Ecotoxicological Assessment of the Upper Danube River
Research Articles
Research Articles
Ecotoxicological Assessment of Sediment, Suspended Matter and Water Samples in the Upper Danube River A pilot study in search for the causes for the decline of fish catches * Steffen Keiter1, Andrew Rastall2, Thomas Kosmehl1, Karl Wurm3, Lothar Erdinger2, Thomas Braunbeck1 and Henner Hollert1* 1 Department
of Zoology, Aquatic Ecology and Toxicology Section, University of Heidelberg, Im Neuenheimer Feld 230, 69120 Heidelberg, Germany 2 Department of Hygiene, University of Heidelberg, Im Neuenheimer Feld 324, 69120 Heidelberg, Germany 3 Gewässerökologisches Labor, Tulpenstr. 4, 72181 Starzach-2, Germany * Corresponding author (
[email protected])
DOI: http://dx.doi.org/10.1065/espr2006.04.300 Abstract
Goals, Scope and Background. Fish populations, especially those of the grayling (Thymallus thymallus), have declined over the last two decades in the upper Danube River between Sigmaringen and Ulm, despite intensive and continuous stocking and improvement of water quality since the 1970s. Similar problems have been reported for other rivers, e.g. in Switzerland, Great Britain, the United States and Canada. In order to assess if ecotoxicological effects might be related to the decline in fish catch at the upper Danube River, sediment, suspended matter and waste water samples from sewage treatment plants were collected at selected locations and analyzed in a bioanalytical approach using a battery of bioassays. The results of this pilot study will be used to decide if a comprehensive weight-of-evidence study is needed. Methods. Freeze-dried sediments and suspended particulate matters were extracted with acetone in a Soxhlet apparatus. Organic pollutants from sewage water were concentrated using XAD-resins. In order to investigate the ecotoxicological burden, the following bioassays were used: (1) neutral red assay with RTL-W1 cells (cytotoxicity), (2) comet assay with RTLW1 cells (genotoxicity), (3) Arthrobacter globiformis dehydrogenase assay (toxicity to bacteria), (4) yeast estrogen screen assay (endocrine disruption), (5) fish egg assay with the zebrafish (Danio rerio; embryo toxicity) and (6) Ames test with TA98 (mutagenicity). Results and Discussion. The results of the in vitro tests elucidated a considerable genotoxic, cytotoxic, mutagenic, bacteriotoxic, embryotoxic and estrogenic burden in the upper Danube River, although with a very inhomogeneous distribution of effects. The samples taken from Riedlingen, for example, induced low embryo toxicity, but the second highest 17β-estradiol equivalent concentration (1.8 ng/L). Using the fish egg assay with native sediments, a broad range of embryotoxic effects could be elucidated, with clear-cut dose-response relationships for the embryotoxic effects of contaminated sediments. With native sediments, embryotoxicity was clearly higher than with corresponding pore waters, thus corroborating the view that – at least for fish eggs – the bioavailability of particle-bound lipophilic substances in native sediments is higher than generally assumed. The effect observed most frequently in the fish egg assay was a
308
developmental delay. A comparison of our own results with locations along the rivers Rhine and Neckar demonstrated similar or even higher ranges of ecotoxicological burdens in the Danube River. Conclusions. The complex pattern of ecotoxicological effects caused by environmental samples from the Danube River, when assessed in an in vitro biotest battery using both acute and more specific endpoints, showed that integration of different endpoints is essential for appropriate hazard assessment. Overall, the ecotoxicological hazard potential shown has indeed to be considered as one potential reason for the decline in fish catches at the upper Danube River. However, based on the results of this pilot study, it is not possible to elucidate that chemically induced alterations are responsible for the fish decline. Recommendations and Perspectives. In order to confirm the ecological relevance of the in vitro results for the situation in the field and especially for the decline of the grayling and other fishes, further integrated investigations are required. For linking the weight of evidence obtained by in vitro assays and fish population investigations, the application of additional, more specific biomarkers (e.g. vitellogenin induction, EROD and micronucleus assay) has been initiated in fish taken from the field as well as in situ investigations. Keywords: Bioassay battery; cytotoxicity; Danube; embryotoxicity, European Water Framework Directive; fish decline; genotoxicity; sediment contact assay; sewage water; weight-ofevidence; yeast estrogen screen
Introduction
Over the last two decades, a decline of fish catches has been reported for several streams in Europe, the USA and Canada (Burkhardt-Holm & Segner 2002, Cook et al. 2003, de Lafontaine et al. 2002, Faller et al. 2003, Fischnetz 2004). In certain Swiss rivers and streams, the catch of fish, especially of brown trout, has decreased by approximately 50% over the last two decades (Burkhardt-Holm et al. 2005, * This article has been developed on the basis of presentations held at the annual meetings of the SETAC-GLB 2003 and 2004 in Heidelberg and Aachen
Environ Sci Pollut Res 13 (5) 308 – 319 (2006) © 2006 ecomed publishers (Verlagsgruppe Hüthig Jehle Rehm GmbH), D-86899 Landsberg and Tokyo • Mumbai • Seoul • Melbourne • Paris
Research Articles
Ecotoxicological Assessment of the Upper Danube River
Burkhardt-Holm & Segner 2002). In order to identify factors responsible for the fish decline, the 'Fishnet' project was initiated in 1998. To structure the search for causes, 12 hypotheses were developed. These hypotheses include causeeffect relationships at multiple levels, with some overlap and interaction between them. One of these hypotheses – 'Chemical pollution (both nutrients and micropollutants) is causing harmful effects' – was used to address possible ecotoxicological reasons. As a consequence, three key factors of national or regional importance were identified in more than 60 integrated projects: Fishery management, parasitic, proliferative kidney disease (PKD), and habitat situations (morphology and water quality). Particularly at the regional scale, chemically-induced declines of fish populations seem to be relevant, since adverse effects on reproduction physiology in fish downstream of wastewater treatment plants were found, especially during dry seasons and where effluent dilution was low (Burkhardt-Holm et al. 2005, BurkhardtHolm & Segner 2002). In analogy to Switzerland, a decline of fish catch has been documented for the upper Danube River in Southern Germany since the beginning of the 1990s (Fig. 1). Despite continuous stocking programs (Wurm 2001), significant declines have been reported in a number of various species including chub (Leuciscus cephalus), barbel (Barbus barbus), pike (Esox lucius), brown trout (Salmo trutta fario) and, in particular, grayling (Thymallus thymallus). As seen in various in vivo experiments, the grayling is known to be a very sensitive species regarding chemical pollution (Fjeld et al. 1998, Peuranen et al. 2003). Within the same period, the water quality in the upper Danube River was improved, as could be shown by chemical analyses (heavy metals and organic priority pollutants) as well as by monitoring hydrological (pH, nitrate, ammonium, phosphate, temperature, oxygen) and biological parameters (saprobic index; LfU-Baden-Württemberg 2002). The extensive modernization of wastewater treatment plants since the late 1970s is thought to be the major contributing factor to these improvements. Nevertheless, surface waters still receive a considerable load of nutrients as well as anthropogenic chemicals and their metabolites from incomplete elimination in wastewater treatment plants (Ahel et al. 1996, Garcia-Reyero et al. 2001, Hollert et al. 2005, Lee et al. 2004, Murk et al. 2002), at-
Number of catches
1200 Sigmaringen Riedlingen Ehingen
1000 400 200 0 1980
1985
1990
1995
2000
Fig. 1: Number of grayling caught in the upper Danube river between 1980 and 2000 (redrawn from Wurm 2001)
Environ Sci Pollut Res 13 (5) 2006
mospheric deposition (Erdinger et al. 2005, Hellmann 1996) and runoff from both agricultural areas (Berenzen et al. 2005, Pedersen et al. 2003, Schulz & Liess 1999) and urban surfaces (Fries & Puttmann 2003, Maltby et al. 1995). Several studies revealed that not only nutrients and water-soluble priority pollutants, but also (1) non-priority pollutants (e.g. genotoxic and endocrine disrupting substances) and (2) particle-bound chemicals may contribute significantly to both ecotoxicological effects in specific bioassays and alterations in fish health (Faller et al. 2003, Vos et al. 2000, White et al. 1998). Especially sediments have been recognized as both a major sink and a potential source for persistent toxic chemicals in aquatic systems (Ahlf et al. 2002b, Burton 1991, Förstner & Westrich 2005, Heise & Ahlf 2002, Hill et al. 1993, Hollert et al. 2002b, Kammann et al. 2005a) and are well known to negatively affect fish health (de Lafontaine et al. 2002, White et al. 1998). The upper Danube is the first river in Germany with a well documented fish decline over two decades, which cannot be explained by conventional chemical analysis of priority pollutants and the monitoring of hydrological and biological parameters. Chemical analysis is capable of identifying both source substances and their metabolites; however, this method inherently fails to provide data about biological effects to organisms (Ahlf 1995, den Besten et al. 2003, Heise & Ahlf 2002). A complete chemical screening of all substances present in aquatic ecosystems would be much too resource- labor- and time-consuming. Moreover, even the most elaborate chemical analysis would not be able to provide information about synergistic / antagonistic effects. In contrast, ecotoxicological bioassays do provide data about biological effects, although without identifying either the responsible substances or their potential sources. In recent weight-of-evidence studies, acute bioassays and mechanism-based bioassays have been used as a primary line of evidence to evaluate if an ecotoxicological hazard potential of complex environmental compartments is either existent or negligible (Chapman et al. 2002, Chapman & Hollert 2006). However, previous studies have repeatedly documented that a comprehensive ecotoxicological hazard assessment cannot be made on the basis of only one bioanalytical system; instead, a combination of biological test systems has been recommended (Ahlf et al. 2002a, Ahlf & Heise 2005, Ahlf et al. 2002b, Burton 1995, den Besten et al. 2003, Heise & Ahlf 2002, Hollert et al. 2002a, Hollert et al. 2002b, Wenzel et al. 1997). Thus, the objective of this pilot study (tier 1) was to use a comprehensive in vitro bioassay battery covering acute and mechanism-based toxicity to evaluate if an ecotoxicological hazard potential could be found in the upper Danube River for the compartments of sediment, suspended particulate matter and effluents from sewage treatment plants. Only if a significant, ecotoxicological hazard potential were evident at tier 1, a more comprehensive weight-of-evidence study would have to be carried out in the future to elucidate whether the decline of fish is induced by chemicals or by other non-toxic parameters. Thus, positive results at tier 1 would trigger a comprehensive weight-ofevidence study using a multiple line of evidence (tier 2).
309
Ecotoxicological Assessment of the Upper Danube River Accordingly, we used a bioassay battery covering both acute toxicity (cytotoxicity, toxicity to bacteria and embryo toxicity) as well as specific endpoints (genotoxicity, mutagenicity, endocrine disruption and teratogenicity) for an investigation of the hazard potential of sediments, suspended particulate matter as well as water and the particle phase of effluents. In order to not only address the hazard potential, but also the actually bioavailable portion of the hazard potential, we used acetonic extracts of sediments, suspended particulate matters and the particle phase of the effluents, as well as XAD water extracts for the water phase of effluents as a worst case scenario. Bacterial toxicity was assessed using native, untreated effluent samples. 1
1.1
Research Articles
Sample processing and extraction
1.1.1 Sediment and SPM samples
Near-surface sediment samples (0–5 cm depth) were taken from four sites immediately downstream to the local STP outlets (see Fig. 2). SPM samples were collected from two sites with the aid of two SPM traps, as described by Hollert et al. (2000). Both sediment and SPM samples were shockfrozen at –30°C immediately after return to laboratory and freeze-dried (Christ Alpha 1–4 freeze dryer, Osterode, FRG). Two replicates of 20 g freeze-dried sediment and SPM samples were separately extracted with 300 ml acetone (Fluka, Buchs, CH) for twelve hours using standard reflux (Soxhlet) extractors at approximately 7 cycles per hour. The extracts were allowed to cool and reduced in volume to approximately 5 ml using a rotary evaporator (WB 2001; Heidolph, Kehlheim, FRG; 400 mbar, 36–38°C) before being reduced close to dryness under a nitrogen stream. Residues from individual sample replicates were re-dissolved in either 2 ml ethanol (Fluka) for the yeast estrogen screen (YES) or 2 ml dimethyl sulfoxide (DMSO; Sigma, Deisenhofen, FRG) for all other bioassays. Extracts were stored at –20°C until testing. Empty extraction thimbles extracted and processed in parallel served as process controls.
Material and Methods
In order to gain a broad overview of the ecotoxicological risk potential of the upper Danube River, samples of sediment (4), suspended particulate matter (SPM, 2), pulp mill effluent (PME, 1) and sewage treatment plant effluents (STPE, 5) were collected from selected locations along the river (Fig. 2) and subjected to the bioassays detailed in Table 1.
Sediment, STPE, PME
Sediment, STPE
SPM
Sediment, STPE
Sediment, STPE SPM
Fig. 2: Sampling sites along the upper Danube River between Sigmaringen and Öpfingen. SPM = suspended particulate matter; STPE = sewage treatment plant effluent; PME = pulp mill effluent
Table 1: Bioassays used to assess the ecotoxicological potential of sediment, suspended particulate matter and water (effluent) samples from the upper Danube River
Sigmaringen
Scheer
Riedlingen
Rottenacker
Öpfingen
Ehingen
1, 3, 4
2
1, 3, 4
1, 3, 4
2
1, 3, 4
3
1, 3, 4, 5
2
1, 3, 4, 5
1, 3, 4, 5
2
1, 3, 4, 5
3, 4, 5
Ames test
1
2
1
1
2
1
–
Yeast estrogen screen
3
–
3
3
–
3
3
Fish egg test
6
7
6
6
7
6
–
Neutral red assay Comet assay
Pulp mill
Dehydrogenase assay 5 – 5 5 – 5 5 1 = sediment extract; 2 = suspended particulate matter extract; 3 = effluent water phase extract; 4 = effluent particle phase extract; 5 = native effluent; 6 = native sediment; 7 = native suspended particulate matter
310
Environ Sci Pollut Res 13 (5) 2006
Research Articles
Ecotoxicological Assessment of the Upper Danube River
1.1.2 Effluent samples
was expressed as a percentage of the process controls (i.e. negative controls) and the data were plotted as concentration-response curves. Non-linear regression analysis was performed using SigmaPlot 9.0 (Systat, Erkrath, FRG), and the extract concentrations inducing 50% mortality after 48 h (48 h NR50) were calculated accordingly.
Sample volumes of 20 L were collected from the effluents of four sewage treatment plants (STP) and one pulp mill (see Fig. 2). All samples were chilled in a refrigerator to 4°C immediately after return to laboratory and filtered trough a 0.4 µm fiberglass filter (Sartorius, Göttingen, FRG) using nitrogen at 1 bar. The effluent particle phase (EPP) was extracted as detailed above for sediment and SPM samples. In each case, the sample filtrate was adjusted to pH 2. 1 L methanol (Fluka) was added to each sample before organic compounds were extracted by using a 1:1 (v:v) mixture of Amberlite-XAD 4 and XAD 7 resins (Serva, Heidelberg, FRG). With a flow rate of approximately 1 L per hour, the samples were extracted. To dry the resin, a nitrogen stream was used. Absorbed chemicals were eluted first with 50 ml of methanol (Fluka) and then with 50 ml of acetone (Fluka). Subsequently, the volume of the sample was reduced by means of a rotary evaporator (WB 2001; 400 mbar, 36–38°C) and fluid/fluid extracted using dichloromethane (Fluka) according to Hollert et al. (2005). The remaining solvated compounds of the effluent samples (effluent water phase = EWP) were concentrated close to dryness using a rotary evaporator (WB 2001; 400 mbar, 80–100°C) and resolved in acetone. The acetone and dichloromethane phases were combined and the extracts were divided into two sub-samples of 1/3 and 2/3 of the total weight proportions before being reduced almost to dryness under a nitrogen stream. The residues were then redissolved in either 0.667 ml ethanol (1/3 sub-sample) for the YES and in 1.333 ml DMSO (2/3 sub-sample) for all other bioassays. As a process control, 20 L of double-distilled water were extracted and processed. All samples and control extracts were stored at –20°C in the dark. Due to the reduction of the volume (20 L to 2 ml), a concentration factor (CF) of 10,000 is achieved for each sample. 1.2
Biotests
1.2.1 Neutral red assay (acute cytotoxicity)
The neutral red assay was performed according to Borenfreund & Puerner (1984) with slight modification detailed in Klee et al. (2004). RTL-W1 cells (a gift from Dr. Niels C. Bols, University of Waterloo, Canada) were cultured at 20°C in 75 cm2 plastic culture flasks (TPP, Trasadingen, CH) without additional gasing in Leibovitz medium (L15) supplemented with 8% fetal calf serum (Sigma), 1% penicillin/ streptomycin and 1% neomycin (Serva, Heidelberg, FRG). Sediment, SPM (20 g sediment and SPM equivalents (SEQ)/ ml DMSO) and EWP extracts were serially diluted with L15 medium to give a concentration range from 3.1 to 200 mg dry sediment and SPM (3.1–200 mg SEQ/ml) and 1.5 to 100 concentration factor per ml medium for the EWP extracts, respectively. The maximum concentration of DMSO in any extract dilution was below its non-observable effect concentration (NOEC) for RTL-W1 cells (
