Quality Assessment of Three Types of Drinking Water Sources in Guinea-Bissau

International Journal of Environmental Research and Public Health Article Quality Assessment of Three Types of Drinking Water Sources in Guinea-Bissau Aducabe Bancessi 1,2,* , Luís Catarino 2, Maria JoséSilva 3, Armindo Ferreira 4, Elizabeth Duarte 5and Teresa Nazareth 6 1Nova School of Business and Economics, Nova University of Lisbon, Campus de Carcavelos, Rua da Holanda, n.1, 2775-405 Lisbon, Portugal 2 Centre for Ecology, Evolution and Environmental Changes (cE3c), Faculty of Sciences, University of Lisbon, Campo Grande, 1749-016 Lisbon, Portugal; [email protected] 3Plant-Environment Interactions & Biodiversity Lab (PlantStress&Biodiversity), Linking Landscape, Environment, Agriculture and Food Unit (LEAF), Institute of Agronomy ISA, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal; [email protected] 4 National Laboratory of Public Health, National Institute of Public Health (INASA), Avenida Combatentes da Liberdade da Pátria, Bissau 1004, Guinea-Bissau; armindoferr[email protected] 5Department of Sciences and Engineering of Biosystems, Institute of Agronomy ISA, University of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal; [email protected] 6Global Health and Tropical Medicine, Institute of Hygiene and Tropical Medicine, Nova University of Lisbon, Rua da Junqueira 100, 1349-008 Lisbon, Portugal; teresa.lobo.nazar[email protected] *Correspondence: [email protected] Received: 20 August 2020; Accepted: 30 September 2020; Published: 4 October 2020   Abstract: The lack of access to safe drinking water causes important health problems, mainly in developing countries. In the West African country Guinea-Bissau, waterborne diseases are recognised by WHO as major infectious diseases. This study analysed the microbiological and physicochemical parameters of drinking water in the capital Bissau and its surroundings. Twenty-two sites belonging to different water sources (piped water, tubewells and shallow wells) were surveyed twice a day for three weeks, in both dry and wet seasons. Most of the microbiological parameters were out of the acceptable ranges in all types of water and both seasons and tended to worsen in the wet season. Moreover, in Bissau, the levels of faecal contamination in piped water increased from the holes to the consumer (tap/fountain). Several physicochemical variables showed values out of the internationally accepted ranges. Both well sources showed low-pH water (4.87–5.59), with high nitrite and iron levels in the wet season and high hexavalent chromium concentration in the dry season. The residual chlorine never reached the minimum recommended level in any of the water sources or seasons, suggesting a high risk of contamination. Results reveal a lack of quality in the three water sources analysed, coherent with the high number of diarrheal cases in the country. There is an urgent need to improve sanitarian conditions to reduce the disease burden caused by these waterborne illnesses. Keywords: West Africa; water quality; E. coli; physicochemical; microbiological Int. J. Environ. Res. Public Health 2020,17, 7254; doi:10.3390/ijerph17197254 www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2020,17, 7254 2 of 15 1. Introduction The access to safe drinking water should be of main concern in any society since water is a basic need for human development, health and well-being [ 1 ], but approximately 1.1 billion people in rural and peri-urban communities of developing countries do not have that access [ 2 ]. In rural areas, the lack of adequate safe water and sanitary infrastructures leaves millions with water of doubtful quality, increasing the harshness of daily life [ 3 , 4 ]. The mortality from diarrhoea-related diseases worldwide extents to 2.2 million people each year [ 5 – 7 ]. With a rapidly growing global population, increasing environmental degradation and the multifaceted impacts of climate change, water demand is expected to increase dramatically by nearly one-third in all major use sectors by 2050 [ 8 – 10 ]. In sub-Saharan Africa, the situation (water quantity and quality) is particularly acute due to global warming, the expansion of the Sahara Desert, civil unrest and poor governance, population growth, migration and poverty [ 11 ]. In Guinea-Bissau, West Africa, one of the poorest countries in the world according to the Human Development Index, ranking 178th among 189 countries in 2018, life expectancy at birth is 58 (male) and 61 years (female), with over 50% of the population living below the poverty line, and diarrhoea diseases being the third leading cause of premature death in 2016 [ 12 – 14 ]. Health infrastructures are poor and were greatly affected by the civil war in the late 1990s. For the majority of the rural population, the only source of water for daily needs, including drinking water, is shallow hand-dug wells (51%). No public wastewater treatment is available in the country and in 2016, the access to improved water sources was limited to 53% of the population, and only 19% had access to piped water [ 15 ]. Just Bissau and Bafat á cities currently have piped water distribution systems that, however, do not cover the entire population of these cities, particularly in the most peripheral areas. According to [ 3 ], each person has only 21 L of water for daily personal needs in Guinea-Bissau. This value is well below the 50 L minimum water requirement for human domestic use [16]. In 2019, 702,974 cases of diarrhoea diseases were reported in Guinea-Bissau [17]. Outbreaks of cholera from contaminated water through food and drinking water are common. Between 1994 and 2013, 83,635 cases and 1895 deaths from cholera were reported in the country. These epidemics occurred mainly during the wet season and lasted for more than six months. Cholera cases were reported in Bissau city (1996, 2002, 2005, 2012), Bijag ó s islands (1994, 2004) and Tombali (2007, 2013), which are all coastal regions, and this disease is considered endemic in the country [ 18 , 19 ]. However, there have not been documented outbreaks since 2013. The current study aimed to assess the quality of drinking water in Bissau city and its surroundings, and Quinhámel (hotspots of waterborne diseases such as diarrhoea including cholera) during the dry and wet seasons, and also to assess the potential correlation between distribution systems and waterborne diseases. 2. Materials and Methods 2.1. Study Area In 2018, the population of Guinea-Bissau was 1.87 million, with an annual growth of about 2.5% [ 20 ]. The climate is tropical sub-humid, with a mean annual temperature of 26.5 ◦ C and two seasons (dry and wet): a dry season (December to May) with north-easterly Harmattan winds and a monsoonal-type rainy season (June to November) with south-westerly winds. The most representative soil groups in the country are Ferrallisols, Plinthosols, Gleysols, Fluvisols and Arenosols [ 21 ]. A more detailed characterisation of Guinea-Bissau is presented elsewhere [ 22 ]. The cities of Bissau (including both urban and peri-urban areas) and Quinh á mel (rural area) were selected for the present study (Figure 1) because they are hotspots of waterborne diseases. Int. J. Environ. Res. Public Health 2020,17, 7254 3 of 15 Bissau has a total area of 77.5 km 2 and a population of c. 350,000 inhabitants, where just 13% of the population has access to the piped water distribution system. The piped water system is based on a set of holes in different locations in the city, each with a pump and a tank with an upstream chlorine dosing system. Each reservoir supplies one of the five distribution areas in the city (DA): Hospital 3 de Agosto (DA1); Bandim (DA2); Hospital Central Sim ã o Mendes (DA3), Queije (DA4) and Hospital Santo Egidio (DA5). In the peri-urban areas of Bissau and rural areas such as Quinh á mel with a surface of 451 km 2 , 37 km away from Bissau [ 23 ], water for human consumption is obtained in two main ways: deep tubewells with hand-pumps and open shallow wells in which water is extracted with a bucket. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 3 of 15 Figure 1. Location of the sampling points in Bissau and Quinhámel: urban (L1, L2, L3, L4, L5, L6, L16, L17, L18, L19, L20, L21, L22), peri-urban (L7, L8, L9, L10) and rural (L11, L12, L13, L14, L15). Bissau has a total area of 77.5 km2 and a population of c. 350,000 inhabitants, where just 13% of the population has access to the piped water distribution system. The piped water system is based on a set of holes in different locations in the city, each with a pump and a tank with an upstream chlorine dosing system. Each reservoir supplies one of the five distribution areas in the city (DA): Hospital 3 de Agosto (DA1); Bandim (DA2); Hospital Central Simão Mendes (DA3), Queije (DA4) and Hospital Santo Egidio (DA5). In the peri-urban areas of Bissau and rural areas such as Quinhámel with a surface of 451 km2, 37 km away from Bissau [23], water for human consumption is obtained in two main ways: deep tubewells with hand-pumps and open shallow wells in which water is extracted with a bucket. 2.2. Location of the Sampling Points and Sample Collection Six water samples were collected from each of the 22 different points (Figure 1) representing three types of drinking water sources available in Bissau and Quinhámel: piped water distribution systems (piped water, urban area), deep tubewells with hand-pumps (tubewells, rural area) and shallow hand-dug wells (shallow wells, peri-urban area). In the urban area, the distribution system was sampled at four key segments: (i) hole, (ii) reservoir outlet, (iii) tap (people’s houses) and (iv) fountains (Table 1). In the peri-urban area, which lacks piped water, samples were collected from the 4 major shallow wells, which supply most of the population. In the rural area, samples were collected from the 5 major tubewells, which supply less than half of the population with water for domestic use. A total of 132 water samples were obtained in the dry (Abril and May) and wet (August and September) seasons of 2019. The collection was performed twice a day (at 7 am and 1 pm) at each sampling point, once a week for three weeks, following the collection guidelines [24]. The exact position of each sampling point was obtained through GPS (Garmin GPSMAP 64s) and mapped using GIS software (QGis 3.10 Girona). Water samples were collected using 500 mL plastic sterile flasks. After collection, they were refrigerated with locally produced shredded ice and transported to the laboratory for analysis within 6 h. Figure 1. Location of the sampling points in Bissau and Quinh á mel: urban (L1, L2, L3, L4, L5, L6, L16, L17, L18, L19, L20, L21, L22), peri-urban (L7, L8, L9, L10) and rural (L11, L12, L13, L14, L15). 2.2. Location of the Sampling Points and Sample Collection Six water samples were collected from each of the 22 different points (Figure 1) representing three types of drinking water sources available in Bissau and Quinh á mel: piped water distribution systems (piped water, urban area), deep tubewells with hand-pumps (tubewells, rural area) and shallow hand-dug wells (shallow wells, peri-urban area). In the urban area, the distribution system was sampled at four key segments: (i) hole, (ii) reservoir outlet, (iii) tap (people’s houses) and (iv) fountains (Table 1). In the peri-urban area, which lacks piped water, samples were collected from the 4 major shallow wells, which supply most of the population. In the rural area, samples were collected from the 5 major tubewells, which supply less than half of the population with water for domestic use. A total of 132 water samples were obtained in the dry (Abril and May) and wet (August and September) seasons of 2019. The collection was performed twice a day (at 7 am and 1 pm) at each sampling point, once a week for three weeks, following the collection guidelines [ 24 ]. The exact position of each sampling point was obtained through GPS (Garmin GPSMAP 64s) and mapped using GIS software (QGis 3.10 Girona). Water samples were collected using 500 mL plastic sterile flasks. After collection, they were refrigerated with locally produced shredded ice and transported to the laboratory for analysis within 6 h. Int. J. Environ. Res. Public Health 2020,17, 7254 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. - Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. DA2 Bandim Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. DA3 Hospital Central Simão Mendes - Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. - Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. DA4 Queije - - Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. - DA5 Hospital S. Egidio Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. - Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 4 of 15 Table 1. Sampled segments of the piped water of each of the five distribution systems in Bissau (urban area). Name Distribution Area Sampled Segment Hole Reservoir Outlet Tap Fountain DA1 Hospital 3 de Agosto - DA2 Bandim DA3 Hospital Central Simão Mendes - - DA4 Queije - - - DA5 Hospital S. Egidio - - sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO3−), nitrite (NO2−), hexavalent chromium (CrVI), iron (Fe2+), sulphates (SO₄2-), total phosphorus (TP), total alkalinity (TA), copper (Cu2+), sulphite (SO₃2⁻), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 μm pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 °C for 24 h (FC) or 48 h (IE), or 37 °C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 °C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [25], the European Council Directive 98/83/EC [26] and the WHO recommendations [27] were adopted in this study. Table 2 and 3 shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. — sampled; - not sampled. 2.3. Analytical Procedures All the equipment used belongs to the National Laboratory of Public Health (LNSP) in Bissau. The consumables needed for bacteriological analyses were acquired in Portugal. 2.3.1. Physical and Chemical Parameters Temperature (T), electrical conductivity (EC), turbidity, dissolved oxygen (DO), oxidation–reduction potential (ORP), salinity, total dissolved solids (TDS) and pH were measured in situ (sampling point), using a multiparameter meter (HI9829, Hanna Instruments, Woonsocket, RI, USA). Nitrate (NO 3− ), nitrite (NO 2− ), hexavalent chromium (CrVI), iron (Fe 2+ ), sulphates (SO 42− ), total phosphorus (TP), total alkalinity (TA), copper (Cu 2+ ), sulphite (SO 32− ), hardness and residual chlorine (RC) were assayed with Palintest water analysis kits (Photometer 7100, Palintest Instruments, Halma Company, UK) according to the standard methods supplied by the manufacturer (https://www.hanna.pt and https://www.palintest.com/). 2.3.2. Bacteriological Analysis Subsamples for enumeration of faecal coliforms (FC), intestinal enterococci (IE) and Vibrio spp. were filtered onto sterile cellulose nitrate membranes (0.22 µ m pore size, 47 mm diameter, GE Healthcare Life Sciences, Little Chalfont, UK), through a hand-pump, placed in Chromogenic Coliform Agar (Biokar diagnostics, Fr), Slanetz-Bartley Agar (Oxoid, Waltham, MA, USA) and (Vibrio Chrome agar medium, Fr) plates, respectively, and incubated at 44.5 ◦ C for 24 h (FC) or 48 h (IE), or 37 ◦ C for 24 h (Vibrio spp.). Aerobic mesophilic microorganisms (AMM) were determined according to ISO 4833: 2003, using the incorporation technique, by pipetting 1 mL of sample onto each plate with the addition of mFc-agar Yeast Extract Agar (Biokar diagnostics, Fr), and incubation at 37 ◦ C for 24 h. For FC, E. coli and IE, a volume of 100 mL was used, while for Vibrio spp., it was 300 mL. Typical colonies were counted, and the result was expressed as colony forming units (CFU/100 mL). 2.4. Statistical Analysis Spatial and seasonal statistically significant differences among samples were evaluated by one-way ANOVA analyses of variance, followed by post hoc Tukey honestly significant difference (HSD) multi-comparison tests, using Statistica 8.0 (Stat Soft, Tulsa, OK, USA). Euclidean distances, principal components and classification analysis (PCA) were used to assess relationships between the different microorganism species and physicochemical parameters. The physicochemical parameters were chosen based on the eigenvalues test and the co-correlation among multiple variables, using the software Statistica 8.0 (Stat Soft, Tulsa, OK, USA). The significance level used for all tests was 0.05. In the absence of water quality standards for Guinea-Bissau, the Nigerian standard for drinking water quality (NSDWQ) [ 25 ], the European Council Directive 98/83/EC [ 26 ] Int. J. Environ. Res. Public Health 2020,17, 7254 5 of 15 and the WHO recommendations [ 27 ] were adopted in this study. Tables 2and 3shows the skeleton ANOVA that includes the expected mean squares and variable under analysis. Table 2. Skeleton analysis of variance for Table 5—season ×distribution areas. ANOVA Univariate Tests of Significance Effective Hypothesis Decomposition DF FC E. coli FE AMM Intercept 1 0.000 *** 0.000 *** 0.000 *** 0.000 *** Season 1 0.000 *** 0.091 *** 0.091 ns 0.000 *** Distribution Area 4 0.000 *** 0.000 *** 0.000 *** 0.000 *** Season ×Distribution Area 4 0.901 ns 0.401 ns 0.401 ns 0.002 ** Error 146 DF—degrees of freedom; FC—faecal coliforms; FE—faecal enterococci; AMM—aerobic mesophilic microorganisms; ** p≤0.01; *** p≤0.001; ns—not significant. Table 3. Skeleton analysis of variance for Table 6—season ×water distribution system segment. ANOVA Univariate Tests of Significance Effective Hypothesis Decomposition DF FC E. coli FE AMM Intercept 1 0.000 *** 0.000 *** 0.000 *** 0.000 *** Season 1 0.000 *** 0.024 * 0.024 * 0.000 *** System/segment 3 0.000 *** 0.000 *** 0.000 *** 0.000 *** Season ×System/segment 3 0.091 ns 0.422 ns 0.422 ns 0.000 *** Error 148 DF—degrees of freedom; FC—faecal coliforms; FE—faecal enterococci; AMM—aerobic mesophilic microorganisms; *p≤0.05, *** p≤0.001, ns—not significant. 3. Results The water quality was analysed seasonally for each of the 22 sampling points representing three water source types: piped water, tubewells and shallow wells. The main microbiological and physicochemical results are described below. 3.1. Physicochemical Parameters Most of the physicochemical parameters varied according to the type of water source and the season (p<0.05). The majority of the values were above the acceptable limits for drinking water, with the higher values found in the wet season (Table 4), particularly in tubewell and shallow well samples. For parameters such as DO, ORP, TP and alkalinity, no established ranges are available to allow a comparison. The pH of piped water was within the acceptable range for human consumption in both seasons, contrasting with the water samples from shallow wells and tubewells, always acidic (pH 4.9–5.6) and clearly below the recommended limit of 6.5. In shallow wells, the turbidity values varied from 14.43 in the dry season to 19.96 NTU in the wet season, well above the maximum value of 5 recommended by NSDWQ. In most shallow wells, the water is taken out with buckets, which increases the turbidity. Nitrite exceeded the maximum recommended levels of 0.50 (WHO) and 0.2 mg L −1 (EU and NSDWQ) in the wet season, for the three water sources analysed. Piped water in Bissau averaged 0.62 mg L −1 , while extremely high nitrite concentrations of 2.84 and 4.66 mg L −1 were found in tubewell and shallow well waters, respectively. The levels of hexavalent Cr(VI) exceeded the limits proposed by WHO and NSDWQ, in the dry season, for the three water sources, with shallow wells revealing extremely high values up to 0.37 mg L −1 in the wet season. The recommended maximum iron content is 0.2 mg L −1 L (WHO and EU), but higher values were observed in the wet season in tubewell (1.80 mg L −1 ) and shallow well (3.51 mg L −1 ) water samples. The residual chlorine never reached the levels recommended for drinking water by WHO and the EU in either season or water Int. J. Environ. Res. Public Health 2020,17, 7254 6 of 15 source analysed. Further, the values found for piped water in Bissau were not different from the ones for untreated water (see Supplementary Materials. Raw data tables). Table 4. Mean values (n=6) of the major water physicochemical parameters analysed during the dry and wet season at Bissau (urban and peri-urban area) and Quinh á mel (rural area), Guinea-Bissau. Limits for drinking water are indicated, as recommended by WHO, the EU and NSDWQ. Values in bold are out of range. Parameter Unit Dry Wet WHO/EU NSDWQ Water Sources Piped Water Tubewell Shallow Wells Piped Water Tubewell Shallow Wells pH - 8.02 4.89 5.59 8.23 5.46 4.87 ≥ 6.5– ≤ 9.5 ≥ 6.5– ≤ 8.5 T◦C 29.9 28.2 27.5 30.5 28.0 27.7 - b Salinity ppm 0.25 0.07 0.04 0.25 0.07 0.06 - - Turbidity NTU 0.84 1.91 14.43 1.56 1.95 19.96 a <5 EC (µs cm−1)521 118 151 531 152 125 <2500 <1000 DO mgL−13.91 4.65 5.21 0.00 0.00 1.06 - - ORP mV 121.24 225.90 187.87 24.98 43.20 42.47 - - TDS mgL−1224.1 61.1 73.7 269.7 76.4 9 - <500 Nitrite (NO2−)mgL−10.01 0.01 0.02 0.62 2.84 4.66 <0.5 <0.2 Nitrate (NO3−)mgL−10.89 0.95 4.40 0.00 0.04 0.14 <50 <50 Chromium (CrVI) mgL−10.12 0.07 0.12 0.00 0.01 0.37 <0.05 <0.05 Iron (Fe2+) mgL−10.08 0.01 0.20 0.17 1.80 3.51 <0.2 <0.3 Sulphate (SO42−) mgL−11.07 1.81 3.03 0.61 0.21 0.18 - <100 Sulphite (SO32−) mgL−115 14 13 17 15 15 <250 - P (TP) mgL−113.34 0.18 0.11 0.61 0.21 0.18 - - Alkalinity (TA) mgL−1208.9 6.2 22.6 234.3 8.5 28.0 - - Copper (Cu2+) mgL−10.04 0.01 0.10 0.04 0.01 0.09 <2<1 Hardness mgL−18.58 26.49 18.08 5.19 27.96 18.49 - <150 RC mgL−10.16 0.10 0.13 0.14 0.08 0.15 0.2–1 0.2–0.25 EU—European Union parametric values for drinking water [ 26 ]; WHO—World Health Organization guideline values [ 28 ]; NSDWQ—Nigerian Standard for Drinking Water Quality [ 25 ]; T—temperature; EC—electrical conductivity; DO—dissolved oxygen; ORP—oxidation–reduction potential; TDS—total dissolved solids; RC—residual chlorine; a—acceptable to consumer; b—room temperature. WHO/EU and NSDWQ—the maximum recommended values for drinking water. 3.2. Microbiological Assessment 3.2.1. Microbiological Quality of the Three Sources of Drinking Water: Piped Water, Tubewells and Shallow Wells For the three drinking water sources available to the population, the results showed that the majority of the microbiological variables failed to meet the drinking water standards recommended by WHO and the EU; the only exception was IE in piped water, during the dry season, and in tubewells, during the wet season (Table 5). Overall, wells showed higher values than piped water in Bissau, in both seasons, although shallow wells exhibited very high microbial contamination in both seasons. No significant differences (p>0.05) were detected between the two main types of water delivery in Bissau, public fountain and house tap. Despite the higher AMM concentration observed in piped water, the FC in both seasons was lower than in tubewells. FC, E. coli and AMM were generally higher during the wet season in all the water sources available for human consumption. Only IE detected in tubewells and shallow wells increased in the dry season, with high values observed in shallow wells (see Supplementary Materials. Raw data tables). Int. J. Environ. Res. Public Health 2020,17, 7254 7 of 15 3.2.2. Piped Water Distribution Systems Table 5presents the values of microbiological parameters recorded in the dry and wet seasons, according to the segments of the piped water distribution system: pre-consumer water (hole, reservoir outlet) and consumer water (house tap, public fountain). In all the sampled points, quality varied with the season (p<0.05) with the wet season presenting the worst condition. FC, IE and E. coli were present in all the sampling points, with exception of the holes. However, AMM detected in both seasons exceeded the WHO and EU acceptable limits for drinking water and values tended to increase significantly in the wet season. Overall, the reservoir outlets, taps and fountains were the sampling points with the highest contamination levels, and the values for consumer water are not significantly different. It is important to notice that all the microbiological parameters appeared better in the dry season (p<0.05) than in the wet season and that the microbiological quality of water decreased from the source to the final consumer. In addition, all the recorded values were beyond the acceptable limits for drinking water according to the EU and WHO. IE, Vibrio cholerae, V. parahaemolyticus and V. vulnificus were not detected in any sample (see Table 3). Table 5. Means ± standard deviations (n=6) and ANOVA results for the comparison of microbiological parameters between the segments of the water distribution system in Bissau, in the dry and wet seasons. Values (CFU/100 mL) in bold are out of range. Season Parameter Hole Reservoir Outlet Tap Fountain WHO/EU Dry E. coli 0±0 bA 0.7 ±0.8 aA 0.6 ±0.8 aA 0.6 ±0.6 aA 0 FC 0 ±0 bA 6.3 ±4.2 aA 9.2 ±6.8 aA 7.5 ±2.6 aB 0 IE 0 0 0 0 0 AMM 52.9 ±38.9 bA 138.8 ±46.0 aB 126.2 ±53.7 aB 156.4 ±79.2 aB <20 Vibrio spp. 0 0 0 0 0 Wet E. coli 0±0 bA 1.0 ±0.7 aA 0.8 ±0.7 aA 1.1 ±0.8 aA 0 FC 0 ±0 bA 9.0 ±4.3 aA 12.9 ±8.1 aA 12.7 ±4.0 aA 0 IE 0 0 0 0 0 AMM 96.3 ±63.9 bA 269.7 ±42.0 aA 282.3 ±35.7 aA 240.0 ±39.7 aA <20 Vibrio spp. 0 0 0 0 0 FC—faecal coliforms; FE—faecal enterococci; AMM—aerobic mesophilic microorganisms; WHO—World Health Organization guideline values [ 27 ], EU—European Union parametric values for drinking water [ 26 ]. Mean values followed by the same letter do not differ significantly at p ≤ 0.05 (lower-case letter in a row and upper-case letter in a column). WHO/EU—the maximum values for drinking water recommended. 3.2.3. Comparison between Piped Water Distribution Areas in Bissau In Table 6, it is possible to compare the microbiological parameters between the five distribution areas (DA) of Bissau city. The majority of the values exceeded the maximum recommended by WHO, in both seasons, despite the trend of worse microbiological water quality in the wet season. The measured values only complied with the international standards in the case of E. coli during the dry season at DA1. DA1 is the distribution area with lower values of microbiological contamination in the dry season but they greatly increase in the wet season, namely for AMM. DA1 was the least contaminated system, whereas the other four sampled distribution areas presented similarly high FC levels. Three of the four bacteriological indicators showed higher counts than the acceptable limits for drinking water in all samples. For the three analysed microbiological parameters in both seasons, there were significant differences (p<0.05) between the distribution areas (Table 6). Int. J. Environ. Res. Public Health 2020,17, 7254 8 of 15 Table 6. Mean values (n=6) of the microbiological parameter of drinking water in each distribution area, during the dry and wet seasons in Bissau. Values are in CFU/100 mL; values in bold are out of range. Piped Water Distribution Areas Parameter Season DA1 DA2 DA3 DA4 DA5 WHO/EU FC Dry 3.5 cA 8.5 bcA 10.3 bA 17.5 bA 5.7 aA 0 Wet 6.4 cA 13.6 bA 16.2 bA 22.3 aA 11.7 bA E. coli Dry 0.0 cA 0.6 bcA 0.8 abA 1.3 aA 1.0 aA 0 Wet 0.8 aA 1.1 aA 0.6 aA 1.3 aA 1.5 aA AMM Dry 70.6 cB 105.8 bcB 259.0 aA 197.7 aA 195.5 aA <20 Wet 298.4 aA 289.2 aA 292.8 aA 232.2 aA 240.2 aA Vibrio spp. Dry 0 0 0 0 0 0 0 Wet 0 0 0 0 0 DA1—Hospital 3 de Agosto; DA2—Bandim; DA3—Hospital Central Sim ã o Mendes; DA4—Queije; DA5—Hospital Santo Egidio. FC—faecal coliforms; IE—intestinal enterococci; AMM—aerobic mesophilic microorganisms. WHO—World Health Organization guideline values [ 27 ], EU—European Union parametric values for drinking water [ 26 ]; WHO/EU—maximum values for drinking water recommended. For each pathogen and season, means followed by the same letter are not significantly different at p ≤ 0.05. (lower-case letter in a row and upper-case letter in a column). 3.2.4. Correlation/Relationship between Physicochemical Parameters and Microorganism Abundance The possible interaction between physicochemical variables was analysed through principal components analysis (PCA) based on Euclidean distance. Figure 2shows the PCA plots showing the relationship between those parameters for each of the three water sources studied in the dry and wet seasons. For piped water, in the dry season, it seems that the development of faecal coliforms and AMM can be associated with high NO 2− levels as well as by high Cr(VI) levels, while for the wet season, the multivariate analysis suggests that the FC concentration can be associated with the levels of SO 32− as well as NO 2− and NO 3− . TDS and AMM, whose concentrations increase with pH, probably due to a greater number of bacteria, can survive at a higher pH. Concerning tubewells’ water, in the dry season, results suggest a link between AMM development and the levels of TP, SO 42− and Fe 2+ , and no relation between nitrogen compounds (NO 2− and NO 3− ), while in the wet season, FC seems to be related to the levels of particles in suspension (TDS and turbidity). AMM is correlated with Fe 2+ , Cr(VI) and NO 3− , while E. coli is correlated with NO 3− . For shallow wells in the dry season, the results suggest a relationship between all the microbiological parameters and EC, TDS and SO 32− but also NO 2− and SO 42− , while in the wet season, all but AMM microbial composition parameters, in particular, show a direct relation with Fe2+, SO42−and TP, but also with SO32−, NO2−and TA. In an overall appreciation, the stronger and more frequent relationships between physicochemical and microbiological parameters were found between faecal coliforms (including E. Coli) and IE with NO 2− , NO 3− , SO 32− and SO 42− and the variables linked to the particles in suspension (TDS and turbidity). Int. J. Environ. Res. Public Health 2020,17, 7254 9 of 15 Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 9 of 15 Figure 2. PCA plots based on Euclidean distances showing the relationship between microbiological and physicochemical parameters for piped water, tubewells and shallow wells in the dry and wet seasons. The microbiological parameters are indicated by vectors; only active variables were considered. Trb—turbidity; TA—total alkalinity; E. c.—Escherichia coli; AMM—aerobic mesophilic microorganisms; IE—intestinal enterococci; FC—faecal coliforms TP—total phosphorus; TDS—total dissolved solids. 4. Discussion Since there are no guidelines for drinking water in Guinea-Bissau, the key water quality parameters were compared to the recommendations from the European Union [26], World Health Organization [27] and Nigerian [25] guidelines for drinking water. The herein recorded values for physicochemical and microbiological parameters confirm the critical situation of this resource in Figure 2. PCA plots based on Euclidean distances showing the relationship between microbiological and physicochemical parameters for piped water, tubewells and shallow wells in the dry and wet seasons. The microbiological parameters are indicated by vectors; only active variables were considered. Trb—turbidity; TA—total alkalinity; E. c.—Escherichia coli; AMM—aerobic mesophilic microorganisms; IE—intestinal enterococci; FC—faecal coliforms TP—total phosphorus; TDS—total dissolved solids.