The new HGCA booklet 'Nitrogen for winter wheat - management guidelines' contains a lot of information on nitrogen use. But how easy is it to use to come up with a recommendation?

To find out Crops asked the guide's author Roger Sylvester Bradley to last year's Cereals host farmer Tim Whitehead, who manages Velcourt's Vine Farm, near Royston to talk him through the guide and to come up with a recommendation for both his feed and milling wheat crops.

The major difference between the guide and the yet-to-be published Fertiliser Manual and RB209, which the Fertiliser Manual replaces, is that instead of using the familiar look-up tables to find the crop's nitrogen requirement, it calculates it.

The calculation is done in three parts. First, the crop's N demand is estimated. From that, controversially for some parts of the industry, the soil N supply is deducted. The shortfall in nitrogen is how much applied N is required, after allowing for the fact that the recovery of applied N - from either fertiliser or manures - is never 100%.

But a key part of the new guidelines is the overall philosophy that N management is both uncertain and often imprecise. Factors, such as soils, varieties, rotations and weather all affect N management, and often the processes affected and their interactions cannot be easily assessed, or accurately predicted.

Therefore it is imperative to review N use after harvest to judge the success of the season's strategy, and to help shape the following season's plan.

And that's where Prof Bradley started with Mr Whitehead: How successful had his N strategy been in the past?

Fortunately Mr Whitehead keeps records of grain protein levels for all his crops, whether it is feed or milling wheat. Grain analysis provided the best indicator of optimal N management, although it was not infallible, Prof Bradley explained.

For feed wheats growers should be looking for around 11% protein content and 12% for breadmaking wheats, although the latter figure is less helpful because breadmaking wheats obviously often receive non-optimal N applications for yield, to boost protein levels to reach the market specification.

In addition, investigations suggest grain analysis is about 70-80% successful in identifying crops that have been over- or under-fertilised with N, so it is better to average results over several fields or years before drawing conclusions.

Mr Whitehead's figures for the past three seasons were impressive, particularly for feed wheats. The low was 10%, the high 12%, with the vast majority very close to the target figure of 11%, as evidenced by an average of 10.9%.