Hydraulic Boundary Conditions Determine Biofilm Growth Patterns and Permeability in Groundwater

Biofilms are naturally occurring consortia of bacteria that colonize porous media such as streambed sediments, soils, and aquifers. Biofilm growth leads to clogging (i.e., bioclogging), which reduces the porosity and permeability of the porous medium and directly influences how nutrients and contaminants are transported and transformed in groundwater. To improve our understanding of how bioclogging regulates nutrient and contaminant fluxes, we must better characterize how flow conditions influence the spatial and temporal progression of biofilm growth within natural porous media geometries, as well as understand how this progression alters fluid flow. In this study, I conducted microfluidic experiments to test the hypothesis that hydraulic boundary conditions control the spatial distribution, timing, and extent to which bioclogging restricts groundwater flow. I manufactured custom microfluidic chambers (micromodels) that were modeled after a natural soil geometry, which I used to monitor the growth of a model biofilm-forming bacterium, Bacillus subtilis. I performed two experiments, with each experiment having identical initial conditions but different imposed boundary conditions: constant flow rate (Q) vs. constant pressure gradient (ΔP) across the micromodel. For each imposed boundary condition, I monitored the corresponding hydraulic variable (Q or ΔP) as it changed over time, and I used Darcy’s Law to relate these metrics to the bulk permeability of the micromodel. I also monitored spatial patterns of biofilm growth within the micromodels using an inverted microscope. For both boundary conditions, biofilm grew uniformly within the micromodel over an initial bioclogging interval, followed by a rapid sloughing event that removed the majority of biomass. Similarities in the initial bioclogging and sloughing events across experiments suggest that sloughing occurred in response to chemical or biological controls rather than by differences in forces experienced by biofilm across the two different boundary conditions when the micromodel was bioclogged. Biofilm regrew until the micromodels reached a pseudo-steady state regime characterized by stable measures of total biomass and bulk permeability, but this progression occurred faster under constant Q than constant ΔP. Steady state biofilm occupation of the pore network was higher at constant Q (74% porosity reduction compared with 54% reduction at constant ΔP), with an observable network of preferential flow paths (PFPs). Despite the differences in porosity reduction, the bulk permeability decreased by 97% in both experiments, approaching a similar steady state value. These results show that less biomass accumulates under constant ΔP, but this biomass more effectively clogs the porous medium because fluid stresses are insufficient to slough biofilm and maintain a connected network of open pores. This work demonstrates that domain-scale hydraulic conditions control the spatial and temporal patterns of bioclogging, thereby influencing the transport and transformation of nutrients and contaminants in groundwater.