Crop-weed offspring show hardy streak.

Genetically engineered crops have yet to hit grocery stores. But companies have inserted foreign genes into many plants to make them tastier and more insect-resistant, for example. In fact, a genetically altered tomato awaits Food and Drug Administration approval for large-scale field trials.

These advances worry some scientists. They fear that pollen containing new genetic material, or transgenes, will fertilize nearby wild, weedy relatives. This could result in a hardier weed. Crops will mate with their wild siblings, even if these plants grow a kilometer or more away, write Terrie Klinger and Norman C. Ellstrand of the University of California, Riverside, in the February ECOLOGY APPLICATIONS.

But for the transgene to spread, the crop-weed hybrids must reproduce. Some scientists argue that the engineered traits would hinder a plant's survival in the wild; others say these hybrids are often robust, Klinger and Ellstrand write. But little research has looked at how crop-weed hybrids reproduce.

Ellstrand and Klinger compared the fitness of wild radishes with that of hybrids of wild and crop radishes. They decided against using genetically engineered plants, in part because they feared the transgene might spread into other species. However, genetically engineered plants would act in ways similar to nonengineered crops, Ellstrand predicts.

The weed-crop hybrids produced about 15 percent more fruit and seeds than their wild siblings, they found.

"These results suggest that, in at least this system, neutral or advantageous transgenes introduced into natural populations will tend to persist," they write. The fitness of the hybrids will depend on the transgene used and on what sort of competition the plants face, they say,

This is "quite an important result," says Johanna Schmitt, an ecologist at Brown University in Providence, R.I. who works with Calgene, the Davis, Calif.-based company that developed the genetically altered Flavr Saver tomato (SN: 11/28/92 p.376). The findings indicate that no "intrinsic barrier" prevents a transgene from moving beyond a farmer's fields, she says.

"It confirms something we in industry have assumed we'd have to deal with," says Ron Meeusen, director of biotechnology for Northrup King Co. in Stanton, Minn. But in some cases, the domesticated crops and their wild relatives live too far apart to breed.

Klinger and Ellstrand note that the "hybridization of crops and weeds already has played an important role in the evolution of several weed species [such as johnsongrass] and has been implicated in the extinction" of the wild relatives of date, chili pepper, and hemp plants.

In a separate study, Schmitt and her colleagues recently found that seeds from a hybrid of a genetically altered canola plant and a wild mustard plant grew as well as the seeds from the hybrid made without the transgene. Neither her nor Ellstrand's study examined how well the crop-weed hybrids fare when they face competition from other wild plants, the authors acknowledge.