How does animal selection work? What levers should be used to make selection effective? Is it possible to select for any trait?
The aim of this article is to try and answer these questions in an accessible way, so that everyone can gain a better understanding of the principles of selection.
First of all, what is animal selection? More or less everyone has heard of Darwin and natural selection, which allows species to evolve by adapting to their environment. In the livestock sector, however, selection is geared towards adapting species to their rearing environment and to the performance expectations of the market. So it’s not a question, as many people imagine, of intervening directly at DNA level through genetic manipulation, but simply of selecting, on the basis of their performance in given conditions, the best animals to produce the next generation.
What data is needed for selection? The classic model of optimised selection requires knowledge of both the genealogy of each individual and its performance. The breeder must therefore implement a precise traceability of the family origins of each animal, in other words know its parents and ancestors, as well as its brothers and sisters and descendants. All these animals must also be measured individually. Once all this data has been obtained, mathematical equations are used to calculate the genetic values of each individual. These genetic values are then used to classify the animals and select the best ones. Genetic values make it possible to take into account not only each animal’s own performance but also that of its entire family, enabling a more accurate estimate to be made of its potential for improvement. This process is called indexing.
SET OF INDIVIDUALS ASSESSED = SELECTION BASE
Now that we know how selection works, let’s look at the different ways of increasing its effectiveness. The best way to do this is to look at the equation for annual genetic progress and consider its various components:
The numerator includes :
– Selection intensity (i): this is directly linked to the proportion of improvers retained at each generation.
– The accuracy of the estimate (ρ): we will not be able to classify individuals correctly if the measurement of their performance is not accurate.
– The variability of traits (σA): this component of genetic progress depends very much on the trait measured (live weight, yield, other) and also on the populations considered. The more different the individuals, the easier it will be to select the best. On the other hand, it should be noted that this is the only criterion on which the breeder cannot intervene.
The denominator is the interval between two generations (T). The faster the selected populations reproduce, the faster progress will be made.
From this we can deduce that there are several levers that enable selection to be effective:
– Breed a large number of animals to keep a small number of breeding stock.
– Use equipment and technologies to measure performance as accurately as possible.
– Maintain good variability in the population, taking care to keep as many families as possible.
– Produce the next generation as soon as possible.
All these factors mean that selection will be more or less effective depending on each trait. The first point to consider is the ability to measure the trait of interest accurately. In the case of egg weight, for example, all you need to do is measure the weight of each egg to obtain a precise measurement of the trait. But in the case of disease resistance, there are a number of complex issues to consider: how do you define resistance? What measurements need to be taken? How should these measurements be carried out? Once these questions have been answered, the variability of the trait needs to be studied. If all the individuals have exactly the same performance, it will be complicated to hope to make any genetic progress on this trait, as the offspring will have the same performance as their parents. Finally, the last point to study is the heritability of the trait. This heritability corresponds, at a population level, to the proportion of variability in performance that is genetic in origin and that can be passed on to the next generation. Of course, some performances do not depend, or depend very little, on the genes carried by the individuals themselves, but rather on their environment and the conditions in which they are reared. The choice of traits to select for is therefore essential, and must also take into account the expectations of the industry and, more generally, the expectations of society. Selecting animals with good butchery qualities helps to produce quality meat, and plays a major role in improving the feed efficiency of animals, which in turn helps to reduce the area of cereals grown for animal feed, in favour of human food.
