Tough test for thresholds

16 February 2001




Tough test for thresholds

For some weeds control

decisions are clear-cut. But

with margins under pressure

and environmental objectives

rising, could some sprays be

left out? Andrew Swallow

reports on research that

could give growers the

confidence to do just that

WEED competition indices promise to take the guesswork out of herbicide decisions, say scientists from IACR-Rothamsted and ADAS.

But first year results of the MAFF-funded weed competition work they are doing have highlighted the difficulties of drawing up control thresholds that are valid every year, on every site.

"Over eight trials our predictions of crop yield loss based on these indices proved to be correct on average," says IACR-Rothamsteds John Cussans. "But on some sites yield loss was slightly under predicted while on others we over predicted."

Where yield losses exceeded estimates, such as with poppies, competition indices, and hence control thresholds, may have to be raised. But last summer was very favourable to weed growth, he says.

"We need to let the project run for three years. It is not clear on the basis of one years results whether last year was just a very bad year, and our thresholds are about right, or whether we really do need to be make more conservative estimates of competitiveness."

At Boxworth, for example, some yield losses were higher than predictions, probably due to competition from poppies (see table). But at Bridgets, yield losses did not reach predicted levels, suggesting some species such as field pansy or speedwells, may have been given higher competition indices than they really justify.

"Poppies are an example of where we may need to revise our competition index upwards. If anything an initial look at this data suggests an even greater separation between the competitive species and those which are really insignificant."

Leaving such weeds below the cereal canopy could be of enormous ecological benefit, which is one of the reasons MAFF is funding the work, he says.

"Farmland weeds are the base of the bio-diversity pyramid. Farmland vertebrates and birds may be at the top of MAFFs list of desirable species but you cant have these if the base of the pyramid is not there."

Economic decision making must go hand in hand with the ecological objectives, he stresses.

"To me the two goals are totally joined together. We are not going to have any farmland unless farmers can make money out of it and you have got to have farmland if you want these birds and animals in the countryside."

Lower grain prices and more costly herbicides mean that for more and more weeds the density warranting control is higher, and therefore more easily identified, he adds.

The competition indices seem to work equally well whether applied in autumn, or spring, provided weeds can be adequately controlled. &#42

2000 results

Density (plants/m sq) Comp Index Yield loss (%)

Boxworth

Poppy 18.4 0.4 7.36

Mayweed 3.2 0.4 1.28

Speedwell 5.6 0.08 0.448

Chickweed 0.8 0.2 0.16

Predicted yield loss 9.3%

Actual yield loss 14.3%

Bridgets

Pansy 67.2 0.02 1.34

Speedwell 26 0.08 2.08

Poppy 0.4 0.4 0.16

Red dead nettle 2.0 0.08 0.16

AMG 0.8 0.1 0.08

Parsley piert 0.4 0.02 0.008

Predicted yield loss 3.83%

Actual yield loss 1.55%

Estimated % yield loss = competition index x species density.

Seed return

The thresholds derived so far are purely based on yield loss in that season, with a spray warranted if predicted yield loss value exceeds the cost of control. Future work aims to estimate the weed seed production risk of leaving low populations. "There are opportunities to cut herbicide costs in the short-term, but we must be sure that in doing so we are not setting a time bomb for future crops," says Mr Cussans.


See more