Shifting prospects for wheat production: Long-term simulation-based insights under best management in Australia

CONTEXT Wheat production in Australia remains vulnerable to climate variability, particularly to drought, late spring frost, and early heat during reproductive stages. Understanding how temperature extremes affect yield trends under best practice management is critical for improving production under changing conditions. OBJECTIVE To assess long-term changes in frost and heat exposure, quantify their impact on wheat yield trends, and evaluate the effectiveness of current adaptation strategies, including reduced frost sensitivity. METHODS APSIM Next Generation simulations were conducted across 83 locations using long-term weather records (1970–2024) and genotypic variation in phenology and the optimal flowering period (OFP) methodology. Simulations were run under best management conditions, with non-limiting nitrogen supply and CO₂ fixed at 350 ppm to isolate temperature-driven effects. Analyses focused on shifts in the timing of last frost and first heat, trends in yield potential and water-limited yield, characteristics of frost events and scenario analysis modifying frost damage thresholds in an empirical function. RESULTS AND CONCLUSIONS Shifts towards later frost and earlier first heat increased the likelihood of thermal stress during the OFP. Yield potential and water-limited yield declined over time across regions, even when flowering occurred within the OFP, indicating that exposure to interacting climatic stresses persists despite phenological optimisation. Frost events were more frequent than heat events between booting and early grain filling, with up to 9.5 frost days recorded in the worst 10% of seasons. Most frost events occurred under high humidity and dew point temperatures below 0 °C, consistent with dew condensation preceding freezing. Scenario analyses showed that empirically reducing frost sensitivity by 1–2 °C increased yields by 7.3% and 13.4% respectively in adverse seasons, with minimal impact on OFP timing. This suggests yield gains under best management arise primarily from reduced crop sensitivity, with additional benefit from improved alignment between critical reproductive stages and reduced exposure to drought and heat. SIGNIFICANCE Model simulations demonstrate that climate variability is increasing spring frost risk in Australian wheat farming systems, exposing the limits of adaptation strategies based solely on phenological optimisation and highlight opportunities to enhance wheat resilience through agronomic innovation, targeted breeding, and improved modelling, supported by better understanding of frost type, canopy-level microclimate, and ice nucleation processes. Together, they show that improving wheat system resilience will depend not only on when crops flower, but on reducing how vulnerable they are to frost during critical reproductive stages.