Search for a command to run...
There is a large gap between the refined approaches to characterise genotypes and the common use of location and season as a coarse surrogate for environmental characterisation of breeding trials. As a framework for breeding, the aim of this paper is quantifying the spatial and temporal patterns of thermal and water stress for field pea in Australia. We compiled a dataset for yield of the cv. Kaspa measured in 185 environments, and investigated the associations between yield and seasonal patterns of actual temperature and modelled water stress. Correlations between yield and temperature indicated two distinct stages. In the first stage, during crop establishment and canopy expansion before flowering, yield was positively associated with minimum temperature. Mean minimum temperature below ~7°C suggests that crops were under suboptimal temperature for both canopy expansion and radiation-use efficiency during a significant part of this early growth period. In the second stage, during critical reproductive phases, grain yield was negatively associated with maximum temperature over 25°C. Correlations between yield and modelled water supply/demand ratio showed a consistent pattern with three phases: no correlation at early stages of the growth cycle, a progressive increase in the association that peaked as the crop approached the flowering window, and a progressive decline at later reproductive stages. Using long-term weather records (1957–2010) and modelled water stress for 104 locations, we identified three major patterns of water deficit nation wide. Environment type 1 (ET1) represents the most favourable condition, with no stress during most of the pre-flowering phase and gradual development of mild stress after flowering. Type 2 is characterised by increasing water deficit between 400 degree-days before flowering and 200 degree-days after flowering and rainfall that relieves stress late in the season. Type 3 represents the more stressful condition with increasing water deficit between 400 degree-days before flowering and maturity. Across Australia, the frequency of occurrence was 24% for ET1, 32% for ET2 and 43% for ET3, highlighting the dominance of the most stressful condition. Actual yield averaged 2.2 t/ha for ET1, 1.9 t/ha for ET2 and 1.4 t/ha for ET3, and the frequency of each pattern varied substantially among locations. Shifting from a nominal (i.e. location and season) to a quantitative (i.e. stress type) characterisation of environments could help improving breeding efficiency of field pea in Australia.