Application of Nonlinear Analysis Methods for Identifying Relationships Between Microbial Community Structure and Groundwater Geochemistry

Abstract

The relationship between groundwater geochemistry and microbial community structure can be complex and difficult to assess. We applied nonlinear and generalized linear data analysis methods to relate microbial biomarkers (phospholipids fatty acids, PLFA) to groundwater geochemical characteristics at the Shiprock uranium mill tailings disposal site that is primarily contaminated by uranium, sulfate, and nitrate. First, predictive models were constructed using feedforward artificial neural networks (NN) to predict PLFA classes from geochemistry. To reduce the danger of overfitting, parsimonious NN architectures were selected based on pruning of hidden nodes and elimination of redundant predictor (geochemical) variables. The resulting NN models greatly outperformed the generalized linear models. Sensitivity analysis indicated that tritium, which was indicative of riverine influences, and uranium were important in predicting the distributions of the PLFA classes. In contrast, nitrate concentration and inorganic carbon were least important, and total ionic strength was of intermediate importance. Second, nonlinear principal components (NPC) were extracted from the PLFA data using a variant of the feedforward NN. The NPC grouped the samples according to similar geochemistry. PLFA indicators of Gram-negative bacteria and eukaryotes were associated with the groups of wells with lower levels of contamination. The more contaminated samples contained microbial communities that were predominated by terminally branched saturates and branched monounsaturates that are indicative of metal reducers, actinomycetes, and Gram-positive bacteria. These results indicate that the microbial community at the site is coupled to the geochemistry and knowledge of the geochemistry allows prediction of the community composition.

Depth-Resolved Quantification of Anaerobic Toluene Degraders and Aquifer Microbial Community Patterns in Distinct Redox Zones of a Tar Oil Contaminant Plume

“…However, correlations between aquifer geochemistry and microbial communities are extremely complex, and statistical tools are needed to unravel these relations (1,39,47). Here, we were able to actually identify distinct aquifer microbiota characteristic for the resolved plume compartments.…”

Phospholipid fatty acid biomarkers in a freshwater periphyton community exposed to uranium: discovery by non-linear statistical learning

“…However, because neither specific PLFA molecules or categories of PLFAs have been consistently assigned to particular ecological categories, it can be difficult to precisely ascribe PLFA biomarkers to community responses (Zelles, 1997;Zelles et al, 1992), thus motivating biochemical research to resolve this (Zelles, 1997). Examples of these correlations with varied ecological categories include: cy19:0 is indicative of a condition, microbial stress (Dickens and Anderson, 1999;Li et al, 2007); branched fatty acids are associated with a physiological trait, Gram positive microorganisms (Haubert et al, 2006;Lindahl et al, 1997;Nakamura et al, 2003), a physiological requirement, anaerobes (Keith-Roach et al, 2002;Nakamura et al, 2003), and a functional capability, metal reducers Schryver et al, 2006); and under specific growth conditions, branched odd-chain fatty acids (i15:0, a15:0, 15:1u6) may indicate narrower taxonomic groups of microorganisms, Desulfococcus or Desulfosarcina (Webster et al, 2006). Using a PLFA profile to describe the microbial community is also a daunting task, given that many PLFA biomarkers are poorly validated or may be valid only under particular conditions.…”

“…The poor specificity of PLFAemicroorganism associations is a confounding factor in the statistical classification/clustering approaches used to interpret PLFA profiles (Zelles, 1999), which do not reliably detect community responses to perturbations (Schryver et al, 2006;Small et al, 2008). The development of new data analysis methods to detect community responses to stress from PLFA profiles is a necessity; tools that can identify the subset of discriminating PLFAs from the full set of PLFAs measured will allow researchers to focus their ecological interpretations and validation efforts to the most relevant PLFA.…”

Longitudinal Changes in the Bacterial Community Composition of the Danube River: a Whole-River Approach

“…The major advantages of ANNs are that the data do not need to fit a predefined model (e.g., the normal distribution), the linear and nonlinear relationships can be modeled simultaneously, and they are tolerant to noisy data due to their high parallelism. Previous studies demonstrate the usefulness of ANNs in microbial ecology (10,32,36,42), and a detailed introduction to ANNs is given by Basheer and Hajmeer (2). …”