Have you wondered why the milk level in the bulk tank suddenly was not as high as expected? Or, why the fat or total solids analysis was not as good as it was a month ago. Could the culprit be the new load of TMR, new haylage, or new silage? Forage and feed analysis may be able to help determine the root of the problem. But more importantly, regular sampling and analysis can help prevent problems from occurring.
What is Feed Evaluation?
Feed evaluation entails three steps, which are often confused (Murray, 1996). The first step is to ensure a feed material is what it is stated to be (qualitative analysis). The second step is to quantify its ranking among like feeds (quantitative analysis). The third step is to determine (or predict) the likely animal performance which may be expected from feeding the material to a target animal (nutritional analysis). All three steps are important and may apply to individual feed ingredients or to mixed feeds. Traditional feed analyses (wet chemistry) are usually used to evaluate step three, but can only be used if the analyst knows the identification of the feed material. For instance, certain analyses (such as fiber and lipid) are different for different kinds of feed materials. There is one method of analysis that can perform all three steps in feed evaluation quickly and accurately. That method is near infrared reflectance spectroscopy (NIRS).
Qualitative analysis can be accomplished using NIRS by identifying different feedstuffs based on their spectral fingerprint. NIRS has been used to determine the percentage composition of hay and clipped pasture samples. Quantitative analysis with NIRS has been used effectively in the wheat industry to determine hardness and baking quality (Williams and Sobering, 1992). It should be very effective in the feed industry to identify heat damaged feed ingredients, and to determine excessive moisture or other adulterants.
The importance of nutritional analysis of feedstuffs has been recognized since the Weende system was developed in Germany in the 1800s. Feeds were characterized for moisture, ash, crude protein, crude fat, and crude fiber content. Nitrogen-free extract was calculated as the residual. Since then, various feeds have been analyzed for chemical composition using this system and modifications such as those introduced by Van Soest (1963). However, for a nutrient to be useful, it must be digested and metabolized by the animal. The Weende system was further developed to include digestible fractions of each nutrient, and the sum of those was called total digestible nutrients or TDN. This description has given way to the more accurate net energy system which partitions the digested energy into energy available for maintenance and either growth or lactation. However, prediction equations for net energy for the new highly-digestible varieties of corn silage need to be evaluated.
One may think it is too expensive and too time consuming to obtain an analysis of feedstuffs. Dairy producers soon find out if the lactation ration is deficient in a key nutrient via the milk tank. However, by then production losses have occurred. Ten to $20 for an analysis is cheap compared to the losses incurred due to ration deficiencies or imbalances. Frequent forage analyses equate to a more accurate, consistent ration. In general, cows fed consistent rations will milk better.
Proper Sampling
Proper sampling is difficult. Many producers may think they do not have the resources to take the recommended number of samples to properly characterize feedstuffs. There was a time when the solution to nutrient variability, i.e., crude protein, was to over formulate for the nutrient. This thinking is no longer economically nor environmentally feasible. With economic and environmental concerns, one needs to ensure diets furnish not only the proper levels of individual nutrients, but that each is balanced so that maximum efficiency is achieved.
A good sampling philosophy for providing the proper nutrients and in the proper ratios is to start with farm-grown and purchased feedstuffs, such as forages and grains. Strict controls on the most variable ingredient will provide the greatest reduction in variability in the ration. Normally, the most variable feed will be farm-grown or purchased forages and grains. Commercial feeds should be less variable in nutrient content for the nutrients listed on the label guarantee.
However, the nutrient content in some commercial feeds can vary depending on quality control measures at the manufacturing plant. When formulating rations from individual feedstuffs, uncertainties associated with variability are cumulative and therefore, are a very real concern. The only way to manage the increased variability is with more frequent sampling and analysis (Murray, 1996). However, a feed ingredient with an expected protein of 19%, which only contains 14% can have serious implications on animal performance. Improper sampling, including too few samples to represent a lot, can be just as bad or worse than failure to analyze. Both can lead to a loss in production or to extra costs for unnecessary nutrients.
How many samples should be taken? The samples must represent the silo, lot of hay, or batch of feed in question. A representative sample is defined as a small fraction from a lot taken in such a way that a determination of any particular characteristic of this fraction will represent the mean value of the characteristic of the lot. Therefore, the number of samples depends on the size and variability of the lot. Ten is the minimum number of samples. For most purposes, the average composition is all that is required. In this case, the samples may be bulked together and a subsample taken for analysis. The subsample is best obtained by quartering. Store the final sample for analysis in a container that will prevent changes either by contamination or by degradation until the analysis can be performed.
Are there Different Methods of Analysis?
The answer to this question is yes. Traditional methods of analysis included chemical procedures designed to isolate the nutrient of concern (e.g. protein) and measure the amount in a given quantity of feed. Traditional methods are laborious, time consuming, and produce toxic chemical wastes. An alternative procedure is NIRS, a chemometric procedure introduced three decades ago for the analysis of agricultural products.
NIRS is a fast, nondestructive method of analysis that can be accurate and reliable. In operation, monochromatic light is focused on a feed sample (usually powdered). The feed particles either absorb, reflect, or scatter the light depending on the chemical bonding of the individual constituents. Different wavelengths are absorbed when the frequency of light matches the vibrational frequency of the chemical bond. Thus, a very specific analysis of that particular bond and its association with other bonds can be measured. Most absorbances in the near-infrared region are generated by hydrogen bonded to carbon, nitrogen, or oxygen, the basic building blocks of all organic molecules of which feedstuffs are composed.
NIRS has been criticized because there are errors associated with the predicted value for nutritional value (or other evaluation). Errors are defined as the difference between computed or measured values and the true or correct values. The truth is that all methods have errors associated with them, including the traditional methods used for reference. Since NIRS is a predictive method, it inherently has built in all the errors associated with the reference method plus other errors that could occur. Typically, the combined errors for NIRS analysis are about twice that expected for the traditional reference method. The greatest error is that of the sample representing the entire lot of hay or feed, which is actually a sampling error, not an error in the ability of NIRS to accurately determine feedstuff composition. Analytical errors are small in comparison. NIRS has proven that it is capable of producing repeatable results for all three steps in the evaluation process--and it can do it in about one minute. Instruments can be networked so that all give the same results. Traditional laboratories often give very different results on the same sample.
In summary, prevention of nutritional problems due to nutrient imbalances requires frequent sampling and analysis of feedstuffs and/or mixes. The method is not as important (so long as it is standardized and approved) as obtaining a representative sample and using the results. NIRS is a technique that can provide multiple analyses in one pass, quickly and reliably.
