Heritabilities and genetic analysis of milk yield and components in crossing project of Saudi rabbits with Spanish V-line

K A Al-Sobayil, A H Al-Homidan, M H Khalil and M A Mehaia

Department of Animal Production and Breeding, College of Agriculture and Veterinary Medicine, Al-Qassim University,
Buriedah P.O.Box 1482, Saudi Arabia
maherhkhalil@yahoo.com

Abstract

A four-year crossbreeding project involving Spanish maternal line called V-line (V) and Saudi Gabali (G) rabbits was carried out to produce six genetic groups of V, G, ½V½G, ½G½V, ¾V¼G and ¾G¼V. Inter se matings for genetic groups of ½V½G, ½G½V, ¾V¼G and ¾G¼V were also practiced. Milk yields (MY) at intervals of 0-7 days (MY7), 7-21 days (MY21), 21-28 days (MY28), and 0-28 days (TMY) and milk components (MC) at 14 days of lactation (fat, protein, lactose, ash, and total solids) were evaluated for 2141 litters of 854 does fathered by 142 sires and mothered by 351 dams. A repeatability animal model was used to estimate the corresponding parameters as the heritabilities, the differences between line V and Gabali in additive direct effects (G^I_V-G) and maternal additive effects (G^M_V-G). The individual (H^I) and maternal (H^M) heterosis, and direct recombination effect (R^I) were also estimated.

Heritabilities for MY traits were moderate, ranging from 0.18 to 0.22, while they were low or moderate and ranging from 0.09 to 0.28 for MC. The positive estimates of G^I_V-G for MY (5.6-14.5%) and MC (4.0-18.7%) were significantly high and in favour of V-line does. Estimates of G^M_V-G were in favour of V-line dams; being 222 g, 0.67% and -0.08 for MY21, total solids in milk, and fat in milk, respectively. All estimates of H^I for MY and MC were positive and most of them were significant; ranging from 9.7 to 22.7 % for MY traits (P<0.05-0.001) and 6.0 to 15.8% for MC traits (P<0.05-0.01). Similar to the trend of H^I, the estimates of H^M for MY and MC were positively moderate and ranging from 7.4 to 15.2 % for MY traits and 1.9 to 8.3% for MC. The ranges in percentages of reduction in direct heterosis were negligible and ranging from 2.2 to 5.3% for MY traits and 3.4 to 9.6% for MC. In practice, crossbred does and dams involving V-line genes in their constitutions gave favourable heterotic effects on milk traits and therefore these crossbred does and dams can produce and lactate efficiently under hot climatic conditions.

Key words: Animal model, crossbreeding, heritability, milk yield and components, rabbits

Introduction

Genetic diversity of rabbit breeds in terms of milk production offers the opportunity to increase the efficiency of doe productivity through crossbreeding. To date, genetic analysis concerning milk yield (MY) and components (MC) for crossbred rabbits raised in hot climate countries are scarce. Further, milk production and milk components are expensive to record. Other traits that are associated with them, are more directly related with production and are easier to record, such as litter size at birth or at weaning (Estany et al 1989; Rochambeau et al 1998; Gomez et al 1996; Capra et al 2000; Garcia et al 2000), litter size at weaning and individual weight at weaning (Moura et al 2001), litter size at weaning and individual weight gain (Gomez et al 2000), and litter weight at weaning (Salaun et al 2001; Garreau and Rochambeau 2003). These traits, in a great part, can replace milk yield as criteria for selection. However, selecting does directly for yields and components of milk should be of considerable interest and importance (Garreau and Rochambeau 2003; Baselga 2004; Garreau et al 2004).

The objective of our study was to estimate direct (G^I_V-G) and maternal (G^M_V-G) additive effects, direct heterosis (H^I), maternal heterosis (H^M) and direct recombination effects (R^I), as well as heritabilities, for MY and MC traits in a crossbreeding project involving Spanish V-line rabbits and Gabali Saudi rabbits.

Materials and methods

The four-year crossbreeding project was started in October 2000 in the experimental rabbitry, College of Agriculture and Veterinary Medicine, El-Qassim region, King Saud University, Saudi Arabia.

Breeding plan, management, farm environment and feeding

Rabbits used in this study represent one desert Saudi breed (Gabali, G) and one exotic breed (Spanish V-line, V). The breeding plan permitted simultaneous production of ten genetic groups. Distribution of number of litters weaned in these genetic groups across different years of kindling is presented in Table 1.

A total number of 2541 litters were born by 854 does, fathered by 142 sires and mothered by 351 dams. The bucks were randomly assigned to mate the does naturally with the restriction to avoid the matings of animals with common grandparents. Young rabbits were weaned at four weeks of age. Rabbits were raised in a semi-closed rabbitry. Breeding does and bucks were housed separately in individual wired-cages. All cages are equipped with feeding hoppers and drinking nipples. In the rabbitry, the environmental conditions were monitored; temperature ranged from 20 to about 32 °C, the relative humidity ranged from 20 to 50 % and photoperiod was 16L: 8D. Rabbits were fed a commercial grower pelleted diet during the whole experimental period, which lasted 16 weeks of age. On dry matter (DM) basis, the diet contained 18.5% crude protein (CP), 8.0% crude fiber (CF), 3.0% ether extract (EE) and 6.5% ash. Feed and water were available ad libitum.

Milk yields and components

Milk yields (MY) were recorded during the first seven days (MY7), 7-21 days (MY21), 21-28 days (MY28) and for the total 0-28 days (TMY). Litter weight at birth (LWB) and litter weight at weaning (LWW) were also recorded. Milk yield of does was recorded using weigh-suckle-weigh method. MC of fat, protein, lactose, ash and total solids (g/100g) were also estimated. In the evening of the day prior to collection, the kits were separated from their mothers to prevent suckling for a period of 12 hours before sample collection in the next morning. Milk samples were collected manually by gently massaging the mammary gland after two minutes of injection with 0.1 ml of oxytocin hormone to enhance maximum contraction of myoepithelial cells. Samples were taken per doe per litter in the morning of the 15^th day of lactation. The samples were cooled and transferred immediately to the laboratory for chemical analysis. Milk sample for each litter born per doe was analysed for total solids, and ash according to procedures outlined in AOAC (1980). Fat was determined by Gerber method as described by Case et al (1985), nitrogen by the standard micro-Kjeldahl method (AOAC 1980). A nitrogen conversion factor of 6.38 was used to calculate protein content. Lactose was determined by subtraction.

Statistical analysis

Repeatability animal model (in matrix notation) used for analysing milk traits was (Boldman et al 1995):

y= Xb + Z_au_a + Z_pu_p + e

Where
y = vector of observed lactation trait for does,
b= vector of fixed effects of genetic group of doe (ten levels; see Table 1), year-season of kindling (one year season every three months), and physiological status of the doe (five levels depending on the parity order and lactation state at the moment of insemination: 1 for nulliparous, 2 for primiparous lactating, 3 for multiparous lactating, 4 for primiparous non-lactating, 5 for multiparous non-lactating);
u_a= vector of random additive effect of the does and sires,
u_p= vector of random effects of the permanent environment (permanent non-additive effect);
X, Z_a and Z_p are the incidence matrices relating records to the fixed effects, additive genetic effects, and permanent environment, respectively; and e= vector of random residual effects.

Variance components of direct additive effects, permanent environmental effects and errors were estimated by DFREML procedure using the animal model (Boldman et al 1995). The inverse of the numerator relationship matrix (A^-1) was considered; Var(u_a)= As²_a, Var(u_p)= Is²_p and Var(e)= Is²_e. Heritabilities for different traits were computed from variance components using the following equations:

The variance components estimated before should be used to solve the model and get solutions for the ten genetic groups (D_i, i= 1 to 10) and the corresponding variance-covariance matrix of the errors. This information should be used to compute later the estimable functions D_i-D₂ and their variance-covariance matrix of errors.

Genetic model and estimation of crossbreeding effects

The Dickerson's genetic model (Dickerson 1992) was used to explain the performances of the different genetic groups. The genetic model take into account the direct additive effects of the line or breed (e.g. G^I_G, for Gabali), the maternal additive effects (e.g. G^M_V, for line V), the individual (H^I) and maternal (H^M) heterosis between Gabali and line V and the direct recombination effect (R^I). Because of linear dependence between the equations of the fixed factors, estimable functions used are D_i-D₂ that allow to estimate the following combinations of parameters of the Dickerson's genetic model: G^I_V-G = G^M_V - G^I_G; G^M_V-G =G^M_V- G^M_G, H^I, H^M and R^I. Table 2 shows the coefficients of the previous combination of parameters for the functions D_i-D₂ that are used with their variance-covariance matrix of errors to get generalised least square estimates of G^I_V-G, G^M_V-G, H^I, H^M and R^I, their standard errors and the corresponding Student's t-test.

Results and discussion

Actual means and variations

To characterize the experiment phenotypically, means, standard deviations, minimum and maximum values for milk traits are presented in Table 3.

However, wide phenotypic variations in all traits were observed. >From producers point of view, milk yield and milk components showed moderate lactational performances particularly for hot climate areas. In hot countries, little lower values for milk yield were reported by Khalil and Afifi (2000) and much lower values by Lahari and Mahjan (1984), Khalil (1994) and Abd El-Aziz et al (2002). Cowie (1969) reported that Dutch does (as a small breed) produced less milk in the first six weeks of lactation than New Zealand White does (3820 gram v. 6940 gram). Lukefahr et al (1983) in USA found that New Zealand White was superior to Californian rabbits in lactational yield. Lahari and Mahajan (1984) in India reported that differences in daily milk yield at 21 days of age among Grey Giant (GG), Soviet Chinchilla (SC), White Giant (WG), New Zealand White (NZW) and Russian Angora (RA) were not significant but GG had the highest yield (189 g per day) followed by RA (147g), SC (139g), WG (134g) and NZW (116g). El-Sayiad et al (1994) with New Zealand White (NZW) and Californian (CAL) rabbits in Egypt stated that the differences between the two breeds in fat, protein, lactose, ash and energy of milk were not significant; the estimates were 14.0, 13.6, 1.9, 2.1% and 87.9 kJ/100g in NZW and 14.0, 14.3, 2.0, 2.2% and 89.9 kJ/100g in CAL for fat, protein, lactose, ash and energy of milk, respectively. Such discrepancies among reports may be due to the strain of breed used, breed x environment interaction or different experimental methods.

Total differences between genetic groups

Deviations of each genetic group from Gabali (D_i-D₂) for different milk traits are presented in Table 4. These deviations are interesting to show the global performance of the V-line, the Gabali breed and their different crosses in order to identify their possibilities to be used as pure stock or as a simple cross or to be used as synthetic line. For all milk traits, V-line rabbits recorded better lactational performance compared to Gabali rabbits (Table 4). Clear differences among the ten genetic groups were notified for fat, protein and total solids in milk.

The highest percentages of total solids (fat, lactose and protein) were recorded for (¾G¼V)² group (31.0 %), while the least values were recorded for G group (27.3%). In most cases (MY7, MY21, MY28 and TMY), genetic group of (¾V¼G)² or (¾G¼V)² gave higher milk yield and components compared to the other genetic groups (Table 4). The above mentioned results indicate that involving V-line genes in crossbreeding program with Gabali rabbits was associated with an improvement in the lactational performance of the crossbred does obtained.

Additive and permanent environmental effects and heritability estimates

Ratios of variance components of direct additive effect (heritabilities, h²) and permanent environment (p²) to the phenotypic variances are presented in Table 5.