Estimation of aboveground biomass recovery through chronosequence in forests degraded by fire in the Legal Amazon

The Amazon biome is under constant pressure from deforestation and fire occurrence, one of the most active forest degradation processes. The advance of deforestation leads to the increase of forest edge effects. Thus, agricultural management based on slash-and-burn practices can lead to fire escaping into native vegetation, leading to forest degradation, impacting biodiversity, forest structure, carbon stocks and emissions. Maranhão state, located in northeastern Brazil and part of the Legal Amazon, encompasses a  transition from the Amazon rainforest to Cerrado. Attention to this region is urgent due to growing pressures related to fire and deforestation mainly within Protected Areas (PA), threatening the conservation and functioning of this unique ecosystem. An up-to-date spatial explicit diagnostic of disturbances such as fire, deforestation and edge effects is important for formulating protective measures for these areas.Methodologies for quantifying Greenhouse Gas (GHG) emissions and removals, analyzing trends, attributing sources and sinks are key to support the establishment and reporting of national GHG inventories. Brazil has legal tools, like the National Climate Change Policy, aligned with Paris Agreement goals, emphasizing the Reduction of Emissions from Deforestation and Degradation (REDD+). Quantifying carbon losses from degradation is challenging due to uncertainties in estimating degraded forest areas and disturbance impacts. About 61% of carbon removals occur in protected native vegetation, yet estimates may be overestimated due to unaccounted forest degradation processes, like burn emissions from non-deforested native vegetation, untracked in National Inventories. These uncertainties, however, can be reduced by combining field measurements with an ever-increasing range of datasets and remote sensing methods. This study aims to enhance understanding of post-fire biomass growth dynamics and recovery potential, emphasizing the pivotal role of carbon removal by vegetation.Our analysis covers the heterogeneous spatial and temporal patterns of vegetation growth in fire-degraded forests, where we combined a satellite dataset tracking fire disturbances with the fusion of 3 products widely used in other studies (MCD64A1, Fire_CCI and Mapbiomas Collection 2, fusion product with 30 m resolution), with a global above-ground biomass (AGB) product (Biomass_CCI, 100 m resolution) in a space-for-time substitution approach to model accumulated AGB as a function of the Years Since the Last Fire Disturbance (YSLF). Over 20 years of recovery (2001 - 2020), regeneration rates in areas degraded by forest fires ranged from 2 to 12% per year, totalling up to 80% biomass recovery, compared to old forests that were never burned. Degraded forests are most severely disturbed after the first YSLF, where AGB is reduced to 58% of the median AGB of old-growth forests (113.86 Mg/ha), resulting in a 42% loss of biomass. In 2016, fires breached Maranhão's protected areas for the first time in two decades. Even after a single fire event, the areas did not fully recover in terms of biomass, indicating a potential reduction in carbon storage capacity.Extreme fire events amplify these occurrences, affecting protected areas and decreasing the carbon storage potential of forests. Urgent measures are needed to protect and restore these areas, recognizing the lasting impacts of forest fires on biodiversity, forest structure and carbon emissions.