2.4.2. Changes in fatality and fertility rates
For the fatality and fertility rates, the cause of change can be
either the reduction of rates through effects of environmental
change or disease outbreaks or the improvement of rates
through technology or better management. Concerning disease
outbreaks, diseases and disease control have always
been central elements of livestock production (Rorı´guex-
Sanchez et al., 2012). The manager is often faced with alternative
decisions (i.e., medication, vaccination, culling,
replacement) during an outbreak. Any decisions will affect
piglet production activity and determine the future productivity
of the herd and of the entire chain. In this particular
case, assume that the epidemic causes irregular preweaning
death. The farm manager can simulate the lowest and highest
death rates that might likely occur, and there could be a
disease outbreak at any time during pig production. Hence,
the manager may conveniently estimate the effect of the
outbreak in relation to time and production in the herd and in
subsequent herds. The shortage of pigs can be quantified and
prepared for by finding alternative sources of fatteners or by
altering replacement policies in the breeding herds to
accommodate such shortages.
To demonstrate the application of the model in the case of
a disease outbreak, we simulate an epidemic in the chain by
assuming that an epidemic occurred. Farm managers can
simulate the effect of an outbreak on the chain and on
fattening production using the proposed model. The occurrence
of an outbreak on any breeding farm as well as its
duration and severity can also be simulated to review their
effects on the herd. For this particular simulation, we
assumed that the manager was not certain concerning the
severity of preweaning death. With the developed system
dynamics model, the manager is equipped with a visualisation
tool and is able to foresee the effects of preweaning
death at any rate, on any farm, and in any occurrence timeframe.
Under this scenario, we demonstrate the simulation of
two preweaning death rates of 40% and 80%, compared with
the normal death rate of 7%. The outbreak was assumed to
last 5 weeks (starting from week 5 and ending during week 11).
Because each farm has a different disease control policy and is
in a different location, the outbreak might affect some of the farms, but not all of the farms. The manager can simulate its
effect on the breeding farms (GGP, GP, and parental) where the
outbreak might be likely to occur separately. This simulation
can be conveniently conducted by increasing weaner death to
specific levels and by specifying the occurrence time of the
shock accordingly. As shown in Fig. 8, for this particular
breeding pyramid, the numbers of sows at the GGP, GP, and
parental levels are quite high. Hence, the fattening units
appear to be guarded against unpredictable events, and a
simulated high preweaning death rate has minimal impacts
on the herd. Therefore, the effects of the irregular event will
appear only for the most extreme cases. For strategic decisions,
the model facilitates the incorporation of alternative
strategies in the model by the manager. The effects of medication,
vaccination, culling, and the replacement policy can be
included, although these effects are not presented in this
study.
In the case of improvement of these rates because of better
farm management or technological changes, examples of
striving to improve productivity include the following.
Improvement of the housing system was examined in a study
by Suriyasomboon, Lundeheim, Kunavaongkrit, and
Einarrsson (2006). These authors stated that the reproductive
efficiency of sows depends on several factors, such as their
parity and breed as well as the season, temperature, photoperiod,
and nutrition. A high ambient temperature, humidity,
and/or changes in the photoperiod were found to decrease the
litter size during some parts of the year because high ambient
temperature leads to heat stress and has been associated with
seasonal infertility. Hence, improving the housing system to
cope with heat stress is always taken into consideration. One
way to reduce heat is by constructing a floor cooling system.
As evaluated by Silva et al. (2006), floor cooling under a
lactating sow improves her productive and reproductive performance
as well as the weight gain of her litter.
Another example concerning genetic improvement is an
attempt to reduce wean-to-service days to increase the
number of piglets per sow and year, as described by Lundgren
et al. (2010), because the number of piglets produced per sow
and year is an economically important trait in pig production.
With the aim of increasing piglet production per sow and year,
some researchers have focused on breeding sows that produce
a large number of heavy piglets at weaning (Lundgren
et al., 2010), as piglet growth is regulated by both the genes
of the sow and the genes of the piglet. Heavy piglets are
considered to yield a shorter wean-to-service period, hence
increasing the number of piglets per sow per year.
Another example is in improvement of the feed system. As
stated by Kongsted (2005), the number of group-housed nonlactatingsowshas
been increased to complywith the European
legislation initiative arising from elevated public concern for
animal welfare. However, some studies have found that group
housing systems cause impaired reproduction, which is characterised
by a reduced litter size and pregnancy rates in grouphoused
sows compared with individually housed sows during
the non-lactating period. This impairment has been in the
range of 0.3e0.6 fewer piglets born per litter, a 0.9% lower farrowing
rate and a 3.4% higher repeat breeding rate in some
periods. In group-feeding systems, individual rationing is not
possible, and variation in energy intake may occur because an
individual sow is not protected during feeding from displacements
by other sows. In Kongsted's report, the use of a new feeding system with a high-energy intake was suggested to
improve the pregnancy rate and litter size.
To demonstrate the application of setting a scenario for
management changes, let us assume that the farm manager
engages in three projects. The first project is improvement of
the housing system, with the successful implementation of
floor cooling, resulting in a reduction of the number of weaner
deaths by 2%. The second project is a genetic improvement
project, successfully leading to a 7-day shorter wean-toservice
breed time. The last project assumes that the farm
commits to improve the feed system in group-housed nonlactating
sows and is successful in implementing the new
feeding system, which in turn results in an average of 0.5 more
piglets born per sow and a 0.9% increase in the farrowing rate.
With the studied model, the manager can simulate the effects
of these improvements as follows. The first project, installing
a cooling floor, was assumed to reduce weaner death by 2%.
This case can be simulated by modifying the weaner death
rate similar to the first scenario, which will directly reduce
PWDi in Equation (10). PWi will rise as a result. The second
project, genetic improvement, was assumed to decrease the
average wean-to-service age by 7 days. This situation can be
simulated by reducing the transit time of the conveyor boxes
in the replacement stage by 7 days. The last project, the
establishment of a new feeding system, was assumed to result
in an average of 0.5 more piglets born per litter and a 0.9%
increase in the farrowing rate. This simulated case can be
achieved by increasing ABAi in Equation (9) by 0.5 and by
reducing either one or a combination of two or three of the
following improvements during the gestation stage: FCSi, Ai,
and NIPi. The effects of these improvements on the herds are
shown in Fig. 9. Because of the fixed target policy regarding
replacement and culling, although the performance of the
breeding farm was improved, the number of sows on the farm
did not change. The improvement is likely to yield approximately
100 more piglets per week. Of note, these simulation
results are for this particular dataset. The simulation results
are not applicable to any other farms because this simulation
uses farm-specific data and situations. Although the total
number of sows, death rate, culling rate, conception rate, and
other farm policies are identical for any two farms, the
simulation results may differ depending on the number of
sows in each stage and parity. Hence, the results should be
interpreted with caution because different farms are likely to
experience different results.