Occurrence of perchloroethylene in surface water and

Occurrence of perchloroethylene in surface water and fish in a river ecosystem affected by groundwater contamination Zdena Wittlingerová, Jiřina Macháčková, Anna Petruželková & Magdalena Zimová Environmental Science and Pollution Research ISSN 0944-1344 Environ Sci Pollut Res DOI 10.1007/s11356-015-5806-7

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Environ Sci Pollut Res DOI 10.1007/s11356-015-5806-7

RESEARCH ARTICLE

Occurrence of perchloroethylene in surface water and fish in a river ecosystem affected by groundwater contamination Zdena Wittlingerová 1 & Jiřina Macháčková 2 & Anna Petruželková 1 & Magdalena Zimová 1

Received: 15 August 2015 / Accepted: 11 November 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract Long-term monitoring of the content of perchloroethylene (PCE) in a river ecosystem affected by groundwater contamination was performed at a site in the Czech Republic. The quality of surface water was monitored quarterly between 1994 and 2013, and fish were collected from the affected ecosystem to analyse the content of PCE in their tissue in 1998, 2011 and 2012. Concentrations of PCE (9–140 μg/kg) in the tissue of fish collected from the contaminated part of the river were elevated compared to the part of the river unaffected by the contamination (ND to 5 μg/kg PCE). The quality of surface water has improved as a result of groundwater remediation during the evaluated period. Before the remedial action, PCE concentrations ranged from 30 to 95 μg/L (1994– 1997). Following commencement of remedial activities in September 1997, a decrease in the content of PCE in the surface water to 7.3 μg/L (1998) and further to 1 μg/L (2011) and 1.1 μg/L (2012) led to a progressive decrease in the average concentration of PCE in the fish muscle tissue from 79 μg/kg (1998) to 24 (2011) and 30 μg/kg (2012), respectively. It was determined that the bioconcentration of PCE does not have a linear dependence because the decrease in contamination in the fish muscle tissue is not directly proportional to the decrease in contamination in the river water. The observed Responsible editor: Hongwen Sun * Jiřina Macháčková [email protected] 1

Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Kamýcká 129, Prague 6, Suchdol 165 21, Czech Republic

2

Technical University of Liberec, Institute of Nanomaterials, Advanced Technologies and Innovation, Studentská 1402/2, Liberec 1 461 17, Czech Republic

average bioconcentration factors were 24 and 28 for the lower concentrations of PCE and 11 for the higher concentrations of PCE in the river. In terms of age, length and weight of the collected fish, weight had the greatest significance for bioconcentration, followed by the length, with age being evaluated as a less significant factor. Keywords Fish tissue contamination . PCE . Bioconcentration . Groundwater contamination . Surface water contamination . Non-linear BCF . Groundwater clean-up

Introduction Chlorinated ethenes, especially perchloroethylene (PCE) and trichloroethylene (TCE), were widely used in the second half of the twentieth century as industrial degreasers and extraction agents. Due to a lack of knowledge of the fate of these substances in the environment, little attention was paid to the handling of these substances during the disposal of spent solvents and operational spills. In many areas, this has led to the development of long-term contamination of the subsurface because chlorinated ethenes (CE) are substances that degrade very slowly in the subsurface environment (Ellis et al. 2000; Hendrickson et al. 2002; Bradley 2003; Kueper et al. 2003; Ademola et al. 2004; Grandel and Dahmke 2004; Löffler and Edwards 2006; Moran et al. 2007; Aulenta et al. 2010; Barnes et al. 2010; Walter et al. 2011; Tobiszewski and Namieśnik 2012; Guan et al. 2013; Koenig et al. 2014) Natural attenuation processes mainly take place in an anaerobic environment, in areas with a content of other organic compounds sufficient enough to stimulate anaerobic microbial communities in which microorganisms are present with dehalorespiring metabolic pathways. Conditions

Author's personal copy Environ Sci Pollut Res

supporting complete and quantitative dehalorespiration or other natural attenuation processes do not occur very often in the natural subsurface environment, and so the contamination of soil and groundwater is often part of the challenge of remedial geology. CE can migrate in the subsurface and contaminate groundwater and therefore potential sources of drinking water (Maymó-Gatell et al. 1997; Wiedemeier et al. 1999; Ademola et al. 2004; Grandel and Dahmke 2004; Dong et al. 2009; Tobiszewski and Namieśnik 2012). CE can also follow the groundwater flow and enter into other environmental components, i.e. surface water, and negatively influence ecosystems. PCE and TCE can bind to receptors on aquatic organisms and terrestrial animals. Toxicity concentrations for PCE in aquatic organisms are in the range of 1 to 100 mg/L (US EPA 1994). The LC50 (the concentration causing lethal complications in 50 % of test subjects) during a 96-h test with fathead minnow (Pimephales promelas), bluegill (Lepomis macrochirus) and rainbow trout (Oncorhynchus mykiss) was in the range of 5 to 21.4 mg/L. During a 48-h test, the LC50 for Daphnia magna was 9.1 to 18.0 mg/L (US EPA 1994). Bioconcentration factors (BCF) measured in fish are 39 (rainbow trout, Oncorhynchus mykiss) and 49 (bluegill, Lepomis macrochirus) (US EPA 1988). These values show that the bioaccumulation of PCE in fish is not significant but that environmental exposure may be harmful to the health of aquatic organisms (US EPA 1998). Due to the high volatilization of PCE and TCE and photochemical decomposition in the air, significant toxic effects on terrestrial organisms outside areas of chronic exposure are not expected (US EPA 1994). The site selected for studying the effects of groundwater contamination by PCE on the contamination of surface water ecosystems was a rendering plant in Northern Bohemia, the Czech Republic, which is an area with extensive groundwater contamination, sometimes referred to as a mega-site. PCE and its degradation products migrating from the groundwater into the surface water of a flowing river were monitored over a 20-year period at the studied site. The site has undergone several scientific investigations, including a study of the transfer of chlorinated ethenes into wood (Trapp et al. 2007; Wittlingerova et al. 2013), monitoring of the changes in the isotopic composition of contaminants (Wiegert et al. 2012), and the possibilities for spatial evaluation of the areal extent of contamination using modern statistics (Wahyudi et al. 2012). This study focuses on the bioconcentration of PCE in the fish living in the river, which is influenced by the drainage of contaminated groundwater. During the investigation, we monitored the content of PCE in the groundwater, surface water and in the tissues of fish living in the affected freshwater ecosystem.

Material and methods Experimental site The experimental site is located near the town of Mimoň in Northern Bohemia, the Czech Republic (Fig. 1). The SAP factory is situated on a floodplain on the right bank of the Ploučnice River, at an altitude of 272 m above sea level. The area of interest is part of the Czech Cretaceous Basin. Middle Turonian sandstones, developed to a thickness of 65–75 m, form the upper part of the Cretaceous strata. In the top part of the profile, fine- to coarse-grained sandstone alternates to a depth of 30 m. The grain size decreases towards the base of the formation, and the coarse-grained sandstone gradually changes to fine-grained sandstone (Macháčková et al. 2008). The Middle Turonian sandstone is covered in the floodplain by Quaternary alluvial deposits of the Ploučnice River in the form of sands and gravels. The maximum thickness of the alluvial deposits is 6 m. The geological profile in the area of the site is enclosed by a layer of anthropogenic backfill with a thickness of up to 3 m. The deposited material is mostly sandy loam with some municipal waste, slag or ash. The unconfined Middle Turonian aquifer reaches a thickness of 60 m and is hydraulically connected to the Quaternary aquifer, which reaches a saturated thickness of 3–5 m. The Middle Turonian aquifer has fracture/inter-granular permeability with a transmissivity coefficient varying in the range of 500–700 m2/day and hydraulic conductivity of 13–70 m/day. The quaternary aquifer has inter-granular permeability with similar parameters. The groundwater level in the area of interest was encountered by a network of monitoring wells at a depth of 0.14 to 2.99 m below ground level. The area is drained by the Ploučnice River with an average flow of 4.92 m3/s (1995– 2010, data from the Czech Hydrometeorological Institute). The majority of the natural groundwater flow occurs in the upper part of the Middle Turonian and Quaternary aquifers to a depth of approximately 20 m below ground level. The aquifers are recharged by infiltrated rainwater and drain into the Ploučnice River (Prokšová et al. 2009; Zachař 2011). The SAP factory focussed on rendering activity, during which PCE was used for material processing (slaughterhouse waste, carcasses) as a fat extracting agent between 1965 and 1986. PCE was consumed and decayed during production. The consumption was high—in total 4250 t was consumed over the 20 years of operation, with an annual volume of 160 to 200 t. The high turnover of this material led to frequent operational leaks, and PCE was also a constituent of the wastewater that was discharged freely into the floodplain (Vargová 1994, Prokšová et al. 2009). The result of this activity was extensive contamination of the subsurface at the site. The contamination from the site did not spread in the direction of the groundwater flow but was influenced by pumping of groundwater from the Boreček


Fig. 1 The site location with sampling profiles

waterworks, which supplied water with a yield of 20 L/s to the adjacent Soviet Army air base. The water source was situated on the opposite side of the Ploučnice River (Fig. 2). The first manifestation of contamination (66 mg/L of PCE) was determined 400 m away in an extraction well of the Boreček waterworks in 1988. However, the SAP site was not identified at that time as the source of the contamination due to its location on the opposite river bank and the inappropriate location of the investigation wells, which were situated in the assumed groundwater flow direction. The influence of pumping of groundwater at the waterworks on the groundwater flow was not taken into consideration. A laundry of the former Soviet Army in the nearby military base, which was still operating during the 1980s, was considered to be the source. But, the military laundry used TCE, whereas the contamination at the

Boreček site was caused by PCE. This was only determined later when the Soviet Army was made to leave Czechoslovakia in 1991, and more detailed investigations were possible in the formerly occupied zone. Further investigations in 1993 showed that the source of pollution is actually the SAP site (Černý et al. 1994). More extensive investigations and subsequent remediation began in the second half of the 1990s. The areal extent and cross-sectional depth of the contamination prior to the remedial work in 1997 are shown in Fig. 2. The contamination at the site was very extensive, with a 500-m plume covering an area of 10.1 ha. The total amount of PCE determined in the subsoil in 1997 was estimated in the range of 149 to 246 t using the Monte Carlo method (confidence interval of 0.8) (Macháčková and Soukup 1998). The maximum levels of contamination were detected in a 2–20-m layer



Fig. 2

The contamination extent in 1998—a aerial view and b cross-

section

of quaternary alluvial sediment and weathered Middle Turonian sandstone; migration to deeper strata was not verified by zonal wells. The average concentration of total chlorinated ethenes in the surface water of the Ploučnice River prior to the remediation ranged from 49.7 to 61.7 μg/L (1994–1996), with approximately 16 kg of chloroethenes, predominantly PCE, entering the river each day as a result of groundwater drainage. Extensive remediation of the groundwater was performed at the site from September 1997 to April 2015, using methods of pump and treat, venting, air sparging and the application of chemical oxidation with ozone. At the start of the remediation, contaminated groundwater was pumped at a volume of 20 L/s; between 2008 and 2015, the pumped volume was gradually reduced due to a decrease in the extent of the contamination plume to a volume of 3.6 L/s. An extensive site investigation was performed between 2006 and 2009, which explained the geological structure and its effect on the distribution of contamination at the site in more detail (Fig. 3). This was performed based on the results of membrane interphase probing, which revealed a layer of hardened sandstone at the base of the Quaternary sediments, which worked as a trap for DNAPL (Wahyudi et al. 2012). A combination of different survey methods (macroscopic description of drill cores, grain analysis and geophysical measurements) was used to verify in detail the geological structure of the upper cross-section of the Middle Turonian and Quaternary sediments. It was determined that the presence of two continuous layers with low permeability has an effect on the distribution of PCE in the rock environment. The extent of the contamination in the Quaternary aquifer in 2011 and the newly defined geological structure are illustrated in Fig. 3 (Macháčková et al. 2008; Prokšová et al. 2009; Zima 2011). In 2012, there was still 4.4 to 10.7 t of PCE in the subsurface at the site, calculated using the Monte Carlo method with a confidence interval of 0.8 (Prokšová and Fadrhons 2014). Investigation of the river ecosystem Monitoring surface water contamination Contamination of the surface water has been monitored since 1994 in the profiles upstream and downstream the SAP site location. The monitoring points are illustrated in Fig. 1. Surface water was collected from a distance of 1 m from the bank, at one third of the depth of the water column at the point of flow using a sampling device on a rod. Water samples were collected in 250-mL glass sample containers with a Teflon seal and transported in a cooler on ice to the laboratory. In the laboratory, the PCE content was analysed using headspace gas chromatography coupled with mass spectrometry.

Aliquot proportions of water samples were transferred into 20-mL headspace vials. After the equilibration between liquid and gaseous phases was achieved, a defined volume of gaseous phase was transferred by the carrier gas flow through a transfer capillary to a chromatographic column for separation and measurement. Analytical methods were based on EPA 601 and EPA 624 standard methods. The laboratory conducting the measurements was accredited by the Czech Institute of Accreditation according to the standard ČSN EN ISO/IEC 17025. The quality of the laboratory work is repeatedly tested by analysing duplicate samples and samples with standards, when a known amount of contaminant is added to a clean environmental matrix both to find the lowest detection limit and to control accuracy and correctness of analyses. The quality of the laboratory work during the research was also repeatedly controlled by analysing duplicate field samples (5 % of the samples). The measured control values were in the range of the confidence interval of the analysis, with variation being less than 15 %. Participation of the laboratory in interlaboratory comparison tests was another part of the internal quality control system. The accuracy of measurements was expressed as an upper and lower limit of the confidence interval of the detected value, which is based on the analysis of laboratory replicates. The confidence interval for the measured values at P 90 was ±30 %. As the study covered 20 years of surface water monitoring, the technical equipment of the laboratory and details of the analytical method have changed several times. The quality of work was assured as described above. Surface water monitoring was conducted quarterly as a part of the water quality monitoring at the site. Collection of fish The first evaluation of the river ecosystem in terms of the influence of long-term contamination by chlorinated ethenes was performed in 1998. The evaluation included stream morphology, sampling and evaluation of benthic organisms to obtain a saprobic index, and collection of fish for ichthyological evaluation and analysis of the content of chlorinated ethenes. Fish were collected again in 2011 and 2012. Lengths of the river measuring 150 m were selected at two locations (A, B, see Fig. 1) for the investigation, and fish were caught using an electric unit (Honda 2.2 kW, voltage 220 V, current 0.5 A). The collected fish were evaluated ichthyologically, determination of the caught species was performed, including the age of the fish (microscopic analysis of scales and otoliths in the case of Barbatula species) and their growth characteristics (length, weight) according to standard statistical methods (Pivnička 1981; Gulland 1983). In order to determine the PCE content in muscle tissue and liver tissue, the fish were sacrificed directly at the site and transported to the laboratory on ice, where they were subsequently frozen. The section between the anus and tail fin was used for the analysis. The


Fig. 3 The contamination extent in 2011—a aerial view of quaternary aquifer, b aerial view of cretaceous aquifer, and c cross-section


Fig. 3 continued.

frozen samples were sliced with a cold knife, disintegrated and extracted with methanol. Between 1 and 3 g of tissue were placed in 5 mL of methanol and sonicated for 10 min. The extracts were transferred into 2-mL vials and after liquid injection further processed in the same manner as the water samples. The confidence interval for the measured values at P90 was +40 %. Only one sample per sampling event from an individual fish was analysed because of the limited amount of tissue. The muscle tissue of several individuals was pooled in the case of young fish in their first year of life to obtain the desired amount of the sample. The accuracy of the determination was verified using duplicate analysis in one case; the results are given in the Table 1. The difference of the measured values was 11 %. Statistical analysis Arithmetic means and medians were used for the statistical analysis, and box plots were processed in the software environment STATISTICA 12. Pearson’s correlation coefficient was used to evaluate possible dependence (Pentecost 1999). The measured data—length, weight and age of the fish—were correlated with the concentrations of PCE in the fish muscle tissue. The values of the PCE content that were below the detection limit were taken in the calculation to be values on the detection limit—for the values of the detection limit in fish

muscle tissue see Table 1, the detection limit in surface water was 0.5 μg PCE/L. The correlation coefficient r ranges from −1 to 1, the closer to one, the more significant the dependence to the investigated phenomena; however, close correlation of data does not necessarily indicate functional dependence. Absolute values of r in the range of 0 to 0.3 indicate a low linear correlation dependence, from 0.3 to 0.5 a slight correlation dependence, from 0.5 to 0.7 a mean correlation dependence, from 0.7 to 0.9 a high degree of correlation dependence, from 0.9 to 1 a very high degree of dependence and 1 means functional dependence. Reliability, with which the correlation function the selected correlation curve describes, is shown by the coefficient r2 (range of 0 to 1, the closer to one, the higher the degree of reliability of the correlation equation). The statistical significance and the correlation strength are given by n—the number of measurements from which the correlation is calculated (Pentecost 1999; Hanousek and Charamza 1992). Data were only correlated from profile B; data from profile A were not assessed in this way due to the large number of analysis results below the detection limit both for the fish and the water. Because a non-linear BCF dependence on the concentration of PCE in the water was determined during the data evaluation, the data was divided into two sets—data from 1998 and data from 2011 to 2012. The individual species and overall group of fish were evaluated.

Author's personal copy Environ Sci Pollut Res Table 1

Concentration of PCE and ichytological parameters of fish

Fish species

Sample date

Sample location

Age of fish

Length of fish (mm)

Weight of fish (g)

Concentration of PCE (μg/kg), fish tissue

Gobio gobio—8

27.5.1998

A

4

100

17