Maxwell Bloomfield1, James Hunt1, Ben Trevaskis2, Kerrie Ramm2, Jessica Hyles2, 3,1 La Trobe University, AgriBio Centre for Agribioscience, 5 Ring Rd, Bundoora, VIC, 3083, M.Bloomfield@latrobe.edu.au,
2 CSIRO Agriculture and Food, GPO Box 1700, Canberra, ACT, 2601,
3 The Plant Breeding Institute, University of Sydney, 107 Cobbitty Rd, Cobbitty, NSW, 2570.
Ensuring wheat flowers at an optimal time minimises combined damage from frost, drought and heat stresses and maximises yield. Predicting flowering time of diverse cultivars across varying sowing dates and environments is crucial for attaining optimal yields, however there is no reliable way to do this for newly released cultivars. Photoperiod1 and Vernalisation1 genes are the major drivers of development in wheat and molecular markers have been developed to identify alleles at these loci. Allelic information has been used to parameterise phenology models, but it remains uncertain how much variation in flowering time can be explained by alleles of these major genes. We grew 13 elite commercial wheat cultivars, selected for diverse phenology and thus allelic variation of the major genes, and 13 near-isogenic lines (NILs) with matching alleles to quantify how much of the variation in time to flowering could be explained by the major genes. The experiment was conducted in four controlled environments (17 or 8-hour photoperiod, ± vernalisation) at a constant temperature of 22°C. NILs explained 97% of variation in time to flowering of elite cultivars under long days without vernalisation, 62% under short days without vernalisation, and less under short and long days with vernalisation (51% and 47%). Long days significantly accelerated time to flowering in all genotypes, while vernalisation hastened flowering in 17 genotypes. Results indicate allelic information of the major genes is not enough to parameterise an accurate model to predict flowering time under field conditions. Further investigation into genetic drivers of development and their interactions with environment are required before a genetically derived parameter estimate model can accurately simulate flowering time.