The Settlement Pattern Evolution And Cropland Change In Hehuang Valley In Ming Dynasty

Settlement is the place where people reside, live, rest and work. Human activities were often carried around settlements in historical period. Influenced by natural and social factors, the locations of settlements were selected by human, and the land around them were reclaimed. It would lead to the division or migration and formed a new settlement system when the settlements enlarged even caused the deficiency of the surrounding environment bearing capacity, so the distribution and evolution of settlements could reflect the land use/cover change process and occupied an important position in the global environmental change. Hehuang valley is in the northeast of Qinghai-Tibetan plateau which is an important component of agriculture and pasturage interlaced zone, the pattern evolution of settlements and the cropland change could provide basis for studying regional land use/cover change. Especially since Ming dynasty, the settlement change in Hehuang valley was relatively obvious affected by the military defense and recuperat policy, the cropland change influenced by settlement change is also deep. Hence, using settlement distribution as a factor to rebuild the cropland distribution and change is a better way, it could provide reference for cropland reconstruction methods in historical period.The settlement name database and cropland area database of Hehuang valley in Ming dynasty were established by extracting the information related to the settlements and croplands from historical literatures. Using methods of literature analysis and GIS spatial expression, combined with modern maps and data calibration, this paper reconstructed the spatial distribution of settlements and analysed the temporal and spatial evolution characteristics. Taked the settlement distribution as the main factor combined with other natural factors that affecting the distribution of croplands, the multivariate logistic regression model was established. Through it we determined the main factors and their coefficients, calculated the probability of cropland distribution in different periods, rebuilted and analysed the distribution changes of croplands. Finally the following conclusions were obtained:(1) The method that rebuild the cropland distribution in Hehuang valley in Ming dynasty by using settlement factor is appropriate. Previous grid method can only calculate the reclamation rate of unit grid, while the regression model through this method can simulate the specific distribution of croplands within unit grid. The reconstruction results are more accurate and consistent with the historical facts of Ming dynasty.(2) The growth trend of settlements can be divided into two stages: formed expansion stage(1373~1411) and defense expansion stage(1480~1596). At the view of the spatial distribution, alluvial-diluvial plains have been the most intensive areas where settlements were distributed and the settlements in the Huangshui River basin were more than that in the Yellow River basin. The characteristics of spatial evolution process are that the settlements in alluvial-diluvial plains were more intense and that in loess mountain ridge areas and medium relief mountains increased significantly. The settlements in high elevation relief mountains also growed out of nothing. The development process of settlements were influenced by many factors which included the regional natural geographical conditions, establishing system of Ming dynasty, military defense, reclaiming wastelands and the distribution of the post roads. Based on this, different types of settlements produced. Because of the closely combination with the process of wastelands, Baozhai and post stations became the main settlement types that affected the cropland distribution.(3) In the determinate regression model, the settlement is the most influential factor which influenced on the cropland distribution, the rest in turn are altitude, slope, distance from the rivers, precipitation, accumulated temperature and soil types. The distance from the settlements, altitude, slope and the distance from the rivers are negatively correlated with probability of cropland distribution, precipitation, accumulated temperature and soil types are positively correlated with probability of cropland distribution.(4) The following results could be obtained by calculating the probability of cropland distribution: the possible cropland distribution areas whose probability value more than 0.5 are gradually expanded, in which the areas that probability value more than 0.9 reached more than half, mainly distributed in concentrated areas of settlements which located in alluvial-diluvial plains along Huangshui river and the Yellow River.With the expansion to the periphery, the probability decreased. The characteristics of cropland change can be summarized as follow: the croplands only concentrated around the settlements in the early Ming and have block distribution. Huangshui river and some large tributary valleys are the areas settlements distributed. With the increase of settlements, the small tributary valleys have been reclaimed and croplands have been gradually linked together, forming sheet. The average elevation rose and the elevation range expanded, which indicated that lower areas suitable for reclaiming was gradually reclaimed. The altitude upper limit of croplands is slightly lower than the settlements, which indicated that the settlement locations were more affected by military defense, while the croplands distributed in the areas with better natural conditions.