Mapping Aquifer Zones Based on Microbial Ecology and Geochemistry in a Landfill Leachate Plume with a Self Organizing Map
A. R. Pearce,
P. J. Mouser,
G. K. Druschel,
D. M. Rizzo,
We implemented a self-organizing map to delineate aqueous geochemistry and microbial ecology in landfill leachate. In subsurface ecosystems microorganisms mitigate a myriad of chemical processes in the environment (including contaminant degradation and immobilization, redox cycling and nutrient transport). Resident microbial communities depend on geochemical energy for their metabolism. Thus, microbial diversity and survival depends on geochemical and contaminant variations in groundwater; yet it is difficult to explicitly include their relevance in site characterization. In many environmental systems there are benefits to including microbial diversity information in characterizing that system; yet traditional multivariate statistical methods are not suited to process multi-dimensional datasets. A self-organizing map, also known as SOM or Kohonen Map, is a non-linear clustering algorithm. The SOM reduces high-dimensional data to a lower dimension that is then grouped using clustering techniques. The SOM is effective with multiple data types (e.g.
including microbial phylogeny and the environmental parameters that describe their habitat). For proof-of-concept, we test this clustering algorithm on data collected from monitoring wells in a shallow landfill leachate-contaminated groundwater aquifer that we sampled for hydrogeochemistry (petroleum byproducts, halogenated volatile organic compounds and inorganic species) and microbial community members. Due to high concentrations of organic carbon in landfill leachate, groundwater sampled from these contaminated aquifers typically exhibits a large range of redox processes, from aerobic to methanogenic conditions. The in situ
microbial community is directly related to available nutrients in each zone. The dataset available from the Schuyler Falls Landfill in Schuyler Falls, NY includes detailed site-wide apparent conductivity as well as hydrochemical and microbiological data from 28 different monitoring wells. Groundwater samples were analyzed for temperature, pH, redox potential, turbidity, specific conductance and a suite of organic and inorganic contaminants. Microbiological ecology is described with 16S rRNA gene surveys using primer sets specific for Bacteria, Archaea and Geobacteraceae
and DNA sequences were identified as operational taxonomic units (OTUs) for further analysis. We used the SOM to cluster and delineate the hydrochemical and microbial data identifying redox zones in the subsurface. The SOM can be modified to account for spatial auto-correlation that exists within most groundwater datasets. Identification of different zones using this clustering algorithm is an important step in linking microbial activity to biogeochemical processes that are important for site characterization and long-term monitoring stewardship (i.e.
delineating groundwater plumes, identifying changes in redox condition, types of contamination or potential for biodegradation or immobilization). Subject Headings: Microbes
| Groundwater pollution
| North America
| United States
| New York Services: Buy this book/Buy this article
Return to search