Evaluation of Land Degradation Neutrality in Inner Mongolia Combined with Ecosystem Services

Abstract

Currently, the internationally recognized land degradation neutrality (LDN) effort is evaluated using three indicators: land use/cover, land productivity, and carbon stocks. However, these three indicators may not completely capture the factors influencing LDN, and some evaluation rules are not in line with the land restoration goals of China. Therefore, this study introduces the ecosystem service value (ESV) indicator, assesses the differences in connotation and evaluation methods between ESV and LDN, and puts forward an evaluation rule that integrates their advantages, so as to carry out an evaluation of LDN in Inner Mongolia. The conclusions are as follows: (a) The ESVs of the Inner Mongolia Autonomous Region in 2000, 2005, 2010, 2015, and 2020 were USD 287.49, 286.04, 285.72, 286.38, and 287.90 billion, respectively, which presents a slight trend of decrease and then increase over time. (b) The modified LDN evaluation rule mainly includes the following changes to the LUCC evaluation rule: (1) the original degradation of cropland to grassland is considered as restoration, (2) water bodies participate in the transformation evaluation between land use types, and (3) the evaluation of transformed secondary land use types is added. The evaluation of net primary productivity (NPP) and soil organic carbon (SOC) still follow the method formulated by the United Nations Convention to Combat Desertification (UNCCD). (c) The proportion of degraded, stable, and restored land area within the LUCC were 11.31%, 77.34%, and 11.35%, respectively. The proportion of restored area is greater than the proportion of degraded land, which indicates that LDN has been achieved in Inner Mongolia according to the LUCC evaluation. The areas of degradation, stability, and restoration for NPP accounted for 0.10%, 40.52%, and 59.38% of the total area, respectively, with the restored area being much larger than the degraded area. The areas of SOC degradation, stability, and restoration accounted for 13.06%, 74.82%, and 12.11% of the total area, respectively, and the degraded area was slightly larger than the restored area. (d) The LDN evaluation results showed that the proportions of degraded, stable, and restored areas were 21.80%, 27.25%, and 50.96%, respectively. From these results, it is clear that Inner Mongolia has achieved the LDN target. Compared with the rules formulated by the UNCCD, for the LDN evaluation results implementing the modified rule, the proportion of degraded land increased by 2.44%, the proportion of stable land decreased by 1.52%, and the proportion of restored land decreased by 0.92%. In the future, Inner Mongolia should strengthen the implementation of a series of ecological restoration projects to obtain greater ecological benefits.

1. Introduction

In September 2015, the United Nations General Assembly adopted 17 sustainable development goals, among which sustainable development goal (SDG) 15.3 was to “By 2030, combat desertification, restore degraded land and soil, including land affected by desertification, drought and floods, and strive to achieve a land degradation-neutral world”. Land degradation neutrality (LDN) refers to “a state whereby the amount and quality of land resources necessary to support ecosystem functions and services and enhance food security remain stable or increase within specified temporal and spatial scales and ecosystems,” and is used as a measure of SDG 15.3 [1,2]. In October 2015, a decision of the 12th session of the Conference of the Parties (COP 12) of the United Nations Convention to Combat Desertification (UNCCD) adopted this definition of LDN. Today, the definition has been widely recognized internationally [3], and as of March 2022, 129 countries have committed to setting LDN voluntary targets, and at least 84 countries, including China, have achieved their LDN target [4]. LDN is mainly evaluated using indicators recommended by the UNCCD, including land use/cover (LUCC), land productivity, and soil organic carbon (SOC).

LDN results are obtained by integrating the three indicators using a “one out all out (1OAO)” principle, where the degradation of any indicator means the degradation of LDN. For example, Guo et al. used LUCC, net primary productivity (NPP), and SOC indicators to evaluate global LDN. Their results suggest that from 2000 to 2015, the area of restored land was generally larger than the area of degraded land, globally. Moreover, China has become the first nation to achieve their 2030 LDN target, with a net restoration area accounting for 18.24% of the total degraded land, globally [5]. Cong et al. [6] drew lessons from the “1OAO” principle of LDN and used normalized difference vegetation index (NDVI), SOC, and dust/sandstorm (DSS) indicators to evaluate the ecological quality of Hunshandak Sandy Land. Their results showed that the ecological quality of the region improved by one fifth. Other studies [7] have investigated the economics of LDN in Ethiopia, and the results show that there was an increasing trend in agricultural land degradation in Ethiopia during the study period of 2003–2016, and the annual aggregate crop production loss amounted to 104 million tons, with a market value of USD 48.35 billion. For the Karst region of China, Zhang et al. [8] combined LDN and ecosystem services to analyze the effects of ecosystem restoration projects and found that the rate of soil loss has decreased significantly since 1999. In contrast, Chappell et al. [9] showed that wind erosion had accelerated arid-zone SOC loss between 2001 and 2016.

However, current methods for evaluating LDN are not always appropriate for their area of application and need to be reviewed. LUCC is evaluated through the conversion of land use types, which may not reflect land restoration goals. For example, the conversion of cropland to grassland is defined as degradation, which is in disagreement with the efforts of the Grain for Green project (the main goal of which is to return cropland to grassland and forest land) in China. IPCC (Tier 1 method) and other studies [10] assume that the conversion of cropland to grassland will increase SOC stocks (land restoration), which is inconsistent with LUCC’s assessment rules, and as such, the LDN results obtained through the 1OAO principle may not be accurate. Additionally, conducting LUCC assessments without considering water bodies may be detrimental to sustainable land management. Water bodies are ecosystems embedded in the land, and like other land use types, they may experience changes in both area and quality. Water is also essential for human survival and is crucial for the development of sustainable land management. Water supports unique animal and plant communities and is also associated with great ecosystem service value (ESV). Moreover, the role of water resources in arid and semi-arid areas is even more prominent. Given water’s importance to land management, the evaluation of LUCC should also assess changes in water bodies. The current LUCC assessment is limited to seven land types, which urgently need to be revised. Finally, conversions of forest to wetland and other land to wetland are also considered restoration. However, the UNCCD does not provide clear ways to determine differences in degrees of restoration, and therefore, some modification may benefit the LUCC evaluation rule formulated by the UNCCD as it is applied to other regions.

Ecosystem services refer to the direct and indirect benefits of ecosystem functions to humans. Land degradation is associated with a reduction in ecosystem services, and in general, the supply of ecosystem services depends on the quality of land [11,12]. Ecosystem services can be effectively captured through the evaluation of ESV. Costanza et al. [13] estimated the value of global ecosystem services using “unit area values for ecosystem services”. Similarly, Xie et al., 2015 obtained an “ESV equivalent coefficients table” in China [14]. An evaluation of ESV from 2000 to 2020 in the Inner Mongolia Autonomous Region has not previously been conducted. This paper aims to compare the connotation and evaluation methods of LDN and ESV and present a set of integrated evaluation methods that are suitable for the region. Such methodological integration will allow for a more realistic LDN assessment while incorporating ecological restoration goals.

Inner Mongolia straddles the three northern regions of China and is composed of various ecosystems. From east to west, these ecosystems transition from forests to grasslands to deserts. The region serves as an important ecological barrier in northern China and is home to a number of ecological projects such as the Three North Shelterbelt project and the Grain for Green project. Moreover, Inner Mongolia comprises both arid and semi-arid areas and has historically experienced frequent land degradation, affecting people’s wellbeing, living environment, and quality of life [15]. For these reasons, it is necessary to evaluate the LDN in Inner Mongolia.

2. Materials and Methods

2.1. Data Sources

Gridded land use data from 2000, 2005, 2010, 2015, and 2020 used to estimate ESV and assess LDN were sourced from the Resource and Environment Science and Data Center (RESDC) of the Chinese Academy of Sciences (CAS) (resolution: 1000 m). A vector map of Inner Mongolia Autonomous Region was downloaded from the RESDC of the CAS. The areas of grain crops for estimating ESV were collected from the statistical yearbook of Inner Mongolia Autonomous Region for 2021, and the net profits of grain were collected from the “Compilation of national cost-benefit data of agricultural products” for 2001, 2006, 2011, 2016, and 2021.

Remotely sensed MOD17A3H data were downloaded from the USGS official website, https://e4ftl01.cr.usgs.gov/MOLT/MOD17A3HGF.006/ (accessed on 11 October 2019), and processed using “MODIS reprojection tool (version 4.1)” software, resulting in NPP data at 500 m resolution from 2000 to 2020.

Soil organic carbon (SOC) layers were generated using the TREND-EARTH plugin of QGIS 3.12, as recommended by the UNCCD. Gridded land use data for 2000 and 2020 were downloaded from the RESDC of the CAS (with a resolution of 1000 m).

2.2. Study Method

2.2.1. Estimation of ESV

The land use data in Inner Mongolia were divided into 6 primary land use classifications and 22 secondary land use classifications [16]. Two of the secondary land use classifications, “bottomland” and “wetland”, are regarded as “wetlands” by UNCCD, and the other 6 primary land use classifications correspond to the remaining 6 primary land use classifications of UNCCD. The ESVs for 2000, 2005, 2010, 2015, and 2020 were estimated using the “equivalence factor per unit area” method. Here, ESV is calculated as the product of a constant (D) and an equivalence coefficient table [14], where in this case, D is the average of Inner Mongolia’s grain net profit in 2000, 2005, 2010, 2015, and 2020 and is calculated as 268.6 USD/ha. The method of assigning an ESV to each land use was adopted from the literature [17]. The change trend of ESV was evaluated using a slope formula, as shown below [18].

$${\theta }_{ESV}=\frac{(\mathrm{n}\times {\displaystyle \sum _{i=1}^{n}i{\times \mathrm{ESV}}_{\mathrm{i}}}-{\displaystyle \sum _{i=1}^{n}i\times {\displaystyle \sum _{i=1}^n{\mathrm{ESV}}_{\mathrm{i}}}})}{n\times {\displaystyle \sum _{i=1}^n{i}^2-{({\displaystyle \sum _{i=1}^{n}i})}^2}}$$

(1)

Here, i is the year serial number, n is the total number of years (5 years), ESV_i is the ESV of the ith year, and θ_ESV is the slope of ESV change trend. Negative θ_ESV values indicate that ESV is decreasing, while positive values indicate an increase in ESV.

2.2.2. Comparison of ESV and LDN

In this study, we aim to compare the connotation and evaluation methods of ESV and LDN, analyze their advantages and disadvantages, integrate the advantages of both methods, and develop an evaluation method more suitable for Inner Mongolia.

2.2.3. Evaluation of LDN in Inner Mongolia

• Evaluation of LUCC

Using the “union” tool in ArcGIS 10.6, with the land use vector data from 2020 and 2000 as inputs, a land use transfer matrix from 2000 to 2020 was obtained. The land use type conversion results were defined according to the ESV per unit area of different land use types, and the evaluation results of LUCC were obtained by measuring the degraded, stable, and restored areas.

2.

Evaluation of NPP

According to the methods formulated by UNCCD [19], the change trends of NPP from 2000 to 2020 were calculated using a slope formula, as shown below.

$${\theta }_{NPP}=\frac{(\mathrm{n}\times {\displaystyle \sum _{i=1}^{n}i{\times \mathrm{NPP}}_{\mathrm{i}}}-{\displaystyle \sum _{i=1}^{n}i\times {\displaystyle \sum _{i=1}^n{\mathrm{NPP}}_{\mathrm{i}}}})}{n\times {\displaystyle \sum _{i=1}^n{i}^2-{({\displaystyle \sum _{i=1}^{n}i})}^2}}$$

(2)

Here, i is the serial number of the year, n is the total number of years (21 years), NPP_i is the NPP of the ith year, and θ_NPP is the slope of NPP change trend. A T-test was used to test for significant trends, and the test results were divided into significant reduction (θ_NPP < 0, p < 0.05), significant increase (θ_NPP > 0, p < 0.05), and no significant change (p > 0.05), corresponding to degradation, restoration, and stability of NPP, respectively.

3.

Evaluation of SOC

SOC stock layers for 2000 and 2020 were generated by the TREND-EARTH plugin of QGIS 3.12 using the “land use conversion coefficient” [19] method (

Table 1. Land use conversion coefficients for evaluating SOC.

2020	Forest	Grasslands	Croplands	Wetlands	Artificial Areas	Other Lands
2000
forest	1	1	f	1	0.1	0.1
grasslands	1	1	f	1	0.1	0.1
croplands	1/f	1/f	1	1/0.71	0.1	0.1
wetlands	1	1	0.71	1	0.1	0.1
artificial areas	2	2	2	2	1	1
other lands	2	2	2	2	1	1

Note: red—degradation, yellow—stability, green—restoration

). The f value of the temperate humid area in eastern Inner Mongolia is 0.69, which is mainly concentrated in some areas of Hulunbuir City, and the f value of other parts in Inner Mongolia is 0.8. Climate boundary data provided by QGIS 3.12 were subsequently used for the scope of climate area.

Specifically, the 250 m resolution SOC grid layer was downloaded using the “download raw data” option in the TREND-EARTH plugin of QGIS 3.12, where the downloaded SOC grid data and LUCC data for 2020 were used as inputs in the “use customized data-SOC” option to obtain the 2 years of SOC stock data.

In ArcGIS 10.6, the “Raster Calculator” tool was used to compare the difference between the SOC grid layer data from 2020 and 2000. Here, if the change in stocks at the pixel level exceeded 10%, it was recognized as either SOC degradation or restoration, while other areas were considered stable.

4.

Evaluation of LDN

At the pixel level, if at least one of the 3 indicators was degraded, the LDN result was considered degraded; if all 3 indicators were stable, the result was considered stable; and if no indicator was degraded and at least one indicator was restored, the LDN result was restored (Figure 1) [19].

3. Results

3.1. ESV in Inner Mongolia

3.1.1. Changes in ESV

From 2000 to 2020, the minimum ESV of the Inner Mongolia Autonomous Region was USD 285.72 billion (in 2010), and the maximum was USD 287.90 billion (in 2020), showing a slight trend of decrease and then increase (

Table 2. ESV of different land use types in Inner Mongolia over time.

Land Use Type		ESV (USD/ha)	Total ESV (Billion USD)
			2000	2005	2010	2015	2020
forest	forested land	5692.1	69.99	70.18	70.14	70.30	70.24
	open forest	5692.1	10.87	11.03	11.02	11.05	11.07
	other woodland	5692.1	1.12	1.44	1.44	1.44	1.44
	shrubland	4088.4	8.28	8.41	8.40	8.35	8.42
	subtotal		90.26	91.06	91	91.14	91.17
grasslands	low-coverage grassland	1361.9	14.28	14.24	14.33	14.44	14.43
	mid-coverage grassland	1361.9	25.39	25.07	25.10	25.02	24.44
	high-coverage grassland	3070.4	73.07	72.80	72.69	72.42	73.79
	subtotal		112.74	112.11	112.12	111.88	112.66
croplands	rainfed cropland	1077.2	11.89	11.88	11.91	11.89	11.88
	irrigated cropland	1044.9	0.10	0.12	0.12	0.13	0.13
	subtotal		12.00	12.00	12.04	12.02	12.02
wetlands	bottomland	13,973.8	6.96	6.33	6.67	6.64	6.60
	wetland	13,973.8	28.23	28.44	28.37	27.79	27.75
	subtotal		35.20	34.77	35.04	34.43	34.35
water bodies	river	33,741.8	6.13	6.91	6.89	7.37	7.78
	lake	33,741.8	21.85	19.81	18.92	19.46	19.79
	reservoirs and pits	33,741.8	2.02	2.02	2.38	2.78	2.84
	subtotal		30.00	28.74	28.19	29.62	30.41
other lands	Gobi Desert	295.5	2.09	2.09	2.09	2.09	2.09
	bare land	53.7	0.02	0.02	0.02	0.02	0.02
	bare-rock stony ground	53.7	0.24	0.24	0.24	0.24	0.24
	desert	295.5	4.25	4.31	4.29	4.25	4.25
	saline–alkali soil	295.5	0.70	0.70	0.70	0.67	0.67
	subtotal		7.30	7.37	7.34	7.28	7.27
artificial areas	urban land	0	0.00	0.00	0.00	0.00	0.00
	rural residential area	0	0.00	0.00	0.00	0.00	0.00
	other construction lands	0	0.00	0.00	0.00	0.00	0.00
	subtotal		0.00	0.00	0.00	0.00	0.00
total			287.49	286.04	285.72	286.38	287.90

Note: the canopy closure of “forested land” and “open forest” are greater than 0.3 and between 0.1 and 0.3, respectively; “other woodland” refers to nurseries, various gardens, etc.; and “shrubland” refers to low woodland and shrub woodland with a canopy density ≥40% and height below 2 m.

). The average annual ESVs of different land use types, ordered from highest to lowest, were grasslands (USD 112.30 billion), forests (90.93 billion), wetlands (34.76 billion), water bodies (29.39 billion), croplands (12.01 billion), other lands (7.31 billion), and artificial areas (0). According to the ESV slope results over time, the ESV of grasslands, wetlands, and other lands followed a decreasing trend, while the ESV of forests, croplands, and water bodies displayed an increasing trend.

3.1.2. Evaluation of ESV

From 2000 to 2020, the areas of increased and decreased ESV in the Inner Mongolia Autonomous Region were 661,699 km² and 484,226 km², respectively, accounting for 57.74% and 42.26% of the total area (Figure 2).

3.1.3. Proposal of Integrated LUCC Evaluation Rule

The three common indicators of LDN used here aim to reflect ecosystem functions and services from different angles [1,19], but they cannot reflect the ecosystem services overall. However, ESV can comprehensively evaluate ecosystem services, and can be quantified in the form of money. In this way, the evaluation scope of ESV reflects the evaluation scope of these three indicators.

The “equivalence factor per unit area” method is based on different land use types and assumes that the ESV per unit area of each land use type is consistent. LDN comprehensively evaluates three indicators, and the minimum evaluation unit is the pixel level, which considers more indicators and is more accurate than the evaluation scale of ESV.

The LUCC evaluation criteria formulated by the UNCCD are determined through the conversion of land use types, which are less appropriate given China’s national land use goals. For example, the interpretation of conversions of cropland to grassland conflict with the Grain for Green project in China. The theoretical basis of ESV evaluation in this study uses the differences in the ESV of various land use types, which provides a quantitative basis for the modification of LUCC evaluation criteria. The UNCCD recommends considering 7 land use types to evaluate LUCC, but with our modification, we increased our scope by considering 22 secondary land use classifications, which provides a more comprehensive and detailed evaluation of LUCC (Figure 3).

In changing this method, the transformation relationship between the seven land use types was also modified. The conversion of cropland to grassland changed from degradation (as defined by UNCCD) to restoration, and water bodies were also considered in the evaluation of LUCC (

Table 3. Comparison of modified and UNCCD LUCC evaluation rules.

2020	Forest	Grasslands	Croplands	Wetlands	Artificial Areas	Other Lands	Water Bodies
2000
forest	0	−1	−1	−1	−1	−1	1(0)
grasslands	1	0	−1(1)	−1	−1	−1	1(0)
croplands	1	1(−1)	0	−1	−1	−1	1(0)
wetlands	−1	−1	−1	0	−1	−1	1(0)
artificial areas	1	1	1	1	0	1	1(0)
other lands	1	1	1	1	−1	0	1(0)
water bodies	−1(0)	−1(0)	−1(0)	−1(0)	−1(0)	−1(0)	0

Note: values of 1, 0, and −1 represent restoration, stability, and degradation, respectively. The numbers without brackets represent the modified evaluation method, while the evaluation method of the UNCCD is denoted with brackets. For example, a table entry of 1(−1) indicates that UNCCD defines the transformation as degradation, while the modified change rule defines this transformation as restoration.

, Figure 3).

3.2. Evaluation of LDN in Inner Mongolia

According to the modified LUCC evaluation rule in this study, from 2000 to 2020, the area of degraded land in Inner Mongolia was 128,958 km², the area of stable land was 881,389 km², and the area of restored land was 129,418 km², accounting for 11.31%, 77.34%, and 11.35% of the total study area, respectively. Based on these results, it is clear that Inner Mongolia achieved its LDN target during this period.

From 2000 to 2020, the areas where NPP decreased and increased significantly accounted for 0.10% and 59.38% of the entire region, respectively, while the area where NPP remained stable accounted for 40.52% of the total area. Therefore, in terms of NPP, Inner Mongolia has achieved LDN, and the area of restored land is significantly larger than that of degraded land.

From 2000 to 2020, the area of degraded, stable, and restored SOC in Inner Mongolia accounted for 13.06%, 74.82%, and 12.11% of the total land area, respectively. In terms of SOC, the Inner Mongolia Autonomous Region failed to achieve the LDN target. The SOC stocks of Inner Mongolia in 2000 and 2020 were 9.03 Tg and 8.98 Tg, respectively. Compared with 2000, the SOC stocks in 2020 decreased by 49.06 million tons, with a reduction ratio of 0.54%.

Based on the above research results for LUCC, NPP, and SOC, the LDN evaluation of Inner Mongolia from 2000 to 2020 could be obtained. The evaluation showed that the area of degraded land in Inner Mongolia during the study period was 247,006 km², accounting for 21.80% of the total area of the region; the area of stable land was 308,739 km², accounting for 27.25% of the total area of the region; and the area of restored land was 577,389 km², accounting for 50.96% of the total area (Figure 4).

To evaluate LDN, the “union” of the degradation of the three indicators was considered, where the degradation of all three was required to determine the failure of LDN. In this way, a degradation LDN result is more inclusive than a degradation result for each contributing indicator. Areas of stable LDN are considered the “intersection” of the stable areas of the three indicators. Here, once an area experiences degradation or restoration, the stable state transitions to other states, and for our study area, the minimum proportion of stable LDN was only 27.25%. The restoration proportion of LUCC and SOC was relatively small, while the restoration proportion of NPP was relatively large, and the LDN results indicated a large proportion of land restoration, which highlights the impact of NPP on the evaluation of LDN. Finally, the proportion of stability assessed by LUCC and SOC was more than 70%, while the proportions of restoration and degradation were less than 30%, illustrating a “one-pole stability” state. The proportion of NPP degradation was less than 1%, while the proportion of restoration and stability accounted for more than 40%, showing that restoration and stability account for the main “two poles” (Figure 5). The areas of failed and successful LDN accounted for 8.25% and 91.75% of the whole region, respectively (Figure 5).

3.3. Difference Analysis of Evaluation Results

According to the LUCC evaluation rule formulated by the UNCCD, the degraded, stable, and restored areas were 94,145 km², 953,058 km², and 92,562 km², respectively, accounting for 8.26%, 83.62%, and 8.12% of the total region. Using the modified rule, the areas of degradation, stability, and restoration increased by 35,273 km², decreased by 71,669 km², and increased by 36,396 km², respectively, and their area proportions increased by 3.09%, decreased by 6.28%, and increased by 3.19%, respectively (

Table 4. Comparison of LUCC and LDN evaluation results before and after modification.

		LUCC			LDN
		Degradation	Stability	Restoration	Degradation	Stability	Restoration
area (km2)	modified rule	129,418	881,389	128,958	247,006	308,739	577,389
	rule formulated by the UNCCD	94,145	953,058	92,562	219,407	325,948	587,779
area proportion (%)	modified rule	11.35	77.34	11.31	21.8	27.25	50.96
	rule formulated by the UNCCD	8.26	83.62	8.12	19.36	28.77	51.87
differences	area (km2)	35,273	−71,669	36,396	27,599	−17,209	−10,390
	area proportion (%)	3.09	−6.28	3.19	2.44	−1.52	−0.92

).

These differences in LUCC evaluation results may be partly attributed to the conversion of cropland to grassland. The conversion of cropland to grassland is defined as restoration in the modified rule, which is more in line with the national restoration goals of China. The differences may also be partly attributed to transformations between secondary land use classifications. The UNCCD divides land into wetland, forest, grassland, cropland, other lands, artificial areas, and water bodies. Grassland in the LUCC data used in this study included high-coverage grassland, mid-coverage grassland, low-coverage grassland, and shrubland; cropland included rainfed cropland and irrigated cropland; and other lands included bare land and the Gobi Desert. This transformation between secondary classifications was also evaluated. Finally, the differences may be partly attributed to the conversion of water bodies. Water bodies are not considered in the UNCCD framework. However, due to the irreplaceable role of water bodies in land management and human wellbeing, especially the role of water bodies in the arid and semi-arid areas of Inner Mongolia, the modified rule also considered water bodies for evaluation.

Compared with the rule formulated by the UNCCD, the evaluation of LDN using the modified rule increased the area of degraded land by 27,599 km² and reduced the area of stable and restored land by 17,209 km² and 10,390 km², respectively. Proportionally, degraded land increased by 2.44%, stable land decreased by 1.52%, and restored land decreased by 0.92%. The difference in the evaluation results was first examined using the modified LUCC rule, followed by the difference in the LDN results due to the “one out all out” principle when the LUCC evaluation results were superimposed onto the NPP and SOC evaluation results (Table 4).

3.4. LDN and ESV

Through the comparative analysis of the LDN and ESV evaluation methods, 41 counties achieved LDN and experienced decreased ESV, covering an area of 480,890 km² and accounting for 41.97% of the whole region. However, 50 counties achieved LDN and experienced increased ESV, with an area of 568,510 km², accounting for 49.61% of the whole region. There were seven counties that did not achieve the LDN target and displayed decreased ESV, with an area of 3336 km², accounting for 0.29% of the whole region. Finally, four counties did not achieve LDN and exhibited increased ESV, covering an area of 93,189 km², accounting for 8.13% of the whole region (Figure 6).

4. Discussion

4.1. Evaluation Rule

Studies have shown that the process of urbanization [20] and land use change [21] lead to declines in cropland productivity. In China, the Grain for Green project is a measure aimed at restoring land ecosystem services and productivity. As it stands, the project has demonstrated great achievements, with the total value of its ecological benefits reaching CNY 1.42 trillion, its economic benefits reaching 0.26 trillion, and its social benefits reaching 0.73 trillion, for a total of 2.41 trillion [22]. In 2000, the Inner Mongolia Autonomous Region implemented the Grain for Green project on a large scale [23], and as the region is also home to one of the largest grassland areas in China, it is very important to accurately determine the value conversion factor between cropland and grassland ecosystems. The rule formulated by the UNCCD is based on the perspective of food security [1]. According to the report of the World Food and Agriculture Organization, about 10% of the global population (1.18 billion people) will face a food shortage crisis in 2020. Therefore, the rule is reasonable and scientifically based, and is especially applicable to areas where food security is threatened, such as Africa and the arid areas of Central Asia. However, the Inner Mongolia Autonomous Region has implemented the policy of Grain for Green for many years, and the sufficient supply of grain provides increased food security to the region. The LUCC evaluation rule formulated by the UNCCD defines the conversion of cropland to grassland as degradation, which is in contradiction to the aim and benefits of the Grain for Green project. The tier 1 method of IPCC (although a rough method) for SOC evaluation also holds that converting cropland to grassland will increase SOC stocks and restore land. Therefore, to avoid the conflict with the benefits of the Grain for Green project and SOC evaluation results, the LUCC evaluation rules should be adjusted.

“The 14th Five-Year Plan for water security in Inner Mongolia Autonomous Region” emphasizes the need for national economic and social development to “Plan development according to the amount of water” available. Similarly, land greening should follow the principle of “Water Determines Greening” [24] to avoid interventions that squander water resources and are not sustainable. Therefore, compared with the evaluation rules of UNCCD, analyzing changes in water bodies is helpful to achieve high-quality human development and sustainable land management in Inner Mongolia. The future LDN should also take water body changes into account.

This study modifies the evaluation rule of LUCC in the Inner Mongolia Autonomous Region Based on the ESV value per unit area. In the evaluation of ESV, the ESV of grassland per unit area is higher than that of cropland. Considering the ecosystem services of both land cover types, the conversion of grassland to cropland should be considered degradation. The ESV per unit area of water bodies was the largest, so the transformation of water bodies into other land use types was defined as degradation. After modifying the rule, the difference in the degree of land restoration or degradation could be determined by measuring the change in ESV. For example, the ESV of 1 hectare of forest converted into wetland increased by USD 8284.3, and the ESV of 1 hectare of other land (desert) converted to wetland increased by USD 13,678.3. Similarly, the ESV restoration coefficient of other land is 1.65 times that of forest restoration.

The conversion rule of land use types should be inclusive and open, and appropriate adjustments can be made to accommodate different regions and national conditions. The modified rule in this study is more suitable for areas where ecological benefits are the main priority. For areas dominated by production ecosystem functions, the model should be adjusted, such as areas with high grain yield or areas where food security is at risk, and more attention should be paid to cropland. Moreover, the evaluation rules in different regions of Inner Mongolia vary. In the future, the characteristics and differences of various regions of Inner Mongolia should be analyzed to formulate more detailed evaluation rules.

4.2. LUCC, NPP, and SOC

Mu et al. [25] found that the net increase in grassland area in Inner Mongolia from 2001 to 2009 was 77,993 km², mainly due to the conversion of desert and farmland. Li et al. [26] identified degraded land by evaluating land-cover change in Inner Mongolia in 2004 and 2014, and their results indicated that the urban area has expanded five-fold, and 35.3% of the farmland has been converted into grassland, which may have been caused by the “Grain to Green” policy. In this study, we found that 20.32% of the cropland in Inner Mongolia has been converted to grassland, while only 4.65% of the grassland has been turned into cropland, and the proportion of new artificial land has reached 38.66%. Xu et al. [27] found that the water storage of tectonic lakes continues to decline, but the water storage of oxbow lakes has not changed significantly in Inner Mongolia over the past 30 years. The area of water bodies in this study increased by 0.91%, which may be attributed to the decrease in irrigation water for cropland and the increase in precipitation. The increased forest area in Inner Mongolia accounts for 0.99% of the forest area, which might be attributed to ecological restoration projects such as the Three North Shelterbelt and Grain for Green projects. Referring to the statistical yearbook of the Inner Mongolia Autonomous Region, the average annual afforestation area from 2000 to 2020 was 6871 km², within which the average annual afforestation areas from 2000 to 2005, 2006 to 2010, 2011 to 2015, and 2016 to 2020 were 7392 km², 6552 km², 7086 km², and 6474 km², respectively. This could also explain the substantial increase in forest ESV in the period 2000–2005. Overall, ecological restoration projects have played a great role in land restoration, and it is necessary to strengthen the restoration and protection of natural ecosystems and continue to implement ecological restoration projects in the future.

The restoration area of NPP was larger than the degraded area, which could be attributed to both natural and human factors. According to other studies, the increase in NPP in Inner Mongolia is mainly driven by precipitation [28]. The eastern cities are affected by the Pacific monsoon, and the precipitation is higher than that in the western cities, so the NPP growth is relatively significant. In this study, the proportions of unchanged and changed land use types in NPP degraded areas were 45.04% and 54.96%, respectively, and in the area where NPP was restored, the proportions were 82.12% and 17.88%, which indicates that the degradation of NPP can be attributed to the joint action of natural and man-made factors, and the restoration of NPP is mainly due to natural factors. Wang et al. [29] studied the spatial pattern and changes in NDVI in Inner Mongolia from 1982 to 2020 and found that the surface vegetation coverage in Inner Mongolia exhibited an upward trend. The research of Wu et al. [30] on grasslands in Inner Mongolia suggests that the climate is always the most important factor in determining biomass production in all grassland types compared with soil and plant factors. However, Mu et al. [25] found that the increased total NPP of Inner Mongolia grasslands from 2001 to 2009 was mainly due to human activities (80.23%), with the restoration of large-scale ecological protection projects likely being the internal driving force of this phenomenon. In conclusion, human activities play an important role in promoting land productivity, and ecosystem restoration projects need to continue.

Presently, there is no mature and stable method for estimating SOC stocks. The evaluation of SOC change in this study used the method developed by the IPCC [1]. However, this method does not consider factors such as land use and land management measures, likely due to the limitation of only considering LUCC change data. In order to obtain stable and accurate SOC stock monitoring data, a scientific layout of monitoring sample points and periodic data acquisition are relatively more reliable methods [1,2].

4.3. LDN and ESV

At present, many studies have focused on the value of monitoring ecosystem services or natural capital globally, but it is difficult to compare the results of different studies. On the one hand, there is a lack of unified methods, while on the other hand, the statistical scope and indicators used by researchers are inconsistent [31]. The “equivalence factor per unit area” method used in this study is relatively simple and effective. Here, the evaluation of LDN is based on pixel-level remote sensing data. However, although LDN and ESV both reflect changes in ecosystem functions and services, their results are not completely consistent. The correlation analysis of the ESV and LDN evaluations indicated that there was no significant negative correlation (r = −0.1718), which suggests that the two indicators represent information relating to different aspects of the ecosystem. Therefore, it is meaningful to comprehensively evaluate LDN and ESV as two separate indicators.

From 2000 to 2020, the ESV in Inner Mongolia changed, presenting a slight pattern of decrease and then increase but showing an overall increasing trend. Land use changes were caused by ecological restoration projects, urbanization, and cropland reclamation, while climate change may have also played an important role. The decrease in ESV in Inner Mongolia from 2000 to 2010 may be attributed to the fact that the area of degraded land was larger than the area of restored land, and the increase in ESV from 2010 to 2020 may be due to the larger area of land restoration. The increase in the ESV of forests and croplands and the decrease in the ESV of grasslands, other lands, and wetlands may have been more dependent on human actions, while the increase in the ESV of water bodies could be due to precipitation. Inner Mongolia has a large east–west span, and therefore the ESV of the same land use type in different regions should also be assessed and distinguished in more detail, so as to make the assessment method of LUCC more accurate.

5. Conclusions

The LDN evaluation rules formulated by UNCCD are not always applicable to different regions. This study adjusted the LUCC indicator and ESV evaluation rules to make them consistent with the actual situation in Inner Mongolia and the LDN and ecological restoration goals of China. The conclusions are as follows: (a) The ESVs of the Inner Mongolia Autonomous Region in 2000, 2005, 2010, 2015, and 2020 were USD 287.49, 286.06, 285.72, 286.38, and 287.90 billion, respectively, which represents a slight pattern of decrease and then increase. (b) The modified LDN evaluation rule mainly includes the following changes to the LUCC evaluation rule: (1) the original degradation of cropland to grassland is considered restoration, (2) water bodies are included in the transformation evaluation between land use types, and (3) the evaluation of transformed secondary land use types is added. The evaluation of net primary productivity (NPP) and soil organic carbon (SOC) still follow the method formulated by the United Nations Convention to Combat Desertification (UNCCD). (c) The proportions of degraded, stable, and restored land area for LUCC, NPP, and SOC were 11.31%, 77.34%, and 11.35%; 0.10%, 40.52%, and 59.38%; and 13.06%, 74.82%, and 12.11%, respectively. Only the SOC indicator failed to achieve the land restoration goal. (d) The LDN evaluation results showed that the proportion of degraded, stable, and restored areas were 21.80%, 27.25%, and 50.96%, respectively. From these results, it is clear that Inner Mongolia has achieved the LDN target. Compared with the rules formulated by the UNCCD, the LDN evaluation results using the modified rules showed that the proportion of degraded land increased by 2.44%, the proportion of stable land decreased by 1.52%, and the proportion of restored land decreased by 0.92%. In the future, Inner Mongolia should strengthen the implementation of a series of ecological restoration projects to obtain greater ecological benefits, and assessments of ESV should be specifically designed to be accurate in different regions.