Post-release monitoring and long-term effects of LMOs released into the environment

Reflecting on the ecological consequences of the gene flow: one should consider the potential exposure and hazard. The likelihood of gene flow occurring is often considered as the probability of exposure in the risk assessment equation. Gene flow can occur, through the pollen or through admixtures of seed, or volunteer plants.  For gene flow to occur each of the following events must take place:
1. Pollen (male) must effectively be transferred onto a sexually compatible pistil (female);
2. Fertilization must occur and a viable seed must develop;
3. The viable seed must dehisce and  in an acceptable environment for germination;
4. The seed must germinate and establish a fertile plant;
5. The fertile plant must be pollinated and in turn produce a viable seed;
6. The viable seed must germinate and produce a plant.
If any of these events fails, pollen-mediated gene flow will not happen, and seed-mediated gene flow will not occur if events 3 to 6 are not met. The likelihood of gene flow is affected by sexual compatibility, the distance of sexually compatible plant species, pollen viability/longevity, coincidence of flowering, presence of pollen vectors (wind or insect), relative size of populations and gene frequencies within the populations.
Research trials have a very low risk as exposure is dramatically reduced compared to commercial releases, mainly because of their small size and the imposed isolation measures. In the context of regulated environmental release, field trials make up a very small portion of total land area.  Furthermore confined field trials are conducted in such a manner to reduce risk of escape. By this argument even the so called “high” risk outcrossing plant has near zero  risk regardless of trait.
A transgene can under a combination of certain conditions: - be transferred; and be expressed;  and  persist and disseminate to a related wild relative through further hybridization and introgression. This is not different from the case of traditional domestication-traits that were either unconsciously or intentionally selected and changed by man during thousands of years. When domestication traits including loss of seed dispersal, synchronized pod shattering, loss of seed dormancy, failure of protection against herbivores (lowering of toxic substances such as in cassava, potato…), herbicide resistance,  rapid growth, early flowering, increased grain yield, etc. are spread to wild relatives, ecological consequences are possible. It is important to keep in mind that certain traits developed through modern biotechnology have also been produced through traditional breeding (e.g. resistance to herbicides, herbivores, pathogens and adaptation to new environments). It is useful to compare the knowledge gained on the effects of conventionally bred, i.e. non-transgenic crop genes that passed to the wild to the effects of LMOs. Decades of breeding in multiple crops has shown little effect from gene flow. Even for traits that one would suspect would have a selective advantage such as disease resistance, wild species have not uniformly become disease resistance.  This could be because consistent disease pressure/i.e a consistent natural selective advantage, is not present. This is the kind of information to consider when assessing the risk from gene flow, whether from GM or conventional breeding.
The possibility of introgression of a transgene might increase when there is a certain combination of conditions (such as significant possibility of hybridization between GM and wild relative and a resulting fertile and fit progeny).  It is part of the risk assessment (as described in Annex III of the CPB) to consider the likelihood of a transgene to spread and to persist and to consider the potential environmental consequences (the hazard) associated with gene flow from an LMO to compatible plants. This is the reason to consider several points such as biological characteristics of the recipient organism, the introduced trait, characteristics of the receiving environment such as centres of origin and genetic diversity, description of the habitat where organisms may persist or proliferate, etc.  If the resulting plant is likely to cause a negative impact to the specific ecosystem, measures or activities to manage the potential risk should be considered. It is in the case of related weedy species in agroecosystems: con-specific weeds, or at least sharing one genome of amphiploid (for example with rice/weedy rice, sorghum/shattercane and Johnsongrass, sunflower/weedy sunflower, wheat/Aegilops cylindrica, beets/wild beets) when sexually compatible that the risk of gene flow is higher. It is in these cases that efficient containment and mitigation strategies should be considered, as the weedy species are typically adapted to maximize gene flow, especially when there is a selective advantage. The consequences of gene flow is trait dependent. Traits that do not alter the fitness of the plant in any way may not be likely to have an impact even if the recipient plant is a weed.
Documents describing the biology of crops (for example the OECD consensus documents) are important to decide on the introduction to a specific environment and on the risk management options. There are already some on major crops  (such as maize, soybean, rice,….) and the development of such documents on other, including major agricultural crops (such as  sugar cane, Phaseolus, cowpea, Cassava, ….), as well as minor crops or specialty crops that will be the subject of a regulatory decision, is very important.
Concerns on the effects of gene flow in specific cases is not limited to genetically engineered crops, but in the case of transgenic plants, new technologies including molecular tools (e.g. GURTs, targeted transformation, etc.) are under development. The potential risks associated with gene flow could be reduced even further.