STUDY ON THE EXTRACTION OF MARGINAL LAND IN CHINA AND ITS POTENTIAL EVALUATION FOR CORN PLANTING

Bo Yi, Xiao Han, Lanna Cheng, Ping Zhang, Heng Tang, Chao Nie, Yongjun Du, Ye HuaApplied Ecology and Environmental Research 2025 · DOI

TL;DR

Researchers have mapped 2.27 million km² of "marginal land" across China—areas currently underutilized due to environmental or economic constraints—and found that 30% of it is highly suitable for corn cultivation. If fully leveraged, these lands could boost national corn production by up to 463 million tons, providing a massive strategic reserve for food security.

What Problem Does This Paper Address?

China faces a classic "land vs. food" dilemma: its high-yield arable land is already under intense cultivation, leaving little room for expansion. To stabilize food supplies and address potential shortages, the country needs to look at "marginal land"—areas with low agricultural productivity or significant topographical and climatic challenges (like saline-alkali soil or steep slopes).

The challenge is that "marginal land" is a fuzzy concept. Previous research often focused only on bioenergy crops or lacked a unified framework for identifying these lands at scale. This paper establishes a clear classification system and uses multi-source spatial data to evaluate exactly where this land is and whether it can actually grow corn, one of China’s most critical staple crops.

Methods

The authors employed a multi-step geospatial approach to move from raw satellite data to a "suitability map":

  1. Classification & Extraction: They defined eight types of marginal land (e.g., low-quality forest, cultivable saline-alkali land, mudflats) and extracted them using 30-meter resolution Land Use/Cover Change (LUCC) data.
  2. Normalization: They used min-max scaling to unify disparate data ranges (like precipitation in mm vs. slope in degrees).
  3. The Entropy Weight Method: To avoid the subjective bias of human experts deciding which factor is "more important," the authors used this objective mathematical approach to determine weights for nine indicators across five categories: topography, soil, water resources, climate, and transportation accessibility.
  4. Weighted Linear Combination (WLC): Finally, they overlaid these weighted factors to calculate a comprehensive suitability score for every 1km grid cell.

Datasets

The study integrated a wide array of high-resolution spatial datasets from 2020:

  • Land Use: China 30-m LUCC remote sensing data.
  • Vegetation: Fractional Vegetation Cover (FVC) and MODIS Net Primary Productivity (NPP).
  • Topography: Digital Elevation Model (DEM) data.
  • Climate: Annual precipitation and $\ge 10^\circ\text{C}$ accumulated temperature data.
  • Infrastructure: China river network and road data (Tianditu).
  • Soil: A specialized dataset of soil properties for land surface modeling over China.

Key Insights and Ideas

  • The Northwest is the "Strategic Reserve": The Northwest region holds nearly 30% of China's marginal land, dominated by cultivable grasslands. While arid, the study suggests that with improved water management (like drip irrigation), this region has the highest potential for yield increases.
  • Infrastructure is a Bottleneck: The suitability model didn't just look at soil; it included "distance to roads." In many regions, the land is biophysically capable of growing corn, but the lack of transportation infrastructure makes it economically "unsuitable."
  • Grassland Dominates: Non-arable land (like cultivable grassland and low-quality forest) makes up 73.7% of the total marginal land identified. This suggests that the biggest gains in food security won't come from fixing "bad" farms, but from converting specific types of unused natural land.
  • Patchy Soil Quality: Unlike climate factors, which show smooth geographic trends, soil suitability is highly "patchy" across China, requiring very localized management strategies.

Key Results

  • Total Area: 2.27 million km² of marginal land identified.
  • Suitability Breakdown: 30.0% is "Highly Suitable," 30.7% is "Moderately Suitable," and 24.7% is "Lowly Suitable."
  • Yield Potential: If only the highly suitable marginal lands are utilized, corn production could increase by 353 to 463 million tons.
  • Regional Winner: The Northwest region alone could contribute between 25% and 34.7% of the total potential yield increase.

Limitations and Future Work

The authors are transparent about several hurdles to turning these maps into actual farms:

  • Nature Reserves: The study did not exclude protected ecological areas. This likely leads to an overestimation of available land, as many of these "marginal" areas are under strict conservation laws.
  • The Human Factor: The model doesn't account for land rent, farmers' willingness to cultivate, or public acceptance of land-use changes.
  • Irrigation Complexity: While the study used "distance to rivers," it didn't account for existing large-scale canal systems (like those in Inner Mongolia), which can make distant land much more productive than the model suggests.
  • Future Direction: The authors suggest that future models should incorporate socio-economic variables and more precise hydrological data to determine the "actual" plantable area versus the "theoretical" maximum.

Summary updated

Abstract

Marginal land refers to areas with low agricultural productivity and economic returns due to soil limitations, severe constraints on water and thermal resources, and significant topographical challenges, making these areas ecologically fragile.Accurate identification and evaluation of the crop cultivation potential of marginal land are crucial for ensuring food security.Based on the concept of marginal land, this study classified and extracted eight types of land use: low-quality arable land, low-quality forest land, lowproductivity orchards, cultivable grassland, cultivable saline-alkali land, cultivable mudflats, cultivable sandy land, and cultivable bare land.Considering factors such as topography, soil, water resources, climate, and transportation accessibility, we assessed the suitability of marginal land in China for maize cultivation.The results showed that the total area of marginal land in China was 2.27 10 km, of which low-quality arable land accounted for 26.3%, while non-arable land (including cultivable grassland, low-quality forest land, low-productivity orchards, cultivable saline-alkali land, cultivable mudflats, cultivable sandy land, and cultivable bare land) accounted for 73.7%.The Northwest region contains the largest proportion of marginal land, representing 29.9% of the national total.Among the marginal land, areas highly suitable for maize cultivation account for 30.0%.If fully utilized for maize production, the potential yield increase could range from 35.3 to 46.3 10 tons, compared with the average corn yield of each province over the past five years.These findings provide a theoretical frameworks and scientific data to support the sustainable utilization of marginal land and the optimization of crop planting patterns in China.
BibTeX
@inproceedings{yi2025study,
  author = {Bo Yi and Xiao Han and Lanna Cheng and Ping Zhang and Heng Tang and Chao Nie and Yongjun Du and Ye Hua},
  title = {{STUDY ON THE EXTRACTION OF MARGINAL LAND IN CHINA AND ITS POTENTIAL EVALUATION FOR CORN PLANTING}},
  booktitle = {Applied Ecology and Environmental Research},
  year = {2025},
  doi = {10.15666/aeer/2304\_72617278},
}
Figure