Increasing proportion of mildly aged population in rural mitigates farmland abandonment in the farming-pastoral ecotone of northern China.

Rural population aging has emerged as a widespread phenomenon, which can lead to farmland abandonment and pose unprecedented challenges to agricultural production and ecological sustainability in the farming-pastoral ecotone of northern China (FPENC). The dynamic changes in farmland abandonment from 2000 to 2020 were systematically explored using a trajectory-based land use change detection approach. Binary logit regression models were employed to analyze the driving mechanism of the current farmland abandonment based on 1,195 questionnaires, and then random forest models were used to predict future farmland abandonment trends under various scenarios. The results showed that (1) rural population aging had emerged as a significant challenge for Ulanqab, with the proportion of people aged 60 and above increasing from 12.3% in 2000 to 44.9% in 2020. Members of mildly aged households (60-69) were identified as the main agricultural labor force, accounting for 46.6%; (2) an overall downward trend of farmland abandonment was observed, decreasing from 295 km2 in 2000 to 273 km2 in 2020. However, the abandonment rate increased slightly from 3.03% to 3.66%, with higher abandonment rates concentrated in northern Ulanqab; (3) an increase in the proportion of the mildly aged population mitigated farmland abandonment, while an increase in the proportion of the severely aged population (≥70) exacerbated it. When other conditions remained unchanged, a 5% and 10% increase in the proportion of the mildly aged population corresponded to a decrease in farmland abandonment rates to 11.4% and 10.5%, respectively; (4) the mechanisms underlying abandonment behavior differed between young and elderly households. For each additional year of age for elderly households, the probability of farmland abandonment increased by 4.1%. Elderly households with higher levels of education, fewer farm laborers, smaller per capita farmland area, a higher proportion of dryland, larger quantities of farmland, and non-contracted farmland were more likely to abandon farmland; (5) under three Shared Socioeconomic Pathway scenarios (SSP1, SSP2, and SSP5), the proportion of abandoned farmland by farmers was projected to rise to 58.8%, 20.1%, and 58.8%, respectively. These findings provided new insights into the issue of abandoned farmland, particularly from a demographic perspective.