Genomic selection of resistance to acute hepatopancreatic necrosis disease in Litopenaeus vannamei

Abstract

Acute hepatopancreatic necrosis disease (AHPND) is a major bacterial disease caused by Vibrio parahaemolyticus (VP_AHPND) that results in significant economic losses to Pacific white shrimp Litopenaeus vannamei aquaculture. Breeding of disease-resistant broodstocks is regarded as a fundamental strategy for solving the disease problem. This study evaluated the feasibility of genomic selection (GS) for enhancing the AHPND resistance of L.vannamei. Following the VP_AHPND challenge, the survival time was utilized as a phenotypic measure to assess the AHPND resistance. The reference population consisted of 447 genotyped and 288 no-genotyped individuals with phenotypes. The candidate population consisted of 206 uninfected and genotyped individuals. The liquid chip “Yellow Sea Chip No.1” with 40 K single nucleotide polymorphism (SNP) generated genotypes for the reference and candidate populations. Heritabilities of survival time obtained using the average information REML method with a relationship matrix (A or H) and an animal model were 0.16 ± 0.06 and 0.22 ± 0.07, respectively. Predictive accuracies of pedigree-based best linear unbiased prediction (PBLUP), single-step genomic BLUP (ssGBLUP), and single-step Bayesian regression (ssBR) were evaluated using random, full-sib and between-family cross-validation. Compared to PBLUP, the predictive accuracies of ssGBLUP and ssBR for the survival time increased by 15.00 % and 67.50 % in random cross-validation, and 13.33 % and 53.33 % in full-sib cross-validation, respectively. All methods exhibited a reduction in predictive accuracy when moving from the random cross-validation or full-sib cross-validation to the between-family cross-validation. The predictive accuracies of Genomic BLUP (GBLUP) and Bayesian A (BayesA) were evaluated using full-sib cross-validation. Compared to GBLUP, the predictive accuracy of BayesA increased by 33.33 %. ssBR was used to predict the candidates' genomic estimated breeding values (GEBVs). 80 candidates with high GEBVs, 86 with medium GEBVs, and 40 with low GEBVs were selected to breed offspring with a spectrum of resistance levels. In the VP_AHPND challenge test, the high resistance offspring had mean survival times that were 10.15 % longer than the moderate resistance offspring and 102.01 % longer than the low resistance offspring. This study validates the practicality and success of GS in improving resistance to AHPND in L.vannamei.