Publications

Predicting grain protein to meet market requirements for breadmaking and minimise diffuse pollution from wheat production


HGCA PROJECT REPORT 483

Predicting grain protein to meet market requirements for breadmaking and minimise diffuse pollution from wheat production


by
Richard Weightman1, Laura Fawcett2, Roger Sylvester-Bradley1, Steve Anthony2, Dhan Bhandari3 and Colin Barrow4

1ADAS UK Ltd, Battlegate Rd Boxworth, Cambs CB23 4NN
2ADAS UK Ltd, Woodthorne, Wergs Rd, Wolverhampton WV6 8TQ
3Campden BRI, Chipping Campden, Glos GL55 6LD
4Bruker Optics Ltd, Coventry CV4 9GH

October 2011

Abstract

The aim of this project was to develop a system to aid decisions on the use of foliar sprays of urea during grain filling to boost grain protein of milling wheat crops.

Reference data totalling 1,210 measurements from six N response experiments and 246 commercial fields in East Anglia over three harvest years 2007-2009 augmented 219 data points from a previous project; these were used to calibrate Near Infra Red (NIR) assessments of moisture and nitrogen (N) in ears and whole plants at flowering and milky ripe (MR) stages.

Plant N% at and after anthesis related clearly to grain N%, and hence to grain protein content at harvest. Relationships were better at the MR stage than at anthesis, and they were as good for ears alone as for whole plants. For high yielding varieties (e.g. Solstice, Einstein, Xi19) 2.0% ear N was indicative of grain with 2.28%N (13% protein), and differences from 2% ear N indicated equivalent differences in grain N%, hence 1.8% ear N related to ~12% grain protein.

For low yielding varieties (e.g. Hereward) 1.8% ear N indicated 13% grain protein. Plant N could be analysed rapidly and directly by NIR (e.g. in a local laboratory) or by the Dumas method after posting samples to a remote laboratory. The predictive precision of both methods was similar. The Dumas method is widely used in many labs, so little extra capital investment may be necessary. The NIR method can be used for numerous applications, and can provide the quick turn-around required for fertiliser decisions. After testing and discarding semi-mechanistic models that took account of weather and yield forecasts, a 'best' grain protein forecasting system was developed.

This system accounted for measured ear N at the MR stage, a variety factor to distinguish older varieties (e.g. Hereward) from modern higher yielding varieties, and a further factor that accounted for regional and rotational differences between trial conditions and farm conditions. Cost-benefit analysis of late urea spray strategies were conducted with or without 'best' predictions of grain protein, hence taking account of whether extra premium was expected due to a spray, and considering premium levels (plus possible deductions), expected grain production (yield x hectares) relating to the spray decision, and the cost of fertiliser (including application costs).

Because of imprecision, results showed only a few circumstances in which a strategy of applying late N according to ear N analysis field-by-field and year-by-year proved better than strategies of never applying late N (when premiums are less than £20 per tonne) or always applying late N (when premiums are more than £20 per tonne).

However, the benefits of ear N analysis improved when predictions were applied across a group of growers over a number of seasons (from £6 to £61/ha with different scenarios).Thus ear analysis should best be used strategically (several fields in one year, or several years on one farm), rather than tactically (for single fields in single seasons).

Indeed, the farms studied here showed consistent differences in protein achievement; these may be inherent and unavoidable, or they may indicate persistent on-farm inaccuracies in N management. In either case, ear N analysis appears to offer a useful additional diagnostic tool, to augment measuring soil N and grain yield in support of good N management.

 

HGCA Project Number: 3211
Price: £17.64

Related Publications

Document downloads

View printer friendly versions of these publications

Download this publication PDF
Download this publication PDF
Project_report_thumb