Can we use heat flows to quantify microbial traits in soils?

Knowledge of functional traits such as the maximum substrate uptake rate and growth efficiency of microorganisms is crucial in understanding the turnover and storage of soil organic carbon. In addition to CO2 measurements, heat dissipation from organic matter decomposition is also a well-recognized proxy for microbial activity in soils. However, only a few attempts have been made to utilize heat signals for quantifying microbial traits.

To leverage high-resolution heat dissipation data, a coupled mass-energy balance model is proposed and used to estimate microbial traits encoded in model parameters. Our underlying question was whether heat dissipation data alone would be sufficient to quantify key microbial traits, or whether respiration rates were also necessary to constrain the model. To this aim, we parametrized four variants of the model using heat dissipation and respiration rate data at different time scales: during the initial lag-phase (5 hours) and throughout the growth-phase until substrate depletion (48 hours) in an isothermal calorimeter combined with a gas analyzer. The four different variants of the model were: (i) a complex physiological model (including active and inactive biomass), (ii) a simplified physiological model (only active microbial biomass), (iii) a model describing only the lag-phase (no growth, only maintenance), and (iv) a model describing only the growth phase (growth under substrate-abundant conditions). Microbial traits were determined using three combinations of data: A) only the heat dissipation rate, B) only the respiration rate, and C) both heat dissipation and respiration rates. We assumed that the ‘best’ parameter estimates would be obtained when using all the available data (i.e., option C).

Our results show that all model variants were able to fit the observed heat dissipation and respiration rates at the respective time scales. Parameters shared among different model variants were generally comparable, indicating that our model simplifications led to structurally sound models. The parameters estimated using only heat dissipation data were similar to the ‘best’ estimates compared to using only respiration rate data, suggesting that the observed heat dissipation rate can be used to constrain microbial models and estimate microbial traits.