|Title||Managing inbreeding in selection and genetic conservation schemes of livestock|
|Source||Wageningen University. Promotor(en): E.W. Brascamp; T.H.E. Meuwissen. - S.l. : S.n. - ISBN 9789058086631 - 175|
Animal Breeding and Genomics
|Publication type||Dissertation, internally prepared|
|Keyword(s)||vee - inteelt - selectief fokken - genetische bronnen van diersoorten - selectiecriteria - paring - veredelingsprogramma's - livestock - inbreeding - selective breeding - animal genetic resources - selection criteria - mating - breeding programmes|
|Categories||Races, Selection, Genetics|
This thesis deals with the definition of selection and mating criteria for animal breeding populations under selection and for genetic conservation populations, especially emphasizing on populations with small effective sizes that have known pedigrees.
The thesis can be divided into four main parts. Firstly, Chapters 2 and 3 deal with selection algorithms that manageDF for populations under selection with overlapping generations and random mating. Secondly, Chapters 4 and 5 deal with non-random mating schemes in combination with selection algorithms for discrete and overlapping generation structures, respectively. Thirdly, Chapters 6 and 7 deal with algorithms that minimizeDF in small and endangered populations. In Chapter 6,DF is minimized for populations with overlapping generations. In Chapter 7,DF is further reduced by using frozen semen of sires from the less related base population. Fourthly, Chapter 8 deals with methods to select against genetic defects while restrictingDF in populations with increased frequency of diseased alleles.
CHAPTERS 2 AND 3
In Chapter 2, a method is presented that maximizes the genetic merit of the selected animals while limiting the average coancestry of a population with overlapping generations after the current round of selection. For populations with overlapping generations, account has to be taken for previous and future use of animals of certain age-classes. Contributions within and over age-class were found by iteration. Inputs are Best Linear Unbiased Predicted (BLUP) breeding values of the selection candidates, and the relationship matrix of all animals. Output is the optimal number of offspring of each candidate. Computer simulations of dairy cattle nucleus schemes showed that the predefined rate of inbreeding was achieved. At the same rates of inbreeding, the dynamic selection method obtained up to 44% more genetic gain than truncation selection for BLUP breeding values. The advantage of the dynamic method over BLUP selection decreased with increasing population size and with less stringent restriction on inbreeding. In Chapter 3, the method of Chapter 2 was compared to a similar method that firstly optimized the distribution of parents within and thereafter over age classes per sex. It yielded significantly lower annual genetic gain, fewer animals selected and longer generation intervals, but maintained the rate of inbreeding closer to its constraint. The use of conventional relationships and of augmented relationships, which do not depend on the level of inbreeding, resulted in very similar breeding schemes, but the use of augmented relationships avoids correction of the current level of inbreeding. When optimising per generation, the generation interval was shorter compared to a scheme where an analogous annual restriction was in place and the annual rate of genetic gain was higher.
CHAPTER 4 AND 5
In Chapter 4, the effect of non-random mating on genetic gain was compared for populations with discrete generations. Mating followed a selection step where the average coancestry of selected animals was constrained, while genetic gain was maximized. Minimum coancestry (MC), Minimum coancestry with a maximum of one offspring per full-sib family (MC1) and Minimum variance of the relationships of the offspring (MVRO) mating schemes resulted in a delay in inbreeding of about two generations compared to Random, Factorial and Compensatory mating. At the sameDF , genetic gain was up to 22% higher for the MC1 and MVRO schemes compared to Random mating schemes. The effect of non-random mating was largest for small schemes or for schemes with a stringent restriction onDF.MC1 yielded the highest genetic gain in almost all selection schemes, with a lower computational cost than MVRO. In Chapter 5, MC1 mating scheme was compared with random mating schemes for populations with overlapping generations and a restriction onDF . When sires were progeny tested, these progeny tested bulls were selected instead of the young bulls, which lead to increased generation intervals, increased selection intensity of bulls and increased genetic gain (35% compared to a scheme without progeny testing). The effect of MC1 decreased for schemes with progeny testing. MC1 mating increased genetic gain 11-18% for overlapping and 1-4% for discrete generations, when schemes with similar rate of inbreeding and genetic gain per generation were compared.
CHAPTER 6 AND 7
In Chapter 6, a method that minimizes the increase of coancestry of parents and optimizes the contribution of each selection candidate for populations with overlapping generations is presented. When survival rate equalled 100%, only animals from the oldest age class were selected, which maximized the number of parents per generation, slowed down the turn over of generations and minimized the increase of coancestry across sublines. However, the population became split into sublines separated by age classes, which substantially increased inbreeding within sublines. Sublines were prevented by a restriction of selecting at least one sire and one dam from the second oldest age class, which resulted in an L times lowerDF , where L equals the average generation interval of sires and dams. Minimum coancestry mating resulted in lower levels of inbreeding than random mating, butDF was approximately the same or somewhat higher. For schemes where only the oldest animals were selected,DF increased with 18-52% compared with the proposed method. In Chapter 7, the advantage on the average coancestry level of selecting not only the least related sires from the oldest age-class, but also sires from the second oldest age-class is presented. By selecting sires from generation zero only, all genes will eventually descend from the founder sires and all genes from the founder dams are lost. By selecting sires from generation zero and one alternatively, also some genes of the founder dams will be conserved and the average coancestry level was approximately 20% lower than for a scheme, where only the oldest sires were used. TheDF was zero at equilibrium for both schemes. Dams could be used for one generation and sires unlimited, because the amount of frozen semen very large relative to the small population sires. Population size was 6, 10 and 20 and the schemes were symmetric with respect to the sexes.
Increased inbreeding will result in increased frequency of detrimental alleles. In Chapter 8, different genetic models and evaluation systems to select against a genetic disease in populations with discrete generations are compared. When using optimum contribution selection with a restriction onDF of 1.0% to select against a single gene, selection directly on DNA-genotypes needed 2.0 generations to half the frequency of the disease allele with additive effects and a population with 100 new-born animals. When only phenotypic records were available, selection on BLUP or on genotype probabilities calculated by segregation analysis (SEGR) needed 1.0 or 2.0 generations longer to half the frequency of the disease allele when allele effects were additive or recessive, respectively. Smaller schemes or schemes with a more stringent restriction onDF needed more generations to half the frequency of the diseased allele or the fraction of diseased animals. SEGR and BLUP were approximately equally efficient under both single gene and polygenic inheritance models, suggesting that efficient selection against a disease is possible without knowing its mode of inheritance.
In conclusion, by taking account of the relationships of the selected group of selected animals, inbreeding is controlled in the selection and mating criteria presented in this thesis. This principle was extended to populations with overlapping generations, which makes the methods useful for practical selection and genetic conservation populations with known pedigree. Only in well-controlled breeding schemes, the optimum contributions of each selection candidate will be realized, some deviations may be corrected in later rounds of selection. The optimized contributions and thus also the restriction onDF affected the structure of the breeding schemes, e.g. whether progeny tested animals were selected or not. In general, older and more parents were selected with a more stringent selection onDF . Due to such dynamic adaptations of the breeding schemes, genetic gain reduced only little whenDF was lowered. Non-random mating can improve the family structure of the population under selection, thereby further increase the genetic gain. For the genetic conservation schemes, the contributions per family could not become completely equalized, because of the overlapping structure of the generations, resulting in a level of average coancestry being somewhat higher than the theoretical minimum. Use of frozen semen from sires of the oldest generations could reduceDF to zero and reduce the average level of coancestry in combined in situ / ex situ conservation schemes.