B061-0021
TerraBio: informing adaptive management for conservation success
Abstract:
TerraBio is an innovative tool being developed to address this need. Key research questions include: How and to what extent do different PSE approaches support biodiversity conservation? and, How do sustainable interventions compare with business as usual practices? Currently, TerraBio is being trained to identify different land use and land cover systems at two pilot sites using deep machine learning algorithms and an iterative approach to discriminate between degraded forest edge and agroforestry. We will then quantify pre- and post- intervention landscape patterns, and evaluate biodiversity conservation potential from sustainable systems compared to control sites. We will use multivariate statistics and habitat mapping to integrate in-situ measurements of diversity and assess which initiatives are the most effective in restoring and maintaining biodiversity. We will also create a catalog of context and intervention specific biodiversity monitoring methods and analyses for the evaluation of other PSE supported initiatives.
Here, we will present the need for tools that can assess biodiversity using satellite technologies, TerraBio’s conceptual model, methods, and findings. Early results suggest improved agroforestry classification resolution compared with previous methods. Observed changes in land cover and land use suggest reductions in forest loss and degradation and increases in reforestation from PSE initiatives. Following this pilot, TerraBio will be applied to additional private sector impact investment initiatives. Findings from TerraBio will help distinguish between more and less effective PSE approaches for biodiversity conservation and will guide future PSE investments.