Study of the Influence of Physicochemical Parameters on the Water Quality Index (WQI) in the Maranhão Amazon, Brazil

Abstract

Water quality is mainly assessed using traditional water quality assessment methods that measure chemical parameters against established standards. The water quality index is used worldwide for water quality assessment. The main parameters evaluated include the total dissolved solids, electrical conductivity, nitrite, and nitrate. In this study, the WQI combined with microbiological analyses was used to assess the water quality of two rivers, Munim and Iguará. Data obtained in this study were then correlated using multivariate statistical analysis. Principal component analysis grouped the monitored sampling points into three clusters and identified temperature, Escherichia coli, and turbidity, as features correlated to the rainy season, while phosphorus, total dissolved solids, and biochemical oxygen demand are associated with the dry season. Four principal components explained 81.20% of the data variance during the studied seasons. The evaluated correlations indicated that in the rainy season, E. coli (~443.63 CFU/100 mL) and turbidity (~36.51 NTU) levels were the highest. However, in the dry season, the levels of phosphorus (~4.25 mg·L⁻¹), total dissolved solids (145.46 mg·L⁻¹), and dissolved oxygen (~9.89 mg·L⁻¹) were the highest.

1. Introduction

The pollution of groundwater is intensified in large urban centers because of the occupation of the soil by humans. As a result, various effluents are generated that return to the environment, interfering with water quality and, to a lesser extent, inducing seasonal changes. Therefore, monitoring groundwater through chemical, physical, and biological analysis is a reliable measure to assess its quality, as such measurements serve as an indicator to possible sources of contamination. Contamination can significantly change the chemical properties of water, compromising the overall balance of the system, causing economic losses, and making its consumption impractical [1,2].

The most common cause of compromised rivers and lakes is the demographic and industrial growth that has occurred in recent years and the inappropriate use of these resources [3]. The prevailing contemporary scenario is that of water misuse, which causes shortages and quality degradation, and impairs water availability for recreation, among other purposes [4].

Water is crucial for human sustenance. In fact, an adequate, affordable, and safe supply of water is indispensable for the population. The supply of clean and safe water has substantial health benefits. In the current scenario, untreated water is problematic. Generally, the most important infectious threat is water contaminated with human or animal waste. Many diseases are caused by microorganisms, including bacteria, pathogens, and viruses, which can survive, replicate, and spread in water [5,6]. According to the World Health Organization (WHO), in 2011, nearly 1.7 billion children with diarrhea induced by the consumption of contaminated water died in developing countries. In addition, globally, 525,000 children with infectious diarrhea died in 2018 due to poor water quality, sanitation, and hygienic conditions. A total of 1.9 billion people worldwide use water contaminated with fecal coliforms [7].

Escherichia coli is a Gram-negative, nonsporulating, rod-shaped pathogenic bacteria that generate gas in the culture medium after fermentation for 48 h at 35 °C [8]. E. coli was first recognized as a human pathogen in 1982 [9] and is categorized into three classes: commensal, diarrheal, and extraintestinal [10]. The bacteria in the fecal coliform group include E. coli, Citrobacter, Enterobacter, Hafnia and Klebsiella, with E. coli being the most common bacteria that normally survive in the gastrointestinal tract of warm-blooded animals [11]. E. coli can survive for 4–12 weeks in water, depending on the environmental conditions [12]. Fecal matter is a source of disease-causing agents in water, and E. coli is a bacterial indicator of water contamination that can impact rivers, marine beaches, streams, groundwater, surface water, natural water, and various activities associated with water use [13]. According to the WHO [14,15], E. coli should not be detectable in 100 mL of samples.

Water quality affects human activities and, consequently, the health of the population. Accordingly, it is necessary to quantitatively assess the characteristics of river water. The traditional method of water quality assessment involves analyzing the chemical parameters and comparing them with existing standards [16]. Each of the parameters that form the WQI has a certain weight relative to the measure of its contribution to water quality. The water quality index (WQI) is not an instrument for evaluating compliance with environmental legislation, but for communicating the environmental conditions of the waters to the public; it is calculated from the weighted product of the water quality parameters, corresponding to the variables that make up the index. Once polluted, water can be recovered and reused for various purposes. However, the quality of the water used and the specific purpose of reuse will determine the recommended levels of treatment [17].

In the evaluation of Brazilian rivers, emphasis is placed on general chemical parameters, such as water temperature, turbidity, electrical conductivity (EC), hydrogenic potential (pH), biochemical oxygen demand (BOD), nitrite (NO₂⁻), dissolved oxygen (DO), nitrate (NO₃⁻), total dissolved solids (TDS), total phosphorus (P), calcium and magnesium hardness, and thermotolerant coliforms, in comparison to organic pollutants [18,19].

Multivariate statistical analyses, such as principal component analysis (PCA), were used to study the interactions between multiple factors. PCA is widely used to analyze water quality data to obtain useful information, such as a small number of pollution factors, and to assess spaciotemporal variations [6,17,20].

The municipality of Nina Rodrigues is found in the Cerrado biome and belongs to the administrative region of Itapecuru, which belongs to the Munim watershed. The income source of the community is mainly trade, regional and national chain stores, and the fabrication of several brick ceramics using clay, which is abundant in the region. In recent years, the opening of enterprises of chicken production chains has been a highlight and topic of conversation; this is due to a modern feed manufacturing center that was established by a company, Frango Americano. However, the slaughterhouse was not constructed in the city as planned, as there were claims of problems related to water supply [21].

The innovation of this research, in general, is in the evaluation of the variation in nutrients, according to the seasonality of the region as a projection of a climate model. The results of this research may be relevant for future research on environmental impacts, water quality, seasonality, and the development of the region, taking into account the siltation and conservation of riparian forests. Basic sanitation in the region of Nina Rodrigues is scarce, and the waste from homes, businesses, and hospitals reaches the sewage system, resulting in contamination with chemical agents and various microorganisms [21]. Therefore, owing to the intense anthropic action in local rivers, the aim of this study was to assess the WQI in the Munim and Iguará rivers, based on the spatiotemporal dynamics of the physicochemical and microbiological parameters.

2. Materials and Methods

2.1. Study Area Description

The hydrographic basin of the Munim River, which merges with the Iguará River, is located in the northeast of the state of Maranhão, with its source found in the municipality of Aldeias Altas and its mouth in São José Bay, between the municipalities of Axixá and Icatu. The Munim River basin covers an area of 15,817.4 km² (03°27′58″ S, 43°54′18″ W). The study area and sampling points are presented in Figure 1.

Sampling was realized at points where there was a higher concentration of human activity in the municipality of Nina Rodrigues. Samples were taken at six points along the Munim and Iguará rivers: P1 (2°45′7.35″ S, 44° 4′49.41″ O), P2 (3°6′35.13″ S, 44°0′20.25″ O), and P5 (3°38′1.81″ S, 43°33′18.35″ O) collected along the Munim river; P3 (3°28′4.53″ S, 43°54′2.31″ O), P4 (3°45′54.44″ S, 43°49′36.12″ O), and P6 (4°2′49.24″ S, 43°35′5.11″) collected along the Iguará river. The sites were sampled four times during 2020 in the rainy (January and April) and dry (September and November) seasons (Figure 2).

The samples were taken in 500 mL polyethylene vessels, rinsed with sample water, and stored in a thermal box for transport to the Laboratory of Environmental Sciences (LACAM). The samples were refrigerated (4 °C) for subsequent filtration in 0.45 mL cellulose acetate membranes (Millipore) and analyzed.

2.2. Monitored Parameters

Thirteen physicochemical parameters (pH, EC, turbidity, salinity, TDS, NO₂⁻, NO₃⁻, P, hardness (magnesium and calcium), DO, BOD, and temperature) were selected. The samples were analyzed according to APHA (American Public Health Association) (APHA, 2012) [23] and National Water Agency [19]. The nutrients P, NO₂⁻, NO₃⁻, Ca and Mg were determined via Uv-Vis spectrophotometry. Phosphorus was determined via reaction with ammonium molybdate and ascorbic acid as a reductant, measuring the absorbance at 883 nm with a spectrophotometer. Nitrite and nitrate were determined via reaction with sulfanilamide at a wavelength of 540 nm. Calcium and magnesium were determined via reaction with EDTA and their respective complex measured at wavelengths of 525 nm and 545 nm, respectively. pH was measured using a digital pH meter (KASVI—model K39-2014B, São José do Pinhais, Brazil), while turbidity, TDS, temperature, salinity, electrical conductivity, and dissolved oxygen were measured in loco using a multiparameter probe (Horiba—model U52G, Kyoto, Japan). BOD was determined using the respirometric method. The tests for BOD5 started on the same day of collection and were monitored daily for 5 days. At this stage, 5 Winkler flasks were incubated for each point evaluated, in order to monitor the daily consumption of oxygen, making it possible to calculate the decay coefficient k1. NaOH tablets were placed in each flask to calculate the pressure difference.

Of these parameters, only pH, turbidity, total waste (estimated from total dissolved solids), phosphorus, dissolved oxygen, BOD, temperature, thermotolerant coliforms, and total nitrogen (estimated from nitrite and nitrate content) were used to calculate the WQI. According to CONAMA 357/05 legislation, maximum values allowed for each parameter: pH 6.0–9.0; electrical conductivity < 100 µS·cm⁻¹; TDS < 500 mg·L⁻¹; turbidity < 100 NTU; salinity < 0.5 ppt; TDS < 500 mg·L⁻¹; nitrite < 1 mg·L⁻¹ and nitrate < 10 mg·L⁻¹; phosphorus < 1 mg·L⁻¹; calcium and magnesium hardness 40–170 mg·L⁻¹; dissolved oxygen < 5 mg·L⁻¹; BOD 2–5 mg·L⁻¹; and temperature 20–33 °C.

The microbiological assays were realized by COLItest^® kit. The water samples were taken in sterile flasks. The samples were subsequently placed in an incubator at 37 °C for 24 h. The presence of E. coli was confirmed by seeding on Eosin Methylene Blue (EMB) agar medium [17,24]. The results were compared with CONAMA Resolution 357/05 and with the Ministry of Health Ordinance 518/04, as the riverside population uses the water for consumption [25].

2.3. Multivariate Statistical Method: PCA

Multivariate statistical analysis was realized using classical statistics. This comprehensive analysis method can analyze multiple objects and indices under the condition that they are interrelated. Multivariate statistical techniques have been widely used to analyze water quality parameters [26,27,28,29,30,31]. These tools help to simplify and organize large datasets to explain the observed relationships among several variables [32]. In this study, analysis of variance (ANOVA), and Tukey test (p < 0.05) and Fisher LSD test were used to analyze the results in Origin Pro 8.0 version 80724-B724 software (OriginLab Corporation, Northhamton, MA, USA). PCA was applied to the experimental data to identify the differences between the parameters within the seasons under study. The results were analyzed using Minitab 17 version 17.3.1 software (State College, PA, USA. For the physicochemical analyses, the results are expressed as the mean ± standard deviation (SD). The physicochemical parameters evaluated were temperature, electrical conductivity, TDS, turbidity, salinity, nitrite, nitrate, DO, BOD, total phosphorus, magnesium, and calcium hardness, and E. coli bacteria via microbiological analysis in the seasons studied.

2.4. Water Quality Index (WQI)

The WQI is the simplest and most widely used index for assessing the overall quality of water and groundwater [33,34,35,36]. In our study, the WQI was calculated as

$$\mathrm{WQI}={{\displaystyle \prod }}_{\mathrm{i}=1}^9{\mathrm{Q}}_{\mathrm{i}}^{\mathrm{w}}$$

where Q_i is the quality value of the i-th parameter, a number between 0 and 100, obtained from the respective average quality variation curve as a function of its concentration or measurement, and wi is the weight corresponding to the i-th parameter set according to its importance for the overall conformation of the quality, that is, a number between 0 and 1. Each of the parameters that make up the WQI has a certain weight relative to the measure of its contribution to water quality. The values used were those presented in the ANA [18]. The quality of the water is a function of the IQA value obtained, which can be very poor (WQI < 25), poor (26 < WQI < 50), regular (51 < WQI < 70), good (71 < WQI < 90), or excellent (91 < WQI ≤ 100) (ANA, 2015). Munim River was classified according to the WQI (

Table 1. Classification of the Water Quality Index (WQI).

WQI Range	Type	Classification
<100	Excellent	I
51–79	Good	II
36–51	Regular	III
19–36	Poor	IV
<19	Very poor	V

). Parameters used for WQI calculation and their weights are presented in

Table 2. Parameters used for WQI calculation and their weights.

Parameters	Weight
pH	0.12
Turbidity	0.08
Total waste *	0.08
Phosphorus	0.10
D.O	0.17
BOD	0.10
Thermotolerant coliforms	0.15
Total nitrogen **	0.10
Temperature	0.10

* Estimated from total dissolved solids content. ** Estimated from nitrite + nitrate content. DO (dissolved oxygen), BOD (biochemical oxygen demand).

.

3. Results

3.1. Descriptive Measures of River Water Quality Data

To compare the significant differences of the mean values at p < 0.05, ANOVA, and Tukey’s test and Fischer LSD’s multiple range test were employed (

Table 3. Physicochemical analysis at the six sampling points.

Parameters	Rainy Season		Dry Season
	January/20	April/20	September20	November/20
pH	6.01 ± 0.21 a	5.96 ± 0.62 a	5.73 ± 0.10 a	5.74 ± 0.15 a
E.C (µS·cm−1)	119.00 ± 27.43 a	70.98 ± 21.19 b	142.22 ± 38.00 a	148.80 ± 22.82 a
Turb. (NTU)	36.51 ± 15.36 a	16.16 ± 2.22 b	15.51 ± 2.97 b	15.42 ± 3.88 b
Sal. (ppt)	0.02 ± 0.008 a	0.01 ± 0.005 a	0.01 ± 0.005 a	0.01 ± 0.005 a
TDS (mg·L−1)	48.53 ± 21.38 a	45.50 ± 0.14.98 a	152.15 ± 24.98 b	138.78 ± 32.44 b
NO2− (mg·L−1)	0.73 ± 0.09 a	0.52 ± 0.11 a	0.55 ± 0.35 a	0.75 ± 0.23 a
NO3− (mg·L−1)	4.93 ± 2.96 a	4.50 ± 1.98 a	5.88 ± 1.15 ab	8.33 ± 1.18 b
P (mg·L−1)	1.26 ± 0.07 a	2.86 ± 1.09 b	3.65 ± 0.85 bc	4.85 ± 0.55 c
Mg (mg·L−1)	6.24 ± 1.71 a	11.16 ± 2.04 abc	14.08 ± 5.58 bc	15.35 ± 6.41 c
Ca (mg·L−1)	374.85 ± 39.39 a	313.00 ± 13.85 b	321.00 ± 11.90 b	344.33 ± 11.10 ab
D.O (mg·L−1)	1.64 ± 0.37 a	10.88 ± 0.35 b	10.01 ± 0.40 c	9.77 ± 0.08 c
BOD (mg·L−1)	6.88 ± 1.19 a	7.91 ± 1.96 a	8.56 ± 1.07 a	8.45 ± 1.49 a
Temp. (°C)	33.74 ± 0.75 a	19.82 ± 2.09 b	24.12 ± 0.52 c	22.86 ± 0.45 c
E. coli (CFU/100 mL)	802.5 ± 58.81 a	408.3 ± 31.25 b	458.3 ± 96.04 b	464.1 ± 80.45 b

Mean values (n = 6) are presented. Means followed by the same letter in the same line do not differ statistically from each other according to Tukey’s test (p < 0.05). Electrical conductivity (EC), Turb. (turbidity), Sal. (salinity), total dissolved solids (TDS), NO₂⁻ (nitrite), NO₃⁻ (nitrate), total phosphorus (P), magnesium (Mg), calcium (Ca), dissolved oxygen (DO), biochemical oxygen demand (BOD), and Temp. °C (temperature), respectively.

), using Origin Pro 8.0 version 80724-B724 software (OriginLab Corporation, Northhamton, MA, USA). PCA was applied to the experimental data to visualize the differences among the samples, and the results were analyzed using the Minitab 17 statistical software program for Windows (version 17.3.1 (State College, PA, USA). The data were evaluated using ANOVA, according to the physicochemical parameters (temperature, electrical conductivity, TDS, turbidity, salinity, nitrite, and nitrate) and microbiological parameters during studied periods, where the means differed at each point (P1, P2, P3, P4, P5, and P6); different letters (a, b, and c) were used to ensure that at the points where the means have equal letters, no statistical difference was found between the data.

A comparison of the averages of all parameters was realized using Tukey’s test (p < 0.05). For the set of analyses (physical, chemical, and microbiological), PCA was applied to the mean values of the replicates (n = 6) to identify possible correlations among the data and to group them according to seasonal influence.

PCA was used to investigate the possible correlations between the studied variables and to evaluate hypothetical models for the rating of the sampled points. Initially, an assessment of the relationships between the nine variables that were correlated to the two studied periods was performed using PCA, based on a correlation data matrix, where the entire dataset was auto-scaled for all variables.

3.2. Microbiological Assay

Figure 3 presents box plots of the individual water quality parameters of E. coli that illustrate the temporal variations related to the two seasons. The plots were generated by combining the data of six determinations corresponding to each season. The median, lowest, and highest values for a given period were determined by analyzing the data for specific periods. The line across the box indicates the median concentration. The vertical lines extending from the bottom and top of the box correspond to the lowest and highest observations, respectively.

3.3. WQI

The indices were used to classify six sampling points from January to November 2020 (

Table 4. WQI and classification of the sampling points.

Months	Sampling Points	WQI 1	Classification
January-2020	P1	55.18	II
	P2	60.80	II
	P3	61.50	II
	P4	61.50	II
	P5	55.20	II
	P6	60.70	II
April-2020	P1	39.08	III
	P2	44.11	III
	P3	46.02	III
	P4	46.02	III
	P5	43.05	III
	P6	46.23	III
September-2020	P1	63.20	II
	P2	63.50	II
	P3	63.20	II
	P4	63.11	II
	P5	53.10	II
	P6	63.08	II
November-2020	P1	55.20	II
	P2	62.40	II
	P3	60.30	II
	P4	62.50	II
	P5	62.05	II
	P6	61.05	II

¹ Water quality index.

). The points sampled in the month of January presented Class II, with an average of 59.14; April, Class III, average = 44.08; September, Class II, average = 61.53; November, Class II, average = 60.58. Water quality in the study area can be classified as good and regular.

3.4. PCA

PCA is a mathematical approach for dimensionality reduction. Using PCA, the original 13 indicators are recombined into several groups of new comprehensive indicators that are unrelated to each other to replace the original indicators. The information contained in each group of indicators is expressed by variance; that is, the higher the variance, the higher the information contained. Each set of indicators is called a principal component. Principal component 1 (PC1) contains most of the information; thereafter, the amount of information contained decreases. In the process of extracting the principal components, we selected those whose initial eigenvalues were greater than one.

The changes in the concentrations of physicochemical parameters presented a correlation with each station evaluated. Such a finding indicates that the high correlation between the parameters associated with a given period has the same sources of pollution, and may have the same trends in changes. Thus, a variation in the concentration of one index may indicate changes in other highly correlated pollutants [37]; this can be examined using PCA. The total variance of the four principal components is shown in

Table 5. Loadings of the variables for the first four principal components.

Variables	PC1	PC2	PC3	PC4
pH	0.211694	−0.339744	0.281397	−0.234741
E.C	−0.089208	0.491282	−0.250585	−0.399001
Turb.	0.308774	0.151709	0.176607	0.265387
Sal.	0.221851	−0.182013	−0.120232	−0.608899
TDS	−0.247860	0.387794	0.130499	−0.065474
NO2−	0.113668	0.237902	0.440053	0.225207
NO3−	−0.099400	0.246317	0.552196	−0.133110
P	−0.355449	0.199129	0.055225	−0.054660
Mg	−0.252059	−0.026302	0.211142	−0.462613
Ca	0.266181	0.332497	0.086386	−0.077411
DO	−0.379364	−0.190284	0.116333	0.037602
BOD	−0.245595	0.195371	−0.429805	0.165634
Temp	0.363751	0.231854	−0.034696	−0.140967
E.coli	0.348072	0.200962	−0.205469	0.0085396
Eigenvalue	5.9228	2.3698	1.8236	1.2480
T.V (%) *	42.30	16.90	13.00	8.90
C.V (%) **	42.30	59.20	72.30	81.20

* Total variance. ** Cumulative variance. Significant correlation (>0.3). Electrical conductivity (EC), Turb. (turbidity), Sal. (salinity), total dissolved solids (TDS), NO₂⁻ (nitrite), NO₃⁻ (nitrate), total phosphorus (P), magnesium (Mg), calcium (Ca), dissolved oxygen (DO), biochemical oxygen demand (BOD), and Temp. °C (temperature), respectively.

.

The variance contribution rate of PC1 was 42.30%, that of principal component 2 (PC2) was 16.90%, and the cumulative variance contribution rate of the first four principal components was 81.20%. According to the statistical correlation coefficients, we can classify them into ‘strong’ (>0.75), ‘moderate’ (0.75–0.50), or ‘weak’ (0.50–0.30) for the absolute values. PC1 showed a weak correlation (|0.37|) for DO and temperature. In PC2, EC had a weak correlation (0.49), along with TDS. In PCs 3 and 4, a moderate correlation of the parameters, nitrate (NO₃⁻) and salinity, respectively, indicating a possible relationship between nitrate from local agriculture and the flow of rainwater through nitrogen fertilizer-rich soils. Among these, nitrate ion, which is one of the limiting nutrients of aquatic life, indicates the possibility of eutrophication of the water and salinity in the concentration of dissolved ions, which directly contributes to the parameters of electrical conductivity and dissolved solids. Figure 4 presents loading and score plots of all parameters and season studied.

In Figure 4, the axes were strongly correlated with the variables. Based on the distribution of points along PC1, three clusters are highlighted: one related to the points analyzed in the month of January (rainy season); one related to the month of April (rainy season), where we can infer a state of transition between the seasons by the distribution of points along the negative and positive axes of PC1; one cluster between the months of September and November (dry season).

4. Discussion

According to Table 1, for pH, all measurements were statistically similar throughout the year, presenting an annual average of 5.86, below the values found by George and Ngole-Jeme [38], 6.77 in their research on the WQI for community use. The electrical conductivity presented different results in April (70.98 µS·cm⁻¹). The highest turbidity value (36.51 NTU) was found in the rainy season (January); for the subsequent months, there was a decrease in turbidity (average of 15.69 NTU). Salinity was constant throughout the year, with no statistical difference found between the measurements. TDS showed an increase from the rainy period (average of 47.01 mg·L⁻¹) to the dry period (average of 145.46 mg·L⁻¹), with data similar to the minimum and maximum values found by Zhang et al. [39], 7 mg·L⁻¹ (minimum) and 239 mg·L⁻¹ (maximum).

In terms of nitrite and nitrate, the concentration did not vary throughout the year; however, an increase in nitrate concentration was observed from the rainy season (average of 4.71 mg·L⁻¹) to the dry season (8.33 mg·L⁻¹). Phosphorus content increased during the year, from 1.26 mg·L⁻¹ in January to 4.85 mg·L⁻¹ in November.

An increase in magnesium concentration was observed from the rainy period (6.24 mg·L⁻¹) to the dry period (average of 13.53 mg·L⁻¹); these values were below the acceptable limits, based on the legislation. Calcium concentration decreased in the months of April and September, with an average of 317 mg·L⁻¹, and its highest concentration was observed in the rainy season (374.85 mg·L⁻¹), where all values were above the limits recommended by the legislation (>170 mg·L⁻¹).

The highest concentration of dissolved oxygen was found in April (10.88 mg·L⁻¹), and in the dry period, there was a decrease in oxygen content (average of 9.89 mg·L⁻¹). The measurements of BOD did not show any statistical difference throughout the year, with an average content of 7.96 mg·L⁻¹. The data obtained for DO and BOD, above 3 to 5 mg·L⁻¹ (CONAMA, 357/05), can be correlated with the action of discharges, containing substances with low biodegradability that are not normally found in domestic sewage. The temperature decreased in April (19.82 °C) and increased in the dry period, with an average of 23.49 °C. Hernández-Mena et al. [40] obtained values of 5.23 mg·L⁻¹ during the dry season and 4.44 mg·L⁻¹ during the rainy season. As for the BOD, they obtained averages of 3.70 mg·L⁻¹ in the dry period and 11.52 mg·L⁻¹ in the rainy period.

In general, the nutrients, nitrate, total phosphorus, dissolved oxygen, and biochemical oxygen demand exceeded the expected margins according to CONAMA Resolution 357/2005 in the seasonal periods. According to Silva et al. [24], predominantly acidic pH data were obtained during the dry period. From data found in our study, at all points, the pH of the water was approximately neutral, showing no significant variation (neither acidic nor alkaline). The pH value influences the distribution of free and ionized forms of several chemical compounds. The pH values can be explained by the influence of garbage and the deforestation of riparian forests, which has been occurring in the region. Deforestation causes a strong silting up, which unprotects the area along the margins of the Munim and Iguará rivers, at points P1 to P4. The results were compared according to the CONAMA Resolution 357/2005, as the river was classified as class 3.

According to Souza et al. [19], this phenomenon occurs because electrical conductivity is inversely proportional to the value of the rainfall index. Phosphorus is the main limiting factor of productivity in water bodies and has been highlighted as the main factor responsible for the artificial eutrophication of these ecosystems; that is, there is a greater production of organic matter than its consumption and decomposition. Phosphorus can originate from natural sources (present in the composition of rocks, carried by surface runoff of rainwater, particulate material present in the atmosphere and resulting from the decomposition of organisms of allochthonous origin) and artificial sources, such as domestic sewage, removal of sand from the riverbed, and deforestation of riparian forests, thereby having a very large impact on aquatic biota.

For the microbiological tests, shown in Figure 3, the presence of E. coli was high in the first month (January, rainy period) and remained at lower levels until November (dry period). In January, there was a significant level of E. coli (802.5 CFU/100 mL), but in April, which was characterized by heavy rains, there was a lower growth of the bacteria (408.3 CFU/100 mL), despite being within the parameters allowed according to CONAMA Resolution 357/05. With the data of bacteria incidence and the association with the significant interference of cattle-raising activity and interference of domestic sewage on water quality, it is possible to infer that the waters at the merging of the Munim and Iguará rivers are inappropriate for use. Our results were similar to Choque-Quispe et al. [41], in their research on the Chumbao river, where they found values of 710.0 in the rainy period and 290.0 in the dry period.

According to previous studies [42,43,44,45], bacteria of the coliform group, which indicate fecal pollution, are employed to evaluate the sanitary conditions of water. Cavalcanti [6] recorded an increase in the number of bacteria in the dry period, from 86.5 to 389 CFU/100 mL and 95.5 to 439 CFU/100 mL, respectively.

By evaluating the quality of surface water of the Munim River, by means of the modified WQI, Silva et al. [24] obtained the classification of regular and good, from 2014 to 2017. Compared to the current study, the Munim Basin maintains the same classification. At the confluence of the Munim and Iguará rivers, the water quality in the region is determined by natural processes (precipitation intensity, weathering, vegetation cover) and anthropogenic influences (agriculture, urban concentration, gravel and sand removal activity, animal breeding, bathing, clothes washing, and leisure); that is, activities with intense use of water. Notably, changes in the aquatic system lead to economic losses in the region, ranging from reduced fishing catches to increased costs of water acquisition and treatment.

According to Souza et al. [19], the determination of both WQI and coliform bacteria is relevant for water classification, and their use in combination is a good basis for decision making. Although not widespread, it is evident that measuring the flow rate and determining the pollutant load are of fundamental importance for water quality studies.

5. Conclusions

A more in-depth discussion about the impacts of pollutants in water bodies, the relation mass of pollutants, and the biodiversity exposed to this quantity of chemical substances is necessary. WQI is a method used to assess the possible deterioration of water resources in spatiotemporal way.

Actions for the collection and treatment of domestic effluents, supervision of irregular disposal and awareness campaigns, concentration of pollutants, and the preservation and recovery of riparian forests, are presented as measures of the remarkable potential for improving the quality of the water upstream and downstream of the Munim and Iguará rivers.

The results obtained for the physical and chemical parameters of phosphorus, magnesium, calcium, dissolved oxygen, and biochemical oxygen demand remained above the maximum limits allowed by CONAMA nº 357/05. In fact, only the levels of nitrite and nitrate remained within the established standards and were, thus, classified as suitable for aquatic life. Principal component analysis showed that the physicochemical parameters, dissolved oxygen, electrical conductivity, nitrate, and salinity were the most sensitive to seasonality. As for E. coli, it was possible to correlate the highest incidence of this contaminant to the rainy season, indicating a high level of contamination of the river by fecal coliforms.

Altogether, the combination of physicochemical and biological analyses with statistical analysis provides a consistent foundation for formulating water resource management strategies. The results of this research may be relevant for future research on water quality and urban planning in the region, which directly affects the environment. From the data in this research, it is clear that there is a need for public policies that enable the conservation of the water resources of the region, since the influence of human activity on the water quality of the Iguará and Munim rivers is clear.