Use of thresholds for insect control decisions has become a generally accepted practice for many farmers. But while there are proponents of thresholds for weed management, many weed science professionals contend applying the concept to weed control is pretty much an exercise in futility.

Looking at the few pros and the many cons, Bruce Maxwell of Montana State University says: “Bottom line, thresholds are detrimental to weed management.”

With insects, he told members of the Western Weed Science Society at their annual meeting at Coeur d'Alene, Idaho, “You can look at development stage and numbers and you can make a fairly accurate prediction of how much damage will be done to the crop over a certain period.

“But, when you apply this process to weeds, it just isn't applicable. Weed growth and populations depend on rainfall, temperatures, and other conditions that can't be reliably predicted. Insect populations won't persist for more than one or two cycles, but weed seeds can be viable in the soil for years.

“The large number of seeds that can be produced by a single weed make the whole threshold concept very difficult,” Maxwell says.

Farmers, he says, are “too programmed with the idea that they have to get rid of all weeds, when they should be asking: Do I need to control X weed? What are the economics of it? What are the environmental concerns?”

‘Abundance thresholds’

A better step, he says, might be to identify “weed abundance thresholds” and base management on economic returns to the producer.

Factors that make the threshold concept difficult to determine, Maxwell says, are variations between crop values, the size and impact of future weed populations, the stability of weed demographics and damage caused by those weeds, weed management costs and the stability of those costs, and the costs of weed detection (farm size).

“There's a lot of information about weed biology we don't know,” he says, “and we need to have it. We have a broad range of data, a lot of it not very scientific, which makes threshold predictions more difficult.”

Cost of detection is also “very important,” Maxwell notes. “These costs are minuscule in small test plots, but the bigger the space, the more prohibitive it becomes to detect and eliminate the last weed. Is there some point where the value of the crop itself would make this worthwhile? We don't know. We just don't have enough information to determine economic thresholds for weeds.

Own or rent land

Whether a producer owns or rents the land is also a consideration in how aggressively weeds are controlled, it was pointed out.

And down the road, sensor-based weed control may enable farmers to achieve near-zero populations and reduce concerns about development of resistance. Already, farmers can use electronic technology to tailor fertilizer applications to specific field areas. In the future, the scientists say, sensors may be able to detect weeds as spray rigs move over the field, applying herbicide only to those weeds.