Assessing effects of the Returning Farmland to Forest Program on vegetation cover changes at multiple spatial scales: The case of northwest Yunnan, China

Chinese state authorities have successfully accelerated afforestation over large areas through the Returning Farmland to Forest Program (RFFP). However, variation in the implementation of the RFFP among villages may lead to uncertainty in forest transitions at multiple spatial levels. Here, we combined remote sensing analysis with field quadrats and questionnaire surveys to assess the RFFP's effectiveness in increasing vegetation cover at the Lancang watershed, township and natural village levels in Weixi County, a biodiversity hotspot in northwest Yunnan. From 2000 to 2010 and 2010 to 2014, forest coverage increased, and areas of shrubland and agricultural land decreased at both watershed and township levels, which reflected a positive impact of the RFFP. However, the new forests established under the RFFP in the study area mostly consisted of monocultures of walnut and pine trees, a situation that could threaten local biodiversity. Changes in forest and shrub coverage varied among natural villages-gain and loss trends both occurred, indicating a limited impact of the RFFP. Nonparticipating RFFP households also abandoned cropland, and agricultural land area in these villages declined over the study period. At the watershed and township levels, RFFP greatly promoted increases in forest coverage. Elevation, the behavior of village officials and household livelihoods were often the most relevant factors at the natural village level. Interactions among these multiple drivers resulted in cross-scale heterogeneity of vegetation cover changes in northwest Yunnan. Future forest policies need to be tailored to specific regions and be based on assessments of local natural and socioeconomic conditions.