How to Start a Fish Breeding Programme

Av Marte Holstmark og Gunnar Klemetsdal (Translation: Ane Skjelfjord)

Broad genetic variation for breeding traits is necessary to ensure the success of fish breeding programmes. Simulations show that using fish from at least four sub-populations provides the greatest genetic gain.

Photo: Hans Magnus Gjøen

Fish farming in Southeast Asia and China can look back on 1000 years of history. However, selection on fish is a more recent practice. The first fish breeding programme was initiated as late as 1972, on Atlantic salmon in Norway, translating knowledge from general animal breeding into fish breeding. Since then, breeding programmes for numerous aquatic species (fish, prawns and shellfish) have been started in many countries world-wide. So far, however, there have been no studies of how such breeding programmes should be started. This was the objective of Marte Holstmark’s PhD thesis.

Typically, a fish breeding programme is started by forming a so-called base population, which is based on wild fish. A wild fish population often consists of several sub-populations, which may have different genetic properties. These traits are generally not well known. Furthermore, the kinship between the sub-populations is also uncertain. Under these circumstances, we used stochastic simulations to study the effect of increasing the number of sub-populations contributing to the base population, as well as the effect of different mating strategies in the base, on short and long-term genetic gain in breeding populations.

The simulations showed that sampling of fish from additional sub-populations increases the genetic variation for the breeding goal trait, both at early and later stages of the breeding programme. Thus, this increased the genetic gain obtained in the breeding program (genetic gain is proportional with genetic variation). The largest improvement was achieved by increasing from one to two sub-populations, and the effect decreased rapidly. Most of the potential genetic gain was achieved when sampling fish from at least four sub-populations, which is thus what we recommend when establishing new breeding programmes.

The simulations also showed that mating fish both within and between the sub-populations in the base gave less variations of the long-term genetic contribution, compared to mating only within sub-populations in the base. This implied a reduced risk of losing contributions from the sub-populations, which in turn could be important for future breeding goal traits.

Published: 01.05.07
Updated: 25.02.11
Printerfriendly version

Del med en venn:

Department of Animal and Aquacultural Sciences

P.O. Box 5003
N-1432 Ås

Phone: +47 64 96 51 00
Fax: +47 64 96 51 01

E-mail: iha@umb.no

Webmaster: Janne Karin Brodin

Additional information