Sensitivity of productivity and deep drainage of wheat cropping systems in a Mediterranean environment to changes in CO₂, temperature and precipitation

Both anticipated climate change and dryland salinity pose a strategic threat to the sustainability of about 6 Mha of agricultural land in Western Australia (WA). These phenomena require an integrated analysis to estimate their potential impacts and initiate strategic thinking about adaptation. Water loss below the root zone, i.e. deep drainage, is the primary cause of sub-soil salt mobilisation leading to surface soil salinity in areas cleared of natural vegetation in Australia; hence deep drainage is an important externality of agricultural production. The purpose of this paper is to show how changes in CO₂ concentration, temperature and precipitation may affect agricultural production and deep drainage. Results are presented of a simulation experiment with the Agricultural Production Systems Simulator (APSIM)-Nwheat model in which we explored sensitivity of (1) wheat production and quality, and (2) deep drainage, for three sites in WA differing in average precipitation, two soils, various nitrogenous fertiliser rates and present wheat cultivars. Since results of Global Circulation Models (GCM) are still largely inconsistent for this region, we have opted for a factorial approach in adapting 90 years of historical weather data. This set-up enabled separation of effects of CO₂, temperature, precipitation and their interactions, and unravelling of the complex interactions between water and nitrogen availability, phenological development and climate change factors, in the extremely variable Mediterranean climate in WA. Elevated CO₂ concentration increased yields, particularly if nitrogen fertilisation was sufficient and conditions were relatively dry. Higher temperatures had non-linear effects, with initial (up to 3°C) benefits on clay soils, but not on sandy soils, and then substantial yield declines. Both elevated CO₂ concentrations and temperatures (+3°C) decreased grain protein, but in financial terms this was more than offset by the increase in yield in most cases. If, in addition, precipitation was decreased, financial returns dropped below present levels, particularly in the low precipitation regions. Deep drainage tended to be slightly higher under elevated CO₂ concentrations but when higher temperatures were also simulated this was reversed. Deep drainage was greatly reduced in the low precipitation scenarios. Evidently climate change is not only likely to affect productivity, but also deep drainage and hence dryland salinity. The impact can vary in direction such that both 'win-win' and 'lose-win' outcomes may occur, particularly depending on the relative change in precipitation.