Reconstructing spatial pattern of historical cropland in karst areas of Guizhou, Southwest China

Karst regions are exceptionally responsive to global change with their harsh natural environment, fragile ecology, and acute human-land conflicts. The reconstruction of cropland spatial pattern in karst areas during the historical period is typical for studying human-land relations in karst areas and has important practical significance for climate study. The ecological environment changes at regional and global scales, primarily to provide essential data and a theoretical basis for studying the inverse evolution of rock desertification in karst areas. Guizhou province, a typical karst area, was selected as the research area in 1820. Based on the correction of historical population data and cropland data, a reconstruction model of cropland spatial pattern in karst areas during the historical period was constructed by selecting factors such as elevation, slope, soil types, organic matter content, climatic productivity potential, distance to river and distance from settlements to reconstruct the spatial pattern distribution of cropland in 1820 of Guizhou. The results show that the data on cropland recorded in Guizhou during the Qing dynasty is too low, mainly due to Yin-Ni and the policy of Tu-Di-Mian-Ke. In 1820, the total area of revised cropland in Guizhou was 1,851,792 hm², with the highest proportion of 14.32% in Dading Fu and the lowest in Songtao Ting at 1.6%. Only 30% of the grid in Guizhou has a cropland distribution. It is mainly concentrated in the central part of Qianzhong District (Anshun and Guiyang Fu), the southern part of Qianbei District (Pingyue Fu and southern Zunyi Fu), the western part of Qiandongnan District, the central and eastern parts of the Qiandongbei District. The overall average reclamation rate of land in Guizhou is 10.93%, the highest reclamation intensity in Qianzhong District, with 8.5% of grids ≥ 50%, and the smallest in Qianxinan District, with only 1.65% of grids ≥ 50%. The analysis is validated by comparing the reconstruction model and the reconstruction results. It can be seen that the reconstruction model and research results of this paper can more objectively reflect the distribution of cropland in karst areas during the historical period, and the reconstruction model is suitable for karst areas with low productivity levels.