Where to Send Samples for Analysis

Once you have gone to the effort of correctly collecting a sample, how can you be sure that the results you receive from the testing laboratory are accurate?
Where to Send Samples for Analysis - Articles


Frequently, concerns about laboratory testing focus on the methods used in determining forage quality. However, the focus of concern should be on the accuracy of results and not the technique of obtaining the results. To help you determine if the test results are accurate or not, we have outlined below some questions to ask the laboratory manager.

  1. Is the lab certified or does it participate in a check-sample program (also called proficiency testing program)? The National Forage Testing Association (NFTA) certification program monitors the performance of a lab against other labs to alert them of potential problems in their accuracy. The American Association of Feed Control Officials conduct a check-sample program that insures consistently and accuracy amoung participating labs. Involvement in either of these programs indicates that the laboratory is concerned with the accuracy of its results.
  2. Does the lab include duplicate samples or quality control check samples in each group of samples analyzed? One of the easiest ways for a laboratory to monitor results is by analyzing replicates of a sample. If the analyses for replicates are not similar, then there is a problem in the testing procedure. In addition, the inclusion of standards or check samples (material of known quality) in each group of samples analyzed can indicate if the analytical procedure is working correctly or not. Standards or check samples can also alert the laboratory technician of small changes in results over time and allow corrective steps to be taken.
  3. What analytical methods are used by the laboratory? There is more than one method of analyses for most plant constituents. Laboratories should be using methods of analysis which are well validated and approved by the Association of Official Agricultural Chemists (AOAC).
  4. Laboratories which use NIRS can be asked three additional questions that will help determine if the results are accurate. Like other laboratory methods, NIRS analysis is sophisticated and should be conducted and monitored by trained personnel.
  5. How frequently are the NIRS instrument and the calibration equations monitored? NIRS instruments should be monitored by running a check sample daily or after every 25th sample, whichever is more frequent. Calibration equations should be monitored by conducting laboratory analyses on every 25th sample. Again, this additional monitoring adds additional costs which will increase the fee charged for each sample.
  6. Does the laboratory do chemical analysis in addition to NIRS? NIRS methods are based on calibrations derived from chemical methods. NIRS labs which have no chemical analytical capability have no method within their lab to monitor the reliability of their calibration equations. It is not impossible for a NIRS-only lab to have a good monitoring program. But it is much more difficult since all of the monitoring samples would have to be sent to another lab for chemical analysis.
  7. How does the lab identify and analyze inappropriate samples received for NIRS analysis? Each NIRS calibration is specific for a particular type of sample. For example, corn silage is most accurately analyzed with a calibration equation developed for corn silage and not a calibration equation developed for alfalfa haylage. How does the lab identify samples that are inappropriate for the calibration equation and then does it have a protocol for analyzing these samples?

Keep in mind that laboratory monitoring practices increase the cost of the analysis. Asking these 6 questions will help evaluate a laboratory and is one way to become more knowledgeable about purchasing analytical services. Laboratories generally report results of analyses as a single number. This does not mean that hay which tested at 20% CP is exactly 20.0% CP. Instead, it means that the hay is 20.0% crude protein plus or minus some variation. The amount of this variation will differ from lab to lab and from method to method. A variation of about 3% can be expected between labs for measurements of crude protein. In other words, a hay sample which tested 20% CP at one lab would be expected to test anywhere from 19.4 to 20.6% CP at another lab or at the same lab if the analysis was repeated. Variation is usually much higher for fiber measurements than for crude protein measurements.