Restoring degraded ecosystems is a "growth" industry throughout the USA. Many cost-share and subsidy programs administered by state and federal government agencies provide funding for the design and execution of environmental restoration projects. In recent years, the voters of various states have passed bond measures for restoration, despite relatively poor support for most increases in government spending. A central thrust of the Clinton Administration's plan for northwest forests is watershed restoration on National Forests to reverse past management practices. There are literally thousands of people working to restore salmon habitat, vernal pools, oak woodlands, old-growth forest, etc. Groups such as the Society for Ecological Restoration and Salmon and Steelhead Restoration Federation have formed to meet the need for these people to educate each other about what they do. Journals such as Restoration Ecology and Wetlands are filled with reports on projects from throughout the USA.
In site-specific restoration activities, we focus on obvious problems such as eroding roads, downcutting creeks in meadows, or replacing habitat lost through development projects. In these cases, projects can be reasonably well-defined although results are often difficult to measure in the short term. There are many reports of site-specific restoration in the literature and there are many skilled and experienced practitioners.
In the era of threatened and endangered salmon stocks and other issues requiring landscape-scale solutions we are confronted with a new restoration problem: how do we restore an entire landscape?
Take for example, the California Sierra Nevada. There are people who would argue that Sierra Nevada forests and streams require restoration. The problem is defined differently as "not enough big trees" or "too much build-up of fuels" or "undesirable tree species composition" or "degraded fisheries." Restoration of the entire Sierra Nevada, or even a single moderate-sized watershed is an immense task. If we assume that people agree that restoration is needed (by no means a given), we are actually presented with two challenges, the first being to decide what our restoration objectives are, and the second, to design and implement a process by which we begin restoration. We must begin any large-scale restoration effort with the understanding that we may not see positive results in our lifetimes. We also must accept that our efforts can be completely reversed by disasterous natural events or other things completely out of our control, such as global climate change.
Are we still ready to take on landscape-scale restoration?
There is a distinction between ecological functions and the values that people place on those functions. We must admit that restoration of landscapes for its own sake is not appealing to everyone. Some of the most active restorationists are the people who hunt and fish in wildlands. In evaluating our readiness to take on the restoration of entire watersheds we must examine our political will. Restoring old-growth forest begins with planting trees or protecting existing stands for centuries. Our uncertainty about political will applies not only to ourselves but to future generations.
Establishing priorities and setting restoration targets for the Sierra Nevada or any other landscape or ecosystem must be based on present and anticipated social, ecological and climatic conditions. What is the Sierran landscape (or salmon run, or riparian zone) that is appropriate to 1990s populations and cultures? How will that landscape be affected by short- and long-term climate changes?
The first concern we face is technical: what are the broad methodological outlines of landscape-level restoration planning?
My colleagues and I have recently completed an EPA-funded project in which we developed a system for defining where restoration should occur in a large watershed. The results will soon be published in a special issue of Restoration Ecology dealing with landscape-level restoration planning. The process we developed uses geomorphic and land use information in a decision analysis based on expert judgment to categorize portions of a stream as restorable, in need of protection, or too degraded to restore. Places that are considered restorable are mostly in a natural state but at risk from land uses. Using landscape measures such as vegetation cover and degree of fragmentation, places with completely natural conditions and little risk are recommended for protection. Completely developed places, perhaps with residual patches of natural riparian vegetation are considered too risky and expensive to restore. This approach differs radically from approaches to restoration and protection that have been used to date driven by regulatory processes such as the "404" permit. Under that project-specific approach, protection and restoration focusses on places that are being developed, where there is a regulatory nexus. In our approach, there would be mitigation for site-specific activities but watershed-level planning would shift first to securing protection where needed in the watershed.
The nature of watershed or landscape-scale planning is such that the use of extensive (i.e., low resolution) data is mandatory. One such approach is called "gap analysis." In this technique, remote sensing data and geographical information systems are used to locate and determine the status of places with high species diversity. Places with high diversity that are unprotected are termed "gaps." In our approach, we use stereoscopic aerial photographs and published maps to obtain the data necessary for establishing restoration priorities. We then do intensive field work in places which we determine, at the landscape scale, to be the best candidates for restoration. Critics of extensive approaches commonly argue that more detail is needed, however, given the limited funding available for such studies, that criticism is sufficient to completely stall a planning project. Low resolution data are accurate for decisionmaking at the scale that they are presented. Using low resolution data for site-specific decisions is just as flawed as using high resolution data for landscape-scale decisions.
From a technical viewpoint, if we are to proceed with landscape-level restoration we must begin by accepting the use of extensive data for such studies. We need not accept this on faith because statistical measures of accuracy are available. We should demand the same scientific quality that we demand from intensive, site-specific studies.
Conducting landscape-level planning is not an objective process, like counting fish in a stream. Judgment-based decision analysis, such as we use in our research, means that an expert or group of experts interprets data and decides on classification. In the fields of remote sensing and geographical information systems, this is referred to as "supervised" classification as opposed to "unsupervised" classification conducted by a computer or artificial intelligence. We are familiar with the use of these "expert systems" in situations such as medical diagnosis but it has only been the past decade or so that this has become popular in resource management and ecological research. A predictable reaction to studies of this kind has been skepticism and concern over the biases of the "experts."
Researchers or managers conducting landscape-level planning studies are not infallible. In the interests of proceeding on this urgent task, I would suggest that as long as the rules and assumptions of the experts are clear, the results should be understandable. In our research, for example, we decided that if a patch of riparian vegetation is adjacent to an upland that is undeveloped, that connectivity and ecological functions would be enhanced relative to a boundary that was urban or agricultural. That is clearly a judgment call which might be refuted at a specific site. At the landscape scale, it is a hypothesis with a track record of proof in other locations. In landscape-level studies we know that expert judgment will be used. We should demand that researchers and managers disclose the basis for their judgments.
We need to establish site-specific restoration priorities within the context of landscape-scale planning studies. Those studies should consider not only what needs restoration but also, what needs protection and what cannot be restored or protected. In the case of site-specific proposals within areas that are not restorable, we should insist upon mitigations but also expect implementation of the landscape-scale program. We need to accept the use of extensive data to do extensive studies and not expect the same level of detail that we would expect for a site-specific restoration project. Finally, we should acknowledge that expert judgment is an essential element in any planning study. We must demand explicit documentation of the values and assumptions that underly interpretation of landscape-level data. I would add that any landscape-scale plan that includes restoration and protection must have a scientifically-sound monitoring program that allows us to know how things are progressing. The plan should be modified as necessary based on findings of monitoring.
Now to the second question, once we have decided where, what is our target for restoration and where do people fit in?
There is no such thing as "pristine" wildlands. Much of the US was either intensively or extensively managed by indigenous people before Europeans arrived on the scene. What was out of reach by them has since been affected either directly or indirectly over the past 300 years. Even the most remote corners of the country have been affected by atmospheric changes over the past 100 years and will experience climatic changes in the future.
Although some restorationists and environmental activists consider people and their activities only as "stressors" on ecosystems and ecological processes, people and their uses must be incorporated into restoration programs and projects in a positive way. That is, we should design and restore ecosystems and landscapes that are "culturally-appropriate." I define this as the landscape that people and other organisms require for life. It is a landscape based on collaboration and ethical agreement between people and other species. People create cultures, cultures create land uses, and land uses create landscapes. We face restoration needs today because the land uses of the past 300 years have caused harm to other species. Do we have the insight and ethics to reverse that harm over the next 300 years?
Human population growth and consumerism are of course, large enough issues to thwart any attempts at restoring ecosystems to a more balanced condition. There are examples of large groups of people coexisting in relatively natural settings but overconsumption is by definition not sustainable. Even if we do not draw from our own local resource base, we create deficits elsewhere. Our habits of consumption are a serious ethical problem for all species, everywhere in the world.
Gary Snyder, in his book of essays, Practice of the Wild, describes a time in Europe when people jointly used communal lands in sustainable ways. There were traditional rules of behavior that kept individuals from overexploiting the resources. Indigenous people had similar rules which defined limits. We may condescendingly refer to these as "taboos" or superstitions but in fact, they were the laws of highly socialized cultures that rarely had controlling central governments or "police." These models of behavior have all but disappeared. After a recent visit to Europe, I returned with the opinion that conditions of the natural environment are far worse there than here. Indigenous communities still exist in the US but their cultures are fragmented, their populations are often too large for the reduced territory, and poverty is ubiquitous. In seeking a culturally-appropriate landscape we must look to our culture, realizing that other cultures can give us help. Most people of European, African or Asian heritage living in the US would not think of themselves as indigenous. On the contrary, many people may feel an uneasiness or "out-of-placeness."
I was born in California but raised on the east coast. As a child, I remember that in many neighborhoods the backyards were unbroken by fences. They were continuous ribbons of grassy field, sometimes undulating, with occasional trees or patches of landscape shrubs. Good habitat for play. My friends and I could travel throughout our town without ever hitting the street, only now and then we had to hop a fence. There were unspoken rules in these common backyards. For example, children did not damage plants, use swimming pools without invitation or break windows. In exchange, there were few if any (I recall none) aggressive dogs, few ornery older kids and few restrictions on our play. It was not until I returned to California that I truly came to appreciate the damage done by parceling of land into private, defended spaces. Defenses include fences, dogs, and even more.
Is there a relationship between our increasing defensiveness and the environmental problems we now confront?
Some of the bioregional thinkers would argue for disaggregating the landscape into watershed or bioregional parts in seeking cultural appropriateness. That is probably reasonable except that transportation, trade and political linkages violate natural boundaries. In Europe and elsewhere, linguistic boundaries can be a basis for evaluating cultural ecology. For example, the Catalan language is spoken in Spain and southern France, defining a historic Catalan territory that has various physical and ecological properties. This is not the case in the US. The foods and products of a region are not a sure descriptor for place in this era of modern agriculture and communications. There is a need for overarching cultural principles that recognize the diversity of peoples and places and that when applied to a specific landscape, transmute into appropriate behavior for that place.
In thinking about these principles, we immediately face the demands that we place on nature and confront the real challenge to retrofit cultural appropriateness into an exceedingly confused landscape. Land uses that now exist in "wildlands" are not limited to forestry, farming, recreation and mining but also include ex-urban development, reservoirs and various