A weed control threshold for cotton production was first released to the Australian industry in 2008. It has been widely used as a guide to the level of weed pressure that can be tolerated in fields without economic damage occurring. However, the model behind the threshold was based on a limited data set, derived using the ‘mimic weeds’ (common sunflower, mungbeans and Japanese millet), based on the assumption that the level of competition from these ‘mimic weeds’ could be readily related to real competition levels from real weeds.

Project DAN1601 focused on testing the underlying assumption that the competition results from mimic weeds could be related to real weeds, and, assuming the approach was valid, using a much larger data set to derive a more accurate threshold.

Data from 3 years was used to test the validity of using mimic weeds to define competition in irrigated cotton. Competition levels were compared between differing population densities of: fierce thornapple vs. common sunflower; bladder ketmia vs. mungbean; and, awnless barnyard grass vs. Japanese millet. Many similarities and differences were found between the mimic and real weeds, but an underlying relationship was derived that showed that as far as the cotton was concerned, the only real differences between the six ‘weeds’ were related to weed height and biomass – in other words, bigger weeds are more competitive! So, competition experiments using mimic weeds are valid, and a lot easier to conduct than experiments using real weeds. Also, the findings of this work showed that it is possible to extrapolate the results from a few mimic weeds to a wide range of real weeds.

Based on this finding, data from field experiments between 2002 and 2015 was collated and analysed to develop much more accurate relationships between the three mimic weeds weeds’ (common sunflower, mungbeans and Japanese millet) and irrigated cotton. Three papers were developed, one for each weed type, developing dynamic weed control thresholds for these weed types. The first of these papers has been accepted for publication and the second paper has been submitted to the publishing journal. The final paper will be submitted when the third paper is accepted.

Results from the first paper showed that for a large weed (eg. fierce thornapple, Noogoora burr, sesbania), densities of 1 per m of row caused more than 5% yield loss if present in the crop between 43 and 615 GDD. Higher densities caused higher levels of damage. Where the weed can be controlled with glyphosate, a 1% yield loss threshold is more realistic, extending the critical period for weed control for 1 weed per m row to between planting and 836 GDD.

It is hoped that the next two scientific papers will be accepted for publication later this year, and then the results from the work can be extended to the Australian industry, giving growers a firm threshold for controlling weeds in irrigated cotton.

In addition to this core work, the project has continued to support the cotton industry through articles, meeting and input into the weeds sub-committee of the TIMS panel, as well as in other areas.

Work on a pupae busting experiment has continued, preliminary results indicating that pupae busting does deliver additional benefits through improved weed control.

A study into 2,4-D damage, prompted by the heavy damage experienced in the Walgett area in the 2017/18 season, highlighted the very low levels of 2,4-D that could cause damage to cotton. A series of tissue tests determined that damage at ACRI observed in this same season was caused by 2,4-D concentrations below the level of detection by instruments.

At the request of the CottonInfo team, all main articles in WEEDpak have been updated, and short (4-page) CottonInfo sheets developed from the longer articles.

This project has also had a large input, as lead, of a parallel project, DAN1805, the “Biological control and taxonomic advancement for management in the Noogoora burr complex”, which also finished in June 2019. The project involved a number of different facets and was exceptionally successful in achieving its objectives in spite of many issues. A final report on this project has separately been submitted to CRDC. Outcomes included: re-evaluating the taxonomy and distribution of the Noogoora burr complex, showing that based on genetics, there appears to be only two groups of burrs (species), with many hybrids occurring; successfully developing a bioherbicide which is effective across all the Noogoora burr species; and determining that all plants in the Noogoora burr complex are hosts of the Verticillium wilt pathogen, with multiple defoliating and non-defoliating strains of the pathogen isolated from the burrs, often with more than one strain present in a single burr. Four scientific papers are currently being developed from this project.

Following the success of the technology developed around the bioherbicide for Noogoora burr, steps towards commercialisation of a bioherbicide for Bathurst burr have been commenced.