The student, Daniel Donato, found that forests recover from wildfires best when they are not logged, a notion that conflicts with the administration's decision to log trees that burned in southwest Oregon's Biscuit fire and with a bill in Congress that would speed logging after wildfires.

Donato's work was published this month in the journal Science, even though the Oregon State scientists asked the publication not to print the research.

The dean, Hal Salwasser, in a letter to the university Thursday, said he should have told the professors to voice their criticism through open scientific debate. He added that he should have congratulated Donato.

"Few faculty, let alone graduate students, get their work published in this prestigious journal," wrote Salwasser, who testified in favor of the bill that would accelerate logging after fires.

The episode created concern that researchers at the respected forestry school face a backlash if they reach conclusions that clash with the timber industry and leading faculty. The College of Forestry gets about 10 percent of its funding from a tax on logging.

Donato and his research team concluded that logging slows forest recovery. They found that logging after the Biscuit fire destroyed seedlings and littered the ground with highly flammable tinder.

As is customary, Science's editors had independent scientists review Donato's research.

The findings differed from a report written earlier by John Sessions, a distinguished professor of forest engineering at OSU, and Professor Emeritus Michael Newton. That report suggested aggressive logging would restore forests after 2002's Biscuit blaze.

Newton and Sessions were among the nine OSU scientists and professors, plus the U.S. Forest Service, who asked Science editors to delay publication of the study until it addresses their criticisms.

They said the conclusions were premature and that the true test of efforts to restore forests will require decades.

(Copyright 2006 by The Associated Press. All Rights Reserved.)