Combining ALOS/PALSAR derived vegetation structure and inundation patterns to characterize major vegetation types in the Mamirauá Sustainable Development Reserve, Central Amazon floodplain, Brazil

Abstract

Remote sensing studies of vegetation cover and hydrologic dynamics in Amazonian wetlands have been mostly limited temporally or spatially, and the distribution and spatial configuration of Amazonian várzea habitats remains poorly known. This study uses multitemporal PALSAR L-band radar imagery combined with object-based image analysis, data mining techniques and field data to derive vegetation structure and inundation patterns and characterize major vegetation types in várzea forests of the Mamirauá Sustainable Development Reserve. Our results show that the combination of vegetation cover and inundation extent information can be a good indicator of the complex gradient of habitats along the floodplain. The intersection between vegetation and flood duration classes showed a wider range of combinations than suggested from field based studies. Chavascal areas—chacaracterized as a dense and species-poor shrub/tree community developing in old depressions, abandoned channels, and shallow lakes—had shorter inundation periods than the usually recognized hydroperiod of 180–240 days of flooding, while low várzea—a diverse community that have fewest and smallest species, and highest individual density and that tolerate 120–180 days of flooding every year—was distributed between flood duration ranges that were higher than reported by the literature. Forest communities growing at sites that were never mapped as flooded could indicate areas that only flood during extreme hydrological events, for short periods of time. Our results emphasize the potential contribution of SAR remote sensing to the monitoring and management of wetland environments, providing not only accurate information on spatial landscape configuration and vegetation distribution, but also important insights on the ecohydrological processes that ultimately determine the distribution of complex floodplain habitat mosaics. © 2014, Springer Science+Business Media Dordrecht.