Response to five generations of selection for growth performance traits in rainbow trout (Oncorhynchus mykiss)

Authors: Leeds, Timothy - Tim; Vallejo, Roger; Weber, Gregory - Greg; GONZALEZ-PENA, DIANELYS - Oak Ridge Institute For Science And Education (ORISE); Silverstein, Jeffrey

Submitted to: Aquaculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/25/2016
Publication Date: 9/4/2016
Publication URL: https://handle.nal.usda.gov/10113/63157
Citation: Leeds, T.D., Vallejo, R.L., Weber, G.M., Gonzalez-Pena, D., Silverstein, J. 2016. Response to five generations of selection for growth performance traits in rainbow trout (Oncorhynchus mykiss). Aquaculture. 465:341-351.

Interpretive Summary: Growth performance is an economically-important trait in rainbow trout aquaculture. Typically, domestic rainbow trout are marketed at approximately 500 grams, but there is interest in marketing the fish at larger sizes (greater than or equal to 1 kilogram) and before sexual maturation (~2 years of age). We selectively bred a pedigreed line of rainbow trout for improved growth performance to and beyond traditional market size for five generations, and compared growth performance of the selected line to that of randomly-mated control lines. As a result of selection, age-specific body weights increased approximately 11.3% per generation, and select-line fish had greater condition factor (i.e., were more rotund) at market weight compared to rainbow trout from randomly-mated control lines. Body weights of fifth-generation rainbow trout from the selected line approached 1.4 kilograms at 13 months post-hatch. This study demonstrates the large and sustained potential for genetic improvement of growth performance in domesticated rainbow trout with only modest accumulation of inbreeding, and provides empirical selection response estimates that have direct inference to commercial breeding programs aiming to select for growth to larger market sizes.

Technical Abstract: A pedigreed rainbow trout population (~100 families per generation) was selected for five generations to improve growth performance to the standard ~500-gram US market weight and beyond (greater than 1 kg). Body weights (BW) were recorded each generation at 5, 8, 10, and 13 months post-hatch. Selection was applied to a two-trait index of estimated breeding values for 10-month body weight (i.e., approximate market weight) and thermal growth coefficient between 10 and 13 months, and both traits were given equal weighting in the index. Objectives of this study were to estimate direct selection responses for BW traits and thermal growth coefficient, and correlated response for condition factor traits. Heritability estimates were 0.28 plus or minus 0.08 to 0.43 plus or minus 0.14 for BW between 5 and 13 months post-hatch, 0.12 plus or minus 0.03 for thermal growth coefficient, and 0.40 plus or minus 0.04 and 0.36 plus or minus 0.04 for condition factor at 10 and 13 months post-hatch, respectively. Genetic correlations among the four BW measures were highest between adjacent BW measures (0.54 plus or minus 0.19 to 0.82 plus or minus 0.05) and decreased as time between measures increased. Genetic correlation between thermal growth coefficient and 10-month BW was low (0.18 plus or minus 0.13), and genetic correlations between condition factor and BW measured at the same age were moderate (0.39 plus or minus 0.09 to 0.45 plus or minus 0.08). Compared to randomly-mated control lines, selection response in the selected line averaged 10.6% (5-month BW), 9.8% (8-month BW), 12.7% (10-month BW), 11.9% (13-month BW), 2.1% (thermal growth coefficient), 3.4% (10-month condition factor), and 3.3% (13-month condition factor) per generation. The rate of inbreeding averaged 0.86% per generation in the selection line, with a cumulative inbreeding estimate of 4.5% in the last generation. The low heritability of thermal growth coefficient, and low genetic correlation with 10-month BW, resulted in only modest selection response for thermal growth coefficient. Larger heritabilities of BW measured at fixed ages, and large genetic correlations between serial BW measures, suggest that direct selection on BW will be more favorable than selection on TGC. This study demonstrates the large and sustained potential for genetic improvement of growth performance in domesticated rainbow trout with only modest accumulation of inbreeding, and provides empirical selection response estimates that have direct inference to commercial breeding programs aiming to select for growth to larger market BW.