Shortly after manure or fertilizer is spread on the land, a portion of the nitrogen that goes unused by crops takes a turn that could affect the course of nutrient trading in the watershed.

Instead of being flushed downhill and into a stream during a rainstorm, some of the excess slowly leaches through the soil until it reaches slow-moving groundwater.

The groundwater will also carry the nitrogen to a stream—and ultimately the Bay—but at a more leisurely pace: It often takes years, and sometimes decades, to complete the journey.

How much nitrogen takes the slow road to the Bay varies by a host of site-specific factors such as soil type. On average, U.S. Geological Survey studies in the Bay region estimate that 48 percent of the nitrogen reaching streams from nonpoint sources, such as farming or lawn fertilizer application, travels through groundwater.

Some of that may reach the stream quickly—in a matter of weeks or months. But the USGS estimates that on average, it takes a decade before half of the nitrogen that sinks into groundwater seeps into a stream. The rest takes even longer.

This groundwater bypass is no secret.

Policy makers have known for years that the full effect of some nutrient reductions efforts today—such as planting a cover crop or promoting conservation tillage—would not be realized for years because much of the nitrogen being affected is carried through groundwater.

In the meantime, older groundwater with higher nitrogen levels that predates various nutrient control actions continues to seep out into the stream.

One reason that nutrient reductions have been pushed for wastewater treatment plants is their discharges are delivered directly into streams—there is no groundwater delay. In December 2004, all of the states in the Bay watershed agreed to set strict new annual nitrogen discharge limits on wastewater treatment plants.

Now, as states create nutrient trading programs that allow wastewater dischargers to meet new permit limits by purchasing nutrient reduction “credits” from actions taken by farmers and others, the groundwater “lag time” raises scientific, legal and policy issues.

Specifically, can the states and the EPA—which are ultimately responsible for ensuring that discharge permits comply with the Clean Water Act—approve a trade if it would delay, or even allow temporary increases in the amount of nitrogen entering a waterbody because of groundwater lag time?

“It is not an easy issue for sure,” said Jon Capacasa, director of the water protection division for EPA Region III, which covers most of the Bay watershed. “It is not a new issue for the Bay region, but one we will closely review in the context of developing trading programs.”

The issue illustrates one of the uncertainties that have slowed the adoption of water quality trading programs nationwide. Touted as a cost-effective way to reduce nutrients, trading has not become widespread. Most trades have been so-called “point-to-point” trades between wastewater dischargers, where levels of pollution are easily monitored.

Trades become more problematic when they are between point sources such as wastewater treatment plants and nonpoint sources such as farms or other land-based sources of runoff. In part, that’s because it’s difficult to quantify exactly how much pollution is being controlled by a streamside forest buffer, a cover crop or other nonpoint source control technique.

All of the states in the region are looking at establishing trading programs. Both Pennsylvania and Virginia recently created programs that allow point-to-nonpoint source trades. To help determine the value of nutrient reductions from various nonpoint source control actions, both use the Bay Program’s Watershed Model.

The Watershed Model is the common “accounting system” used in the Bay cleanup effort. It was used to set the maximum annual nutrient load limits or “caps” for each major tributary to meet Chesapeake cleanup goals. States, in turn, used the model to develop river-specific tributary strategies which determine the level of wastewater treatment plant reductions—and the number of runoff control practices—required to meet annual caps.

But the model does not directly account for the groundwater lag time—it assumes nutrient reductions take place when a runoff control practice is implemented.

That’s where the groundwater “time lag” becomes an issue.

Suppose a new wastewater treatment plant is built and is required to offset all new nitrogen discharges—as is often the case. If it discharges 100 pounds a year, and offsets it with a 100-pound trade from a nonpoint source, it is possible that only 50 pounds of the nonpoint source control action is immediately realized in the surface water that same year. Because this is a new plant, and thus a new discharge, the result is a 50-pound increase in the nitrogen load into the water.

Capacasa said that the Clean Water Act does not allow a permitted facility to increase discharges that affect an “impaired” waterbody—one that does not meet its water quality standards—such as the Bay. “New sources are pretty clear,” he said. “You need to have a net reduction in pollutants to that waterbody. You cannot allow new and increased discharges to an impaired waterbody. That is clear to us.”

That means the full reduction achieved through a trade would need to be in the surface water, not the groundwater. Such a trade would need to involve nutrient reductions efforts that avoid groundwater lag times—perhaps hauling manure out of the watershed. Or, it could mean that a “trading ratio” is used which may, for example, require 2 pounds of nonpoint source credits to offset 1 pound of point source discharge to achieve greater reductions in the surface water.

The matter becomes more complex for existing facilities which are not adding new discharges, but merely reducing existing releases. Trading might slow the pace of reductions, but would not result in an increased discharge.

For example, suppose a wastewater treatment plant is required to reduce nitrogen discharges by 100 pounds. In a trading program it may opt to achieve that reduction by purchasing an equivalent amount of credits from a farmer who may be planting a nitrogen-absorbing cover crops.

If just 50 pounds are prevented from entering the stream—and the rest potentially delayed that year because it travels through groundwater—annual nutrient levels are going down, but has the permit limit been met?

Opinions vary as to exactly where in the hydrologic system a nutrient reduction must be to meet the obligation of a discharge permit that contains a maximum annual nitrogen discharge limit.

“The credit needs to be earned and available [from the nutrient control action] at the time the permit is approved, or at the time the trade takes effect,” Capacasa said. “We will be looking at the individual permitting action as well as the design of the overall trading program to ensure this objective is met.”

Many clean water advocates generally say it’s unfair for a trade to delay better Bay water quality just so people upstream can make the most cost-effective trade. If a permit sets a maximum annual load limit for a waterbody, they say, the full reduction has to be in the impaired waterbody.

“If you have annual cap loads limits in a permit, and if you meet them by an offset, there should be adequate assurance that the project which generated the offset will result in actual reductions,” said Matt Royer, an attorney with the Chesapeake Bay Foundation. But, he added, “determining how much of a local reduction is delayed by groundwater losses, for what period of time and how this impacts credit valuation are complex science and policy questions that need a closer look.”

Others say that farmers are penalized if their reduction efforts do not count equally with those of wastewater treatment plants.

“When calculating nutrient reduction credits, we apply various discount factors, utilizing the best science available,” said Andrew Zemba, who oversees the trading program for the Pennsylvania Department of Environmental Protection. “For us, it is a question of balancing what’s practical. If we discount reductions more than necessary, we run the risk of the trading program not taking off.

“We are trying to design something that folks are going to participate in, which we think will get us more reductions in the long run, than having a program out there where folks won’t participate. Participation is important because we feel the program is an important tool in our efforts to focus on building long-term sustainable programs that promote innovative ideas.”

The issue is likely more problematic for Pennsylvania; Virginia’s program initially emphasizes trades between point source dischargers, where groundwater is not an issue. Also, Virginia’s program calls for a 2-to-1 trading ratio between nonpoint and point sources; It requires two pounds of nitrogen reductions from a nonpoint source to offset one pound from a point source to compensate for uncertainties, including groundwater, associated with nonpoint source control programs.

Pennsylvania’s program, announced in late December, more aggressively pushes trades between point and nonpoint sources—officials touted it as the first statewide program in the nation to do so. Its program allows 1-to-1 trades between point and nonpoint sources—an aspect that has been criticized by many scientists and environmental groups. Ten percent of the value of each trade is subtracted to establish a credit “reserve”—essentially a bank of nonpoint source practices to provide a margin or error and serve as a backup for nonpoint source controls that fail. That adjustment makes the trade 1.1 to 1.

Facing some criticism, DEP officials in recent months have contended that—because farmers must take certain basic nutrient control actions, such as not applying manure near streams before they are eligible to trade—that the actual trading ratio is 1.6 to 1. Nonetheless, many say that is still insufficient to cover issues such as groundwater.

“With all that uncertainty around it, I would have to say that you would have to have a greater buffer than 1.6 to 1,” said Bill Gerlach, a former DEP attorney who is now works for the Waterkeeper Alliance, a New York-based advocacy group.

Capacasa said that EPA guidance allows for the use of various safeguards to address uncertainties, including those involving groundwater, when deciding if a permit meets its obligations. Trading ratios are one option but other exist, such as stream monitoring to verify actual pollutant loadings; the use of a credit reserve pool; using conservative assumptions about the effectiveness of best management practices; or other site-specific discount factors for practices.

“We will look at individual permits as well as the design of the entire program in determining whether the effort complies with the Clean Water Act—that is our objective,” he said.

But, Capacasa said, it is time-consuming to account for all the uncertainties in such trades within an individual permit. He said trading would be easier, and more accountable, if it were done on a watershed basis through a through a separate watershed permit.

Under a watershed permit, a total nitrogen limit is set for a group of wastewater treatment plants. The facilities can then trade among themselves or purchase credits from farmers or other nonpoint sources to meet the overall cap. On a larger scale, Capacasa said, it is easier to see whether adequate safeguards, such as pooled reserves, are in place than in an individual permit.

Also, in the case of nitrogen, Capacasa said watershed permits make sense because impairment is a concern downstream in the Bay, and not typically an issue with the local river.

“What we have here is a scale issue,” Capacasa said. “This all makes sense on a watershed basis. The load caps are done on a watershed basis. The reduction of pollutants needs to happen on a watershed basis. We are getting hung up on site-specific aspects with individual permits which do not mesh so well with that watershed approach.”

Exactly how the issue plays out may not be clear for a while. Point to nonpoint trades in Virginia’s program are not expected until sometime in the future, and the state has adopted watershed permits. In Pennsylvania, interest in trading is picking up.

But Capacasa said there is time for programs to adapt to resolve these concerns. The new nitrogen discharge limits being placed in permits will likely be phased in over a period of years, allowing some of the difficult issues to be worked out—as well as a potential shift to a watershed permit program.