Identifying the acute toxicity of contaminated sediments using machine learning models.

Ecological risk assessment of contaminated sediment has become a fundamental component of water quality management programs, supporting decision-making for management actions or prompting additional investigations. In this study, we proposed a machine learning (ML)-based approach to assess the ecological risk of contaminated sediment as an alternative to existing index-based methods and costly toxicity testing. The performance of three widely used index-based methods (the pollution load index, potential ecological risk index, and mean probable effect concentration) and three ML algorithms (random forest, support vector machine, and extreme gradient boosting [XGB]) were compared in their prediction of sediment toxicity using 327 nationwide data sets from Korea consisting of 14 sediment quality parameters and sediment toxicity testing data. We also compared the performances of classifiers and regressors in predicting the toxicity for each of RF, SVM, and XGB algorithms. For all algorithms, the classifiers poorly classified toxic and non-toxic samples due to limited information on the sediment composition and the small training dataset. The regressors with a given classification threshold provided better classification, with the XGB regressor outperforming the other models in the classification. A permutation feature importance analysis revealed that Cr, Cu, Pb, and Zn were major contributors to toxicity prediction. The ML-based approach has the potential to be even more useful in the future with the expected increase in available sediment data.

Assessing the activity and diversity of fumarate-fed denitrifying bacteria by performing field single-well push-pull tests.

In situ biological denitrification has been proposed as an important metabolic activity in the remediation of nitrate-contaminated groundwater. In this study, the effects of fumarate, an electron donor for biological denitrification, on the in situ denitrifying activity were determined by using three types of single-well push-pull tests; transport, biostimulation and activity tests. During the tests, changes in microbial community composition were also investigated using denaturing gradient gel electrophoresis (DGGE) of 16S rRNA genes. Transport test demonstrated that non-reactive tracer and biologically reactive solutes behaved similarly. A biostimulation test was conducted to stimulate the denitrifying activities of native microorganisms, which were monitored by detecting the simultaneous production of CO(2) and drastic degradations of both nitrate and fumarate after the injection of fumarate as an electron donor and/or carbon source, with nitrate as an electron acceptor. A phylogenetic analysis suggested that the taxonomic affiliation of the dominant species before biostimulation was γ-Proteobacteria, including Acinetobacter species and Pseudomonas fluorescens, while the dominant species after biostimulation were affiliated with ß-Proteobacteria, cytophaga-Flavobacterium-Bacteroides and high G+C gram-positive bacteria. These results suggest that the analyses of groundwater samples using a combination of single well push pull tests with DGGE can be applied to investigate the activity, diversity and composition shift of denitrifying bacteria in a nitrate-contaminated aquifer.

Determination of electron donors by comparing reaction rates for in situ bioremediation of nitrate-contaminated groundwater.

Groundwater contaminated by nitrates occurs frequently. In this research, fumarate, acetate, formate, lactate, propionate, ethanol, and methane were evaluated as a potential electron donor and carbon source by comparing the denitrification rate for the in situ bioremediation of nitrate contaminated groundwater. The denitrification rate for each substance was the quickest in the order of: fumarate > hydrogen > formate/Lactate > ethanol > propionate > methanol > acetate. Microcosm studies were performed with fumarates and acetates. When fumarates were used as a substrate, nitrates were removed completely at a rate of 0.66 mmol/day, while the conversion rate from nitrate to nitrogen gas and other by-products was 87%. For the microcosm test, 42 mg of fumarates were needed to remove 30 mg of NO(3)--N/L. When using acetate as a sole carbon source, 31% of nitrates were removed during the initial adjustment period. Among the removed fractions, however, 83% of the nitrates were removed by the cell growth. Overall, the nitrate removal rate was 0.37 mmol/day when acetate was used as a sole carbon source. The acetate showed longer lag time before denitrification occurred, which implied that fumarate would have been a better carbon source compared to acetate as more amounts were utilized for nitrate removal than cell growth.