Moderación de costes y superficies mínimas

In rabbit production, as in pig or poultry production, the scheme for genetic improvement and its diffusion is pyramidal (Figure 4).

Figure 4. Rabbit industry breeding scheme

In the peak of the pyramid, called the nucleus of selection, is situated the specialized lines that are the base for obtaining the superior animals for production. The step immediately below corresponds to multiplication. This step serves as a link between the nucleus and the production farms. Unlike pigs, the multiplication companies in rabbits are not very common, being farmers who more often produce the does for production (Baselga and Blasco, 1989). Thus, these farmers have the advantage of reducing the costs of animals and diminishing the health and adaptation problems (Gómez et al., 1998).

In general, rabbit lines used for genetic improvement can be considered as maternal or paternal lines, which are used for meat production, following a three-way crossbreeding scheme. The first cross is between two maternal lines for the production of the crossbred female, which is used as doe stock for production on commercial

farms. The crossbred does are mated or inseminated to bucks of a paternal line to obtain the rabbits for fattening (Baselga, 2004).

Selection for rapid growth rate has been largely introduced to develop paternal lines to modify the whole pattern of growth, feed efficiency, and tissue composition, thus affecting carcass and meat quality traits (Khalil and Al-Saef, 2008).

Selection in paternal lines is done or should be done to improve the feed conversion ratio (FCR) because it is a critically economically important trait in meat rabbit production (Armero and Blasco, 1992; Cartuche et al., 2013). However in practice, post-weaning growth is an effective criterion of selection because it is very easy to record and it has a negative and favourable genetic correlation with feed conversion ratio (Piles et al., 2004a). The average daily gain is the preferred trait for selection during the post-weaning period, because this trait is less affected by common litter effects than individual weights at specific ages (Khalil and Al-Saef, 2008), which is also moderately correlated with FCR. Moura et al. (1997) stated that selection based on an index including both growth rate and feed conversion ratio would be more efficient for improving feed efficiency than selecting solely for growth rate. Other selection criteria considered recently in genetic improvement programs in paternal lines are the selection for residual feed consumption (Larzul and Rochambeau, 2005) and post-weaning growth under food restriction (Garreau et al., 2008). Selection for growth traits has made the growing period shortened, and the degree of maturity of rabbits is lower at slaughter weight (Pascual, 2007), whichpossibly can have a negative effect on carcass and meat quality indicators. In this light, other selection criteria related to carcass quality have been proposed. For example, Milisits and Levai (2002) applied the technique of TOBEC (Total Body Electrical Conductivity) to improve carcass

composition. A technique of X-ray computerized tomography was utilized to assess in vivo body composition

The selection criteria usually considered in selection programs of maternal lines are commonly related with reproductive traits, such as the number weaned or number born alive (Rochambeau et al., 1988; Estany et al., 1989; Gómez et al., 1996; Capra et al., 2000; El-Raffa, 2000; Baselga and García, 2002). In some cases, selection criteria included litter size at birth and weight at nine weeks (Bolet and Saleil, 2002) in order to counterbalance the negative effect of large litter sizes on individual weights. The number of teats (Rochambeau et al., 1988) has been considered as another selection criteria. Attention has been also posed on traits directly related with the ability of the doe for lactating and nourishing of the progeny, such as weight at weaning (Garreau and Rochambeau, 2003), litter weight at weaning or total milk production (Al-Saef et al., 2008; Iraqi et al., 2008; Youssef et al., 2008). Also, selection for ovulation rate and uterine capacity has been successfully performed as indirect ways for improving prenatal survival and litter size in rabbits (Ibañez et al., 2004, 2006; Blasco et al., 2005; Mocé et al., 2005; Santacreu et al., 2005).

The selection methods in maternal lines are more complicated than for paternal lines because males do not affect litter size traits, and heritability of reproduction traits is so low that it is necessary to consider as many records as possible on the individual and on relatives during the genetic evaluation of the does and bucks (Baselga, 2004). Also, purebred performance in the nucleus is, in some cases, a poor predictor of future crossbred performance on commercial farms (Ibañez-Escriche et al., 2011) due to the existence of genotype by environment interaction. To overcome these limitations, Wei and van der Steen (1991) and Lo et al. (1993) proposed combining crossbred and purebred selection (CCPS), in which phenotypic data collected on crossbred relatives

are used for selection of purebreds. Thus, the application of a CCPS system, that until now has not been applied in rabbits, offers much potential as follows (Ibañez-Escriche and Noguera, 2013):

a) The selection response is proportional to the precision of the estimated breeding value (Falconer and Mackay, 1996). The adequate incorporation of crossbred animals can improve the precision of estimated breeding value, and therefore the selection response.

b) CCPS can reduce the generational interval (Wei and Van der Werf, 1994), whereupon the selection response per unit of time is higher.

c) Crossbred data permit for estimation of dominance effects and the combinatory aspects of candidates for selection. This information allows selection of candidates not only for their estimated breeding value, but also for their dominance or combinatory values.

d) CCPS enables the incorporation of important economic traits such as meat quality in genetic programs. Moreover, it is also important to note that these traits can be controlled in animals under commercial conditions, being different than the environmental conditions in the nucleus.