Editor's note: The following item was provided by Dairyland Laboratories in Arcadia, Wis. It is the second of a three-part series.
In the first part of this series, we discussed the basics of how NIR works, how drying and grinding affect the spectral properties of the sample, and what spectral ranges are covered by various NIR instruments that are commercially available. In Part 2, we will look at how these factors affect the accuracy of NIR measurements, and then finish up with Part 3, which will propose some possible use cases for NIR on the farm.
All of the current marketing for NIR on the farm, that we are aware of, promotes the use of samples that are neither dried nor ground, and instruments that cover 400-1000 wavelengths. For this type of analysis, we can say that it has the potential to measure moisture with reasonable accuracy and detect changes in the value of other nutrients like protein and fiber.
The dominance of the moisture peak in NIR spectra makes it a fairly easy value to measure. Most of the published studies on this type of feed analysis show moisture measurements with a standard error of approximately 2%. In comparison, most laboratory methods for moisture would have a standard error of 0.25-0.5%.
Moving beyond moisture is much more difficult with on farm NIR. The spectral peaks for nutrients like protein, fiber and fat are tiny in comparison to the moisture peaks. In high moisture feeds like ensiled forages, high moisture corn, and wet commodities like beet pulp or distillers grains, this severely limits the ability to measure individual nutrients. Still, on farm NIR can be used to detect large changes in nutrient values.
For example, if a haylage pile went from 16% protein up to 22%, an on-farm NIR instrument would be able to show us that a change has happened. Alternatively, we could mount the NIR instrument on our forage harvester and make field maps that show which parts of the field yielded high amounts of starch. While these measurements are not nearly as precise as we could have received from a sample submitted to a laboratory, that’s really not the point. The major advantage of having NIR on the farm is the ability to measure many samples in a short amount of time.
Ohio State University recently published these findings from a study it did about the variation of feeds on a farm. The accompanying charts show how much variation was observed on an “average” farm over the course of 2 weeks. The same study found that, over a 2-week period, the average lot of corn silage varied 12 points in starch, and haylage varied 8.5 points in NDF.
So we know that feeds vary, and we know that on farm NIR can detect large changes in nutrient content. Where do we go from here? In Part 3 of this series, we’ll look at some cases where on farm NIR makes a lot of sense, and some other cases where its use is limited.
Read Part 1