Source and Ecological Risk Characteristics of PAHs in Sediments from Qinhuai River and Xuanwu Lake , Nanjing , China

In order to investigate the residual characteristics, sources, and ecological risk of PAHs in sediment fromurban rivers, the sediments of 15 typical sites from Qinhuai River and Xuanwu Lake, which are typical urban rivers and lake, were collected from October 2015 to July 2016; the sources of PAHs in sediment were also identified by several methods. Results showed that ∑PAHs concentration in sediment ranged from 796.2 ng/g to 10,470 ng/g with an average of 2,713.8 ng/g. High molecular weight PAHs with 4-5 rings were most prominent in the sediment during all four seasons. Source characterization studies based on the analysis of diagnostic ratio (triangular plot method), cluster analysis, and positive factor matrix analysis suggested that the PAHs of Qinhuai River Basin were mainly from pyrogenic origin (biomass and coal combustion and vehicular emission), and the petroleum source also cannot be ignored (specially in summer). Most individual PAHs occasionally affect the aquatic organisms. The highest benzo[a]pyreneequivalent doses (BaPeq dose) appear at the sites of sewage discharge and heavy traffic. So, the PAHs pollution sources of urban water body have obvious seasonal-dependent and human activities-dependent characteristics.


Introduction
Polycyclic aromatic hydrocarbons (PAHs), a group of typical chemicals containing two or more aromatic rings, are widely distributed in water [1], soil, sediment [2][3][4][5], and air [6].PAHs have been attracting increasing attention, because many are known to be carcinogenic, mutagenic, or teratogenic [7,8], which are serious health threats to organisms.Hence, United States Environmental Protection Agency (US EPA) has listed sixteen PAHs as priority pollutants.In recent years, PAHs research has focused on the distribution and source apportionment of sediments.PAHs may come from natural sources, but anthropogenic activity is generally considered the major source [9].Anthropogenic PAHs primarily originate from pyrogenic and petrogenic sources.Pyrogenic PAHs are mainly from incomplete combustion of various fossil fuels (such as coal, oil, and natural gas), residential heating, industrial activity, wood, tobacco, and other hydrocarbons [10].Petroleum sources also generate PAHs [11], including crude oil and refined petroleum products such as oil leakage [12].
PAHs, once produced, enter the aquatic environment via wastewater, surface runoff, atmospheric deposition, and oil leakage [13].Due to their low water solubility, PAHs are easily adsorbed by particles and colloids when transferred into the water and sediment [14].After settling into the sediment, PAHs will rarely photochemically and biologically oxidize and, therefore, will accumulate to high concentrations in sediments [15].The concentration of PAHs in sediments is related to sediment type and size, geographical location, water migration, and other factors.When environmental conditions change, adsorbed PAHs may be rereleased into the water by chemical and biological processes, leading to secondary pollution of the surrounding environment [16,17].Sediments act both as a reservoir and a secondary source for PAHs contamination [18].
Until recently, researches on surface sediment PAHs mainly concentrated in the estuary area of some rivers and lakes [19][20][21][22][23][24][25].Sofowote et al. (2008) investigated the profiles and relative contributions of pollution sources to the Hamilton harbor using principal component analyses (PCA) and positive matrix factorization (PMF) methods, and results showed both methods identified four factors and gave excellent correlation coefficients between predicted and measured levels of 25 aromatic compounds; both methods predicted similar contributions from coal tar/coal combustion sources to the harbor (19 and 26%, resp.), but PMF afforded better source identification than PCA [26].Feng et al. (2007) studied the distribution, source, and ecological risk assessment of PAHs in suspended solids and sediment from the Yangtze River in Wuhan [27].Chen et al. (2007) have researched the Qiantang River in the Yangtze River Delta region, mainly to analyze the temporal and spatial distribution and sources of PAHs in its water and sediments [28].So far, relatively little research has focused on the seasonal distribution, source analysis, and ecological risk assessment of PAHs in a typical urban river [29].Qinhuai River, a tributary of the Yangtze River, is located in the southwest of Jiangsu Province.With a total length of 110 km and the basin area of 2630 km 2 , it is the largest regional river in Nanjing.Internal Qinhuai River is a typical urban river, flowing across the Nanjing city, which directly affected the sustainable economic development of Nanjing city.Xuanwu Lake is the largest shallow lake of Nanjing (3.7 km 2 ), a typical urban shallow lake, which has been subjected to a series of human disturbances, such as the inflow of domestic sewage, surface rain runoff from the surrounding residents and mountain, water diversion from drink water treatment plant (1.8 × 10 5 t/d), tourist entertainment, and rubbish in lake park.The objective of this article is to explore the applicability of source appointment technologies and the effects of city activities on the distribution and ecological risk of PAHs in surface sediments, provide theoretical support for the pollution sources controlling and desilting treatment, and make most of PAHs removed from the river, to reduce the ecology risk of urban river and lake.

Sampling Region and Sites.
A map showing the location of the sampling sites is provided in Figure 1.Corresponding functional regional characteristics are shown in Table 1.Sediment sampling was carried out along the inner Qinhuai River and Xuanwu Lake in four seasons.The sampling time, based on local seasons, is grouped as follows: autumn (October, 2015), winter (January, 2016), spring (April, 2016), and summer (July, 2016).The middle month of different season was chosen as the typical month of sampling.According to the statistical data estimated by local government, the sedimentation rate is about 10 cm-60 cm a year for different river.Moreover, the sampling tool is the grab sampler, so the sediment samples can be used as representative samples of different seasons.Surface sediment sampling was collected by grab sampler.Each sample consisted of central parts of 5 subsamples randomly collected at 5 points within the study area.The samples were placed into the sampling bottles, sealed and stored at −20 ∘ C in the refrigerator for further analysis.

PAH Extraction, Cleanup, and Determination.
For PAHs analysis, sediments were freeze-dried, crushed into fine powders, and passed through a 100-mesh sieve.Five grams of sieved samples was mixed with 2 g anhydrous sodium sulfate and ultrasonic-extracted for 10 min with 15 mL n-hexane: acetone (2 : 1 V/V) after soaking for 1 h.The supernatant was transferred to a pear-shaped flask; then the residual sediment was ultrasonic-extracted again with 15 mL n-hexane: dichloromethane (3 : 1 V/V) as described above.All the supernatant was combined to a 50 mL pear-shaped flask and concentrated to about 1 mL by vacuum rotary evaporation apparatus (RE-3000, Yarong Biochemical Instrument Plant, Shanghai, China).For purification, the concentrate was transferred to solid phase extraction column (upper: 1.0 g anhydrous sodium sulfate; the second layer is 0.5 g copper, the third layer is 1 cm silica gel, and 1 cm neutral alumina is at bottom) (Borui Co., Ltd., Tianjin, P. R. China) rinsed with 5 mL n-hexane, which had been rinsed with 5 mL nhexane with a drop-wise flow rate.The column was eluted with a mixture of 25 mL n-hexane: dichloromethane (3 : 1, V : V).The eluent was collected into a pear-shaped flask and concentrated to 1 mL for HPLC determination.
Among the reagents, acetone, n-hexane, and copper powder were analytical reagent; anhydrous sodium sulfate was guaranteed reagent grade (Nanjing Chemical Reagent Limited Company, Nanjing, China); neutral alumina was chromatographically pure (Shanghai Ludu Chemical Reagent Plant, Shanghai, China).
known PAH levels spiked with a mixture consisting of 1 g each of PAHs), and procedural blanks were used as quality control procedures.Analyses were run in batches of 10 samples plus four quality controls (QCs) including one reagent blank, one matrix blank, one matrix spikes sample, and one random sample in duplicate.No detectable amounts of target PAHs are detected in the blanks.The recovery efficiency of PAHs ranged from 89% to 111% for HS-5 and from 78% to 105% for the spiked samples with relative standard deviation (RSD) 3.2%-10.6%.The method detection limits were in the range of 0.5-2.0ng g −1 dry weight.The accuracy and precision with PAHs are within 8% variability for both 10 ng/ml and 200 ng/ml concentrations.PAHs standard solution was diluted with acetonitrile; six concentrations of the mixed standard were configured to quantify peak area by establishing a working curve with external standard method, and the correlation of the standard curve was above 0.99.All of the measurements were performed in triplicate, the results were not corrected according to the percent recovery of HS-5, spiked samples, and blanks, and the means were used for calculations.

Positive Matrix Factorization (PMF).
The PMF model, developed by Paatero and Tapper [30], is one of the typical factor analysis models.PMF is a useful factorization method that can calculate source profile and contribution.Unlike the traditional factor analysis models, PMF model utilizes nonnegativity constraints for obtaining physically realistic meanings [31].
The principle of the PMF model is to factor the initial n × m data matrix X (n: number of samples; m: number of species) into (n × p) matrices G (source contributions) and (p × m) matrices F (source profiles), as well as the "residual matrix" E (n × m) firstly, as follows: where G is the source contribution matrix, F is the source profile matrix, and p is the number of sources [31].Certainly, the elements of G and F are constrained to nonnegative values.
The elements   of matrix E ( × ) indicate the residual value not accounted for by the modeled data value   .
where   is the jth species concentration measured in the th sample,   is the th source's contribution to sample , and   is the jth element's concentration in source k.   is the residual associated with the jth species concentration measured in the th sample.The objective function () related to the residual and uncertainty is minimized using weighted least-squares by PMF, which is defined as where n and m are the number of samples and species, respectively; p represents the number of factors extracted;  = 1, 2, . . .,  samples;  = 1, 2, . . .,  species;  = 1, 2, . . .,  sources.  is the difference between the observations and the model;   is the uncertainty for each observation.The PMF solution minimizes the object function  based upon the given uncertainty  [32].The uncertainties for each sample were calculated using measurement uncertainties (MU percent) and method detection limits (MDLs).A discussion of calculating uncertainties is provided by Reff et al. ( 2007) [33] and USEPA (2008) [21].If the concentration ⩽ MDL, the uncertainty u is calculated as [34] and when the concentration > MDL,  is calculated as In this study, the EPA PMF 5.0 model was used [21].3), mean concentration of ∑PAHs in sediments from Qinhuai River was found to be lower than that from Calabar River (9,370 ng/g), Nigeria [35], and Haihe River, Tianjin, China (27,074.1 ng/g) [36].The maximum concentration in this study was lower than that in sediments of Haihe River (16,901 ng/g) in 2014 [29], but mean concentration was higher than other rivers.Owing to extensive economic and industrial development in Nigeria, the concentration of ∑PAHs in Calabar River was higher than  that in Qinhuai River [35].The average content (27,074.1 ng/g) of ∑PAHs in the sediment of Haihe River is about ten times that of Qinhuai River (2,713.8 ng/g); this result could be attributed to the fact that Tianjin was an old and established industrial city with the largest port in north China [37], along with rapid development of the petrochemical industry, the iron and steel industry, and other heavy industries and rapid urbanization [36].In addition to Luan River as a water source with an average of 115.3 ng/g, ∑PAHs levels of other rivers were relatively high, which were closely related to the economic development and industrialization.Qinhuai River is a typical urban river through the densely populated city of Nanjing; sewage discharges and industrial pollution emissions lead to higher mean concentration of ∑PAHs than other rivers apart from Calabar River and Haihe River.

Results and Discussion
As shown in Table 2 and Figure 2(b), average concentrations of ∑PAHs in sediments were found to be as follows: autumn (4020.1 ng/g) > spring (3,169.1 ng/g) > winter (2244.9ng/g) > summer (1421.1 ng/g), indicating that the worst sediment contamination by PAHs appeared in autumn.The uneven seasonal distribution of PAHs in sediments could be attributed to the difference of rainfall and temperature in different seasons.Strong rainfall can scour the surface, carrying the PAHs into the river [38]; high temperature could lead to their subsequent evaporation and degradation [39].So the least concentration of PAHs appeared in summer.
From Table 1 and Figure 2(a), PAHs of different sampling sites were quite spatially different.High levels of PAHs at sampling sites 1#, 3#, 4#, 5#, and 7# may be related to the inputs of surrounding pollution [49].For example, 7# is located in Yunliang River, receiving the tail water from eastern-city's sewage treatment plant, while 3# and 4#, located in the middle of Internal Qinhuai River, received polluted incoming water from eastern section of Internal Qinhuai River, contributing to the high PAHs concentrations detected.High concentration PAHs was also found at site 5#, which was in the southern section of Internal Qinhuai River, and the discharges of sewage and markets sewage were observed during the sampling period.The low concentration in site 8#, located in Qiqiaoweng Wetland Park between the downstream of Yunliang River and External Qinhuai River, might be related to the clean incoming water of upper reaches from External Qinhuai River and natural self-cleaning function of wetlands in park.Overall, mean concentrations of ∑PAHs in sediments from Internal Qinhuai River (from sites 1# to 5#) were higher than that from External Qinhuai River, and because Internal Qinhuai River flows through the urban center, frequent human activities lead to more pollution input and the deposition of PAHs in sediments caused by the low-flow rate due to the gate-control and stagnation.It is worth mentioning that the relatively low concentration ∑PAH found at the estuary (site 14#) may be attributed to the fact that this site is far from anthropogenic pressure and the probable tidal flushing of pollutants adsorbed on sediment particles.So, human activities are the main pollution source of PAHs for urban water body.This result is consistent with the view of Oyo-Ita et al. [35].In the present study, the ratios of Ant/(Ant + Phe), Fla/(Fla + Pyr), and BaA/(BaA + Chr) were used to identify the PAHs sources in sediment.Ant/(Ant + Phe) ratios of >0.1 are the indicator of pyrogenic origin of PAHs, while <0.1 ratios are petrogenic origin of PAHs [51].Meanwhile, a Fla/(Fla + Pyr) < 0.4 ratio indicates a petroleum source, a ratio between 0.4 and 0.5 indicates liquid fossil fuel combustion, and a ratio of >0.50 suggests the combustion of biomass and coal [53,54].In addition, BaA/(BaA + Chr) ratio is considered as a specific marker for coal combustion [55,56].The petroleum source is demonstrated if the ratio of BaA/(BaA + Chr) is lower than 0.20; a source from combustion coal, grass, and wood is shown if the ratio is higher than 0.35; and petroleum combustion (especially liquid fossil fuel and vehicle and crude oil spillage) is demonstrated if the ratio is between 0.20 and 0.35 [51].For a more detailed analysis of sources, the diagnostic ratios of Ant/(Ant + Phe), Fla/(Fla + Pyr), and BaA/(BaA + Chr) for the four seasons are plotted in Figure 3.As shown in Figure 3, the source of each site was significantly different in summer, while the other three seasons were relatively convergent and consistent.These characteristic ratios demonstrated the PAHs in sediments of Qinhuai River mainly originated from petroleum and combustion of biomass and coal in spring, autumn, and winter is attributed to anthropogenic activities like straw burning before rice cultivation and after harvest, automobiles, and heating in winter and spring.Sources of PAHs in summer were relatively more diverse.Among the sampling sites, M14 was typical petroleum source, M13 and M15 were mixed sources of petroleum source, biomass, coal combustion, and petroleum combustion, and, additionally, M1 and M3 were biomass and coal combustion.Other sampling sites had a mixed source from pyrogenic sources and petroleum sources.

Source Estimates from Cluster Analysis.
Cluster analysis was performed to identify the homogeneous groups of individual PAHs in the Qinhuai River sediments.The result of cluster analysis is shown in the hierarchical dendrogram (Figure 4), which distinguishes the 14 individual PAHs into three major groups.The first group is subdivided into two subgroups.The first subgroup contains BaP, BkF, and BghiP, which are the high molecular weight PAHs with 5-6 rings, usually detected in pyrogenic sources, for example, combustion of biomass, coal, wood, and petroleum.The second subgroup consists of Flu and Phe, which are 3-ring PAHs, from petrogenic sources mainly caused by petroleum spills, for example, fresh or used crankcase oil, crude, and fuel oil [57][58][59][60][61].The results indicate the first group is mixed pyrogenic sources and petroleum sources.The second major group, which includes Nap, Ace, Ant, and Fla belonging to the low molecular weight PAHs with 2-3 rings, is usually detected in petrogenic sources.The third major group contains Pyr and Chr with 4 rings and BaA, BbF, and DahA with 5 rings, the same as the first subgroup of the first group, usually detected in pyrogenic sources, for example, combustion of coal, wood, vehicle fuel, and waste tire [56,62,63].The results of the second and third groups are consistent with the study by Liu et al. [12], which is attributed to the fact that both Qinhuai River and Huangpu River are urban rivers affected by anthropogenic activities (heating, cooking, and vehicles).

Source Estimates from Positive Matrix Factorization (PMF).
For further ascertaining the validity of PAHs sources, fifty-two ( 52) objects (samples) and fourteen ( 14) variables (PAHs) together with their uncertainties were used as the input data.Factor numbers ranging from three to eight were considered in order to choose the "optimal solution.""Robust" mode was adopted to simulate for eliminating the effect of individual extreme value [21,30,[64][65][66][67][68][69].Three factors were eventually chosen since it gave  (robust) value ranging from 21041.1 to 23360.0, which converged.The slope and  2 values indicated that the predicted and actual concentrations were well fitted and three factors could adequately explain the data based on the background of PAH sources in the sediment.
The source identification was based on the molecular markers in different sources of PAHs.As shown in Table 4 and Figures 5 and 6, the first factor ( 1 ) is responsible for 33.4% contribution rate, and it is heavily weighted in Ant, BaA, BbF, BkF, BaP, DahA, and BghiP with 4-6 rings.Venkataraman et al. (1994) thought that BkF was typical pollutant of diesel emissions [70].Ant, BaA, Chr, and BaP were thought to be typical pollutants of biomass or coal combustion [71,72].Indeno[1,2,3-cd]pyrene (InP), BghiP, and DahA were typical markers of traffic emission [73,74].Hence, this factor represents the traffic-related source > biomass and coal combustion of PAHs, which accounts for the larger proportion of all the sources; Nanjing has a large population and also many private cars and buses.Figure 7  illustrates the seasonal contributions of vehicular emission are spring > winter > autumn > summer.The results are the same as those studied by Liu et al. [12], indicating the choice of transport is temperature-related and people tend to choose cars in cold seasons.Factor 2 ( 2 ) is predominately composed of PAHs with 3-6 rings, including Flu, Phe, Pyr, Chr, BkF, BaP, and BghiP, which mainly come from biomass and coal combustion [75].Some researches showed that Phe, Ant, Fla, and Pyr were indicative of coal combustion [76].According to Khalili et al., gasoline combustion was shown to predominantly produce Nap, Flu, Acy, and Pyr [77].Meanwhile, BghiP and BkF have been identified as tracers for gasoline engines, with Pyr found in both diesel-and gasolinepowered engines [71].And BaA, Chr, and BaP were thought to be typical pollutants of biomass and coal combustion [71].Therefore, factor 2 ( 2 ) suggests a mixed source of biomass and coal combustion equal to or >diesel and gasoline combustion sources, which is responsible for 25.3% of the total contribution rate throughout the year.Figure 7 shows that winter has the biggest contribution from mixed sources, followed by spring, autumn, and summer.Because winter and spring temperature are relatively low, people will use coal combustion, air conditioners, and other electrical facilities for heating.Nanjing municipal electric power supply is mainly derived from coal-burning power plants.Assigning this factor to coal combustion is demonstrated and shows a seasonal variation.The third factor ( 3 ) is predominately composed of Nap, Ace, Phe, Ant, Fla, and Chr with higher loadings.The higher level of Nap and Ace relative to other PAHs suggests a petroleum source [77].Existing research showed that Ace and acenaphthylene (Acy) are the foremost product of coke burning [78].And Phe, Ant, Fla, and Chr were indicative of coal combustion [71,76].Therefore, factor 3 ( 3 ) was identified as petroleum source > coal combustion [79,80].Petroleum source and coal combustion are a major source in the Qinhuai River basin, contributing about 34.9% of PAHs throughout the year.From Figure 7, we can see its seasonal contributions: autumn > spring > winter > summer.
Petroleum source may be related to the Yangzi Petrochemical Plant located in the main urban area of Nanjing city.
From above we can see that source characterization studies based on the analysis of diagnostic ratio (triangular plot method), cluster analysis, and positive factor matrix analysis [81] suggested that the PAHs of Qinhuai River Basin were mainly from pyrogenic origin (biomass and coal combustion and vehicular emission), and the petroleum source also can not be ignored (specially in summer).

Sediment Quality Guidelines (SQGs).
PAHs often accumulate in various environmental systems including coastal estuaries and marine sediments [82].Sediments release PAHs into water as an intrinsic source, which harms aquatic organisms, and then it is harmful to human health, so PAHs of sediments are a potential threat to aquatic ecosystems.Therefore, ecological risk assessment of PAHs in surface sediments from the city-river like Qinhuai River is very important to protect human health and for ecological environment security.The need for chemical guidelines that could be used to predict adverse biological effects in contaminated sediments led to the development of sediment quality guidelines (SQGs) during the last decade [83].SQGs developed on the basis of Biological Effects Database for Sediments (BEDS) are very useful for sediment assessments in freshwater and marine environments.BEDS contain data that can be used to establish links between the concentration of a given chemical and its biological effect.One of the most widely applied SQGs for estuarine and marine stations is effects range-low (ERL) and effects range-median (ERM) [84].Threshold effects levels (TELs) and probable effects levels (PELs) are also considered to be widely applicable [85].These two sets of SQGs can be used to identify three ranges of chemical concentrations, including a low range, within which adverse biological effects are unlikely to occur: below the TEL or ERL values; a middle range, within which adverse biological effects are possible: between the TEL or ERL and PEL or ERM values; and an upper range, within which adverse biological effects are likely to occur: above the PEL or ERM values [42,84,85].As shown in Tables 5 and 6, all the sampling sites are below ERL values except for 13# and 9# in spring.However, concentrations of Nap, Flu, Phe, Ant, and

Sediment Potential Human Toxicity and Biological
Effects Based on PAHs.Many PAHs, especially their degradation products, are carcinogenic, while benzo[a]pyrene is the only PAH with sufficient toxicological data for derivation of a carcinogenic potency factor [86].Therefore, the toxic equivalency factor (TEF) was used to quantify the carcinogenicity of other PAHs relative to benzo[a]pyrene.In the present study, the potential toxicity of sediment was evaluated using the total toxic benzo[a]pyrene equivalent (TEQ) of five PAHs and calculated using the following equation: where   is the concentration of the PAH congener and TEF  is the toxic equivalency factor.According to US EPA [87], TEFs of BaA, BbF, BkF, BaP, and Chr are 0. The results showed that winter had largest impact on the ecological health and the maximum potential toxicity.This is probably due to people's choice of cars for travelling because of low temperatures and incomplete combustion of fossil fuels.The maximum TEQ value was found at site 7# (Yunliang River) which received the tail water of sewage treatment plant, during autumn and winter.Therefore, direct sewage pollution was the main reason for the distribution of PAHs during autumn and winter in this study.It is reasonable to find that the maximum TEQ value in spring was at site 4#, which received incoming water from eastern section of Internal Qinhuai River, where there are much vehicle traffic and domestic sewage.The maximum TEQ value was found at site 1# in summer; this is attributed to the fact that site 1# is on the main roadside of Nanjing with constant heavy traffic, indicating that the distribution of PAHs of site 1# was affected by traffic.Therefore, we can

Figure 2 :
Figure 2: Temporal (b and c) and spatial variation (a) of ∑PAHs in sediments.

Figure 5 :
Figure 5: Factor fingerprints of three PMF factors in each of the individual PAHs.

Table 5 :Table 6 :
The first set of SQGs values (ERL and ERM) for PAHs (ng/g d.w.) and relative percentage of sites including the range of Sediment Quality Guidelines.The second set of SQGs values (TEL and PEL) for PAHs (ng/g d.w.) and relative percentage of sites including the range of Sediment Quality Guidelines.

Table 1 :
Sampling sites and the corresponding regional characteristics.

Table 2 :
Mean PAHs concentrations (ng/g), range in surface sediments from Qinhuai River during various seasons.

Table 3 :
Sediment PAHs concentrations reported from elsewhere.

Table 4 :
Factor profiles ( matrix) of individual PAHs in each of the three PMF factors identified in the analysis of Qinhuai River sediment data set.
Chr in autumn; Flu and Phe in winter; Nap, Ace, Flu, Phe, Ant, and BaP in spring; and Nap in summer are all above the PEL values, indicating that these sampling sites are likely to have adverse −1, with an average of 94.1 ng TEQ•g −1 .Seasonal mean TEQ values ranged winter > spring > autumn > summer.