Microbe may kill fish by skinning, not poisoning. (Pfiesteria's Bite).

Arguments have taken a strange turn over how to isolate toxins from the Pfiesteria microbes accused of killing fish by the millions and threatening human health. Two research teams now say that in the Pfiesteria strain they've examined, there's no toxin to find. Instead, explains one of the teams, the single-celled dinoflagellate alga kills by swarming over a fish and eating away its skin.

A laboratory sample of Pfiesteria shumwayae, one of two named Pfiesteria species, slays fish only by direct contact, says Robert Gawley of the University of Miami in Coral Gables. By centrifuging a solution of P. shumwayae growing in the lab, the scientists removed the microbes from the mix. The leftover liquid, which would contain any released poisons, proved harmless, Gawley's team reports in an upcoming Proceedings of the National Academy of Sciences.

Results from other tests prompted the same conclusion from Wolfgang Vogelbein of the Virginia Institute of Marine Science in Gloucester Point and his colleagues. They report that fish stayed healthy when bathing in the same liquid as P. shumwayae, as long as filters prevented direct contact.

Also, microscope images of fish in contact with P shumwayae showed algal cells gouging the fish's skin, Vogelbein's team reports in an upcoming Nature.

Debates over Pfiesteria biology have heated up since 1992 when JoAnn Burkholder of North Carolina State University at Raleigh and her colleagues blamed this genus for filling North Carolina rivers with dead fish with bloody skin ulcers (SN: 9/6/97, p. 149). In the late 1990s, alarm rose that breathing spray from Pfiesteria-laden waters, in the lab or outdoors, impairs people's mental functioning.

Gawley says that when he started his current work, he intended to find the toxins causing such havoc. Besides testing centrifuged samples, his team looked for genes encoding enzymes that other dinoflagellates seem to use to make toxins, but none turned up in P. shumwayae.

Vogelbein and his colleagues tested the same strain of P. shumwayae that Gawley did. "The only time we saw fish dying was when they were in contact with the Pfiesteria," say's Vogelbein's collaborator Jeffrey Shields of the Virginia institute.

Burkholder criticizes both papers for not using a sufficiently toxic strain of Pfiesteria. The ones in her lab knock out the fish much faster, within a few hours instead of a day or two. She says that when her team exposed shellfish larvae to Pfiesteria contained within a dialysis sac, the larvae died rapidly even though they had no direct contact with the algae. Burkholder adds that the chemists with whom she works have found a toxin.

The two new papers join other recent work to suggest that fears of Pfiesteria have been overblown, says Wayne Litaker of the National Oceanic and Atmospheric Administration in Beaufort, N.C. He's seen Pfiesteria feeding directly on fish in laboratory tanks, but he cautions that wild, free-swimming fish may avoid lethal densities of these algae.

Some 40 other menaces--from low oxygen concentrations to a water fungus--can give fish bleeding skin ulcers, Litaker says. In June, his lab reported a 7-stage lifecycle for Pfiesteria, instead of the unusually complex sequence of more than 20 stages that Burkholder has described.

Litaker says that the microbe once seen as an extraordinary poisoner may turn out to be just "a normal, everyday dinoflagellate."