Maize silage NIRS


Results Summary:

  1. A total of 90 samples of maize silage were collected from commercial farms in the UK over two years. Silage DM contents ranged from 211 to 436 g/kg as fed, crude protein from 64 to 112g/kg DM, starch contents from 98 to 367 g/kg DM and neutral detergent fibre (NDF) contents from 349 to 646 g/kg DM. The large variation of the samples covered most of the extremes of UK maize silage and allowing a wide range for equation NIRS development.

  2. The predicted metabolisable energy contents estimated in sheep ranged from 9.1 to 12.0 MJ/kg DM. The range of values were similar to those used in the study by Givens and Deaville (2001), suggesting that there has not been a substantial change in the last ten years in the composition and nutritive value of maize silages.

  3. The in situ rate of DM degradation was lower than silage values in the current Feed into Milk (FiM) database and therefore the calculated effectively rumen degraded DM content (using the FiM model) was low. This may reflect a shortcoming of the FiM model for maize silage. To overcome it is proposed that for maize silage it is a more rational to assume that a large proportion of the starch particles (i.e. the a fraction) degrade at the same rate as the soluble components (i.e. the s fraction).

  4. The high N solubility coupled with the low N content of maize silage meant that the N degradability data from most of the silage samples could not be fitted to the Orskov and McDonald (1979). If single values for a, b and c are accepted (calculated from the mean of the 90 samples), the estimate of effectively degraded N content of a maize silage would then be moderated by variation in its N content and N solubility.

  5. Starch in the maize silages was highly degradable and was completely degraded within 24 h. Although starch degradability is not required by the FiM model, application of the FiM assumptions to starch gave very low effective degradability. If a modified FiM model is used where the a fraction was treated as soluble, more realistic estimates of effectively degradable starch are produced.

  6. Acceptable wet silage NIRS equations were developed for DM, crude protein, starch, NDF, pH, DM a and soluble N. No NIRS prediction equations could be generated for the FiM b and c terms for DM and starch.

  7. The Bioparametrics gas production method may be able to predict the FiM DM degradability parameters a plus b, but was not able to separate the a term from the b term. The c term could not be predicted by the gas production method.


Planned activity:

  • The information gathered on the maize silage type and quality will be used to better inform farmers on the quality of feed they might expect from the ensiling maize.

  • The digestibility results from this study will be combined by the FAA group with a further 45 silages in an attempt to improve the prediction of maize energy value. The wet chemistry components will be combined with the existing FAA database for maize silage to improve the prediction of these parameters. The FAA group plans to carry out this work during the summer for implementation of these equations during the autumn of 2012. The implementation of appropriate and reliable wet NIRS equations by the UK FAA group will allow fast and accurate estimation of the nutritive value of maize silages.

  • The Bioparametrics gas production technique does not predict the FiM degradation characteristics of maize silage.

  • The suggested modification of the Feed into Milk model for the dry matter and nitrogen degradation of maize silages will be presented to the UK ruminant feeding industry. Further work by them will be required to test the effect of these modifications in practical rations.

  • Results of the study will be circulated via consortium member’s staff, scientific and industry papers (Maize Growers Association (February, 2012), Society of Feed Technologists (April 2012)) and a series of farmer and industry articles.

  • EBLEX will work with DairyCo to communicate the implementation of the new calibration equations to industry when they are in use in commercial laboratories across the country.

Beef & Lamb
Project code:
01 December 2009 - 31 October 2011
AHDB Beef & Lamb, AHDB Dairy, Defra LINK, Feed into Milk, MGA, ABAgri, David Bright Ltd, Richard Keenan & Co, Syngenta Seeds Ltd, Biotal
AHDB sector cost:
Total project value:
Project leader:
University of Reading


74202 Gas Production Report 2012 74202 Executive Summary Apr 2012 74202 Eurofina Equation Report 2012 74202 University of Reading Consortium Report 2012 74202 University of Reading Appendix 2012

About this project

The Problem:

Currently there is no agreed UK equation to predict energy content of maize silage. Consequently poor estimates of nutritive value prevent the formulation of accurate and balanced diets containing maize silage. Furthermore, work is needed to identify the level of variation in maize silages produced on farm and investigate methods to predict their degradation characteristics using NIRS.


Project Aims:

  1. Generate accurate and repeatable NIRS equations to predict maize silage organic matter digestibility and hence ME value
  2. Generate accurate and repeatable NIRS equations to predict DM, starch and N degradability of maize silage
  3. Evaluate a cheap, rapid and reliable gas production system to replace the in situ method for the estimation of DM, starch and N degradability



Over two years (2010 & 2011), a wide range of 90 maize silages were collected from various UK farm sites.  All silages were wet and dry NIR scanned and analysed by wet chemistry methods as appropriate. Organic matter digestibility (DOMD) of each silage was also measured in sheep.

Dry matter, starch and nitrogen (N) degradability was measured in situ in cows using the polyester bag technique at the University of Reading, and by the gas production technique by Bioparametrics Ltd.

The digestibility results were combined with existing FAA data to generate a new NIRS prediction equation for DOMD. Wet silage NIRS degradability predictions will be developed from the polyester bag results. In addition, results from the gas production technique will determine whether this method could replace the in situ approach to measuring degradability.