Risk assessment and risk management of specific receiving environments

From a scientific/ecological perspective, any import of any domesticated species poses a risk when introduced into a fragile ecosystem such as a small island with limited or specialized biodiversity, whether the species is transgenic or not.  The risk is less, but not insignificant, if the ecosystem is robust.  Thus, patches of inadvertently released imported species are often
observed in ruderal, human disturbed areas along roadsides leading from ports of entry to processing plants.  Such patches usually remain limited and/or disappear with time, and do not invade agricultural or natural ecosystems.

Such historical introductions of non-transgenic material serve as an  appropriate baseline for predicting whether an LMO will establish and become a problem.  For example, if a country has been importing soybeans for decades, one can ascertain whether there have been established patches, and if there have been patches, whether fertile hybrids have established with indigenous species.  Then, to compare transgenic glyphosate resistant soybeans to this baseline for a risk analysis, one must know from the historical evidence whether there have been problems from introduced soybeans.  If glyphosate is not used along those roadsides, the risk from transgenics is no different than the past, acceptable risk.  It is only when the crop itself poses a risk that one need ascertain whether the risk is greater due to the traits encoded by a transgene.  In other words, special considerations for LMOs are needed only where there is a known historical risk from non-LMO introductions, and where the transgene encodes a strong selective advantage to the imported LMO.