Weed Seed Wizard: A tool that demonstrates the value of integrated weed management tactics such as harvest weed seed destruction

Harvest weed seed destruction (HWSD) can be included in an integrated weed management (IWM) program to help control weed populations and combat herbicide resistance. However, it may not be possible or practical to use this technique in every year. This research utilised a computer model, the Weed Seed Wizard, to investigate the impact of employing HWSD every second year. Data from a long term trial in Merredin Western Australia demonstrated the value of residue burning (an early method of HWSD) compared to residue retention on the destruction of Lolium rigidum seed from 2003 to 2013. The agronomic practices utilised in this trial, soil type and rainfall at the field site, and crop yield data were used to parametrise two scenarios in the model. Scenario 1 was based on data from the field trial plots where residue was retained and scenario 2 was based on data from the plots where residue was burnt (i.e. HWSD). A third hypothetical scenario was based on scenario 2, but only incorporated HWSD in every second year. The model gave reasonable predictions of L. rigidum seed production each year (when compared to the actual seed production in the field trial), and accurately predicted when L. rigidum seed numbers would reach very low levels in scenario 2, due to annual HWSD. Scenario 3 indicated that HWSD in every second year could not reduce L. rigidum seeds to the same extent as annual HWSD, but the L. rigidum population was reduced to and maintained at less than one plant m−2 at harvest within four years. The model indicated that the total cost of weeds ranged from $85 ha−1 in scenario 1 to $32 and $27 ha−1 in scenario 3 and 2. However, income was greatest in scenario 1 as HWSD via burning residue caused a yield reduction in scenario 2 and 3. The results highlight both the benefits of HWSD as a weed management tactic and the value of the Weed Seed Wizard as a tool to investigate different IWM programs.