Distribution and environmental risk evaluation of heavy metal in core sediments from Lake Çıldır (NE Turkey)

Distribution and environmental risk evaluation of heavy metal

in core sediments from Lake Ç

ıldır (NE Turkey)

Serkan Kükrer&Ahmet Evren Erginal& SebahatŞeker&Mustafa Karabıyıkoğlu

Received: 6 February 2015 / Accepted: 9 June 2015 / Published online: 21 June 2015 # Springer International Publishing Switzerland 2015

Abstract This study examined the vertical distri-bution of heavy metals in core sediments taken from Lake Çıldır, Turkey, and their potential eco-logical risks together with organic carbon content and chlorophyll degradation products. Samples were collected from six sampling stations deter-mined along two main transections aligned in north–south and east–west directions. The enrich-ment (EF) and contamination factor (CF), potential ecological risk (PER) index, and pollution load index (PLI) were calculated from the obtained results. For the elements Pb, As, and Cd, a mod-erate level of contamination was detected, whereas a moderate-to-high concentration level was obtain-ed for Mn. The highest contamination level was found for Hg. A pollutant accumulation exists particularly in the surface sediments. Cd and Hg are the only two metals considered to be a poten-tial risk factor in the lake.

Keywords Lake Çıldır. Heavy metal pollution . Potential ecological risk . Sediment

Introduction

Comprising one of the key components of water masses, bottom sediments both provide nutrients to benthic or-ganisms and function as reservoirs for noxious chemical substances (Bodog et al.1997; Wang et al.2012). As is well known, various chemical compounds unknowingly or unscrupulously released into nature lead to environ-mental pollution. In coastal areas, heavy metal pollution, amongst other pollutants, constitutes a serious risk of global importance (Kamala-Kannan et al.2008). These pollutants are permanent, are toxic, and can be hazard-ous to human health when they enter the food chain (Kishe and Machiwa2003; Luo et al.2010; Bing et al.

2013; Çevik et al.2009).

Even though major and trace elements constitute micronutrients for organisms (Vrhovnik et al. 2013), over-accumulation of heavy metals may cause ecologi-cal damage (Karadede and Ünlü 2000). The elements Cu and Zn, for example, found in natural waters and sediments, play an important role in aquatic life as a micronutrient, albeit their toxic characteristics are at high concentrations (Bai et al.2011). Since they cannot easily be metabolized within living organisms, heavy metals accumulate within soft tissues and result in toxic effects (Suresh et al.2012).

In aquatic systems, metals are found as dissolved ions and complexes, as suspended ions, and as solids (Özmen et al. 2004). The metal ions may bind onto organic and inorganic complexes and accumulate as metal hydroxides and sulfides by attaching themselves to minerals or soluble metal salts. In consequence of DOI 10.1007/s10661-015-4685-1

S. Kükrer (*)

:

:

S.Şeker

Department of Environmental Engineering, Faculty of Engineering, Ardahan University, Ardahan, Turkey

chemical degradation of particles deposited in oxic sedi-ment layers, the contaminants become remobilized (Bodog et al.1997). Heavy metals accumulated within sediments can be released into the water depending on changes in sediment characteristics, such as oxidation and reduction, pH, and organic and inorganic carbon, as well as dissolved oxygen (Wang et al.2012; Çevik et al.2009).

Heavy metals in lake-water reservoirs are derived from different sources of natural or anthropogenic origin. The weathering of rocks provides one source of sediments containing heavy metals while industrial and domestic wastes comprise anthropogenic inputs (Ochieng et al.

2007). Having different sedimentological attributes in terms of grain size distribution, mineral content, and or-ganic compounds associated with variances in erosion and transportation agents as well as in the aquatic system, sediments may have considerable heavy metal composi-tion, equaling that of anthropogenic sources (Liu et al.

2010). On the other hand, all lakes do not have point sources (Rippey et al.2008). The deposition of metals in the form of atmospheric particles may be either anthropo-genically induced due to fossil fuels and mining activities (Hu et al.2011) or of volcanic origin with regard to gases dispersed during eruptions (Siegel and Siegel1987; Sy-monds et al.1987). Algae accumulate heavy metals from water and sediments, which produces an adverse effect on both algae communities and aquatic life (Makundi2001). Contaminants absorbed by the algae are also transported into sediments (Özkan2012).

In this study, the vertical distribution of heavy metals within core samples taken from Lake Çıldır, and their potential ecological effects, were examined together with organic carbon content and chlorophyll degrada-tion products.

Materials and methods

Study area

Situated at an elevation of 1959 m above the present sea level, freshwater lake Çıldır (41.0425° N 43.2552778° E) is the second largest lake in the Eastern Anatolian region of Turkey with a surface area of 115 km2. The lake has a roughly triangular shape with longest and shortest axes of 18.3 and 16.2 km, respectively. Previous data regarding maximum depth of the lake waters are controversial. Atalay (1978) suggested a maximum depth of 42 m; however, notably lower values

ranging between 10 and 14 m were achieved from our six sampling sites, referred to here as stations St1–St6 (Fig.1). The lake is surrounded by volcanic mountains to the east (Mt. Akbaba, 3026 m) and west (Mt. Kısır, 3197 m). To the north, the lake is separated from Çıldır Plain by an E-W trending volcanic ridge or lava flow lying at altitudes between 2000 and 2150 m, which has been attributed by Atalay (1978) as the main cause of the formation of the lake. A small tributary of the Arpaçay Stream (Çarçı St.), one of the main tributaries of the Kars River, discharges excess waters of the lake from the south, where the lake becomes narrower.

The environs of Lake Çıldır are composed totally of volcanic rocks of Neogene to Quaternary age with the exception of Pliocene lacustrine deposits cropping out to the south and north (Lahn1951; Demirsu1954). Given the lack of a meteorology station around the lake, we used data obtained from the nearest station at Arpaçay (1688 m) district, 10 km south of the lake. The lake area has an average annual precipitation of 492.1 mm, much of which is received during the period between May and June.

Even though the average annual temperature is 5.5 °C, the lowest and the highest values recorded up to the present are −29.2 and 35.1 °C, respectively. During the winter season, snow cover reaches a thick-ness of 90 cm and the upper 70 cm of the lake waters is frozen for 5–6 months during harsh winters.

Sampling and analyses

Samples were collected in August 2013 from six different sites determined along two main transections aligning in north–south and east–west directions (Fig. 1), using a Kajak gravity core sampler.

The length of cores varied between 70 and 20 cm due to the physical attributes of the lake bottom as well as currents and waves. The depth of subsample interval was 5 cm, from which two sets were separated for analysis of heavy metal and organic carbon (Corg) content and determination of chlorophyll degradation products (CDPs). To measure the Corg% within the collected sediments, subsamples dried for Corgwere first crushed in a mortar to obtain a fine-grained powder, and analysis was carried out using the Wakley–Black titration method (Gaudette et al.1974). Wet sediments were extracted for 24 h with acetone to ascertain CDP. After 24 h, the supernatant liquid obtained was read using a visible spectrophotometer (Lorenzen1971) and the CDP concentration was calculated from the absorbance of the solution. Heavy metal analyses were performed using

ICP-MS in ACME Analytical Labs, Canada. Quality con-trol was evaluated based on duplicates, method blanks, and internal standard reference material (STD OREAS45EA) from ACME Analytical Labs. The following results (μg/g) were acquired from the reference sample:

Elements Observed values Expected values Detection limits

Cu 685.66 685.55 0.01 Pb 14.29 13.56 0.01 Zn 27.1 27.6 0.1 Ni 375.2 376.6 0.1 Mn 348 362 1 Fe 24.20 % 23.64 % 0.01 % As 9.0 9.6 0.1 Cd 0.03 0.04 0.01 Cr 756.9 800.3 0.5 Al 3.08 % 3.08 % 0.01 % Hg <0.005 0.007 0.005 In order to determine metal enrichment in the sedi-ments, two different methods were adopted. One was

the enrichment factor (EF), which is used to determine the anthropogenic contribution in heavy metal concen-trations (Zhang et al.2007). This value is obtained from the ratio between the measured concentration of metal/ Al and that in background values within the lake sediments. In this study, the background values of lake sediments from Kükrer et al. (2014) were used and the following scale of Sutherland (2000) was considered in assessing the enrichment factor:

EF<2, deficiency to minimal enrichment EF=2–5, moderate enrichment

EF=5–20, significant enrichment EF=20–40, very high enrichment EF>40, extremely high enrichment

Another method used to ascertain contamination levels was the contamination factor (CF), which is the ratio of metal concentration to the background metal concentration. This was classified into four divisions (Hakanson1980), as follows:

CF<1, low contamination 1≤CF<3, moderate contamination 3≤CF<6, high contamination CF>6, very high contamination

To determine the environmental quality of the sedi-ments, pollution load index (PLI) was used (Suresh et al.

2011), which is calculated considering all metals present to ascertain the level of toxicity in the material being examined. The following formula was used:

PLI ¼ C Fð 1 C F2 ……: C FnÞ1=n

To determine any toxic effects of the metals, the potential ecological risk index (PERI) developed by Hakanson (1980) was used. The risk factor for an indi-vidual metal is calculated according to the following formula:

Eri¼ Cif Tri

where Triis the toxic-response factor of the individual metal (or response coefficient) and Cifis the contamina-tion factor. The coefficients corresponding to the toxic-ity of metals are as follows:

Hg = 40, Cd = 30, As = 10, Cu = Pb = Ni = 5, Cr = 2, Zn=1 (Guo et al.2010).

The following formula, on the other hand, was used to calculate the ecological risks caused by the combined effects of the metals:

PER ¼ ∑Eri

The following classification was used to evaluate the risk factor (Hakanson1980):

Eri<40, low potential ecological risk 40≤Eri

<80, moderate potential ecological risk 80≤Eri<160, considerable potential ecological risk 160≤Eri

<320, high potential ecological risk Eri≥320, very high ecological risk

PER<150, low ecological risk

150≤PER<300, moderate ecological risk 300≤PER<600, considerable ecological risk PER≥600, very high ecological risk

Pearson’s correlation test was applied to determine correlations between variables. Cluster analysis was used to designate distance between variables. In order to determine the principal causes of variation, factor analysis was carried out.

Results and discussion

Heavy metal content

The vertical distribution of measured heavy metals in core sections at each sampling station is shown in Fig.2. In addition, the minimum, maximum, and average values (±standard error) are presented in Table1.

The concentration of Cu, except at St1, does not show any marked variations with regard to depth. In the case of St1, despite the minor increases noted in the lower layers, a general trend of decrease is apparent. At all stations, Pb reaches its maximum value in the surface sediments and shows a decrease toward the base of the cores. Although the elements of Zn and Ni appear to have higher concentrations in the upper 10 cm, signifi-cant increases are also noted at the lower levels of the sediments. It was established that Mn shows a decreas-ing trend from the surface to the base. Fe shows higher concentrations in the surface sediments and a declining trend toward the lower levels in general, except in the cases of St1 and St2, where it exhibits minor increases at the lower levels. Except at St6, As appears to reach its maximum concentration in the upper 10 cm, but it is interesting to note that it shows irregular distribution throughout the cores. In all stations, it was determined that Cd reaches its maximum concentration in the 0–5-cm interval and shows a declining trend downward, except with some increases at the lower levels. Having higher amounts in the upper 10 cm, Cr shows varied concentrations in the deeper sections. Al has a tendency to decrease toward the deeper parts with the exception of subordinate increases. Unlike the others, Hg shows an irregular distribution in the vertical section, represented by increased concentrations at different levels.

In order to assess the sediment quality, mean values of the heavy metals in each station were compared with the values proposed by MacDonald et al. (2000). It should be noted here that adverse biological effects are rarely expected to occur below the threshold effect concentration (TEC) and are usually expected to be above the probable effect concentration (PEC). In this respect, Pb, Zn, As, Cd, Cr, and Hg values were deter-mined to be smaller than the values of TEC, whereas Cu value at St4 lies between the TEC and PEC values. Ni appears to be over the PEC value (Table1).

A comparison of values obtained from the surface sediments of Lake Çıldır with other lakes in Turkey and around the world is shown in Table2.

Enrichment factor

Enrichment factor (EF) is an important means used in evaluating probable sources of heavy minerals. Thus, it is possible to obtain a clear idea of the source of the heavy metal in terms of whether it is of anthropogenic origin or derived from elements existing in the bedrock and soils of the surrounding area. Box and Whisker diagrams of the EF values obtained from Lake Çıldır sediments are shown in Fig. 3. Calculated EF values show a decrease in the following order, as Hg>Cd>

As>Mn>Pb>Zn>Cu>Cr>Fe>Ni. The maximum en-richment for Hg is noted in the 20–25 interval of St5. In regard to EF, no distinct enrichment is noted in other metals. At St1, Hg enrichment occurs in the 5–10-cm interval, but this enrichment is slightly less than the surface value of St2. At St2, it is only the surface sediment that appears to be rich in Hg content. At St3, Hg enrichment appears to be limited to the upper 25-cm interval and reaches its highest value in the interval of 20–25 cm. St4 is represented by a short core, only 25 cm long, but a marked Hg enrichment is noted throughout Fig. 2 Vertical distribution of heavy metals in core samples

the core. A similar trend of Hg enrichment, except for the upper 0–5-cm interval, is also present in the 35-cm-long core obtained at St5. At St6, Hg enrichment occurs only in the upper 5–10-cm interval.

Contamination factor

A useful guideline for sediment contamination assess-ment is the contamination factor (CF). The critical value for CF is 1, and values greater than 1 are classified further to define the degree of contamination. Box and Whisker diagrams for the calculated CF values are shown in Fig.4.

The calculated CF values present a decrease in the following order: Hg>Cd>As>Al>Mn>Pb>Zn>Cu> Cr>Fe>Ni, respectively. In considering the mean values for Cu, Zn, Ni, Fe, and Cr, they appear to be very close to the limit of their CF values. Among these elements, it is only Zn which lies slightly above the critical value at some levels, whereas the others lie below the critical value. Pb values show moderate contamination through-out all stations, but slightly higher values appear in the

surface sediments. In considering the overall Mn values, significant contaminations appear at stations 3 and 4 and partially at station 6. In particular, Mn enrichment reaches its maximum value in the surface sediments, and a mod-erate contamination is also noted in the upper 20-cm interval. At St4, significant Mn enrichment is recorded in the upper 10 cm, with a CF value reaching above 3, representing a significant level. At St6, similarly, a max-imum concentration occurs in the 5–10-cm interval. As concentrations show a similar trend with those of Mn concentrations. A moderate level of As concentration occurs in the upper 20- and 10-cm intervals at St3 and St4, respectively, but a marked concentration detected in the 40–50-cm interval of St3 is close to the surface value. At St6, important peaks are noted only in the 5–10- and 30–33-cm intervals. At all stations, Cd concentration appears to be markedly higher in the surface sediments. The highest CF values were recorded for Hg. At St1, the highest CF values are recorded in the upper 10-cm inter-val of the core and the contamination is moderate. At St2, higher contamination appears in the 0–5-cm interval, while moderate contamination was detected at the lower Table 1 Descriptive statistics of variables measured in sediments from Lake Çıldır and comparison with toxicological references

St1 St2 St3 St4 St5 St6 Background TEC-PEC Cu 20.56–30.88 25.28±0.75 28.46–30.65 29.66±0.53 28.34–33.12 30.86±0.38 30.44–33.35 31.99±0.52 29.03–31.35 30.1±0.31 27.5–31.15 29.62±0.43 29.92 31.6–149 Pb 8.1–18.72 11.76±0.92 14.43–18.77 16.29±0.99 14.07–20.39 17.24±0.78 14.34–20.36 17.15±1.20 11.55–17.49 14.55±0.92 13.27–18.17 14.96±0.78 11.42 35.8–128 Zn 33.6–65 49.01±2.26 56.6–61.2 59.8±1.09 57.6–76.8 62.67±1.81 52.4–69.7 61.32±3.43 47.0–69.10 57.27±2.75 51.3–66.3 60.57±2.23 53.83 121–459 Ni 33.6–48.2 39.87±1.35 46.8–55.0 49.55±1.92 45.3–53.1 49.39±0.72 47.1–52.4 49.68±1.01 36.5–49.6 44.69±1.9 45.0–52.8 49.3±1.11 48.5 22.7–48.6 Mn 297–705 417.29±30.24 379–756 609±83.61 447–905 643±65.46 414–1415 809.6±188.4 365–696 530.43±49.45 382–946 586.43±79.87 409.75 – Fe 13,100–19,800 15,507.14±575.1 17,300–20,500 18,575±745.4 17,700–22,400 19,690±551.9 17,400–24,500 20,560±1348.2 15,700–20,900 18,871.43±746.3 18,800–22,700 20,085.7±610.8 19,225 – As 1.5–3 2.47±0.12 2.6–3.2 2.85±0.13 2.3–5.0 3.49±0.27 1.9–4.9 3.28±0.55 2.5–3.3 2.71±0.11 1.9–3.8 2.87±0.26 1.77 9.79–33 Cd 0.13–0.36 0.21±0.02 0.2–0.31 0.24±0.03 0.2–0.41 0.32±0.02 0.17–0.44 0.29±0.05 0.14–0.44 0.28±0.03 0.21–0.33 0.26±0.02 0.15 0.99–4.98 Cr 24–35.1 28.66±0.87 32.8–37.8 34.13±1.23 36.0–41.9 38.82±0.68 33.5–41.4 38.34±1.40 32.8–41.4 36.37±1.22 36.7–44.2 39.6±1.03 36.44 43.4–111 Al 12,500–21,000 15,792.86±662.53 18,500–21,400 19,850±597.9 20,000–22,600 21,250±312.43 19,900–24,000 21,700±887.13 16,400–22,500 19,942.86±951.9 21,000–24,900 22,871.4±462.2 13,380 – Hg ND–0.059 0.030±0.004 0.045–0.07 0.053±0.005 0.045–0.095 0.068±0.005 0.067–0.096 0.078±0.005 0.06–0.11 0.08±0.007 0.02–0.08 0.05±0.008 0.02 0.18–1.06 ORGC 1.24–2.54 1.83±0.11 2.4–2.49 2.45±0.02 2.12–2.47 2.31±0.03 1.93–2.52 2.32±0.1 2.26–2.63 2.49±0.04 2.23–2.72 2.44±0.06 – – CDP 3.6–74.29 20.62±4.77 17.98–69.37 19.88±8.38 8.84–161.03 45.43±14.2 8.35–81.6 32.82±13.30 23.79–125.87 52.5±12.99 22.57–84.89 40.2±8.34 – –

TEC threshold effect concentration, PEC probable effect concentration, ORGC organic carbon, CDP chlorophyll degradation products, ND not detected

levels. At St3, while higher contaminations are noted in the 25- and 45–50-cm intervals, respectively, moderate contamination is recorded at other levels. At St4, a high concentration is noted throughout the core sediment. At St5, while a moderate contamination is noted at the surface, a higher contamination is recorded between 5

and 30 cm, and is characterized by a moderate contam-ination again in the 30–36-cm interval. At St6, while high contamination occurs at the surface sediment, this is replaced by moderate concentration at the lower levels.

The fact that heavy metal enrichment is confined only to the surface sediment of the lake bottom can Table 2 Comparison of heavy metal content obtained from lake sediments in Turkey and other countries

Lake Cu Pb Zn Ni Mn Fe As Cd Cr Al Hg

Uluabat, Turkeya _{0.75 1.42 3.89 – – – – 0.078 2.95 –}

Atatürk Dam lake, Turkeyb 14.57– 22.7 ND 59.14– 60.79 43.69– 139.69 73.6– 514.07 12,587– 19,265 ND ND

Seyhan Dam, Turkeyc 19.80 39.09 803.63 39,350 2.15 118.95

Sapanca, Turkeyd _{26.68 15.20 62 26.72 337.81 – – 0.29 19.09 –} Hazar, Turkeye 10–64 ND 46–210 38–130 85–625 3650– 30,000 – – 17–79 – Baihua, Chinaf _{102.71 184.3 1780 5.39 3.08} Longgan, Chinag 26.4– 51.4 26.1– 64.0 59.5– 124.3 29.2–58.5 68.6– 123.4 47.4–101.8 (mg/g) Veeranam, Indiah _{94.12 30.06 180.08 63.61 – – – 0.81 88.20 –} Chapala, Mexicoi 29.26 81.74 102.75 32.24 – 3.97 % – – 66.12 4.5 % – Kapulukaya, Turkeyj 5–29.3 8.6–34 14.8– 124.2 24.7–127.1 326.6– 1053 0.92– 3.48 % 9.1– 69.7 0.5– 1.8 98– 11-16 1.47– 4.64 -% 1–1.6 Çıldır 30.11 18.85 64.52 50.12 781.5 21,000 3.27 0.38 39.57 22,583 0.07 ND not detected a Barlas et al.2005 b Karadede and Ünlü2000 c_{Çevik et al.2009} d Duman et al.2007 e Özmen et al.2004 f Wang et al.2012 g Bing et al.2013 h_{Suresh et al.2012} i Rosales-Hoz et al.2000 j Kankılıç et al.2013

Fig. 3 Box and Whisker diagrams of EF values obtained from Lake Çıldır sediments

reasonably be explained in terms of consumption of fossil fuels, mainly lignite, as an energy source in the settlements around the lake for most of the year. However, it is difficult to interpret metal enrichment at the lower intervals purely in terms of anthropogenic origin related to fossil fuel use and industrial waste in the study area. Volcanic origin is suggested as an alternative hypothesis for the source of the concentra-tion in lower level sediments since the study area is located in the Eastern Anatolian Volcanic Province where there are references in the literature for volca-nic eruptions of mounts of Nemrut, Süphan, and Ağrı during the historical times (Tchalenko 1977; Smithsonian Institution National Museum of Natural History Global Volcanism Program2015). Lack of any distinct tephra layers in the studied lake sediments and evidence of recent volcanism in the nearby area suggest that the source of the concentration might have originated from distant sources. Symonds et al. (1987) noted that the volcanic gases of Meropi Volcano in Indonesia contained a rich array of elements comprising Se, Re, Bi, Cd, Au, Br, In, Pb W, Mo, Cl, Cs, S, Sn, Ag As, Zn, F, Rb, Cu, K, Na, Sb, Ni, Ga, V, Fe and Mn as well as Li. It has also been noted (Siegel and Siegel

1987) that as a result of volcanic eruptions, Kilauea Volcano (Hawaii) produces 270 t of Hg gas every year, which can be transported as far as 320 km away from the source. Guedron et al. (2006) developed the equation given below to determine the origin of Hg accumulation in soil samples to see whether it originated from the decomposition of soil or was deposited as fallout due to atmospheric circulation:

Hg

½ _lithogenic;s¼ Hg½ p= k½ p* k½ s

where k represents the reference element and Nb, U, Zn, or Fe could be used as a reference element, whereasBs^

represents the sediment sample and Bp^ is the main reference material. In this study, Fe was used as the reference element because of lack of significant Fe enrichment in the sediments. Background values calcu-lated for Fe were used as reference.

The exogenic Hg value was estimated by subtracting the calculated lithogenic Hg value from the total Hg value. Following these calculations, the probable sources of Hg enrichment in the lake sediments are given in Fig.5. It is clear from the figure that although Hg concentration at St1 appears to be exogenic at the surface, lithogenic Hg shows an increasing trend down-ward as a percentage. At the lowermost interval, the exogenic Hg value also appears to be negative, probably due to over-estimation of lithogenic Hg resulting from low Hg concentration. This negative value occurred at only one of the sampling sites. At other stations, Hg concentration appears to be of exogenic origin.

Pollution load index

Pollution load index (PLI) values are shown in Fig.6. Suresh et al. (2011) note that the ideal PLI value is 0, a value which indicates no contamination. The PLI value 1 represents the base line for sediment, and values greater than 1 indicate higher contamination. In this respect, as is clear from the PLI values in Fig. 6, it is possible to state that in all stations, the contamination appears to be high in the surface sediments of the lake bottom.

At St1, while the highest PLI value occurs in 0–5 cm, this value is close to 1 in 5–10 cm; then, this value becomes closer to 0 at the lower levels, indicating a decrease in contamination. At St2, while the maximum value is reached in the uppermost 5 cm, the PLI value appears to decrease to 1 in the lower 5–10 cm, and then, f u r t h e r d o w n w a r d , i t f a l l s o u t s i d e t h e Fig. 4 Box and Whisker

diagrams of CF values obtained from Lake Çıldır sediments

contamination limit. At St3, the uppermost 20 cm is considered to be highly contaminated, whereas the contamination value is slightly above the base level in 20–25 cm and then becomes less than 1 at the lower levels. At St4, the highest PLI values are in the uppermost 10 cm. At St5, the uppermost 10 cm lie within PLI values indicating contamina-tion, but this value is slightly above 1 in the 10– 15-cm interval, which may be regarded as a tran-sitional zone of contamination. At St6, similar to other stations, contamination increases in the sur-face sediment with a particular concentration in the uppermost 10 cm.

Potential ecological risk

Potential ecological risk (PER) indices for each metal and their integrated risk indices are given in Figs.7and

8. Accordingly, the calculated Erivalues for Cu, Pb, Zn, Ni, As, and Cr are less than 40 (<40), and therefore, the PER indices for these metals are at a low level. It was determined that only two metals in the sediments, Cd and Hg, might constitute a potential risk in the lake. The risk potential of Cd at St1 is moderate in the upper 20 cm, but the highest risk potential becomes particu-larly evident in the uppermost 0–5 cm; at the lower levels, some increase of risk potential occurs in the intervals of 40–45, 50–55, and 65–68 cm, respectively. However, these potentials are not as significant as the uppermost one. The core sediments of St2 show mod-erate risk potential. At St3, while the potential risk of Cd reaches a considerable level in 0–5-cm interval, the level of potential risk is only moderate at the lower levels of the core. At St4, similar to St3, the surface sediment shows a considerable level of risk potential, but at the lower levels, except for the interval of 15–20 cm, only a moderate level of risk potential exists. St5 is also char-acterized by a considerable risk potential in the surface sediment which is, in turn, followed by a moderate risk down to 30 cm and then declines downward into a lower risk value. At St6, the core sediment shows a moderate risk while the highest risk potential was detected in the surface sediment.

The potential risk indices detected for Hg are higher than the Cd values. At St1, the considerable potential risk that occurs in the upper 10 cm is generally followed by a moderate level at the lower intervals. At stations 2, 3, 4, and 5, respectively, considerable-high levels of risk are noted throughout the core sediments. At St6, the

increasing risk factor at the surface sediment becomes less at the lower levels.

In analyzing the integrated potential ecological risks of the metals, a moderate level of risk is recognized in the upper 15-cm interval of St1, whereas moderate levels of risk are detected throughout core sediments at stations 2, 3, and 6. However, an increased risk level is noted in the core sediment at St4, where the risk level lies above the considerable level in its surface sediment but remains at a moderate level at the lower intervals. At St5, the risk level appears to be moderate in general, except in the interval of 20–25 cm, where the risk level is considerable.

Factor analysis

Factor analysis was performed to summarize a large number of variables together with a smaller number of factors in order not to lose too much information. Two factorial components, represented by eigenvalues great-er than 1, wgreat-ere distinguished. The first factor, which consists of Cu, Ni, Cr, Al, Hg, and organic carbon, accounts for 67.88 % of total variability. This represents transportation through alumina silicates (clay minerals) and organic carbon. The second component mainly consists of Pb, Zn, Mn, Fe, As, Cd, and CDP and accounts for 9.1 % of variability. This factor represents transportation through phytoplankton and iron compounds.

Alumina silicates are of terrestrial origin, and they can lose some of their characteristic trace elements in lake waters from wind- and wave-induced resuspension. Accumulation of Fe on the sediment surface originates from precipitation and oxidation of dissolved Fe spe-cies, atmospheric dust, and terrestrial sources, as well as from the mobilization of reduced iron from deeper sed-iment layers (Özkan2012).

Cluster analysis

Cluster analysis was performed to determine the relationships between variables and clusters of var-iables. The results of the cluster analysis indicate that Hg is an important metal for PER. The most important contribution for PLI results from Mn, while a cluster of Fe, Cr, Al, Ni, and Pb appear to be the other significant elements that contribute to PLI. A triple cluster of Fe, Cr, and Al suggests that they came from similar sources of terrestrial

origin. Ni and Pb are located near/next to the triple cluster. As, Cd, organic carbon, and CDP

show an entirely different clustering from the other compounds (Fig. 9).

Fig. 5 Probable sources of Hg enrichment in Lake Çıldır sediments

Fig. 6 Box and Whisker diagrams of PLI values obtained from Lake Çıldır sediments

Pearson’s correlation

Pearson’s correlation coefficient test was performed to evaluate the relationship between variables (Table 3). The test suggests that the majority of the metals form a close relationship with each other. However, the relation of As with other metals, except Zn, Mn, and Cr, is relatively low; among these metals, Mn has compara-tively the highest correlation coefficient. Although the relation between Zn and Hg, compared with other metals, appears to be slightly low, a strong correlation between Zn and other metals is apparent. The correla-tions of Ni with As and Hg are also relatively low. While Cd shows low correlation with As and Hg, it indicates high correlation with the others. In general, the correla-tion of the metals with organic carbon and CDP is also high, but, in exceptional cases, it appears that the rela-tion of organic carbon with Mn, As, and Cd, and the relation of CDP with Cu and As, is relatively low. However, in considering the correlation coefficient be-tween the metals and organic carbon and CDP, it appears that Cu, Ni, Cr, Al, and Hg are mainly related to organic carbon, whereas Pb, Zn, Mn, Fe, As, and Cd are mostly related to CDP. These results are consistent with results obtained from factor analysis. The highest contribution

to PER appears to be provided by Hg. The strong correlation between organic carbon and CDP indicates a phytoplankton contribution to the organic carbon.

One-way ANOVA analysis

One-way analysis of variance (ANOVA), used to test for significant variations between stations with regard to PLI, indicates some statistically significant differences (F, 12.26; p<0.05). Fisher’s least significant difference (LSD) test was also used to see at which specific station, or between which stations, significant differences existed. Station 3 located at the center of the lake formed a group both with northern stations 2, 5, and 6 and with station 4 to the south. In this respect, it is suggested that pollution is higher in the southern part of the lake and decreases northward, with station 3 representing a tran-sition zone.

Conclusions

Regarding CF values determined for Cu, Zn, Ni, Fe, Al, and Cr, it is interesting to note that the determined values are quite close to the critical value. Therefore, it may be Fig. 7 Box and Whisker

diagrams of Eifvalues obtained

from Lake Çıldır sediments

Fig. 8 Box and Whisker diagrams of PER values obtained from Lake Çıldır sediments

considered that CF values for these elements are not at significant levels. For the elements of Pb, As, and Cd, a moderate level of contamination was detected, whereas a moderate-to-high concentration level was obtained for Mn. The highest contamination level was determined for Hg. An atmospheric origin is considered to be the most likely source for Hg contamination, since the stud-ied area lies in an extensively volcanic region. There-fore, it may further be elaborated that even an enrich-ment of Hg concentration could be considered as normal in the lowermost layers of sediments deposited at the bottom of Lake Çıldır.

In terms of PLI values, a pollutant accumulation exists particularly in the surface sediments comprising the upper 10-cm interval of all stations. In considering the potential ecological risk (PER) indices of each ele-ment, the calculated Erivalues are <40 for Cu, Pb, Zn, Ni, As, and Cr, representing a low ecologic risk in the lake.

Cd and Hg are the only two metals considered to be a potential risk factor in the lake. Although a natural origin for Hg concentration in the lower layers of the lake bottom deposits is related to accumulation from Hg-rich dissolved gases released directly into the Fig. 9 Cluster dendogram of all

variables obtained from Lake Çıldır sediments

Table 3 Pearson’s correlation matrix of all variables

Cu Pb Zn Ni Mn Fe As Cd Cr Al Hg CDP ORGC PER PLI

Cu 1 Pb .759** 1 Zn .765** .795** 1 Ni .845** .826** .799** 1 Mn .585** .836** .688** .714** 1 Fe .828** .868** .816** .871** .863** 1 As .453** .501** .609** .480** .661** .639** 1 Cd .626** .782** .705** .692** .672** .744** .474** 1 Cr .850** .829** .797** .825** .688** .893** .566** .695** 1 Al .842** .851** .812** .888** .727** .949** .532** .705** .950** 1 Hg .651** .725** .460** .556** .588** .674** .345* .540** .668** .668** 1 CDP .445** .730** .555** .528** .659** .639** .386** .625** .628** .631** .630** 1 ORGC .652** .690** .532** .585** .369* .567** .160 .439** .727** .708** .744** .652** 1 PER .743** .829** .627** .671** .706** .794** .471** .725** .768** .762** .958** .682** .708** 1 PLI .807** .930** .873** .874** .886** .956** .699** .837** .890** .912** .663** .697** .588** .809** 1 ORGC organic carbon, CDP chlorophyll degradation products, PER potential ecological risk, PLI pollution load index

atmosphere during volcanic eruptions, a marked con-centration of Cd and Hg in the surface sediments of the lake bottom could only be interpreted in terms of an-thropogenic origin, resulting both from lignite burning as a home energy source in the studied area for most of the year and from the discharge of domestic waste directly into the lake from nearby settlements.

It appears that the only way to reduce present-day contamination is to take necessary steps for the preven-tion of household waste discharge into the lake.

Acknowledgments This study was supported financially by the Scientific and Technological Research Council of Turkey (TÜBİTAK; project number: 113Y205), for which we are thankful.

Compliance with ethical standards We as the authors are aware of research and publication ethics of the journal EMAS and declare that we have no competing interests.

References

Atalay, İ. (1978). Geomorphology of Çıldır Lake and around. Journal of Geomorphology, 7, 23–34.

Bai, J., Cui, B., Chen, B., Zhang, K., Deng, W., Gao, H., & Xiao, R. (2011). Spatial distribution and ecological risk assessment of heavy metals in surface sediments from a typical plateau lake wetland, China. Ecological Modelling, 222, 301–306. Barlas, N., Akbulut, N., & Aydoğan, M. (2005). Assessment of

heavy metal residues in the sediment and water samples of Uluabat Lake, Turkey. Bulletin of Environmental Contamination and Toxicology, 74, 286–293.

Bing, H., Wu, Y., Nahm, W. H., & Liu, E. (2013). Accumulation of heavy metals in the lacustrine sediment of Longgan Lake, middle reaches of Yangtze River, China. Environmental Earth Sciences, 69(8), 2679–2689.

Bodog, I., Polyak, K., & Hlavay, J. (1997). Determination of heavy metals in lake and river sediments by selective leaching. International Journal of Environmental Analytical Chemistry, 66(2), 79–94.

Çevik, F., Göksu, M. Z. L., Derici, O. B., & Fındık, Ö. (2009). An assessment of metal pollution in surface sediments of Seyhan dam by using enrichment factor, geoaccumulation index and statistical analyses. Environmental Monitoring and Assessment, 152, 309–317.

Demirsu, A. (1954). Çıldır-Posof-Şavşat-Kemalpaşa bölgesinin jeolojik etüdü hakkında mecmua. MTA Rapor No. 2377, Ankara.

Duman, F., Aksoy, A., & Demirezen, D. (2007). Seasonal vari-ability of heavy metals in surface sediment of Lake Sapanca, Turkey. Environmental Monitoring and Assessment, 133, 277–283.

Gaudette, H. E., Flight, W. R., Toner, L., & Folger, W. (1974). An inexpensive titration method for the determination of organic

carbon in recent sediments. Journal of Sedimentary Petrology, 44, 249–253.

Guedron, S., Grimaldi, C., Chauvel, C., Spadini, L., & Grimaldi, M. (2006). Weathering versus atmospheric contributions to mercury concentrations in French Guiana soils. Applied Geochemistry, 21, 2010–2022.

Guo, W., Liu, X., Liu, Z., & Li, G. (2010). Pollution and potential ecological risk evaluation of heavy metals in the sediments around Dongjiang Harbor, Ttianjin. Procedia Environmental Sciences, 2, 729–736.

Hakanson, L. (1980). Ecological risk index for aquatic pollution control, a sedimentological approach. Water Research, 14, 975–1001.

Hu, X., Wang, C., & Zou, L. (2011). Characteristics of heavy metals and Pb isotopic signatures in sediment cores collected from typical urban shallow lakes in Nanjing, China. Journal of Environmental Management, 92, 742–748.

Kamala-Kannan, S., Batvari, B. P. D., Lee, K. L., Kannan, N., Krishnamoorthy, R., Shanthi, K., & Jayaprakash, M. (2008). Assessment of heavy metals (Cd, Cr and Pb) in water, sedi-ment and seaweed (Ulva lactuca) in the Pulicat Lake, South East India. Chemosphere, 71, 1233–1240.

Kankılıç, G. B., Tüzün, İ., & Kadıoğlu, Y. K. (2013). Assessment of heavy metal levels in sediment samples of Kapulukaya Dam Lake (Kirikkale) and lower catchment area. Environmental Monitoring and Assessment, 185(8), 6739– 6750.

Karadede, H., & Ünlü, E. (2000). Concentrations of some heavy metals in water, sediment and fish species from the Atatürk Dam Lake (Euphrates), Turkey. Chemosphere, 41, 1371– 1376.

Kishe, M. A., & Machiwa, J. F. (2003). Distribution of heavy metals in sediments of Mwanza Gulf of Lake Victoria, Tanzania. Environment International, 28, 619–625. Kükrer, S.,Şeker, S., Abacı, Z. T., & Kutlu, B. (2014). Ecological

risk assessment of heavy metals in surface sediments of northern littoral zone of Lake Çıldır, Ardahan, Turkey. Environmental Monitoring and Assessment, 186, 3847– 3857.

Lahn, E. (1951). Bazı Türkiye Göllerinin Jeolojisi ve Jeomorfolojisi. MTA Dergisi, 41, 71–83.

Liu, E., Shen, J., Yang, L., Zhang, E., Meng, X., & Wang, J. (2010). Assessment of heavy metal contamination in the s e d i m e n t s o f N a n s i h u L a k e C a t c h m e n t , C h i n a . Environmental Monitoring and Assessment, 161, 217–227. Lorenzen, C. J. (1971). Chlorophyll-degradation products in

sed-iments of Black Sea. Woods Hole Oceanographic Institution Contribution, 28, 426–428.

Luo, W., Lu, Y., Wang, T., Hu, W., Jiao, W., Naile, J. E., Khim, J. S., & Giesy, J. P. (2010). Ecological risk assessment of arsenic and metals in sediments of coastal areas of northern Bohai and Yellow Seas, China. AMBIO, 39, 367–375. MacDonald, D. D., Ingersoll, C. G., & Berger, T. A. (2000).

Development and Evaluation of Consensus-Based Sediment Quality Guidelines for Freshwater Ecosystems. Archives of Environmental Contamination and Toxicology, 39, 20–31. Makundi, I. N. (2001). A study of heavy metal pollution in Lake

Victoria sediments by energy dispersive x-ray fluorescence. Journal of Environmental Science and Health, Part A: Toxic/ Hazardous Substances and Environmental Engineering, 36(6), 909–921.

Ochieng, E. Z., Lalah, J. O., & Wandiga, S. O. (2007). Analysis of Heavy Metals in Water and Surface Sediment in Five Rift Valley Lakes in Kenya for Assessment of Recent Increase in Anthropogenic Activities. Bulletin of Environmental Contamination and Toxicology, 79, 570–576.

Özkan, E. Y. (2012). A new assessment of heavy metal contamina-tions in an eutrophicated bay (Inner Izmir Bay, Turkey). Turkish Journal of Fisheries and Aquatic Sciences, 12, 135–147. Özmen, H., Külahcı, F., Çukurovalı, A., & Doğru, M. (2004).

Concentrations of heavy metal and radioactivity in surface water and sediment of Hazar Lake (Elazığ, Turkey). Chemosphere, 55, 401–408.

Rippey, B., Rose, N., Yang, H., Harrad, S., Robson, M., & Travers, S. (2008). An assessment of toxicity in profundal lake sedi-ment due to deposition of heavy metals and persistent organic pollutants from the atmosphere. Environment International, 34, 345–356.

Rosales-Hoz, L., Carranza-Edwards, A., & Lopez-Hemandez, M. (2000). Heavy metals in sediments of a large, turbid tropical lake affected by anthropogenic discharges. Environmental Geology, 39(3–4), 378–383.

Siegel, B. Z., & Siegel, S. M. (1987). Hawaiian volcanoes and the biogeology of mercury. In R.W. Decker, T.L. Wright and P.H. Stauffer (eds.), Volcanism in Hawaii (pp. 827–839). U.S. Geological Survey Professional Paper 1350, v. 1.

Smithsonian Institution National Museum of Natural History Global Volcanism Program.http://www.volcano.si.edu/. Accessed 1 June 2015.

Suresh, G., Ramasamy, V., Meenakshisundaram, V., Venkatachalapathy, R., & Ponnusamy, V. (2011). Influence of mineralogical and heavy metal composition on natural

radionuclide contents in the river sediments. Applied Radiation and Isotopes, 69, 1466–1474.

Suresh, G., Sutharsan, P., Ramasamy, V., & Venkatachalapathy, R. (2012). Assessment of spatial distribution and potential eco-logical risk of the heavy metals in relation to granulometric contents of Veeranam lake sediments, India. Ecotoxicology and Environmental Safety, 84, 117–124.

Sutherland, R. A. (2000). Bed sediment associated trace metals in an urban stream, Oahu. Hawaii Environmental Geology, 39, 611–627.

Symonds, R. B., Rose, W. I., Reed, M. H., Lichte, F. E., & Finnegan, D. L. (1987). Volatilization, transport and subli-mation of metallic and non-metallic elements in high tem-perature gases at Merapi Volcano, Indonesia. Geochimica et Cosmochimica Acta, 51(8), 2083–2101.

Tchalenko, J. S. (1977). Reconnaissance of seismicity and tecton-ics at northern border of Arabian plate (Lake Van region). Revue de Géographie Physique et de Géologie Dynamique, 19, 89–207.

Vrhovnik, P., Smuc, N. R., Dolenec, T., Serafimovski, T., & Dolenec, M. (2013). An evaluation of trace metal distribution and environmental risk in sediments from the Lake Kalimanci (FYR Macedonia). Environmental Earth Science, 70, 761–775.

Wang, Y., Hu, J., Xiong, K., Huang, X., & Duan, S. (2012). Distribution of heavy metals in core sediments from Baihua Lake. Procedia Environmental Sciences, 16, 51–58. Zhang, L., Ye, X., Feng, H., Jing, Y., Ouyang, T., Yu, X., Liang,

R., Gao, C., & Chen, W. (2007). Heavy metal contamination in western Xiamen Bay sediments and its vicinity, China. Marine Pollution Bulletin, 54, 974–982.