This means genetic traits that are developed in crops - such as resistance to insect pests - can become a permanent part of the weed population, in turn posing possible risks to crops.
These results suggest that biotech companies should steer clear of developing transgenic radish varieties with beneficial traits that could be passed on to weeds, said Allison Snow, a study co-author and a professor of ecology at Ohio State University.
Transgenic crops are crops engineered with specialized traits such as resistance to viral diseases, insect pests and herbicides.
While the new hybrid weeds may not be as fit at first as their wild parents, they seem to regain reproductive fitness quickly, Snow said.
"It's inevitable that these and other fitness-related traits will make their way into weed populations," Snow said. "The result may be very hardy, hard-to-kill weeds."
In California, the crop itself has become a highly successful and very damaging weed. Scientists suspect that this transition was aided by genes from the wild radish. In fact, wild radish is considered one of the 100 most economically damaging weeds worldwide.
"Gene movement from crops to their wild relatives is an ongoing process that can spur rapid evolutionary adaptation in weeds that will be ultimately harmful to crops," Snow said.
Snow presented the findings August 9 in Madison, Wisc. at the annual Ecological Society of America meeting.
The researchers studied four populations of hybrid and wild radish for six years in Michigan. At the outset, each field consisted of 100 first-generation crop-wild radish hybrids and 100 wild radishes. To monitor the continuation of crop radish genes in the populations, the researchers looked for four genetic traits: two enzymes, flower color and pollen fertility.
On average, the wild radishes reached peak flowering one month before the hybrid plants. The hybrids also produced fewer seeds per fruit than wild plants and fewer viable pollen grains. A large portion of hybrids never produced fruits (60 to 78 percent), while 92 to 97 percent of the wild plants did.
Even so, traits from the original crop - such as white flower color - persisted in subsequent generations of hybrid radishes.
"Even though the effects of delayed flowering and reduced fertility inhibited the movement of certain crop traits to later generations, we did find evidence of crop genes in every generation," Snow said.
In order to compare the lifetime fertility of wild and hybrid varieties, the researchers also grew one population of potted radishes.
Viable pollen averaged 63 percent in the potted hybrids, as compared to 92 percent in wild plants. The hybrids produced only half as many seeds per fruit as wild plants. Although the potted hybrid plants flowered on average half a month later than the potted wild ones, the hybrid radishes still produced enough viable pollen with enough time left in the growing season for pollination to occur.
"The hybrids were capable of ecologically significant levels of reproduction," Snow said. "The second hybrid generation was still at a fitness disadvantage, but to a lesser extent than the first hybrid generation. This indicated that their reproductive fitness was increasing."
The next step is to figure out if crop genes benefit weed populations and also to gauge how easily crops can become wild.
"The constant gene flow between crops and weeds is a subtle process that no one may notice, but evolution can happen very quickly," Snow said.
Snow conducted her research with Kristen Uthus, a graduate student in Ohio State's Evolution, Ecology, and Organismal Biology graduate program and Theresa Culley, a researcher at the University of California, Irvine.
Snow's research was supported by a grant from the National Science Foundation, Ohio State University and the University of Michigan Biological Station.