The regulatory bottleneck for biotech specialty crops

Speciality crops. Image from Colorado State University.

We often hear that there are only two genetically engineered traits on the market – Roundup Ready and Bt. And, for the most part, that’s correct. There are a few other commercialized traits, such as virus resistant papaya and squash, but why aren’t there more? We see all sorts of papers about awesome genetically engineered traits, from nemotode resistance to nutritional enhancement to really specialized traits like nicotine free tobacco and allergen free peanuts. There are so many traits that we even started a list at Biofortified in an attempt to keep track of them all.
These traits have been developed and tested for efficacy, often with public funding, but haven’t make it to the commercialization stage. All that’s needed is a little breeding to get these traits ready for market. Why don’t we see them in the grocery store? UC Davis researchers Kent Bradford and Jamie Miller have collected a huge amount of data on genetically engineered traits and presented it in their recent paper The regulatory bottleneck for biotech specialty crops. It’s a short but insightful piece. Don’t forget to check out the supplementary material that has lists of all the different traits they found described in the literature.
Here’s the first two paragraphs to whet your appetite:

Specialty crops, which include fruits, vegetables, nuts, turf and ornamental crops, are important components of human diets and provide environmental amenities. In 2007, such crops represented ~40% of the $140 billion in total agricultural receipts, despite being cultivated on just 4% of the total cropped area. Although tomato was the first genetically modified (GM) food crop to be commercialized in 1994, the only GM specialty crop traits currently marketed are virus-resistant papaya and squash, insect-resistant sweet corn and violet carnations. All of these received initial regulatory approval over 10 years ago. As a group, GM specialty crops have garnered limited market share (the exception is GM papaya resistant to papaya ringspot virus, which now produces 90% of Hawaii’s crop). In contrast, GM field crops, such as soybean, maize, cotton and canola, have come to dominate the markets in countries where they have been released. What is responsible for this disparity in the commercialization of GM field crops versus specialty crops?
One possibility is that the dearth of GM specialty crops indicates a lack of current research or of beneficial traits for crop improvement through genetic engineering. Alternatively, research may have continued but progression through the regulatory process to the marketplace may have failed. Anticipated lack of market acceptance could have stopped either research or regulatory submissions. To find out why specialty crops with GM traits have fared so poorly, we have analyzed the research, regulatory and market pipeline to determine which steps in the process may be responsible for the limited range of commercially available products.

The researchers conclude that the primary barrier to genetically engineered specialty crops is steep regulatory costs. These costs can be met by companies producing traits for commodity seeds because there’s a lot of commodity seeds to be sold. This isn’t the case for speciality crops. Additionally, they argue that it doesn’t make sense to require regulation for traits that have been produced with genetic engineering that could be produced with other methods because there isn’t any evidence of harm coming from the genetic engineering process itself.
ResearchBlogging.orgMiller JK, & Bradford KJ (2010). The regulatory bottleneck for biotech specialty crops. Nature biotechnology, 28 (10), 1012-4 PMID: 20944582

Science, ,
Anastasia Bodnar

Written by Anastasia Bodnar

Anastasia Bodnar serves as the Policy Director of Biology Fortified, Inc. She is a science communicator and multidisciplinary risk analyst with a career in federal service. She has a PhD in plant genetics and sustainable agriculture from Iowa State University.

12 comments

  1. Roundup Ready and Bt. And, for the most part, that’s correct.

    I call shennanigans.
    Bt isn’t a single trait, it is multiple traits all of which offer different insect resistances, as well say there’s one insecticide on the market.
    Also given that Smartstax (which is on a not inconsiderable acreage this year – record breaking for 1st year trait coverage in corn if I am not mistaken (given my track record this could be a 50:50 shot) has RR and liberty-link the whole false “two traits” is only true if you consider things at the high level of what the traits do (IR and HT) and set your cutoff at massive acreage.
    I think :-
    We often hear that there are only two genetically engineered traits on the market – Roundup Ready and Bt. And, for the most part, that’s correct. That’s a bald faced lie concoted by those too lazy to keep up with GMOs but all too willing to paint the field as stagnant in an effort to bolster their own preconceived ideas.
    /snark

  2. Agreed, there are many different Bt genes that produce slightly different proteins, but can we really call them different traits?
    For some reason I though Liberty Link wasn’t genetically engineered.
    Anyway, I suppose I could correct the post to say:

    We often hear that there are only two types of genetically engineered traits on the market – herbicide resistance and Bt. And, for the most part, that’s correct.

  3. I’d say yes different Bt genes are different traits – the Cry gene family is pretty large and specificity varies for different insect types (specificity of Cry proteins extends beyond insects once you get to different Cry proteins (remembering the worm killing version highlighted here some time ago)) – you’d have to at least look at above and below ground insect protection as different traits in my opinion if not individual proteins with different MOAs.
    Although your best course of action would be to ignore my whining =p everyone knows what you mean!

  4. Clearfield is the herbicide tolerance trait that is not genetically engineered. What’s different about Liberty Link (besides that it is glufosinate tolerance instead of glyphosate) is that they do not restrict you from saving and replanting the seeds. This is because the patent on the herbicide has a long way to go and they recoup their costs through herbicide sales instead of seed patents.

  5. Ewan-
    The Cry genes are a fascinating bunch, no doubt. But I’m inclined to lump ’em together for this aspect of the discussion. So I would tend toward your earlier classification of IR and HT for the current set of commercially important GMO trait categories.
    Consider too that there are two different commercially important RR traits in soybean (and corn I believe) RR1 and RR2, which for both biological and commercial reasons need to be technically separate, but ultimately to a producer mean that you can spray the crop with glyphosate and not kill it.
    Under this system then Liberty Link gets lumped with RR traits and the newer herbicide traits in the pipeline will end up lumped here also as HT traits.
    I have even seen both IR and HT traits lumped together as “input traits” – suggesting that the grower is the primary beneficiary of the trait (changing their input mix). Potential “output trait” GMOs would then be something like a modified oil in the seed which can then benefit a seed consumer. I’m not as excited about this latter level of lumping. The splitter in me imagines that folks involved in the debate should be smart enough to discern a difference between IR, HT, and other general catagories of plant traits.

  6. I guess if you need to be a lumper rather than a splitter that is potentially a way to go, but I think it runs the risk of supporting the asinine arguement that GM is stalled b/c there are only 2 commercialized traits.
    I’m thinking producers see the traits as individual. Which is why they’ll pay more for smartstax than for VT triple than for VT double – and which is why these individualized products exist at all.

    Consider too that there are two different commercially important RR traits in soybean (and corn I believe) RR1 and RR2, which for both biological and commercial reasons need to be technically separate, but ultimately to a producer mean that you can spray the crop with glyphosate and not kill it.

    Producers will look at the RR1 and RR2 traits in different ways – sure, RR1 allows you to spray with glyphosate and not kill your crop, but RR2 extends your window of application (as far as I recall) – which means there could even be an arguement to be made that the RR1 and RR2 traits should be looked at as individual rather than as a single lump (although in terms of the protein expressed there is no difference unlike for Cry proteins)
    It obviously depends to an extent on the discussion at hand – the recent “GM is stalled because we only have 2 traits and nothing new since 1996” raised my heckles and as such I’m not willing to lump any more than I have to.

    The splitter in me imagines that folks involved in the debate should be smart enough to discern a difference between IR, HT, and other general catagories of plant traits.

    Should be yeah, but when you deal with dishonesty at a level which can’t discern the difference between a graph with an upwards trend and one with a downwards trend then assumptions of smarts may need to be dialled back a bit.

  7. Ewan:
    Not so fast…

    I’m thinking producers see the traits as individual. Which is why they’ll pay more for smartstax than for VT triple than for VT double – and which is why these individualized products exist at all.

    The individual products exist for good reasons. First, not all producers have all the same needs. So unless and until traits that one doesn’t need are given away for free there will be folks with sharp pencils who will choose to purchase only what they need. And this is, to my mind at least, exactly what we would hope will happen. We shouldn’t want to deploy insect resistance technologies just to be doing it. Sound technology stewardship would suggest that a particular gene only be deployed where it is needed. Secondly, as there is value to the individual traits there is no marketing incentive to mask this value by giving it away… or supplying only one choice at the highest cost.
    As for dealing with folks who can’t discern extreme differences in a dataset… you can allow your hackles to be raised, or you can take a deep breath and choose a more rational response (recalling that no response is often a wise course). So I’ll leave with the observation that a Bt proponent who is easily miffed by arguing with dishonest folks might be seen as a ‘Cry’ baby. (sorry all).

  8. Cry baby is worthy of at least a golfclap – well played sir.

    So unless and until traits that one doesn’t need are given away for free there will be folks with sharp pencils who will choose to purchase only what they need.

    My point exactly – within the IR family of traits exist different individual traits – rootworm protection and corn borer protection are both IR traits, both employ Cry proteins, but are still different traits. I fully concur that traits should only be used where required, but as such think that traits should be defined at the right level (I’m a splitter, you lumpers can burn in hell! (or something less confrontational as the whim takes you))

  9. IR traits are also drought tolerance traits. In a roundabout way. Turns out, when a plant is stressed by drought, the effects of attack by insects are multiplied. Especially with corn rootworms.
    This has come out in field trials where rootworm pressure is considered to be quite low. As I recall, it’s on the order of an extra 3 bushels per acre.
    Of course, then the question involves very sharp pencils indeed — essentially an insurance actuary. The extra 3 bpa only shows up in drought conditions, and is otherwise not necessary when insect pressure is low.
    What do you pay for a trait you might not need?

  10. Makes sense that root based IR increases drought resistance – less root means less surface to absorb available water means more drought stress – I would assume the same applies to essentially any nutrient stress situation. I’d also guess that you may lose some water through tissue damaged by insects (nevermind losing water to the insects as they munch through the tissue) but am also guessing the effects here would be considerably lower than a root trait.

  11. I like the use of ‘family’ in this sense. So we can lump like traits into families and split out families for those with either shorter attention spans, or with too little appreciation for the detail (or for situations where there is no need for this degree of detail).

  12. I just stopped by to mention that, apart from the traits you mention, in 2010 another trait appeared on the market, at least in Europe, with the industrial potato Amflora (BASF) with an altered starch composition, in order to be used by the paper industry. The first GM potatoes have been processed just a few weeks ago in Germany, Sweden and (I guess) the Czech republic
    Very nice blog, btw, and very high quality discussions in the comments 🙂
    Dario

Comments are closed.

Scroll Up