Abstract
<title>Abstract</title> <p>Forest restoration is a crucial natural climate solution, yet its long-term carbon benefits remain uncertain. Globally, three main strategies are adopted: natural regeneration (NR), assisted natural regeneration (ANR), and active restoration (AR). We integrated 36 years of aboveground carbon (AGC) dynamics derived from lidar, Landsat, and forest monitoring plots across East Africa and applied a a quasi-experimental framework to quantify the additionality of active interventions (including AR and ANR) for AGC change over time. We found active interventions commonly exhibit early-stage setbacks in AGC accumulation relative to NR, reflecting site preparation and seedling establishment. However, their impact strengthens over time, becoming pronounced > 9 years after restoration. Long-term AGC gains attributable to ANR and AR range from 0.88 to 3.31 Mg C ha-1 yr-1, with AR performing best in environmentally constrained dry and montane systems and ANR in productive forests facing land-use competition. Optimizing restoration across 14.45 million hectares in East Africa—2.1% of global potential restoration areas—could enhance carbon removal by 2.18 ± 0.36 Gt C by 2050, although less than one-quarter would be realized by 2030. Restoration benefits are therefore strongly time-dependent, and sustained commitment is essential to realize their climate mitigation potential and enhance restoration efficiency.</p>