Revealing the spatio-temporal dynamics of tropical forest communities using airborne LiDAR data
Mr YIP Ka Hei Anson
Living in an era of rapid global change, conservation of tropical (and sub-tropical) forests have become a global agenda to address contemporary environmental challenges, including climate change and habitat degradation. Tropical forests, despite covering only 12% of the Earth’s land surface, host more than half of the global biodiversity and terrestrial biomass. Notwithstanding their ecological importance and ecosystem services, tropical forests are severely threatened by accelerating deforestation, whereas large-scale and long-term monitoring is crucial for the sustainable management of tropical forests. Forest inventories, which have been conventionally relied on field surveys, are highly tedious and laborious. In recent years, with the integration of remote sensing technologies, especially airborne light detection and ranging (LiDAR, also known as airborne laser scanning), forest inventory becomes feasible and cost-effective at larger spatial and temporal scales. LiDAR, which produces 3D point clouds comprising of heigh-density, high-accuracy, and multi-return laser pulses, offers previously unseen information, especially related to canopy structure, which enables novel vegetation analysis techniques. For instance, individual tree segmentation (ITS) based on LiDAR point clouds allows further ecological studies at individual level and from a spatial perspective, such as species identification, community assessment, and species and functional diversity estimation. Although demonstrating robust performance in coniferous forests in temperate and boreal regions, ordinary ITS methodologies, such as the watershed algorithm, show major deficiencies in structurally complex tropical forests, including biodiverse and multi-layered woodlands and densely clustered mangroves. Despite some existing studies demonstrate that post-segmentation processing, which demands extensive knowledge of forest attributes, or the use of spatially restricted terrestrial laser scanning technology could potentially enhance the ITS performance in tropical forests, the notion of attaining efficient forest inventories at larger scales would be diverted, and thus remains challenging. Therefore, by employing Tai Po Kau Nature Reserve and Mai Po Nature Reserve as the representatives of tropical woodlands and mangroves respectively, this thesis aims to 1) evaluate the association between canopy structure and biodiversity in dense and species-diverse woodlands, 2) quantify the functional traits of mangroves, and thus the spatial patterns of functional diversity and composition along the zonation, and 3) investigate the spatio-temporal process of biological invasion in mangrove ecosystems, and its impacts on the native community.