triplicate tanks and were assigned randomly. Each tank was filled with
dechlorinated freshwater. During the experimental period, one third
of the water was exchanged daily in the control tanks. Water was
not exchanged in the BFT group, but evaporation losses were compen-
sated with dechlorinated freshwater. In order to well determine the
variation of the biofloc and acquire optimal experimental conditions,
a pre-experiment was conducted with a control group and two BFT
treatments (C/N ratio of 15 and 20, respectively), with three replicate
tanks for each group. The design and arrangement used here were
similar to this trial. Due to the more stable condition of the BFT group
(C/N ratio of 15), we selected it as the experimental group after the
pre-experiment. Before the fish were stocked, the BFT group was
inoculated with concentrated biofloc (250 mL per tank) collected
from an indoor biofloc tank (C/N ratio of 15) during the pre-
experiment. The tank water was aerated and agitated continuously
using air-stones connected to an air pump. The diameter of the aeration
tube was 6 mm, and that of the air-stone was 20 mm. The laboratory
where the trialwas conducted had large glasswindows and the sunlight
was very strong. Consequently, all tanks were continuously covered
with shade cloth to block out the sunlight. In addition, the shade cloth
prevented the fish from jumping out of the tanks.