method, the apparent ileal digestibility
values of amino acids in the test protein
supplement (canola meal) were over-
estimated (P< 0.05 or 0.10) when a low-
protein feedstuff (barley) was the basal feed
ingredient and when the inclusion of the
test protein supplement (canola meal) was
low in the assay diet; however, this over-
estimation was eliminated through includ-
ing a high level of the protein supplement
(canola meal). Therefore, it is recom-
mended that a protein supplement rather
than a low-protein feedstuff (e.g. cereal
grains) is the basal feed ingredient when
the difference method is used. On the other
hand, in the case of some protein supple-
ments of poor palatability (e.g. feather meal
and blood meal), the ileal digestibility
values of amino acids should be deter-
mined by the difference rather than by the
direct method (e.g. Knabe et al., 1989;
Sauer et al., 1989). Because of the low level
of inclusion of these protein supplements
in the assay diets, the ileal digestibility
values of amino acids determined by the
difference method may not be reliable,
especially for the limiting amino acids. For
example, Knabe et al. (1989) reported that
the ileal digestibility values of isoleucine
calculated by the difference method for
three samples of blood meal were 60, 70
and 80%, respectively. As pointed out by
these authors, these values do not repre-
sent the variation in ileal digestibility
values of isoleucine and may simply be a
reflection of experimental error. Blood
meal provided only 15% of the isoleucine
content in the assay diets. Therefore, the
regression rather than the direct or differ-
ence method is only suitable for these pro-
tein supplements.
In conclusion, the regression method
is, in principle, the most accurate approach
for determining apparent ileal digestibility
values of amino acids in protein supple-
ments. In order to apply the regression
method successfully for all amino acids,
the basal and the assay feed ingredient
should be paired in such a way that differ-
ences in ileal digestibility values of amino
acids between the two feed ingredients are
large enough to create linear variations.
With respect to the difference method, it is
recommended that a protein supplement
rather than a low-protein feedstuff (e.g.
cereal grains) is the basal feed ingredient
and that the inclusion level of the test
protein supplement is high in the assay
diet. However, for some protein supple-
ments of poor palatability, the regression
rather than the direct or the difference
method should be used.Legume seeds
In further studies with peas (Fan and
Sauer, 1995b), five barrows, as described
previously, were fed five maize starch-
based diets according to a 5 5 Latin
square design. Diet 1 contained 885 g kg^1
wheat providing the sole source of amino
acids. In diets 2, 3 and 4, peas were
included at three levels at the expense of
maize starch, 168, 336 and 504 g kg^1 ,
respectively. Diet 5 contained 671 g kg^1
peas providing the sole source of amino
acids. All diets were formulated to contain
160 g kg^1 crude protein.
The apparent ileal digestibility values of
crude protein and amino acids in peas were
determined by the difference method
(Equation 13.7) at three inclusion levels
(diets 2, 3, and 4). There were no differences
(P> 0.05) in ileal digestibility values of
crude protein and amino acids in peas deter-
mined by the difference method between the
three inclusion levels. Although the ileal
digestibility values appeared to be higher at
the high (504 g kg^1 ) rather than at the low
inclusion level (168 g kg^1 ) for some amino
acids, the differences were not significant
(P> 0.05), for example, for histidine,
methionine and serine. The level of
inclusion of peas did not affect the
digestibility values of crude protein and
amino acids when these were determined by
the difference method from diets 2, 3, and 4.
Furthermore, these results also showed that
increases in trypsin inhibitor activity up to
2.32 TIU mg^1 of diet, resulting from
increasing the inclusion of peas in the diets,
did not affect the ileal digestibility values of
crude protein and amino acids when these
were determined by the difference method.
In addition, there were decreases in theMethods for Measuring Ileal Amino Acid Digestibility 299