Reflections on the Pond

By Sarah DeWeerdt

Illustration © Tomek Olbinski/SIS

Ask a child to draw a wetland and you’ll likely get a pond. Ask a planner to restore a wetland and you’ll likely get a pond. The pond is the nearly universal icon for wetlands. They are easy to build, meet regulatory guidelines, and they are clearly wet .

But to Joy Zedler, the pond is the ecological equivalent of a fast-food chain or a cookie-cutter suburb, an emblem of the homogenization of the contemporary landscape. It’s also a symptom of how our efforts to restore long-lost wetlands have gone wrong.

Zedler, a botanist by training who holds the Aldo Leopold Chair in Restoration Ecology at the University of Wisconsin, literally wrote the book on wetlands restoration. She is the editor of the Handbook for Restoring Tidal Wetlands (1), and she was a member of the National Research Council (NRC) panel that produced an influential 1992 report on wetland mitigation projects around the U.S. (2)

Despite these established credentials, Zedler does not hesitate to challenge the status quo. She argues that the recent wetland restoration and mitigation boom is having unintended consequences: the conversion of a landscape of diverse wetland types into a monoculture of ponds, and the creation of “generic wetlands,” where weedy, generalist species thrive and invasives crowd out native biodiversity. In short, we’re often creating the wrong kinds of wetlands, in the wrong places, that support the wrong kinds of species.

Working Hypotheses

Our mistake is thinking that we can simply “build it and they will come.” But that mentality runs counter to experience. The question of what makes an ecologically functioning wetland is a lot less straightforward than it seems. Even sites that look like healthy wetlands often lack some subtler quality. To illustrate, Zedler tells the story of how a well intentioned effort to restore a California salt marsh failed to provide habitat for its intended resident, the endangered light-footed clapper rail (Rallus longirostris levipes).

As mitigation for a highway project that had destroyed a parcel of the rail’s coastal marsh habitat, the California Department of Transportation (Caltrans) agreed in the 1980s to restore 12 hectares of coastal wetland in the Sweetwater National Wildlife Refuge, about eight kilometers south of downtown San Diego. Restorationists transplanted cordgrass (Spartina foliosa) from an existing marsh nearby to create the stands of vegetation that the rail favors for nesting.

But it didn’t work. After ten years of study and experimentation, Zedler—then at San Diego State University—and her colleagues concluded that the new marsh would never provide adequate nesting habitat for the rail. “The basic problem was that the plants wouldn’t grow tall enough,” Zedler says. The rail won’t nest unless the cordgrass is at least 90 centimeters high—tall enough to be woven into a canopy over the nest that provides protection from flying predators and acts as an anchor during high tide. But the site’s sandy dredge spoils couldn’t hold enough nitrogen to enable the plants to grow to that height.

When the research team added nitrogen fertilizer to experimental plots, the cordgrass grew tall enough, but in larger trials, the boost of nutrients caused another marsh plant, annual pickleweed (Salicornia bigelovii), to out-compete the cordgrass. And after the researchers stopped adding fertilizer, the nutrients washed out of the porous soil.

In 1998, the U.S. Fish and Wildlife Service ruled that Caltrans had failed to comply with its mitigation agreement, and Caltrans was asked to restore an additional area of filled wetland. Today, the rail is limited to small remnants of disturbed salt marsh habitat, and it’s no surprise that the bird remains on the endangered species list.

Meanwhile, Zedler had been working on the 1992 NRC report, in which she and her fellow panelists analyzed evaluations of 25 restoration projects around the country. They found that all of the projects fell short of their goals in one way or another. This convinced her that the shortcomings of wetland restoration were widespread, not just confined to her southern California research sites.

So what was going on? Zedler began teasing apart the problems behind our current efforts and realized that we need to start thinking about restoration projects not only as engineering feats but also as scientific experiments. And therein lies her simple but powerful message.

Zedler and her colleagues put this idea into practice with a hybrid restoration/research project in the Tijuana Estuary, near San Diego. The project was designed to simultaneously vegetate the site and test hypotheses about the best restoration techniques to use.

An ongoing debate within the field of wetland restoration concerns whether to invest time, energy, and money in planting vegetation at restoration sites. Some argue that rest-orationists must painstakingly insert seedlings (usually cordgrass or pickleweed in southern California) one by one into the sediment, like pointillist gardeners, to render a “natural” wetland. The other camp says that “Nature knows best”; we need simply restore the hydrology of a site and a wetland will develop of its own accord.

To Zedler, the question wasn’t whether to plant, but what to plant—and when. So instead of simply leaving the site bare or planting pickleweed over the entire site—the traditional approach—the group marked off 87 plots to test which marsh plant species could establish best, alone and in various combinations.

It turned out that pickleweed, which restorationists often spend so much time planting, was the only one of eight common marsh plants that could establish without help. The plots that were planted with a diverse array of native species looked and functioned the most like natural marsh, with more biomass, taller plants, and more complex canopies, which presumably made them better habitat for wildlife.

Zedler and her colleagues later used what they learned from this restoration experiment to design the nearby eight-hectare Friendship Marsh, planting seedlings of five marsh plant species in mixed clusters to provide a variety of habitat types. They’re continuing the experimental approach at this new site, evaluating the growth of this species-rich vegetation in relation to the presence of tidal creeks, the use of soil amendments, and spacing of transplants.

A Critical Mass of Wetlands

Still, Zedler cautions that just developing techniques for the restoration site itself isn’t enough. We need to start working at the landscape scale. She points to Bill Crumpton and his colleagues at Iowa State University who are doing just that. They’re linking restoration and evaluation at the watershed scale. They’re working with the Iowa Department of Agriculture and Land Stewardship on a Conservation Reserve Enhancement Program (CREP) that will remove excess nutrients from agricultural runoff using restored wetlands.

Nutrient-rich runoff is a particular problem in Iowa and the other Corn Belt states of Illinois and Indiana because millions of hectares of farmland in the region are drained by networks of buried clay tiles and perforated pipes. And the rich, wet soil holds its nitrogen as nitrate, the most soluble form. “You can’t keep nitrate if the water goes. And you can’t grow row crops if you don’t drain the water,” Crumpton says.

The consequences of this Catch-22 are enormous. Agriculture contributes about 65 percent of the nutrients that wash into the Mississippi River. In turn, when these excess nutrients reach the Gulf of Mexico, they cause eutrophication and a “Dead Zone” of oxygen-poor water up to 20,000 square kilometers in area.

But properly designed wetlands can become giant ecological sponges that can absorb up to 80 percent of the nitrate that flows into them. The current CREP, which is open to farmers in 37 counties in the north-central portion of Iowa, builds on a decade’s worth of research by Crumpton and his collaborators about just how and why wetlands accomplish this nitrogen removal. One key to the process is creating the right type of wetland: shallow, vegetation-rich swales much like those that characterize the native landscape. In these wetlands, anaerobic bacteria break down dead plant matter in a process called denitrification and release nitrogen gas (N2) to the atmosphere.

“In order for these wetlands to be effective, they have to be in the right place in the landscape,” Crumpton says—that is, downstream of a substantial area of tile-drained farmland. By contrast, previous wetland restoration projects in the region have focused on restoringprairie potholes scattered within the agricultural landscape. Those wetlands provide valuable wildlife habitat, but they don’t contribute much to nitrogen removal.

For this reason, Crumpton deliberately designed his project differently than conventional restoration efforts. Land enrolled in the CREP must be downstream of tile-drained agricultural land at least 200 hectares in area. Other enrollment criteria, such as the requirement that the wetland be 0.5-2 percent of the watershed area, are designed to facilitate evaluation of the CREP. Success will be measured not by the number of hectares enrolled, as in many restoration programs, but by the amount of nitrogen removed by the wetlands.

If the wetlands or the watersheds are too small, the measurements of nitrates flowing in and out will be too “noisy” and unreliable. And if the wetlands are too large, they will remove nitrogen too thoroughly. So, one of the things Crumpton’s team is trying to determine is where the point of “efficiency” or balance is—removing a good amount of nitrogen but not taking too much land out of production.

Eventually, Crumpton hopes that using wetlands to remove excess nutrients from agricultural runoff will catch on throughout the Corn Belt. And his big thinking doesn’t stop there. Many of the older drainage tiles and pipes in the region will have to be replaced in the next few decades, and the research group’s next task is to figure out how to integrate wetlands into agricultural drainage systems themselves, and how to design tile drainage systems to help the wetlands work better.

Picking and Choosing

Clearly, we can’t restore all 47 million hectares of lost wetlands in the U.S.—let alone the many more hectares that have been lost around the world. The 1992 report of the NRC panel that Zedler participated in recommended that ten percent of the wetland area lost between 1780 and 1980 needs to be restored to improve water quality and deter invasives over the landscape as a whole. Zedler admits that the panel was shooting from the hip when they chose that ten percent figure. But the real question is Which ten percent?

A few pioneers are already starting to think along these lines. Rather than starting with a restoration site and asking what to do, they’re starting with a landscape and asking what to restore.

Tom Bernthal and his colleagues in the Wisconsin Department of Natural Resources (DNR) are working on an answer. Using GIS, they looked at potential restoration sites in the Milwaukee River Basin and projected habitat quality before and after restoration. They then created a color-coded map that highlights how much each restoration project would add to the existing matrix of wetlands, streams, forests, and fields in the 2,000-square-kilometer basin. The restoration “hotspots”—areas that would produce the biggest gains in wildlife habitat if wetlands were restored—pop out in deep purple.

Bernthal, GIS expert Kate Barrett, and their colleagues managed to put the map together without undertaking any big new studies. Instead, they mined existing data sources, such as an inventory of current wetlands, a statewide habitat survey (including both uplands and wetlands), and local land use maps. There was no map of where wetlands used to be, so they used data from soil surveys, assuming that areas with hydric soils represented former—and thus potentially restorable—wetlands.

Although the map still needs ground truthing (i.e., Are the sites that the map identifies as high priority actually restorable?), it has the potential to be a powerful local planning tool. For example, the map shows clusters of several closely spaced hotspots where restoration could be concentrated.

Meanwhile, Bernthal’s DNR colleagues Joanne Kline and Marsha Burzynski are taking charge of the next phase of the project. Using a similar GIS-based process, they’re creating maps of where restoration would improve flood control and water quality the most, which will complement the habitat map that Bernthal has spearheaded. The group expects to find some areas where the hotspots for habitat, water quality, and flood control all overlap. “These overlapping areas could be the highest priority for restoration,” Burzynski says.

A Few Well Placed Incentives

The landscape-scale focus and experimental approach that Zedler advocates have great intuitive appeal. But are these ideas practical?

Zedler approaches this question with a characteristic calm, matter-of-fact optimism. Although the philosophical shift may be huge, she says that implementing the approach on the ground may be just a matter of a few well placed incentives and tweaks to existing programs.

For example, mitigation bankers could be given reduced performance bond requirements and longer evaluation times on projects that create native wetland types, which are often riskier than merely creating ponds. Zedler tells the story of a wetland mitigation banker in Oregon whose work on a native marsh has been frustrated by several years of drought. Once the drought breaks, his plants will probably do fine. But in the meantime, he’s having trouble getting his site certified as a wetland so he can sell mitigation credits. What if he received extra mitigation credits for trying to create native wetland types? What if extra mitigation credits were offered to projects designed as experiments, like Zedler’s Tijuana Estuary site? Granting agencies could play their part by giving funding priority to experimentally designed restoration projects, as the National Science Foundation has done by supporting Zedler’s research in southern California. “The critical step is designing the site to accommodate multiple trials. The benefits of learning which trials work (and why) greatly outweigh any added cost,” Zedler emphasizes.

It all makes sense, I’m convinced. But still, we have to shake the human preference for ponds. A recent survey of participants in the U.S. Fish and Wildlife Service Partners for Wildlife program, which funds farmers and other landowners to undertake small wetland restoration projects, found that landowners were only satisfied if the project produced open water on their property.


Wetland Joy Zedler has used data from a study by Robert Costanza and his colleagues (3) to calculate that although wetlands cover only about 1.5 percent of the planet’s surface, they contribute nearly 40 percent of the global renewable ecosystem services, worth US$33 trillion per year. Wetlands absorb rainfall and snowmelt after big storms, damping the effects of floods. They filter out excess nutrients from agricultural and urban runoff, preventing eutrophication of downstream waters. They trap loads of sediment that might otherwise degrade vital waterways. Tropical rainforest is often compared to the Earth’s lungs; if that’s so, then wetlands are the planet’s kidneys. Just as rainforests make clean air, wetlands make clean water.

Yet, these vital ecosystems were regarded as unproductive areas to be diked, drained, and filled. Between 1780 and 1980, the lower 48 United States lost an average of 53 percent of their wetland area, and some areas were harder hit than others. In California, 91 percent of the state’s wetland area is gone. In the upper Midwest, the loss amounts to over ten million hectares. Internationally, the timeline is different but the story is much the same: according to Wetlands International, about 50 percent of the world’s wetland area has been lost since 1900.

In recent years, growing appreciation of the ecosystem services provided by wetlands and the biodiversity they harbor has fueled efforts to halt and even reverse this loss. Today, the federal Clean Water Act in the U.S. mandates that developers and government agencies create nearly one hectare of new wetland as mitigation for every half-hectare destroyed in the course of subdivision, shopping mall, and highway construction. Meanwhile, a host of voluntary programs encourage farmers and other landowners to restore former wetland areas on their property. And private nonprofit groups, ranging from small, local land trusts to international organizations such as The Nature Conservancy, are also working to bring wetlands back. Wetland restoration efforts are also increasingly prominent in countries around the world, from the marshes of southern Iraq to the Doñana National Park in Spain.

Literature Cited

1. Zedler, J.B. ed. 2001. Handbook for Restoring Tidal Wetlands. Marine Science Series. CRC Press LLC, Boca Raton, FL.

2. Restoration of Aquatic Ecosystems: Science, Technology, and Public Policy. 1992. Commission on Geosciences, Environment and Resources.

3. Costanza, R. et al. 1997. The value of the world’s ecosystem services and natural capital. Nature 387:253-260.

About the Author

Sarah DeWeerdt is a freelance science writer based in Seattle, Washington.

Further Reading

Maurer, D.A. et al. 2003. The replacement of wetland vegetation by reed canarygrass ( Phalaris arundinacea ). Ecological Restoration 21(2):116-119.

Zedler, J.B. 2000. Progress in wetland restoration ecology. Trends in Ecology and Evolution 15(10):402-407.

Zedler, J.B. 2003. Wetlands at your service: reducing impacts of agriculture at the watershed scale. Frontiers in Ecology and the Environment 1(2):65-72.

Zedler, J.B, J.C. Callaway, and G. Sullivan. 2001. Declining biodiversity: Why species matter and how their functions might be restored in Californian tidal marshes. BioScience 51(12):1005-1017.