Spectral Induced Polarization (SIP) of Denitrification‐Driven Microbial Activity in Column Experiments Packed With Calcareous Aquifer Sediments

• Published in: Journal of geophysical research. Biogeosciences Vol. 128; no. 1
• Main Authors: Strobel, C., Abramov, S., Huisman, J. A., Cirpka, O. A., Mellage, A.
• Format: Journal Article
• Language: English
• Published: 01.01.2023
• Subjects: cation exchange; competitive sorption; denitrification; microbial activity; reactive transport modeling; spectral induced polarization
• ISSN: 2169-8953; 2169-8961
• DOI: 10.1029/2022JG007190

Spectral Induced Polarization (SIP) has been suggested as a non‐invasive monitoring proxy for microbial processes. Under natural conditions, however, multiple and often coupled polarization processes co‐occur, impeding the interpretation of SIP signals. In this study, we analyze the sensitivity of SIP to microbially‐driven reactions under quasi‐natural conditions. We conducted flow‐through experiments in columns equipped with SIP electrodes and filled with natural calcareous, organic‐carbon‐rich aquifer sediment, in which heterotrophic denitrification was bio‐stimulated. Our results show that, even in the presence of parallel polarization processes in a natural sediment under field‐relevant geochemical conditions, SIP is sufficiently sensitive to microbially‐driven changes in electrical charge storage. Denitrification yielded an increase in imaginary conductivity of up to 3.1 μScm−1 ${\upmu }\mathrm{S}\,{\mathrm{c}\mathrm{m}}^{-1}$ (+140%) and the formation of a distinct peak between 1 and 10 Hz, that matched the timing of expected microbial activity predicted by a reactive transport model fitted to solute concentrations. A Cole‐Cole decomposition allowed separating the polarization contribution of microbial activity from that of cation exchange, thereby helping to locate microbial hotspots without the need for (bio)geochemical data to constrain the Cole‐Cole parameters. Our approach opens new avenues for the application of SIP as a rapid method to monitor a system's reactivity in situ. While in preceding studies the SIP signals of microbial activity in natural sediments were influenced by mineral precipitation/dissolution reactions, the imaginary conductivity changes measured in the biostimulation experiments presented here were dominated by changes in the polarization of the bacterial cells rather than a reaction‐induced alteration of the abiotic matrix. Plain Language Summary To better predict the contribution of microbes to groundwater clean‐up it is important to locate microbes in the ground that are actively removing contaminants and measure how fast they are doing so. Our ability to do so, however, is limited by the difficulty in visualizing underground processes. Electrical methods such as spectral induced polarization (SIP) have been applied to monitor microbes and provide an alternative to visualize them underground. SIP, however, has so far only been shown to work in controlled environments and its sensitivity in natural systems remains a question. In this study, we conducted experiments with sediment collected from an underground aquifer, in which we stimulated microbial activity through the addition of nitrate, a widespread groundwater contaminant. Our results show that microbial consumption of nitrate causes a distinct SIP signal that is similar to SIP signals of bacteria in previously studied well‐controlled systems. Furthermore, we propose an approach to separate the SIP signal of microbes from that of other processes that occur in natural groundwater. This enables us to quickly asses where the microbes are active and can potentially improve future experiments through the localization of microbial hot‐spots and SIP‐guided sampling for more detailed microbiological analysis. Key Points Microbial denitrification in natural sediments can be tracked using spectral induced polarization (SIP) σ″ ${\sigma }^{{\prime\prime}}$ spectra of a natural microbial community match reported spectra from single strains and the level of microbial activity affects σ″ ${\sigma }^{{\prime\prime}}$ Superimposing Cole‐Cole terms provides a framework for separating microbial and abiotic contributions from SIP signals