Recovery of Miombo Woodland in Abandoned Lands Post-Agriculture, Charcoal Production and Logging in Central Mozambique

Agricultural activities, charcoal production and commercial logging are the primary drivers of deforestation and forest degradation in Miombo woodlands (MWs), yet the abandonment of these activities can allow for ecological recovery. While previous research has largely focused on MWs recovery following agricultural disturbances, the regrowth pathways and ecological impacts of charcoal production and logging remain poorly understood. This study fills that gap by analysing vegetation recovery within a common framework across smallholder agriculture, charcoal production and commercial logging in central Mozambique, and examining the environmental and socio-economic factors that influence recovery across these disturbance types. Using remote sensing data, we identified three sites for each disturbance type and conducted forest inventories in abandoned lands on trees with diameter greater than 5 cm, applying a chronosequence approach, along with undisturbed MWs. We assessed vegetation structure parameters, species diversity metrics and conducted plot pairwise similarity comparisons for each disturbance type. A generalised additive model was used to evaluate the effects of abandonment age and external factors on vegetation structure, species diversity and composition. Our findings show that MWs recover more quickly in vegetation structure (in 30 years) than in species diversity and composition. Species composition trajectories differed among disturbance types, with Sørensen dissimilarity showing significant temporal change only in charcoal stands; although species turnover and nestedness remained largely stable over time, turnover contributed more to overall compositional dissimilarity than nestedness, suggesting that shifts were more associated with species replacement than richness differences. Species richness and the effective number of species showed no significant change with abandonment age across any disturbance type. Charcoal and logging stands displayed greater species diversity convergence compared to agricultural stands. External factors, such as proximity to temporary rivers, villages and secondary roads, as well as fire frequency, influenced recovery, with impacts varying based on disturbance type. Our findings emphasise the complexity of Miombo woodland recovery and highlight the need for long-term monitoring and management to support the resilience of these ecosystems in the face of ongoing disturbances. Future research should focus on longer timeframes, larger sample sizes, and include the natural regeneration to better understand the mechanisms behind vegetation recovery in disturbed MWs.