Arab J Geosci (2017) 10:143 DOI 10.1007/s12517-017-2877-4

ORIGINAL PAPER

Assessment of variations in water quality using statistical techniques: a case study of Işıklı Lake, Çivril/Denizli, Turkey Fatma Aksever 1 & Seher Büyükşahin 1

Received: 7 April 2016 / Accepted: 26 January 2017 # Saudi Society for Geosciences 2017

Abstract A study of the hydrochemical evaluation of waters in the Işıklı Lake and surrounding area was carried out with the objective of identifying the geochemical processes and their relation with water quality in the region. The multivariate statistical techniques were used in the hydrochemical evaluation of waters. Statistical analysis of water quality parameters was made to seeing the interrelationship between different variables in order to explain the water quality and pollution status of study area. For this purpose, water samples were taken from lake, river, stream, and springs which are represented by investigated area and water qualities were evaluated. Generally, Ca2+, Mg2+, and Cl−, HCO3− ions are dominant within surface water and water sources. Arsenic concentration increase is determined in Işıklı spring and Kufi stream water samples. Also, aluminum concentration is high level in the Kufi stream water samples. This increase was related to igneous rocks as geogenic origin. Also, geogenic contamination was identified in R-mode factor and cluster analyses. There is high correlation between electrical conductivity and major ions of waters.

Keywords Işıklı Lake . Water sources . Water quality . Multivariate statistical techniques

* Fatma Aksever [email protected] Seher Büyükşahin [email protected] 1

Department of Geological Engineering, Süleyman Demirel University, Isparta, Turkey

Introduction Surface water systems are seen as natural receiving media all over the world. Therefore, the surface water quality is influenced by many factors (Hem 1985; Feller 2007; Hussain et al. 2008; Raymond et al. 2008). Nowadays, treatment of contaminated surface water is increasing under the natural and anthropogenic conditions. Human activities such as the discharge of industrial and domestic effluents, the use of agricultural chemicals, land use, and cover changes are the anthropogenic factors that influence surface water quality. The natural processes are geology, climate change, erosion, and weathering of crustal materials degrade surface waters (Carpenter et al. 1998; Jarvie et al. 1998; Peters and Meybeck 2000; Buck et al. 2004; Alam et al. 2006; Zhang et al. 2009; Hussain et al. 2008). It is important to control and to determine the surface water pollution due to the surface waters that play a major role in water cycle. The many available graphical and statistical methodologies are used to interpret the water quality. The use of graphical techniques and all the methods are compared with similarity and difference of water data. So, water type, quality, and pollution parameters are classified. In the recent years, there are many studies in the literature on these subjects. Multivariate statistical methods were used for assessment and classification of water chemistry data (Kazi et al. 2009; Yidana 2010; Güler et al. 2012; Raiber et al. 2012; Zhang et al. 2012; Salem et al. 2015; Alamgir et al. 2016). Işıklı Lake is a fresh water lake and it is located in the Dinar-Çivril graben system from the western part of Turkey. The maximum water depth of the lake is approximately 7.0 m, water area of 365.85 km, and its surface area of 65 km2 (Ismael 2009). Işıklı Lake has an open lake system; the lake is fed at the origin by the Kufi stream from the northeast and the Dinarsuyu stream from the southeast. However, the lake is

143

Page 2 of 17

discharged to the outside via the Büyük Menderes River in the southwest (Fig. 1). Kufi stream originates from the Küçük Sincanlı plain and it arises as the Mahman stream. Karadirek creek and Hamam creek flow through the Sandıklı plain and to Mahman, stream is added (Fig. 1). Kufi stream has approximately 36 km length and 1–1.5 km, and it is located in the Çivril plain (Ceylan and Eskikurt 2001). The amount of water transferred to the Işıklı Lake is 60% from the Kufi stream in March (http://www.turkiyesulakalanlari.com/isikli-civrilgolu-denizli/) and 45.7% from the Dinarsuyu stream in May, respectively (Ceylan 1998). Büyük Menderes River basin is the largest basin of the Aegean region in Turkey. The Büyük Menderes River, a typical meandering river, is the dominant water body of the basin. The total length is approximately 530 km. Its average annual flow volume is about 3.020 billion m3. The average annual discharge is 110 m3/s. The surface area of the river is 200 km2. Its sea level height changes between 50 and 70 m (Koç 2010). A dam was constructed on the Büyük Menderes River bed in 1968 for irrigation. In this way, lake water is to be controlled. 72. 300 km2 of agricultural area is irrigated with lake water. Also, lake is recharged by the springs (Işıklı, Yuva, Gökgöl, Bektaş). Işıklı spring is the most efficient water source for

Fig. 1 Location map of investigation area

Arab J Geosci (2017) 10:143

lake. The spring is connected to the Kufi stream with drainage channel in 1963. The amount of water is 22.9% incoming from the lake. In addition to, the amount of water supplied to the lake from Yuva and Bektaş springs are 6.8 and 1%, respectively (Ceylan 1998). The main objective of this study was to investigate the water quality in the Işıklı Lake, surface waters, and springs feeding the lake and to define a hydrochemical evaluation of the waters with many of the available graphical and statistical methodologies.

Material and methods Sampling and chemical analysis A total of ten location water samples were collected from lake, river, stream, and springs. Coordinate and elevations of water sampling points were recorded with a Global Positioning System (GPS) unit, portable Magellan Triton 500. Geological–hydrogeological location map of the investigation area (Fig. 2) was prepared. The main physicochemical parameters, including water temperature (T), redox potential (Eh), total dissolved solids (TDS), pH, and electrical conductivity

Arab J Geosci (2017) 10:143

Page 3 of 17 143

Fig. 2 Geological and hydrogeological map of the investigation area (modified from Öztürk 1981; Konak 2002; Konak and Senel 2002; Şenel 2002; Turan 2002; Gürbüz et al. 2011) and locations of the water points

(EC), were determined in situ with a portable instrument (multiparameter YSI (US) 556–01 MPS). The water samples were collected, transported in clean polyethylene bottles (100 ml), and dispatched for analysis to the laboratory in an ice-filled box. All the samples were transported to the laboratory within 24 h of collection and stored in a refrigerator at 4 °C for further analyses. Major ions (K+, Na+, Ca2+, Mg2+, Cl−, SO42−, HCO3−, and CO32−) were measured in the laboratory of the Süleyman Demirel University, Research and Application Center for Geothermal Energy, Groundwater and Mineral Resources. Na+, Ca2+, Mg2+, K+, Cl−, SO42−, NO3, and NO2 concentrations were measured with Ion Chromatography (ICDionex ICS-3000). NH4 concentration was determined with reagent kit (Spectroquant Merck 14752) by Spectroquant Merck Nov. 60. HCO3− and CO32− concentrations were determined by reagent kit (Aquamerck Test Alcalinite Merck 11,109). Heavy metal concentrations were determined by Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) in the Bureau Veritas Minerals Laboratories (BVML)/ (Canada). AquaChem 3.70 software was used for plotting the Piper (1944), Pie, Schoeller (1955, 1962) semi-logarithmic, Schoeller, Wilcox (1955) and U.S. Salinity Laboratory (Richards 1954) diagrams.

Statistical methods The multivariate statistical techniques such as cluster analysis (CA), principal component analysis (PCA), and factor analysis (FA) have widely been used in analysis of water-quality data. (Vega et al. 1998; Perona et al. 1999; Belkhiri et al. 2010; Garizi et al. 2011; Palma et al. 2010; Oliva González et al. 2011; Zhang et al. 2012; Wu and Kuo 2012; Varol et al. 2012; Zhao et al. 2012; Arslan 2013; Oketola et al. 2013; Park et al. 2014; Marinović and Ruždjak 2015). Cluster analysis is an unsupervised pattern recognition technique that uncovers intrinsic structure or underlying behavior of a data set without making a priori assumption about the data, in order to classify the objects of the system into categories or clusters based on their nearness or similarity (Vega et al. 1998). Recently, PCA has been widely used to evaluate a variety of environmental issues (Facchinelli et al. 2001). PCA provides information on the most meaningful parameters which describe the whole data set interpretation and data reduction and summarizes the statistical correlation among constituent in the water with minimum loss of original information (Helena et al. 2000). Factor analysis, which includes PCA is a very powerful technique applied to reduce the dimensionality of a data set

143

Arab J Geosci (2017) 10:143

Page 4 of 17

Results and discussion

The study area has different aquifer systems that are widely spread around the lake. In particular, alluvium constitutes significant groundwater reservoirs. Quaternary alluvium is porous aquifer system and is the main drinking water resource for the region. Alluvial fan is potentially good reservoirs for groundwater, and it is terrestrial aquifer system. The Hamamçay formation is a productive aquifer and it is common aquifer system with highly permeability. Homa mélange is a fractured rock aquifer system. The groundwater is hosted in the fractures and within the voids of the mélange unit. Oligocene conglomerate and Kestel formation is impermeable units, and the units are defined as aquifuge mediums. Hydrologically, the investigation area is drained by two main hydrographic units, respectively, the Büyük Menderes River watershed that flows to the southwest, and the Dinarsuyu stream watershed to the southeast. Theoretical analyses of groundwater flow patterns under varying hydrogeologic conditions provide an important basis for evaluating chemical interactions between groundwater and streams (Hill 1990). So, it is important to know the groundwater flow direction of the region for evaluating chemical of water. The groundwater flow paths of the region along the alluvial body are generally towards the Büyük Menderes River. Also, groundwater flow direction is parallel to the Dinarsuyu stream and from southeast to Işıklı Lake (Fig. 2, FugroSİAL 2014).

Geology and hydrogeology

Hydrogeochemistry

The basement of the area comprises Precambrian age metamorphic rocks. The metamorphic bedrock as Kestel formation is composed of quartz, sericite schist, albite, quartzite, calcschist, phyllite, and metabasalt. Kestel formation is tectonically overlain by ophiolitic mélange and allocthonous units. The highlands in the study area are represented by Homa mélange of Cretaceous age. Mélange composed of serpentinized ultramafic rocks together with limestone, chert, and radiolarite blocks. Oligocene conglomerate is located on the mélange. The Pliocene Hamamçay formation is situated on Oligocene conglomerate. The Hamamçay formation is intercalated gravel, clay, sand, silt, sandstone, claystone, and loosely attached conglomerate. The alluvial fan and alluvium are the youngest sediments around the investigation area. The alluvial fan consists of lacustrine sediments of sandstone, claystone, mudstone, and conglomerate. Quaternary alluvium deposits composed of uncemented clay, sand, silt, and gravel levels that overlie above another units. The Alluvium deposits are especially well developed along the Büyük Menderes River, the Dinarsuyu stream, and the flat areas around the Işıklı Lake (Öngür 1973; Öztürk 1981; Öztaş 1989; Çakmakoğlu 1986; Afşin 1991; Gürbüz et al. 2011). Geological map and stratigraphic columnar section of the investigated area is shown in Figs. 2 and 3.

The water samples were collected from ten locations (spring, stream, river, and lake water) in March 2014. Sampled locations were homogeneously distributed over the study area. The water samples are Dinarsuyu stream (2), Büyük Menderes River (1), Kufi stream (3), Işıklı spring (1), Yuva spring (1), Gümüşsu spring (1), and Işıklı Lake (1). All sample locations are shown in sampling location map (Fig. 2). Geochemical analyses were performed to determine physical and chemical properties of waters in the study area. pH, temperature, redox potential, electrical conductivity, and total dissolved solids of water samples were measured in situ. Also, the analysis results of 26 parameters, including chemical, physical, and pollution parameters (nutrient compounds and heavy metals) were determined in ten water samples taken from a study area, and their results were presented statistically in Table 1. The pH significantly affects the treatment and use of water. Also, the pH of water is a measure of its reactive characteristics (Todd and Mays 2005). The water samples of the investigation area have pH values ranged from 7.20 to 8.70 with a mean value of 8.15 which indicates that the waters in the study area are alkaline (pH >7). The temperature ranged from 8.60 to 15.90 °C with a mean value of 10.53 °C. Total dissolved solids (TDS) is a measure

consisting of a large number of inter-related variables, while retaining as much as possible the variability present in data set. This reduction is achieved by transforming the data set into a new set of variables, the principal components (PCs), which are orthogonal (non-correlated) and are arranged in decreasing order of importance. Mathematically, the PCs are computed from covariance or other cross-product matrix, which describes the dispersion of the multiple measured parameters to obtain eigenvalues and eigenvectors. Principal components are the linear combinations of the original variables and the eigenvectors (Wunderlin et al. 2001). Chemical analysis results of waters in the investigation area are evaluated statistically. All the statistical computations were carried out using the statistical software SPSS 15.0 (Statistical Package for the Social Sciences). In this statistical evaluation, multivariate statistical techniques, such as cluster analysis (CA), principal component analysis (PCA), and factor analysis (FA) were applied to the data set on water quality of the Işıklı Lake and surrounding area, at ten different locations for 26 parameters. Basic statistics of data set on water quality is summarized in Table 1. The results of obtained statistical analysis were evaluated by R-mode analysis, scree plot and dendrogram based on Ward (Joe and Ward 1963) method.

Arab J Geosci (2017) 10:143 Table 1 Summary statistics of physical, chemical, and pollution parameters of waters

Page 5 of 17 143

Variable

Units

Minimum

Maximum

Mean

Standard deviation

WHO (2011)

7.20 8.60 −97.40 230.00

8.70 15.90 −16.00 860.00

8.15 10.53 −67.30 475.00

0.49 1.99 26.86 229.79

6.5–8.8

TDS mg/l Major elements Na+ mg/l Ca2+ mg/l K+ mg/l Mg2+ mg/l Cl− mg/l 2− SO4 mg/l HCO3− mg/l CO32− mg/l Nutrient parameters NO3− mg/l − NO2 mg/l NH4+ mg/l Trace elements Al μg/l As μg/l Ba μg/l

117.00

454.00

246.60

123.02

600–1000

1.68 36.48 0.37 4.19 42.00 1.96 91.53 0.00

75.04 109.76 8.21 62.00 246.00 122.54 457.65 48.00

25.71 68.78 4.29 25.93 185.00 38.75 269.71 16.80

27.55 22.78 2.58 20.02 99.87 45.33 127.50 18.50

200

1.43 0.01 0.00

9.31 0.01 0.00

5.31 0.01 0.00

3.09 0.01 0.00

2.00 1.10 6.55

77.00 30.50 150.30

60.10 10.39 72.89

147.34 9.05 43.19

100 10 700

μg/l μg/l μg/l μg/l μg/l μg/l μg/l

0.05 0.90 0.50 10.00 0.12 0.10 0.50

0.18 3.90 4.50 214.00 70.35 1.50 13.30

0.07 2.21 1.98 61.80 15.50 0.25 3.07

0.05 1.15 1.42 126.09 20.92 0.44 3.89

3 50 2000 300 400 10 3000

In situ measurements pH o T C Eh mV EC μS/cm

Cd Cr Cu Fe Mn Pb Zn

of the total amount of minerals dissolved in water and a very useful parameter in the evaluation of water quality. Water containing less than 500 mg/L is preferred for domestic use and for industrial processes (Todd and Mays 2005). TDS values of waters were measured in situ, and these values vary from 117 to 454 mg/L with a mean value of 246.60 mg/L (Table 1). The dissolved solid concentration (0–1000 mg/L) of water is classified as Bfresh water^ by Freeze and Cherry (1979). The redox potential (Eh) value of samples ranged from −97.4 to 16.0 mV. The electrical conductivity (EC) values ranged from 230 to 860 μS/cm at 25 °C. Waters hardness values were found in the range of 7.05–23.95 °F. Hydrochemical facies The hydrogeochemical facies of the waters in the investigation area were determined on the basis of the results of the chemical analyses. The order of abundance of the major cations in the water samples is Ca 2 + > Mg 2 + > N a + > K + and m ajor anions are

250 500

50 3 −

HCO3− > Cl− > SO42− > CO32−. Chemical composition of Işıklı, Yuva, and Gümüşsu spring samples in the study area have not exceeded the desirable limit for drinking water (WHO 2011). The major cations and anions for the analyzed water were plotted on a piper diagram. Hydrochemical diversity among the samples is illustrated in the piper diagram (Piper 1944) of Fig. 4. According to the diagram; Dinarsuyu stream has Ca–HCO3–Cl (1 and 2), Büyük Menderes River has Ca–Mg–Cl–HCO3 (3), Kufi stream has Na–Mg–Ca– HCO3–Cl (4 and 5) and Ca–Cl (6), Işıklı spring has Ca– HCO3–Cl (7), Yuva spring has Ca-HCO3–Cl (8), Gümüşsu spring has Ca–HCO3–Cl (9), and Işıklı Lake has Na–Mg– Ca–HCO3–Cl (10) facies (Fig. 4). Increasing chlorine concentration of spring waters may come from chlorapatite, mica, and feldspathoids group minerals. The mineral group is located in igneous rocks within Homa mélange in the study area. Also, increasing chlorine concentration of surface waters is an anthropogenic origin that is especially use as fertilizer for

143

Page 6 of 17

Arab J Geosci (2017) 10:143

as anions are dominant position in water samples (Fig. 5). The Schoeller semi-logarithmic diagram (Schoeller 1955, 1962, Fig. 6) allows the total concentration of major ions in logscale. In the diagram, water samples give similar peaks. The majority of water samples show Mg2+, Ca2+, Cl− and HCO3− concentrations are dominant ions (>7 meq/L) based on Schoeller semi-logarithmic diagram. Drinking water quality The analytical results were evaluated to ascertain the suitability of water of the study area for drinking use. Firstly, total dissolved solid values of waters are evaluated for drinking water quality. According to WHO (2011) specification TDS up to 600 mg/L is the highest desirable and up to 1000 mg/L is maximum permissible (Table 1). The TDS values range from 117 to 454 mg/L, and waters are suitable for WHO (2011) international drinking water standards. Also, pH values ranged from 7.20 to 8.70, and the waters are usable for allowable limit values (6.5–8.8) of WHO (2011) standards. Besides, the waters in the investigation area were evaluated with the Schoeller diagram. The waters are suitable in terms of drinking water, and the waters are classified as Bvery good and good quality waters^. Işıklı, Yuva, Gümüşsu springs, Büyük Menderes River, Dinarsuyu stream (1 and 2) and Işıklı lake samples are determined as Bvery good quality waters,^ and Kufi stream samples (4, 5, and 6) are identified as Bgood quality waters.^ These surface water samples have the high electrical conductivity and hardness values (Fig. 7). Fig. 3 Stratigraphic columnar section of the investigation area (not to scale)

agricultural purposes. Pie charts was plotted in order to see the interpretations regarding ionic concentrations (in mg/L) and the distribution of the major elements in waters. The chart shows that Ca2+, Mg2+ ions as cations and Cl−, HCO3− ions

Fig. 4 Chemical facies of water samples in Piper (1944) diagram

Irrigation water quality USA Salinity (Richards 1954) diagram is used to determine salinity and alkalinity hazard of irrigation water. The SAR vs EC values for groundwater samples of the study area were plotted in the USA Salinity diagram of irrigation water. Based on the diagram, the water quality shows that majority of the samples fall in the “C1S1,” “C2S1,” and “C3S1” (Fig. 8). Işıklı (7), Yuva (8), and Gümüşsu (9) spring water samples fall in the field of BC1S1,^ and the irrigation waters are suitable for most plants. Dinarsuyu stream (1 and 2), Büyük Menderes River (3), Kufi stream (6), and Işıklı Lake (10) water samples fall in the field of BC2S1^ and indicating medium salinity and low sodium water, which can be used for irrigation on all types of soil without danger of exchangeable sodium. Kufi stream water samples (4 and 5) fall in the field of BC3S1^ and indicating water of medium-high salinity and low sodium, which can be used for irrigation in almost all types of soil with little danger of exchangeable sodium (Fig. 8). Sodium concentration plays an important role in evaluating the groundwater quality for irrigation because sodium causes an increase in the hardness of soil as well as a reduction in its permeability (Tijani 1994). The Wilcox (1955) diagram relating sodium percentage (%Na-meq/L) and electrical conductivity was used to determine the suitability of irrigation water. The diagram shows that 80% of the total samples fall in the field of Bexcellent to good,^ and 20% of the groundwater

Arab J Geosci (2017) 10:143

Page 7 of 17 143

Fig. 5 Pie charts of waters, 1, 2 Dinarsuyu stream, 2 Büyük Menderes River, 3, 4, 5 Kufi stream, 6 Işıklı spring, 7 Yuva spring, 8 Gümüşsu spring, 10 Işıklı Lake

samples fall in the field of Bgood to permissible^ for irrigation (Fig. 9). Sources of water pollution Nitrogen compounds contamination Study area is one of the intensive farming regions which artificial and natural fertilizers are used extensively in agricultural activities. Generally, the fertilizers are the origin of nitrogen. Nitrogen fertilizer application rates may increase the potential

Fig. 6 Schoeller (1955, 1962) semi-logarithmic diagram

groundwater nitrogen pollution. Nitrogen compounds contamination is one of the main causes of health problems in human beings. Therefore, nitrogen compounds (NO3, NO2, and NH4) were analyzed in the investigation area (Table 1). Nitrite and ammonium concentration values of water samples are 0.01 and 0.00 mg/L, respectively. These values are suitable for the WHO (2011) and SWQR (2015) water quality standards. The concentration of the nitrate in water samples ranged from 1.43 to 9.31 mg/L with mean value of 5.31 mg/ L. The results obtained analysis were evaluated according to

143

Arab J Geosci (2017) 10:143

Page 8 of 17

Fig. 7 Classification of the water analyses in the drinkable water diagram by Schoeller

the WHO (2011) and SWQR (2015) water quality standards. The spring water values are above the WHO (2011) drinking water standard limit (10 mg/L) (Table 1). So, effects of

nutrient pollution in drinking water have not been observed in the study area. The surface water samples are suitable for the Class I and II limit (>5–10 mg/L) according to the SWQR

Arab J Geosci (2017) 10:143

Page 9 of 17 143

Fig. 8 Salinity and alkalinity hazard of irrigation water in USA Salinity (Richards 1954) diagram Fig. 9 Suitability of water for irrigation in Wilcox (1955) diagram

(2015), and the waters are good quality^ (Table 2).

classified as Bvery good and

Heavy metal contamination Heavy metals were analyzed for water samples in the investigation area. The Ba, Cd, Cr, Cu, Fe, Mn, Pb, and Zn concentrations in water samples ranged from 6.55–150.30 μg/L, 0.05–0.18 μg/L, 0.90–3.90 μg/L, 0.50–4.50 μg/L, 10 to 214 μg/L, 0.12–70.35 μg/L, 0.10– 1.50 μg/L, and 0.50–13.30 μg/L, respectively (Table 1). Generally, there is not negative effect for water samples except for aluminum and arsenic concentration increases in the some water samples. Aluminum (Al) The Al contents in the spring water samples were determined from 2 to 17 μg/L (Table 1). Spring water samples in the study area are suitable to the permissible limit value (200 μg/l) according to WHO (2011) drinking water standard. The Al concentration values in surface water samples ranged from 2 to 477 μg/L with a mean value of 60.10 μg/L (Table 1). Maximum permissible limit of Al concentration for the surface water is ≤300 μg/L for SWQR (2015). The extreme value of Al concentration (477 μg/L) is determined in sample 6 (Kufi stream). The water sample is classified as

Bmedium quality^ according to the SWQR (2015), (Table 2). This surface water sample could be related to geology due to the interaction of water with igneous rocks, the Homa mélange. The increase of Al content is originated from geogenic sources with water–rock interaction as related to feldspar, kaolin, and mica minerals within igneous rocks (Eriksson 1981).

Arsenic (As) The As concentration values in water samples ranged from 1.10 to 30.50 μg/L (Table 1). The permissible limit of As for drinking water is 10 μg/L according to WHO (2011) standard. As concentration of Yuva and Gümüşsu spring samples is low, but As concentration of Işıklı spring water sample (7) is high (30.05 μg/L). The increase could be geogenic origin due to the interaction of water with igneous rocks in Homa mélange. As concentration of groundwaters average is 1–2 μg/L except for areas affected with igneous rocks and sulfide minerals (Şahinci 1991). But, Işıklı spring is a drinking water source. The long-term consumption of waters containing high As concentration is hazardous to human health. High As concentrations in drinking waters is known to increase cancer risk in humans (Chatterjee and Mukherjee 1999).

143

Arab J Geosci (2017) 10:143

Page 10 of 17

Table 2 The surface water quality criteria (SWQR 2015)

Water quality parameters

General conditions Temperature (°C) pH EC (μS/cm) Trace elements Al (μg/l) As (μg/l) Ba (μg/l) Cd (μg/l) Cr (μg/l) Cu (μg/l) Fe (μg/l) Mn (μg/l) Pb (μg/l) Zn (μg/l) Nutrient parameters NO3− NO2− NH4+

Water quality classes Class I Very good quality

Class II Good quality

Class III Medium quality

Class IV Bad quality

≤25 6.5–8.5 <400

≤25 6.5–8.5 1000

≤30 6.0–9.0 3000

>30 <6.0 or >9.0 >3000

≤300

≤300

1000

>1000

≤20 ≤1000 ≤2 ≤20 ≤20 ≤300 ≤100 ≤10 ≤200

50 2000 5 50 50 1000 500 20 500

100 2000 7 200 200 5000 3000 50 2000

>100 >2000 >7 >200 >200 >5000 >3000 >50 >2000

<5 <0.01 <0.2

10 0.06 1

20 0.12 2

>20 >0.3 >2

The As concentration values in surface water samples ranged from 4.00 to 21.10 μg/L. As concentration of Kufi stream, sample (5) is 21.10 μg/L. The increase may be for the Işıklı spring feeds to Kufi stream (Fig. 2). The water sample is classified as Bgood quality,^ and the other surface water samples are as Bvery good quality^ for SWQR (2015), (Table 2).

A dendrogram of sampling sites in region is obtained by Ward’s (Joe and Ward 1963) clustering methods. There are two major clusters for the clustering procedure generated as shown in Fig. 10. The generated a dendrogram, where all the ten sampling sites were grouped into two clusters at (Dlink/ Dmax) × 100 < 20. Cluster I included sites 1, 2, 3, 7, 8, 9, and 10. These sites are Dinarsuyu stream, Büyük Menderes River,

Multivariate statistical analysis Hierarchical agglomerative cluster analysis was performed on the normalized data set (mean of observations over the whole period) by means of the Ward’s method using squared Euclidean distances as a measure of similarity (Kazi et al. 2009). Ward’s method is distinct from other linkage rules because it uses an analysis of variance approach to evaluate the distances between clusters (StatSoft Inc 2004). Cluster significance was determined using the criterion of 0.66 Dmax (Simeonov et al. 2003a) Cluster analysis was applied to the river water quality data set with a view to group the similar sampling sites (spatial variability) spread over the river stretch and in the resulted dendrogram, the linkage distance is reported as Dlink/Dmax, which represent the quotient between the linkage distance for a particular case divided by the maximal distance, multiplied by 100 as a way to standardize the linkage distance represented on y-axis (Simeonova et al. 2003b; Singh et al. 2004; Alberto et al. 2001).

Fig. 10 Dendrogram of the hierarchical cluster analysis using the Ward (Joe and Ward 1963) method

1.000 −0.617 −0.078 0.411 0.036 0.377 0.313 0.209 0.465 0.199 0.269 0.425 0.071 −0.136 −0.208 0.706* −0.656 0.147 0.127 0.491 −0.029 −0.264 −0.659 −0.616 −0.669 −0.217

1.000 0.334 −0.248 −0.635 −0.107 −0.451 0.232 −0.204 0.358 −0.009 −0.580 0.617* 0.136 0.117 −0.408 0.963* 0.243 0.336 −0.267 −0.368 0.320 0.961* 0.950* 0.956* −0.027

T

1.000 0.776* −0.403 0.849* 0.591* 0.820* 0.810* 0.571* 0.903* 0.588* 0.588* −0.147 0.676* 0.205 0.271 0.173 0.932* −0.286 −0.019 0.378 0.239 0.454 0.245 −0.250

EC

1.000 0.072 0.970* 0.807* 0.777* 0.971* 0.450 0.934* 0.840* 0.233 −0.071 0.501* .627 −0.239 0.333 0.803* 0.130 0.104 0.195 −0.270 −0.075 −0.260 −0.218

TDS

1.000 −0.124 0.596* −0.299 −0.123 −0.261 −0.212 0.556* −0.845 0.276 −0.181 0.200 −0.439 0.199 −0.406 0.403 0.313 −0.139 −0.432 −0.514 −0.414 0.280

Eh

1.000 0.658* 0.806* 0.983* 0.506* 0.987* 0.700* 0.397 −0.028 0.605* 0.548* −0.132 0.287 0.861* 0.037 0.045 0.318 −0.172 0.029 −0.160 −0.172

Na+

1.000 0.488 0.652* 0.256 0.581* 0.943* −0.212 0.070 0.149 0.618* −0.316 0.499 0.450 0.393 0.233 0.088 −0.329 −0.236 −0.314 −0.019

Ca2+

1.000 0.786* 0.432 0.782* 0.517* 0.471 −0.289 0.473 0.664* 0.202 0.498 0.891* 0.160 0.044 0.189 0.198 0.384 0.185 −0.348

K+

1.000 0.466 0.959* 0.730* 0.360 −0.141 0.556 0.589 −0.247 0.212 0.832* 0.031 0.050 0.159 −0.283 −0.069 −0.276 −0.307 1.000 0.509* 0.160 0.621* 0.499 0.142 0.154 0.397 0.587* 0.662* 0.279 −0.655 0.132 0.362 0.433 0.360 −0.257

Mg2+ Cl−

Correlation matrix of variables taken for waters from investigation area

*represent r > 0.500

pH T EC TDS Eh Na+ Ca2+ K+ Mg2+ Cl− SO42− HCO3− CO32− NO3− NO2− NH3− Al As Ba Cd Cr Cu Fe Mn Pb Zn

pH

Table 3

1.000 0.608* 0.471 −0.014 0.654* 0.418 −0.049 0.215 0.882* −0.088 0.027 0.392 −0.094 0.112 -.078 −0.141 1.000 −0.301 −0.079 0.274 0.714* −0.499 0.335 0.417 0.346 0.242 −0.068 −0.507 −0.373 −0.504 −0.167 1.000 0.001 0.156 −0.063 0.519* 0.178 0.683* −0.179 −0.346 0.327 0.488 0.572* 0.491 −0.198 1.000 0.061 −0.276 0.289 0.383 −0.080 0.236 −0.341 0.418 0.237 0.118 0.278 0.503 1.000 0.059 0.072 −0.140 0.479 −0.277 0.078 0.419 0.041 0.223 0.043 0.012

1.000 −0.377 0.515* 0.380 0.651* 0.169 −0.150 −0.357 −0.262 −0.380 −0.248

1.000 0.412 0.305 −0.123 −0.317 0.355 0.993* 0.949* 0.998* 0.062

SO42− HCO3− CO32− NO3− NO2− NH3− Al

1.000 0.435 0.643* −0.113 0.144 0.398 0.352 0.417 0.005

As

1.000 −0.060 −0.098 0.257 0.268 0.449 0.280 −0.364

Ba

1.000 −0.234 −0.071 −0.078 −0.110 −0.126 0.084

Cd

1.000 0.091 −0.351 −0.397 −0.277 0.213

Cr

1.000 0.330 0.337 0.352 0.777*

Cu

Mn

Pb

Zn

1.000 0.961* 1.000 0.992* 0.937* 1.000 0.055 −0.070 0.080 1.000

Fe

Arab J Geosci (2017) 10:143 Page 11 of 17 143

143

Page 12 of 17

Fig. 11 Relationship between electrical conductivity (EC) and major ions

Arab J Geosci (2017) 10:143

Arab J Geosci (2017) 10:143

Işıklı Lake, and springs. Hydrogeochemical relations among sampling locations in the Cluster I is observed clearly by dendrogram. Dinarsuyu stream (1 and 2) and springs (7, 8, 9) feeds to the Işıklı Lake (10). Also, Işıklı Lake (3) discharges into the Büyük Menderes River (10). Cluster II included sites 4, 5, and 6, which are samples of Kufi stream. The Cluster II corresponds to the high pollution region. The Cluster II has different properties due to the Kufi stream that arises from the Küçük Sincanlı plain. Because 4 and 5 (Kufi stream) samples were taken through a single line along the river, they have the same hydrochemical characteristic. Also, the Kufi stream sample (5) has high arsenic concentration value; the Kufi stream sample (6) has high the aluminum concentration value due to the geogenic origin. Principal component analysis (PCA) Correlation analysis was carried out for the water quality parameters. Table 3 presents the matrix of correlation analysis for hydrochemical data in study area. The analysis results indicated that significant positive (underlined values) and negative correlations existed between parameters of water quality. Between hydrochemical variables, it confirms the high interdependence. The resultant matrix shows up the positive correlation of total dissolved solids with electrical conductivity (r = 0.776, Table 3). TDS concentration is influenced by surface water runoff from point and nonpoint sources (Shin et al. 2013). Electrical conductivity of water is a direct function of its total dissolved salts (Harilal et al. 2004) and is used as an index to represent the total concentration of soluble salts in water (Purandara et al. 2003; Gupta et al. 2008). Comparison of the relationship between EC and major ions (Fig. 11) showed that the high correlations are obtained for water samples (r > 0.500, Fig. 12 Scree plot of the eigenvalues of principal components

Page 13 of 17 143

Table 3). The relationship is useful to interpret the hydrogeochemical properties of water. The cations and anions exhibited high and low correlation. There is a high correlation between Mg2+, Na+, and EC (R2 = 0.94, Fig. 11a, b), but the correlations of K+, Ca2+, and EC are low, with R2 = 0.60 and 0.54, respectively (Fig. 11c, d). The high correlation of Mg2+, Na+ cations in the waters can indicate contribution of the geogenic sources in the catchments. The effect of geological formations on groundwater quality is mainly due to the influence of EC in terms of TDS, Mg2+, Ca2+, and SO42−, rather than other parameters (Moncaster et al. 2000). There is a high correlation between SO42− and EC (R2 = 0.86, Fig. 11e), followed by the low correlation between Cl−, HCO3− and EC (R2 > 0.50, Fig. 11f, g). The SO42− anion of waters can reach high levels as a result of anthropogenic activities in the study area. The sulfate in groundwaters circulating through sedimentary rocks may be naturally derived from the dissolution of soluble sulfate minerals, the oxidation and solution of reduced sulfur minerals, or from a variety of both organic and inorganic atmospheric and soil sources which are combined in recharge to the system (Moncaster et al. 2000). The EC value increases with TDS, and the relationship exhibited the highest correlation with R2 = 0.99, Fig. 11h. TDS in water originates from natural sources, sewage, urban runoff, and industrial wastewater (WHO 2003). Factor analysis (FA) Although developed as a tool in the social sciences, R-mode factor analysis has proven highly effective in studies of groundwater quality (Love and Hallbauer 1998; Olmez et al. 1994; Reghunath et al. 2002; Subbarao et al. 1995). R-mode analysis method in factor analysis has been done for the data generated.

143

Arab J Geosci (2017) 10:143

Page 14 of 17

Table 4 Eigenvalues, percent of variance, and percent of cumulative for the factor analysis of parameters in study area Component

Eigenvalue

% Total Variance

Cumulative % Variance

1 2

8.748* 7.641*

33.648* 29.389*

33.648* 63.037*

3 4 5 6 7 8 9

3.070* 2.546* 1.679* 1.187* 0.644 0.376 0.108

11.809* 9.792* 6.458* 4.565* 2.478 1.445 0.417

74.846* 84.637* 91.095* 95.660* 98.138 98.583 100.000

*represent eigenvalues > 1 (6 factors)

R-mode analysis reveals the interaction among the variables studied. R-mode analysis is applied to ten samples collected from surface waters and springs in the investigation area.

The Scree plot was used to identify the number of PCs to be retained in order to comprehend the underlying data structure (Vega et al. 1998; Jackson 2005). PCA was carried out to extract the various factors. The Scree plot is shown in Fig. 12, which also includes the percentage variances explained by each component and gives an idea on how the different principal components were extracted. According to Fig. 12, after the sixth principal component, starting the elbow in the downward curve, other components can be omitted. Variations in water quality parameters were evaluated through R-mode factor analysis. The R-mode analysis was performed on the 26 parameters. Correlations among 26 hydrochemical parameters (in situ measurement, major elements, nutrients, and trace elements) are statistically examined to determine hydrochemical relations. The rotated loadings, eigenvalues, percentage of variance, and cumulative percentage of variance of all the six factors are given in Table 4. The R-mode analysis, performed for the six factors (indicated by PCA) using the varimax rotation (Table 5). In PCA,

Table 5

Factor structure and loading for varimax rotated factor matrix of six-factor in study area

No

Variable

1 2 3 4 5 6 7 8 9

pH T Eh EC TDS Na+ Ca2+ K+ Mg2+

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Cl− SO42− HCO3− CO32− NO3− NO2− NH3− Al As Ba Cr Cd Cu Fe Mn Pb Zn

Units

Communalities factor loading

Component 1

2

3

4

5

6

C mV μS/cm mg/l mg/l mg/l mg/l mg/l

0.990 0.991 0.992 0.993 0.995 0.998 0.934 0.978 1.000

0.201 0.001 0.926* 0.947* −0.158 0.977* 0.627 0.836* 0.957*

−0.730 0.923* 0.296 −0.203 −0.333 −0.135 −0.222 0.260 −0.240

0.540 −0.115 −0.159 0.193 0.222 0.115 0.399 0.353 0.105

−0.305 −0.333 −0.139 0.124 0.890* −0.051 0.575 −0.152 −0.070

−0.124 0.023 −0.032 −0.022 0.133 0.066 0.034 −0.167 −0.078

0.132 0.117 −0.022 0.035 −0.037 0.071 −0.016 −0.190 0.757*

mg/l mg/l mg/l mg/l mg/l mg/l mg/l μg/l μg/l μg/l μg/l μg/l μg/l μg/l μg/l μg/l μg/l

0.985 0.998 0.992 0.936 0.925 0.691 0.942 0.997 0.952 0.981 0.870 0.814 0.980 0.998 0.969 0.994 0.978

0.464 0.983* 0.665 0.401 −0.094 0.679 0.458 −0.028 0.244 0.898* −0.084 0.079 0.310 −0.066 0.151 −0.053 −0.230

0.282 −0.062 −0.391 0.387 0.155 0.093 −0.315 0.971* 0.413 0.303 −0.130 0.875* 0.277 0.975* 0.947* 0.975* −0.004

0.285 −0.006 0.313 0.026 0.068 −0.421 0.743 0.003 0.795 0.134 0.875 −0.047 −0.061 0.031 −0.025 0.012 0.004

−0.141 −0.099 0.532 −0.751 0.285 0.058 0.039 −0.145 0.221 −0.202 0.179 0.227 −0.119 −0.145 −0.182 −0.126 0.125

−0.058 0.107 −0.108 0.024 0.570 0.200 −0.180 0.102 0.071 −0.124 0.057 0.193 0.888* 0.072 −0.014 0.109 0.950*

0.765 0.081 −0.066 0.246 0.694 −0.035 −0.216 0.150 0.189 0.093 0.213 −0.820 −0.035 0.127 0.124 0.112 −0.087

o

*Represents strong loadings

Arab J Geosci (2017) 10:143

eigenvalues indicate the significance of each factor, and eigenvalues of 1.0 or greater are considered to indicate significance, with higher values indicating greater significance (Kim and Mueller 1987). R-mode analysis of the entire data set evolved six factors with eigenvalues >1 explaining about 95.66% of the total variance in the water-quality data set. The first eigenvalue is 8.75 which accounts for 33.65% of the total variance, and this constitutes the first and main factor. The second, third, fourth, fifth, and sixth eigenvalues are 7.64, 3.07, 2.55, 1.68, and 1.19, and these account for 29.39, 11.81, 9.79, 6.46, and 4.56%, respectively, of the total variance (Table 4). Liu and Wu (2003) classified the factor loadings as terms Bstrong,^ Bmoderate,^ and Bweak^ that refer to absolute loading values of >0.75, 0.75–0.50, and 0.50–0.30, respectively. The first factor is characterized by strong loadings of Na, K, Mg, SO4, Ba, Eh and EC, and moderate loadings of Ca, HCO3, and NO2. The strong loading (loading >0.75) of Na, K, Mg, SO4, Ba, Eh and EC parameters indicates carbonate weathering reactions and ion-exchange processes in the waters. The second factor is mainly associated with very strong loadings of Al, Fe, Mn, Pb, and T. The strong loading of these ions and temperature indicates anthropogenic input in the waters. The third factor is related with strong loadings of As and Cr, and moderate loadings (loading 0.75–0.50) of NH3 and pH. The strong loading of As and Cr ions represents water– rock interaction, weathering process, and geogenic origin. The fourth factor is involved by strong loadings of TDS and moderate loadings of Ca and HCO3. This factor can be interpreted as the physiochemical source of variability. The fifth factor is characterized by strong loadings of Cu and Zn and moderate loadings of NO3. This factor indicates an anthropogenic input in the waters due to leaching of fertilizer from agriculture land. The sixth factor is interrelated with strong loadings of Cl and moderate loadings of NO3. The strong loadings of Cl ion may be sign of anthropogenic pollution (Table 5).

Conclusion In this study, the water quality data for the Işıklı Lake, surface waters, and springs feeding the lake were evaluated for variations of the relationship between physical and chemical parameters. This study finds that the major water type in the investigation area is the Ca–HCO3–Cl water type which gradually degrades into Na–Mg–Ca–HCO3–Cl water types towards the northwest. As and Al concentrations are high in the Kufi stream samples and Işıklı spring sample due to the interaction with igneous rocks. Also, various multivariate statistical techniques were successfully applied to evaluate variations in water quality parameters of investigation area. There is high correlation (R2 > 0.85) between electrical conductivity and TDS, Mg2+, Na+, SO42− ions of waters in the correlation analysis. The high correlation coefficients of waters showed

Page 15 of 17 143

the impact of anthropogenic activities and geogenic sources in the region. Cluster analysis grouped ten sampling sites into two clusters of similar water quality characteristics according to the Ward method. The results obtained from the Ward method indicate that the parameters responsible for water quality variations are mainly related to heavy metals (Al and As) and geogenic pollution. Also, R-mode factor analysis helped in identifying the factors responsible for water quality variations in six different regions. The results of the PCA suggested parameters responsible for water quality variations in region was mainly related to water–rock interaction, weathering process, geogenic origin, anthropogenic pollution, etc. Thus, in this study, the multivariate statistical techniques were provided significant benefits to get better information about the quality of water sources. Acknowledgements This study has been achieved under the scope of The Scientific and Technological Research Council of Turkey (TÜBİTAK/2209-A), the project of University Students Research Projects Support Programme-2015.

References Afşin M (1991) Afyon Sandıklı Kuruçay ovası ve Hüdai kaplıcasının hidrojeoloji incelemesi Ankara Üniversitesi Doktora tezi yayımlanmamış 330 s Ankara Alam MN Elahi F Alam DU (2006) Risk and water quality assessment over view of River Sitalakhya in Bangladesh Acad Open Internet J http://www.acadjournal.com/2006/v19/part2/p1 Alamgir A, Khan MA, Schilling J, Shaukat SS, Shahab S (2016) Assessment of groundwater quality in the coastal area of Sindh province Pakistan. Environ Monit Assess 188(2):1–13 Alberto WD, del Pilar DM, Valeria AM, Fabiana PS, Cecilia HA, de los Ángeles BM (2001) Pattern recognition techniques for the evaluation of spatial and temporal variations in water quality a case study: Suquı́a River Basin (Córdoba–Argentina). Water Res 35(12):2881– 2894 Arslan H (2013) Application of multivariate statistical techniques in the assessment of groundwater quality in seawater intrusion area in Bafra plain Turkey. Environ Monit Assess 185(3):2439–2452 Belkhiri L, Boudoukha A, Mouni L, Baouz T (2010) Application of multivariate statistical methods and inverse geochemical modeling for characterization of groundwater—a case study: Ain Azel plain (Algeria). Geoderma 159(3):390–398 Buck O, Niyogi DK, Townsend CR (2004) Scale-dependence of land use effects on water quality of streams in agricultural catchments. Environ Pollut 130:287–299 Çakmakoğlu A (1986) Çivril-Banaz-Sandıklı-Dinar Arasındaki Bölgenin Jeolojisi (K23-c1c2c3c4d3; K24-d3d4; L23-a2a3b1b2b3b4; L24a1a2a3a4) MTA Rap Derleme 8062 Ankara Carpenter SR, Caraco NF, Correll DL, Howarth RW, Sharpley AN, Smith VH (1998) Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol Appl 8(3):559–568 Ceylan MA (1998) Baklan-Çivril havzası ve yakın çevresinin hidrojeomorfolojik etüdü Doktora Tezi Marmara Üniversitesi Sosyal Bilimler Enstitüsü 275 s İstanbul (yayınlanmamış) Ceylan MA, Eskikurt A (2001) Kufi Çayı boğazının doğal ve tarihî coğrafyası. Marmara Coğrafya Dergisi 3(1):123–152

143

Page 16 of 17

Chatterjee A, Mukherjee A (1999) Hydrogeological investigation of ground water arsenic contamination in South Calcutta. Sci Total Environ 225:249–262 Eriksson E (1981) Aluminium in groundwater possible solution equilibria. Nord Hydrol 12:43–50 Facchinelli A, Sacchi E, Mallen L (2001) Multivariate statistical and GISbased approach to identify heavy metals sources in soils. Environ Pollut 114:313–324 Feller MC (2007) Forest harvesting and streamwater inorganic chemistry in western north American: a review. J Am Water Resour Assoc 41: 785–811 Freeze RA,Cherry JA (1979) Groundwater Englewood cliffs: PrenticeHall, Upper Saddle River FugroSİAL (2014) Yukarı Büyük Menderes havzası (Küfi) yeraltısuyu planlama (hidrojeolojik etüt) raporu s:396 Garizi AZ, Sheikh V, Sadoddin A (2011) Assessment of seasonal variations of chemical characteristics in surface water using multivariate statistical methods. Int J Environ Sci Technol 8(3):581–592 Güler C, Kurt MA, Alpaslan M, Akbulut C (2012) Assessment of the impact of anthropogenic activities on the groundwater hydrology and chemistry in tarsus coastal plain (Mersin SE Turkey) using fuzzy clustering multivariate statistics and GIS techniques. J Hydrol 414:435–451 Gupta S, Maheto A, Roy P, Datta JK, Saha RN (2008) Geochemistry of groundwater Burdwan district West Bengal India. Environ Geol 53: 1271–1282 Gürbüz A, Boyraz S, Ismael MT (2011) Plio-Quaternary development of the Baklan–Dinar graben: implications for cross-graben formation in SW Turkey. Int Geol Rev 54(1):33–50 Harilal CC, Hashim A, Arun PR, Baji S (2004) Hydro geochemistry of two rivers of Kerala with special reference to drinking water quality. J Ecol Environ Conserv 10(2):187–192 Helena B, Pardo R, Vega M, Barrado E, Fernandez JM, Fernandez L (2000) Temporal evolution of groundwater composition in an alluvial aquifer (Pisuerga River Spain) by principal component analysis. Water Res 34:807–816 Hem JD (1985) Study and Interpretation of the water chemical characteristic of natural water Science for a changing world USGS http:// pubsusgsgov/wsp/ wsp2254/pdf/intropdf Hill AR (1990) Ground water flow paths in relation to nitrogen chemistry in the near-stream zone. Hydrobiologia 206(1):39–52 Hussain M, Ahmed SM, Abderrahman W (2008) Cluster analysis and quality assessment of logged water at an irrigation project eastern Saudi Arabia. J Environ Manag 86:297–307 Ismael MT (2009) Işıklı Göl (Çivril/Denizli) Tortullarının Sedimantolojik İncelenmesi Ankara Üniversitesi Fen Bilimleri Enstitüsü Yükseklisans Tezi 80s Ankara Jackson JE (2005) A user’s guide to principal components, vol 587. Wiley, Hoboken Jarvie HP, Whitton BA, Neal C (1998) Nitrogen and phosphorus in east coast British rivers: speciation sources and biological significance. Sci Total Environ 210:79–109 Joe H, Ward JR (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58:236–244 Kazi TG, Arain MB, Jamali MK, Jalbani N, Afridi HI, Sarfraz RA, Shah AQ (2009) Assessment of water quality of polluted lake using multivariate statistical techniques: A case study. Ecotox Environ Safe 72(2):301–309 Kim JO, Mueller CW (1987) Introduction to factor analysis: what it is and how to do it quantitative applications in social sciences series. Sage University Press, Newbury Park Koç C (2010) A study on the pollution and water quality modeling of the River Buyuk Menderes Turkey. Clean–Soil Air Water 38(12):1169– 1176 Konak N (2002) Türkiye’nin 1/500000 ölçekli jeoloji haritası İzmir bölgesi MTA Ankara

Arab J Geosci (2017) 10:143 Konak N Senel M (2002) 1/500000 ölçekli Türkiye Jeoloji Haritasi Denizli paftası MTA Ankara Liu GD, Wu WL (2003) The dynamics of soil nitrate leaching and contamination of groundwater in high-yielding farmlands. Chin J EcoAgric 11(1):91–93 Love D Hallbauer DK (1998) Characterisation of groundwater contamination in a dolomitic aquifer: a case study from the Far West Rand Geocongress 98 Geological Society of South Africa Pretoria South Africa July 1988 Extended abstracts pp 272–275 Marinović RA, Ruždjak D (2015) Evaluation of river water quality variations using multivariate statistical techniques: Sava River (Croatia): a case study. Environ Monit Assess 187(4):4393–4393 Moncaster SJ, Bottrell SH, Tellam JH, Lloyd JW, Konhauser KO (2000) Migration and attenuation of agrochemical pollutants: insights from isotopic analysis of groundwater sulphate. J Contam Hydrol 43: 147–163 Oketola AA, Adekolurejo SM, Osibanjo O (2013) Water quality assessment of river Ogun using multivariate statistical techniques. J Environ Prot 4:466–479 Oliva González S, Almeida CA, Quintar S, Mallea MA, González PS (2011) Application of multivariate statistical techniques to evaluate organic pollution on a river in Argentina Ambiente & Água-an Interdisciplinary. J Appl Sci 6(3):27–42 Olmez I, Beal JW, Vilaaume JF (1994) A new approach to understanding multiple-source groundwater contamination: factor analysis and chemical mass balances. Water Res 28:1095–1101 Öngür T (1973) Sandıklı (Afyon) Jeotermal Araştırma Bölgesine İlişkin Jeolojik Durum ve Jeotermal Enerji Olanakları Maden Tetkik ve Arama Genel Müdürlüğü Ankara 5520 Öztaş Y (1989) Homa-Akdağ yöresi Sandıklı-Şuhut kuzeylerinin jeolojisi ve petrol olanakları TPAO Arşivi 2584 Ankara Öztürk A (1981) Homa-Akdağ (Denizli) Yöresinin Stratigrafisi TJK Bülteni Ankara 24(1):75–84 Palma P, Alvarenga P, Palma VL, Fernandes RM, Soares AM, Barbosa IR (2010) Assessment of anthropogenic sources of water pollution using multivariate statistical techniques: a case study of the Alqueva’s reservoir Portugal. Environ Monit Asses 165(1–4):539– 552 Park S, Kazama F, Lee S (2014) Assessment of water quality using multivariate statistical techniques: a case study of the Nakdong River basin Korea. Environ Eng Res 19(3):197–203 Perona E, Bonilla I, Mateo P (1999) Spatial and temporal changes in water quality in a Spanish River. Sci Total Environ 241(1):75–90 Peters NE, Meybeck M (2000) Water quality degradation effects on freshwater availability: impacts of human activities IWRA. Water Int 25(2):185–193 Piper AM (1944) A graphical procedure in the geochemical interpretation of water analysis. Am Geophys Union Trans 25:914–928 Purandara BK, Varadarajan N, Jayashree K (2003) Impact of sewage on ground water: a case study. Pollut Res 22(2):189–197 Raiber M, White PA, Daughney CJ, Tschritter C, Davidson P, Bainbridge SE (2012) Three-dimensional geological modelling and multivariate statistical analysis of water chemistry data to analyse and visualise aquifer structure and groundwater composition in the Wairau plain Marlborough District New Zealand. J Hydrol 436:13–34 Raymond PA, Oh NH, Turner RE, Broussard W (2008) Anthropogenically enhanced fluxes of water and carbon from the Mississippi River. Nature 451:449–452 Reghunath R, Sreedhara Murthy TR, Raghavan BR (2002) The utility of multivariate statistical techniques in hydrogeochemical studies: an example from Karnataka India. Water Res 36:2437–2442 Richards LA (1954) Diagnosis and improvement of saline alkali soils: agriculture handbook, vol 160. US Department of Agriculture, Washington DC, p 60 Şahinci A (1991) Geochemistry of natural waters The Reform Printing Office Izmir (in Turkish)

Arab J Geosci (2017) 10:143 Salem ZE, Atwia MG, El-Horiny MM (2015) Hydrogeochemical analysis and evaluation of groundwater in the reclaimed small basin of Abu Mina Egypt. Hydrogeol J 23(8):1781–1797 Schoeller H (1955) Géochemie des eaux souterraines. Rev Inst Fr Pétrol 10:230–244 Schoeller H (1962) Les eaux souterraines Hydrologie dynamique et chimique Recherche exploitation et évaluation des ressources Masson & Cie Paris France 642 Şenel M (2002) Geological Map of Turkey in 1/500000 Scale: Konya Sheet Publication of Mineral Research and Exploration Directorate of Turkey (MTA) Ankara Shin JY, Artigas F, Hobble C, Lee YS (2013) Assessment of anthropogenic influences on surface water quality in urban estuary northern New Jersey: multivariate approach. Environ Monit Assess 185: 2777–2794 Simeonov V, Stratis JA, Samara C, Zachariadis G, Voutsa D, Anthemidis A, Kouimtzis T (2003a) Assessment of the surface water quality in northern Greece. Water Res 37(17):4119–4124 Simeonova P, Simeonov V, Andreev G (2003b) Water quality study of the Struma river basin Bulgaria (1989–1998). Open Chem 1(2):121– 136 Singh KP, Malik A, Mohan D, Sinha S (2004) Multivariate statistical techniques for the evaluation of spatial and temporal variations in water quality of Gomti River (India)—a case study. Water Res 38(18):3980–3992 StatSoft Inc (2004) STATISTICA (Data Analysis Software System) Version 6 Subbarao C, Subbarao NV, Chandu SN (1995) Characterisation of groundwater contamination using factor analysis. Environ Geol 28:175–180 SWQR (2015) Surface Water Quality Regulation Official Gazette Republic of Turkey Date: 15042015 No: 29327 (in Turkish) Tijani MN (1994) Hydrochemical assessment of groundwater in Moro area Kwara state Nigeria. Environ Geol 24:194–202 Todd DK, Mays LW (2005) Groundwater hydrology, Third edn. Wiley, Hoboken

Page 17 of 17 143 Turan N (2002) Geological Map of Turkey in 1/500000 Scale: Ankara Sheet Publication of Mineral Research and Exploration Directorate of Turkey (MTA) Ankara Varol M, Gökot B, Bekleyen A, Şen B (2012) Water quality assessment and apportionment of pollution sources of Tigris River (Turkey) using multivariate statistical techniques—a case study. River Res Appl 28(9):1428–1143 Vega M, Pardo R, Barrado E, Deban L (1998) Assessment of seasonal and polluting effects on the quality of river water by exploratory data analysis. Water Res 32:3581–3592 WHO (2003) Nitrate and Nitrite in Drinking Water Background Document for Preparation of WHO Guidelines for Drinking Water Quality World Health Organization Geneva WHO (2011) World Health Organization Edition F Guidelines for Drinking-water Quality WHO chronicle 38:104–8 Wilcox LV (1955) Classification and use of irrigation water. USDA Circular, Washington DC, p 969 Wu EMY, Kuo SL (2012) Applying a multivariate statistical analysis model to evaluate the water quality of a watershed. Water Environ Res 84(12):2075–2085 Wunderlin DA, Diaz MP, Ame MV, Pesce SF, Hued AC, Bistoni MA (2001) Pattern recognition techniques for the evaluation of spatial and temporal variations in water quality a case study: Suquia river basin (Cordoba-Argentina). Water Res 35:2881–2894 Yidana SM (2010) Groundwater classification using multivariate statistical methods. Southern Ghana J Afr Earth Sci 57(5):455–469 Zhang Q, Li ZW, Zeng GM, Li JB, Fang Y, Yuan QS, Wang YM, Ye FY (2009) Assessment of surface water quality using multivariate statistical techniques in red soil hilly region: a case study of Xiangjiang watershed China. Environ Monit Assess 152:123–131 Zhang B, Song X, Zhang Y, Han D, Tang C, Yu Y, Ma Y (2012) Hydrochemical characteristics and water quality assessment of surface water and groundwater in Songnen plain Northeast China. Water Res 46(8):2737–2748 Zhao Y, Xia XH, Yang ZF, Wang F (2012) Assessment of water quality in Baiyangdian Lake using multivariate statistical techniques. Procedia Environ Sci 13:1213–1226