A bioenergetics-kinetics coupled modeling study on subsurface microbial metabolism in a field biostimulation experiment

Microorganisms in nature conserve energy by catalyzing various geochemical reactions. To build a quantitative relationship between geochemical conditions and metabolic rates, we propose a bioenergetics-kinetics coupled modeling approach. This approach describes microbial community as a metabolic network, i.e., fermenting microbes degrade organic substrates while aerobic respirer, nitrate reducer, metal reducer, sulfate reducer, and methanogen consume the fermentation products. It quantifies the control of substrate availability and biological energy conservation on the metabolic rates using thermodynamically consistent rate laws. We applied this simulation approach to study the progress of microbial metabolism during a field biostimulation experiment conducted in Oak Ridge, Tennessee. In the experiment, ethanol was injected into a monitoring well and groundwater was sampled to monitor changes in the chemistry. With time, concentrations of ethanol and SO42- decreased while those of NH4+, Fe2+, and Mn2+ increased. The simulation results fitted well to the observation, indicating simultaneous ethanol degradation and terminal electron accepting processes. The rates of aerobic respiration and denitrification were mainly controlled by substrate concentrations while those of ethanol degradation, sulfate reduction, and methanogenesis were controlled dominantly by the energy availability. The simulation results suggested two different microbial growth statuses in the subsurface. For the functional groups with significant growth, variations with time in substrate concentrations demonstrated a typical S curve. For the groups without significant growth, initial decreases in substrate concentrations were linear with time. Injecting substrates followed by monitoring environmental chemistry therefore provides a convenient approach to characterize microbial growth in the subsurface where methods for direct observation are currently unavailable. This research was funded by the NABIR program, DOE, under grant No. DE-FG02-04ER63740 to CZ. We thank J. Istok, David Watson, and Philip Jardine for their help. The views and opinions of authors expressed herein do not necessarily state or reflect those of the DOE.