Field and glasshouse experiments
In 2017, individual lines from the six DH populations were planted as 1-sqm plots in Katanning (Kat), Wongan Hills (WH) and South Perth, while four DH populations, namely BW, SpM, SpB, and Sp7A were also planted in 4-litre pots in a glasshouse at Murdoch University. To identify the young microspore stage, the parental lines, together with other varieties, were grown in Muresk in 2019. Field-grown plants were rainfed under standard agronomic management practices, whereas adequate water and fertilizer were provided to plants grown in pots in the glasshouse (Zhang et al., 2013).
  1. The field experiments with the six DH populations were conducted at three locations across Australia representing distinct environments including 1716 plots at Narrabri (NSW), 1884 plots at Muresk (WA) and 2316 plots at Williams (WA). Based on the seed availability, a total of 1060 DH lines represented six DH populations including 96, 180, 257, 161, 202 and 165 lines for the DH populations BW, G7A, Sp7A, SpB, SpM and St3B, respectively. The majority of the DH lines (>96%) were replicated two times at one or more than one of the three locations, more than 77% of DH lines were replicated two times at two or more than two of the three locations, and a very small proportion of the DH lines (< 4%) were unreplicated at the three locations due to seed availability. Eight parental lines and three control varieties were also utilised in each of the field experiments. The partially replicated experiments were designed using the experimental design tool DiGGer in R (Coombes, 2018).
  2. Frost and growth stage phenotype
Given the importance of establishing the plant growth stage during the frost events, anthesis date and plant growth stages during ear emergence were recorded for all populations in Williams and Muresk during the 2018 trials. In the Narrabri trials, the Zadoks stages around 50-60 were recorded. Days to anthesis data were also recorded in the glasshouse, in Wongan Hills and Katanning, in 2017. In 2018, days to maturity were estimated for each plot at the Williams site as follows: on 31 October, based on the plant maturity performance, the plant maturity scoring was entered as 0 (mature) to 15 (based on an estimate of 15 days to maturity).
In 2018, probably because of high rainfall and low soil temperatures in August, the average anthesis time was delayed by 15 to 20 days in the WA field trials. Flowering time ranged from 25 September to 12 October in Williams, and from 13 September to 6 October in Muresk. The lowest temperatures (-1.1 and 0.3oC, respectively) were recorded on 14, 15 and 16 September in Williams, and 15 and 16 September in Muresk (Fig. 1). Those temperatures lasted 300 and 130 mins in Williams and Muresk, respectively. The low temperatures during heading and anthesis caused frost damage in all six DH populations. The highest frost damage was to spikes (Fig. 1). Ten heads were picked from each plot for frost assessment around 20 days after anthesis. Three parental lines (Bethlehem, Mace, and Westonia) and two control lines (Wyalkatchem, Yitpi) from three anthesis windows were used, and 7 to 14 plots for each line were randomly selected for floret sterility assessment (Supplementary Fig. 1a). Frost damage was scored visually by a single person using a 0-15 scale (0: no frost damage; 1-2: a proportion of 0.5-5% of spikes damaged by frost within a plot (95-99% of grain remaining); 3-4: 5-20% damaged (80-95% of grain remaining); 5-6: 20-40% damaged (60-80% of grain remaining); 7-8: 40-60% damaged (40-60% of grain remaining); 9-10: 60-80% damaged (20-40% of grain remaining); 11-12: 80-90% damaged (10-20% of grain remaining); 13-15: 90-100% damaged (0-10% of grain remaining) (Supplementary Table 5). The correlation coefficient between visual frost damage and floret sterility was 0.79 (P <0.01) (Supplementary Fig. 1b).