Joint Modeling of Wheat Head Blast and Water Deficit Risks for Climate-Informed Sowing Calendar Decisions for Rainfed Wheat in the Brazilian Cerrado

Triticum lineage, is a major constraint to wheat expansion in the Brazilian Cerrado. Delayed sowing is often recommended for disease avoidance; however, late sowing can increase the risk of water deficit. This study quantifies the joint risks from head blast and water deficit in the Brazilian Cerrado to inform sowing-date recommendations. Using daily weather data from a 62-year period (1961-2023), crop phenology was simulated based on an accumulated growing degree day approach. Heading and maturity dates were estimated for six sowing dates defined at 10-day intervals across the sowing window, starting on February 25. Epidemic probability was calculated using a logistic model, while water-deficit events were identified based on an accumulated effective precipitation threshold (150 mm) during crop growth. For each pixel within the potential wheat area in the Cerrado and for each sowing date, we empirically estimated the probability of at least one hazard, fitted smooth pixel-wise probability curves, and identified the optimal sowing date. A spatially smoothed 7-day bin classification was used to generate a practical sowing-recommendation map. In addition, we assessed the effects of the El Niño-Southern Oscillation on epidemic probability. Results showed that mid- to late-March sowings generally minimize joint risk. Spatial heterogeneity highlighted regions suitable for earlier sowing and potential expansion. El Niño/La Niña events amplified/suppressed epidemic probability, particularly for intermediate sowing dates where sensitivity was greatest. These findings support region-specific, multi-hazard sowing recommendations and the incorporation of climate-variability signals into decision support for tropical rainfed wheat.