【摘要】 研究土地利用变化机制，是保障生态脆弱区生态环境良性发展的必然要求。基于土地利用信息图谱构建新指标——土地利用转换度，借助土地利用转移矩阵和土地利用转换度刻画1990—2021年研究区土地利用转移的宏观特征，构建渔网提取土地利用转换度，分析其空间自相关性，并用地理加权回归(GWR)分析研究区土地利用变化的影响因素，为林牧交错区社会经济发展服务。研究发现：(1)1990—2021年，研究区土地利用类型的宏观格局变化不大，林地和草地是该区域的主要用地类型，林地主要分布在东部和北部的山区，草地主要分布在西南部，耕地主要分布于山地和草原交接的地区及河谷地带。(2)1990—2021年林地呈现增加趋势，草地逐年减少，耕地和建设用地缓慢增加。(3)草地、耕地和林地是该地区土地利用变动的主要类型，以草地转出最多，主要转换为林地和耕地。(4)土地利用转换度有显著空间自相关，存在明显的高高聚集和低低聚集。转换频繁区呈倒“C”字形，分布在大兴安岭西部山区与草地过渡带，即海拔700～900 m的地带。(5)GWR分析发现坡度、气温变化、人均GDP变化是影响该区土地利用的主要因素，人均GDP增加会降低土地利用转换度，而其他因素增加会导致土地利用转换加速。(6)土地利用转换度是一种可靠的土地利用变化度量指标，能够将空间和时间融合反映土地利用状态的变化强度。更多还原

【Abstract】 Studying the mechanism of land use change is an inevitable requirement for ensuring the healthy development of the ecological environment in ecologically fragile areas. A new index, land use conversion degree, was constructed based on land use information Atlas, and the macro characteristics of land use transfer in the study area from 1990 to 2021 were depicted by means of land use transfer matrix and land use conversion degree. A fishing net was constructed to extract land use conversion degree, and its spatial autocorrelation was analyzed. The influencing factors of land use change in the study area were analyzed to serve the social and economic development of the forest-grazing area. The findings are as follows:(1) From 1990 to 2021, the macro pattern of land use types in the study area has little change. Woodland and grassland are the main land types in the study area. Woodland is mainly distributed in the mountainous areas in the east and north, grassland is mainly distributed in the southwest, and cultivated land is mainly distributed in the area where mountain and grassland intersect and river valleys.(2) From 1990 to 2021, the forest land showed an increasing trend, the grassland decreased year by year, and the cultivated land and construction land increased slowly.(3) Grassland, cultivated land and forest land are the main types of land use change in this region, and grassland is the most transferred, mainly converted to woodland and cultivated land.(4) There is significant spatial autocorrelation of land use conversion degree, and there are obvious high aggregation and low aggregation. The frequent conversion area is in an inverted “C” shape, and it is distributed in the transition zone between mountain and grassland in the west of the Greater Khingan Mountains, that is, the zone with an altitude of 700～900 m.(5) GWR found that slope, temperature change and per capita GDP change were the main factors affecting land use in the region, and the increase of per capita GDP would reduce the land use conversion degree, while the increase of other factors would also lead to the acceleration of land use conversion.(6) Land use conversion degree is a reliable measure of land use change, which can integrate space and time to reflect the intensity of land use change.更多还原