UC Santa Cruz

Thesis Statement: Managed aquifer recharge (MAR) is a set of tools and protocols that purposefully collects water to replenish underground aquifers. Types of MAR vary in their capacity to recharge, the price to implement them, and the water quality of infiltrating water. Distributed stormwater collection MAR uses hillslopes and gravity to direct stormwater into collection basins where it then infiltrates back down into the aquifer. This dissertation uses techniques from microbiology, water analysis, carbon characterization, and metagenomic sequences to elucidate how the design of a distributed stormwater collection MAR basin impacts the microbial community, thus impacting the geochemical cycling and water quality. The addition of a carbon-rich permeable reactive barrier (PRB) layered on top of a collection basin promotes the removal of nitrates from agricultural stormwater. The removal of nitrates, along with most of the geochemical cycling in the subsurface soil, is primarily done by the native microbial community. However, prior to this research, it was unclear how the microbial was being impacted by either infiltration or infiltration with a carbon-rich PRB. There was also little research into how different PRB materials could have different results in both water quality and microbial community composition. This research’s overarching hypothesis was that the addition of a carbon-rich PRB would provide carbon to the native subsurface microbes and power their denitrification. We also hypothesized this carbon shift would provide a trophic benefit to a common consort of microbes and the type of carbon being provided would have different outcomes. This research gives first insights into the soil-microbe-water interactions that occur during MAR and helps provide a framework for a predictive understanding of the carbon and nutrient cycling occurring during MAR infiltration. We first looked at how the soil and water characteristics influence the microbial communities of three pilot-scale field experiments simulating shallow MAR infiltration. We found there was a common shift towards Proteobacteria in sites that had coarser soil profiles (mainly sand) after infiltration, while sites with finer soil profiles (more clay and silt) did not exhibit significant changes in microbial community composition. This soil characteristic trend continued when we compared soil communities below a carbon-rich PRB of woodchips to communities below a plot with no PRB. The finer soil did not show major differences in genera with the addition of a PRB, but the coarse soils had similar increases in bacteria capable of complex carbon degradation and decreases in genera capable of nitrification. The next question we asked was: how is the PRB influencing the metabolism of the underlying microbial soil community? To answer this, we took the leachate from two PRB materials: almond shells and woodchips, and created anaerobic microcosms with soil to mimic fully saturated MAR conditions. We observed the almond shell leachate removed nitrate as quickly as the positive control microcosm amended with lactate, while woodchip leachate was a little slower but faster than the unamended microcosm. This corresponded to the metagenomic analysis where we saw an increase in denitrifying genes such as napA and nosZ in the after samples. Overall the division of genes into different subcategories of metabolism remains unchanged, although the assigned taxonomy of the samples varied greatly between pre- and post- incubation. The final question we asked was: How can we characterize the carbon being leached off the different PRB materials? In order to answer this, we used a biodegradable organic carbon assay, an assimilable organic carbon assay, ultraviolet-visible absorption, and nuclear magnetic resonance to characterize the leachates. Our results show that microbes are able to use the carbon leached off of the almond shells better than the woodchips. The woodchips were composed of more complex carbon which could provide a long-term solution for a MAR basin that is experiencing wet-dry/ oxic/ anoxic cycling. These results taken together give a greater understanding of how the design of a MAR system is impacting the native communities in the subsurface soil. It also provides the foundation of important predictive factors for future models and analyses of MAR infiltration.