The developments in Norwegian sheep breeding since the early 1990s are reviewed. For the largest breeding population, the Norwegian White Sheep, results are presented for both genetic and phenotypic changes. Of the nine traits that make up the aggregate genotype, the largest gain per year, in per cent of the corresponding phenotypic average, was found for carcass grade (1.66%) and carcass weight (0.99%), number of lambs born at 1, 2 and 3 years of age (0.32% to 0.60%) and the maternal effect on weaning weight (0.26%). For fat grade, a genetic deterioration was estimated. This may be due to the too small weighting of this trait in the aggregate genotype and the true genetic parameters being somewhat different from the estimates in the prediction of breeding values. For lamb as well as ewe fleece weight, genetic change was close to zerointerpreted as mainly a correlated response to other traits in the aggregate genotype. Data for the two traits of fleece weight were, respectively, selected and few. Thus, phenotypic change was calculated for all traits except for fleece weight, and in addition for number of lambs at weaning, being indirectly selected for through number of lambs born. For all traits, with the exception of fat grade, advantageous phenotypic change was estimated. For weaning and carcass weight, the phenotypic change was less than the genetic change, while the opposite was observed for carcass and fat grade and number of lambs born. The latter traits can be more easily controlled by environmental actions, and the results thus exemplify the interdependency between environmental and genetic change.

Bias and inflation in genomic evaluation with the single-step methods have been reported in several studies. Incompatibility between the base-populations of the pedigree-based and the genomic relationship matrix (G) could be a reason for these biases. Inappropriate ways of accounting for missing parents could be another reason for biases in genetic evaluations with or without genomic information. To handle these problems, we fitted and evaluated a fixed covariate (J) that contains ones for genotyped animals and zeros for unrelated non-genotyped animals, or pedigree-based regression coefficients for related non-genotyped animals. We also evaluated alternative ways of fitting the J covariate together with genetic groups on biases and stability of breeding value estimates, and of including it into G as a random effect. In a whole versus partial data set comparison, four scenarios were investigated for the partial data: genotypes missing, phenotypes missing, both genotypes and phenotypes missing, and pedigree missing. Fitting J either as fixed or random reduced level-bias and inflation and increased stability of genomic predictions as compared to the basic model where neither J nor genetic groups were fitted. In most models, genomic predictions were largely biased for scenarios with missing genotype and phenotype information. The biases were reduced for models which combined group and J effects. Models with these corrected group covariates performed better than the recently published model where genetic groups were encapsulated and fitted as random via the Quaas and Pollak transformation. In our Norwegian Red Cattle data, a model which combined group and J regression coefficients was preferred because it showed least bias and highest stability of genomic predictions across the scenarios.

No abstract

Alternative Norwegian sheep breeding schemes were evaluated by stochastic simulation of a breeding population with about 120 000 ewes, considering the gain for an aggregate genotype including nine traits and also the rate of inbreeding. The schemes were: a scheme where both young unproven rams (test rams) and proven rams (elite rams) are used in artificial insemination (AI scheme), a scheme with test rams in natural mating in ram circles and elite rams (from one and a half years of age) in AI across all flocks in the country (NMAI2 scheme), a scheme where, in addition to testing rams, the youngest elite rams (one and a half years of age) are also used in natural mating in ram circles, while older elite rams are used in AI (NMAI1 scheme), and a scheme, acting as a control, where both test and elite rams are used in natural mating (NM scheme). Within the NMAI-and AI-schemes, experimentation was performed for percent ewes inseminated to elite rams v. test rams (EM%), numbers of ewes inseminated per elite ram (EAIn), and numbers of ewes mated per test ram by natural service (TNMn) or by AI (TAIn), respectively. With a restriction on the rate of inbreeding (<0.8% per generation), the AI scheme gave similar gain to the NMAI2 scheme (and about 40% more than did the NM scheme). Less gain was generated by the NMAI1 scheme, but it was still considerably more than for the NM scheme (about 25%). In the AI scheme, relatively few ewes (200/300) should be inseminated to each test/elite ram, and a low EM% should be chosen (10%). In the NMAI schemes, TNMn should be relatively high (40 to 50), combined with average and somewhat larger than average EAIn (NMAI2: 700 ewes, NMAI1: 900 ewes), and EM% medium (30%).

BackgroundThe main aim of single-step genomic predictions was to facilitate optimal selection in populations consisting of both genotyped and non-genotyped individuals. However, in spite of intensive research, biases still occur, which make it difficult to perform optimal selection across groups of animals. The objective of this study was to investigate whether incomplete genotype datasets with errors could be a potential source of level-bias between genotyped and non-genotyped animals and between animals genotyped on different single nucleotide polymorphism (SNP) panels in single-step genomic predictions.ResultsIncomplete and erroneous genotypes of young animals caused biases in breeding values between groups of animals. Systematic noise or missing data for less than 1% of the SNPs in the genotype data had substantial effects on the differences in breeding values between genotyped and non-genotyped animals, and between animals genotyped on different chips. The breeding values of young genotyped individuals were biased upward, and the magnitude was up to 0.8 genetic standard deviations, compared with breeding values of non-genotyped individuals. Similarly, the magnitude of a small value added to the diagonal of the genomic relationship matrix affected the level of average breeding values between groups of genotyped and non-genotyped animals. Cross-validation accuracies and regression coefficients were not sensitive to these factors.ConclusionsBecause, historically, different SNP chips have been used for genotyping different parts of a population, fine-tuning of imputation within and across SNP chips and handling of missing genotypes are crucial for reducing bias. Although all the SNPs used for estimating breeding values are present on the chip used for genotyping young animals, incompleteness and some genotype errors might lead to level-biases in breeding values.