Remote sensing of subtropical tree diversity: The underappreciated roles of the practical definition of forest canopy and phenological variation

作者：Yongchao Liu, Ruyun Zhang, Chen‐Feng Lin, Zhaochen Zhang, Ran Zhang, Kankan Shang, Mingshui Zhao, Jingyue Huang, Xiaoning Wang, You Li, Yulin Zeng, Yunpeng Zhao, Jian Zhang, Dingliang Xing · 发表于：Forest Ecosystems · 年份：2023 · DOI：10.1016/j.fecs.2023.100122 · 被引用次数：15 · 研究领域：Remote Sensing in Agriculture、Species Distribution and Climate Change、Ecology and Vegetation Dynamics Studies

Tree species diversity is vital for maintaining ecosystem functions, yet our ability to map the distribution of tree diversity is limited due to difficulties in traditional field-based approaches. Recent developments in spaceborne remote sensing provide unprecedented opportunities to map and monitor tree diversity more efficiently. Here we built partial least squares regression models using the multispectral surface reflectance acquired by Sentinel-2 satellites and the inventory data from 74 subtropical forest plots to predict canopy tree diversity in a national natural reserve in eastern China. In particular, we evaluated the underappreciated roles of the practical definition of forest canopy and phenological variation in predicting tree diversity by testing three different definitions of canopy trees and comparing models built using satellite imagery of different seasons. Our best models explained 42%–63% variations in observed diversities in cross-validation tests, with higher explanation power for diversity indices that are more sensitive to abundant species. The models built using imageries from early spring and late autumn showed consistently better fits than those built using data from other seasons, highlighting the significant role of transitional phenology in remotely sensing plant diversity. Our results suggested that the cumulative diameter (60%–80%) of the biggest trees is a better way to define the canopy layer than using the subjective fixed-diameter-threshold ...