Farmers are finding themselves on the front line of climate change. General climate trends point to warmer temperatures and less rainfall – critical factors for broadacre crops such as wheat.
In these adverse growing conditions, crop yield projections have become an important factor in agricultural decision making across the production chain.
They are produced using crop simulation models that take into account the impact of weather, soil conditions and crop management practices. As a result of their effectiveness, models have become a key tool in a modern farmer’s kit that informs them about what to grow, when to plant, and when to fertilise.
In policy making, crop projections can inform policy development and government interventions to secure food availability, reliability of supply, and affordability. And in research, crop growth and yield simulations assist in breeding selections for crops with suitable traits, as well as identifying ‘climate smart’ crop management practices.
However, many different crop models are used worldwide to simulate yields, and individually they can produce very different results, leading to uncertainty in simulated yields. Such uncertainty becomes particularly striking under elevated temperatures.
A new study published in Nature Plants led by CSIRO researcher Dr Enli Wang and Dr Pierre Martre of the French National Institute for Agricultural Research (INRA) found a way to improve the simulation of key physiological processes in wheat crop models in response to temperature, which decreased around half of the uncertainty in simulated yield across a wide range of in-season temperatures of 14oC to 33oC.
“We developed new algorithms that improve the accuracy of yield simulations by as much as 50 percent, and 42 percent on average, which is a significant development,” Dr Wang says.
He says that the problem was caused by crop models being developed using historic local data, but they are increasingly being applied to simulate crop yield under a changing climate worldwide.
“Under future climate scenarios, different regions are characterised by higher temperatures and more extreme weather, i.e., beyond the environmental conditions where the current models were developed,” he says.
“We hope these improvements will provide decision makers with increased confidence in using simulated yield for targeted interventions.”
Heading off a global problem
The study formed part of the Agricultural Model Intercomparison and Improvement Project (AgMIP), an international collaboration focussed on improving global crop simulations by incorporating physical, biophysical, and socioeconomic factors.
In this particular work the researchers analysed and compared 29 widely used wheat models against experimental field data, which involved 60 wheat crop modellers and field crop experimentalists, as well as a climate scientist and a biostatistician from 15 countries.
The research revealed that the inaccuracy of simulating key physiological processes of wheat in response to temperature change is the major cause for uncertainty in simulated yields, which is larger than the uncertainty caused by using 16 different global climate change models.
While temperatures of around 20oC are optimum for wheat growth, wheat is being grown under hotter and suboptimal temperatures under the changing climate. The newly developed algorithms will allow more accurate simulations of wheat yield under such conditions.
“We improved accuracy of yield simulations for wheat crops grown under contrasting sowing dates and temperature environments and across the major wheat growing regions of the world, including Australia,” he says.
“Projections with a higher degree of accuracy under these conditions will provide farmers with increased confidence, and allow for better informed decision making to maximise their yields, or avoid making costly mistakes.
“We hope advancements in crop projections, such as those provided in this work, will encourage more decision makers to adopt crop projections as part of their regular practices.”
A growing dependence on crop projections
Dr Cynthia Rosenzweig leads the Climate Impacts Group at the NASA Goddard Institute for Space Studies and Columbia University, and is an AgMIP Principal Investigator.
She says climate change will impact all cropping systems, making the improvement of crop projections an important focus area.
“Climate change will affect food security by altering regions where crops are grown, productivity levels, and nutritional content. Increases in extreme events such as heatwaves, droughts, and floods will affect the stability of food delivery systems,” Dr Rosenzweig says.
While these changes will put increased pressure on government and intergovernmental developmental agencies, improved crop projection modelling will provide a clearer understanding of how climate change will affect agricultural systems in their regions and countries.
Dr Rosenzweig says that the model intercomparison work that AgMIP has done so far shows that many crop models are ready for improved response functions, including the temperature formulations reported in this study.
“Not only is there room for improvement in the ‘big four’ crops (i.e., wheat, maize, soybean, and rice), but there is so much work to do for improving models for crops of importance in developing countries including sorghum, millet, sugarcane, and groundnut,” Dr Rosenzweig says.
In Australia, the findings are already in the process of being implemented. Co-developed by CSIRO, the Agricultural Production Systems sIMulator (APSIM) is a leading crop and cropping systems simulation model. It is increasingly being used throughout the agricultural decision making process in Australia and internationally, and Dr Wang says a future update will incorporate the improved temperature response algorithms.
“We are already starting to see rising temperature and extreme weather conditions appearing during cropping seasons, so the sooner changes are implanted the better.”