Genetic diversity and gain : the concept of a status number
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
A trade-off always tends to exist involving genetic gains and selection intensity, on the one hand, and the remaining effective population size (usually known as Ne), on the other. A new approach is presented and analysed for different breeding situations, using stochastic simulations, in terms of mating designs and subline sizes, guiding breeders through a new concept of status number (Ns) and its trade-off with gain. Status number is defined as half the inverse of the average coancestry and depicts the current state of the population. The status number concept can easily be applied to deployment of different genotypes with unequal representation. Breeding schemes with small breeding groups are slightly more efficient in preserving status number through multiple generations than breeding schemes with large groups. Medium- to large-size breeding groups showed a comparatively small reduction in aggregated status number over generations but showed greater increases in gain compared with small groups. Inbreeding in small elites becomes so great that it is likely to cause fertility problems and disturb selection considerably. Small breeding groups will probably not be useful for a sustainable long-term breeding strategy. Substantial benefits on status number for subdividing the population into small breeding groups will only be seen after numerous generations. Selection schemes that maximise gain by unrestricted combined index selection will result in rapid inbreeding, and may not be sustainable in the long term. Selection procedures that place less emphasis on family information would best meet long-term diversity targets. However, gains may be too low for mating systems and selection procedures that do not include a between-family component, especially with low heritabilities. This is a good reason for using a large number of families as founders of the breeding population. Going from selection within only 0.5 or 1 available cross per parent per generation (made equivalent to within-family selection) to 2.5 crosses per parent (restricting the number of individuals chosen per full-sib family) resulted in substantial increases in genetic gain, depending on heritability. However, increasing the number of crosses per parent up to 2.5 does carry a modest penalty of increased coefficient of inbreeding and reduced status number. Higher levels of gain per unit of status number loss are obtained with a conservative within family selection strategy but to reach the same level of gain more cycles of breeding will be required. Effects of departures from assumptions (zero inbreeding coefficient and coancestry for the founders, genes being independently assorted, no mutation and interactions, or combinations from departures of the neutrality assumption) , singly and in various combinations will occur, meaning that calculations and predictions based on pedigrees will be biased. Future work will require modelling the effects for departures from the idealised assumptions and laboratory-based quantification of departures from some key assumptions.