THE INTEGRATION OF FIELD SURVEY AND REMOTE-SENSING FOR BIODIVERSITY ASSESSMENT - A CASE-STUDY IN THE TROPICAL FORESTS AND WETLANDS OF SANGOBAY, UGANDA

Field surveys of plants and animals were combined with satellite remot e sensing of broad vegetation types to map biodiversity and thereby he lp plan conservation in the Sango Bay area, some 30 by 100 km borderin g Lake Victoria in Uganda. A statistical classifier applied to satelli te images identified 14 land-cover classes including water, swamp, dry grasslands, degraded woody vegetation, semi-natural forest classes an d intensive land uses. Validation, using 240 sample sites, recorded 86 % correspondence between field and map data. Intensive land use makes up 23% of the area, water and swamps 27%, dry grasslands 29%, woody ve getation 21%, with semi-natural forests covering 15% of the area. The species data from sample-based field surveys included flowering plant species, dragon/lies, butterflies, fish, amphibians, reptiles, birds a nd mammals. The species data were used to generate biodiversity rating s, based on species 'richness' and 'rarity', which could be related to the vegetation cover. This inter-relation helped to generate a biodiv ersity map of the Sango Bay area which has since been used to aid cons ervation planning. (C) 1998 Elsevier Science Ltd. All rights reserved.