Defining agro-ecological regions for field crops in variable target production environments: A case study on mungbean in the northern grains region of Australia

Abstract

The northern grains region (NGR) of Australia, which includes the state of Queensland and the northern half of the New South Wales, has highly variable climate leading to heightened production risk for all rainfed crops. Characterisation of the production environment of this region can assist in exploration of potential opportunities for reducing this risk. In this case study on mungbean (Vigna radiata L. Wilczek.) we demonstrate how this region could be characterised using the Agricultural Production Systems sIMulator (APSIM) model. The model was first evaluated for variety Crystal grown widely in the region, and then applied to simulate a water stress index (the daily supply and demand ratio) and yield from 1889 to 2012 at 28 locations. The model was run using location specific as well as three generic soils of 136, 166 and 204. mm plant available water holding capacities (PAWC). Two complementary characterisations were performed using the simulated output, one based on clustering of supply demand ratio averaged for every 100. °Cd to and from flowering, and another on clustering of percentile rankings of seasonal yield variation at different locations. Clustering of supply demand ratio revealed four drought patterns (i.e., target production environments) which commenced at different times from flowering. Seasonal frequencies of these drought patterns, which differed due to major location effects and relatively smaller soil effects, accounted for significant (~84%) variation in simulated yield. Clustering of percentile ranks corresponding to simulated yield in different seasons identified seven meaningful yield clusters. Location memberships of these yield clusters were geographically contiguous and were only slightly influenced for the lowest PAWC generic soil. All locations within these yield clusters showed a tendency to have similar seasonal drought patterns and their frequencies. Locations within different yield clusters could therefore be considered as part of distinct agro-ecoregions. These model defined agro-ecoregions could be used as selection environments for their dominant target production environment(s) to develop new genotypes and their agronomy for better adaptation and yield under variable climatic conditions. © 2014.