塞罕坝地区植被景观格局时空尺度效应

Spatiotemporal Scale Effect of Vegetation Landscape Pattern in Saihanba Area

为探讨植被景观格局时空尺度效应,运用随机森林分类方法解译塞罕坝地区3期(1989、2000和2016年)遥感影像,并利用简单聚合法将景观类型栅格数据进行空间粒度上推,计算基于斑块类型水平和景观水平的景观格局指数;采取尺度图谱分析不同时空尺度草地、天然次生林、灌木及人工林等植被类型的景观尺度效应,通过信息熵模型确定了景观格局分析的最佳空间粒度。结果表明:(1)4种植被景观在空间尺度上具有相同的变化特征,而在时间尺度上表现出不同的变化趋势;(2)不同景观指数对时空尺度的响应结果不同,边缘密度、斑块密度、景观形状指数、聚集度指数及蔓延度指数在时空尺度上具有较强的尺度效应,且90 m为研究区植被景观格局的最佳分析粒度;(3)草地、天然次生林在1989—2016年期间,破碎化程度增加,对研究区的生物多样性造成威胁。

In order to explore the spatiotemporal scale effect of vegetation landscape pattern, the random forest classification method was used to interpret the remote sensing images of three periods(1989, 2000 and 2016) in Saihanba area. The spatial grain of the landscape type grid data was scaled up by simple aggregation method. Landscape pattern indices based on patch type and landscape level were calculated. Landscape scale effects of four vegetation types(grassland, natural secondary forest, shrub land and plantation) were studied by using the scale spectrum analysis. The information entropy model was adopted to determine the optimum spatial grain size of landscape pattern. The four vegetation types had the same change characteristics in spatial scale, whereas showed different trends in temporal scale. Different landscape metrics responded differently to the spatiotemporal scale. Edge density, patch density, shape index, aggregation index and contagion index had strong scale effects on the spatial and temporal scales. The optimal spatial grain size threshold was 90 m. The fragmentation of grasslands and natural secondary forests increased from 1989 to 2016, which posed a threat to the biodiversity of the study area.