RESUMEN
Despite their low water solubility, hydrophobic pollutants are widespread in the aquatic environment and could represent a threat for living organisms. EU regulations on chemicals require accurate and reliable data on chemical toxicity. Current normalised fish toxicity assays, in particular those advocated by OECD guidelines, do not allow reliable toxicity assessment of hydrophobic compounds due to their low water solubility. In order to accurately evaluate the toxicity of this kind of compounds, a new spiked sediment assay using embryos of the Japanese medaka was developed. It consists of directly exposing fertilised eggs, during their entire embryonic development, onto the reference sediment spiked with the test compound. A large set of lethal or sublethal effects in embryos and newly hatched larvae, including non-invasive endpoints is analysed in order to maximise the sensitivity of the test. The approach was validated using four model pollutants with different modes of action: DMBA, PCB126, PCB153 and 4-nonylphenol (NP). All compounds, except PCB153, induced a dose-dependent increase in toxic effects. In fact, lethal effects only occurred at the highest tested concentration. In contrast, sub-lethal effects including skeletal deformations, cardiac activity modulation, body length reduction and hatching delay were observed at low to moderate concentrations of DMBA and PCB126. NP induced subtle effects in embryos, altering cardiac activity and hatching success but only at high concentrations. Although a few more improvements would make it a fully standardised assay, this spiked sediment assay using medaka embryos proves to be sensitive enough to measure hydrophobic chemical toxicity using an environmentally realistic mode of exposure.