Glyphosate resistant crowsfoot in the Northern Territory.

Integrated management of glyphosate-resistant weeds in cotton

Glyphosate resistant grass populations won’t stop showing up in Australian cotton, and no new herbicide silver bullet is on its way to save us. Fortunately, the industry has become very good at adapting existing modes of action for new use patterns, and developing integrated strategies to make the most of available tools.

Glufosinate and dicamba are becoming available as over-the-top applications, through Bayer’s XtendFlex cotton technology. High rates of glyphosate and glufosinate can also now be applied on sparse populations through optical spot spray equipment. Both offer new chances to get control of challenging glyphosate-resistant summer grass populations, and maintain profitability while doing it.

In order to look at the effectiveness of the latest herbicide resistance management strategies from an agroeconomic perspective, we used the digital decision support tool BYGUM to develop and analyse simulations of XtendFlex and optical patch spraying on glyphosate-resistant barnyard grass.

The setup

In each of these simulations, we assume that each barnyard grass population is around 7x resistant to glyphosate, compared to wild type (a figure based on what’s been found in field research), and that maximum optical spot spray rates of glyphosate (equivalent to 5.5 L/ha of Crucial or 7.0 L/ha of Weedmaster DST) and glufosinate (equivalent to 10L/ha of Biffo) provide reliable control of glyphosate resistant barnyard grass at this level of resistance.

We’ve used 2024 prices for herbicides and application costs from the February 2025 edition of NSW DPI’s Weed Control in Winter Crops guide to develop gross margin estimates. Other crop costs (water, nutrients, insect and disease management, insurance, etc) are consistent between simulations.

Herbicide costs for optical applications are based on an assumed 15 per cent coverage of the total field, which is a realistic but somewhat favourable assumption. Optical applications can be used on up to 30 per cent of total field area, so in some weedy years herbicide costs would be somewhat higher than shown here (and conversely, somewhat lower in the most successful systems). We’ve used an average cotton price of $620 per bale, and a weed-free maximum yield of four bales per hectare dryland and 12 bales per hectare irrigated.

Simulated weed management

Throughout this analysis, we’re comparing between three different integrated weed management (IWM) strategies in crop:

A Best Management Practice (BMP) strategy from the CRDC/CottonInfo Cotton Pest Management Guide;
A strategy using XtendFlex over-the-top herbicides plus residuals (Xtend+R); and ,
Low input glyphosate-reliant practices.

In summer fallows, we’re comparing between a low-input, glyphosate-centric strategy and a moderate-input strategy relying on optical applications of glyphosate.

Note that only systems that include BMP or Xtend+R strategies in crop would reach the HRMS target diversity for herbicide modes of action and survivor control.

Results

We varied the weed’s emergence percentage and seed production by up to 20 per cent to cover a range of conditions. These figures are the average of four simulations. Each scenario starts with a somewhat problematic population of 15 weeds/m2 setting seed in the year prior to simulation start.

1. Dryland scenarios

One crop in two years dryland

Low-input summer fallows are prone to huge seed bank replenishment of glyphosate-resistant barnyard grass in these scenarios, even when using BMP in crop. The BMP in-crop scenario is able to produce reasonable returns even when summer fallows are not of great quality, but the XtendFlex system relies on better fallow weed management to do more than break even.

The XtendFlex system, with the use of residuals in crop, can return the system to low weed seed density and positive profitability, although somewhat slower than the BMP system, and only when good control is obtained in
summer fallows.

Low-input, glyphosate-centric cropping strategies clearly fail to control resistant barnyard grass and lead to crop failure. The use of highly effective summer fallow strategies does appear to lead to sufficient slowdown in seed bank replenishment to provide some profitable years, but spray timings would need to be very carefully managed to ensure that optical applications stayed under the 30 per cent of field application registration limit.

One crop in three years dryland

In these scenarios (Tables 5 and 6), both the first year and the preceding, unsimulated year are fallow, which predicts good soil moisture for the first crop in year 2. Where good fallow control of the weed population is obtained in the first year (the optical SF strategy), very good returns are possible in year 2 with either XtendFlex or BMP in-crop strategies. Where weed control is poor in the first fallow, the BMP system provides much more reliable returns in year 2. BYGUM predicts that good weed control in fallows can quickly translate into future profitability.

Two years of summer fallow with highly effective weed control are sufficient to almost reset the seed bank for awnless barnyard grass, but low-input crop tactics do result in increases in the seed bank. Therefore vigilance in non-cropped years is especially critical if using lower-input in-crop tactics. It is important to note that these results in particular rely on our assumption that very high rates of glyphosate or glufosinate through optical spray equipment can control barnyard grass with the levels of glyphosate resistance commonly observed in the field to date.

The XtendFlex plus residual scenario was as successful as the BMP system over the five simulated years of one crop every three years, when combined with moderate input summer fallows.

2. Irrigated cropping

Three crops every four years irrigated

Both the XtendFlex plus residual and the BMP in-crop strategies are predicted to be highly effective in irrigated cotton. It should be noted that there is a substantial reliance here on pre-planting cultivation and an early residual, and the loss or failure (through poor timing or other problems) of either of those tactics results in the gross margin being reduced by up to two-thirds (not shown), an effect that cannot be overcome with multiple glufosinate applications in crop.

Low input cropping, relying on glyphosate and a single residual, is a poor choice in the high-cost environment of irrigated cotton with a resistant weed. Summer fallows with very good weed control can bring such a system back into profitability quickly, but without changes to the in-crop system, we return straight back into weed blow-outs.

Continuous irrigated cropping

A long-term continuous scenario in irrigated cropping suggests that the BMP and XtendFlex strategies are approximately equal over time. The lower cost of the XtendFlex system leads to slightly higher gross margins overall, and the higher intensity of the BMP strategy results in more rapid and complete weed seedbank depletion. As expected, doing very little other than glyphosate in a glyphosate-resistant situation (the low input strategy) is consistently unsuccessful both for gross margin and weed seedbank control.

Takeaways

Optical spot spray technology is a powerful option for retaining control of patchy infestations while maintaining gross margins. This is true even at very high herbicide concentrations, as long as the total area sprayed remains low.

Opportunities for using optical technology are more frequent in fallow—and so summer fallows in dryland systems should be seen as a space for getting on top of burgeoning resistant weed numbers, to support clean fields in the next crop. What happens this year can have its largest effects next year, when it comes to weeds and this is strongly reflected in predicted gross margins.

Results from BYGUM suggest that XtendFlex systems can be slightly more economically efficient than a very intensive BMP program, but only when they include supporting options:

fallows with robust weed control;
effective residuals in crop;
high levels of crop competition; and/or,
cultivation where possible.

If weed control in summer fallows is poor, XtendFlex may not provide enough protection from weed blowouts in crop, especially for dryland growers. BMP weed management is still the highest standard system for herbicide diversity and long-term weed seed bank control.

Finally, these scenarios do not consider what happens if the weed population were to evolve resistance to glufosinate. Herbicide diversity (which mitigates against resistance evolution) is highest in the BMP system, implying both that glufosinate resistance would be slower to evolve, and that it would be more readily managed if it did evolve. More work is needed on predicting the evolution of resistance to glufosinate in XtendFlex cotton, and the effects on crop economics that would occur in the presence a multi-resistant summer weed in our systems.