Genetic variation for carcass quality traits in cultured sea bass (Dicentrarchus labrax)

¹ IFREMER, Laboratoire de Recherche en Pisciculture marine, chemin de Maguelonne, 34250 Palavas-les-Flots, France
² Ecloserie marine de Gravelines, voie des Enrochements, 59820 Gravelines, France
³ INRA, UMR 1313, Génétique Animale et Biologie Intégrative, 78350 Jouy-en-Josas, France
⁴ SYSAAF, Section aquacole, Station SCRIBE, Campus de Beaulieu, 35042 Rennes, France

Corresponding author: eric.saillant@usm.edu

Received: 29 August 2008
Accepted: 6 February 2009

Abstract

Genetic parameters for carcass quality traits were estimated in 27 families of sea bass (3 dams × 9 sires factorial mating design), raised mixed in the same tanks starting before hatching. Offspring parentage was determined a posteriori using 6 microsatellite loci. Carcass quality traits were recorded at 818 days post fertilization (mean standard length: 32.6 ± 3.1 cm). Genetic parameters were estimated from the sire half sib variance and covariance components. Heritability of body weight (BW) and carcass processing traits (standardized to body weight) percent head weight (Head%), percent viscera weight (Viscera%) and percent visceral fat weight (VisceFat%) were relatively high ranging from 0.48 ± 0.15 (Viscera%) to (Head%); the estimate of heritability for fillet yield (Fillet%) was lower (0.25 ± 0.10) but was significantly greater than zero. Body weight was positively correlated to Fillet%, Viscera%, and VisceFat%, and negatively correlated to Head%. These results indicate that these carcass processing traits can be modified by directional selection and that selection for greater BW would lead to an increase of Fillet%, Viscera% and VisceFat% and a decrease of Head%.

Muscle lipid content (MuscleLipid%) was determined using two indirect methods: measurements with a Torry Fish Fatmeter^® (TorryLipid) and determination of the percentage of dry matter content (MuscleDry%) via desiccation. Both measures were highly correlated to chemical measurements of MuscleLipid%. Regression analysis indicated a superior predictive value of TorryLipid suggesting that MuscleLipid% may be evaluated via rapid, non lethal measurements with a Torry Fish Fatmeter. Heritability estimates of TorryLipid and MuscleDry% differed significantly from zero (TorryLipid: 0.28 ± 0.12, MuscleDry%: 0.36 ± 0.14) indicating that MuscleLipid% could be lowered by directional selection. TorryLipid and MuscleDry% were weakly correlated to body weight and carcass processing traits suggesting that simultaneous improvement of MuscleLipid% and other carcass quality traits may require definition of multi-trait selection indices.