Identification of environment similarities using a crop model to assist the cultivation and breeding of a new crop in a new region

Context Rainfed crop-growing environments are known for their high yield variability, especially in the subtropics and tropics. Improving the resilience of crops to such environments could be enhanced with breeding and agronomy research focusing on groups of similar environments. Aim This study presents a framework for developing these groups using the Agricultural Production Systems Simulator (APSIM, ver. 7.10) model. Methods As a case study, the framework was applied for pigeonpea (Cajanus cajan L. Millsp.) as a potential new pulse crop for the Australian northern grains region. The model was first validated and then used to simulate yield, compute heat and drought stress events and analyse their frequencies for 45 locations over 62 seasons from 1960 to 2021. Key results The model performed satisfactorily compared to field trial data for several sowing dates and locations. The simulated yield varied greatly across locations and seasons, with heat-stress events (maximum temperature ≥35°C) and rainfall showing highly significant associations with this variability. The study identified seven groups of locations after converting the simulated yield into percentiles, followed by clustering. Drought-and-heat stress patterns varied across these groups but less so within each group. Yield percentiles significantly declined over the seasons in three of the seven groups, likely due to changing climate. Conclusions The framework helped identify pigeonpea’s key production agroecological regions and the drought and heat constraints within each region. Implications The framework can be applied to other crops and regions to determine environmental similarity.