Acute and mechanism-specific toxicity of oils and oil spill response actions - Adaption of relevant bioanalytical tools and evaluation of ecotoxicological effects

Due to an increasing demand and consumption the worldwide oil production is still rising and further predicted to grow within the next decades. Hence, continuing exploration and transportation leads to a constant risk for oil spills in the aquatic environment. Each oil spill provides a unique scenario affected by individual characteristics of the oil and environmental conditions. In order to accomplish a comprehensive risk assessment of the individual situation minimizing the adverse impact, important knowledge about regional ecology, the key oil compounds as well as the ecotoxicological risk of the spilled oil is required. Crude oils and refined fossil fuels are complex mixtures consisting of thousands of compounds with widely varying physical and chemical properties. As a consequence, the variety of characteristics might lead to a range of potential exposure scenarios and resulting strength of toxicity for aquatic biota. To assess the hazardous potential of complex environmental samples, targeted chemical analysis can imply great uncertainties and hence should be complemented with effect-based bioanalytical tools. Against this background, the scope of the present thesis was to acquire relevant knowledge about the environmental effects of oil spills and response actions within in the framework of the EU Horizon 2020 funded project GRACE. A comprehensive bioassay battery on acute and mechanism-specific toxicity was optimized and applied to three different petroleum products varying in their degree of processing from crude oil to distillate. Focusing on laboratory model species such as zebrafish (Danio rerio) embryos and adults as well as microscale in vitro assays, water-accommodated fractions (WAFs) of untreated and chemically dispersed oils as well as in situ burn resides were investigated. The overall aims were to (a) understand modes of action of crude and refined petroleum toxicity, (b) establish novel biomarkers for oil and oil spill effects, (c) evaluate the environmental effects of different oil spill response measures, and (d) establish an effect-based toolbox of oil toxicity testing. In zebrafish petroleum toxicity was addressed on different levels of biological organization from transcriptome over enzymatic biomarkers up to behavior and survival. Overall, typical morphological effects in zebrafish were successfully linked to biomarker responses and gene expression changes. Major findings include the suggestion of oculotoxicity in fish early life stages as an important endpoint, which might be comparably relevant for petroleum embryotoxicity as cardiotoxicity. Evidence of impairments of the visual system was found on transcriptional, histological, morphological and behavioral level. In this context, especially the larval reaction to alternating photoperiodic stimuli in a swimming behavior assay was suggested as a very sensitive endpoint. Microscale assays focused on potential modes of action like general cell viability, oxidative stress, genotoxicity and endocrine disruption. In light of today’s very limited application of microscale assays in oil risk assessment, their successful adaption and significant responses should not only indicate the relevance of those endpoints for the (eco-)toxicological characterization of oil contamination but also promote their valuable use as cost- and time efficient screening tools. The present thesis showed that in situ burn residues from a large-scale field experiment at least did not increase the toxicity compared to the initial oil. In contrast, the application of a chemical dispersant strongly increased the acute toxicity across all investigated organisms and cell lines. However, dispersant characteristics and changing partitioning kinetics have to be carefully considered for the interpretation of the dispersant’s role in toxicity. Results of the present thesis indicate that the dispersants promoted the bioavailability of crude oil constituents but also contributed to the observed toxicity itself. Especially the influence of the particulate oil fraction in the water column is controversially discussed across the scientific community and is one of the major research needs revealed within the current thesis. The present thesis was able to identify specifically small oil droplets in the nm range, which were calculated to be bioavailable throughout the entire exposure period. This indicates a high relevance of particulate oil fractions for the fate and behavior of oil in relevant exposure scenario. Since in contrast to native and chemically dispersed oil dispersants alone did induce acute- but not mechanism-specific toxicity in model organisms and in vitro assays, the hypothesis was derived that the dispersant cause of effects might mainly be associated with baseline toxicity (narcosis). The ecotoxicological characterization of petroleum contamination needs to include particular tools relevant for the marine environment. In this context, the present thesis successfully established a set of biological endpoints in the marine medaka (Oryzias melastigma), providing a promising marine laboratory alternative to zebrafish. Likewise, a yeast-based reporter gene assay on estrogenicity, which is characterized by high salinity tolerance, was adapted to petroleum WAF testing. Evaluating a unique data set from a comprehensive bioassay battery by means of a multivariate statistical approach, a sensitive fingerprinting toolbox for oil contamination was elaborated. Especially the invertebrates Calanus finmarchicus and Mytilus trossulus were the most sensitive species towards petroleum WAF exposure. Furthermore, the thesis showed that endocrine disruption is of major concern for the contamination of the aquatic environment with oil constituents. Focusing on genotoxicity, especially the micronucleus assay using the fish cell line ZF-L seem to provide a rapid and low-cost alternative to animal experiments. Overall, it is recommended to combine different effect-based methods to support a reliable risk assessment, since not every pathway initiated on low levels of organization such as transcriptional responses will manifest on higher orders. This toolbox in general presents a first step for the design of efficient oil toxicity studies. Subsequent discussions about details of the toolbox such as the transformation of effect data into a shared classification system are encouraged in order to further improve the strategies