Anda di halaman 1dari 10

Breed and Environmental Effects on Postweaning

Growth of Rabbits

J. I. McNitt*J and S. D. Lukefahrt

*Center for Small Farm Research, Southern University and A&M
College, Baton Rouge, LA 70813 and +International Small Livestock
Research Center, Department of Food Science and Animal Industries,
Alabama A&M University, Huntsville 35762

ABSTRACT Growth records of 4,270 weanling NZW tended to be less affected by the environmental
rabbits born between March 1985 and December 1989 extremes than the other breeds. The effects on GAIN
were studied to evaluate the effects of breed and of mean monthly temperature and daylength and the
month of birth on postweaning growth performance of interrelationships of these with estimated milk
four medium-sized breeds: Californian (CALI, New production and litter size at weaning were evaluated
Zealand White (NZW), Palomino (PAL), and White by regression methods for the 2,100 NZW fryers.
Satin (WS). Traits examined were 28-d weaning Temperature and daylength had significant effects on
weight (WW), postweaning gain (GAIN), attainment GAIN, with lowest GAIN in the summer, but the
of 1,600-g market weight by 76 d of age (MKT), and individual contributions t o the variance were small
approximate age at 1,600 g (AGE). Least squares
because of some redundancy when month, tempera-
models included breed, month of birth, sex, and year of
ture, and(or) light were included in the same model.
birth as fixed effects and litter within breed by month
and by year and the residual as random variables. The Curvilinear trends were observed that favored GAIN
NZW had significantly higher GAIN and MKT and as estimated milk production increased but decreased
lower AGE than the other three breeds. White Satin GAIN as litter size at weaning increased. In the hot,
had the highest WW, followed by CAL, NZW, and humid climate of southern Louisiana important breed
PAL. White Satin had higher GAIN and lower AGE differences were noted. There were also indications
than PAL or CAL but did not differ for MKT. Poorer that daylength may be an important factor in post-
performance was seen during the summer, but the weaning fryer performance.
Key Words: Rabbits, Breeds, Growth, Environmental Temperature, Photoperiod

J. Anim. Sci. 1993. 71:1996-2005

Introduction 56-d weight, whereas McNitt and Lukefahr (19911, in

Louisiana, and Roberts and Lukefahr (19921, in
Several factors influence postweaning growth of Alabama, detected significant month effects on in-
rabbits. Climatic variations on a seasonal or shorter dividual market weight traits. Comparisons of these
basis have been shown to have definite effects on doe studies must be done with caution because the rabbits
productivity (Ross et al., 1961; Sittmann et al., 1964; in the California studies were weaned at 56 d and
McNitt and Moody, 19901, but few studies have been were all purebred New Zealand White (NZW),
conducted in the United States of the effects of climate whereas in the latter three reports rabbits were
on fryer production. Rollins and Casady (19601, in weaned at 28 d and represented several breeds and
California, found no effect of season on individual crosses.
weaning weight at 56 d; however, season did affect Breed comparison studies conducted in the United
total litter weight at 56 d (Casady et al., 1962). More States involving the medium-sized NZW, Californian
recently, Lukefahr et al. (19861, in Oregon, observed ( CAL) , Champagne D'Argent, and Palomino ( PAL)
that season had a significant effect on individual breeds generally have demonstrated small to nonsig-
nificant differences for postweaning growth traits
(Lukefahr et al., 1983b; Grobner et al., 1985; Ozimba
'To whom correspondence should be addressed.
and Lukefahr, 1991; Roberts and Lukefahr, 1992). In
Received November 30, 1992. contrast, crossbred litters from large-sized sire breeds
Accepted April 6, 1993. (e.g., Flemish Giant and Terminal Sire Line) have

been reported to achieve more rapid gains, attaining 35 -
market weight earlier than medium-sized breeds or
30 - -
their crosses (Lukefahr et al., 1983b, 1984; Ozimba
and Lukefahr, 199 1). Development and general 25 --
recommendations for purebreeding or crossbreeding u* 20 --
programs will require information on genetic and E
environmental parameters for growth performance $ 15--

traits of both purebred and crossbred rabbits reared in 10-
a variety of climates. To date, no postweaning growth I-

studies using the white variety of the Satin (WS) 5 ._

breed have been reported. - ,

The objectives of this investigation were 1 ) to
compare four medium-sized meat rabbit breeds for
postweaning growth traits, 2 ) to examine breed x
month trends for growth traits during a 5-yr period, Figure 1. Monthly mean maximum and minimum
3 ) to determine the influence of mean monthly temperatures (fSE) and daylength in the rabbitry from
temperature on ADG in NZW purebred fryers, 4 ) to March 1985 to December 1989.
determine the effect of daylength on ADG in NZW
purebred fryers, and 5) to study response surfaces
involving climatic factors (daylength and tempera- office located at an airport 3 km from the rabbitry.
ture) and maternal and(or) litter characters (esti- Because of daylight savings time, the lengths of the
mated milk yield and litter size at weaning) and ADG daily light periods reached a minimum in October and
in NZW purebred fryers reared in southern Louisiana. a maximum in April (Figure 1).
A commercial, alfalfa-based, pelleted rabbit diet
(Petrus Feed and Seed, Alexandria, LA) with a
Materials and Methods guaranteed analysis of 18% CP, 18%crude fiber, and
2.5%crude fat was available t o the rabbits fresh daily
Stock Housing and Environment. The data used in on an ad libitum basis. Water was available continu-
this analysis were collected from 4,270 rabbits reared ously from automatic valves.
in the Southern University rabbitry between March Breeds and Traits Studied. The fryers were of four
1985 and December 1989. The herd comprised com- medium-sized breeds of US.origin (American Rabbit
mercial stock used for production research at the Breeders Association, 1991) and included 720 CAL,
Center for Small Farm Research at Southern Univer- 2,100 NZW, 691 PAL, and 759 WS. The breeding stock
sity and A&M College, Baton Rouge, LA (latitude 30” were selected as commercial replacement animals on
32”). The rabbits were housed in groups of up to four the basis of gains and body conformation. Matings
in suspended, all-wire cages (76 cm x 76 cm x 46 cm) were unplanned, although sire-daughter and full-sib
inside a building with opened side panels that matings were avoided. No fostering of kits at birth to
provided protection from rain and sun. Fans were used equalize litter size was practiced. The does were
for air circulation when ambient temperatures ex- usually rebred 8 d postpartum. The distribution
ceeded approximately 23°C. among breeds of sires, dams, litters, and fryers is
Climate in the rabbitry was monitored by a shown in Table 1. Of the 4,270 fryers included, 3,915
recording hygrothermograph that provided continuous reached market weight within the specified time.
records of the temperature and humidity. As far as In all years, there were more NZW than rabbits of
possible, the hygrothermograph was maintained in an other breeds. In the earlier years of the study this was
empty cage or at cage height to provide a record of the a reflection of the greater productivity of the NZW in
temperatures actually experienced by the rabbits. terms of the numbers of fryers that actually reached
Monthly means of daily maximum and minimum market weight. The breed comparison study was
temperatures were used for statistical analyses. The phased out by attrition after 1988 and the herd
mean monthly maximum and minimum temperatures reverted t o purebred NZW. As a result, there was no
in the rabbitry for the period of this study are shown replacement of breeding does of the other breeds,
in Figure 1. Of particular interest are June, July, which resulted in the decline in numbers.
August, and September, when the mean maximum All rabbits were weaned at 4 wk of age. They were
temperatures exceeded 30°C and the mean minimum weighed and tagged at weaning and then weighed
temperatures did not fall below 21°C. once weekly, usually on Tuesday, until they were
Lights were turned off daily at 2200. Daylength removed from the herd. Rabbits were generally
(total length of natural light plus added hours of marketed on a weekly basis and a shipment included
incandescent lighting) thus varied according to the all animals that weighed 2 1,600 g live weight at the
time of sunrise. Those times were obtained from the latest weighing. Data were included in this study for
National Oceanic and Atmospheric Administration all fryers that reached 2 1,600 g or had not been
Table 1. Numbers of rabbits of each breed and classification
included in the study

~ ~~~~

Category CAL NZW PAL ws Total

Sires 8 9 8 8 33
Dams 29 76 30 31 166
Litters 157 382 151 180 870
Fryers 720 2,100 69 1 759 4,270
aCAL = Californian, NZW = New Zealand White, PAL = Palomino, WS = White Satin.

removed from the herd before the sixth weighing after litter accounts for more than just additive genetic
weaning ( a maximum of 76 d). Using the weekly variance. For the sex source and interactions inclusive
weights, a linear regression coefficient of BW on day of of sex, the residual was the error term used. In
age during the postweaning growth period was preliminary analyses, except for breed x month and
computed for each fryer as an estimate of the breed x sex, no other interactions influenced the
individual postweaning rate of gain (Liu et al., 1990). growth traits investigated, so these sources were
Traits included in the analysis were individual eliminated from the models. Also, age and(or) parity
weaning weight at 28 d (WW), postweaning rate of of dam were not included in the above model because
gain ( GAIN), age at which the 1,600-g market weight preliminary analyses showed these effects to be
was first recorded (AGE), and attainment (0, 1) of nonsignificant. The Bonferroni t-test was used to
market weight ( MKT) at or before the sixth weighing separate breed means at the P < .05 probability level
postweaning. (Gill, 1978). The sex mean difference with a single df
Statistical Procedures. To compare the postweaning was tested based on results from ANOVA.
performance of the four breeds, data were analyzed To examine breed x month trends for growth traits,
using the General Linear Mixed Model (GLMM) a statistical package (Harvey, 1990) with polynomial
program of Blouin and Saxton (19901, according to regression capabilities was used. The second mathe-
the following mathematical model: matical model was the same as above, except that
equations for litters were absorbed using the ratio of
residual to litter variances obtained from GLMM
analyses (litter effect accounted for 75.3, 37.5, 21.2,
and 47.2% of total random variation [o$ = 0; + 0 2~ 1
for WW, GAIN, AGE, and MKT, respectively), as
where Yijumn = observed value of the dependent
variable; p = overall mean; Bi = fxed effect of the ith
breed; Mj = fxed effect of the jth month of birth; YRk =
fixed effect of the kth year of birth; (BM)ij = fixed
effect of the breed x month interaction; (BYR)ik =
fixed effect of the breed x year interaction; (MYR)p = A generalized least squares solution for a fixed effects
fixed effect of the month x year interaction; model was then obtained using Harvey's program. The
(BMYR)ijk = fixed effect of the breed x month x year sums of squares for breed, month, and breed x month
interaction; ll$ = random effect of the kth litter nested interaction were partitioned into those due to each
within breed x month x year of birth subclass, degree in orthogonal polynomials (limit was 5th
assumed to be NID (0, 4); Gm = fixed effect of the mth degree). A step-down, P c .05 probability level was
used to yield best fit polynomial regression models.
sex class; (BG)im = fixed effect of the breed x sex
interaction; (MG)jm= fixed effect of the month x sex To carry out a more detailed analysis of the effects
interaction; (YRG)I, = fxed effect of the year x sex
of daylength and temperature on growth, only the
data from the purebred NZW were used because of the
interaction; and qumn= random errors, assumed to be
larger numbers ( n = 2,100). After absorbing the
NID (0, 4). random litter effect, solutions for fixed effects were
The above full model was employed to obtain obtained from two generalized least squares models
generalized least squares means for breed, month, and consisting of month, year, and sex and first-order
year and their first- and second-order interactions interactions, excluding or including mean monthly
where the random litter source served as the error temperature (TI as a covariate, as follows:
term. Despite the obvious genetic relationships among
litters, this was considered the appropriate error term
because of the limited number of sires and because
Table 2. Generalized least squares breed and sex means, SE, and variances for growth characteristics

Age a t reaching
Weaning Rate of gain, market wt, market wt
Item wt, g ( W W g/d (GAIN) d (AGE) (MKT)
Californian 525 f 9.6b 35.0 k .35a 61.1 f .2gC .90 f .02a
New Zealand White 539 k 6.2c 39.9 f .23c 58.4 f .laa .94 f .Olb
Palomino 496 f 34.5 f .36a 61.9 f .30d .89 ir .02a
White Satin 550 k 9.1d 36.0 f .34b 60.1 f .28' .92 f .02a
Male 529 f 4.6 36.7 f .19 60.2 ir .16 .92 f .01
Female 526 k 4.6 36.0 i .19 60.5 ir .16 .90 f .01
Male-female 3 k 2.1 .7 f .17** -.3 f .16** .03 f .01**
Among-litter, a1 11,590 12.0 14.2 .02
Within-litter, uc 3,810 19.7 15.9 .06
a,b,c,dBreed means within a column lacking a common superscript letter differ ( P c .05).
**P c .01.

of the litter at 21 d (Lebas, 1970; Lukefahr et al.,

1983a). Model 4b was augmented by the additional
covariates litter size at weaning ( L S ) and estimated
milk production ( M P ) to produce Model 5:
where all effects are as defined for Model 1, except for
the linear, quadratic, and cubic regression of GAIN on
mean monthly temperature. In preliminary analyses,
interactions were not significant and so were elimi-
nated from the final models. A step-down, P < .05
probability level was used for testing temperature to
yield the best-fit regression model. Using multiple regression analysis step-down proce-
To determine the effect of mean monthly daylength dures, covariates were first tested at the cubic level for
on GAIN in NZW purebred fryers, month as a main significance at the P e .05 probability level. Partial
effect was replaced by daylength ( L ) (the number of regression coefficients obtained from the final, best-fit
classes being 10) in the Models 3a and 3b described regression prediction equation were used to plot three-
above to produce Models 4a and 4b: dimensional GAIN response surfaces.

Results and Discussion

Breed Comparisons. According to Model 1 results,

least squares means for WW differed significantly
among all four breeds (Table 2). The WS had the
heaviest WW, whereas the GAL had the lightest. The
NZW and PAL breeds were intermediate. Small breed
differences for WW were associated with a high
experimental sensitivity (residual df = 4,214 for WW,
Interactions were not important in preliminary ana- GAIN, and MKT and 3,609 for A G E ) . Previous
lyses and were therefore not included in the final comparative breed studies (Carregal and Lui, 1984;
models. A polynomial regression analysis was per- Grobner et al., 1985; Coudert and Brun, 1989; Ozimba
formed for daylength and the effect of the temperature and Lukefahr, 1991; Roberts and Lukefahr, 1992)
covariate was again evaluated to determine the best- involving fewer observations did not report differences
fit regression model. for this trait.
The final objective was to study interrelationships The NZW had higher GAIN than the other three
between climatic (daylength and temperature) and breeds ( P c .05). The WS had significantly better
maternal (litter size at weaning and estimated milk GAIN than GAL and PAL, whereas the latter two
production) factors on GAIN in NZW purebred fryers. breeds did not differ. Grobner et al. (1985) observed
Milk production was estimated from the total weight no difference between NZW and PAL purebred fryers
in gain performance from 28 t o 56 d, although the Breed x Month Trends. Based on Model 2 results,
former breed was numerically superior by 3.8 gid. there were definite breed trends in relation to month,
Ozimba and Lukefahr (1991) found no difference in although breed did not interact with month in
gain from 28 to 70 d in purebred CAL and NZW influencing WW (Figure 2a). All the breeds showed
litters. In a crossbreeding study, Masoero et al. similar declines in WW through the summer months
(1985) reported more rapid postweaning gains in and had the highest values in December and January.
fryers sired by CAL than in those sired by NZW bucks, The increase in WW in the autumn tended to be
but no difference between fryers reared by CAL and slower for the WS than for the NZW and CAL.
those reared by NZW dams. No differences in gain Weaning weight at 28 d is strongly influenced by the
from 28 to 56 d between CAL and NZW sires or dams milk production of the dam and the size of the litter
in purebred or crossbred litters were observed by (Lebas, 1970; Lukefahr et al., 1983a). The latter
Lukefahr et al. (1983b). Roberts and Lukefahr authors reported that 89% of the variation in total
(1992) reported that PAL x NZW crossbred litters 21-d litter weight was directly attributable to milk
gained more slowly from 28 to 70 d than did CAL x production of the dam. Heavy WW is important
NZW crossbred litters, whereas NZW purebred litters because the time periods when the WW was greatest
did not differ from these two crossbred groups. were also the periods when AGE was lowest.
Overall, there was no clear trend for postweaning gain The GAIN of the NZW was more rapid than that of
performance involving the medium-sized breeds in the other breeds for all birth months (Figure 2b). The
this study compared with that reported in the CAL and the NZW had similar GAIN response early in
literature. Generally, use of nonstandardized breed the year with a decline in the summer and an increase
populations and limited control of environmental later in the year. The NZW, however, had much
factors (e.g., climate, diet, housing, and management) higher values than the CAL for all months. The PAL
may explain this disparity. and WS showed a decline in GAIN early in the year
The breed means for AGE were different ( P c .05). with an increase in the early autumn and a secondary
The NZW reached market weight earliest (58.4 d ) , decline in late autumn. This seemingly unique re-
followed by WS (60.1 d ) , CAL (61.1 d), and PAL sponse may have been in some way related to the
(61.9 d). Rouvier (1973) reported that NZW purebred fluctuating daylengths in those months. A significant
fryers reached 2,200-g market weight 6 d earlier than breed (cubic) x month (cubic) interaction was de-
CAL purebred fryers, compared with the approximate tected in GAIN response. For all breeds, the least
3-d difference to reach 1,600 g observed in our study. favorable production occurred during the summer and
Proportion of marketable fryers was significantly early autumn months and the most favorable during
higher for NZW than for the other three breeds, which the winter.
did not differ from each other. Obviously, the superior The mean age at marketing was lowest early and
GAIN and AGE performance for NZW fryers directly late in the year for all breeds, reached a maximum for
account for the better MKT results, even though the kits born in the summer, and then declined through
breed mean range was only .05. Roberts and Lukefahr the autumn (Figure 2c). Generally, the changes in
(1992) observed no differences in the proportion of this growth characteristic closely follow as a reverse of
marketable fryers by 70 d among NZW purebred and the curve for GAIN. A significant breed (quadratic) x
CAL x NZW and PAL x NZW crossbred fryers. month (linear) interaction was observed in AGE
Numerical rank order of sire breed means did response. The NZW showed a more symmetrical trend
correspond, however, to the ranking in the present in market age from winter and spring and from
experiment. summer and fall than the other breeds.
No sex difference existed in WW, but males had An increase MKT was seen throughout the year for
slightly better GAIN, AGE, and MKT performance all breeds in nearly parallel fashion (Figure 2d).
than females (by .7 g/d, -.3 d, and .02 proportion, Breed did not interact with month in affecting MKT.
respectively). Paradoxically, due to the experimental For all months, the NZW had the highest MKT and
sensitivity involved in this large data set, these the PAL the lowest.
differences were statistically significant, although the Generally, poorer performance was seen during the
biological importance is questionable. Unlike in most summer and early autumn, although there were
domestic livestock species, the sex effect does not different breed trends noted in response to month.
strongly influence weaning-to-market growth and These results show the serious decline in production
carcass characters in rabbits (Rollins et al., 1963; during the summer months when the temperatures
Vogt, 1979; Lukefahr and Ozimba, 1991; Ozimba and are the highest. It is also apparent from these results
Lukefahr, 1991). In addition, a significant breed x sex that, among the breeds studied, production and
interaction was detected for GAIN. Females had growth performance of the NZW tends to be less
higher GAIN than males (but by only .07%) in the sensitive to the climatic extremes. The WS seemed to
NZW, whereas the opposite trend was observed for the follow the NZW closely. That may be a reflection of the
other three breeds. The implications of this minor relatively high proportion of NZW breeding that was
interaction can be disregarded. originally included when the WS was developed.

..' 0

5 0 . 1 : : : : : : : : : : : I


86 .m4: : : : : : : : : : : I

Figure 2. Breed x month responses for fryer growth performance. (CAL = Californian, NZW = New Zealand
White, PAL = Palomino, and WS = White Satin; WW = weaning weight at 28 d, GAIN = postweaning rate of gain,
AGE = age at which the 1,600-gmarket weight was first observed, MKT = proportion attaining market weight at or
before the sixth weighing postweaning].

Year was a significant source of variation for WW, also the years when the highest proportion of fryers
GAIN, and AGE, although there were no apparent were marketed at the lowest ages. The heaviest WW
trends in the year means during the 5-yr period of the were observed in 1986, although those in 1987 were
study (Table 3 ) . The highest rates of gain were in nearly the lowest.
1986 and 1987 when, incidentally, the lowest mean Temperature and Daylength Effects on Postweaning
maximum temperatures were recorded. These were Weight Gain o f New Zealand White Fryers. As shown

Table 3. Mean values (+ SE) by year for production characteristicsa

Item 1985 1986 1987 1988 1989

Production trait least squares means

ww, g 539 k 12.0 557 f 8.8 514 f 8.2 509 i 8.1 518 f 9.75**
GAIN, g/d 35.5 k .46 37.4 k .33 37.6 f .30 36.0 f .29 35.3 f .35**
AGE, d 60.4 f .37 59.6 f .27 59.7 k .25 61.7 f .24 60.6 f .28**
MKT .89 f .03 .94 i .02 .92 f .02 .92 f .02 .87 f .02
Covariate overall means
TEMP, "C 27.5 f .39 25.0 i .38 24.0 f .38 25.9 i .38 24.8 f .41
MILK, g 2,060 k 20.9 1,993 2 18.2 2,066 f 17.7 2,129 f 15.2 2,213 f 17.4
LS28, n 6.2 i .09 5.8 f .07 6.3 k .06 6.8 i .06 6.9 f .07
aWW = weaning weight at 28 d, GAIN = postweaning rate of gain, AGE = age at which the 1,600-gmarket weight was first observed, MKT
= attainment ( 0 , 1 ) of market weight a t or before the sixth weighing postweaning, TEMP = mean monthly maximum temperature, MILK =
estimated doe milk production, and LS28 = litter size at weaning a t 28 d.
**Indicates year was a significant ( P < .01) source of variation for these variables.

50 - summer. Addition of temperature as a covariate

(Model 3b) resulted in a quintic response for month,
with the highest GAIN in the summer and the lowest
45 -- in mid-winter, suggesting a somewhat different re-
sponse due to month. Although temperature was a
\ significant source of variation for GAIN in Model 3b,
U, 40.-
its marginal contribution to the total variance was
z4 small (4.45 - 4.06% = .39%) in relation to Model 3a
0 (Table 4). This probably occurred because month
35 -- d
already reflects most of the variance due to this
source. Several authors have associated elevated
3 0 4 : : : : : : I I I I 4 temperatures with reduced feed intake and rate of
gain (Simplicio et al., 1988; Jin et al., 1990). Others
MONTH have compared seasonal performance and attributed
the lower performance during the summer to higher
Figure 3. Postweaning rate of gain (GAIN) of New temperatures (Lebas and Ouhayoun, 1987; Simplicio
Zealand White fryers for the Models 3a and 3b as et al., 1988; Ismail, 1992). This assumption is not
described in the text. strictly true, however, because of other seasonal
factors such as daylength and precipitation, which
may also affect performance.
in Figure 3, Model 3a, which included month, year, Daylength also significantly influenced GAIN.
and sex as fixed effects, predicted that the response for Model 4a, which included daylength replacing month
month would be quartic with the highest GAIN in the as a main effect, showed an increase in GAIN as
early spring and late autumn and the lowest in the daylength increased up to 15.0 h, with a decline

Table 4. Prediction equations for average daily gain (GAIN] in New Zealand White rabbits according to en-
vironmental and(or) maternal characters

Item 3a 3b 4a 4b 5
Intercept 37.68 46.85 39.68 41.75 41.17
Month of birth
Linear .2593 2.6870
Quadratic ,4346 -.5179
Cubic -.0029 -.2166
Quartic -.0113** ,0056
Quintic .0046**
Daylength, h
Linear -.06770 -.06772 -.06400
Quadratic .00011 -.00036 -.00036
Cubic .00001* .00001* .00001**
Temperature, "C
Linear -1.0699 ,3383 .3868
Quadratic -.0488* -.0498 -.0479
Cubic - -.0062* -.0065**
Litter size a t 28 d, n
Linear -.7558
Quadratic .1638
Cubic -.0297*
Estimated milk yield, kg
Linear 2.8407
Quadratic - - - - -2.9063**
Residual df 2,083 2,081 2,085 2,082 2,077
Residual R2,% 4.06 4.45 3.84 4.42 6.30
Total R2, %b 48.66 49.05 45.31 45.89 47.77
aRefer to text for full description of models.
bMultiple coefficient of determination (R2) values represent the sum of partial contributions of model fixed effects only from a generalized
least squares analysis after absorbing variation accountable to random among-litter effects. The difference between total and residual R2
values is due to the litter effect.
*P < .05.
**P< .01.
crease feed intake in rabbits (Cheeke et al., 1987;
Papp et al., 1987; Simplicio et al., 1988; Jin et al.,
1990). Further, the partial linear, quadratic, and
cubic regression coefficients for temperature and
daylength were similar in both Models 4b and 5
(Table 41, even with the additional covariates of litter
size at weaning and estimated milk yield. Statisti-
cally, this finding might imply that milk production
and litter size at weaning tended to be independent of
environmental effects of temperature and daylength.
37 J 4 GAIN was clearly maximized when litter size at
14.80 15.00 15.20 15.40 15.60 15.80 16.00 weaning was at a minimum (one fryer), whereas
DAYLENGTH, h temperature had a more consistent cubic response
over the litter size range (Figure 5b). Interestingly, a t
Figure 4. Postweaning rate of gain [GAIN) of New intermediate temperatures, the competition-related
Zealand White fryers for the Models 4a and 4b as stress of a large litter size on GAIN seemed to be less
described in the text. marked. Estimated milk production of does had a
quadratic tendency on their own fryer GAIN, but in
association with a cubic effect of temperature (Figure
thereafter, with a slight increase in growth rate from 5c). This implies that fryers that suckled does with
15.9 to 16.0 h (Figure 4). A similar pattern was seen lower milk production may have been more sensitive
when temperature was included (Model 4b), although to cold and heat stresses. GAIN was maximized at the
predicted GAIN was at a higher level, reaching a peak milk production level of approximately 3.2 kg.
at approximately 15.1 h. The marginal contribution of Daylength imparted only a small effect on GAIN,
temperature was only 5 8 % (Table 4). The similarity but, as litter size at weaning increased, GAIN sharply
in the patterns between models with and without decreased in a quadratic manner (Figure 5d). Milk
temperature could also be explained on the basis of production influenced GAIN to a greater extent than
natural confounding among temperature, daylength, daylength (Figure 5e); the higher milk production
and month variables. The difference between the level gave a favorable quadratic response in GAIN. As
“month” models (3a and 3b) that excluded or included expected, fryers from the largest litters and nursing
temperature and the “light” models (4a and 4b) was from the poorest-lactating does achieved the slowest
only .22 and .03% of total variation in GAIN, GAIN (Figure 5 0 . GAIN declined substantially as
respectively. Either would seem appropriate, there- litter size at weaning increased from one to four
fore, depending on whether a month- or light-constant fryers, and thereafter showed a steady negative linear
reference is preferred. trend to a litter size of 11 fryers. Hardman et al.
The unusual annual variation in daylength that (1970) and McNitt and Moody (1988) demonstrated
arose from not controlling the morning light adds to that kits responded positively to an increased milk
the difficulty of drawing precise conclusions regarding
supply (i.e., double-suckling). Lukefahr et al. (1990)
the effect of daylength on fryer performance in the
recommended that, to avoid poor gains, litter size at
commercial situation. Changes in daylength of as little
birth should be adjusted by fostering so as not to
as 1 h are sufficient to induce marked alterations in
reproductive behavior of does (Hudson and Distel, exceed nine kits.
1990). Although the effect is probably not as obvious In Table 4, the joint marginal contribution due to
in growing rabbits, such changes may nonetheless litter size at weaning and estimated milk production
affect growth performance. From our investigations, it in addition to the variation accounted for by the
seems that there are real effects of daylength on fryer random litter effect (47.77 - 6.30% = 41.47%) is
growth rate that need to be investigated in much more small. This finding is despite the marked response
detail under more controlled conditions. surfaces portrayed for GAIN involving these two
Postweaning Weight Gain Response Surfaces. From independent variables. The litter source may reflect
Model 5, response surfaces with GAIN as the depen- certain proportions of direct genetic variance (addi-
dent variable and daylength, temperature, litter size tive, dominance, and epistasis), plus maternal genetic
at weaning, and estimated milk yield as independent and(or) environmental variance, direct by maternal
continuous variables were studied (Figure 5). The genetic covariance, and environmental effects (e.g.,
cubic response in GAIN associated with temperature competition, litter size, pen effects) apart from mater-
was much more pronounced than that due to day- nal environmental effects being shared in common
length (Figure 5a). More feed energy is expended t o among littermates. Further research is also warranted
maintain body temperature at colder ambient temper- to determine the relative importance of these compo-
atures, whereas higher ambient temperatures de- nents of variance in meat rabbit production.

Figure 5. Response surfaces for postweaning rate of gain [GAIN) in relation to daylength, temperature, litter size
at weaning, and estimated milk production.

Implications Literature Cited

. This study demonstrates important breed responses American Rabbit Breeders Association. 1991. Official Guidebook to
t o the hot, humid climate of southern Louisiana. Of Raising Better Rabbits and Cavies. Am. Rabbit Breed. Assoc.,
Bloomington, IL.
the four breeds studied' the New White had Blouin, D. C., and A. M. Saxton, 1990. General Linear Mixed Models
the best postweaning growth performance under these (GLMM) User's Manual. Louisiana State Univ., Baton Rouge.
conditions. Therefore, the New Zealand White is Carregal, R. D., and J. F. h i . 1984. Effect of mothering ability and
recommended when a purebred fryer production sys- general combining ability in crossbred rabbits. Rev. SOC.Bras.
tem is preferred. Month, temperature, and(or) day- Zootec. 13(4):440.
Casady, R. B., W. C. Rollins, and D. B. Sittmann. 1962. Effect of
length clearly have pronounced effects on fryer season and age of dam on individual weaning weights, number
growth, although these climatic variables are as- weaned, and total litter weight of hutch raised domestic rab-
sociated with the time of year and are naturally bits. Small Stock 46:7.
confounded to some extent. Further work is needed to Cheeke, P. R., N. M. Patton, S. D. Lukefahr, and J. I. McNitt. 1987.
examine more closely the effects of these factors on Rabbit Production (6th Ed.). Interstate Printers and Publish-
ers, Danville, IL.
fryer production.
Coudert, P., and J. M. Brun. 1989. Productivity and morbidity of
rabbit breeding does: A comparison of 4 genotypes. Genet. Sel. weaning and post-weaning traits in rabbits. h i m . Prod. 38:
Evol. 21:49. 293.
Gill, J . L. 1978. Design and Analysis of Experiments in the Animal Lukefahr, S. D., and C. E. Ozimba. 1991. Prediction of carcass merit
and Medical Sciences. Vol. 1. pp 175-177. Iowa State Univer- from live body measurements of rabbits of four breed types.
sity Press, Ames. Livest. Prod. Sci. 29:323.
Grobner, M. A,, K. L. Robinson, N. M. Patton, and P. R. Cheeke. Masoero, G., A. Ubertalle, P. Mazzocco, and L. M. Battaglini. 1985.
1985. Comparison of the Palomino and New Zealand White Terminal crossing of New Zealand White and Californian rab-
breeds for commercial rabbit production. J. Appl. Rabbit Res. 8: bits: 1. Characteristics on the live animal. Ann. 1st. Sper.
62. Zootec. 18:93.
Hardman, J. J.,D. Hull, and J. Oyesiku. 1970. The influence of birth McNitt, J. I., and S. D. Lukefahr. 1991. Postweaning growth charac-
weight and nutrition on postnatal growth of rabbits. Biol. Ne- teristics of four breeds of rabbit in Louisiana. J . Anim. Sci.
onate 16:306. 69(Suppl. 1):8 (Abstr.).
Harvey, W. R. 1990. User's Guide for LSMLMW and MIXMDL (PC- McNitt, J. I., and G. L. Moody, J r . 1988. Milk intake and growth
2 Version). Ohio State Univ., Columbus. rates of suckling kits. J. Appl. Rabbit Res. 11:117.
Hudson, R., and H. Distel. 1990. Sensitivity of female rabbits to McNitt, J . I., and G. L. Moody, Jr. 1990. Effects of month, breed, and
changes in photoperiod as measured by pheromone emission. J . parity on doe productivity in southern Louisiana. J . Appl.
Comp. Physiol. A 167:225. Rabbit Res. 13:169.
Ismail, F.S.A. 1992. Effect of environmental temperature, season of Ozimba, C. E., and S. D. Lukefahr. 1991. Comparison of rabbit breed
the year, on performance of growing NZW rabbits. Agrartud. types for postweaning litter growth, feed efficiency, and sur-
Egyetem. 4:19. vival performance traits. J. Anim. Sci. 69:3494.
Jin, L. M., E. Thompson, and D. J. Farrell. 1990. Effects of tempera- Papp, Z., F. Kovacs, and P. Rafai. 1987. Impact of the climate on
ture and diet on the water and energy metabolism of growing large scale rabbit farming. 111. Effect of environmental temper-
rabbits. J . Agric. Sci. 115:135.
ature on the nest temperature and growing of suckling rabbits.
Lebas, F. 1970. Alimentation et croissance du lapereau sous la mere.
Magy. Allatorv. Lapja 42:209.
Rec. Med. Vet. 146:1065.
Roberts, J. D., and S. D. Lukefahr. 1992. Evaluation of Californian,
Lebas, F., and J. Ouhayoun. 1987. Incidence du niveau proteique de
Champagne D'Argent, New Zealand White and Palomino as
l'aliment, du milieu d'elevage, et de la saison sur la croissance
potential sire breeds: I. Postweaning litter traits. J. Appl. Rab-
et les qualites boucheres du lapin. Ann. Zootech. 36:421.
bit Res. 15:274.
Liu, M. F., M. Makarechian, and A.K.W. Tong. 1990. Relationships
Rollins, W. C., and R. B. Casady. 1960. A genetic analysis of
between measures of absolute growth rate and between meas-
ures of relative growth rate in growing meat animals. J . Anim. weaning weight of fryer rabbits. J. Anim. Sci. 19:1226 (Abstr.).
Breed. Genet. 108:187. Rollins, W. C., R. B. Casady, K. Sittmann, and D. B. Sittmann. 1963.
Lukefahr, S. D., P. R. Cheeke, and N. M. Patton. 1990. Prediction Genetic variance component analysis of litter size and weaning
and causation of litter market traits from preweaning and weight of New Zealand White rabbits. J. Anim. Sci. 22654.
weaning characteristics in commercial meat rabbits. J. Anim. Ross, S., P. B. Sawin, V. H. Denenberg, and M. X. Zarrow. 1961.
Sci. 68:2222. Maternal behavior in the rabbit: Yearly and seasonal variation
Lukefahr, S. D., P. R. Cheeke, N. M. Patton, and M. A. Grobner. in nest building. Behaviour 18:154.
1986. Genetic, maternal and seasonal factors affecting 56-day Rouvier, R. 1973. L'apport des croisements dans l'amelioration de
market weights in rabbits. J. Anim. Sci. 63(Suppl. 1):195 l'elevage du lapin de chair. In: Journ. Rech. Avicoles Cunicoles
(Abstr.). France. December 12-14.
Lukefahr, S., W. D. Hohenboken, P. R. Cheeke, and N. M. Patton. Simplicio, J. B., C. Cervera, and E. Blas. 1988. Effect of two different
1983a. Characterization of straightbred and crossbred rabbits diets and temperatures on the growth of meat rabbit. Proc. 4th
for milk production and associative traits. J . Anim. Sci. 57: World Rabbit Congr., Budapest, Hungary. 3:74.
1100. Sittmann, D. B., W. C. Rollins, K. Sittmann, and R. B. Casady. 1964.
Lukefahr, S., W. D. Hohenboken, P. R. Cheeke, and N. M. Patton. Seasonal variation in reproductive traits of New Zealand White
1983b. Breed, heterotic and diet effects on postweaning litter rabbits. J. Reprod. Fertil. 8:29.
growth and mortality in rabbits. J. h i m . Sci. 57:1108. Vogt, D. W. 1979. Selection experiment with domestic rabbits: I.
Lukefahr, S. D., W. D. Hohenboken, P. R. Cheeke, and N. M. Patton. Heritabilities of and genetic correlations between post-weaning
1984. Genetic effects on maternal performance and litter pre- average daily gain and gross feed efficiency. J . Hered. 70:421.