Take-all model reveals true threat

18 May 2001




Take-all model reveals true threat

Everybody wants to tackle

take-all. But how do you

know what risk your crops

face? A new take-all model

from Monsanto aims to

provide some answers.

Charles Abel takes a closer

look and reports the views of

a leading agronomist and a

Lincs farmer involved in its

development

HOW would you like to know how much take-all pressure your crops are likely to experience next year?

And how about checking what impact a change of drilling date could have?

Answering those questions and a host of what-if scenarios is the goal of Monsantos new take-all decision-making tool. But every agronomist worth their salt can recite anti-take-all tactics. So why go to the trouble of creating a sophisticated computer model?

David Leaper, technical development manager with Monsanto, has reasons aplenty. Top of his list is the complexity of take-all. "It is affected by the soil environment, how the crop is grown and subtle biological effects, including take-all decline."

Unlike most foliar diseases it can only ever be managed, never controlled, he adds. And a host of tactics must be deployed. Furthermore, fungicides can only be applied to seed before the disease is seen. So a tool to improve targeting and justify fungicide applications is vital. It will help allay concerns from regulators and the public about routine use of pesticides and ensure growers optimise take-all control and get value for money.

The FARModel has been developed for use across northern Europe, using data collected from over 800 individual trials over the past six years.

Detailed evaluation on 77 sites across the UK last year suggests it is accurate. Field details were loaded into the model and the resulting forecast compared with actual take-all levels based on plant samples collected at mid to late grain fill by distributor agronomists. On average the model gave a similar picture to actual results.

To achieve that accuracy the model considers the four most important factors influencing take-all. In order of importance, those are climate; rotation; sowing date; and soil characters.

For each character Monsanto has analysed its extensive trials database to come up with a corresponding take-all risk (see panel).

As details for each field are entered into the FARModel, the software draws on that data to generate a take-all risk rating.

The model provides great scope for "what if" scenarios, says Mr Leaper. By varying sowing date and expected climate risk the resulting take-all indices can be compared.

But at this stage no account has been taken of the likely yield impact. That is determined by comparing the take-all risk index with yield results from the 800 trials. The model displays the expected yield taking into account the forecast take-all level, with a baseline of 8t/ha, which was the average from Monsanto trials.

Once Monsantos development take-all fungicide is approved, the yield recovery to be expected in such a situation will also be displayed.

For growers expecting a higher, or lower, yield from a take-all prone site the average result from the top 25% and bottom 25% of trials results is also shown.

But by themselves the take-all risk indices and likely yield effects are far from the full answer, stresses Mr Leaper. "We are not trying to produce definitive answers.

"It is vital that other factors are considered, such as N timing, variety characteristics, weed control, drainage and compaction, which can all influence take-all. That is why it is important for farmers to use the model with a trained distributor agronomist who can help investigate results further."

Take-all modellers Steve Portas of Dalgety (left) and Lincs grower Vernon Stuffins reckon the new software provides a useful guide.

Rotation risk


Risk of take-all build-up

A (host) B (moderate host) C (poor host)

Crop wheat/ grass/set-aside OSR/potatoes/sugar beet/ barley/ rye beans/oats/linseed/maize/ triticale peas/fodder beet

Crop harvested 2001 Crop harvested 2000 Take-all risk (for wheat drilled autumn 2001)

C C Zero

B C Low

A C Medium

A A High

TAKE-ALLMODEL

&#8226 Uses over 800 trial results.

&#8226 Identifies take-all threat.

&#8226 Field specific.

&#8226 Allows what if checks.

&#8226 Needs interpretation.

&#8226 Highlights role of drilling date.

&#8226 Could aid overhead decisions on drilling capacity.

&#8226 Will support seed treatment decisions in future.

&#8226 Service will be available from trained distributor agronomists.

KEYRISKFACTORS

Climate Biggest factor and hardest to quantify, but the factor with the greatest impact on take-all. In the absence of robust long-term weather forecasting, historic data is used as a guide to the season. Soil temperatures and moisture levels, which are the main drivers of take-all development, have been collated for each UK county over the past 21 years. Results are grouped according to high, medium or low take-all risk and shown as % likelihood. So Lincs crops, for example, have been at high climatic risk of take-all in 14% of the past 21 years, medium risk in 52% of years and low 33%. The model uses the most frequent level of take-all risk by default. But users can over-ride that depending on their weather expectations for the season, or level of risk they want to insure against.

Rotation The two previous crops are entered and risk calculated according to the rotation risk table shown above. This has been validated against 100 UK sites each year over three years. Local knowledge, such as a previous long-term ley, needs to be accounted for once the model has calculated its risk.

Sowing date Earlier drilling markedly increases risk. Data is compared with results from 800 trials.

Soil characters Texture is most important, sand content significantly raising risk. Soil pH and % organic matter are also logged. Results from 450 trials used to calculate risk.


See more