Agriculture Reference
In-Depth Information
observed values. To avoid this bias the re-
search studies may be replicated to validate
the applicability of alternatives for different
locations, breeds and genders. As a result of
these studies a meta-analysis may be per-
formed to compare the results and validate
Research studies should be performed
along with model simulations until the model
has been fully evaluated. Over time, fewer
validation trials will be required as model
simulations will progressively substitute for
research studies, while the model's advan-
tages and limitations will be more clearly
understood and precision and accuracy
improved.
Implementation
Once the simulated alternative has been se-
lected and validated it can be commercially
implemented. A reasonable strategy would
220
200
180
160
140
120
92%
100%
AME level (% of standard recommendations)
108%
D49-sim
D49-obs
D105-sim
D105-obs
Fig. 8.5.
Simulated and observed average daily gains of male turkeys subjected to different metabolizable energy
contents (AME, % standard recommendations).
110
100
90
Study
1,
males
Study
1,
females
Study
2,
males
Study
2,
females
80
70
85.00%
90.00%
95.00%
100.00%
105.00%
110.00%
Amino acid level (% standard recommendations)
Fig. 8.6.
Meta-analysis of the effect of amino acid levels on average daily gain in male and female
growing turkeys (i.e. two studies × two genders).
