• Question: My main worry about GM is that modified genes could leak or escape into non-GM species with unpredictable and possibly damaging ecological consequences. For example (and it is just an example), a crop is engineered to be resistant to weedkiller, then the weedkiller resistance transfers to other plants. Is this at all plausible, and if so what steps are taken to reduce the risks?

    Asked by tomatosquid to Andy, Cathie, Jules, Les, Ricarda on 24 Jun 2012. This question was also asked by isabelnoble.
    • Photo: Cathie Martin

      Cathie Martin answered on 24 Jun 2012:

      The simple answer to this problem is to use another weedkiller, to remove the plants with the transferred resistance. There are examples of herbicide resistance selected in a few crops by using non-GM methods. Exactly the same concerns could be raised about using these crops with herbicide resistance. However, there has never been any concern raised about using these ‘naturally resistant’ varieties, because farmers could always eliminate any unwanted escapes by using alternative herbicides.
      Unless the herbicide is being used in the ecologically sensitive site (which seems unlikely) there would be no selection pressure for the trait to be maintained in the new species, and in the longer term it would be lost.

    • Photo: Les Firbank

      Les Firbank answered on 24 Jun 2012:

      Gene transfer from only plant species to another only happens if the plants are very closely related. There was a study a few years ago looking at gene transfer from domestic oilseed rape to its wild relatives in the UK, and found only a handful of cases even though it’s grown on many farms. BUt this risk still has to be taken seriously. I’m on the committee that advises Government on whether or not to allow a GM crop to be grown, and we consider whether such gene transfer is possible, and what the implications would be. If there is a risk that the modified gene might create a plant that might be somehow dangersou to the environment, we don’t support the application to grow it.

      Have a look at the ACRE website: all of our meetings and decisions and recorded, along with why.

      http://www.defra.gov.uk/acre/

    • Photo: Julian Little

      Julian Little answered on 26 Jun 2012:

      Hi tomatosquid, I can understand the concern, and it is one that worried a number of ecologists when commercialisation of GM crops was first being considered. Take your example of herbicide tolerance. The theory goes that the genes from the crop might cross-pollinate with a similar species and make the resultant weed what was termed a “superweed”, prompting cartoons with triffids wandering down the farm track. The reality is somewhat different, especially since most of the GM crops thusfar commercialised on any scale tend not to have many close relatives to “outcross” with. It was however looked at as part of the Farm Scale Evaluation, a series of field trials to look at this and many other aspects of growing a GM crop here in the UK – Les was part of the team involved in this programme.

      And the GM oilseed rape trials looked like a suitable candidate since oilseed rape does have wild relatives growing in the UK. However, to the best of my knowledge (and Les may be able to comment on this), only in one or two cases were weeds found that had been transformed – they were extremely sickly looking weeds that failed to shed viable seeds themselves.

      It would appear that whilst weeds like charlock will pollinate oilseed rape, the reverse, ie oilseed rape pollinating charlock, is not nearly as common and does not seem to result in a very good weed at all, let alone a “super”-weed!!

      So a concern in theory, but not one borne out yet in reality, to my knowledge.

    • Photo: Ricarda Steinbrecher

      Ricarda Steinbrecher answered on 27 Jun 2012:

      This is a valid concern, which is also held by many scientists, farmers, foresters and regulators. In the case of genetically modified trees for example it is very difficult to assess the risks and consequences should such GM trees find their way into the forests. The impact on global forest ecosystems could be serious, considering that genetically modified trees are being genetically modified for traits such as cold or salt tolerance, insecticidal properties, altered lignin content.
      There is presently no technology or reliable method in place to prevent the spread of GM trees, in particular forest trees, as trees and their genes spread both sexually (pollen, seed) and vegetatively, eg via roots, shoots, suckers, broken off branches. And all these propagules are very often capable of travelling very long distances.

      One attempt made to reduce the risks partially is to modify the GM tree to also be male sterile. Whilst this is not a reliable method (there is leakage as well as the potential for gene silencing), it will not reduce or prevent any of the other pathways of spread (gene flow), eg via seed. Making the tree female sterile or block its sexual reproduction completely may reduce the problem, but leakage still occurs and onset of gene silencing in just one or a few trees years down the line would result in gene flow via seed/pollen. In the meantime vegetative reproduction may already have resulted in the spread of the modified tree.

      Concerning annual GM crops, the spread of a trait (and other unintended characteristics) can happen in a number of ways, with risks for biological diversity and ecosystem function and services.

      Genes can move between related species – this can be a problem if the GM crops are being grown in areas where those crops originated such as maize in Mexico, oilseed rape in northern Europe, beet in Europe, soy beans in China and brinjal (egg plant) in India. Wheat genes can cross to some grass species. Currently, gene transfer discussion often focuses on herbicide tolerance genes spreading – making the recipients resistant to particular herbicides as noted above. However, any GM genes/traits can spread, and to the extent that more and different genes are being placed into crops, these genes may in future be transferred to weeds that already have other GM traits. This could lead to these plants getting better adapted to particular habitats – the arable field or field margins for example. The outcomes of this would be very difficult to predict or model.

      It should be remembered that the current rapid development of herbicide resistance in GM crops in the US due to over use of glyphosate (RoundUp) has come about through weeds evolving resistance and not gene transfer. This was meant to be a perfect way to deal with weeds ¬- spray the herbicide and the crops would survive while the weeds would die. The problem is that the weeds quite quickly acquire resistance to the herbicide. Every living thing obviously has a great interest in being able to continue to live and reproduce. The pressure to adapt or die is likely to result in adaptation at some point, meaning that the weeds become tolerant to the herbicide. There are already estimated to be around 15 million acres of land affected with glyphosate resistant weeds in the US. This is because of over-reliance on just one herbicide, glyphosate, plus the vast areas planted to glyphosate-tolerant GM crops.

      The solution proposed by industry is more GM herbicide tolerant traits and mixtures of herbicides – thus promises of easier and cheaper weed control were short lived and now US farmers are faced with a difficult decision every time they spray. For one weed, Palmer Amaranth in cotton, the wrong decision can mean that the only option is to pull up the weeds by hand to stop them seeding. The Southeast Farm Press for example reported in July 2010, that in Georgia 500,000 acres of cotton were weeded by hand and that weed control increased from $25 to $60-$100/acre.

      Thirteen weeds are now resistant to glyphosate in the US – mostly in soy, maize and cotton belts. Now farmers are having to use additional, often more toxic herbicides to try and control those weeds, including 2,4-D. Unfortunately this problem could have been – and was – predicted, but no measures were put in place at the time to address it.

Comments

  • Photo: imcmartin

    imcmartin commented on :

    Is there any data as to how closely related 2 species would have to be for gene transfer to occur between GM and non GM plants? Also is the relatedness restricted to certain genes, such as for photosynthesis or respiration?

  • Photo: fern

    fern commented on :

    Yes it is possible.
    Weeds such as Johnson grass are already showing resistance to glyphosate (Round up) in both the USA and Latin America. Although resistance may not necessarily require gene transfer to emerge! Using such GM herbicide resistance plants means that a lot more of this particular herbicide is used, which can encourage resistance through selecting for naturally resistant mutants. While the use of diverse herbicides may previously have controlled this emergence, the use of one dominant one encourages the problem! And while it may be true that farmers can always use a different herbicide to control those weeds, it is often the case that they have to turn to a more toxic and dangerous herbicide, which is not good. In latin america particularly, they tend to use chemicals that are banned in Europe because they are considered so dangerous…but then we import the food of course!

  • Photo: tomatosquid

    tomatosquid commented on :

    Dear Scientists,

    Thank you for these generous, thoughtful and thorough responses. The consensus seems to be that it is possible for GM genes/traits to “leak” into related species. It’s less clear whether this could pose a risk to the wider ecology, but it’s not hard to think of cases where human interventions (such as the introduction of domestic species) have had pretty drastic effects, so it seems wise to me to be cautious.

    I am not against GM in principle and I can easily imagine many beneficial applications, but these benefits need to be balanced against the risks, which seem extremely hard to quantify. Some proponents of the technology seem rather gung-ho about dangers to the ecosystem, but these concerns seem to be perfectly valid from the responses above. The public can’t give GM researchers a blank cheque to decide for themselves whether the benefits of a given GM application outweigh the dangers. Each application has to be subject to an ethical review which should include an independent evaluation of the potential ecological impacts. The message that has come from previous consultations is “don’t worry your pretty little head about it, Science knows best”. I am a scientist myself, and I tend to find that type of appeal to authority completely unpersuasive. I think a lot of the scepticism about GM is due to a past failure to engage fully with such concerns. It seems to me that IASGM is at least a step in the right direction.

  • Photo: getthetruth

    getthetruth commented on :

    Either the GM crop is as safe as or safer than the existing crop or it is not. If it is a hazard then it should be grown. Then it is considered safe we shouldn’t have to worry about “spread”. Most crops have new seed planted every year and there is a very low likihood that an approved GM crops will overtake all that type of crop. The seed can be checked routinely to minimize any spread. Also there is a massive amount of movement of genes between the crops already and you can’t track that. So why worry about the small effects when there is a higher level of mixing going on without GM traits