Proceedings of the World Congress on Genecs Applied to Livestock Producon, 11. 923 Using GBLUP selection for increased Piscirickettsia salmonis (SRS) resistance in Atlantic salmon (Salmo salar) T. Kristjánsson 1 , K. G. Dodds 2 , J. C. McEwan 2 , R. Brauning 2 , R. M. Anderson 2 , T. C. van Stijn 2 , & S. M. Clarke 2 1 Stofnfiskur, Staðarberg 2-4, Hafnarfjörður IS-221, Iceland; theodor@stofnfiskur.is (Corresponding Author) 2 AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel, New Zealand; Summary In 2013, Atlantic salmon juveniles were challenge tested for Piscirickettsia salmonis (SRS) in VESO in Norway. The juveniles belonged to year class 2012 (YC-2012) comprising of 100 full and half sib families. In 2016 a sib-group of YC-2012 (non-challenged) was selected as breeding candidates to create an offspring generation year class 2016 (YC-2016) which was then challenge tested in VESO in 2017. Tissue samples from both challenged and non- challenged groups of YC-2012 were sent to AgResearch, New Zealand, for DNA extraction and genotyping by sequencing (GBS) using PstI. The KGD method (Dodds et al., 2015) was used for estimating the genomic relationship matrix (GRM) allowing a GBLUP estimation of genomic estimated breeding value (GEBV) of SRS resistance in YC-2012 breeding candidates. Two models were applied: Test survival (TS) and Day of death (DD). Estimated heritabilities for the models differ from 0.28 (TS) to 0.41 (DD). Accuracy in the non- challenged group was estimated as 0.66 and 0.68 for TS and DD model respectively. GEBV of YC-2012 breeding candidates was correlated with the families’ performance measured on offspring in 2016 challenge test in VESO. The correlation was 0.62 and 0.68 for TS and DD models respectively. Keywords: Genomic selection, genotyping by sequencing, disease resistance, Salmon salar: Introduction SRS is caused by the intracellular Gram-negative bacterium, Piscirickettsia salmonis, first identified in Chile and later in Canada and several European countries (Corbeil et al., 2005). SRS causes high losses in Chilean Atlantic salmon and Coho salmon industry (Cvitanich et al., 1991). Mortality rates have been reported to be 30-90% among Coho salmon (Corbeil et al., 2005). Treatments with antibiotic have provided a partial solution to the problem, but do not give control of the disease. However, genetic selection could be a long-term solution to the problem. Genomic selection (GS) has been shown to increase genetic gain and accuracy of selection substantially even with sparse SNPs (4K) (Sonesson & Meuwissen 2009; Ødegård et al., 2014) in Atlantic salmon populations. Therefore, selecting for traits with low heritability and relatively few records per trait measured (e.g. carcass and disease resistance traits), can benefit from GS (Villanueva et al., 2011). In aquaculture populations where large full-sib families are produced, GS captures within family variation allows prediction of the Mendelian sampling term. Several platforms (e.g. beadarrays) are available for single nucleotide polymorphism (SNP) genotyping, but genotyping by sequencing (GBS) is a low-cost approach to genotyping. Since SNP information from the GBS approach can be sparse, methods have