Mineral properties, microbes, transport, and plant-input profiles control vertical distribution and age of soil carbon stocks

The role of organo-mineral interactions, microbial dynamics, and vertical plant input profiles are hypothesized to be important in controlling soil organic matter (SOM) stocks and dynamics. To test this hypothesis, we enhanced and applied a model (Biotic and Abiotic Model of SOM – BAMS1) that represents microbial dynamics and organo-mineral interactions integrated with a multiphase reactive transport solver for variably saturated porous media. The model represents aqueous chemistry, aqueous advection and diffusion, gaseous diffusion, sorption processes, bacterial and fungal activity, and archetypal monomer- and polymer-carbon substrate groups, including dead cell wall material. This model structure is fundamentally different, and produces different SOM dynamics, than the pseudo-first order relationships that are used in most site- and global-scale terrestrial SOM models. We simulated two grasslands, a seven-site chronosequence (∼3.9–240 Ky) in Northern California, and a Russian Chernozem site where soils were sampled 100 years apart. We calibrated the model's vertically-resolved soil bulk specific surface area (SBSSA) using observed bulk SOM content, and then tested the model against observed Δ14C profiles. The modeled microbial processes, organo-mineral interactions, and vertical aqueous transport produced realistic vertically-resolved predictions of bulk SOM content, Δ14C values of SOM, lignin content, and fungi-to-aerobic bacteria biomass ratios. Using sensitivity analyses, we found that vertical carbon input profiles were important controls over the Δ14C depth distribution. Shallower carbon input profiles lead to older carbon at depth. In addition, the SBSSA was the dominant control over the magnitude and vertical distribution of SOM stocks. The findings of this study demonstrate the value of explicitly incorporating microbial activity, sorption, and vertical transport into land models to predict SOM dynamics