During recent work on the Chesapeake Bay, David Kidwell discovered what was long known by watermen: If you want to find oysters, look for ducks.

The wintering ducks float on the surface, periodically diving to pick off mussels, clams and other prey associated with oyster bars.

"The watermen knew this, but the scientists didn't," said Kidwell, who quickly learned that lesson while conducting duck surveys on the Bay for the U.S. Geological Survey's Patuxent Wildlife Research Center.

While other studies have shown that oyster reefs provide important habitat for a variety of aquatic life under the water, Kidwell's research suggests they are also important for ducks floating on the surface.

"What happens underwater is definitely affecting the ducks," he said.

That may be bad news. The demise of the Bay's oysters may have affected surf scoters and other sea ducks, according to Kidwell.

While they don't eat oysters, the ducks do dive into the Bay's murky waters to pull off hooked mussels-their preferred food-which thrive mainly on oyster beds.

"My guess is that we have a lot less food," said Matthew Perry, the principal investigator at the center for Atlantic seaduck research. "We only have 1 percent of the historic oyster beds in the Bay, so do we also have only 1 percent of the hooked mussel, which is the major food of the scoter? You don't have to run a lot of statistics on data like that. That's a major change."

The study was part of a broader effort by scientists from the United States and Canada to learn more about seaducks. Little is known about these waterfowl because they nest in areas of Canada that are difficult to reach. They also overwinter offshore in the Bay and other coastal waters, making them hard to spot-aerial surveys used to count ducks, geese and swans often can't find seaducks.

While many other duck species are rebounding from historic lows, concern has grown that the 15 species of sea ducks have dwindled in recent decades, spurring the creation of the Sea Duck Joint Venture, a coalition of U.S. and Canadian agencies and scientists, to learn more about the birds.

Throughout its range, the surf scoter population is thought to have declined by about half since the 1950s. Scientists at Patuxent are trying to determine whether changes in the Bay may have played a factor in the decline of scoters and a handful of other sea ducks that overwinter on the estuary.

Oyster populations are at all-time lows because of disease and overfishing. Not only have oysters largely vanished, but so have their reefs, which gradually become silted over as they lose the water-filtering capacity of oysters.

Those bars are habitat for a host of organisms, including the hooked mussel-a cousin of the blue mussel commonly consumed by humans.

"You are not going to find hooked mussels in sand bottom," Kidwell said. "You are typically finding them associated with hard bottoms, especially oyster bars or degraded oyster bars."

To study the link between bottom habitats and surf scoters, Kidwell examined a 16-square-mile section of the Bay, roughly bounded by the West River and North Beach on Maryland's Western Shore, and the southern tip of Kent Island and Tilghman Island on the Eastern Shore.

During the winter, he repeatedly surveyed the area, counting birds, documenting their locations and sampling the benthic communities. Over time, he noticed that large groups-or "rafts"-of scoters repeatedly congregated at three sites.

Divers confirmed that two locations were remnant oyster reefs with a sprinkling of live oysters. The third was largely sand. The scoters noticeably avoid sites that were silted over, which accounted for much of the study area. "They were selecting against mud," Kidwell said.

Foodwise, there's a big difference in sand and hard bottom. Sand habitats tend to have fewer species of clams, and clams and mussels are less abundant overall. Further, they do not have hooked mussels.

Because areas with hard bottoms are usually small, they are in close proximity to sandy areas, giving ducks a choice between hooked mussels and dwarf surfclams.

Although the ducks eat both, Kidwell said, hooked mussels appeared to be the favorite. During the winter, his surveys showed a dramatic drop in the density of hooked mussels in the areas where the ducks were feeding. "With the dwarf surfclam, I didn't see anything like that," he said.

The sites that he examined generally had hooked mussel densities that were only in the dozens per square meter. But studies by Kennedy Paynter, an oyster researcher with the University of Maryland, have found densities of 3,000 hooked mussels per meter on restored oyster sites.

"You can just imagine what that dramatic increase would do for the birds," Kidwell said. "That's a lot of food."

Researchers can't say whether duck populations are limited by the food supply today, but it's likely the Bay had the capacity to support more ducks in the past, when there were more oysters-and oyster bars.

Also, other species of waterfowl have successfully switched from one food item to another. That happened with the canvasback duck, once the most common wintering duck on the Bay. It used to eat underwater grasses, but as grasses disappeared from the Chesapeake, the canvasbacks that continued to overwinter on the Bay switched to clams.

But studies by Perry and other researchers at Patuxent also showed that clams were not as nutritious, and required canvasbacks to exert more energy to get their meals.

That could also be the case with surf scoters.

"They seem to be exhibiting this shift to feeding more on the dwarf surfclam, and I think that is because of availability," said Alicia Berlin, a research assistant at Patuxent.

Berlin is trying to learn what this might mean for the scoters. Using a pair of 8-foot-tall concrete tanks filled nearly to the top with water, she is learning which food the ducks prefer to eat, and which provide the best energy.

She's also looking to see how much energy the ducks have to spend to get their meals. Dwarf surfclams are smaller, so the ducks have to consume more of them-and likely make more dives-to get the same energy value.

The clams are typically buried in sand, while the mussels attach themselves to solid surfaces, forcing the ducks to yank them free. In her experiments, Berlin places trays on the bottom with mussels glued to a fake oyster bed-oysters set in concrete-and trays of sand with buried clams. Different trays have different densities.

Surf scoters won't dive in public-"they're too shy," Berlin said-so she videotapes the dives to watch later. She studies how long it takes them to forage, and how many mussels and clams they consume over a period of time.

If they aren't getting enough high-quality food, it may mean that surf scoters will likely continue to decline in the Bay. That could also hold true for other seaducks that winter on the Bay and include hooked mussels in their diets-such as the black scoter and white-winged scoter.

Perry said it's unlikely that the decline of oysters-and the likely related decline of hooked mussels-is solely responsible for the decline of seaducks.

"The food is definitely one factor," Perry said. "but you have all the other anthropogenic factors-oil and contaminants and hunting and loss of habitat, higher nutrients in the Bay. There are a lot of factors."

But one conclusion, Perry said, is that if the Bay had more oysters, it should help seaducks. "I can't imagine that it wouldn't."

Seaducks' life far from land makes it difficult to study them

The 15 species of seaducks found in North America are the most poorly understood group of North American waterfowl. Sea ducks spend much of their time a mile or more from land, diving to the bottom to find food, which makes it difficult to find, let alone observe, them.

There are few reliable population indices or estimates of annual productivity for any of the species. One problem is that surveys are not designed to accurately estimate the harvest of most seaduck species.

The life cycles of seaducks may also be problematic for their long-term well-being.

Seaducks have long life spans. Most species do not reach sexual maturity until they are 2 or 3 years old. Their reproductive success is low even in relatively stable environments.

Changes occurring in environments inhabited by seaducks do not bode well for these birds.

Many traditional wintering areas have become urbanized and industrialized, and breeding habitats are being lost or degraded. Thriving predator populations are placing increased pressure on the annual production of a number of seaduck species, and many indirect factors, such as the bioaccumulation of chemical contaminants, may be negatively affecting survival and production in some populations.

Continentally, seaducks represent less than 5 percent of the total waterfowl harvest. As a result, their conservation has historically been a low priority among waterfowl managers.

Seaducks received no special consideration under the original 1986 North American Waterfowl Management Plan, an international agreement aimed at improving management, because populations were thought to be stable.

Very little data existed at that time, though, and since then, the eastern population of harlequin ducks has been listed as endangered in Canada, and spectacled eiders and the Alaskan breeding population of Steller's eiders have been listed as threatened in the United States.

Analysis of existing survey and harvest data along with new surveys and studies conducted during the 1990s indicate population declines in 10 of the 15 species of North American sea ducks.

Sea ducks found on the Chesapeake Bay include: black scoter, bufflehead, common merganser, hooded merganser, common goldeneye, long-tailed duck, red-breasted merganser, surf scoter, white-winged scoter.

Sources: Sea Duck Join Venture, Chesapeake Bay Program