Quantitative genetic variation of resistance against Piscirickettsia salmonis in Atlantic salmon (Salmo salar)
Artículo de publicación ISI Piscirickettsiosis (Piscirickettsia salmonis) is one of the diseases that cause large economic losses in Chilean salmon industry. Genetic improvement of disease resistance represents one strategy for controlling infectious diseases in farmed fish. However, knowledge of wh...
| Published in: | Aquaculture |
|---|---|
| Main Authors: | , , , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
Elsevier
2013
|
| Subjects: | |
| Online Access: | https://doi.org/10.1016/j.aquaculture.2013.08.009 https://repositorio.uchile.cl/handle/2250/120218 |
| Summary: | Artículo de publicación ISI Piscirickettsiosis (Piscirickettsia salmonis) is one of the diseases that cause large economic losses in Chilean salmon industry. Genetic improvement of disease resistance represents one strategy for controlling infectious diseases in farmed fish. However, knowledge of whether genetic variation exists for piscirickettsiosis resistance is needed in order to determine the feasibility of including this trait into the breeding goal. Using data from a challenge test performed on 2601 Atlantic salmon (Salmo salar) from 118 full-sib groups (40 half-sib groups) we found significant genetic variation for resistance to piscirickettsiosis. We used a cross-sectional linear model (CSL) and a binary threshold (probit) model (THR) to analyze the test-period survival, a linear model (LIN), Cox (COX) andWeibull(WB) frailty proportional hazardmodels to analyse the day at death, and a survival score (SS) model with a logit link to analyze the test-day survival. The estimated heritabilities for the different models ranged from 0.11 (SS) to 0.41 (COX). The Pearson and Spearman correlation coefficients between fullsib families estimated breeding values (EBVs) from the six statistical models were above 0.96 and 0.97, respectively. We used different data subsets, splitting the entire dataset both at randomand by tank, in order to predict the accuracy of selection for eachmodel. In both cases COX (0.8 and 0.79) and CSL (0.76 and 0.71)models showed the highest and the lowest accuracy of selection, respectively. These results indicate that resistance against P. salmonis in Atlantic salmon might be genetically improved more efficiently by means of using models which take both time to death and data censoring into account in the genetic evaluations. |
|---|