Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques

STOTEN-19906; No of Pages 11 Science of the Total Environment xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques Golshan Shirneshan a, Alireza Riyahi Bakhtiari a,⁎, Mahmoud Memariani b a b

Department of Environmental Sciences, Faculty of Natural Resources and Marine Science, Tarbiat Modares University, P.O. Box 46414-356, Noor, Mazandaran, Iran Geosciences Division, Research Institute of Petroleum Industry, Tehran, Iran

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Multi-marker approach efficiently distinguished the source of tar balls. • Turkmenistan oil was identified as the most probable source of the tar balls. • Photo-oxidation, biodegradation and evaporation occurred in tar balls.

Results of PCA of the 8 source-specific biomarker parameters versus oil samples from Turkmenistan, Azerbaijan and Anzali and 6 tar ball samples.

a r t i c l e

i n f o

Article history: Received 23 December 2015 Received in revised form 29 April 2016 Accepted 30 April 2016 Available online xxxx Editor: D Barcelo Keywords: Tar balls Fingerprint Biomarkers Alkanes Hopanes Caspian Sea

a b s t r a c t In 2012, a significant number of tar balls occurred along the Southwest coasts of the Caspian Sea (Iran). Several oil fields of Turkmenistan, Azerbaijan and Iran might be sources of oil spills and lead to the formation of these tar balls. For source identification, 6 tar ball samples were collected from the Southwest beaches of the Caspian Sea and subjected to fingerprint analysis based on the distribution of the source-specific biomarkers of pentacyclic tri-terpanes and steranes. Comparing the diagenic ratios revealed that the tar balls were chemically similar and originated from the same source. Results of double ratio plots (e.g., C29/C30 versus ∑C31–C35/C30 and C28 αββ/(C27 αββ + C29 αββ) versus C29 αββ/(C27 αββ + C28 αββ)) in the tar balls and oils from Iran, Turkmenistan and Azerbaijan indicated that the tar balls might be the result of spills from Turkmenistan oil. Moreover, principle component analysis (PCA) using biomarker ratios on the tar balls and 20 crude oil samples from different wells of Azerbaijan, Iran and Turkmenistan oils showed that the tar balls collected at the Southwest beaches are highly similar to the Turkmenistan oil but one of the Azerbaijan oils (from Bahar field oils) was found to be also slightly close to the tar balls. The weathering characterizations based on the presence of UCM (unresolved complex mixture) and low/high molecular weight ratios (L/H) of alkanes and PAHs indicated the tar ball samples have been significantly influenced by natural weathering processes such as evaporation, photo-degradation and biodegradation. This is the first study of its kind in Iran to use fingerprinting for source identification of tar balls. © 2016 Elsevier B.V. All rights reserved.

⁎ Corresponding author. E-mail address: [email protected] (A.R. Bakhtiari).

http://dx.doi.org/10.1016/j.scitotenv.2016.04.203 0048-9697/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

2

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Fig. 1. Locations of Turkmenistan, Azerbaijan oil field, Anzali well and the sampling sites of tar balls collected from the Southwest coast of Caspian Sea.

1. Introduction The Southwest coast of the Caspian Sea is under the threat of petroleum pollution. In 2009 and 2012, this coast received about 14 and 3 tones of tar balls, respectively. Deposition of tar balls attracts a huge concern for the residents living in this region. Since these tar balls affect the health of the coastal waters and threat the marine life, Iran's Environmental Protection Organization has a major concern about this environmental problem. Therefore, tracing of the sources of tar balls is an important environmental management problem. Tar balls may originate from many petroleum sources. They can be formed during the weathering of oil films in the sea mostly derived from tanker washing and routine shipping operations (Clark, 2002). Operational losses of fossil fuel hydrocarbons from drilling, petroleum platforms and tanker derived oil spills are also the main contributors in tar ball generation (Asuquo, 1991). The Caspian Sea, the biggest enclosed body of water on the Earth bordered by five countries namely Kazakhstan, Azerbaijan, Turkmenistan, Russia and Iran, has attracted the attention of the international oil and gas industry, especially since the break-up of the Soviet Union in 1991 (Effimoff, 2000). The nearest possible spill sources to Iran coast in the Caspian Sea are offshore oil fields located off Azerbaijan and Turkmenistan. The Southwest coast of the Caspian Sea is covered by the majority of oil fields in Azerbaijan where petroleum exploration and

production has been continuing since the early 1900s. GuneshliChirag-Azeri (GCA), Bahar, and Gum Adasi fields are the most important oil fields in Azerbaijan. Currently, proved reserves of these oil fields are estimated at 18–35 billion barrels. This significant production and exploration could cause the Southwest beaches in Iran to be highly susceptible for stranding of tar balls originating from Azerbaijan. Also there are potential oil resources in Turkmenistan, 30–40% of which are located offshore (Tolosa et al., 2004). Petroleum activities from these areas could have caused some spills which led to the formation of tar balls. Another possible source of tar balls in the Southwest coasts of the Caspian Sea can be Anzali oil well located 30 km away from Northwest of Anzali Port (Fig. 1). This well is an offshore well drilled and completed in April 1991 in Southwest region of Caspian Sea. In spite of existence of Table 1 Positions and sampling date and codes of tar balls collected from the Southwest coast of Caspian Sea. Station

Location

Latitude (N)

Longitude (E)

Sand content (%)

TB1 TB2 TB3 TB4 TB5 TB6

Kiashahr City Asgarabad City Chaf City Chamkhaleh City Hasansara City Kalachy City

37°28′35.74″ 37°24′43.58″ 37°21′30.05″ 37°15′54.06″ 37°8′32.69″ 37°4′21.56″

49°56′20.40″ 50°3′58.38″ 50°13′42.49″ 50°15′9.03″ 50°19′9.70″ 50°26′5.46″

5 7 8 6 9 5

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx Table 2 Compounds of standard mixtures of PAH, hopane and sterane. Compounds of PAH standard mixtures

Abbreviation

Naphthalene 1M-Naphthalene 2M-Naphthalene 2,6DM-Naphthalene Acenaphthylene Acenaphthene 2,3,5TM-Naphthalene Fluorene Dibenzothiophen Phenanthrene Anthracene 3-Methylphenanthrene 2-Methylphenanthrene 9-Methylohenanthracene 1-Methylphenanthrene 3,6DM-Phenanthrene Fluoranthene Pyrene Benzo(a)fluorine Benzo(a)anthracene Chrysene Benzo(b)fluoranthene Benzo[k]fluoranthene Benzo(a)pyrene Benzo(e)pyrene Perylene Indeno[1,2,3-cd]pyrene Dibenzo[a,h]anthracene Benzo[ghi]perylene

Na 1MNa 2MNa DMNa Acy Ace TMNa Fl DBT P Ant 3MP 2MP 9MP 1MP DMP Flu Py BaF BaA Chy BbF BkF BaP BeP Per IP DBA BP

Compounds of hopane standard mixtures 17β(H),21α(H)- norhopane 17α(H)-22,29,30-trisnorhopane 17β(H),21α(H)-hopane

C29 17βα Tm C30 17βα

Compounds of sterane standard mixtures 5α (H)-cholestane 24-methyl-5α (H)-cholestane 24-ethyl-5α (H)-cholestane

C27α C28α C29α

3

some physical characteristics of transported oil and its products is important for post-spill assessments (Radović et al., 2012). These ratios were used to categorize the sources and weathering of stranded tar balls (Kvenvolden et al., 1993; Wang et al., 1995; Zakaria et al., 2000, 2001; Suneel et al., 2013). Kvenvolden et al. (1993) used a collection of molecular markers consisting of terpanes, steranes and monoaromatic sterenes for determining the source of oil residues from the beaches of six islands in Prince William Sound, Alaska. They found that 6 out of the 8 residues from the mentioned beaches have originated from the Exxon Valdez oil spill in 1989. Wang et al. (1998), using terpanes, steranes and polycyclic aromatic hydrocarbons, determined that tar ball samples collected from the coasts of Vancouver Island, Canada, and Northern California, USA, have been highly weathered, and that none of the samples had come from Alaska North Slope oil or California Monterrey Miocene oil. Chandru et al. (2008) based on the alkane, hopane and PAH data, found that tar ball stranded on coastal beaches in peninsular Malaysia originated from Middle East crude oil and had undergone different phases of weathering and biodegradation. Suneel et al. (2015) used a multi-criteria approach of hopanes, pentacyclic terpanes and regular steranes for fingerprinting the tar ball samples collected from 9 beaches of Goa coast. They showed that the multi-criteria approach was effective in identifying the source of tar balls. Despite that deposition of high volume tar balls has become a serious issue along the Southwest coast of the Caspian Sea, no investigation has been carried out to understand the chemical characteristics of the tar balls. The aims of this study were to apply the fingerprint techniques of diagnostic ratios of hopanes and steranes biomarkers using principle component analysis (PCA) to identify the sources of tar balls deposited along the Southwest coast of the Caspian Sea. The alkane and PAH content of tar balls were also analyzed to discover the extent of weathering in the samples.

2. Materials and methods 2.1. Sample collection

good reservoir quality sediments, only minor quantities of oil and gas were collected. Tracing the sources of tar balls is highly important and essential. Several studies have been conducted worldwide for tracing the source of tar balls deposited along the beaches since the 1970s (Nair et al., 1972; Dhargalkar et al., 1976; Macko and Parker, 1983; Zakaria et al., 2001; Diez et al., 2007; Mulabagal et al., 2013; Suneel et al., 2015). Fingerprinting of biomarkers is a potent tool for identifying the source of petroleum-derived contaminants in the marine environments (Wang et al., 2007). In the past decade, biomarker compounds such as terpanes, steranes and polycyclic aromatic hydrocarbons (PAHs) have been proposed to forensic investigations on some oil spill accidents (Boehm et al., 2001; Wang et al., 2006; Mulabagal et al., 2013; Bernabeu et al., 2013). Investigating the diagnostic ratios of these compounds and

In total, six tar ball samples, representing six locations in the Southwest Caspian Sea, were used for analysis in the present study (Fig. 1 and Table 1). The tar ball samples were collected by hand using clean plastic gloves in August 2012. The samples were kept in clean aluminum foil packets and transported to the laboratory for further analysis. In order to consider whether the collected tar balls were floating or sinking in the water column, we took a subsample and dropped into the beaker of sea water with a salinity of 13 psu and temperature of 29 °C (Suneel et al., 2015). Based on our observations, all the tar balls were floating at the surface. Two crude oil samples from the Anzali oil well in Iran and Turkmenistan were provided by the Research Institute of Petroleum Industry in Iran. An oil sample from Azerbaijan was also included in the

Table 3 Data of alkanes, hopanes and steranes in tar ball and oil samples.

Azerbaijan oil Turkmenistan oil Anzali oil TB1 TB2 TB3 TB4 TB5 TB6

CPI

n-C18/ phytane

n-C17/ pristane

L/H-alkane

C23/ C30H

Ts/Ts + Tm

OL/C30

C29H/ C30H

C31HS/C31H (S + R)

∑C31_ C35/C30

Sterane C28αββ/ (C27 αββ + C29 αββ)

Sterane C29 αββ/ (C27 αββ + C28 αββ)

0.98 0.99 1.1 1.02 1.03 1.02 0.99 1.01 1.02

1.53 1.19 1.77 0.063 0.07 0.08 0.11 0.082 0.09

1.34 1.22 1.46 0.18 0.22 0.15 0.32 0.12 0.27

3.91 4.64 3.61 0.81 0.89 1.04 0.92 0.72 0.78

0.06 0.055 0.14 0.052 0.053 0.055 0.054 0.054 0.053

0.47 0.46 0.52 0.48 0.48 0.48 0.48 0.48 0.48

0.075 0.062 0.14 0.061 0.062 0.063 0.061 0.062 0.064

0.44 0.39 0.41 0.39 0.40 0.39 0.40 0.39 0.40

0.56 0.6 0.56 0.61 0.61 0.6 0.62 0.61 0.62

1.3 1.52 1.34 1.52 1.53 1.51 1.52 1.51 1.51

0.53 0.47 0.55 0.47 0.48 0.47 0.47 0.46 0.47

0.65 0.73 0.61 0.72 0.72 0.73 0.72 0.73 0.72

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

4

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

analysis. Also we used the data of 20 crude oil samples from different wells of Azerbaijan reported in the study by Gurgey (2003).

2.2. Estimation of oil percentage in the tar ball samples To determine the amount of residual oil fraction remaining in the tar ball samples, about 1 g of each tar ball was extracted with 10 ml of dichloromethane four times. The remaining solid (sand) fraction was dried and weighted. The average value of oil content in the tar ball samples varied from 91% to 95% (Table 1).

2.3. Sample preparation Twenty mg of the tar ball and crude oil samples were accurately weighed and dissolved in 2 ml of DCM/n-hexane (1:3, v/v) together with 200 μl of PAH SIS (the surrogate internal standard). The surrogate internal standard for PAHs consists of 5 ppm of each of naphthalene-d8, acenaphthene-d10, phenanthrene-d10, chrysene-d12 and perylene-d4. The solution was purified and fractionated by the methods described in Zakaria et al. (2000). The solution was transferred on the top of a 5% H2O deactivated silica gel chromatography column. The alkanes, PAHs and biomarkers were eluted with 20 ml of DCM/n-hexane (1:3, v/v). The

Fig. 2. Terpane patterns in Turkmenistan, Azerbaijan and Anzali oils and six tar ball samples collected from the Southwest coast of Caspian Sea (m/z = 191). Ts: 18α(H),21β(H),22,29,30trisnorhopane; Tm: 17α(H),21β(H),22,29,30-trisnorhopane; C29: 17α(H),21β(H)-norhopane; C30: 17α(H),21β(H)-hopane; C31 to C35: Homohopanes consisting of C31 to C35 carbons with S and R stereoisomer at 22 carbon, respectively; IIS: Internal Injection Standard (17β(H),21βhopane).

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

DCM/n-hexane eluate was further fractionated with fully activated silica gel column chromatography to obtain aliphatic and PAH fractions. The alkanes and biomarkers (terpanes and steranes) were eluted with 4 ml of hexane. The PAHs were eluted with 14 ml of dichloromethane/ hexane (1:3, v/v). The fractions were evaporated and transferred to a 1.5 ml glass ampoule and blow-down with N2 to almost dryness. All compounds were determined by GC–MS, using an Agilent Technologies (Palo Alto, CA, USA) instrument having a gas chromatograph

5

(GC), model 7890A equipped with a fused silica capillary HP-5MS column (30 m × 0.25 mmi.d., 0.25 μm film thickness) with splitless injection and the following instrument parameters for biomarkers: initial temperature, 70 °C (maintained for initial 1 min) with ramps of 30 °C/min to 150 °C and then with ramps of 5 °C/min to 290 °C and held for 10 min. The injector's temperature was 290 °C. For PAHs, the oven temperature was kept at 70 °C during the injection (for 2 min). Then the temperature was increased at a rate of 30 °C/min to 150 °C.

Fig. 3. Steranes patterns in Turkmenistan, Azerbaijan and Anzali oils and six tar ball samples collected from the Southwest coast of Caspian Sea (m/z = 218). C27 R: C27 5α(H),14β(H),17β(H)-cholestane (20R); C27 S: C27 5α(H),14β(H),17βH)-cholestane (20S) C28 R: C28 5α(H),14β(H),17β(H)-ergostane (20R); C28 S: C28 5α(H),14β(H),17β(H)-ergostane (20S); C29 R: C29 5α(H),14β(H),17β(H)-stigmastane (20R); C29 S: C29 5α(H),14β(H),17β(H)-stigmastane (20S).

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

6

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Fig. 4. C29/C30 vs. ∑C31–C35/C30 cross-plot diagram for tar ball and oil samples.

Next the temperature was raised to 290 °C at a rate of 4 °C/min and was fixed at 290 °C for 10 min. The injector's temperature was 300 °C. Helium was used as the carrier gas. Characteristic ions were analyzed in single ion monitoring (SIM) mode: m/z 191 for tri-and and

pentacyclic terpanes, m/z 217 for ααα-steranes and m/z 218 for αββsteranes. P-Terphenyl-d14 was used as an Internal Injection Standard (IIS). The n-alkanes were identified by comparison of the retention times with those of known standards of n-alkanes ranging from n-C14

Fig. 5. Double ratio plots using biomarkers. C28 αββ/(C27 αββ + C29 αββ) vs. C29 αββ/(C27 αββ + C28 αββ).

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

7

Fig. 6. Results of PCA of the 8 source-specific biomarker parameters versus oil samples from Turkmenistan, Azerbaijan and Anzali and 6 tar ball samples.

to n-C32. The standard mixtures of hopane, sterane and PAH used are shown in Table 2. A ten-point calibration curve of different concentrations of PAHs in the range of 10 to 50,000 ng/ml was used. For hopanes, the calibration solution concentrations were 20, 50, 100, 200, 500 and 1000 ng/ml. The concentration levels used for the calibration curves of steranes were: 10, 50, 100, 200 and 500 ng/ml. For the Internal Injection Standard (IIS) purposes, 17β, 21(H)β-hopane and n-tetracosane-d50 were used for biomarkers and n-alkanes, respectively. All authentic standards for alkanes, PAHs, hopanes and steranes were purchased from Chiron, Norway. Recoveries were calculated by spiking a known concentration of SIS mixture into the sample followed by performing the entire analytical procedure. The Recoveries of individual spiked SIS were more than 83% for alkanes and more than 87% for hopanes. The recoveries for PAHs, ranging from 81% to 86%, were used for the recovery correction calculations.

3. Results and discussion 3.1. Sources of tar ball samples A total of 6 tar ball samples were collected from the Southwest coast of the Caspian Sea. Three crude oil samples were analyzed using the molecular marker approach to fingerprint the source of the tar balls. In addition, individual n-alkanes (C12–C36), isoprenoid hydrocarbons (pristane and phytane) and PAHs were investigated to characterize the tar ball samples. The chemical composition of a tar ball can give primary information on its origin. A CPI (carbon preference index, defined as the sum of odd over even-carbon-number n-alkanes) is often used to

2.4. Determination of PAH depletion levels For estimating the degree of weathering of a target PAH, the following equation was used: Actual % depletion of PAH = (1- PAH in weathered sample/PAH in reference oil∗(Hoil/H weathered)∗100 where, Hoil and H weathered are the concentrations of C30αβ hopane in the source oil and weathered sample, respectively (Radović et al., 2012; Yin et al., 2015). C30αβ-hopane was used as a conservative reference for normalization. 2.5. Statistical analysis The source identification of the tar ball samples was also performed by using principal component analysis (PCA). PCA is a multivariate a statistical procedure that is widely used in the interpretation of oil spill fingerprinting (Suneel et al., 2014). Other statistical test applied was the clustering test. Cluster analysis with relative Euclidean distances and Ward's linkage method was conducted using the petroleum biomarker ratios to classify the samples in cluster groups. Cluster analysis divides the data into groups (clusters) that share similar values across a number of variables (Silva and Bicego, 2010). Cluster analysis was applied after PCA in order to mark the oil samples with higher similarity to the tar ball samples. All statistical evaluations were performed using SPSS 15 for windows and PC ORD 6.

Fig. 7. Results of PCA of the 11 source-specific biomarker parameters versus 20 oil samples from Azerbaijan (Gurgey, 2003), oil samples from Turkmenistan, Azerbaijan and Anzali and 6 tar ball samples. Oil Samples clustering with tar balls are highlighted.

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

8

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

estimate the thermal maturity of the petroleum, and it is normally 1.0 for oil and petroleum products (Wang et al., 1998). CPI values of the present samples were near 1.0 (0.98 to 1.02) (Table 3). This indicates that the present tar ball samples contain petroleum. The distribution chromatograms of highly degradation-resistant biomarker terpanes at m/z 191 for the tar ball samples are presented in Fig. 2. It shows that C27-C35 pentacyclic terpanes (e.g. hopanes and homohopanes) were more abundant than the lower molecular weight C19-C29 tricyclic terpanes, as reflected by the low C23 tricyclic/C30 hopane pentacyclic ratios. Predominance of C30 hopane over C29 norhopane, which is a typical characteristic of oils derived from clay-rich source rocks, was observed in all the tar balls. The concentrations of Ts, Tm and C31–C35 22S/22R homohopane epimers in these samples are relatively low compared to the C30 αβ-hopane levels. A peculiarity of the m/z 191 terpane chromatograms of all tar ball samples is the presence of 18α (H) oleanane. This compound which is usually observed in the crude oils produced by plant materials deposited in the deltaic environment, is an indicator of Tertiary or late Cretaceous source rock with terrigenous influence (Ekweozor et al., 1979). This specific property is helpful in identifying the source of anonymous oil spills. The chromatographic fingerprints of steranes in the oil and tar ball samples were also developed using a GC/MS method acquired in SIM mode (m/z of 217 and 218) (Fig. 3). For steranes, in addition to C21 and C22 5α(H),14β(H),17β(H)-sterane, the dominance of C27, C28, and C29 20S/20R steranes, particularly the

epimers of αββ-steranes, is apparent in the tar ball samples. Although the above results show that the tar balls contained C30 αβ-hopane and C27, C28 and C29 20S/20R αββ-steranes as the major biomarkers, crude oils can vary in their hopane and sterane ratios and have a unique source-specific fingerprint. To demonstrate the distribution patterns of source-specific biomarkers, C23/C30 (ratio of C23 tricyclic terpane relative to 17α,21β(H)-hopane), Ts/Ts + Tm (ratio of 17α-22,29,30trisnorhopane relative to 17α-22,29,30-trisnorhopane + 18α22,29,30-trisnorhopane), C29/C30 (ratio of 17α,21β(H)-30 norhopane to 17α,21β(H)-hopane), C31HS/C31H(S + R), C32HS/C32H(S + R), ∑ C31_C35/C30 (ratio of sum 17α,21β(H)–C31 homohopane to 17α,21β(H)–C35 homohopane relative to 17α,21β(H)-hopane), C28 αββ/(C27 αββ + C29 αββ) and C29 αββ/(C27 αββ + C28 αββ) ratios were calculated for all the tar ball samples (Table 3). These ratios have been widely used for the purpose of source identification and differentiation of oils (Bence and Burns, 1995; Wang et al., 1994; Wang and Fingas, 1997; Wang et al., 1998). Similarity of these data shows that the relative distributions of various steranes and terpanes in the tar ball samples are rather similar, showing that they have originated from the same source. Identifying the source of tar ball samples were distinguished using the cross-plot diagram of the diagnostic ratios of C29/C30 versus ∑ C31_C35/C30 for hopanes and C28 αββ/(C27 αββ + C29 αββ) versus C29 αββ/(C27 αββ + C28 αββ) for steranes between all the tar ball and oil samples of Turkmenistan, Azerbaijan and

Fig. 8. Full-scan mass chromatograms of aliphatic hydrocarbons in six tar ball samples collected from the Southwest coast of Caspian Sea. The scan mass range was 50–500 a.m.u.

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

9

Table 4 Concentration of parent PAHs, alkylated PAHs and hopanes (μg/g) in tar balls collected from the Southwest coast of Caspian Sea.

Naphthalene 1M-Naphthalene 2M-Naphthalene 2,6DM-Naphthalene Acenaphthylene Acenaphthene 2,3,5TM-Naphthalene Fluorene Dibenzothiophen Phenanthrene Anthracene 3-Methylphenanthrene 2-Methylphenanthrene 9-Methylohenanthracene 1-Methylphenanthrene 3,6DM-Phenanthrene Fluoranthene Pyrene Benzo(a)fluorine Benzo(a)anthracene Chrysene Benzo(b)fluoranthene Benzo[k]fluoranthene Benzo(a)pyrene Benzo(e)pyrene Perylene Indeno[1,2,3-cd]pyrene Dibenzo[a,h]anthracene Benzo[ghi]perylene ∑PAH (μg/g) L/H-PAH ∑Hopane (μg/g)

Azerbaijan oil

Turkmenistan oil

Anzali oil

TB1

TB2

TB3

TB4

TB5

TB6

52.5 118.8 96.9 144.4 59.9 4.4 87.5 42.4 37.6 260.5 29.4 45.7 48.7 158.5 54.2 36.0 6.4 9.3 9.5 3.8 4.1 3.5 3.0 4.3 9.9 4.9 8.6 3.7 5.9 1336.5 17.7 202.3

84.5 174.8 176.9 164.4 89.9 1.9 77.5 32.4 18.1 200.5 34.1 77.0 76.5 126.5 90.9 26.0 4.4 18.2 3.5 2.8 5.0 2.4 4.9 5.7 24.3 1.0 4.1 3.7 7.7 1539.8 16.7 287.3

74.7 79.5 76.9 336.9 311.9 176.7 913.5 421.1 130.3 1466.9 25.9 639.6 717.9 1170.8 784.1 191.8 23.5 34.9 63.8 2.8 61.1 4.9 28.0 74.9 34.3 2.0 5.1 6.3 16.1 7876.2 22.6 209.2

0.9 3.0 11.9 63.2 66.6 1.6 70.0 26.0 17.7 90.3 35.2 68.5 40.6 19.9 81.9 21.5 4.1 28.3 5.7 4.5 8.7 3.6 5.3 5.1 30.1 1.6 4.3 3.3 9.8 733.4 5.6 672.2

0.4 1.4 6.7 30.3 31.9 1.6 57.5 17.7 11.2 80.0 22.4 61.3 19.5 9.6 39.3 10.3 3.5 21.9 5.4 3.9 9.0 2.7 3.4 3.0 26.0 1.2 2.9 2.2 7.9 494.0 4.5 735.4

0.5 1.8 8.2 37.3 39.3 1.6 63.2 21.8 13.8 92.0 27.6 60.4 23.9 11.8 48.3 12.7 3.8 26.9 5.7 4.2 8.8 3.3 4.1 3.7 27.6 1.4 3.6 2.8 8.1 568.0 4.7 661.7

1.1 2.3 13.4 75.1 80.4 1.7 70.2 26.1 17.0 91.3 35.7 73.0 40.5 24.1 98.9 26.0 4.3 28.1 5.7 4.5 8.7 3.7 5.6 5.6 31.0 1.7 4.4 3.5 10.0 793.5 6.1 713.0

0.6 2.0 9.2 41.9 44.1 1.6 67.9 24.5 15.5 91.3 31.0 64.8 26.9 13.2 54.3 14.3 3.9 28.3 5.6 4.3 9.0 3.7 4.4 4.1 28.1 1.6 3.8 2.9 8.7 611.4 4.8 692.1

0.7 2.3 10.6 48.0 50.5 1.6 65.4 24.7 15.4 90.1 35.5 65.1 30.8 15.1 62.1 16.3 3.9 28.0 5.8 4.3 8.9 3.6 4.8 4.5 27.6 1.6 4.0 3.0 9.8 644.2 5.1 746.8

Anzali (Figs. 4 and 5). The double ratio plot used in some studies can be a reliable source identifier (Yim et al., 2011; Suneel et al., 2014; Zakaria et al., 2001). In both of the double ratio plots using biomarkers, Turkmenistan crude oil falls among the tar ball samples. Table 5 Percentage depletion of parent PAHs and alkylated PAHs in tar balls collected from the Southwest coast of Caspian Sea.

Naphthalene 1M-Naphthalene 2M-Naphthalene 2,6DM-Naphthalene Acenaphthylene Acenaphthene 2,3,5TM-Naphthalene Fluorene Dibenzothiophen Phenanthrene Anthracene 3-Methylphenanthrene 2-Methylphenanthrene 9-Methylohenanthracene 1-Methylphenanthrene 3,6DM-Phenanthrene Fluoranthene Pyrene Benzo(a)fluorine Benzo(a)anthracene Chrysene Benzo(b)fluoranthene Benzo[k]fluoranthene Benzo(a)pyrene Benzo(e)pyrene Perylene Indeno[1,2,3-cd]pyrene Dibenzo[a,h]anthracene Benzo[ghi]perylene

TB1

TB2

TB3

TB4

TB5

TB6

100 100 97 80 62 55 54 59 50 77 47 55 73 92 54 58 52 21 18 17 12 24 45 54 37 19 46 54 35

100 100 98 92 84 61 66 75 72 82 70 63 88 97 80 82 63 45 30 36 18 49 68 76 51 48 67 72 53

100 100 98 89 79 58 61 68 64 78 61 62 85 96 74 77 59 29 23 27 16 35 60 69 46 34 58 64 49

100 100 96 76 53 52 53 58 51 76 53 51 72 90 43 48 50 20 16 16 11 20 41 49 34 15 44 50 32

100 100 98 88 77 59 59 65 60 79 57 60 83 95 72 74 58 27 25 26 16 28 58 66 46 27 56 63 47

100 100 97 86 72 57 59 63 58 78 45 59 80 94 66 69 56 25 20 23 14 26 52 61 44 23 52 60 38

To better separate the characteristics of the reference oils of Anzali, Turkmenistan, Azerbaijan and the tar ball samples, a PCA was carried out using 8 parameters viz. C23/C30, OL/C30 (18α(H)-oleanane/C30 17α,21β-hopane), Ts/Ts + Tm, C29/C30, C31S/(C31S + R), ∑ C31_C35/ C30, C28 αββ/(C27 αββ + C29 αββ) and C29 αββ/(C27 αββ + C28 αββ). The aim was to identify the source of those samples through a more reliable evaluation using a multi-criteria approach. Two factors that were chosen as the principal factors explained about 93.99% of the total variability in the diagenic ratios of the tar ball and crude oil samples. The biplot between these two components is shown in Fig. 6. It is observed that the Anzali and Azerbaijan crude oils are far away from the remaining samples. This clearly suggests that the Anzali and Azerbaijan crude oils cannot be the source candidates. Evidently, the tar balls and Turkmenistan crude oil fell close together. This strongly suggests that the tar ball samples were originated from the Turkmenistan oil. Since there are many oil wells in Azerbaijan and we obtained only a sample from one of its oil wells, for further investigation, we used data from the study by Gurgey (2003) who categorized 20 sample oils from Azerbaijan wells using PCA by 11 parameters. The 20 sample oils included G-1, G-2, G-3, G-4, C-1, C-2, C-3 and A-1 oils from GuneshliChirag-Azeri, B-1, B-2, B-3, B-4, B-5 oils from Bahar, and GA-1, GA-2, GA-3, GA-4, GA-5, GA-6, GA-7 oils from Gum Adasi fields. The 11 parameters were compared in 6 tar ball and 3 oil samples of this study and above mentioned 20 sample oils. These source-specific parameters were C23/C30, C24/C26 (C24 tetracyclic terpane/C26 tricyclic terpane), OL/C30, Ts/Ts + Tm, C30 ∗ H/C29Ts (17 α(H)-diahopane/18 α(H)30-norneohopane), HHI (homohopane index, (C35/ C35 + C34 + C33 + C32 + C31), %C27, %C28, %C29, C27/C29 (5α,14α,17α,20R-cholestane/5α,14α,17α,20R-24-ethyl-cholestane), and C28/C29 (5α,14α,17α,20R-24-methyl-cholestane/5α,14α,17α,20R24-ethyl-cholestane). The variation for the first two principal components, F1 and F2, accounted for about 62.72% of the total variance in the data. In order to set objective boundaries for distinguishing the oil

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

10

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

samples with similarities to the tar ball samples, a clustering analysis was performed. Most of the Azerbaijan oils differed from the tar ball samples. The highlighted areas in Fig. 7 show the oil samples clustering with the tar ball samples. Among the Azerbaijan oils, only B-1oil from Bahar field oils was clustered with the tar balls. Though B-1oil fell close to the tar balls, the Turkmenistan oil is closer to the tar balls. It shows that Turkmenistan oil has more similar characteristics to the tar ball samples. The cross-plot diagram and PCA (Fig. 6) indicated that the tar ball samples deposited along the Southwest Caspian Sea beaches during 2012 are most likely originated from the Turkmenistan oil. 3.2. Weathering of tar-ball samples Analyses of PAH, n-alkanes and hopanes have been previously used to evaluate the weathering of tar-balls and oil spill samples (Barakat et al., 2001; Hegazi et al., 2004; Zakaria et al., 2001; Diez et al., 2007). Table 3 shows the significant ratios for tar ball samples and reference oils. The n-C17/pristane and n-C18/phytane ratios are useful indices of the biodegradation processes of oil spilt in the environment. The nC17/pristane and n-C18/phytane ratios of the tar balls are lower than those of the crude oils, suggesting that the samples had biodegraded significantly (Table 3). The low ratios of n-C17/pristane and n-C18/phytane in the biodegraded samples result from higher degradation of straight chain alkanes than branched ones (pristine and phytane) by bacterial communities (Radović et al., 2012). There is the familiar “hump” of the unresolved complex mixture (UCM) in the present tar balls indicating the oil has undergone weathering (Fig. 8). UCM, defined as the unresolved complex mixture of hydrocarbons detected by GC, is a characteristic of heavily weathered and biodegraded oils (Barakat et al., 2001; Dutta and Harayama, 2000; Atlas, 1995; Peters and Moldwan, 1993). The tar ball samples showed a very similar shape of the UCM, which implies that they may be from the same source and have undergone the same degree of weathering (Wang et al., 1998). Presence of UCM may suggest that the tar balls have a long residence time on the sea, which is enough for severe exposure to weathering and biodegradation processes (Zakaria et al., 2000). As shown in Table 4, Turkmenistan, Azerbaijan and Anzali oils exhibit a substantially higher value of L/H-alkane (low/high molecular weight) ratio ranging from 3.61 to 4.64 as compared to the tar ball samples (0.78 to 1.04). The L/H ratio indicates the relative depletion of lower molecular weight (LMW) (i.e. n-C16–n-C26) compounds relative to the higher molecular weight (HMW) (i.e. n-C27–n-C36) compounds in the samples. This is consistent with low alkane contents in the 6 tar balls (93 to 140 mg g− 1) compared with those in the crude oil samples (210 to 500 mg g−1). Since evaporation usually favor the short chained alkanes compared to the longer chained ones, severe depletion of LMW alkanes has been ascribed to dissolution or evaporation. This is proved with the studies by Dutta and Harayama (2000); Wang and Stout (2007) and Zakaria et al. (2000). Generally, all the samples exhibited the presence of UCM and depletion of LMW alkanes, which may suggest that the tar balls are not originated from tanker washes. The tar balls derived from tanker washes are in abundance of HMW n-alkanes with the absence of UCM (Zakaria et al., 2001). Petroleum is a complex mixture of many thousands of components, including PAHs, alkylated PAHs and their configured isomers (Table 4). The PAHs in the tar ball samples also showed familiar signs of evaporation and biodegradation similar to the alkanes. The lack of LMW compounds such as dibenzothiophine, phenanthrene, 3-methylphenanthrene, 2-methyl-phenanthrene, 9-methyl-phenanthrene, 1methyl-phenanthrene, 2-methylanthracene and fluoranthene in the tar balls indicates that these samples have been highly weathered (Suneel et al., 2013). Although there are LMW PAH components in all tar balls, but based on the L/H-PAHs ratio, these samples had lost a greater fraction of LMW PAHs compared to the crude oil samples. It has been suggested that the ratio of PAH degradation due to evaporation

and biodegradation decreases with the number of rings, and increases in the degree of branches within a homologous series (Elmendorf et al., 1994). As discussed in the Material and methods section, to quantify the true percentage depletion level for the selected parent PAHs and alkylated PAHs, the measured concentrations have been normalized using Eq. (1). The hopane normalization factors for various tar ball samples are estimated to be in the range 0.49–0.52 (mean = 0.48). Using these factors, the true percentage depletion levels of PAHs were estimated to be in the range 73–85% (mean = 79%). The percentage depletion levels for the selected parent PAHs and alkylated PAHs were computed and the results are summarized in Table 5. Data showed that both the light and heavy PAHs have been weathered in the tar ball samples. Although some heavy components concentrated more in the tar balls than in the oil sample (Turkmenistan oil), after hopane normalization, their percentage depletion was higher than zero; for example, Chrysene concentration was concentrated from 503.77 ng/g in the source oil to 874.01–898.10 ng/g in the tar ball samples but, the true percentage depletion level after hopane normalization was 11%–18%. As the heavy PAHs were unaffected by evaporation (Yin et al., 2015), this indicates that in addition to evaporation, photo-oxidation was the dominant weathering process. Increasing alkyl substitution and HMW PAHs accelerates the sensitivity of aromatic compounds to photooxidation, which could have caused the lower degree of depletion of H-PAHs in the samples (Diez et al., 2007). In general, the presence of UCM and low L/H ratio for both alkanes, and PAHs for all the tar balls revealed that these samples were significantly weathered, and depleted in alkanes and PAHs through microbial degradation, evaporation or dissolution, and photo-oxidation. A supporting method for estimating the degree of weathering in the target oils and tar balls has been developed using C30 hopane as an internal conservative reference in the following equation (Wang et al., 1998): P ¼ ð1−cs =cw Þ  100

ð1Þ

where, P is the percentage of weathering level with respect to the source crude, and cs and cw are the concentrations of C30αβ hopane in the source oil (Turkmenistan oil) and weathered sample, respectively. The weathered percentages were estimated for the tar ball samples TB1–TB6 were in the range 48–54% (mean = 51%). According to the alkane, hopane and PAH data, the tar ball samples from this study had undergone similar phases of weathering and biodegradation and have been highly weathered after the spill. Long-rang transport could facilitate the weathering, and consequently, these tar balls could have been derived from one spill from the offshore oil platform from Turkmenistan. As these tar balls are floating in the water column, then surface currents and tides are the main forces transporting them to the Southwest coasts of the Caspian Sea. The winds' character seems to play a major role in forming the Caspian Sea currents. The most stable and dominant winds over the Caspian Sea during the major part of the year are winds with northerly-northwesterly and southeasterly directions during summer and winter, respectively (Kosarev, 2005); however, large-scale anti-cyclonic winds prevailing over the Caspian Sea with south-southwest-ward winds occur in February–July. For the dearth of required instrumental current observations, the majority knowledge on the general water circulation in this sea comes from the diagnostic simulations by numerical hydrodynamic models (Tuzhilkin and Kosarev, 2005). According to the model presented by Ibrayev (2010), during February–July, the South-Southwestward Ekman drift currents are dominated in the deep-sea regions, and superimposed on the Southward coastal currents along the Eastern and Western shelf regions (Ibrayev, 2010). These currents can deliver tar balls from Turkmenistan to the Southwest coasts of the Caspian Sea.

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203

G. Shirneshan et al. / Science of the Total Environment xxx (2016) xxx–xxx

4. Conclusions Based on the diagnostic ratios of n-alkane and hopane, and steranes analysis, it is concluded that the tar balls that were deposited along the Southwest coast of the Caspian Sea in 2012 have the characteristics of floating tar balls. The tar ball samples were all remarkably similar and consistent with the same source. The results of source identification by using cross plot and principle component analysis of the diagnostic ratios of Azerbaijan, Anzali and Turkmenistan oils confirm that the source of the tar ball samples analyzed in this study is most likely from the Turkmenistan oil. PCA and cluster analysis using 20 Azerbaijan oils and 3 samples from Azerbaijan, Anzali and Turkmenistan oils and tar balls indicated that the tar ball samples are most likely from the Turkmenistan oil even though the oil sample from Bahar field oils may also be a source of tar balls. Further evaluation is needed to confirm a definite source for the tar balls. The alkanes, PAHs and hopane data showed that the tar balls have been severely weathered since their release. Acknowledgement The support by the Ministry of Science and Technology (Iran) that provided partial funding for this project is highly appreciated. References Asuquo, F.M., 1991. Tar-balls on Ibeno–Okposo beach of South-East Nigeria. Mar. Pollut. Bull. 22, 150–151. Atlas, R.M., 1995. Petroleum biodegradation and oil spill bioremediation. Mar. Pollut. Bull. 31, 178–182. Barakat, A.O., Qion, Y., Kim, M., Kennicutt, M.C., 2001. Chemical characterization of naturally weathered oil residue in arid terrestrial environmental in Al-Alamein. Egypt Environ. Sci. Technol. 27, 291–310. Bence, A.E., Burns, W.A., 1995. Fingerprinting hydrocarbons in the biological resources of the Exxon Valdez spill area. In: Wells, P.G., Butler, J.N., Hughes, J.S. (Eds.), Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters. ASTM, Philadelphia, PA, pp. 85–140. Bernabeu, A.M., Fernández-Fernández, S., Bouchette, F., Rey, D., Arcos, A., Bayona, J.M., Albaiges, J., 2013. Recurrent arrival of oil to Galician coast: the final step of the prestige deep oil spill. J. Hazard. Mater. 250-251, 82–90. Boehm, P.D., Page, D.S., Burns, W.A., Bence, A.E., Mankiewicz, P.G., Brown, J.S., 2001. Resolving the origin of the petrogenic hydrocarbon background in Prince William Sound, Alaska. Environ. Sci. Technol. 35 (3), 471–479. Chandru, K., Zakaria, M.P., Anita, S., Shahbazi, A., Sakari, M., Bahry, P.S., Mohamed, C.A., 2008. Characterization of alkanes, hopanes, and polycyclic aromatic hydrocarbons (PAHs) in tar-balls collected from the East Coast of Peninsular Malaysia. Mar. Pollut. Bull. 56, 950–962. Clark, R.B., 2002. Marine Pollution. fifth ed. Oxford University Press, Oxford. Dhargalkar, V.K., Kureishy, T.W., Bhandare, M.V., 1976. Deposition of tar balls (oil residue) on beaches along the west coast of India. Mahasagar Bull. Natl. Inst. Oceanogr. 10 (3 $_amp_$amp; 4), 103–108. Diez, S., Jover, E., Bayona, J.M., Albaigés, J., 2007. Prestige oil spill. III. Fate of a heavy oil in the marine environment. Environ. Sci. Technol. 41, 3075–3082. Dutta, T.K., Harayama, S., 2000. Fate of crude oil by the combination of photo oxidation and biodegradation. Environ. Sci. Technol. 34, 1500–1505. Effimoff, I., 2000. The oil and gas resource base of the Caspian region. J. Petrol. Sci. Eng. 28, 157–159. Ekweozor, C.M., Okogun, J.I., Ekong, D.E.U., Maxwell, J.R., 1979. Preliminary organic geochemical studies of samples from Niger Delta (Nigeria) I. Analysis of crude oils for triterpanes. Chem. Geol. 27, 11–28. Elmendorf, D.L., Haith, C.E., Douglas, G.S., Prince, R.C., 1994. Relative rates of biodegradation of substituted polycyclic aromatic hydrocarbons. In: Hinchee, R.E., Leeson, A.E., Semprini, L., Ong, S.K. (Eds.), Bioremediation of Chlorinated and Polycyclic Aromatic Hydrocarbon Compounds. Lewis Publishers, Ann Arbor, MI, pp. 188–202. Gurgey, K., 2003. Correlation, alteration, and origin of hydrocarbons in the GCA, Bahar, and Gum Adasi fields, western South Caspian Basin: geochemical and multivariate statistical assessments. Mar. Petrol. Geol. 20, 1119–1139.

11

Hegazi, A.H., Andersson, J.T., Abu-Elgheit, M.A., El-Gayar, M.S., 2004. Source diagnostic and weathering indicators of tar ball utilizing acyclic, polycyclic and S-heterocyclic components. Chemosphere 55, 1053–1065. Ibrayev, R.A., Özsoy, E., Schrum, C., HI, S., 2010. Seasonal variability of the Caspian Sea three-dimensional circulation, sea level and air-sea interaction. Ocean Sci. 6, 311–329. Kosarev, A.N., 2005. Physico-geographical conditions of the Caspian Sea. In: Kostianov, A.G., Kosarev, A.N. (Eds.), The Caspian Sea Environment (Handbook of Environmental Chemistry)Springer 5–31. Kvenvolden, K.A., Hostettler, F.D., Rapp, J.B., Carlson, P.R., 1993. Hydrocarbons in oil residue on beaches of islands of Prince William Sound Alaska. Mar. Pollut. Bull. 26, 24–29. Macko, S.A., Parker, L.P., 1983. Stable nitrogen and carbon isotope ratios of beach tars on South Texas barrier islands. Mar. Environ. Res. 10, 93–103. Mulabagal, V., Yin, F., John, G.F., Hayworth, J.S., Clement, T.P., 2013. Chemical fingerprinting of petroleum biomarkers in deepwater horizon oil spill samples collected from Alabama shoreline. Mar. Pollut. Bull. 70, 147–154. Nair, A.V.P., Devassy, S.N., Dwivedi, R.A., Selvakumar, 1972. Tar ball pollution in Central west coast of India. Curr. Sci. 41, 766–767. Peters, K.E., Moldwan, M.J., 1993. The Biomarker Guide. 21. Prentice Hall, New Jersey, pp. 303–312 Sosrowidjojo, I.B., Alexander, R., Kagi, P. 1994. The biomarker composition of some crude oils from Sumatra. Org. Geochem. Radović, J., Domínguez, C., Laffont, K., Díez, S., Readman, J.W., Albaigés, J., Bayona, J.M., 2012. Compositional properties characterizing commonly transported oils and controlling their fate in the marine environment. J. Environ. Monit. 14, 3220–3229. Silva, A.M.D., Bicego, M.C., 2010. Polycyclic aromatic hydrocarbons and petroleum biomarkers in So Sebastio Channel, Brazil: assessment of petroleum contamination. Mar. Environ. Res. 69, 277–286. Suneel, V., Vethamony, P., Zakaria, M.P., Naik, B.G., Prasad, K.V.S.R., 2013. Identification of sources of tar balls deposited along the Goa coast, India, using fingerprinting techniques. Mar. Pollut. Bull. 70, 81–89. Suneel, V., Vethamony, P., Naik, B.G., Vinod Kumar, K., Sreenu, L., Samiksha, S.V., Tai, Y., Sudheesh, K., 2014. Source investigation of the tar balls deposited along the Gujarat Coast, India, using chemical fingerprinting and transport modeling techniques. Environ. Sci. Technol. 48, 11343–11351. Suneel, V., Vethamony, P., Naik, B.G., Krishna, M.S., Jadhav, L., 2015. Identifying the source of tar balls deposited along the beaches of Goa in 2013 and comparing with historical data collected along the west coast of India. Sci. Total Environ. 527-528, 313–321. Tolosa, I., Mora, S.D., Sheikholeslami, M.R., Villeneuve, J.P., Bartocci, J., Cattini, C., 2004. Aliphatic and aromatic hydrocarbons in coastal Caspian Sea sediments. Mar. Pollut. Bull. 48, 44–60. Tuzhilkin, V.S., Kosarev, A.N., 2005. Thermohaline structure and general circulation of the Caspian Sea waters. In: Kostianoy, A.G., Kosarev, A.N. (Eds.), The Caspian Sea Environment (Handbook of Environmental Chemistry)Springer 33–57. Wang, Z.D., Fingas, M., 1997. Development s in the analysis of petroleum hydrocarbons in oils, petroleum products and oil-spill-related environmental samples by gas chromatography. J. Chromatogr. A 774, 51–78. Wang, Z., Stout, S.A. (Eds.), 2007. Oil spill environmental forensics: fingerprinting and source identification. Academic Press, London, pp. 229–256. Wang, Z.D., Fingas, M., Sergy, G., 1994. Study of 22-year-old arrow oil samples using biomarker compounds by Gc/Ms. Environ. Sci. Technol. 28, 1733–1746. Wang, Z.D., Fingas, M., Sergy, G., 1995. Chemical characterization of crude oil residue from an Arctic beach by GC/MS and GC/FID. Environ. Sci. Technol. 29, 2622–2631. Wang, Z., Fingas, M., Landriault, M., Sigouin, L., Castle, B., Hostetter, D., Zhang, D., Spencer, B., 1998. Identification and linkage of tar balls from the coast of Vancouver Island and Northern California using GC/MS and isotopic techniques. J. High Resolut. Chromatogr. 21, 383–395. Wang, Z., Stout, A.S., Fingas, M., 2006. Forensic fingerprinting of biomarkers for oil spill characterization and source identification. Environ. Forensic 7, 105–146. Yim, U.H., Ha, S.Y., An, J.G., Won, J.H., Han, G.M., Hong, S.H., Kim, M., Jung, J., Shim, W.J., 2011. Fingerprint and weathering characteristics of stranded oils after the Hebei Spirit oil spill. J. Hazard. Mater. 197, 60–69. Yin, F., John, G.F., Hayworth, J.S., Clement, T.P., 2015. Long-term monitoring data to describe the fate of polycyclic aromatic hydrocarbons in Deepwater horizon oil submerged off Alabama's beaches. Sci. Total Environ. 508, 46–56. Zakaria, M.P., Horinouchi, A., Tsutsumi, S., Takada, H., Tanabe, S., Ismail, A., 2000. Oil pollution in the straits of Malacca, Malaysia: application of molecular markers for source identification. Environ. Sci. Technol. 34, 1189–1196. Zakaria, M.P., Okuda, T., Takada, H., 2001. Polycyclic aromatic hydrocarbon (PAHs) and hopanes in stranded tar-balls on the coasts of Peninsular Malaysia: applications of biomarkers for identifying sources of oil pollution. Mar. Pollut. Bull. 42, 1357–1366.

Please cite this article as: Shirneshan, G., et al., Identification of sources of tar balls deposited along the Southwest Caspian Coast, Iran using fingerprinting techniques, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.04.203