Shaking mud from her hands, U.S. Geological Survey botanist Cairn Krafft makes her way toward experimental plots designed to examine the grazing impact of Canada geese on marsh vegetation. Credit: Erica Goldman.

American lotus plant droop in Anacostia mud at low tide. Volunteers worked hard to transplant the lotus to Kingman Marsh, burying each tuber by hand a foot beneath the mud. The lotus, with its signature yellow flowers, once flourished in the Anacostia’s historic­ marshes. Credit: Erica Goldman.

WEARING ARMY GREEN HIP WADERS, CAIRN KRAFFT CLOMPS ACROSS the manicured fairway grass of the Langston Golf Course. She moves through the first-cut rough, then the uncut rough, then steps off the course completely — where she sinks deep into the mud. She lunges forward, shifting her body weight to her left quadricep so she can extract her right leg. Thuuwwuucck. The mud closes noisily around the hole left by her foot. She lunges forward again, sinking in past her knee. Momentum is the key to not getting stuck.

With the practiced walk of someone who's moved through mudflats many times before, she surges steadily forward through the thick, gray ooze of Kingman Marsh, an expanse of freshwater tidal wetland in the center of Washington, D.C. She's not looking for golf balls.

It's an unlikely place for a freshwater marsh. And to the golfers playing a hole at the adjacent course, Krafft, with her windblown, curly hair and oversized waders, must seem an unlikely sight. Metro's orange line roars across a bridge nearby. Just beyond lie some of Northeast D.C.'s most troubled urban neighborhoods. But since 2000, efforts to restore historic marshland to the Anacostia River have transformed these waters that run beneath the overpass on Benning Road.

Krafft shifts forward through the mud, which sucks at her waders with each step. Through dark sunglasses over her prescription lenses, she sees the nearshore vegetation thin out, giving way to a span of exposed mud. Thin patches of green plants punctuate the barren expanse — low-standing spadderdock and arrow arum, with their wide, heart-shaped leaves. She makes her way carefully toward a fenced area, part of an experimental study she leads along with biologist Mikaila Milton of the National Park Service and other scientists from the U.S. Geological Survey.

Inside the fence, lush marsh vegetation grows some eight feet in the air. Krafft, a USGS botanist at the Patuxent Wildlife Research Center in Laurel, Maryland, picks out native cattails, wild rice, and burr marigolds.

Nearly all the Anacostia River once looked like this. Historical accounts describe extensive marshes, dense with wild rice, cattails, and reeds, lining a deep-water channel. The marshes were home to muskrat, beaver, and turtles. They were rich in bird life, with abundant kingfishers, herons, and Sora rails. Their waters teemed with shad, pike, perch, bass, and herring. Fishermen plucked giant sturgeon from the depths. (See Making Mud of the Ancient Anacostia)

Crouching down, Krafft heaves at a heavy log propped against the fencing. She levers it upwards with the weight of her body and casts it to one side, her bare hands now covered in mud. The eight-inch opening at the base must remain clear for fish, turtles, and other small animals to enter the area. Only Canada geese should be excluded. A fallen log or any other obstacle would introduce a confounding factor to the study.

Just outside the fenced plot, two PVC pipes mark the dimensions of an identical unfenced parcel. Here no plants grow, save a single gnawed-off shoot. Dimpled by puddles trapped by a receding tide, the unfenced area is barren — a flat expanse of mud.

The goal was to bring back some of the ecological power of Anacostia's historic wetlands. Create habitat for wildlife. Filter water. Give urban D.C. a glimpse of what a freshwater marsh can offer. That's why the U.S. Army Corps of Engineers, the D.C. government, and the National Park Service set out to restore 33 acres of wetlands in Kingman Lake, an area once dominated by prolific marshes. Building on earlier restoration work at Kenilworth Marsh upstream, the Army Corps in 2000 worked to elevate the planned marsh to a height suitable for the right plant species. They filled the mudflats with 179,400 cubic yards of dredged material removed from the Anacostia's Federal Navigation Channel. Then they planted. And planted. Almost 700,000 plants representing 7 different species. In total, this effort cost more than $6 million.

Nearly none of these plants now remain. The resident Canada goose population has eaten almost everything that's not fenced in. Krafft is not surprised. She's been watching the geese feast on plants in Kingman Marsh for almost ten years. As the workers began planting in 2000, geese followed on their heels. Soon the birds had devoured nearly 50 percent of the new marsh or between 300,000 and 400,000 young plants. The geese, over 600 at last count, would have consumed every last shoot had a contractor working for the Army Corps not rapidly erected fencing to protect new vegetation.

The Canada geese in question are a non-native subspecies imported from the Midwest. The hunting community intentionally introduced the geese to the area on two separate occasions, first in the early 1900s and again in the 1960s. The first introduction was to serve as a "live decoy," to lure the high-flyers closer to the ground, according to Steve McKindley-Ward from the Anacostia Watershed Society.

Hunters brought the introduced subspecies of Canada geese near extinction in subsequent years, both in the Chesapeake region and around the country. In the 1960s, perhaps to right past wrongs, hunters reintroduced this subspecies to the area from a small remaining flock in Minnesota. Due to genetic differences from native Canada geese (so-called "Carolina high-flyers"), the imported geese don't have a strong urge to migrate. No longer hunted extensively, the imported geese stayed and reproduced, becoming a burgeoning resident population. Each year, this resident subspecies remains in the D.C. area during the plant growing season, feeding on marsh vegetation, when the migratory birds have already flown north to breed.

By 2001 the National Park Service was confident that the replanted Kingman Marsh had firmly established itself — and it removed the fencing around the new plantings. The action turned out to be premature. The geese came back in large numbers and feasted on the young shoots, effectively eliminating much of the new planting from the year before.

Fences quickly went back up, built by various groups, and plants began to grow again inside these enclosed areas. Now the National Park Service cannot take down the fences, not until they come up with a plan for controlling the resident Canada goose population. If the fences come down too soon, the geese may graze years of restoration work right back to mudflat.

Lush grass on the Langston Golf Course lures Canada geese to Kingman Marsh, which borders greens and fairways (top). Leaving their telltale footprints in mud (above), the geese have consumed nearly all of the marsh plants not protected in fenced plots (right). Credit: Erica Goldman.

Krafft maneuvers across an open expanse of mudflat. It's slow going. Lunge. Stabilize. Extract back foot. Lunge again. In spots the mud is knee-deep.

She's striking out toward another set of fenced plots across the marsh. There she'll repeat her surveillance of the fence perimeter. This surveillance is part of the monitoring protocol for a large-scale goose herbivory (grazing) study at Kingman Marsh. Krafft is the lead biologist for the study, whose goal is to quantify the impacts of grazing by Canada geese on the vegetation in Kingman Marsh. Working with her on the surveillance is Peter Hill, who works with the Watershed Protection Division of the District Department of the Environment (DDOE). Krafft and Hill are checking each plot to make sure that the eight-inch gap at the bottom is free of trash or woody debris.

Few doubt that the geese are responsible for denuding Kingman Marsh of vegetation. But without hard numbers, collected in a rigorous way, the evidence remains anecdotal. It will take the weight of scientific data collected in this study to push through the public process needed to develop a Goose Management Plan.

Side-by-side fenced and unfenced plots (the two PVC poles at left mark the unfenced plot), help scientists quantify the grazing impact of Canada geese on Kingman Marsh. Credit: Erica Goldman.

The goose plan would be part of an Environmental Impact Statement (EIS) that the National Park Service (NPS) has been working on since 2007, explains Stephen Syphax, who heads the NPS local Resource Management Division. The study is part of the assessment looking at a whole suite of possible impacts on marsh vegetation — including invasive species and other grazers like turtles and fish. The paired site design, with fences designed to exclude geese specifically, are necessary to prove that geese are the culprits, explains Syphax.

Several measures for controlling geese have already been tried over the past four years, under the direction of a multi-agency Resident Goose Management Committee, a body composed of the D.C. government, U.S. Army Corps of Engineers, National Park Service, Anacostia Watershed Society, and USGS's Patuxent Wildlife Research Center. All along the tidal river corridor, from Bladensburg to Poplar Point, workers have coated eggs of resident geese with corn oil, a technique that destroys an egg by preventing gas exchange through the shell.

While this method has drastically reduced the number of geese being added to the Anacostia flock through reproduction, it can't reduce the existing population enough to relieve overgrazing pressure on Kingman Marsh. Urban geese, it turns out, live longer than their migratory counterparts — surviving up to 15 years. This is thanks to a lack of hunting or natural predators and easy access to ample food. And even if 95 percent of all eggs in a local population could be found and destroyed each year, the population would only see a 25 percent reduction over 10 years, according to calculations by the Canada Goose Committee of the Atlantic Flyway Council.

The most effective control option may be culling the geese. Culling, of course, is usually the technical term for killing, and any plan that includes killing geese is likely to upset some citizens and advocacy groups, says Syphax. The Humane Society, for example, faults the construction of wetlands next to a golf course as intentionally serving up a "salad bar" for the geese. Since Canada geese eat grass as a primary staple of their diet, they gravitate toward large expanses, such as playing fields, cemeteries, and golf courses, particularly if near a water source.

The draft EIS, anticipated in the summer of 2010, will present a game plan of alternatives for managing Canada geese and other threats. Based on public feedback that may prove contentious, the final record of decision will shape a course for Kingman Marsh. Without a control plan, however, the restoration effort in this section of Kingman Marsh could be doomed.

Making Mud of the Ancient Anacostia
EXTENSIVE MUDFLATS IN THE ANACOSTIA RIVER were pretty rare hundreds of years ago. Some tidal flats probably fringed the dense stands of vegetation, providing important habitat for certain shore birds and invertebrates. But based on historical maps and archival accounts, they were few and far between. The Anacostia River was quite deep, some 40 feet in the channel, according to historian John Wennersten's book, Anacostia: The Death and Life of an American River. It was navigable for ocean-going vessels all the way up to the port of Bladensburg in Maryland. Highway for communication and commerce, the river drove economic growth in the region beginning in the late 17th century — at a time when tobacco was king. [more]

In 1993, soon after wetland restoration projects began in the Anacostia River, Peter May started making first-hand observations in another marsh. Just upstream from Kingman Marsh was a wetland called Kenilworth Marsh, and it was the first Anacostia mudflat targeted for restoration.

In 1992 the National Park Service collaborated with the Baltimore District Army Corps of Engineers to hydraulically pump sediment out of the tidal Anacostia River and deposit it in the new marsh area, raising surface elevations to levels favorable for marsh vegetation. This massive effort in the spring of 1993 put 290,000 plants in the ground at Kenilworth, foreshadowing what would later be tried at Kingman Marsh.

May had just gone to work for the Watershed Protection Division of the D.C government, and his job included collecting data on birds, fish, and invertebrate life at Kenilworth Marsh in response to the new plantings. He was tasked with evaluating the success of the restoration effort by comparing it with another well-established reference marsh — Dueling Creek.

One observation really piqued his interest. The plantings seemed to be flourishing at Kenilworth. But one small section of the restored marsh reverted back to mudflat. Why? What were the drivers that favored the mudflat state in this spot? These were questions that would change his professional course. They would also lead him to develop an idea that geese and mudflats may both have a place in an urban marsh ecosystem.

A New Beginning for Kingman Island A WOODEN PEDESTRIAN BRIDGE LEADS VISITORS to a lookout over Kingman Lake, a platform large enough for school groups to assemble for educational programs or to dangle fishing rods into the Anacostia River below. The bridge and platform overlook the patches of restored marsh of Heritage Island and Kingman Lake that have managed to thrive in spite of hungry geese. [more]

May's curiosity about mudflats led him to apply to a master's program at the University of Maryland to study with ecological engineer Patrick Kangas. There he could test his ideas in an academic setting. While still working with the D.C. government, May began his graduate work in 1996. For nearly a decade, he worked to develop an idea that marshes, mudflats, or a mix of the two reflect different "alternate states" of the marsh ecosystem. In two different marshes, he explored how factors like goose grazing can play a pivotal role in "flipping the switch" between these different states.

When planning efforts began for restoration work at Kingman Marsh, the Anacostia again shaped May's course. He realized he would have a singular chance to study an ecological system as it underwent transformation. At Kingman, he could conduct formal before-and-after comparisons of the hoped-for transition from mudflat to marsh. He could monitor which plants would thrive in the restored marsh and what unexpected challenges might arise that might be unique to this urban environment. Excited by this new opportunity, May decided to expand his master's work into a Ph.D. dissertation.

From his master's work May saw that geese had not destroyed marsh restoration work at Kenilworth in the same way they later did at Kingman Marsh. And he thought he knew why. The geese were present at Kenilworth, but a combination of factors made the ecosystem transformation play out differently. At Kenilworth, dense stands of trees prevent the kind of easy access to grass and water that geese later found at Kingman, where the marsh sits next to a golf course. The sediment at the base of Kenilworth marsh had been elevated higher prior to planting — and that variable proved highly significant. The high elevation allowed more vegetation to establish and flourish, even in the face of some grazing pressure. Why did one area revert to mudflat? It was located at a lower elevation.

Hindsight is always 20-20. Before the success at Kenilworth and the struggles at Kingman, no one anticipated how sensitive the marsh plants would be to modest variations in elevation. Nor did anyone truly appreciate how much the location near a golf course would attract Canada geese. Only later would the U.S. Geological Survey's monitoring efforts, along with May's field research, with its carefully designed experimental modules surveyed over periods of years, provide some of the proof necessary to explain why the restoration at Kenilworth went so much more smoothly than the one at Kingman Marsh.

Intellectual curiosity led Peter May to consider how restoration ecology can help natural systems heal themselves. Here he pages through Ecological Engineering, a book by his Ph.D. advisor, Patrick Kangas. From his Baltimore-based office in the environmental design firm Biohabitats, May works to incorporate ideas by Kangas and others, including famous ecologist Howard T. Odum. Credit: Erica Goldman.

From the outset, the vision for Kenilworth and Kingman was much the same. Restore the ecological function of the marshes of the Anacostia's past. Repair a piece of this ruined river so people could understand the significance of what had come before — lush emergent marsh vegetation, willow, wild rice, native cattails. Mudflats were not part of this original vision, nor were hungry geese or invasive plant species.

But the reality of the stressors in this urban ecosystem has led May to develop an alternate vision for Kingman Marsh. May thinks that mudflats have their place there too. They provide a key ecological niche, says May, especially for certain shorebirds who rely on an open field of view to scan for predators. Marshes almost always have some mudflat habitat at the fringe, he explains. He suggests that given the set of challenges, a more realistic ideal for Kingman Marsh might be an "equilibrium state," a mix of mudflat and emergent marsh vegetation.

What might this equilibrium look like? Without a solid plan for managing Canada geese, it would be hard to justify the expense or effort of further plantings of the desired plant species in the low to mid-elevation levels of Kingman Marsh. After the initial plantings were devoured and fencing went up for a second time, many of the new plants that grew did not come from those planted the year before. This new growth came from seeds already present in the seed bank, or seeds that had been borne by wind and water, so-called volunteer species.

Volunteer species would be an inevitable part of a restored Kingman Marsh. To May, this fact speaks to the need to shift people's "palate" toward a new "design acceptance" for different plant species. This philosophy helps guide his current work as an environmental scientist for Biohabitats, an ecological engineering and environmental planning design firm in Baltimore. In areas subject to intense grazing pressure or other disturbances, according to May, a restoration effort should be designed to include species present in the seed bank, perhaps even invasive species. A key aspect of ecological engineering is to let the natural system work for you, May suggests, not to fight it.

"I'm totally for restoration work in the Anacostia, but it is very expensive," says May. "We need to be looking at ways to achieve restoration goals by getting more for less."

May believes that "getting more for less" will require understanding the "natural way" of the system. This disturbed, urban environment, to some extent, needs to self-organize and take care of itself, he says. That is, in the continued presence of variables like geese and invasive species.

"Hopefully we are not so arrogant to think that we will build the system and it will stay this way," May says. "At some time you have to let it go."

Cairn Krafft steps around the sign that marks the end of the trail on Kingman Island. Picking her way carefully down a steep incline, she bushwhacks through overgrown vegetation. A thick, thorny stem of a blackberry bush, as wide as her forearm, blocks her path. She steps down on it hard, swinging her leg high to avoid getting jabbed by the thorns.

She's heading down towards "Fringe A," a restoration area near the southern tip of Kingman Island, located along the bank of the Anacostia main stem.

Stepping off the edge of the sea wall, Krafft has more bushwhacking to do. As she lowers herself into the marsh, cattails and willows, some eight-feet-tall, swallow her whole. The marsh assails her with smells of rotten eggs and mint. Coarse blades of rice cutgrass tangle in her curly hair and sting her arms with dozens of tiny slices. She wishes she'd worn long sleeves. Mud splashes up her hip waders but she doesn't sink too deeply here. The dense vegetation binds the mud, anchoring the ooze.

Krafft makes her way with difficulty through the dense vegetation. From where she stands she can't even see the edge of Fringe A, some 60 meters distant. Deep inside the marsh, the buzz of insects almost drowns out the rumble of Metro's orange line.

Dwarfed by dense plants, Cairn Krafft plunges into Fringe A, a restored marsh near Kingman Island on the banks of the Anacostia River. Here plastic barricades sheltered plants from hungry geese during restoration work, and now some 60 plant species flourish, 98 percent of them native to the region. Credit: Erica Goldman.

Here the Army Corps drew from lessons learned in Kingman Marsh, where May's research and USGS monitoring efforts demonstrated that marsh elevation was a critical issue. When the Corps undertook the Fringe project in 2003, they attempted to build the elevations higher than the lowest areas of Kingman Marsh. To protect the plantings from the high-energy flow of the main stem Anacostia, they also drove interlocking plastic barriers (so-called sheet piling) into the riverbank, which had an ancillary benefit of protecting against grazing geese.

When the barricade is removed, the team hopes that Fringe A will remain safe from hungry geese. The marsh plants grow so dense and high that a goose shouldn't be able to land in the thick of it. More than 60 species of marsh plants now thrive in Fringe A. Diverse species grow at high densities. And, at the end of five years of monitoring, 98 percent of the marsh is native plants. To Krafft, current conditions at Fringe A represent a "holy grail" for tidal freshwater marsh restoration on the Anacostia.

But for nearby Kingman Marsh, the fate of this massive wetland restoration effort remains uncertain, hinging on the outcome of the ongoing Environmental Impact Statement. The fences that shelter the vegetation won't stay up forever. They're difficult and expensive to maintain, says Krafft. Keeping the fences intact takes tremendous volunteer time and labor. More centrally, she says, permanent fencing is not compatible with the vision of a restored marsh.

So without a plan for controlling the number of Canada geese at Kingman Marsh, the plants at low to mid-elevations may be lost once the fences come down. And when the geese devour all the palatable marsh plants, they'll move on — leaving an expanse of mudflat in their wake.