When Maryland biologists responded to a report of dead and dying fish on Fishing Creek this June, they quickly found evidence that pointed to a microscopic killer.

Most fish kills are the result of suffocation when oxygen is depleted from the water. In this case, tests showed oxygen levels were high enough to suggest that the 1,700 dead menhaden — along with a handful of other species floating in the water — should have been able to escape.

But investigators of the June 17 incident noticed that the water in the slow-moving creek near Chesapeake Beach was stained with a reddish tinge caused by dense algae populations.

Lab tests showed the organism discoloring the water was the Karlodinium micrum — a known fish-killer in other places. But the smoking gun was a lab analysis that showed the water was filled with toxins produced by K. micrum.

“We were actually able to isolate the toxin in the water from the fish kill,” said Allen Place, a professor at the University of Maryland’s Center of Marine Biotechnology. “And it’s the same one that you get from [K. micrum] cells in culture. That’s about as strong evidence as you can get.”

The Fishing Creek event was the first time the species was shown to produce a toxin in the Bay that could kill. But recent detective work by Maryland Department of Natural Resources biologists suggests that the species might be a serial killer responsible for a number of previously unexplained die-offs over the last two decades.

“It is fairly common in the Bay,” said David Goshorn, the DNR’s chief of living resource assessments. “It is not a rare species, but generally it is found at pretty low levels.”

The Chesapeake is filled with hundreds of types of phytoplankton, of which only a handful are known or suspected of producing toxins harmful to other species.

Place and other scientists had become increasingly suspicious that the dinoflagellate (a free-swimming type of phytoplankton) K. micrum may be one of those bad actors in recent years. Concern was raised when Dan Terlizzi, a scientist with the Maryland Sea Grant College Program, began investigating kills involving thousands of fish at the HyRock Fish Farm on the Eastern Shore starting in 1996.

At first, scientists thought the kills might be caused by Pfiesteria piscicida, which was present in water samples.

A closer look, though, showed that blooms in the hybrid striped bass ponds appeared to be dominated by another dinoflagellate initially identified as Gymnodinium estuariale, a species known to be common in the Bay — but not toxic. Ultimately, scientists realized they were not looking at G. estuariale at all — it was Gymnodinium galatheanum (which was renamed Karlodinium micrum in 2000).

That changed the picture. G. galatheanum was first observed in South Africa the 1950s where it was associated with periodic fish kills which, at times, resulted in beaches being “strewn with thousands of tons of dead fish,” according to a fisheries report from the country.

At times, Gymnodinium populations were so dense, they turned the water red. Scientists in Norway subsequently confirmed that the species, when present in high numbers, could have toxic effects on fish.

It is also closely related to two other known toxic dinoflagellates: Gymnodinium breve, the organism responsible for red tides and neurotoxic shellfish poisoning in the Gulf of Mexico, and Gyrodinium mikimotoi, which has been associated with kills in aquaculture.

In the last several years, work by Place, his graduate student Jonathan Deeds, and a handful of other scientists along the East Coast have linked K. micrum to fish kills at a variety of aquaculture sites, and even a kill in a large runoff retention pond.

In laboratory tests using toxins obtained from kill sites, and directly from K. micrum cells, they have shown that the toxin kills small fish by eroding the epidermis, or mucus layer, that covers them. But in larger fish, the route of toxicity is most likely through the gills, Deeds said.

Recent research is revealing clues as to why K. micrum sometimes turns deadly. But the picture is complex, in part because K. micrum — like many other dinoflagellates — is a mixotrophic species. That is, it can get energy both from photosynthesis, or through the consumption of other single-cell organisms.

Place and Deeds believe the toxin is mainly a self-defense mechanism to help prevent K. micrum from being eaten. Work in collaboration with Diane Stoecker and her graduate student Matt Johnson at the University of Maryland Center for Environmental Science has suggested that tiny aquatic predators known as copepods are less likely to graze on high toxin-producing strains of K. micrum than low toxin-producing forms, Place said.

Beyond that, K. micrum, when acting as a predator, may also use a toxin to immobilize other singe-cell organisms that it is about to consume — some of which may be 80 percent of its own size, according to Place.

That hypothesis is bolstered by the observation that K. micrum cells multiply much faster when grazing than when living on sunlight.

While the dinoflagellate typically retains the toxin in its cell, the material is easily released when the cell is agitated, possibly to aid in self defense. In lab tests, the more the cells are disturbed, the more toxin they release. “If you shake a culture hard enough, you may get a twofold increase of the toxin in the water,” Place said.

Normally, those toxin releases would happen in small amounts to protect individual cells. But when the right set of conditions come together, they can create deadly results for fish. K. micrum can form dense blooms with up to 200,000 cells per milliliter (about 0.03 fluid ounces) of water.

If those cells are disturbed, it can result in a large release of toxin into the water.

It may be possible that a school of menhaden passing through a calm backwater may be enough to cause the release of the toxin, Place said. Likewise, filters and pumping systems in aquaculture can easily trigger a toxin release.

As a result, Place and Deets say, fish are probably not the target of the K. micrum toxin. Rather, they are probably innocent bystanders who happened to stir up the wrong group of dinoflagellates at the wrong place at the wrong time.

Like other dinoflagellates, K. micrum appears to thrive in nutrient-rich conditions, so the Bay has become a place more conducive to blooms in recent decades, and surveys suggest it is common in the Chesapeake.

In general, though, dense blooms with fish-killing potential would likely take place mainly in coves or other poorly flushed areas — such as Fishing Creek or aquaculture facilities — where large populations can build up. “We don’t expect this to be a problem out in the main area of the Bay,” Deets said.

Place and Deets believe the species has been in the Bay for years but — as was the case at HyRock — was often misidentified as the nontoxic G. estuariale.

As a result, Goshorn has been reviewing fish kill records dating to the early 1980s to see if G. estuariale was reported as being present in water samples. Out of about two dozen kills at which K. micrum was present, he said it looked like the dinoflagellate was a possible killer in at least two or three — and it may have been associated with another half dozen.

“When you look back over time, there have been some associations between fish kills and Karlodinium,” Goshorn said. “So you can’t rule it out as a contributor, especially in light of what we saw this year in Fishing Creek.”