Putting problem formulation at the forefront of GMO risk analysis

Figures & data

Figure 1. Steps in the risk analysis strategy used at ICGEB training courses in risk analysis for the unconfined release of GMOs.

Table 1. Prioritization of a selection of the risk hypotheses developed for the cultivation of drought-tolerant (DT) white maize in Kenya (Nairobi 2010 and Trieste 2012 workshops)

Risk hypothesis
	Requiring consideration	Prioritization with small consequences	Why consequences are small
DT maize will lead to increased pest and disease populations when it is planted off-season (dry season	X
DT maize cultivation would lead to increased greenhouse gas emissions.		X	Growing maize contributes little to greenhouse gas production, so any conceivable increase would be minor.
Extension of DT maize hectarage will lead to loss of protected species.	X
DT maize will decrease populations of beneficial soil organisms.	X
DT maize would increase eutrophication of water sources due to increased biomass.		X	No significant increase of biomass expected, and little maize residue is transported to water bodies.
Increased levels of aflatoxins in DT maize will lead to increased toxicity for consumers	X
DT maize consumption will result in increased obesity levels as a result of higher carbohydrate intake in diet.		X	Higher carbohydrate consumption is a desired effect for the poor. No increase in maize consumption among the wealthy is expected, where obesity could be a problem.
DT maize consumption will result in reduced human reproductive health.		X	The transgene-encoded proteins are already abundant in the human diet.
DT maize stems will have lower nutritional value, and thus have a negative effect on humans who chew the stems, particularly children.	X
Supplemental irrigation allows increased cultivation of DT maize, and will lead to increased soil salinity.	X
Consumption of drought tolerant maize will compromise antibiotic treatments.	X
DT maize will not yield higher than conventional maize under water stress conditions, resulting in unexpected crop failure and lack of contingency planning.	X
DT maize will compromise the production of non-GM maize and hence affect maize exports.	X
DT maize cultivation will be extended to new areas, leading to land use conflict with pastoral communities.	X
The adoption of DT maize cultivation will result in the reduction of genetic diversity of landraces	X

Figure 2. An example of how a risk hypothesis can be made explicit by developing a pathway to harm. The three examples presented are taken from a hypothetical case study of drought-tolerant maize for release in Kenya. (A) The participants concluded this to not be a significant problem, since resistance to kanamycin is already common in bacteria in the environment (breaking step 6) and kanamycin is not used as an antibiotic for humans in Kenya (breaking step 7). (B) Green maize stems are a complementary source of nutrients for small children in Kenya and elsewhere in Africa. The workshop participants considered that compositional analysis of maize stems at the stage of development where they are consumed was essential. (C) The workshop participants considered that the best solution to this potential problem would be to set up a system for mediation between agriculturalists and pastoralists regarding land use, in order to avoid conflicts over land use.