Food, Agriculture and Natural Resources Policy Analysis Network (FANRPAN) Food, Agriculture and Natural Resources Policy Analysis Network (FANRPAN)
Site map|Contact us  


International Seed Testing Association (ISTA) progress on testing Genetic Modification (GM) seed presence

Genetic trueness-to-type is at least as important to seed quality as physical purity, germination, vigour, and freedom from seed-borne diseases. Modern biotechnology, specifically genetically modified varieties, has now brought a new requirement for distinguishing GM seed from non-GM. This requires more complex diagnostics at the molecular level. ISTA - the International Seed Testing Association - is the international body that sets guidelines, procedures and standards for seed testing, and has been running capacity building through workshops and GM testing since 2002. Its recent report highlights the complexities involved in testing methodology.

Sometimes different varieties can be visually identified by their seed characteristics, at other times by growing the plants to determine plant phenotype. In most cases GM seed cannot be visually distinguished from its conventional counterparts and molecular diagnostics have to be applied. Two basic detection methods are used: checking presence of the specific gene DNA or testing its specific protein. The test may be qualitative of quantitative. Mostly, seed merchants want to know what percentage of GM seed is present in non-GM. The converse argument is also true. GM seed needs to have negligible presence of conventional seed otherwise the latter will suffer damage from herbicide application or will succumb to attacks by insect larvae. Unintentional GM seeds in non-GM, is generally referred to as adventitious presence or AP.

The ISTA has conducted international proficiency testing over 6 years by more than 40 accredited national seed laboratories, based on carefully blended samples of maize and soybean seed at levels of 0.1, 0.5 and 1% level of AP of GM seed. Based on this experience, ISTA has developed the following draft set of positions on GM testing.
  • The so-called dipstick method is quick and cheap but only indicates whether or not GM material is present in a sample and cannot show percentage. It is also less reliable at low levels of presence.
  • Real-time PCR is still considered the best quantitative test. This is based on calibration curves using Certified Reference Material and requires exact proficiency. However, there are also other methods based on linking protein-based or end-point PCR results to bio-assays of plants or seeds. Following prescribed sampling protocols in seed lots is essential. Results can best be described in terms of % of seeds.
  • Units of expression can be % number of seeds, % mass of seeds, or % copies of DNA. The latter may be ambiguous as reference material may have different number of sets of chromosomes. Calibration curves should be based on single sets.
  • Biological factors for both reference and GM material may have an impact. The endosperm has three sets of chromosomes and makes up 80-90% of kernel weight, while the endosperm has two sets, and pollen only one. This affects the number of DNA copies.
  • A single test on a seed lot cannot fulfill all needs. Sub-sampling allows testing for single and stacked gene events. Real-time PCR cannot do this on a single bulk sample.
  • There is no conversion factor between % number of seeds and % DNA copies. Therefore, more than one unit should be acceptable in reporting. Variability in lot sampling must also be taken into account.
The ISTA experience shows that seed testing labs need to go through proficiency and referee sample testing until a high standard of proficiency is attained, and that modern diagnostic equipment, facilities and trained staff are required.

An incident in Kenya is a case in point. A certain party alleged that a maize seed shipment from South Africa to Kenya was co-mingled with GM seed. The authorities (KEPHIS) investigated the case but could not obtain any sample from the party or details of the test method. The several seed lots were then extensively sampled according to protocol and tested by KEPHIS, and no presence of GM seed could be detected.

More information on GM seed detection can be accessed on the ISTA website at:

Top of page   -   Home   -   Contact us   -   Disclaimer
Food, Agriculture and Natural Resources Policy Analysis Network
FANRPAN Remote Access FANRPAN Webmail
Octoplus Information Solutions