Agriculture Reference
In-Depth Information
the models have low accuracy in predicting
feed intake (FI), and base their predictions
of FI on theoretical energy needs for an ex-
pected genetic potential for growth and
some environmental constraints to FI and
growth, or on regulation of thermogenesis
and thermolysis. Very few models include
effects of stocking density, diet presentation
and bird activity influenced by light or
space availability. The low accuracy of FI
predictions has resulted in issues with re-
gard to observations when predicting other
traits. This problem is mainly due to lack of
equations to estimate the effects of other
factors also known to affect FI. Very few
models include the interactive effects of
ambient temperature, relative humidity and
wind speed, which determine the effective
body temperature of birds, as well as minor
factors such as air quality, lighting or effects
of dietary calcium levels. The widely spread
utilization of phytases and other enzymes
has resulted in more variability in dietary
levels of calcium and phosphorus, and in
the levels of energy used. This generates more
unevenness in FI and animal responses re-
lated to the effects of other nutrients.
The growth models developed to date
for broilers and turkeys employ very similar
functions to describe growth. All models
basically have used the Gompertz (1825)
function to describe growth or development
of body components at the tissue (viscera,
carcass and feathers) or the chemical com-
ponents (ash, protein, lipid and water).
After these components are described, nu-
trients needed to either maintain or deposit
more energy or nitrogen are estimated using
efficiency factors for dietary nutrient util-
ization, body component allometry and ra-
tios among nutrients such as ideal protein
profiles to estimate all amino acids involved
with the animal's lysine needs. However, ly-
sine is not the main limiting amino acid for
poultry and other presumptions of the ideal
protein concept are not applicable to poultry.
Consequently, many ideal protein profiles
are observed worldwide. Growing birds re-
spond to increasing levels of crystalline
lysine, while other amino acids can remain
constant (Si
et al
., 2001). It would be useful
if biological models allowed users to modify
Advances in Modelling Applications for
Poultry Nutrition and Production
In poultry nutrition the most common ap-
proach has been to model growth or egg
production of a single bird and, on a very
few occasions, generate small populations.
Based on the mathematical description of
the biological system constituted by the ani-
mal biology, the feed composition and the
environmental factors involved and all their
interactions, such models are able to simu-
late diverse conditions, and estimate out-
puts of the system.
The modelling techniques employed
so far may vary from empirical to mechanis-
tic. Models that have been developed for
poultry are deterministic. Stochastic models
are not currently available for modelling the
biological responses of poultry. Unfortu-
nately, the intrinsic variability of data,
which is natural in all biological systems, is
oftentimes disregarded for both experimen-
tal and commercial conditions. The avail-
ability of such data is, however, extremely
important for detecting the problems and
inefficiencies in any system to be modelled.
New computer modelling advancements
could make it possible for such stochas-
ticity to be incorporated into more efficient
models that are able to repeatedly simulate
all the natural variability and detect the crit-
ical limits that cause some of the issues
often observed in commercial poultry pro-
duction.
From a couple of published reviews
that this author has been involved with
on models developed for poultry nutrition
(Oviedo-Rondón and Waldroup, 2002;
Oviedo- Rondón
et al
., 2002a, 2013), it can
be concluded that each modeller has taken
different approaches, but almost all models
have very similar outputs. The majority of
these models are able to determine body
weight (BW), BW gain, feed conversion
ratio (FCR), and energy, protein and amino
acid needs, or other biological responses,
such as tissue, chemical component accre-
tion or egg output of broilers, turkeys, layers
or breeders to different dietary nutrient
levels or environmental conditions. Most of
