Agro-landscapes under armed conflicts: global research trends and recovery priorities (2000–2026)

Armed conflicts are among the most disruptive forces acting on agro-landscapes, yet their impact on land-use systems and ecosystem services remains insufficiently integrated into landscape ecology and agroecological research. This paper aims to systematise the international body of knowledge on the mechanisms of war-induced impacts on agro-landscapes, to identify reproducible causal patterns of ecosystem service degradation, and to propose methodological priorities for comprehensive assessment and monitoring of agro-landscapes under active armed conflict and post-conflict recovery. The study is based on a bibliometric analysis of 1,502 publications (WoS/Scopus, 2000–2026) and a comparative analysis of five literature clusters reflecting the evolution of research paradigms from foundational ecological concepts to conflict-specific environmental impact studies. The synthesis shows that conflicts disrupt agro-landscapes through direct destruction, forced displacement, land abandonment, and institutional collapse, triggering cascading degradation of provisioning, regulating, supporting, and cultural ecosystem services. Resilience theory and socio-ecological systems frameworks are identified as essential but systematically underutilised analytical lenses. A pronounced geographic bias favouring East Africa and Latin America was detected, with Eastern Europe and the Middle East comparatively understudied — a gap being rapidly addressed by the Ukrainian case post-2022. Remote sensing has emerged as the dominant methodology for assessing conflict-induced land-use change, yet integration with socio-economic and institutional dimensions remains limited. The paper argues that the next priority is the development of coupled models — Earth observation + field validation + ecological-economic assessment + scenario analysis — capable of underpinning evidence-based agro-landscape recovery policy under conditions of prolonged uncertainty.