Genomic Selection in a Multi-Generation Breeding Population of Maritime Pine

The maritime pine breeding programme follows a recurrent selection scheme from a base population of 370 plus-trees (G0) selected during the 1960ies by exploring the forest in South-Western France. The main population has cycled through three generations (G0, G1 and G2), with more than 5,000 families tested over 500 ha of field trials. Genomic selection (GS) is of first interest for tree breeding as a way to reduce selection cycle length, to increase the selection intensity and better monitor genetic diversity along generations.

Our main objective is to develop a GS proof-of-concept consisting in evaluating the possibility to build a prediction equation for mandatory traits in order to explore key factors that will drive the implementation of GS in maritime pine. Given the rapid drop of linkage disequilibrium in this outbred species (within few hundreds bp), the size of its genome (24 Gb/C) and the number of markers available (few thousands), we have designed a reference population of a limited effective size (about 25) which is composed of 710 G2 trees as well as all their ancestors: 46 G0 and 62 G1 trees. This multi-generation population was genotyped with 4,100 informative SNPs. Breeding values for growth and wood quality were available from a global analysis of the breeding population (more than 350,000 trees phenotyped).

In this poster we detail the methodology considered for designing the reference population and present results regarding the accuracy of GS we can expect.