A Framework for Modeling Soil Health through Microbial-Soil Physics Coupling

Thesis Abstract

Soil health is increasingly recognized as a multifaceted system governed by interactions between microbial communities and physical soil processes, yet integrative frameworks that quantify these couplings remain underdeveloped for predictive management. The study aims to develop and validate a framework for modeling soil health through microbial-soil physics coupling, by linking microbial functional traits with soil physical properties to predict emergent soil health indicators under variable management and climate scenarios. Specific objectives include (i) to identify key microbial functional groups and their soil-physical process mediators (aggregation, porosity, water retention, and diffusion); (ii) to formulate a coupled mechanistic-empirical model integrating microbial kinetics with porous media physics; (iii) to calibrate and validate the framework using field and controlled-environment data; and (iv) to evaluate the framework’s predictive capacity for soil carbon stabilization, nutrient cycling, and aggregate stability under different tillage, moisture regimes, and organic amendments. Methodologically, the research adopts a mixed-methods design grounded in system theory and the microbial-ecology of soil physics, drawing on the theory of microbial by-products shaping pore-scale hydraulics and the Soil Health Index framework. The study population comprises temperate agroecosystems with diverse soil orders. A stratified sampling design will select 12 farms representing silt loam, clay loam, and sandy loam textures, with 20 plots per farm (n = 240 plots). Within each plot, soil cores (five per plot) will be collected seasonally across two years for physical, chemical, and microbiological analyses. Data collection instruments include high-resolution X-ray computed tomography for pore-scale architecture, gas chromatography for microbial metabolite profiling, qPCR and metagenomic sequencing for microbial functional genes, soil respiration chambers, and standard soil tests for bulk density, particle size distribution, organic carbon, and moisture content. Analytical techniques will comprise regression analysis and structural equation modeling to quantify coupling pathways, ANOVA to test treatment effects, and machine learning approaches (random forest and gradient boosting) to enhance predictive accuracy. A coupled microbial-physical model will be developed by integrating Monod-type microbial growth kinetics with Darcy-scale soil flow and transport equations, augmented by pore-network modeling to translate microbial activity into changes in porosity, tortuosity, and hydraulic conductivity. Model calibration will utilize a Bayesian hierarchical framework to accommodate multiscale data and uncertainty, with cross-validation across farms and years. Sensitivity analysis will identify dominant drivers of soil health outcomes, and scenario simulations will assess responses to tillage intensity, irrigation regimes, and organic amendments. Expected findings include (i) identification of core microbial functional groups (e.g., lignocellulose-degrading taxa, nitrifiers, denitrifiers, and extracellular polymeric substance producers) whose activity modulates pore structure and water retention; (ii) quantified relationships between microbial metabolism and changes in aggregate stability, soil organic carbon stabilization, and nutrient mineralization; (iii) a validated coupled model capable of predicting soil health indices with higher accuracy than conventional soil-physical or microbial-only models; and (iv) delineation of management practices that optimize beneficial microbe-physical interactions to enhance soil resilience under climate variability. The study contributes to knowledge by operationalizing a formalized framework that bridges microbial ecology and soil physics into a predictive tool for soil health, enabling scenario-based management and informing policy on sustainable soil stewardship. The main conclusion anticipates that incorporating microbial functional dynamics within physical soil processes markedly improves the predictability of soil health indicators under heterogeneous agroecosystems, and it recommends integrating the framework into precision agriculture platforms, expanding trials across tropical and semi-arid systems, and refining microbial proxies for rapid field assessment.