Transforming Agriculture, Perennially

Issue No. 70
Summer 2001


The Emperor’s New Chromosomes

BY Stan Cox
Senior scientist Stan Cox and his daughter, Sheila, cut and bind the perennial wheat that he is breeding through traditional crosses and selection.

Senior scientist Stan Cox and his daughter, Sheila, cut and bind the perennial wheat that he is breeding through traditional crosses and selection.

The widespread adoption of transgenic plants now under way promises to accelerate the degradation of human health, rural life and the environment. But too many critics are neglecting to zero in on transgenic technology’s Achilles heel: its inherent inability to deliver on its promises. Longstanding theory and practice predict, and growing evidence confirms, that transgenes cannot dramatically accelerate plant breeding, let alone revolutionize agriculture, save the family farm or feed the world. In the the 55th Annual Corn and Sorghum Seed Research Conference proceedings published this year by the American Seed Trade Association, Drs. Major M. Goodman and Martin L. Carson of North Carolina State University add to growing evidence that transgenic technology isn’t all it’s cracked up to be. Taking a hard look at transgenic corn, the authors — highly respected in genetics and plant pathology, respectively — conclude that it will not speed development of new hybrids, and that its costs vastly exceed those of breeding through sexual hybridization. Not faster, much costlier Corn hybrids grown in the Midwest set new yield records almost every year. Most farmers and breeders regard the U.S. Corn Belt gene pool as closer to agricultural perfection than just about any other species bred by humanity. But Goodman has spent the past 25 years demonstrating that there remains much to be gained from the vast array of corn varieties grown across the tropics of Central and South America. He has used them to breed competitive, genetically diverse inbred lines — the parents of hybrids — adapted to the United States. His work is considered to be long-range, basic research with strictly long-term payoffs. Some would liken it to taking an Indy 500 car and substituting parts from a 1938 Ford sedan to improve its speed. Goodman and Carson cite the example of NC296, an inbred line adapted to North Carolina but developed from all-tropical parentage. Released in 1990, it has been used to produce commercial hybrids in the United States and at least two other countries. NC296 took 15 years to develop and five more years for hybrid seed production and distribution — a long process, typical of breeding that uses so-called exotic germplasm, with parents adapted to another part of the world. But compare that with the timetable for Bt corn, the collection of transgenic hybrids that have been making headlines in recent years because they carry a bacterial gene coding for an insecticidal toxin. Goodman and Carson write,

Bacillus thuringiensis (Bt) was used as an insecticide by the 1950s. The first gene encoding the Bt toxin was cloned by 1981. … Bt gene regulation was known by 1986. … Bt was [inserted] into corn in 1990. … Bt hybrids were first sold in 1997. BecauseBt was a well-known entity with a long history of use as an “organic” insecticide, little toxicity and allergenicity testing were required for its initial use as a transgene. Even so, its transgenic use took 17 years.

Of course, Bt was one of the very first transgenes commercialized. But the great advances made in biotechnology over the past two decades won’t make gene discovery, cloning and transfer faster and more efficient. No matter how quickly one can carry out laboratory procedures, a certain number of plant generations are needed to accomplish any genetic manipulation, and the life cycles of crop plants can be speeded up only so much. Goodman and Carson list the steps that must occur before a transgenic strain of corn — with a truly novel gene, not just another version of Bt — can even be tested in yield trials:

  1. Discovery of the gene.
  2. Modification, producing what is known as a “construct” that can be transferred to a new species and, one hopes, perform as expected.
  3. Efficacy testing.
  4. Transformation of model species.
  5. Construct comparison.
  6. Transformation of maize plants.
  7. Backcrossing the gene into best inbred lines.

These steps occupy nine seasons, more or less. Then, the authors point out, at least as much time is needed to bring the gene to the farmer. That process includes applying for experimental permits, three years of small-plot trials in different hybrid combinations, Environmental Protection Agency clearance, two years of large-plot trials, inbred and hybrid seed production, and sales. Even with the use of winter nurseries in the tropics to achieve two generations per year, and even if no unforeseen delays occur, Goodman and Carson estimate 10 to 15 years for development and deployment of a hybrid with a new transgene. This is similar to the timetable for developing a hybrid with new germplasm through traditional sexual methods. Many of the steps required to produce the two types of hybrids are the same. But there is a big difference between the two methodologies: the transgenic hybrid costs at least 25 times as much to develop and release to farmers — 28 times when the current $150,000 in federal permit and clearance fees are figured in. A table shows itemized costs estimated by Goodman and Carson. Their million-dollar estimate for discovering a new gene is based on the assumption that discovery is “a one-in-10-year event by a $100,000-a-year postdoc or equivalent (including salary and lab costs.)” In other words, we are assuming that for every ten postdocs or scientists searching for new genes to clone, one gene per year will be discovered and eventually utilized successfully. The authors don’t estimate the number of postdocs and scientists worldwide engaged in such activity, but it is huge, with only a handful of useful genes discovered. So, to date, the cost of a transgenic hybrid has been much, much more than 28 times the cost of other hybrids. One gene vs. manyGenetic engineering doesn’t speed up the breeding process, and it costs a lot more, but it produces plants with new traits that we can’t get any other way. Ifthe new trait is one that improves the lot of the farmer, and if it gives us more or better food on our table, and if it protects or restores the rural environment, then something might be accomplished. But the only genes that have been deployed to date are ones that are expected to provide a return on investment for the companies holding the patents. They have not increased farmers’ yields or profits, enhanced food quality or improved the environment. Indeed, transgenic technology — that is, single-gene technology — is not equipped to solve complex problems. For decades, basic textbooks on plant breeding have included a section on something called backcross breeding, a traditional technique for moving a gene from Parent No. 1 into Parent No. 2 while keeping most of the other thousands of genes of Parent No. 1 intact. Sound familiar? Transgenic technology is just a high-tech form of backcross breeding, the only difference being that it can import genes from more distant branches of the evolutionary tree. Textbooks also tell us that backcrossing is a useful adjunct to a breeding program, but that it is limited to producing updated versions of yesterday’s crop varieties — nothing truly new. The forces that do produce new crop varieties:

  1. Genetic diversity.
  2. Recombination, the shuffling of the entire genetic deck that occurs in the production of every egg or sperm.
  3. Selection.

These are the forces behind evolution in natural populations as well. Sexual recombination in diverse crosses almost always produces some offspring with unexpected expression of traits and unprecedented trait combinations. Breeders must sort through large populations to find the “keepers,” but the effort is rewarded when unique trait combinations are identified and new varieties developed. Almost all new crop varieties, traditional or modern, have arisen from cycles of hybridization and selection in diverse gene pools, with widespread exchange of seeds, cuttings, tubers, etc. among breeders. Without diversity, recombination and selection, breeding grinds to a halt. The sacrifice to engineering Genetic engineering is not simply being superimposed on healthy, well-funded breeding programs, it is undermining them. To understand how, consider the economic tradeoff, based on Goodman and Carson’s estimates. To produce as many transgenic hybrids as non-transgenic but exotic ones, a breeding program would need a 28-fold increase in funding. (And, even then, the resulting hybrids would embody far less genetic diversity.) That kind of increased investment is rare. More often, 28 non-transgenic hybrids or varieties will be sacrificed to produce one transgenic product. Here, we should quote Goodman and Carson at length:

Once the euphoria over the promise of transgenics fades, the closing of so many quality breeding programs, the loss of valuable sales staff, and the centralization of decision-making at company headquarters are almost certain to be regarded as tragic, even by stockholders interested in short-term profits. There are few good investments that are more long-term than rational plant breeding. Repeated studies have shown that very high returns on investment are available from expenditures on [non-transgenic] breeding … but the returns are not the instantaneous sort favored by the five-year funding plans currently in vogue. The usefulness of a breeding program is probably more dependent on continuity than ingenuity. The probability of great success by any one breeder is small, but the odds of success of a group of reasonably competent breeders working independently and continuously [and, we might add, sharing seed] is high. At present, the evidence that these same rules apply to biotechnology is almost nonexistent.

The seas of corporate and venture capital on which plant biotechnology has floated for two decades will indeed begin to dry up sooner or later. As Goodman and Carson point out, genetic engineering has followed the classic trajectory of all the bandwagons that have come and gone in the history of plant and animal breeding, such as mutagenesis, polyploidy, haploidy, somaclonal variation and ideotypes. It has lasted a bit longer than most fads, maybe because it has a more pronounceable, non-Latin name, but probably because of its patent potential and the flood of investment that it has brought. But before this bandwagon rumbles off into the sunset, it will have dealt serious blows to science, to the environment and to our food supply.

How Costs Stack Up

The expense of developing an exotic inbred corn line vs. a transgenic inbred line, not including federal fees, as estimated by Major M. Goodmand and Martin L. Carson of North Carolina State University

Choice of source/discovery of gene$14,000$1,000,000
Efficacy testing50,000
Transformation of model species50,000
Construct comparisons50,000
Maize transformation50,000

What it Would Take to get Transgenes in Natural Systems Agriculture

Drs. Goodman and Carson’s paper was prepared for and presented to a meeting of the American Seed Trade Association, private-sector breeders of corn, sorghum, soybeans and other crops for industrial agriculture. But all of the foregoing analysis applies equally to breeders at The Land and other institutions who are working to develop perennial crops for Natural Systems Agriculture and other systems. Proven techniques, including interspecific hybridization, embryo rescue, chromosome identification, recurrent selection and, of course, extensive field trials, are our preferred methods. Transgenic technology simply isn’t necessary. We could speculate on the potential for transforming annual into perennial plants by gene insertion, but with the meager state of knowledge on the subject, we can go no further. No research to date suggests that perenniality is governed by a single gene, or even two or three genes, in any crop or crop relative. In rye, triticale, sorghum, maize, soybean and sunflowers, it is often observed that unless 50 percent or more — i.e., tens of thousands — of the plant’s genes are inherited from a perennial parent, that plant is not perennial. This, along with breeders’ failure to backcross a single gene or chromosome conditioning perenniality into any annual crop genotype, attests to the complexity of the trait. It is not impossible that a gene might be isolated that conditions the perennial growth habit when transferred to an annual plant. But if a perenniality gene is identified in a particular species and cloned, its effect when transferred to any but very closely related species is entirely unpredictable. Transgenic technology conceivably could be used, once perennial grain crops have been developed, to improve their pest resistance, food quality or other more simply inherited traits. But other breeding and ecological methods will always be available, and preferable. Before we at The Land Institute would consider utilizing transgenic technology, all of the following conditions would have to be met:

  1. Any gene to be transferred would be in the same botanical family as the target species and govern expression of a necessary trait that we could not introduce via any other practical method.
  2. Resulting varieties would not be patented or burdened by any other intellectual property agreements, so they would be open to public use.
  3. In our judgement and that of our Scientific Advisory Team members, the gene and its carrier DNA, and the methodology used to insert it, would be thoroughly tested and represent no threat to gene pools, the environment or human health.

It is clear that none of these conditions come close to fulfillment today for any gene, and we do not expect all of them to be fulfilled for many years, if ever.

Share On: Share on Facebook Share on Twitter Select other ways to share