Can biotechnology ‘feed the world’? Not likely, says genetic engineering expert Doug Gurian-Sherman

istock_droughtsoyWith food shortages provoking riots in recent years, and the world’s population increasing exponentially, Congress will soon be debating the next big U.S. aid package for developing countries. America currently spends about $2.5 billion on food aid, most of it to buy surplus U.S.-grown commodity grains to donate — a policy that may fill bellies but is wildly inefficient and by undercutting local farmers, has hampered aid-receiving communities in feeding themselves.

And now there's a new development aid policy in the pipeline that may prove equally short-sighted, but also far riskier for small farmers around the world.

Officials at USAID, the agency that carries out U.S. development work abroad, have been positioning hunger as a problem of agricultural yield — poor countries just aren’t producing enough food — and promoting genetically engineered seeds as an important part of the solution. It’s an argument uncannily similar to the massive ad campaign recently mounted by Monsanto, the biotech company that controls some 90% of the global market for genetically engineered (GE) seeds. Secretary of State Hillary Clinton is expected to release a blueprint for development aid that features such seeds prominently as early as this week. It could become legislation soon after.

[July 9, 2009 update: In anticipation of this announcement, CREDO Action — the activist arm of Working Assets — is circulating a petition titled "Tell your senators Monsanto can't feed the world.]

We asked Doug Gurian-Sherman, a plant pathologist and molecular biologist currently working for the Union of Concerned Scientists and an expert in genetic engineering (see bio), to discuss the science of GE crops and whether they are appropriate tools with which to combat world hunger. Below is the first in a two-part Q&A. Part two is now available here.

The Ethicurean: Tell us a little about your background and work on GE issues.

Doug Gurian-Sherman: From as early as I can remember, I’ve been interested in biology. I got interested in molecular biology in the ‘70s; this was the field that would become genetic engineering. New advances in molecular biology, cloning and other things, were just being developed then. I found it fascinating. So I went to grad school. I wanted to use molecular biology and genetic engineering to study ecological problems, to get a better understanding of the ecology of microorganisms that grow on plants. The lab I worked at did some of the first field tests allowed by the Environmental Protection Agency on a GE organism in the environment. I was doing classical GE work: if I changed a gene, how did it change the properties of the organism?

I did my postdoc with the USDA in one of the first labs to successfully develop GE wheat and rice. Only one other lab, I believe, had developed GE wheat at that time. From there, I ended up in the U.S. Patent Office, in the biotechnology group, examining biotech patents. Then I worked at the EPA evaluating the safety of GE organisms, among other things. So I’ve both regulated these things and developed them.

When we talk about promoting GE technology in developing countries, what crops and traits do we mean exactly?

So far, it’s the same crops and traits that are in use in the U.S.--corn , soybeans, and cotton. Two main traits have been developed. The first is insect resistance: to date they’ve used a few different genes from Bt bacteria — which are selectively toxic to some insect pests — to protect against a handful of lepidoptera, moth-like pests like the European corn borer. There are a few other GE traits used in corn in the U.S. aimed at several beetle species called rootworms, but those haven’t been used in developing countries.

The other category of GE crops are those with herbicide tolerance — like tolerance to Roundup, Monsanto’s herbicide — that allow farmers to spray herbicide onto crops without killing them. It’s convenient for the farmers because they can control weeds with minimal labor.

So how come when Monsanto and members of Congress talk about the potential of GE crops in developing countries, they talk about yield-increasing potential? Do they mean the yield increases you’d get because your plants aren’t being attacked by pests or overwhelmed by weeds?

You’re right: there’s confusion when companies talk about increasing yield.

There are two ways you can influence the yield of a plant. The first is by boosting the plant’s intrinsic yield, the yield potential provided by the genetics of the plant itself. You boost that yield by optimizing the genetic ability, and this is best observed when of the crop is growingunder favorable conditions. The second way is by boosting operational yield. These are the increases you get when you prevent losses due to pests or stresses. Losses from pests or stresses like drought can be thought of as reductions from the intrinsic yield. Bt and herbicide tolerance are this second kind — operational yield gains. In the U.S., the main type of Bt crops prevent damage from the pests corn borer and rootworm on corn, or bollworm on cotton. So whatever losses you’d get from corn borer, to the extent they’re prevented because you’re using Bt seeds, that’s operational yield gain.

There have been no GE crops yet commercialized that have provided intrinsic yield gain or yield potential. That should be contrasted with conventional breeding and enhanced ways of breeding using genomic information, both of which have increased intrinsic yield considerably. Genetic engineers have tried to do it, but they haven’t succeeded. These current crops weren’t designed to increase intrinsic yield. They’re talking about operational yield.

OK, so how effective would operational yield gains be in developing countries?

In developing countries, pests often tend to cause higher losses than they do in the US. This is often because the farmers are trying to grow the same export crop year after year or the same subsistence crop year after year. Those aren’t good agricultural practices, but farmers have been pushed into undesirable situations due to poverty.

The impact of existing GE crops on yield is really variable in developing countries. Bt cotton has been studied in developing-country settings but there’s been less study of Bt corn. Yield increases can often vary from about 10 to 40%, and sometimes more. That sounds like a lot, but usually these farmers are starting at very low yield levels, so these kinds of yield increases are usually not going to be enough to lift them out of poverty. For herbicide-tolerant crops, grown mostly in South America, studies don’t show any benefit to yield compared to other available technologies. That’s because even for conventional farmers, the herbicides they already had worked just as well as Roundup, although several different herbicides had to be used.

And how have these crops worked here in the U.S.?

In our recent report [“Failure to Yield,” PDF here and Ethicurean summary here], we looked at the public record of field trials on GE soy and corn, the two major food/feed crops in the U.S. We found that in the U.S., as I said, GE technology hasn’t increased intrinsic yields at all. Herbicide tolerance hasn’t increased operational yields either, and Bt traits have only increased operational yields a little bit, much less than other technology. Although some individual farmers can see yield increases of 10 or 12%, we found that Bt corn was responsible for only a 3 to 4% yield gain nationally since it became available 13 years ago. But over that same period of time, corn yields overall have gone up 28%. So only 14% of the total yield increase we’ve seen in corn in the U.S. is due to Bt traits. And again, none of the crops we looked at have increased intrinsic yield.

The companies have tried; we found that there have been many field trials — over 3,000 — for various types of yield improvement, including for intrinsic yield.

If these companies have been working for so long to boost intrinsic yield with GE technology, how come it’s proving so difficult?

The traits to increase intrinsic yield are much more complex than for adding Bt. For Bt, you have a gene that codes for a single protein — the Cry protein — and it directly attacks the insect. It’s genetically very simple. Herbicide tolerance is similar: you either have a gene that produces an enzyme that inactivates the herbicide when it goes into the plant, or, as in Roundup Ready seeds, you find an enzyme that’s immune to the herbicide and replace a target enzyme in the plants with that. They’re genetically simple systems.

Yet despite that, these genes have caused unintended changes in the genetics of the plants. Bt corn, for example, has increased lignin — a component of plants’ cell walls — and no one knows why. You’re not putting a gene into an empty box; it interacts with the genes that are part of the genome of the organism.

Traits for intrinsic yield and drought tolerance are much more complex and are controlled by multiple genes. Different plants use different genes under different conditions; drought tolerance comes from one combination of genes in corn, a different combination in cactii. If there are many things going on in the environment — drought stress, salinity stress, insects — the plant may have a whole other set of genes involved to help it get through this. These genes may not be the same as the ones that help it weather drought alone.

On top of that, because the genetics of this kind of GE crop are more complex, you also have more unintended effects through interactions with other genes in the plant. You can get “downstream effects” that are pretty far removed from the trait you’re intending to get. Some of those unintended effects could be harmful ecologically or in terms of human health — but they’re also harmful agriculturally.

Most of the plants we’re working with have been bred for decades to improve various qualities, including yield. With genetic engineering, you’re adding additional changes to the plant, but you could also screw something up that’s been optimized already. I suspects that what’s happening in the field trials with a lot of these crops is that they try something and then get unintended effects. They just don’t work as well as they hoped or an unintended effect interferes with some other important property of the crop, so they can’t market them.

Can you give an example of these unintended effects?

In the report we cover an interesting case. One problem with some drought-tolerant crop varieties is that under normal moisture conditions, the variety doesn’t yield as well as varieties without drought tolerance. The New York Times recently covered a potential breakthrough with a particular gene that reportedly conferred drought tolerance but didn’t show that downside. But then a few months later, another lab working on the gene for different reasons found that it made plants more susceptible to various plant diseases. So the same gene that confers drought tolerance makes plants more susceptible to disease. Farmers may have to use pesticides to control these diseases if this drought tolerance gene is approved. How will this balance out in terms of benefit and risk?

Such unintended effects are not publicized because companies don’t like to talk about failures. The bottom line is that there has been a huge amount of effort to produce a lot of crops over the years with success of only a few traits: Bt and herbicide tolerance. They have not resulted in significant yield gains at all in the U.S. And we also have to put any yield gains in the context of the expense and other factors and compare GE technology to other technologies and production methods.

Some of these genes may pan out in the future, but overall, how will GE stack up compared to other methods and technologies that have been more successful in the past, or that have the multiple benefits? GE gets undue attention in the form of media hype, research investment, and political encouragement when in fact other methods show much more promise.

In Part Two, Gurian-Sherman shares the results of studies on alternative production methods in developing countries and talks about what the GE approach means for poor farmers’ livelihoods.

4 Responsesto “Can biotechnology ‘feed the world’? Not likely, says genetic engineering expert Doug Gurian-Sherman”

  1. Kat says:

    Wow.  For all of the confusion around this issue, you spelled it out very simply.  I just went on an agricultural exchange program to France with the Dean of our Agricultural College and a group of ag students.   Bringing up these issues was like experiencing the inquisition.  We have to get past all of the ideological and self-interested POV's and look at the science.  Agriculture is our bedrock, our foundation for life now and our future.  Thank you for this great series.  I look forward to the next one.

  2. Some questions for Dr. Gurian-Sherman:  GMO corn, soy and cotton have been around for quite a few years, and with patents having terms of 20 years from the date of application (in most cases), have any important GMO seed patents already expired?  Are any expiring soon?  If so, do you foresee that seeds produced using the innovations contained in the expired patents will have enough value to be made and sold as "generics"?  It seems to me that having a suite of GMO products that can be sold without royalties and freely saved by farmers changes the equation quite a bit, as monopoly and patent enforcement issues fade away.  Not having to worry about intellectual property rights might also lead to more independent health effects research.

  3. Good point, Marc. When the first GE trait patents expire (I think the time is 23 years), they will be royalty-free and easily available for people in developing countries.

    There is another point to be made here, completely overlooked in the commentary about GE technology and poor farmers in developing countries - and that is the assumption that all crops with GE traits will require those farmers to pay a premium for those seeds. But when it comes to GE projects such as Golden Rice, Super Cassava, and Super Sorghum, the goal is to release them to those people to grow and regrow without owing anyone any money.

    There is also the prospect of an 'open source' genetic engineering community that would also sidestep that issue.

  4. In response to Marc and Inoculated Mind (AKA Karl), I agree that the patent issue is an important one. And it certainly helps to have genes for GE crops in the public domain. Of course, for hybrid seed, as some of the maize and cotton (in India) may be, farmers still must buy seed every year regardless of patent status. I don't know whether important patents are expiring soon, but I suspect that patents on the first commercialized genes, like several of the Bt genes, would be getting close to expiration. U.S. patents at that time (e.g. 80s or early 90s) were for 17 years from the time that the patent was approved. The crops have been commercialized for 13 years, and I suspect that the patents were obtained several years before that. In addition, I have heard that some of the genes are being donated to some of the poorest countries. Open sourcing is also a good idea.

    But the first genes that may be coming close to public availability, while they may provide some benefits, are very limited--only a few Bt and herbicide tolerance genes--and the latter require additional inputs (herbicide) to be effective. And in addition to the genes, the promoters that allow the genes to function (express), and the methods to make the GE crops are also patented. But some of the earlier patents for those may also expire in the next few years.

    Even for these early genes, there are limitations in their use. For example, most Bt crops in the U.S. are now being "stacked" with two different Bt genes which should slow the rate that resistance develops in the target insects, which would make the Bt genes ineffective. But those second or third Bt genes were commercilaized about 5 or more years after the first ones, so their patents may also have been later, and won't expire for some time. In the mean time, the first resistance by stem borer to Bt in maize in South Africa has already been reported (in 2007--whether it will become a practical problem is not yet clear). The point here is that biology is not static (especially where pests are concerned), and so the cheap availability of older genes may not always be as useful as they first appear.

    Other genes that may be effective in the future have not even been commercialized yet, so the expiration of their patents is still a long way off.

    But when (in my second posting) I say GE technology is expensive, I was not talking primarily about the cost to farmers (or researchers) from patents. The infrastructure and trained personnel needed to produce GE crops is very expensive, and addresses (typically) only one trait at a time. So, when it comes to how much value one can get for a given amount of investment, I don't think that GE is often the best approach. Agro-ecological approaches, as I discuss in my second post, are often better for several reasons, but are being neglected.

    That does not mean that improved crop varieties are not important. They are and will be critically important for increasing yield potential and other crop properties.  But these varieties need to be grown under ecologically sound systems--not the current industrial model of farming as in the U.S. And I think the evidence so far suggests that conventional breeding, and especially conventional breeding enhanced by increased understanding of crop genomes, is and will continue to be more effective than GE for virtually all types of crop improvement, from pest resistance to yield potential to drought tolerance.

    In some cases, GE may be the best approach, and if it is available to poor farmers for free, and its safety has been assured, then those applications of GE may be worth pursuing if the farmers and public want it. But in my opinion, adequate regulatory systems are not in place, and need to be functioning before lots of new genes are widely deployed.