Next Article in Journal
Optimal Return Policies and Micro-Plastics Prevention Based on Environmental Quality Improvement Efforts and Consumer Environmental Awareness
Next Article in Special Issue
Sediment and Nutrient Retention in Ponds on an Agricultural Stream: Evaluating Effectiveness for Diffuse Pollution Mitigation
Previous Article in Journal
Status of Endemic Freshwater Fish Fauna Inhabiting Major Lakes of Turkey under the Threats of Climate Change and Anthropogenic Disturbances: A Review
Previous Article in Special Issue
The Role of Ponds in Pesticide Dissipation at the Agricultural Catchment Scale: A Critical Review
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Eco-Compensation Schemes for Controlling Agricultural Non-Point Source Pollution in Maoli Lake Watershed

1
School of Ecology and Nature Conservation, Beijing Forestry University, Beijing 100083, China
2
Business School, Central South University of Forestry and Technology, Changsha 410004, China
3
Academy of Engineering Innovation, China Construction Fifth Engineering Division Corporation Limited, Changsha 410004, China
*
Author to whom correspondence should be addressed.
Water 2021, 13(11), 1536; https://doi.org/10.3390/w13111536
Submission received: 22 April 2021 / Revised: 23 May 2021 / Accepted: 24 May 2021 / Published: 30 May 2021

Abstract

:
Maoli Lake is the water source for local residents and a national nature protected area. However, due to intensive agriculture development, the water quality has deteriorated over the past decades. An effective measure to improve water quality is to control the agricultural non-point source (NPS) pollution through elaborate schemes based on eco-compensation. In order to develop such eco-compensation schemes, three scenarios of agricultural activity adjustment were designed: S1 (halving fertilization every year), S2 (fallow every other year), and S3 (returning agricultural land to forest). A Soil and Water Assessment Tool (SWAT) model was adopted to simulate runoff, total nitrogen, and total phosphorus. Based on SWAT results, a multi-criteria spatial evaluation model considering the environmental, economic, and social effects of eco-compensation was created for best scenario decision. The results reveal the following: (1) the total nutrients loss of agricultural land reduces in all scenarios, but S2 has more reduction compared to S1 and S3; (2) from the comprehensive perspective of environment–economy–society effects, S2 is the best scenario for rice land and dry land; (3) the comprehensive effect of eco-compensation at the grid scale has a significant spatial difference, and therefore, we highlight the necessity and significance of controlling agricultural NPS pollution by eco-compensation on a precise spatial scale. This study can broaden the application field of the SWAT model and provide a scientific basis and experience for the evaluation and spatial design of agriculture eco-compensation.

1. Introduction

According to data published by the United States Environmental Protection Agency in 2000, approximately 30–50% of surface water in the world was affected by non-point source (NPS) pollution at the end of the last century. The NPS pollution caused by agricultural activities has gradually become the main cause of regional water environment deterioration [1,2]. The composition and the mechanism of agricultural NPS pollution are complex; its types are diverse, and its migration path is not clear. Moreover, water pollution has many influence factors, such as space, time, climate, soil topography, hydrology, and management measures [3,4,5]. At present, the comprehensive and quantitative description on the complex NPS pollution process in the whole watershed has been being considered as one of the best ways to quantitatively estimate NPS pollution, analyze the influence factors of pollution, and evaluate the effects of pollution control measures [6,7,8].
The NPS pollution models [9] mainly include two categories: empirical model and mechanism model. The representative empirical models include the soil conservation service (SCS) model [10], universal soil loss equation (USLE) model [11], and output coefficient model [12]. These models mainly use pollutant output coefficients to estimate the NPS pollution load output and are mostly used to evaluate the relationship between land use and lake eutrophication [13]. The mechanism model considers the internal mechanism of NPS pollution occurrence, migration, transformation, and environmental impact [14]. The CREAMS (chemicals, runoff, and erosion from agricultural management systems) model [15], the ANSWERS (areal non-point source watershed environment response simulation) model [16], the HSPF (hydrological simulation Program–Fortran) model [17], and the SWAT (Soil and Water Assessment Tool) model [18] have been widely used. Among these, the SWAT model [19], developed by the Agricultural Research Institute of the United States Department of Agriculture in the early 1990s, can simulate the effects of climate change, land use change, and agricultural management measures on the water cycle, sediment, nutrients, and crops. As a typical distributed hydrological model, it is widely used for long-term prediction and to simulate NPS pollution at the watershed scale [20,21,22].
Implementing best management practices (BMPs) at the watershed scale is one of the most effective measures for controlling NPS pollution [23,24]. The eco-compensation approach [25] in BMPs has a wide range of utility and can appropriately address problems of dispersion, randomness, scale, hysteresis, and difficulty of monitoring NPS pollution [26], which make it an efficient measure for control agricultural NPS pollution [27]. “Eco-compensation” is an institutional arrangement that focuses on the protection and sustainable utilization of ecosystem services and mainly adjusts the interest relationship of stakeholders through economic means [28], and in this paper, it refers to a series of approaches composed of “adjustment measures on agricultural land use” and “economic compensation to stakeholders”. Since the second half of the last century, eco-compensation, also known as the payment for environment and ecosystem services, has gradually become an important policy tool to deal with the contradiction between the development of human society and the resource environment [29,30]. Agricultural land is a semi-natural ecosystem and a major subsystem in the agricultural ecosystem. It is not only restricted by the laws of nature but also affected by the social and economic activities of humans [31]. Agricultural eco-compensation can externalize the value of agricultural land ecosystem services by compensating the providers of the ecological value of agricultural land [32]. Eco-compensation measures for farmers have been widely implemented in the practice of protecting and restoring agricultural land ecosystems and reducing the environmental impact of agricultural land use [33,34,35,36]. The current research of eco-compensation on preventing and controlling agricultural NPS pollution mainly includes the evaluation of household willingness to improve the water environment quality by eco-compensation [37,38], the decision-making of a compensation plan in the best management practice of watershed agricultural land [39], and the comparative analysis of “cost–benefit” of eco-compensation [40]. There were also some interdisciplinary studies based on SWAT, such as scenario assessment of environmental impact of ecological engineering [41], impact of climate change on ecosystem habitats [42], and impacts of land-use conversion on the water cycle [43]. However, there are very few studies with comprehensive consideration of the impacts of eco-compensation scenarios on nutrient response, regional society, and economy. There is even less research on the spatial evaluation of eco-compensation scenarios.
Maoli Lake is the reservoir lake in Hunan Province, China. It is also a drinking water source for tens of thousands of residents and a site of a national wetland park. The agriculture in the Maori Lake Watershed is mainly family-scale traditional farming, including rice, vegetables, cotton, etc. Since the 1990s, due to the continuous high-intensity and intensive development of agriculture in this region, the eutrophication of lakes and streams continued to intensify, and the water quality of Maoli Lake declined to Class V around 2013 [44]. According to the Chinese standard GB3838-2002 “Surface Water Environmental Quality Standard”, the water quality is divided into five grades; the larger the number, the worse. Monitoring results in 2018 showed the water quality of MLW had not been fundamentally improved, and the problem of water eutrophication was still prominent. These problems have been seriously affecting the supply of ecosystem services in this region and the sustainable development of this watershed. The aims of this study include (1) assessing improvement of water quality within the MLW by simulating nutrient process under different agricultural eco-compensation scenarios with a verified SWAT model; (2) evaluating the comprehensive effect of the eco-compensation scenarios on precise grid scale and conducting the optimum scenarios selection.

2. Materials and Methods

2.1. Study Site

The Maoli Lake Watershed in Changde City, Hunan Province, China was selected as the research area (Figure 1). It belongs to the Dongting lake system and has a control area of 387.63 km2 (111°51′–111°58′ E, 29°20′–29°29′ N). The watershed belongs to the humid monsoon climate zone at the transition from the middle subtropical zone to the north subtropical zone. Its average annual precipitation is 1164.3 mm, and the average annual temperature is 16.6 degrees Celsius. Since the 1990s, due to the sustained and high-intensity development of agriculture within this area, the nutrient levels of streams and lakes have been increasing, and the water quality has been classified as Class Ⅴ [43]. Since 2011, the local government has implemented a series of measures for preventing and controlling NPS pollution in the watershed. However, the watershed usually has high rainfall, and the agricultural land area accounts for more than 70% of the total area. The main agricultural land types in the study area include rice land, dry land, orchard, forest, and grassland. The rice land is mainly used for interplanting rice with rapeseed, vegetables, and other crops. The dry land is mainly used for rapeseed, cotton, vegetables, wheat, and other crops.

2.2. SWAT Model

The SWAT model is a distributed hydrological model based on a physical mechanism, which was developed by the Agricultural Research Center of the United States Department of Agriculture in 1994 based on the Simulator for Water Resources in Rural Basin (SWRRB) model [19]. During later development processes, some modules in the CREAMS model, environmental policy-integrated climate (EPIC) model, and other similar models were integrated into the SWAT model. The SWAT model has been widely used to simulate hydrological process [45], sediment yield process [46], and transfer process of agricultural chemical pollutants load [47]. According to the formation and occurrence mechanism of NPS pollution, the SWAT model is mainly composed of a hydrological process simulation module, soil erosion module, and pollution load module [48]. According to the differences among land use types, soil types, and slopes in different regions, the watershed is divided into several sub-watersheds, and each sub-watershed is divided into several hydrological response units (HRUs).
The SWAT model separate hydrologic simulation into land phase and routing phase. The driving force in the land phase is the water balance equation:
S W t = S W 0 + i = 1 t R d a y Q s u r f a c e E a W s e e p Q g w
In Equation (1), S W t represents final soil moisture content, (mm); S W 0 represents soil moisture of the previous day, (mm); t represents total time, (day); R d a y represents the precipitation in the i-th day, (mm); Q s u r f a c e represents the surface runoff in the i-th day, (mm); E a represents the evapotranspiration in the i-th day, (mm); W s e e p represents the amount of percolation and bypass flow exiting the soil profile bottom in the i-th day, (mm); Q g w represents the recharge water in the i-th day, (mm).
After finishing the calculation of the land phase, the water transport enters the routing phase, where sediments, nutrients, and other organic chemicals are transported via water. The SWAT model provides two routing models [48] for users: characteristic river length method and Muskingum method. The characteristic river length method was adopted in this research. The universal soil loss equation [49] is used to calculate erosion caused by rainfall and runoff in the SWAT model of this research:
S E D = 11.8 Q s u r f · Q p e a k · A R E A H R U 0.56 · K U S L E · C U S L E · P U S L E · L S U S L E · C F R G  
In Equation (2), SED is the sediment yield in a given day, (tons); Qsurf is the surface runoff volume, (mm/ha); Qpeak is the peak runoff rate, (m3/s); AREAHRU is the area of the HRU, (ha); KUSLE is the USLE soil erodibility factor, (0.013t·m2·hr/(m3·t·cm)); CUSLE is the USLE cover and management factor; PUSLE is the USLE support practice factor; LSUSLE is the USLE topographic factor; CFRG is the coarse fragment factor.
The entire nutrient cycling process of nitrogen and phosphorus can be simulated by the SWAT model. In this model, partial nitrogen is absorbed by the plant, and nitrate and organic nitrogen may be removed from the soil via water flow. The amount of nitrate nitrogen contained in runoff, lateral flow, and percolation is estimated by the product of water volume and the average concentration of nitrate in the layer. Organic nitrogen transported by sediment is calculated with an empirical loading function [50,51]. As for the phosphorus, the transport mechanism of organic phosphorus is similar to that of organic nitrogen. The amount of soluble phosphorus removed in runoff is predicted by the solution phosphorus concentration in the top 10 mm of soil, the runoff volume, and a partitioning factor [48,52].

2.3. Description of Input Data for SWAT

In order to construct the SWAT model, digital elevation model, soil, land use map, weather, measured runoff, and water quality data were utilized. For better water quality simulation, local agricultural management information was also required. These data are detailed in Table 1.

2.4. Parameters Sensitivity Analysis, Calibration, and Validation

In this research, the MLW was defined as 80 sub-watersheds and further divided into 2460 HRUs. The SWAT calibration and uncertainty program (SWAT-CUP) [53] was chosen to analyze the uncertainty and calibrate the parameters. The previous studies [21,53,54] have demonstrated its applicability. The sequential uncertainty fitting algorithm integrated in SWAT-CUP was used for parametric sensitivity and calibration. Owing to limited data, the calibration and validation periods of runoff were set as October 2014–April 2016 and May 2016–November 2017, respectively; the calibration and validation periods of nutrients were January 2014–December 2015 and January 2016–December 2017, respectively. The sampling site was located in Baiyian Creek, which is near the entrance of Maoli Lake (111°52′24.58″ E, 29°28′7.21″ N).
According to the relevant research [55,56,57] and fully considering the characteristics of MLW, 12 parameters for runoff and 11 parameters for nutrients with higher sensitivity were selected to analyze uncertainty and perform calibration (more details are shown in Appendix A).The Nash–Sutcliffe efficiency (NSE) coefficient, determination coefficient R2, and relative volume error (RVE) were used to evaluate the performance of the SWAT model. According to Moriasi [58] and Mamo [59], for monthly runoff simulation, if NSE > 0.5, −10 < RVE < 10, and R2 > 0.6, the simulation performance can be satisfactory; for monthly nutrient simulation, if NSE > 0.35, −10 < RVE < 10, and R2 > 0.3, the simulation performance can be satisfactory.
Table 2 shows the simulated and observed monthly runoff, total nitrogen, and total phosphorus from October 2014 to November 2017. Although the NSE coefficient of the calibration period is only 0.619, the validation period is 0.712, which is close to the application requirement of 0.70. The RVE of calibration period is 4.22 and the validation period is −1.77, which indicates that the simulation deviation is relatively small and meets the application requirement. According to the research of Rahbeh [60], it is qualified for runoff simulation.
In line with Figure 2, it can be inferred the model produces an obvious error in the simulation of 2016 (wet year), which affects the fitting effect of the simulation results in the whole evaluation period to a certain extent. On the whole, the simulation effect is acceptable. The simulation results also can be seen in Figure 3 and Figure 4. These values are qualified for nutrient simulation according to Rahbeh [54] excluding the validation period of TOT_N, which may be caused by the fact that the measured TOT_N and TOT_P data are not dense enough. Due to the limitation of the number of monitoring equipment and the distance between monitoring points, only one sample was used to calculate the concentration of TOT_N and TOT_P per month. This may be the main reason for such accuracy. However, the relative error between the simulated value and the measured value in each period is small, which indicates that the yearly simulations of TOT_N and TOT_P are accurate and can be used to evaluate the nutrient response on a planning scale.

2.5. Scenarios Designs

Generally, the design of the eco-compensation scenario is based on the comprehensive consideration of multiple factors, such as the government policy background, the local social and natural status, and the investigation of stakeholders [61]. Since the beginning of the 20th century, China has successively implemented many ecological restoration projects closely related to eco-compensation [62], including the Grain for Green project, fallow, and manure management. These measures have been implemented in many regions of China and achieved significant ecological and environmental benefits [63]. Based on the background of Chinese environmental policy, this research takes “nutrient input control” [64] as the key of mitigation of agricultural NPS pollution and designs 3 agricultural adjustment scenarios for the MLW. All scenarios are described in detail as follows, and the planting and fertilization management design of each scenario are displayed in Appendix B.
S0: Current situation. Under this scenario, the setting of fertilization and farming activities is maintained according to current practices.
S1: Halving fertilization every year. Under this scenario, the control of agricultural land fertilization in the watershed is emphasized. The amount of fertilizer applied to all rice land, dry land, and orchard is reduced in the whole watershed. This scenario is based on the assumption that farmers are fully publicized and reliably supervised by our village-level organization “village committee”, which can guarantee that the fertiliser reduction measure can produce the expected results that is consistent with the SWAT calculated result.
S2: Fallow every other year. Under this scenario, the impact of fallow measures on the pollution load in the watershed is emphasized. This scenario was applied to all rice land and dry land in the watershed. Following normal farming in the first year, a grass-planting measure is implemented on agricultural land in the second year. Weeding is carried out at the end of the year, and the resulting biomass is kept as fertilizer. Normal farming resumes in the third year, and so on.
S3: Returning agricultural land to forest. Under this scenario, the impact of conversion measures on the pollution load in the watershed is emphasized. There are many lakes and rivers throughout the watershed, which may cause fertilizer to enter the watershed circulation to a greater extent. Thus, the condition for returning agricultural land to the forest is formulated: rice land and dry land within 200 m from the lake and 100 m from the main water system can be changed from agricultural land to forest.

2.6. Construction of Evaluation Model for Ecological Ccompensation Scenarios

The AHP combined with GIS, one of the most effective methods for multi-attribute ranking/weighting problem, has been widely used in the field of environmental policy evaluation and decision making [65]. In order to evaluate eco-compensation scenarios, a spatial multi-criteria evaluation (SMCE) model is constructed with the analytic hierarchy process (AHP) as the core component [66] in this paper.

2.6.1. Construction of Hierarchy Structure of Comprehensive Evaluation Model

The previous research studies show that the expression for the comprehensive effect of environment policy is usually based on a collaborative perspective of social, economic, and environmental dimensions [67]. Therefore, we build the SMCE to evaluate the comprehensive effect of eco-compensation scenarios from these three dimensions. The “nitrogen and phosphorus reduction” is choosen as the indicator of environmental effect [68]. The “farmland income” is choosen as the indicator of economic effect [69], and the “participation willingness of eco-compensation” is choosen as the indicator of social effect [36]. Taking spatial agricultural land grid as the basic computing unit [70], the comprehensive effect evaluation of farmland use based on these indicators can be carried out using the GIS platform. Figure 5 shows the framework and index system of the SMCE model.

2.6.2. Determination of Data Layer

The 10-m square farmland grid is the basic unit of the model operation, and a total of 1,350,727 grids of agricultural land in MLW were divided by GIS software. Only two existing land use types, rice land and dry land, are calculated, because the area of orchards only accounts for 2.35% of the total watershed area, and its compensation measure is only in S1. The spatial distribution map of farmland and the S3 target region is shown in Figure 6.
The input data for the SMCD model were derived from SWAT simulation results of HRU scale, and the eco-compensation questionnaire survey involves 370 householders in MLW between 2018 and 2019. It is assumed that there is a total of j grids for eco-compensation, and the values of B2 (low-nitrogen degree of water) and B3 (low-phosphorus degree of water) of the grid can be obtained by normalizing the difference of the maximum value of all grids in the SWAT simulation results minus the value of this grid (Equations (3) and (4)). The input values of B4 (farmland income) and B5 (participation willingness) are obtained from the questionnaire survey, and different types of farmland are assigned different values.
B 2 i s = max T O T _ N k s T O T _ N i s / max T O T _ N k s ;   k = 1 , 2 , 3 , , j ;   s S 0 , S 1 , S 2 , S 3
B 3 i s = max T O T _ P k s T O T _ P i s / max T O T _ P k s ;   k = 1 , 2 , 3 , , j ;   s S 0 , S 1 , S 2 , S 3
The annual compensation (COST) in different scenarios and the corresponding input values of indicator layer B are shown in Table 3. The COST is determined according to the current policies launched by the local government in recent years and the relevant references [71,72].

2.6.3. Determination of Weights

In order to determine the weight of indicators of each level, government staff, stakeholders, and experts in various fields are organized to conduct participatory discussion [73,74] where the AHP is used to score their opinions, and the average value of all their weights is used for the final weight value. The weight of AHP is determined by the way of stakeholder discussion, including government officials, stakeholders (farmers), and experts in various fields. The Balanced Scorecard method was used to collect and process data. The government officials came from the relevant departments that made the eco-compensation policy, including the local Ministry of Agriculture, the local Ministry of Environmental Protection, and the local Ministry of Finance. Each department appointed 1 representative. Farmers came from three town-level administrative regions in the study area, and each town appoints 1 person as its representative. The expert group included 1 expert each from agricultural environment science, economics, and sociology.

2.6.4. Model Decision

The SMCE model is used to calculate D9 (comprehensive effect of land use) of each grid, and the spatial differences of all grids can be analyzed. By comparing the changes of D9 before and after eco-compensation, E10 (the comprehensive effect increments of per unit compensation cost) of each grid can be calculated by Equation (5), and its spatial distribution can be mapped for decision-making analysis. Finally, the optimal eco-compensation scenario can be determined by comparing E11, which is the mean value of E10 of all grids (Equation (6)).
E 10 i s = D 9 i s D 9 i S 0 / COST i s ; s S 1 , S 2 , S 3  
E 11 i s = 1 / j · i = 1 j E 10 i s ; s S 1 , S 2 , S 3  

3. Results

3.1. Characteristics of NPS Pollution Load in MLW

3.1.1. Temporal Distribution

Based on the simulation results of S0, the basic characteristic of NPS pollution load in the study area can be analyzed. Figure 7 describes the NPS pollution load of TOT_N and TOT_P in the 8-year simulation period. The interannual difference of TOT_N and TOT_P is obvious. The TOT_N in the simulation period is (1377.50 ± 374.28) t each year, and the TOT_P is (200.95 ± 78.96) t each year. The highest NPS nitrogen pollution load is 1583.14 t in 2015, and the highest NPS phosphorus pollution load is 235.83 t in 2016. The lowest values of nitrogen and phosphorus pollution are respectively 1003.22 t and 121.99 t in 2011.
Figure 8 shows the nutrient output of the months with abundant precipitation (Mar. to Jun.) is greater than that of other months throughout the year, and the nutrient output of the months with the least rainfall (Dec. to Feb. of the next year) is the least. The NPS pollution load is mainly concentrated in the wet season from Mar. to Jun. In 2011–2018, the nitrogen pollution load from Mar. to Jun. accounts for 66.51–78.38% of the whole year, and the phosphorus pollution load accounts for 61.29–78.69% of the whole year. According to the time distribution characteristic of NPS nitrogen and phosphorus pollution load in the MLW, Mar. to Jun. is the key period for the reduction of pollution. In this period, the management measures of fertilizer application, such as reducing fertilizer application and fallow, can effectively reduce the impact of NPS pollution in theory.

3.1.2. Spatial Distribution

During the simulation period, the annual average of the TOT_N nutrient loss load in different sub-watersheds ranges from 0 to 77.69 t, and the annual average of TOT_P nutrient loss load in different sub-watersheds ranges from 0 to 10.2 t. The distribution of TOT_N and TOT_P nutrient loss exhibits a similar differentiation law. The high value of nutrients was mainly concentrated in the 73th, the 75th, and the 77th sub-watersheds in the southwest of MLW. The contribution of pollution in the southwest mountainous area and near the central and western lake entrance is greater than that in the eastern flat area.

3.2. Pollution Distribution under Eco-Compensation Scenarios

3.2.1. Temporal Distribution of Pollution under Eco-Compensation Scenarios

Figure 9 shows that the annual variation trend of nitrogen and phosphorus NPS pollution load is similar under each scenario. S1 is slightly better than S3 in decreasing the total amount of nitrogen and phosphorus, but its effect is not significant enough comparing to the fallow year of S2. S2 can reduce more nitrogen and phosphorus during the fallow year than other scenarios, but its nitrogen and phosphorus in the first year are nearly the same as S0. Except for the first year, the amount of nitrogen and phosphorus of S2 in subsequent farming years still has a certain reduction comparing to S1. This indicates that S2 can reduce nutrients not only in the fallow year but also in the farming year. Comprehensively considering the yearly scale, S2 works better than S1 and S3 in mitigating total nitrogen and phosphorus. Moreover, the monthly simulation results of TOT_N and TOT_P load (Appendix C) show that the decreased nutrient load of S2 is the largest comparing to the decrease of S1 and S3.

3.2.2. Spatial Distribution of Pollution Load under Eco-Compensation Scenarios

The spatial distribution of TOT_N and TOT_P in the S0 are shown in Figure 10 and Figure 11. Figure 10 describes the distribution of annual NPS TOT_N load in the study area from 2011 to 2018 under each scenario. Taking 80 sub-watersheds as the control unit, the TOT_N load of each sub-watershed reduces in different degrees under different design scenarios compared to S0. The average annual loss load of TOT_N in the S1 is 0–62.21 t in different sub-watersheds; the average annual loss load of TOT_N in the S2 is 0–47.92 t, and that in the S3 is 0–77.69 t. Similar to S0, the high values of each scenario are mainly concentrated in the 57th, 59th, 62th, 63th, 76th, and 78th sub-watersheds, and the highest values are concentrated in the 73th, 75th, and 77th sub-watersheds in the southwest.
Figure 11 describes the average annual NPS TOT_P load distribution of each scenario from 2011 to 2018. The figure shows that the TOT_P load of each sub-watershed reduces in different degrees under different scenarios compared to S0. The average annual loss load of TOT_P in different sub-watersheds varied from 0 to 8.53 t under S1, from 0 to 7.34 t under S2, and from 0 to 10.2 t under S3. The spatial distribution of each scenario is similar to S0, and the high value of each scenario is mainly concentrated in the similar sub-watershed with S0.

3.2.3. Characteristic of NPS Pollution Load in Different Types of Agricultural Land

Table 4 shows the nutrient loss of different types of agricultural land in the study area under S0. When the HRU is used as the control unit, the average annual nitrogen loss of agricultural land (including rice land, dry land, orchard, forest, and grass land) was 1162.4 t, and phosphorus loss was 134.6 t. Rice land and dry land are the main types of agricultural land in the study area and main sources of nitrogen and phosphorus loss. The total area of rice land accounts for 28.87% of the watershed area, and its annual TOT_N loss is 672.1 t, which accounts for 57.82% of the total loss. The total area of dry land accounts for 16.53% of the watershed area, and its annual TOT_N loss is 436.25 t, which accounts for 37.53% of the total loss. The orchard accounts for 2.36% of the total agricultural land but contributes 3% of the total nitrogen loss. The forest land accounts for 19.68% of the total agricultural land but only contributs 0.8% of nitrogen loss and 1.41% of phosphorus loss. The percentages of contribution of grass land to nitrogen and phosphorus loss are 0.11% and 0.8%, respectively.
Taking the agricultural land per unit area as the control unit, the comparison of TOT_N and TOT_P NPS load of different types of agricultural land in each scenario is illustrated in Table 5. Taking rice land as an example, under S0, the nitrogen loss rate reaches 62.33%, and the phosphorus loss rate is 11.06%. Under S1, the nitrogen use in rice land reduces by half, and the nitrogen loss rate reduces to 59.21%, which was 150.85kg·a−1 per hectare; the total nitrogen loss reduces by 52.51%, and the nitrogen loss rate reduces by 5.01%. Under S1, the use of phosphorus fertilizer in rice land also reduces by half, and the non-point phosphorus load reduces to 20.11 kg·a−1 per hectare. Compared with S0, the non-point phosphorus load reduces by 34.23%, but the phosphorus loss rate increases by 31.54%. The average annual nitrogen and phosphorus input, which is half of S0, in S2 is consistent with that in S1, but S2 exhibits better NPS pollution reduction efficiency than S1; in S2, the nitrogen loss rate reduces to 47.14%, and the phosphorus loss rate reduces to 7.62%. The values of total amount of nitrogen and phosphorus are respectively 120.11 kg·a−1 and 10.53 kg·a−1 per hectare and reduce by 62.18% and 65.57% comparing to S0, while the loss rates of nitrogen and phosphorus respectively reduce by 24.37% and 31.13% comparing to S0. However, under S3, the loss of nitrogen and phosphorus nutrients per unit area of rice land increased slightly, which were 328.23 kg·a−1 and 32.41 kg·a−1 per hectare, respectively. Compared with S0, the loss rates of nitrogen and phosphorus increase by 3.34% and 6.06%, respectively.
In each scenario as shown in Table 5, the load of the orchard is medium; the load of rice land and dry land is relatively high, and the load of forest and grass land is relatively low. The order of annual average loss load of nitrogen and phosphorus of agricultural land per unit area is: dry land > rice land > orchard > grass land > forest.
The loads of nitrogen and phosphorus of rice land and dry land in S0 are the same as those in S3; the nitrogen load of the two types of land are similar, while the phosphorus load of dry land is slightly higher than that of rice land. In S1, the nitrogen and phosphorus load of rice land, dry land, and orchard significantly reduce, and the load difference among these three types of agricultural land reduces to a certain extent. The total NPS load of S2 is lower than that of S1 because the loads of rice land and dry land in S2 are slightly lower than that of S1, and the load of the orchard is higher than that of rice land and dry land and becomes the highest load. Generally, rice land, dry land, and orchard are the main land of NPS pollution in the MLW. However, the NPS pollution in different areas of the watershed has different characteristics. Therefore, it is necessary to consider the probability of load loss of different nutrient in farmland when carrying out accurate prevention and control of NPS pollution load.

3.3. Results of SMCE Model and Evaluation of Eco-Compensation Scenarios

The SMCE model was used to evaluate the comprehensive effect of farmland utilization under different eco-compensation scenarios. The evaluation was carried out in each target farmland grid within the watershed, and corresponding quantitative values were obtained. In scenario S0/S1/S2, the grid counts of rice land and dry land were 839,207 and 511,520, respectively. In scenario S3, the grid count of rice land and dryl and were 199,223 and 89,215, respectively.

3.3.1. Spatial Distribution of D9 and E10 under Different Scenarios

The spatial distribution of D9 and E10 under each eco-compensation scenario is plotted as shown in Figure 12 and Figure 13. It can be seen from the figures that the distribution of D9 and E10 in different scenarios is greatly different, and the significant spatial distribution difference also appears on the grid scale under a same scenario. As shown in Figure 13, based on the perspective of the whole watershed, the order of increment of comprehensive effect per unit compensation cost is S2 > S1 > S3.

3.3.2. Calculation Results of Indicators of Each Layer of SMCE Model

Table 6 shows the averages of all target grids of nodes of the SMCE model under different scenarios. By comparing C6, the order of environmental effects of eco-compensation scenarios is S3 > S2 > S1. By comparing C7, the economic effect of eco-compensation scenarios is S1 > S2 > S3. By comparing C8, it can be seen that the social effect of eco-compensation scenarios is S2 > S1> S3.
Through the analysis of the results of D9 in Table 6, the comprehensive effects of all eco-compensation scenarios are improved to a different extent compared to those of S0. By comparing E10, it can be found that the E10 value of dry land is greater than that of rice land under each eco-compensation scenario. By combining the scenario with land type, it can be found that the E10 value in “S2 on dry land” is the highest, followed by “S2 on rice land”, and the low E10 values appears in “S1 on rice land” and “S3 on rice land”.

3.3.3. Evaluation and Optimum Selection of Eco-Compensation Scenarios

By comparing the results of E10 (Table 6), the “S2 on dry land” can be regarded as the best eco-compensation measure in MLW, followed by the “S2 on rice land”. If only one of the above two measures is chosen, with the consideration of irregular distribution of the rice land and dry land within the watershed, the management difficulty and implementation cost will increase. Therefore, we think that the fallow scenario (S2) both on rice land and dry land is the best eco-compensation scenario in MLW for NPS pollution control.
It is also worth noting that the environmental effects (C6) of S3 are relatively optimal, but its C8 and D9 are lower due to its relatively lower compensation COST standard of current policy. If the compensation COST standard of S3 can be improved or phased compensation (higher compensation standard before the trees become timber and then lower) can be implemented in future, S3 can be further compared with S2.

4. Discussion

4.1. Qualitative Evaluation of Simulation Inaccuracy under Eco-Compensation Scenarios

After the calibration of the SWAT model, the performance of this SWAT model in the MLW is qualified for runoff simulation and TOT_N and TOT_P simulation but not as good as other similar studies that simulate hydrological processes by the SWAT model [21,22,46]. This may be caused by the applicability of the SWAT model. The SWAT model is a distributed hydrological model and suitable for hydrological simulation of large and medium watersheds [19], but the MLW is a small watershed with an area of 387.63 km2. For small watershed simulation, the input data need to be extremely accurate. For the simulation of large and medium scale, human activities can be assimilated in hydrological simulation. However, in a small watershed, these factors of human activities can not be ignored. This may bring some uncertainty to the calibration. In addition, the measured data used for calibrating TOT_N and TOT_P parameters is monitored once per month, which may cause some uncertainty in calibration work. Moreover, the planting and fertilization schedule for rice, dry crops, and orchards within the watershed are generalized to the same period. Although they are obtained by field investigation, they are still not completely consistent with the actual situation. These uncertainties are associated with the objective limitation of the study subjects and methods.
Furthermore, for most hydrological models, equifinality exists among different parameters [75,76]. This means that there are many groups of parameters that could lead to good simulation results, but parameter settings vary in each compositional simulation [77]; this may lead to uncertainty for different scenario simulations. The scenarios set in this research are based on three different measures that may have a positive effect on decreasing nutrient loss. However, the detail for each scenario is set subjectively. The simulation results under each scenario also yield some uncertainties. However, the qualitative analysis for nutrient responses under different scenarios is influenced less by these uncertainties [78]. When considering equifinality for different parameters, more appropriate methods for analyzing nutrient responses under different scenarios require further investigation.
For the simulation of S3, considering the long-term character of the implementation of eco-compensation measures and the difficulty in obtaining the data for the simulation of the long growth process of forest, the calculation of the response is based on the simulation results in scenario with mature forest, rather than the simulation across two scenarios. Therefore, the simulation results are based on the assumption that the measures have been successfully carried out. These may also bring some uncertainty in scenario simulation.

4.2. Advice on NPS Pollution Control within MLW

The results from scenarios analysis indicate that agricultural eco-compensation measures aimed at reducing the application of chemical fertilizers are effective for reducing NPS pollution in the MLW. However, farmland eco-compensation will bring some economic and social effects on local householders and community. So, it is necessary to combine environmental, social, and economic effects into the scenario evaluation of eco-compensation.
The result of the SMCE model on the 10 m × 10 m grid scale indicates that the comprehensive effect of eco-compensation in different regions of the watershed has significant spatial differences, as well as shown in Figure 12 and Figure 13. Obviously, a higher value of comprehensive effect (E10) can be achieved by selectively compensating for those grids that have a relatively high comprehensive effect. In each compensation scenario, the economic effects (C7) or social effects (C8) of all grids with the same land use are of the same value, which is the mean value calculated by the results of the questionnaire survey. Therefore, the spatial difference of the comprehensive effect of the same land use mainly comes from the difference of environmental effect (C6). As a consequence, the spatial and differentiated compensation can be further carried out in line with the spatial difference of C6, and the optimal compensation area can be determined by combining different land use types. If equipped with appropriate technological and management measures, eco-compensation measures on the grid scale will obtain better comprehensive effect and cost-effectiveness, which is both the necessity and significance of controlling agricultural NPS pollution by eco-compensation on a precise spatial scale.
In general, a good fertilization plan should be able to fulfill the crop requirements with low fertilizer. In this case, the widespread implementation of “optimal fertilizer use” [79] is very likely to be the best NPS control measure. However, the extent of pollution reduction by the method of “optimal fertilizer use” is usually not very large and has an upper limit. This method is not particularly suitable because the research area in this article is a national natural protected area, which needs to greatly reduce pollution. However, for the areas without developed economic conditions or nature reserve, they are significantly different from the study area of this paper, so it is necessary to fully study the weights of environmental, economic, and social effects and then recalculate the comprehensive effect according to the framework of this paper to determine the optimal compensation measures.

5. Conclusions

Three agricultural activity adjustment scenarios were designed in this research: S1 (halving fertilization every year), S2 (fallow every other year), and S3 (returning agricultural land to forest). The SWAT model constructed in this paper could basically meet the requirements of the simulation application of runoff and pollution in Maoli Lake Watershed. The following major conclusions can be drawn:
(1)
The nutrients loss of the agricultural land reduced in all scenarios, and S2 had more of a reduction compared to S1 and S3. The implementation of each scenario will have an impact on the farmers’ farmland income, and the impact on S2 and S3 is relatively large. As the result of different compensation cost standards, farmers’ willingness to participate in different scenarios is different. Their willingness to participate in S2 is the highest, followed by S1, and that of S3 is the lowest.
(2)
With the combined perspective of environment–economy–society effects, the rice land and dry land fallow every other year (S2) is the best eco-compensation scenario in MLW for NPS pollution control in the near future.
(3)
The comprehensive effect of eco-compensation at grid scale has significant spatial difference. When an eco-compensation scheme is used to control NPS pollution, a better result can be obtained by accurate and differentiated compensation on the precise spatial scale.
(4)
This study was carried out from the perspective of combining SWAT and an eco-compensation scheme, which could broaden the application fields of the SWAT to some extent and provide a scientific basis and experience for the comprehensive evaluation and spatial design of agricultural eco-compensation.
(5)
For the areas without developed economic conditions or nature reserve, they are significantly different from the study area of this paper, so it is necessary to fully study the weights of environmental, economic, and social effects and then recalculate the comprehensive effect according to the framework of this paper to determine the optimal compensation measures. The widespread implementation of “optimal fertilizer use” in this case is very likely to be the best NPS control measure.

Author Contributions

All the authors made significant contributions to the work. G.L. supervised the study. Y.Z. collected data, developed analysis models, and drafted the paper. P.Y. was responsible for data analysis, parallel calculation, and results verification. All authors made a contribution to writing, revising, and proofreading the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Key Research and Development Program of China (Grant Number 2017YFC0405303), the Foundation of Humanity and Social Science of China Ministry of Education (Grant Number 20YJC630231), the Scientific Research Project of Education Department of Hunan Province of China (Grant Number 18K059), and the Science and Technology Research and Development Program of China State Construction Engineering Corporation Ltd. (Grant Number CSCEC-2021-Z-26 and CSCEC-2020-Z-43).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Informed consent was obtained from all stakeholders involved in the study.

Data Availability Statement

The data are contained within this article and references.

Acknowledgments

The authors would like to thank Ting Liu (Management School, Swan College of Central South University of Forestry and Technology, China) for many helpful discussions and computer programming regarding research process and manuscript preparation.

Conflicts of Interest

The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Appendix A

Table A1. Ranges and calibrated values of the sensitive parameters selected in this SWAT model.
Table A1. Ranges and calibrated values of the sensitive parameters selected in this SWAT model.
Parameters 1DescriptionInitial ValueRangeAdjusted Value
R__CN2.mgtSCS runoff curve number67(−0.5, 0.5)0.12
V__SURLAG.bsnSurface runoff lag time (1, 24)3.59
R__SOL_K.solSaturated hydraulic conductivity (mm/hr)264.6(−0.5, 0.5)0.08
R__SOL_AWC.solAvailable water capacity of the soil layer (mm/mm)0.344(−0.5, 0.5)−0.03
V__CANMX.hruMaximum canopy storage (0, 30)4.29
V__GW_REVAP.gwGroundwater “revap” coefficient (0.02, 0.2)0.09
V__REVAPMN.gwThreshold depth of water in the shallow aquifer for “revap” to occur (mm) (0, 1000)49.92
V__GW_DELAY.gwGroundwater delay (days) (0, 500)173.12
V__ALPHA_BF.gwBaseflow alpha factor (days) (0, 1)0.39
V__GWQMN.gwThreshold depth of water in the shallow aquifer required for return flow to occur (mm) (0, 500)91.32
V__ESCO.hruSoil evaporation compensation factor (0, 1)0.19
V__EPCO.hruPlant uptake compensation factor (0, 1)0.35
V__SPCON.bsnLinear parameter for calculating the maximum amount of sediment that can be re-entrained during channel sediment routing (0.001, 0.01)0.015
V__SPEXP.bsnExponent parameter for calculating sediment reentrained in channel sediment routing (1, 1.5)1.38
R__USLE_P.mgtUSLE equation support practice factor0.53(−0.5, 0.5)−0.35
V__CH_N2.rteManning’s “n” value for the main channel (0, 0.3)0.11
V__SOL_ORGN.chmInitial organic N concentration in the soil layer, (mg/kg). (0, 100)72.2
V__PHOSKD.bsnPhosphorus soil partitioning coefficient. (100, 200)125
V__NPERCO.bsnNitrogen percolation coefficient (0, 1)0.35
V__SOL_ORGP.chmInitial organic P concentration in surface soil layer, (mg/kg) (0, 100)85.9
V__CDN.bsnDenitrification exponential rate coefficient (0, 3)0.72
V__PSP.bsnPhosphorus sorption coefficient (0.01, 0.7)0.41
V__RCN.bsnConcentration of nitrogen in rainfall (0, 15)2.3
1 R__ means the existing parameter value is multiplied by (1 + an given value); V__ means the existing parameter value is to be replaced by the given value.

Appendix B

Table A2. Planting and fertilization management of each scenario in SWAT operations. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Table A2. Planting and fertilization management of each scenario in SWAT operations. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Farmland TypeMeasureMeasure TimeLand CoverFertilization Management of Each Scenario kg/(ha·yr−1)
S0S1S2(c)S2(d)S3
Dry landSowing1 Jan. -
Fertilization15 Mar.Dry land149.99 (b)74.99 (b)0149.99 (b)149.99 (b)
Fertilization15 Apr.Dry land112.49 (b)56.24 (b)0112.49 (b)112.49 (b)
harvest and removing15 May -
Sowing20 May -
Fertilization25 MayDry land899.96 (a)449.98 (a)0899.96 (a)899.96 (a)
Fertilization15 Jun.Dry land262.49 (b)131.24 (b)0262.49 (b)262.49 (b)
Fertilization15 Jul.Dry land224.89 (b)112.44 (b)0224.89 (b)224.89 (b)
Fertilization15 Aug.Dry land187.49 (b)93.74 (b)0187.49 (b)187.49 (b)
harvest and removing20 Oct. -
Sowing25 Oct. -
Fertilization30 Oct.Dry land674.97 (a)337.48 (a)0674.97 (a)674.97 (a)
Fertilization15 Nov.Dry land187.49 (b)93.745 (b)0187.49 (b)187.49 (b)
harvest and removing25 Dec.Dry land -
Rice landSowing1 Jan. -
Fertilization5 Jan.Dry land149.99 (b)74.99 (b)0149.99 (b)149.99 (b)
Fertilization10 Mar.Dry land112.49 (b)56.24 (b)0112.49 (b)112.49 (b)
harvest and removing20 Apr. -
Sowing25 Apr. -
Fertilization30 Apr.Rice land749.63 (a)374.81 (a)0749.63 (a)749.63 (a)
Fertilization15 MayRice land149.93 (b)74.96 (b)0149.93 (b)149.93 (b)
Fertilization15 Jul.Rice land149.93 (b)74.96 (b)0149.93 (b)149.93 (b)
harvest and removing25 Sep. -
Sowing1 Oct. -
Fertilization5 Oct.Dry land674.97 (a)337.48 (a)0674.97 (a)674.97 (a)
Fertilization10 Dec.Dry land187.49 (b)93.74 (b)0187.49 (b)187.49 (b)
harvest and removing25 Nov. -
OrchardFertilization15 Mar.Orchard149.92 (a)74.96 (a)149.92 (a)149.92 (a)149.92 (a)
Fertilization15 MayOrchard374.81 (b)187.40 (b)374.81 (b)374.81 (b)374.81 (b)
harvest15 Nov.
Fertilization25 Nov.Orchard899.55 (b)449.77 (b)899.55 (b)899.55 (b)899.55 (b)
Forest--Forest00000
Grass land--Pasture00000
(a): Synthetic fertilizer; (b): Urea; (c): Fallow year of S2; (d): Farming year of S2.

Appendix C

Figure A1. Simulation of monthly TOT_N load in MLW under different scenarios.
Figure A1. Simulation of monthly TOT_N load in MLW under different scenarios.
Water 13 01536 g0a1
Figure A2. Simulation of monthly TOT_P load in MLW under different scenarios.
Figure A2. Simulation of monthly TOT_P load in MLW under different scenarios.
Water 13 01536 g0a2

References

  1. Cherry, K.; Shepherd, M.; Withers, P.; Mooney, S. Assessing the effectiveness of actions to mitigate nutrient loss from agriculture: A review of methods. Sci. Total Environ. 2008, 406, 1–23. [Google Scholar] [CrossRef]
  2. Smith, L.E.; Siciliano, G. A comprehensive review of constraints to improved management of fertilizers in China and mitigation of diffuse water pollution from agriculture. Agric. Ecosyst. Environ. 2015, 209, 15–25. [Google Scholar] [CrossRef] [Green Version]
  3. Xueman, Y.; Wenxi, L.; Yongkai, A.; Dong, W. Assessment of parameter uncertainty for non-point source pollution mechanism modeling: A Bayesian-based approach. Environ. Pollut. 2020, 263, 114570. [Google Scholar] [CrossRef]
  4. Lu, H.; Xie, H. Impact of changes in labor resources and transfers of land use rights on agricultural non-point source pollution in Jiangsu Province, China. J. Environ. Manag. 2018, 207, 134–140. [Google Scholar] [CrossRef] [PubMed]
  5. Liu, G.; Wu, W.; Zhang, J. Regional differentiation of non-point source pollution of agriculture-derived nitrate nitrogen in groundwater in northern China. Agric. Ecosyst. Environ. 2005, 107, 211–220. [Google Scholar] [CrossRef]
  6. Haregeweyn, N.; Yohannes, F. Testing and evaluation of the agricultural non-point source pollution model (AGNPS) on Augucho catchment, western Hararghe, Ethiopia. Agric. Ecosyst. Environ. 2003, 99, 201–212. [Google Scholar] [CrossRef]
  7. Lai, Y.; Yang, C.; Hsieh, C.; Wu, C.; Kao, C. Evaluation of non-point source pollution and river water quality using a multimedia two-model system. J. Hydrol. 2011, 409, 583–595. [Google Scholar] [CrossRef]
  8. Zhang, P.; Liu, Y.; Pan, Y.; Yu, Z. Land use pattern optimization based on CLUE-S and SWAT models for agricultural non-point source pollution control. Math. Comput. Model. 2013, 58, 588–595. [Google Scholar] [CrossRef]
  9. Xiang, C.; Wang, Y.; Liu, H. A scientometrics review on nonpoint source pollution research. Ecol. Eng. 2017, 99, 400–408. [Google Scholar] [CrossRef]
  10. Chowdary, V.; Kar, S.; Adiga, S. Modelling of non-point source pollution in a watershed using remote sensing and GIS. J. Indian Soc. Remote Sens. 2004, 32, 59–73. [Google Scholar] [CrossRef]
  11. Kinnell, P. AGNPS-UM: Applying the USLE-M within the agricultural non point source pollution model. Environ. Model. Softw. 2000, 15, 331–341. [Google Scholar] [CrossRef]
  12. Wu, L.; Gao, J.; Ma, X.; Li, D. Application of modified export coefficient method on the load estimation of non-point source nitrogen and phosphorus pollution of soil and water loss in semiarid regions. Environ. Sci. Pollut. Res. 2015, 22, 10647–10660. [Google Scholar] [CrossRef]
  13. Borah, D.K.; Bera, M. Watershed-scale hydrologic and nonpoint-source pollution models: Review of applications. Trans. ASAE 2004, 47, 789. [Google Scholar] [CrossRef] [Green Version]
  14. Adu, J.T.; Kumarasamy, M.V. Assessing Non-Point Source Pollution Models: A Review. Pol. J. Environ. Stud. 2018, 27. [Google Scholar] [CrossRef]
  15. Gikas, G.D.; Yiannakopoulou, T.; Tsihrintzis, V.A. Modeling of non-point source pollution in a Mediterranean drainage basin. Environ. Modeling Assess. 2006, 11, 219–233. [Google Scholar] [CrossRef]
  16. Polyakov, V.; Fares, A.; Kubo, D.; Jacobi, J.; Smith, C. Evaluation of a non-point source pollution model, AnnAGNPS, in a tropical watershed. Environ. Model. Softw. 2007, 22, 1617–1627. [Google Scholar] [CrossRef]
  17. Wang, H.; Wu, Z.; Hu, C. A comprehensive study of the effect of input data on hydrology and non-point source pollution modeling. Water Resour. Manag. 2015, 29, 1505–1521. [Google Scholar] [CrossRef]
  18. Cao, D.; Cao, W.; Fang, J.; Cai, L. Nitrogen and phosphorus losses from agricultural systems in China: A meta-analysis. Mar. Pollut. Bull. 2014, 85, 727–732. [Google Scholar] [CrossRef]
  19. Arnold, J.G.; Williams, J.R.; Maidment, D.R. Continuous-time water and sediment-routing model for large basins. J. Hydraul. Eng. 1995, 121, 171–183. [Google Scholar] [CrossRef]
  20. Yang, S.; Dong, G.; Zheng, D.; Xiao, H.; Gao, Y.; Lang, Y. Coupling Xinanjiang model and SWAT to simulate agricultural non-point source pollution in Songtao watershed of Hainan, China. Ecol. Model. 2011, 222, 3701–3717. [Google Scholar] [CrossRef]
  21. Wang, Q.; Liu, R.; Men, C.; Guo, L. Application of genetic algorithm to land use optimization for non-point source pollution control based on CLUE-S and SWAT. J. Hydrol. 2018, 560, 86–96. [Google Scholar] [CrossRef]
  22. Abbaspour, K.C.; Rouholahnejad, E.; Vaghefi, S.; Srinivasan, R. A continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model. J. Hydrol. 2015, 524, 733–752. [Google Scholar] [CrossRef] [Green Version]
  23. Geng, R.; Sharpley, A.N. A novel spatial optimization model for achieve the trad-offs placement of best management practices for agricultural non-point source pollution control at multi-spatial scales. J. Clean. Prod. 2019, 234, 1023–1032. [Google Scholar] [CrossRef]
  24. Sharpley, A.N.; Kleinman, P.J.; Flaten, D.N.; Buda, A.R. Critical source area management of agricultural phosphorus: Experiences, challenges and opportunities. Water Sci. Technol. 2011, 64, 945–952. [Google Scholar] [CrossRef]
  25. Maringanti, C.; Chaubey, I.; Popp, J. Development of a multiobjective optimization tool for the selection and placement of best management practices for nonpoint source pollution control. Water Resour. Res. 2009, 45. [Google Scholar] [CrossRef] [Green Version]
  26. Zhong, H.; Qing, P.; Hu, W. Farmers’ willingness to participate in best management practices in Kentucky. J. Environ. Plan. Manag. 2016, 59, 1015–1039. [Google Scholar] [CrossRef]
  27. Link, J.; Graeff, S.; Batchelor, W.D.; Claupein, W. Evaluating the economic and environmental impact of environmental compensation payment policy under uniform and variable-rate nitrogen management. Agric. Syst. 2006, 91, 135–153. [Google Scholar] [CrossRef]
  28. The Research Group on Mechanism and Policy of Eco-compensation in China. The Research on Mechanism and Policy of Eco-compensation in China; Science Press: Beijing, China, 2007. (In Chinese) [Google Scholar]
  29. Wunder, S. Payments for Environmental Services: Some Nuts and Bolts. 2005. Available online: http://hdl.handle.net/10919/66932 (accessed on 1 January 2005).
  30. Wunder, S. When payments for environmental services will work for conservation. Conserv. Lett. 2013, 6, 230–237. [Google Scholar] [CrossRef]
  31. Firbank, L.; Bradbury, R.B.; McCracken, D.I.; Stoate, C. Delivering multiple ecosystem services from Enclosed Farmland in the UK. Agric. Ecosyst. Environ. 2013, 166, 65–75. [Google Scholar] [CrossRef]
  32. Cong, R.G.; Smith, H.G.; Olsson, O.; Brady, M. Managing ecosystem services for agriculture: Will landscape-scale management pay? Ecol. Econ. 2014, 99, 53–62. [Google Scholar] [CrossRef]
  33. Jack, B.K.; Kousky, C.; Sims, K.R. Designing payments for ecosystem services: Lessons from previous experience with incentive-based mechanisms. Proc. Natl. Acad. Sci. USA 2008, 105, 9465–9470. [Google Scholar] [CrossRef] [Green Version]
  34. Kerr, S.; Greenhalgh, S.; Simmons, G. The Taupo Nitrogen Market: The World’s Only Diffuse Source Trading Programme. MOTU Note #20. Motu Public Policy and Research: Wellington, New Zealand, 2015. Available online: http://www.motu.org.nz/assets/Documents/our-work/environment-andresources/nutrient-trading-and-water-quality/Motu-Note-20-Taupo-Nitrogen-Market (accessed on 16 January 2015).
  35. Aguilar, F.X.; Obeng, E.A.; Cai, Z. Water quality improvements elicit consistent willingness-to-pay for the enhancement of forested watershed ecosystem services. Ecosyst. Serv. 2018, 30, 158–171. [Google Scholar] [CrossRef]
  36. McGurk, E.; Hynes, S.; Thorne, F. Participation in agri-environmental schemes: A contingent valuation study of farmers in Ireland. J. Environ. Manag. 2020, 262, 110243. [Google Scholar] [CrossRef]
  37. Ma, S.; Swinton, S.M.; Lupi, F.; Jolejole-Foreman, C. Farmers’ willingness to participate in Payment-for-Environmental-Services programmes. J. Agric. Econ. 2012, 63, 604–626. [Google Scholar] [CrossRef]
  38. Beharry-Borg, N.; Smart, J.C.; Termansen, M.; Hubacek, K. Evaluating farmers’ likely participation in a payment programme for water quality protection in the UK uplands. Reg. Environ. Chang. 2013, 13, 633–647. [Google Scholar] [CrossRef]
  39. Talberth, J.; Selman, M.; Walker, S.; Gray, E. Pay for Performance: Optimizing public investments in agricultural best management practices in the Chesapeake Bay Watershed. Ecol. Econ. 2015, 118, 252–261. [Google Scholar] [CrossRef]
  40. Saarikoski, H.; Mustajoki, J.; Barton, D.N.; Geneletti, D. Multi-Criteria Decision Analysis and Cost-Benefit Analysis: Comparing alternative frameworks for integrated valuation of ecosystem services. Ecosyst. Serv. 2016, 22, 238–249. [Google Scholar] [CrossRef]
  41. Tian, F.; Yi, H.; Fu, B.; Yang, Y.; Qiu, G.; Zhang, L. Effects of ecological engineering on water balance under two different vegetation scenarios in the Qilian Mountain, northwestern China. J. Hydrol. Reg. Stud. 2015. [Google Scholar] [CrossRef] [Green Version]
  42. Basheer, A.K.; Lu, H.; Omer, A.; Ali, A.; Abdelgader, A. Impacts of climate change under CMIP5 RCP scenarios on the streamflow in the Dinder River and ecosystem habitats in Dinder National Park, Sudan. Hydrol. Earth Syst. Sci. 2016, 20, 1331–1353. [Google Scholar] [CrossRef] [Green Version]
  43. Hu, J.Y.; Wu, Y.; Wang, L.; Sun, P.; Lian, Y. Impacts of land-use conversions on the water cycle in a typical watershed in the southern Chinese Loess Plateau. J. Hydrol. 2021, 593, 125741. [Google Scholar] [CrossRef]
  44. Cao, X. Maoli Lake: A Transformation from Turbid to Limpidly. Hunan Daily. Available online: https://hnrb.voc.com.cn/hnrb_epaper/html/2015-10/20/content_1023619.htm?div=-1 (accessed on 20 October 2015).
  45. Alibuyog, N.R.; Ella, V.B.; Reyes, M.R.; Srinivasan, R.; Dillaha, T. Predicting the effects of land use change on runoff and sediment yield in Manupali River subwatersheds using the SWAT model. Int. Agric. Eng. J. 2009, 18, 15–25. [Google Scholar]
  46. Xu, Z.X.; Pang, J.P.; Liu, C.M.; Li, J. Assessment of runoff and sediment yield in the Miyun Reservoir catchment by using SWAT model. Hydrol. Process. 2009, 23, 3619–3630. [Google Scholar] [CrossRef]
  47. Lin, Z.; Radcliffe, D.E.; Risse, L.M.; Romeis, J.J.; Jackson, C.R. Modeling phosphorus in the Lake Allatoona watershed using SWAT: II. Effect of land use change. J. Environ. Qual. 2009, 38, 121–129. [Google Scholar] [CrossRef] [PubMed]
  48. Neitsch, S.L.; Arnold, J.G.; Kiniry, J.R.; Williams, J.R. Soil and water assessment tool theoretical documentation. Comput. Speech Lang. 2011, 24, 289–306. [Google Scholar] [CrossRef]
  49. Williams, J.R. Chapter 25; The EPIC Model. In Computer Models of Watershed Hydrology; Water Resources Publications: Highlands Ranch, CO, USA, 1995; pp. 909–1000. [Google Scholar]
  50. McElroy, A.D.; Chiu, S.Y.; Nebgen, J.W.; Aleti, A.; Bennett, F.W. Loading Functions for Assessment of Water Pollution from Nonpoint Sources; EPA 600/2-76-151; US Environmental Protection Agency, Office of Research and Development: Washington, DC, USA, 1976. [Google Scholar]
  51. Williams, J.R.; Hann, R.W. Optimal Operation of Large Agricultural Watersheds with Water Quality Constraints; Texas A&M Univ., Tech. Rept. No. 96; Texas Water Resources Institute: Dallas, TX, USA, 1978. [Google Scholar]
  52. Long, Y.; Tang, R.; Wang, H.; Jiang, C. Monthly precipitation modeling using Bayesian non-homogeneous hidden Markov Chain. Hydrol. Res. 2019, 50, 562–576. [Google Scholar] [CrossRef] [Green Version]
  53. Abbaspour, K.C. SWAT-Cup 2012: SWAT Calibration and Uncertainty Programs—A User Manual; Eawag, Swiss Federal Institute of Aquatic Science and Technology: Duebendorf, Switzerland, 2014. [Google Scholar]
  54. Abbaspour, K.C.; Vaghefi, S.A.; Srinivasan, R. A Guideline for Successful Calibration and Uncertainty Analysis for Soil and Water Assessment: A Review of Papers from the 2016 International SWAT Conference. Water 2017, 10, 6. [Google Scholar] [CrossRef] [Green Version]
  55. Abbaspour, K.C.; Jing, Y.; Maximov, I.; Siber, R.; Srinivasan, R. Modelling of Hydrology and Water Quality in the Pre-Alpine/Alpine Thur Watershed Using SWAT. J. Hydrol. 2007, 333, 413–430. [Google Scholar] [CrossRef]
  56. Niraula, R.; Norman, L.M.; Meixner, T.; Callegary, J.B. Multi-gauge Calibration for modeling the Semi-Arid Santa Cruz Watershed in Arizona-Mexico Border Area Using SWAT. Air Soil Water Res. 2012, 5, 41–57. [Google Scholar] [CrossRef]
  57. Li, D.; Liang, X.; Wu, J. Water pollution risk simulation and prediction in a drinking water catchment. J. Zhejiang Univ. (Agric. Life Sci.) 2018, 44, 75–88. [Google Scholar] [CrossRef]
  58. Moriasi, D.N.; Gitau, M.W.; Pai, N.; Daggupati, P. Hydrologic and water quality models: Performance measures and evaluation criteria. Trans. Asabe 2015, 58, 1763–1785. [Google Scholar] [CrossRef] [Green Version]
  59. Mamo, K.H.M.; Jain, M.K. Runoff and sediment modeling using SWAT in Gumera catchment, Ethiopia. Open J. Mod. Hydrol. 2013, 2013. [Google Scholar] [CrossRef] [Green Version]
  60. Rahbeh, M.; Chanasyk, D.; Miller, J. Two-way calibration-validation of SWAT model for a small prairie watershed with short observed record. Can. Water Resour. J. Rev. Can. Des Ressour. Hydr. 2011, 36, 247–270. [Google Scholar] [CrossRef]
  61. Chuanzhun, S.; Lin, Z.; Chao, W. Biodiversity simulation of Poyang Lake wetlands by InVEST model under different scenarios. Resour. Environ. Yangtze Basin 2015, 24, 1119–1125. [Google Scholar]
  62. Xie, H.; Cheng, L.; Lv, T. Factors influencing farmer willingness to fallow winter wheat and ecological compensation standards in a groundwater funnel area in Hengshui, Hebei Province, China. Sustainability 2017, 9, 839. [Google Scholar] [CrossRef] [Green Version]
  63. Liu, J.; Li, S.; Ouyang, Z.; Tam, C.; Chen, X. Ecological and socioeconomic effects of China’s policies for ecosystem services. Proc. Natl. Acad. Sci. USA 2008, 105, 9477–9482. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  64. Wu, Y.; Liu, J.; Shen, R.; Fu, B. Mitigation of nonpoint source pollution in rural areas: From control to synergies of multi ecosystem services. Sci. Total Environ. 2017, 607, 1376–1380. [Google Scholar] [CrossRef]
  65. Omid, G.; Bakhtiar, F.; Thomas, B. An interval matrix method used to optimize the decision matrix in AHP technique for land subsidence susceptibility mapping. Environ. Earth Ences 2018, 77, 584. [Google Scholar] [CrossRef]
  66. Bennett, D.E.; Gosnell, H. Integrating multiple perspectives on payments for ecosystem services through a social–ecological systems framework. Ecol. Econ. 2015, 116, 172–181. [Google Scholar] [CrossRef]
  67. Huber, R.; Briner, S.; Peringer, A.; Lauber, S.; Seidl, R.; Widmer, A.; Gillet, F.; Buttler, A.; Le, Q.B.; Hirschi, C. Modeling social-ecological feedback effects in the implementation of payments for environmental services in pasture-woodlands. Ecol. Soc. 2013, 18, 41. [Google Scholar] [CrossRef] [Green Version]
  68. Moore, R.; Provencher, B.; Bishop, R. Valuing a Spatially Variable Environmental Resource:Reducing Non-Point-Source Pollution in Green Bay, Wisconsin. Land Econ. 2011. [Google Scholar] [CrossRef]
  69. Cao, S.; Xia, C.; Xian, J.; Guo, H.; Zheng, H. Payoff of the Grain for Green policy. J. Appl. Ecol. 2020, 57. [Google Scholar] [CrossRef]
  70. Yuan, P.; Li, D.J.; Cai, C.S.; Xu, G.J. A novel decoupling dynamic method with third-order accuracy and controllable dissipation. Comput. Struct. 2021, 249, 44–56. [Google Scholar] [CrossRef]
  71. LI, G.; SHI, H. The Payment of Grain to Green Project, the Behavior Choice of Peasants and Their Gains and Losses. China Popul. Resour. Environ. 2015, 25, 152–161. [Google Scholar] [CrossRef]
  72. Liu, D.; Hu, Z.; Jin, L. Study on compensation rate for fallow program in groundwater over-exploited area based on rural households’ willingness to accept. China Popul. Resour. Environ. 2019, 29, 130–139. [Google Scholar]
  73. Gonzalezredin, J.; Luque, S.; Poggio, L.; Smith, R.; Gimona, A. Spatial Bayesian belief networks as a planning decision tool for mapping ecosystem services trade-offs on forested landscapes. Environ. Res. 2016, 144, 15–26. [Google Scholar] [CrossRef] [PubMed]
  74. Liu, J.W.; Li, D.J.; Yu, P. Study on optimization algorithm of tuned mass damper parameters to reduce vehicle-bridge coupled vibration. PLoS ONE. 2019, 14, e0215773. [Google Scholar] [CrossRef] [Green Version]
  75. Khatami, S.; Peel, M.C.; Peterson, T.J.; Western, A. Equifinality and flux mapping: A new approach to model evaluation and process representation under uncertainty. Water Resour. Res. 2019, 55, 8922–8941. [Google Scholar] [CrossRef]
  76. Yuan, P.; Li, D.J.; Cai, C.S.; Xu, G.J. An efficient decoupling dynamic algorithm for coupled multi-spring-systems. Comput. Struct. 2018, 209, 44–56. [Google Scholar] [CrossRef]
  77. Teweldebrhan, A.T.; Burkhart, J.F.; Schuler, T.V. Parameter uncertainty analysis for an operational hydrological model using residual-based and limits of acceptability approaches. Hydrol. Earth Syst. Sci. 2018, 22, 5021–5039. [Google Scholar] [CrossRef] [Green Version]
  78. Moges, E.; Demissie, Y.; Larsen, L.; Yassin, F. Sources of Hydrological Model Uncertainties and Advances in Their Analysis. Water 2020, 13, 28. [Google Scholar] [CrossRef]
  79. Che, S.G.; Zhao, B.Q.; Ting, L.Y.; Yuan, L.; Lin, Z.; Wen, H.; Shen, B. Nutrient uptake requirements with increasing grain yield for rice in China. J. Integr. Agric. 2016, 15, 907–917. [Google Scholar] [CrossRef]
Figure 1. Research area description.
Figure 1. Research area description.
Water 13 01536 g001
Figure 2. Comparision between simulated and observed runoff.
Figure 2. Comparision between simulated and observed runoff.
Water 13 01536 g002
Figure 3. Comparison between simulated and observed result of TOT_N.
Figure 3. Comparison between simulated and observed result of TOT_N.
Water 13 01536 g003
Figure 4. Comparison between simulated and observed result of TOT_P.
Figure 4. Comparison between simulated and observed result of TOT_P.
Water 13 01536 g004
Figure 5. Hierarchy structure of comprehensive evaluation model (where the number after “|” is weight).
Figure 5. Hierarchy structure of comprehensive evaluation model (where the number after “|” is weight).
Water 13 01536 g005
Figure 6. The spatial distribution of farmland (a) and S3 target region (b) at grid scale.
Figure 6. The spatial distribution of farmland (a) and S3 target region (b) at grid scale.
Water 13 01536 g006
Figure 7. TOT_N and TOT_P of MLW from 2011 to 2018.
Figure 7. TOT_N and TOT_P of MLW from 2011 to 2018.
Water 13 01536 g007
Figure 8. Average monthly nutrient load from 2011 to 2018.
Figure 8. Average monthly nutrient load from 2011 to 2018.
Water 13 01536 g008
Figure 9. TOT_N (a) and TOT_P (b) pollution load of each scenario.
Figure 9. TOT_N (a) and TOT_P (b) pollution load of each scenario.
Water 13 01536 g009
Figure 10. Spatial distribution of TOT_N loss in MLW under different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Figure 10. Spatial distribution of TOT_N loss in MLW under different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Water 13 01536 g010
Figure 11. Spatial distribution of TOT_P loss in MLW under different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Figure 11. Spatial distribution of TOT_P loss in MLW under different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Water 13 01536 g011
Figure 12. Spatial distribution of D9 under different scenarios. S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Figure 12. Spatial distribution of D9 under different scenarios. S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Water 13 01536 g012aWater 13 01536 g012b
Figure 13. Spatial distribution of E10 under different scenarios. S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Figure 13. Spatial distribution of E10 under different scenarios. S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Water 13 01536 g013
Table 1. Main input data of SWAT model of Maoli Lake Basin.
Table 1. Main input data of SWAT model of Maoli Lake Basin.
Data DescriptionSourceResolution
Digital elevation modelComputer Network Information Center of Chinese Academy of Sciences: Geospatial Data Cloud Platform30 m
Soil typeNational Earth System Science Data Platform: Cold and Arid Regions Scientific Center1 km
Land use mapGovernment sector10 m
Weather dataNational Centers for Environmental Prediction Climate Forecast System ReanalysisDaily
Measured precipitationLixian Meteorological Bureau, Anxiang Meteorological Bureau in Hunan ProvinceDaily
Measured runoff 1Government dataDaily
Measured water quality 2MonitoringMonthly
Agricultural managementInvestigation
1 The Doppler flowmeter and no-contact ultrasonic water gauge were used to measure velocity and water level of runoff; the common bottle sampler was used for sediment sampling, and the samples were treated by drying method. 2 The water quality indicators are total nitrogen (TOT_N) and total phosphorus (TOT_P).
Table 2. Statistic of evaluation indices of the SWAT model. NSE: Nash–Sutcliffe efficiency; RVE: relative volume error.
Table 2. Statistic of evaluation indices of the SWAT model. NSE: Nash–Sutcliffe efficiency; RVE: relative volume error.
FactorPeriodNSER2RVE
RunoffCalibration period (October 2014–April 2016)0.6190.684.22
Validation period (May 2016–November 2017)0.7120.71−1.77
TOT_N 1Calibration period (January 2014–December 2015)0.450.51.51
Validation period (January 2016–December 2017)0.250.451.92
TOT_P 1Calibration period (January 2014–December 2015)0.480.586.50
Validation period (January 2016–December 2017)0.450.3−0.98
1 The number of measured data of TOT_N is 36, and so is TOT_P.
Table 3. Input values of indicator layer of different types of land use in different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest; B2: low-nitrogen degree of water; B3: low-phosphorus degree of water; B4: farmland income; B5: farmers’ participation willingness.
Table 3. Input values of indicator layer of different types of land use in different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest; B2: low-nitrogen degree of water; B3: low-phosphorus degree of water; B4: farmland income; B5: farmers’ participation willingness.
ScenarioFarmland TypeCOST 1B2B3B4B5
S0Rice land0The value of each grid is calculated by Formula (2)The value of each grid is calculated by Formula (3)10.667
S0Dry land00.9050.667
S1Rice land7150.8570.786
S1Dry land7150.810.813
S2Rice land5250.7620.895
S2Dry land5250.6671
S3Rice land5000.3450.482
S3Dry land5000.3450.478
1 The unit of COST is yuan/(mu·yr), and the “mu” is a Chinese unit of land measurement that is commonly 666.7 square meters.
Table 4. Nutrient loss of different types of farmland in MLW.
Table 4. Nutrient loss of different types of farmland in MLW.
Land TypeDataArea/km2TN/tOrganic N/tNitrate N/tTP/tOrganic P/tInorganic P/t
Rice landValue112.00672.1099.70572.4063.817.2456.57
Percentage28.87%57.82%42.28%61.78%47.41%43.43%47.97%
Dry landValue64.15436.25122.50313.7565.257.7757.49
Percentage16.53%37.53%51.95%33.86%48.48%46.57%48.75%
OrchardValue9.1534.870.1134.760.480.010.48
Percentage2.36%3.00%0.05%3.76%0.36%0.03%0.41%
ForestValue76.359.305.593.711.900.681.22
Percentage19.68%0.80%2.36%0.40%1.41%4.05%1.03%
Grass landValue5.281.281.060.210.360.140.23
Percentage1.36%0.11%0.46%0.02%0.27%0.80%0.19%
Table 5. Nutrient input and loss per unit area in different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Table 5. Nutrient input and loss per unit area in different scenarios. S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
ScenarioLand TypeNitrogen Input
(kg/ha)
Phosphate Input
(kg/ha)
Nitrogen Load
(kg/ha)
Phosphate Load
(kg/ha)
Nitrogen Loss Rate
(%)
Phosphate Loss Rate
(%)
Change Rate (%)
NitrogenPhosphateNitrogen Loss RateHosphate Loss Rate
S0Rice land509.57276.37317.6130.5862.3311.06////
Dry land700.34305.54315.0347.7744.9815.64////
Orchard293.18203.60196.572.7967.051.37////
Forest007.131.48//////
Grass land0013.563.75//////
S1Rice land254.79138.19150.8520.1159.2114.55−52.51−34.23−5.0131.54
Dry land350.17152.77139.2730.6339.7720.05−55.79−35.89−11.5828.21
Orchard146.59101.8099.421.1967.821.17−49.43−57.321.15−14.65
Forest007.131.48//00//
Grass land0013.563.75//00//
S2Rice land254.79138.19120.1110.5347.147.62−62.18−65.57−24.37−31.13
Dry land350.17152.77110.0917.2231.4411.27−65.05−63.96−30.11−27.92
Orchard293.18203.60196.572.7967.051.370000
Forest007.131.48//00//
Grass land0013.563.75//00//
S3Rice land509.57276.37328.2332.4164.4111.733.34 5.98 3.34 6.06
Dry land700.34305.54324.0349.1446.2716.082.86 2.87 2.87 2.81
Orchard293.18203.60210.412.4271.771.197.04 −13.26 7.04 −13.14
Forest006.101.11//−14.45 −25.00 //
Grass land0013.463.72//−0.74 −0.80 //
Table 6. Averages of all target grids of nodes of the SMCE model. B2: low-nitrogen degree of water; B3: low-phosphorus degree of water; C6: environmental effect of farmland; C7: economic effect of farmland; C8: social effect of farmland; D9: comprehensive effect of farmland; E10: comprehensive effect increment of per unit compensation cost of each grid; S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
Table 6. Averages of all target grids of nodes of the SMCE model. B2: low-nitrogen degree of water; B3: low-phosphorus degree of water; C6: environmental effect of farmland; C7: economic effect of farmland; C8: social effect of farmland; D9: comprehensive effect of farmland; E10: comprehensive effect increment of per unit compensation cost of each grid; S0: current situation; S1: halving fertilization every year; S2: fallow every other year; S3: returning agricultural land to forest.
ScenarioFarmland TypeAverages of All Target Grids
B2B3C6 C7 C8 D9E10
S0Rice land0.3735 0.7549 0.4822 1.0000 0.6670 0.6602 -
S0Dry land0.3796 0.6970 0.4701 0.9048 0.6670 0.6356 -
S1Rice land0.7144 0.8396 0.7501 0.8571 0.7860 0.7860 1.760 × 10−4
S1Dry land0.7257 0.8040 0.7480 0.8095 0.8130 0.7858 2.100 × 10−4
S2Rice land0.7633 0.9138 0.8062 0.7619 0.8953 0.8315 3.262 × 10−4
S2Dry land0.7720 0.8920 0.8062 0.6667 1.0000 0.8523 4.126 × 10−4
S3Rice land0.9966 0.9957 0.9964 0.3452 0.4823 0.6638 1.365 × 10−4
S3Dry land0.9952 0.9940 0.9948 0.3452 0.4780 0.6615 2.021 × 10−4
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Zheng, Y.; Lei, G.; Yu, P. Eco-Compensation Schemes for Controlling Agricultural Non-Point Source Pollution in Maoli Lake Watershed. Water 2021, 13, 1536. https://doi.org/10.3390/w13111536

AMA Style

Zheng Y, Lei G, Yu P. Eco-Compensation Schemes for Controlling Agricultural Non-Point Source Pollution in Maoli Lake Watershed. Water. 2021; 13(11):1536. https://doi.org/10.3390/w13111536

Chicago/Turabian Style

Zheng, Yumei, Guangchun Lei, and Peng Yu. 2021. "Eco-Compensation Schemes for Controlling Agricultural Non-Point Source Pollution in Maoli Lake Watershed" Water 13, no. 11: 1536. https://doi.org/10.3390/w13111536

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop