Sources and transports of polycyclic aromatic hydrocarbons in the Nanshan Underground River, China

The Laolongdong Underground River System (LURS) area is located at the center of Chongqing municipality, Southwest China (Fig. 1). The underground drainage area of the system is approximately 12.6 km². The elevation of LURS is between 350 and 685 m above average sea level. The climate is primarily subtropical monsoonal with a mean annual precipitation of 1,100 mm and a mean air temperature of 18.7 °C. The monsoonal climate results in a rainy season from May to October and a dry season from November to April. The population of the Nanshan karst valley is over 50,000 and is associated with rapid urbanization, apart from farming. Residential areas that are concentrated in an internal trough valley produce various effluents, which are discharged into an artificial channel which flows southward to join the Laolongdong Underground River, and then into a tributary of the Yangzi River in Chongqing. There are no treatment plants for the waste water. Even worse, some sinkholes are used to discharge waste water residues. By the end of the 1990s, surface ecosystems of the karst valley had been disturbed by human activities and local groundwater has been deteriorating for many years.

It is a transitional zone where the Nanwenquan anticline in the longitudinal structural system turns to the Tongluoxia anticline in NE–SW structural system. The outcrops are principally carbonated rocks (limestone), with Triassic strata being widespread. The strata of the anticlinal axis are carbonate rocks of the Lower Triassic Challenging Formation (T₁ j), with limestone being a major lithology, whereas anticlinal wings are carbonate rocks of the Middle Triassic Leikoupo Formation (T₂ l) and sandstones of the Upper Triassic Xujiahe Formation (T₃ xj). Yellow-green calcareous clay rocks of the lower leikoupo (T₂ l) overlie the Chialingchiang Group (T₁ j). No sulfate evaporites (gypsum and anhydrite) are exposed in the study area. Due to the banded distribution of carbonate rocks and the presence of relatively impermeable sandstone at the two wings combined with a vertical portals cranny which was well developed in the anticlinal core, basic conditions exist for a formation of a karst trough valley and the development of a karst groundwater system. Precipitation is the major recharge source of the groundwater that ensures abundant groundwater resources. The Laolongdong Underground River originates from the core of the karst trough valley and flows along anticlinal axis (Lower Triassic Chialingchiang formation (T₁ j)) in a NE–SW direction with a total length of 6 km.

A mixture of subtropical evergreen broadleaf needle forests is dominant, including Masson pine, fir, cedar, camphor, and ficus virens. The dominant types of soils are mainly limestone soils and yellow earth derived from carbonate rocks and sandstones, respectively. There are four land use categories (Fig. 1): land use for residential construction with an area of 4.5, 2.5 km for factories, 2 km for agriculture, 3 km of forest and unused land, respectively. Residential areas are built in the center of the northern trough or on either side of the roads. Industries are mainly cement and quarry plants centered among relict mountains along the valley whereas farmlands are scattered around the bottom of the valley. Underground rivers and springs used to be the only one major source of water supply for the residents.

Water samples were collected monthly in individual 1 L cleaned glass bottle between 2011 and 2012 from the underground river and its surface system (Fig. 1); 100 g of sediment samples was collected in plastic bags from the underground river and surface system; 500 g of soil sample was collected in plastic bags from the surface system. Water samples were filtered with Whatman GF/F (0.45 μm effective pore) pre-combusted at 450 °C for 5 h. Surrogate standards (2 μL) were added into water samples after the filtration. The filtered water was then passed through solid phase extractor (SPE-DEX controller 4,700/4,790, Horizon Technology). Soil and sediment samples were air dried before analysis.

Sixteen PAHs standards (16 compounds specified in EPA Method 610) in a mixture, perdeuterated PAHs (naphthalene-d₈; acenaphthene-d₁₀; phenanthrene-d₁₀; chrysene-d₁₂; perylene-d₁₂) in a mixture used as surrogate standards were obtained from Dr. Ehrenstorfer, Germany. An internal standard hexamehylbenzene solution was purchased from Supelco, USA. Silica gel (80–100-mesh) and neutral alumina (100–200-mesh) were extracted with dichloromethane for 72 h by using a Soxhlet extractor. Upon drying under room temperature conditions, silica gel and alumina were baked at 180 and 250 °C for 12 h, respectively. Sodium sulfate was baked at 550 °C for 8 h and stored in sealed containers (Luo et al. 2004).

About 10 g soil and sediment dried samples were kept in a cleaned filter paper (extracted for 72 h) and were spiked with surrogate standard and extracted for 24 h with dichloromethane. Activated Cu was added for desulphurization. The extract for each sample was concentrated and solvent-exchanged to hexane, and further reduced to approximately 1 mL under vacuum rotary evaporator. Concentrated extracts were fractionated with a 1:2 alumina:silica gel glass column and 15 mL n-hexane, 70 mL 3:7 dichloromethane: n-hexane, successively. It was finally concentrated to 0.2 mL. The extracts of water samples were dried by anhydrous sodium sulphate and concentrated to 0.2 mL with a gentle stream of purified nitrogen. Known quantities of internal standard were added to the sample prior to instrumental analysis.

PAHs were quantified (Agilent 7890A-5975C) gas chromatography coupled to mass spectrometer operating in electron impact and selective ion monitoring modes (SIM). A 30 m × 0.32 mm i.d. HP-5MS capillary column (0.25 μm) was used. Additional details of the chromatographic and spectrometric conditions are provided elsewhere (Mai et al. 2002). Quantification was performed using the internal calibration method based on five-point calibration curve for individual component. Five perdeuterated PAHs (naphthalene-d8, ace-naphthene-d10, phenanthrene-d10, chrysene-d12 and perylene-d12) were added in samples prior to extraction in order to quantify procedural recoveries.

For each sampling period, two glass bottles of de-ionized (cleaned) water were carried on board and exposed to the ambient environment during the course of field operation. This water was brought back to the laboratory and treated as a field blank. Field blanks were filtered and extracted in the same procedure as the field samples. For each batch of field samples, a procedural blank (solvent with clean GF/F filters), a spiked blank (16 PAHs standards into solvent with clean GF/F filters) and a National Institute of Standards and Technology 1,491 reference standard sample were processed. The field blank and procedural blank samples contained undetectable amounts of target analyses. The reported results were surrogate corrected. Recoveries of all the PAHs ranged from 56 to 115 % of the certified values. Detection limits ranged from 0.09 to 0.25 ng L⁻¹.

The mean concentration of ΣPAHs is 1,699 ng L⁻¹ in the groundwaters. The mean concentration of ΣPAHs is 3,612 ng g⁻¹ in the sediments of the underground river (Table 1). The level of ΣPAHs is two times higher in the sediments than that of the waters. According to the report of Luo et al. (2004) the waters of the underground river of this study are polluted or near to be polluted.

The concentrations of ΣPAHs in underground waters ranged from 353 ng L⁻¹ in December to 13,203 ng L⁻¹ in September. The content of ΣPAHs in the underground sediments ranged from 169 ng g⁻¹ in October to 12,038 ng g⁻¹ in May. The concentrations of ΣPAHs in surface waters ranged from 272 ng L⁻¹ in February to 13,248 ng L⁻¹ in September while surface sediments and soil showed the variations from 467 ng g⁻¹ in October to 4,275 ng g⁻¹ in December (Table 1). The seasonal variations suggest that the underground system affected by the surface system substantially.

Ormah et al. (2008) and Wang et al. (2009) found PAHs ranged from 16.93 to 190 ng g⁻¹ in soils in Chinese karst area. This was 1–5 times lower than the PAHs in soils in this study. These results might be related to the karst watershed conditions. Schwartz et al. (2011) reported 50 ng g⁻¹ of PAHs in sediments, 532 ng g⁻¹ in soils, 1–371 ng L⁻¹ in groundwaters in a highly vulnerable karst area. These values were lower than those of this study by several magnitudes.

The parent PAH ratios have been widely used to detect the sources of PAHs (Budzinski et al. 1997). The ratio calculations are usually restricted to PAHs with a given molecular mass in order to minimize the other perplexing factors such as the differences in volatility, water solubility, adsorption, etc. Ranges of PAHs ratios with molecular weight of 178 (anthracene, phenanthrene) and 202 (fluoranthene, pyrene) are commonly used to distinguish between combusted and non-combusted petroleum, coal, and biomass sources (Soclo et al. 2000; Yunker et al. 2002; Li et al. 2012). The anthracene to anthracene plus phenanthrene (ant/ant + phe) ratio of <0.1 is taken as an indicator of direct input of petroleum, biomass, and coal without combustion while the ratio of >0.1 indicates dominant of combusted petroleum, coal, wood, grass, and bush fire. Fluoranthene to fluoranthene plus pyrene (fla/(fla + pyr) ratio of 0.5 which is defined as non combustion of petroleum, coal, and biomass or combustion transition point of petroleum, grass, wood, and coal (Yunker et al. 2002; Li et al. 2012). Therefore, ant/(ant + phe) with fla/(fla + pyr) could be used to identify the sources of PAHs accurately. Ant/(ant + phe) ratios in the waters of the underground river ranged from 0.10 to 0.50 and in the sediments ranged from 0.40 to 0.50 while concerned fla/(fla + pyr) ranged from 0.50 to 0.80, 0.40 to 0.70, respectively (Table 1). The cross plot in ratios suggests that PAHs were delivered in the groundwaters from combusted grass, wood and coal (Fig. 2). The cross plot in ratios indicate that PAHs input in the sediments of the underground river were mixture of combusted grass, wood, coal and non combusted of petroleum, grass, wood, coal (Fig. 2).

The ant/(ant + phe) and fla/(fla + pyr) ratios in the surface waters ranged from 0.10 to 0.60 and 0.40–0.60, respectively (Table 1). The cross plot in ratios in the surface waters suggests that PAHs were delivered mainly from combusted grass, wood, coal (Fig. 3). The ant/(ant + phe) ratios in the surface sediments and the surface soils ranged from 0.02 to 0.60 while fla/(fla + pyr) ranged from 0.03 to 0.70 (Fig. 3). The cross plot in ratios in sediments and soils indicate that PAHs were mainly input in the surface sediments from combusted and non-combusted grass, wood, and coal slightly with combusted and non-combusted of petroleum (Fig. 3). China uses 18.3 % of petroleum, 3.4 % of natural gas for its energy sources (CESY 2008). Coal consumption in Chongqing climbed steadily with economic growth from 2005–2007 to a reported 42.9 million tonnes. Power plants including industries, reported by a number of different sources to have consumed about 15 million tonnes in 2007, are the most significant end-users, at approximately 35 % of the total coal consumption (Chongqing Economic Commission 2007; CESY 2008). It was also reported that Chongqing’s steel mills purchased an additional 1–2 million tonnes of cooking coal from outside the province. Some areas of China are becoming more urban, more than 40 % of the populations are still rural, most of which still use biomass (mainly wood and crop residues) and coal fuels that produce substantial pollution in simple stoves. In 2003, approximately 80 % of the energy consumed by rural households was in the form of biomass and almost 10 % as coal. Furthermore, although most Chinese cities have plans to eliminate coal for households, many urban communities continue to rely on coal. The combustion of biomass and coal (collectively called “solid fuels”) is the dominant source of indoor air pollution (IAP) in the country and contributes significantly to the total burden of ill health (Zhang and Smith 2007). We have no data on the usages of coals, fuels, and wood in the study area. During samplings, it was informed by the inhabitants of the area that wood, fuels, and coals are used by people and some of small factories located in the area. The similarities in sources between a surface system and an underground river suggest that PAHs in the underground waters and sediments transported through similar pathways.

The lower mass ΣPAHs of 2–3 rings were dominant in the waters and sediments of the underground river. Similarly, the lower mass ΣPAHs of 2–3 rings were dominant in the waters, sediments, and soils of the surface system (Figs. 4, 5, 6, 7, 8, 9). The orders of the transport of ΣPAHs are 2–3 rings > 4–5 rings > 6 rings in the waters and 2–3 rings > 4–5 rings = 6 rings in the sediments and soils both in the underground river and surface system. The higher mass InP, DaA, and BgP were non-detectable (ND) in the waters both in the underground river and surface system (Table 1). However, these were detectable in the sediments and soils in the two systems. The higher mass PAHs might be accumulated and retained for a long time in the sediments and soils which account for the increases in values found in the sediments and soils. These results are similar to those of Wang et al. (2009). Aislable et al. (1999), Blanchard et al. (2004), and Luo et al. (2004) found similar higher concentrations of 2–3 rings of PAHs in soils, sediments and waters. The surface waters could be influenced by the water–gas exchange, and higher concentrations of dissolved PAHs of lower ring sizes might be related to the net input from the air (Luo et al. 2004). In addition, it is necessary to study the PAHs contamination conditions in the air in the study area (Luo et al. 2004). This study did not find PAHs data in the concerned atmosphere. However, Golomb et al. (2001) reported the atmospheric deposition of PAHs to surface waters. In addition, Readman et al. (1982) proposed volatilizations is an important process in determining the fate of lower molecular weight PAHs, including naphthalene, phenanthrene, and anthracene, which accounted for the higher concentrations of lower masses PAHs in the waters, the soils and the sediments. Readman and Mantoura (1984) showed that PAHs with molecular weight >200 mostly occur in the particulate phase, which also accounted for the abundance of higher molecular PAHs in the sediments of the underground river and the sediments and the soils of the surface system.

The similarities in sources and the variations of different rings sized PAHs indicate the pathways and the role of the surface system in the karst. Surface waters, sediments, and soils could transport into the underground river through leaching and percolation, because leaching and percolations of organic materials are common phenomena in surface systems (Thurman 1985). The karst features might allow the easy leaching and percolation processes (Fig. 10). Seepage waters and drip waters commonly found in the underground cave are evidences of the transport of surface waters and soils into underground river (Schwartz et al. 2011).