Wheat production from sub-tropical, irrigated or high rainfall environments is a major contributor to world food security. The Northern region of Australia, with subtropical climate, summer rainfall and availability of tactical irrigation or high rainfall (contingent on seasonal variability) provides an opportunity to stabilise wheat production and income when most of the cropping-belt is under drought. Australian germplasm is generally developed in and for low rainfall environments. Lines with high yield potential and low lodging likelihood need to be developed. This paper discusses the complexity of finding variation for traits underpinning lodging in high yielding germplasm across the different stages of a pre-breeding program in the Northern region of the Australian cropping belt, from multi-environment testing to population development and marker discovery.

Screening local and imported germplasm in multi-location multi-year experiments in irrigated environments (above 7t/ha), we identified low-lodging lines with consistently higher-yield than currently-grown commercial cultivars. To identify these lines, the lodging-score threshold was set to represent a crop with 80% of the plot area at 70° or closer to the vertical. A subset of 50 low-lodging lines was subsequently phenotyped with a physiological approach to standability. Those with useful genotypic variation for traits related to stem and root failure moment and plant and shoot leverage were selected as donor parent for new populations in adapted commercial backgrounds. At a dedicated lodging nursery, we evaluated the BC1 progeny with high-throughput phenotyping methods newly-developed for this purpose. Overall family trends in lodging score were consistent across years and recurrent commercial parents. Two donors were identified as good at delivering low-lodging lines and also as better parents, despite their contrasting heights (ca. 80 vs. 105 cm). This trend could already be predicted from an evaluation of BC1 progeny in single rows bordered by the same short cultivar. A large number of lines was identified as having good lodging scores over two years and a proportion of those also out-yielded the recurrent parents in a one year evaluation (without undergoing prior selection for yield per se). Molecular marker development using an independent population segregating for lodging (Seri x Babax) has shown evidence of four significant (P<0.01) QTL for lodging score on chromosomes 3A-S, 4A, 4B and 5D, with Babax transmitting the reduced lodging allele. Further studies QTL validation studies are underway. The presence of Seri x Babax based QTL will be checked in a proportion of lines from the new populations.