International Soil and Water Conservation Research 4 (2016) 284–292

HOSTED BY
Contents lists available at ScienceDirect

International Soil and Water Conservation Research

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

Original Research Article

Assessment of spatial and seasonal water quality variation of Oum Er Rbia MARK
River (Morocco) using multivariate statistical techniques☆
⁎
Ahmed Barakata, , Mohamed El Baghdadia, Jamila Raisa, Brahim Aghezzafb, Mohamed Slassib
a
Georessources and Environment Laboratory, Sultan My Slimane University, Béni-Mellal 23000, Morocco
b
Hydraulic Basin Agency of Oum Er Rbia (ABHOER), Béni-Mellal 23000, Morocco

A R T I C L E I N F O A BS T RAC T

Keywords: The aim of this study is to assess the spatial and temporal water quality variation and to determine the main
Oum Er Rbia basin contamination sources in the Oum Er Rbia River and its main tributary, El Abid River. The water quality data
Temporal/spatial assessment were collected during 2000–2012 from fourteen sampling stations distributed along the river. The water quality
Water quality indicators used were TEMP, pH, EC, turbidity, TSS, DO, NH4+, NH3–, TP, BOD5, COD and F. coli. The water
Multivariate statistics
quality data was analyzed using multivariate statistical methods including Pearson's correlation, PCA, and CA.
The results showed that in some stations the water quality parameters were over Moroccan water standards.
PCA applied to compare the compositional patterns among the analyzed water samples, identified and four
factors accounting for almost 63% of the total variation in the data. This suggests that the variations in water
compounds’ concentration are mainly related to point source contamination (domestic and industrial waste-
water), non-point source contamination (agriculture activities), as well as natural processes (weathering of soil
and rock). CA showed relatively spatial and seasonal changes in surface water quality, which are usually
indicators of contamination with rainfalls or other sources. Overall, this study showed that the water was
potentially hazardous to health of the consumers and highlighted the need to treat industrial and municipal
wastewater and to encourage sustainable agricultural practices to prevent adverse health effects. We therefore
suggest wise management of anthropogenic activities in the catchment of Oum Er Bia River and their
tributaries.

1. Introduction evaluating the quality of river water are required for integrated
management of these water resources (Singh et al., 2005). For
The rivers represent the major source of water used for human monitoring the quality of river water, sampling networks seem to be
consumption, culture irrigation, and industrial purposes. Efficient an excellent source of information for local and temporal vision of the
management of these water resources requires information about the state of the water of the river. These networks provide an overview of
river water quality and its variability. This need is quite marked in the temporal condition as well as seasonal and geographical evolution
semi-arid countries such as Morocco whose water resources are of the ecosystem (Berzas et al., 2000; Simeonov et al., 2003). To assist
becoming increasingly difficult to renew, due to their over exploitation in the processing and analyzing these data that have been increasing
by rapidly growing population. The deterioration of river water quality over time, the more appropriate and most used methods are the
water can result from natural processes and more recently due to multivariate statistics (i.e. López‐López et al., 2014; ; Wang et al.,
anthropogenic activities through the discharge of industrial and 2014; Osman et al., 2012; Garafa et al., 2011; Bouza-Deano et al.,
domestic wastewater as well as agricultural drainage to the rivers. 2008; Idris, 2008; Mendiguchía et al., 2004).
However, the big bulk of rivers’ pollution comes from industrial and Recently, the multivariate statistical approach becomes popular for
domestic wastewater and agricultural drainage (Carpenter et al., 1998; a better understanding of water quality and ecological status, due to
Jarvie et al., 1998). Seasonal variations in both of these anthropogenic their ability to treat large volume of spatial and temporal data from a
and natural processes such as temperature and precipitations, affect variety of monitoring sites. In the scientific literature, different
the quality of river water and lead to different attributes between statistical techniques, including a cluster analysis (CA), principal
seasons (Vega et al., 1998). Therefore, regularly monitoring and component analysis (PCA), factor analysis (FA) and discriminate

Peer review under responsibility of International Research and Training Center on Erosion and Sedimentation and China Water and Power Press.
⁎
Correspondence to: Faculty of Sciences and Techniques, P.O. Box 523.
E-mail address: a.barakat@usms.ma (A. Barakat).

http://dx.doi.org/10.1016/j.iswcr.2016.11.002
Received 18 May 2016; Received in revised form 23 November 2016; Accepted 26 November 2016
Available online 09 December 2016
2095-6339/ © 2016 International Research and Training Center on Erosion and Sedimentation and China Water and Power Press. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

Fig. 1. Oum Er Rbia river basin and water quality stations.

analysis (DA), were used for this kind of studies because they are oil mills, phosphate extraction, livestock farming, and sugar beet
capable to assess temporal and spatial variations in river water quality processing. Because of the deteriorating water quality at the Oum Er
and to identify potential sources of water contamination (i.e. Phung Rbia basin, the Hydraulic Basin Agency of Oum Er Rbia (Ministry of
et al., 2015; Khan et al., 2016; Sharma et al., 2015; Varekar et al., 2015; Energy, Mines, Water and Environment of Morocco), started an
Kumarasamy et al., 2014; Thuong et al., 2013; Razmkhah et al., 2010b; environmental monitoring program for surface water and groundwater
Kazi et al., 2009; Kumar & Dua, 2009; Varol & Sen, 2009; Zhang in the basin since over two decades. From this long-term monitoring of
et al., 2009). For instance, Phung et al. (2015) used CA, PCA, FA and the water physico-chemical and bacteriological parameters, huge and
DA to evaluate temporal/spatial variations of surface water quality in complex amounts of data were generated whose treatment is difficult,
Can Tho City, a Mekong Delta area of Vietnam. Kumarasamy et al. and requires methods of reduction in dimensionality of the data,
(2014) used CA and PCA/FA to investigate the hydrochemistry of a namely statistical methods.
Tamiraparani river basin, Southern India. Khan et al. (2016) applied The objective of the present study is to analyze and interpret a large
CA to hydrochemical data to assess spatial variability in the water data set obtained during a 12-year (2001–2012) monitoring program
quality of Ramganga River and its tributaries (Ganga Basin, India). in Oum Er Rbia River and its tributary Oued El Abid. To achieve this
Correlation analysis and PCA and CA components were used by objective, matrix correlation, PCA and CA multivariate techniques were
Sharma et al. (2015) to study the seasonal variation, identify potential applied to (1) determine the similarities and differences between the
sources of pollution, and clustering of monitoring stations of Ganga sampling stations, (2) evaluate the contribution of water quality
and Yamuna Rivers in Uttarakhand State (India). Other works were parameters to temporal variations in surface water quality, and 3)
conducted using PCA and CA (Simeonov et al., 2003, Bouza-Deano identify the contamination affecting water quality and their potential
et al., 2008, Kazi et al., 2009, Hai et al., 2009, Razmkhah et al., 2010b). sources. The results are expected to help evaluate the spatial–temporal
Furthermore, other studies (Boyacioglu & Boyacioglu, 2007; Zhang evolution of Oum Er Rbia water quality and consequently enable
et al., 2009; Zhou, Liu, & Guo, 2007) applied the PCA and CA methods managers to understand the main sources of pollution at the different
to classify the sampling sites and to identify the latent pollution locations along the river basin.
source.. All above studies showed that the PCA and CA statistical
methods are important tools to determine underlying relationships 2. Material and methods
between the water quality parameters, identify sources of pollution,
and group similar monitoring stations into clusters with similar 2.1. Study area
characteristics. For these reasons, we opted for the use of these two
of these statistical methods, PCA and CA, to study the water quality in Oum Er Rbia River basin is located in the Center-Western of
Oum Er Rbia River. Morocco (Fig. 1), at 31°19.33′−33°22.21″N lat. and
In Morocco, several studies on identifying the sources of contam- 5°8.55′−8°22.53′W Long., and covering an area of 48.070 km2.
ination in many rivers have been conducted over the last decade, Annual rainfall in the Oum Er Rbia basin varies between 1100 mm
showing that their water quality has been deteriorating with various on the Middle-Atlas and 300 mm in the downstream area of the river,
anthropic activities (i.e. Bader et al., 2015; El Baghdadi et al., 2015; with a mean of 550 mm (fUSAID, 2010). It is concentrated from
Hayzoun et al., 2015; Nadem et al., 2015; Mourhir et al., 2014; Barakat October to April with maximum between December and February
et al., 2013; El Morhit et al., 2013; Barakat et al., 2012; Iavazzo et al., (Lionello et al., 2006). It snows on average 20 days/year over 800 m.
2012; Oufline et al., 2012; Derwich et al., 2011; Naoura & Temperature varies between 5 and 50 °C with mean minimum and
Benaabidate, 2011; Alla et al., 2006; Taoufik & Dafir, 2005; Koukal maximum values of 3.5 °C (January) and 38 °C (August) (USAID,
et al., 2004). In Oum Er Bia basin, the vulnerability of water resources 2010). Evaporation can reach 2500 mm/year with a monthly max-
pollution is due to strong agricultural and industrial activity including imum of 300 mm in July and August. Water supplies of the Oum Er

285
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

Rbia basin are valued at 3360 million m3/year, ranging from a high of Table 2
8300 Mm3 and a minimum of 1 300 Mm3. Contributions from many The water quality parameters and their analytical method in this study.
sources associated with those of snowmelt guarantee sustained low
Parameters Analytical method Unit
flows for the Oum-Er-Rbia making it the most consistent of the
Moroccan rivers. Temperature Temperature probe °C
The investigated Oum Er Rbia River covers a length of 550 km from pH Potentiometry/pH probe
Turbidity Turbidimetry NTU
the Middle Atlas at 1800 m, through the Tadla plain and the coastal
Total suspended solids Gravimetric mg/l
plateau, and flows into the Atlantic Ocean at Azemmour city. The main Electrical conductivity Conductometry µs/cm
tributaries of the river are those of the left bank namely Oude Derna, El Ammonia Spectrophotometry mg/L
Handek, Tessaout, Lakhdar and El Abid Rivers (Fig. 1). Many dams Nitrate Spectrophotometry mg/L
and reservoirs have been built on the Oum Er Rbia River to generate Dissolved oxygen Oximeter Mg/l
Total phosphorus Spectrophotometry mg/L
hydroelectric power, to provide water for irrigation, and to supply
Biological oxygen demand BOD meter mg/l
domestic and industrial water. The Oum Er Rbia River and their Chemical oxygen demand COD meter mg/l
tributaries crosses some regions characterized by strong agricultural Fecal coliforms Spread plate methods UFC/100 ml
(large irrigated areas and livestock), industrial (oil mills, phosphate
extraction, and sugar beet processing). In addition, untreated munici-
pal sewage ends up in the Oum Er Rbia River and its tributaries. The respectively. These both sites are considerate relatively non-affected
industrial effluents connected to the wastewater system are negligible. area by domestic pollution. The monitored water quality parameters,
The daily sewage effluents are estimated to be 44.000 m3. The total their units and the methods of analysis are summarized in Table 2. The
daily pollution produced by urban population and transported by temperature (TEMP), pH, electrical conductivity (EC), turbidity and
sewage is approximately 50 t of oxidizable matter, 8.4 t of crude dissolved oxygen (DO) of each water sample were measured in situ. The
protein and 1.25 t of phosphorus material. other parameters, namely total suspended solids (TSS), ammonia
(NH4+), nitrate (NO3–), total phosphorus (TP), biological oxygen
demand (BOD5), chemical oxygen demand (COD), and fecal coliforms
2.2. Sampling and monitored pollutants (F. coli), were analyzed in laboratory using the same analytical
techniques following the standard methods (Rodier et al., 2009). The
As an effort to monitor changes in the Our ErRbia water quality, the obtained data are compared to the Moroccan quality standards
Hydraulic Basin Agency of Oum Er Rbia, Ministry of Energy, Mines, (Table 3) adopted by the Ministry of Energy, Mines, Water and
Water and Environment of Morocco, has installed a program of Environment of Morocco (2002).
measurement of multiple parameters at periodic times in different In order to facilitate consistent evaluation of the dataset of these
monitoring stations along the Oum Er Rbia River and its tributary multiple variables monitored during various periods at different
Oued El Abid. The present study was conducted to analyze and sampling stations, multivariate analyses of the water quality dataset
interpret data obtained from this monitoring program in order to were performed through Pearson's correlation, PCA and CA. All
assess the evolution of water quality of Oum Er Rbia River and its statistical computations in this study were made using Microsoft
tributaries. Water samples were collected from 14 sampling stations Excel 2007 and IBM SPSS 20.
distributed alongside the Oum Er Rbia and El Abid Rivers (Fig. 1, Dataset analysis was performed using the Pearson correlation
Table 1), during different periods over 12 years (2000–2012). The analysis in order to evaluate the relationship between water quality
sampling was done on seasonal basis viz., during winter (December to variables. This minimizes the effect of between-stations correlations
April) and low-water (April to November) seasons. The sampling and between-sampling campaigns relationships. To interpret the
stations were selected on the basis of their contamination vulnerability, coefficients’ correlation, a common source, uniform distribution and
i.e., the stations near to the river source were supposed to be less similar behaviors must be assigned to the correlated variables (Azaza
impaired by urban contamination and the stations downstream to the et al., 2011; Parizi & Samani, 2013). A correlation coefficient near −1
city were supposed to be more impaired by this contamination type. or 1 means a strongest or negative or positive relationship between two
The stations located just upstream and downstream from the urban variables and its value closet to 0 means no linear relationship between
and rural zones are S2, S3, S4, S8 and S10, and just after the confluence them at a significant level of p < 0.05 (Kumar et al., 2006). However, it
of the Oum Er Rbia River with its tributaries are S6, S7, S11 and S12. should be noted that with larger samples, a low strength of correlation,
The rest of the monitoring stations are located on areas vulnerable to for example r < 0.50, can be highly statistically significant at p < 0.01.
the nonpoint source pollution along the river (S13 and S14). S1 and S9 The data were further checked for normality by conducting Kaisere
stations are near the source of Oum Er Rbia and El Abid Rivers, Meyere Olkin (KMO) and Bartlett's Sphericity tests in order to check if
our measured variables can be factorized efficiently. The KMO index
Table 1 compares the values of correlations between variables and those of the
Water quality monitoring stations of the Oum Er Rbia River and its tributaries. partial correlations. When KMO index is close to 1, the PCA of the
Stations Rivers
variables is suitable; but when it is close to 0, the PCA is not relevant.
Generally, this index should be greater than 0.5 for satisfactory factor
S1 Taghat Oum er Rbia analysis. In our analysis, the KMO had a value of 0.551. Bartlett's
S2 Downstream of Khénifra discharge Oum er Rbia Sphericity test was used to check the null hypothesis that the inter-
S3 Tadla dam Oum er Rbia
S4 Downstream of Tadla discharge Oum er Rbia
correlation matrix comes from a population in which the variables are
S5 Kasba Zidania Oum er Rbia uncorrelated (Shrestha & Kazama, 2007). The null hypothesis was
S6 Mechraa Ed Dahk confluence Oum er Rbia rejected at the significance level of 0.05, but in our study, the level value
S7 Oued Day confluence Oum er Rbia is 0 which is small enough to reject the null hypothesis. In fact, the PCA
S8 Downstream of Ould Zidouh discharge Oum er Rbia
was applied efficiently on our data to identify underlying interrelation-
S9 Ait Ouchen El Abid
S10 Ouaouirinth El Abid ship amongst the parameters.
S11 Ouled Sidi Driss Oum er Rbia PCA was applied on reduced standardized data sets to extract
S12 Bssi Bissa Tassaout information about correlation among variables analyzed in the water
S13 Tassaout confluence Oum er Rbia samples (Mahlknecht et al., 2004; Srivastava and Ramanathan, 2008).
S14 Upstream of Sidi Said Mâachou dam Oum er Rbia
Z-scale standardization with mean and variance of zero and one, was

286
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

Table 3
Moroccan surface water guidelines (2002).

Quality classification TEMP pH EC Turbidity TSS DO BOD5 COD NH4+ NO3- TP F. coli
°C (µS/cm) (NTU) (mg/l) (UFC/100 ml)

Excellent < 20 6.5–8.5 100–750 < 15 < 50 >7 <3 < 20 < 0.1 < 10 < 0.1 < 20
Good 20–25 6.5–8.5 750–1300 35–70 50–200 7–5 3–5 20–25 0.1–0.5 10–25 0.1–0.3 20–2000
Moderate polluted 25–30 6.5–9.2 1300–2700 35–70 200–1000 5–3 5–10 25–40 0.5–2 25–50 0.3–0.5 2000–20000
Polluted 30–35 < 6.5 or > 9.2 2700–3000 70–100 1000–2000 3–1 10–25 40–80 2–8 > 50 00.5–3 > 20000
Highly polluted > 35 > 3000 > 100 > 2000 <1 > 25 > 80 >8 >3

used in order to minimize the influence of difference on variance of summer in stations close to the confluence of Oum Er Rbia River with
variables and to adjust for the disparity in the variable sizes and in the their tributaries which are characterized by a summer floods and high
measurements units (Simeonov et al., 2003; Liun et al., 2003). flow regime. This could be explained by the presence in water, besides
Calculation was achieved based on the correlation matrix of measured microorganisms, of colloidal materials such as silt, clay, organic and
parameters, and the PCA scores were obtained from the standardized inorganic material into fine particles (WHO, 2008). The TSS concen-
variables data. The varimax rotation is used to maximize the variance trations, ranging between 1.4 and 37043 mg/l in summer season,
of the squared loadings. The Kaiser's criterion (Kaiser, 1960) was used peaked in confluence areas of tributaries and recipient river (Oum Er
to determine the number of factors to retain. Rbia), and decreased when moving away from these areas. This could
To investigate similarity and dissimilarity in composition between be particularly attributable to runoff waters after storm events
sampling stations (temporal and spatial variability), CA analysis was (Kistemann et al., 2002). Since the TSS was low in stations located
employed on the normalized data by means of the Ward's method, far from confluence areas especially in winter, there were probably
using squared Euclidean distances as a measure of similarity (Massart other factors that contributed to the water turbidity. At winter season,
& Kaufman, 1983; Singh et al., 2005; Zhao et al., 2011). The Ward's the significantly higher mean TSS values recorded at S2, S3, S4 and S5
method says that the proximity between two clusters is the increase in stations compared to downstream stations could be attributed to urban
the squared error. This is the most common method to categorize more wastewater discharges.
accurately groups (Willet, 1987). For all parameters tested, distribu- EC high level is observed in all stations of Oum Er Rbia River, and
tions were centered and reduced prior to clustering. The result is ranged between 314 and 3500 µS/cm. It's due to significant amount of
illustrated by a dendogram presenting the clusters and their proximity dissolved salt in the major springs supplying Oum Er Rbia River. Water
(Forina et al., 2002) with a reduction in dimensionality of the original EC shows a slight decrease moving downstream, despite increasing or
data. decreasing discharge. The mean EC values were 1990 µS/cm at S3 and
1295 µS cm−1 at station S14. At El Abid River, the EC the level is low
3. Results and discussion and ranges between 521 µS cm−1 at S9 and 714 µS cm−1 at S10. The net
increase in conductivity level at station S10 which lies downstream of
3.1. Water quality evaluation El Abid River, especially near the Ouaouirint village, would be linked to
both agricultural and domestic human activity. According the
The summary of the range, mean value, standard deviation of Moroccan guidelines (2002), these measurements also indicated that
measured variables in the river water samples at 14 stations for the the major water samples ranked as moderate to highly polluted class in
summer and winter seasons, and the Moroccan quality standards terms of water quality. EC showed significant positively correlated with
(2002) are provided in Tables 3, 4, respectively. Pearson's correlation COD but negatively correlated with TSS, NO3- and TP.
coefficient (r) gives an idea about the possible relationships between NH4+is a water-soluble gas that exists at low levels (0.1 mg/l) in
biophysical variables. Table 5 presents the values of Pearson's correla- natural waters. NH4+comes from the nitrogen-containing organic
tion coefficient (p < 0.01) for pairs of variables at all sampling stations material and gas exchange between the water and the atmosphere
and campaigns. (Chapman & Kimstach, 1996). It also comes from the biodegradation
A high standard deviation indicates that the data is widely spread, of waste and inputs from domestic, agricultural and industrial, and that
due to the presence of temporal variations caused likely by natural and/ is why it is a good indicator of contamination of water ways. The
or anthropogenic polluting sources. measured NH4+values vary between 0 and 3.59 mg/l in winter and
The water TEMP causes significant environmental impacts between 0 and 2.40 mg/l in summer, and showed a significant positive
(Leynaudn, 1968) by influencing the physical, chemical and biochem- correlation (p < 0.01) with F. coli (r=0.26), turbidity (r=0,147), TSS
ical (WHO, 1987). In the present study, the TEMP values recorded (r=0,103) and BOD5 (r=0,144), and a significant negative correlation
ranged between 9.5 and 39.2 °C. The recorded TEMP showed higher with DO (r=−0,124). NH4+ values remain below the normal rate set by
level in summer, and substantial increase towards downstream. The the Moroccan standards in several stations showed normal values,
measured pH values varied between 6.91 and 9.25, and they are except in S2, S4 and S8 stations that are located downstream of
generally within the recommended range 6.5–8.5 of Moroccan stan- wastewater discharges of Khenifra, Kasba Tadla and Dar Ould Zaidouh
dards, except for few levels significantly higher recorded at stations S6, cities, respectively. In these stations higher values ( > 50% of the
S10, S11 and S12. It was higher in summer compared to winter season. samples) were recorded, indicating moderate to highly polluted river
The lower pH values are recorded at stations close to wastewater water. This showed the negative impact of discharges from cities on the
discharge points (S3, S4). It was negatively correlated with TSS water quality of the studied rivers.
(r=0.14) and turbidity(r=0.20). NO3- concentrations in natural waters are in the order of 1–10 mg/
The turbidity concentration shows a wide ranging between 0 and l. At excessive concentrations, NO3– impacts negatively fauna and flora.
61107 NTU, indicating that the river water had moderately to highly The average concentrations of NO3- in this study ranged from 0 to
polluted quality according to Moroccan standards (over 60% the 24.2 mg/l with around 80% of analysis are lower than 10 mg/l in
samples). There was a positive correlation between turbidity and water summer and around 75% higher than 15 mg/l in winter. The measured
TSS (r=0.77), TP (r=0.16) and NH4+ (r=0.15), but no apparent NO3– concentrations showed significant negative correlation with
correlation between turbidity and F. coli. The highest values were TEMP (r=−0.50) and EC (r=−0.20), and positive correlation with
recorded during the winter season in upstream stations, and during COD (r=0.73). This indicates that the NO3-concentrations are likely

287
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

due to the leaching from agricultural land. All of the NO3- samples were

16284.62 32819.25

20863.64 38090.79

4046.61 13722.15
1006.92 1591.84

1122.31 2003.97

1414.58 4775.79
well below the level of 25 mg/l, qualifying the water quality as excellent

1372.66 6495.6
926.92 1934.37
890.13 3382.46

1081.6 3801.12
643.5 1267.84

413.08 357.76

151.15 135.68

331.19 881.83
(UFC/100 ml)
to good quality class.
DO content is an essential parameter that maintains the equili-
F. coli

brium of aquatic ecosystems. It is commonly used to assess the water

resource quality (Qadir et al., 2002; Sanchez et al., 2006). The river
running water contains about 10 mg/l. Mean concentrations of DO in
study are generally level of below 7 mg/l, but they remain significant
0.28 0.23

0.54 0.58
0.24 0.38
0.21 0.22
0.26 0.37
0.19 0.15

0.22 0.28

0.21 0.46
0.44 0.71
0.11 0.11
0.21 0.3

0.5 1.31

0.5 1.26
0.31 1

with values of 2.80–21.80 mg/l in summer season and 5.92–14.02 mg/

TP

l in winter season, qualifying the water quality as quality good to

excellent. Spatially, lower values were observed at stations downstream
of wastewater discharges from cities (S4 and S8), due to local domestic
11.28 6.14
10.36 7.25
12.62 6.77
8.45 5.81
6.17 6.76
7.92 6.96
6.29 7.04
7.51 7.51
9.11 6.94
5.61 7.13
9.49 8.37
5.67 5.54

8.91 8.11
10.68 7.1
wastewater. In correlation matrix, it was found that DO was negatively
NO3-

correlated to TEMP. As documented by Kumari et al. (2013), , this was

explained by the fact that increasing temperature reduces the dissolu-
tion of ambient DO in river water. Also, Yang et al. (2007) hypothesized
0.05 0.05
0.74 0.34

0.91 1.03
0.11 0.08
0.09 0.12
0.17 0.14

0.09 0.08
0.08 0.09
0.09 0.09

0.11 0.14
0.09 0.08

that the low DO values in summer were linked to the high activities of
0.14 0.1

0.08 0.1
0.4 0.96
NH4+

microorganisms requiring large amounts of oxygen for metabolizing

activities and for organic matters degradation. DO in our study present
also negative correlation with NH4+, indicating that low DO concentra-
17.63 10.86

22.89 15.62

18.44 14.85

20.95 11.56
25.97 13.56
17.61 14.47

23.79 18.74

21.71 14.34
21.57 15.15

tion decreases nitrification rate (ammonia oxidation) (Ruiz et al.,

19.98 7.89

20.6 11.72

29.8 73.08

17.9 13.61

19.9 21.58

2006). Consequently, this explained the positive correlation between

(mg/l)

DO and NO3–.
COD

TP includes organic and inorganic phosphorus. In surface water, it

content is less than 0.1 mg/l. The discharge of domestic and industrial
1.08
2.17
4.45
1.02
1.82
1.12
0.69
2.22

2.01
1.99
1.71
2.12

water and the drainage of agricultural land fertilized contribute to

2.3

0.9
BOD5

increase the concentration. TP shows concentrations of from 0.01 to

1.59
1.98
1.46
3.85
0.84
2.05
1.66
1.26
1.88
1.18
2.08
1.56
2.23
2.04

8.85 mg/l, with more than 50% of the samples show values above the
level of the good quality (0.3 mg/l) in winter season, and less than 14%
of samples monitored in summer season have values above said level.
9.21 0.88

8.39 0.68
6.71 2.06
8.21 0.93
8.99 1.36
8.29 1.03

8.54 0.95
8.37 1.29
8.75 2.05
8.76 2.09
8.43 1.48
8.64 1.05
7.4 1.22
8.37 1

The highest values were recorded in upstream stations (close to

DO

discharge points) in winter and in downstream stations in summer

period, indicating a mixed origin due to a likely effect of leaching and
runoff from fertilized farmland (Groupement ADI/CACG, 2010) and
498.46 1605.94

457.74 2087.47
373.23 1071.65

805.91 2424.85
207.84 643.23

950.65 3637.3
901.7 2528.75

63.95 284.06

effluent from certain industries (e.g. food processing) and domestic

42.44 43.88
40.35 54.25
36.66 56.79

63.44 55.35
86.48 54.95
58.5 77.7

waste. TP correlated reasonably well with turbidity, TSS, BOD5 and F.

coli suggesting that TP originated from natural and anthropogenic
TSS

sources.
BOD5 and COD are two parameters used to quantify the organic
contamination load (Galal-Gorchev et al., 1993). BOD5 in the various
1001.76 4331.88
711.17 1750.21
556.43 1050.12

837.61 1746.09
546.15 1917.67

548.78 1672.39

573.07 1982.24
1213.55 1771.7
105.18 165.76

345.35 758.56
417.11 808.47
280.25 445.17

156.75 541.86

stations shows values between 0.10 and 15 mg/l with nonsignificant

250.1 616.03

variation between seasons. These values indicated that the water

Turbidity

samples ranked as excellent to polluted quality with respect to

(NTU)

Moroccan standards (2002). BOD5showed positive correlation with

TP and F. coli indicating that the wastewater effluent input contribute
to BOD5 in studied river basin. COD is widely used for determining
2276.03 470.41

1692.71 406.71
1673.46 414.63
1590.58 403.48
1646.74 578.86
1569.38 390.93

2155.17 310.46
1079.62 322.31
1543.64 451.09
1413.37 389.27
1530.45 488.47
1646.06 271.36

waste concentration (Kazi et al., 2009). COD shows contents between

548.62 64.12
1690 392.47

0.02 and 149 mg/l, with some variations between seasons, especially
(µS/cm)
Statistical descriptive for the parameters analyzed at all stations.

for stations situated downstream the confluence of El Abid River. Also,

EC

higher values are monitored at stations (S2, S3, S6, S8, S9) located
downstream of wastewater effluents. The high COD values compared to
BOD5 values and the non-correlation between them, indicate that the
8.22 0.26

8.09 0.39

8.04 0.41

8.14 0.21
8.11 0.23
8.01 0.24
8.18 0.22
8.18 0.26
8.03 0.21
8.22 0.25
8.1 0.42

8.2 0.22
8.03 0.4

8.26 0.3

major part of organic material is not biodegradable. All of the above

indicate that the COD could be related to the leaching and transport of
pH

natural, domestic sewage, agricultural and industrial pollutant.

Compared to Moroccan standards, different stations with more than
16.53 3.33

19.74 4.21
19.91 5.06
20.95 4.16
20.67 4.98

22.26 4.07
19.46 5.98

21.87 4.63
22.12 4.23

21.98 4.07
21.8 4.56
18.01 3.6

21.49 4.5

21.15 4.9

50% of samples show a moderately to highly polluted quality, especially

TEMP

in summer.
°C

The main sources of F. coli are untreated wastewater, animal

husbandry, grazing areas, and the application of manure. The presence
of F. coli is an indicator of the bacteriological quality of water. The
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD
MeanSD

monitored concentrations of F. coli ranged between 0 and 120,000

CFU/100 ml, and do not satisfy the Moroccan standards of the surface
water quality in many stations, especially in winter. These stations are
Table 4

S10
S11
S12
S13
S14

located downstream urban liquid discharge points. As also shown in

S1
S2
S3
S4
S5
S6
S7
S8
S9

Table 5, there was a significant positive correlation between F. coli and

288
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

Table 5
Correlation matrix of water quality parameters (Pearson correlation coefficients (r)).

TEMP pH EC Turbidity TSS DO BOD5 COD NH4+ NO3- TP F. coli

T° 1
pH 0.02 1
EC −0.08 −0.06 1
Turb 0.08 −0.20** −0.08 1
TSS 0.01 −0.14** −0.13** 0.77** 1
DO −0.33** 0.00 0.02 −0.03 −0.07 1
BOD5 0.04 0.02 0.04 0.03 0.05 0.02 1
COD −0.67** 0.03 0.27** −0.03 −0.03 0.23** 0.00 1
NH4+ −0.02 −0.05 −0.08 0.15** 0.10** −0.12** 0.14** −0.02 1
NO3– −0.50** −0.01 −0.20** −0.01 0.02 0.17** −0.01 0.73** 0.02 1
TP −0.08 −0.06 −0.21** 0.16** 0.26** 0.04 0.31** 0.03 0.06 0.16** 1
F Col 0.05 −0.04 −0.07 0.09 0.06 −0.08 0.38** −0.10 0.30** 0.02 0.26** 1

**
Correlation is significant at p < 0.01 level (2-tailed)

BOD5 (r=0.38), NH4+(r=0.30) and TP (r=0.26), confirming that they where inorganic nitrogen fertilizers are in common use. The second
have an origin, nevertheless, in common part. factor (PC2) explained 17.80% of the total variance. It had negative
weak loading on pH, and strong positive loading on TSS and turbidity,
3.2. Source identification of monitored variables and could be attributable to the physical and chemical properties of the
water, and to natural weathering of the basin. These three variables
The PCA based on the correlation matrix was performed to under- originated primarily from run-off with high load of solids and wastes
stand the underlying relationships between the water quality variables from point sources of pollution like agricultural fields, and domestic
of all monitoring stations, and to identify their characteristics. The areas (Gazzaz et al., 2012). The studied rivers receive large amount of
Kaiser's criterion (Kaiser, 1960) was used to determine the number of debris due to continuous erosion of soil from the river watersheds and
PCs to retain. The PC loadings were classified as ‘strong’, ‘moderate’, of the Atlas Mountain. The third factor (PC3), explaining 15.43% of the
and ‘weak’ corresponding to absolute loading values of > 0.75, 0.75– total variance, contains a large positive loading on F. coli, TP and
0.50 and 0.50–0.30, respectively (Liu et al., 2003). The scree plot BOD5, and weak positive loading on NH4+. This factor including
graphs are used widely to identify the number of PCs of the basic data organic and nutrient variables may be associated to influences from
structure (Liu et al., 2003). The scree plot (Fig. 2) showed the sorted municipal and industrial point-source discharges, agricultural non-
eigenvalues from large to small as a function of the PC number. This point sources, livestock operations, and/or domestic sources
figure shows a pronounced change of slope after the 4th eigenvalue; (Simeonov et al., 2003). Hence, it can be regarded as the contamination
four components were retained. Loadings of four retained PCs are index of river water. The fourth factor (PC4) accounting for the lowest
expressed in Table 6. total variance (10.31%), has a strong positive loading on EC, weak
In this study, eigenvalues greater than 1 were extracted from the loading on BOD5, and weak negative loading on NO3–, which is
PCs having. Four principal components were obtained with eigenva- indicative of mixed source of contamination comprising of natural
lues > 1 (Kaiser Normalization) that explained almost 63% of the total processes as well as anthropogenic inputs.
variance in the water dataset (Table6). The variables with eigenvalues These PCA analyses identified the potential contamination sources
lower than 1, were removed due to their low significance (Kim and of Oum Er Rbia river water. This contamination is due to a mixed
Mueller 1978). source including natural processes and anthropogenic activities. The
The first factor (PC1), accounting 21.76% of the total variance, discharging untreated urban wastewater into surface water system
showed high positive loadings of COD, NO3-, weak positive loading of constitutes the major point anthropogenic contamination source. The
DO, and high negative loading of TEMP. This factor can be attributed non-point source which also contributes immensely in Oum Er Rbia
to biogenic and anthropogenic (urban wastewater) pollution sources. water contamination is from agricultural activities; especially the river
NO3– may additionally derive from agricultural areas in the region runs through agricultural areas known by intensive culture and live-
stock farming.

3.3. Spatial and temporal variation in river water quality

CA was applied to categorize spatial and temporal similarity group

of all monitoring stations according to their water quality similarities.
The dendrograms of stations obtained by Ward's method in the
summer and winter seasons are shown in Fig. 3.
Spatial CA result showed that all 14 monitoring stations were
classified according to three clusters in summer (Fig. 3) and winter
(Fig. 4) seasons. The stations in each group have similar water
contamination types and greatness.
In summer season, the cluster 1 included the stations S1, S3, S5, S6
and S7 that are located upstream part of the Oum Er Rbia River. The
stations S10 and S14 located at the downstream of El Abid and Oum Er
Rbia Rivers, respectively, showed the similar quality characteristics
with the previous stations. The cluster 2 is composed by the stations S8,
S9, S11, S12, and S13. These two clusters 1 and 2 correspond to
moderate contaminations. The cluster 3 included stations S2 and S4
Fig. 2. Scree plot of the eigenvalues. that are located just downstream of the urban wastewater discharge

289
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

Table 6
Loadings of experimental variables (12) on principal components for the whole datasets.

PCs COD NO3- TEMP DO TSS Turbidity pH F. coli TP BOD5 NH4+ EC eigenvalues % of variance

PC1 0.90 0.82 −0.81 0.43 −0.06 −0.06 0.00 −0.09 0.08 −0.14 0.08 0.19 2.42 20.13
PC2 0.01 −0.04 0.08 −0.01 0.95 0.95 −0.45 −0.04 0.06 −0.02 0.21 0.03 2.05 17.12
PC3 −0.02 0.03 0.02 −0.06 0.15 0.10 0.08 0.76 0.72 0.71 0.46 −0.05 1.85 15.43
PC4 0.15 −0.30 −0.06 0.22 −0.14 −0.10 −0.18 0.03 −0.29 0.34 −0.26 0.87 1.24 10.31

In winter season, the dendrogram shows that all monitoring

locations cluster into three clusters. The cluster 1 corresponding to
relatively high levels of water contamination compared to the others
clusters. This cluster is formed by the uppermost monitoring stations
S2, S3, S4, S5, S7 and S8, located at the Oum Er Rbia river upstream
section. Important anthropogenic sources can be detected in this river
section. This upstream section received urban and rural wastewater
effluents discharged directly into Oum Er Bia River (S2, S3, S4, S8) and
via their tributaries. Also, at this zone, pollutants from leaching of soil
by intensive agricultural and livestock activities constitute the common
non-point contamination sources. The cluster 2 is formed by the
stations S1 and S9 that is situated upstream of Oum Er Rbia and El
Abid River, which consisted to less contaminated areas. Rock and soil
alteration represent the only source of potential contamination. The
cluster 3 composed by the rest stations mainly located downstream of
the Oum Er Rbia river basin, it corresponded to relatively moderate
levels of river contamination. Stations of cluster 3 received pollutants
from domestic wastewater and mostly from non-point sources, such as
animal waste and manure and soil leaching and runoff.

Fig. 3. Dendogram based on hierarchical clustering (wards method) for complete 4. Conclusion
stations in summer season and winter season.

This study was conducted to identify the sources of contamination

on the water quality in the Oum Er Bia River and its tributary Oued El
Abid. Multivariate statistics including Pearson's correlation, PCA and
CA were employed to evaluate spatial and seasonal variations of surface
river water quality data. According to obtained physicochemical an
biological results, some variables at most stations showed average
concentration exceed those recommended by the guide levels allowed
by the Moroccan Directive concerning the quality of river water. This
indicated that the river water is affected from the pollutants in the river
catchment area. The PCA techniques helped in identifying the factors
or sources causing the degradation of water quality. The cause of water
quality variations is mainly related to contamination load from diffuse
(non-point) sources due to weathering of mineral and soil (PC1, PC2
and PC4 factors), and from urban wastewater discharge (PC3). CA
applied to categorize spatial and temporal similarity group of all
monitoring stations according to their water quality similarities was
performed, and revealed the presence of nonpoint and point sources of
contamination and the temporal variations that are controlled by the
precipitation and water runoffs. In summer season, the contamination
is linked to the stations (S2 and S4) located downstream of the urban
wastewater discharge points. These stations were identified as affected
Fig. 4. Dendogram based on hierarchical clustering (wards method) for complete by organic contamination, and corresponded to relatively high levels of
stations in winter season. river contamination. In the stations situated at the mid and down-
stream of Oum Er Rbia River, the water contamination it remains
points of Kasba Tadla and Khenifra Cities, respectively. This cluster moderate due to lack of leaching and runoff. The water quality in
corresponds to highly pollution, compared to the other clusters (1−2), winter is controlled by mixed origin of natural, wastewater discharges
due to the urban wastewater discharge. During the summer, due to the and surface runoff.
lack of precipitation, tributaries become dry and therefore urban We conclude that the management of domestic and industrial
wastewater discharges do not reach the Oum Er Rbia River. Another wastes is required to lower the accumulation of pollutants in water
effect of lack of precipitation is the decrease of the diffuse input of and soil, and to minimize environmental degradation. This should be
contaminants through leaching of watershed soils. This explains why achieved by installing municipal solid waste landfills, and proper
the contamination during the summer is mostly located in the stations treatment of municipal and industrial wastewater before being released
located immediately downstream of the great cities, especially Kasba to the environment, and also the improvements in agricultural
Tadla and Khenifra. practices. The results showed also the importance of multivariate

290
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

statistical assessment of large datasets to get better information about Jarvie, H. P., Whitton, B. A., & Neal, C. (1998). Nitrogen and phosphorus in east coast
British rivers: Speciation, sources and biological significance. Science of the Total
the quality of surface water which can help the environmental Environment, 210–211, 79–109.
managers to make better decisions regarding action plans. Kaiser, H. F. (1960). The application of electronic computers to factor analysis.
Educational and Psychological Measurement, 20, 141–151.
Kazi, T. G., Arain, M. B., Jamali, M. K., Jalbani, N., Afridi, H. I., Sarfraz, R. A. Shah, A. Q.
Acknowledgements (2009). Assessment of water quality of polluted lake using multivariate statistical
techniques: A case study. Ecotoxicology and Environmental Safety, 72(20),
We would like to express our thanks to the Hydraulic Basin Agency 301–309.
Khan, M. Y. A., Gani, K. M., & Chakrapani, G. J. (2016). Assessment of surface water
of Oum Er Rbia, Ministry of Energy, Mines, Water and Environment of quality and its spatial variation.A case study of Ramganga River, Ganga Basin, India.
Morocco, for their cooperation and for providing physicochemical and Arabian Journal of Geosciences, 9(1), 1–9.
biological data. We also thank our colleague from the University of Kistemann, T., Classen, T., Koch, C., Dangendorf, F., Fischeder, R., Gebel, J. Exner, M.
(2002). Microbial load of drinking water reservoir tributaries during extreme rainfall
South Dakota, Abdelali Barakat, for his editing the manuscript.
and runoff. Applied and Environmental Microbiology, 68 (2188-2187).
Koukal, B., Dominik, J., Vignati, D., Arpagaus, P., Santiago, S., Ouddane, B., &
References Benaabidate, L. (2004). Assessment of water quality and toxicity of polluted Rivers
Fez and Sebou in the region of Fez (Morocco). Environmental Pollution, 131(1),
163–172.
Alla, A. A., Mouneyrac, C., Durou, C., Moukrim, A., & Pellerin, J. (2006). Tolerance and Kumar, A., & Dua, A. (2009). Water quality index for assessment of water quality of River
biomarkers as useful tools for assessing environmental quality in the OuedSouss Ravi at Madhopur. Global Journal of Environmental Sciences, 8(1), 49–57.
estuary (Bay of Agadir, Morocco). Comparative Biochemistry and Physiology – Part Kumarasamy, P., James, R. A., Dahms, H. U., Byeon, C. W., & Ramesh, R. (2014).
C: Toxicology & Pharmacology, 143(1), 23–29. Multivariate water quality assessment from the Tamiraparani river basin, Southern
Azaza, F. H., Ketata, M., Bouhlila, R., Gueddari, M., & Riberio, L. (2011). India. Environmental Earth Sciences, 71(5), 2441–2451.
Hydrogeochemical characteristics and assessment of drinking water quality in Zeuss- Kumarin, M., Tripathin, S., Pathakn, V., & Tripathin, B. D. (2013). Chemometric
Koutine aquifer, southeastern Tunisia. Environmental Monitoring Assessment, 174, characterization of river water quality. Environmental Monitoring and Assessment,
283–298. 185(4), 3081–3089.
Bader, A., El Morhit, M., El Blidi, S., Yahyaoui, A., & Fekhaoui, M. (2015). Impact of Leynaudn, G. (1968). Les pollutions thermiques, influence de la température sur la vie
anthropogenic and natural processes on the degradation of water and sediment aquatique. BTI Ministere de l’agriculture, 224–881.
quality of the Mghohga River estuary (northern Morocco). International Journal of Liun, C. W., Linn, K. H., & Kuon, Y. M. (2003). Application of factor analysis in the
Aquatic Science, 6(1), 3–14. assessment of groundwater quality in a black foot disease area in Taiwan. Science of
Barakat, A., El Baghdadi, M., Meddah, R., Rais, J., Nadem, S., & Afdali, M. (2013). The Total Environment, 313(1–3), 77–89.
Evaluation of water quality in open channels flowing through Beni-Mellal City López-Lópezn, J. A., Mendiguchían, C., García-Vargasn, M., & Morenon, C. (2014).
(Morocco). Journal of Water and Land Development, 19, 3–11. Multi-way analysis for decadal pollution trends assessment: The Guadalquivir River
Barakat, A., El Baghdadi, M., Rais, J., & Nadem, S. (2012). Assessment of heavy metal in estuary as a case study. Chemosphere, 111, 47–54.
surface sediments of Day River at Beni-Mellal region, Morocco. Research Journal of Forina, M., Armanino, C., & Raggio, V. (2002). Clustering with dendrograms on
Environmental and Earth Sciences, 4(8), 797–806. interpretation variables. Analytica Chimica Acta, 454, 13–19.
Berzas, J. J., Garcia, L. F., Rodriguez, R. C., & Martin-Alvarez, P. J. (2000). Evolution of Kumar, M., Ramanathan, A., Rao, M. S., & Kumar, B. (2006). Identification and
the water quality of a managed natural wetland: Tablas de Daimiel National Park. evaluation of hydrogeochemical processes in the groundwater environment of Delhi,
Water Research, 34, 3161–3170. India. Environmental Geology, 50, 1025–1039.
Bouza-Deaño, R., Ternero-Rodríguez, M., & Fernández-Espinosa, A. J. (2008). Trend Lionello, P., Malanotte, P., & Boscolo, R. (2006). Mediterranean climate variability
study and assessment of surface water quality in the Ebro River (Spain). Journal of Amsterdam: Elsevier.
Hydrology, 361, 227–239. Massart, D. L., & Kaufman, L. (1983). The interpretation of analytical chemical data by
Boyacioglu, H., & Boyacioglu, H. (2007). Water pollution sources assessment by the use of cluster analysis New York: John Wiley Sons.
multivariate statistical methods in the Tahtali Basin, Turkey. Environmental Mahlknecht, J., Steinich, B., & De León, I. N. (2004). Groundwater chemistry and mass
Geology, 54, 275–282. transfers in the Independence aquifer, central Mexico, by using multivariate
Carpenter, S., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N., & Smith, V. statistics and mass-balance models. Environmental Geology, 45(6), 781–795.
H. (1998). Nonpoint pollution of surface waters with phosphorus and nitrogen. Mendiguchı́a, C., Moreno, C., Galindo-Riaño, M. D., & Garcı́a-Vargas, M. (2004). Using
Ecological Applications, 8(3), 59–568. chemometric tools to assess anthropogenic effects in river water: A case study:
Chapman, D., & Kimstach, V. (1996). Selection of water quality variables. , in: Chapman Guadalquivir River (Spain). Analytica Chimica Acta, 515(1), 143–149.
(Ed.). (1996). Water quality assessments: A guide to the use of biota, sediments and Moroccan Standards for surface water Quality (2002). Official Bulletin No 5062. http://
water in environment monitoring (2188-2187). 2nd ed. London: E FN Spon, www.sgg.gov.ma/BO/bulletin/Fr/2002/BO_5062_fr.PDF. Accessed 01.01.16
59–126. Mourhir, A., Rachidi, T., & Karim, M. (2014). River water quality index for Morocco
Derwich, E., Benziane, Z., & Benaabidate, L. (2011). Diagnostic of physicochemical and using a fuzzy inference system. Environmental Systems Research, 3(1), 1–12.
bacteriological quality of fez wastewaters rejected in Sebou River: Morocco. Nadem, S., El Baghdadi, M., Rais, J., & Barakat, A. (2015). Evaluation de la
Environmental Earth Sciences, 63, 839–846. contamination en métaux lourds des sédiments de l′estuaire de Bou Regreg (Côte
El Baghdadi, M., Oumeskou, H., Barakat, A., Nadem, S., & Rais, J. (2015). Effet de la atlantique, Maroc). Journal of Materials and Environmental Science, 6(11),
Décharge publique de la ville de Béni-Mellal sur les Sédiments et les Sols au niveau 3338–3345.
d′Oued Sabeq. Journal of Materials and Environmental Science, 6(11), 3371–3381. Naoura, J., & Benaabidate, L. (2011). Monitoring of heavy metals in the sediments of the
El Morhit, M., Fekhaoui, M., Serghini, A., El Blidi, S., El Abidi, A., & Yahyaoui, A. (2013). Inaouene River, Morocco. Journal of the Black Sea / Mediterranean Environment,
Typology of water quality in the Loukkos river estuary (Morocco). LARHYSS 17(3), 193–202.
Journal, 12, 1112–3680. Osman, R., Saim, N., Juahir, H., & Abdullah, M. (2012). Chemometric application in
Galal-Gorchev, H., Ozolins, G., & Bonnefoy, X. (1993). Revision of the WHO guidelines identifying sources of organic contaminants in Langat river basin. Environmental
for drinking water quality. Annalidell'IstitutoSuperiore di Sanità, 29, 335–345. Monitoring and Assessment, 184(2), 1001–1014.
Garafa, R., Faggiano, L., Real, M., Munné, A., Ginebreda, A., Guasch, H. von der Ohe, P. Oufline, R., Hakkou, R., Hanich, L., & Boularbah, A. (2012). Impact of human activities
C. (2011). Water toxicity assessment and spatial pollution patterns identification in a on the physico-chemical quality of surface water and groundwater in the north of
Mediterranean River Basin district. Tools for water management and risk analysis. Marrakech (Morocco). Environmental Technology, 33(18),
Science of the Total Environment, 409(20), 4269–4279. 2077–2088 (doi: 101080/095933302012660644).
Gazzaz, N. M., Yusoff, M. K., Ramli, M. F., Aris, A. Z., & Juahir, H. (2012). Parizi, H. S., & Samani, N. (2013). Geochemical evolution and quality assessment of
Characterization of spatial patterns in river water quality using chemometric pattern water resources in the Sarcheshmeh copper mine area (Iran) using multivariate
recognition techniques. Marine Pollution Bulletin, 64(4), 688–698. statistical techniques. Environmental Earth Sciences, 69, 1699–1718.
Groupement CACG/ADI (2010). Etude d′actualisation du PDAIRE de la zone d′action de Perrin, J. L., Raïs, N., Chahinian, N., Moulin, P., & Ijjaali, M. (2014). Water quality
l′Agence du Bassin Hydraulique de l′Oum Er Rbia assessment of highly polluted rivers in a semi-arid Mediterranean zone Oued Fez and
Hai, X. U., Lin-Zhang, Y. A. N. G., Geng-Mao, Z. H. A. O., Jia-Guo, J. I. A. O., Shi-Xue, Y. Sebou River (Morocco). Journal of Hydrology, 510, 26–34.
I. N., & Zhao-Pu, L. I. U. (2009). Anthropogenic impact on surface water quality in Phung, D., Huang, C., Rutherford, S., Dwirahmadi, F., Chu, C., Wang, X. Dinh, T. A.
Taihu Lake region, China. Pedosphere, 19(6), 765–778. (2015). Temporal and spatial assessment of river surface water quality using
Hayzoun, H., Garnier, C., Durrieu, G., Lenoble, V., Le Poupon, C., Angeletti, B. Mounier, multivariate statistical techniques: A study in Can Tho City, a Mekong Delta area,
S. (2015). Organic carbon, and major and trace element dynamic and fate in a large Vietnam. Environmental Monitoring and Assessment, 187(5), 1–13.
river subjected to poorly-regulated urban and industrial pressures (Sebou river, Qadir, A., Malik, R. N., & Husain, S. Z. (2008). Spatio-temporal variations in water
Morocco). Science of the Total Environment, 502, 296–308. quality of NullahAik-tributary of the river Chenab, Pakistan. Environmental
Iavazzo, P., Ducci, D., Adamo, P., Trifuoggi, M., Migliozzi, A., & Boni, M. (2012). Impact Monitoring and Assessment, 140, 43–59.
of past mining activity on the quality of water and soil in the High Moulouya Valley Razmkhah, H., Abrishamchi, A., & Torkian, A. (2010). Evaluation of spatial and temporal
(Morocco). Water Air Soil Pollution, 223(2), 573–589. variation in water quality by pattern recognition techniques: A case study on Jajrood
Idris, A. M. (2008). Combining multivariate analysis and geochemical approaches for River (Tehran, Iran). Journal of Environmental Management, 91, 852–860.
assessing heavy metal level in sediments from Sudanese harbors along the Red Sea Rodier, J., Legube, B., & Merlet, N. (2009). L′Analyse de l′Eau Dunod, Paris
coast. Microchemical Journal, 90(2), 159–163. Rudolph, A., Ahumada, R., & Perez, C. (2002). Dissolved oxygen content as an index of

291
A. Barakat et al. International Soil and Water Conservation Research 4 (2016) 284–292

water quality in San Vicente Bay, Chile” (36◦45_s). Environmental Monitoring and barriers to trade and investment
Assessment, 78, 89–100. Varekar, V., Karmakar, S., J. R., & Ghosh, N. C. (2015). Design of sampling locations for
Ruiz, G., Jeison, D., Rubilar, O., Ciudad, G., & Chamy, R. (2006). Nitrification– river water quality monitoring considering seasonal variation of point and diffuse
denitrification via nitrite accumulation for nitrogen removal from wastewaters. pollution loads. Environmental Monitoring and Assessment, 187(6), 1–26.
Bioresource Technology, 97(2), 330–335. Varol, M., & Sen, B. (2009). Assessment of surface water quality using multivariate
Sanchez, E., Colmenarejo, M. F., Vicente, J., Rubio, A., García, M. G., Travieso, L., & statistical techniques: A case study of Behrimaz Stream, Turkey. Environmental
Borja, R. (2006). Use of the water quality index and dissolved oxygen deficit as Monitoring and Assessment, 159, 543–553.
simple indicators of watersheds pollution. Ecological Indicators, 7(2), 315–328. Vega, M., Pardo, R., Barrado, E., & Deban, L. (1998). Assessment of seasonal and
Sharma, M., Kansal, A., Jain, S., & Sharma, P. (2015). Application of multivariate polluting effects on the quality of river water by exploratory data analysis. Water
statistical techniques in determining the spatial temporal water quality variation of Research, 32(12), 3581–3592.
Ganga and Yamuna Rivers present in Uttarakhand State, India. Water Quality, Wang, Y. B., Liu, C. W., Liao, P. Y., & Lee, J. J. (2014). Spatial pattern assessment of river
Exposure and Health, 7(4), 567–581. water quality: Implications of reducing the number of monitoring stations and
Shrestha, S., & Kazama, F. (2007). Assessment of surface water quality using multivariate chemical parameters. Environmental Monitoring and Assessment, 186(3),
statistical techniques: A case study of the Fuji river basin, Japan. Environmental 1781–1792.
Modelling & Software, 22(4), 464–475. WHO (World Health Organization) (2008). Guidelines to drinking water quality (3rd
Simeonov, V., Stratis, J. A., Samara, C., Zachariadisb, G., Voutsac, D., Anthemidis, A. ed.) Geneva: World Health, Organization, 1–666.
Kouimtzisc, T. (2003). Assessment of the surface water quality in Northern Greece. Willet, P. (1987). Similarity and clustering in chemical information systems New York:
Water Research, 37, 4119–4124. Research Studies Press, Wiley.
Singh, K. P., Malik, A., & Sinha, S. (2005). Water quality assessment and apportionment Yang, H., Shen, Z., Zhang, J., & Wang, W. (2007). Water quality characteristics along the
of pollution sources of Gomti river (India) using multivariate statistical techniques - course of the Huangpu river. China) Journal of Environmental Sciences, 19(10),
A case study. Analytica Chimica Acta, 538(1–2), 355–374. 1193–1198.
Srivastava, S. K., & Ramanathan, A. L. (2008). Geochemical assessment of groundwater Zhang, Q., Li, Z., Zeng, G., Li, J., Fang, Y., Yuan, Q., & Ye, F. (2009). Assessment of
quality in vicinity of Bhalswa landfill, Delhi, India, using graphical and multivariate surface water quality using multivariate statistical techniques in red soil hilly region:
statistical methods. Environmental Geology, 53, 1509–1528. A case study of Xiangjiang watershed, China. Environmental Monitoring and
Taoufik, M., & Dafir, J. E. (2005). A study of water quality in lower OumRabiaa Assessment, 152(1–4), 123–131.
BasinWater Quality Research Journal of Canada, 40 Morocco: Speciation of Zhao, J., Fu, G., Lei, K., & Li, Y. (2011). Multivariate analysis of surface water quality in
nitrogen and phosphorus, 120–129. the Three Gorges area of China and implications for water management. Journal of
Thuong, N. T., Yoneda, M., & Matsui, Y. (2013). Does embankment improve quality of a Environmental Sciences, 23(9), 1460–1471.
river? A case study in To Lich River inner city Hanoi, with special reference to heavy Zhou, F., Liu, Y., & Guo, H. (2007). Application of multivariate statistical methods to
metals. Journal of Environmental Protection, 4(4), 361–370. water quality assessment of the water courses in north western new territories Hong
USAID (United States Agency for International Development). Moulouya and OumEr- Kong. Environmental Monitoring and Assessment, 132,
Rbia watersheds: organizational it and management system needs. Submitted to 1–13 (doi: 101007/s10661-006-9497-x).
USAID/Morocco, Economic Growth Office — Assistance Objective 3: Reduced

292