Can We Prevent Irreversible Decline? A Comprehensive Analysis of Natural Conditions and Quality Factor Thresholds of Cultivated Land in China

: Over the past two decades, China’s spatial distribution of cultivated land has been stable, yet there remains an urgent need to amplify grain productivity. The central scientiﬁc question addressed in this study is: How can we effectively evaluate the natural resource thresholds of cultivated land at a regional scale? Despite existing systems, there is a noticeable gap, particularly in this area of inquiry. Recognizing the present state of these resources and introducing more efﬁcient management strategies is pivotal. This paper aims to research the restrictive situation of the natural resources background quality (NRB-quality) of China’s cultivated land by developing an innovative classiﬁcation method and a short-board identiﬁcation method and adopting cluster analysis and other technical methods. The results showed that (1) China mainly maintains medium-quality land (57.30%). The soil texture displayed a ﬁne average grade of 1.39, while the biodiversity, soil pH, secondary land types, and cropping systems indicated poor conditions, with average scores of 2.01, 2.06, 2.26, and 2.33. (2) A notable difference in the restrictive factors of cultivated land quality emerged, with secondary land types of cultivated land (including paddy ﬁeld, irrigated land, and dry land) being the only national short-board factor. Regionally, the cropping system, secondary land types of cultivated land, and natural region were identiﬁed as short-board factors in 31, 24, and 23 natural regions. (3) The regional difference in cultivated land quality leads to six different management strategies. Eastern regions emphasized stabilizing land distribution and infrastructure enhancement, while China’s western areas advocated cautious development and heightened ecological protection. The ﬁndings demonstrated regional differences in the status of cultivated land quality factors; implementing classiﬁed management approaches for cultivated land quality factors based on their characteristics is vital to preventing irreversible decline. The study serves as an important basis for the protection and utilization of cultivated land in the new era to clarify the NRB-quality differences of cultivated land in China.


Introduction
Cultivated land refers to the land for planting crops, including paddy fields, irrigated land, and dry land (also known as dry farmland), forming the bedrock of national food security. While China has a larger total cultivated land area, the per capita availability is relatively lower. Additionally, a significant area of middle-and low-yield fields leads to a lack of high-quality cultivated land [1]. Presently, the quantity and space distribution of cultivated land resources in China remain largely stable [2][3][4][5][6], and few cultivated land reserve resources exist in China [7,8], which means that the natural resources background quality (NRB-quality) of cultivated land in China is basically unchanged. Figure 1 shows China's current spatial distribution of cultivated land [9]. At the same time, there are natural quality differences in cultivated land resources in China. The vast territory and diverse natural geography of China contribute to variations in the natural resources background and carrying capacity. These elements significantly impact the distribution and quality of cultivated land resources [10].
Land 2023, 12, x FOR PEER REVIEW 2 of 17 reserve resources exist in China [7,8], which means that the natural resources background quality (NRB-quality) of cultivated land in China is basically unchanged. Figure 1 shows China's current spatial distribution of cultivated land [9]. At the same time, there are natural quality differences in cultivated land resources in China. The vast territory and diverse natural geography of China contribute to variations in the natural resources background and carrying capacity. These elements significantly impact the distribution and quality of cultivated land resources [10]. Currently, China is in a tight balance of basic food security, socioeconomic stability, and ecological health. China's grain yield has stood firm at 650 million tons annually for eight consecutive years [11], and the grain possession has been stable above the internationally recognized food security standard line of 400 kg per capita, which ensured food security. The nation's demographic trend, economic evolution, and globalization capability are on a steady trajectory [12]. There are prominent contradictions between the population distribution, economic development, and the layout and structure of resources and the environment in China. However, with the gradual transformation of China's economic and social development mode and residents' consumption mode, population growth, energy consumption, and resource consumption growth are expected to slow down gradually, ensuring ecological stability [13].
In the future, there is still a need to further improve the grain production capacity in China, which puts forward higher requirements for cultivated land protection. Under the background of stable cultivated land space distribution, it is of great significance to analyze the NRB-quality of cultivated land, clarify the restrictive factors to improve cultivated land quality, and then put forward measures to control cultivated land resources reasonably. There are few countries in the world like China with less cultivated land per capita, significant differences in the NRB-quality of cultivated land, and great food demand. The study of the NRB-quality of China's cultivated land is significant for other regions in a tight balance worldwide. Currently, China is in a tight balance of basic food security, socioeconomic stability, and ecological health. China's grain yield has stood firm at 650 million tons annually for eight consecutive years [11], and the grain possession has been stable above the internationally recognized food security standard line of 400 kg per capita, which ensured food security. The nation's demographic trend, economic evolution, and globalization capability are on a steady trajectory [12]. There are prominent contradictions between the population distribution, economic development, and the layout and structure of resources and the environment in China. However, with the gradual transformation of China's economic and social development mode and residents' consumption mode, population growth, energy consumption, and resource consumption growth are expected to slow down gradually, ensuring ecological stability [13].
In the future, there is still a need to further improve the grain production capacity in China, which puts forward higher requirements for cultivated land protection. Under the background of stable cultivated land space distribution, it is of great significance to analyze the NRB-quality of cultivated land, clarify the restrictive factors to improve cultivated land quality, and then put forward measures to control cultivated land resources reasonably. There are few countries in the world like China with less cultivated land per capita, significant differences in the NRB-quality of cultivated land, and great food demand. The study of the NRB-quality of China's cultivated land is significant for other regions in a tight balance worldwide. Cultivated land quality evaluation supports cultivated land protection policies such as cultivated land balance and high-standard farmland construction [14,15]. China has established national standards for evaluating cultivated land quality to support its management. In 2012, the Ministry of Land and Resources (MOLR) of the People's Republic of China published the Regulation for gradation on agriculture land quality (GB/T 28407-2012) [16]. In 2016, the Ministry of Agriculture and Rural Affairs (MARA) of the People's Republic of China published the Cultivated land quality grade (GB/T 33469-2016) [17]. While these standards provide scientific indicators and methods for investigating, monitoring, and evaluating cultivated land quality, they have limitations. They focus on comprehensive evaluation and fail to address the specific constraints of cultivated land resources or meet the requirements for protection in the current situation. In 2019, the Ministry of Natural Resources launched the third national land survey, which included ten indicators that can represent the NRB-quality to comprehensively evaluate cultivated land quality [18,19].
Internationally, many scholars are concerned about the restrictive research based on factors. Some scholars have put forward and followed up the conceptual framework of planetary boundaries, delineating global safety thresholds around ecological determinants like climate change, biodiversity loss, a change in land use, and other factors [20][21][22][23]. Liu and Chen [24] hold that the tipping elements are irreversible, and once the tipping point is activated, the earth system will not return to its original stable state. In evaluating cultivated land quality, the limitation of factors is also more prominent. For example, the Land Evaluation and Site Assessment (LESA) system of the United States Department of Agriculture analyzes the structure, function, and obstacles of the land and site system, and the land sub-level obtained by land use potential classification indicates the types of land restrictive factors [25]. Targeting cultivated land degradation issues, such as soil pollution, decreased soil organic matter content, and biodiversity, the European Union has established a problem-oriented evaluation and monitoring standard for the soil environment [26]. Some studies have also explored the threshold of cultivated land quality factors, with some suggesting a Soil Organic Carbon (SOC) content threshold of 2% in temperate regions and 1.1% in tropical areas [27][28][29]. However, threshold determination methods lack uniform standards, primarily relying on the breakpoint method and empirical approaches [30]. Gao [31] used the natural breakpoint method to classify ecological sensitivity factors. Yang [32] divided NPP into four categories by the natural breakpoint method and extracted the threshold of NPP classification. Furthermore, restrictive study on the factors of cultivated land system security remains elusive.
This research was designed to evaluate the restrictive situation of cultivated land quality factors in China. The hypothesis that will be tested is whether the cultivated land quality factors in China have reached the threshold. Based on the cultivated land resources quality classification data in the third national land survey in 2019, we evaluated the grade of cultivated land quality by the comprehensive evaluation method and analyzed the comprehensive quality of cultivated land resources in China. We analyzed the NRBquality of China's cultivated land by developing the identification method of short-board factors of cultivated land, defining the restrictive threshold of each factor, and identifying the short-board factors of cultivated land quality. Finally, based on the comprehensive evaluation results of cultivated land quality and the restrictive situation of cultivated land quality factors, we analyzed the regional differences in cultivated land resource quality and advanced the countermeasures of the classified management of cultivated land resources quality.

Data Sources and Preprocessing
The study took cultivated land in China as the research scope. According to the Main Data Bulletin of the Third National Land Survey, there were 127.86 million hectares of cultivated land in China in 2019, of which paddy fields, irrigated land, and dry land accounted for 24.55%, 25.12%, and 50.33%, respectively [33]. In our study, the digital elevation model (DEM) data and land use remote sensing monitoring data came from the Resource and Environment Science and Data Center [34]. The statistical data of the classification and grading of cultivated land resources quality, and the vector data of natural regions and administrative divisions, were all derived from the results of the third national land survey and analyzed by ArcGIS 10.7 and Excel 2021. The cultivated land resources quality classification in the third national land survey established a classification method system including six levels and ten indicators, including the natural region, slope, soil thickness, soil organic matter content, soil texture, soil pH value, biodiversity, soil heavy metal pollution, cropping system, and secondary land types of cultivated land. Among them, the natural region reveals the regional differentiation law of nature in China, reflects the eco-geographical relationship of various factors of the natural environment (such as temperature, moisture, and landform), and can represent the resource background of cultivated land. In order to unify the evaluation criteria, 49 natural regions were divided into three levels through expert discussion. And the five grades of the slope were divided into three grades, including ≤6 • , 6~15 • , and 15~25 • . Other indicators in the system were divided into three levels, of which the first-level indicator is the best, and the third is the worst. In our study, only 46 natural regions within the scope of data investigation were analyzed, and three natural areas were not analyzed, including the VIA1 Mountain Plain Region in North-Central Taiwan, VIIA1 Low Mountain Region of Southern Taiwan, and IXA1 Nansha Islands Region.

Comprehensive Evaluation Method of Cultivated Land Quality
We took the cultivated land map of the third national land survey in 2019 as the evaluation unit. The ten indicators of each evaluation unit were divided into one to three grades. The seven indicators were selected as evaluation indicators, including slope, soil thickness, soil organic matter content, soil texture, soil pH value, biodiversity, and soil heavy metal pollution [9]. Each evaluation unit's cultivated land quality grade was determined by the number of indicators at each grade. See Equation (1) for the specific evaluation method.
where the grade represents the quality index of cultivated land quality evaluation, n represents the number of indicators, i represents the ith evaluation unit, and j represents the indicator level. i takes 1, 2, 3. . .n, and j takes 1, 2, 3.
According to the comprehensive evaluation method of cultivated land quality, we divided cultivated land into four grades: superior cultivated land, good cultivated land, medium cultivated land, and poor cultivated land, and the corresponding cultivated land quality indexes are 1, 2, 3, and 4, respectively. According to the evaluation results, the evaluation units were summarized according to the county, and statistics were obtained on the area of four cultivated lands in each county. Then, according to the analysis needs, the grading results and comprehensive evaluation results of each index were summarized by natural regions or administrative regions.

Identification Method of Short-Board Factors of Cultivated Land
We found an unreasonable cultivated land use status using the Natural Breaks. The Natural Breaks is a classical analysis tool used to present and identify the characteristics of Land 2023, 12, 1669 5 of 17 regional spatial differentiation in geographical research [35]. The method holds that there are some turning points and breakpoints set by nature rather than human beings between any series, and these breakpoints are good boundaries for classification.
Our study viewed that, in a certain region, when a factor' third grade (the worst grade) exceeds a certain proportion, the factor negatively impacts the cultivated land quality in this region and is a short-board factor. Taking each county as a sample, R software 4.2.2 was used to carry out binary natural breakpoints on the third-level ratio of all sample points of each factor, and the breakpoint value between the two groups was taken as the threshold. Then, the restrictive degree of a factor in a certain region was characterized by the ratio of the proportion of the region of the third-grade cultivated land in the region to the threshold. When the degree of restriction is greater than one, it is considered that the factor is a short-board factor in this region.
According to the calculation results, each factor was divided into five categories: no obvious restriction factor, weak restriction factor, strong restriction factor, short-board factor, and strong short-board factor. The corresponding restriction degree values are 0~0.4, 0.4~0.8, 0.8~1, 1~2, and >2, respectively.

Analysis Method of NRB-Quality of Land Quality Characteristics Based on Clustering
According to the evaluation results of cultivated land resources and the NRB-quality difference reflected by various factors in various natural regions, 46 natural regions within the analysis scope were classified, which lays the foundation for the classified management of cultivated land quality ( Figure 2). The K-means clustering algorithm classified the natural regions based on cultivated land evaluation results and the restrictive degree of each factor in 46 natural regions. Natural regions with similar cultivated land quality will be divided into one category. The cluster analysis results were used as the basis for natural area classification.
the grading results and comprehensive evaluation results of each index were summarized by natural regions or administrative regions.

Identification Method of Short-Board Factors of Cultivated Land
We found an unreasonable cultivated land use status using the Natural Breaks. The Natural Breaks is a classical analysis tool used to present and identify the characteristics of regional spatial differentiation in geographical research [35]. The method holds that there are some turning points and breakpoints set by nature rather than human beings between any series, and these breakpoints are good boundaries for classification.
Our study viewed that, in a certain region, when a factor' third grade (the worst grade) exceeds a certain proportion, the factor negatively impacts the cultivated land quality in this region and is a short-board factor. Taking each county as a sample, R software 4.2.2 was used to carry out binary natural breakpoints on the third-level ratio of all sample points of each factor, and the breakpoint value between the two groups was taken as the threshold. Then, the restrictive degree of a factor in a certain region was characterized by the ratio of the proportion of the region of the third-grade cultivated land in the region to the threshold. When the degree of restriction is greater than one, it is considered that the factor is a short-board factor in this region.
According to the calculation results, each factor was divided into five categories: no obvious restriction factor, weak restriction factor, strong restriction factor, short-board factor, and strong short-board factor. The corresponding restriction degree values are 0~0.4, 0.4~0.8, 0.8~1, 1~2, and >2, respectively.

Analysis Method of NRB-Quality of Land Quality Characteristics Based on Clustering
According to the evaluation results of cultivated land resources and the NRB-quality difference reflected by various factors in various natural regions, 46 natural regions within the analysis scope were classified, which lays the foundation for the classified management of cultivated land quality ( Figure 2). The K-means clustering algorithm classified the natural regions based on cultivated land evaluation results and the restrictive degree of each factor in 46 natural regions. Natural regions with similar cultivated land quality will be divided into one category. The cluster analysis results were used as the basis for natural area classification. The K-means clustering algorithm is an iterative unsupervised learning algorithm. Its working mechanism is to group similar objects into a cluster so that the sample points The K-means clustering algorithm is an iterative unsupervised learning algorithm. Its working mechanism is to group similar objects into a cluster so that the sample points in each cluster have high similarity. In contrast, the sample points in different clusters have low similarity. The cluster center is calculated by the average of all points in the cluster. According to the principle of minimizing the clustering performance index, the commonly used clustering criterion function is that the Sum of the Squared Error (SSE) between each Land 2023, 12, 1669 6 of 17 sample point in the cluster and the center of the cluster is the minimum [36]. The criterion function of SSE is shown in Equation (2).
where k is the number of clusters, Ci is the data set of the ith cluster, µi is the center point of the cluster Ci, and x is the data sample in cluster Ci. According to the clustering results of the geographical location of natural regions, we adjusted to obtain the classification management model of cultivated land resources quality based on natural regions.

Evaluation Results of Cultivated Land Quality in China
The quality of cultivated land resources in China was not high. Specifically, the proportion of superior, good, medium, and poor cultivated land in China was 8.47%, 27.03%, 57.30%, and 7.20%, respectively ( Figure 3a). Due to the unmatched spatial distribution of water and soil resources and the large area of mountains and hills, the quality of cultivated land resources in China is generally not high. It was shown that the superior cultivated land and good cultivated land with excellent quality accounted for 35.50% of the cultivated land area in China, which were mainly distributed in the northeast region with good cultivated land site conditions and fertile soil and the middle and lower reaches of the Yangtze River with a flat terrain and sufficient water and heat. Medium cultivated land and poor cultivated land accounted for 64.50% of the cultivated land area in China, accounting for a relatively large proportion, which was mainly distributed in the Yun-Gui Plateau and Loess Plateau with a steep slope and serious soil erosion and the Qinghai-Tibet Plateau with a high altitude and fragile ecology.
in each cluster have high similarity. In contrast, the sample points in different clusters have low similarity. The cluster center is calculated by the average of all points in the cluster. According to the principle of minimizing the clustering performance index, the commonly used clustering criterion function is that the Sum of the Squared Error (SSE) between each sample point in the cluster and the center of the cluster is the minimum [36]. The criterion function of SSE is shown in Equation (2).
where k is the number of clusters, Ci is the data set of the ith cluster, µi is the center point of the cluster Ci, and x is the data sample in cluster Ci. According to the clustering results of the geographical location of natural regions, we adjusted to obtain the classification management model of cultivated land resources quality based on natural regions.

Evaluation Results of Cultivated Land Quality in China
The quality of cultivated land resources in China was not high. Specifically, the proportion of superior, good, medium, and poor cultivated land in China was 8.47%, 27.03%, 57.30%, and 7.20%, respectively (Figure 3a). Due to the unmatched spatial distribution of water and soil resources and the large area of mountains and hills, the quality of cultivated land resources in China is generally not high. It was shown that the superior cultivated land and good cultivated land with excellent quality accounted for 35.50% of the cultivated land area in China, which were mainly distributed in the northeast region with good cultivated land site conditions and fertile soil and the middle and lower reaches of the Yangtze River with a flat terrain and sufficient water and heat. Medium cultivated land and poor cultivated land accounted for 64.50% of the cultivated land area in China, accounting for a relatively large proportion, which was mainly distributed in the Yun-Gui Plateau and Loess Plateau with a steep slope and serious soil erosion and the Qinghai-Tibet Plateau with a high altitude and fragile ecology.   There are significant differences among different factors according to the classification and the proportion of different factors of cultivated land resource quality in China (Figure 3b). The soil heavy metal pollution situation was the best among all the factors. The average grades of soil texture, slope, and soil organic matter content were 1.39, 1.47, and 1.49, and the overall situation was relatively good. The area proportions of the best quality grades of the three factors were 66.15%, 61.93%, and 55.47%. The overall situations of the natural region, soil thickness, biodiversity, soil pH value, secondary land types of cultivated land, and cropping system were poor, with average grades of 1.75, 1.79, 2.01, 2.06, 2.26, and 2.33. Among them, the cultivated land area with the best quality of the cropping system and soil pH value was very small, accounting for only 14.73% and 14.77% of the cultivated land area in China.

Background Status of Cultivated Land Resources Quality in China
The different background conditions of cultivated land quality factors in China lead to different restrictive degrees of each factor. According to the calculation results of the identification method of short-board factors in cultivated land, there was no strong shortboard factor (>2) at the national level (Figure 4a). The restrictive degree of the secondary land types of cultivated land was 1.09, which was a short-board factor (1~2), indicating that the dry land proportion in China was high and the water resources utilization conditions of cultivated land were poor. The restrictive degrees of the cropping system and biodiversity were 0.96 and 0.87, which were strong restriction factors (0.8~1), and the cropping system was close to the threshold, reflecting that there was more cultivated land that crops once a year in China and was greatly limited by light and temperature conditions. The restrictive degrees of the soil thickness and soil pH value were 0.57 and 0.51, which were weak restriction factors (0.4~0.8). The restrictive degrees of the natural region, slope, soil texture, and soil organic matter content were 0.39, 0.33, 0.15, and 0.14, which belonged to no obvious restriction factors (0~0.4) at the national level, and the restrictive degree of the natural region was relatively high, close to a weak restriction factor.
Our study further analyzed the restrictive degree of each factor in different natural regions, and the results showed obvious differences (Figure 4b). The factors of the cropping system, secondary land types of cultivated land, and natural region were very restrictive in most natural regions. The cropping system was a short-board factor in 31 natural regions in China, a strong restriction factor in 1 natural region, and no restriction in other regions. The secondary land types of cultivated land was a strong short-board factor in VIIIA1 and a short-board factor in 23 natural regions. The natural region factor was a short-board factor in 23 natural regions, but the cultivated land area only accounted for 17.23% of the national cultivated land. The factors of soil thickness, slope, and soil pH value had strong restrictions in some areas. Soil thickness was a strong short-board factor in two natural regions, IA1 and HIID3, and it was a short-board factor in nine natural regions, mainly distributed in northeast and southwest China. The slope was a strong short-board factor in two natural regions, IVA2 and VIIA3, and it was a short-board factor in seven natural regions, mainly distributed in southwest China. The soil pH value was a strong short-board factor in VIIIA1 and a short-board factor in six natural regions, mainly distributed in southeast China. Biodiversity was a short-board factor in 9 natural regions but a strong restriction factor in 23 natural regions. The soil texture and soil organic matter content were less restrictive in various natural regions of China, and they were only short-boards in two and one natural regions, respectively. Still, future construction may have an impact on these factors.  The restrictive conditions of 10 factors in each natural region, including the natural region, slope, soil thickness, soil organic matter content, soil texture, soil pH value, cropping system, secondary land types of cultivated land, biodiversity, and soil heavy metal pollution. The red circle represents the threshold boundary 1, and the column represents the restrictive degree of the factor in different natural regions. If the column exceeds the red line 1, the factor is a short-board factor in the natural region.

Analysis of Regional Differences in Cultivated Land Resources Quality in China
Based on the evaluation results of cultivated land resources and the restrictive conditions of each factor, 46 natural regions were divided into six groups by the K-means clustering algorithm, and the total sum of squares was 281.1829 and the Clustering adequacy was 55.2%, which showed that the clustering effect was good (Table 1). Combined with the geographical location of natural regions, they were finally divided into six cultivated land resources quality management zones: Northeast High-quality Cultivated Land Protection Zone, Huang-Huai-Hai Modern Agricultural Development Zone, Southeast High-quality and Efficient Utilization Zone, Northwest Drought Moderate Development Zone, Southwest Ecological Vulnerable Zone, and Qinghai-Tibet The restrictive conditions of 10 factors in each natural region, including the natural region, slope, soil thickness, soil organic matter content, soil texture, soil pH value, cropping system, secondary land types of cultivated land, biodiversity, and soil heavy metal pollution. The red circle represents the threshold boundary 1, and the column represents the restrictive degree of the factor in different natural regions. If the column exceeds the red line 1, the factor is a short-board factor in the natural region.

Analysis of Regional Differences in Cultivated Land Resources Quality in China
Based on the evaluation results of cultivated land resources and the restrictive conditions of each factor, 46 natural regions were divided into six groups by the K-means clustering algorithm, and the total sum of squares was 281.1829 and the Clustering adequacy was 55.2%, which showed that the clustering effect was good (Table 1) Zone ( Figure 5). Significant differences existed in the comprehensive quality of cultivated land and NRB-quality in different districts. The three regions in eastern China had large cultivated land areas, few restrictive factors, and good overall quality. The Northeast High-quality Cultivated Land Protection Zone had a large cultivated land area, accounting for 28.14% of the national cultivated land area. At the same time, superior cultivated land accounted for a large proportion and poor cultivated land accounted for a small proportion, accounting for 9.59%, and 5.02%. The restrictive factors were mainly the cropping system and secondary land types of cultivated land, and the average restrictive degrees were 2.00 and 1.54, that is, the area of one crop a year and dry land was relatively large, and the overall quality was good. Huang-Huai-Hai Modern Agricultural Development Zone had a large area of good cultivated land, accounting for 41.01%. The main restrictive factor in this zone was the secondary land types of cultivated land, with an average restrictive degree of 1.11, that is, there was more dry land. The proportion of medium cultivated land in the Southeast High-quality and Efficient Utilization Zone was as high as 75.84%, and the superior cultivated land, good cultivated land, and poor cultivated land accounted for 9.29%, 9.22%, and 5.65%, respectively, with a low proportion. The soil pH value was a common short-board factor in this zone, with an average restrictive degree of 1.66, and the problem of soil acidification was serious.
There was a lot of poor cultivated land in the three zones of western China, and the quality of cultivated land was relatively poor because of the strong restrictions of some factors. The proportion of good cultivated land and poor cultivated land in the Northwest Drought Moderate Development Zone was high, accounting for 37.08% and 9.42%. The primary short-board factors were the natural region and cropping system, with average restrictive degrees of 1.71 and 1.93, respectively. This means that the natural geographical conditions in this zone were poor, and the annual cropping area was widely distributed. The poor cultivated land in the Southwest Ecological Vulnerable Zone accounted for 20.71%. The main restrictive factors were the slope, soil thickness, and secondary land types of cultivated land, and the average restrictive degrees were 1.58, 0.95, and 1.48. This indicates that there was a greater proportion of sloping cultivated land, a thinner soil layer, and dry land. There was less cultivated land in the Qinghai-Tibet Ecological Protection Zone, accounting for only 0.31% of the total cultivated land in China. The proportion of medium cultivated land was as high as 69.10% in this zone. There are many short-board factors, such as the natural region, soil thickness, cropping system, and biodiversity, and the average restrictive degrees were 1.85, 1.08, 2.00, and 1.12, respectively. Hence, the quality of cultivated land was poor in the zone.

Comparison between Classification Results of Cultivated Land Resources Quality and Related Results
The classification system of cultivated land resources quality was improved on the national standard of cultivated land quality evaluation of MOLR and MARA [37]. It can better meet the new requirements of natural resources management. The evaluation system of MOLR has comparable characteristics in the whole country. Therefore, the classification results of cultivated land quality are converted into the comparable cultivated land quality grade in the control area of the balance of cultivated land in the whole country. Figure 6 shows that the classification system is reasonable. The classification results of the cultivated land resources quality coincided with the updated evaluation results of the national cultivated land quality in 2016 by MOLR and the classification results of the national cultivated land quality in 2019 by MARA. However, there was a certain gap between the classification results of the cultivated land resources quality and the evaluation results of two cultivated land qualities, especially for the evaluation results of cultivated land with general quality.

Comparison between Classification Results of Cultivated Land Resources Quality and Related Results
The classification system of cultivated land resources quality was improved on the national standard of cultivated land quality evaluation of MOLR and MARA [37]. It can better meet the new requirements of natural resources management. The evaluation system of MOLR has comparable characteristics in the whole country. Therefore, the classification results of cultivated land quality are converted into the comparable cultivated land quality grade in the control area of the balance of cultivated land in the whole country. Figure 6 shows that the classification system is reasonable. The classification results of the cultivated land resources quality coincided with the updated evaluation results of the national cultivated land quality in 2016 by MOLR and the classification results of the national cultivated land quality in 2019 by MARA. However, there was a certain gap between the classification results of the cultivated land resources quality and the evaluation results of two cultivated land qualities, especially for the evaluation results of cultivated land with general quality.  The difference is mainly due to the different evaluation index systems besides the different evaluation years. The regulation for gradation on agriculture land quality (GB/T 28407-2012) [16] issued in 2012 by MOLR not only considers natural factors such as light, temperature, moisture, and soil conditions but also considers utilization efficiency and production benefits. The Cultivated land quality grade (GB/T 33469-2016) [17] issued by MARA comprehensively considers the cultivated land fertility, soil health status and field infrastructure from the perspective of agricultural production. However, the classification system of cultivated land quality highlights the natural resources background, For classification evaluation, it selects ten stable indicators from three angles: natural geographical, natural background conditions, and harmonious coexistence between man and nature. In addition, the classification system can not only reflect the comprehensive quality of cultivated land resources and evaluate the quality grade but also reflect the status of various cultivated land quality factors and the restrictive factors to serve the management of cultivated land resources better [19].

Relationship between the NRB-Quality of Cultivated Land Resources and Grain Productivity
The essence of cultivated land protection is to protect the productive capacity [38,39], and the NRB-quality of cultivated land is the basic condition of cultivated land productivity and sustainable utilization [40]. Taking each province as a sample, the restrictive degree of factors represents the NRB-quality of cultivated land, and the grain yield per unit area represents the grain productivity of cultivated land. The relationship between the NRBquality of cultivated land and the grain yield was calculated by a stepwise regression algorithm. The average grain output per unit area from 2018 to 2020 was calculated, which is the ratio of the total grain output to the grain sown area in each province. The statistics came from the National Bureau of Statistics (stats.gov.cn, accessed on 20 April 2023). The model fitting is shown in Table 2. The model correction R 2 is 0.5711, and the p-value (0.0001132) is less than 0.05, which shows that the model fitting is good and has a high degree of interpretation. The results showed that the natural region, cropping system, and secondary land types of cultivated land had significant effects on cultivated land grain productivity, and the p-values of all factors were less than 0.05. However, the other factors had little influence on cultivated land grain productivity ( Table 2). The natural region reflects the differences in the natural environment, such as the temperature, moisture, and landform. Agricultural production has a strong dependence on natural conditions [41]. The weaker the restriction of natural conditions, the higher the grain productivity of cultivated land. The analysis results of our study showed that the greater the restriction of the cropping system, the higher the grain yield per unit area. This may be due to the low utilization intensity, high organic matter content, and higher productivity of cultivated land in the one-year cropping area [42]. The restrictive degree of secondary land types of cultivated land hurt grain productivity, that is, the greater the proportion of dry land, the lower the grain productivity. Dryland had a low irrigation rate, weak drought resistance, and poor stability of grain production [43,44]. The results proved that the natural background status of cultivated land resources significantly impacted the grain productivity of cultivated land, and the classification system of cultivated land resources quality had a strong practical significance.

Classified Management Countermeasures of Cultivated Land Resource Quality
The current measures for cultivated land protection strongly guarantee cultivated land quantity, but the protection for cultivated land quality needs to be stronger. Studies have shown that China is one of the countries with more cultivated land degradation worldwide, facing multiple land degradation pressures such as drought, vegetation decline, and soil erosion [45]. From the 1980s to 2000, the pH value of soil in the black soil region decreased by 0.33 units, with 91.6% of the paddy soil region experiencing a decrease in the pH value [46]. Soil acidification has become a widespread issue in Chinese cultivated land, activating soil pollutants and increasing ecological and environmental risks. Compared with the second national soil survey in the 1980s, in 2019, it was observed that the soil organic carbon storage in the plow layer (0-20 cm) in Northeast China decreased by 0.41 Mg C/hm 2 , which led to the decline in soil water and fertilizer conservation capacity and further degraded the ecological functions of cultivated land such as the black soil biodiversity and balanced nutrient supply, which seriously threatened food security and the sustainable utilization of cultivated land [47]. More effective measures must be taken to prevent the irreversible decline in China's cultivated land quality. The NRB-quality of cultivated land in China had different characteristics and restrictions in different zones. Implementing differentiated management countermeasures is more helpful in improving the NRB-quality (Table 3). Table 3. Classification and management countermeasures of cultivated land resources.

Type Natural Regions Restriction Management Countermeasures
Northeast High-quality  Cultivated Land  Protection Zone   IA1, IIA1, IIA2, IIA3,  IIB1, IIB2, IIB3,  IIC1, IIC2, IIC4 Soil erosion and degradation are serious in the black soil area of Northeast China, the farmland infrastructure is weak, and the effective irrigation area is small.
Stabilize the space distribution, implement conservation tillage, and strengthen infrastructure construction. Modern Agricultural  Development Zone   IIIA1, IIIB1, IIIB2,  IIIB3, IVA1 The level of farmland infrastructure is not high, and the utilization efficiency of irrigation water is low.

Huang-Huai-Hai
Stabilize the space distribution, improve the efficiency of water resources utilization, and improve infrastructure construction.
Southeast High-quality and Efficient Utilization Zone VA1, VA2, VIA2, VIIA2, VIIIA1 Non-agriculture and soil acidification of cultivated land is serious.
Carefully develop and strengthen ecological protection.
Eastern China serves as the primary region for grain production, benefiting from relatively favorable natural conditions for cultivated land. Maintaining stability in the space distribution of cultivated land, preventing the trend of non-agricultural and non-grain production on cultivated land, and mitigating the issue of north-to-south grain transport are essential. The farmland infrastructure in the Northeast High-quality Cultivated Land Protection Zone needed to be stronger, with a limited effective irrigated area for cultivated land. Additionally, there was a pressing issue of black soil degradation. To address these challenges, it is crucial to focus on stabilizing the existing cultivated land space distribution and preventing cultivated land's expansion into marginal areas to ensure local environmental sustainability [48]. On this basis, we should promote conservation tillage, strengthen the construction of farmland infrastructure, and maintain a high crop yield while protecting and improving soil quality [49]. Huang-Huai-Hai Modern Agricultural Development Zone was the main grain-producing area in China. However, the level of farmland infrastructure construction was not optimal, and the efficiency of irrigation water utilization was low. The North China Plain is also recognized as the largest groundwater "hopper" area globally. To address these challenges, prioritizing infrastructure improvements and implementing intelligent water-saving irrigation measures are crucial. These actions can enhance the water resource utilization efficiency and promote agricultural modernization. The Southeast High-quality and Efficient Utilization Zone had sufficient water and heat, making it a high-yield and stable grain production zone. However, the phenomenon of non-agriculture and the problem of soil acidification are serious [50]. To ensure the sustainability of the zone, it is vital to stabilize the distribution pattern of cultivated land and curb its decrease within the zone to improve the current situation of north-to-south grain transportation [51]. Additionally, adopting a rational approach to the use of chemical fertilizers and pesticides, implementing soil acidification improvement measures, and enhancing infrastructure construction are necessary steps to achieve sustainable development in the zone.
The ecological environment in Western China was relatively fragile. Water resources were in short supply, the terrain was relatively high, and the environmental cost of grain production was high. Therefore, it must be carefully developed to prevent the phenomenon of the marginalization and upslope of cultivated land. The Northwest Drought Moderate Development Zone faced significant challenges, including a serious shortage of water resources, poor irrigation conditions, a low soil organic matter content, and a weak grain production capacity. To prevent a worsening of the water deficit, soil wind erosion, and fragile ecology in this zone, which would endanger regional ecological security, it must utilize water-saving irrigation methods, improve soil fertility measures, and carefully develop marginal cultivated land [48]. A lot of slope farmland was in the Southwest Ecological Vulnerable Zone, and the soil layer was thin, resulting in serious soil erosion. The occupation of construction land and the development of cultivated land promoted the upslope of cultivated land [52], which aggravated the marginalization of cultivated land. To halt the marginalization and upslope trend of cultivated land, engineering actions must be taken to support the restoration and management of areas with substantial soil erosion and secure the ecological security of the cultivated land in this area. With its high altitude, low temperature, sluggish soil layer development, thin soil layer, and unique geographical environment, the Qinghai-Tibet Ecological Protection Zone possessed a delicate ecological environment [53]. The infrastructure has to be properly created, and local ecological protection needs to be increased.

Conclusions
Based on data on cultivated land quality classification in the third national land survey in 2019, our study analyzed the natural background status and regional differences of cultivated land resources in China using comprehensive evaluation, short-board factor identification, and cluster analysis. We also proposed classified management measures for cultivated land resources. The main findings are as follows: (1) China's cultivated land quality mainly falls within the medium range (57.30%), with superior and good cultivated land making up minor portions, accounting for 8.47% and 7.20%, respectively. (2) At a national level, the secondary land types of cultivated land were identified as a short-board factor, with a restrictive degree of 1.09. The restrictive factors varied across different natural regions. The cropping system, secondary land types of cultivated land, and natural region were identified as short-board factors in 31, 24, and 23 natural regions in China. (3) China's natural regions can be divided into six classified management modes for cultivated land resources. The three eastern zones prioritize stabilizing cultivated land's spatial distribution and improving infrastructure construction. Meanwhile, the three western zones cautiously develop cultivated land resources and strengthen the ecological protection of cultivated land. The findings highlight the variations in the background status of cultivated land quality factors in China. To prevent the irreversible decline in cultivated land quality, it is essential to enhance the management of factors such as the biodiversity, soil pH value, secondary land types of cultivated land, and cropping system. Each factor's restrictive degree differed in different zones, emphasizing the importance of implementing classified management measures tailored to specific factors. These targeted measures can effectively curb the deterioration of cultivated land quality. By assessing the limitation degree of cultivated land quality factors nationwide, our research lays the groundwork for better identifying insufficient factors and providing a more scientifically rigorous foundation for cultivated land resource management, including improving the quality of cultivated land, classification management, and the protection of cultivated land. This includes initiatives aimed at enhancing the cultivated land quality, classification management, and cultivated land protection. In forthcoming studies, we plan to further scrutinize the NRBquality of China's cultivated lands by employing scenario-based approaches and verify the relationship between the NRB-quality of cultivated land and grain productivity to ensure the safety of cultivated land resources and food security more scientifically and effectively.