Alterações no pulso de inundação: como reagem os igapós da Amazônia Central?

Abstract

Igapós are floodplain forests inundated by acidic and nutrient-poor waters and occur along the black-water rivers, such as the Negro River and its tributaries, the Uatumã River and others in the Amazon Basin, covering more than 140,000 km2. Under natural conditions, the flood-pulse is monomodal and predictable, which drives the specialization and dependence of the flora to semiaquatic conditions. This work sought to understand how the igapós of two sub-basins in Central Amazonia perform under 1) normal conditions; 2) under the influence of extreme hydroclimatic events which increased dramatically during the last three decades (Jaú River, Negro River tributary); and 3) under flood-pulse disturbances (Uatumã River, Amazon River tributary) where the Balbina hydroelectric power plant was installed in the 1980s. For this, the igapó forests downstream the Uatumã River were classified and mapped to identify areas with massive tree mortality and where mortality is still occurring (first chapter). The mapping was done using 56 synthetic aperture radar images (ALOS/PALSAR) acquired at different flood levels between 2006 and 2011. The results showed that the application of object-based image analysis (OBIA) methods and the random forests supervised classification algorithm generated an overall accuracy of 87.2%. The mortality of 12% of the igapó forest was observed after the first 43 km downstream, along a stretch of 80 km. Based on the images and hydrological data, we detected 29% of the living igapó forests might be currently suffering mortality. Low topographies in the floodplains characterized by macrohabitats dominated by the tree species Eschweilera tenuifolia (O. Berg) Miers (Lecythidaceae) were the most impacted. To further study the ecology of this species, chapters 2 and 3 compare the macrohabitats in the two study areas to understand how and when flood pulse disturbances affected the growth and mortality of populations of this species, highly adapted to prolonged flooding conditions with flood durations of up to 10 months per year on average. Therefore, we collected discs of 30 dead individual trees and cylindrical cores (two per tree) from 62 living individuals in the Uatumã River igapós. In forests adjacent to the Jaú River, with natural tree mortality, we collected 21 discs from dead individuals and 31 cores from living trees. The second chapter discusses how the relationship between tree growth and climatic factors changed with the intensification of the hydrological cycle in natural environments (Jaú National Park - JNP) and the disturbed system (Uatumã Sustainable Development Reserve - USDR). For that, tree-ring chronologies were constructed at each study site, which presented robust statistics for the periods 1927-1999 (JNP) and 1920-2006 (USDR). The chronologies were related to regional parameters of hydrology (monthly water level) and climate (monthly precipitation) and with sea surface temperature anomalies (SST) of the Equatorial Pacific and Tropical Atlantic. The results suggest strong changes in the responses of Eschweilera tenuifolia trees to the intensification of the hydrological cycle, characterized by the increase in the frequency and magnitude of severe floods in the natural system (JNP) and by the increase of the minimum water level during the operational period of the Balbina dam (USDR). In the JNP, the trees indicate an increasing strength in the responses to the SST anomalies of both oceans, which reflects the described mechanisms that results in the intensification of the hydrological cycle, caused by the warming of the Tropical Atlantic and the simultaneous cooling of the Equatorial Pacific resulting in an intensification of Walker's circulation, strong cloud convection and increased precipitation in the northern and central sections of the Amazon Basin. In the USDR, the impacts of changes in the hydrological regime induced by the Balbina dam resulted in opposite responses of tree growth compared to the pre-dam period. The third chapter shows that the mean diameter increment (MDI) and mean age for trees in the JNP are 2.04 ± 0.39 mm and 201 ± 103 years, respectively, while trees at the USDR has MDI of 2.28 ± 0.69 mm and a mean age of 213 ± 103 years. The cumulative growth trajectory is similar between both areas for a 500-year period of growth modeling and the MDI analysis by periods indicates that overall the tree growth did not differ between both sites in the period before the construction of the Balbina hydroelectric power plant. In the same way, tree growth at JNP did show any significant difference between the periods before and after the intensification. Opposingly, the alteration of the flood pulse, caused by the operation of the dam, resulted in the MDI differentiation between the two periods at the USDR. The trees at USDR died about ten years after the beginning of the dam construction, whereas those from the JNP did not show a clear pattern throughout the evaluated periods. The living trees of USDR showed an increase in growth between 1982 and 2000 and an abrupt and strong growth decline after the year 2000, continuing until present day induced by the permanent flooding conditions leading to growth suppression and mortality. The change in the flood-pulse, more than 30 years after the disturbance, continues to cause growth anomalies and mortality in Eschweilera tenuifolia, which is highly sensitive to changes in flood amplitude and duration. In summary, the findings indicate the excessive mortality of trees is still occurring in the floodplains of the Uatumã River basin affecting especially macrohabitats dominated by this species, causing losses of ecosystem services and regional extinction of an extraordinary tree species. It is necessary that laws for the installation and norms for the operation of hydroelectric power plants in the Amazon should be carefully reviewed based on studies done in wetland areas downstream of dams.