Where is herbicide resistance taking our farming systems

Christopher Preston, School of Agriculture, Food & Wine, University of Adelaide

GRDC code: UA00113, UA00124

How did we get here?

The most important factor driving the rise of herbicide resistance is herbicide use. The more an individual herbicide is used the more likely resistance to that herbicide will arise. However, not all herbicides are equal in terms of getting resistance. Some herbicides are more resistance prone than others. Some years ago, I developed a set of rules of thumb about when herbicide resistance might be expected. An updated version of that is in Table 2. What is obvious from Table 2 is that for some herbicide modes of action (Groups A and B) resistance occurs rapidly, but takes longer for other groups. 

Other factors that have contributed to the recent rise in resistance are changes to farming systems. In particular has been the move to more continuous cropping, the increase in farm size and the reduction in tillage. All of these, particularly the adoption of no-till, have contributed to increased reliance on herbicides for weed control. In the Northern Region, chemical fallows have been particularly problematic for resistance selection. As fallows have no other plants growing in them, any survivor of herbicide application has full use of available resources and will produce vast amounts of seed. For this reason resistance to herbicides can occur quickly in fallows.

Where is resistance taking our farming systems?

The experience from the southern grain regions of Australia, where resistance has been a major problem for more than 2 decades, is that resistance will make grain production more complicated. Rotations will need to change to introduce or increase the frequency of crops where control of the most problematic weeds can be maximised. Thinking about weed control will also change. The focus will be on population management and seed set control, rather than killing weeds after they have emerged. Reducing seed banks will become an important strategy.

Figure 1 shows the results of a trial conducted on managing annual ryegrass across years to reduce the weed seed bank. The details of the treatments used are not important, because they were treatments appropriate for southern Australia, but the principles are. What this trial showed is that a large intervention in 2009 followed by business as usual, MS1, leads to weed seed banks rebounding quickly to high numbers. A strategy where weed seed set control is implemented in the years after the large intervention, MS3, results in a continuing decline in the weed seed bank. A single intervention may not be sufficient to manage resistant weeds.

Figure 1. Weed seed banks of annual ryegrass in a long-term management trial at Roseworthy in SA. Three management strategies were adopted following a large intervention. MS1 was the usual weed management after 2009, MS2 had additional weed management and MS3 included seed set control tactics in 2010 and 2011. λ is ratio of the final seed bank to the original seed bank. A λ of 1 means no change in seed bank size.

Figure 1 text description: All three management strategies had the same results in 2009 and 2010, showing a decline from approximately 5000seeds/m in 2009 to approximately 700seed/m in 2010. In 2011, MS1 increased to approximately 8000seeds/m and then decreased again in 2012 to approximately 3500seeds/m. MS1 had a ratio of the final seed bank to the original seed bank of 0.64. MS2 stayed steady after 2010 in the ARG seed bank at approximately 700seed/m in both 2011 and 2012. MS2 had a ratio of the final seed bank to the original seed bank of 1.18. MS3 continued to decline after 2010 until there was approximately 0seeds/m in 2012. The ratio for MS3 of the final seed bank to the original seed bank was 0.02.

In the Northern Region the fallow system is currently a weak link in weed management. Tactics such as double knocking and using weed detection sprayers are helping manage resistance in fallows. However, these practices are reliant on herbicides and resistance is inevitable. The fallow phase is going to be the greatest challenge to sustainable weed management in the Northern Region.

What other techniques are available? In the short term there are not many. Tillage is the obvious tactic that can be re-introduced for fallow management. However, tillage is very much a two-edged sword. While tillage can effectively remove troublesome weeds it will inevitably bury weed seed through the soil profile. One of the advantages of no-till systems is that they leave weed seed sitting on the soil surface where it is often easier to deal with, particularly with pre-emergent herbicides. Burying the weed seed also increases the probability of staggered germination events. Cultivating an area that was cultivated in the past few years will bring weed seed to the surface and stimulate germination. Therefore, while strategic tillage may have a place in weed management, it may lead to a requirement for ever more tillage.

Other physical tactics, such as gas, steam or even microwaves, are not available in a format that could be used in large area grain production and may never be practical. Harvest weed seed control tactics, such as currently employed in Western Australia, require the weed seed to enter the harvester at harvest. Such techniques will only work for some weeds and cannot be usefully employed in a fallow. Strategies such as mowing or rolling could be used for weed control, but a cover crop is required to provide the necessary competition against weeds. This sacrifices a potential cropping phase to weed management.

One technology that is likely to arrive is robotic technology. There is considerable research and development in robotics for many industries including agriculture. Some aspects of the technology are already in use in agriculture, for example auto steer and weed detection sprayers. The value of robotics for weed management will be the ability to conduct weed management at any time, regardless of what other activities are being conducted. At present, most of the focus in robotics for weed control centres on detecting individual weed seedlings and applying a herbicide to kill them. While this will reduce the cost to growers of the herbicides they are using, resistance management requires stopping seed set. Robotics for weed control will become much more valuable if non-chemical weed destruction tactics can be coupled with it.

There is no obvious silver bullet that will solve the herbicide resistance issue. Even if there were, use of a single tactic would lead to resistance just as reliance on herbicides has resulted in herbicide resistance. Growers in southern Australia have understood that a diversity of practices are required to manage weed populations and there must be a focus on reducing seed banks. In the absence of a new ground-breaking weed control technology, herbicides will continue to be a major contributor to weed management for at least the next decade. To preserve this valuable tool, growers need to use it more wisely. In my opinion, for the Northern Region residual herbicides have to become a bigger part of the equation to take pressure of the post-emergent products, and practices that stop weed seed set have to be developed and adopted.

Contact details

Christopher Preston

University of Adelaide

Ph: 08 8313 7237

Email: christopher.preston@adelaide.edu.au