Agriculture Reference
In-Depth Information
Table 21.4.
Ideal protein patterns based on literature.
Baker and Han
(1994)
Baker
et al
.
(2002)
Roth
et al
.
(2001)
Mack
et al
.
(1999)
Rostagno
et al
. (2011)
Amino acids
8
to 21 days
8
to 21 days
8
to 28 days
1
to 21 days
22
to 56 days
21
to 42 days
Lysine
100
100
100
100
100
100
Met+Cys
72-
75
a
nd
70
72
73
75
Trypto-
phan
16-17
17-19
14
17
18
19
Threonine
67-70
56-62
66
68
68
63
Arginine
105-108
nd
108
105
105
112
Valine
77-80
77-87
81
79
80
81
Isoleu-
cine
67
60-72
63
67
68
71
Leucine
109
nd
108
107
108
nd
Phe+Tyr
105
nd
121
115
115
nd
Gly+Ser
nd
b
nd
nd
150
137
nd
Histidine
32-35
nd
38
37
37
nd
a
Ranges are due to differences in the criterion optimized and in the model fitted data.
b
nd = not determined.
Additionally, in the whole grower period
the ideal ratios differed with the broken line
model, but using the Goettingen approach
the amino acid ratios in periods II and III
were similar (
P
= 0.287). Studies indicate
that the ideal dietary amino acid profile for
birds scarcely undergoes change in the whole
grower period (Baker and Han, 1994) but the
differences between periods I to II and III
(
22
to 53 days) indicate the opposite. This
difference is due to the total requirement
pattern depending on the relative contribution
of maintenance and growth requirements
(Zhen and Li, 1999) in which the mainten-
ance contribution is smaller in young birds
and increases with age.
In the present study, the N deposition
measured by the N balance technique was
overestimated compared to nitrogen depos-
ition estimated by the comparative slaughter
technique. This explains in part the small
difference in EAA profiles determined by the
two techniques, since the variation between
the two estimates can represent approxi-
mately 0.16 (Just
et al
., 1982). Compared
with N balance technique, the differential of
N in carcass analysis by the comparative
slaughter technique has an additional advan-
tage of not overestimating the N gain due to
potential unrecorded N losses (Heger and
Frydrych, 1985) but the N balance technique
is mostly preferable because of the associ-
ation with animal welfare.
The model proposed by Rollin
et al
.
(2003) relies on two key assumptions. First,
the equation assumes that the response to
EAAs is well described by the 'broken line'
regression approach. However, some authors
(Gahl
et al
., 1991, 1994) have shown a con-
tinuously diminishing approach to a max-
imum, whereas other researchers (Campbell
et al
., 1984, 1985; Dunkin
et al
., 1986) de-
scribe the responses of their animals as lin-
ear models. In broilers most of the reported
requirements have been estimated accord-
ing to the broken line model (Baker
et al
.,
2002). Curvilinear models have also been
proposed (Robbins
et al
., 1979; Rodehutscord
et al
., 1997), but the broken line approach
does not generally give a worse fit than the
non-linear models with regard to the stand-
ard deviation of the residuals (Rodehutscord
and Pack, 1999). In addition, inflection points
of best fit broken lines are objectively estab-
lished and predict minimal requirement
values, and this is viewed as desirable
for calculating EAA ratios (Baker, 2003).
Second, the equation assumes that EAAs
are utilized with similar efficiencies. Accord-
ing to Fuller (1994), the results of some
