Create a free Feed Strategy account to continue reading

New directions for near-infrared reflectance spectroscopy in the feed industry

Five decades have passed since the earliest attempts to assess cereals by spectral analysis and the first predictive equations were derived from reflectance spectra. It is already more than 40 years ago that research showed how to employ spectroscopy in estimating the quality and digestibility of forages for feeding to ruminants.


Five decades have passed since the earliest attempts to assess cereals by spectral analysis and the first predictive equations were derived from reflectance spectra. It is already more than 40 years ago that research showed how to employ spectroscopy in estimating the quality and digestibility of forages for feeding to ruminants. Yet new applications in feed manufacturing still continue to appear for the spectroscopic technique known as near-infrared reflectance (NIR).

Those most closely involved in developing these novel applications quickly dismiss any suggestion that the NIR story is almost complete. They say that its recent move from the laboratory into the feed mill for online scanning promises to revolutionise the way in which mills operate, and insist that the value of the technique will be extended beyond its current role.

While the flour industry most often uses NIR to determine the milling characteristics of wheat along with its protein level and moisture content, expect grain analysis by spectroscopy to have a more profound impact on the animal feeding of the future.

Greater understanding  

A widespread view until recently had been that the natural variability of farm livestock would frustrate attempts to predict grains’ nutritive value with any accuracy. However, a report in the journal British Poultry Science offered a different perspective when it indicated that NIR data on whole wheat could support a prediction of what the grain would mean for the daily growth rate and gain-to-feed ratio of broiler chickens.

“We found that our NIR calibrations of samples of whole wheat were good for predicting the live weight gain and feed conversion efficiency of broilers,” reports research team member Dr Elizabeth Ball from the Agri-Food and Biosciences Institute in Northern Ireland. In studies, 94 wheat samples were selected for inclusion in three trials involving chickens grown in metabolism cages to 28 days of age. The wheat came from British and Canadian locations and represented different varieties as well as various years. It was scanned in both whole and milled forms.

“Predictive equations for growth performance have been produced from this study relating to whole wheat,” she continues. “The requirement now is to find a commercial company who will help us to validate them across more birds.

“Today, NIR is firmly established as a rapid way of finding the chemical composition of grains. The calibrations of most companies for this purpose have been fully tested and validated. Predictions of nutritive value are less well developed by comparison. It is an area of growth or opportunity, but also of some difficulties because it involves more than straight chemistry – there is the animal factor to consider.

“Solving this by research takes a long time and large numbers of animals or birds in testing. However, it is within reach and ultimately it could save feed manufacturers time and money.”


The latest generation of NIR instruments has advanced by becoming more accurate and easier to use, notes Stefan Tordenmalm at equipment supplier Perten Instruments in Sweden. These advances have made them an excellent investment also for smaller plants to use in determining the composition of ingredients for cost-efficient formulation and monitoring the process for maximum output.

Additionally, modern diode-array instruments are fast (needing only a few seconds to perform an analysis, with no grinding of samples required) and are robust, attributes which allow them to be moved into the mill and used in the processing line to give direct real-time online measurements.

“When it comes to new applications we have noticed a substantial interest in our recently launched Online NIR instrument for feed optimisation. If there is a trend right now, it would be online feed analysis. Something else which we have noticed is an increased interest in linking the NIR with the least-cost formulation software. The integration can work both ways – getting NIR results directly into the ingredient database, but also getting the specifications from the formulation into the NIR so operators can see target values and actual values in the same graph in the NIR instruments.”

Greater role  

It has become fairly common practice for feed manufacturers to use NIR in monitoring supplies of raw materials, but Australia has seen calls for energy values from rapid NIR measurement to be made the industry standard in valuing feed grains at the time of trading or marketing. Research management consultant Dr John Black says that current measurements used for trading the grain (protein content, test weight and screenings percentage) do not give a good representation of either the available energy content or the available energy intake that is primarily responsible for the level of animal performance.

Under the research programme Premium Grains for Livestock that ran from 1996 to 2008 evaluations were checked on over 350 cereal grains that had been selected or acquired from some 3,500 samples of wheat, barley, oats, triticale and sorghum. The energy value of individual grain samples was shown to vary widely between and within grain species and species. Batches of grain could vary by 3-4 MJ/kg in their available energy content for pigs and poultry, Dr Black has reported, and by about 1 MJ/kg for ruminants.

Moreover, when the same grains were fed to cattle, pigs, broilers and layers, the results for available energy values suggested that individual grain samples were often more suitable for feeding to one animal type than to another – probably because of differences in modes of digestion.

Current methods for assessing grain quality cannot readily distinguish between grains of higher energy value for livestock and those of lower value, Dr Black argues, whereas near-infrared calibrations newly developed in Australia can help make these distinctions. In addition to ones for predicting the available energy content and available energy intake index for cereal grains for some common livestock types, other NIR calibrations have been produced to predict the likelihood of the grain causing acidosis in ruminants.

“These NIR calibrations are forecast to be the new method for measuring grain quality and to form a more rational basis for trading grains for livestock than the measurements used currently,” says Dr Black.

Long-term benefits  

Their commercialisation in Australia is being handled by AusScan, which is a business arm of the Pork Co-operative Research Centre. “I see the access to NIR energy determination for grains as providing the feed industry with a significant stimulus to better utilise NIR technology,” AusScan project manager John Spragg has said.

“To date, NIR use has been for QA purposes, assessing ingredient and finished feed moisture and protein. The use of available energy content and available energy intake has a greater impact upon end-use livestock performance. The greatest benefit can therefore be derived by integrated companies, although there are benefits through the supply chain that each participant can derive.

“For feed mill application, we are in the early stage of mills working out how they can utilise the additional information on available energy. The immediate advantage they see is in fine-tuning of feed formulations.

“There will be a longer-term benefit, too. It will come in grain breeding, as breeders screen genetic lines for available energy.”

Page 1 of 74
Next Page