GMO Testing Guidelines

• Strip tests, which analyze the protein expressed by the DNA, can have an important place in the production system. They are the only rapid, on-site method for GMO screening.
• Strip tests may be used for an initial screen, for instance when unloading a truck and before dumping that truckload into a bin.  Use of strip tests can prevent the accidental mingling of a truckload of GMO beans with a bin of non-GMO beans.
• While a producer may use strip tests as an initial screen, it is essential to have a strong identity preservation (IP) system that incorporates the more rigorous and sensitive PCR testing method at critical points in the production process.

• The PCR (polymerase chain reaction) test method, which analyzes the DNA directly, is conducted in a laboratory. PCR is recognized as being more sensitive and reliable than any other GMO test method.
• The PCR test method can detect all commercialized GMOs in agricultural products, can precisely quantify the amount of genetically modified DNA in a sample, can detect GMO content in a wide range of processed foods, and can detect GMOs in extremely low concentrations.
• PCR is the industry standard used worldwide to verify contracts, ensure regulatory compliance, and validate non-GMO claims.

• PCR testing gives the most useful information and is most cost-effective when the sample to be tested is a single-ingredient product in which the DNA is still largely intact.
• Examples of good sample types to test: soybeans, soy flakes, soy flour, soy grits, corn grain, corn flour, corn grits, corn meal. These sample types have DNA that is sufficiently intact to allow for valid quantitative analysis by PCR.
• Certain processing methods (involving heat, change in pH, etc.) may degrade the DNA in highly-processed products.
• Examples of highly-processed products in which the DNA may be significantly degraded: lecithin, soy sauce, soy oil, dextrose, corn syrup.
• To know the true GMO level of a highly-processed product, use this rule-of-thumb: test the input(s) from which the product was produced, and then maintain traceability to the end product.
• Example:  In order to know the true GMO level of a lot of soy sauce, the producer first tests the lot of soybeans or soy grits from which the soy sauce is made. He may then also test the final soy sauce product as an additional step in his quality system.
• Example: In order to know the true GMO level of a highly-processed product such as lecithin, test the lot of soybeans from which the lecithin is made, rather than relying solely upon testing of the lecithin itself.
• The intention of the Non-GMO Project is to drive testing to the most efficient and critical points in the production chain, namely the consolidation of the lot of the single-ingredient product (e.g., the crop) and the seed from which it was produced.
• Even though the emphasis of the Non-GMO Project is on testing of single-ingredient products early in the production chain, multi-ingredient and/or highly-processed products may (and likely will) be tested anywhere in the production chain by any participant or by the Project’s own surveillance testing program.
• Variance #7 to the Non-GMO Project Standard allows for inputs to be verified as compliant with the Standard based on testing alone at any stage of the production chain if
1. PCR testing indicates that the GMO content of the input in question is below the relevant action threshold set by the Standard; and
2. Appropriate laboratory controls indicate that the DNA of the input is sufficiently intact to allow valid quantitative analysis by PCR.

Inputs that do not meet this criteria, and are therefore not “testable” in this manner, must be verified by lot-specific traceability back to precursors to the input that are testable.

• The sample sent to the laboratory should be taken in a manner that insures that it is statistically representative of the larger lot of material.
• To comply with the Non-GMO Project Standard, sampling plans must be designed to achieve 90% confidence in quantification of GMO at the action threshold set by the Standard.
• Technical guidance on sampling procedures can be obtained from GIPSA, ISO, GAFTA, and other international sources. The GIPSA Grain Inspection Handbook is available at this link: http://www.gipsa.usda.gov/Publications/fgis/handbooks/gihbk1_insphb.html

• The recommended sample size for testing whole beans and grain contains at least 10,000 beans/kernels. Samples of this size range are not too large to ship and process yet they offer a high degree of statistical strength (much greater statistical strength than can be achieved, for example, using a smaller sample size such as 500 beans/kernels).
• As a general rule, a 100-gram sample size is sufficient for ground products such as flour.
• See table below for examples of recommended sample sizes.

• Plant species with risk of GMO contamination: alfalfa, canola, corn, cotton, papaya, soy, squash, sugar beet, etc.
• When testing a sample using PCR analysis, the at-risk species in the sample determines which primer(s) must be used.  For example, a canola seed sample is tested using different primers than is a soybean sample.
• See the Real Time PCR Test Tables for examples of appropriate PCR tests for different crop species.

• The GMO testing laboratory should be accredited to ISO 17025, the international standard for laboratory accreditation, and should use methods that are included within the scope of their ISO 17025 accreditation, for the crops/inputs in question.

• If a sample is not expected to contain any GMOs, a Qualitative (yes/no) PCR test may be used.
• If a sample may contain GMOs, a Quantitative PCR test will probably be more useful.

• A Qualitative (yes/no) test result will show that GMOs are “Detected” or “Not detected” at the limit of detection of 0.01%.
• If GMOs are detected, the sample can be upgraded to a Quantitative test.  If GMOs are not detected, no further testing is needed for that sample.
• A Qualitative test is generally less expensive than a Quantitative test.

• A Quantitative test result will show the actual % GMO in the sample.  For example, the result might be “0.3% GMO”.