4.2. Determination of SMR, duration, and total SDA
When calculating the total SDA as the area under the MO2 curve for each fish, it is obvious that this measure is influenced by potential errors in estimation of SMR and the duration of the SDA response. Regarding SMR, we chose to use the lowest value obtained during the entire measurement (note that this value is already an average of three points). As this deviates from the normal methods for calculation of SMR (as an average of the 5 or more lowest point or the 10th quantile), we compared the values used here with those found by using the average of the 5 lowest values, and found no significant difference. Furthermore, the values obtained in hypoxia and normoxia are extremely similar (see Table 1), which is also expected in an air-breathing fish, despite the severe level of hypoxia, and compare well with previous studies on other Channa species (Ojha et al., 1979). We therefore feel confident that the SMR values obtained and used in further calculations are valid. The SMR will in turn influence the duration of the SDA. We defined the SDA as terminated when MO2 for each fish had returned to within 2 standard errors of the overall SMR estimate. In our case, it turned out that MO2 in some fish was still elevated above SMR after 23.5 h, which was the shortest time measured. In order to calculate the most comparable values of the area under the MO2 curve, i.e. the total SDA, we chose 23.5 h as the time span, except in fish where the SDA was finished earlier.