Effectiveness of agronomic practices in dealing with climate change impacts in the Australian cotton industry - A simulation study

•Cotton yield, water use and WUE would increase or decrease under future climate•Changing sowing time would have little effects on irrigated cotton yield•Changing sowing time would have substantial impacts on rain-fed cotton yield

Climate change is threatening the viability of cotton production in one of the world's leading cotton regions but we do not know the magnitude of the threat and how effective the common management options are in responding to the threat. This study aims to answer these two questions by using a system modelling approach under various climate change and management scenarios. To construct local climate change scenarios (CCSs) for key cotton production areas in eastern Australia daily outputs from the CSIRO Conformal Cubic Atmospheric Model driven by four general circulation models (GCMs) were used in a stochastic weather generator, LARS-WG. These CCSs were then linked to a process-oriented cotton model (CSIRO OZCOT) to quantify the effect of changing planting times and irrigation scheduling triggers on cotton lint yield, water use, and water use efficiency (WUE) in 2030. We considered four planting times: normal planting, 15 days early, and 15 and 30 days later in both irrigated and rain-fed cotton. For irrigated cotton, we considered two soil water deficits (50 and 70 mm) below which an irrigation event was triggered in nine cotton production areas (Emerald, Dalby, St George, Goondiwindi, Moree, Bourke, Narrabri, Warren and Hillston). For rain-fed cotton, we considered three row configurations (solid, single skip and double skip) in four production areas: Emerald, Dalby, Moree and Narrabri.Simulations for irrigated cotton (access to full water, no nitrogen limitation), when compared with baseline (current climate and atmospheric CO2 concentration with normal planting), show that lint yield, water use and WUE decreased in 2030 in the majority of study locations with both irrigation triggers under normal planting. The impact of planting time could not offset the negative impact of future climate change on cotton lint yield in 2030 across both irrigation triggers at all locations except for Dalby. The 30 day late planting and irrigation at 50 mm deficit performed better than their respective counterpart(s).Simulating rain-fed cotton (assuming full soil water profile at planting), when compared with baseline, lint yield and water use increased at Emerald and Narrabri and decreased at Dalby and Moree across row configurations in 2030 with normal planting; and WUE increased by 15.5% under future climate scenarios with normal planting. Early planting decreased lint yield by 8 to 18% at Dalby and Moree while late planting increased lint yield by up to 30%; the later the planting, the greater increase in lint yield at Emerald, Dalby and Moree.The simulation results indicate that the response of cotton to climate change is variable and region specific, which suggests that different management strategies may be required for each situation. Cotton growers will need to consider which management options best suit their circumstances.