Introduction

The Mediterranean Sea is among the most specific and vulnerable marine ecosystem on the earth. It is, therefore, likely that global change will affect this semi-enclosed basin more rapidly and intensively than the world ocean. In fact, many persistent and bio-accumulative chemicals are sources of serious health problems that affect people and the entire ecosystem.

Some authors have demonstrated the negative effects that hydrocarbon has on the ecosystem and, therefore, on marine biodiversity (Clark 1992). However, the presence of hydrocarbons does not necessarily indicate pollution, as these compounds may also have terrestrial (vascular plants) or marine (algae and/or phytoplankton) biogenic origin (Colombo et al. 1989). Among hydrocarbons, polycyclic aromatic hydrocarbons (PAHs) are a widespread class of environmental pollutants that are carcinogenic and mutagenic. They arise from the incomplete combustion of organic material, especially fossil fuels (pyrolytic origin), from the discharge of petroleum and its products (petrogenic origin) and from the post-depositional transformation of biogenic precursors (diagenetic origin). Terrestrial plant waxes, marine phytoplankton, volcanic eruptions, biomass combustion and natural oil seeps contribute natural inputs of hydrocarbons, including aliphatic and aromatic hydrocarbons (Saliot 2005).

For many decades, industrial, agricultural and urban wastes have been discharged into the Mediterranean Sea via coastal outfalls, rivers and the atmosphere. Consequently, this phenomenon causes a considerable increase in pollution and a progressive degradation of the marine ecosystem (Saliot 2005).

Anthropogenic origin hydrocarbons such as petroleum and its derived products are spilled into the environment as a result of some activities related to oil operation, petroleum transport, harbor activities, domestic and industrial effluents. PAHs have been determined worldwide in many environmental matrices and associated with the toxicity of marine sediment (Volpi Ghirardini et al. 1999; Macias-Zamora et al. 2002; Mai et al. 2002; Maskaoui et al. 2002; Eggleton and Thomas 2004; Illou 1999; Zaghden et al. 2005).

These endocrine-disrupting compounds are persistent in the environment and, therefore, both their accumulation on the sediments and possible remobilization, as well as their uptake by organisms, pose a significant threat of the aquatic life (Zaghden et al. 2007).

Furthermore, hydrocarbons of biogenic origin naturally occur at low concentrations in different substrates, such as water and sediment, and they are part of the natural hydrocarbon baseline of an ecosystem (Volpi Ghirardini et al. 1999). For this reason, it is imperative to investigate the origin and the distribution of hydrocarbons in the aquatic environment. In previous works, the determination of the complex assemblage of these hydrocarbons coming from various sources had been assessed (Macias-Zamora et al. 2002; Mai et al. 2002; Maskaoui et al. 2002; Eggleton and Thomas 2004).

In the south of Tunisia and the Mediterranean Sea, the industrialized region of Sfax, located in Gabes Gulf, is vulnerable and sensible due do its shallow area and its high ecological value characterized by an important biodiversity. Therefore, a lot of work is needed to study these specific systems to understand the common and specific processes taking place there and also to produce a background of information and data that shall be very useful for the sustainable management of this ecosystem (Illou 1999; Zaghden et al. 2005, 2007). This region, second city in Tunisia, is marked by a significant expansion of the urban land, and it also has important industrial and commercial activities. These activities generate an important chemical pollution. The impact of this anthropogenic pressure can be observed in summer with some phenomena of red waters resulting from eutrophication and disequilibrium process. In fact, it is necessary to study and quantify this chronic pollution that affects the coastal and marine ecosystem. As part of an ongoing investigation of Sfax coastal area, we have investigated 16 superficial sediment samples from sites throughout the coast for their hydrocarbon content, distribution and probable origin (anthropogenic and/or biogenic) of hydrocarbon.

Materials and methods

Study area and sampling

The area of this study concerns the coastline of Sfax city located in the southeast of Tunisia in the northern shore of Gabes Gulf (34°44′N, 10°46′E) in the southern Mediterranean Sea (Fig. 1). It is characterized by a shallow area and an important biodiversity with endemic seagrass Posidonia oceanica. This area is a sensible model site for hydrocarbon analysis. It is exposed to high entropic pressure due to its commercial harbor and the presence of intense maritime traffic. For many decades, the urban evolution of Sfax city has been marked by a significant expansion of the urban land which is characterized by an important harbor board, a fishing port and industrial activities. The main industries of the region are food processing, olive oil, soap, paint production, chemical and textile industries, pulp factories, metal and phosphate processing.

Fig. 1
figure 1

Map of Sfax coast showing sampling locations

Full size image

In this study, intertidal superficial sediment samples are collected from 16 sites that stretch to about 20 km along the coastal area of Sfax city. Eight superficial sediment samples are collected from the northern coast of Sfax and the other eight are collected from the southern coast of Sfax (Fig. 1; Table 1). This area is characterized by a high tide with +1.60 m and a low tide with +0.30 m. Samples are collected manually, at low tide in an area of 40 × 40 cm.

Table 1 Sample site description of Sfax costal area
Full size table

The first centimeter [Surficial sediments (0–5 cm)] of near shore sediments is removed by scraping with a clean spatula.

Chemical analysis

Sediments are freeze-dried immediately after collection. After elimination of largevegetal fragments, sediments are sifted to keep the fine fraction <63 μm. Approximately 30 g of each sediment is used for Soxhlet extraction for 12 h using a mixture of dichloromethane/methanol (2:1 v/v). The extraction of hydrocarbons yield is 90 ± 6 %. The lipid extracts are submitted to adsorption chromatography on a column (2 cm i.d., 20 cm length) filled with florisil (60 mesh). The florisil column elimination of highly polar compounds, resins and extracts obtained after precipitation asphaltenes. This results in a more purified fraction containing a mixture of neutral compounds (total hydrocarbons: alkanes, alkenes and aromatics). Total hydrocarbons are eluted with dichloromethane and polar compounds with methanol.

The amount of total hydrocarbons is determined by gravimetry and by Fourier transformed infrared spectroscopy (FT/IR) using a Nicolet apparatus (Madison, USA). The calibration is performed with the Ashtart crude oil produced in Sfax area and n-eicosene. The extract is concentrated by a rotary evaporator and fractioned by silica–alumina gel chromatography using a three-step scheme, providing three fractions:

  • Non-aromatics (F1), elution with hexane,

  • Aromatics (F2), elution with hexane/dichloromethane 4:1 v/v,

  • Polar compounds (F3), elution with methanol.

Therefore, it seems more interesting to use the Soxhlet extractor than the ultrasonic method. Concentrations obtained for all PAHs with the Soxhlet extractor are equivalent to those recommended by IAEA for the reference material IAEA-408 (Mzoughi et al. 2002).

Fractions are evaporated to 150–200 μL under a stream of dry nitrogen. Aliphatic and alicyclic hydrocarbons (F1, data base of this study) are analyzed by a Delsi Model DI 200 gas chromatograph (Perichrom, France). The analysis is performed on a 50-m fused silica column (0.32 mm i.d.) coated with CPSil5-CB (0.25 μm film thickness). Helium at a flow rate of 1.5 mL/min (pressure 0.8 kPa) is used as the carrier gas. The temperature of the column is ramped at 4 °C/min from 70 to 280 °C. Injector and detector temperature is held at 320 °C.

Detection is monitored by flame ionization detector (FID). Hydrocarbons are identified in comparison with retention time by those of known standards of n-alkanes ranging from n-C14 up to n-C35. The structure of several hydrocarbons is confirmed by gas chromatography/mass spectrometry (HP6890-HP5973MSD Agilent Technologies, Wilmington, DE, USA). The GC is used with a 30-m fused-silica column (0.25 mm i.d.) coated with 5 % phenyl methyl siloxane. Helium is used as the carrier gas at a flow rate of 1.4 mL/min. The following temperature program is used: 80–290 °C with ramping at 4 °C/min. The samples are injected in the splitless mode by a temperature injector of 280 °C. Analyses are run in the electron impact mode at 70 eV with a 2.9-s scan time over a 50–550 a.m.u. range resolution. Interpretation of GC/MS spectra is based on mass chromatogram at m/z: 191, 217, 218 and 249.

Evaluation indices

Diagnostic techniques such n-alkane indices, unresolved complex mixture (UCM) presence and n-alkane homologous series are used to help identify potential sources (biogenic and/or anthropogenic) of hydrocarbons. Carbon preference index (CPI) is used to to assess hydrocarbon origin. The carbon predominance index (CPI) represents the relative abundance of odd numbered linear alkanes in front of even numbered linear alkanes. This CPI is around 1 for the first four stations, which characterizes petroleum hydrocarbons (Le Dréau et al. 1997).

$$ {\text{CPI}} = \, {{\left( {\sum n{\text{C}}17 \, + \, n{\text{C}}19 \, + \cdots + \, n{\text{C}}27} \right)} \mathord{\left/ {\vphantom {{\left( {\sum n{\text{C}}17 \, + \, n{\text{C}}19 \, + \cdots + \, n{\text{C}}27} \right)} {\left( {\sum n{\text{C}}16 \, + \, n{\text{C}}18 \, + \cdots + \, n{\text{C}}26} \right)}}} \right. \kern-0pt} {\left( {\sum n{\text{C}}16 \, + \, n{\text{C}}18 \, + \cdots + \, n{\text{C}}26} \right)}} $$

The CPI has been frequently used as a source indicator of n-alkanes in marine sediments (Colombo et al. 1989). It represents the relative abundance of odd-numbered linear alkanes versus even-numbered linear alkanes. n-Alkanes derived from terrestrial vascular plant usually have CPI values ranging from 3 to 6, while petrogenic hydrocarbons show CPI values close to 1. Lower CPI indicates microbial sources of hydrocarbons (Blumer et al. 1963). Pristane/phytane (Pr/Ph), n-C17/Pr, n-C18/Ph and the ratio between unresolved complex mixture over resolved aliphatic hydrocarbons (U/R). For the aromatic fraction, ratios of selected PAH concentration are calculated.

Results and discussion

Extractible organic matter and hydrocarbons

Extractible organic matter (EOM) concentrations vary from 765 to 9,100 μg/g of sediment dry weight (Table 2). Total values obtained in this study are considered as the highest level in comparison with many studies of different regions in the world (Eggleton and Thomas 2004; Le Dréau et al. 1997; Maljevic and Balac 2007). The highest values of this area are found in the stations 4, 10 and 11 with an average 8,000 μg/g.

Table 2 Characteristic of extractable organic matter and hydrocarbons in surface sediments of Sfax coastal area
Full size table

These concentrations correspond, respectively, to the first and second channel of urban and industrial discharge. The second group of concentration values of EOM is found in stations 5, 6, 7 and 8. These stations correspond to the industrial area and harbor port with values varying between 3,000 and 5,000 μg/g of sediment dry weight. The other stations located in the northern and the southern coastal present relatively less concentration of EOM. Therefore, the concentrations of total hydrocarbon (THC) and their relatively concentration to EOM show a significant result. The concentrations of THC are varying between 482 and 4,087 μg/g (Table 2). The low proportion of THC varying between 30 and 50 % of EOM is found in stations 4, 5, 6, 7, 10 and 11 corresponding to the stations with a high concentration of EOM. This is in relation to the sites’ location in front of the channel and to the industrial area with important organic discharge. The accumulated pollution has difficult dispersion and dilution in this shallow area. The south area is more polluted than the north area. The highest values are found in front of the southern channel (location name Sidi Salem) with an average of 4,000 μg/g and in front of the northern channel (location name PK4) with 2,366 μg/g.

n-Alkanes and isoprenoid hydrocarbon

Non-aromatic (or aliphatic and alicyclic) hydrocarbons (NAH) present a high concentration with a range varying from 180 to 1,400 μg/g of dry sediment (Table 2). Comparable to the THC, the two channels of Sfax area show the highest values on NAH with an average proportion to THC of around 35 % (Fig. 2). The high proportion (83 %) of NAH to THC is found at station 13 that is characterized by the effluents of urban wastewater and fertilizer factory. These concentrations are generally high compared with other marine areas that receive important anthropogenic inputs such as Hon Kong (Gao et al. 2008), Patagonia, Argentina (Zheng and Richardson 1999), Bay of Fort de France (Commendatore et al. 2000), Tianjin, China (Mille et al. 2006) and Todos os Santos Bay, Brazil (Bixiong et al. 2007).

Fig. 2
figure 2

Total hydrocarbon (THC) and non-aromatic hydrocarbon (NAH) concentrations of the surface sediment of Sfax coastal area

Full size image

GC traces of NAH of samples from Sfax coastal sediments show comparable patterns as illustrated by sample 10 in Fig. 3. The n-alkanes range from C15 to C32 without predominance of odd/even carbon numbered chains. Petroleum usually shows a wide distribution range of n-alkanes and the even/odd shows no predominance. In contrast, in most plant waxes, odd-chain alkanes are 8–10 times more abundant than even-chain n- alkanes (Venturini et al. 2008).

Fig. 3
figure 3

Gas chromatogram of aliphatic and alicyclic hydrocarbons of surface sediment at site 10. S sulfur compound; Pr pristine; Ph phytane; C thiosterane; UCM unresolved complex mixture (Zaghden et al. 2007)

Full size image

Compared with n-alkanes, UCM is more resistant to biodegradation and it has a greater tendency to remain and to accumulate in the environment (Volkman et al. 1992). Unresolved complex mixture/sum of the resolved aliphatic hydrocarbons (U/R) index is used as a criterion to assess the anthropogenic input and to estimate the relative degradation degree. Some researchers suggest that a U/R value ≥2 reflects significant contamination by petroleum products (Aboul-Kassim and Simoneit 1995), while others suggest that U/R value >1 can be taken as a threshold of such significant contamination (Simoneit 1986), whereas low values suggest fresh oil presence. The UCM magnitude is related to the degree of anthropogenic contribution. UCM presence is usually associated with petroleum hydrocarbons. It is suggested that UCM can result from bacterial degradation of organic matter (Aboul-Kassim and Williamson 2003). Anyway, the fact that most of the stations of Sfax show U/R higher than 2 indicates a high degree of anthropogenic contribution (Table 2). The CPI index of Sfax sediment ranges between 0.80 and 1.60 (Table 2) and indicates petrogenic hydrocarbons with a highly degraded organic matter (Venkatesan and Kaplan 1982). Therefore, the contribution of uncontaminated sediment and vascular plants ranges from 3 to 6 (Colombo et al. 1989).

Another useful indicator of hydrocarbons’ origin is the ratio of isoprenoids pristine and phytane (Pr/Ph). The detection of a large number of aliphatic isoprenoid hydrocarbons in oils, coals, shale and dispersed organic materials is thought to be one of the most important discoveries in petroleum chemistry and organic geochemistry (Simoneit et al. 1991). Pristane and phytane are the two most abundant isoprenoid hydrocarbons and are identified in all samples of this study. These two compounds are considered to originate primarily from the phytyl side chain of chlorophylls during diagenesis. In uncontaminated recent sediments, phytane is practically absent and high pristane content can be derived from zooplankton and some other marine animals (Venturini et al. 2008), leading to the pristane/phytane (Pr/Ph) ratio higher than 1, typically between 3 and 5 (Wang et al. 2006). A Pr/Ph value close to 1 or lower than 1 suggests petroleum contamination (Blumer et al. 1963). The Pr/Ph ratios obtained from Sfax surface sediments are between 0.60 and 1.10 with an average of 0.73 indicating petroleum contaminations (Table 2).

In addition to their use as petroleum contamination markers, the Pr/n-C17 and Ph/n-C18 ratios are often used to evaluate relative biodegradation of n-alkanes (Mai et al. 2002). Comparatively, isoprenoid hydrocarbons are more resistant to biodegradation than n-alkanes. High values of these indices suggest the presence of degraded oil, while lower indices suggest low degradation (Steinhauer and Boehm 1992). When hydrocarbon concentrations are also high, lower indices indicate fresh oil inputs (Colombo et al. 1989). Petroleum contamination in Sfax coastal sediments is relatively fresh with both ratios (Pr/n-C17 and Ph/n-C18) around to 1. These ratios determined in the stations (4, 5, 9, 10 and 11) show values between 0.59 and 0.8 for Pr/n-C17 and 0.36 and 1.25 for Ph/n-C18. The distribution of Pr/n-C17, Ph/n-C18 and Pr/Ph ratios for all sample sites shows a common petroleum contamination source in Sfax coastal area (Table 2).

Hopanes and steranes

Alicyclic hydrocarbons are known to be more resistant to biodegradation than aliphatic hydrocarbons. Hopanes and steranes are representatives of these compounds and are considered as petroleum biomarkers. They can be used as source and/or maturity indicators to identify the nature of the fossil materials from which petroleum originates (González-Vila et al. 2003). Such molecules are characterized by their restricted occurrence, source specificity, molecular stability and suitable concentration of analytical detection (Scholz-Böttcher et al. 2008). The majority of the shelf surface sediment yields a complex hopane and sterane. These compounds are converted from corresponding biogenic precursors, and they exist as many stereoisomers with different α/β and/or R/S configurations that have different thermodynamic stability. During sedimentary burial, the thermodynamically unstable isomers are gradually replaced by the 35 geologically stable isomers reaching a known equilibrium point and providing a measure of the maturity of organic matter (Simoneit and Mazurek 1982; Seifert and Moldowan 1986). These processes are a combination of a series of bacteriological actions and low-temperature reactions generally referred to as diagenesis, catagenesis and metagenesis (Simoneit et al. 1991).

Figures 4 and 5 show, respectively, the representative mass fragmentograms of m/z 191 (hopane) and 217 (sterane) of non-aromatic hydrocarbon fraction from Sfax surface sediment at site 10. A series of C27–C35 hopanes are identified in all sediment samples. In Sfax surface sediment two hopanes with 29 and 30 carbon atoms with configuration 17α(H), 21β(H) are found as predominant in these series. Hopanes with the 17α, 21β-configuration in the range of 27–35 carbon atoms are characteristic of petroleum because of their greater thermodynamic stability compared with other epimeric series (Peters and Moldowan 1991).

Fig. 4
figure 4

Mass hromatogram of m/z = 191(hopanes) of Sfax sediment at site 10. Ts 18α(H)22,29,30-tris-norhopanes; Tm 17α(H) 22,29,30-tris-norhopanes; 13 17α(H), 21β(H), 30-norhopane; 14 18α(H)-30-norneohopane; 15 17α(H), 21β(H)-hopane; 16 17α(H)-30-nor-29-homohopane; 17 17α(H), 21β(H)-homohopane(22S); 18 17α(H), 21β(H)-homohopane(22R); 19 17α(H), 21β(H)-bishomohopane(22S); 20 17α(H), 21β(H)-bishomohopane(22R); 21 17α(H), 21β(H)-trishomohopane(22S); 22 17α(H), 21β(H)-trishomohopane (22R); 23 17α(H), 21β(H)-tetrakishomohopane(22S); 24 17α(H), 21β(H)-tetrakishomohopane(22R); 25 17α(H), 21β(H)-pentakishomohopane(22S); 26 17α(H), 21β(H)-pentakishomohopane(22R)

Full size image
Fig. 5
figure 5

Mass chromatogram of m/z = 217 (steranes) and m/z = 218 (diasteranes) of Sfax sediment at site 10. 1 C27ααα-cholestane(20S); 2 C27αββ-cholestane(20R); 3 C27αββ-cholestane(20S); 4 C27ααα-cholestane(20R); 5 C28ααα-ergostane(20S); 6 C28αββ-ergostane(20R); 7 C28αββ-ergostane(20S); 8 C28ααα-ergostane(20R); 9 C29ααα-stigmastane(20S); 10 C29αββ-stigmastane(20R); 11 C29αββ-stigmastane(20S); 12 C29ααα-stigmastane(20R)

Full size image

Steranes in sediment or oil are derived from the transformation of biological sterol precursors. In general, C27 and C29 steranes are indicative of algae and higher plant source of organic matter, respectively. So, the ratio of C27/C29 steranes >1 specifies the predominance of organic matter input from marine algae, while when <1 it indicates a preferential higher plant input (Wang et al. 2006; Peters and Moldowan 1993). In Sfax surface sediment, the ratio of C27/C29 is higher than 1 indicating an abundance of organic matter and marine origin of the petroleum products.

PAH concentrations and compositions

Total PAH concentrations in sediments are reported as the sum of 16 priority PAHs. Ratios of selected PAH concentrations were used to study the possible sources of pollution. The total concentrations of PAH, determined in sediment from the southern coast of Sfax, ranged from 602.5 to 5608.2 ng/g dw (Table 3). Thyna bay and harbor of Sfax city present the highest concentrations in sediments from these sites. The use of ratios between alkyl substituted and unsubstituted PAH homologs provides information about their anthropogenic sources. Generally petrogenic hydrocarbons are characterized by the dominance of alkylated compounds over their parent homologs and by the dominance of low condensed (2–3 rings) PAHs over high condensed (4–6 rings) PAHs (De Luca et al. 2004). Four- and five-ring PAHs were the most abundant compounds in coast line of Sfax city sediments. On average, phenanthrene, fluorene, pyrene and chrysene were the most abundant PAH compounds in coast line of Sfax city, together accounting for 90 % of PAHs at these sites. The discharge of industrial wastewater and the emission of atmospheric particles might be the pyrolytic source in this study area. Values of ratio alkyl-substituted/parent compound are very high ranging from 1.5 to 13.4 for the phenanthrene series and from 0.28 to 1.56 for the pyrene series (Table 3). Thus, PAH compositional patterns and diagnostic ratios reflect a mixture of both petrogenic and pyrolytic sources in most of the sampling sites in sediments of Sfax coastal area.

Table 3 Individual and total concentrations of the PAHs (ng/g) and ratios for origin identification for the sediments collected in Sfax coastal zone
Full size table

Conclusions

The study of sediment contamination by hydrocarbons from the coastal area of Sfax indicates that hydrocarbon levels are generally high compared with other maritime areas. Anthropogenic hydrocarbon inputs are more apparent at sites associated with industrial discharges, shipping activities and sewage outfalls. Therefore, the concentrations of total hydrocarbon (THC) non-aromatic hydrocarbon (NAH) and PAH are relatively high near the urban area, whereas the concentrations of these compounds decrease going further from urban area in the north and the south of Sfax city. The sampling sites present CPI values, unresolved complex mixture/resolved aliphatic hydrocarbons (U/R) and pristine to phytane (Pr/Ph) ratios as indicative of petroleum products. Petroleum biomarkers (Hopanes and steranes) indicating higher maturity are found in all analyzed samples confirming the importance of oil inputs and its derivatives in this area. The PAH compositional patterns and diagnostic ratios reflect a mixture of both petrogenic and pyrolytic sources. The present study shows evidence of a common petroleum contamination source in Sfax coastal area. Future studies should include a more exhaustive evaluation of the impact of anthropogenic inputs. In addition, strict regulations are necessary to improve the environmental quality of this region and to avoid the impact of hydrocarbon contamination on fishing activities. This will contribute to the development of management and of control measures to preserve the coastal and marine environment of Gabes Gulf.