Answers are grounded in this paper and cite their sources.
Vegetation monitoring can inform pastoral land management only when changes in greenness, seasonal vegetation and mapped cultivation are distinguished. This study reanalysed Landsat-derived vegetation summaries for 2002, 2013 and 2024, five Sentinel-2 monthly vegetation summaries for 2024, and land-cover statistics for ten grazing reserves in Adamawa State, Nigeria. Relative change, annualised loss, loss-to-gain ratios, area-weighted farmland prevalence, concentration and deterministic classification-error scenarios were calculated. High-NDVI cover declined from 337,020.66 to 213,629.85 ha, a 36.61% reduction rather than the previously reported 50%. About 93.76% of this net loss occurred during 2002–2013. Within the pairwise change domain, decreases exceeded increases by 19.57 times over 2002–2024. The ten reserves contained 6,446.07 ha of mapped farmland, equivalent to 17.76% of their reported combined area. Dauchi, Guyaku and Ladde Dan-Kiffi accounted for 61.07% of mapped reserve farmland while occupying 44.26% of assessed reserve area. A conditional reserve-level permutation test found limited size–prevalence association (rho = 0.406; p = 0.248), while concentration equalled 5.78 equally contributing reserves. Monthly vegetation presence ranged from 20.19% in May to 96.74% in November. However, mismatches between pixel counts, area conversion and land-cover summaries prevent the seasonal classification from being treated as an independently validated cultivation map. The results support targeted verification of concentrated cultivation and monitoring of high-NDVI loss, while demonstrating why greenness change alone cannot establish deforestation, unlawful encroachment or conflict causation
No takes yet. Share an insight, caveat, or question.
Sarkinzango et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: