Madison, Wisconsin
July 23, 2003

In a river valley just southwest of Mexico City stands a small patch of teosinte - a wild, weedy grass thought to be the ancient ancestor of corn. As a gentle breeze blows gene-carrying pollen from a nearby crop of maize to its wild relative, the genetic integrity and even survival of this ancient plant and others could be jeopardized, according to new mathematical models.

The models, described in the July 23 online edition of the Proceedings of the Royal Society of London and developed by scientists at the University of Wisconsin-Madison and the University of Minnesota-St. Paul, show that genes from crops rapidly can take over those in related wild plants. The end result, say the researchers, could be major changes in the genetic make-up of wild plants, decreases in their population size and the permanent loss of natural traits that could improve crop health.

Although gene flow from crops to wild relatives has occurred ever since humans started farming, few studies before the 1980s examined the effects of this evolutionary process in a scientific manner. Most of the people concerned up until then were farmers, not researchers, says Ralph Haygood, a UW-Madison postdoctoral fellow and lead author of the paper.

But, as genetic engineering developed and emerged as both a biological and political issue, gene flow from crops containing transgenes - genetic information from other species that's artificially inserted - to wild plants gained more scientific attention.

"Most of the concern about crop-wild gene flow," says Haygood, "is driven by concern about transgene escape," the idea that these artificially inserted genes in a crop plant can leak into the genomes of wild relatives. According to Haygood, growers around the world have planted 145 million acres of transgenic crops.

Conserving the genetic integrity of wild plants, explains Haygood, is important for two reasons: protecting the survival of the plants themselves and maintaining their repository of advantageous traits. These traits, he adds, can be used to improve crop health: "The fact is that most genes for crop improvement have come from wild relatives of those same crops."

To begin to understand the effects of gene flow from crop to wild plant populations, Haygood and his colleagues Anthony Ives from UW-Madison and David Andow from UM-St. Paul, developed mathematical models based on fundamental principles of population genetics.

"The key to the models," says Ives, "is that they summarize fundamental properties of evolutionary change. They show what is likely to happen."

Specifically, the models examine how rates of pollen flow and how the selective effects of crop genes on wild plants alter two evolutionary processes: genetic assimilation, wherein crop genes replace genes in wild populations, and demographic swamping, wherein wild populations shrink in size because crop-wild hybrids are less fertile.

"Genetic assimilation and demographic swamping could change a wild plant in some way that might be important for its survival in some habitats or for other organisms that depend on them for their survival," says Haygood. "The potential ramifications are huge and diverse."

The research team starts with a simple model, where a wild population of large and constant size receives pollen from a crop that differs genetically by only one gene. They then add complexity, or, as Ives says, "more realism." That is, they consider a crop that is more different genetically and a wild population that is small or varies in size.

The researchers are quick to point out that the models do not distinguish between crops developed through traditional breeding and genetic engineering. "How the genes get in the crops doesn't matter," explains Haygood. "What's important is what they do once they're there."

In both the basic and expanded models, the researchers find that crop genes rapidly can take over wild populations and, sometimes, just a small increase in the rate of pollen flow can make a big difference in the spread of a crop gene. When this happens, says, Ives, "There's no going back. It's irreversible."

The findings, explains Haygood, show that few conditions are needed to enable genetic assimilation and demographic swamping. "You don't need high rates of pollen flow or strongly favored traits," he says. "Crop genes, even fairly deleterious ones, can easily become common in wild populations within 10 to 20 generations."

At the same time, the combined forces of these two processes on the wild populations can change their genetic make-up in unfavorable ways and drastically shrink their population size, leading to what evolutionary biologists call a "migrational meltdown."

Although the models look at gene flow from a crop plant to a wild relative, the researchers say that the models probably also could apply to gene flow from a commercial to a landrace crop raised each season from the previous year's seed. But they add that more investigation is needed.

The goal of the gene flow models, explain the researchers, is to provide qualitative insight that they hope will enhance the public dialogue on gene flow from crop to wild plants.

"Gene flow from crops to wild relatives is one of a host of environmental issues that humans must deal with," says Haygood. "These models are a resource that can contribute to the discussion."