As expected, the mean of the index scores for the calibration data set did not differ from the expected value of 0.5 (0.505, p = 0.416) and was not significantly different than that of the validation data set (0.504) (t = 0.067, p = 0.947), whereas the mean of the index scores for the disturbed data set (0.339) was significantly lower than that of the calibration (t = −13.707, p < 0.0001) and validation (t = −10.157, p < 0.0001) data sets, indicating clear discrimination between reference and impaired sites (Fig. 2)