Smithsonian researchers develop models to assess wetland health

In a special issue of the journal Wetlands, Smithsonian scientists report a promising method of wetland assessment that will help environmental managers quickly take stock of wetlands across an entire watershed. Tools for this kind of rapid watershed-scale assessment—relying on a few easily measurable key factors—have been previously unavailable to managers.

In three papers, Dennis Whigham, Donald Weller and Thomas Jordan of the Smithsonian Environmental Research Center and their colleagues present the results of a large-scale study that combines field studies and remote-sensing data to assess the ecological functioning of wetlands in a landscape.

The researchers based their study on an approach previously developed for assessing individual sites, called the Hydrogeomorphic (HGM) approach, in which ecological conditions are inferred from readily observable indicators, such as plant species present and the degree of human disturbance.

“We took these methods for assessing wetland functions and expanded them to a whole-landscape scale, which is something that has not been effectively done before,” said Whigham, who coordinated the project. “These days, most land managers are not asking how to understand what is going on in an individual wetland, they want to manage resources at a much larger scale.”

Wetlands are important buffers for flood control, can absorb pollutants and excess nutrients and provide critical habitats for many plants and animals, including some threatened and endangered species.

For this study, the researchers focused on non-tidal wetlands in the Nanticoke River watershed of Maryland and Delaware. Draining into the Chesapeake Bay, the Nanticoke system is one of the most biologically important and wetland-rich watersheds in the mid-Atlantic region. Wetlands are found along streams (riverine wetlands) and in poorly drained uplands called “flats.”

During the first year of the project, the researchers visited wetlands of both types, taking field measurements and observations according to the HGM protocol at more than 100 sites. They used the data to formulate models to rate the condition of the sites, which ranged from nearly undisturbed to highly degraded. The sites were chosen according to a statistical procedure developed by the Environmental Protection Agency to ensure that they were representative of the entire landscape.

For a subset of the sites, the researchers took a closer look at one important ecological function of wetlands: the cycling of nutrients, particularly nitrogen. Many watersheds are overloaded in nutrients due to runoff from agricultural fields and other sources. The result is diminished water quality. But soils in healthy wetlands contain bacteria that remove excess nitrogen by a process called denitrification and can restore water quality.

“We found that you can predict denitrification potential from some fairly easy-to-measure properties of the soil, such as percent organic matter or pH,” said Jordan, who led this portion of the study.

As a final step, the researchers took the results of the field assessments and compared them with digital maps and remotely sensed data, such as satellite land cover images.

“The idea was to develop statistical models that would successfully predict what was observed in the field,” said Weller, whose lab performed the analysis. “Once you’ve developed the models, you then can assess additional wetlands without having to go out and sample them,” he added. While the models cannot predict the precise conditions at a given site, they can provide enough information to identify potentially degraded areas and help guide management priorities in a watershed.