Áreas prioritárias para prevenção de incêndios florestais na caatinga

Abstract

The Caatinga, an exclusively Brazilian biome located in the northeast of the country, is the largest seasonally dry tropical forest in the world. In this environment, climatic and environmental factors, associated with extensive human exploitation, contribute significantly to the degradation of the ecosystem, making it one of the most threatened on the planet. Although historically considered independent of the action of fire, over the years, the Caatinga has been constantly affected by forest fires, which can transform it into a sensitive biome. In this context, this study was subdivided into two chapters with the following objectives: Chapter I: to analyze the behavior of forest fires in the Caatinga over a 10-year period (2011-2020), associating these events with the environmental characteristics that can increase or reduce the risk; Chapter II: to propose a forest fire risk zoning for the biome and, based on this, delimit areas with the greatest need for preventive actions against forest fires, in order to provide conservation of local biodiversity. To achieve these objectives, geotechnology, statistics and computational tools were used. Over the years analyzed, 2015 was the year with the largest burned areas. The state of Piauí was spatially the most vulnerable to fire. Regarding environmental characteristics, regions with altitudes below 600 m; air temperature above 24 °C; rainfall index below 900 mm; drought severity ranging from mild to extreme; potential evapotranspiration above 1,500 mm; vegetation index by normalized difference of low to moderate vigor; leaf area index below 3.5 and Savanna Formation are the ones that most drove the occurrence of fires in the Caatinga. The topographic variables, slope and orientation of the relief alone, did not demonstrate characteristics capable of indicating greater or lesser propensity for fire. Regarding risk modeling, the use of Fuzzy Logic associated with the Analytical Hierarchy Process (AHP) technique allowed for accurate modeling for the studied biome, with 96.6% accuracy, classifying the Caatinga as having a high to very high risk of forest fires. The northern and western portions of the Caatinga, especially the borders with the Cerrado biome, were the regions where the model indicated a very high risk of fires. As observed in the history of occurrence, the state of Piauí had the greatest representation of areas for fire prevention, followed by Bahia and Ceará. In general, vegetation located on private properties has a greater need for protection (92.02%) due to fragmentation and proximity to agricultural activities. Conservation units and indigenous lands encompassed 7.98% of the areas where forest fire prevention measures should be adopted, with emphasis on the sustainable use category, Environmental Protection Area modality. In this context, these results indicate the need for public management to prevent the occurrence of fires in a Biome previously considered independent of fire, especially in private areas and in the Caatinga-Cerrado transition.