Introduction

Polycyclic aromatic hydrocarbons (PAHs) are high-priority semi-volatile organic pollutants that are receiving attention because of their toxicity, probable mutagenic and carcinogenic effects and the potential health risk they pose to urban residents (Colombo et al. 2006; Szabová et al. 2008). They have been listed by the US Environmental Protection Agency (US-EPA) and the European Community as priority pollutants (Zhang et al. 2012).

In soil, PAHs are present both as a result of natural processes and anthropogenic release mainly through fossil fuel combustion, petroleum refining, vehicle emission, biomass burning, industrial processing, oil spills, etc. (Peng et al. 2011). Soil has been identified as an effective sink for medium to low volatility atmospheric PAHs throughout the year. As such, it has to be regarded as a long-term archive of atmospheric deposition rather than an indicator of the current input. Due to their persistence and hydrophobicity, PAHs accumulate in the soil where they are retained for a long time (Wilcke 2000; Kaya et al. 2012; Liu et al. 2012).

The unintentional ingestion of soil particles is a potential route of exposure to environmental chemicals, including PAHs, particularly for children, who may ingest significant quantities of soil by playing on the ground outdoors, hand-to-mouth movements, or by eating food that has dropped on the floor (US-EPA 2011). Moreover, humans are exposed to soil PAHs by direct contact with the particles and their inhalation. Estimates of the contaminant intake by unintentional ingestion, inhalation, and dermal contact with soil particles contribute to understanding the overall exposure to environmental chemicals. Therefore, studies on soil contaminations by PAHs enable the risk assessment through human exposure to soil particles (Holoubek et al. 2009).

Soil contamination by PAHs in the Serbian urban zones has been investigated several times (Škrbić et al. 2005; Crnković et al. 2007; Škrbić and Đurišić-Mladenović 2009). However, the corresponding health risks in these urban areas and wider in the region have never been assessed. Thus, the main aims of the study were to assess the levels and spatial distribution of 16 PAHs in soil samples from Novi Sad, the second largest city in Serbia; to indicate the dominant emission sources by selected diagnostic ratios; and to estimate the corresponding potential cancer risks to humans via ingestion, dermal contact, and inhalation of soil particles. As the first paper studying the cancer risk of the inhabitants of an urban zone in Serbia, it could provide a baseline for future studies on human health protection and risk management in order to control residential exposure to surface soil contamination.

Materials and methods

Soil sampling and preparation

Surface soil samples were collected from 15 different urban locations (coded I-XV, Fig. 1) in Novi Sad during the winter season of 2014–2015. The urban area of Novi Sad is about 130 km2; there is a population of 250,439 according to the last census conducted in 2011 (Statistical Office of the Republic of Serbia 2011). The locations were chosen in order to consider places with different anthropogenic activities within the area (recreational, residential, and industrial parts of the city, schools). At each location, soil samples were taken from two sites, one along the nearest road (NR) and the other away from the road (e.g., in the schoolyard) in order to investigate spatial, short-range variability of the PAH occurrence at the chosen locations. The soil samples were coded numerically from 1 to 30, with even number for “NR” samples. The sample representative for each site consisted of five sub-samples taken randomly within approximately 50 × 50 m, while in case of park locations, it was approximately 100 × 100 m. The additional details of the soil sampling are given in the Supporting Information. Organic matter (OM) was determined for each soil sample measuring sample loss after 2 h at 550 °C as described in Škrbić and Čupić (2004).

Fig. 1
figure 1

Map of Novi Sad showing study area and sampling locations

Full size image

Chemicals and materials

A mixed standard solution of 16 EPA priority PAHs (naphthalene (Nap), acenaphthylene (Acy), acenaphthene (Ace), fluorene (Fl), phenanthrene (Phe), anthracene (Ant), fluoranthene (Fly), pyrene (Pyr), benzo[a]anthracene (BaA), chrysene (Chr), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[a]pyrene (BaP), indeno[1,2,3-cd]pyrene (InP), dibenzo[ah]anthracene (DB), benzo[ghi]perylene (Bghi)) in methylene chloride/benzene (1:1) was obtained from LGC Standard, Germany. Deuterated BaP (d12-BaP) in acetonitrile, obtained from Chiron AS, Norway, was used as a surrogate standard; 1,3,5-triphenylbenzene (Supelco, USA) was used as an internal standard. Prior to use, alumina (Sigma-Aldrich, Switzerland) and silica gel (CARLO ERBA Reagents, France) for chromatography and diatomaceous earth (Thermo Scientific, USA) were heated on 350 °C for 4 h. n-Hexane (Fisher Chemical, UK) and methylene chloride (AppliChem Panreac ITW Reagents, Spain), both of HPLC grade, were used as solvents.

Sample preparation, cleanup, and analysis

Sample preparation followed the method of Zhang et al. (2012) with slight modification. The method is based on simultaneous extraction and cleanup procedure, the so-called in cell pressurized liquid extraction of the samples, using Accelerated Solvent Extractor ASE 350, Dionex 350, Thermo Scientific, USA. PAHs were analyzed using a gas chromatograph with mass selective detector (GC MS-Agilent 7809B-5977A, Germany) equipped with an HP-5MS column (30 m × 0.25 mm i.d., 0.25 μm film thickness), using helium as the carrier gas (2 mL min−1). All the compounds were identified and quantified in selected ion monitoring (SIM) mode by their characteristic ions and using internal standard calibration method. The details about the sample preparation, cleanup, and instrumental analysis are given in the Supporting Information.

Quality assurance/quality control

Quality assurance and quality control (QA/QC) procedures included procedural blanks, matrix-spike, and calibration verification. Special attention was given to determine linearity, limits of detection (LODs), recovery, and precision in accordance to Sanchís et al. (2013). The QA/QC parameters were obtained in accordance to procedures described in the Supporting Information, and they are given in Table 1. In every analytical batch made up of 6 samples (extracts), 1 blank sample and 1 control standard mixture solution were inserted in order to identify any contamination during sample preparation and analysis and to check instrument response against the calibration curve, respectively. Each sample was analyzed in duplicate. For further calculation, all the analytical results were corrected for the relevant recovery values.

Table 1 Validation parameters: instrumental linearity, the instrumental limits of detection (ILOD), the intra- and inter-day accuracy, the method limits of detection (MLODs), recoveries, and relative standard deviations (RSDs)
Full size table

Compliance with the relevant guidelines

In order to rank the soil samples in accordance to the obtained PAHs levels, the results were compared with the national and some international guidelines on PAHs in soils. Many countries have developed regulatory guidance values for PAHs regarding soil protection (Carlon 2007; Jennings 2012a, 2012b). These guidelines are widely variable because of differences in the set objectives, derivation methods, and regulatory applications (Cachada et al. 2016).

According to the Serbian national limits for PAHs in soil (Official Bulletin of the Republic of Serbia 2010), the maximum allowed concentration (MAC) and remediation concentration (RM) for sum of 10 PAHs (∑10 PAHs), i.e., Ant, BaA, BkF, BaP, Chr, Phe, Fly, InP, Nap, and Bghi, in soil samples with organic matter (OM) less than 30% are 1000 and 40,000 μg kg−1, respectively. MACs are defined as the maximum values at which functional properties of the soil are still fully achieved and indicate level at which a sustainable soil quality is reached, while RM represent values at which basic functions of land are threatened or seriously disrupted, requiring remediation and other measures (Official Bulletin of the Republic of Serbia 2010). These limits are rather similar to the Dutch guidelines, which were the first to be developed in Europe for soils with 10% of organic matter: soil with ∑10 PAHs level below the Dutch target value (1500 μg kg−1) is considered clean, and no further action is needed; those with the contents between the target and the intervention (40,000 μg kg−1) values are considered slightly contaminated, and a sustainable soil management is needed; whereas soils with ∑10 PAH concentrations above the intervention value are seriously contaminated, and they need urgent remediation (Škrbić et al. 2005; Cachada et al. 2016). Furthermore, the Dutch government introduced “National Maximal Values,” which are related with the land use (Cachada et al. 2016): a maximum value for a residential land use of 6800 μg kg−1 and a maximum value for an industrial land use of 40,000 μg kg−1 have been established, and according to these maximums, three categories of soil are defined—always suitable (below target value of 1500 μg kg−1), suitable for residential (<6800 μg kg−1), and suitable for industrial (≤40,000 μg kg−1) and not suitable (>40,000 μg kg−1). Concerning the guidelines that refer to the levels of individual PAHs, the precaution level of BaP defined in Germany was also considered for the comparison with the results from this study: values above the BaP precaution level (300 and 1000 μg kg−1 for soil with organic matter content ≤8 and >8%, respectively) indicate a certain chance of future soil problems, which needs to be addressed in order to avoid upcoming damages (Cachada et al. 2016).

Ecological risk-based numerical standards for PAHs in soil have not yet been established in Serbia. Therefore, soil quality guidelines (SQGs) proposed on non-carcinogenic effects of PAHs by the Canadian Council of Ministers of the Environment (CCME 2010) for 14 PAHs (Nap, Acy, Ace, Fl, Phe, Ant, Fly, Pyr, BaA, BbF, BkF, BaP, DB, and InP) were used for the assessment of ecological risk of PAHs in this study. The environmental SQGs were developed in a way to maintain important ecological functions that support activities associated with the identified land uses (e.g., agricultural, residential/parkland, commercial and industrial land uses). These SQGs could be further used in a form of mean risk quotient (m-RQ) (Wang et al. 2015) calculated as a mean ratio of the obtained concentrations and the respective SQG for evaluating the ecological risk of PAHs. The value of m-RQ higher than 1 indicates medium to high risk, those in the range 0.1–1 low risk, while lower than 0.1 negligible risk.

Risk assessment procedure

In order to evaluate the carcinogenic potency of simultaneous occurrence of all 16 PAHs (∑16 PAHs) detected in surface soil, the toxic equivalency factors (TEFs) set for each compound were used to estimate the toxicity equivalent quotient concentrations (BaPeq) of the soil samples (Tsai et al. 2004; Jiang et al. 2014; Wang et al. 2015). BaPeq was calculated by summing the individual PAH concentrations multiplied with the corresponding TEF values (Tsai et al. 2004; Jiang et al. 2014; Wang et al. 2015), which is defined in relation to the TEF of BaP (assumed to be 1). BaP has been chosen since it is a well-studied compound, important to environmental toxicology.

The ratio of the sum concentration of seven carcinogenic PAHs (∑carcPAHs = Chr + BaA + BkF+ BbF + BaP + DB + InP) and the sum concentration of the ∑16 PAHs was calculated, too. The closer the value of ∑carcPAH/∑16 PAHs to 1, the more hazardous ∑16 PAHs were to humans (Kozielska et al. 2015). It should be pointed that this ratio only shows the contribution of the sum of seven carcinogenic PAHs to the total PAHs, while the toxicity of the sample depends on the exposure dose.

Potential cancer risk (CR) imposed on inhabitants of Novi Sad as a result of being in contact with urban soil PAHs via accidental ingestion (CRingestion), dermal contact (CRdermal), and inhalation of the soil particles (CRinhalation) were estimated using Eqs. (1), (2), and (3), respectively (US-EPA 2009), presented in the Supporting Information. Similar approach has been used in several recent studies that reported cancer risk assessment results attributable to PAHs in soil (Peng et al. 2011; Kumar et al. 2013; Man et al. 2013; Jiang et al. 2014; Kumar et al. 2014a, 2014b; Yang et al. 2014; Wang et al. 2015). This calculation estimates a theoretical excess cancer risk expressed as the proportion of population that may be affected by carcinogen during a lifetime of exposure (Wang et al. 2015). It might be regarded as a point estimate that is compared to an acceptable level giving a “negative” or “positive” result on the risk assessment for an exposure scenario (Kang et al. 2015). It is noted that different numbers of the individual PAHs was considered in these studies, either seven carcinogenic (Yang et al. 2014), eight (Kumar et al. 2014b) possible indicators of the carcinogenic PAHs potency (i.e., PAH8) proposed by the European Food Safety Authority (EFSA 2008), 16 EPA PAHs (Peng et al. 2011; Kumar et al. 2013; Jiang et al. 2014; Kumar et al. 2014a), or 24 PAHs (Wang et al. 2015). In this study, cancer risks were assessed for (a) group of 7 carcPAHs, (b) for group of 7 carcPAHs + Nap which is classified as possible carcinogenic to humans (IACR group 2B (Wang et al. 2015)), and (c) for PAH8. Each of CRs was estimated for two age groups: children and adults. The total lifetime carcinogenic risk (TLCR) was obtained summing the cancer risks calculated for all three exposure routes, for children and adults (Kumar et al. 2013, 2014a: Wang et al. 2015).

The values of parameters used to calculate CRs by Eqs. (1)–(3) are taken from US-EPA (1991, 2009, 2011, 2012), and they are presented in Table S3 of the Supporting Information. These parameters represent the extreme conditions, e.g., maximal exposure frequency, exposure time, and ingestion rates.

Statistical analysis

All data were processed using Microsoft Excel 14.0.4734.1000 (2010 Microsoft Corporation). One-way ANOVA analysis of variance using F-test at P < 0.05 was performed to test the significant differences between the PAH concentrations, ∑carcPAH/∑16 PAHs, and TLCR values among four areas of soil collection.

Results and discussion

Levels and compositional profiles of PAHs

Descriptive statistics for concentration of the individual and the sum of 16 US-EPA priority PAHs (∑16PAHs) measured in 30 composite soil samples from Novi Sad, as well as for the soil level of 2-, 3-, 4-, and 5-ring PAHs, are presented in a form of box-whiskers graph in Fig. 2, as well as in Table S3 in the Supporting Information. The organic matter content in all soil samples was less than 10%. As it can be seen, all 16 PAHs were detected in all surface soil samples. The highest concentration of ∑16 PAHs was detected in sample 30 (2247 μg kg−1), followed by samples 9 (1207 μg kg−1), 10 (1052 μg kg−1), 16 (860 μg kg−1), and 18 (686 μg kg−1). Sample 30 was collected from a location full of pine needles on the ground, which could possibly explain such extreme values. Similar results were encountered in other pine needle litter (Yang et al. 2014). Namely, it was observed that in soil rich in natural dissolved organic matter derived from pine needles, the availability and mobility of PAHs like Phe and Fly can be promoted due to the observed significant increase in desorption of these compounds from soil organic matter. Hence, the presence of pine needle-dissolved organic matter in soil sample no. 30 probably influenced better desorption of these PAHs and their transfer to the extract. The other two samples (nos. 9 and 10) where the ∑16 PAH content over 1000 μg kg−1 was noted were both taken from the same location of a school in the wider city center with low traffic frequency. It can only be speculated that specific features of this location could be the reason behind the elevated contents. Namely, the school is located in the vicinity of cobble road and old, damaged buildings on one side and with high residential buildings on other sides; being such closely surrounded, it might be presumed that there is a larger quantity of street dust and particles “trapped” here, particularly taking into account that washing of cobble road is not practiced in the city, contrary to the asphalted roads. It is well documented that street dust represents a source of pollutants (Martuzevicius et al. 2011; Mostafa et al. 2009) as it receives varying inputs of the compounds from a variety of mobile or stationary sources, such as traffic, building construction materials, industrial plants, residential sources, waste, construction and demolition activities, and resuspension of surrounding soil particles (Lu et al. 2009). Thus, it might be presumed that higher loads of street dust on the soil surface at these sites led to the accumulation of the higher PAH quantities. The ∑16 PAH content at site no. 18 (considered also within the “residential” group of samples), known for heavy traffic and presence of many heavy-duty vehicles (city and intercity busses), was 686 μg kg−1. The mean and median values calculated for all (30) samples were 363 and 200 μg kg−1, respectively, while after the exclusion of the extreme value, they were 298 and 190 μg kg−1, respectively. Ninety percent of soil samples was characterized by the ∑16 PAH content <1000 μg kg−1.

Fig. 2
figure 2

Box-whisker graph of the PAH contents (in μg kg−1 except the unitless ratio) detected in the soil samples and related derived values

Full size image

In order to compare samples from the same sampling locations but within different distances from the nearby road (NR), ratios of ∑16 PAH values detected in samples coded with odd and even numbers (e.g., 1 and 2 from location I; 3 and 4 from location II, etc., where odd-numbered samples are those from the sites distant from the road, while the even numbered are from NR) were calculated and presented in Fig. 3. At most locations, the ratio lower than 1 was observed, indicating higher concentration of PAHs in the NR samples. This was not surprising since traffic emission is a well-known source of these contaminants, particularly in urban areas (Nadal et al. 2007). Only at the three sampling locations, III (school), V (school), and X (residential), the calculated ratios were 1.3, 1.1, and 1.8, respectively. Even though at these locations rather different ∑16 PAH contents were found (at III and X in the range from 56 to 122 μg kg−1, while at V about 10 times higher), common for these samples was the higher contents of almost all 16 PAHs in the samples away from the road, suggesting the importance of the aerial transport of PAHs either of low molecular weight (LMW) compounds in gaseous phase or those with heavier molecular weights (HMW) being bound to particulate matter, probably due to some specific features for these locations (e.g., the specific features of location V, i.e., sites 9 and 10, have already been previously discussed).

Fig. 3
figure 3

Ratio of the ∑16 PAH values detected in samples collected distant and nearby road, NR, at each of 15 locations

Full size image

The differences in levels of PAH contamination among four functional areas (recreational, schools, industrial, residential) were investigated using one-way ANOVA (after excluding sample no. 30 characterized by the extreme ∑16 PAH value). Mean values of ∑16 PAHs in school, industrial, residential, and recreational areas were 452, 281, 348, and 188 μg kg−1, respectively. The highest variability of ∑16 PAH value, represented by a so-called spread ratio (ratio of the respective maximum and the minimum values, Table S3) was seen in the “recreational” and “residential” groups of samples (spread ratios of 17 and 15, respectively), followed by “schools” group (spread ratio of 12.5 not taking into account the outlier—sample no. 30), indicating that these locations (from different parts of the city) were affected by variable intensities of the PAH emissions. The results of the ANOVA indicated that there were no significant differences (p < 0.05) observed between the four functional areas.

Compared with literature data on the presence of PAHs in urban soil, the mean ∑16 PAH concentration of 298 μg kg−1 calculated in this study after excluding the extreme value of 2247 μg kg−1 was lower than the value reported for an urban soil in Torino, Italy (1990 μg kg−1; Morillo et al. 2007), which is one of the most polluted European cities (Europe: Pollution Index 2015). It has to be noted that our results were also lower than the mean soil ∑16 PAH content reported for Shanghai (1970 μg kg−1; Wang et al. 2015)—one of the main industrial centers of China. In fact, the mean ∑16 PAHs obtained for Novi Sad was among the lowest values reported for soil of the urban areas worldwide and lower than Korba, India (385 μg kg−1; Kumar et al. 2014a); Tarragona, Spain (438 μg kg−1; Nadal et al. 2007); Ljubljana, Slovenia (898 μg kg−1; Morillo et al. 2007 ); Ulsan, South Korea (960 μg kg−1; Kwon and Choi 2014); Tianjin, China (814 μg kg−1; Jiao et al. 2009); Beijing, China (1083 μg kg−1; Liu et al. 2010 ); Lisbon, Portugal (1544 μg kg−1; Cachada et al. 2012 ); Bergen, Norway (6780 μg kg−1; Haugland et al. 2008); Delhi, India (6839 μg kg−1; Kumar et al. 2014a), etc.; it was also lower than previously published results for urban locations in Novi Sad in 2001 (667 μg kg−1; Škrbić et al. 2005) and 2007 (980 μg kg−1; Škrbić and Đurišić-Mladenović 2009). A comparison of PAH contents in urban soil around the world might be found in a recent study of Jiang et al. (2016). The mean value from this study was however higher than in Estarreja, Portugal (98 μg kg−1; Cachada et al. 2012 ), and Viseu, Portugal (169 μg kg−1; Cachada et al. 2012 ).

For all samples, except one, the concentration of ∑10 PAHs was below the Serbian MAC indicating no contamination. The highest content of ∑10 PAHs found in soil sample 30 was 1.7 times higher than MAC, and it was slightly higher than the Dutch target value. It could be concluded that the soil from the urban area of Novi Sad was “suitable as residential” and that no further action or intervention would be needed. The levels of BaP in all samples were less than the German precautionary level, suggesting that there would be no future soil problems or upcoming damages if the current levels retained. This was further confirmed by comparison of the individual PAH concentrations with the respective SQGs (CCME 2010) for all 30 soil samples; as the m-RQs values were always below 0.1, the occurrence of PAHs in the analyzed soil samples indicated a negligible ecological risk in Novi Sad.

When assessing the profile of individual 16 analyzed PAHs (Table 2), Chr, Fly, Bghi, and Pyr were the major components in all soils samples, accounting for 14, 12, 12, and 11% of the ∑16 PAH concentrations, respectively. Group of 4-, 5-, 6-, 3-, and 2-ring PAHs accounted for 42, 23, 19, 13, and 2.8% of the ∑16 PAHs, respectively; thus, HMW PAHs, with 4 to 6 rings were dominant. The mean and median percentage concentrations of HMW PAHs were 84 and 85%, respectively, which coincided with the results obtained in Beijing (Peng et al. 2011), Shanghai (Wang et al. 2015), Viseu, and Lisbon (Cachada et al. 2012 ) where the HMW PAHs were also the most abundant in the urban soil. This predominance is typical in urban areas and may be due to (i) the higher persistence of these compounds in soils, (ii) the predominance of the combustion over petrogenic sources, and (iii) the tendency of HMW PAHs to accumulate in soils that are close to emission sources (Cachada et al. 2012 ).

Table 2 Shares of PAHs in % relative to the∑16 PAHs in composite soil samples collected in Novi Sad
Full size table

Diagnostic ratios for identification of possible PAH sources

The PAH emission profile derived from different sources can vary significantly depending on the conditions of the processes producing these compounds (Tobiszewski and Namiesnik 2012). PAH diagnostic ratios—the ratios of defined pairs of individual compounds—have been often used in identification of the PAH sources (Katsoyiannis et al. 2011; Peng et al. 2011; Jiang et al. 2014).

The ratios calculated in this study were Ant/(Ant + Phe), Fly/(Fly + Pyr), BaA/(BaA + Chr), InP/(InP + Bghi), BaP/Bghi, ΣLMW/ΣHMW, and ΣCOMB/ΣPAH, where ΣCOMB is the sum of Fly, Pyr, BaA, Chr, BkF, BbF, BaP, InP, and Bghi; ΣPAHs is the sum of 16 analyzed PAHs; ΣLMW and ΣHMW are sums of two- and three-ring PAHs and of four- and five-ring PAHs, respectively. The mean, median, minimum, and maximal values of these ratios (except for BaP/Bghi) are presented in Table 3 together with the respective limit values that point the pyrogenic sources (Tobiszewski and Namiesnik 2012). In majority of the analyzed soil samples, the calculated ratios indicated pyrogenic sources. The limit value of 0.6 for BaP/Bghi ratio separates two kinds of emissions: non-traffic (<0.6) and traffic (>0.6); the mean, median, minimum, and maximal values calculated here for this ratio were 1, 0.6, 0.1, and 8, respectively, pointing traffic emission as the PAH source in general.

Table 3 PAH diagnostic ratios calculated for the analyzed soil samples and their limit values indicating the pyrogenic sources
Full size table

Risk assessment procedure

In order to compare and quantify the toxic potency of soil samples, BaPeq values were calculated based on TEF (Nisbet and LaGoy 1992). In this present study, BaPeq ranged from 2 to 202 μg kg−1 with mean and median value of 39 and 22 μg kg−1, respectively (Fig. 2 and Table S3). This mean BaPeq value was rather similar to the mean BaPeq value obtained for Viseu (24 μg kg−1; Cachada et al. 2012 ) but lower than mean BaPeq values from Beijing (181 μg kg−1; Liu et al. 2010 ), Lisbon (229 μg kg−1; Cachada et al. 2012 ), Shangai (236 μg kg−1; Wang et al. 2015 ), and Guwahati in Northeast India (116–557 μg kg−1; Hussain and Hoque 2015 ).

Values of the ∑carcPAH/∑16 PAH ratio (Fig. 2) were from 0.2 to 0.7, with both mean and median values of 0.5, indicating generally medium carcinogenic influence. There were no significant differences in ∑carcPAH/∑16 PAHs in all four groups of the samples.

The calculated CR and TLCR values of 7 carcPAHs over three exposure pathways for four sampling zones are shown in Table 4. When two other groups of PAHs were considered (carcPAHs + Nap and PAH8), the median and mean TLCRs of the same orders of magnitude were obtained, except in the case of median for recreational zones, which was changed from 10−5 for 7 carcPAHs to 10−4 for carcPAHs + Nap and PAH8. US-EPA sets the acceptable cancer risk to be in the range 10−6 to 10−4 (US-EPA 2001), while the risk estimated to be above 10−4 is considered to be of concern and requires additional action for the reduction; risk below 10−6 denotes virtual safety not requiring further action (US-EPA 2008). However, there is no consistency with interpreting some ranges of cancer risks, e.g., Canadian soil quality guidelines for the protection of environmental and human health consider cancer risk of ≤10−5 as insignificant (CCME 2010), while New York State Department of Health sets qualitative descriptions of lifetime cancer risks as follows: very low when the estimated value is ≤10−6; low from <10−6 to <10−4, moderate from ≤10−4 to <10−3, high from ≤10−3 to <10−1, and very high when the value is ≥10−1 (as quoted in Man et al. 2013).

Table 4 Cancer risks (CRs) via ingestion, inhalation, and dermal contact with seven carcinogenic PAHs in soil and total lifetime cancer risk (TLCR) for children and adults for four sampling zones
Full size table

The CRs of three exposure pathways were in the order: ingestion > dermal contact > inhalation for both age groups. Ingestion and dermal CRs were either the same order of magnitude or one order different (with lower order of magnitude found for dermal CRs). Compared with dermal and ingestion pathways, CRs by the inhalation of soil particles was negligible (for both, children and adults), not representing health risk, being approximately of order of 10−8 to 10−11. This coincided with the finding of Man et al. (2013), who investigated the Hong Kong soils contaminated by PAHs.

The highest (max) CR value (Table 4) was calculated for children through soil ingestion based on the maximum PAH level found in this study. When the corresponding mean and median values of CRs estimated for children and adults were compared, children seemed to be under greater health risk. This could be also the conclusion when TLCRs are compared (Table 4): the higher TLCR mean and median values were calculated for children. Children as the most sensitive subpopulation may ingest significant quantities of soil owing to their hand-to-mouth activity. In addition, the contaminant intake by a child is greater than by an adult due to its body weight (see Eq. 1 in the Supporting Information) (Jiang et al. 2014). In this study, TLCR values were within the acceptable range according to EPA (10−6 to 10−4) at 30 and 47% of the sampling sites for children and adults, respectively, while for the rest of the samples, TLCR values were >10−4 indicating risk of concern (Table 4). If eight compounds with possible carcinogenic potency were considered (either carcPAHs + Nap or PAH8), cancer risk of concern was calculated for even more sites (83 and 63% for children and adults, respectively). Mean and median TLCRs for children were 2.8 × 10−4 and 1.6 × 10−4, while mean and median TLCR values for adults were 1.8 × 10−4 and 1.0 × 10−4, respectively (Table 4). The group of samples from “school” locations had the highest TLCR values (Table 4); however, the differences between four collection areas were not statistically significant, as it was also the case with ∑16 PAH contents.

It should be emphasized that the CRs and TLCRs calculated here represented risks for extreme exposure (Jiang et al. 2014). Firstly, the maximal exposure time of 24 h per day was taken with exposure frequency of 365 days per year. Then, a conservative soil ingestion rate of 200 mg day−1 for children and 100 mg day−1 for adults was considered, which are 4 to 5 times higher than the proposed general population central tendency values (US-EPA 2011). For the average body weights, 18.6 kg for children and 80 kg for adults were taken into account; nonetheless, if the lower average weight for adults was applied (e.g., 55–60 kg, which is the weight used for pollutant exposure assessment in other studies (Škrbić et al. 2012; Man et al. 2013; Perez et al. 2014; Škrbić et al. 2014; Wang et al. 2015)), the same orders of magnitude for each type of CRs and TLCRs would be obtained (data not shown). Furthermore, it should be mentioned that conducting risk assessment by means of total pollutant concentrations always results in the overestimation of the health risk exerted on humans, particularly knowing that bioavailability of HMW PAHs from soil is likely to be <100% (CCME 2010). Using bioavailable pollutant concentrations to evaluate health risk is commonly regarded as the most accurate way, as only the bioavailable portion of the contaminants will impose adverse effects on the body (Oomen et al. 2000). However, the method based on bioavailable contents usually involves experiments conducted with the use of animals, raising ethical concerns and introducing uncertainties with using animal data in estimating human health risk. Moreover, development of in vitro and in vivo testing methods for bioaccessibility of organic compounds has not progressed as far as for metals and are unlikely to be ready for use in bioavailability adjustment in risk assessment (Man et al. 2013). Thus, conservative estimation of the health risk based on the use of the total PAH content might be regarded as a possible drawback of this study and also of the similar ones conducted previously (Man et al. 2013; Jiang et al. 2014; Kumar et al. 2014a; Yang et al. 2014) but overestimation is usually better than underestimation during risk assessments in order to provide greater protection of human health (Man et al. 2013).

Conclusions

The present study has reported the concentrations of ∑16 PAHs in 30 surface soil samples taken from 15 different locations in Novi Sad, Serbia. Generally, the obtained levels were lower than those reported in literature for other similar size cities. Only in one soil sample, concentration of ∑10 PAHs was slightly higher than the maximum allowable limit set by the Serbian legislation; nevertheless, all the samples could be classified as suitable as residential according to the Dutch guidelines. Among different functional groups of surface soil samples (schools, industrial, recreational, and residential), there was no significant difference concerning ∑16 PAH levels; consequently, no significant differences were found in the derived values (BaPeq, ∑carcPAH/∑16 PAHs, and TLCR). The composition of PAHs was characterized by high molecular weight PAHs, indicating combustion processes as possible predominant sources of PAHs in Novi Sad, particularly from traffic emission, which was confirmed by the calculated diagnostic ratios. The estimated cancer risks in the case of extreme exposure through unintentional ingestion, inhalation, and dermal contact with PAHs in soil indicated a potential cancer risk of concern for residents of Novi Sad including both sub-populations considered, i.e., adults and children.

These observations point out the importance of assessing health risks in parallel to regular analytical monitoring of PAHs in the urban soils. The contribution of the unintentional soil ingestion to the overall exposures via the environmental chemicals, including PAHs, is important and should not be disregarded particularly in the urban areas. Additionally, this study brings the first results on cancer risk assessment in Serbia, bridging the gap between the existing PAH monitoring studies and the lack of data on the risks posed by unintentional exposure to PAHs present in soil.