Statistical analysis of biological data from preoperational-postoperational industrial water quality monitoring

An application of methods for analysis of biological data from preoperational-postoperational industrial surveys is presented. Repeated observations at the same sampling station form a time series in which the observations are not statistically independent, and the usual forms of statistical analysis do not apply. If the data do not indicate a trend, control charts can be applied to analyze this form of data, and control charts are easy to use. A careful analysis of biological monitoring data requires a well designed sampling scheme and removal of seasonal fluctuations from the data. It is shown that stream monitoring can detect only relatively large changes that occur in aquatic populations.