Selection for more than one trait - created from Mind Map

Relationships between traits We find that some traits are related in a positive manner. Other traits are related in a negative manner. We measure the degree to which traits are related with a correlation coefficient This is a phenotypic correlation.

Genetic Correlations In animal breeding we are interested in how breeding values for different traits are related to one another, Usually genetic correlations and phenotypic correlations are similar in sign and magnitude. This means when we select for one trait we also make genetic changes in another. However, we often find that same genotypes will perform differently in different environments. This is not always the case. Known as genotype by environment interaction. Before we start breeding, we need to know if one trait is going to negatively correlate another. Calculate the correlated response to selection

Indirect selection In some situations it is difficult to measure the trait we wish to improve. It may be useful to select on a second, correlated trait, which is easy to measure. For example - prolificacy is only expressed in females. For example, scrotal circumfrance is genetically correlated to prolificacy and can easily be measured in males.

Methods of selecting for multiple traits Tandem Selection Breeder selects for one trait at a time. After selecting for one trait for a number of generations - selection is changed to another trait and so on Simple to use method. Not efficient if traits are negatively related. Independent culling levels A culling level is set for each trait that is to be improved. However, we do not take advantage of all the good genes in the population. Only animals that exceed the level for each trait are selected. An animal that exceeds in one trait but not the other would be culled and so we lose the trait that she exceeded in. Index Selection Most efficient method Selection upon an index score which combines the measure of performances of each trait, weighted optimally Allows animals that are extremely good in one trait to make a contribution to the breeding population I = A x (trait 1) + B x (trait 2)

Animals that are good in one trait tend to be good in the other related trait.

I = index score A is the index weight for trait 1 B is the index weight for trait 2

align_clippings_on_this_page_please

Selection for more than one trait

Annotations