The research, published in Frontiers in Forests and Global Change, was led by Zhe Zhu and colleagues at the Wuhan Institute of Technology. The team drew on the China Land Cover Dataset, a 30-meter-resolution satellite record spanning 1985 to 2025, to identify plots that slipped out of cultivation between 2012 and 2021 and then tracked which of them returned to farming between 2022 and 2025. Using land-use trajectory analysis, kernel density mapping, and spatial autocorrelation statistics, the study evaluated recultivation across three dimensions: scale, efficiency, and stability. The findings arrive at a moment when global food systems are under strain. The 2026 Global Report on Food Crises, produced with the Food and Agriculture Organization and the World Food Programme, estimated that roughly 266 million people still faced severe food insecurity in 2025.

The stakes for China are considerable. The country’s grain output reached about 714.88 billion kilograms by 2025, yet with a population above 1.4 billion and annual consumption near 830 million tons, a supply-demand gap of more than 100 million tons persists. Demand for feed and industrial grain keeps climbing as diets upgrade, and China remains heavily dependent on imports for soybeans. Against this tight balance, farmland abandonment has become one of the most intractable problems in cultivated land protection. Previous research cited in the study estimates that abandoned farmland in China covers 20.79 percent of the total, representing a loss of 49.23 million tons of grain production capacity—enough to feed 123 million people.

The temporal results are striking. In January 2021, the Ministry of Agriculture and Rural Affairs issued national guidelines on coordinated utilization of abandoned land, and the 2023 No. 1 Central Document imposed a rigid requirement to intensify recultivation. The satellite record shows the response was immediate. The recultivation rate in the Yangtze River Economic Belt rose from 16.63 percent in 2022 to 26.49 percent in 2025, a total gain of 9.86 percentage points over four years, averaging about 2.5 percentage points annually. In 2023 alone, the increase reached 3.51 percentage points, the fastest single-year jump in the record. The team also traced each year’s recultivated area back to the abandoned-farmland stock of three periods—2012-2015, 2015-2018, and 2018-2021—and found that early recultivation relied heavily on the oldest stock, but by 2025 the contributions from all three periods had become far more balanced, with the most recently abandoned land absorbing resources fastest.

Geographically, the pattern follows a pronounced west-high, east-low gradient. The upper reaches—Chongqing, Sichuan, Guizhou, and Yunnan—hold the largest abandoned stock and posted the biggest absolute gains, with recultivated area expanding from 8,047.42 to 12,923.91 square kilometers and the recultivation rate climbing 10.03 percentage points to 26.59 percent. The middle reaches—Jiangxi, Hubei, and Hunan—achieved an almost identical 10.02-point rise despite a much smaller stock base, expanding from 3,055.37 to 4,925.80 square kilometers. The lower reaches—Shanghai, Jiangsu, Zhejiang, and Anhui—started from a higher initial rate of 19.53 percent but grew only 5.77 points to 25.30 percent, constrained by an intrinsically tiny abandonment stock and strong pressure for non-agricultural conversion.

Stability tells its own story. Taking the plots recultivated in 2022 as an observation sample, the team found that 91.12 percent of middle-reach plots remained farmland continuously through 2025, and 90.28 percent in the upper reaches. In the lower reaches, only 85.36 percent held for three years, with 2.95 percent reverting after a single year and 8.86 percent after two. Kernel density analysis confirmed the clustering: high-density and medium-high-density recultivation zones concentrated in the upper reaches, with density declining stepwise eastward. Over time, low-density zones expanded while medium-density and above contracted, indicating that recultivation is diffusing from hotspot areas toward a more balanced spatial pattern.

Spatial autocorrelation statistics reinforced the gradient. Moran’s I values ranged from 0.754 in 2022 to 0.761 in 2025, all significant at the 99 percent confidence level, confirming strong positive clustering. Local analysis identified 31 to 32 cities in high-high agglomeration zones, nearly all in the upper reaches, while a stable 27 cities formed contiguous low-low clusters in the lower reaches. Most middle-reach cities showed no significant spatial association. The structure remained remarkably inert across four years, suggesting that the local geography of recultivation is deeply rooted and unlikely to shift without targeted intervention.

The driving forces form a nested, interacting system. Natural geography comes first: the upper reaches are dominated by mountains and plateaus where sloping farmland accounts for 72 percent of cultivated land, and abandoned area in 2018-2021 represented 72.5 percent of the basin total. International evidence cited in the study shows mountainous abandonment rates in Slovakia ran twice those of lowlands, and Central and Eastern Europe held 52.5 million hectares of abandoned agriculture concentrated in marginal areas. Economic development shapes opportunity costs: per capita GDP in Jiangsu reached 160,694 yuan in 2024 against 67,473 yuan in Yunnan, meaning the same subsidy pulls far harder in the west, where non-farm wages are lower and land attachment stronger. Labor structure adds another layer—research shows aging farmers with limited off-farm employability can actually stabilize recultivation once it begins, while in the lower reaches recultivation depends heavily on large-scale operators whose succession is uncertain.

Policy timing explains the temporal acceleration. The 2021 guidelines launched administrative accountability, fiscal incentives, and multi-departmental joint supervision simultaneously, converting recultivation from spontaneous household action into government-led campaigns. This effect proved most pronounced in the upper reaches, where a large stock base and high household flexibility produced the biggest gains. The middle reaches benefited from mature land-transfer markets and well-developed new agricultural operators, while the lower reaches saw policy effects channeled mainly into preventing new abandonment. The 2025 No. 1 Central Document’s call to promote reclamation by category signals a shift from digesting historical stock toward categorized governance of newly abandoned land.

The study’s authors propose region-specific priorities: in the upper reaches, suitability assessment to avoid one-size-fits-all mandates and post-recultivation consolidation mechanisms; in the middle reaches, expanded land-transfer markets and full-chain agricultural socialization services; in the lower reaches, a pivot from expanding recultivation to preventing new abandonment through high-value agriculture. The research also acknowledges limitations—the 30-meter satellite data may misclassify subtle transitions, and the three-year stability window cannot capture long-term re-abandonment risk. Still, the message is clear: recultivation outcomes depend not just on how much land was abandoned, but on the interplay of terrain, economics, demographics, and institutional design. As China’s grain balance stays tight and global food insecurity persists, the Yangtze River Economic Belt offers a natural laboratory for understanding whether the world’s farmland can be won back—one terrace, one plain, one policy at a time.

Subject of Research: A satellite-based study of the Yangtze River Economic Belt shows abandoned farmland recultivation rising about 2.5 percentage points per year since 2021 policies, with strong west-to-east gradients in scale, efficiency, and stability.

Article Title: Spatiotemporal differentiation and determinants of abandoned farmland utilization in the Yangtze River Economic Belt

Article References: Zhu, Z., Chen, S., Shao, L., Li, P., Zou, H., & Sheng, Y. (2026). Spatiotemporal differentiation and determinants of abandoned farmland utilization in the Yangtze River Economic Belt. Frontiers in Forests and Global Change, 9, Article 1975795. https://doi.org/10.3389/ffgc.2026.1975795

Image Credits: AI Generated

DOI: 10.3389/ffgc.2026.1975795

Keywords: Spatiotemporal, differentiation, determinants, abandoned, farmland, utilization, Yangtze, River, Economic, Belt, scientific research