Characterizing seasonal dynamics of Amazonian wetlands for conservation and decision making

Abstract

Abstract In many wetlands the timing and duration of inundation determine ecological characteristics and the provision of ecosystem services; however, wetland conservation decisions often rely on static maps of wetland boundaries that do not capture their dynamic hydrological variability and connectivity. The Amazon River basin contains some of the world's most extensive wetlands, many of which are floodplains where seasonal flood pulses result in a temporally varying inundation area and hydrological connectivity with river systems. This study classified Amazon wetlands according to the timing and duration (months per year) of inundation detected by remote sensing, and also investigated the contribution of precipitation regimes in affecting wetland distribution and hydrological dynamics. Permanently inundated wetlands account for the largest area and are mainly floodplains located in the lowlands of the catchment. Seasonally inundated wetlands varied greatly in the duration of inundation over the course of the year, ranging from 1 to 9 months. Distinct seasonal timing was detected among the large wetland complexes, reflecting rainfall regimes as well as time lags for drainage and drying. For example, inundation in the extensive Llanos de Moxos region of the southern Amazon was protracted and lasted well after the rainy season, compared with the Roraima region of the northern Amazon, where inundation was shorter and tracked the rainy season. The integration of inundation dynamics into wetland classification captures regional differences in timing and duration of inundation in the major wetlands of the basin that should be considered for conservation planning and other ecological applications. This information can aid regional wetland management and planning, especially with regards to minimizing the effects of dam and waterway construction that can directly affect the natural wetland dynamics. The use of global remotely sensed inundation data makes this approach easily transferable to other large tropical wetlands.