Breeding Design for Atlantic Cod (Gadus morhua L.)

This thesis summarizes the development of the Icelandic family breeding program for Atlantic cod (Gadus morhua L.). The development of the base population began in the year 2003 when eggs were collected from wild brood fish from different spawning sites around Iceland. Approximately 350 viable families have been established and were later used to form the base population for the cod family selection program. Comparison of growth and early maturity of cod originating off the Northern and Southern coast of Iceland did not show significant differences. The eggs were hatched at the Marine Research Institute at Staður in Grindavík and reared in land based facilities. At the age of 12 months the juveniles were transferred to sea cages in Berufjörður at the Eastern coast and Ísafjarðardjúp at the Western coast of Iceland. Throughout the rearing an estimation of genetic (co)variation of economically important traits and genotype by environment interactions (G×E) among the two rearing locations were explored. Atlantic cod in Iceland shows high additive genetic variance for body weight (h2 = 0.34), suggesting that selection for increased body weight is likely to be successful similar to what has been shown in Atlantic salmon (Salmon salar L.) farming. Genetic correlation between body weight in Ísafjarðardjúp and Berufjörður was high (rG = 0.95), which indicates low G×E for the trait. However, low heritability for process yields such at fillet yields (h2 = 0.04) and the hepatosomatic index (h2 = 0.061) is less promising for obtaining genetic improvement. The strong phenotypic and genetic correlation (rG = 0.9) in body weight and maturation suggests that an increased growth rate will consequently lead to a higher proportion of mature individuals in the population. As a consequence, genetic manipulations to simultaneously increase growth and delay maturation may present a challenge. DNA profiling was used at the start of the cod breeding program for constructing pedigree. The result revealed highly imbalanced representation of individual families. The survival rate among full-sib families in the early stages was highly skewed. In this study, the effect of an imbalanced family structure on breeding was explored. A stochastic model was used to simulate genetic gain in an imbalanced versus a balanced family structure. Balanced design had a significantly higher gain for normally distributed traits and binary trait for all comparisons. The accuracy of estimated breeding values was not significantly different between balanced and imbalanced design. However the estimated breeding values were significantly higher for selected parentage fish in balanced design.

Author dc:creator

• Theódór Kristjánsson 1966-

Contributors dc:contributor

• Landbúnaðarháskóli Íslands

• Fiskeldi
• Þorskur
• Erfðatækni
• Kynbætur

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/1946/19399

OAI identifier oai:identifier

oai:skemman.is:1946/19399