Advances may make fish farming a healthy prospect, even for inner cities.

An economic renaissance is apparent in much of Baltimore, including the cobblestoned, 225-year-old Fells Point neighborhood that hugs a northern edge of the city's Inner Harbor. Some 18th-century storefronts display antiques, others feature homemade ethic cuisine. Construction on towering new hotels booms overhead.

Inside a nondescript, greenish-beige waterfront warehouse, marine biologists are laboring over their own contribution to this area's economic revitalization. The ambitious goal of these researchers at the university or Maryland's Center of Marine Biotechnology (COMB) is to redesign aquaculture, bringing teeming tanks of seafood to urban centers. Though they're starting with Baltimore, the next site could be Detroit, Iowa City, or even Phoenix.

At the Virginia Institute of Technology in Blacksburg, researchers are spearheading similar efforts to move aquaculture to nontraditional venues--in this case, depressed farm towns.

"For folks used to the commitment that a farm takes, such as milking dairy herds twice a day, fish may represent a welcome break," observes George J. Flick. A full-scale prototype that his Virginia Tech team is constructing in rural Saltville, Va., should "take at most only a couple hours a day to run," he says.

Key to the efforts of both groups is an emerging technology known as recirculating-tank aquaculture.

Conventional aquaculture raises fish in the open. Fish farmers usually corral marine species--sometimes in bays, other times within pens floating in the open ocean. Growers of freshwater species usually rear animals to market size in outdoor artificial ponds. These systems all rely on large volumes of clean water flowing to the fish and carrying waste away.

Over the past decade, such operations have been coming under intense criticism for fouling the environment.

Moving to tanks that recycle their water would not only limit the input of clean water but also reduce--and perhaps eliminate--discharges of waste streams. To achieve this, fish farmers have been adapting some of the technology that keeps finned specimens alive in home aquariums and zoo tanks. However, because the purpose of fish farming is not display, the aquarists' glass-walled tanks have given way to huge, opaque fiberglass pools--many holding at least 25,000 gallons of water.

Moreover, as the aquatic analog of cattle feedlots, aquaculture's recirculating tanks must nurture a far denser population of fish than standard aquariums do. They must also foster quick growth--while not compromising fish health. Overall, it's a tall order.

Yet if that order can be filled affordably, this technology could open aquaculture to new species and applications, notes Yonathan Zohar, COMB's director.

For instance, scientists have begun rearing fish for restocking the wild. Among them are species threatened with extinction. For some, their early life stages require more protection or more tightly controlled conditions than open pens or pools offer.

Zohar's team is breeding a tasty marine fish, known as bream, that doesn't naturally inhabit U.S. waters. The COMB researchers weren't permitted to import the fish until they assured federal regulators that none of the finned stock could escape, even into sewers.

Though a few commercial fish farms have already begun investing in recirculating-tank technology, such operations remain niche players in the United States' $900 million aquaculture industry, observes Jim McVey, director of aquaculture programs for the Commerce Department's National Sea Grant College Program in Silver Spring, Md. However, he argues, advances in this emerging technology leave it poised to assume a major role in satisfying the nation's appetite for freshwater fish and seafood.

Each year, the United States imports fish and shellfish valued at some $14 billion, McVey notes. Led by $2 billion worth of shrimp, this industry has evolved into one of the largest import sectors. "In terms of natural products," he explains, "it's second only to oil."

While precise numbers aren't available, McVey estimates that 60 to 80 percent of those aquatic imports have been farmed. He'd like to see the United States reclaim much of the market.

Because the young of marine species tend to be "smaller and more delicate than their freshwater cousins," he observes, they have been difficult to raise in captivity. Moreover, to be attractive to growers, stocks must spawn on demand, so that commercial-size fish can come to market year-round. Yet even in conditions emulating their spawning environment, the captive females of many marine species don't ovulate.

The coddling that recirculating systems permit makes this technology the best hope for growing marine fish, McVey says. With it, growers can precisely control temperature, wave action, oxygen supplies, nutrients, and daylight. There is a catch, however. Such operations today cost more than ordinary fishing or open aquaculture.

That's why growers who've begun investing in this technology tend to raise only high-value fish. Zohar acknowledges that economics provides one reason that his team works with gilthead bream. U.S. imports of bream now command up to $14 a pound.

In the mid-1980s, several fish reproductive endocrinologists, including Zohar, closed in on the problem of ovulation in marine species. The pituitary gland inside the brains of captive females had plenty of a particular gonadotropin, a hormone important to reproduction. In healthy spawners, the pituitary releases this hormone into the blood, but the hormone wasn't found in the blood of captive females. The finding suggested that these fish might be deficient in another class of hormones, called gonadotropin releasing hormones, or GnRHs.

Injecting captive bream with GnRH jump-started a cascade of events, including a release of the pituitary's gonadotropins into the blood. However, the injected GnRH broke down before it had its full effect. For instance, a bream might spawn for only a day or two following an injection--not daily for months, as it would in the wild.

So, Zohar, who was then at the Israeli National Center for Mariculture in Eilat, and his colleagues developed scores of synthetic GnRHs that could resist breakdown. Yet even with the best, Zohar notes, "we needed to inject fish more than once, which is not only labor-intensive but also stresses the fish."

Eventually, Zohar spent a sabbatical in Massachusetts working with chemists who specialize in the development of materials for the controlled release of drugs. There, he homed in on a polymer that could be used to deliver GnRH to the fish's body, meting out the hormone over a period of months.

When he moved to Baltimore a year later, in 1990, he began testing GnRH implants in U.S. fish that were infertile in captivity. These included Maryland's beleaguered striped bass.

COMB's Fells Point warehouse contains two 3,200-gallon recirculating tanks of 7-year-old progeny from that experiment--broodstock averaging 15 pounds apiece. Zohar's group is currently looking for the genetic controls that prevent GnRH production in the captive fish.

Together with commercial hatcheries, "we have now tested the [GnRH-implant] technology on more than 20 important species from around the world," Zohar notes. When dosages are tailored for each species, the technique works in them all.

These implants are crucial to the operation of George Nardi's 4-year-old commercial, recirculating-tank flounder farm at GreatBay Aquafarms in Portsmouth, N.H. He has long run a hatchery and is now looking to grow some of his young fish to market size.

The State of Maryland also relies on GnRH implants to trigger spawning of captive striped bass, shad, and Atlantic sturgeon--all endangered species that state biologists are rearing to restock rivers feeding the Chesapeake Bay.

In February, COMB announced a millennial milestone. By implanting each fish with GnRH and then modulating the conditions in the recirculating tanks late last year, the center's biologists fooled female bream into thinking they'd reached their Mediterranean spawning ground. In late December, the fish began a 3-month wave of daily egg laying. Their first young hatched on New Year's Day.

Though larval survival in traditional hatcheries can be as low as 5 percent, COMB biochemist Allen R. Place notes, "our rates are probably already above 70 percent." He credits this success to nurturing methods, disease-free recirculating tanks and a fat-enriched additive that he's developed to feed the hatchling's live prey.

Initially, the tiny fish feed on microscopic animals called rotifers. After 2 weeks, they graduate to Artemia, somewhat bigger species, better known as sea monkeys. A month or so later, scientists wean the larval fish onto an adult diet of pellets--an artificial blend of nutrients.

The problem with rotifers and Artemia, Place observes, "is that they are not nutritionally balanced for the [young] fish." So, his team has been "playing games" with the fingerling's tiny prey, tinkering with their fats, he says.

Docosahexaenoic acid (DHA), one of the primary omega-3 fish oils, plays important roles in neurological function and reproduction across a range of species, including humans. Fish and other vertebrates also rely on arachidonic acid, another fat, to make compounds for managing stress and immunity.

Most rotifers and Artemia have little DHA, Place notes, and some possess far less arachidonic acid than fingerlings need. So, he's developing feed to enrich these small animals in those fats.

However, the quantities of each fat that are right for bream might not serve another hatchling well. They might not even serve bream well later in their development, observes Moti Harel, also at COMB.

His data indicate that at first, fish need a high-DHA diet. Larvae without enough arachidonic acid grow well until metamorphosis and other periods of stress, such as when fish are moved between tanks or shipped to growers. Without this fat, the fish die in large numbers the first time they're stressed.

Harel's currently working to identify the optimal ratio of the two fats at various stages of a cultured fish's life. With that information, Place and others will be able to further improve the nutritional value of the prey.

Scientists are investigating not only what to feed delicate young fish but also how much.

With recirculating tanks, Flick notes, "you can see what the fish are eating. You can deliver just the right amount." That's important because feed may amount to 60 percent of a fish farm's fixed costs.

Moreover, excess food pollutes the water and can tie up its oxygen. Well-fed fish produce large amounts of feces, too.

Unlike open-aquaculture systems, Place notes, where wastes whisk away with the ocean current or effluent stream, recirculating tanks can quickly develop lethal concentrations of nutrients and wastes if the filters become overwhelmed.

The more crowded the fish, the bigger the challenge to a filtering system. Yet a high fish density is required for aquaculture to turn a profit. With freshwater tilapia, "we've been able to get up to 60 kilograms of fish per cubic meter," Place notes. "Essentially one-third of the tank is fish."

Spending a year or more in essentially the same water, these fish rely as much on filters as astronauts living aboard a spacecraft do.

The filters are large vessels--separate from the fish tanks--filled with water, a mixture of naturally occurring waste-degrading bacteria, and plastic beads. Fish farmers buy the components from supply houses, combine them on site, and then wait days or weeks for the bacteria to grow and coat the beads.

The bacteria in this biofilter can persist for weeks, months, or years, then die in a matter of days for reasons that remain a mystery, observes Harold Schreier, a microbial molecular biologist at COMB. It's a costly puzzle since a crash wipes out all the fish in the tank.

Schreier's group and others are now developing tests to identify signs of an impending crash. They are also considering techniques--such as changing the water's pH or salinity--to head off a filter's breakdown.

Despite their finicky nature, recirculating tanks can offer big advantages over conventional aquaculture, Nardi says. Summer flounder in the wild or in open pens take 3 or 4 years to grow to a commercial size of several pounds. His pilot-scale operation, producing 10 tons of fish per year, can bring those flounder to market in just 2 years.

Keeping these fast-growing fish healthy, he says, requires tight control of temperatures and day length. Recirculating systems liberate growers from vicissitudes of climate and need for copious water.

Nardi would like to see research deliver more reliable components for recirculating tanks--especially biofilters--energy-saving techniques for the power-hungry systems, and a better understanding of what it takes to scale up components. Quintupling the size of a tank, for instance, may require substantially more than five times as much filtering.

At least as important as these technical issues, Flick believes, are the financial ones. Virginia Tech's 12-year-old recirculating-tank aquaculture program has brought in economists to estimate the costs of each process employed in its new 80,000-square-foot Saltville facility. Sometime next year, it should begin producing 7.5 tons of yellow perch every 9 months--this freshwater species' period of maturation in recirculating tanks.

By detailing all of the system's costs, Flick's group hopes to help investors evaluate where this technology might succeed commercially. For instance, farmers with an empty barn might be able to set up an aquaculture operation without buying additional land or building a new structure. Others might shave costs by tapping waste heat emitted by an industrial neighbor or exploiting the low-cost water and shelter offered by an abandoned mine.

If transportation or labor costs turn out to be high, investors might save money by siting tanks within low-tax urban-revitalization areas, near a ready source of labor and consumers of their product.

A totally closed system might be engineered to meet even stringent urban environmental standards. Most recirculating tanks today still release a few percent of their volume daily to sewer systems as disinfected waste water. The goal, Flick and others note, is tanks that don't discharge any liquid wastes.

Although the federal government has been funding research into recirculating aquaculture for 30 years, Flick cautions that "it's still a new and developing science"--one with plenty of kinks to work out.

"We have tended to fund research on individual components," McVey says, like a new filter or better fish food. "Now, we need to see what it takes to integrate all of these into standardized systems" that prove reliable and affordable on the farm or in the city. Where these systems finally take off and when, he says, "will, in the end, all come down to economics."