**3. Results**

#### *3.1. Land-Use Change Matrix*

From 2010 to 2020, the increased area of cultivated land in the four towns mainly resulted from shrub-grassland and forestland, while the decreased area of cultivated land was mainly converted to shrub-grassland, forestland, and built-up land. The transition area between cultivated land, forestland, and shrub-grassland in the four towns was large, whereas the transition area in water bodies and unused land was small. The increased area of built-up land and roads mainly resulted from cultivated land (Table 2).

#### *3.2. Changes in the Non-Agriculturalization of Cultivated Land*

Based on Equations (3)–(5), the rate of conversion from cultivated land to noncultivated land was obtained, as reported in Table 3. The total rate of conversion from cultivated land to non-cultivated land in Xianchang with a karst trough valley landform was found to be higher than that in Minxiao with a karst low hilly landform. The total rates of conversion from cultivated land to non-cultivated land in Longchang with a karst mid-mountain landform and Liuguan with a karst basin landform were found to be lower than those in the other two towns. The rate of conversion from cultivated land to ecological land, in descending order, was found to be that of Xianchang, Minxiao, Longchang, and Liuguan. The rates of conversion of cultivated land to living land in Minxiao, Xianchang, and Liuguan were found to be higher than that in Longchang (Table 3).


**Table 2.** The land-use change matrix from 2010 to 2020 (hm2).

*Land* **2022**

, *11*, 1727


**Table 3.** The overall characteristics of the rate of conversion from cultivated land to non-cultivated land from 2010 to 2020 (%).

Note: Ecological land includes forestland, shrub-grassland, water bodies, and unused land. Living land includes built-up land and roads. Non-cultivated land includes ecological land and living land.

With the increase of the slope gradient (from gradient I to V), the rate of conversion of cultivated land to non-cultivated land was found to increase in Longchang with a karst midmountain landform and Liuguan with a karst basin landform but decreased in Xianchang with a karst trough valley landform and Minxiao with a karst low hilly landform. The rate of conversion of cultivated land to living land in the four towns with different landforms was found to decrease from slope gradient I to V. The change characteristics of the rate of conversion from cultivated land to ecological land in Longchang and Liuguan were found to be contrary to those in Xianchang and Minxiao from gradient I to V (Table 4).

**Table 4.** The slope gradient characteristics of the non-agriculturalization of cultivated land from 2010 to 2020 (%).



**Table 4.** *Cont.*

Note: Ecological land includes forestland, shrub-grassland, water bodies, and unused land. Living land includes built-up land and roads. Non-cultivated land includes ecological land and living land. The slope is divided into five gradients (from gradient I to V) according to the slope value from lowest to highest.

The rate of conversion from cultivated land to forestland in Xianchang with a karst trough valley landform and Minxiao with a karst low hilly landform was found to be higher than those in Liuguan with a karst basin landform and Longchang with a karst mid-mountain landform. The rates of conversion from cultivated land to shrub-grassland in Minxiao and Liuguan were found to be lower than those in Longchang and Xianchang. The rates of conversion from cultivated land to water bodies and unused land in the four towns were found to be low. Moreover, the rates of conversion from cultivated land to built-up land and roads in Minxiao, Xianchang, and Liuguan were found to be higher than those in Longchang (Table 5).

**Table 5.** The rate of conversion from cultivated land to each non-cultivated land type from 2010 to 2020 (%).


The high-value area of the non-agriculturalization rate (>20%) in Longchang is concentrated in the western part, while the low-value area (<10%) is concentrated in the eastern and southern parts (Figure 3a). The high-value area of the non-agriculturalization rate (>20%) in Liuguan is widely distributed and mainly located in the western, northern, and central parts, while the areas with a non-agriculturalization rate of less than 20% are scattered (Figure 3b). The high-value area of the non-agriculturalization rate (>20%) in Xianchang is mainly distributed in the central part, and the non-agriculturalization rate in most of the eastern and western regions is less than 20% (Figure 3c). In most areas of Minxiao, the non-agriculturalization rate is less than 20%, and these regions are mainly located in the western, central, and southern areas (Figure 3d).

**Figure 3.** The spatial patterns of the non-agriculturalization of cultivated land from 2010 to 2020. (**a**) Longchang, (**b**) Liuguan, (**c**) Xianchang, (**d**) Minxiao.

#### *3.3. Changes of Landscape Ecological Risk*

Overall, the amounts of change of the landscape ecological risk in Longchang with a karst mid-mountain landform and Minxiao with a karst low hilly landform decreased from 2010 to 2020, while those in Liuguan with a karst basin landform and Xianchang with a karst trough valley landform increased. With the increase of the slope gradient, the amount of change of the landscape ecological risk index in the four towns was found to gradually decrease (Table 6).


**Table 6.** The changes of landscape ecological risk from 2010 to 2020.

Note: The slope is divided into five gradients (from gradient I to V) according to the slope value from lowest to highest.

Except for the southern part, most parts of Longchang experienced a decrease in landscape ecological risk, and the central part exhibited a significant decrease (Figure 4a). The landscape ecological risk increased in most parts of Liuguan, and the amount of increase of the landscape ecological risk in the southeastern part was higher than that in other parts (Figure 4b). The landscape ecological risk increased in the central part of Xianchang but decreased in the eastern and western parts (Figure 4c). The landscape ecological risk in Minxiao decreased in the western and eastern parts but increased in the southeastern and northern parts (Figure 4d).

**Figure 4.** The spatial patterns of changes in landscape ecological risk from 2010 to 2020. (**a**) Longchang, (**b**) Liuguan, (**c**) Xianchang, (**d**) Minxiao.

#### *3.4. Correlation between the Non-Agriculturalization of Cultivated Land and Landscape Ecological Risk*

A negative correlation was found between the non-agriculturalization of cultivated land and landscape ecological risk in the four towns. The degrees of correlation between the non-agriculturalization of cultivated land and landscape ecological risk in Liuguan with a karst basin landform and Xianchang with a karst trough valley landform were higher than those in Longchang with a karst mid-mountain landform and Minxiao with a karst low hilly landform. Except for Liuguan, the degrees of correlation between the non-agriculturalization of cultivated land and landscape ecological risk in the other three towns were found to gradually decrease from slope gradient I to V (Table 7).


**Table 7.** The correlation coefficient (global Moran's index) between the non-agriculturalization of cultivated land and landscape ecological risk.

Note: \*\* indicates a significant correlation at the 0.01 level, \* represents a significant correlation at the 0.05 level. The slope is divided into five gradients (from gradient I to V) according to the slope value from lowest to highest.

The high-high and low-high areas of Longchang are concentrated in the western and southern parts, while the low-low and high-low areas are concentrated in the central part (Figure 5a). The high-high and low-high areas in Liuguan are mainly distributed in the southern part, while the low-low and high-low areas are mainly distributed in the northwestern part (Figure 5b). The northern and southern parts of Xianchang are dominated by high-high and low-high areas, while the western and eastern parts are dominated by low-low and high-low areas (Figure 5c). The eastern and central parts of Minxiao are dominated by low-low and high-low areas, while the southeastern and northern parts are dominated by high-high and low-high areas (Figure 5d).

**Figure 5.** The spatial patterns of the types of agglomeration between the non-agriculturalization of cultivated land and landscape ecological risk. (**a**) Longchang, (**b**) Liuguan, (**c**) Xianchang, (**d**) Minxiao.

#### **4. Discussion**

#### *4.1. Comparison with Previous Research Results*

This study found that the karst mountainous areas in western China face the serious non-agriculturalization of cultivated land, which is consistent with the research results of scholars in Africa, Europe, and eastern China [4,9,10]. However, regarding the formation factors of the non-agriculturalization of cultivated land, the results of this study differ from those conducted in other regions around the world. Ecological management policy (namely, the Grain for Green Project) is the key factor causing the non-agriculturalization of cultivated land in the karst mountainous areas of western China, while non-agriculturalization in other regions of the world is mainly caused by urbanization and industrialization [31,32]. In addition, this study found a negative correlation between the non-agriculturalization of cultivated land and landscape ecological risk, i.e., the non-agriculturalization of cultivated land has a positive impact on the ecological environment, which is contrary to the research results of Yang [33] and Yang [34]. The reason for this is that the cultivated land in the karst mountainous areas considered in this study has been mainly converted to ecological land. In contrast, Yang [33] and Yang [34] found that cultivated land was mainly converted to built-up land.

#### *4.2. Formation Mechanism of the Non-Agriculturalization of Cultivated Land in Karst Mountainous Areas*

From the 1950s to the end of the century, a large amount of sloping land in karst mountainous areas in China was reclaimed to cope with the increased demand for food caused by the surging population. To deal with the serious ecological problems (such as rock desertification and soil erosion) caused by excessive land reclamation, the Grain for Green Project was implemented in the 21st century, which resulted in the conversion of a large amount of sloping cultivated land to ecological land [35]. Therefore, although the project is helpful for the restoration of the ecological environment, it causes the serious conversion of cultivated land to non-agriculturalization land-use types in karst mountainous areas. In addition, although economic development due to China's western development policy has led to the conversion of a portion of cultivated land in karst mountainous areas to built-up land and roads in the past ten years [36,37], the resulting rate of conversion of the non-agriculturalization of cultivated land has been far lower than the impact of the Grain for Green Project; this is related to the low demand for artificial land resulting from population losses and a low economic level. It should be noted that the cultivated land in karst mountainous areas is mainly converted into ecological land, and the increase of the natural landscape reduces the degree of landscape fragmentation and vulnerability. Therefore, there is a negative correlation between the non-agriculturalization of cultivated land and landscape ecological risk in karst mountainous areas.

Different landform regions have different natural conditions and human activities, which lead to differences in the distribution and utilization of cultivated land [38]. Under the interference of complex human activities, the changes of cultivated land in different karst landforms are bound to be different, which causes variations in the nonagriculturalization characteristics of cultivated land in different landforms. For example, Minxiao with a karst low hilly landform is located in the surrounding area of the Fanjing Mountain Nature Reserve. To protect and restore the ecological environment, a large amount of sloping cultivated land has been converted to forestland and shrub-grassland, which has caused a high non-agriculturalization rate in this town. However, the terrain of Liuguan is relatively flat, and there is little sloping farmland. The Grain for Green Project has had a relatively small influence on the conversion of cultivated land to non-agricultural land in this town, resulting in a lower non-agriculturalization rate in this town as compared to that in other towns with different landforms.

#### *4.3. Land Management Policy*

In view of the serious situation of the non-agriculturalization of cultivated land in karst mountainous areas, the following measures are suggested. (1) It is suggested that a balanced relationship between economic development, ecological restoration, and cultivated land protection be coordinated. First, the traditional mode of economic development in karst mountainous areas must be changed to reduce the amount of cultivated land occupied by built-up land. Second, it is necessary to reasonably arrange the implementation planning of the Grain for Green Project and prevent a large amount of cultivated land from being converted into ecological land. Finally, the red line of cultivated land protection should be delimited, and priority protection should be given to high-quality and concentrated contiguous cultivated land. The non-agriculturalization of high-quality cultivated land should be avoided. (2) It is necessary to improve the level of the intensive utilization of builtup land, and the government should establish an economic penalty mechanism to prevent the excessive conversion from cultivated land to built-up land. (3) It is necessary to improve the irrigation conditions of cultivated land, increase the usage degree of agricultural machinery, and improve the agricultural production efficiency. In addition, it is necessary to increase planting subsidies for cultivated land to prevent the abandonment of cultivated land, especially sloping cultivated land.

#### *4.4. Limitations*

The spatial resolution of the remote sensing images is one of the factors affecting the reliability of the research results. In view of the small spatial scale of the four selected towns, high-precision remote sensing images were selected for use in this study. However, due to the difficulty in obtaining long-term historical remote sensing images with high precision, only the past decade (between 2010 and 2020) was selected as the research period. Thus, the short research period was a limitation of this study.

#### **5. Conclusions**

The issue of the conversion of cultivated land into non-cultivated land in karst mountainous areas in China has been very serious. The conversion is mainly manifested as the conversion of cultivated land to forestland and shrub-grassland. The rate of conversion from cultivated land to ecological land is significantly higher than that from cultivated land to living land. There are differences in the non-agriculturalization of cultivated land in different slope gradients of different karst landforms. The Grain for Green Project has led to the conversion of a large amount of cultivated land into ecological land in karst mountainous areas in China and has played a key role in the non-agriculturalization of cultivated land in this area. The increase of ecological land and the decrease of cultivated land resulting from the non-agriculturalization of cultivated land have reduced the degree of landscape disturbance, which has led to a negative correlation between the non-agriculturalization of cultivated land and landscape ecological risk. It is worth noting that the conversion of cultivated land into non-cultivated land has not caused food risk, as revealed by the grain yield data on the national scale. This study is of great value to the formulation of protection strategies for cultivated land. Future research should predict the future conversion of cultivated land to non-agricultural land under different development scenarios based on a mathematical spatial model. In addition, in future research, questionnaires will be used to analyze the impacts of various economic factors on the non-agriculturalization of cultivated land at the scale of farmers.

**Author Contributions:** Data curation, H.P.; methodology, J.Y. and Z.Y.; resources, S.L.; writing original draft preparation, H.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the National Natural Science Foundation of China (41861035), the Humanity and Social Science Youth foundation of the Ministry of Education (19YJC760135) and the Natural Science Research Project of Education Department of Guizhou Province (KY[2021]075).

**Data Availability Statement:** The data used to support the findings of this study are available from the corresponding author upon request.

**Conflicts of Interest:** The authors declare no conflict of interest.
