Revisiting the progress of natural afforestation on abandoned farmlands using long-term satellite time observations

Farmland abandonment is among the most widespread land-use change processes in Europe and is often assumed to promote natural afforestation through vegetation succession, particularly in temperate regions. However, empirical evidence quantifying the extent and controls of afforestation on abandoned farmlands remains limited. Using Lithuania as a case study of post-Soviet land-use transition, satellite Landsat, Sentinel-2 time-series were combined with spatial analysis to quantify farmland abandonment between 1990 and 2000 and subsequent land-cover trajectories through 2025. Further, it assessed the extent to which abandoned farmland reverted to forest and identified key biophysical and socio-economic determinants shaping these dynamics. The results indicated that approximately 25% of farmland was abandoned during the early post-Soviet period, yet only a portion of this land remained abandoned by 2025. Among the remaining abandoned areas, only a small fraction exhibited spectral convergence with adjacent natural forest, suggesting limited progression toward mature forest states. Reversion to forest was strongly conditioned by accessibility, socio-economic factors, seed dispersal potential, and biophysical constraints. In contrast, recultivation of abandoned farmland was promoted by favorable cultivation conditions, agricultural subsidies, and land-use interventions such as the designation of hunting grounds. Overall, the findings challenge the assumption of widespread, spontaneous forest recovery on abandoned farmland and demonstrate the value of long-term Earth observation data for disentangling land-use trajectories and their controlling factors, as well as the formation of novel scapes, neither representing agriculture nor newly established forests. The proposed analysis framework is expandable and transferable to other regions experiencing land-use transitions, thus helping better quantify natural and socio-economic potentials.