Queimadas e incêndios florestais na região do Parque Estadual do Jalapão no período de 2000 a 2024

Abstract

The Cerrado, a biodiversity-rich biome historically shaped by fire, was subjected for many years to a policy of total fire exclusion ("Zero Fire"), in force until 2013, which, by disregarding the inevitability of fire, favored fuel accumulation and larger, more severe fires. From 2014 onward, Integrated Fire Management (IFM) introduced prescribed burns to redistribute fire seasonally. This study characterized the fire regime, its relationship with interannual climate variability, and the effect of the transition between the two management periods across six protected areas of the Jalapão Mosaic and their surrounding Buffer zone (Tocantins), from 2000 to 2024. Monthly data on burned area (MapBiomas Fire, Collection 4), precipitation (CHIRPS), and land surface temperature (MODIS) were used, from which four fire metrics were derived (burned area, ignition density, mean scar size, and the size of the largest events). Analyses were conducted on a protected area-year basis using the Mann-Kendall test with Sen's slope, Spearman correlation and Mann-Whitney tests with Benjamini-Hochberg (FDR) correction, bootstrap confidence intervals, a mixed-effects linear model pooling the protected areas, and Interrupted Time Series (ITS) analysis. Burned area showed strong interannual variability and was concentrated in the late season (August–October), which accounted for 67% of the total. At the protected area scale, interannual climate control proved weak and spatially inconsistent: of 140 associations tested, only three survived correction, restricted to APAJA and SPI-6. However, the mixed-effects model revealed a consistent, though modest, climate signal on the extent and size of fire, absent for ignition density, indicating that climate modulates fire spread rather than occurrence. Regarding management, IFM was associated with a reduction in late season share in APAJA (r = −0.49; p = 0.040) and in EESGT (r = −0.51; p = 0.035), with an average reduction of about 13% estimated by the mixed-effects ITS model and no effect detected in the Buffer zone. We conclude that fire at the local scale is dominated by management and land-use factors rather than by interannual climate variability, and that IFM shifts fire away from the most critical period, although the low power of individual time series recommends caution and continued monitoring.