Introduction

Water is a crucial component of life, so the increasing deterioration of the water resources is a severe risk to health and people’s well-being, as well as a serious environmental problem. In Brazilian urban areas, these problems are mostly related to the lack of planning on the expansion of cities and its precarious infrastructure for basic sanitation (Tundisi 2008). For instance, a survey of the Ministry of Cities from Brazil revealed that 42% of Brazilian households are still not served by sewerage system (BRAZIL 2016). The same survey also shows that only 41% of the sewage generated receives treatment before discharge on water bodies, which is very detrimental in terms of environmental healthiness.

The disposal of untreated sewage in watercourses generally entails in high loads of nutrients (nitrogen-N, and phosphorous-P) and biodegradable material (BM). The excess of nutrients promotes the eutrophication of the receiving water body, while the high load BM is associated with the depletion of dissolved oxygen (DO), which is rapidly consumed in the metabolic processes of stabilization of organic matter (Wetzel 2001; Von Sperling 2007). Therefore, the DO is a sensitive indicator of water quality and is a key parameter to characterize the effects of pollution by biodegradable wastes (Lewis 2006) on self-depurative studies (Yu et al. 2010; Bayram et al. 2015; Nemati et al. 2015).

The self-depurative capacity of a watercourse is generally evaluated using the Streeter and Phelps (1925) model. In this simulation, the hydraulic and physical–chemical characteristics of the river section that receives the effluent comprise the input variables of the model, together with the constant of deoxygenation (K 1), which reflects the rate of DO consumption during BM stabilization.

The risks associated with the presence of heavy metals in water are another environmental concern, because they may lead to accumulation of high levels in fish and its consumers (Protano et al. 2014; Lü et al. 2011).

The watercourses of São Mateus River (SMR) watershed receive a significant amount of pollutants from their headwaters to their mouth, mostly in the form of domestic wastewater released by more than 430,000 residents of the 25 cities distributed over their drainage area (IBGE 2015). The city of São Mateus-ES has a population of 124,575 inhabitants (IBGE 2015) and a high human development index (PNUD 2013). Despite that, the city still has no sewage treatment plant, so the domestic wastewater and other effluents are directly launched in the SMR and in its tributary, the River Abissínia (RA). The situation is worrying because the SMR waters are collected for public supply.

In this work, water samples from SMR and two of its tributaries (Abissínia and Preto do Sul) were collected and analyzed in order to identify and quantify the impacts caused by the direct disposal of municipal effluents and to assess whether the river self-depurative capacity is respected or exceeded. A respirometric test is proposed to determine the constant of deoxygenation (K 1), which was used to evaluate the depurative capacity of the river. The results were compared with those obtained from the classic BOD test. Moreover, the presence of heavy metals in these waters and its lanthanides patterns were studied.

Materials and methods

Study area

The SMR watershed has approximately 13,500 km2. There are two main rivers: at north Cotaxé, which is 244 km long, and at south Cricaré with 188 km of extension. The confluence of these rivers takes place in the lower course of the basin, in the municipal area of São Mateus City. The drainages of the SMR watershed run through different lithologies, mostly on Araçuaí orogen of granitic composition, and the junction of the arms takes on Abrolhos formation, comprised by ignimbritic rocks of rhyolitic composition with varying degrees of weathering (CPRM 2015).

According to the classification of Köppen, the climate of the area of study is classified as tropical hot and humid, with rainy summers from November to February, concentrating 70% of the annual precipitation, and dry winters, from May to September. The main economic activities are agriculture, granite extraction, alcohol and wood production. The geomorphology of the watershed is of recognized risk due to the occurrence of precipitations with high erosive impact, causing great loss of soil and siltation of watercourses (Mello et al. 2012). Near the city of São Mateus, the river course has low declivity (≈0.1 m/km) and therefore presents a slow flow speed, strongly influenced by the tide regime.

The sub-basins of the rivers Abissínia and Preto do Sul are about 101.6 and 291.5 km2, both formed by kaolinitic sediments of the Barreiras Group (CPRM 2015).

The Abissínia sub-basin drains the urban area of the municipality and has its use granted for dilution of the municipal wastewater. Therefore, most of the sewage generated within the municipal boundaries (≈13,500 m3/day, according to the municipal Water and Wastewater Company) is directly discharged into the Abissínia River without any treatment, which causes serious impacts, as will be presented in this work.

The Preto do Sul sub-basin, on the other hand, is of agricultural use, but the lack of environmental protection resulted in the removal of its riparian vegetation and the construction of several dams to attend the demand for irrigation.

The study was conducted by monitoring a set of parameters (pH, conductivity, temperature, DO, BOD, total suspended solids, the fractions of soluble and total phosphorus (Psol and Ptot, respectively)) at four locations (Fig. 1), one at Abissínia River (2: 18°43′41.73″S and 39°49′08.08″W), another at Preto do Sul River (3: 18°44′8.94″S and 39°47′47.31″W) and two in SMR, one located approximately 2 km upstream of the city (1: 18°41′41.55″S and 39°52′5.06″W) and other downstream, after the discharge of the municipal effluents (4: 18°43′44.52″S and 39°46′31.21″W), close to the access road to the district of Guriri.

Fig. 1
figure 1

Location of the study area and sampling points (1–4)

Full size image

Estimative of the minimum reference flow for the SMR and effluent discharge

The minimum reference flow is the basis for the management process of a water body. This can be represented by Q min,7.10, which corresponds to the average of the smallest flow rates maintained in a watercourse for 7 consecutive days and with a return period of 10 years. For the segment of the SMR under study, a drainage area (A) of 12,700 km2 was considered, with an average precipitation (P), of 1100 mm/year (INCAPER 2011). The Q min,7.10 was estimated for a period (D), of 7 days, using the flow rate regionalization curve (Q min,7.10 = 0.0008 × A 1.0659 × D 0.1659 × P −0.1109), established by CPRM (2002). The contribution of the Preto do Sul River was disregarded because it represents less than 2% of the total flow.

In the sampling point of Abissínia River, the water flow rate was determined three times in one-channeled section, measuring its width and depth and the water speed (float method), to estimate the total discharge over the SMR.

Sample collection and monitored parameters

Water samples from the SMR (upstream and downstream of the city) and Abissínia River were collected and analyzed, quarterly, in order to identify and quantify the possible impacts caused by the release of untreated sewage on these waters between 2012 and 2015. Sampling occurred in the morning period, respecting a period of 3 days without rainfall. The study of the Preto do Sul River aimed its hydrogeochemical characterization, since it does not receive urban effluents.

For the determination of BOD, Psol, Ptot, total solids, major cations, anions and trace elements, water samples were taken with flasks of 1 L, which were submerged to a depth of 20–30 cm, rinsed with the sample and then completely filled. All bottles were previously acid (HNO3 5%) cleaned and washed with distillated water. Samples were collected at the morning and immediately transported to the laboratory in a box with ice.

The measurement of physicochemical parameters was carried out in situ, with specific sensors. For DO and temperature, the HANNA meter HI 9146 oximeter and the por meter HANNA HI 98130 were used for pH and conductivity. Vertical DO profiles (at 0.25, 1, 2, 3, 4 and 5 m of depth) were recorded at different locations to validate the self-depuration model. The same oximeter was used in the determinations of BOD5,20, with incubation without seed for 5 days at 20° C (APHA method 5210 B 1999). The total solids content was obtained by gravimetry (EPA method 1684 2001), and phosphorus content was determined by the spectrophotometric method with quantification at 880 nm (APHA method 4500-P 1999).

Major cations and anions were determined by ion chromatography (EPA method 300.0 1993) for water classification purposes, and the presence of trace elements (Ag, Al, As, B, Ba, Be, Bi, Cd, Co, Cr, Cs, Cu, Fe, Ga, Ge, Hf, Li, Mn, Mo, Nb, Ni, Pb, Rb, Sb, Sc, Se, Sn, Sr, Ta, Th, Ti, Tl, U, V, W, Y, Zn, Zr and lanthanides) was determined by ICP-MS (Cotta and Enzweiler 2009) on samples previously filtered (0.45 µm membrane) and acidified to 1% with sub-boiling purified HNO3. Samples from Abissínia River were not analyzed for trace elements due to the high load of wastewater.

The quality control was performed through the analysis of the certified reference material (SLRS-5: river water reference material for trace metals). The obtained results for the tested water samples were compared against the reference values of the Brazilian legislation (CONAMA 357 2005) and examined for anomalous concentrations.

Assessment of the self-depuration capacity of the São Mateus River

In order to evaluate the depurative capacity, a simulation with the program AD’ÁGUA 2.0 (Braga and Santos 2010) was carried out considering the minimum water flow of reference, the hydraulic characteristics of the SMR and the effluent characteristics.

The proposed model was also used to estimate the required efficiency for the treatment of the effluents, not to exceed the river’s depurative capacity, which is essential information for the design of a sewage treatment plant.

The respirometric assay

A system composed of an air pump connected to the lateral orifice of a 1-L kitasate (reactor) through a flexible hose, with a porous stone at the end to improve sample aeration, was set up to perform respirometric measurements (Spanjers et al. 1998).

One dedicated plastic lid (leaked cap) was made to close the main opening of the reactor during measurements. During the daytime, the consumption of DO was monitored with the oximeter fixed through a hole in the leaked cap. This connection was tightly sealed using rubber rings. During the night, no measurements were taken and the reactor was maintained open. With this procedure, the calibration of the sensor could be carried out each day, and possible deviations associated with the permanence of the electrode in the sample solution for long periods were avoided.

For each assay, nutrients and phosphate buffer solution were added to the sample, in a similar fashion to the BOD determination (APHA 5210 B 1999), before filling the reactor flask with the sample collected in Abissínia River, which represents the urban effluents discharged on the river. During the measurements, the reactor was maintained at 20 °C and out of the light for 3 days. This period corresponds to the time that the water in SMR takes to travel the 40 km stretch from the city to its mouth.

The respirometric assay starts by aeration of the sample up to 100% of DO saturation. After 5 min of rest, DO concentrations were recorded at each 2–3 h intervals under magnetic stirring. The aeration and resting time were repeated always when DO concentration dropped to approximated 3 mg/L. These data sets were used to calculate the deoxygenation velocity constant, K 1 (Zanoni 1967). Concomitantly, standard BOD tests (APHA 5210 B 1999) were performed with daily monitoring of the DO consumption to compare with the results of the proposed respirometric assay.

Results and discussion

Reference flow for the SMR and the impacts of the urban effluents

The minimum water flow of reference (Q min 7.10) calculated to the SMR was of 12.0 m3/s. This value is similar to the permanence flow rates (Q 90 = 16.4 m3/s and Q 95 = 11.7 m3/s) estimated by Uliana et al. (2011) and corresponds to a specific yield of about 1.0 L/s km2.

The water flow rate measured on the sampling point of the Abissínia River ranged between 0.5 and 0.7 m3/s, considering an interval of 3 days without rain. If just the regionalization curve for this sub-basin was considered, the expected water flow would be around 0.2 m3/s. Therefore, more than 50% of the recorded flows were due to wastewater, which makes the Abissínia River, an open sewage course. Table 1 summarizes the range of values for some parameters registered in the scope of this investigation. These results are compared with the maximum permissible values established by the Brazilian Environmental Council (CONAMA357 2005). The SMR and its tributaries have not yet been frameworked and therefore are evaluated as belonging to class 2.

Table 1 Comparison between the data obtained and permissible limits from CONAMA 357
Full size table

The results confirm that Abissínia River is severely impacted by the discharge of the urban effluents. The verified DO values were always close to 0 mg/L, revealing that this watercourse is completely anoxic. It also had an elevated BOD5.20, ranging from 45 to 120 mg/L (median ≈100 mg/L), which corresponds to a load of 4.3 T of BOD/day. On these waters, the Psol values ranged from 3.6 to 4.5 mg/L, exceeding fairly the levels recorded in the SMR.

The sampling site of Abissínia River presents strong odors, fairly turbid waters, with elevated values of total solids (that surpass 500 mg/L), conductivity around 550 μS/cm and neutral pH. Photographs that illustrate the degradation of water quality on the Abissínia River are displayed in Fig 1S in supplementary material (Online Resource 1).

At the sampling point 1, upstream of the city, the SMR exhibited DO levels ranging from 5.5 to 7.2 mg/L, which corresponds to more than 80% of saturation, BOD5.20 from 1 to 2.6 mg/L and low levels of Psol between 0.01 and 0.09 mg/L.

At point 4, after receiving the municipal effluents, a deterioration on water quality was verified, with less DO availability, saturation between 50 and 80%, higher BOD5,20 and Psol values (median of 0.06 mg/L).

Samples from point 4, downstream of the city of São Mateus, also presented higher concentrations of Ptot (median of 0.12 mg/L), but with lower contribution of the soluble fraction, contrasting with point 1, upstream, which had higher proportions of soluble fraction. This difference indicates that the waters of these sampling sites are receiving P from different sources. The higher proportion of Psol recorded upstream may be associated with the runoff from agricultural areas (Yu et al. 2010), while downstream the Ptot is elevated by the organic species of phosphorus present on the wastewater. It should be noted that the average and maximum values of Ptot downstream are several times higher than that registered upstream and well above the limit proposed by CONAMA Resolution 357.

The obtained results of Ptot are similar to values registered in rivers such as Preto and Grande (0.01–1.5 mg/L), Guamá (0.06–0.65 mg/L), Pirapó (0.05–1.09 mg/L), Faria (0.19–2.0 mg/L) and Cunha Canal (0.09–5 mg/L), all of which receive the majority of the domestic sewage of the municipalities of São José do Rio Preto-SP and the Metropolitan regions of RBelém-PA, Maringá-PR, and Rio de Janeiro-RJ, respectively, (Campanha et al. 2010; Lima et al. 2015; Bortoletto et al. 2015; Borges et al.2015).

On the basis of the values recorded for Ptot, the trophic state of the SMR was assessed. It varied from oligotrophic (0.013–0.035 mg/L) to mesotrophic (0.035–0.137 mg/L) for most (≈90%) of the samples taken upstream of the city, while 30% of the samples collected downstream achieved the eutrophic (0.137–0.296 mg/L) and supereutrophic (0.296–0.640 mg/L) levels, according to Carlson`s (1991) classification modified by Lamparelli (2004) for tropical and subtropical climates. Thus, the SMR is susceptible to the eutrophication process with algal boom, and water quality impairment, as a direct consequence of the indiscriminate release of phosphorus-rich effluents.

According to reports of water quality monitoring generated by the Environmental Sanitation Technology Company of the State of São Paulo, Brazil (CETESB), conductivity values above 100 μS/cm are indicative of impacted environments (CETESB 2009). On the headwaters of the SMR, the physicochemical parameters are monitored by IGAM (Water Management Institute of the State Minas Gerais, Brazil); since 2009, and the registered values of conductivity are about 100 μS/cm (IGAM 2015). Thus, as the mean conductivity recorded on the monitored points of the SMR (≈300 ± 50 μS/cm) at the end of the basin is three times higher, it reveals that these waters (even upstream the city) present an integrated result of several diffuse sources of pollution distributed throughout the basin.

The respirometric characterization of the effluents

Regarding the respirometric assays performed, the use of the proposed system allowed to register the DO consumption during the degradation of the BM present in the sample (Fig. 2). Figure 3 shows the BOD variation during the three days of assays with samples collected in Abissínia River on three different occasions.

Fig. 2
figure 2

Dissolved oxygen values (OD) recorded during a test with sample

Full size image
Fig. 3
figure 3

Exerted BOD and calculated K 1 for three different samples, to exemplify the range of obtained values

Full size image

Equation (1) describes the progression of the exerted BOD (y) at time (t), which is related to the initial BOD (L o ) of the sample through the deoxygenation velocity constant (K 1).

$$y = L_{o} \cdot \left( {1 - {\text{e}}^{ - K1.t} } \right)$$
(1)

To calculate the K 1, the recommendations of Von Sperling (2007) were followed, with the aid of the “solver” tool (of the Excel spreadsheet software).

The value of K 1 depends on the characteristics of the sample. For concentrated sewage, it varies from 0.35 to 0.45/day (Von Sperling 2007). The values of K 1 calculated from the data obtained with the constructed prototype are between 0.33 and 0.38/day, being compatible to that expected for untreated wastewater. These K 1 values matched those obtained by the classical BOD test of 0.35–0.40/day, demonstrating the reliability of the proposed assay.

Modeling the self-depuration capacity of the São Mateus River

The self-depuration capacity of SMR was evaluated considering its hydraulic and physicochemical characteristics, as well as those of the effluents transported by the Abissínia River. Table 2 summarizes the input variables of the model and Fig. 4 the results.

Table 2 Input data for the self-depuration model
Full size table
Fig. 4
figure 4

Estimated DO concentration for SMR after the discharge of the municipal effluents (full line), obtained with the self-depuration model. Minimum limit of 5 mg/L, according to CONAMA 357 (straight dotted line). Estimative of DO profile for effluent treatment with 60% efficiency in BOD removal (dashed line). The points (square) represent the mean DO concentrations measured and the bars one standard deviation of the mean

Full size image

As shown in Fig. 4, the SMR is significantly impacted by the release of untreated municipal wastewater rich in BM, since the minimum value of 5 mg/L is disregarded. The comparison among the registered DO values, in the vertical profiles, and the predicted values shows that the model was able to describe the impact of the effluents. The mixing zone among effluents and the river water is heterogeneous, as the name itself suggests, and therefore, the data of DO have great amplitude.

The critical DO value depicted (≈3.4 mg/L) reproduces the mean values recorded for the river during the dry season, which validates the developed model and confirms that the river’s self-depurative capacity is exceeded, threatening the maintenance of aquatic life and confirms the immediate need of preservation of this natural resource.

With the model calibrated, it was possible to estimate that an efficiency of 60% in the removal of the BOD from the effluents is necessary to maintain the DO above 5 mg/L (Fig. 4), and this can be achieved with primary effluent treatment (Chin 2012).

The situation of SMR resembles that experienced in many other Brazilian waterways. For instance, studies on the rivers Lavapés, (Valente et al. 1997), Meio (Sardinha et al. 2008) and Preto and Grande (Campanha et al. 2010), impacted by the discharge of untreated sewage, revealed that in all, the DO is intensely consumed to the point of impairing aquatic life, which confirms the spread of the problem and the need for investments in basic sanitation.

Hydrogeochemical assessment

The water samples were classified by its cationic and anionic composition using the Piper diagram, generated with QualiGraf program (Möbus 2002). Figure 2S in supplementary material (Online Resource 1) shows that all samples have a sodium-chlorinated composition.

In comparison with the values of CONAMA Resolution 357 (2005), trace element concentrations recorded for SMR are below the stipulated limits, except for Fe (Table 1S, Online Resource 2). The main alteration registered among the upstream and downstream of the discharge of effluents was perceived on Ni values, whose increase exceeded one order of magnitude with 0.2 μg/L at upstream and 2.5 μg/L at the downstream point, but even so, below the limit of 25 μg/L (CONAMA 357 2005).

Concentrations of As and Pb in natural waters are generally lower than 1 μg/L (Smedley and Kinniburgh 2002; Yoshinaga 2012). High concentrations of Pb and As (17 and 8 μg/L, respectively) were recorded in the sample taken from the Preto do Sul River, in which Pb exceeds limit foreseen in CONAMA Resolution 357, which is stipulated to 10 μg/L (for As and Pb), being also much higher than that registered for the SMR, of about 2.0 and 0.3 μg/L for Pb and As, respectively.

The high values of Pb and As in the Preto do Sul River may have geogenic origin or to be associated with the use of herbicides and insecticides in the agricultural regions, or the wood treatment activity, which involves the manipulation of formulations with these elements. Such activities are developed in the drainage area of the Preto do Sul River. Lü et al. (2011) and Zhang et al. (2015) pointed out that high concentrations of toxic elements in the water may cause the contamination of fish and molluscs and that these elements can reach humans through biomagnification.

The value of Al in the Preto do Sul River (184 μg/L) also exceeds the limit of CONAMA Resolution 357, which is of 100 μg/L, and is up to four times higher than the values registered for SMR (44 and 67 μg/L, upstream and downstream). The higher concentration of Al in waters of the Preto do Sul River is probably due to the lower pH (≈6.0) of these waters, compared to the neutral waters (pH ≈ 7) of SMR, and to the fact that they flow through the sediments of the Barreiras formation that are rich in kaolinite of low crystallinity (Melo et al. 2002).

The waters of the Preto do Sul River are distinguished because they are naturally rich in autochthonous organic material, originated from its riparian vegetation. This material, beyond imparting a very dark coloration to the water, can act in the complexation of metals such as Fe2+, which has great affinity for humic substances (Botero et al. 2014). This explains the higher concentration of Fe in Preto do Sul River waters (≈5 mg/L) compared to the SMR (≈0.35 mg/L).

The pattern of lanthanides, after MUQ normalization, following Kamber et al. (2005), is presented in Fig. 5. The water samples present a pattern enriched in light lanthanides that resemble that registered for samples of rhyolitic ignimbrites of the Abrolhos formation (CPRM 2015). This fact suggests that the pattern of lanthanide observed in the waters was acquired from the lithologies in which they flow.

Fig. 5
figure 5

MUQ-normalized pattern of lanthanide for the SMR samples, upstream (upward triangle) and downstream (downward triangle) the city, and Preto do Sul River (ο). The Abissínia River was not analyzed due to the high load of wastewater

Full size image

A notable negative anomaly of Eu reveals a contrast between the waters of SMR and Preto do Sul River. Negative Eu anomalies are characteristic of waters flowing through granitic rocks or in sediments formed by their weathering, which are widespread in the SMR basin (CPRM 2015). The absence of Eu anomaly in the waters of the SMR may result from the fact that these waters interact with different types of lithologies and substrates, dispersed on a large drainage area, which causes them to assume a medium lanthanide pattern, without distinctive characteristics. In contrast, the Preto do Sul River drains a smaller and more homogeneous area, and so the characteristics of its lanthanide pattern are preserved. Another parameter that indicates that there is significant contrast among the substrates of these drainages is the Zr contents, once the value recorded in the Preto do Sul River (162 μg/L) is 18 times higher than the value of the SMR (≈9 μg/L).

Pourret et al. (2007) argue that organic compounds, such as humic and fulvic acids, significantly act in the complexation of lanthanides and therefore contribute to defining the pattern developed. The complexation effect is more expressive while higher the organic content, as in the case of the Preto do Sul waters.

Conclusions

The obtained data for rivers Abissínia and Preto do Sul, whose small outflow makes them susceptible to respond immediately to negative anthropogenic changes in their watersheds, reveal that for the first the launching of an excessive load of wastewater, and in the second a possible leaching of toxic trace elements, compromises the quality and uses of these waters. For SMR, despite its higher water flow, the obtained data warn to the disregard of it depuration capacity and to the risk of eutrophication, both impacts being directly associated with the lack of treatment of municipal effluents. The minimum processing efficiency was estimated and can be achieved with primary treatment.

The hydrochemical characterization revealed similarities between the samples analyzed, in terms of its cationic and anionic compositions, and distinctive characteristics when the contents of Fe, Al, Zr and pattern of lanthanide are considered. The compositional contrasts result from interactions with the substrates of drainage and from the action of a wide system of complexation and precipitation reactions. Anomalous values of As and Pb require further investigation to trace their source.