Young American shad will soon begin their perilous fall migration out of Bay tributaries and into coastal waters where, if they're lucky, they will live the next four or five years until they return as adults to spawn.

As they swim downstream, they will have to find their way past predators their ancestors never had to deal with, such as blue catfish in Virginia rivers and the Potomac, largemouth bass-even snakeheads. In some rivers, the young shad will have to make their way through power-producing turbines at hydroelectric dams.

When they make it to the Bay, they'll have to get past a hungry and abundant population of striped bass. Not only is the population large, individual fish on average are bigger than they have been in decades, meaning they need more to eat.

Along the coast, the shad that survive will face more predators and a new hazard: an array of fishing gear targeting other fish that often snares shad as well.

By various estimates, only one out of every 300-400 larval shad survives to return. But the odds may be getting even worse.

In recent years, the shad population in almost every major Bay tributary except the Potomac has been dropping for reasons that are unclear. Others, at best, are holding their own.

And the Chesapeake is not alone. On the Hudson River, the annual shad festival was shad-free this year-to conserve fish.

Last year, the Atlantic States Marine Fisheries Commission, a panel of state and federal fishery managers responsible for managing migratory fish, completed an American shad stock assessment that concluded that shad stocks along the East Coast are "currently at all-time lows."

No one is certain why shad, which have been the target of large-scale stocking efforts, fish passage projects and a host of fishing regulations along the coast, have not rebounded.

"That's a very good question and one that everyone is struggling with," said Jack Travelstead, fisheries director for the Virginia Marine Resources Commission, and a member of the ASMFC. "The status of the population does not appear to be the result of the current shad fishery because the amount of harvest is miniscule."

Few rivers have fisheries that target shad; those that did have closed in recent years, or are being ratcheted back.

Suspects in the mystery include increased predation by other fish both in estuaries and along the coast, degraded water quality from pollution, loss of habitats such as underwater grass beds along migration routes and the loss of shad as bycatch in other fisheries operating along the Atlantic Coast.

Shad once supported one of the largest fisheries along the coast, with landings peaking at about 50 million pounds at the beginning of the 20th century. That had fallen to 3.8 million coastwide in 1980. That year, landings in Maryland dropped to 25,000 pounds, and the state imposed a moratorium on shad fishing. The Potomac followed suit in 1989, and Virginia closed its rivers to shad fishing in 1994.

While fishing was cut back, massive efforts were undertaken to restore fish. Pennsylvania, Virginia, Maryland and-most recently-Delaware launched hatchery operations to rear shad for release in their rivers. Hundreds of millions of larval shad "fry" have been stocked in recent decades.

Meanwhile, major efforts were made to reopen historic spawning areas to returning fish. On the Susquehanna alone, which historically had the most spawning habitat of any river on the East Coast, utilities invested roughly $100 million to build passages at four hydroelectric dams in the lower part of the river.

On the James, a passage at Boshers Dam near Richmond, along with other projects, have opened almost the entire drainage to fish migration. On the Rappahannock, the Embrey dam outside Fredericksburg was blown up in 2004 to help clear the way for spawning shad.

Similar strategies were pursued on other rivers along the East Coast. To further help those efforts, the ASMFC in 2000 began phasing out commercial ocean fishing that targeted shad. Because shad from many rivers may migrate together along the coast, managers were concerned that shad fisheries in the ocean indiscriminately caught fish, thwarting recovery efforts in many rivers.

That fishery was fully phased out in 2005, and effectively left shad management on a river-by river basis so areas with seemingly healthy shad runs could have small fisheries, while others could protect their fish.

For a while, it looked like all of those efforts were paying off. Shad runs almost everywhere around the Bay-although low compared with historic levels-seemed to be edging up.

But in recent years, that trend has halted or reversed. Biologists acknowledge there is no single "smoking gun" in the decline, and say several factors may be to blame.

In fact, the ASMFC's three-volume, 1,200-page stock assessment acknowledged that it "may not provide definitive answers to all the questions plaguing management of Atlantic coastal American shad" but did offer "insight to management on the complexity of the issues."

Nowhere around the Bay is the the story more dismal than on the Susquehanna. This spring, only 19,912 fish were counted at the Conowingo Dam, the lowest number since a $10 million elevator began operating in 1997 to lift fish over the 100-foot-high obstruction.

Initially, shad numbers returning to Conowingo showed encouraging signs. In 1997, 90,971 passed the dam; by 2000 the number reached 193,574. Since then, it has trended strongly down.

No single factor seems to explain the decline.

But getting fish past all four dams and into spawning waters has been problematic. On average, only about 2 percent of the fish that are lifted over Conowingo make it past the fourth dam at York Haven. This year, only 21 fish made it past all four.

"Probably 80 to 90 percent of the quality nursery and spawning habitat lies upstream of the hydroelectric dams," noted Larry Miller, who heads the Mid-Atlantic Resource Office of the U.S. Fish and Wildlife Service.

Ironically, more fish got upstream before the passages were built. Prior to the completion of the fourth and final passage in 2000-which theoretically opened the Susquehanna to migrating shad-utilities paid to truck as many as 50,000 migrating adult shad a year past the dams and release them in spawning habitat.

"Since we are not trucking fish anymore, we rely on fish passage to get them in up river where they spawn on their own," said Mike Hendricks, a fisheries biologist with the Pennsylvania Fish and Boat Commission. "Since fish passage isn't working as well as it needs to, it's been a number years since we've had good production of juvenile wild fish."

The owners of the Holtwood Dam, which is immediately upstream of Conowingo, have proposed changes aimed at improving passage at the dam, which in most years passes less than half of the fish that make it over Conowingo. Fisheries officials hope to negotiate improvements to passages at other dams as their federal licenses come up for renewal.

But passage is not the only issue on the Susquehanna. Hatchery production has been poor since 2001 because of a variety of problems. This year, 2.5 million fry were reared in the Fish and Boat Commission's hatchery on the Juniata River, a Susquehanna tributary. That was better than most recent years, but far below the 10 million stocking goal.

Hendricks doesn't expect that goal to be met anytime soon because of shad problems elsewhere. The Hudson, once the main egg source for the Susquehanna hatchery, is closed because of its own plummeting population. "You can't politically justify allowing some other state to take fish on a stock that is declining," Hendricks said. "No one would do that." The Delaware, another source, has also been closed for Susquehanna use because of concerns about its stock.

The lack of spawning adults, and lack of hatchery production, means few juvenile shad leave the river. "If they don't leave, they can't come back," Hendricks said.

Elsewhere, surveys by the Virginia Institute of Marine Science show that shad runs on both the James and York rivers-once important shad spawning areas-are trending down. The index on the James has fallen sharply since 2003, and this year's index was the lowest since 1998, when the survey began. The survey shows the York River has been declining for a decade.

The James decline is especially disappointing because the river has been the subject of an intense stocking effort dating to 1992. Since then, biologists have stocked 104 million shad fry.

"Things looked great until 2003," said Dean Fowler, a biologist with the Virginia Department of Game and Inland Fisheries. "We were trending up. The runs appeared to be increasing in abundance every year."

Trends in many other rivers are less clear. In Maryland, the lack of surveys directed at adult shad make it difficult to determine trends. The state has heavily stocked the Patuxent River, once devoid of shad, for more than a decade and more recently has started stocking the Choptank and Nanticoke rivers.

While those stocking efforts have brought shad back to the rivers, Brian Richardson, a biologist with the Maryland Department of Natural Resources, said "we should be seeing more adults."

"We have all these shad restoration programs up and down the coast, and you can do everything right," Richardson said. "But if there is natural mortality or some other mortality out there on the coast, you're kind of spinning your wheels."

Yet the story is not universal. In Virginia, the VIMS survey shows that the Rappahannock has had a small but stable-and perhaps slightly increasing population-over the last decade. The Rappahannock has not been the target of restoration efforts until the last few years when the Embrey Dam near Fredericksburg was removed and large-scale stocking efforts began. But that was too recent to influence the 10-year trend.

Across the river, surveys in the Nanticoke also suggest the stock may be small, but stable, although recent years seem to be far below peaks seen in 2003 and 2004.

Meanwhile, the Potomac, almost alone along the East Coast, has shown a marked increase over the past 15 years. An annual survey during the spring shad run caught 23.6 shad per net this year, slightly more than the 15-year average of 22.3 and significantly higher than the 13.4 average seen in 1995.

"It was a decent run," said Jim Cummins, a biologist with the Interstate Commission on the Potomac River who has been working for years to restore the river's shad.

Cummins can tick off reasons why he thinks the Potomac shad run has bounded back. Water quality in the river has improved dramatically since the 1960s, when the Potomac was considered a national disgrace. Underwater grass beds, which provide habitat, have rebounded. Fishing moratoriums, both in the river and along the coast, have helped to protect the shad. A fishway completed at the Little Falls dam in 2000 reopened 10 miles of spawning habitat that Cummins believes were historically important and largely protected spawning shad from predators. Stocking efforts have likely helped as well, he said.

But the main reason, Cummins said, is that the Potomac has some unique feature, yet to be identified, which makes it especially productive, even though it has relatively little spawning habitat. Great Falls just upstream from Washington blocks the river to migration-in other systems, shad sometimes have hundreds of miles of habitat for spawning.

Despite the seeming lack of habitat, historic records indicate "almost unbelievable" runs on the river, Cummins said. In the early 1800s, as many as 20 million shad were caught per year, with little apparent impact on the population: The fishery did not peak until the late 1800s.

Nonetheless, Cummins sees some cautionary signs on the horizon. In fact, he was disappointed in this year's run. In 2004, the survey averaged 39 fish per net-its highest number ever. Fish produced from that record run began reaching spawning age this year. "I was expecting to see more of them entering the system and was expecting a big run," he said.

Another concern is that the five-year survey average from 1995-99 was 16.7 fish per net. It leaped to 30.1 from 2000-04. But from 2004-08, the average dipped slightly to 27, which may suggest the rebound may have started to plateau.

Like many others biologists, he strongly suspects that something not fully understood is taking a coastwide toll on shad. The ASMFC stock assessment said that total shad mortality in the ocean is unknown, and can come from a variety of sources, including predation by other species and being caught as bycatch in fisheries targeting other species. The lack of information hampers efforts to explain what's driving the trends.

Rebounding striped bass numbers have been cited as a suspect by some. In places such as the Connecticut River, there appears to be a correlation between rising striped bass numbers and declining shad. It's unclear if that's a coastwide problem, according to the stock assessment, but some suggest that could be the case in the Bay, as well, where the striped bass numbers are high, and the availability of other forage fish, such as menhaden and bay anchovy, is thought to be low. "It's not hard to imagine that it would be the case in the Chesapeake, where we believe the resident stripers are limited by forage," said Bill Goldsborough, a fisheries scientist with the Chesapeake Bay Foundation and member of the ASMFC.

Coastal waters are also filled with a host of other species that eat shad, such as bottlenose dolphins, sharks, seals, other fish and birds. It's unknown how many shad are eaten by other species, or whether those trends have changed over time. "They're spending nine-tenths of their lives in the ocean, and they are principally a food fish for other things," Cummins said.

While fisheries targeting shad have been closed in the ocean, there's growing concern that large numbers of the fish may be caught in other fisheries.

Species- specific bycatch is often unrecorded because of the difficulty of distinguishing between fish. Many fisheries along the coast harvest small bait fish. While the bycatch of more lucrative species, such as striped bass are recorded, young shad and other species are listed as "unclassified bait," according to ASMFC.

The same problem exists with some fisheries that target larger fish in coastal waters, such as Atlantic herring. Many species are caught as bycatch, but are often not identified. "We need to resolve the extent to which bycatch exists, and then we need to correct it if it does exist," Goldsborough said.

But he cautioned, that is not as easy as it sounds. "Fish in the herring family [which includes shad] can look awful familiar, and there are a lot of different ones, some of which might be open to harvest," he said. Trying to understand how many shad might be caught means sending trained biologists on fishing boats to identify what's being caught before the fish are discarded, a potentially expensive endeavor.

If bycatch is a problem, it would also mean other fisheries would have to be managed differently to protect shad. That can be problematic because fisheries are typically managed independent of each other.

Further, shad don't stop swimming at the U.S. border when they migrate north during summer months, and it's unknown whether Canadian fisheries might have some impact.

Other factors could be playing a role, such as changing ocean currents, and even changing climate. In Virginia, Fowler said surveys by DGIF biologists on the Pamunkey River, one of two major tributaries to the York, show that survival of young shad is poor in years when July is drier than normal.

"It seems we've had a lot of drought years the last 10 years," Fowler said. "Whether that is the only driver behind this decline, we don't know."

If it is a major factor, it bodes poorly for this year's stocking effort. Although 6.9 million shad fry were stocked in the James and its tributaries, making it a better than average year, July was also drier than normal. "We can stock fish until we're blue in the face." Fowler said, "but if the rearing conditions aren't there, the fish aren't going to survive. Mother Nature has to cooperate, too."

In addition, concerns linger about the quality of spawning habitat in rivers themselves.

The ASMFC is exploring management options and is expected to make recommendations this fall, which likely will include setting river-specific population targets to guide management; gathering better information about various factors affecting the shad stock, including predation and bycatch; and further restricting any harvests as the first steps toward restoring shad to its past glory.

The Latin name for shad is Alosa sapidissima -sapidissima means "most delicious"-and the fish was once valued all the way from the estuaries to the headwaters of major rivers. But in many areas, people have not savored that taste for years.

"The shad are the silver thread that binds us," said Miller, of the USF&WS. "They bind the folks in the Upper Susquehanna to the folks in the Bay to the resources off the state of Maine. All of them have to be clicking right for anyone to get the benefit of having shad."