Thomas Fisher hiked carefully down a slope and into the water at South Forge, a small, barely there waterway that includes a culvert running under a busy Caroline County road.

Fisher, an ecologist with the University of Maryland Center for Environmental Science, took a sample of the murky water and pondered its contents. He noted the depth, the flow conditions and the location in the stream. Then, he capped the clear plastic jar and stored it with more than a dozen other samples in a gray bin about the size of a municipal garbage can, for later analysis in a laboratory.

This routine, repeated multiple times a month over five years in four different spots in the Choptank River watershed, will help to answer a vexing question in Chesapeake Bay restoration work: Which farm runoff control measures work the best, and how can it be ensured that farmers are doing them correctly?

“My intention is, when we finish this project, to take the practical implications to the agencies,” Fisher said. “I’m hoping it will make a difference in how best management practices are used, and how they’re paid for.”

The Choptank River, where Fisher has worked for close to three decades, is an ideal place to test such farming practices. While many river systems in the Chesapeake Bay have stayed the same or improved, the Choptank River water quality has been getting worse every year. Almost half of the Choptank’s basin is in agricultural land, and much of that is in the upper part of the river where Fisher is working. The lower river, near Cambridge and Easton, has shown some improvement, perhaps because of flushing with the Bay.

Runoff control measures, or best management practices, are a conundrum for farmers and regulators alike. Because crop agriculture is largely exempt from the Clean Water Act, the government must rely on farmers to voluntarily reduce pollution. The U.S. Department of Agriculture and state agencies provide money to help farmers pay for conservation measures believed to reduce nitrogen and phosphorus running off the land. In most cases, the government covers more than 80 percent of the costs, with the farmers picking up the remainder so they have a stake in the outcome.

The computer model used to track Bay pollution trends counts these best management practices, or BMPs, when it assesses how the estuary is meeting its cleanup goals. But in recent years, some farmers and regulators have acknowledged it is an imperfect science. In some cases, models overestimate the effectiveness of certain practices; in others, the state doesn’t do enough monitoring to make sure the farmer has maintained the practice.

Farmers have long complained the model isn’t giving them enough credit for the practices they’ve put in place. In contrast, environmental advocates assert that enforcement is lax and that inspectors aren’t making sure farmers put in the practices for which they’re being paid. And while some practices, such as cover crops for corn, are well-established, farmers and advocates struggle to acquire funding for some of the newer practices, which are still proving themselves.

The questions about BMPs have intensified in recent years because while farmers are putting more of the practices in the ground, the pollution from agriculture is not dropping as much as expected. The best way to reduce the loads from agriculture is to convert the farm to another use. Even housing developments, scientists say, often produce less runoff than farms.

In some cases, it’s a matter of not maintaining practices, like stream-bank fences that get breached so livestock can get into waterways. In others, it’s not installing the practice properly in the first place. In still others, it’s not using the practices as intended: parking tractors instead of poultry litter in taxpayer-funded manure storage sheds.

The National Academy of Sciences issued a report in 2011 faulting the process, concluding that “the overall accounting of BMPs in the Bay watershed cannot be viewed as accurate.” The Bay Program’s Scientific and Technical Advisory Committee has also been saying for years that there needs to be better verification of how well BMPs work.

Fisher, whose work is being funded by the National Science Foundation, has selected three sites in the upper Choptank to test some management practices. For comparison, he’s also monitoring one control field with none of the recommended practices.

Among the BMPs he’s reviewing are:

  • tile drainage management structures, which allow farmers to control how quickly rainfall runs or seeps from their fields, thus limiting how much nitrogen and phosphorus is lost;
  • cover crops for soybeans and corn, which are planted after the cash crops are harvested to take up any excess nitrogen left in the soil so it isn’t washed off by winter or spring rains; and
  • gypsum “curtains,” in which the soil on the edge of farm fields is lined with a mineral that can filter much of the phosphorus out of runoff before it reaches drainage ditches and streams.

Fisher spent more than a year talking to farmers and landowners, trying to persuade them to participate in the study. It was not an easy sell, said Jim Lewis, a Caroline County extension agent who farms nearly 400 acres and is part of Fisher’s study.

Lewis said area farmers had been working well with environmental groups a decade ago. Notably, farmer William Collier has several water-control structures on his ditches to manage flow. Partnerships with environmental groups, including the Midshore Riverkeeper Conservancy, paid for three of them.

But after the Waterkeeper Alliance sued Berlin poultry farmer Alan Hudson and Perdue Farms in 2010 over an alleged poultry waste discharge on his farm, farmers became reluctant to let environmental groups anywhere near their ditches. Though Hudson and Perdue prevailed and some local environmentalists criticized the lawsuit, the case left some growers with lingering distrust and resentment.

“There are people we worked with before who say, ‘Stay off the farm. I don’t want you back,’” Lewis said. “People don’t forget that fast.”

With Lewis’ help, Fisher did persuade both farmers and landowners to participate in the study. The NSF has paid for the monitoring, equipment and staff, but Fisher has had to raise money for the BMPs he’s evaluating. He secured $80,000 from the National Fish and Wildlife Foundation, which largely paid for everything from a small $700 drainage structure to bioreactors costing several thousand dollars each. Choptank Riverkeeper Matt Pluta, who’s with the Midshore conservancy, has been helping to put in some of those bioreactors. Pluta said he’s excited about Fisher’s effort to assess the efficacy of various practices in the field.

“I don’t think it’s ever been done before,” Pluta said. “Right now, we’re just in this big mindset to get as much stuff in the ground as we can, when in reality, we need studies to tell us what are the best things to put in the ground that will give us the best outcome.”

If the study leads to better runoff controls, Pluta said, that could pay off in improved water quality in the Choptank. And, it also ought to help protect the significant investments in oyster restoration in the lower part of the river, where the U.S. Army Corps of Engineers and Maryland have spent millions of dollars building new reefs and stocking them with hatchery-reared bivalves.

One example of a change that Lewis and Fisher would like to see happen is best management practices that factor in real conditions. For example, farmers can get up to $90 an acre from the state to plant a cover crop after harvesting corn regardless of weather; but under some conditions cover crops perform much better than in others.

In contrast, some farmers are reluctant to plant a cover crop after soybeans. They say the deadlines they have to meet are too rigid or the government payments are not enough to offset the potential complications of trying to juggle the bean harvest and cover crop planting. Even though soybeans pull most of their nitrogen from the atmosphere rather than from fertilizer, they can still leave a lot of the nutrient in the field, where it may run off or seep into drainage ditches and streams via shallow groundwater

Fisher said it’s too early to draw conclusions about the practices, as he is only entering year three of a five-year study. But, he said, he’ll know something soon.

“We’re on the verge of getting enough data to tell some interesting things,” he said.