On This Page:

What is a weed?
Why are most crop plants not weedy?
How could genetic engineering make plants more weedy?
In a nutshell

The American Heritage dictionary defines a weed as:

"A plant considered undesirable, unattractive, or troublesome, especially one growing where it is not wanted, as in a garden."

By that definition, technically any plant could be considered to be a weed, as long as it is growing where people don't want it to grow. But some plants are more likely to become weeds than others. Why do plants like kudzu (southern US) and purple loosestrife (northeast US) spread rapidly through the countryside, choking out other plants and covering every open surface in sight, while others, like corn plants, never manage to escape from their unenclosed fields? There are certain characteristics that make some plants more troublesome than others, and much more likely to grow where they are not wanted. These traits that tend to make a plant more "weedy" include (among others):

• long-lived seeds that don't all germinate at the same time
• rapid seedling growth
• high tolerance to changes in environment, and ability to grow in different environments
• competes aggressively with other plants
• produces new seeds continuously
• produces a large number of seeds
• can disperse its seeds long distances

Although weeds don't always have all of these characteristics, most weeds do have some combination of them. Weeds also tend to be plants that are happy growing without help in areas disturbed by man (as opposed to natural habitats), such as gardens, fields, along highways, and in vacant lots.
[top of page]

It is true that most crop plants do not act like weeds. Like kudzu and purple loosestrife, most crop plants have been brought to the US from other parts of the world. Crop plants, however, aside from the occasional "volunteer" plant sprouting in a field from last year's crop, are very rarely found growing outside of a farm field, despite that no one is actively stopping them from doing so. Despite that it is planted every year on millions of acres, never grows wild along the sides of the highway. Why?

After thousands of years, crop plants have become extremely dependent on humans to disperse their seeds, and can no longer survive in nature without our help. Many of the traits that humans prize are actually detrimental to a plant's ability to "act like a plant in the wild." We prefer large, starchy seeds with thinner coats (easier to digest); seeds that stay attached to the plant (easier to harvest); seeds that don't go dormant and that sprout almost as soon as they hit the ground (easier to plant); larger fruit with fewer (or no) seeds (easier to eat); etc. Many plants eaten as "vegetables" make an exaggerated effort to produce leafy parts, and only feebly make flowers and seeds. In addition, humans probably chose in the first place the crop plants we use today in part because they are so well behaved in human society and only grow where they are supposed to. There are some exceptions, however: sugarbeets, for example, brought to the US from Europe, have in some areas of California escaped from farmer's fields and now grow in small "wild" populations.
[top of page]

It is conceivable that if, through genetic engineering, a crop plant was inadvertently given one of the "weedy" characters described above, the plant might be more successful at growing outside the confines of an agricultural field. Such changes might include (but are not limited to): increased seed viability, number, dispersal, or reduced seed dormancy; altered plant growth habit, such as an ability to survive winter, produce more generations per year, or produce seeds over a longer time period; traits that make the plant more aggressively competitive with other weeds; etc.

There is an important distinction between increased weediness and increased fitness, however. Fitness is the ability of a plant to respond better to its environmental stresses and to be more successful at making viable seeds. Many of the traits presently genetically engineered into plants do increase the fitness of the plant-- such as resistance to insects, viral disease, and herbicides-- but do not affect the weediness of the plant. Bt corn, for example, is much more resistant to certain insects than non-Bt corn, but this improvement does not help corn overcome all of the other seed dispersal and growth habit traits that prevent corn plants from spreading out into the wild.

How then can we know ahead of time if a new GE trait will affect the weediness of a plant, even if we know it is likely to improve its fitness? It's an educated guess, really. The USDA's Animal and Plant Health Inspection Service (APHIS) treats all new GE plants as potential weeds, the same way it treats all new plant varieties imported into the US from other countries. Furthermore, APHIS requires developers to observe all experimental GE plants (usually for a few years) in small, controlled test plots and to submit data addressing the potential change in weediness before they are approved for widespread commercial release. APHIS also requires immediate notification if there is any evidence that a GE plant has behaved like a weed (that is, if it ever grows outside the field where it is planted).

None of the traits currently engineered into plants appear to alter the plants' ability to overcome their built-in escape barriers. A recent ten-year study of most commercially released GE varieties to date determined that the GE varieties are no more likely to grow outside an agricultural field than their non-GE counterparts. However, if in the future the traits engineered into plants become more complex, it may be difficult to predict their weediness from the trait alone. More complex GE plants may require careful and cautious pre-market field-testing, and additional vigilance once the plants are released commercially.
[top of page]

Most crop plants have significant limitations in their plant growth and seed dispersal habits which prevent them from surviving long without constant care from humans, much less thriving in the wild as weeds. Genetic engineering could conceivably improve a plant's ability to "escape" into the wild. A recent ten-year study suggested that current GE plants are no more likely to grow in the wild than are non-GE plants, but traits engineered into future GE plants could possibly make a crop plant more likely to grow where it is not wanted.
[top of page]

Baker, H. G. 1965. Characteristics and modes of origin of weeds. In: The genetics of colonizing species, pp. 147-168. Baker, H. G., and Stebbins, G. L. (eds.), Academic Press, New York.

Crawley, M. J., Brown, S. L., Hails, R. S., Kohn, D. D. & Rees, M. 2001. Transgenic crops in natural habitats. Nature 409:682-683.

de Wet, J. M. J., and Harlan, J. R. 1975. Weeds and domesticates: Evolution in the man-made habitat. Economic Botany 29:99-107.

Keeler, K. 1989. Can genetically engineered crops become weeds? Biotechnology 7:1134-1139.

Tiedje, J. M., Colwell, R. K., Grossman, Y. L., Hodson, R. E., Lenski, R. E., Mack, R. N., and Regal, P. J. 1989. The planned introduction of genetically engineered organisms: Ecological considerations and recommendations. Ecology 70:298-314.