Frost and heat limits to the wheat flowering ‘window’ in present and future climates

Abstract

To pollinate and set grain, Australian wheat needs to flower in spring in a ‘window’ when risks of frost or heat stress are low. Other stresses (e.g. early and late drought conditions) are also important in affecting the window, but here the focus is temperature and the window itself. Across the Australian wheat belt, Global Climate Models (GCMs) predict changes of 1 to 4°C within the next 20 years. We estimate current and ‘future’ flowering windows to determine the impacts these have on sowing time and variety choice at sowing and particularly where ‘new’ maturity types may be needed. Analysis of 50 years of weather data from >2800 stations in the wheat belt allowed mapping of the last frost day (LFD = P(Tmin<0°C)<10%) and first heat day (FHD = P(Tmax>35°C) > 30%). The earliest ‘end of frost’ is in the northern parts of WA, QLD and along the southern coastal strips of wheat in WA and SA. In parts of the central WA wheat zone, the LFD is >one month later in 1985-2010 cf 1960-1985. Over 50 years, the occurrence of FHD is similar to that of LFD, excluding the coastal zones. However, during 1986 to 2010, the FHD has become earlier almost everywhere. Using the existing weather for 22 stations across the wheat belt, and adjusting for the predictions of several GCMs and global emission scenarios, we found that the optimal flowering window is expected to shift earlier by two to four weeks by 2030. As the warmer temperatures through winter will also accelerate crop development, the consequence is that earlier sowing with current cultivars cannot compensate for the shift. This predicted ‘shortening’ of the growth season, simply due to temperature response, is more extreme in SA and the Mallee region. If we are to maintain our current season (and the use of radiation, water and nutrients to match it), we will need cultivars that are ‘later-maturing’ in today’s climate, but that will suit future climates. These results for temperature effects on the flowering window are now being extended using crop simulation to determine the consequent changes in patterns of seasonal water use and growth, and what cultivar characteristics in terms of maturity, tillering, and other adaptive traits are likely to be needed within the next 20 years.