LEAH, Ga. -- Susan Wilde is all too familiar with Thurmond Lake's deadly secret.
"It has the perfect environment for something we don't want," she said. "The problem is finding a way to fix it."
Wilde, an assistant professor at University of Georgia's Warnell School of Forestry, is part of a team working to unravel the mystery behind a tiny algae bloom blamed for hundreds of bald eagle deaths across the Southeast, including at least 54 at Thurmond Lake.
The disorder, known as avian vacuolar myelinopathy, or AVM, is believed to be spread from algae that grows on hydrilla -- an exotic weed from the aquarium trade that was first detected in the lake in 1995.
Hydrilla is a choice food source for small waterbirds called coots, which in turn are eaten by bald eagles lured by the lake's erratic shoreline and abundant nesting trees, said Army Corps of Engineers biologist Ken Boyd.
"It causes lesions in their brain," he said. "They lose equilibrium and soon they can't fly. As you move into a group of coots, you can pick out the sick ones that get left behind."
So far this winter, two bald eagles have been found dead at the lake. One was too badly decomposed to test. The other, found in Lincoln County near Camp Daniel Marshall, was confirmed to be a victim of AVM. They were the 53rd and 54th dead eagles found at the lake since 1999.
"You also have to consider that when you find a few dead ones, there are lots of others that no one ever finds," said wildlife biologist Tom Murphy, who managed South Carolina's bald eagle recovery program for three decades.
Last week, Wilde and other scientists returned to the lake to gather samples of the unwanted hydrilla -- and to discuss ways to control it. "We're mainly looking at algae growth on the hydrilla," she said.
The slimy plants are taken to Athens, Ga., and mounted on slides to be scanned for signs of the deadly algae. Of the 17 lakes from North Carolina to Texas where AVM and bird mortality have been identified, the algae densities at Thurmond are among the greatest.
"You consistently have consistently high leaf coverage here," she said. "There's something perfect about this environment."
Wilde and her colleagues have identified an algae species in the order Stigonematales as the most likely suspect in the bird deaths.
In addition to eagles, AVM has also been confirmed in owls, Canada geese and ducks, but there is no evidence -- so far -- that it can be spread to mammals. In one experiment, infected coots were fed to pigs to see if they developed the fatal brain lesions; they didn't.
The hydrilla's presence at Thurmond Lake -- and its ability to spread rapidly to other waterways -- is a concern among resource agencies hoping to stem the loss of bald eagles, which have slowly recovered after being poisoned to near extinction by pesticides.
One possible remedy being explored for Thurmond Lake involves the introduction of sterile, grass eating carp that would consume the hydrilla -- and possibly eradicate AVM.
The fish, which are sterilized to prevent them from reproducing, were successfully used at South Carolina's Lake Murray, where 64,400 were stocked in June 2003. At that time, the lake's hydrilla infestation covered an estimated 3,880 acres. By 2005, the hydrilla was gone -- and no confirmed AVM cases have occurred there since.
It is a potentially controversial step. Hydrilla and aquatic weeds are welcomed by fishermen because they provide structure and cover. Adding a new species would also require environmental assessments, and -- very likely -- public hearings.
NEITHER GEORGIA NOR South Carolina has come forward with a specific request for such a program, corps spokesman Billy Birdwell said. "We would be supportive if the states came and requested it," he said. "But we would certainly not do it on our own without the states walking hand-in-hand with us."
One concern is carp would eat much more than the unwanted hydrilla -- including native vegetation -- and could change the lake's ecology in unforeseen ways.
Scott Hyatt, Thurmond Lake's project manager, summed it up this way in an e-mail to residents inquiring about such a program:
"Carp are like goats. Yes, goats will eat the kudzu, but they will also eat your grass, your shrubs, and sometimes your neighbor's laundry if they leave it out on the line," he said. "Carp are not particularly discriminatory in the vegetation they eat, so we have to recognize that their use may also wipe out native aquatic plants and have an impact on other fish species that rely on them for forage or cover."
Georgia Wildlife Resources Division fisheries biologist Ed Bettross said Georgia doesn't have an official position on carp. "There has been no hard decision one way or the other because there is really no proposal for us to review at this time," he said.
Wilde acknowledged the use of carp could be controversial.
"While very few in both natural resource agencies are adamantly opposed to grass carp, no one really wants to take the heat if they get public opposition," she said. "We are developing a public opinion survey for the Corps of Engineers that we will help them distribute. The first step that all have agreed on is that we do need to educate the public about the concerns and include them in the evaluation process."
Thurmond Lake's hydrilla mats are concentrated along shallow shorelines and -- at one time or another -- have affected about 7,300 acres, said Jamie Sykes, a corps fisheries biologist. This year, fewer hydrilla patches have been found, although higher water levels make detection of the weed more difficult.
Wilde said a recent experiment involved placing sterile grass carp in a pond packed with algae-contaminated hydrilla. "They ate every bit if it," she said.
Oddly enough, the carp developed brain lesions after eating the algae-encrusted hydrilla, she said. "They got the same lesions, but they didn't die," she said.
The experiment was carried a step further when the infected fish were fed to chickens, which did not develop AVM.
"There are lots of other pathways we haven't fully explored," Wilde said, noting that any solution to prevent eagle mortality will have both costs and consequences.
"We wish there was a magic plan to fix it without any side effects," Wilde said. "But there isn't."