Exploring the relation between remotely sensed vertical canopy structure and tree species diversity in Gabon

Suzanne Mariëlle Marselis,David Kenfack,Alfonso Alonso,Simon L Lewis,Nicolas Labrière,Seung-Kuk Lee,Ralph Dubayah,Kathryn Jeffery,Hervé Memiaghe,Hao Tang,John Armston,John R Poulsen,Nicolas Barbier,Katharine Abernethy,Lee White,Pulchérie Bissiengou

doi:10.1088/1748-9326/ab2dcd

Abstract

Mapping tree species diversity is increasingly important in the face of environmental change and biodiversity conservation. We explore a potential way of mapping this diversity by relating forest structure to tree species diversity in Gabon. First, we test the relation between canopy height, as a proxy for niche volume, and tree species diversity. Then, we test the relation between vertical canopy structure, as a proxy for vertical niche occupation, and tree species diversity. We use large footprint full-waveform airborne lidar data collected across four study sites in Gabon (Lopé, Mabounié, Mondah, and Rabi) in combination with in situ estimates of species richness (S) and Shannon diversity (H′). Linear models using canopy height explained 44% and 43% of the variation in S and H′ at the 0.25 ha resolution. Linear models using canopy height and the plant area volume density profile explained 71% of this variation. We demonstrate applications of these models by mapping S and H′ in Mondah using a simulated GEDI-TanDEM-X fusion height product, across the four sites using wall-to-wall airborne lidar data products, and across and between the study sites using ICESat lidar waveforms. The modeling results are encouraging in the context of developing pan-tropical structure-diversity models applicable to data from current and upcoming spaceborne remote sensing missions.

Highlights

Spatial information on tree species diversity is important to enable effective conservation and biodiversity management (Turner et al 2003) and allow for a better understanding of scale dependent relationships between forest composition and productivity (Luo et al 2019)
Huston (1979) pointed out that there is a large body of literature regarding the relation between forest structure and the diversity of different taxa, while only few studies explore the relation between forest structure and tree species diversity itself
We first provide the results of the linear models using canopy height alone to predict S and H′

Summary

Introduction

Spatial information on tree species diversity is important to enable effective conservation and biodiversity management (Turner et al 2003) and allow for a better understanding of scale dependent relationships between forest composition and productivity (Luo et al 2019). Those are time-consuming and expensive, which often prevents extensive and spatially representative coverage Using such inventories in combination with remotely sensed data is one approach to overcome some of these limitations and relationships between remotely sensed environmental data (e.g. mean annual temperature, annual precipitation, dry season length, etc) and vegetation species have been developed to map Amazonian and Global vascular plant diversity (Ter Steege et al 2003, Mutke and Barthlott 2005). These approaches incorporate little information on the vegetation itself and the data products are provided at spatial resolutions (typically ∼100 km grid cells) not optimized for conservation management.

Objectives

Methods

Results

Discussion

Conclusion