Joint linkage and association mapping of complex traits in shrub willow (Salix purpurea L.)

Abstract

Background and aims: Increasing energy demands and the necessity to reduce greenhouse gas emissions are key motivating factors driving the development of lignocellulosic crops as an alternative to non-renewable energy sources. The effects of global climate change will require a better understanding of the genetic basis of complex adaptive traits to breed more resilient bioenergy feedstocks, like willow (Salix spp.). Shrub willow is a sustainable and dedicated bioenergy crop, bred to be fast-growing and high-yielding on marginal land without competing with food crops. In a rapidly changing climate, genomic advances will be vital for the sustained improvement of willow and other non-model bioenergy crops. Here, joint genetic mapping was used to exploit genetic variation garnered from both recent and historical recombination events in S. purpurea.

Methods: A panel of North American naturalized S. purpurea accessions and full-sib F2S. purpurea population were genotyped and phenotyped for a suite of morphological, physiological, pest and disease resistance, and wood chemical composition traits, collected from multi-environment and multi-year replicated field trials. Controlling for population stratification and kinship in the association panel and spatial variation in the F2, a comprehensive mixed model analysis was used to dissect the complex genetic architecture and plasticity of these important traits.

Key results: Individually, genome-wide association (GWAS) models differed in terms of power, but the combined approach, which corrects for yearly and environmental co-factors across datasets, improved the overall detection and resolution of associated loci. Although there were few significant GWAS hits located within support intervals of QTL for corresponding traits in the F2, many large-effect QTL were identified, as well as QTL hotspots.

Conclusions: This study provides the first comparison of linkage analysis and linkage disequilibrium mapping approaches in Salix, and highlights the complementarity and limits of these two methods for elucidating the genetic architecture of complex bioenergy-related traits of a woody perennial breeding programme.

Keywords: Biomass; GWAS; QTL; bioenergy; breeding; phenology; wood composition; yield.

Fig. 1.

Significant GWAS hits by trait class in the association panel. Highlighted points signify significant SNP–trait associations (P < 1 × 10⁻⁶) within each class, and are coloured according to the corresponding trait class in the key (top left).

Fig. 2.

LOD support intervals of QTLs for biomass-related traits anchored to the F₂S. purpurea linkage map. Linkage groups are labelled according to respective S. purpurea v1.0 physical chromosomes. Bars to the right of each linkage group represent LOD support intervals for respective trait QTLs. Only QTLs for a sub-set of traits with significant LOD support (LOD >4.1) are shown. Horizontal lines within support intervals indicate the peak LOD position. Support interval bars are coloured according to their phenotypic class (see the key).