Perspectives on genome mapping and marker-assisted breeding of eucalypts

Abstract

In Eucalyptus and forest tree breeding in general, the generalised application of molecular markers for directional selection is still an unfulfilled promise. In highly heterogeneous eucalypts, while conventional quantitative trait loci (QTL) mapping has revealed useful markers that are currently exploited in within-family selection tactics, only a more direct linkage disequilibrium mapping approach will likely uncover population-wide applicable marker–trait associations. Due to the very low-range linkage disequilibrium seen in Eucalyptus, a whole-genome association approach does not seem immediately feasible. Association studies based on candidate genes have been started in Eucalyptus, but only small proportions of the variation have been accounted for by such genes to be exciting news to breeders. Selection of candidate genes for direct manipulation or association studies based on their presumed biochemical role is not an easy task even for well-defined phenotypes and/or known metabolic pathways. Going from phenotypes to genes by a forward genomics approach based on an integrative expression–QTL mapping route, should prove to be a powerful way to choose target genes for marker-assisted selection. In this context at least two possibilities have recently emerged. The first one is to use high-performance genotyping technologies that would allow sufficient throughput and low cost for association genetic analysis of thousands of genes at a time. The second one would be to have access to a whole genome sequence so that candidate genes in a fine mapping interval delimited by flanking markers could be mined, reannotated and then analysed in association mapping. A fully public draft of the E. grandis genome will be sequenced by the US Department of Energy within the next few years following a proposal submitted by the International Eucalyptus Genome Network (EUCAGEN). As this genome project advances and more powerful tools become accessible, the true challenge for understanding and manipulating the complex nature of important traits in Eucalyptus will depend to a large extent on our ability to accurately phenotype trees, analyse the overwhelming amount of genomic data and translate this into truly useful molecular tools for breeding.

Keywords: Eucalyptus; genomics; molecular breeding

Southern Forests 2008, 70(2): 69–75