BREEDING AND GENETICS

Comb Size and Estrogen Levels Toward the Onset of Lay in Broiler and
Layer Strain Females Under Ad Libitum and Restricted Feeding
YOAV EITAN,* MORRIS SOLLER,*,1 and ISRAEL ROZENBOIM†

*Department of Genetics, The Silberman Life Sciences Institute, The Hebrew University of Jerusalem, 91904 Jerusalem,
Israel, and †Department of Animal Sciences, The Faculty of Agriculture, Rehovot, Israel

ABSTRACT The time course of comb development
and estrogen levels were compared in broiler (BX) and
layer (LX) females that consumed feed ad libitum (A) or
were subjected to quantitative feed restriction (R). The
chicks were reared under short photoperiod [(6 h light
(L):18 h dark (D)] until 22 wk of age. At this time,
photoperiod was increased in one step to 8L:16 D, and
then gradually increased until 14L:10D at 34 wk. There
was a significant interaction between genetic type and
feeding treatment, such that entry into lay of the LX-R
and BX-R females was delayed by 1 and 4 wk,
respectively, relative to the LX-A and BX-A birds. Mean
comb size of LX-A birds began to increase while still
(Key words: comb size, estrogen, onset of lay, feed restriction, photoperiod)
under 6L:18D photoperiod; that of the other treatment
groups did not begin to increase until photoperiod was
shifted to 8L:16D. Comb size of individual LX-A, LX-R,
and BX-A birds began to increase about 8 wk prior to
individual onset of lay; that of BX-R birds about 11 wk
prior to onset of lay. In all groups, estrogen levels
remained low until 3 to 4 wk prior to onset of lay, when
they showed a sharp increase. Following onset of lay,
estrogen levels of all groups remained high. Critical day
length of LX-A birds appears to be lower than that of
BX-A birds. In addition, feed restriction per se appears to
decrease photoperiod responsiveness of BX birds.
1998 Poultry Science 77:1593–1600

INTRODUCTION this window, an increase in photoperiod to some

In previous studies, it was found that decreasing
photoperiods had a strong delaying effect on onset of
lay in broilers, but relatively little effect in layers (Eitan
and Soller, 1991, 1994, 1996). The greater sensitivity to
suboptimal photoperiod of broilers was interpreted to
imply that broilers respond in a less powerful manner to
photostimulation than layers, so that a suboptimal
photoperiod resulted in a longer period from onset of
photosensitivity to onset of lay. It is of interest to
determine whether broilers and layers differ in other
aspects of their responses to photostimulation. In
particular, once minimum BW and age thresholds are
attained, entry into lay ensues promptly in females
provided with an adequate photoperiod. Entry into lay
is deferred, but ensues eventually, in birds kept in total
darkness or short day length (Wilson and Woodward,
1958; King, 1961). Thus, a photostimulation “window”
can be defined, bounded by the earliest age at which
females enter lay under adequate photostimulation, and
the age they enter lay under short day length. Within
minimum length, termed the “critical day length”
(CDL), stimulates earlier onset of sexual maturity;
whereas increase in photoperiod beyond some maximal
length, termed the “saturation day length” (SDL), does
not result in further decrease in age at onset of sexual
maturity (Dunn and Sharp, 1990). Layer and broiler
females have been reported to differ in CDL and SDL
(Dunn and Sharp, 1990); and from data presented in
Robinson (1994) there appears to be an effect of feed
restriction on response to photoperiod of broilers.
The purpose of the present study was to compare the
effect of short photoperiod on entry into sexual maturity
and lay in layer and broiler females, under ad libitum
and restricted feeding. The approach was to compare
the time course of comb size development and estrogen
levels in broiler and layer females under gradually
increasing photoperiod; comb size is commonly used as
a surrogate measure for development of reproductive
function in chickens (Lowry, 1958; Snapir et al., 1969;
Rozenboim et al., 1986, 1993; Jacoby et al., 1992). The
basic assumption was that under this treatment, differ-

Received for publication October 24, 1997. Abbreviation Key: A = ad libitum feeding; BX = broiler cross; CDL =
Accepted for publication June 24, 1998. critical day length; D = h of darkness; L = h of light; LH = luteinizing
1To whom correspondence should be addressed: soller@vms.huji. hormone; LX = layer cross; R = restricted feeding; SDL = saturation day
ac.il length.

1593
1594 EITAN ET AL.
ences in CDL, for genetic or nutritional reasons, would
be expressed as a differences in the time of initiation of
the increase in comb size and estrogen levels normally
observed prior to onset of lay.
MATERIALS AND METHODS
Stocks and Rearing Procedures
The experiment included 50 chicks each of two
commercial lines: Arbor Acres feather-sexed broiler
breeders, denoted BX (Kvutzat Yavne Hatchery, Israel),
and Lohman layers, denoted LX (Poultry Breeders Union,
Netanya, Israel). The chicks were hatched on June 7, 1995,
and vaccinated the same day against Marek’s disease
using the Rispens vaccine, and against infectious bronchi-
tis and Newcastle disease virus using live vaccine. The
chicks were reared at the North Talpiot Experimental
Farm of the Hebrew University in Jerusalem. At 14 d of
age, the birds were vaccinated against infectious bursal
disease and Newcastle disease virus using an inactivated
vaccine.
All chicks were brooded for 21 d in electrically heated
batteries. Following the brooding period, the layers were
beak trimmed to avoid pecking, and all chicks were
transferred to floor pens until 41 d. Broiler chicks were not
beak trimmed, as picking is not a problem with broiler
chicks under our rearing conditions. Beak trimming is not
known to affect onset of sexual maturity. At 41 d, all chicks
were transferred to individual cages with individual
feeding troughs, in a controlled-light dark room. A
proprietary starter mash (210 g CP and 3,150 cal ME/g)
was fed until 14 d; a broiler mash (170 g CP and 3,250 cal
ME/g) until 42 d; a grower mash (140 g CP and 2,850 cal
ME/g) until 157 d; and a layer mash (150 g CP and 3,100
cal ME/g), thereafter.
Chicks were reared under naturally increasing daylight
until 41 d; under 22 h light (L):2 h dark (D) from 42 to 55 d,
using incandescent light (35 to 50 lx, light intensity varied
between upper and lower cage tiers); and under 6L:18D
from 56 d to 157 d (5 to 10 lx). Photoperiod was increased
to 8L:16D at 157 d, and thereafter increased by 0.5 h
weekly, reaching 14L:10D at 34 wk. Simultaneously, light
intensity was increased, from 7 to 22 lx at 22 wk, to 30 to 50
lx at 28 wk. Previous work has shown that differences in
response to photoperiod between genetic types are
magnified under conditions of low photostimulation
(Eitan and Soller, 1991, 1994, 1996). The gradual increase
in light intensity, starting from a low level, was intended
to provide minimal photostimulation consistent with
achieving CDL. In was anticipated that this would
provide full expression of differences in photoperiod
responses of the various genetic types and feeding
treatments.
2Cayman Chemical Co., Ann Arbor, MI 48108.
Experimental Treatments
Chicks were allowed to consume feed ad libitum until 56
d. The chicks of each genetic type were then divided into
two groups: BX-A and LX-A, which continued with ad
libitum feeding; and BX-R and LX-R, which were changed
to quantitative feed restriction. The feeding schedule of
the BX-R and LX-R birds was adjusted at intervals to
maintain approximately linear growth to 25 wk. Thus, the
feeding schedule of the BX-R birds was as follows: 50 g/d
from 56 to 67 d; 60 g/d from 68 to 114 d; 80 g/d from 115 to
157 d; 110 g/d from 158 to 164 d; 120 g/d from 165 to 171
d; 130 g/d from 172 to 178 d; 140 g/d from 179 d to the end
of the experiment. The feeding schedule of the LX-R birds
was as follows: 30 g/d from 56 to 67 d; 40 g/d from 68 to
157 d; 80 g/d from 158 to 171 d; 90 g/d from 172 d to the
end of the experiment (equivalent to ad libitum).
Data Collection
Chicks were weighed every 2 wk. Weighing was by a
digital scale accurate to 1 g until 41 d, and by a spring scale
accurate to 20 g thereafter.
Comb height and width were measured by a straight
edge ruler, scored at 1-mm intervals. Height was taken as
the distance from the highest point on the second blade
from the rear to the base of the skull. Width was taken
across the entire width of the comb from the attachment of
the front blade to the base of the skull, to the rearmost
curve of the rear blade. Comb measurements were taken
at approximately 2-wk intervals from 14 wk to the end of
the experiment.
Heparinized blood samples were drawn at 2-wk
intervals from 6 to 22 wk (from 14 wk for LX-R birds),
except for Weeks 12 and 16; and at weekly intervals from
23 to 36 wk, except for Weeks 28 and 36. Blood was taken
between 0800 and 1200 h (noon). Plasma was stored at –20
C until assayed for estrogen level. The estrogen measured
was 17b-estradiol. All plasma estrogen levels were
measured on the same day, using an Eliza Immuno Assay
Kit.2 The sensitivity of the assay was 17 pg/mL at 80%
binding. The inter- and intra-assay coefficients of varia-
tion were 7 and 5%, respectively.
Statistical Analyses
Following Lowry (1958), comb size was taken as the
product of comb height and comb width, omitting
division by 2, as this would not change any of the statistics.
Data were analyzed by the SAS GLM procedure (SAS
Institute, 1985), according to the following model
Yijk = m + bi + tj + btij + eijk
where Yijk is the variable being analyzed of the kth
individual of the ith genetic stock and jth feeding
treatment; m is the overall mean; bi is the effect of the ith
genetic stock (i = 1 or 2); tj is the effect of the jth feeding
 COMB SIZE AND ESTROGEN TOWARD ONSET OF LAY 1595
TABLE 1. Mean, CV, and correlation (r) of age and BW at first
egg according to genetic type and feeding treatment1

Age BW
Group Number x CV x CV r
(d) (%) (g) (%)
BX-A 19 197.0b 9.7a 4,781.6a 11.8a 0.90a
BX-R 19 225.4a 3.0b 4,416.8b 7.1b 0.74b
LX-A 18 175.8c 8.4a 1,492.2c 13.0a 0.80a,b
LX-R 18 182.8c 9.0a 1,472.2c 12.8a 0.72b
a–cValues in the same column with no common superscript differ significantly by t test (P < 0.05).
1BX = broiler cross; LX = layer cross; A = ad libitum feeding; R = restricted feeding.

treatment (j = 1 or 2); btij is the interaction between genetic
stock and feeding treatment; and eijk is the residual error
term. Statistical significance of differences between
specific lines and treatments were determined by t test
(Walpole and Myers, 1978). The CV was calculated as SD/
x. Statistical significance of differences between two CV
were tested by calculating (CV1/CV2)2, and comparing to
F(n1, n2) where n1 and n2 are df of larger and small CV,
respectively. Statistical significance of correlation coeffi-
cients was tested by z transformation (Walpole and
Myers, 1978).
In analyzing the time course of BW, comb size, and
estrogen levels, the data for each measurement date were
first plotted against the “chronological age” of the birds at
that date (all birds were hatched on the same date).
Because at a given chronological age some birds may be in
lay, whereas others have not yet entered lay, changes in
mean trait values of an experimental group with age
primarily reflect the changing proportions of birds that
are in lay or not in lay in the various groups. In order to
account for this, a second approach was used in which the
data were plotted against “physiological age”. In this case,
the date at which each bird individually entered lay was
taken as the origin (zero) point for that bird. Measurement
dates for each bird were then assigned positive (+) or
negative (–) “physiological age” values, measured in
number of weeks preceding or following entry into lay.
For each variable, the mean value for the birds of each
genetic type and treatment at each physiological age, was
calculated and plotted against physiological age. As
compared to chronological age, physiological age in the
present experiment has two weaknesses: 1) Data points
were taken at somewhat irregular intervals; consequently,
not all birds are represented at each physiological age. For
example, with BW measurements taken at 2-wk intervals,
only half the birds will have BW measurements at any
given physiological age. This procedure introduces a
degree of experimental sampling in moving from point to
point along the physiological age curves, which does not
occur when using chronological age, for which all birds
are included in all points. 2) Photoperiod of all birds was
increased systematically with chronological age. When
compared at the same chronological age, all birds in a
given genetic by treatment group are at the same
photoperiod. However, at a given physiological age, the
birds can be at very different photoperiods, depending on
the effects of their genetic type and feeding treatment on
age at first egg.
Estrogen levels fluctuated rather widely in consecutive
weeks, producing a somewhat confusing picture. To
moderate these fluctuations, a 2-wk moving average was
calculated, and plotted against mean chronological or
physiological age of the assay days included in each
average. In order to test for statistical significance of
differences in estrogen levels between treatments and
ages, measurements made in the same general period (6 to
10, 14 to 20, 22 to 24, and 33 to 36 wk) were pooled, and
tested for differences by t test (Walpole and Myers, 1978),
assuming different SD in each cell.
RESULTS
Table 1 shows age and weight at first egg for the
various genetic and treatment groups. Body weight and
age of the BX birds at first egg were greater than those
of the LX birds, irrespective of treatment. Body weight
and age of the LX-A and LX-R birds at first egg did not
differ significantly; BW of the BX-A birds was signifi-
cantly greater and age was significantly less than for the
BX-R birds. The interaction effects of genetic type by
feed treatment on both age and BW at first egg were
highly significant by ANOVA (P < 0.01). This result is
due to the 4-wk delay in onset of lay of the BX-R birds
as compared to the BX-A birds.
The CV for age and BW at first egg was similar in BX-
A, LX-A and LX-R, but significantly lower for BX-R
(Table 1). Thus, the BX-R birds came into lay more
synchronously than the other groups, both with respect
to age and with respect to BW. The correlation between
age and BW at first egg was high in all groups.
Figure 1A shows BW plotted against chronological
age. From 16 wk until the end of the experiment, BW of
BX-A exceeded that of BX-R by about 1,000 g (from 50 to
30%); mean BW of both groups continued to increase
until the end of the experiment. From 16 to 20 wk, BW
of LX-A exceeded that of LX-R by about 200 g (20%).
The difference diminished to 50 g (5%) by 22 wk, and
disappeared by 30 wk. Mean age at first egg of the
various genetic by treatment groups is indicated by
arrows. Figures 1B and 1C show BW plotted against
physiological age for BX and LX birds, respectively. In
all groups there was a steady linear increase in BW from
1596 EITAN ET AL.
about 15 wk prior to first egg. With the onset of lay,
increases in BW immediately ceased. The very different
time course of BW on a physiological basis as compared
to a chronological basis clearly illustrates the difference
between the two modes of analysis.
Figure 2A shows comb size plotted against chrono-
logical age. Within treatment groups, comb size of LX
birds was twice that of the BX birds throughout the
experimental period. Comb size of the LX-R, BX-A, and
BX-R birds remained low and virtually constant until 22
wk, when photoperiod was increased to 8L:16D. In
contrast, comb size of the LX-A birds showed a steady
FIGURE 1. Body weight plotted against age. A) BW plotted against
chronological age. B) BW plotted against physiological age for BX-A
and BX-R birds. C) BW plotted against physiological age for LX-A and
LX-R birds. ∫ = BX-A; π = BX-R; o = LX-A; ÿ = LX-R. BX = broiler
cross; LX = layer cross; A = ad libitum feeding; R = restricted feeding.
Arrows show mean age of birds at onset of lay.
increase from 16 to 22 wk, at which time the birds were
still under 6L:18D photoperiod. During this period,
comb size of BX-A and LX-R birds was the same, but
twice that of BX-R birds. From 22 wk, comb size of LX-R
birds increased rapidly, achieving equality with the LX-
A birds by 26 wk. Similarly, comb size of the BX-A and
BX-R birds also began to increase from 22 wk. Until 25
wk, comb size of BX-A birds increased more rapidly
than that of the BX-R birds. From 25 to 31 wk, the
increase was parallel. From 31 wk comb size of the BX-A
birds plateaued, whereas that of the BX-R birds
continued to increase, although even at the very end of
the experiment the gap between the two groups was still
present.
Half of the total increase in comb size was attained by
24 wk for LX birds, and by 26 and 28 wk for the BX-A
and BX-R birds, respectively (Figure 2). At this time, the
correlation coefficient of comb size with age at onset of
lay was –0.65 and –0.31 for BX-A and BX-R, respectively;
and –0.76 and –0.75 for LX-A and LX-R, respectively.
The negative sign of the correlation coefficients indicates
that during the period of rapid comb development comb
size was a measure of relative sexual development, in
that the birds with the larger combs subsequently enter
lay earlier than those with smaller combs. The correla-
tion coefficients for LX-A, LX-R, and BX-A birds did not
differ significantly; that for the BX-R birds was signifi-
cantly less than for the other groups. This reflects the
more synchronous entry into lay of the BX-R group, as
noted above. At the attainment of half final comb size,
the correlations of comb size with BW at first egg were
less than with age (–0.61, –0.43, –0.48, and –0.22 for LX-
A, LX-R, BX-A, and BX-R, respectively), indicating that
the primary correlation was with age, rather than BW.
Again, the correlation for the BX-R birds was signifi-
cantly less than for the other groups.
Figures 2B and 2C show comb size plotted against
physiological age. The LX-A and LX-R birds show a
plateau from –13 to –9 wk; the BX-A birds show a
plateau from –15 to –9 wk. All three types show first
signs of an increase in comb size about 8 wk prior to
onset of lay. In contrast, BX-R birds appear to have
initiated an increase in comb size from 11 wk prior to
onset of lay. When plotted against physiological age,
comb size of LX-A and LX-R birds did not differ; but
even on this basis, comb size of the BX-A birds was
greater than that of the BX-R birds, throughout the
period preceding lay.
Figures 3A and 3B show the 2-wk moving average of
estrogen levels plotted against chronological age for BX
and LX birds, respectively. Estrogen levels were gener-
ally low from 6 wk through 18 to 20 wk. During this
period, estrogen levels of the BX-A and LX-A birds were
significantly higher than that of BX-R and LX-R groups
(Table 2). The BX-A and LX-A birds showed a significant
increase from the 6 to 10 wk period to the 14 to 20 wk
period; but BX-R birds remained unchanged. The LX-A
birds showed a particularly strong increase, so that
 COMB SIZE AND ESTROGEN TOWARD ONSET OF LAY 1597
although levels of the LX-A birds were significantly
below those of the BX-A group in the 6 to 10 wk period;
they were significantly greater in the 14 to 20 wk period.
Following the increase in photoperiod at 22 wk,
estrogen levels during the 22 to 24 wk period, increased
significantly for BX-A and LX-R birds, but not signifi-
cantly for LX-A; and remained virtually unchanged for
BX-R birds (Table 2). At this time, BX-A and LX-A had
significantly higher estrogen levels than BX-R and LX-R;
but BX-A and LX-A did not differ significantly, whereas
LX-R was significantly greater than BX-R. All groups
show a significant increase in estrogen levels in the
period following onset of lay (33 to 36 wk) as compared

FIGURE 3. Estrogen levels plotted against age. A) two-week

moving average of estrogen levels plotted against chronological age
for BX-A and BX-R birds. B) two-week moving average of estrogen
levels plotted against chronological age for LX-A and LX-R birds. C)
FIGURE 2. Comb size plotted against age. A) comb size plotted
two-week moving average of estrogen levels plotted against physiolog-
against chronological age. B) comb size plotted against physiological
ical age for BX-A and BX-R birds. D) two-week moving average of
age for BX-A and BX-R birds. C) comb size plotted against
estrogen levels plotted against physiological age for LX-A and LX-R
physiological age for LX-A and LX-R birds. ∫ = BX-A; π = BX-R; o =
birds. ∫ = BX-A; π = BX-R; o = LX-A; ÿ = LX-R. BX = broiler cross; LX
LX-A; ÿ = LX-R. BX = broiler cross; LX = layer cross; A = ad libitum
= layer cross; A = ad libitum feeding; R = restricted feeding. Arrows
feeding; R = restricted feeding. Arrows show mean age of birds at
show mean age of birds at onset of lay.
onset of lay.
1598 EITAN ET AL.
TABLE 2. Mean and SD (in parentheses) of estrogen levels by compared to 49, 34, and 37 d for BX-A, LX-A, and LX-R,
age according to genetic type and feeding treatment1,2
respectively.
Age Figure 4B shows rate of lay plotted against physiolog-
ical age. In this case, the increase represents an increase
Group 6 to 10 wk 14 to 20 wk 22 to 24 wk 33 to 36 wk
in the rate of lay of individual birds from week to week
(pg/mL) (Bornstein and Lev, 1982). All groups show very similar
BX-A 161.0a,z 220.4b,y 354.9a,x 629.8a,w behavior in the 1st mo of lay, following which BX-A
(59.6) (58.1) (166.7) (125.0)
leveled off at 4 eggs per week, whereas the LX birds
BX-R 141.2a,z 145.4d,z 153.1c,z 542.4a,y
(19.0) (16.5) (24.8) (105.8) continued to increase to 6 eggs per week at 10 wk.
LX-A 122.6b,z 361.6a,y 446.5a,y 577.2a,x
(6.6) (134.3) (137.8) (99.8)
DISCUSSION
LX-R ND 177.3c,z 246.5b,y 677.5a,x
ND (29.1) (52.0) (27.0) The results of the present study showed an effect of
a–dValues in the same column with no common superscript differ genetic type and feeding treatment on BW and age at
significantly by t test (P < 0.05). first egg, and on the time course of comb size and
w–zValues in the same row with no common superscript differ
estrogen levels prior to onset of lay. These effects will be
significantly by t test (P < 0.05).
1Number of measurements per cell, 13 to 15; all t tests on the basis considered in relation to threshold BW and age
of 14 measurements per cell. requirements for responsiveness to photostimulation,
2BX = broiler cross; LX = layer cross; A = ad libitum feeding; R = CDL, SDL, and photoperiodic responsiveness. The
restricted feeding. salient observations, and their interpretations, are as
3ND = not done.
follows:

Age and BW at First Egg

to the 22 to 24 wk period; but did not differ among Onset of lay in LX-A birds preceded that of the BX-A
themselves at this time. birds and onset of lay in LX-R birds preceded that of BX-R
Figures 3C and 3D show the 2-wk moving average of
estrogen levels plotted against physiological age. In all
birds, levels basically remain low with some fluctuations
until about 3 or 4 wk prior to onset of lay, when they
show a sharp increase. In both BX-A and LX-A birds
there was a tendency for greater fluctuations in estrogen
levels in the period 15 to 5 wk preceding lay, as
compared to the BX-R and LX-R groups; and the LX-R
birds appear to show an upward step in estrogen levels
in the period 8 to 4 wk preceding lay. In the absence of
replication, however, it is difficult to know how
representative these group and treatment-specific be-
haviors are. Following the onset of lay, estrogen levels
of all groups remain high with wide fluctuations.
Figure 4A shows rate of lay (eggs per week) plotted
against chronological age. In this case, the increase in
rate of lay with age, represents primarily an increase in
the proportion of birds in lay from week to week
(Bornstein and Lev, 1982). The LX-A and LX-R laid at
virtually identical levels, beginning at about 1.5 eggs per
week at 25 wk, increasing linearly to 6 eggs per week at
33 wk. There was a slight advantage to LX-A until 30
wk, but this advantage is due entirely to the 1 wk earlier
onset of lay in this group. The BX-A group follow a
parallel curve, but at generally lower levels, beginning
at 0.5 eggs per week at 25 wk, and increasing linearly to
4.2 eggs per week at 33 wk, and 5 eggs per week at 35
wk. In remarkable contrast, the BX-R birds did not
initiate lay until 31 wk, and then increased rapidly to 5 FIGURE 4. Rate of lay plotted against age. A) rate of lay plotted
eggs per week at 35 wk. The more synchronous entry against chronological age. B) rate of lay plotted against physiological
age. ∫ = BX-A; π = BX-R; o = LX-A; ÿ = LX-R. BX = broiler cross; LX =
into lay of the BX-R birds is shown by the fact that for layer cross; A = ad libitum feeding; R = restricted feeding. Arrows show
this group, 90% of the birds entered lay within 18 d, as mean age of birds at onset of lay.
 COMB SIZE AND ESTROGEN TOWARD ONSET OF LAY
birds (Table 1). Yet onset of lay in BX-A and BX-R birds
was later than recommended for this strain under normal
commercial conditions (26 to 27 wk); and BW were much
higher than the recommended weight at the start of lay
(2,930 to 3,150 g) (Arbor Acres, 1994). Thus, it seems
reasonable to attribute the differential response of the two
genetic types to the photoperiod treatment, to differences
between them in CDL and SDL, or in photoperiodic
responsiveness.
Within genetic types, feed restriction caused a marked
delay in onset of lay in BX-R birds as compared to BX-A
birds. As pointed out above, the BX-R birds entered lay at
a BW much higher than recommended for this strain at the
start of lay. Consequently, the later onset of lay in the BX-R
birds cannot plausibly be attributed to a failure to meet
BW threshold requirements. It seems reasonable, there-
fore, to attribute the delay to a direct effect of feed
restriction on CDL or photoperiodic responsiveness. A
similar delaying effect of feed restriction on onset of lay,
much smaller in magnitude and statistically nonsignifi-
cant, was found for LX-R birds as compared to LX-A birds.
In this case, information on usual BW of LX birds at onset
of lay was not available to us. Thus, it is not possible to
exclude the possibility that the delay was due to failure of
the LX-R to meet threshold BW requirements.
Comb Size
The increase in comb size for LX-A birds, prior to 22 wk,
and the lack of increase in comb size for the BX-A birds
during this period, indicates that a photoperiod of 6L:18D
was above CDL for the LX-A birds; and below CDL for the
BX-A birds. The lack of response of the BX-R and LX-R
birds may also be due to a higher CDL for restricted birds,
or may be due to lack of attainment of threshold BW.
The gradual increase in comb size of the LX-A group
prior to 22 wk, may reflect the fact that prior to 22 wk some
of the LX-A birds had not yet reached threshold BW;
alternatively it may mean that a 6L:18D photoperiod,
although above the CDL for these birds, is below their
SDL.
The very rapid increase in comb size in the LX-A and
LX-R groups following 2 h of added light at 22 wk may
reflect attainment of threshold BW by all birds; alterna-
tively, it may mean that a photoperiod of 8L:16D was
above their SDL.
The more moderate response of the BX-A birds to the
increase in photoperiod at 22 wk, may reflect the fact that
the 8L:16D photoperiod was above the CDL but some-
what below the SDL for these birds. Alternatively,
photoperiodic responsiveness of BX-A birds may be
below that of LX-A birds.
The very slow initial response, and more rapid later
response of the BX-R birds admits many possibilities:
some of the BX-R birds may not yet have reached
threshold BW for photoresponsiveness at 22 wk; the 8L:
16D photoperiod may have been much below their SDL,
so that the continued increase in photoperiod following 22
1599
wk only somewhat later reached the SDL for these birds;
the photoperiodic responsiveness of BX-R birds may be
less than that of BX-A birds.
When plotted against physiological age, LX-A, LX-R,
and BX-A birds showed first signs of a comb size increase
about 8 wk prior to onset of lay, whereas the BX-R birds
showed an increase from 11 wk prior to onset of lay.
Taking comb size to represent the integration of the
response of the hypothalamic-pituitary-gonadal axis to
photostimulation, these results suggest that the time
course of response to photoperiod was essentially the
same in LX-A, LX-R, and BX-A, birds, but weaker in BX-R.
This conclusion is consistent with the conjecture that LX-A
birds were at or above SDL even at 6L:18 D; that LX-R and
BX-A birds were at or close to SDL at 8L:16D; but that BX-
R birds were initially further from their SDL than the other
groups. A lower photoperiodic responsiveness for the BX-
R group is also a possibility.
Estrogen Levels
With the increase in photoperiod at 22 wk, estrogen
levels of BX-R birds remained unchanged at low levels;
whereas that of all other groups increased modestly. This
result is consistent with the view that at 8L:16D, LX-A and
LX-R birds now exceed their SDL; BX-A birds attain their
CDL, but remain below their SDL; and the BX-R remained
below their CDL. The low estrogen levels of the LX-R
birds in the 14 to 20 wk period, and subsequent increase in
22 to 24 wk period can also be interpreted as an effect of
feed restriction on CDL. However, in this case the
interpretation is equivocal, as it is also possible that BW
thresholds were involved.
When plotted against physiological age, all groups
showed an equivalent increase in estrogen levels starting 3
to 4 wk prior to onset of lay. This effect indicated that once
the estrogen response was established, it reached similar
levels irrespective of genetic type and feeding treatment.
This result is in contrast to the reported relationship of egg
production and luteinizing hormone (LH) levels. Selection
of females for a change in the rate of egg production has
resulted in increased plasma LH concentrations in sibling
males around the onset of sexual maturation (Gowe et al.,
1991). In addition, the duration of the period of rapid
comb growth in birds was closely related to prepubertal
LH peak. Differences in mean plasma LH concentrations
in individual birds of either sex before the onset of puberty
also appeared to be related to subsequent reproductive
performance (Sharp, 1975).
The results of this study show some degree of
dissociation between comb size and estrogen levels. Thus,
although estrogen levels of BX-A and LX-A are greater
than those of BX-R and LX-R in the 14 to 20 wk period;
comb size of the LX-A groups shows a distinct increase
from 14 to 20 wk, which is not seen for the BX-A birds.
Similarly, comb size of BX-A birds begins to increase only
after the increase in photoperiod at 22 wk. Although
estrogen levels also increase modestly at this time, the
1600 EITAN ET AL.
modest increase does not seem sufficient to explain the
marked increase in comb size. In addition, comb size of
LX-A birds increased dramatically after 22 wk, whereas
estrogen levels do not change at all; comb size of BX-R
birds increase from 22 wk, although estrogen levels did
not begin to increase until 29 wk.
Summarizing the above discussion, leads to the
following conclusions. The BX-A and BX-R groups
entered lay at much higher BW and at later ages than
usual for this strain under commercial conditions. Thus,
the delay in onset of lay in the BX groups as compared to
the LX groups, and in BX-R as compared to BX-A, cannot
be explained as a result of the interplay of threshold
weight requirements alone, and some effect of the
photoperiod treatments must be involved. For LX-A and
BX-A, the time course of comb and estrogen responses
against physiological age, are very similar. This similarity
suggests that the major difference between the two genetic
types is in CDL and SDL, rather than in photoperiod
responsiveness, with broilers having a higher CDL and
SDL than layers, which confirms the previous results of
Dunn and Sharp (1990). For BX-A and BX-R, the results
lead to the unexpected conclusion that restricted feeding
per se affects the photoperiodic responsiveness or CDL of
BX birds, independently of an effect on BW. Results that
can be interpreted in a similar manner were reported by
Robinson (1994). Considering the ancestral environment
of the domestic chicken, it may be speculated that in the
event of food shortages in early spring, it would be
adaptive to delay the onset of lay in the anticipation of
better conditions later in the season. Alternatively, the
effect of feed restriction on photoperiod responsiveness
could be a simple result of nutritional imbalance in-
troduced by feed restriction (pers. comm. John Brake,
Department of Poultry Science, North Carolina State
University, Raleigh, NC 27695). Because levels of feed
restriction (relative to ad libitum consumption) have been
continually increasing in heavy breed birds, as a conse-
quence of the continued increase in juvenile growth rate
while keeping mature BW constant, either hypothesis
provides an explanation for the observed decrease in
photoperiodic responsiveness in heavy breed birds.
ACKNOWLEDGMENTS
This research was supported by a grant from the
Israel Poultry Council. We thank Tikva Geno for
dedicated technical assistance.
REFERENCES
Arbor Acres, 1994. Rearing Manual for Breeder Hens and
Males. Arbor Acres Farms, Inc., Glastonbury, CT.
Bornstein, S., and Y. Lev, 1982. The energy requirements of
broiler breeders during the pullet-layer transition period.
Poultry Sci. 61:755–765.
Dunn, I. C., and P. J. Sharp, 1990. Photoperiodic requirements
for LH release in juvenile broiler and egg-laying strains of
domestic chickens fed ad libitum or restricted diets. J.
Reprod. Fertil. 90:329–335.
Eitan, Y., and M. Soller, 1991. Two way selection for threshold
BW at first egg in broiler strain females. 2. Effect of
supplemental light on weight and age at first egg. Poultry
Sci. 70:2017–2022.
Eitan, Y., and M. Soller, 1994. Selection for high and low
threshold BW at first egg in broiler strain females. 4.
Photoperiodic drive in the selection lines and in commer-
cial layers and broiler breeders. Poultry Sci. 73:769–780.
Eitan, Y., and M. Soller, 1996. Selection for high and low
threshold BW at first egg in broiler strain females. 7. Effect
of photoperiod on BW and age at onset of mature semen
production in males of the selection lines, and in
commercial broiler and layer males. Poultry Science 75:
828–832.
Gowe, C. B., R. J. Scaramuzzi, N. B. Carter, and P. J. Sharp,
1991. Plasma concentrations of luteinizing hormone and
BWs during somatic maturation in intact and castrated
Australorp cockerels from a line of hens selected for
increased ovulation rate. Br. Poult. Sci. 32:799–808.
Jacoby, S., N. Snapir, I. Rozenboim, E. Arnon, R. Meidan, and
B. Robinzon, 1992. Tamoxifen advances puberty in the
White Leghorn hen. Br. Poult. Sci. 33:101–111.
King, D. F., 1961. Effects of increasing, decreasing, and
constant light treatments on growing pullets. Poultry Sci.
40:479–484.
Lowry, M. M., 1958. Compensatory testicular adjustments
following the injection of small quantities of androgen into
unilaterally castrated White Leghorn cockerels. Poultry
Sci. 37:1129–1136.
Robinson, F. E., 1994. Ovarian form and function in chickens of
varying reproductive status. Final Report, Alberta Agricul-
tural Research Institute Project Number #AAR1920202.
University of Alberta, Edmonton, AB Canada.
Rozenboim, I., G. Gvaryahu, B. Robinzon, N. Sayag, and N.
Snapir, 1986. Induction of precocious development of
reproductive function in cockerels by Tamoxifen adminis-
tration. Poultry Sci. 65:1980–1983.
Rozenboim, I., N. Snapir, E. Arnon, R. Ben-Aryeh, W. H.
Burke, P. J. Sharp, Y. Koch, and B. Robinzon, 1993.
Precocious puberty in tamoxifen treated cockerels:
hypothalamic gonadotropin-releasing hormone-I and
plasma luteinizing hormone, prolactin, growth hormone
and testosterone. Br. Poult. Sci. 34:533–542.
SAS Institute, 1985. SAS User’s Guide: Statistics. Version 5
Edition. SAS Institute Inc., Cary, NC.
Sharp, P. J., 1975. A comparison of variation in plasma
luteinizing hormone concentrations in male and female
domestic chickens (Gallus domesticus) from hatch to sexual
maturity. J. Endocrinol. 67:211–223.
Snapir, N., I. Nir, F. Furuta, and S. Lepkovsky, 1969. Effect of
administered testosterone propionate on cocks function-
ally castrated by hypothalamic lesions. Endocrinology 84:
611–618.
Walpole, R. E., and R. H. Myers, 1978. Probability and
Statistics for Engineers and Scientists. 2nd ed. Macmillan
Publishing Co., Inc., New York, NY.
Wilson, W. O., and A. D. Woodward, 1958. Egg production of
chickens kept in darkness. Poultry Sci. 37:1054–1057.