Skip to main content

More experiments needed to advance environmental restoration

October 1, 2008 By Madeline Fisher

Across Wisconsin and the nation, scientists, land managers and conservationists are trying to restore large tracts of wetlands degraded by pollution, sedimentation and invasive plants to a more natural state. Although these efforts are laudable, Joy Zedler sees in most of them a missed opportunity.

“Very few restoration ecologists are taking advantage of large restoration sites by conducting large-scale experiments,” says Zedler, who is Aldo Leopold Chair of Restoration Ecology at UW–Madison. “Most people wouldn’t buy a new shirt without trying on several different kinds to see which fits best and looks right. It’s similar with restoration; we want to find the best fit between the methods we use and the outcomes we want.”

To achieve this, Zedler promotes an approach called “adaptive restoration,” which sets up restoration attempts as large experiments, compares several methods at once, and then analyzes them to see which has been most effective at meeting project goals. She herself has practiced adaptive restoration for well over 20 years, but she wondered how many of her fellow restoration ecologists were following suit. So she had four graduate students — Kelly Wagner, Sally Gallagher, Matthew Hayes and Beth Lawrence — dig into the recent wetland restoration literature as a seminar project.

Their analysis, which appears in the current issue of Restoration Ecology, revealed that of the 311 published wetland restoration papers they examined, fewer than half used any kind of experimentation whatsoever. What’s more, just 13 percent employed large experiments of one hectare or more.

Instead, researchers often used just one restoration method throughout a site, making it difficult to “learn while restoring,” Zedler says. And while small experiments do give scientists some idea of how to restore a bigger area, testing still needs to go on at the larger scale because “you sometimes get a different answer,” she adds.

She’s speaking from experience. About 20 years ago, Zedler was involved in trying to restore a certain plant species needed by an endangered California bird. Starting out with tests in carefully controlled 2-by-2-meter plots, she and her colleagues were encouraged to find that additions of nitrogen boosted the plant’s growth. “But when we went to 20-by-20-meter plots where we couldn’t control everything, it turned out that the nutrients allowed a different species of plant to come in and displace the one we wanted,” she says.

At the same time, Zedler also knows from her own work that large experiments not only require more resources to carry out, but arranging the necessary funds and people can take months or years.

“You have to be very persistent and willing to stick with it for a long time,” she says. “But these experiments are very, very useful because you’re testing, with replication, how an entire system responds to the restoration at a realistic scale.”