Heritability estimates and genetic correlations for post-weaning body weight traits in Sudanese rabbits

K M Elamin, I A Yousif^*, M A Elkhairey and D M Mekki^**

Faculty of Animal Production, University of Gezira, Sudan
khalid1130@yahoo.com
* Faculty of Animal Production, University of Khartoum, Sudan
** Faculty of Natural Resources and Environmental Studies, University of Kordofan, Sudan

Abstract

Data on 233 progeny from two consecutive generations of Sudanese local rabbits were used to estimate heritability, and genetic, phenotypic and environmental correlations for post-weaning weekly body weights between 6 and 15 weeks of age.

In the first generation, estimates of heritability based on paternal half-sib analyses were moderate, ranging from 0.211 to 0.372. The results for the second generation showed that the estimates of heritability were low to moderate, ranging from 0.085 to 0.295. The genetic correlations in the first generation among growth traits were all positive, but were low to moderate for the correlations of body weights traits at early age with other weight traits, and fairly high for the correlations of body weight traits at late ages with the remaining traits.  Heritability estimates for post weaning growth traits were generally moderate, so improvement of this breed should be realized through genetic selection. The genetic correlations among growth traits revealed that selection can be practiced at any stage in the post-weaning period since improving body weight at any stage should improve the trait at other stages.

Key words: Baladi, environmental, growth, phenotypic

Introduction

Development and evaluation of sound breeding programs depend largely upon accurate knowledge of both environmental and genetic parameters. Environmental effects are responsible for an important portion of the rabbit’s productivity and these effects must be considered by researchers in the analysis of their data (Ferraz et al 1991). Heritability, which is a function of variance components, provides information about the genetic nature of a trait and is needed for genetic evaluation and selection strategies (El-Raffa 2005). Phenotypic variance for growth traits in rabbits is commonly reported to be overwhelmed by environmental effects attributable to the dam and/or litter, which may be due to the short time between weaning and marketing age (McNitt and Lukefahr 1996). This phenomenon may also account for the low heritability estimates and slow rates of genetic responses to selection for growth traits that have been reported (Lukefahr et al 1996). The genetic potential for improvement of rabbits is dependent largely on the heritability of the trait measured and its relationship with other traits of economic importance.

A better understanding of sources of random environmental variation that influence growth, and that can be controlled by management, is required to enhance effectively the amount of genetic variation that can be exploited through selection in rabbit populations.The application of an animal model is useful to partition trait variances due to direct additive genetic, common litter and within-litter (residual) environmental effects (Yossef et al 2009). Most of the discrepancies between estimates of heritabilities from different studies may be attributed to the differences in breed types of rabbit reared under particular environmental conditions during definite period of time, size of the data and variations in the statistical methods used (El-Zanfaly 1996). Narayan (1985) attributed these discrepancies to small sample size per generation, while Khalil and Soliman (1989) suggested that they are due to non-randomness in the distribution of daughters within sire group. Misztal (1990) pointed out that the accuracy of estimates of variance components is dependent on the choice of the data, methods and models. Estimates of heritability and repeatability for litter traits have a broad range literature estimates and the dissimilarity of estimates could reflect real differences among populations and environmental conditions, but it could be partly due to different methods and models used (Sorensen et al 2001).
 

Estimates of heritability for individual body weights tend to increase with age (Rochambeau 1988 cited by Rochambeau 1997). In contrast, Nofal et al (2005) concluded that low heritability values for litter size traits (h2 less than 0.2) indicate that litter size could be improved more readily by crossbreeding (between breeds or lines within breeds) as opposed to selection.  Khalil et al (1987) attributed higher heritability values in local rabbits compared to exotic breeds to the fact that exotic breeds were generally subjected to previous selection while local breeds were not.  Our study was conducted to estimate heritability, and genetic, phenotypic and environmental correlations between growth traits of local rabbits to determine the suitability of selection for achieving genetic improvement. 

Materials and Methods

Study location

This study was conducted in a rabbitry within the premises of the Extension and Rural Development Centre, Faculty of Animal Production, University of Gezira in Managil town, Gezira Province. The town is about 76 km west of Wad-Medani, which is located between the Blue and White River Niles (14.25N- 32.99 E). The average air temperature during the experimental period ranged from 16.88±.47 to 40.23±.64°C, whereas the relative humidity ranged from 34.3±2.18 to 66.1±2.18%. 

Experimental animals

The foundation stock consisted of 65 female, 25 male mature local rabbits. They were bought from five different localities in central Sudan to assure a high level of genetic variation. These animals were divided randomly into groups, each consisting of three dams and one sire; however, mating animals from the same area was avoided. Twenty sire families were used and the remaining animals were left to substitute mortalities. A total of 449 kits were reproduced in the first generation from 129 litters, of which only 168 reached the weaning age at 42 days and only 135 were used to provide data of  body weights for genetic analysis. The second generation was reproduced by dams and sires chosen at random from the first generation.  Sib matings were avoided. Data consisted of 216 kits born alive of which 106 reached weaning age at 42 days, but only 98 rabbits survived to 4 months of age and were used for matings and subsequent genetic analyses. 

Animal management

Experimental animals were housed in two buildings which were divided into 30 rabbit cages each of 1×1 m dimensions. Each pair mate was housed in a cage until pregnancy was confirmed. Animals were fed ad libitum a diet formulated from local ingredients with 16% crude protein and 2500 kcal/kg of DE.  Clean water was available throughout the experimental period.  Traits measured weekly were body weight..

Statistical analysis

Data consisted of 230 observations on animals with full records (133 from the first generation and 97 from the second generation), sired by 36 bucks and reared by 67 dams, which were analyzed to estimate genetic parameters using Harvey's (1990) LSMLMW program (Least Squares and Maximum Likelihood Procedure) to estimate variance and covariance components.

Y_ij = u + a_i+e_ij

Where:

Y_ij = observation on the j^th progeny of the i^th sire;

μ = overall population mean;

a_i = random effect of the i^th sire; and

e_ij = residual random error.

The program computed estimates of heritability and genetic correlations as: 

(h²) =         and rG =       

Where:

h² = heritability;

r_G = genetic correlation;

  = sire variance component;

  = error variance component; and

Cov (xy) = sire covariance component between traits x and y 

Results and Discussion

Estimates of heritability, and genetic, phenotypic and environmental correlations are provided in Tables 1 and 2 for  6 to 15 weeks of age for the first and  second generations combined. The results for the first generation revealed that the estimates of heritability were moderate and in the range of .211 to .372. These results agree with the findings of Akanno and Ibe (2005), Larzul et al (2004), and Enab (2001) involving New Zealand White (NZW) rabbits and Garcia and Baselga (2002) involving line V. Higher estimates were obtained by Estany et al  (1992) and Youssef (2004) involving NZW and Red Baladi (RB) rabbits and Umesh singh et al (2005) in India  but were lower than estimates reported by Lukefahr et al (1992) involving a crossbred population and El-Faky et al (2001) involving RB rabbits in Egypt. Genetic, phenotypic and environmental correlations between body weight traits were generally high with the exception of the genetic correlations of WT6 and WT7 with other weights at other ages. Further, genetic correlations in the first generation among growth traits were all positive, but were low to moderate for the correlations of body weights traits at early age with other weight traits, and fairly high for the correlations of body weight traits at late ages with the remaining traits. 

The results for the second generation (Table 2) showed that the estimates of heritability for body weights at all ages were low to moderate, in the range of .085 to .295 (the highest estimate was for WT6 and the lowest estimate was for WT12). Heritability estimates from sire components in the second generation for Wt 6-Wt11 were low, which agrees with Testik et al (1998), Ahmed et al (2007), Orunmuyi et al (2006), Sorensen et al (2001), Ferraz and Eler (1994), and Estany et al (1992). These same estimates are lower than the findings of Iraqi et al (2008), Ibrahim et al (2007), and Akanno and Ibe (2005). Estimates for weight at 12 to 15 weeks of age were moderate, which is in agreement with the findings of Estany et al (1992), Iraqi et al (2008), Abdou  et al (2006), and Youssef (2004) in NZW and Black Baladi (BB) and Enab (2001) and Elfaky et al (2001) in RB rabbit, but are lower than those reported by Akanno and Ibe (2005) and Hassan (2004). On the other hand these correlations were generally high in the second generation. This is in agreement with the results of Youssef et al (2009), Iraqi (2008) and Ibrahim et al (2007). Heritabilities were lower in generation two, this may be due to decreased additive variance  caused by inbreeding .

The genetic correlations in the first generation among the growth traits were all positive, but were low to moderate between the weights at earlier ages and fairly high among weights at older ages. On the other hand, these correlations were generally high in the second generation. This is in agreement with the results of Youssef et al (2009), Iraqi (2008) and Ibrahim et al (2007). 
 

The phenotypic correlations among the growth traits in both generations were high (> 0.5). These estimates are in agreement with the findings of Iraqi (2003) and Enab (2001) (0.98-0.79, but are higher than the estimates of Youssef et al (2009) (0.49- 0.30) and Umesh Singh et al (2005) (0.37- 0.22).  The environmental correlations among the growth traits in both generations ( between brackets below diagonal) were positive and exceedingly high, approaching one. These results agree with Iraqi (2008), but they were much higher than the results reported by Umesh Singh et al (2005) and Iraqi (2003).  

 

In conclusion, heritability estimates for post weaning growth traits tended to be high, therefore genetic improvement is possible through selection. The genetic correlations among growth traits indicated that selection can be practiced at any stage in the post weaning period since improving body weight at any stage leads to improvement of growth traits at other stages. 

References