legal system to ensure that the compensation work for farmland protection has legal basis. In addition, building a sound financial security system is key to ensuring the long-term effective operation of the FC mechanism [56]. This can be achieved by establishing an investment and financing system for FC funds, starting from the perspective of bidirectional collaboration, and gradually establishing a government led, market-regulated, and socially involved investment and financing system for FC funds. Reducing the financial burden on the government can help promote long-term stable implementation of policies. The current fallow areas are designated according to administrative regions based on tasks issued by higher-level governments, rather than calculated through scientific monitoring, and their scientificity needs to be improved. Soil quality testing and groundwater level monitoring play an important role in scientifically delineating fallow areas and implementing fallow policies more accurately to alleviate the phenomenon of groundwater overexploitation in the basin [57,58].

Due to differences in soil quality, irrigation, and management methods among farmland, there are variations in their returns. A unified FC standard can easily cause dissatisfaction among some farmers whose farmland income exceeds the compensation standard, which is not conducive to the continuous implementation of policies. In addition, the current compensation standards are generally higher than the losses caused by fallow, and the widespread implementation of fallow policies will increase the government’s financial burden. It is suggested that local governments determine differentiated FC standards based on their resource endowments, which can more accurately implement policies and reduce government financial burdens. Secondly, the compensation for groundwater overexploitation in the TRB is currently issued as a lump sum, based on the fallow area. A tiered compensation scheme can be implemented to incentivize the active participation of the beneficiaries. It is vital to implement differentiated compensation standards to ensure the basic survival needs of beneficiaries and set performance-driven compensation incentives to encourage long-term fallow behavior, reducing the phenomenon of “returning to cultivation”. Finally, it is necessary to strengthen the implementation of policies such as vocational skills training, as most existing policies mainly rely on single monetary compensation. The results of this study indicate that 8.57% of farmers are more inclined towards technical compensation. Monetary compensation is not the only goal for farmers to participate in farmland protection [59]. Therefore, we can learn from the practical experience of FC in foreign countries, implement diversified compensation methods, and organically combine monetary compensation, physical compensation, technical compensation, etc. [60], increase diversified survival skills training for farmers, popularize environmental knowledge, develop green industries, reduce dependence on land, promote the transfer of rural labor to nonfarm industries, increase farmers’ income, and promote rural revitalization and sustainable agricultural development.

4.4. Limitations and Prospects

Due to the fact that the calculation of farmers’ losses in opportunity cost caused by fallow is based on questionnaire survey data, although we have verified the confidence level of this value and the results are statistically significant, the error is still difficult to completely eliminate [61]. The research results indicate that there are significant differences in FC standards in different counties, and affecting factors may include soil quality, land capital investment, etc. However, the main affecting factors have not been further analyzed. Secondly, the compensation standards obtained in this study are based on the county level and have not been further divided into compensation standards for fallow land of different levels. Finally, the current compensation standard calculation is based on “historical data”, but fluctuations in grain prices, extreme weather events, and policy adjustments may significantly affect opportunity costs. Therefore, in future research, in-depth analysis will be conducted on the factors that affect the FC standards, in order to identify the main influencing factors; for this study, the quality level of cultivated land was used as the classification standard to calculate the FC standards for different levels, and thus in future research we should shift from “static calculation” to “scenario prediction simulation”, and propose an FC standard elastic adjustment threshold to provide reliable theoretical support for further refining the implementation of fallow policies.

5. Conclusions

This study takes the area suffering from groundwater overexploitation in the Tarim River Basin (TRB) as the research area, constructs a bidirectional collaborative framework for FC, uses the ecological compensation quantification method and the opportunity cost method to determine the FC standard, and analyzes the influencing factors of the compensation method selection for farmers. The main conclusion is as follows: the FC standards exhibit significant spatial heterogeneity. Due to differences in resource endowment, economic development level, and infrastructure, the range of FC standards in each county in the TRB is 5540.70–7770.60 CNY/hm² (769.50 to 1079.24 USD/hm²), which is generally lower than the current standard of 7500 CNY/hm² (1041.67 USD/hm²). This indicates that the current standard adopts a single economic compensation, including insurance premiums to prevent reclamation. There are three main FC methods chosen by farmers, ranked in descending order of selection ratio: monetary compensation (72.06%) > physical compensation (19.37%) > technical compensation (8.57%). The dependency ratio is the main influencing factor in the choice of monetary compensation (β = 0.738), and the higher the dependency ratio, the greater the pressure on family life, and the higher the demand for funds, and monetary compensation can be directly used for living expenses. The evaluation of economic conditions has a significant negative correlation with the choice of physical compensation (β = −0.562), meaning that the worse the economic situation, the higher the demand for food by the compensated individuals to alleviate their concerns about survival issues. Nonfarm household income is the main influencing factor for choosing technical compensation (β = 0.747), and the higher the proportion of household nonfarm income, the less dependent farmers are on land and the more inclined they are to obtain technical compensation in order to better achieve nonfarm income growth.