Characterisation of chickpea cropping systems in Australia for major abiotic production constraints

Abstract

To develop higher yielding and better adapted chickpeas, various breeding programs currently use a limited number of multi-location trials as a surrogate for the target population of environments (TPE). These TPEs have, however, not been adequately characterised, resulting in some uncertainty about the true representativeness of these surrogate locations as selection environments. We used the Agricultural Production Systems sIMulator (APSIM) model to characterise the Northern Grains Region of Australia, which is a major TPE of chickpea, for drought and thermal regimes. The model was first evaluated for its ability to simulate phenology, dry matter and yield of three new commercially-relevant chickpea varieties including PBA Boundary, PBA HatTrick and PBA Seamer. The model was then used to simulate dynamic changes in water stress quantified through the supply demand ratio, and yield of the highest yielding genotype PBA Boundary from 1900 to 2014 at 45 locations within the region. Water stress, and maximum, minimum and mean temperature patterns were derived through cluster analysis of respective averages computed for every 100 °Cd from 900 °Cd before flowering, to 900 °Cd after flowering. The Northern Grains Region TPE was characterised by four types of water stress patterns and five types each of maximum, minimum and mean temperatures patterns. Ward’s cluster analysis of the percentile ranks of simulated seasonal yield resulting from agro-climatic variability of the different locations enabled identification of eight unique agro-ecological regions within the TPE. Locations within each agro-ecological region were geographically contiguous and had highly harmonised annual variability in yield compared to locations of other agro-ecological regions. Overall, the identified agro-ecological regions were fairly homogenous with respect to drought and thermal regimes and could be treated as separate sub-TPEs. We argue that selecting locations within an agro-ecological region should assist breeding for locally adapted genotypes. In contrast, selecting locations distributed across agro-ecological regions could improve the broad adaptation of chickpea.