Concepts of Progeny Test Analysis
Forest tree breeders establish progeny tests primarily to predict breeding values of parents based on the performance of their offspring. The breeder then ranks parents according to their predicted breeding values (or some function of their predicted breeding values) to make selections. Progeny tests are generally replicated over a number of years and planting locations with the objective of sampling a wide range of environmental conditions. Because of large numbers of parents being evaluated, economic incentives to establish tests quickly, differential survival, etc. there is often a great deal of unbalance across all progeny tests in forest tree breeding programs. For example, all families are not always planted in the same year or location, and some families may be represented at only a few sites (Hatcher et al. 1981, Cotterill et al. 1983). This can make the goal of accurately and precisely ranking parents quite complex as data from a large number of progeny tests with diverse characteristics must be combined into one value representing the relative genetic worth of the parent. It is common in forestry to calculate a family mean of all offspring of a particular parent in a given test, and then somehow average those means across all tests in which the parent is represented. Different approaches to calculating the average, and different transformations of the data are sometimes utilized to deal with the problems associated with ’messy’ progeny test data.
KeywordsGenetic Correlation Growth Trait General Combine Ability Volume Growth Progeny Test
Unable to display preview. Download preview PDF.