A geospatial data integration framework for mapping and monitoring tropical landscape diversity in Costa Rica's San Juan-La Selva Biological Corridor

Abstract

Landcover change has substantially reduced the amount of tropical rain forests since the 1950s. Little is known about the extent of remaining forest types. A multivariate analysis of 144 forest plots across Costa Rica's San Juan-La Selva Biological Corridor resulted in eight floristically defined old-growth forest categories. Spectral separability was tested between categories using Landsat TM bands and vegetation indices for old-growth types, palm swamps, tree plantations and regrowth. Image filtering and NDVI increased spectral separability among categories by 30 per cent. Separability tests resulted in seven well-discriminated forest categories. Factors driving forest beta-diversity are not well quantified for wet tropical environments. We examined the relationship between rain forest composition and environmental variation for a 3000 km2 area in northeastern Costa Rica.