Agriculture Reference
In-Depth Information
a flock is photostimulated between
10
and
20
weeks. A flock of broiler breeders with
typical feed restriction starts to mature
spontaneously under the influence of the
initial photoperiod from about
25
weeks.
The broiler breeder model is fully described
in Lewis
et al
. (2007).
These empirical models for commercial
laying pullets and for broiler breeders en-
able the prediction of AFE for individuals
making up a laying flock. Using appropriate
means and standard errors for each of the
parameters in the models it is possible to al-
locate randomly an AFE to each bird in the
simulated flock, which contributes to its po-
tential rate of laying, as will be described in
the next section.
Gous, 2003). Such a population of birds
would generate a range of ovulation times,
the distribution of which is unimodal and
positively skewed in young hens, becoming
bimodal with age. Reproductive senescence
in hens manifests as an increase in the intra-
sequence ovulation and oviposition inter-
vals with time, as well as an increase in the
number of pause days.
Different approaches have been used
to model the decline in rate of lay over
time. Most of these have been empirical in
nature (Gavora
et al
.,
1971; McNally, 1971;
McMillan
et al
.,
1986; Foster
et al
.,
1987;
Yang
et al
.,
1989; Koops and Grossman,
1992; Fialho and Ledur, 1997), all of which
are severely limiting when making use of a
mechanistic approach for describing the
decay in the rate of laying of an individual
hen over time. Emmans and Fisher (1986)
suggested that the hen's internal cycle
length increased with time from first egg,
resulting in a linear decline in the rates of
ovulation and oviposition with age. They
suggested that, at the start of the laying
period, some hens had the capacity to lay
at a rate greater than one egg in
24
h, but
that laying performance of these birds was
constrained by the external cycle length.
Eventually, the internal cycle length would
become longer than the external cycle
length, when ovulation rate would begin to
decline. However, there is evidence to
show that sequence length tends to rise ini-
tially (Lewis and Perry, 1991; Johnston,
2004), with most hens exhibiting a single
characteristically long (prime) sequence
about the time of peak egg production,
which then declines at different rates be-
tween individuals (Robinson
et al
.,
1990),
so the model of Emmans and Fisher (1986)
is unsatisfactory in describing the change
in ovulation rate over time.
In order to reproduce these changes in
sequence length over time, the internal cycle
length initially needs to be long (usually
more than
24
h), before decreasing with ad-
vancing time from first egg to close to, or
below
24
h, and subsequently increasing. In-
ternal cycle lengths longer than the external
cycle length will cause the time of lay to be
later each day, whereas those shorter than
Modelling Potential Egg Output
Describing the potential rate of lay of a lay-
ing hen is complex because of the number
of interacting factors involved, and the fact
that the potential varies over time within
each individual. The mathematical model
of Etches and Schoch (1984), based on the
theory of Fraps (1955), demonstrated that
two functions, representing two independ-
ent but interacting systems of the hen's
asynchronous ovulatory cycle, were able to
predict realistic ovulation times and intra-
sequence ovulation intervals. Johnston
and Gous (2003) extended this model by
defining a set of continuous functions, rep-
resenting the changes required to the val-
ues of the different parameters, such that
the prediction of any sequence length is
possible.
Mean rate of lay in a flock of hens at a
particular age is determined by the individ-
ual patterns of sequential laying at that
time. Within a population of birds, individ-
uals of the same age show considerable vari-
ation about a mean sequence length, which
may be due to variation in the length of the
open period for luteinizing hormone re-
lease, or variation in follicular dynamics.
This variation may be accounted for using
mean values and standard errors for each of
the parameters in the model (Johnston and
