Take home messages

• The changing Australian climate is shifting rainfall later in the cropping season
• The development of a short-season, winter-sown commercial wheat would provide growers with greater flexibility, sowing, and agronomic options
• Winter-sown wheats have the potential to increase yield and grower profitability in mild climatic years
• Early vigour and phenologically-quick wheats are expected to outperform slower growing and developing wheats when sown late.

Background

The Australian climate is trending toward a hotter drier future, with less rainfall in autumn (Bureau of Meteorology 2022). As wheat growers traditionally wait until the colloquially termed ‘autumn break’ rainfall to sow their crops, a delay in its arrival is resulting in an increase in the number of crops that are dry sown instead of waiting (Unkovich 2010). Dry sowing refers to the practise of sowing crops into soil without adequate moisture to germinate the seeds. Alternatively, growers are having to delay sowing for an eventual rainfall. In either scenario, this can result in a crop that germinates outside of its optimum window. As a result of this delayed germination, the crop will not reach anthesis within the optimum flowering period (OFP). The OFP is the period of time after winter when the crop is at least risk of frost or heat and drought stress damaging sensitive floral parts and impacting final grain yield (Flohr et al., 2017). The OFP varies yearly and spatially depending on local climatic conditions but remains a sound concept to maximise grain yield from a phenological perspective. Furthermore, double knock herbicide strategies to combat the increasing number of herbicide resistant weeds consume large amounts of time at the start of the season which can also delay sowing (Widderick and McLean 2018). The wheat industry in Australia does not possess any wheat varieties that are suited to winter sowing, as all current varieties are too phenologically slow to time flowering in the OFP. There have been few Australian studies that have focused on shorter season wheats, with the majority investigating earlier sowing for a longer growing season (Fletcher et al., 2016). A late-sown, short-season wheat would provide growers with more flexibility at sowing time, and aid in mitigation of late breaks and the control of herbicide-resistant weeds. This paper summarises a combined analysis of four sites and three trial years to identify the traits of late-sown wheats that were contributing positively to yield in these environments.

Methods

Field experiments were conducted at Narrabri in 2019, Merredin and Yanco in 2019 and 2020, and Wagga Wagga in 2021 and 2022. The aim of all experiments was to evaluate a set of highly vigorous pre-breeding wheat genotypes, international lines, and commercial varieties for their suitability to winter sowing. The second and main aim of the study was to identify the plant traits that were contributing positively to yield from late sowing, especially in relation to performance in earlier-season sowing dates. All experiments were sown as completely randomised block designs with up to four sowing dates at an experiment location. Water for crop growth and yield was plentiful with experiments irrigated at Narrabri, Merredin, and Yanco while Wagga Wagga experiments received annual rainfall totals of 769mm and 914mm for 2021 and 2022, respectively.

Throughout the season, essential metrics including biomass at key stages, phenology, ground cover, light interception, and tiller number were recorded. The inner four rows of plots were hand harvested at maturity to provide final grain yield and other derived harvest metrics. The data were analysed by linear mixed models fitted using ASReml-R (The VSNi Team 2023). Spatial variation was accounted for within the residual component of the model. Principal component analyses were performed using the package stats 4.3.2 (R Core Team 2023) and visualisations using the package factoextra 1.0.7 (Kassambara and Mundt 2020).

Results and discussion

The key finding from these experiments was as sowing date was shifted later in the year, grain yield declined. For example, at Wagga Wagga in 2021, plots sown on June 7 yielded on average 5.18t/ha in comparison to those sown approximately a month later on July 12 which yielded 4.47t/ha (LSD = 0.27t/ha). This is consistent with other studies and can be attributed to less time for growth and development of spikes key to achieving larger grain number and yield (Fischer 1979). Yet despite the later sowing, grain yield was only reduced an average 14% across all genotypes.

The Principal Component Analysis (PCA) biplots in Figure 1 contrast early and late sowing in Wagga Wagga 2021. This analysis was performed to identify associations for traits contributing to yield with early and later sowing. Strong positive correlations were observed in both plots between grain yield and grain number (GN), and with harvest index. Notable differences between times of sowing included a lack of correlation for days to anthesis and yield for the June sowing, to strongly negatively correlated for the July sowing. Similarly, days to maturity was positively correlated with yield in the June sowing but strongly negatively correlated in the July sowing. Later sowing times negatively impacting grain yield is consistent with the findings of Sharma et al., (2008) and Hunt et al., (2012). As both sowings occurred relatively late in the season, well past the normal sowing window, Z31 biomass and ground cover were both positively correlated with yield (Figure 1) confirming the importance of early vigour to grain yield.

Figure 1. Two principal component analysis biplots of 144 wheat genotypes evaluated in 2021 at the Wagga Wagga Agricultural Institute for traits in a field experiment. The plot on the left displays data from the June 7 sowing, and the one on the right from the July 12 sowing. Abbreviations: 65GCP DPS, days to 65% ground cover; Z31 BM, amount of biomass at Zadok’s growth stage 31.

Note on simplified interpretation: Arrows at an angle of less than 90° to one another indicate positively correlated traits, while greater than 90° indicates a negative correlation. Length of each arrow refers to the contribution of that trait to the overall model, longer indicating a greater contribution.

Another pair of PCA biplots are presented in Figure 2 summarising trait relationships with grain yield for the earliest (May 12) and latest (July 9) sowing dates at Yanco in 2020. Similar to the Wagga Wagga experiment, days to anthesis was negatively correlated with grain yield in the later July sowing but also the earlier May sowing. This again suggests that the phenology and later flowering of the wheat lines used in this experiment were not well suited to any of the sowing dates, and mistimed anthesis negatively impacted yield. Numbers of tillers and grain were strongly positively correlated to grain yield at both sowing dates. Interestingly, between sowing dates, anthesis biomass shifted from strongly negatively correlated with grain yield in the earlier May sowing to no correlation in the later July sowing. These associations suggest that too much vegetative growth can be a detriment to yield when sown early consistent with Hunt et al., (2012). Biomass at Zadok’s score of 31 had very little impact on either PCA model but was slightly more positively correlated with grain yield in the July sowing than the May.

Figure 2. Two principal component analysis biplots of wheat genotypes evaluated in 2020 at the Yanco Agricultural Institute for traits in a field experiment. The plot on the left displays data from the May 12 sowing, and the one on the right from the July 9 sowing. AbbreviationsZ31 BM, amount of biomass at Zadok’s growth stage 31.

Conclusions

A late-sown, short-season wheat for the southern and western Australian wheat zones would provide growers with greater flexibility with time of sowing. Such a wheat variety would aid in the mitigation of climate change affects (frosts and late breaks), in addition to providing growers with more time at the start of the season to control herbicide-resistant weeds with double knock strategies. Together, greater flexibility should contribute to greater profitability and reduced risk in rainfed cropping systems. Experiments performed at Wagga Wagga, Yanco, Merredin, and Narrabri from 2019 to 2022 have demonstrated the potential of some traits in breeding an improved short-season wheat variety. In all experiments, at all times of sowing, grain and tiller number positively contributed to grain yield. Increase in tiller number can either be achieved through selection of highly vigorous wheat genotypes such as used here, or through changes in agronomy to increase plant population. In addition to these physiological traits, a short-season wheat will also need to rapidly reach anthesis to time it during the optimum flowering period. These experiments have demonstrated some of the key traits a short-season wheat will require when sown mid-winter; what remains is their introgression into commercial varieties and their eventual release to market.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the authors would like to thank them for their continued support. We also would like to thank the managers and teams at the Managed Environment Facilities in Merredin (WA DPIRD), Narrabri (USYD) and Yanco (NSW DPIRD).

References

Bureau of Meteorology (2022) State of the climate 2022. No. 1486317707.

Fischer R (1979) Growth and water limitation to dryland wheat yield in Australia: a physiological framework. Journal of the Australian Institute of Agricultural Science

Fletcher A, Lawes R, Weeks C (2016) Crop area increases drive earlier and dry sowing in Western Australia: Implications for farming systems. Crop and Pasture Science 67, 1268-1280.

Flohr BM, Hunt JR, Kirkegaard JA, Evans JR (2017) Water and temperature stress define the optimal flowering period for wheat in south-eastern Australia. Field Crops Research 209, 108-119.

Hunt J, Fettell N, Midwood J, Breust P, Peries R, Gill J, Paridaen A (2012) Optimising flowering time, phase duration, HI and yield of milling wheat in different rainfall zones of southern Australia, Proceedings of the 16th Australian Society of Agronomy Conference: Capturing Opportunities and Overcoming Obstacles in Australian Agronomy. (Australian Society of Agronomy Armidale)

Kassambara A, Mundt F (2020) '_factoextra: Extract and Visualize the Results of Multivariate Data Analyses_'. R package version 1.0.7.'

R Core Team (2023) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.

Sharma DL, D’antuono MF, Anderson WK, Shackley BJ, Zaicou-Kunesch CM, Amjad M (2008) Variability of optimum sowing time for wheat yield in Western Australia. Australian Journal of Agricultural Research 59, 958-970.

The VSNi Team (2023) asreml: Fits Linear Mixed Models using REML.

Unkovich M (2010) A simple, self adjusting rule for identifying seasonal breaks for crop models, Proceedings of 15th Australian Agronomy Conference.

Widderick M, Mclean A (2018) Optimal intervals differ for double knock application of paraquat after glyphosate or haloxyfop for improved control of Echinochloa colona, Chloris virgata and Chloris truncata. Crop Protection 113, 1-5.

Contact details

Timothy Green
Charles Sturt University
Email: tigreen@csu.edu.au

Date published
February 2025