Can weeds adapt to harvest weed seed control?

Once weed seed enters the header, there is very little opportunity to escape from harvest weed seed control tools. But, what happens if a few weeds find ways to evade capture in the first place?

Dr Mike Ashworth, research agronomist with the Australian Herbicide Resistance Initiative (AHRI), says there is evidence that long-term, consistent use of harvest weed seed control can result in patches of weeds that possess an evasion trait, such as the ability to shed their seeds early, before harvest.

“It takes a long time, but all weed control tactics apply selection pressure to weed populations for any trait that allows individuals to evade or survive the control measure, allowing them to proliferate,” he says. “We have studied a population of wild radish plants that sheds its seed pods before wheat harvest and distinctly earlier than other wild radish populations in the area, in what seems like a response to 20 years of consistent and effective harvest weed seed control.”

Harvest weed seed control (HWSC) provides excellent control – up to 90 per cent – of many key weeds in Australian grain systems, including annual ryegrass, wild radish, wild oats and brome grass. Although these species are all well suited to HWSC, variability in seed dormancy, early flowering and early seed or pod shedding can enable a portion of the population to avoid being collected by the header at harvest. These seeds are then able to carry this trait into the next generation.

“Diversity is always the answer,” says Mike. “If patches of weeds are flowering early, chances are they are also germinating early. This means early weed control and strong crop competition will help keep weed numbers low. There may also be an opportunity to choose crops or varieties that mature early, allowing the harvest weed seed control tool to do its job on these early populations.”

Diversity is the central component of the WeedSmart Big 6 strategy – diversity in farming systems, diversity in herbicide use, and diversity in cultural controls combine to keep weed numbers low.

Has evasion of HWSC been seen in the field?

Yes, in 2016, a grower in the central wheatbelt of Western Australia and his agronomist noticed that persistent wild radish patches were shedding pods well before the wheat crop matured and before wild radish plants on neighbouring farms or roadsides. This grower has consistently and efficiently implemented harvest weed seed control for over 20 years, along with following a diverse crop rotation and a regionally appropriate herbicide program. However, when the farm was purchased in the mid-1990s, it had a high density of wild radish.

We visited the farm at the beginning of the wheat harvest and found that 86 per cent of the wild radish pods in these patches had fallen to the ground, compared to 0 to 3 per cent pod shedding from populations on a neighbouring farm and along roadsides. The neighbour had not used harvest weed seed control, suggesting that a selection event may have occurred.

We collected seed pods from the ground and planted the seeds in a space with no crop competition to observe their unrestricted growth and maturity patterns. Wheat was also planted at the site so the wild radish plants could be harvested when the wheat reached the soft dough stage. We observed that these plants definitely shed their pods earlier and formed pods closer to the ground than the other populations.

Does this mean that HWSC has failed?

Not really. The critical factor in this case seems to be the high weed numbers when HWSC was first implemented. When weed numbers are high, even a 90 per cent kill rate results in many weed seeds returning to the seed bank. The grower’s diverse weed management program, including HWSC, has successfully reduced the overall weed burden on the farm. Still, the selection pressure has enabled a shift in the population towards the early shedding trait.

To keep HWSC effective, look for ways to mix up your harvest date as much as possible. Diversity in crop type and variety helps with different sowing dates and crop maturity, and you might also consider methods for bringing harvest forward, such as swathing and desiccating early.

How can I avoid selecting for early maturity, prostrate growth habit and the like?

When weed numbers are high, do everything you can, as early as possible, to drive down the weed seed bank. Reducing the number of weeds minimises the chance of certain traits gaining an advantage.

Ensure early germinations are controlled. For wild radish particularly – spray small plants twice. For all weeds, consider using effective pre-emergent herbicides to provide early weed control and maximise early crop competition to reduce weed seed production.

Look for and record changes in weed growth and maturity. Remote sensing or yield data management can give early indications that weed growth habits are shifting. Have samples assessed to determine exactly what has changed and implement patch control with alternative chemicals or treatments.

Other resources

AHRI Insight – Can wild radish evade HWSC?

AHRI Insight – Early shedding of wild radish