Statistical analysis was carried out using SPSS. A
Kolmogorov–Smirnov test was applied to the water chemistry
data. The data was not found to conform to normality
therefore non-parametric analysis was used to compare different
conditions. As a measure of the performance of each of the
conditions, Mann–Whitney U tests (for independent samples)
were used to determine whether the outlet concentration in the
flood/drain condition was significantly different to the submerged
condition. Since this experiment involved treatments carried out
in sequence, statistical comparisons were only made between the
flood/drain and the submerged within each treatment but not
between treatments. To identify whether the outlet concentration
was significantly different to inlet concentration, Wilcoxon
signed-rank tests (for paired samples) were performed. Pearson’s
correlation coefficient (r) was calculated as a measure of the effect
size (Cohen, 1992).