关键词: 环境异质性, 异质性度量, 植被结构, 植被高度, 植被丰富度, LiDAR数据, 尺度依赖, 景观

Abstract: Vegetation heterogeneity, is a key driver of biodiversity by providing niches and refuges and facilitating coexistence and diversification. Because heterogeneity is scale-dependent, quantifying how it varies across spatial scales is essential for understanding biodiversity patterns and guiding ecosystem management. However, empirical evidence describing how heterogeneity changes with scale remains limited. We systematically analysed 60 relationships between vegetation heterogeneity and spatial scale across three common landscape types, using six vegetation structural and one plant diversity measures and three calculation methods. We identified dominant forms of heterogeneity–scale relationships and assessed how vegetation measures, calculation methods, and landscape types influenced model support, rate of change, and characteristic spatial scales. Most relationships (88%) were nonlinear, and logarithmic models were the most common best-supported form. Maximum rates of change occurred mainly at finer spatial scales, whereas levelling-off points in logarithmic relationships were concentrated primarily below 10 km. Both slope magnitude and levelling-off scale varied significantly among vegetation heterogeneity measures, and levelling-off scale also differed among landscape types, with urban landscapes tending to level off at finer scales than agricultural and semi-natural landscapes. Overall, our results show that vegetation heterogeneity usually changes nonlinearly across spatial scales and that the scale dependence of these patterns varies among heterogeneity measures and landscape contexts. Our framework provides a useful workflow for quantifying heterogeneity–scale relationships and guiding the selection of ecologically relevant metrics and spatial scales for biodiversity monitoring, conservation planning, and ecosystem management.

Key words: Environmental heterogeneity, heterogeneity measurement, vegetation structure, vegetation height, vegetation richness, LiDAR data, scale-dependent, landscape