Feeding the Future: Agroecological Pathways to Climate Adaptation and Food Security

Soil erosion by water remains one of the most consequential forms of land degradation, and projections of intensifying rainfall erosivity have renewed interest in tree-based farming systems as a means of protecting soils and stabilising catchment hydrology. Agroforestry is frequently presented as a dual-purpose intervention that reduces soil loss while improving the water-regulating capacity of watersheds, yet the evidence supporting this proposition is uneven in quality, strongly concentrated at the plot scale, and rarely tested under the extreme rainfall conditions that dominate long-term sediment budgets. This review critically evaluates the state of knowledge on agroforestry as an erosion-control and watershed-resilience intervention, drawing on peer-reviewed literature retrieved through open scholarly databases and indexes and appraised for methodological adequacy, consistency and relevance. The synthesis is organised around four analytical problems: the mechanistic basis of erosion control in tree-based systems; the dependence of measured effectiveness on system configuration; the transferability of plot-scale effect sizes to catchment sediment yield; and the trade-off between improved infiltration and increased evapotranspiration in water-limited settings. The available evidence indicates that agroforestry reliably reduces interrill erosion and surface runoff where a dense ground layer and continuous litter cover are maintained, with quantitative syntheses reporting approximately halved soil loss relative to annual monocultures in humid and sub-humid tropical settings. Confidence declines sharply for concentrated-flow erosion, catchment sediment yield, low-flow maintenance and performance under high-magnitude storms, where measurements are scarce, methodologically heterogeneous and frequently confounded by simultaneous changes in tillage, residue management and grazing pressure. The persistent identification of tree canopies as the operative mechanism is not well supported; ground-layer cover and soil structural modification appear more decisive. Priority research needs include paired-catchment and long-term monitoring of tree-based systems, event-resolved measurement across storm magnitudes, explicit representation of sediment connectivity in evaluations, and quantification of the conditions under which infiltration gains outweigh transpiration losses. Agroforestry warrants continued policy support as a soil-conservation practice, but claims regarding catchment-scale water security remain conditional and require context-specific verification.