Genetic mapping of biomass yield in three interconnected Miscanthus populations

Improving biomass yield is a major goal of Miscanthus breeding. We conducted a study on one interspecific Miscanthus sinensis × Miscanthus sacchariflorus F₁ population and two intraspecific M. sinensis F₁ populations, each of which shared a common parent. A field trial was established at Urbana, IL during spring 2011, and phenotypic data were collected in 2012 and 2013 for fourteen yield traits. Six high-density parental genetic maps, as well as a consensus genetic map integrating M. sinensis and M. sacchariflorus, were developed via the pseudotestcross strategy for noninbred parents with ≥1214 single-nucleotide polymorphism markers generated from restriction site-associated DNA sequencing. We confirmed for the first time a whole-genome duplication in M. sacchariflorus relative to Sorghum bicolor, similar to that observed previously for M. sinensis. Four quantitative trait locus (QTL) analysis methods for detecting marker-trait associations were compared: (1) individual parental map composite interval mapping analysis, (2) individual parental map stepwise analysis, (3) consensus map single-population stepwise analysis and (4) consensus map joint-population stepwise analysis. These four methods detected 288, 264, 133 and 109 total QTLs, which resolved into 157, 136, 106 and 86 meta-QTLs based on QTL congruency, respectively, including a set of 59 meta-QTLs common to all four analysis methods. Composite interval mapping and stepwise analysis co-identified 118 meta-QTLs across six parental maps, suggesting high reliability of stepwise regression in QTL detection. Joint-population stepwise analysis yielded the highest resolution of QTLs compared to the other three methods across all meta-QTLs. Strong, frequently advantageous transgressive segregation in the three populations indicated a promising future for breeding new higher-yielding cultivars of Miscanthus.