Predicting recreational water quality and public health safety in urban estuaries using Bayesian Networks.

To support the reactivation of urban rivers and estuaries for bathing while ensuring public safety, it is critical to have access to real-time information on microbial water quality and associated health risks. Predictive modelling can provide this information, though challenges concerning the optimal size of training data, model transferability, and communication of uncertainty still need attention. Further, urban estuaries undergo distinctive hydrological variations requiring tailored modelling approaches. This study assessed the use of Bayesian Networks (BNs) for the prediction of enterococci exceedances and extrapolation of health risks at planned bathing sites in an urban estuary in Sydney, Australia. The transferability of network structures between sites was assessed. Models were validated using a novel application of the k-fold walk-forward validation procedure and further tested using independent compliance and event-based sampling datasets. Learning curves indicated the model's sensitivity reached a minimum performance threshold of 0.8 once training data included ≥ 400 observations. It was demonstrated that Semi-Naïve BN structures can be transferred while maintaining stable predictive performance. In all sites, salinity and solar exposure had the greatest influence on Posterior Probability Distributions (PPDs), when combined with antecedent rainfall. The BNs provided a novel and transparent framework to quantify and visualise enterococci, stormwater impact, health risks, and associated uncertainty under varying environmental conditions. This study has advanced the application of BNs in predicting recreational water quality and providing decision support in urban estuarine settings, proposed for bathing, where uncertainty is high.

Molecular microbiological approaches reduce ambiguity about the sources of faecal pollution and identify microbial hazards within an urbanised coastal environment.

Urbanised beaches are regularly impacted by faecal pollution, but management actions to resolve the causes of contamination are often obfuscated by the inability of standard Faecal Indicator Bacteria (FIB) analyses to discriminate sources of faecal material or detect other microbial hazards, including antibiotic resistance genes (ARGs). We aimed to determine the causes, spatial extent, and point sources of faecal contamination within Rose Bay, a highly urbanised beach within Sydney, Australia's largest city, using molecular microbiological approaches. Sampling was performed across a network of transects originating at 9 stormwater drains located on Rose Bay beach over the course of a significant (67.5 mm) rainfall event, whereby samples were taken 6 days prior to any rain, on the day of initial rainfall (3.8 mm), three days later after 43 mm of rain and then four days after any rain. Quantitative PCR (qPCR) was used to target marker genes from bacteria (i.e., Lachnospiraceae and Bacteroides) that have been demonstrated to be specific to human faeces (sewage), along with gene sequences from Heliobacter and Bacteriodes that are specific to bird and dog faeces respectively, and ARGs (sulI, tetA, qnrS, dfrA1 and vanB). 16S rRNA gene amplicon sequencing was also used to discriminate microbial signatures of faecal contamination. Prior to the rain event, low FIB levels (mean: 2.4 CFU/100 ml) were accompanied by generally low levels of the human and animal faecal markers, with the exception of one transect, potentially indicative of a dry weather sewage leak. Following 43 mm of rain, levels of both human faecal markers increased significantly in stormwater drain and seawater samples, with highest levels of these markers pinpointing several stormwater drains as sources of sewage contamination. During this time, sewage contamination was observed up to 1000 m from shore and was significantly and positively correlated with often highly elevated levels of the ARGs dfrA1, qnrS, sulI and vanB. Significantly elevated levels of the dog faecal marker in stormwater drains at this time also indicated that rainfall led to increased input of dog faecal material from the surrounding catchment. Using 16S rRNA gene amplicon sequencing, several indicator taxa for stormwater contamination such as Arcobacter spp. and Comamonadaceae spp. were identified and the Bayesian SourceTracker tool was used to model the relative impact of specific stormwater drains on the surrounding environment, revealing a heterogeneous contribution of discrete stormwater drains during different periods of the rainfall event, with the microbial signature of one particular drain contributing up to 50% of bacterial community in the seawater directly adjacent. By applying a suite of molecular microbiological approaches, we have precisely pinpointed the causes and point-sources of faecal contamination and other associated microbiological hazards (e.g., ARGs) at an urbanised beach, which has helped to identify the most suitable locations for targeted management of water quality at the beach.