R Culvenor1, K Smith2,3, J Forster3, N Cogan3, L Pembleton3 and J Sewell4
1CSIRO Agriculture and Food, GPO Box 1700, Canberra ACT 2601, firstname.lastname@example.org
2Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Private Bag 105, Hamilton, Victoria 3300
3AgriBio, Department for Economic Development, Jobs, Training and Resources, 5 Ring Road, Bundoora, Victoria 3083
4PGGWrightson Seeds, Ballarat Research Station, PO Box 1425, Ballarat MC, Victoria 3354
Genomic selection (GS) is a new form of marker-assisted selection which offers the potential for more efficient selection of quantitative traits such as yield in species with long selection cycles. GS requires accurate genotyping and phenotyping of a ‘training population’ which is then used to select on genotype in a breeding population. We describe a training population developed to test GS in phalaris and present seasonal herbage yield data from one of three evaluation sites. The population contains 290 seed-retaining, winter-active families based Holdfast GT, Advanced AT, Landmaster and a Holdfast-derived population. Progeny were evaluated with commercial cultivars from 2014 -16 at Canberra for herbage DM yield in rows and a subset also evaluated in swards. A high level of phenotypic variation was observed with high correlations between observations in the main growing season within and between years. Correlations between row and sward yield were 0.7-0.8 in the growing season and lower at other times when stem material was present. The data from this and other sites will be combined with marker information to assess the feasibility of using GS in phalaris breeding.