Carbon Sequestration in Agricultural Soils through Agronomic Management: Mechanisms, Evidence, Accounting Boundaries and Research Priorities

Abstract

Agricultural soils have lost substantial organic carbon through land conversion, intensive cultivation and insufficient organic inputs, creating a technically important but highly context-dependent opportunity for carbon restoration. This critical narrative review evaluates whether agronomic management can produce additional, measurable and sufficiently durable soil organic carbon (SOC) gains while maintaining agricultural production. Literature published from 1 January 1990 to 6 July 2026 was considered, with earlier or later-dated online records excluded from the evidentiary synthesis when outside this interval. The evidence indicates that SOC responses are governed principally by changes in plant-derived carbon inputs, microbial transformation, mineral association, aggregation, erosion and vertical redistribution rather than by any agronomic label alone. Cover crops, diversified rotations that increase biomass inputs, residue retention, perennial phases and agroforestry generally provide credible pathways for endogenous carbon accrual, although effect sizes vary with biomass production, soil mineralogy, climate, starting SOC and duration. Reduced and no-tillage frequently increase near-surface SOC but do not consistently increase whole-profile stocks when equivalent soil mass and deeper sampling are used, making tillage a particularly sensitive case for carbon accounting. Manure, compost and biochar can substantially increase measured soil carbon, but their climate value depends on carbon-source counterfactuals, processing emissions, displacement of organic resources and non-CO₂ greenhouse gases. Nutrient and water management can enhance SOC indirectly by increasing crop carbon inputs, yet gains must be evaluated against fertiliser-related nitrous oxide and energy emissions. Across practices, finite sink capacity, reversibility, spatial heterogeneity and measurement uncertainty constrain claims of permanent sequestration. The most defensible strategy is therefore not a universal practice hierarchy, but context-specific agronomic design that increases photosynthetic carbon inputs, protects soil from loss, verifies stock change across an appropriate soil mass and depth, and evaluates whole-system greenhouse-gas consequences. Future research should prioritise long-term, profile-scale, carbon-source-aware experiments and monitoring frameworks that connect total SOC change with particulate and mineral-associated fractions, crop performance, permanence and full greenhouse-gas accounting.