Plant Genome III Conference Plant Genome III Conference
Wood density is an important determinant of the quality of Pinus radiata. Classical tree breeding approaches have been most successful in delivering gains for traits Uc stem diameter, which can be inexpensively measured on the 8-12,000 trees typically planted in progeny trials. Wood property traits, which are expensive to measure, have not been as amenable to genetic manipulation (even though heritability is higher) because the cost of accurately measuring thousands of trees has genetically been prohibitive. QTL analysis and marker aided selection show promise for increasing genetic gain in such traits over conventional tree breeding approaches. If markers linked to genes controlling wood density can be found, not only would the selection process be less expensive than conventional assessment, but selection could be carried out immediately after propagation, rather than at a mature age after substantial wood formation. This very early selection would make it possible to capture all the gain from conventional selection of parent clones for establishment in seed orchards, plus an additional gain in wood density from a tandem selection among potential seed orchard offspring. Selected offspring would be cloned, vegetatively multiplied and planted in operational forests. We have measured wood density of 927 offspring of the full-sib cross 850-55 x 850-96 and are using 188 which are selected from each tail of the approximately normal distribution to search for RAPD markers linked to wood density. We have identified 71 reliable markers (55 in a 1:1 Contest cross and 16 in a 3:1 configuration) from screening 525 primers. Of the 32 markers analysed so far, three markers showed wood density differences of one standard deviation (31.7 kg/m3) between trees with and without bands. On an independent set of 94 trees, however, none of these three markers showed significant differences, although the difference between rifts with and without bands was in the predicted direction. Until more markers are analysed and a map is constructed, it cannot be determined whether these putative linkages are small effect QTL'S, weak linkages to larger effect QTLs, or statistical artifacts.