Introduction

Hydrocarbons as important components of organic input have received much attention because most of them are toxic, mutagenic and carcinogenic. In aquatic environment, due to their low water solubility and hydrophobic properties, hydrocarbons are commonly found to be transported to the sediments. Hydrocarbon in sediments reflects the relative contributions from multiple sources: biogenic, diagenetic, petrogenic and pyrogenic (Yunker et al. 1993; Hostettler et al. 1999). Thus, the chemical characteristics and distribution of hydrocarbons in the environment can vary greatly from one area to another. A number of indicators such as hydrocarbon concentrations, n-alkane profiles, polycyclic aromatic hydrocarbon (PAH) composition and petroleum biomarkers have been proposed to estimate the pollution level and find out the pollution sources (Volkman et al. 1992; Witt 1995).

River Beiluohe, the second-order tributary of the Yellow River, is located in northern Shaanxi Province, geographically belonging to the Loess Plateau. It has a total length of 680.3 km with a catchment basin of 26,905 km2. In Yan’an areas, the Beiluohe River flow throughout six counties including Zhidan, Wuqi, Ganquan, Fuxian, Luochuan and Huangling, which have large oil and gas reserves. Oil production and transportation are among the most important economic activities in these regions. However, these activities have generated large quantities of toxic waste, runoff and oil spills to the Beiluohe River. Moreover, located in the arid region of China, the Beiluohe River is of fundamental importance for agricultural activities and water supply for human consumption. However, due to a long history of wastewater drainage and platform runoff from nearby petroleum industries, many locations in Beiluohe River waters are no longer considered fit for their designated beneficial uses. Pollution has threatened the economic development and human health in this region (Mu 1994; Lin et al. 2005). However, to date, information regarding the local pollution level and its effects is very limited. Especially, reliable data on concentrations and origin of petroleum hydrocarbon in the river sediments is scarce. As a result, an integrated geochemical, hydrochemical and ecological survey of the Beiluohe River Basin, was implemented by the local government during the period October 2005–April 2006.

In this paper, we report the analysis of 12 sediment samples obtained from the Beiluohe River for petroleum hydrocarbons and selected PAHs. Our objectives were to assess the level of contamination in river sediments by petroleum hydrocarbons due to the oil production platforms and drilling activities undertaken in the study area, and to find out the source of hydrocarbon pollution using n-alkane profiles, PAH composition and petroleum biomarkers.

Materials and methods

Study area and sampling

Figure 1 shows the studied site—the Beiluohe River, which is located between 35°24′, 37°15′ North latitude and 107°40′, 110°17′ East longitude. To determine the hydrocarbon contamination, a total of 12 surface sediment samples throughout the Beiluohe River and its tributary were collected in October 2005 and they were extracted and detected by gas chromatography with flame ionization detector (GC/FID). Among these 12 samples, we selected 6 samples for gas chromatography/mass spectroscopy (GC/MS) detection and 6 samples for PAHs analysis. The sample localization is shown in Fig. 1. Sediments were obtained at a depth 0–10 cm utilizing a stainless-steel grab and placed in pre-cleaned brown glass jars and frozen immediately until laboratory analysis.

Fig. 1
figure 1

Study areas and sampling location in the Beiluohe River, northern Shaanxi

Full size image

Extraction and fractionation

In the laboratory, about 100 g dry sediment samples were Xoxhlet-extracted with chloroform for a period of 72 h. Elemental sulfur was removed from the extracts using activated copper. The extract was fractionated into aliphatic and aromatic hydrocarbons by adsorption liquid chromatography using a column of alumina and silica-gel, and gradient solvents as eluent: ligarine and ligarine/dichloromethane (3:9, v/v) or aliphatic and aromatic fractions, respectively. The aliphatic and aromatic fractions were concentrated again on a rotary evaporator, transferred to a vial, and then the volume was adjusted to 1 ml exactly using a stream of filtered N2 gas. An aliquot of 1 μl of each extract was subjected to GC/FID and GC/MS analysis.

Instrumental analysis

Normal and isoprenoid alkanes analyses were conducted on a Hewlett-Packard (HP) 6890 II GC-FID. The samples were analyzed using a fused silica capillary column (30 m × 0.32 mm, 0.25 μm film thickness) with helium as carrier gas. The GC conditions were: injector 290°C, oven temperature program: 40°C for 1 min; 6°C min−1 to 290°C. Petroleum biomarkers were analyzed by GC/MS (HP 6890 II GC coupled to a Platform II mass spectrometer). The capillary column used was coated with DB-5 (30 m × 0.32 mm, 0.25 μm film thickness) with helium as carrier gas. The GC conditions were the same as described earlier. The MS operating conditions were: electron energy 70 eV, scanned from 50 to 600 amu. The samples were analyzed in the splitless mode. The 16 PAHs in the concentrated and purified extracts were quantified with a gas chromatograph (HP-6890 model) with a mass spectrometer (HP-5975 MSD). A capillary GC column (30 m × 0.25 mm, 0.25 μm) was used. The GC oven temperature program was: 60°C for 2 min, then 3°C min−1 to 290°C, held constant for 25 min. The PAHs were determined by selected ion monitoring (SIM). SIM acquisition was carried out by comparison of the base peak of each PAH.

Results and discussion

Concentrations of total petroleum hydrocarbons

Concentrations of total hydrocarbons ranged between 12.1 and 3,761.5 μg g−1 dry sediment, with an average of 1,886.8 μg g−1 dry wt (Table 1). Petroleum hydrocarbon concentrations in sediments are typically a few μg g−1 in unpolluted coastal areas and from 50 to >1,000 μg g−1 in contaminated areas (Clark and Macleod Jr 1977). Insufficient documentation exists for total hydrocarbon contents of Beiluohe sediments. Results obtained from the present studies show that total hydrocarbon concentrations in the uncontaminated sediments of Beiluohe Basin range between 6.4 and 15.6 μg g−1. These low total hydrocarbon concentrations may be regarded as the natural background levels in the Beiluohe area. As a result, total hydrocarbon concentrations <15 μg g−1 are considered to denote unpolluted sediments, whereas values >200 μg g−1 are generally indicative of significant pollution.

Table 1 Concentrations (μg g− 1 dry wt) and characteristics of petroleum aliphatic hydrocarbons in surface sediments from the Beiluohe River
Full size table

The highest TPH concentration was found at Station S87, which was ascribed to the effect of the direct discharge water from the production platform. The lowest values found for stations S15 and S48 ranging from 7.9 to 14.4 μg g−1, were typical of uncontaminated sediments because of less petroleum activities in the neighborhood. Station S72 was very special for its low hydrocarbon concentration although it was located near the many production platforms. Stations such as S17, S39 and S43, had medium concentration ranging from 57.5 to 147.3 μg g−1, which might be a consequence of illegal discharge of petroleum wastewater from refinery factories or platforms. The station S64, situated near the garbage dumps of Danba, also presenting a medium TPH concentration, might result from anthropogenic activities. Other stations S38, S40, S45 and S47, presented a slight pollution with TPH concentration ranging from 16.4 to 40.1 μg g−1. In addition, the stations located in the tributary due to limited water circulation presented a higher value than those in the mainstream. Except the sample from S87 that was significantly contaminated, the levels of TPH concentration in Beiluohe were relatively low compared to those at worldwide industrial locations. The studied area, geographically located in the Loess Plateau, easily experienced erosion of soil materials resulting in larger inputs of sediments to rivers and the river was relatively shallow in depth so it had not accumulated most pollutants in sediments. As a result, we suggest that higher inputs of sediments in Beiluohe could contribute to the lower TPH concentrations in Beiluohe sediments.

Aliphatic hydrocarbons

The n-alkane found in sediment could be originated from anthropogenic or natural sources in aquatic environment. n-Alkanes GC profiles and some indicators such as ratio of odd/even carbon numbers, pristane/phytane ratio and the presence of the unresolved complex mixture (UCM) can be used to distinguish the two origins.

Generally, the aliphatic hydrocarbon fraction isolated from the sediment extracts showed the presence of a homologous series of n-alkanes extending from C15 to C34. Typically, abundance of C27, C29 and C31 with odd carbon predominance is an indicator of terrestrial plant, whereas a dominance of the n-alkanes in the range C18–C24 without odd carbon predominance, suggests relatively oil input (Sultan and Sameer 1997). In this study, n-alkanes, representing two distinct sources, are evident on the basis of their distribution pattern in the samples collected. As can be seen in Fig. 2a, the GC profile shows a dominance of n-alkanes in the range C19–C22 and has very prominent humps indicating the UCM of petroleum-origin hydrocarbons. This profile is very similar to that extracted from the crude oil in this region and as a result is characteristic of petroleum inputs. Stations S17, S40, S43 and S87 can be ranged in the group of predominant petroleum input. In contrast, stations S39, S45 and S64 had both petroleum and terrestrial plant inputs showing a predominance of odd on even numbered n-alkanes in the range C25–C31 and the presence of UCM (Fig. 2b).

Fig. 2
figure 2

a The aliphatic hydrocarbon profile of a typical sediment sample at site S40. The dominance of C21 without odd carbon predominance suggests oil input. b The aliphatic hydrocarbon profile of a typical sediment sample at site S39. The dominance of long chain n-alkanes (C27, C29 and C31) suggests high plant input

Full size image

The carbon predominance indices (CPI), which represent the ratios between the odd and the even carbon-homologues were around 1 at stations S17, S43, S38, S40 and S87. These CPI values and the FID/GC profiles suggest oil contribution to these sediments. Higher values of CPI were found for stations S39, S45 and S64 and a strong dominance of odd-numbered linear alkanes in the range C25–C31 could be noted, which was typical of the origin of higher plants. As a result, besides petrogenic contribution, there is also a significant terrestrial plant input at stations S39, S45 and S64. In spite of low hydrocarbon concentration, the FID/GC fingerprints and geochemical indices indicated the trace of oil contamination at station S72.

Another useful indicator of the origin of the hydrocarbons is the ratio of isoprenoids, pristane and phytane. In uncontaminated sediments, the ratio Pr/Ph is typically between 3 and 5 (Steinhauer and Boehm 1992). However, ratio of pristane to phytane was relatively low (0.73–1.32) in all samples in this study, indicating petroleum contribution to sediments.

Petroleum markers

Six samples (S38, S40, S43, S47, S64 and S72) were selected to identify the hopane and sterane series detected in GC/MS analysis by monitoring m/z 191 and 217, respectively. Hopanes (Fig. 3) comprised a series of 17α(H),21β(H)-compounds, maximizing at the C30 homologue, and less prominent 17β(H),21α(H)-hopanes. Such isomeric configurations are thermodynamically stable and occur in crude oils and mature rocks (Perters and Moildovan 1993). They are derived from the diagenic interconversion of the 17β(H),21β(H) homologues of bacterial origin. Extended hopanes (C ≥ 31) were also present as pairs of the C-22 diastereoisomers (22R, 22S) with a 22S/22S + 22R ratio of about 0.6, which also typifies petrogenic material (Mackenzie 1984). Gammacerane, a non-hopanoid C30 triterpane, was also present in the analyzed sediments. This compound is a biomarker common to most Chinese crude oils derived geologically from lascustrine sources (Fu and Sheng 1989). As shown in Fig. 3, in addition to the hopanes, a series of tricyclic hopanes were also observed, including a homologous series of tricyclic hopanes and the trisnorhopanes (Ts and Tm). In all analyzed samples, the ratio Ts/(Ts + Tm) vary in the range 0.44–0.71, which confirms the presence of mature petroleum in sediments.

Fig. 3
figure 3

Representative mass fragmentograms of m/z 191 for terpanes from sediment at site S43. 1–7 Tricyclic terpane; Ts 18α(H)22,29,30-tris-norhopanes; Tm 17α(H)22,29,30-tris-norhopanes; 8 17α(H),21β(H),30-norhopanes; 9 17α(H),21β(H)-hopanes; 10 17β(H),21α(H),30-hopane (moretane); 11 17α(H),21β(H)-homohopane (22S), 17α(H),21β(H)-homohopane (22R); 12 Gammacerane; 13 17α(H),21β(H)-bishomohopane (22S), 17α(H),21β(H)-bishomohopane (22R); 14 17α(H),21β(H)-trishomohopane (22S), 17α(H),21β(H)-trishomohopane (22R); 15 17α(H),21β(H)-tetrakishomohopane (22S), 17α(H),21β(H)-tetrakishomohopane (22R); 16 17α(H),21β(H)-pentakishomohopane (22S), 17α(H),21β(H)-pentakishomohopane (22R)

Full size image

The steranes were detected in all samples analyzed and comprised mainly the 5α,14β,17β and minor 5α,14α,17α configurations, both occurring as 20S and 20R epimers (Fig. 4). The configuration of steranes reflects an origin from crude oil and mature rocks (Perters and Moildovan 1993). The relative abundances of three regular sterane series are in an order of C29 > C27 > C28, suggesting a lacustrine sourced crude oil from China that contaminate the sediments (Sheng et al. 1993). Maturity-indicators based on steranes such as C29αααS/(S + R) ration (0.5 at equilibrium) show only minor variations among the analyzed samples (0.51–0.67) and have the high maturity level. And as a result, it is likely that the petroleum-derived hydrocarbons are introduced into the river through illegal dumping of petroleum waste.

Fig. 4
figure 4

Representative mass fragmentograms of m/z 217 for steranes from sediment at site S43. 1 C27ααα-cholestane(22S); 2 C27αββ-cholestane(22R); 3 C27αββ-cholestane(22S); 4 C27ααα-cholestane(22R); 5 C28ααα-ergostane(22S); 6 C28αββ-ergostane(22R); 7 C28αββ-ergostane(22S); 8 C28ααα-ergostane(22R); 9 C29ααα-stigmastane(22S); 10 C29αββ-stigmastane(22R); 11 C29αββ-stigmastane(22S); 12 C29ααα-stigmastane(22R)

Full size image

Both hopanes and steranes provided convincing evidence that oil-derived hydrocarbons are widespread in the sediments of the study area.

Polycyclic aromatic hydrocarbons

PAHs contents in sediments and toxic assessment

Polycyclic aromatic hydrocarbons are pollutants that are ubiquitous in aquatic environments and a product of combustion and fossil fuels as well as from oil pollution. Total PAH concentrations (sum of the 16 EPA priority pollutants) for the six analyzed samples ranged from 17.7 to 407.7 ng g−1 dry wt with the mean value of 212.7 ng g−1 (Table 2). If the relatively high PAH value for S15 (407.7 ng g−1) sourced from combustion (discussed later), PAH levels in Beiluohe were <100 ng g−1, which is indicative of low pollution (Baumard et al. 1998). Total concentration of potentially carcinogenic PAHs (PCPAH7) including BaA, Chry, BbF, BkF, BaP, Indeno, DBahA (Savinov et al. 2003) ranged from 5.2 to 232.9 ng g−1 dry wt in this study. The geometric mean of PCPAH7 is 119.8 ng g−1 dry wt. The greatest concentration of PCPAH7 was also detected at location S15 (232.9 ng g−1).

Table 2 Concentrations (ng g− 1 dry wt) and characteristics of PAHs in surface sediments from Beiluohe River
Full size table

Concentration levels of total 16 PAHs in surface sediments from the Beiluohe River were relatively low in comparison with many other aquatic systems in China (Ma et al. 2001; Wu et al. 2003; Zhang et al. 2004; Qiao et al. 2006). In order to assess whether sediments in Beiluohe River will cause toxic effect, the PAH levels in sediment were compared against effect-based guideline values such as the effect range-low (ER-L) (4,002 ng g−1) and effect range-median (ER-M) (44,792 ng g−1) developed by the US Natural Oceanic and Atmospheric Administration (Kim et al. 1999). The concentrations below the ER-L value represent a minimal-effect range, i.e., adverse biological effect would rarely be observed below the ER-L. On the other hand, if the concentration was higher than the ER-M value, adverse effects on biological systems will frequently occur. The concentration of total 16 PAHs is significantly lower than the ER-L value. Of the individual compounds, all PAHs in Beiluohe River did not exceed their respective ER-L values (Fig. 5). As a result, it may be concluded that PAHs will not cause immediately adverse biological effects in Beiluohe River sedimentary environment.

Fig. 5
figure 5

Composition of PAHs (ng g−1 dry wt) in sediments from the Beiluohe River. Nap naphthalene; Acy acenaphthylene; Ace acenaphthene; Flu fluorene; Phe phenanthrene; Ant anthracene; Fluor fluoranthene; Pyr pyrene; BaA benzo[a]anthracene; Chry chrysene; BbF benzo[b]fluoranthene; BkF benzo[k]fluoranthene; Bap benzo[a]pyrene; IP indeno[1,2,3-cd]pyrene; DBahA dibenzo[a,h]anthracene; BghiP benzo[ghi]perylene

Full size image

PAHs composition and possible sources

The study of the compositional patterns of PAH mixtures can provide useful information regarding their sources and transport pathways. In general, two types of PAHs are commonly found in aquatic sediments. Pyrolytic PAHs derived during the combustion of all the organic materials and petrogenic PAHs, which are contained in petroleum and its products. Pyrolytic PAH mixtures are characterized by predominance of parent compounds with four or more aromatic rings, whereas in petrogenic PAHs, the two to three aromatic ring compounds are most abundant components (Soclo et al. 1999). Several molecular ratios such as naphthalene/fluoranthene (Nap/Fluor), phenantharene/anthracene (Phe/Ant), fluoranthane/pyrene (Fluor/Pyr), benzo[a]anthracene/chrysene (BaA/Chry) and pyrene/benzo[a]anthracene (Pyr/BaA) have been developed for interpreting PAH compositions and inferring the possible sources. The LMW/HMW is the ratio of the sum of the low molecular weight PAH concentrations to the sum of high molecular weight PAH concentrations, i.e., Nap + Acy + Ace + Flu + Phe + Ant/Pyr + BaA + Chry + BbF + BkF + BaP + IP + DBahA + BghiP. Since no alkylated PAHs derived from oil were determined in this study, concentration ratios of parent PAH including Phe/Ant, Fluor/Pyr, BaA/Chry and LMW/HMW, were used to identify the possible PAH origins in sediments from Beiluohe River (Table 2).

Petroleum often contains more thermodynamically stable compounds such as naphthalene, fluorine, phenanthrane and chrysene, while fluoranthene and pyrene are usually the most abundant compounds for pyrolytic PAHs. Therefore, the Fluor/Pyr ratio is obviously related to pyrolytic origins as the values greater than 1, while values less than 1 are attributed to petrogenic sources (Baumard et al. 1998). The Phe/Ant ratio is observed to be very high in PAH petrogenic pollution, but low in pyrolytic contamination cases (Gschwend and Hites 1981; Soclo et al. 2000). Budzinski et al. (1997) suggested that sediments with Phe/Ant > 10 were mainly contaminated by petrogenic inputs and Phe/Ant < 10 was typical of pyrolytic sources. At site S15, the ratio with Phe/Ant < 10 and Fluor/Pyr > 1, clearly indicates a pyrolytic source in sediment. In addition, the low ratio of LMW/HMW (0.17) supports its pyrolytic origin. At sites S17, S39 and S43, ratios of Phe/Ant were higher than 10 and ratios of Flu/Pyr were lower than 1, clearly indicating that PAH inputs were mainly from the petroleum contamination. Finally, sites S40 and S64 with the ratio of Phe/Ant > 10 and Fluor/Pyr > 1 were characteristic of a mix pattern of pyrolytic and petrogenic contamination. Similar results were observed for the BaA/Chry. Chrysene and benzo[a]anthracene are both derived from the combustion processes with BaA/Chry ratio lower than 1 (Gschwend and Hites 1981). The suggested ratio values for crude and fuel oil range from 0.24 to 0.40. The petrogenic PAH inputs at four sites (S17, S39, S40 and S43) in Beiluohe River are also confirmed by the BaA/Chry ratio, ranged between 0.18 and 0.32. But at sites S15 and S64, the ratio of BaA/Chry, which is 0.86 and 0.46, respectively, confirms their pyrolytic input.

Conclusions

The Beiluohe River showed to be only slightly or moderately contaminated by hydrocarbons in most surface sediments. Concentrations of “total” hydrocarbons in Beiluohe River sediments are shown to be substantially higher than those from pristine environments. The most contaminated stations (i.e., S87, concentrations >200 μg g−1) are shown to be associated with illegal discharges of petroleum or petroleum wastewater from platforms and refinery factories. Samples taken from locations far from petroleum activities show comparatively little contamination (total hydrocarbon concentrations <15 μg g−1 dry wt). Gas chromatographic distribution patterns of n-alkanes show a strong UCM with a small predominance of odd on even numbered n-alkanes in most samples revealing a petroleum-related source. Besides oil-related hydrocarbon inputs at this region, and aliphatic hydrocarbons analysis showed also the presence of hydrocarbons of terrestrial origin at some sampling site, characterized by high CPI values. Both hopanes and steranes in analyzed samples provided convincing evidence that oil-derived hydrocarbons are widespread in the sediments of the study area.

In general, the concentrations of PAHs (sum of 16 isomers) in Beiluohe River sediments are low by comparison with those observed in other regions. The highest concentrations of total PAHs were observed at site S15 (407 ng g−1 dry wt). Total and individual PAHs in Beiluohe River did not exceed their respective ER-L values. It may be concluded that PAHs will not cause immediate adverse biological effects in Beiluohe River sedimentary environment. The molecular ratios indicate a pyrolytic source from the site S15. Petrogenic contributions were recorded at sites S17, S39 and S43. And a mixture of pyrolytic and petrogenic PAHs was observed in the sediments at sites S40 and S64. Except for site S15, good correlation between total hydrocarbons and PAH was observed indicating similar primary sources for the two classes of compounds.