Twenty one barrows averaging 19.63±0.37 kg BW were
used in a randomized complete block design experiment,
with seven replications per treatment and one animal per experimental
unit (metabolism cage). The pigs were housed
randomly and individually in stainless steel metabolism
cages, which allowed for the separate collection of urine
and feces, as described by Pekas (1968).
The experimental period consisted of 7 days adaptation
period, followed by 5 days collection period of feces and
urine, using protocols described by Sakomura and Rostagno
(2007). The treatments included a reference diet (Table 1)
that was formulated and based on the nutrient requirement
table of weanling pigs (Rostagno et al., 2005) and two test
diets (RO and SO). The test diets had a composition of 90% reference
diet (as-fed basis) and 10% RO or SO. During the collection
period, the daily feed intake wasmaintained constant, and
the amount of feed provided for each pig was based on a lower
feed intake per kg of metabolic BW (BW0.75) observed during
the adaptation period in each block. The diets were provided
twice daily, and water was provided ad libitum during the entire
experimental period. To determine the beginning and end
of a collection period, ferric oxide (Fe2O3) at a dietary level of
2% was used as a fecal marker.