Assessing the Grain for Green Program and its Conclusion

The Grain for Green Program (GFG) in China stands as the largest ecological restoration and rural development effort globally, marked by its extensive public participation. This research leverages MODIS satellite land use data to examine the spatial and temporal dynamics of changes in forest land, grassland, and cropland for all of China following the implementation of GFG. Overall, the satellite image data show that there have been large increases in forest land and grassland during GFG, but these increases have been offset by concurrent landscape transformations. Specifically, we find that, in aggregate, from 2001 to 2020, approximately 109,498 km² of cropland was converted to forest land and 43,963 km² of cropland was converted into grassland. However, this period also saw the conversion of 104,402 km² of forest land and 109,042 km² of grassland for cropland, resulting in an almost negligible net change in cropland to forest land conversion, and a significant net loss of about 65,000 km² of grassland. Based on our data, the critical food security threshold of 1.8 billion mu (around 120 million km²) China claims for arable land has not been compromised. This study also uncovers a multifaceted interaction in land use conversions among cropland, forest land, and grassland. Significant shifts from cropland to forests occurred in Yunnan, Sichuan, and Heilongjiang. Meanwhile, the largest forest-to-cropland land conversions were observed in Sichuan, Hunan, and Guizhou. This occurrence of concurrent afforestation and reclamation, particularly noticeable in regions like Sichuan, suggests either the influence of local government strategic adjustments based on geographical and soil quality factors, or the varied impacts of regional policies on “food security” and “environmental protection,” inviting further investigation. Furthermore, the results indicate low preservation rates for land converted from cropland to forest. Eighteen years after the start of the program, only 37% of the lands converted from cropland to forest in 2001 were preserved as forest, with 50% reclaimed for cropland and 3% degraded to grassland. Preservation rates are similar for forests planted during other years of the policy. We also attempt to quantify the efficiency of the distribution of cropland and forest land in areas that alternate between the two land use types, and we observe sub-optimal distributions with regards to soil retention and food production. Together, these results imply that while subsidies may help decrease the rate of forest reclamation for agricultural use, they may also promote the conversion of lands unsuitable for agriculture into forests, highlighting the adverse natural conditions for afforestation.