Recent studies have shown that the U.S. population, as a whole, is becoming more overweight. Physicians suggest people eat better diets.

Increasingly, scientists say much of the same is true for the nation's rivers, lakes and estuaries. Waterways throughout the United States are being overfed with nutrients, which - like other foods - are crucial to life but a bad thing when overdone.

Instead of a stomach bulge, excess nutrients in waterways can show up as excess algae and bacteria, lower levels of oxygen in the water and a variety of other effects.

The EPA's solution is to put waterways on a diet.

Its prescription for doing that is the establishment of nutrient "criteria." A criterion is a number - or specific amount - of a particular substance that can be in a water body and still allow it to support its designated use, such as recreation, drinking supply, etc.

Once the EPA establishes criteria, all states are expected to set enforceable nutrient "standards" for their own waters. Ultimately, those standards will translate into discharge limits in permits for industry and wastewater treatment plants.

In addition, they may begin turning voluntary initiatives into regulatory programs. For example, states are required to review the health of their waterways every two years and identify any water segment that is "impaired" by any chemical. For each impaired segment, they have to write a plan, known as a "Total Maximum Daily Load," which describes actions that will keep that chemical from exceeding "safe" levels.

"Once you have a number and set a standard, you're putting the wheels in motion for TMDLs and permit limits," said Robert Cantilli, the EPA's national nutrient program coordinator. "There may be battles when we start to implement these criteria."

Because nutrients are the leading cause of pollution for most water bodies, new nutrient standards could result in an increased number of "impaired" waterways that require TMDLs. That could result in more regulation, not just for wastewater treatment plants, but also for municipal stormwater systems, animal feedlots and other nutrient sources.

For example, the EPA and U.S. Department of Agriculture have proposed requiring permits for large animal feedlots. But the strategy also requires that smaller feedlots on impaired waterways get permits as well. New nutrient criteria, therefore, could increase the number of feedlots that need permits nationwide.

"The EPA regulatory process is finally beginning to catch up and bring nutrients where we thought they should have been all along - under the regulatory umbrella, like toxic chemicals" said Mike Hirshfield, vice president of the Chesapeake Bay Foundation.

What's less clear, though, is a pivotal question - just how should the "right" number be selected? It's an issue many are watching with caution: The number selected for a particular water body will ultimately determine the level of nutrient reduction that is necessary - and possibly how much money will have to be spent to clean it up.

Some states, under certain circumstances, have chosen to regulate nutrients, in the past. Those efforts have mostly focused on phosphorus and were aimed at correcting a specific localized water quality problem, such as low dissolved oxygen or algae blooms.

The new criteria would, in most cases, control phosphorus and nitrogen in tandem. And instead of regulating nutrients to resolve a specific local problem in a water body, it seeks to control excess nutrients because they can contribute to a whole host of water quality problems, either locally or downstream.

But establishing nutrient criteria is a daunting task. Most water quality criteria are for toxic materials. Their numbers - or permissible concentrations - are based on lab work that can relate harmful effects on human or aquatic health with a specific level of chemical concentration; a single number that can be used nationwide.

Nutrients, by contrast, are essential to aquatic health. Determining when their amount exceeds the point of being essential and becomes a problem is, in itself, a problem. The "natural" amount of nutrients can vary dramatically in different places, and among different water bodies.

The relative importance of the phosphorus and nitrogen also vary: Phosphorus is typically more important in freshwater, while nitrogen is of more concern in saltwater. But unless nitrogen is controlled upstream - where it may not pose a water quality problem - excess amounts may reach saltwater, as is the case in the Chesapeake.

"Conceptually, I think nutrient criteria have some merit," said Mike Haire, of the Maryland Department of the Environment and a member of a regional team working with the EPA on criteria development. "It would be wonderful if it all worked out and everyone is happy with the end result. But I think it's going to be a difficult task, not that it's undoable."

To deal with this, the EPA last year issued a "National Strategy for the Development of Regional Nutrient Criteria." It envisions a range of numbers for at least 14 "eco-regions" around the country to reflect differences in such factors as geology and climate, which influence the effect of nutrients.

Within those regions, there would also be a range of numbers for different water body types - rivers, lakes, wetlands, and coastal waters and estuaries - which have different nutrient needs. The idea is to have a nutrient criteria tailored for specific water bodies rather than a single number for the whole country.

To do that, the strategy calls for finding relatively undisturbed "reference conditions" in each eco-region to serve as the basis for determining target ranges that the criteria should fall within. Based on that research, along with input from scientists and information from any historical data that may be available, states would identify what they consider to be the proper number for individual water bodies and set standards by the end of 2003. If they don't, the EPA says it will do it for them.

Some water bodies may have different numbers for different places. And, complicating the picture even more, some areas may have more than one "right number." For example, salinity levels in parts of the Bay fluctuate widely between spring, when freshwater flows from rivers drives saltwater down the Chesapeake, and fall, when dry conditions allow salinity levels to increase. As a result, those areas may need different criteria for different seasons.

But some details of the strategy are the subject of debate. One of the main problems is finding a "reference" area in a place like the Chesapeake Bay, which has been impacted by human activity for centuries and is daily bombarded with additional nutrients from air pollution.

"The idea of a reference condition breaks down here," said Rich Batiuk, associate director for science with the EPA's Bay Program Office. "We would have a difficult time, if not an impossible time - at least in the Chesapeake - of finding an undisturbed reference condition."

Establishing a near-pristine condition in the Bay, agreed CBF's Hirshfield, would essentially mean converting the entire watershed to forest and ending air pollution. "I don't think anyone is proposing that as a realistic endpoint, even us," Hirshfield said. Tying criteria to an abstract number, he said, could result in disputes that only delay the program.

Instead, Batiuk envisions criteria based on the restoration needs of key resources, such as underwater grasses, which are harmed by nutrient enrichment. Excessive amounts of nutrients spur algae blooms that block sunlight to the plants, and can fuel the growth of sun-blocking parasites directly on their leaves.

Setting criteria based on grasses would also show the public - and those who face potential regulation - tangible results as the resource recovers. "What we've said is instead of focusing on nitrogen and phosphorus, let's focus on those things that have an impact and which the public and the stakeholders can make a connection to," Batiuk said.

Another factor nutrient criteria could be tied to, Batiuk said, is the amount of dissolved oxygen - crucial for habitat - in the water. Oxygen is depleted as algae die and are decomposed by bacteria.

But one EPA official questioned whether that approach would achieve the clean water goal. "That's pragmatic," he said, "but that is trying to take away the worst conditions incrementally, and trying to improve the situation incrementally. We're trying to find the situation in the Bay that represents high quality, and strive for that in terms of developing our criteria."

Nutrient reductions suitable for one resource may not be adequate for another, he noted. "Whenever you're developing a criteria, you're trying to come up with something that is protective of water quality - of the best use of the water," he said. "We're trying to elevate the situation, not maintain a bad situation."

George Gibson, a scientist with the EPA's Office of Water, is presently studying the potential of finding a reference site in a section of the the Chester River, which appears to have few human impacts, though he said conditions would need to be adjusted for atmospheric deposition.

But Gibson said criteria based on grasses as a major factor may have merit as long as the connection to nutrients was clearly made - and results from the reductions could be shown.

"I don't see a conflict there," he said. "That's a lot better measurement than an esoteric number." He noted that other factors, such as sediment, also affect grasses and that it would not be acceptable to adopt a underwater grass-based criteria that dealt with sediment and not nutrients. "You couldn't ignore nitrogen and phosphorus as part of the response," he said.

Others question whether nutrients should be the focus of criteria - and the immense amount of time and expense it takes to develop them - unless there are clear indications of problems.

For example, said Jonathan Phinney, a scientist with the Center for Marine Conservation, nutrient levels may be elevated coming into San Francisco Bay but they create no problems because that bay has been overrun with foreign Asiatic clams which effectively filter algae out of the water before it can cause oxygen or other problems. "There you have a nutrient concentration issue, but there is no problem," he said.

Instead of setting criteria in areas with no visible problems, Phinney said the EPA would be better off focusing on areas with clearly identified nutrient-enrichment problems, such as waters with low oxygen conditions. "I think it's crazy to look at a number if there is not a problem," he said.

Another concern is the scale at which nutrient issues should be addressed, especially when dealing with such questions as how far upstream one should go to address downstream problems. It's not realistic, some say, to set a nutrient standard in Minnesota to reduce low-oxygen conditions in the Gulf of Mexico.

Such issues may also raise equity issues. Because of differences between regions - in some cases, even between river segments - what needs to be done in one area may be different from what is needed in another. But, Batiuk said, it's logical that the "prescription" for cleaning one river system would not be the same as another, for a host of physical and biological reasons. "All ecological systems are not created equal," he said. "If we're going to do it right in an ecological system, we've got to move away from equity."

Whatever the outcome, the process almost certainly guarantees that the nutrient diet of waterways in the future will be determined increasingly by regulations, rather than voluntary actions, which have guided many nutrient control efforts in the past.

While that raises concerns for some, Batiuk said regional officials have time to minimize the impact - such as having to develop more TMDLs - by vigorously stepping up the pace of voluntary nutrient control efforts before the new criteria become effective. "These," he said, "are things people should be doing anyway."