22 November 1996


Silage analysis techniques must be standardised to avoid current confusion and discrepancies between different labs.

Jonathan Riley reports

GRASS silage can account for up to 70% of a winter diet and so accurate analysis is vital to produce cost effective rations. But the use of different methods of analysis is leading to confusion and discrepancies in results between different labs.

"The problem arises because the metabolisable energy content of a producers silage sample can only be measured by offering the sample to an animal and assessing the level of energy retained," says ADAS consultant Bruce Cottrill.

"So several methods of predicting energy based on the components of the silage sample have been developed.

"Traditionally a method called modified acid detergent fibre – MADF – has been used. This makes a prediction of energy content based on the level of fibre in a sample," says Dr Cottrill.

He explains that theoretically as fibre content increases, energy increases. But that in practice as grass matures a form of fibre called lignin is deposited in the stem. Lignin is indigestible and so does not yield energy, but as it is a fibre that the MADF system includes it in the calculation and so overestimates the energy content.

"This leads to inaccuracies and Scottish Agricultural College studies have shown that 45% of samples analysed in this way are inaccurate by 0.5MJ/kg ME.

"This method cannot be relied upon but is still used by some labs using wet chemistry and by some labs to calibrate near infrared spectroscopy (NIRS) machines. Samples sent to these labs are prone to huge variations in predicted energy content," he says.

He advises producers who are concerned to ask labs which wet chemistry method is used and how NIRS machines are calibrated.

"Any NIRS machine is totally dependent on the way it has been calibrated for its accuracy. In simple terms the machine shines a light on a sample and measures the wavelengths of the reflected light," explains Dr Cottrill.

It then interprets the different wavelengths from the sample by comparing them with wavelengths from different silage samples programmed into its memory. From this information the machine can provide a breakdown of the silage components.

"Therefore, the higher the number and wider the range of samples used to calibrate the machine, the more accurate it is at interpreting the silage components," explains Dr Cottrill.

But different data sets used for different machines could lead to discrepancies in results between labs presented with the same silage sample and so a single standard, the ME Tick system, is being established.

A major advantage of NIRS is that results can be returned to producers more quickly than with wet chemistry. And the machine does not require specialist operators to run it making it potentially very cost effective and useful for producers wishing to know feed values quickly, says Dr Cottrill.

Most NIRS machines are calibrated using dried and milled samples of silage but BOCM Pauls are now using NIRS on fresh silage samples.

"This cuts time taken and costs needed to analyse a sample still further," says BOCM Pauls nutritionist Bruce Woodacre. "Wet scanning can also measure some components which would be burnt off in any drying procedure used for dry scanning," says Mr Woodacre.

However, Genus management consultant Derek Gardner says that as a relatively new method, the data set of silage samples is still small and wet scanning is less able to interpret samples which produce extreme results. For example samples that have been contaminated with soil.

"The most accurate method of assessing ME would be to use rumen liquor as a digester, but because this involves the use of fistulated animals it is unacceptable.

"The closest method to this is to use wet chemistry using the enzymes cellulase and gamminase – responsible for digestion in the rumen – mimicking digestion in the animal.

"The technique can be used to spot errors in samples such as soil contamination and can assess butyric and lactic acid levels which indicate how good the fermentation process has been. But because a skilled operator is needed and because the technique takes longer than NIRS costs can be higher," says Mr Gardner.

With grass silage making up to 70% of a winter ration for lactating dairy cows, accurate analysis is vital.