parameters measured, only NH3 concentrations were significantly
different between low and high oxygen tanks. The
highest difference in NH3 concentrations among the treatments
(0.007 mg L−1) was very small and the highest NH3 concentration
(0.011 mg L−1) was far below the lowest level
(0.080 mg L−1) which may affect feed intake of Nile tilapia
(Shelton and Popma, 2006). El-Shafai et al. (2004) showed that
feed intake of Nile tilapia was not influenced by NH3–N
concentrations within the range of 0.004–0.434 mg L−1
(corresponding to 0.005–0.527 mg NH3 L−1); however, for
best growth, they recommended an NH3–N concentration
below 0.10 mg L−1 (corresponding to 0.12 mg NH3 L−1). Muir
et al. (2000) even considered an NH3 level below 0.2 mg L−1
ideal for flowing-water tilapia culture. However, Soderberg
(2006) doubted that the chronic toxicity level of NH3 might be
lower than the level recommended by Muir et al (2000). Based
on studies on channel catfish, rainbow trout (Oncorhynchus
mykiss) and blue tilapia (Oreochromis aurea), he estimated
chronic toxicity levels of NH3 ranging from 0.035 to 0.092 mg
L−1 for tilapias. The lower bound of this range is higher than the
highest NH3 concentration in the present study.