The development of genetically modified organisms has been a potentially explosive topic for those who see themselves as scientists, environmentalists, or just plain eaters-of-food. 

Although some evidence suggests that GMOs can be enormously beneficial, particularly in developing countries (1), others have suggested that most of the gains are attributable to other factors than the engineered traits (2).  Ever since I first learned about genetic engineering and the court battles between farmers and biotech giants, I have generally thought of the controversy of GMOs as more socially than scientifically controversial.

Yet sometimes the lack of scientific debate has more to do with a lack of data than anything else.  Indirect effects, while the very meat of systems thinking, can be notoriously difficult to demonstrate in practice without a full survey of an ecosystem's matter or energy flows.  For this reason, there has been little work considering the indirect ecological effects of genetically modified organisms, even though scientists have recognized the need for such research (3).

An article in this week's issue of Science has now opened up an examination of indirect effects of GMOs, although the results are more convoluted than the article's title might initially suggest.  Researchers in China have tracked mirid bug populations in Heibei province for a decade now, and overall no difference in mirid populations were found between varieties of cotton with or without Bt, a gene used widely in transgenic crops as a built-in insecticide.  This is pretty much what you might expect, since Bt does not affect mirids -- it is engineered into cotton to control cotton bollworm. 

What's important here, though, is not the genetic background of the crop but the difference in pesticide regimes.  While no difference between mirid populations was observed between
no-spray Bt cotton and no-spray conventional cotton, the populations were lower in sprayed cotton than in no-spray Bt cotton.  Since cotton bollworm can be managed well with the Bt gene alone, the Bt cotton is not treated with the insecticide sprays that were used to control bollworms before the transgenic cotton became available.  Yet the very benefit of Bt cotton -- the fact that insecticide sprays are not needed to control cotton bollworm -- seems to be its downfall when it comes to mirids.  The sprays are less selective than Bt, and they will typically take out mirids as a sort of chemical "collateral damage" when farmers try to limit bollworm on conventional cotton.  The mirids take no collateral damage from the Bt cotton. 

As the authors note, this means that fields that previously acted as mirid sinks are now acting as mirid sources.  The cotton fields where mirid populations were once limited by insecticides have now become areas where mirids can thrive, then spread to a range of other crops, necessitating greater levels of insecticide sprays.

Yes, you read that correctly.  The indirect effect of an environmentally-friendly, theoretically-spray-reducing transgenic crop has actually led to increased levels of sprays for a non-target pest that wouldn't really be a problem if we just used the nasty, broad-spectrum insecticides that take out both cotton bollworms and mirids.  As the proportion of Bt cotton planted increases, so does the mirid infestation on other crops, and so does the use of mirid sprays.

A person who considers him- or herself a biotechnologist just out to make some money may not be bothered much by this.  It's job security, right?  A new indirect effect pops up, and that becomes an opportunity for the next round of crop development.  On the other hand, for an ecologist this situation may seem rather distasteful.  We have now spent a great deal of our dwindling material resources (think of all those pipette tips and petri plates!) developing a crop that had the promise of reducing our use of chemical sprays, but that promise retreats the more we use the new supercrop.  Instead of shifting resource use from chemicals to plant biotechnology, we are now using resources on both of these. 

This paper is a good reminder that in ecology, nothing happens in a vacuum.

References:
(1) Brookes and Barfoot.  April 2010.  GM Crops:  Global Socio-Economic and Environmental Impacts 1996-2008.
(2) Gurian-Sherman.  April 2009.  Failure to Yield:  Evaluating the Performance of Genetically Engineered Crops.  Cambridge, MA:  UCS Publications.  44 pp.
(3) National Research Council.  2008.  Genetically Engineered Organisms, Wildlife, and Habitat:  A Workshop Summary.  Washington, D.C.:  The National Academies Press.  72 pp.
(4) Lu et al.  28 May 2010.  Mirid Bug Outbreaks in Multiple Crops Correlated with Wide-Scale Adoption of Bt Cotton in China.  Science 328(5982): 1151-1154.