Animal breeding goals are normally based on traits and characteristics linked to long-term profitability. To reach a breeding goal, it is important to assess each breeding animal’s ability to produce offspring that will perform as closely as possible to the goal. Genetic merit can be described as the animal’s breeding value, or its value as a parent in the desired traits.
Early livestock farmers based their predictions on visual observation. Simply put, the biggest calves were more likely to breed the biggest or heaviest progeny. A refinement came later with the measurement of accurate weights; calves were weighed and the heaviest kept back for breeding.
This was followed by assessing animals in contemporary groups. The effects of specific non-genetic factors such as age at measurement and sex and age of the dam were taken into account. In such a contemporary group, each animal was rated with its contemporaries and the performance differences were expressed in terms of an index, the norm being 100 with variation above or below the norm numerically rated.
Dr Japie van der Westhuizen
However, using indices to predict breeding value has limitations. Firstly, it offers no real guarantee that superiority within a specific contemporary group will reflect the situation in another group. A young bull born in the spring of 2012 that outperforms his peers on a specific farm in the 2013 weaning season is not guaranteed to do the same in a different herd, on a different farm or in any other contemporary group.
Other obvious variations could be linked to his dam’s ability to care for him. His superior weaning weight could be the result of his genetic merit for growth, or it could be due to the nurturing ability of his dam. Similarly, when averaging the progeny performance of a bull, his performance could be linked to the top cows to which he was mated or to his own genetic merit.
These problems led to the development of more sophisticated methods to predict the genetic merit of breeding stock. The first real breakthrough in prediction accuracy was the development of BLUP breeding values. The principle of BLUP (usually called EBVs) relies on the approach of selecting individuals that change the mean production level of the population from which they are selected. BLUP maximises the correlation between predicted values and ‘true’ genetic merit.
Related animals share the same parts of DNA – Related animals share parts of their genetic code that may be responsible for genetic value. Trait differences are expressed in phenotype and performance. For example, a herd sire will contribute and share 50% of his genetic material with each of his offspring. If this sire is used widely over many herds and produces progeny over several years, he serves as a link between different contemporary groups. His progeny in each group will represent half his genetic worth. Taking this to its logical extreme, virtually all the animals in a breed are somehow related and share some genetics with others.
Only part of the performance superiority can be transferred to progeny – All recorded differences among animals, even in the same peer group, are the result of two important influences. The first relates to differences due to variations in genetic merit – the genes and gene interactions on their chromosomes. The second is the influence of different environments.
A simple example would be to list the known non-genetic influences, from a beef calf’s perspective, on weaning weight. These include data on age, sex, age and parity of the dam, season and year of birth, farm or location, and any treatment differences from those of the other animals in the peer group. Taking all this and the possibilty of random genetic superiority into account, only the genetic superiority transferable to the progeny can be expressed. This is achieved by using the heritability estimate for each trait when predicting its breeding value.
Traits are ‘linked’ through common parts of the DNA
All traits recorded in livestock are correlated in one way or another. Selection for one trait in a population will generally affect genetic merit for the other traits. Modern BLUP models take such genetic correlations into account. The models have an advantage in breeding value prediction accuracy in sequential culling (when different animals are culled at different ages), sex limited traits (scrotal circumference, female fertility, maternal behaviour and milk production), animals that were still too young to be measured for specific traits, or when the relevant trait (such as feed intake, marbling or eye muscle) is difficult or expensive to measure for all animals.
A herd sire contributes 50% of his genetic material to each of his offspring. A widely used bull siring progeny in different years serves as a link between groups.
These factors make BLUP breeding values reliable and accurate in genetic merit prediction. Many examples exist where this methodology has been used to achieve genetic change in different livestock breeds and production systems. Genomic selection is a step up in enhancing the accuracy of breeding value prediction. BLUP breeding values rely on the concept of ‘family selection’, as the genetic relationship between animals is based on average values.
Two collateral half-siblings are assumed to have 25% of their genes in common. However, fragments of DNA in the genes separate relatively independently and randomly when sperm and egg cells are formed. Some half-siblings, in our example, can have a relationship of more than 25% while others have less genetic material in common. By using the additional information in the genetic code of every animal, the genome can establish the ‘true’ relationship between all animals.
Where BLUP used family information, genomically-enhanced BLUP adds a new dimension, namely the specific separation of genes within animal families in the population. Obviously this is possible with BLUP alone when breeding animals have numerous measured progeny. Genomic selection accelerates the process, making information available at a younger age.Genomically enhanced breeding value predictions, known as GEBVs, start off with a higher prediction accuracy than those of BLUP. However, the genomic information of a sire with highly accurate EBVs will be less useful to a breeder, as the sire’s progeny would have already contributed to establishing his value as a sire.
The Greatest benefit
Genomic selection works best in combination with BLUP and enhances the accuracy of prediction in specific cases.
The dairy industry benefits enormously from genomic selection. Previously, years would pass before a potential breeding sire’s BLUP could be established because milk production, its components and daughter fertility could be based only on his female offspring. Combining genomic information and BLUP means that the prediction accuracy on a young potential AI sire equals that of an older sire with ten measured daughters.
Genomics has created many new opportunities. Selection for traits in which recording is limited to a single sex, such as milk production, daughter fertility, mastitis resistance, maternal ability (in dairy and beef cattle and sheep), litter size and semen quality, are some examples. Other traits that can benefit include those only measurable on dead animals (carcass and meat properties), those that are expensive or difficult to measure (feed intake, ovulation rate, disease resistance and parasite resistance) and those that can only be established late in the animal’s life (longevity and productive life).
Keeping a level head
Although genomics has taken the livestock breeding world by storm, it does not eliminate the need to record selection traits. This remains crucial, as the relationship between genetics and breeding values changes over time and is usually population-specific within the breed and even within the country.
It is important always to keep defined and logical breeding objectives in mind when using EBVs, whether genomically enhanced or not. In some cases, bigger or heavier might not be better. Establishing and executing these objectives is a science involving knowledge of the genetic correlation between different traits, their economic contribution, and the environmental constraints imposed on the production system.
Contact Dr Japie van der Westhiuzen on 082 331 9923 or email [email protected]