After hauling a 100-foot seine net around a semi-circle from the beach, two biologists dragged the net out of the Mattawoman Creek and onto dry land.

Inside the net, more than a thousand fish - ranging from fingernail-size to several inches in length - flailed about.

A half dozen biologists swarmed around and began dividing up the catch.

Lines of bluegills, piles of striped bass, and groups of bay anchovy began to form on the sandy beach. "Anyone doing white perch?" one sorter called out, holding a few specimens in her hand.

They counted the total number present for each species, measured the largest and smallest among them, and divided them by age or "year class." They shouted the results to a record keeper: "White perch - 25 so far." "Minimum white perch, zero year class, 55." "Maximum white perch, zero year class, 80."

The counted fish were tossed into a cooler of river water so they could later be returned to the Mattawoman.

This would go on for nearly half an hour. Then it would be done again. And then, the whole process would be repeated at four more sites in Mattawoman, a tidal fresh water river that enters the Potomac River about 20 miles south of Washington, D.C.

Offshore at each site, a boat hauled a trawl net along the bottom for five minutes. Then, biologists would inventory whatever turned up - always far fewer than found in seining. The results are interesting nonetheless: In one river, the 10-foot trawl net caught a cownose ray.

This scene, which took place one morning in early September, has been played out by biologists from the Maryland Department of Natural Resources in nearly a dozen rivers almost every summer day since 1989.

What they are fishing for - quite literally - is a way to relate the biological health of a river to the impacts from nearby land uses. That, in turn, would give them a way to measure whether land-based restoration efforts - including the tributa ry nutrient reduction strategies - are having quantifiable benefits on the resources they are aimed at protecting.

"It's not going to be enough to say we've reduced nutrients by 'xU percentage," said Ron Klauda, who heads the DNR's Chesapeake Bay Research and Monitoring Division. "We have to show some linkages among nutrient reductions, better water quality a nd living resources."

From the mass of data collected from fish counts, water-quality monitoring and habitat assessments, scientists hope to develop an "index of biological integrity" for estuarine rivers which can be used, in effect, to rate the river from a fish's point of view.

"Our real goal here is to try to find ways that the ecosystem can speak for itself," said Steve Jordan, a DNR biologist who helped design the survey in the late 1980s. "We're asking the fish to tell us how they're doing."

Similar indexes, based on fish and benthic insect populations, are widely used to evaluate the quality of fresh water streams. But establishing such an index in tidal rivers is far more complex. Tides constantly change water levels; river salinity varies both by proximity to the Bay and by seasonal changes in water flows; and different species of fish are constantly passing through.

"I've been told by a lot of people that for an estuarine system, we will never be able to devise this index, because there are too many variables influencing fish populations which are too difficult to measure in a program like this," said Marga ret McGinty, the DNR biologist who headed up the field work last summer. "You can't standardize for tides, for instance."

Mattawoman Creek, in some ways, illustrates the difficulty. By most measures of fish diversity and abundance, it ranked highest of all rivers surveyed. But it fared poorly when ranked for estuarine spawners, which include such salt water species as menhaden, silversides and blennies. But that most likely reflected the lack of salinity rather than poor habitat.

Likewise, catches of anadromous species - oceangoing fish that swim into fresh water areas to spawn - were low in the Wicomico River, a reflection of its high salt water level, which makes it unattractive to such fish.

And unlike freshwater systems where ecological health can be measured by a single visit, at least three day-long trips are needed to each estuarine river because of all the variables involved.

So people have some basis to be skeptical that a meaningful indicator can be developed for estuarine systems. But in the 1970s, people said the same thing about stream assessments, said Jordan, who now heads the Cooperative Oxford Laboratory, a r esearch facility jointly run by the DNR and the National Marine Fisheries Service.

"A lot of people argued that it wasn't worth taking benthic samples in streams because you could never figure anything out from them," Jordan said. "Now, everyone takes these rapid benthic bioassessments of streams."

DNR is not the only agency interested in the results. The work has also been supported by the National Oceanic and Atmospheric Administration's Coastal Zone Management program, and by the EPA. The EPA is interested in establishing such "biocriter a" to help assess major ecosystems across the nation.

Some of the data show clear patterns. Bottom trawl samples find greater species diversity when taken in rural areas than near urban centers or industrial locations. Trawls from Curtis Creek near Baltimore and the Severn River near Annapolis averaged less than one species per sample. Several times that number were caught, on average, in rural locations.

"The diversity in those trawls seems to be a real strong indicator," Jordan said.

What he and others believe will emerge from this work is not one, but a series of indicators that can be used to help assess health. They hope those will fit together in a way that will provide a measure of habitat health by examining "fish assemblages" - the makeup of the entire fish population in a given area.

To do that, biologists have sorted through the data to determine such things as the total number of species at a site, whether the total population is dominated by one or a few species; the abundance of migratory spawners considered sensitive spe cies; and the proportion of the population that is classified as algae-eating planktivores, fish-eating carnivores or bottom-dwelling benthivores.

The ideal, Jordan said, is to have a wide range of diversity both in the number of species, including the types of species that make up different parts of the food chain. A river dominated by planktivores, such as menhaden, bay anchovy or silversides, tends to have the greatest nutrient problems. Nutrients spur excess algae growth which provides plenty of food for the algae-eaters.

As cleanup progresses toward the Bay Program goal of reducing nutrients 40 percent by the turn of the century - the objective of the tributary strategies - there should be fewer planktivores and a greater proportion of carnivores, such as striped bass and bluefish. And fewer nutrients should mean better water quality - with more oxygen - at the bottom of the rivers, offering more habitat for benthivores, such as spot and croaker.

That, Jordan said, would reflect the trophic system - or food webs - found in coastal bays where there is less nutrient pollution. "The idea is to get more balance in the system," he said.

Surveys in the South and Severn rivers, by contrast, have shown that 80 percent to 90 percent of the population consists of planktivores. And while carnivores accounted for more than a quarter of the catch in the Mattawoman, they made up only 2 p ercent to 4 percent of the total in several other rivers. "That's not very balanced, if you ask me," Jordan said.

While scientists are still trying to sort through those variables to develop an index that can be widely applied, McGinty said the findings produced already offer valuable information for resource managers.

"As far as I'm concerned, the index can be used in large-scale management and in identifying what watersheds need attention," she said. One watershed that may need attention, for example, is the Mattawoman.

The Mattawoman today is in the best shape of any of the rivers being tested. But the Mattawoman watershed is projected to have 20 percent to 40 percent population growth by the year 2020. Given what the surveys are showing about the health of fish populations near urban areas, McGinty said that trend is "pretty disturbing."