Modelling of enteric bacteria concentrations in a river in Akure Nigeria

BOD

Biological oxygen demand

DO

Dissolved oxygen

pH

Potentiale hydrogenae

TSS

Total suspended solids

m-FC

Membrane faecal coliform agar

MLSA

Membrane lauryl sulphate agar

EMB

Eosin methylene blue agar

m-Ent

Membrane Enterococcus agar

m-CP

Membrane Clostridium perfringens agar

BSM

Bifidobacterium selective medium

SSA

Salmonella-Shigella agar

CFU

Colony-forming units

SPSS

Statistical Package for Social Sciences

R

Correlation coefficient

GIS

Geographical Information System

NTU

Nephelometric turbidity Unit

EC

Electrical conductivity;

WHO

World Health Organization

CEU

Council of the European Union

SRC

Sulfite-reducing Clostridium

VWREC

Vancouver Water Resources Education Center

USEPA

United States Environmental Protection Agency

With the growth of human populations, commercial and industrial activities, surface water has received large amounts of pollutants from variety of sources (Rodrigues & Cunha 2017). Enteric bacteria in surface water present an important health risk because approximately 80% of all diseases and death in low-income countries are water related as a result of contact or ingestion of polluted water (Augustyn et al. 2016). Adequate modelling of the survival of enteric bacteria in surface water may provide a valuable tool in assessing water quality. Surface waters are prone to pollution due to poor waste management and anthropogenic activities.

In Nigeria, indiscriminate disposal of municipal wastes remains a major threat to the microbial and physicochemical qualities of many water resources. In most cases, sewage and wastewater from homes are channeled directly into rivers. Jaji et al. (2007) observed elevated water quality parameters (such as turbidity, dissolved oxygen, total dissolved solids etc.) in some sampling locations in Ogun River. These were partly attributed to the activities of abattoir located close to the River at a notable market in Abeokuta metropolis. A study by Faremi et al. (2021) on the impact of sawmill activities on the water quality of River Benin reported high biological oxygen demand (BOD) and low dissolved oxygen (DO) values at the discharge point of the wastes into the river. The impact of point source pollution from sewage treatment oxidation pond on a receiving stream studied by Ogunfowokan et al. (2005) revealed significant elevation of water indices such as pH, BOD, nitrate, phosphate and total suspended solids (TSS).

Oxygen depletion in water bodies may cause fish death while increase in BOD signifies high load of organic matter. Furthermore, organic matter decomposition in surface water may produce inorganic nutrients such as ammonia, nitrate and phosphorus that may eventually lead to eutrophication and other ecological problems (Ogunfowokan et al. 2005). Understanding and monitoring surface water quality of a region remains a better tool towards promoting sustainable development of water resources within the societal economic and conservational contextual need (Ifabiyi 2000). The assessment of human activities capable of affecting the quality of river water within an urban area is important because per capita water demand is increasing while accessibility to available freshwater availability is on the decrease (Francis et al. 2015; Pak et al. 2021).

Linear regression provides information on the most meaningful variables that bring surface water quality variation and examine the relationship between single depended variables and a set of independent variables to best represent relationship in a population. The aim of this study was to develop a risk prediction model for assessment of faecal pollution in River Ala in Akure, Nigeria. The objectives were to determine the loads of enteric bacteria in water samples from River Ala; examine the meteorological and physicochemical characteristics of the water samples; assess the effect of meteorological and physicochemical characteristics on the load of entericbacteria in the water samples; determine the applicability of risk prediction model to River Ala and develop a risk map for River Ala.

The source of River Ala and its many tributaries are from River Ogbese, south-western Nigeria. River Ala with total length of approximately 57 km has a length of about 14.8 km within Akure Township. The river is approximately 3.6–6.5 meters wide and 1.4–2.5 meters deep. It flows downstream from the North-western part of Akure city and flows towards the southeastern part of the city through Oba-Ile to Edo State (Pak et al. 2021).Human population (approximately 120,000) residing around River Ala in Akure, Nigeria depend on the water from the river for various purposes such as domestic, recreation, agriculture etc. The sampling points were selected based on various anthropogenic activities in and around the river. These include dumping of household effluents, livestock waste, municipal waste, runoff of fertilizers and agrochemicals from farmlands, windborne debris and falling leaves.

Membrane filtration method (Anon. (2014) ISO 9308-1) was employed to determine the load of enteric bacteria in water samples from River Ala. The membrane filter was placed on membrane filter setup and 1 ml of water was filtered through on each occasion. The membrane filters were placed on freshly prepared selective media: membrane faecal coliform agar (m-FC), membrane lauryl sulphate agar (MLSA), eosin methylene blue agar (EMB), membrane Enterococcus agar (m-Ent), membrane Clostridium perfringens (m-CP) agar, Bifidobacterium selective medium (BSM) and Salmonella-Shigella agar (SSA). Agar plates were incubated at 37 °C for 24 hours (MLSA, EMB, SSA), 44 °C for 24 hours (m-FC), 37 °C for 48 hours (m-Ent) and 37 °C for 24 (MCP) in anaerobic jar with desiccants. Colonies were counted, calculated and expressed as colony-forming units (CFU) per 100 ml of water.

The physicochemical (temperature and pH) properties of all water samples were determined weekly at the point of collection on-site using a handheld mercury in glass thermometer and digital handheld pH meter (HANNA) pre-calibrated with buffer solutions. Other properties such as dissolved oxygen, electrical conductivity, turbidity, salinity and total dissolved solids were measured using a multi-parameter instrument in Marine Science Department, Federal University of Technology, Akure. Meteorological information such as the amount of rainfall was obtained from Meteorology Department, Federal University of Technology, Akure.

Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) version 20 software. Data obtained were converted to logarithm, reported in colony forming unit per 100 ml of water and subjected to descriptive statistical analysis (95% confidence interval). Correlation coefficient (R) measures the degree of association that exists between two variables, one taken as dependent variable. Correlation between the physicochemical properties of the water samples and the concentrations of enteric bacteria was determined using a 2-tailed Pearson's correlation analysis. Correlations and test of significance were considered statistically significant at 95% confidence interval

Linear regression is one of the most flexible statistical tools that allow the modeling of multiple influences on an outcome. Linear regression models for parameters exhibiting significant correlations were developed by fitting a linear equation to the observed data set. The regression equation was used as a mathematical tool to calculate different dependent characteristics of water quality by substituting the values for the independent parameters in the equations. Furthermore, the Geographical Information System (GIS) was employed and used to determine the representative monitoring points at high risk of disease outbreak, thereafter, risk maps for River Ala were developed.

The mean viable load of enteric bacteria in water samples collected from the representative monitoring points on River Ala showed that Apatapiti and South gate had the highest mean faecal coliforms count, while Alagbaka and Fiwasaye had the lowest mean faecal coliforms count. The highest mean count of E. coli was observed at Leo, while the least mean count of E. coli was observed at Adegbola. Similarly, the highest mean count of Clostridium was observed at Leo, while the least mean count of Clostridium was observed at Adegbola (Table 1). The highest mean count of Bifidobacterium was observed at Oke-Ijebu, while the least mean count of Bifidobacterium was observed at Fiwasaye. Similarly, the highest mean count of enterococci was observed at Apatapiti, while the least mean count of enterococci was observed at Alagbaka (Table 1). The highest mean count of Shigella was observed at Southgate, while the least mean count of Shigella was observed at Fiwasaye. Similarly, the highest mean count of Salmonella was observed at Alagbaka, while the least mean count of Salmonella was observed at Southgate. In general, all the representative monitoring points on River Ala had high levels of enteric organisms ranging from 4 to 5 log (Table 1).

The mean values of temperature of water samples from the monitoring points in River Ala ranged from 26.75 to 27.50 °C while the mean values of pH ranged from 6.50 to 7.31. Likewise, the mean values of salinity ranged from 0.07 to 0.11 ppt, whereas the mean values of turbidity ranged from 25.10 and 36.86 NTU with samples obtained from Oke-Ijebu having the highest turbidity and samples obtained from Alagbaka had the least turbidity. The mean values of total dissolved solids ranged from 61.81 mg/l to 146.06 mg/l. Similarly, the mean values of electrical conductivity ranged from 155.06 μS/cm to 237.12 μS/cm. In addition, the mean values of DO ranged from 1.25 mg/l to 5.04 mg/l. The amount of rainfall in the catchment for the monitoring points for 48, 24, 12 and 1 h showed that the highest mean rainfall value for the monitoring points was 34.4 mm at 12 h before sample collection and the least mean rainfall value was 0.6 mm at 1 h before sample collection (Table 2).

The Pearson's correlation analysis revealed that faecal coliforms and E. coli correlated positively with turbidity (r = 0.76, p < 0.05) and (r = 0.62, p < 0.05), respectively. Similarly, faecal coliforms, E. coli, intestinal enterococci correlated positively with rainfall at 12 h (r = 0.72, p < 0.05), (r = 0.73, p < 0.05) and (r = 0.54, p < 0.05), respectively (Table 3).

The modelling of the concentration of enteric bacteria, physicochemical characteristics and meteorological profile of the water samples from River Ala in relation to seven representative monitoring points was conducted in this study. River Ala is important surface water in Akure, where residents around the river rely on it for several purposes including domestic, recreational and agricultural purposes. Monitoring the microbiological quality of the river is, therefore, a necessary step towards achieving the United Nation's Sustainable Development Goal 6. Escherichia coli is a member of the faecal coliform group and is a more specific indicator of faecal pollution than other faecal coliforms. The presence of E. coli in River Ala may be due to improper management and disposal of sewage and waste from the pigsty sited (i.e pig farming) beside the water body. The mean viable count of E. coli observed in this study appeared to be higher than those obtained by Olalemi et al. (2020) where the authors reported that the highest mean load of E. coli in a river was 4.35 log₁₀ CFU 100 ml⁻¹ in the comparative hazard evaluation of enteric bacteria in two surface water sources in Akure, Nigeria. Jaji et al. (2007) reported that indiscriminate disposal of municipal wastes remain a major threat to surface water pollution in Nigeria. Apatapiti and Southgate appeared to be more faecally-polluted than the other representative monitoring points because of their higher concentration of faecal coliforms. The load of faecal coliforms was above permissible limits for drinking water in all the representative monitoring points where drinking water should contain less than one E. coli or faecal coliform CFU/100 ml (WHO 2017). Whereas, water intended for bathing, recreation, or aquaculture should contain less than or equal to 100 E. coli or faecal coliforms CFU/100 ml (CEU 2006).

In addition, the high load of faecal coliforms in water samples from Apatapiti may likely be as a result of indiscriminate channeling of sewage into the river. The mean concentration of faecal coliforms in water samples from River Ala appeared to be higher than those obtained by Abdul et al. (2007) where the authors reported a mean range of 1.46–2.47 log₁₀ CFU 100 ml⁻¹ in river water samples obtained from Khairpur City, Sindh, Pakistan. This finding is in agreement with Nuru et al. (2019) who reported the presence of pathogenic organisms and indicator organisms in water samples obtained from a river in Gombe State, Nigeria.

Enterococci are usually considered appropriate indicators of faecal contamination in water that may be of human or animal origin. Wu et al. (2011) observed that enterococci were more likely to correlate with pathogens, above other indicators. Intestinal enterococci are very effective in determining water quality as they survive longer in water than E. coli; they do not multiply in water and also resist chlorination and drying (WHO 2006). In this study, the load of intestinal enterococci in the water samples from River Ala were higher than those observed by Olalemi et al. (2020), where the authors reported a lower load of intestinal enterococci of 2.52 log₁₀ CFU 100 ml⁻¹ in River Malaika located in Northgate, Akure, Nigeria. The concentrations were also higher than those observed by Augustyn et al. (2016) where the authors reported an average load of enterococci between 1.88 log₁₀ CFU 100 ml⁻¹ and 2.20 log₁₀ CFU 100 ml⁻¹ in the inflow course of River Ganga, India. The load of intestinal enterococci in water samples from River Ala clearly do not meet the World Health Organization standard of 0 CFU/100 ml of enterococci in water meant for drinking (WHO 2017).

Bifidobacteria have been recommended as potential indicators of human faecal pollution in surface waters (Regina et al. 2008). Bifidobacteria are exclusively of faecal origin and some species occur in human faeces more than E. coli. Due to their physiology and complex growth requirements, Bifidobacteria are unlikely to grow in water, in contrast with E. coli (Rodrigues & Cunha 2017). Despite the advantages of Bifidobacterium spp. to act as potential indicators of faecal contamination in surface waters, there are limited information on their extra-enteric behavior and persistence in water environments (Savichtcheva & Okabe 2006). The concentration of Bifidobacterium in water samples from River Ala may be as a result of anthropogenic activities and surface runoffs following rain and storm events that carry faecal wastes into the river thus, increasing the level of microbial pathogens and ultimately impacting the sanitary quality of the river negatively (Wilkes et al. 2013).

Shigella species have been reported to be the major cause of shigellosis or bacillary dysentery affecting, in particular, immuno compromised individuals, children and the elderly (Shahin et al. 2019). Incidence of shigellosis outbreaks as a result of poor water quality have been reported throughout the world (WHO 2019). Generally, river water harbour a vast majority of enteropathogens derived from municipal sewage discharges, rainfall runoffs from agricultural farms and faecal waste from humans, pets, farm animals and wildlife (Asadullah & Seema 2013). The high level of Shigella in water samples from River Ala may most likely be as a result of high influx of raw sewage into the river and runoff from farmlands, industries and markets located close to the river due to rainfall. This finding is in agreement with Shahin et al. (2019) on the characterization of Shigella species isolated from river catchments in North West province of South Africa where the authors reported high concentration of Shigella in the water samples from the river catchment.

The presence of Salmonella in water sources can be a serious danger for the human and animal community. Huanli et al. (2018) reported that surface waters are more exposed to environmental events such as discharge of sewage, rainfall, animal husbandry, and wildlife, and thus are more susceptible to contamination. River water, however, has been shown to be one of the largest reservoirs of viable Salmonella. Outbreak of diseases caused by Salmonella may occur as a result of ingestion of faecal-impacted water and this may lead to morbidity with great economic losses and mortality (WHO 2006; Fagbayide & Abulude 2018). In this study, water samples from Alagbaka possesses greater risk associated with Salmonella than other representative monitoring points.

Spore-forming, sulfite-reducing Clostridium (SRC) species such as C. perfringens have been used as an indicator of faecal pollution in water because they are the most dominant of all the anaerobes in the gastrointestinal tract of humans and warm-blooded animals (Cabral 2010; Figueras & Borrego 2010). Consequently, the presence of anaerobes in surface water environments are usually linked to poorly treated wastewater effluent from wastewater treatment plants (Marcheggiani et al. 2008). Clostridia do not replicate in surface water, but has been found to be stable in these aquatic environments due to their spore-forming abilities (Cabral 2010). These spores are extremely resistant to harsh environmental conditions such as pH and temperature extremes and UV radiation, and most importantly, disinfection treatment processes (Holcomb & Stewart 2020). The concentrations of Clostridium in water samples from River Ala were above permissible limits of less than or equal to 100 CFU/100 ml for water intended for bathing, recreation, or aquaculture (CEU 2006).

Water temperature is an important factor that influences the rate of all biological activities (Patil et al. 2012). The mean temperature of water samples from River Ala that ranged from 26.6 to 27.5 °C is adequate for microbial growth. This agrees with Okoye & Okoye (2008), who reported that high water temperature enhances the growth of microorganisms and hence, taste, odour, colour and corrosion problem. The pH of water is one of the important indicators in water quality assessment and it is a reflection of hydro-chemical characteristics that can directly reflect the state of water pollution. The pH values of the water samples from River Ala were within the permissible range of 6.5–8.5 (WHO 2017). The concentration of the salts determines whether the water is of high quality (usable for irrigation without need for special precautions) or of low quality (brackish or saline). Salinity becomes a concern when an excessive amount or concentration of soluble salts occurs in the water, either naturally or as a result of mismanagement of surface water (El-Swaify 2000). Small amounts of dissolved salts in natural waters are vital for the life of aquatic plants and animals, higher levels of salinity alter the way the water can be used yet even the most hyper-saline water can be used for some purposes.

The salinity values observed in the water samples from River Alawere within the permissible set limit of 250 mg/l for surface water (WHO 2011). High turbidity values indicate the possible presence of microorganisms, clays, silts and other suspended solids in water, which affect its aesthetic value (Addo-Bediako et al. 2018). The level of turbidity is also dependent on factors such as the presence of humic substances produced through decomposition of certain organic matter and intensity of suspended soil particles (Njoku et al. 2015; Aladese & Pondei 2021). The mean turbidity values of the water samples from River Ala that ranged from 25.10 to 36.86 NTU appeared to be very high. Dissolved oxygen is the oxygen present in water that is available to aquatic organisms as reported by Vancouver Water Resources Education Center (VWREC 2019). It is a fundamental factor for metabolism of the aerobic aquatic organisms, and it is important in determining the natural self-purification capacity or the degree of freshness of a river (Rabee et al. 2011; Naubi et al. 2016).The mean values of DO observed in the water samples from River Ala appeared to be very low compared with acceptable limit of 5 mg/l (WHO 2002; USEPA 2009). Dissolved oxygen content of water is influenced by water source, microbial load and temperature. Depletion of dissolved oxygen in water may lead to reduction of certain nutrients such as sulphate and nitrate.

TDS is the sum total of all of the dissolved substances in a given water body and include hardness, alkalinity, chlorides, bromides, sulfates, silicates and all manner of organic compounds (Islam et al. 2017). Although elevated TDS concentration may not mean that the water is hazardous to health, it does mean the water may have aesthetic hitches or cause nuisance problems (Adekunle et al. 2007). The TDS in water samples from River Ala were lower than the 500 mg/l standard of EPA (2002).

Sudden increase or decrease in electrical conductivity in a body of water may indicate pollution. Agricultural runoff or a sewage leak may be the primary cause of rise in conductivity due to the additional chloride, phosphate and nitrate ions (Mihir et al. 2015). The values of electrical conductivity in water samples from River Ala were below the permissible limit of 1,000 μs/cm for water intended for anthropogenic activities (WHO 2017). In terms of EC, the water from River Ala may be suitable for domestic use.

Heavy rainfall, especially after a long dry period that allow more pollutants accumulated in runoff pathways and available for transport by runoff, could wash-off a lot of pollutants into rivers (Bae 2013; Rostami et al. 2017). The amount of rainfall 12 and 1 h prior collection of samples from River Ala influenced negatively the water quality of River Ala. Tornevi et al. (2014) reported that rainfall elevates microbial risks in surface water and act as the main driver of varying water quality. The authors further stated that heavy rainfall appears to be a better predictor of faecal pollution than water turbidity. Ingun et al. (2014) assessed the impact of rainfall on the hygienic quality of blue mussels and water in urban areas in the Inner Oslofjord, Norway and reported that heavy precipitation of rainfall during the 48 hours prior sampling had a negative effect on the hygienic quality of the water, at all the 5 sampling sites. In another related study, Lafforgue et al. (2018) developed a coupled (terrestrial and marine) model for identifying sources of microbial pollution at Baie des Veys catchment in Normandy, France and observed that rainfall had a major impact on enterobacterial loads entering the seawater.

Pearson's correlation showed high positive relationships between faecal coliforms, E. coli, enterococci, turbidity and rainfall. This observation is in agreement with the finding of Matthews et al. (2012) where the authors observed that turbidity correlated positively with levels of enteric bacteria in water. Many studies have demonstrated the positive relationships between enteric bacteria, turbidity and rainfall (Aladese & Pondei 2021; Mijin et al. 2019; Olalemi et al. 2020, 2021). Linear regression analysis is one of the modelling techniques used to describe and predict relationships between microbial water quality and physicochemical properties by fitting a linear equation to the observed data set. Finding linear correlation between various physicochemical, bacteriological and meteorological water parameters can be treated as a unique step towards water quality management (Bhaswati & Bibhash 2018). In this study, a significant relationship obtained from a systematic correlation among the observed data set that showed strong correlation (i.e., faecal coliforms, enterococci, E. coli, turbidity and rainfall) were subjected to linear regression analysis. The regression analyses carried out for the water quality parameters found to have better and higher level of significance in their correlation coefficient. The linear regression model for turbidity as function of faecal coliforms validation demonstrated satisfactory predictive values based on the R² value compared with the linear regression equation for turbidity as function of E. coli with lesser R² value. Furthermore, rainfall at 12 h before sampling day showed high predictive values with faecal coliforms and E. coli. The method of linear correlation has been found to be a significant approach to get an idea of quality of rivers by determining a few parameters experimentally (Soni & Mohan 2015). It was observed that turbidity and rainfall are important physicochemical and meteorological parameters of river water quality because they correlated with the faecal indicator bacteria.Risk maps for the representative monitoring points in this study showed areas highly prone to faecal pollution. Disease risk mapping showed a visual representation of the geographical distribution of disease within the population, this is a valuable tool in determining the potential source and areas greatly affected by a disease outbreak. The maps revealed that no region in the representative monitoring point is entirely unaffected by human activities and that each river catchment area can be differentiated by human and environmental conditions. The physical and microbiological analysis of water samples confirmed that there was an input of bacterial and other pathogens into River Ala. Studies have shown that the degree of pollution correlates with the level of activity in the catchment areas and with rainfall patterns (Kistemann et al. 2001; Asadullah & Seema 2013).

The findings of this study revealed high levels of faecal contamination at the different representative monitoring points in River Ala in Akure. Physicochemical and meteorological parameters influenced the load of enteric bacteria as well as rainfall at 12 h before sampling showed high predictive values for the levels of faecal coliforms and E. coli in water from River Ala. The risk maps showed areas highly prone to faecal pollution and revealed that no region in the representative monitoring point is entirely unaffected by human activities and that each river catchment area can be differentiated by human and environmental conditions.

The authors are grateful to the Department of Microbiology, School of Life Sciences, The Federal University of Technology, Akure, Ondo State, Nigeria for providing appropriate support in terms of equipment and laboratory used for the study.

Not applicable

Not applicable

None was received

‘O.A.’designed and supervised the study.’O.O.’ developed the methodology and acquired the data, analyzed the data and interpreted the data. ‘O.O.’ wrote the manuscript text, ‘O.A.’ corrected and reviewed the manuscript and provided administrative support. Both authors read and approved the final manuscript.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.