Integrating microbial diversity into soil carbon models: a hybrid modeling approach

AXIS 2: Coupling biogeochemical cycles in a context of global change

Microbes for SOC
Integrating microbial diversity into soil carbon models: a hybrid modeling approach

Understanding soil organic carbon (SOC) and nutrient dynamics is critical to predict climate change impacts. Soil microbial communities drive carbon (C)
decomposition and nutrient cycling. However, due to their complexity, the level and direction of microbial diversity effects on these processes remain unclear. As a result, questions remain on whether and how to incorporate microbial activity in SOC models. This project leverages a hybrid modeling approach, combining process-based with machine learning models, to study the effects of microbial diversity on soil C and nitrogen (N) fluxes. By encapsulating microbial complexity within a deep learning framework, this approach balances interpretability and predictive power, aiming to enhance model performance, advancing our understanding of microbial diversity effects on soil functions, and potentially improving the predictions of C and N dynamics in a context of climate change.