| Land use change is one of the most important componsents of global environmental change, which is received increasing attention. The demand for cropland increased with the growth of the population, and the regional economy was expanding with a high speed, which resulted in the change of the patterns of the land use types. The effects of land use change are various, such as water environmental change, climate change, soil quality, food security, biogeochemical cycles, biological diversity, and sustainable development of the society. Xinjiang is located in the northwestern China with a fragile ecological environment. The ecosystem received severe disturbances and the area of desertification land increased with the intensification of human activities. The unreasonable use of natural resources lead to a series of ecological problems including soil salinization, oasis degradation, and the decline of farmland quality, making the originally fragile ecological environment worse.Based on the domestic and foreign research status of land use change, our study was carried out in Xinjiang at different scales, which has a typical fragile ecological environment, to explore the land use changes and the driving mechanisms. The study area was classified with respect to the relevant variables of cropland, in order to identify the characteristics of cropland disturibution and the socio-economic drivers of each group. The landsat images of 1998, 2006, and 2011 were used to explore the dynamics of the land use types in typical counties selected from different groups, to compare the effects of human disturbance, natural factors, and the socio-economic variables on land use change in different ethnic cultural backgrounds, and to analysis the interrelationships between different driving forces. Based on the results of driving force research, the prediction of future land use change in different scenarios was conducted in the typical counties, which can provide decision basis for regional sustainable development. The main results of this paper are as follows:(1) Cluster analysis can classify agricultural and animal husbandry conties, which laid the foundation for the study of drivers of cropland change in different groups. The counties in Xinjiang were classified into 6 groups according to the percentage of cropland, the population density, and the proportion of the population engaged in animal husbandry. The statistical analysis of the variables of these gropus indicated that the counties of group I and group V belonged to agriculture counties, and the counties of group IV were animal husbandry counties. According to the changes of cropland in the 6 groups, most of the counties were experiencing cropland expansion during 1998 and 2008, and the cropland areas increased dramatically during the last five years. The area of cropland in group II and V increased a lot, and the growth rate of many counties in group V were even greater than 200%. The areas of the cropland in the counties of group III increased slowly except several counties, which also had a dramatically increase during 2003 and 2008.(2) The variables TPAM, GOVA, and CCF were most frequently highlighted as significant driving factors of cropland change in the six models. All of the 14 socioeconomic variables can affect cropland change on the whole. Each group had different drivers of cropland change with others, and the influences of each variable in different groups varied a lot. Besides the most frequently highlighted variables, M-POP, A-PoP, GOVAFO, GOVAH, and Y-OIL were identified as strong driviers only once each in the six models. The results of this research indicated that the patterns and changes of cropland of different regions have their own characteristics, and significant differences existed in the drivers of cropland change among groups. As a result, to classify the counties according to the related attributes is of great importance for further studying land use change.(3) The compositions and patterns of the land uses in Baicheng and tekes have their own characteristics. The area of grassland in Tekes was approximately 5-fold larger than that of cropland, however in Baicheng, the area of grassland was about the same size as the area of cropland. Both of the two counties had large areas of unused land, the proportion of which was above 70% in Baicheng. However in Tekes, the proportion of grassland and unused land were both about 40%. As for spatial distribution, the settlement and cropland were located along rivers. The grassland in Baicheng was mainly located in the hillside and the valley far away from the settlement. However in Tekes, besides the grassland far away, large amount of grassland was located in the gentle slope and the flat ground near to the residents.(4) The analysis of the drivers of land use disturibution and change indicated that, elevation, slope, temperature, precipitation, and the distance to the nearest resident were the main driving factors, and the influences of each variable were different in Baicheng and Tekes. There was a big difference in the composition of ethnic groups in the two counties, which was an important reason for the variation in land use patterens and compositions. Different enthic groups have their own production and living styles, and their needs for each land use differ greatly. The people in Baicheng are mainly engaged in agriculture, so there is a large demand for cropland. With the increase of the population, other types of land use were reclaimed to cropland gradually. As most of the people in Tekes are engaged in animal husbandry, the need for grassland is larger than cropland. When grassland degeneration occurs, the hersmen would find new grazing sites far away from their houses.(5) Variation partitioning was used in our study in order to analyze the interrelationships among different driving factors. Results indicated that anthropogenic disturbance shared more variation with topography than climate in explaining the disturibution and transition of land use types. Anthropogenic disturbance had positive relationships with topography, and higher elevation and steeper slope would limit the production and living lives of the people, as a result, the strength of human activieies may vary with topography. Climate was related to geographical location, topography, and some other complicated factors, so climate had little shared variation with anthropogenic disturbance.(6) The results at different scales were compared and indicated that, the patterns and distributions of the land use types were influenced by the topography, climate, and socio-economic factors at regional scale, and had apparent regional characteristics. Most cropland was located in northwest of Xinjiang, which had abundant rainwater, while in the southeast, which had a drought climate, the cropland was rare. The composition and driving forces of different groups varied a lot, and the main variables included TPAM, GOVA, and CCF. Ethnic culture had significant impacts on land use composition and transition at county-level scale. The ratio of cropland in Baicheng was relatively high, the people of which were mainly engaged in agriculture, however in Tekes, which contained most animal husbandry population, the area of grassland was much larger than cropland. Under the condition of different ethnic cultural backgrouds, the topography, climate, population and so on were still the main driving forces of land use changes, but the influences of human activities were remarkable.(7) 8 driving factors of land use changes such as climate and topography were used to model the land use changes of Tekes in 2020 under different constraint conditions, including trend development scenario, cropland protection scenario, ecological security scenario, and artificial modification scenarios, which have achieved good results. The demands for land use types differed in different scenarios, and the spatial patterns of land use change also varied a lot. However on the whole, as the land use changes were affected by the driving factors, the patterns and the changes of the land use types had the similar characteristics in some degree. Cropland and settlement expansition occurred based on the original cropland, and the gain of forest and grassland were mainly located in the places with gentle slope and high population density near to the cropland, settlement, and rivers. |
PDF Full Text Request