Distribution and Bioconcentration of Polycyclic Aromatic Hydrocarbons in Surface Water and Fishes

To examine spatial distribution and bioconcentration of PAHs, water and fish samples were collected from Pearl River Delta in summer and spring, respectively. Particulate organic carbon, dissolved organic carbon, biodegradable DOC (BDOC), and chlorophyll a were measured. PAHs were dominated by 2- and 3-ring compounds in the water and SPM samples. Aqueous and solid-phase PAHs, respectively, showed significant correlations with total organic matter (TOC) in SPM or dissolved organic matter (DOC) in the water. The in-situ partitioning coefficients (logK oc, mL/g) for the samples were observed to be related to logK ow, implying that the hydrophobicity of PAHs is a critical factor in their distribution. It was also observed that BCF increased with the increasing K ow in the viscera of tilapia (logBCF = 0.507logK ow − 1.368, r = 0.883). However, most of the observed log BCF values in other different fish tissues at first increased with the increasing of log K ow, then reached a maximum value when logK ow is between 5 and 7, and then decreased when logK ow is higher than 7, indicating that the value of BCF may vary due to the diversity of fish species.


Introduction
Polycyclic aromatic hydrocarbons (PAHs), which are listed as US-EPA and EU priority pollutants, are widely distributed in the environment. PAHs are produced primarily as a result of incomplete combustion of fossil fuels and other organic materials, as well as from forest fires [1]. PAHs in seawater depend on their chemical properties. PAHs with low molecular weight can enter atmosphere by evaporation, while nonvolatile PAHs with high molecular weight could contaminate surface water through atmospheric deposition [2]. Due to their carcinogenic and mutagenic effects to both terrestrial and aquatic organisms, PAHs have attracted much attention.
Many investigations focus on the transport and fate of PAHs in aquatic environment [3]. Qiu et al. [4] examined the level of 15 PAHs in seawater, suspended particulate matter (SPM), surface sediment, and core sediment samples of Deep Bay, South China. Recently, distributions, composition, and sources of polycyclic aromatic hydrocarbons (PAHs) in sediments and suspended particulate matter (SPM) from the Pearl River Delta have also been evaluated [5,6]. Major environmental factors in mediating PAH levels in the sediments as well as bioaccumulation patterns in fish were identified at Mai Po Marshes [7].
Although numerous studies have investigated the occurrence of PAHs in various compartments of the PRD, data on fish species are limited [8,9]. PAHs in fish tissues, for example, fish liver, skin, or gills, which could provide more evidence for the bioaccumulation of PAHs and reflect the environmental conditions, have not been investigated. Previous investigations in Pearl River Delta mainly focused on the source, distribution, migration, and fate of PAHs. However, their environmental processes such as the transformation and enrichment of PAHs have rarely been conducted.
The Pearl River Delta (PRD) has three main tributaries, which are the Xijiang (West) River, the Beijiang (North) River, and the Dongjiang (East) River, and flow into the South China Sea. They form one of the largest rivers in China. PRD endures a significant urbanization and industrialization in recent three decades. It is located in the northern subtropical zone, where the climate is characterized by mild temperatures and frequent rainfalls all years around, 2 The Scientific World Journal facilitating the transport of contaminants to the aquatic environments. Owing to high population density, massive use of chemicals, and intensive industrial and agricultural development in this area, significant air and water pollutions occur [10,11]. With dramatic increase in aquatic environment pollutions in this region, the local fishery resource, biomass, and biodiversity decline continuously. For example, fish species in the Pearl River Estuary sharply decreased from more than 200 species in 1970s to 50 species in recent years, and the proportion of the large size fish dropped from nearly 50% in 1980s to lower than 10% in this century [8,12].
The present study aimed to determine the spatial distribution and partitioning of PAHs in water/SPM and the affecting factors, and the distribution and accumulation of PAHs in different species of fishes in order to evaluate the enrichment regularities of PAHs among water, SPM, and fish organisms.

Sample Collection.
Water and SPM samples at 0.5 m below the surface were collected from Pearl River Delta in July 2010 and April 2011, respectively ( Figure 1). Meanwhile fish samples were collected at site D7 (xenocypris davidi, Bleeker) and site D8 (tilapia, blunt snout bream, Cirrhinus mrigala) in July, 2010, and at site DJ-5 (red grass carp, blunt snout bream named as blunt snout bream-2, carp) in April 2011. Water samples were pumped into precleaned 10 L brown glass bottles with a stainless-steel submersible pump. NaN 3 was added to each bottle to inhibit biodegradation of PAHs. pH, conductivity, and salinity were measured immediately at the sites by using a digital pH meter with dissolved oxygen meter and salinometer (MP511,Shanghai). All of the parameters are listed in Tables 1 and 2. Some of the fish samples were also bought from small fishing boats along the river. The water samples were filtered through the 47 mm glass fiber filters (Whatman GF/F, 0.7 um pore sizes) precombusted at 450 • C for 4 h beforehand. Then, the GF/F filters were stored at −20 • C until analysis. Fish samples were dissected carefully to obtain muscles, gills, and viscera. These samples were also stored at −20 • C until analysis.

4
The Scientific World Journal phenanthrene-d 10 , chrysene-d 12 , and perylene-d 12 ). The Envi-C 18 SPE cartridge was sequentially conditioned with 5 mL of ethyl acetate, 5 mL of methanol, and 5 mL distilled water containing 2% methanol. Then, the water sample passed through the preconditioned Envi-C 18 SPE cartridge at a flow rate of 8-10 mL/min. The cartridge was cleaned with 5 mL distilled water, dried under vacuum for 15 minutes, and eluted with 3 × 5 mL of ethyl acetate. Finally the elution was vacuum-evaporated to 1 mL and concentrated to 100 µL under a gentle nitrogen stream. Particle-loaded filters were freeze dried, weighed, and spiked with surrogate standards and Soxhlet extracted for 72 h with 200 mL of dichloromethane (DCM). Each extract was concentrated, solvent exchanged to hexane, and reduced to approximately 1 mL. A 1 : 2 alumina : silica gel glass column was used to purify the concentrated extracts. Then, the column was eluted with 15 mL n-hexane and 70 mL 7 : 3 hexane/DCM (v/v) successively. The second fraction containing PAHs was also finally concentrated to 100 µL under a gentle N 2 stream before GC/MS analysis.
Fish tissue samples were freeze dried, spiked with surrogate standards, and Soxhlet extracted for 72 h with 200 mL of dichloromethane (DCM). Each extract was concentrated to about 5 mL and divided into two fractions. One fraction was used to determine the content of lipid by weight method, and the remaining fraction was used to determine the concentration of PAHs in fish tissue. The remaining fraction passed through a gel permeation column to remove lipid. The elution solvent from 90 to 280 mL was collected and concentrated by a rotary evaporator. Then, the concentration extract was again cleaned by an alumina/silica gel column. And the subsequent analytical procedure was the same as that of SPM. The fraction containing PAHs was also finally concentrated to 100 µL before GC/MS analysis.

Instrumental
Analysis. Sixteen PAHs were quantified by a Hewlette Packard (HP) 6890 gas chromatograph (GC) coupled to a HP 5975 mass spectrometer (MS) with a DB-5 fused silica capillary column (30 m × 0.25 µm × 0.25 mm i.d.). The system was operated in electron impact mode (EI) and detected by using selective ion monitoring mode (SIM) with helium as the carrier gas at a constant flow rate of 1 mL/min. The oven temperature was programmed from 60 • C to 200 • C at 10 • C/min, to 214 • C at a rate of 2 • C/min and to 255 • C at 5 • C/min and held for 2 min and further increased to 290 • C at 20 • C/min and held at 290 • C for 12 min. The concentrations of PAHs in the water and suspended particle matter were quantified by using the isotope dilution method with isotope-labeled internal standards (d8-Nap, d10-Acy, d10-Phe, d12-Chry, and d12-Per). PAHs in fish tissues were quantified with the internal calibration method based on five-point calibration curve.
Ten mL of each water sample passed through the GF/F filter was acidified with HCl to pH = 3 and then used for DOC analysis. TOC analyzer (TOC-VCPH, Shimadzu) was used to measure the DOC concentration. For the determination of TOC in SPM samples, the filters were dried at 60 • C for 12 h after acidification with diluted HCl to remove carbonates. TOC in particle samples was measured using an elemental analyzer (Vario EL III Elementar, Germany) with acetanilide as external standard. For the determination of biodegradable DOC (BDOC), approximately 300 mL of filtered water was contained into 500 mL precombusted brown glass bottles and incubated in the dark at about 20 • C for 30 d. The difference of DOC concentrations measured before and after incubation were regarded as the BDOC concentrations.
Samples for determining Chl a were filtered through 0.45 um cellulose acetate filters, and then the membrane samples were extracted with 90% acetone for 24 h. Chl a was determined by using a UV-VIS spectrophotometer (752, UV-2000, Shanghai). The absorbency at wavelength of 663 nm, 645 nm, 663 nm, and 750 nm were measured. Chl a was calculated by the following equation [15]: where D630, D645, D663, and D750 represent the absorbency of 630, 645, 663, and 750 nm, respectively. The V 1 , V , L, and C represent the volume of water samples (L), the thickness of cuvette (cm), and the concentration of chlorophyll a (µg/L), respectively.

Quality Assurance and Quality Control (QA/QC).
Procedural blanks and spiked blanks were analyzed with field samples, and surrogate standards (d8-Nap, d10-Acy, d10-Phe, d12-Chry, and d12-Per) were also added to all the samples to monitor procedural performance. Except for Nap, 10.05 ng/L of total PAHs was detected on average in water blanks (n = 4), and 17.21 ng/L of total PAHs was detected in particle blanks (n = 5). The recoveries of 16 PAHs in spiked blanks (n = 3) varied from 50.9% (Nap) to 122.7% (BgP). Because of the high background values for Nap, total concentrations of PAHs did not include Nap. Phe was also not considered in the distribution of total PAHs in water and SPM samples due to the possible pollution during the process of experiments. And the reported PAHs concentrations were corrected with the blank values.  (Figure 2), suggesting that DOC was highly biodegradable within the time scale of the incubation (1 month). Chl a concentrations were also related to the PAHs concentrations (Figure 3), which indicated that PAHs can be easily absorbed by algae in the SPM. It was also implied that algae was the dominant composition of POC and governed the distribution of particulate PAHs. An increasing phytoplankton biomass and growth rate was reported to increase the air-water transfer of PAHs [16]. New phytoplankton production contributed to substantially longer times for air-water equilibrium, and the depletion of the dissolved phase by phytoplankton uptake prevented the equilibrium of air and water phases. The gas phase supports the concentrations of organic pollutants such as PAHs in atmospherically driven environments. Increased air-water exchange followed by phytoplankton uptake also seemed to be two of the most relevant processes increasing the vertical flux in the water column [17]. Individual PAHs also showed considerable variances among the samples ( Figure 5). For all the water samples, low molecular weight PAHs were the dominate compounds. The percentage of 3 and 4-ring PAHs ranged from 42.06% to 81.09% with an average of 65.38%. Acenaphthylene, fluorene, fluoranthene, and pyrene were the major constituents of PAHs in the water samples.

Concentrations of
PAHs in the water samples are compared with those of other investigations ( Table 3). The total concentration of PAHs was 2 to 3 orders of magnitude lower than those reported in Daya Bay, China (4181-27507 ng/L) [18], Jiulong River Estuary, and Western Xiamen Sea, China (6840-25620 ng/L) [19] and approximately 2 orders of magnitude lower than those found in Pearl River and the Macao Harbor, China (691-6457 ng/L) [20]. However, high PAHs in the Pearl River and the Macao Harbor [20] were related to the contamination of naphthalene in the laboratory, which accounted for about 90% of the total PAHs in dissolved phase. The aqueous concentrations in this study are similar to those of the previous investigations if Nap is excluded (13.64-106.85 ng/L). On the other hand, PAHs in this investigation were several times higher than those found in Beltic Sea (3.85-14.1 ng/L) and in the North Sea (0.63-3.51 ng/L) [21]. Similar concentrations were found in Xijiang River, China (21.7-138 ng/L) [5], Pearl River Delta (10.8-323 ng/L) [14], and Chesapeake Bay, USA (20-65.7 ng/L) [22].    It was found that PAHs in SPM was higher in Pearl River than in Dongjiang River in spring.
Like the water samples, low molecular weight PAHs in SPM were also the dominant compounds. However, PAHs in this study are at different levels compared with the previous investigation in other areas. PAHs in SPM in this study were 2 orders of magnitude higher than those of the particulate samples collected from other regions (Table 3), such as the Xijiang River (0.17-58.2 ng/L) [5] and six to seven times higher than York River of the VA Estuary (2.09-123 ng/L) [23]. They are at similar level to the concentrations of PAHs in Pearl River and the Macao Harbor, China (150-431 ng/L) [20] and the Seine River and Estuary, France (2-687 ng/L) [24]. found. The highest concentration of PAHs was detected in red grass carp, ranging from 46.85 to 236.14 ng/g dw. It was approximately 2 to 3 times higher than other fishes. And the lowest PAHs levels occurred in tilapia (collected in summer), ranging from 14.70 to 80.51 ng/g dry weight. However, there were no significant differences among the other fish species. In terms of the individual PAHs, low molecular weight PAHs were the major compounds in the fish species, which are similar to those of the water and SPM samples. Compared with PAHs in the muscle (184-194 ng/g dw) and viscera tissues (505-854 ng/g dw) in different sized tilapia reported for Mai Po Marshes by Liang et al. [7], PAHs here both in muscle (14.55 ng/g dw) and viscera (80.51 ng/g) were much lower. This difference might be caused by the feeding habits of different fish species in different aqueous environment.

PAHs in the Fish Species.
Significantly different concentrations of PAHs were also observed among fish tissues. Because the visceras of Cirrhinus mrigala, red grass carp, blunt snout bream-2 collected in April 2011, and carp were mashed, only the data of their muscle and gills were present. The highest concentrations of PAHs were found in the visceras, ranging from 80.51 to 180.87 ng/g dry weight, followed by the concentrations in gills, ranging from 25.43 to 236.14 ng/g dw, and those in muscle (10.52 to 46.85 ng/g dw) are the lowest. The different concentrations of PAHs in fish tissues may be affected by the physical-chemical properties of PAHs, the lipid content, and the uptake capacity of different fish tissues [9].

Association of PAHs with DOC in Water and with POC in SPM.
One of the important factors affecting PAHs in the water and SPM samples was DOC and POC. Correlation analyses between PAHs and DOC or POC were illustrated in Figure 7. Although aqueous PAHs showed no significant correlations with DOC in summer, positive correlations were found between aqueous PAHs and DOC in both the Dongjiang River (r = 0.736, P < 0.05) and the Pearl River (r = 0.78, P < 0.01) in spring. For the particulate samples, PAHs in SPM was significantly related to POC in summer    (r = 0.695, P < 0.05) in Dongjiang River and in both the Pearl River (r = 0.625, P < 0.05) and Dongjiang River (r = 0.783, P < 0.05) in spring. The highly significant correlation between PAHs and organic carbon indicated that both DOC and POC are important to the distribution of PAHs in aquatic environment. Moreover, the slopes in Figure 7 demonstrate the importance of DOC to the association of PAHs. The slopes are −0.788, 12.19, and 7.63 ng/mg for DOC, and 50.71, 37.97, and 22.79 ng/mg for POC in the Dongjiang River in summer, in the Dongjiang River and in the Pearl River in spring, respectively. Hence, PAHs should be greatly affected by POC than by DOC in the targeted river system.
It is also widely acknowledged that K oc is closely related to K ow [25]. Hence, hydrophobic compounds such as PAHs with higher K ow show stronger affinity to POC or DOC. The four dominant PAHs (acenaphthylene, fluorene, fluoranthene, and pyrene) in the dissolved and the particulate phases were normalized by DOC and POC, respectively (Figure 8). It was found that the mean POC-normalized concentrations for Flo, Flu, and Pyr were 21.79, 13.84, and 12.74 µg/g oc, respectively; and the mean DOC-normalized concentrations were 1.57, 1.50, and 2.67 µg/g oc, resp.). The formers were over one order of magnitude higher than the latter ones. And the POC-normalized concentration for Ace (2.25 µg/g oc) was similar to its DOC-normalized concentration (2.25 µg/g oc) ( Figure 5). As Flo, Flu, and Pyr are hydrophobic nature with log K ow of 4.18, 4.90, and 4.88, they are readily associated with POC and accumulated in SPM. The partitioning patterns of PAHs further reveal that POC and DOC are the most important factors in controlling their distribution, transport, and fate in the surface river water.

Distribution Coefficients of PAHs between Water and SPM.
Distribution of PAHs between SPM and water plays a very important role in the mobility and fate of PAHs in aqueous systems. The most frequently used parameter for evaluating their distribution is the organic carbon-normalized particlewater partitioning coefficients K oc , which were calculated as follows: where C s is the solid phase concentration (ng/g), C w is the aqueous phase concentration (ng/mL), and f oc is the mass fraction of organic carbon in the particle. From Figure 9, log K oc mL/g was significantly related to log K ow for the samples collected both from the Dongjiang River (r = 0.577) and the Pearl River (r = 0.897), implying that PAHs with high hydrophobicity can be adsorbed on SPM more easily. The free energy relationship between log K ow and log K oc was established in Figure 9. The observed equation for PAHs is similar to the previous investigation on the log K oc − log K ow regression for PAHs in the water of the PRD [20].
From the slope of the equation in Figure 9, the lipophilicity of SPM relative to the reference octanol/water system may be inferred. The slope in this study is lower than the value listed in Table 4, suggesting that the lipophilicity of SPM in this is relatively low. organisms, since PAHs are easily accumulated in lipid-rich tissue of fish. The lipid contents in different tissues of each fish species are shown in Figure 6. The highest lipid contents were presented in the gill tissues, ranging from 15.3% in blunt snout bream-2 (collected in spring) to 46.5% in tilapia with an average percentage of 27.4% dw, followed by those in viscera tissues ranging from 16.68% in tilapia to 33.93% in xenocypris davidi Bleeker with the average percentage of 26.03%. The lowest lipid contents were found in the muscle tissues, varying from 1.96% in tilapia to 6.79% in blunt snout bream with an average percentage of 4.7%. Relationship between PAHs and lipids in different tissues of fish species was showed in Figure 10. Significant positive correlations between total PAHs and lipids in different tissues of fishes from Dongjiang River were obtained (r = 0.859, P < 0.0001).

Bioconcentration Factors (BCFs) in Fish Tissues.
In order to compare between the bioaccumulation patterns among individual PAHs in fish, BCFs were calculated and plotted. BCF in the viscera of tilapia was observed to increase with the increasing K ow (log BCF = 0.507 log K ow + 1.631, r = 0.883) (Figure 11(a)). This observation is consistent with other investigation on bioconcentration [26], suggesting that the bioaccumulation of organic chemicals in biota increases with the increasing of log K ow . However, measured BCF values tend to decline below the true equilibrium condition as the K ow of the chemical increases. BCF will reach the maximum value when log K ow reaches 5-7 and then decreases when log K ow is higher than 7. Except for the viscera of tilapia, most of the BCF values in different fish tissues follow this trend as demonstrated in Figure 11(b). This difference may depend on living habitat and trophic levels of fish and environmental behaviors of PAHs. The bioavailability, uptake, and fate of PAHs by aquatic organisms from contaminated media (water, sediments, and food) were also affected by a variety of physical (e.g., lipophilicity, temperature, etc.) and biological parameters. As a general rule, water is dominant pathway of exposure for fish if log K ow of organic compounds are lower than 5, while sediment particles can be used for some fish species such as food and can contribute substantially to bioaccumulation for PAHs with log K ow higher than 5 [27]. As Tilapia used to live in the bottom layer of aquatic ecosystem and in sediments PAHs in sediments may contribute to accumulation in tilapia via the dietary route and exchange with water through gills.
The previous result indicates that BCF is quite different due to the diversity of fish species.
The previous result is consistent with previous studies which illustrated that the relationship between K ow and bioconcentration appeared to be relatively complex [28]. For fish with low feeding rates, the bioaccumulation of nonpolar organic compounds with log K ow ranging from 2 to 6.5 is mainly determined by exchange across the gills. However, dietary uptake seems negligible because of poor absorption efficiency and rapid metabolism rates [29,30]. For nonmetabolized PAHs, less bioaccumulation of organic compounds would appear in small fishes due to the higher oxygen uptake, which leads to much loss than uptake of organic compound, than in large ones [29]. It was also suggested that bioaccumulation of low K ow PAHs (such as naphthalene) was high due to higher gill transfer efficiencies, and that of high K ow was low because of enhanced biotransformation and decreased gut assimilation in fish [31,32]. Moreover, the physicochemical properties of a chemical, the physiological components of the uptake process, biotransformation, blood flow, and fatty acid composition and lipid content in aquatic animals could all affect the uptake and accumulation of organic chemicals for fishes [28]. All of those factors contribute to the unpredictability of the bioaccumulation of POPs in fishes.

Conclusion
The spatial distribution and bioconcentration of PAHs in the water, SPM, and fish species from the Pearl River Delta were   examined. Aquatic chemical data were also determined. In both the dissolved and the particulate phases, the low molecular weight PAHs were the dominant components. Positive correlation were found between aqueous PAHs and DOC as well as particulate PAHs and POC, indicating the importance of DOC and POC to the distribution of PAHs in the aquatic environment. The in-situ partitioning coefficients (log K oc , mL/g) for the samples were related to log K ow . The relative lipophilicity of SPM could be evaluated by the slope of the observed regression equation. PAHs showed significant correlations with lipids in different tissues of fishes. BCF in the viscera of tilapia was positively related to log K ow . But BCF values in most of the fish samples were found to reach the maximum value when log K ow reaches 5-7 and then decrease when log K ow is higher than 7. The different distribution of PAHs among the fish species and their tissues were affected by log K ow of PAHs and the lipid contents in fish tissues.