Associated Press

In an unusually bitter and public dispute among scientists, the journal Nature says it should not have published a study concluding that DNA from genetically
engineered corn contaminated native maize in Mexico.

"Nature has concluded that the evidence available is not sufficient to justify the publication of the original paper," editor Phillip Campbell said in a terse message in the
latest issue of the distinguished journal.

However, Mr. Campbell stopped short of retracting the original paper, which was peer-reviewed and published in November. Instead, Nature printed two harsh
new criticisms of the work, as well as a defense by the researchers that included new data.

Mr. Campbell wrote that Nature would allow its readers "to judge the science for themselves." The London-based journal said that Mr. Campbell was on vacation
and that Nature would have no further comment.

The controversy further clouds the debate over the possible risks of genetically modified crops.

In their study, David Quist, a graduate student at the University of California at Berkeley, and microbial ecology professor Ignacio Chapela reported that genes from
laboratory-altered strains of corn had found their way into native corn growing in the rural Oaxaca region.

In 1998, Mexico had banned the planting of genetically modified corn to protect its indigenous maize.

The Berkeley researchers' findings had heightened fears about the possible risks of genetically engineered crops.

Nature's decision enraged the study's authors, who conceded only minor interpretive errors.

"We certainly stand by our original main statement," Mr. Chapela said.

Primarily, critics suggested that the researchers misidentified sequences in the maize genes and mistook them for genetically modified material.

"The Quist and Chapela study is a testament to technical incompetence," said Matthew Metz of the University of Washington, a co-author of one of the two
criticisms. "Evidence for the presence of transgenic DNA in Mexican maize remains dubious and empirical."