Short communication: Including genomic information increases accuracy of breeding value estimation in a small outdoor...

To optimise organic pig breeding, we studied the effect of including genomic information for predicting breeding values (EBVs) in a small organic pig population. The recorded traits were the number of functional teats (n = 16 494), BW at birth (n = 36 995) and on day 10 (n = 29 744), and litter size on day 0, 4, or 11 (n ≈ 5 900 litters). Genomic information from 18 929 SNPs was available for 1 394 pigs, including Landrace sows and boars, and Yorkshire x Landrace crossbred sows. Throughout the study, a purebred- and crossbred correlation of 1 was assumed. Univariate mixed models that either included pedigree information or both pedigree and genomic information were used to estimate parameters and EBV. The prediction accuracy of the EBV was based on a forward prediction of data from the final 11 months of data. For the number of functional teats, the direct additive genetic component was considered, whereas for the other traits, the maternal genetic component was considered. The accuracy of predicting EBV for individuals without their own phenotype increased for all traits when genomic information was included, especially when the heritability was low. In conclusion, including genomic information can improve EBV prediction accuracy, which can optimise the genetic improvement in small breeding programmes.