Post-fire assessment of recovery of montane forest composition and stand parameters using in situ measurements and remote sensing data

Abstract

Tree diversity in the tropical forests face escalating threats from wildfires. This study assessed post-fire impacts and recovery patterns in Tanzanian Eastern Arc Mountains forests from 2017 to 2022 using field measurements and remote sensing techniques. Tree species diversity, composition, and forest stand parameters were compared between burned and unburned forest plots across two reserves. A predictive model utilizing 14 key variables derived from multispectral satellite data was developed to accurately map burned areas and spatial fire patterns. Results revealed significantly lower tree density, aboveground biomass, species richness, and Shannon diversity in burned areas compared to unburned forests. However, compositional analysis showed extensive species overlap between burned and unburned sites, with burned areas containing more indicative pioneer and disturbance-adapted species such as Apodytes dimidiata. Over time since fire events, tree density, basal area, aboveground biomass, species richness, evenness, and diversity increased markedly, evidencing active tree recovery. The remote sensing model effectively delineated approximately 1430 hectares of burned areas concentrated near villages, suggesting prevalent anthropogenic fire ignitions. Although wildfires substantially impacted forest structure and biodiversity, the limited compositional shifts point to resilience of these tropical montane forests. Integration of diverse spectral bands and textural metrics from multispectral satellite data can support precise mapping of fire effects and forest recovery dynamics in these ecologically vital yet threatened ecosystems, aiding conservation and management. Overall, this study provides novel insights into post-fire responses in Eastern Arc Mountain forests using a synergistic field and remote sensing approaches.