Identifying Thresholds

By Erica Goldman

Ranging from pristine to degraded, streams in western Maryland run the gamut. Bob Hilderbrand (top) turns over a rock to look for insect larvae in a pristine stream. Acid mine drainage (below) releases sulfuric acid and iron from the rock, turning the water burnt orange. The acidity of the water can also cause white toxic foam to bubble out of aluminum-rich rocks.

Just off a country road in the coal-mining region of western Maryland, about as far in the state from the Chesapeake Bay as one can get, Bob Hilderbrand scrambles down an embankment and plows through dense vegetation. He carefully avoids the broken bottle shards and litter underfoot until he comes to a spot where the overgrown foliage is just thin enough to see through.

"Look up there," he says, pointing to a small waterfall cascading into a fast-flowing stream, with a dilapidated structure perched atop the cliff. It takes a moment to register that this is not the usual mountain stream tableau. The water pouring down is a vibrant, burnt orange — the color of a rusted pipe.

"Acid mine drainage," Hilderbrand explains. It happens when water and oxygen interact with rock that has been drilled to expose pyrite — an iron sulfide that is not a typical feature of the outermost layer. The reaction releases sulfuric acid and dissolved iron, which are devastating to fish, plants and invertebrates that live in the stream. The switch to a degraded state can happen very suddenly, he says.

Further downhill from the orange stream, called Braddock Run, Hilderbrand pulls the car off the road and points at a rusty rivulet on the ground that flows a short distance and then abruptly turns bubbly white, like Alka Seltzer. "Here, the acidic water is flowing over rock rich with aluminum. The acidity dissolves the aluminum from the rocks and causes it to precipitate out of the water downstream," he says. "Aluminum is also extremely toxic to everything that lives in the stream."

"This stream system cannot be restored without constant chemical treatment. It is an example of an irreversible regime change," he says. "Furthermore, its water will eventually drain into the Potomac River, and from there, into the Bay."

Hilderbrand, a young theoretical ecologist at UMCES Appalachian Laboratory in Frostburg, Maryland, studies thresholds and works to develop quantitative methods to identify them in stream ecosystems. The streams in western Maryland that ultimately drain into the Chesapeake Bay are much simpler ecosystems than the estuary as a whole and present an opportunity to characterize what biological and physical factors can make a system resilient and what can push it to the brink.

Specifically, Hilderbrand is comparing the different ways in which land use has modified streams in order to pinpoint what characteristics make them more or less vulnerable to disturbance. Funded largely by the National Park Service, he is building upon the Maryland Biological Stream Survey (MBSS) dataset, which has sampled over 800 streams in the state, in order to assess the biological composition and condition of streams. Using a Geographic Information System (GIS) and spatial data on land use, agriculture, and percent cover for each, combined with samples of community composition, he is trying to predict how continued land use practices will affect these streams.

"We want to be able to predict how changes in the watershed relate to functional and taxonomic changes in the community structure of the streams. We suspect some streams will be more vulnerable or more resilient to land use based on their size, channel gradient (slope), and location," he says. Using a statistical approach, Hilderbrand will construct clusters of minimally disturbed streams based on structure and function. He hypothesizes that streams falling outside of these clusters will match to GIS data that can be pinpointed as thresholds. Ultimately, he plans to construct a model that predicts state shifts in streams given the form and magnitude of the landscape alteration and stream class.

Identifying thresholds in different ecosystems presents a major challenge, but also offers a potential opportunity for framing management decisions. There are a lot of general features common to major regime shifts but few specifics that can be extrapolated directly across diverse environments. Slowly, on a case-by-case basis, however, scientists are beginning to build a searchable database of thresholds to characterize what conditions cause shifts to take place and to compare across ecosystems. The database, established as a joint activity of the Santa Fe Institute and the Resilience Alliance, contains 64 examples from across the globe and across historical periods. They range from local, recent regime shifts such as the sudden eutrophication of Lake Washington in Puget Sound in the early part of the 20th century, to more distant changes such as climate-induced abrupt switches between vegetation and desert in the Sahara some 14,800 years ago, and again 5,500 years ago.

So far the responses to the project have been good, says database co-founder Brian Walker from the Commonwealth Scientific and Industrial Research Organization (CSIRO) in Canberra, Australia. But it is early in the process, he emphasizes. "There is a bias towards lakes and drylands and we want to increase the range of systems included.... It is important to increase the number and diversity of examples before people begin to analyze and draw conclusions from too small a sample size," he says.

In terms of content, the developing database aims to extend beyond the ecology of regime changes to provide information about the social causes and responses to these shifts and their potential for reversal. The idea that social and ecological systems are coupled is crucial — that is, that the behavior of a society can cause ecological shifts and that those shifts, in turn, can influence the way that a society behaves.

In the extreme, a society-driven ecological change can lead to the subsequent collapse of that society — such was the case described in the database for human society in Easter Island in the Pacific Ocean. Settled around 800 A.D., Easter Island was covered by a tropical forest, with six species of land birds and 37 species of breeding sea birds. Inhabitants cut down trees for firewood, for making gardens and building canoes and for moving the giant statues carved on the island. By 1600 A.D. the population had swelled to an estimated high of around 10,000 people, and all the trees, land birds and all but one of the sea birds had become extinct. Without trees, tropical rains washed the soil away and islanders could no longer build the canoes they needed to fish and hunt. The society resorted to cannibalism and the population of Easter Island collapsed — irreversibly.

In fact, of the 64 examples currently presented in the thresholds database, an impressive 24 have undergone an irreversible regime shift, 8 are unknowns, and 32 appear reversible — but at least 8 of these show signs of permanent changes or hysteresis (see main article).

Although Chesapeake Bay is far from the doomsday scenario experienced by this small island nation over a millennium ago, it is similar in that there is a tremendous feedback between humans and the environment in the way that they effect change in each other. The Bay has already crossed one threshold — when it underwent a shift from bottom-driven to water column processes in the 1970s. The question now is whether another shift lies ahead on the road to recovery.

This page was last modified November 03, 2004

h/V03N3/side3