NO2 concentrations in the experimental tanks stayed within 20% of the nominal concentrations. Exceptions were the 0.1, 0.25 and 0.5 mg L1 N–NO 2 treatments in the B. calamita bioassay in
which NO2 concentrations deviated almost 50% below nominal concentrations. This deviation was possibly the result of an error in the analytical procedure or the occurrence of NO 2 reduction
via bacterial or algal activity, possibly potentiated by a low renewal rate of exposure solution. NO 2 background concentrations of the dilution water used in the bioassays were between 0.000 and 0.021 mg L1 N–NO 2 . Test water dissolved oxygen levels remained above 70% and pH and conductivity remained stable throughout the experimental period. The test water from GM had lower background pH (5.6–6.7) and conductivity (15 lS cm 1) than the water from DNP (8.0–8.2; 422 lS cm 1). In general, average water conditions remained within the natural ranges of the respective breeding sites. The water temperature during the DNP bioassays was considerably higher (in average 4–7 C higher), and with less variation throughout each bioassay than in GM (Table 1 ).