Dinâmica da cobertura florestal e ocorrência de incêndios florestais e suas implicações na gestão da Reserva Extrativista Chico Mendes.

Abstract

The Brazilian Amazon makes up 41% of the world's humid tropical forests and allows Brazil to be a leader in forest issues. However, it also attracts worldwide attention, owing to the continued increase in deforestation, biodiversity loss and ecosystem services. Protected areas have acted as barriers to deforestation and forest degradation, however, this capacity becomes fragile when deforestation agents manage to break the legal protection barrier of these protected areas. In this context, the objective of this work was to understand the dynamics of deforestation and forest degradation in the Chico Mendes Extractive Reserve (RECM), Acre, Brazil, and its implications for the management of this conservation unit. The analysis was divided into two chapters: Chapter I - Dynamics of forest fires in the Chico Mendes Extractive Reserve - Acre and Chapter II - Forest cover dynamics in the Chico Mendes Extractive Reserve: an example of management challenges for conservation units in the Amazon. In Chapter 1historical series of Landsat images from 1984 to 2015 was analyzed and processed by the software Claslite with the application of the BSI equation (Burn Scar Index). An area of 50,363 ha was identified as forest fire scars, an area equivalent to the total deforested area according to Prodes in this conservation unit. The years of peak occurrence of forest fires were 2005 (80%) and 2010 (19%). Eleven polygons greater than 1,000 ha were identified, accounting for 52% of the total burn scar area. The most affected forest types were open forests with bamboo (53%) and alluvial forests (20%). The year 2005 became the milestone in the management of forest fires in the State of Acre and, consequently, for RECM. In years of extreme drought, the forest changes from being the barrier to becoming fuel for forest fires. In Chapter II, we analyzed the available historical series of deforestation data, Prodes (1997 to 2014) and Hansen et al. (2000 to 2014), and the process forest regeneration from 2000 to 2012, measured by Hansen et al. With historical data from Prodes, RECM had 55,948 ha of deforestation by 2014, representing 6% of the RECM area. In an annual analysis with data from Prodes and Hansen, there is a divergence between the years with the highest deforestation rates, with Prodes in 2004 with 4,462 ha/year and Hansen et al. in 2005 with an area of 5,317 ha/year. Three seringals were identified as the most critical to control deforestation in RECM, Nova Esperança (52% of its area deforested), Santa Fe (55% of its area deforested) and Rubicon (42% of its area deforested). These seringals are close to the BR-317 road that connects other Brazilian states to Peru through the Interoceanic road and are being pressured by the expansion of cattle ranches around RECM. Forest regeneration extended to 2,796 ha, representing 10% of the total deforestation by Hansen from 2000 to 2012. RECM has regenerating forests, with 43 of 46 seringals presenting signs of natural regeneration. The analysis of deforestation using two independent methods increases the reliability of deforestation information and assists decision making in controlling land use. This work concludes that forest fires are now an additional threat to the integrity of RECM. In the process of monitoring deforestation, there are still differences between data sets, but the comparison has helped improve the process of monitoring and control, increasing the understanding of the forest cover dynamics of forest cover by using public domain data and software.