Application of ring study:: Water toxicity determinations by bioluminescence assay with Vibrio fischeri

Acceptance of toxicity bioassays as effective analytical tools in environmental areas needs guarantees of standardization but also validation. Ten European laboratories took part in an inter-laboratory study using different commercial devices based on bioluminescence inhibition of bacteria Vibrio fischeri. Reproducibility and stability by short toxicity endpoints, effective concentration that gives 10%, 50% and 80% of inhibition (EC10, EC50 and EC80) is evaluated. Parametric and non-parametric statistic is applied and performance of participant laboratories is addressed by z-scores calculated by non-parametric statistic. z-Score classification was based on harmonised protocol for proficiency testing of analytical laboratories (satisfactory \z\ <= 2; questionable 2 <\z\ <= 3; unsatisfactory \z\ > 3). Tested samples were phenol, 3,5-dichlorophenol and influent wastewater. Based on z-score classification, more than 70% of the laboratories showed a satisfactory performance for phenol, 3,5-dichlorophenol and influent wastewater (86%, 90% and 70%, respectively). Reproducibility and stability was observed in toxicant references and in wastewater samples. EC80 determination appears to be more robust that EC10 and EC50. EC determinations can be considered favorable at 5 and 15 min of exposition, in particular for EC80. The use of different commercial devices can not be considered an additional source of variation. (C) 2005 Elsevier B.V. All rights reserved.