Factors affecting the spatial and temporal distribution of E. coli in intertidal estuarine sediments

• Published in: The Science of the total environment Vol. 661; pp. 155 - 167
• Main Authors: Wyness, Adam J., Paterson, David M., Mendo, Tania, Defew, Emma C., Stutter, Marc I., Avery, Lisa M.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.04.2019
• Subjects: air temperature; bacteria; Bathing water quality; bulk density; carbohydrates; cost effectiveness; data collection; direct contact; Escherichia coli; Estuaries; estuarine sediments; fate and transport models; FIO; indicator species; littoral zone; monitoring; Pathogens; risk; salinity; Sediments; temporal variation; water quality; Estuaries; Pathogens; Bathing water quality; Sediments; FIO
• ISSN: 0048-9697; 1879-1026; 1879-1026
• DOI: 10.1016/j.scitotenv.2019.01.061

Microbiological water quality monitoring of bathing waters does not account for faecal indicator organisms in sediments. Intertidal deposits are a significant reservoir of FIOs and this indicates there is a substantial risk to bathers through direct contact with the sediment, or through the resuspension of bacteria to the water column. Recent modelling efforts include sediment as a secondary source of contamination, however, little is known about the driving factors behind spatial and temporal variation in FIO abundance. E. coli abundance, in conjunction with a wide range of measured variables, was used to construct models to explain E. coli abundance in intertidal sediments in two Scottish estuaries. E. coli concentrations up to 6 log10 CFU 100 g dry wt−1 were observed, with optimal models accounting for E. coli variation up to an adjusted R2 of 0.66. Introducing more complex models resulted in overfitting of models, detrimentally affected the transferability of models between datasets. Salinity was the most important single variable, with season, pH, colloidal carbohydrates, organic content, bulk density and maximum air temperature also featuring in optimal models. Transfer of models, using only lower cost variables, between systems explained an average deviance of 42%. This study demonstrates the potential for cost-effective sediment characteristic monitoring to contribute to FIO fate and transport modelling and consequently the risk assessment of bathing water safety. [Display omitted] •E. coli concentrations up to 6 log10 CFU 100 g dry wt−1 were observed in sediments.•Sediment characteristics explained E. coli abundance to an Adj-R2 of 0.66.•Higher cost predictor variables did not increase model transferability.•Model transferability between datasets explained up to 51.9% of deviance.•Salinity and season were the most important predictor variables.