Claire Teakle
A modern approach to Amazonian tree species identification through leaf-level reflectance spectroscopy
Name: Claire Teakle
Primary Supervisor: Dr Frederick Draper
Year: 2
Discipline: Geography
Presentation type: Poster
Project Title: A modern approach to Amazonian tree species identification through leaf-level reflectance spectroscopy
Abstract:
There are estimated to be more than 15,000 tree species in Amazonia. This huge diversity makes species identification difficult when using traditional botanical techniques, leading to errors and uncertainty. It is common for closely related species to appear almost identical, seeming indistinguishable without the presence of fertile structures. Unfortunately, flowers, fruit and seeds are scarcely available. Spectroscopy is an alternative method for species identification, however no standardised protocol exists for the collection of foliar spectra. Through fieldwork in Peru, Brazil and French Guiana, I am developing a standardised method, whilst collecting spectra from 5 key Amazon hyperdominant genera: Eschweilera, Inga, Protium, Pouteria and Virola. Using a linear discriminant analysis, I am investigating how accurately Amazonian tree species can be identified using their spectra.