Remote sensing and vegetation indices applied to the monitoring of long-term and large-scale biodiversity quality

Abstract

Brazilian tropical forests provide rich system in which to place biodiversity studies into a global context. Accurate assessment, however, is often hampered by the amount of data available over the necessary spatial and temporal scales, which limits the ability to make predictions to aid in management decisions. Remote sensing, specifically satellite imagery, has the potential to drastically increase the number of data that can be reasonably collected over ground level collection alone and can fill gaps over space and time. Land use/land cover (LULC) classification maps generated from these satellite images at varying spatial scales are useful in countless areas of scientific research, land use planning, and management efforts. While this technology has allowed for vast numbers of data points to be recorded, generating accurate classification maps or vegetative health maps of forested areas remains challenging, yet still desirable. The use of field spectral measures provides researchers with tools to generate supervised classification and vegetative ground-truthing when applied to the very same satellite images and will yield quantitative accuracy assessment measures. In the present study, we will combine the expertise of Prof. Milton Ribeiro on landscape ecology, Prof. Bernnatz on remote sensing, Prof. Mauro Galetti on biodiversity and plant-animal interactions, Prof. Maria Luisa Jorge on field ecology and animal movement and Prof. Marina Corrêa Côrtes on spatial genetics of plants to generate accurate classification maps and vegetative "health" maps of distinct landscape scenarios of tropical terrestrial ecosystems and correlate them with several biodiversity indices. Such maps may help to further predict biodiversity patterns in future scenarios of climate and land cover changes. (AU)