ABSTRACT

Over the past decades, genetic improvement has been a very important contributor to improved productivity of pigs, as outlined in the previous chapter. Most of that improvement has been achieved by selection of individuals (young males and females) for breeding based on estimated breeding values (EBVs) derived using phenotypes for important traits that have been recorded on the selection candidates themselves and/ or their close relatives (phenotype-based prediction, see top of Fig. 1). For that purpose, sophisticated statistical methods based on mixed linear models and best linear unbiased prediction (BLUP) have been pioneered in animal breeding. These methods optimally utilize all available phenotypes on the individual itself and its relatives in order to obtain the most accurate EBV of the individual, which can then be used to rank individuals to identify animals that should be used to breed the next generation.