ABSTRACT
The microbiological quality of water plays a crucial role in the relationship among human, animal, and environmental health. This research gives insight into the relationship between concentrations of Escherichia coli bacteria and physiochemical parameters in water, which is captured from the Butoniga reservoir and then used for treatment in the drinking water treatment plant Butoniga. Analysis was carried out using statistical analysis through the Pearson correlation coefficient and supported with PCA. The conducted analysis revealed that turbidity and Fe have the highest correlation coefficients with E. coli bacteria. Turbidity was also identified as a potential indicator for E. coli bacteria. Additionally, parameters such as Mn and UV 254 were also found to be closely related to E. coli bacteria, alongside turbidity and Fe. Furthermore, a relationship between E. coli bacteria and different water intakes was conducted. This shows that higher concentrations of E. coli bacteria were present when water was captured from lower water intakes, characterized by increased water turbidity. Thus, the research results provide important information on influential water quality parameters related to E. coli bacteria, especially in the Butoniga reservoir and related drinking water treatment plant, creating a foundation for future water quality management.
HIGHLIGHTS
Turbidity has been identified as a potential indicator for E. coli bacteria.
Mn, Fe, and UV 254 were also found to be closely related to E. coli bacteria.
Relationship between E. coli bacteria and different water intakes was found.
Research provides a basis for informed management practices and policies to ensure the safety of water intended for human consumption.
INTRODUCTION
Water quality, primarily the microbiological quality of water, is integral to the nexus of human, animal, and environmental health. Ongoing collective initiatives are committed to investigative methodologies for monitoring, predicting, and managing microbiological water quality. The substantial influence of human activities, exemplified by urbanization and industrialization, has significantly disrupted the environmental equilibrium, and aquatic systems are not exempt from these impacts. This phenomenon has endangered the microbiological quality of water through contamination by untreated domestic wastewater and other discharges of anthropogenic origin (Shahid Iqbal et al. 2017).
Today, water quality, especially microbiological water quality is a fundamental determinant of public health, particularly in the context of potable water sources. Among the various indicators used to assess microbiological water quality, Escherichia coli (E. coli) stands out as a crucial indicator of faecal contamination in water systems. The presence of E. coli bacteria indicates that potential pathogenic microorganisms can pose health risks to individuals consuming the water (Odonkor & Ampofo 2013). E. coli bacteria is a member of the faecal coliform group and it specifically indicates faecal contamination in a greater proportion than other faecal coliform species. Contamination of water with E. coli bacteria can indicate the presence of other harmful bacteria that can cause diseases and can refer to the extent as well as the nature of the contaminants. E. coli bacteria are able to survive in water for 4–12 weeks and are considered the more specific indicator of faecal contamination than faecal coliforms since the more general test for faecal coliforms also detects thermotolerant non-faecal coliform bacteria (Odonkor & Ampofo 2013).
The occurrence of pollution indicator bacteria (E. coli, total and faecal coliforms) has also been used as a sanitary parameter for evaluating the quality of drinking water in the legislation of the Republic of Croatia (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023).
Previous research works investigating the relationship between microbiological water quality and other influential parameters were done by Smith et al. (2008), Shahid Iqbal et al. (2017), Ding et al. (2018), Seo et al. (2019), Aram et al. (2021), Khan et al. (2021), Kothari et al. (2021), Sokolova et al. (2022), and Yoneda et al. (2024). In his study, Smith et al. (2008) presents the relationship between water turbidity and E. coli bacteria in water troughs on cattle farms of England and Wales. The relationship between hydro-climatic variables (i.e., air temperature, precipitations, water flow, etc.) and E. coli concentrations in surface and drinking water of the Kabul River basin in Pakistan using basic statistical analysis was done by Shahid Iqbal et al. (2017). Ding et al. (2018), in his study, established the relationship between water quality parameters (temperature, pH, total organic carbon (TOC), total nitrogen, etc.) and E. coli survival potential in well waters from a rural area of southern Changchum City in China. Seo et al. (2019), with the use of basic statistical analysis, investigated the relationship between coliform bacteria and water quality factors (i.e. pH, dissolved oxygen, total nitrogen, TOC, etc.) at weir stations in the Nakdong River in South Korea. Using a nested binary logistic regression model, a comparative assessment of the relationship between coliform bacteria and water geochemistry (i.e. pH, water turbidity, dissolved oxygen, manganese, etc.) in surface and groundwater systems in the Tarkwa mining area in the Western Region of Ghana was done in Aram et al. (2021). The study by Khan et al. (2021) presents a superposition learning-based model for the prediction of E. coli bacteria in groundwater using physicochemical water quality parameters like pH, turbidity, total dissolved solids, etc. Kothari et al. (2021) give a mutual correlation of various water quality parameters (i.e. pH, turbidity, total dissolved solids, iron, magnesium, etc.), which also include microbiological water quality parameters like total and faecal coliforms, with the Water Quality Index (WQI) of districts of Uttarakhand. In the research made by Sokolova et al. (2022), data-driven models were developed for predicting the microbial water quality in the river Göta älv at the water intake of the drinking water treatment plant (DWTP) in Gothenburg, Sweden using E. coli monitoring and hydro-meteorological data like turbidity, precipitations water flow, etc. A comparative experiment to select water quality parameters like temperature, pH, total dissolved solids, etc. for modelling the survival of E. coli bacteria in lakes is presented in the research by Yoneda et al. (2024).
Previous studies regarding the functioning and problems of the Butoniga reservoir and related DWTP were made in studies by Zorko (2017), Hajduk Černeha (2021), Volf et al. (2022, 2023) and Guide for the application of measures for adaptation to climate change (2023). The study made by Hajduk Černeha (2021) describes the first experiences in the usage of DWTP Butoniga for drinking water supply. In addition, this study analyzed raw water quality data from the Butoniga reservoir, and some management guidelines regarding the Butoniga reservoir and the related DWTP were given. In the second study made by Zorko (2017) the impact of the Butoniga reservoir raw water quality on water treatment was considered, and provided some interesting conclusions, mentioning that the main problem in the Butoniga reservoir, and thus the related DWTP, appears in the summer months when water temperature is the most critical parameter. This is because, in order to be suitable for use and for treatment processes, water must not exceed the maximum allowable concentration (MAC) of 25 °C, according to Croatian regulations for drinking water (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023). In research made by Volf et al. (2022) in order to improve the treatment processes of the DWTP, a prediction of the WQI was done using machine learning tools based on physiochemical water quality parameters. Prediction models for manganese, iron, and ammonium were developed to forecast these critical parameters 7 days in advance. These parameters significantly influence the raw water treatment process at the DWTP Butoniga, especially during the summer months, as highlighted in research by Volf et al. (2023). Also, a Guide for the application of measures for adaptation to climate change (2023) was made. In this guide, guidelines regarding the Butoniga reservoir and related DWTP are given for sampling methodology, selection of parameters, optimizing some treatment process units in the DWTP, etc.
The reported microbiological water quality conditions often face challenges related to both timeliness and accuracy. This is primarily due to inadequate monitoring activities on water bodies, which are frequently limited by financial constraints (Aram et al. 2021). Additionally, the 24-h incubation period required for laboratory analysis of E. coli bacteria introduces time constraints that further complicate monitoring efforts (Khan & Gupta 2020). As a result, there is a great need to develop models that can establish correlations between pollutant concentrations and factors that can be more readily measured (Sokolova et al. 2022).
Statistical models provide a valuable tool for determining initial probability distributions of impairment (Shahid Iqbal et al. 2017; Seo et al. 2019; Aram et al. 2021). They guide monitoring initiatives, reduce the amount of necessary data, and facilitate the assessment of potential future microbiological water quality scenarios (Aram et al. 2021). The application of statistical modelling to estimate water contamination, particularly for indicator organisms like total coliform, faecal coliform, or E. coli bacteria, proves advantageous and has the potential to become a standardized approach for evaluating water pollution (Shahid Iqbal et al. 2017; Seo et al. 2019; Aram et al. 2021).
The primary objective of this research is to analyze the relationship between concentrations of E. coli bacteria and physiochemical parameters in water, which is captured from the relatively small and shallow Butoniga reservoir and then used for treatment on DWTP, thus providing a basis for informed management practices and policies to ensure the safety of water intended for human consumption.
The paper is structured as follows: The Introduction section gives a general overview of E. coli bacteria with proper references on what has been done regarding similar research and on problems regarding the Butoniga reservoir and related DWTP. The second section, Methods, givens a description of the study area and data used in the research, with a description of statistical data analysis. The third section, Results and discussion, details about the results of the conducted statistical analysis with proper discussion and cited relevant references. In the final section, Conclusions, relevant conclusions of this research and some guidelines for further work are given.
METHODS
Study area and data description
Given these reservoir attributes, it is noteworthy that Butoniga is very sensitive to eutrophication and degradation processes triggered by both climate change and human activities in the surrounding watershed. Notable pressures within the surrounding watershed involve erosion and the leaching of nutrients, predominantly nitrogen and phosphorus, originating from agricultural lands. Additionally, untreated wastewater from settlements can drain underground to the reservoir from black pits or open sewers, contributing to the challenges faced by the Butoniga reservoir (Hajduk Černeha 2021; Zorko 2017).
The operation of the DWTP is primarily related to the tourist season. Out of the total annual production and distribution of 5,000,000 m3 of water, 3,000,000 m3 is generated and distributed between 15 June and 15 September. During this period, the water quality in the Butoniga reservoir is the worst, but all water samples in the effluent of DWTP, after treatment, are below the MAC according to Croatian regulations for drinking water (Zorko 2017).
The dataset used for statistical analysis (see Table 1) consists of physical, chemical, and microbiological parameters that were measured once a day at the inflow of raw water to the DWTP, from 2011 to 2020.
Dataset used in statistical data analysis
Symbol . | Description . | Measurement unit . |
---|---|---|
Temp. | Water temperature | °C |
pH | pH | – |
Tur | Water turbidity | NTU |
O2 | Oxygen concentration | mg/L |
TOC | Total organic carbon | mg/L |
KMnO4 | Potassium permanganate | mg/L |
NH4 | Ammonium | mg/L |
Mn | Manganese | mg/L |
Al | Aluminium | mg/L |
Fe | Iron | mg/L |
UV 254 | Organic matter in water | 1/cm |
Lake level | Lake level data | masl |
Water intake | Water intake number | – |
E. coli | Escherichia coli | CFU/100 mL |
Symbol . | Description . | Measurement unit . |
---|---|---|
Temp. | Water temperature | °C |
pH | pH | – |
Tur | Water turbidity | NTU |
O2 | Oxygen concentration | mg/L |
TOC | Total organic carbon | mg/L |
KMnO4 | Potassium permanganate | mg/L |
NH4 | Ammonium | mg/L |
Mn | Manganese | mg/L |
Al | Aluminium | mg/L |
Fe | Iron | mg/L |
UV 254 | Organic matter in water | 1/cm |
Lake level | Lake level data | masl |
Water intake | Water intake number | – |
E. coli | Escherichia coli | CFU/100 mL |
Physiochemical parameters include water temperature in the reservoir (Temp.), pH, turbidity (Tur.), oxygen concentration (O2), TOC, potassium permanganate (KMnO4), ammonia (NH4), manganese (Mn), aluminium (Al), iron (Fe), and amount of organic substances (UV 254) whose concentrations were determined in the internal laboratory of the Butoniga DWTP by standard analytical methods according to ISO standards, that is, according to the norm HRN EN ISO 5667-3. In addition to the data measured on the DWTP, statistical analyses were also used such as the reservoir water level data (lake level) and data about selected water intake positions, i.e. water intake numbers. There are four water intake units on the water intake tower (see Figure 2) which can collect water from four different layers in the reservoir depending on the water level in the reservoir. For microbiological parameters, E. coli bacteria were used, which, as stated in the Introduction section stands out as a crucial bacterium, serving as an indicator of faecal contamination in water systems. For the detection and enumeration of E. coli bacteria, the Colilert method, i.e. norm HRN EN ISO 9308-2:2014, was used.
As mentioned briefly in the Introduction section, the main problem in the Butoniga reservoir and thus the related DWTP occurs in summer months when the water temperature is the most critical parameter, so, in order to be suitable for use and also for treatment processes, water temperature must not exceed the MAC of 25 °C according to the Croatian legislation (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023). During this specific period, water is captured from the lowest water intake, which collects water from the deepest water layer in the Butoniga reservoir. This layer has another problem, namely increased concentrations of manganese (Mn), iron (Fe), and ammonium (NH4) under lower pH and oxygen (O2) values. Increased concentration of Mn, Fe, and NH4 under lower pH and O2 values of raw water from the lowest water intake requires enhanced continuous process control and higher consumption of chemicals for the treatment process in the DWTP. Under these conditions, the process in the DWTP is also stable and all water samples at the effluent are in accordance with the Croatian legislation (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023), although the exceeding values of the water temperature, due to the heating of the Butoniga reservoir, cannot be influenced (Zorko 2017).
Also, all water samples throughout the year in the effluent of the DWTP are below the MAC according to Croatian regulations for drinking water (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023).
The mentioned parameters were mainly used in the analysis because they accurately describe the parts of the aquatic ecosystem on which the target variable depends.
Statistical data analysis
Relationship analysis between E. coli bacteria and physiochemical parameters was done using Pearson correlation coefficient and principal component analysis (PCA) using Statistica 14 (Tibco Software Inc. 2017) and software package WEKA (Witten & Frank 2005).
Correlation analysis determines whether the relationship between two variables is present or absent, and that two variables are associated to a certain degree (Hüsser 2017). For correlation analyses in this research, the degree of linearity between two variables was derived using the correlation coefficient (R) through Pearson's correlation analyses, and the significance was based on a p-value (<0.05). The correlation coefficient ranges between −1 and +1. The linearity between variables is ignored when the correlation coefficient is between ±0.00 and ±0.1 (negligible correlation). A correlation coefficient within the range of ±0.1 to ±0.39 indicates a weak positive/negative linear correlation, while a coefficient within ±0.4 to ±0.69 suggests a moderate positive/negative linear correlation. A correlation in the range of ±0.7 to ±0.89 signifies a strong positive/negative linear correlation, and a coefficient between ±0.9 and ±1.0 is indicative of an exceptionally high or very strong positive/negative linear correlation (Schober et al. 2018).
PCA is a linear dimensionality reduction technique with applications in data analysis, visualization, and data pre-processing. This is accomplished by linearly transforming the data onto a new coordinate system (principal components (PCs)) such that the directions capturing the largest variation in the data can be easily identified. In data analysis, the first principal component of a set of p variables, presumed to be jointly normally distributed, is the derived variable formed as a linear combination of the original variables that explains the most variance. The second principal component explains the most variance in remaining variables once the effect of the first component is removed, and then proceed through p iterations until all the variance is explained (Jolliffe & Cadima 2016). In this research, PCA was used to study, verify, and support the relationship between each water quality parameter and E. coli bacteria in a way that it still contains most of the information in the given dataset.
RESULTS AND DISCUSSION
The relationship between physiochemical parameters and E. coli bacteria concentrations in water captured from the Butoniga reservoir and then treated in the DWTP was examined using Pearson correlation coefficient (R) and PCA.
Table 2 presents the Pearson correlation coefficients (R) between E. coli bacteria and various physiochemical parameters. Interpretation of magnitudes of correlation coefficients was done according to Schober et al. (2018). A moderate correlation, denoted by a correlation coefficient of 0.494 and 0.412, is evident among E. coli bacteria, water turbidity, and Fe respectively. Conversely, a weak or low correlation is observed between E. coli bacteria, UV 254, and Mn with correlation coefficients of 0.349 and 0.317, respectively. Notably, there is no significant correlation with the other physiochemical parameters. Statistically significant correlations (p < 0.05) in Table 2 are underlined.
Pearson correlation coefficient (R) between E. coli bacteria and physiochemical parameters
Parameter . | Correlation coefficient with E. coli . | Parameter . | Correlation coefficient with E. coli . |
---|---|---|---|
Water temperature | 0.17 | UV 254 | 0.349 |
O2 | − 0.137 | NH4 | 0.195 |
pH | −0.197 | Mn | 0.317 |
Turbidity | 0.494 | Al | 0.081 |
TOC | 0.153 | Fe | 0.412 |
KMnO4 | 0.134 | Lake level | − 0.13 |
Parameter . | Correlation coefficient with E. coli . | Parameter . | Correlation coefficient with E. coli . |
---|---|---|---|
Water temperature | 0.17 | UV 254 | 0.349 |
O2 | − 0.137 | NH4 | 0.195 |
pH | −0.197 | Mn | 0.317 |
Turbidity | 0.494 | Al | 0.081 |
TOC | 0.153 | Fe | 0.412 |
KMnO4 | 0.134 | Lake level | − 0.13 |
Higher correlations coefficients are bolded, while statistically significant correlations (p < 0.05) are underlined.
Regardless of the moderate correlation between water turbidity and Fe with E. coli bacteria, it can be concluded that in practical terms, the general tendency is that water turbidity along with Fe is related to E. coli bacteria, so a correlation exists.
It can be seen that noticeable correlations exist between Fe and Mn of 0.412 and 0.317, respectively, with E. coli bacteria, and can be explained by the fact that during the summer months, water is captured from the lowest water intakes, which inturn captures water from the deepest water layer in the Butoniga reservoir due to increased water temperature of the Butoniga reservoir. As mentioned MAC is 25 °C, according to Croatian regulations for drinking water (Regulation on parameters compliance, analysis methods and monitoring for water intended for human consumption 2023). This layer has another problem, namely increased concentrations of Mn, Fe, and NH4 under lower pH and O2 values (Zorko 2017). Turbidity and total suspended solids increase with depth because of the settling and re-suspension of inorganic solid particles on the reservoir bottom (Ling et al. 2016). Due to this reason, the lowest layer has increased water turbidity and in this case, increased concentrations of Mn and Fe also have higher concentrations of E. coli bacteria.
Concentrations of E. coli bacteria and turbidity over the observation period (2011–2020).
Concentrations of E. coli bacteria and turbidity over the observation period (2011–2020).
Smith et al. (2008) aimed to develop and test a predictive tool that can be used to estimate the level of E. coli contamination of drinking water. Through conducted research, turbidity was associated with E. coli bacteria concentration, although the association was not linear. Also, in research works made by Sokolova et al. (2022) and Khan et al. (2021), turbidity was chosen to be an important potential predictor for E. coli bacteria.
The above correlation suggests that suspended particles in turbid water may act as carriers for microbial contaminants. However, also other environmental parameters must be taken into account in water quality assessments for a comprehensive understanding of contamination risks (Seo et al. 2019; Khan & Gupta 2020; Aram et al. 2021).
According to the above-mentioned correlation, each system must be observed as a unique entity, which depends on various water quality and other environmental parameters.
PCA-biplot of eigenvectors (PC1 and PC2) loadings plot for the observed parameters.
PCA-biplot of eigenvectors (PC1 and PC2) loadings plot for the observed parameters.
E. coli bacteria concentrations and corresponding water intake numbers during the observation period (2011–2020).
E. coli bacteria concentrations and corresponding water intake numbers during the observation period (2011–2020).
The microbiological water quality is a critical determinant of public health, particularly concerning its use for drinking purposes. Microorganisms found in water, including bacteria, viruses, and parasites, can pose significant health risks due to a lack of control and management. Insurance of microbiological water safety is essential for preventing waterborne diseases and protecting the well-being of individuals and communities (Aram et al. 2021). E. coli, as part of microbiological water quality, is an indicator of faecal contamination of the surfacewater, groundwater, or drinking water. Thus, proper attention should be given to monitoring activities and laboratory experiments on which public health depends (Odonkor & Ampofo 2013). As mentioned before (Seo et al. 2019; Khan & Gupta 2020; Aram et al. 2021), E. coli bacteria depend on various water quality and environmental parameters, therefore, each ecosystem should be observed as a unique entity.
In this study, regarding the Butoniga reservoir and related DWTP, E. coli bacteria were associated with higher levels of turbidity where turbid water may act as a carrier for microbial contaminants. Also, it is evident from Figure 5 that higher concentrations of E. coli bacteria are related to capturing water from different water intakes. As mentioned, higher concentrations of E. coli bacteria are observed when water is captured from the lowest water intake 4, or a combination of intakes 3 and 4, which has higher levels of turbidity.
On DWTP Butoniga, regarding physiochemical and microbiological parameters, all samples at the effluent are in accordance with the Croatian regulations for drinking water (Regulation on parameters compliance, analysis methods, and monitoring for water intended for human consumption 2023).
CONCLUSIONS
In this research, the relationship between E. coli bacteria and physiochemical parameters in water, which is used for the treatment in the DWTP of the Butoniga reservoir, is studied. The relationship analysis, carried out by the statistical Pearson correlation coefficient and PCA, proved to be very useful in conducting the research.
The conducted analysis revealed that turbidity and Fe have the highest correlation coefficients. Turbidity was also identified as a potential indicator for E. coli bacteria. This implies that higher turbidity levels generally indicate higher concentrations of E. coli bacteria. Additionally, parameters such as Mn and UV 254 were also found to be closely related to E. coli bacteria, alongside turbidity and Fe. During the conduction of the monitoring, more attention should be paid to these parameters. Furthermore, a relationship between E. coli bacteria and different water intakes was conducted. It was observed that higher concentrations of E. coli bacteria were present when water was captured from lower water intakes, characterized by increased water turbidity.
Thus, this research provides a basis for informed management practices and policies to ensure the safety of water intended for human consumption. A part of this research also emphasises the need to establish proper monitoring and remediate surface waters for potential microbial pollution, taking into account different water quality and other relevant parameters associated with the observed ecosystem. Therefore, continuous monitoring of water quality is required for general public health.
Further analysis is focused on developing a prediction model for concentrations of E. coli bacteria considering physiochemical parameters at the inlet of of the raw water to the DWTP Butoniga. Such a model can help in devising strategies and management approaches that enhance its efficiency and resource utilization through optimization of the parameters in the treatment processes, while simultaneously mitigating the risks associated with inadequate actions.
ACKNOWLEDGEMENTS
This study was presented at the Waters in Sensitive and Protected Areas (WSPA 2024) conference held in Pula, Croatia. This research was supported by the projects: ‘Sustainable river basin management by the implementation of innovative methodologies, approaches and tools’ (uniri-tehnic-18-129) and ‘Hydrology of water resources and identification of flood and mudflow risk in karst’ (uniri-tehnic-18-54). This research was also supported under the project line ZIP UNIRI of the University of Rijeka, for the projects ZIP-UNIRI-1500-3-22 and ZIP-UNIRI-1500-2-22. The authors greatly appreciate the comments of two anonymous reviewers for their valuable comments and reviewing of the manuscript, which substantially improved the paper.
FUNDING
This research was funded under the project line ZIP UNIRI of the University of Rijeka, for the project ZIP-UNIRI-1500-3-22.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.