Impact of soil core preparation on hydraulic properties and modeled carbon cycling across texture classes

Abstract Soil hydraulic properties regulate microbially mediated carbon (C) cycling, but empirical measurements are labor‐intensive and often rely on homogenized soils that do not encompass the impacts of soil structure on hydraulics. We evaluated how the soil matrix structure influences measured hydraulic properties and the resultant soil C model predictions. We measured water retention and hydraulic conductivity on intact versus repacked soil cores from four textures (sandy loam, loam, silt loam, and silty clay) collected by the Molecular Observation Network (MONet), compared results to the HiHydroSoil v2.0 database, and evaluated their impacts on ecosystem soil C predictions through the Millennial model. Structural differences produced the largest divergences in hydraulic properties for silty textures and the smallest for sandy soils. Relative to intact cores, repacked sandy loam and loam soils exhibited higher simulated respiration and reduced soil organic carbon (SOC) stocks. In contrast, repacked silt loam and silty clay showed lower respiration and increased SOC stocks, consistent with enhanced clay‐associated C protection. Model outcomes also differed depending on whether measured properties or database values were used. Together, these results show that sample structure and parameter source introduce systematic biases into predictions of soil C dynamics, with effects that depend on texture. Standardized, structure‐aware, and higher throughput measurements of hydraulic properties are needed to improve parameterization, reduce uncertainty, and better represent environmental and anthropogenic impacts on soil structure in coupled hydrology–biogeochemistry models.