This seems to be a commonly raised issue (variants include '
jellyfish and tomatoes' and 'monkeys and polio'). It is also less of a pure ethical issue - although possibly a
theological one. My gut feeling is that nothing
surprising should happen, but the consequences of being wrong are too high to dismiss any possiblity.
What happens then if you 'cross the species barrier'? Well, assuming you are transplanting only one gene at a time (to keep the argument simpler) here goes:
- Divide the fish genes into two types: homologs and the rest. A small proportion of the genes of organisms are practically 'universal' (histones, cytochromes etc) - these are the close homologs. They can't necessarily be dismissed, but they are less likely to alter anything.
- With the rest, which ones are the most volatile? That is, which are inert structural proteins and which active enzymes. Pesticides can be dangerous because chemicals often resemble each other, and foul up the works. Equally, therefore enzymes that could make dangerous chemicals would be little 'pesticide factories'.
So, not
all genes are forbidden in fruit; and some genes might well be dangerous. To be
globally destructive, the plant has to put out new, destructive chemicals or kill foraging
herbivores. Of course, plants are engaged in chemical warfare all the time and positively enjoy killing each other - but that's no consolation if they start to kill
cattle or us!
If it is inevitable that genetically modified foods are produced, then two things are needed. Firstly, labelling of products as GM - if people really don't want it, they won't buy it. Secondly, more careful study (before planting large fields) of the ecological net a crop is part of and how the particular gene of interest would affect it. Enclosed, small scale trials mimicing the natural state - with insects and animals - shouldn't be impossible. Just expensive!
It is a difficult question, of how to test the unexpected. Long term change requires long term trials and big business probably doesn't want to do so. Tough?