内蒙古大兴安岭林草交错带景观动态分析

为加强对林草交错带生态系统的科学管理,进一步促林草资源保护与合理利用,迫切需要摸清交错带景观本底并分析其时空动态演化趋势。以大兴安岭林草交错带为研究对象,选取2000、2010年Landsat 5 TM影像和2018年Landsat 8 OLI影像,利用面向对象的决策树分类算法建立3期土地利用数据集,据此分析土地利用动态变化与景观格局演变特征,然后利用状态转换模拟模型STSM模拟研究区2025年的土地利用数据。结果表明：（1）2010年林地、草地、耕地、湿地、人工表面、盐碱地及荒漠和过火区面积占比分别为46.93%、31.66%、5.02%、13.73%、1.08%、1.55%和0.04%;2018年分别为46.89%、31.69%、4.99%、13.72%、1.15%、1.54%和0.02%。（2）景观尺度上,2010-2018年间林地面积减少43.55 km²,破碎化程度加剧、景观完整性降低、景观构成愈发复杂;草地面积增加38.11 km²,其景观完整性升高。（3）在现行趋势下,预测2025年研究区林地、草地、人工表面和过火区面积分别增加92.27、183.21、66.2 km²和10.25 km²;耕地、湿地和盐碱地及荒漠面积分别减少184.2、2.89 km²和164.84 km²。林火频发是导致研究区林地面积减少的主因,模拟的过火区面积增加提醒森林管理部门要严控林区用火风险并增强火灾扑救能力建设。后期"天然林保护"工程和"退牧还草"政策的实施是生态环境改善的主因。在制定区域发展战略时,需要充分平衡农业生产与城市扩张之间的竞争性,满足区域耕地红线的基本要求。

Dynamic analysis and prediction of landscape pattern in Daxinganling forest-grass ecotone in the Inner Mongonia

To strengthen the scientific ecosystem management in the forest-grass ecotone and to further promote forest and grass resources protection and rational utilization, there is an urgent need to ascertain landscape baseline and its spatio-temporal evolution trend. In this study, the forest-grass ecotone of Daxinganling in the Inner Mongonia was chosen as the research area. Based on the Landsat 5 TM images acquired in 2000, 2010 and Landsat 8 OLI images acquired in 2018, the object-oriented decision tree classification algorithm was applied to create the land use datasets first. Next, the dynamic characteristics of land use and landscape pattern evolution were analyzed, followed by a simulation analysis based on the State and Transition Simulation Model (STSM) to project the current land use trend of the study area to 2025. The results showed that:(1) in 2010, the area proportion of forest, grassland, cropland, wetland, urban, saline-alkali lands and desert, and burned area was at 46.93%, 31.66%, 5.02%, 13.73%, 1.08%, 1.55%, and 0.04%, respectively. In 2018, it was at 46.89%, 31.69%, 4.99%, 13.72%, 1.15%, 1.54% and 0.02%, respectively. (2) At the landscape scale, during the period of 2010-2018, the forest dropped by 43.55 km² and its fragmentation severity increased, accompanied by a decreased integrity of forest landscape and a more complex composition of forest landscape, while grassland increased by 38.11 km² and its integrity of landscape improved. (3) Following the current trend, the projected area of forest, grassland, urban and burned area in 2025 from STSM would increase by 92.27, 183.21, 66.2 km² and 10.25 km², respectively, while cropland, wetland, saline-alkali lands and desert would decrease by 184.2, 2.89 km² and 164.84 km², respectively. In the study of large-scale land cover prediction with historical data, the prediction accuracy of STSM model was better than that of Ca-Markov model. Frequent occurrence of forest fire was primarily responsible for the reduction of forest area in the study area. The simulated increase in burned area informed the forest management department of strictly controlling fire risk in the forested area and escalating the capability construction of fire suppression. And the implementation of Natural Forest Protection Project and the Grain for Green Project in the later period was the leading trigger for the improvement of ecological environment. When developing a regional development strategy, the competition between the agricultural production and urban sprawl must be compromised to satisfy the basic requirement of the regional red line for arable lands.