Microbial source tracking in impaired watersheds using PhyloChip and machine-learning classification

Health impacts of poor water quality on an endangered shorebird breeding programme in Aotearoa New Zealand

CASE HISTORY

Two clusters of mortality among endangered tūturuatu/tchūriwat'/shore plover (Thinornis novaeseelandiae) have occurred at captive breeding facilities around New Zealand in recent years. In the first, four chicks died at Pūkaha National Wildlife Centre (Mount Bruce, NZ) in February 2016, and in the second five adult birds at the Cape Sanctuary (Cape Kidnappers, NZ) died in 2022.

CLINICAL FINDINGS

In 2016, four chicks were noted to become weak, have increased vocalisations and closed eyes prior to death. The remaining chicks were treated for 5 days with amoxycillin/clavulanate orally twice daily. Water containers and brooders were cleaned and disinfected with chlorhexidine. No further mortality was seen.In the 2022 cluster, three adult breeding birds died acutely and five others showed inappetence, weight loss and diarrhoea approximately 10 days after heavy rains flooded the local river. The five birds were treated with amoxycillin/clavulanate orally twice daily and oral fluids for 5 days. Two birds died and three survived. No breeding occurred in the aviaries in the following season.

PATHOLOGICAL FINDINGS

In 2016, the chicks showed pulmonary changes ranging from congestion and oedema to heterophilic inflammation consistent with septicaemia.In 2022, the adult birds showed proliferation of bacteria in the distal small intestine associated with mucosal ulceration and heterophilic infiltration. Acid-fast staining of the caecal contents in one bird showed organisms consistent with Cryptosporidium spp.

LABORATORY FINDINGS

Aerobic bacterial cultures of the lung and liver of two affected chicks carried out in 2016 showed heavy growth of Plesiomonas shigelloides. The same organism was cultured from water trays and holding tanks containing water boatmen (Sigara arguta) on which the chicks were fed.In 2022, cultures from the livers of three dead birds each showed a mixed bacterial growth with differing dominant organisms (Aeromonas sobria, Hafnia alvei, Citrobacter freundii and an Enterococcus sp.). PCR and sequencing confirmed Cryptosporidium parvum in the caecum of one bird. Fresh faeces from 24 breeding birds from the captive breeding facilities were negative by PCR for Cryptosporidium spp.The captive breeding facilities obtain water for the aviaries and aquatic invertebrates to feed to the chicks from local freshwater sources. Water quality testing at the Cape Sanctuary revealed concentrations of faecal indicator bacteria in excess of safe drinking water guidelines, with peaks following heavy rainfall.

CLINICAL RELEVANCE

Fluctuations in water quality associated with mammalian faecal bacteria can adversely affect bird health and impact on captive rearing of endangered wildlife.

Digitalizing river aquatic ecosystems

Traditional river health assessment relies on limited water quality indices and representative organism activity, but does not comprehensively obtain biotic and abiotic information of the ecosystem. Here, we propose a new approach to evaluate the ecological and health risks of river aquatic ecosystems. First, detailed physicochemical and biological characterization of a river ecosystem can be obtained through pollutant determination (especially emerging pollutants) and DNA/RNA sequencing. Second, supervised machine learning can be applied to perform classification analysis of characterization data and ascertain river ecosystem ecology and health. Our proposed methodology transforms river ecosystem health assessment and can be applied in river management.

Persistence of location-specific microbial signatures on boats during voyages

Objects collect microorganisms from their surroundings and develop a microbial "fingerprint" that may be useful for determining an object's past location (provenance). It may be possible to use ubiquitous microorganisms for forensics or as environmental sensors. Here, we use microbial communities in the Chesapeake Bay region to demonstrate the use of natural microorganisms as biological sensors to determine the past location of boats. The microbiomes of two boats and of the open water were sampled as these vessels traveled from the Port of Baltimore to the Port of Norfolk, and back to Baltimore. 16S rRNA sequencing was performed to identify microorganisms. Differential abundance and machine learning analyses were utilized to identify microbial signatures and predicted probabilities which were used to determine the vessel's previous location. The work presented here provides a better understanding of how microbes in aquatic systems can be leveraged as utility for object biosensors.

Anthropogenic activity remains the main contributor to fecal pollution in managed tropical watersheds as unraveled by PhyloChip microarray-based microbial source tracking

The spread of disease by enteric pathogens associated with fecal contamination is a major concern for the management of urban watersheds. So far, the relative contribution of natural and anthropogenic sources to fecal pollution in managed tropical watersheds remains poorly evaluated. In this study, the microbiomes of water samples collected from managed watersheds in Singapore were elicited using the PhyloChip, a dense 16S rRNA gene-based DNA microarray, and fecal impairment was inferred using a machine-learning classification algorithm (SourceTracker). The predicted contribution of wildlife fecal sources to environmental samples was generally negligible (< 0.01 ± 0.01), indicating a low likelihood of fecal impairment from natural sources. However, sewage showed considerably higher contribution (0.09 ± 0.05) to microbial communities in a subset of watershed samples from canals and rivers, suggesting persistent impairment of certain areas by anthropogenic activity although being managed. Interestingly, the contribution of sewage microbial communities showed decreasing trends from canals/rivers to the connected reservoirs, indicating meaningful auto-mitigation of fecal pollution in canals and rivers. Notably, exclusion of locally derived fecal samples and source categories from the training data set impaired the predictive performance of the classification algorithm despite a high degree of similarity in the phylogenetic composition of microbiomes in biologically similar but geographically distinct sources.

Advances in machine learning-based bacteria analysis for forensic identification: identity, ethnicity, and site of occurrence

When faced with an unidentified body, identifying the victim can be challenging, particularly if physical characteristics are obscured or masked. In recent years, microbiological analysis in forensic science has emerged as a cutting-edge technology. It not only exhibits individual specificity, distinguishing different human biotraces from various sites of occurrence (e.g., gastrointestinal, oral, skin, respiratory, and genitourinary tracts), each hosting distinct bacterial species, but also offers insights into the accident's location and the surrounding environment. The integration of machine learning with microbiomics provides a substantial improvement in classifying bacterial species compares to traditional sequencing techniques. This review discusses the use of machine learning algorithms such as RF, SVM, ANN, DNN, regression, and BN for the detection and identification of various bacteria, including Bacillus anthracis, Acetobacter aceti, Staphylococcus aureus, and Streptococcus, among others. Deep leaning techniques, such as Convolutional Neural Networks (CNN) models and derivatives, are also employed to predict the victim's age, gender, lifestyle, and racial characteristics. It is anticipated that big data analytics and artificial intelligence will play a pivotal role in advancing forensic microbiology in the future.

Water treatment membranes embedded with a stable and bactericidal nanodiamond material

Filtration has emerged as a critical technology to reduce waterborne diseases caused by poor water quality. Filtration technology presents key challenges, such as membrane selectivity, permeability and biofouling. Nanomaterials can offer solutions to these challenges by varying the membranes' mechanical and bactericidal properties. This research uses nanodiamond particles with facile surface functionality and biocompatibility properties that are added to membranes used for filtration treatments. Scanning and transmission electron microscopy (SEM and TEM) and Fourier transform infrared spectroscopy (FTIR) were performed to study the membrane surface. FTIR spectra confirms an increase in oxygen functional groups onto the ultradispersed diamond's (UDD) surface following acid treatment. SEM images show particle deagglomeration of functionalized UDD at the membrane surface. Tensile strength tests were done to measure the UDD mechanical properties and Coliscan membrane filtration characterization was performed to determine the filter effectiveness. Polyether sulfone (PES) and polyvinylidene (PVDF) membranes expressed a change in their yield point when UDD was incorporated into the porous matrix. A significant microorganism reduction was obtained and confirmed using t-test analysis at a 95% level of confidence. UDD-embedded membranes exhibit a significant bactericidal reduction compared to commercial membranes suggesting these membranes have the potential to enhance current membrane filtration systems.

Tracking bacterial pollution at a marine wastewater outfall site - A case study from Norway

Coastal marine environments are increasingly affected by anthropogenic impacts, such as the release of sewage at outfall sites and agricultural run-off. Fecal pollution introduced to the sea through these activities poses risks of spreading microbial diseases and disseminating antibiotic resistant bacteria and their genes. The study area of this research, Bore beach, is situated between two such point sources, an outfall site where treated sewage is released 1 km off the coast and a stream that carries run-off from an agricultural area to the northern end of the beach. In order to investigate whether and to what extent fecal contamination from the sewage outfall reached the beach, we used microbial source tracking, based on whole community analysis. Samples were collected from sea water at varying distances from the sewage outfall site and along the beach, as well as from the sewage effluent and the stream. Amplicon sequencing of 16S rRNA genes from all the collected samples was carried out at two time points (June and September). In addition, the seawater at the sewage outfall site and the sewage effluent were subject to shotgun metagenomics. To estimate the contribution of the sewage effluent and the stream to the microbial communities at Bore beach, we employed SourceTracker2, a program that uses a Bayesian algorithm to perform such quantification. The SourceTracker2 results suggested that the sewage effluent is likely to spread fecal contamination towards the beach to a greater extent than anticipated based on the prevailing sea current. The estimated mixing proportions of sewage at the near-beach site (P4) were 0.22 and 0.035% in June and September, respectively. This was somewhat below that stream's contribution in June (0.028%) and 10-fold higher than the stream's contribution in September (0.004%). Our analysis identified a sewage signal in all the tested seawater samples.

Domestic wastewater causes nitrate pollution in an agricultural watershed, China

Excessive quantities of nitrates in the aquatic environment can cause eutrophication and raise water safety concerns. Therefore, identification of the sources of nitrate is crucial to mitigate nitrate pollution and for better management of the water resources. Here, the spatiotemporal variations and sources of nitrate were investigated by stable isotopes (δ¹⁵N and δ¹⁸O), hydrogeochemical variables (e.g., NO₃^- and Cl^-), and exogenous microbial signals (i.e., sediments, soils, domestic and swine sewage) in an agricultural watershed (Changle River watershed) in China. The concentration ranges of δ¹⁵N- and δ¹⁸O-NO₃^- between 3.03‰-18.97‰ and -1.55‰-16.47‰, respectively, suggested that soil nitrogen, chemical fertilizers, and manure and sewage (M&S) were the primary nitrate sources. Bayesian isotopic mixing model suggested that the major proportion of nitrate within the watershed (53.12 ± 10.40% and 63.81 ± 15.08%) and tributaries (64.43 ± 5.03% and 76.20 ± 4.34%) were contributed by M&S in dry and wet seasons, respectively. Community-based microbial source tracking (MST) showed that untreated and treated domestic wastewater was the major source (>70%) of river microbiota. Redundancy analysis with the incorporation of land use, hydrogeochemical variables, dual stable isotope, and exogenous microbial signals revealed domestic wastewater as the dominant cause of nitrate pollution. Altogether, this study not only identifies and quantifies the spatiotemporal variations in nitrate sources in the study area but also provides a new analytical framework by combining nitrate isotopic signatures and community-based MST approaches for source appointment of nitrate in other polluted watersheds.

Interfacing Machine Learning and Microbial Omics: A Promising Means to Address Environmental Challenges

Microbial communities are ubiquitous and carry an exceptionally broad metabolic capability. Upon environmental perturbation, microbes are also amongst the first natural responsive elements with perturbation-specific cues and markers. These communities are thereby uniquely positioned to inform on the status of environmental conditions. The advent of microbial omics has led to an unprecedented volume of complex microbiological data sets. Importantly, these data sets are rich in biological information with potential for predictive environmental classification and forecasting. However, the patterns in this information are often hidden amongst the inherent complexity of the data. There has been a continued rise in the development and adoption of machine learning (ML) and deep learning architectures for solving research challenges of this sort. Indeed, the interface between molecular microbial ecology and artificial intelligence (AI) appears to show considerable potential for significantly advancing environmental monitoring and management practices through their application. Here, we provide a primer for ML, highlight the notion of retaining biological sample information for supervised ML, discuss workflow considerations, and review the state of the art of the exciting, yet nascent, interdisciplinary field of ML-driven microbial ecology. Current limitations in this sphere of research are also addressed to frame a forward-looking perspective toward the realization of what we anticipate will become a pivotal toolkit for addressing environmental monitoring and management challenges in the years ahead.

Environmental DNA metabarcoding for benthic monitoring: A review of sediment sampling and DNA extraction methods

Environmental DNA (eDNA) metabarcoding (parallel sequencing of DNA/RNA for identification of whole communities within a targeted group) is revolutionizing the field of aquatic biomonitoring. To date, most metabarcoding studies aiming to assess the ecological status of aquatic ecosystems have focused on water eDNA and macroinvertebrate bulk samples. However, the eDNA metabarcoding has also been applied to soft sediment samples, mainly for assessing microbial or meiofaunal biota. Compared to classical methodologies based on manual sorting and morphological identification of benthic taxa, eDNA metabarcoding offers potentially important advantages for assessing the environmental quality of sediments. The methods and protocols utilized for sediment eDNA metabarcoding can vary considerably among studies, and standardization efforts are needed to improve their robustness, comparability and use within regulatory frameworks. Here, we review the available information on eDNA metabarcoding applied to sediment samples, with a focus on sampling, preservation, and DNA extraction steps. We discuss challenges specific to sediment eDNA analysis, including the variety of different sources and states of eDNA and its persistence in the sediment. This paper aims to identify good-practice strategies and facilitate method harmonization for routine use of sediment eDNA in future benthic monitoring.

Interpretable tree-based ensemble model for predicting beach water quality

Tree-based machine learning models based on environmental features offer low-cost and timely solutions for predicting microbial fecal contamination in beach water to inform the public of the health risk. However, many of these models are black boxes that are difficult for humans to understand, which may cause severe consequences such as unexplained decisions and failure in accountability. To develop interpretable predictive models for beach water quality, we evaluate five tree-based models, namely classification tree, random forest, CatBoost, XGBoost, and LightGBM, and employ a state-of-the-art explanation method SHAP to explain the models. When tested on the Escherichia coli (E. coli) concentration data collected from three beach sites along Lake Erie shores, LightGBM, followed by XGBoost, achieves the highest averaged precision and recall scores. For all three sites, both models suggest lake turbidity as the most important predictor, and elucidate the crucial role of accurate local data of wave height and rainfall in the model development. Local SHAP values further reveal the robustness of the importance of lake turbidity as its SHAP value increases nearly monotonically with its value and is minimally affected by other environmental factors. Moreover, we found an intriguing interaction between lake turbidity and day-of-year. This work suggests that the combination of LightGBM and SHAP has a promising potential to develop interpretable models for predicting microbial water quality in freshwater lakes.

Tracking microeukaryotic footprint in a peri-urban watershed, China through machine-learning approaches

Microeukaryotes play a significant role in biogeochemical cycling and can serve as bioindicators of water quality in freshwater ecosystems. However, there is a knowledge gap on how freshwater microeukaryotic communities are assembled, especially that how terrestrial microeukaryotes influence freshwater microeukaryotic assemblages. Here, we used a combination of 18S rRNA gene amplicon sequencing and community-based microbial source tracking (MST) approaches (i.e., SourceTracker and FEAST) to assess the contribution of microeukaryotes from surrounding environments (i.e., soils, river sediments, swine wastewater, influents and effluents of decentralized wastewater treatment plants) to planktonic microeukaryotes in the main channel, tributaries and reservoir of a peri-urban watershed, China in wet and dry seasons. The results indicated that SAR (~ 49% of the total communities), Opithokonta (~ 34%), Archaeplastida (~ 9%), and Amoebozoa (~ 2%) were dominant taxa in the watershed. The community-based MST analysis revealed that sewage effluents (7.96 - 21.84%), influents (2.23 - 13.97%), and river sediments (2.56 - 11.71%) were the major exogenous sources of riverine microeukaryotes. At the spatial scale, the downstream of the watershed (i.e., main channel and tributaries) received higher proportions of exogenous microeukaryotic OTUs compared to the upstream reservoirs, while at the seasonal scale, the sewage effluents and influents contributed higher exogenous microeukaryotes to river water in wet season than in dry season. Moreover, the swine and domestic wastewater led to the presence of Apicomplexa in wet season only, implying rainfall runoff may enhance the spread of parasitic microeukaryotes. Taken together, our study provides novel insights into the immigration patterns of microeukaryotes and their dominant supergroups between terrestrial and riverine habitats.

Microbial source tracking of fecal contamination in Laguna Lake, Philippines using the library-dependent method, rep-PCR

Laguna Lake is an economically important resource in the Philippines, with reports of declining water quality due to fecal pollution. Currently, monitoring methods rely on counting fecal indicator bacteria, which does not supply information on potential sources of contamination. In this study, we predicted sources of Escherichia coli in lake stations and tributaries by establishing a fecal source library composed of rep-PCR DNA fingerprints of human, cattle, swine, poultry, and sewage samples (n = 1,408). We also evaluated three statistical methods for predicting fecal contamination sources in surface waters. Random forest (RF) outperformed k-nearest neighbors and discriminant analysis of principal components in terms of average rates of correct classification in two- (84.85%), three- (82.45%), and five-way (74.77%) categorical splits. Overall, RF exhibited the most balanced prediction, which is crucial for disproportionate libraries. Source tracking of environmental isolates (n = 332) revealed the dominance of sewage (47.59%) followed by human sources (29.22%), poultry (12.65%), swine (7.23%), and cattle (3.31%) using RF. This study demonstrates the promising utility of a library-dependent method in augmenting current monitoring systems for source attribution of fecal contamination in Laguna Lake. This is also the first known report of microbial source tracking using rep-PCR conducted in surface waters of the Laguna Lake watershed.

Synergistic Application of Molecular Markers and Community-Based Microbial Source Tracking Methods for Identification of Fecal Pollution in River Water During Dry and Wet Seasons

It is important to track fecal sources from humans and animals that negatively influence the water quality of rural rivers and human health. In this study, microbial source tracking (MST) methods using molecular markers and the community-based FEAST (fast expectation–maximization microbial source tracking) program were synergistically applied to distinguish the fecal contributions of multiple sources in a rural river located in Beijing, China. The performance of eight markers were evaluated using 133 fecal samples based on real-time quantitative (qPCR) technique. Among them, six markers, including universal (BacUni), human-associated (HF183-1 and BacH), swine-associated (Pig-2-Bac), ruminant-associated (Rum-2-Bac), and avian-associated (AV4143) markers, performed well in the study. A total of 96 water samples from the river and outfalls showed a coordinated composition of fecal pollution, which revealed that outfall water might be a potential input of the Fsq River. In the FEAST program, bacterial 16S rRNA genes of 58 fecal and 12 water samples were sequenced to build the “source” library and “sink,” respectively. The relative contribution (<4.01% of sequence reads) of each source (i.e., human, swine, bovine, or sheep) was calculated based on simultaneous screening of the operational taxonomic units (OTUs) of sources and sinks, which indicated that community-based MST methods could be promising tools for identifying fecal sources from a more comprehensive perspective. Results of the qPCR assays indicated that fecal contamination from human was dominant during dry weather and that fecal sources from swine and ruminant were more prevalent in samples during the wet season than in those during the dry season, which were consistent with the findings predicted by the FEAST program using a very small sample size. Information from the study could be valuable for the development of improved regulation policies to reduce the levels of fecal contamination in rural rivers.

Next Generation High Throughput Sequencing to Assess Microbial Communities: An Application Based on Water Quality

Traditional techniques to identify different contaminants (biological or chemical) in the waters are slow, laborious, and can require specialized expertise. Hence, the rapid determination of water quality using more sensitive and reliable metagenomic based approaches attains special importance. Metagenomics deals with the study of genetic material that is recovered from microbial communities present in environmental samples. In traditional techniques cultivation-based methodologies were used to describe the diversity of microorganisms in environmental samples. It has failed to function as a robust marker because of limited taxonomic and phylogenetic implications. In this backdrop, high-throughput DNA sequencing approaches have proven very powerful in microbial source tracking because of investigating the full variety of genome-based analysis such as microbial genetic diversity and population structure played by them. Next generation sequencing technologies can reveal a greater proportion of microbial communities that have not been reported earlier by traditional techniques. The present review highlights the shift from traditional techniques for the basic study of community composition to next-generation sequencing (NGS) platforms and their potential applications to the biomonitoring of water quality in relation to human health.

Template length, concentration and guanidine and cytosine content influence on multiple displacement amplification efficiency

Detection of low abundance human health pathogens in environmental samples is a challenge for water monitoring. This limitation can be overcome by the introduction of multiple displacement amplification (MDA) where a minute amount of genetic material can be amplified using a phi-29 DNA polymerase. However, the genetic makeup and the concentration of the polynucleotides might influence the amplification process due to inherent assay bias. Herein, a series of experiments were designed to demonstrate the effect of genome length, guanidine and cytosine content, and template concentration on the efficiency of MDA. Quantitative polymerase chain reaction (qPCR) was performed to quantify pre- and post-MDA concentrations of selected genes. Linear regression between pre- and post-MDA log gene copies L^-1 of both environmental and lab-grown samples showed a positive correlation (F = 77.59, P < 0.001, R² = 0.7, slope = 1.01). Correlation between relative polynucleotide increase after MDA and target organism length and gene target guanidine and cytosine (G + C) content (F = 4.3, P = 0.02) shows that lower G + C and higher genome length is favored in the MDA process. The MDA process was shown to favor a longer genome over a shorter genome (1.19 and 1.04 change in log gene copy L^-1, respectively) and a lower G + C content over a higher G + C content (1.11 and 0.61 change in log gene copy L^-1, respectively). There was no MDA bias observed when polynucleotides had the same G + C and genome length but different initial concentrations. This study highlights the need for increased caution when interpreting relative abundance of organisms amplified by MDA such as in next generation sequencing.

Tracking Major Sources of Water Contamination Using Machine Learning

Current microbial source tracking techniques that rely on grab samples analyzed by individual endpoint assays are inadequate to explain microbial sources across space and time. Modeling and predicting host sources of microbial contamination could add a useful tool for watershed management. In this study, we tested and evaluated machine learning models to predict the major sources of microbial contamination in a watershed. We examined the relationship between microbial sources, land cover, weather, and hydrologic variables in a watershed in Northern California, United States. Six models, including K-nearest neighbors (KNN), Naïve Bayes, Support vector machine (SVM), simple neural network (NN), Random Forest, and XGBoost, were built to predict major microbial sources using land cover, weather and hydrologic variables. The results showed that these models successfully predicted microbial sources classified into two categories (human and non-human), with the average accuracy ranging from 69% (Naïve Bayes) to 88% (XGBoost). The area under curve (AUC) of the receiver operating characteristic (ROC) illustrated XGBoost had the best performance (average AUC = 0.88), followed by Random Forest (average AUC = 0.84), and KNN (average AUC = 0.74). The importance index obtained from Random Forest indicated that precipitation and temperature were the two most important factors to predict the dominant microbial source. These results suggest that machine learning models, particularly XGBoost, can predict the dominant sources of microbial contamination based on the relationship of microbial contaminants with daily weather and land cover, providing a powerful tool to understand microbial sources in water.

Microbial Source Tracking in the Love Creek Watershed, Delaware (USA)

Fecal contamination of waterways in Delaware pose an ongoing problem for environmental and public health. For monitoring efforts, Enterococcus has been widely adopted by the state to indicate the presence of fecal matter from warm-blooded animals and to establish Primary and Secondary Contact Recreation criteria. In this study, we examined sites within the Love Creek watershed, a tributary of the Rehoboth bay, using next-generation sequencing and SourceTracker to determine sources of potential fecal contamination and compared to bacterial communities to chemical and nutrient concentrations. Microbial community from fecal samples of ten different types of animals and one human sample were used to generate a fecal library for community-based microbial source tracking. Orthophosphate and total dissolved solids were among the major factors associated with community composition. SourceTracker analysis of the monthly samples from the Love Creek watershed indicated the majority of the microbial community were attributed to "unknown" sources, i.e. wildlife. Those that attribute to known sources were primarily domestic animals, i.e. cat and dog. These results suggest that at the state level these methods are capable of giving the start for source tracking as a means to understanding bacterial contamination.

Source tracking of antibiotic resistance genes in the environment - Challenges, progress, and prospects

Antibiotic resistance has become a global public health concern, rendering common infections untreatable. Given the widespread occurrence, increasing attention is being turned toward environmental pathways that potentially contribute to antibiotic resistance gene (ARG) dissemination outside the clinical realm. Studies during the past decade have clearly proved the increased ARG pollution trend along with gradient of anthropogenic interference, mainly through marker-ARG detection by PCR-based approaches. However, accurate source-tracking has been always confounded by various factors in previous studies, such as autochthonous ARG level, spatiotemporal variability and environmental resistome complexity, as well as inherent method limitation. The rapidly developed metagenomics profiles ARG occurrence within the sample-wide genomic context, opening a new avenue for source tracking of environmental ARG pollution. Coupling with machine-learning classification, it has been demonstrated the potential of metagenomic ARG profiles in unambiguously assigning source contribution. Through identifying indicator ARG and recovering ARG-host genomes, metagenomics-based analysis will further increase the resolution and accuracy of source tracking. In this review, challenges and progresses in source-tracking studies on environmental ARG pollution will be discussed, with specific focus on recent metagenomics-guide approaches. We propose an integrative metagenomics-based framework, in which coordinated efforts on experimental design and metagenomic analysis will assist in realizing the ultimate goal of robust source-tracking in environmental ARG pollution.

Bacterial community and environmental factors associated to rivers runoff and their possible impacts on coral reef conservation

Rivers potentially conduct important components as result of anthropogenic stressors for coral reefs. Molecular techniques are increasingly being used for monitoring biological and chemical monitoring of rivers and reefs. Here, we use PhyloChips™ to process surface water samples collected along two rivers and associated reefs in an environmental protection area in northeastern Brazil. Our results indicate that a significant part of Operational Taxonomic Units (OTUs) identified were able to survive the transition from freshwater to seawater, several of them belonging to genera implicated in human pathogenesis. The BBC:A ratio and functional prediction suggests that both study rivers are subject to fecal contamination and xenobiotics input and that the bacterial communities were more homogeneous in these environments. We suggest that protection actions adopted for reefs should be broadly extended to the surrounding environment, and that other bacterial group (besides cultivable coliforms) should be included in routine water quality monitoring.

Does environmental pollution inhibit urbanization in China? A new perspective through residents' medical and health costs

Health problems caused by environmental pollution may affect the process of urbanization in China. Therefore, this study, against the backdrop of promoting new-type urbanization, evaluates the level of China's urbanization comprehensively using the fully arranged polygon graphical index method. It uses a dynamic threshold panel model to study the potential non-linear relationship between environmental pollution (wastewater, sulfur dioxide, and solid wastes) and urbanization under different health costs of residents. Our findings show that environmental pollution has inhibited the improvement of comprehensive urbanization, population urbanization, economic urbanization, and living conditions urbanization, but promoted living environment urbanization, in China. It is worth noting that with the rise in residents' health costs, the inhibiting effect of environmental pollution on comprehensive urbanization, population urbanization, economic urbanization, and living conditions urbanization in China has gradually increased, but on living environment urbanization, it has decreased.

Microbial indicators and molecular markers used to differentiate the source of faecal pollution in the Bogotá River (Colombia)

Intestinal pathogenic microorganisms are introduced into the water by means of faecal contamination, thus creating a threat to public health and to the environment. Detecting these contaminants has been difficult due to such an analysis being costly and time-intensive; as an alternative, microbiological indicators have been used for this purpose, although they cannot differentiate between human or animal sources of contamination because these indicators are part of the digestive tracts of both. To identify the sources of faecal pollution, the use of chemical, microbiological and molecular markers has been proposed. Currently available markers present some geographical specificity. The aim of this study was to select microbial and molecular markers that could be used to differentiate the sources of faecal pollution in the Bogotá River and to use them as tools for the evaluation and identification of the origin of discharges and for quality control of the water. In addition to existing microbial source markers, a phage host strain (PZ8) that differentiates porcine contamination was isolated from porcine intestinal content. The strain was identified biochemically and genotypically as Bacteroides. The use of this strain as a microbial source tracking indicator was evaluated in bovine and porcine slaughterhouse wastewaters, raw municipal wastewaters and the Bogotá River. The results obtained indicate that the selected microbial and molecular markers enable the determination of the source of faecal contamination in the Bogotá River by using different algorithms to develop prediction models.

Compositional and temporal stability of fecal taxon libraries for use with SourceTracker in sub-tropical catchments

Characterization of microbial communities using high-throughput amplicon sequencing is an emerging approach for microbial source tracking of fecal pollution. This study used SourceTracker software to examine temporal and geographical variability of fecal bacterial community profiles to identify pollutant sources in three freshwater catchments in sub-tropical Australia. Fecal bacterial communities from 10 animal species, humans, and composite wastewater samples from six sewage treatment plants were characterized and compared to freshwater samples using Illumina amplicon sequencing of the V5-V6 regions of the 16S rRNA gene. Source contributions were calculated in SourceTracker using new fecal taxon libraries as well as previously generated libraries to determine the effects of geographic and temporal variability on source assignments. SourceTracker determined 16S rRNA bacterial communites within freshwater samples, shared taxonomic similarities to that of wastewater at low levels (typically <3%). SourceTraker also predicted occasional fecal detection of deer and flying fox sources in the water samples. No significant differences in source contributions were observed within sequences from current and previously characterized fecal samples (P ≥ 0.107). However, significant differences were observed between previously characterized and newly characterized source communities (ANOSIM P ≤ 0.001), which shared <15% community composition. Results suggest temporal instability of fecal taxon libraries among tested sources and highlight continual evaluation of community-based MST using confirmatory qPCR analyses of marker genes.

Next-generation sequencing reveals fecal contamination and potentially pathogenic bacteria in a major inflow river of Taihu Lake

Taihu Lake is one of the largest freshwater lakes in China and serves as an important source for drinking water. This lake is suffering from eutrophication, cyanobacterial blooms and fecal pollution, and the inflow Tiaoxi River is one of the main contributors. The goal here was to characterize the bacterial community structure of Tiaoxi River water by next-generation sequencing (NGS), paying attention to bacteria that are either fecal-associated or pathogenic, and to examine the relationship between environmental parameters and bacterial community structure. Water samples collected from 15 locations in three seasons, and fecal samples collected from different hosts and wastewater samples were used for bacterial community analysis. The phyla Proteobacteria, Actinobacteria, Bacteroidetes, and Cyanobacteria were predominant in most of the water samples tested. In fecal samples, Bacteroidetes, Firmicutes, and Proteobacteria were abundant, while wastewater samples were dominated by Proteobacteria, Bacteroidetes, Acidobacteria, and Chloroflexi. The cluster analysis and principal coordinate analysis indicated that bacterial community structure was significantly different between water, fecal and sewage samples. Shared OTUs between water samples and chicken, pig, and human fecal samples ranged from 4.5 to 9.8% indicating the presence of avian, pig and human fecal contamination in Tiaoxi River. At genus level, five bacterial genera of fecal origin and sequences of seven potential pathogens were detected in many locations and their presence was correlated well with the land use pattern. The sequencing data revealed that Faecalibacterium could be a potential target for human-associated microbial source-tracking qPCR assays. Our results suggest that pH, conductivity, and temperature were the main environmental factors in shaping the bacterial community based on redundancy analysis. Overall, NGS is a valuable tool for preliminary investigation of environmental samples to identify the potential human health risk, providing specific information about fecal and potentially pathogenic bacteria that can be followed up by specific methods.

A Brief Review of Random Forests for Water Scientists and Practitioners and Their Recent History in Water Resources

Random forests (RF) is a supervised machine learning algorithm, which has recently started to gain prominence in water resources applications. However, existing applications are generally restricted to the implementation of Breiman’s original algorithm for regression and classification problems, while numerous developments could be also useful in solving diverse practical problems in the water sector. Here we popularize RF and their variants for the practicing water scientist, and discuss related concepts and techniques, which have received less attention from the water science and hydrologic communities. In doing so, we review RF applications in water resources, highlight the potential of the original algorithm and its variants, and assess the degree of RF exploitation in a diverse range of applications. Relevant implementations of random forests, as well as related concepts and techniques in the R programming language, are also covered.

Incorporating microbial community data with machine learning techniques to predict feed substrates in microbial fuel cells

The complicated interactions that occur in mixed-species biotechnologies, including biosensors, hinder chemical detection specificity. This lack of specificity limits applications in which biosensors may be deployed, such as those where an unknown feed substrate must be determined. The application of genomic data and well-developed data mining technologies can overcome these limitations and advance engineering development. In the present study, 69 samples with three different substrate types (acetate, carbohydrates and wastewater) collected from various laboratory environments were evaluated to determine the ability to identify feed substrates from the resultant microbial communities. Six machine learning algorithms with four different input variables were trained and evaluated on their ability to predict feed substrate from genomic datasets. The highest accuracies of 93 ± 6% and 92 ± 5% were obtained using NNET trained on datasets classified at the phylum and family taxonomic level, respectively. These accuracies corresponded to kappa values of 0.87 ± 0.10, 0.86 ± 0.09, respectively. Four out of six of the algorithms used maintained accuracies above 80% and kappa values higher than 0.66. Different sequencing method (Roche 454 or Illumina sequencing) did not affect the accuracies of all algorithms, except SVM at the phylum level. All algorithms trained on NMDS-compressed datasets obtained accuracies over 80%, while models trained on PCoA-compressed datasets presented a 10-30% reduction in accuracy. These results suggest that incorporating microbial community data with machine learning algorithms can be used for the prediction of feed substrate and for the potential improvement of MFC-based biosensor signal specificity, providing a new use of machine learning techniques that has substantial practical applications in biotechnological fields.

Accounting for Bacterial Overlap Between Raw Water Communities and Contaminating Sources Improves the Accuracy of Signature-Based Microbial Source Tracking

Microbial source tracking (MST) analysis is essential to identifying and mitigating the fecal pollution of water resources. The signature-based MST method uses a library of sequences to identify contaminants based on operational taxonomic units (OTUs) that are unique to a certain source. However, no clear guidelines for how to incorporate OTU overlap or natural variation in the raw water bacterial community into MST analyses exist. We investigated how the inclusion of bacterial overlap between sources in the library affects source prediction accuracy. To achieve this, large-scale sampling - including feces from seven species, raw sewage, and raw water samples from water treatment plants - was followed by 16S rRNA amplicon sequencing. The MST library was defined using three settings: (i) no raw water communities represented; (ii) raw water communities selected through clustering analysis; and (iii) local water communities collected across consecutive years. The results suggest that incorporating either the local background or representative bacterial composition improves MST analyses, as the results were positively correlated to measured levels of fecal indicator bacteria and the accuracy at which OTUs were assigned to the correct contamination source increased fourfold. Using the proportion of OTUs with high source origin probability, underpinning a contaminating signal, is a solid foundation in a framework for further deciphering and comparing contaminating signals derived in signature-based MST approaches. In conclusion, incorporating background bacterial composition of water in MST can improve mitigation efforts for minimizing the spread of pathogenic and antibiotic resistant bacteria into essential freshwater resources.

Fecal source tracking methods to elucidate critical sources of pathogens and contaminant microbial transport through New Zealand agricultural watersheds - A review

In New Zealand, there is substantial potential for microbial contaminants from agricultural fecal sources to be transported into waterways. The flow and transport pathways for fecal contaminants vary at a range of scales and is dependent on chemical, physical and biological attributes of pathways, soils, microorganisms and landscape characteristics. Understanding contaminant transport pathways from catchment to stream can aid water management strategies. It is not practical, however to conduct direct field measurement for all catchments on the fate and transport of fecal pathogens due to constraints on time, personnel, and material resources. To overcome this problem, fecal source tracking can be utilised to link catchment characteristics to fecal signatures identifying critical sources. In this article, we have reviewed approaches to identifying critical sources and pathways for fecal microorganisms from agricultural sources, and make recommendations for the appropriate use of these fecal source tracking (FST) tools.

Transcriptome Changes of Escherichia coli, Enterococcus faecalis, and Escherichia coli O157:H7 Laboratory Strains in Response to Photo-Degraded DOM

In this study, we investigated gene expression changes in three bacterial strains (Escherichia coli C3000, Escherichia coli O157:H7 B6914, and Enterococcus faecalis ATCC 29212), commonly used as indicators of water quality and as control strains in clinical, food, and water microbiology laboratories. Bacterial transcriptome responses from pure cultures were monitored in microcosms containing water amended with manure-derived dissolved organic matter (DOM), previously exposed to simulated sunlight for 12 h. We used RNA sequencing (RNA-seq) and quantitative real-time reverse transcriptase (qRT-PCR) to compare differentially expressed temporal transcripts between bacteria incubated in microcosms containing sunlight irradiated and non-irradiated DOM, for up to 24 h. In addition, we used whole genome sequencing simultaneously with RNA-seq to identify single nucleotide variants (SNV) acquired in bacterial populations during incubation. These results indicate that E. coli and E. faecalis have different mechanisms for removal of reactive oxygen species (ROS) produced from irradiated DOM. They are also able to produce micromolar concentrations of H₂O₂ from non-irradiated DOM, that should be detrimental to other bacteria present in the environment. Notably, this study provides an assessment of the role of two conjugative plasmids carried by the E. faecalis and highlights the differences in the overall survival dynamics of environmentally-relevant bacteria in the presence of naturally-produced ROS.

Tracking antibiotic resistance gene pollution from different sources using machine-learning classification

BACKGROUND

Antimicrobial resistance (AMR) has been a worldwide public health concern. Current widespread AMR pollution has posed a big challenge in accurately disentangling source-sink relationship, which has been further confounded by point and non-point sources, as well as endogenous and exogenous cross-reactivity under complicated environmental conditions. Because of insufficient capability in identifying source-sink relationship within a quantitative framework, traditional antibiotic resistance gene (ARG) signatures-based source-tracking methods would hardly be a practical solution.

RESULTS

By combining broad-spectrum ARG profiling with machine-learning classification SourceTracker, here we present a novel way to address the question in the era of high-throughput sequencing. Its potential in extensive application was firstly validated by 656 global-scale samples covering diverse environmental types (e.g., human/animal gut, wastewater, soil, ocean) and broad geographical regions (e.g., China, USA, Europe, Peru). Its potential and limitations in source prediction as well as effect of parameter adjustment were then rigorously evaluated by artificial configurations with representative source proportions. When applying SourceTracker in region-specific analysis, excellent performance was achieved by ARG profiles in two sample types with obvious different source compositions, i.e., influent and effluent of wastewater treatment plant. Two environmental metagenomic datasets of anthropogenic interference gradient further supported its potential in practical application. To complement general-profile-based source tracking in distinguishing continuous gradient pollution, a few generalist and specialist indicator ARGs across ecotypes were identified in this study.

CONCLUSION

We demonstrated for the first time that the developed source-tracking platform when coupling with proper experiment design and efficient metagenomic analysis tools will have significant implications for assessing AMR pollution. Following predicted source contribution status, risk ranking of different sources in ARG dissemination will be possible, thereby paving the way for establishing priority in mitigating ARG spread and designing effective control strategies.

Elucidating Waterborne Pathogen Presence and Aiding Source Apportionment in an Impaired Stream

Fecal indicator bacteria (FIB) are the basis for water quality regulations and are considered proxies for waterborne pathogens when conducting human health risk assessments. The direct detection of pathogens in water and simultaneous identification of the source of fecal contamination are possible with microarrays, circumventing the drawbacks to FIB approaches. A multigene target microarray was used to assess the prevalence of waterborne pathogens in a fecally impaired mixed-use watershed. The results indicate that fecal coliforms have improved substantially in the watershed since its listing as a 303(d) impaired stream in 2002 and are now near United States recreational water criterion standards. However, waterborne pathogens are still prevalent in the watershed, as viruses (bocavirus, hepatitis E and A viruses, norovirus, and enterovirus G), bacteria (Campylobacter spp., Clostridium spp., enterohemorrhagic and enterotoxigenic Escherichia coli, uropathogenic E. coli, Enterococcus faecalis, Helicobacter spp., Salmonella spp., and Vibrio spp.), and eukaryotes (Acanthamoeba spp., Entamoeba histolytica, and Naegleria fowleri) were detected. A comparison of the stream microbial ecology with that of sewage, cattle, and swine fecal samples revealed that human sources of fecal contamination dominate in the watershed. The methodology presented is applicable to a wide range of impaired streams for the identification of human health risk due to waterborne pathogens and for the identification of areas for remediation efforts.IMPORTANCE The direct detection of waterborne pathogens in water overcomes many of the limitations of the fecal indicator paradigm. Furthermore, the identification of the source of fecal impairment aids in identifying areas for remediation efforts. Multitarget gene microarrays are shown to simultaneously identify waterborne pathogens and aid in determining the sources of impairment, enabling further focused investigations. This study shows the use of this methodology in a historically impaired watershed in which total maximum daily load reductions have been successfully implemented to reduce risk. The results suggest that while the fecal indicators have been reduced more than 96% and are nearing recreational water criterion levels, pathogens are still detectable in the watershed. Microbial source tracking results show that additional remediation efforts are needed to reduce the impact of human sewage in the watershed.

Toward Forensic Uses of Microbial Source Tracking

The science of microbial source tracking has allowed researchers and watershed managers to go beyond general indicators of fecal pollution in water such as coliforms and enterococci, and to move toward an understanding of specific contributors to water quality issues. The premise of microbial source tracking is that characteristics of microorganisms that are strongly associated with particular host species can be used to trace fecal pollution to particular animal species (including humans) or groups, e.g., ruminants or birds. Microbial source tracking methods are practiced largely in the realm of research, and none are approved for regulatory uses on a federal level. Their application in the conventional sense of forensics, i.e., to investigate a crime, has been limited, but as some of these methods become standardized and recognized in a regulatory context, they will doubtless play a larger role in applications such as total maximum daily load assessment, investigations of sewage spills, and contamination from agricultural practices.

Bacterial community structure transformed after thermophilically composting human waste in Haiti

Recycling human waste for beneficial use has been practiced for millennia. Aerobic (thermophilic) composting of sewage sludge has been shown to reduce populations of opportunistically pathogenic bacteria and to inactivate both Ascaris eggs and culturable Escherichia coli in raw waste, but there is still a question about the fate of most fecal bacteria when raw material is composted directly. This study undertook a comprehensive microbial community analysis of composting material at various stages collected over 6 months at two composting facilities in Haiti. The fecal microbiota signal was monitored using a high-density DNA microarray (PhyloChip). Thermophilic composting altered the bacterial community structure of the starting material. Typical fecal bacteria classified in the following groups were present in at least half the starting material samples, yet were reduced below detection in finished compost: Prevotella and Erysipelotrichaceae (100% reduction of initial presence), Ruminococcaceae (98–99%), Lachnospiraceae (83–94%, primarily unclassified taxa remained), Escherichia and Shigella (100%). Opportunistic pathogens were reduced below the level of detection in the final product with the exception of Clostridium tetani, which could have survived in a spore state or been reintroduced late in the outdoor maturation process. Conversely, thermotolerant or thermophilic Actinomycetes and Firmicutes (e.g., Thermobifida, Bacillus, Geobacillus) typically found in compost increased substantially during the thermophilic stage. This community DNA-based assessment of the fate of human fecal microbiota during thermophilic composting will help optimize this process as a sanitation solution in areas where infrastructure and resources are limited.