A computer model will be used this year to help find specific and practical ways of limiting the development of glyphosate resistance in weeds on Queensland and northern New South Wales farms.

Glyphosate is an extensively used herbicide, and resistance is regarded as the major threat facing our modern “zero-tilled” farming systems, with the potential to make grain farming unprofitable in some areas.

Already, farmers in many overseas countries have large populations of glyphosate- resistant weeds infesting their prime cropping paddocks.

Department of Primary Industries and Fisheries plant modeller, David Thornby of the Leslie Research Centre in Toowoomba, said the computer model had passed the development phase and was now fully functional, using cropping and weed data from local field research.

Dr Thornby said the model would be used this year to investigate a wide range of farming practices and their impact on the evolution rate of glyphosate resistance.

The Grains Research and Development Corporation is supporting the three-year research.

“Gathering such a wide range of data through field experiments would be very time consuming,” Dr Thornby said. “With the computer model, we can come up with the answers very quickly.”

He said there were particular concerns about the development of resistance to glyphosate, a low cost, effective herbicide widely used in grain production.

“We use the plant growth model APSIM to simulate the effects of competition between weeds and crops, and we have added our own sub-models to account for differing levels of annual germination rates, dormancy, and seed bank persistence.

“Our approach takes into account soil and climate variables, as well as crop rotations and other agricultural factors, to predict the likelihood of a particular set of practices resulting in a herbicide resistance problem,” he said.

Dr Thornby said spraying frequency and the complementary use of different mode of action herbicide groups, and other agricultural practices such as tillage, were incorporated into the computer model.

“We are working on producing a model that is able to simulate the growth and population dynamics of a wide range of northern region weed species, rather than a single target species.

“We can then investigate the likely rates of evolution of herbicide resistance for any weed species of interest in the region and the effects of recommended resistance management tactics on these weeds,” he said.

Dr Thornby said the model was presently set up to simulate barnyard grass as an example weed species.

He said the computer simulations were due to be finished late this year and the project findings provided to scientists, agronomists, and growers until June 2008.

“We hope to produce recommendations for agricultural practices that will have a positive effect on the herbicide resistance problem, backed by the model’s predictions,” Dr Thornby said.