| Abstract: | To investigate the reliability of analytical data for tributyltin (TBT) in sea water, split water samples were distributed to ten laboratories in six countries. The sub-surface samples comprised: (i) an offshore (0·5 km) water sample, (ii) the same sample but spiked with an undisclosed quantity of TBT standard compound (175 ng TBT+ liter−1), and (iii) a sample taken from a yacht marina. The seven acceptable data sets were in good agreement for the spiked sample (178 ± 26 ng TBT+ liter−1) but showed a greater variation in concentrations reported for the yacht marina sample (366 ± 93 ng TBT+ liter−1). Atomic absorption and gas chromatographic-flame photometric detection techniques produced results of similar accuracy and precision. Samples acidified with 1 ml of 10% (v/v) acetic acid appeared stable for more than 2 weeks when stored refrigerated and in darkness. Analyses of the offshore seawater sample revealed TBT contamination (9 ± 7 ng TBT+ liter−1) indicating dispersion of the compound to the shelf waters off Monaco. The spread in values reported by the laboratories demonstrates inherent difficulties in obtaining good precision below approximately 20 ng TBT+ liter−1. This observation is discussed with respect to the setting and enforcing of water quality standards. |
