Agriculture Reference
In-Depth Information
with the exception of the AID Lys, which
was reduced in concentration by 60%.
Dietary phosphorus (P) and calcium (Ca)
requirements were estimated according to
Jondreville and Dourmad (2005). Microbial
phytase (500 FTU/kg,
Aspergilus Niger
,
Natuphos) was added to all feeds in which
the Ca:P ratio was maintained constant.
Feeds were steam-pelleted at 4 mm. The
four feeds were blended daily for each pig
to constitute the experimental diets.
Sixty pigs were assigned to treatments at
25 kg of BW (group
25-
55) and
60
others at
70 kg of BW (group
70-
100). Pigs were
housed in groups of
60,
but fed individually
using automatic and intelligent precision
feeders (AIPF) specially developed for this
project (Pomar
et al
., 2009a, 2011) and able
to provide to each pig at each visit the esti-
mated mix of the four experimental diets.
These AIPFs consisted of a single space
feeder in which precision Archimedes' screw
conveyors delivered and blended simultan-
eously volumetric amounts of up to four
diets contained in independent feed contain-
ers. To this end, the AIPF identified each pig
when their head was introduced into the
feeder and the feeds were blended and de-
livered upon the animal's request according
to the estimated optimal Lys concentration
and assigned experimental treatment. A serv-
ing is composed of the amount of feed de-
livered upon each effective serving request.
A time lag was imposed to ensure that pigs
ate each serving before requesting a new one.
Serving size was progressively increased
during the experiment and ranged between
15
and
25
g. A meal includes all the servings
delivered during each feeder visit. Pigs tend
to leave the feeder hopper empty or leave
very small amounts of feed after each visit,
thus ensuring that each pig received the as-
signed amount of blended feed. Feed density
was measured weekly and this information
used to convert feed volumes to feed weights.
Pigs had free access to the AIPFs and water
throughout the experiment. Feeding phases
lasted for 28 days for a total experimental
length of 84 days. Pigs were weighed weekly
in both trials and their body composition
measured at the beginning of each feeding
phase and at the end of the trials by dual
X-ray densitometry (DXA, GE Lunar Prodigy
Advance, GE Healthcare).
In the
25-
55 kg of BW pig group, the
level of Lys did not affect average DFI
(ADFI), but in the
70-100
group, ADFI
showed a quadratic response (Lin:
P
= 0.04,
Quad:
P
= 0.04) with maximal values ob-
served in pigs fed 10% below requirements.
The ADG and average PD (APD) increased
linearly (
P
< 0.01) with the level of dietary
Lys. In the first group, however, ADG and
APD increased linearly (
P
< 0.01) although
maximal ADG (1.00 kg/day) and APD (
174
g/day) were observed in animals fed accord-
ing to their requirements (100%). Similar
linear effects (Lin:
P
< 0.01) were observed
for the second group with maximal ADG
(1.19 kg/day) and APD (
185
g/day) reached
with diets providing 110% of estimated Lys
requirements. The factorial method used in
this study to estimate the dynamic Lys re-
quirements of individual pigs seemed appro-
priate for pigs fed between
25
to 55 kg of
BW, but it appeared to underestimate the
requirements of heavier pigs. The detailed
results of this trial have been described else-
where (Zhang
et al
., 2011, 2012).
Second calibration trial
The objective of the second trial was to val-
idate the mechanistic component of the
mathematical model estimating in real
time the individual Lys requirements after
adjustment based on the previous results.
For this purpose,
75
castrated pigs corres-
ponding to a terminal genetic line and
72
castrated pigs corresponding to a maternal
genetic line were used for this study. This
trial consisted of two
28-
day experimental
phases, one beginning at 25.8 ± 2.5 kg live
weight and the other at 73.3 ± 5.2 kg live
weight. The treatments were randomly as-
signed to the pigs according to a
2
×
4
factor-
ial design with the two genetic lines and the
four Lys levels (70%, 85%, 100% and 115%
of requirements) as the factors. Feeds were
prepared, the animals were cared for and
measurements were taken as indicated for
the previous trial. From
25
to 50 kg of BW,
the amino acid estimation method slightly
