Virginia scientists investigating a mysterious fish kill in the James River last month found a microscopic parasite that they say also could have caused earlier kills blamed on Pfiesteria piscicida.

However, researchers in Maryland said yesterday that they remain sure that pfiesteria, a toxic microbe, was responsible for killing thousands of fish on Maryland's lower Eastern Shore in 1997. Moreover, they said, subsequent research has found a nontoxic form of the organism in many areas of the Chesapeake Bay.

The distinction between fish killed by the Kudoa parasite and those dead because of pfiesteria is significant because Kudoa is not harmful to humans, while pfiesteria is believed to have sickened more than two dozen people two years ago. The newly stricken fish have been mainly Atlantic menhaden, a bony, oily fish used mainly to make meal and fish oil.

"They're never caught for human consumption," said Greg C. Garman, a fish biologist at Virginia Commonwealth University who participated in the James River investigation. "You'd have to be really hungry to eat a menhaden."

While scientists may dispute the immediate cause of the fish kills, there is little doubt about an underlying reason for the affliction: deteriorating water quality brought on largely by pollution from nutrients such as fertilizer runoff and wastewater from treatment plants and septic systems.

The Virginia scientists made their discovery three weeks ago when routine monitoring of the lower James River found hundreds of Atlantic menhaden with "classic pfiesteria-like symptoms," including lesions, Garman said.

But no trace of pfiesteria was found at the site. Instead, about a week later, the scientists determined that all the fish they had collected were heavily infected with Kudoa, a protozoan parasite more commonly found in ponds. Because the fish they caught were still alive, Garman said, they had not yet become infested with "scavenger critters" such as bacteria and fungi that invade a fish within hours of its death and make studying it more difficult.

"We sort of caught an event as it was happening that was classically pfiesteria without the pfiesteria," Garman said. "We're now proposing that Kudoa may be, probably is, an alternative explanation for certainly this most recent pfiesteria-like event and probably some of the others over the last five years."

The discovery "doesn't rule out pfiesteria" in the Maryland cases, he said, "but it means we probably shouldn't be focusing all our efforts in research and monitoring on just pfiesteria. It's at least as likely that the events that occurred two years ago in this region were caused by Kudoa."

But David Goshorn, a senior researcher with the Maryland Department of Natural Resources, said separate laboratories in two states confirmed that toxic pfiesteria was present in Maryland's 1997 outbreaks.

"I'm not disputing what they found this year," Goshorn said of the Virginia scientists. "But you can't draw conclusions on what happened two years ago in a different river based on what they're finding now."

Several things can cause fish health problems, Goshorn said, including parasites, fungi, bacteria, chemical contaminants and pfiesteria. Although there have been no toxic outbreaks in Maryland since 1997, he said, nontoxic forms of the pfiesteria have been found "in a lot of areas in the Chesapeake." The good news, he said, is that the microbe releases toxins only under certain conditions--such as warm water, moderate salinity and high levels of nutrients--that have not recurred in the last two years.