One-factor ANOVA followed by Tukey’s test ( Zar, 1984 ) was
used to compare treatments at a significance level of 0.05. Normality and homoscedasticity were assessed by the Shapiro–Wilk
and Levene tests, respectively ( Zar, 1984 ). Data expressed as
a percentage underwent angular transformation before analysis.
Microbiologicl analysis data showed no homoscedasticity and
were Log 10 transformed. The analysis of changes in ammonia
over time was performed by repeatedmeasures ANOVA.
Treatments were considered to be the main factors, and duration
of culture was the additional factor. Significant differences were
analyzed by Tukey’s test ( Zar, 1984 ) with a significance level of
0.05.