The role of oxalate-extractable aluminum and root inputs in carbon storage in Swedish agricultural soils: ‘beyond the average’

Soil organic carbon (SOC) supports multiple ecosystem services, yet SOC pool estimates remain too uncertain for reliable policy and climate-mitigation measures. Many SOC models inadequately represent soil properties and carbon inputs, often relying on clay content for storage predictions and fixed root-to-shoot ratios for rootderived inputs, neglecting within-field variability. This thesis empirically evaluates these two sources of uncertainty (i.e. soil properties and carbon inputs) and their implications for SOC storage in Swedish agricultural mineral soils in three studies (Papers I–III). In Paper I, roots (0–40 cm depth) and shoots were sampled during the milking/early dough stage in a 50 × 50 cm square at 11 locations within a field in southwestern Sweden over two years. The field was cultivated with spring barley (Hordeum vulgare L.) both years. Root-to-shoot ratios varied (quartile coefficients of variation = 7–18 %), suggesting that the use of fixed ratios may introduce uncertainty in modelled root-derived carbon inputs within a field. Acidic mineral topsoils (n=100; 0–20 cm) from the Swedish soil and crop monitoring program (SMP) were analyzed for physical and geochemical characteristics (Paper II). Dispersion and oxalate extractions (in darkness) were used to determine silt-sized aggregation and reactive mineral phases, respectively. A random forest model using these data and SMP soil physical and management variables, and climate variables, identified oxalate-extractable aluminum (Alox) as a stronger predictor than clay content for SOC content. Partial dependence plots suggested a potential for SOC accrual in soils with higher Alox. Long-term (ca. 500 days) laboratory incubations of sieved soil samples (n=35; 3–10 cm) from the field used in Paper I, analyzed via partial least squares regression, did not support a protective effect of Alox on SOC content (Paper III). At quasi-steady-state, specific respiration rates declined with increasing SOC, total nitrogen (TN), and C:N ratios, suggesting SOC protection and possible effect of C:N ratios and TN in regulating SOC mineralization (Paper III). These results urge models to transcend clay content and fixed ratios, incorporating variable inputs and mineral phases for better SOC estimates in Swedish contexts.