Academic article

Genetic improvement of farmed tilapias: Response to five generations of selection for increased body weight at harvest in Oreochromis niloticus and the further impact of the project

Bentsen, Hans; Gjerde, Bjarne; Eknath, Ambekar E.; de Vera, Marietta S. Palada; Velasco, Ravelina R.; Danting, Jocedel C.; Dionisio, Edna E.; Longalong, Felicisima M.; Reyes, Ruben A.; Abella, Tereso A.; Tayamen, Melchor M.; Ponzoni, Raul W.

Publication details

Journal: Aquaculture, vol. 468, p. 206–217, 2017

Publisher: Elsevier

International Standard Numbers:
Printed: 0044-8486
Electronic: 1873-5622

Open Access: none


Selection for increased bodyweight at harvest in Nile tilapia (Oreochromis niloticus) was carried out for five generations from 1991 to 1996, as a part of the project Genetic Improvement of Farmed Tilapias (GIFT). The base
population for selection was composed of a mixture of various three- and four-way cross individuals descending
from four wild African strains and four farmed Asian strains. Methods for single pair mating, separate rearing of
full-sib groups and individual tagging and pedigree recordingwere developed. The parents for a new generation
were selected based on a selection index including theirown age corrected-bodyweight at harvest aswell as that
of their full- and half-sibs. Across generations, a total of 512 males were mated in a nested mating design to 941
females to produce 81,429 tagged progenies. Of those, 56,633 progeny had their bodyweight recorded at harvest after grow-out testing in a diverse range of farming environments. Estimates of the within generation realized response to selection across farming environments were obtained as the difference between the least squares mean performance of offspring of selected parents and of offspring of parents with average values of the selection index. The average realized response to selection per generation was 13.6% (range 9.0 to 20.1%), resulting in an accumulated response over five generations of 88% relative to the base population. A genetic trend analysis based on BLUP breeding values estimated across generations after the termination of the project suggested an accumulated response of 67%. The genetic composition of the synthetic population also changed during selection.
The proportion of ancestors from three of the wild African founder strains increased from 59.7 to 76.3%. The accumulated
coefficient of inbreeding was 7.1%. After the termination of the GIFT project, selection was continued in the GIFT population. The population has also been used as a genetic source in a number of similar public and private selection programs. A status review of the presently recorded dissemination of the genetic material and methodology is presented.
Statement of relevance: The GIFT project showed that five generations of repeated selection based on testing of
individually tagged and pedigreed individuals from a synthetic farmed population of Nile tilapia gradually increased the bodyweight at harvest. The total increase was 67–88%. Reports from several descending populations shows that the selection response has continued during N10–15 additional generations.