Interprovincial food trade aggravates China's land scarcity

Global food security is a major concern, projected to require a substantial increase in agricultural production by 2050 to meet rising demands. Land, a critical resource for food production, is increasingly scarce, necessitating efficient resource management. While agricultural trade can potentially mitigate land scarcity, its effectiveness remains unclear. This study addresses this gap by focusing on China, the world's most populous country with a relatively limited amount of cropland. The study utilizes a high-resolution approach to analyze the complex interplay between interprovincial food trade and land scarcity in China. The research aims to provide a detailed analysis of the spatial distribution of land use, identify hotspots of land scarcity, and assess the impact of virtual land flows caused by interprovincial trade. This granular analysis is crucial for developing effective policies to address land scarcity and improve resource management in China and potentially other countries facing similar challenges.

Previous studies have examined land footprint (LF) and its role in international trade, highlighting the teleconnections between land supply and demand across regions. However, these studies often neglect the crucial aspect of land scarcity, defined as the imbalance between land availability and demand. The concept of scarcity is critical because the same land area can contribute differently to overall scarcity depending on the local balance of supply and demand. Existing LF measures do not account for this spatial variability in scarcity, leading to an incomplete understanding of land use impacts. This research builds upon previous work by incorporating a land scarcity index (LSI) to create a scarce land footprint (SLF), which more accurately reflects the true impact of land use on scarcity. The incorporation of high-resolution data also represents a significant advance over previous studies that often operated at coarser spatial resolutions, limiting the detail and accuracy of their findings.

This study employed a multi-regional input-output (MRIO) analysis, a powerful tool for capturing economic linkages between different sectors and regions. The model consisted of 31 Chinese provinces and 42 sectors. The fundamental equation of the MRIO model (X = AX + Y = (I − A)−1Y = LY) relates total output (X), final demand (Y), and the technical coefficient matrix (A). The Leontief inverse matrix (L) captures direct and indirect inputs required to satisfy final demand. The study calculated the land footprint (LF) and scarce land footprint (SLF) using the MRIO table and land use data. LF represents the total cropland used directly and indirectly to satisfy final demand, while SLF incorporates a land scarcity index (LSI) to account for the relative scarcity of cropland in different regions. The LSI is calculated as the ratio of total cropland area to suitable cropland area. The study also introduced a trade-related land scarcity index (TLSI) to assess the impact of virtual land trade on land scarcity in each province. High-resolution land footprint and scarce land footprint maps were generated at a 1 km × 1 km resolution using spatial downscaling methods based on population gridded data. Hotspot analysis using Getis-Ord G* statistic identified spatial clusters of high LF and SLF. The analysis distinguished between direct and indirect (virtual) land use, offering insights into interprovincial land flows. Data sources included the MRIO table from Liu et al. (2019), cultivated land area from the China Statistical Yearbook (NBSC, 2013), suitable cultivated land from GAEZ 3.0 model, and gridded population data from Landsat Satellite 7 (MODIS, 2012).

The study's key findings highlight significant spatial heterogeneity in land footprint (LF) and scarce land footprint (SLF) across China. Approximately 70% of China's LF and SLF were concentrated in less than 20% of the land area, primarily in the Jing-Jin-Ji Region, Yangtze River Delta, and Sichuan Basin. Nearly 38% of LF and SLF hotspot clusters crossed provincial boundaries, indicating the significant trans-provincial nature of land use impacts. Virtual land and scarce land flows predominantly moved from less-developed inland regions (net exporters) to developed coastal regions (net importers). Heilongjiang was the largest net exporter of both virtual land and scarce land. Jiangsu was the largest net importer. A striking imbalance was observed in the spatial distribution of LF and SLF; a small fraction of land area accounted for a large proportion of the total LF and SLF. Comparison of the land scarcity index (LSI) and the trade-related land scarcity index (TLSI) showed that while virtual land trade mitigated land scarcity in importing provinces by 50.8%, it disproportionately aggravated land scarcity in exporting provinces by 119.8%. The difference between SLF and LF was negative in 91.5% of China's land area, suggesting an overestimation of land use impacts by traditional LF measures. Analysis also revealed a notable role transition in Anhui and Hunan between net virtual land and net virtual scarce land.

The findings challenge the conventional wisdom that interprovincial food trade automatically mitigates land scarcity. While importing provinces benefit from reduced local pressure on land resources, exporting provinces experience a disproportionate increase in land scarcity. This spatial mismatch highlights the need for policy interventions that address the distributional consequences of food trade. The high-resolution spatial analysis revealed significant cross-provincial interactions, underlining the importance of collaborative governance strategies to manage land resources effectively. The overestimation of land use impacts using traditional LF underscores the critical need to incorporate scarcity metrics into environmental footprint accounting for a more accurate and nuanced understanding of sustainability.

This study provides the first comprehensive assessment of the impact of trade-induced land redistribution on land scarcity in China. The findings demonstrate that interprovincial food trade, while beneficial to importing regions, exacerbates land scarcity in exporting provinces. The integration of LSI and high-resolution data enabled the identification of geographically specific hotspots and the quantification of the uneven distribution of land use pressures. Policy recommendations include improving agricultural productivity, updating land resource management policies, promoting collaborative governance, and incorporating scarcity metrics into existing environmental footprint indicators to support the implementation of the SDGs. Future research should explore time-series analysis, incorporate foreign trade, and develop more geographically explicit scarcity metrics.

The study's limitations include uncertainties associated with the MRIO model selection. Although a Monte Carlo simulation showed consistency among different MRIO databases, utilizing a time-series analysis would strengthen the findings. The aggregated agriculture sector in the MRIO table may also have affected the accuracy of land use efficiency calculations, and neglecting foreign trade limits the global perspective. The land scarcity metrics may represent a simplistic view of land-use intensity. Future work could address these limitations by using more refined data, incorporating a global MRIO model, and developing more comprehensive scarcity metrics.