Using Satellite Data to Map Biodiversity and Support Ecosystem Services in Anguilla
Cameron, Iain; Medcalf, Katie
Environment Systems Ltd, UNITED KINGDOM

The terrestrial environments of the Anguillian archipelago are considered fragile. Since the mid 1980's, the tourism industry has been the main contributor to the socio-economic development in the country. The tourism industry and ad-hoc infrastructure development due to poor legislative framework have contributed to the degradation and loss of valuable habitats in Anguilla. It is therefore, pivotal to have a baseline habitat database in order to monitor the island’s biodiversity and feed into decision making to help ensure the sustainable management of the natural resources of the country. This paper outlines a project conducted by Environment Systems on behalf of the Government of Anguilla to deliver baseline habitat mapping of the country using Earth observation (EO) data.

A rich multi-season dataset was collected that targeted known seasonal vegetation cycles and included SPOT-5, and Landsat-5 images, Lidar and aerial photography. This data was used, together with expert habitat knowledge and field observations, to develop a new hierarchical set of habitat classes that share functional components based on the density and productivity of the vegetation layers. Focusing on functional rather than species composition provides a good starting point to highlight features of pertinence to the ecosystem in general and the green economy. These habitat classes were mapped using object-oriented image analysis and an ecological rule-base classification. Field validation demonstrated that the new habitat classification has an accuracy of 80% for the major classes, with the major errors being observed for habitats which occur at a fine spatial scale, such as coastal ecotones where distinct changes in vegetation assemblage occurs at scales similar to or below satellite resolutions. The operational techniques used in this project are robust and repeatable and can be used to monitor change and to input into further analysis of ecosystem features such as the regulation of surface water runoff.