Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier.com/locate/marpolbul

Quality index of the surface water of Amazonian rivers in industrial areas in

Pará, Brazil
Adaelson Campelo Medeirosa,c,⁎, Kleber Raimundo Freitas Faiala, Kelson do Carmo Freitas Faiala,
Iris Danielly da Silva Lopesa, Marcelo de Oliveira Limaa, Raphael Mendonça Guimarãesb,
Neyson Martins Mendonçac
a
Instituto Evandro Chagas, Seção de Meio Ambiente, Rodovia BR-316, Km 07 S/N, Levilândia, Ananindeua, PA CEP: 67030-000, Brazil
b
Fundação Oswaldo Cruz, Avenida Brasil, 4365 Manguinhos, Rio de Janeiro, RJ, CEP 21041-360, Brazil
c
Universidade Federal do Pará, Faculdade de Engenharia Sanitária e Ambiental, Rua Augusto Corrêa, S/N Guamá, Belém, PA CEP: 66075-110, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: In this study was to evaluate the waters quality of the Murucupi River, located in urban agglomerate area and
Index intense industrial activity in Barcarena City, Pará State. The Arapiranga River in Abaetetuba City was used as
Water quality control area (Background), next to Barcarena. Was used the Water Quality Index (WQI) based on nine variables
Surface water analized. Waters quality of the Arapiranga and Murucupi rivers were regular to good and bad to good, re-
Water pollution
spectively. Anthropogenic influence on the Murucupi River was higher, mainly by the disposal of domestic
Background
effluents from the urban agglomerate and of the industrial waste tailing basins upstream of this river. Due to its
less inhabited environment and further away from the area urban and industrial, the Arapiranga River was more
preserved. Waters pollution of around these area is increasingly intense, and restricted its uses for various
purposes.

1. Introduction
The classification of the water quality based on the degree of purity
and pollution dates back to 1848, in Germany. Over the years, several
water quality indexes (WQIs) and environmental indicators have been
developed based on physical–chemical and microbiological character-
istics (Sladecek, 1973; Ott, 1978; Steinhart et al., 1982; Dojlido and
Best, 1993; Lumb et al., 2011).
In 1908, one of the first WQIs was used, denominated Saprobic
Index (SI), in order to estimate the level of easily degradable organic
matter in running water (Sladecek, 1969; Washington, 1984). The SI
was insufficient and led to the search for a better index in the following
decades. From there, the numerical indices were developed through
mathematical calculations associated with physical, chemical, and
biological variables and were applied to classify and evaluate the sur-
face waters quality (Horton, 1965). In 1950, the scientists Olaf Helmer
and Norman Dalkey created a new method after several meetings with
experts. This technique marked the beginning of a new field of research
called “opinion technology” (Magalhães Júnior, 2007). The surface
water quality is fundamental for water resources management and
multiple uses. Many studies have been carried out to evaluate the
characteristics of the waters used as public supply sources. These
studies investigated important variables, also used for calculation of the
WQI, providing information about rivers water quality to population
(Silva and Jardim, 2006).
The city of Barcarena, in Pará State, Northern Brazil, has its en-
vironment continuously impacted by industrial and port activities. A
significant part of its territory in Barcarena was used for the installation
of industries. In these, large areas were deforested and traditional po-
pulations had to leave their homes (IEB, 2012). Records showed that
environmental accidents occurred in the surrounding industrial area
since 2000, resulting in the contamination of the surface and under-
ground waters (MP-PA/MPF-PA, 2015). Located next to Barcarena,
Abaetetuba is a city in which there are no records of industrial activ-
ities. However, Abaetetuba and Barcarena have the same problems as
other Brazilian cities, such as basic sanitation and health services pre-
carious.
For instance, due to an accident occurred in this region, the in-
fringing company had to sign a Term for the Adjustment of Conduct
(TAC), in which they established mitigating factors including the
transfer of financial resources to research entities and others to sub-
sidize the monitoring activities in the affected area for a period of four
years (2009 to 2012) (IEC, 2009).
Hence, we report a seasonal evaluation of the surface water quality

⁎
Corresponding author.
E-mail address: adaelsonmedeiros@iec.pa.gov.br (A.C. Medeiros).

http://dx.doi.org/10.1016/j.marpolbul.2017.09.002
Received 8 March 2017; Received in revised form 31 August 2017; Accepted 1 September 2017
0025-326X/ © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

Please cite this article as: Medeiros, A.C., Marine Pollution Bulletin (2017), http://dx.doi.org/10.1016/j.marpolbul.2017.09.002
A.C. Medeiros et al.
near the industrial areas were carried out in 2009 using the WQI for
Murucupi river located in Barcarena. As a regional background, the
same evaluation was carried out on the Arapiranga River in Abaetetuba.
In the last years, the Murucupi is one of the most impacted rivers by
domestic and industrial sewage in this region. The Arapiranga River is
farther away from the industrial area in Barcarena, and effluents dis-
charges were not found in it the tool used in this study (WQI) to assess
the waters quality can subsidize the public polices and reduce the socio-
environmental problems observed in this region.
2. Material and methods
2.1. Study area
The Abaetetuba and Barcarena cities are located in the northeast of
the Pará State and next to Belém City. The territory for both cities is
~ 1,610 and 1,310 km2, respectively. In 2016, the population was es-
timated in 151,934 and 118,537 for Abaetetuba and Barcarena, re-
spectively (IBGE, 2017).
The Arapiranga River is ~ 20 km long. No anthropogenic activities
have been reported along its banks. Its sources are located in the
Abaetetuba City, passing the Highway PA 409 near kilometer 3, and its
mouth is adjacent to Vila de Beja on the right bank of the Pará River.
The Murucupi River is ~ 10 km long; its sources are near a sedi-
mentation basin (red mud) in the alumina production area. The rupture
of this sedimentation basin and red mud spill in the Murucupi River bed
have been reported, with abrupt changes of the physicochemical
characteristics of the surface water (Pereira et al., 2007). In its natural
course, this river flows through the urban districts of Vila dos Cabanos
and Laranjal in the Barcarena City. Untreated sewage was observed in
both districts and the riparian forests have been partially destroyed by
human activities (IEC, 2009).
The Arapiranga and Murucupi rivers present characteristics of fresh
water based on the Brazilian legislation that deals with the classifica-
tion of the surface water and associated environmental guidelines
(CONAMA Resolution 357, 2005) regarding the salinity of the water
bodies. The Amazon Rivers can be classified into three categories:
white, dark, and clear water (Horbe and Santos, 2009). Based on the
classification of the water in the Amazon, the visual characteristics of
the Arapiranga and Murucupi rivers changed from dark to clear in the
direction from their sources to mouths.
In case of the Arapiranga River, eight sample points at > 1000-m
distances spread over an area of ~7677 m were evaluated. In case of
the Murucupi River, eight sample points in a region of ~ 5269 m were
analyzed, with a minimum and maximum spacing between the sample
points of 584 m and 875 m, respectively. The sampling points (Fig. 1) in
the Arapiranga and Murucupi rivers are upstream and downstream of
small tributary drains along the evaluated regions, from their source to
mouths, considering the navigability, accessibility, and field staff and
financial cost to carry out the work.
Both rivers are located in the tidal zone of the Amazon Estuary. The
primary land cover along the shores of these water sources comprises
capoeiras, flooding plains, and sandy plains influenced by the flooding
tide. Based on the classification of Köppen, the temperature is of type
Am and hot equatorial, with an annual average of 27 °C. The seasonal
periods are well defined: rainy from January to June and dry during the
remaining months of the year (Amaral et al., 2002; Souza and Lisboa,
2005). Table 1 shows the regional climatic characteristics of the year
2009 according to the public meteorological database and tidal fore-
casts (INMET, 2009 and Marinha do Brasil, 2009).
2.2. Sampling procedures
The sampling for the analysis of the physicochemical variables and
microbiological was carried out at the sampling points using the col-
lection vessel immersion technique at a maximum depth of 30 cm from
2
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
the water column (CETESB, 2001). Polyethylene flasks with a capacity
of 1 L were used for the physicochemical analyses. The samples were
collected, refrigerated ( ± 4 °C) in isothermal boxes, and transported to
the Laboratory of Toxicology of the Environment Section of the
Evandro Chagas Institute in Ananindeua (PA) shortly after sampling.
Amber polyethylene flasks with a capacity of 1 L were used for the
analysis of the biochemical demand of oxygen (BOD) based on the same
conservation criterion discussed above. Furthermore, 100-mL NASCO®
bags were used for the water samples for the microbiological ex-
amination. The samples were conditioned in isothermal boxes ( ± 4 °C)
and transported to the field laboratory in Barcarena, where the neces-
sary procedures were performed.
2.3. Analyzed variables
The following physicochemical and microbiological variables were
used in the calculation of the WQI: temperature, pH, total dissolved
solids, total suspended solids, dissolved oxygen, BOD, thermotolerant
coliforms, total nitrogen, total phosphorus, and turbidity. The sum of
the total dissolved solids and total suspended solids resulted in the total
residue values.
The analytical methods used for the determination of the variables
obeyed the procedures and methodologies recommended in the
Standard Methods for Examination of Water and Wastewater (APHA/
AWWA/WEF, 2005) and manuals of the DR 2000 and 2800 spectro-
photometers (HACH®, 1990 and 2007). The pH, temperature and total
dissolved solids and dissolved oxygen concentrations were determined
at the time of sampling using previously calibrated HANNA® equipment
HI 769828. The total suspended solids, total nitrogen, and total phos-
phorus contents and turbidity were characterized with the DR 2000 and
2800 spectrophotometers (UV–ViS). The determination of the BOD5 at
20 °C was performed on the BODTRAK equipment of HACH®.
2.4. Calculation of the WQI
The calculation of the WQI was carried out following the criteria of
the National Sanitation Foundation (NSF) adapted by the
Environmental Company of the State of São Paulo (CETESB). Was
performed the weighted products calculation from i-th quality (qi) for
each variable raised to respective weights (wi). The WQI calculation is
described in Eq. (1) (CETESB, 2017):
n
WQI = ∏ qi wi
i=1 (1)
The WQI assumes a number between 0 and 100. The WQI can be
divided into five categories and weights: 0 to 19, worst water quality;
20 to 36, bad quality; 37 to 51, regular quality; 52 to 79, good quality;
and 80 to 100, optimum quality. The qi were obtained from the re-
spective average CETESB quality variation curves, as a function of the
concentration of each determined variable. The wi correspond to a
given value of each observed variable as a function of its importance for
the overall quality. The sum of the wi is equal to one based on Eq. (2):
n
∑ wi = 1
i=1 (2)
2.5. Statistical treatment
Descriptive statistics and Principal Component Analysis (PCA) were
applied to the WQIs of the two water bodies. The normality tests for the
confidence level α = 95% were used to identify the sample distribution
(H0 = 0 for normal distribution and H1 = 0 for non-normal distribu-
tion), with a p-value of the normal distribution > 0.05. The t and
Wilcoxon tests were applied to the normal (H0 = 0) and non-normal
(H1 ≠ 0) data, respectively. The Minitab 17 software (license
A.C. Medeiros et al. Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Fig. 1. Study area and respective sampling points.

Table 1 tests of the dry and ebbing tide period and the flooding tide data re-
Sampling months and regional climate characteristics. vealed a normal (p-value = 0.125) and non-normal distribution (p-
value = 0.008), respectively. The hypothesis tests showed differences
Sampling Seasonal period Rainfall Temperature (°C) Tides (m)
(mm)
(H0 hypothesis rejected) between the rainy and dry period data during
Maximum Minimum High Low ebbing tide; the t-test (p-value = 0.005) indicated a good level of sig-
nificance. During flooding tide, the t-test (p-value = 0.244) showed a
January Dry-rainy 354.5 31.3 23.1 3.4 −0.5 low level of significance, indicating the similarity (hypothesis H1 re-
transition
April Rainy 469.9 30.2 23.2 3.1 0.0
jected) of the WQIs. In general, these similarities and differences in the
July Rainy-dry 193.1 32.7 23.1 3.4 −0.4 two rivers are observed in Fig. 2.
transition There is a great influence of the rainy period on the quality of these
October Dry 45.1 33.7 22.9 3.1 0.2 waters, with a considerable difference between the tidal conditions,
where there are very heterogeneous qualities, probably due to the
leaching of surface materials and water dilution through rainfall and
317201144883580098) was used to create time series graphs and
tidal dynamics, specifically in the dry period, it is observed conditions
perform Anderson–Darling Normality, t, and Wilcoxon tests and mul-
of more homogeneous qualities with improvement in the flooding tide.
tivariate analysis. The results for the rainy and dry seasons were com-
Bad quality conditions were determined during ebbing tide in January
pared.
and April at the sampling points MUR 1 and MUR 2 at which the un-
treated sewage of the Village of Cabanos and Laranjal Communities are
3. Results
discharged.
The variables that influence the water quality of the studied water
3.1. Water quality index of the Arapiranga and Murucupi Rivers
bodies the most are the pH, dissolved oxygen, turbidity, BOD, and
mainly thermotolerant coliforms. Some statistical data and the quality
The values of the WQIs of the Arapiranga River reveal a regular
notes (qi) of the variables used in the calculation of the WQI can be
quality of 48.44% of the data and good quality of 51.56% of the data,
observed in Tables 2, 3 and 4.
indicating a quality improvement in the source–mouth direction of the
water body.
The WQIs of the Murucupi River revealed a bad quality of 7.81% of 3.2. Principal component analysis for the rainy period
the data, regular quality of 50% of the data, and good quality of 42.19%
of the data, indicating a quality improvement in the source–mouth di- Based on PCA, three groups were identified: 1, 2, and 3 (Fig. 3).
rection of this water body. Subgroups A (EA1 to EA6, and FA1) and B (EA7, FA2 to FA5, EM5,
Arapiranga River, the Anderson–Darling normality tests revealed a EM6, and FM4) constitute group 1; the sampling points of the Arapir-
normal distribution between the rainy and dry season data during eb- anga River are dominant. Group 2 includes the subgroups C (EA8, FA7,
bing tide (p-value = 0.739) and flooding tide (p-value = 0.071). The FA8, EM7, and EM8) and D (FA6 and FM5 to FM8) and sampling points
hypothesis tests implied similarities (H1 hypothesis rejected) of the at the mouth of both water bodies. Group 3 comprises the two sub-
ebbing tide data of rainy and dry periods; the t-test (p-value = 0.914) groups E (EM1 and EM2) and F (EM3, EM4, and FM1 to FM3), with a
indicated a low level of significance. However, the Wilcoxon test (p- predominance of the sampling points of the Murucupi River.
value = 0.052) for the flooding tide showed good significance (hy- The WQIs of the subgroup A range from 43 to 50 (AI1–EA6 and
pothesis H0 rejected). FA1), indicating regular quality conditions for these sampling points.
Murucupi River, the comparison of the Anderson–Darling normality The values of the subgroup B vary from 45 to 53 (EA7, FA2 to FA5,

3
A.C. Medeiros et al. Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Fig. 2. Seasonal comparison of WQI values in the Arapiranga

Arapiranga River Murucupi River and Murucupi rivers, ebbing and flooding tide, 2009. Note: ARA
70
(Arapiranga River); MUR (Murucupi River).

60
WQI

50

40

30

Rainy Dry Rainy Dry Rainy Dry Rainy Dry

Ebbing Tide Flooding Tide Ebbing Tide Flooding Tide

Table 2
Descriptive statistics for variables in the rainy and dry season.

Season/tide Variables Unit N Arapiranga River N Murucupi River

X G ± SD Max Min X G ± SD Max Min

Rainy ebbing tide Temperature °C 32 26 ± 0.47 28 26 32 28 ± 0.94 29 27

pH – 32 – 6.2 4.6 32 – 7.2 6.1
DO mg·L− 1 32 6 ± 0.70 6.5 4.5 32 5 ± 1.50 6.8 1.4
BOD mg·L− 1 32 5 ± 1.50 9 4 32 12 ± 4.01 18 6
Thermotolerants ThCU 32 *3.87E + 00 ± 0.26 2.01E 3.26E 32 4.38E + 00 ± 0.67 2.42E 1.48E
coliforms 100 mL− 1 + 04 + 03 + 05 + 03
Total nitrogen mg·L− 1 32 0.29 ± 0.15 0.6 0.2 32 0.84 ± 1.00 3.5 0.4
Total phosphorus mg·L− 1 32 0.06 ± 0.02 0.09 0.04 32 0.16 ± 0.17 0.50 0.05
Total residue mg·L− 1 32 18 ± 6.03 33 12 32 42 ± 7.19 53 26
Turbidity FTU 32 23 ± 7.69 38 15 32 32 ± 8.20 47 18
WQI – 32 48 ± 3.66 57 43 32 43 ± 9.35 60 29
Rainy flooding tide Temperature °C 32 26 ± 1.13 29 26 32 28 ± 1.16 29 26
pH – 32 – 8.1 4.8 32 – 7.2 5.7
DO mg·L− 1 32 6 ± 0.55 7.0 5.1 32 5 ± 1.32 6.9 3.1
BOD mg·L− 1 32 4 ± 1.59 8 2 32 10 ± 2.49 14 6
Thermotolerants ThCU 32 *3.75E + 00 ± 0.22 1.20E 2.69E 32 3.62E + 00 ± 0.66 4.35E 7.76E
coliforms 100 mL− 1 + 04 + 03 + 04 + 02
Total nitrogen mg·L− 1 32 0.24 ± 0.13 0.5 0.1 32 0.63 ± 0.49 2.1 0.3
Total phosphorus mg·L− 1 32 0.05 ± 0.02 0.07 0.03 32 0.12 ± 0.13 0.50 0.04
Total residue mg·L− 1 32 13 ± 6.61 29 6 32 38 ± 5.78 49 30
Turbidity FTU 32 15 ± 5.47 25 8 32 28 ± 8.03 44 18
WQI – 32 53 ± 6.08 62 46 32 50 ± 10.30 63 37
Dry ebbing tide Temperature °C 32 27 ± 1.15 29 26 32 29 ± 1.10 30 27
pH – 32 – 7.3 4.7 32 – 6.6 5.6
DO mg·L− 1 32 6 ± 1.22 7.9 4.3 32 6 ± 0.74 7.5 5.3
BOD mg·L− 1 32 8 ± 2.78 13 4 32 10 ± 2.31 14 6
Thermotolerants ThCU 32 *3.70E + 00 ± 0.93 6.49E 9.80E 32 3.87E + 00 ± 0.79 1.73E 4.10E
coliforms 100 mL− 1 + 04 + 00 + 05 + 02
Total nitrogen mg·L− 1 32 0.22 ± 0.05 0.3 0.1 32 0.46 ± 0.30 1.1 0.2
Total phosphorus mg·L− 1 32 0.04 ± 0.02 0.08 0.02 32 0.02 ± 0.02 0.07 0.01
Total residue mg·L− 1 32 21 ± 10.28 40 11 32 40 ± 9.28 57 26
Turbidity FTU 32 18 ± 4.75 25 11 32 27 ± 3.43 34 22
WQI – 32 48 ± 4.61 59 42 32 50 ± 8.26 66 39
Dry flooding tide Temperature °C 32 27 ± 1.06 29 26 32 29 ± 0.94 30 27
pH – 32 – 7.6 4.8 32 – 7.1 5.6
DO mg·L− 1 32 6 ± 0.60 6.8 4.8 32 7 ± 0.69 7.9 6.0
BOD mg·L− 1 32 5 ± 2.73 11 2 32 8 ± 2.17 13 5
Thermotolerants ThCU 32 *3.30E + 00 ± 0.41 7.22E 1.60E 32 3.92E + 00 ± 0.81 2.42E 8.40E
coliforms 100 mL− 1 + 03 + 02 + 05 + 02
Total nitrogen mg·L− 1 32 0.19 ± 0.06 0.3 0.1 32 0.41 ± 0.29 1.2 0.2
Total phosphorus mg·L− 1 32 0.02 ± 0.01 0.04 0.01 32 0.02 ± 0.02 0.08 0.00
Total residue mg·L− 1 32 18 ± 9.72 39 10 32 36 ± 7.72 50 25
Turbidity FTU 32 17 ± 5.00 32 13 32 23 ± 2.02 27 20
WQI – 32 55 ± 5.75 65 47 32 52 ± 8.56 67 41

N: data number; X G : geometric average; Max: maximum; Min: minimum. *- values on log 10 base.

4
 Table 3
Quality parameters of the variables in the Arapiranga and Murucupi rivers during ebbing tide.

Points Variable quality notes (q)

A.C. Medeiros et al.

pH Dissolved oxygen BOD Thermotolerant coliform Total nitrogen Total phosphorus Total residue Turbidity

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

ARA 1 17 25 19 52 30 56 31 57 50 36 30 61 14 9 15 5 98 97 98 98 95 95 96 94 82 82 81 82 66 59 75 64
ARA 2 20 32 19 71 37 59 29 57 50 45 45 54 14 8 13 6 98 97 99 98 95 93 97 96 82 83 81 83 61 54 74 62
ARA 3 19 42 22 79 39 57 34 58 61 50 28 45 12 9 13 5 99 97 98 98 95 94 97 95 82 83 82 83 68 53 72 59
ARA 4 18 37 24 80 35 59 37 69 61 50 40 45 10 10 16 5 99 97 98 98 95 95 95 92 82 83 82 84 69 55 69 57
ARA 5 23 57 20 87 37 60 39 70 50 40 30 50 13 8 13 6 99 95 98 98 95 94 96 94 82 83 82 84 67 52 67 59
ARA 6 26 44 29 79 42 60 41 75 54 50 31 50 11 7 67 5 99 96 98 98 93 91 96 96 82 83 82 85 63 48 67 58
ARA 7 36 65 36 90 56 62 47 76 61 54 30 50 12 7 12 7 98 96 98 98 93 92 96 97 82 84 83 84 61 47 67 58
ARA 8 65 68 64 93 62 64 60 71 61 61 25 50 12 8 13 6 98 96 98 98 92 91 96 94 82 84 83 85 62 46 66 58
MUR 1 67 65 48 60 23 9 48 52 17 21 30 30 3 3 3 8 82 97 92 97 63 91 93 97 85 85 85 84 54 41 60 57
MUR 2 66 71 50 70 32 23 52 56 17 14 28 31 5 3 3 11 70 96 91 97 63 95 94 98 85 85 86 84 52 44 53 53
MUR 3 69 76 56 60 35 31 44 60 17 15 25 25 5 4 4 11 89 96 93 97 75 94 94 97 85 85 86 85 52 40 50 53
MUR 4 73 75 57 42 40 41 49 63 30 14 31 21 7 4 5 14 90 97 93 98 75 92 93 98 85 85 85 84 50 47 52 54
MUR 5 79 85 47 79 52 57 74 70 31 28 36 28 10 5 7 23 90 97 94 98 75 95 95 98 85 85 86 85 52 48 49 54
MUR 6 84 91 55 94 59 59 66 72 31 21 30 28 12 7 8 17 90 96 95 98 75 93 96 98 83 85 85 84 57 43 52 55
MUR 7 92 85 55 74 63 64 76 74 40 36 40 45 17 9 8 22 90 97 96 97 69 93 97 98 84 85 84 84 61 48 56 58
MUR 8 92 87 62 90 67 62 80 75 50 45 50 40 18 4 10 27 90 97 96 99 69 90 98 98 84 85 83 84 64 53 59 58

ARA: Arapiranga; MUR: Murucupi; 1: January; 2: April; 3: July; 4: October; Obs: The constant value of 94 was adopted according to CETESB because the temperature values were in equilibrium and maximally varied by 2 °C between the sampling
points.

5
Table 4
Quality parameters of the variables in the Arapiranga and Murucupi rivers during flooding tide.

Points Variable quality notes (q)

pH Dissolved oxygen BOD Thermotolerant coliform Total nitrogen Total phosphorus Total residue Turbidity

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4

ARA 1 21 21 19 32 41 44 35 51 61 40 40 78 11 12 21 15 99 98 99 98 82 95 98 97 82 81 82 82 70 82 72 71
ARA 2 26 22 20 35 45 44 41 44 61 54 54 78 11 9 19 14 99 98 98 98 82 95 98 96 82 81 82 81 64 80 72 72
ARA 3 22 28 22 44 46 44 46 42 78 61 36 69 11 10 18 13 99 98 98 98 82 95 98 96 82 81 82 82 64 67 70 64
ARA 4 28 33 25 54 42 42 49 45 78 61 42 78 15 9 14 13 99 96 98 98 82 95 95 96 82 82 82 82 63 61 67 64
ARA 5 38 43 34 65 43 44 52 62 61 50 54 36 13 9 10 13 99 96 99 98 82 95 97 97 82 82 81 84 61 66 70 61
ARA 6 68 79 79 78 48 47 56 62 61 54 45 61 14 11 36 13 99 97 99 98 83 93 96 96 83 82 83 84 58 69 66 60
ARA 7 83 91 91 88 68 47 61 59 78 61 50 50 15 10 19 11 99 97 98 98 84 93 97 97 84 81 84 84 57 81 64 59
ARA 8 92 87 90 92 71 51 66 68 61 78 30 54 13 8 21 17 99 97 98 98 84 93 96 98 84 81 84 85 67 81 63 50
MUR 1 68 49 49 57 28 21 55 56 21 31 45 36 9 5 3 7 84 97 91 97 69 95 92 96 85 84 85 85 51 47 61 58
MUR 2 67 48 51 42 33 27 57 57 21 30 25 40 6 5 4 3 89 97 92 98 63 94 94 96 85 85 85 85 52 42 55 59
MUR 3 70 58 50 48 39 35 62 55 25 30 40 40 8 5 6 4 92 97 94 97 82 94 95 96 85 85 85 85 51 42 58 56
MUR 4 75 77 52 50 47 52 61 64 28 31 36 45 13 10 10 7 93 97 96 97 82 95 97 97 85 85 83 84 52 48 60 57
MUR 5 81 86 58 57 64 67 76 72 28 36 30 40 14 15 13 19 93 97 96 98 82 95 96 97 84 85 85 84 58 51 59 57
MUR 6 89 91 67 59 67 66 74 71 30 45 36 31 21 20 12 18 92 96 96 98 82 92 97 98 84 84 85 83 64 56 60 61
MUR 7 92 92 62 94 69 69 80 80 36 50 45 54 23 20 12 22 93 97 96 99 82 92 98 97 84 84 85 84 64 58 61 60
MUR 8 92 92 63 92 70 68 82 83 40 50 54 54 22 20 15 22 93 97 96 98 82 94 97 99 84 84 84 83 64 59 62 61

ARA: Arapiranga; MUR: Murucupi; 1: January; 2: April; 3: July; 4: October; Obs: The constant value of 94 was adopted according to CETESB because the temperature values were in equilibrium and maximally varied by 2 °C between the sampling
points.
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
A.C. Medeiros et al. Marine Pollution Bulletin xxx (xxxx) xxx–xxx

1,0
water bodies.
EM8 The WQIs of the subgroup A (EM1 to EM4, AI2, and EA5) range
C from 39 to 52, indicating conditions of regular and good quality for
Group 1 FA7
0,5 these sampling points. The WQI values of subgroup B (FM1 to FM3, AI1,
EM6 EM7
PC2(5.4%)

EM4 FA4 FA8 EA3, and EA4) vary from 41 to 47 and thus reflect regular quality
E B EM5 FA5 EA8
EM2 FM1 FM3 conditions.
EM1 F FA2
0,0 FM2 FA3 EA7 The WQIs of the subgroups C (EA7, EA8, FA1 to FA5, and FM4) and
EM3 FM4 FA6
FA1 FM8 D (EM5 to EM7, FM5, and FM6) range from 47 to 55 and 48 to 63,
Group 3 EA2 EA6 D FM7 respectively, implying regular and good quality conditions. The quality
EA3
EA1 FM5 FM6
-0,5 A EA5 of groups 3 (EA6, FA6, and FA7) and 4 (FA8, EM8, FM7, and FM8) is
EA4 regular and good, with a variation of 50 to 67, respectively.
Group 2
The PC1 (85.2%) indicates the distinct separation between groups 1,
-3 -2 -1 0 1 2 3
3, and 4 (y-axis) and a complete separation of groups 1 and 3–4. The
PC1(94.6%) PC2 (14.8%) shows the separation between the groups 3 and 4 and
Fig. 3. PCA analysis score plot for the Arapiranga and Murucupi rivers in the rainy season
subgroups A–B and C–D (x-axis).
(January and April); EA: Ebbing tide in Arapiranga; EM: Ebbing tide in Murucupi; FA: To sum up, the rivers show similar characteristics at both their
Flooding tide in Arapiranga; FM: Flooding tide in Murucupi. Subgroups: A, B, C, D, E, and sources and mouths during the dry period, indicating that, in contrast to
F. the rainy season, the rainfall or tidal conditions did not have a great
influence on the water quality of these rivers.
EM5, EM6, and FM4), implying regular and good quality conditions. The pH, temperature and dissolved oxygen were the main variables
The WQIs of subgroup C (AI8, FA7, FA8, EM7, and EM8) range between influencing the formation of groups 1 and 2 in the rainy season (Fig. 3)
48 and 62, indicating regular and good quality conditions. and 2, 3 and 4 in the dry season (Fig. 4), with increase in the source-
The subgroup D (FA6 and FM5 to FM8) represents good quality mouth direction.
conditions, with WQI values between 54 and 63. The WQI of subgroup The anthropic influence is more visible in group 3 (rainy season) for
E (EM1 and EM2) ranges from 29 to 40, presenting bad and regular the closest sampling points of the sources of the Murucupi river (Fig. 3),
quality conditions for the sample points. The water quality of subgroup where higher values were recorded for the thermotolerant coliform
F is regular, with a WQI variation from 37 to 40. variables, BOD, turbidity, total residue, nitrogen Total and total phos-
The PC1 (94.6%) shows a distinct and complete separation between phorus. In the dry period, this behavior was repeated in the waters of
groups 2 and 3 on the y-axis; the PC2 (5.4%) shows the separation the Murucupi River as group 1 (Fig. 4) and also similar characteristics
between subgroups A–B and C–D on the x-axis. were observed at the sampling points near the sources of the Arapiranga
In summary, different quality conditions are observed for the water River, probably due to the increase in the levels of thermotolerant co-
bodies; the quality improves in the source–mouth direction. In the rainy liforms and total phosphorus in this period. The water quality level of
season, the sources of the two rivers have different characteristics the Murucupi River is coherent with the field observations, which de-
(groups 1 and 3), which is probably due to the water dilution and scribe the continuous discharge of sewage of the Vila dos Cabanos and
leaching of surface material. The quality at the mouths of the two rivers Laranjal.
is similar based on group 2, which is possibly related to the great in- For a better understanding, under the aspects of the determined
fluence of the water of the Pará River. variables, the same behavior was observed between the sampling points
in the rainy and dry periods, with different characteristics in the ex-
tensions evaluated (nascente-foz) according to Fig. 5. In the rainy
3.3. Principal component analysis of the dry period season there was a better separation and formation of the groups of the
variables (Fig. 5 A, PC1: 46.6% and PC2: 25.9%), with different char-
Based on PCA, four groups were identified: 1, 2, 3, and 4 (Fig. 4). acteristics in the waters of these rivers. In the dry period, the sampling
Group 1 comprises the subgroups A (EM1 to EM4, EA2, and EA5) and B points of these water bodies were similar in the mouths (Fig. 5 C; PC1:
(FM1 to FM3, AI1, EA3, and EA4) and group 2 includes the subgroups C 42.5% and PC2: 23.5%). The variables that influenced in the formation
(EA7, EA8, FA1 to FA5, and FM4) and D (EM5 to EM7, FM5, and FM6). of groups in the Arapiranga and Murucupi rivers are shown in Fig. 5 (B:
Groups 1 and 2 contain a mixture of sample data from the Arapiranga 1.january/2.april and D: 3.july/4.october), confirming greater influ-
and Murucupi rivers. Groups 3 (EA6, FA6, and FA7) and 4 (FA8, EM8, ence of the Thermotolerant Coliforms (TC), BOD, Turbidity (Turb),
FM7, and FM8) include the sampling points at the mouths of these two Total Residue (TR), Total Nitrogen (TN) and Total Phosphorus (TP) in
the Murucupi River.
1,0
EM5
Group 4
EM2 A 4. Discussion
EM7
0,5 EA5 EM6 EM8 FM7
EM1 EM3
FA4 D FM8
EM4 FA1 The WQI is a mathematical tool, which is used to simplify the in-
PC2 (14.8%)

FA5
FM1 EA2 EA7 FM5
0,0 formation about analytical water quality variables; a single number
EA3 FA2 FA3 FA8
EA1 FM6 represents the water quality level. The WQI is widely used in land use
C EA8
-0,5 FM2 EA4
B FM4 planning and water resources management (Giljanovic, 1999). It is an
Group 3 important indicator of the water quality, which is used by the compe-
FM3
Group 2
FA7 tent authorities to manage the resources based on various purposes. For
-1,0 Group 1 FA6 example, the WQI developed in New Zealand for the management of
EA6
water resources has been applied for more than two decades to four
-1,5 classes waters intended for bathing, water supply, spawning of fish and
-2 -1 0 1 2 3
general uses, considering, above all, the protection of the aquatic life
PC1(85.2%)
(Smith, 1990).
Fig. 4. PCA analysis score plot for the Arapiranga and Murucupi rivers in the dry season The WQIs determined for the water of the Arapiranga and Murucupi
(July and October). EA: Ebbing tide in Arapiranga; EM: Ebbing tide in Murucupi; FA: rivers revealed quality conditions similar to those of other water bodies
Flooding tide in Arapiranga; FM: Flooding tide in Murucupi. Subgroups: A, B, C, and D.
located in the vicinity of large urban centers such as the Sinos River in

6
A.C. Medeiros et al. Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Fig. 5. Principal component analysis of

RAINY SEASON (A) DRY SEASON (C) score plot type (A and C) and loading plot (B
4 Outlie r 4
Group 1 FM1
EM1 Group 3 and D) about data of variables determined
3 Group 1
FM1 3 EM1 EM2 in the rainy season (A–B) and dry (C–D) in
FA1 EA1 EM2 FM2
2 FA3 FA2 2 EA4
EM3 2009. Note: EA (Ebbing Tide Arapiranga);
FA4 EA4 EA2 FM2
EA1 FM3

PC2 (23.5%)
FA (Flooding Tide Arapiranga); EM (Ebbing
PC2 (25.9%)

FA5 EA3 Group 4 EM4

1 1 EA3 EA5
EA5 FM3 EM3 EA2 Tide Murucupi); FM (Flooding Tide
EA6 FA1 EA6
0 EA7 0 FA3 FA4 EM5 Murucupi).
FA6 EM4 EA7
FA2 EA8
FM4 EM6
-1 EA8 FM4 -1 FA5
FM5
EM5 FA6
FM5 -2 FM6 EM7
-2 FA7
EM6 FA7
FA8 EM7 FA8
-3 Group 2 FM8 FM6 Group 3
-3 Group 2 FM7
EM8 FM8 EM8
-4 FM7 -4
-5,0 -2,5 0,0 2,5 5,0 7,5 -5,0 -2,5 0,0 2,5 5,0
PC1(46.6%) PC1 (42.5%)

RAINY SEASON (B) DRY SEASON (D)

0,2 TC2 0,4
TC1 TC3 TP3
BOD2 TN4
0,3
0,1 TURB2 TN1 TC4 TN3
BOD3
TURB1 BOD1 TP4
0,2
0,0 TP1 BOD4
PC2 (25.9%)

PC2 (23.5%)
TN2 TR1 TR4
0,1 TR3
-0,1 TR2 TURB4 TURB3
0,0
TP2
-0,2
-0,1
DO2
-0,3 T2 T4
-0,2
pH2 pH1 pH3
pH4
-0,4 DO1 T1 -0,3 T3
DO3
DO4
-0,5 -0,4
-0,2 -0,1 0,0 0,1 0,2 0,3 -0,2 -0,1 0,0 0,1 0,2 0,3 0,4
PC1 (46.6%) PC1 (42.5%)

southern Brazil that is impacted by domestic and industrial untreated forests (Sioli and Klinge, 1962; Sioli, 1985; Queiróz et al., 2009).
effluents (Steffens et al., 2015). The lack of adequate and effective sa- The thermotolerant coliform variable is relevant for the quality of
nitation services and treatment processes has been the main reason for this water, especially at the sampling points closest to the sources of the
negative quality conditions, which compromise the use of the water. Murucupi River, representing an environment with fecal material
It should be pointed out that the Arapiranga River has a better water characteristic for sanitary sewage. A similar situation was observed in
quality because it is further away from the urban influence. In contrast, the surface waters of Pontal do Paraná (PR) in southern Brazil, which
the urban agglomerate and industrial facilities strongly and negatively shows high values of thermotolerant coliforms, especially in the dry
affect the quality of the water of the Murucupi River due to large daily season when tourism increases (Siqueira et al., 2009). In another study
discharges of domestic effluents and recurrent leaks of industrial ef- carried out in Zaria, Nigeria, high values of thermotolerant coliforms
fluents from tailings ponds such as kaolin tailings in 2007 and red mud were detected in the sources used for public supply, irrigation of
in 2009 (analytical results not discussed in this study), which reached commercial crops, and recreation (Agbogu et al., 2006), a reality ob-
the nearest water bodies, caused fish mortality and great losses to the served in several water bodies close to urban areas, where sanitation
local population, and compromised the quality of the groundwater of conditions are precarious and the treatment of the produced effluents is
wells used for human consumption in the vicinity of the tailings ponds inadequate or non-existent.
and water at the sources of the Murucupi River and other tributaries of Multivariate analysis helped to compile the data and better visualize
the Pará River (Lima et al., 2011 and MP-PA/MPF-PA, 2015). In ad- the water quality of the Arapiranga and Murucupi rivers in the rainy
dition, the frequent deforestation of large areas around these water and dry periods under different tidal conditions. Three quality groups
bodies and other urban activities harm the water quality (IEB, 2012). were identified in the rainy season and four groups were observed in
The water of the Arapiranga and Murucupi rivers is also used by the dry period for the two rivers. This difference is probably related to
riverside families for individual supplies, bathing, and fishing. the lower influence of the rain leaching the surface materials and,
However, most of the water used for human consumption by the ma- consequently, the alteration of the analyzed variables used in the WQI
jority of the population of these municipalities is obtained from un- calculation for the dry period. Irrespective of the rain, the two rivers
derground aquifers. Although these underground sources have soil as a have different and similar quality indexes at their sources and mouths,
protective barrier against the action of pollutants, they are not exempt respectively. Statistical methods, which are widely used in other en-
from the anthropogenic influences in these municipalities based on a vironmental studies, have contributed to the assessment of the water
WQI study of the groundwater of communities in the area used for quality levels, quality groups, and pollution levels such as the physical-
human consumption. The results of this study indicated that 64.28% chemical assessment of Lake Uluabat in Turkey, the water of the Three
(rainy season) and 78.57% (dry period) of the samples analyzed in the Gorges Region in China, and Japan (Iscen et al., 2008; Zhao et al., 2011;
community of Village of Conde in the municipality of Barcarena present Shrestha and Kazama, 2007).
a poor quality index (Medeiros et al., 2016).
The variables pH, dissolved oxygen, turbidity, biochemical oxygen
5. Conclusions
demand and thermotolerant coliforms were the ones that most influ-
enced negatively in the water quality of the studied rivers according to
The IQA was an important tool used to evaluate the water quality of
the quality notes (Tables 2 and 3). Despite having presented lower
the studied rivers. However, it was observed in this study that the
quality note, the pH most acidic in Amazonian rivers is quite common,
thermotolerant coliforms variable presented higher importance on the
due to the presence of humic and fulvic acids of the soils and riparian
calculated indices. The pH values also indicated quality notes (qi) lower

7
A.C. Medeiros et al.
for these rivers, especially at the sampling points closest to the sources.
However, the pH of the water is more acidic in several localities of the
Amazon Region under natural condition due to the organic decom-
position influence and leaching of fulvic and humic acids from micro-
biological activities in soils and riparian forest.
The efficacy of the WQI and statistical instruments used to evaluate
the water quality of the Arapiranga and Murucupi rivers was confirmed
based on the observation of the significant differences between the
rainy and dry seasons. The Murucupi River was more affected by an-
thropogenic activities due the proximity to the Barcarena City, where
the urban influence was more evident, whereas the Arapiranga River
(Background) was less influenced by such activities.
With respect to the possibility of supplying water to the local po-
pulation, the best WQIs of the four sampling campaigns were de-
termined at the rivers mouth. This study indicates that there is an in-
creasing need to generate information about the water quality in the
Amazon Region because the riverside population uses untreated water.
The quality conditions of the studied rivers vary in space and time and
are important information to technically subsidize the local re-
presentatives with respect to possible future water supply and for the
technical–scientific community.
The results of the determined IQAs show a reality for the six (09)
important variables that were used in the calculation methodology
applied in this study, suggesting a better evaluation involving other
important variables such as sulfate, btex (benzene, toluene, ethylben-
zene and xylenes), Aluminum, manganese, iron, lead, cadmium, mer-
cury and other toxic elements. It is known that generally these other
variables can be associated to the emission of domestic effluents and
residues resulting from industrial activities and there is a reality ob-
served in the surroundings of the study area. The addition of these
variables in the IQA calculation would help to improve the degree of
sensitivity for a more thorough assessment of the quality of these wa-
ters. It would be interesting to implement automated daily monitoring
systems for a better view on the influence of tides and pollutant load on
the hydrographic basin of these water bodies.
Despite the large volume water volume existing in the hydrographic
basin of the Pará River which probably contribute to recovery of these
water bodies, the information generated in this study indicates that
greater attention should be given to adequate treatment of the domestic
and industrial sewage in this region, mainly extreme care with the re-
sidues stored in the tailings basins that has already caused great impacts
on these rivers, damaging the aquatic ecosystem and the life of the local
population.
Acknowledgments
Thanks to IEC/FIDESA/MPE-PA Project (Process 001/2007),
Evandro Chagas Institute, Federal University of Pará and Federal
University of Rio de Janeiro for financial and laboratory support.
Thanks Paulo Trindade for help with the map elaboration.
References
Agbogu, V.N., Umoh, V.J., Okuofu, C.A., Smith, S.I., Ameh, J.B., 2006. Study of the
bacteriological and physicochemical indicators of pollution of surface waters in Zaria,
Nigeria. Afr. J. Biotechnol. 5 (9), 732–737.
Amaral, D.D., Bastos, M.N., Silva, A.S.L., Oliveira, J., Lisboa, L.R.C., Rosário, C.S., Gomes,
A., Silva, C.A., Aguiar, J., 2002. Inventário da Flora da região de Barcarena, Pará.
Relatório Final. Ministério da Ciência e Tecnologia, Museu Paraense Emílio Goeldi,
Belém.
American Public Health Association (APHA), 2005. Standard Methods for the
Examination of Water and Wastewater, 21th. Ed. (Washington, 1600 p).
Companhia Ambiental do Estado de São Paulo (CETESB), 2001. Guia de coleta e
preservação de amostras de água. (São Paulo,150 p).
Companhia Ambiental do Estado de São Paulo (CETESB), 2017. Índice de Qualidade das
Águas. http://aguasinteriores.cetesb.sp.gov.br/wp-content/uploads/sites/32/2013/
11/02.pdf.
Conselho Nacional de Meio Ambiente (CONAMA), 2005. Dispõe sobre a classificação dos
corpos de água e diretrizes ambientais para o se enquadramento. Alterada pelas
8
Marine Pollution Bulletin xxx (xxxx) xxx–xxx
Resoluções 410/2009 e 430/2011. In: Diário Oficial da República Federativa do
Brasil, Brasília, DF. 53. pp. 58–63. http://www.mma.gov.br/port/conama/res/
res05/res35705.pdf.
Dojlido, J.R., Best, G.A., 1993. Chemistry of Water and Water Pollution. Ellis Horwood,
London.
Giljanovic, N.S., 1999. Water quality evaluation by index in Dalmatia. Water Res. 3 (16),
3423–3440.
Hach Company, 1990. DR 2000 Spectrophotometer Procedures Manual, 4th ed. S.l.: s.n..
(655 p).
Hach Company, 2007. DR 2800 Spectrophotometer Procedures Manual, 2nd ed. S.l.: s.n..
(814 p).
Horbe, A.M.C., Santos, A.G.S., 2009. Chemical composition of black-watered rivers in the
western Amazon Region (Brazil). J. Braz. Chem. Soc. 20 (6), 1119–1126.
Horton, R.K., 1965. An index number system for rating water quality. J. Water Pollut.
Control Fed. 37 (3), 300–306.
Instituto Brasileiro de Geografia e Estatística (IBGE)-BRASIL, 2017. Informações gerais
sobre as cidades de Abaetetuba e Barcarena no Estado do Pará. http://cidades.ibge.
gov.br/xtras/uf.php?coduf=15&search=para&lang=_EN.
Instituto Evandro Chagas (IEC), 2009. Relatório técnico SAMAM 001: caracterização dos
impactos ambientais, danos ao ecossistema e riscos a saúde decorrentes do
lançamento no rio Murucupi de efluentes do processo de beneficiamento de bauxita,
Barcarena-Pará. Ananindeua, PA, Brasil. 32 p.. http://iah.iec.pa.gov.br/iah/fulltext/
pc/relatorios/barcarena2009.pdf.
Instituto Internacional de Educação do Brasil (IEB), 2012. Posicionamento da rede da
sociedade civil Pró-Fórum em Barcarena. (Belém, PA, 62 p).
Instituto Nacional de Metereologia (INMET), 2009. Banco de dados metereológicos para
ensino e pesquisa (BDMEP). http://www.inmet.gov.br/portal/index.php?r=bdmep/
bdmep.
Iscen, C.F., Emiroglu, Ö., Ilhan, S., Arslan, N., Yilmaz, V., Ahiska, S., 2008. Application of
multivariate statistical techniques in the assessment of surface water quality in
Uluabat Lake, Turkey. Environ. Monit. Assess. 144, 269–276.
Lima, M.O., Santos, E.C.O., Jesus, I.M., Medeiros, A.C., Faial, K.C.F., Alves, C.N., 2011.
Assessment of surface water in two Amazonian rivers impacted by industrial waste-
water, Barcarena City, Pará state (Brazil). J. Braz. Chem. Soc. 22 (8), 1493–1504.
Lumb, A., Sharma, T.C., Bibeault, J.F., 2011. A review of genesis and evolution of water
quality index (WQI) and some future directions. Water Qual Expo Health 3, 11–24.
Magalhães Júnior, A.P., 2007. Indicadores ambientais e recursos hídricos: realidades e
perspectivas para o Brasil a partir da experiência francesa. Bertrand Brasil, Rio de
Janeiro (688 p).
Marinha do Brasil, 2009. Diretoria de Hidrografia e Navegação: Previsões de marés do
porto de Vila do Conde-PA. Centro de Hidrografia da Marinha. Banco Nacional de
Dados Oceanográficos. http://www.mar.mil.br/dhn/chm/box-previsao-mare/
tabuas/index.htm.
Medeiros, A.C., Lima, M.O., Guimarães, R.M., 2016. Assessment of the quality of water for
consumption by river-bank communities in areas exposed to urban and industrial
pollutants in the municipalities of Abaetetuba and Barcarena in the state of Pará,
Brazil. Ciência & Saúde Coletiva 21 (3), 695–708. http://dx.doi.org/10.1590/1413-
81232015213.26572015.
Ministério Público do Estado do Pará (MP-PA)/Ministério Público Federal (MPF-PA),
2015. Inquérito civil público n° 1.23.000.000661/2015-70: Ação civil pública com
pedido de liminar. http://www.mpf.mp.br/pa/sala-de-imprensa/documentos/2016/
acp-agua-potavel-barcarena.
Ott, W.R., 1978. Environmental Quality Indices: Theory and Practice. Ann Arbor Science
Publishers.
Pereira, S.F.P., Lima, M.A., Freitas, K.H., Mescouto, C.S., Saraiva, A.F., 2007. Estudo
químico ambiental do rio Murucupi-Barcarena, PA, Brasil, área impactada pela
produção de alumínio. Revista Ambi-Água 2 (3), 62–82. http://dx.doi.org/10.4136/
ambi-agua.34.
Queiróz, M.M.A., Horbe, A.M.C., Seyler, P., Moura, C.A.V., 2009. Hidroquímica do rio
Solimões na região entre Manacapuru e Alvarães-Amazonas-Brasil. Acta Amazon. 39
(4), 943–952.
Shrestha, S., Kazama, F., 2007. Assessment of surface water quality using multivariate
statistical techniques: a case study of the Fuji river basin, Japan. Environ. Model.
Softw. 22, 464–475.
Silva, G.S., Jardim, W.F., 2006. Um novo índice de qualidade das águas para proteção da
vida aquática aplicado ao Rio Atibaia, região de Campinas/Paulínia-SP. Química
Nova 29 (4), 689–694. http://www.scielo.br.sci-hub.org/pdf/qn/v29n4/30244.pdf.
Sioli, H., 1985. Amazônia: fundamentos da ecologia da maior região de florestas tropicais.
Vozes, Petrópolis (72 p).
Sioli, H., Klinge, H., 1962. Solos, típicos de vegetação e águas na Amazônia. Amazoniana
1, 27–41.
Siqueira, A., Godinho, M.J.L., Kolm, H.E., Machado, E.C., 2009. Evaluation of the water
quality of tidal creeks of Pontal do Paraná, Paraná, Brazil. Braz. Arch. Biol. Technol.
52 (2), 483–492.
Sladecek, V., 1969. The measures of saprobity. In: Verhandlungen der Internationalen
Vereinigung für Theoretische und Angewandte Limnologie. 17. pp. 546–559.
Sladecek, A., 1973. System of water quality from the biological point of view. In: Archiv
fur Hydrobiologie Beihefte. Ergebnisse der Limnologie, (218 p).
Smith, D.G., 1990. A better water quality indexing system for rivers and streams. J. Braz.
Wat. Res. 24 (10), 1237–1244.
Souza, A.P.S., Lisboa, R.C.L., 2005. Musgos (Bryophyta) na Ilha Trambioca, Barcarena,
PA, Brasil. Acta Bot. Bras. 19 (3), 487–492. http://dx.doi.org/10.1590/S0102-
33062005000300008.
Steffens, C., Klauck, C.R., Benvenuti, T., Silva, L.B., Rodrigues, M.A.S., 2015. Water
quality assessment of the Sinos River – RS, Brazil. Braz. J. Biol. 75 (4) (Suppl. 2),
S62–S67.
A.C. Medeiros et al. Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Steinhart, C.E., Schierow, L.J., Sonzogni, W.C., 1982. An environmental quality index for Zhao, J., Fu, G., Lei, K., Li, Y., 2011. Multivariate analysis of surface water quality in the
the great lakes. Water Resour. Bull. 18 (6), 1025–1031. Three Gorges area of China and implications for water management. J. Environ. Sci. 23
Washington, H.G., 1984. Diversity, biotic and similarity indices. A review with special (9), 1460–1471.
relevance to aquatic ecosystems. Water Res. 18 (6), 653–694.