Search

Systematically recognizing spatial patterns and driving factors of cultivated land fragmentation is of great significance for the exploration of locally appropriate path to relieve cultivated land fragmentation. This study aims to estimate cultivated land density, mean patch size and area-weighted mean shape index to respectively indicate the characteristics of cultivated land fragmentation from three dimensions, namely, natural resource endowment, spatial partition, and convenience of utilization. The regional leading factors of cultivated land fragmentation are also analyzed. The results demonstrate that the distribution of cultivated land density is higher in Northern regions compared with those of southern regions. The significant positive correlation between cultivated land density and mean patch size is found to be universal across nearly all cities, exceeding differences in terrain, elevation, climate, soil, and social economic condition. For cities in the southern part of China, cultivated land of regular shape is partitioned to smaller blocks compared with irregular ones; alternatively, intensive and meticulous farming under small-scale agricultural operation leads to clusters with low mean patch size - low area-weighted mean shape index. Random forest model explains the impact of driving factors on cultivated land fragmentation, with an explanatory power ranging from 66% to 95%. The terrain factor emerges as the primary driver negatively affecting cultivated land density. Gross domestic product emerges as the dominant factor with a significant (p < 0.01) negative correlation to mean patch size for nearly all agricultural climatic zones. Terrain, gross domestic product and population is the most important factor affecting area-weighted mean shape index. Rural development degree influences correlation between dominant factors and cultivated land fragmentation. This study is greatly instructive for recognizing the spatial patterns of cultivated land fragmentation at the national scale and for exploring the barriers that impede regionally scaled cultivated land use. • First reveal 1-km grid scale cultivated land fragmentation in mainland China. • Detect the polarizing differences of cultivated land fragmentation between peri-urban and out-of-urban areas. • Analyze China national spatial clustering characteristics of cultivated land fragmentation. • Recognize regional dominant factor to cultivated land fragmentation. • Generate China national 1-km grid-based cultivated land fragmentation maps. [ABSTRACT FROM AUTHOR]

In recent years, the sustainable utilization of China's arable land has been confronted with several challenges. The China government has been very strict in arable land protection, and a package of policies and measures have been promulgated. All these endeavors are of great significance for proposing an innovative policy system for sustainable land use in China. However, above stated policies are all designed from the perspective of space control with the purpose of reducing arable land loss or increasing arable land area, few policies have been designed from the perspective of utilization control, namely guide the actual arable land farming in sustainable ways and constraint unreasonable land use behavior such as overuse, rough use, land abandonment. In this paper, we analyze spatial distribution of average land-use intensity (ALUI) at the county-level in Mainland China, which can be used as a significant index for evaluating the rationality of arable land use and providing effective decision-making supporting information for design of regional arable land protection policy. Based on the experimental results, there is still considerable room for yield improvement as the ALUI of ∼73.1 % counties are lower than 0.7 while the 53.60 % counties are lower than 0.6. Furthermore, the ALUI dataset shows significant global spatial autocorrelation characteristic. Boundaries of regions that aggregated by counties with high ALUI are more consistent with that of provincial administrative districts, comparing with that of sub-standard farming system regions. On the other hand, counties with low ALUI are mostly cluster in mountains, hills, or plateaus, where grain yield is mainly limited by regional hydrothermal conditions. In addition, counties with different ALUI status have been divided into six classes, using k-means clustering algorithm. This will facilitate the understanding of appropriate arable land protection and utilization paths for different regions and the rethinking of current support policies on farmland protection. • Present arable land-use intensity of mainland China at county scale. • Analyze spatial aggregation characteristics of arable land-use intensity. • Key points for regional arable land protection were found among different counties. • Present a new, output-oriented measure for arable land-use intensity. [ABSTRACT FROM AUTHOR]