Ten Adaptive Management Areas (AMAs) were designated in the Northwest Forest Plan (USDA Forest Service and USDI Bureau of Land Management 1994b--hereafter "NWFP 1994") for the management of federal forest lands within the range of the northern spotted owl (Strix occidentalis caurina) in the Pacific Northwest (i.e. the western portions of Washington, Oregon, and northern California). This Plan is the key strategy designed to meet several federal mandates (Including the Forest and Rangeland Renewable Resources Planning Act, National Forest Management Act, Federal Land Policy and Management Act, Endangered Species Act, and the National Environmental Policy Act [NWFP 1994 ROD-5]) for maintaining native species in healthy ecosystems, particularly species associated with late-successional and old-growth forests and species dependent on streams and rivers. The NWFP adopted the concepts of ecosystem management and adaptive management to accomplish these objectives.
 

Adaptive Management Areas comprise 6% of the federal lands in the NWFP and are intended to "develop and test new management approaches to integrate and achieve ecological and economic health, and other social objectives" (NWFP 1994, page A-4). The primary objectives of the AMAs are "development, demonstration, implementation, and evaluation of monitoring programs and innovative management practices that integrate economic and ecological values" and "the provision of flexible experimentation with policies and management" (NWFP 1994, D-3 and D-4). Many people interpret the latter to mean that a primary mission for AMAs should be to test, evaluate, and refine the standards and guidelines of the NWFP itself, but recent guidance from the Regional Ecosystem Office (August 1998 "AMA Work Group Paper on Standards and Guidelines and the Adaptive Management Area (AMA) System"--the Regional Ecosystem Office is an interagency oversight group on NWFP implementation) suggests that attempts at innovation will require substantial preparation and precaution before it is allowed. AMAs were intended to be focal areas of innovation; adaptive management techniques developed in the AMAs would then be applied on all federal lands under the NWFP.
 

The principal objective of adaptive management is to institutionalize learning so that organizations can adjust their management to meet changing goals, conditions, and knowledge. Adaptive management acknowledges that uncertainty and surprise in how ecosystems respond to management policies will always be with us, and therefore explicitly treats policies as experiments (Lee 1993). Each AMA is in a different geographic and ecological province of the Pacific Northwest, and most are associated with areas that have been impacted socially and economically by reduced timber harvest from federal lands. The ten areas together "provide a diversity of biological challenges, intermixed land ownerships, natural resource objectives, and social contexts" (NWFP 1994, page D-1).
 

These recent changes in the focus of forest management are not unique to the Pacific Northwest. The Model Forest Network is a program which originated in Canada and has expanded to other nations, including Russia, Mexico, the Philippines, Malaysia and the United States (three AMAs--Cispus, Applegate, and Hayfork--are part of the Network). The objectives of Model Forests are almost identical to those of AMAs, including promotion of sustainable forestry and integrated resource management, creation of partnerships with different social interests, and testing and demonstrating the best sustainable forestry practices using current technology and science.
 

The United Nations Conference on Environment and Development (UNCED) held in Rio de Janeiro, Brazil in 1992, culminated in the "Rio Declaration on Environment and Development", which offers policies and programs to achieve a sustainable balance between consumption, population and the Earth's life-support capacity. The Rio agreements led to the "Montreal process" and the "Santiago declaration", or "Statement on Criteria and Indicators for the Conservation and Sustainable Management of Temperate and Boreal Forests", signed by the United States and 9 other countries. This document lists indicators which will be measured to assess, among other things, each nation's success at conserving biological diversity, maintaining healthy and productive forest ecosystems, and maintaining forests' contribution to regulating the global carbon cycle. These international initiatives provide a framework and a source of ideas for adaptive management in the Pacific Northwest.
 
 

The Northern Coast Range AMA extends from the Wilson River drainage northeast of Tillamook to the Siletz and Luckiamute River drainages east of Newport, and from the Pacific Ocean to the Willamette valley foothills (Figure 1). The initial emphasis in the NWFP for the Northern Coast Range AMA is "management for restoration and maintenance of late-successional forest habitat..[and]..development of a comprehensive strategy for conservation of the fisheries and other elements of biological diversity in the northern Oregon Coast Ranges" (NWFP 1994, page D-15). The key to success will be to test comprehensive approaches that meld these biological objectives with local social and economic objectives. Flexibility in some of the NWFP management guidelines within AMAs allows for local innovation and experimentation. It is expected that localized approaches to managing AMAs will be developed that "rely on the experience and ingenuity of resource managers and communities rather than [on] traditionally derived and tightly prescribed approaches that are generally applied in management of forests" (NWFP 1994, page D-1).

There are approximately 113,000 hectares (281,000 acres) of federal lands administered by the Forest Service (USFS) and Bureau of Land Management (BLM) in the Northern Coast Range AMA. Although adjacent landowners may have similar interests in protecting forest ecosystems, they are not required to implement the standards and guidelines of the NWFP. Recognizing that relationships between adjacent lands can be important, however, the AMA management team has adopted an "AMA boundary" for its assessments which includes all watersheds containing federal lands in the AMA (approximately 670,000 ha or 1.66 million ac) (Figure 1).

  Figure 1             Figure 2

There are several important federal land allocations from the NWFP that apply in this AMA (Figure 2). Late-Successional Reserves (LSRs) were designated to protect and enhance old-growth forest conditions; management designed to enhance such conditions is allowed in stands up to age 110. These reserves occupy 66% of the federal lands in the Northern Coast Range AMA. Modification of boundaries or management guidelines is allowed under the NWFP given appropriate planning and consideration of the needs of late-successional species. Within the LSRs, spotted owl Reserve Pair Areas (RPAs) must be designated using suitable and "next-best" habitat around confirmed points where spotted owls have been active and most vegetation management is not allowed. Murrelet circles are designated around any stand with evidence of occupation by marbled murrelets; management guidelines are similar to those for LSRs. Riparian reserves of different widths were designated around all streams (perennial and intermittent) and water bodies in the NWFP. These were intended to protect aquatic habitat and provide for productive late-successional forest. Management guidelines are similar to those for LSRs and boundaries designated in the NWFP can be modified following site-specific evaluation. The Cascade Head Scenic Research Area and Experimental Forest is designated for research purposes and there is no management guidance under the NWFP. The mature forests at Cascade Head will most likely be protected, however, and will contribute greatly to the overall goals of the NWFP in this area.
 

Several documents provide additional useful information concerning the Northern Coast Range AMA. Background information, a management process, and initial plans for the AMA are presented in the "Northern Coast Range Adaptive Management Area Guide" (USDI Bureau of Land Management and USDA Forest Service 1997). The "Social Assessment" (Leonard 1997) analyzes social and economic conditions and trends in and around the AMA. The "Late-successional Reserve Assessment" (USDA Forest Service and USDI Bureau of Land Management 1997) describes the resources and issues affecting late-successional habitat and presents objectives and guidance for management of Late-Successional Reserves within the AMA. The USFS and BLM have also completed Watershed Analyses for the Nestucca, North Yamhill, Schooner/Drift, and Upper Siletz watersheds; several more are in progress. A broader perspective of the ecological and social context of the Oregon Coast Range is provided in the "Coast Province Assessment" (USDA Forest Service 1995).
 

An important part of any learning endeavor, which is what AMAs are all about, is having a clear idea of what the primary issues are, what is already known about them, and a compilation of tractable questions to guide future learning. The purpose of this assessment is to provide those building blocks for the Northern Coast Range AMA. A complete synthesis of existing knowledge applicable to this AMA would be a work of encyclopedic proportions; I have therefore tried to summarize the major points of past work and include references for readers wishing greater detail. The scope of the assessment is focused on the northern Oregon Coast Range, but also borrows from relevant work conducted within the region, particularly if local information is not available. The specific goals of this assessment include:

1. Summarize past and current research efforts in and around the AMA,

2. Identify questions relevant to the AMA in need of further learning,

3. Provide some examples of learning designs to answer selected questions,

4. Identify opportunities to match future learning efforts with different areas within the AMA, and

5. Stimulate further discussion on learning priorities, implementation strategies, and coordination between the various partners in the AMA
 

The development of this assessment has benefitted greatly by borrowing ideas from the Research and Learning Assessment for the Central Cascades AMA (Cissel 1995).
 
 

Adaptive management is a formalized process that applies knowledge of the effects of a management action to adjust subsequent actions to better meet goals and expectations (Holling 1978, Lee 1993, Bormann et al. 1994). Adaptive management recognizes that the full ramifications of management actions on ecosystems and society are unknown and that social demands and ecological issues change rapidly. The process is often described as a continuous cycle allowing assessment and adjustment to new information and new conditions (Figure 3). Any management approach with multiple objectives (for example, maintaining many different resources, species, and benefits to humans) is of necessity a "best guess" about what will work. Thus learning about the effects of management becomes an integral goal of adaptive management and is what distinguishes it from many past forms of management.

Management agencies have always learned, but issues were often narrowly defined and learning often occurred slowly and in response to crises. The actions of land management agencies and oversight agencies are often based on selecting the "best management practices" out of an array of possible options. These practices often average or compromise competing viewpoints and objectives to choose a single approach (Bormann et al. 1999). Best management practices were typically applied more or less indiscriminately across the landscape until sufficient evidence of unacceptable results accumulated to lead managers to develop new practices.
 

The most efficient approach to learning is very different from "best management practices", because different approaches to achieve similar goals are applied in management actions and the necessary variables to assess their effectiveness are
 

monitored and evaluated (Figure 4). This is an application of scientific principles of experimental design--which have been developed to increase learning as rapidly and efficiently as possible--to management questions. There will always be a place for carefully controlled experiments to answer specific questions about a limited number of variables. It would be impossible to exercise the same level of control, however, when addressing broad management questions concerning multiple variables (e.g. fish, wildlife, revenue, and recreation). Nevertheless, structured comparisons of multiple management approaches is the fastest way to increase learning, although the statistical power common to traditional experiments may not be achieved. The greater the differences between compared management approaches, the easier (i.e. cheaper) it is to detect different effects through monitoring (Bormann et al. in review). Some approaches will no doubt be seen as "risky", and some may fail to meet objectives, but it is no doubt better to find the most effective approach than to simply choose the momentarily least controversial (and potentially least effective) one.

Once a range of approaches have been identified, a written record of the anticipated outcomes of each action should be kept to document the thinking and assumptions behind a decision. Elimination of bias will be critical to ensure results are useful and accepted by the widest range of stakeholders: objectivity in management studies is best (and possibly only) gained by randomly allocating approaches on units that are believed to be initially similar. Results will be more generalizable and powerful if the sets of approaches are repeated in different areas across the landscape (replication). Other, less powerful learning designs are discussed in Bormann et al. (1999). In general, however, all learning designs (and science) have two key technical elements in common: systematic observation and critical analysis.

Monitoring plans should be developed and incorporated into the design of the management study; monitoring variables should be carefully chosen to evaluate the questions of interest, rather than be an inventory wish-list. A long-term commitment to field monitoring is imperative. "Adaptive management is based on monitoring that is sufficiently sensitive to detect relevant ecological changes. In addition, the success of adaptive management depends on the accuracy and credibility of information obtained through inventories and monitoring" (NWFP 1994, p. E-3). Without monitoring, there can be no adaptive management. Without monitoring, federal land managers may find their management under court injunction again (Seattle Audubon Society v. Lyons 1994).
 

Adaptive management presents some real challenges to the research community. Resources will rarely be available to definitively answer the large-scale, interacting questions facing land management, and management actions will rarely have the level of design and control of variation many researchers would desire. Research specialists may be unable or unwilling to assist with many broad multi-resource issues central to sustainability (e.g. resource consumption, population growth). A particular challenge is how to incorporate different kinds of information about different social and natural resource processes into a systematic learning process (Stankey and Shindler 1997). Perhaps most insidious is a public perception that the conduct of "science" (i.e. structured learning) is best left to a technocracy of academically trained professionals (Cooperrider 1996). AMAs might founder without an involved citizenry if this perception prevails.
 

The topography of the Northern Coast Range AMA consists primarily of highly dissected mountains commonly less than 600 m (2,000 ft) high, but with a few ranging to 1,200 m (4,000 ft) in elevation. Slopes are commonly steep (>40% slope), particularly in the northern portion of the AMA. Soils are generally deep and have developed from volcanic basalts and sedimentary sandstones, siltstones, and shales; landslides are a dominant process shaping the topography (Federal Lands Assessment 1995). The Siuslaw National Forest has developed land-type associations that integrate geologic bedrock, soil type, slope, and climate zone into similar geomorphic units (USDA Forest Service and USDI Bureau of Land Management 1997).
 

Climate in the Coast Range follows the Pacific Northwest pattern of cool, wet winters and warm, dry summers, but there is substantial variability within the AMA, influenced primarily by proximity to the ocean and elevation. Annual precipitation ranges from about 200 cm (80 in) on the coast to as high as 500 cm (200 in) on the highest ridges of the Coast Range, down to 110 cm (45 in) in the Willamette valley foothills. A "fog belt" exists along the coast and extends up to 1.6 km (1 mi) inland, or even farther up valleys, where it is often cool and moist in the summer. Strong winter storms can deliver winds as high as 220 km/hr (140 mi/hr) in exposed areas along the coast. Transient snowpacks may occupy the higher peaks in winter.
 

The AMA includes some of the most productive forests in the world, with aboveground net primary production of mature stands of 10-15 t/ha/yr or wood volume production of 16 m³/ha/yr (229 ft³/ac) (Fujimori et al. 1976, Gholz 1982), and biomass production in young (26-yr old) stands as high as 30 t/ha/yr (Fujimori 1971). Most of the vegetation within the AMA belongs to two basic forest types (Franklin and Dyrness 1973). The Sitka spruce (Picea sitchensis) Zone is found along the coast and up major river valleys from the coast, and the vegetation is dominated by Sitka spruce, western hemlock (Tsuga heterophylla), western redcedar (Thuja plicata), and salmonberry (Rubus spectabilis). Fire occurrence is rare, but wind is an important disturbance to these forests. Forests farther inland belong to the western hemlock zone, and the vegetation is dominated by Douglas-fir (Pseudotsuga menziesii) and western hemlock. Fire tends to be infrequent but severe, killing most of the trees in a stand when it does occur. Forests in the drier valley margin may constitute a grand fir (Abies grandis) zone and are often dominated by Douglas-fir, grand fir, and big-leaf maple (Acer macrophyllum). Due to their proximity to the frequently-burned Willamette Valley grasslands, valley margin forests in pre-settlement times probably experienced fires of moderate frequency and severity which killed patches of trees in some areas, single trees in other areas, and burned lightly on the forest floor in still others. Within each zone characteristic plant communities (or associations) are found which reflect differences in site productivity and moisture status (Hemstrom and Logan 1986). An effort to predict and map "Plant Association Groups" across the coast range is almost complete (Cindy McCain, personal communication).
 

Forest lands adjoining federal lands in the AMA include the Tillamook State Forest at northern end of the AMA, the Confederated Tribes of the Grande Ronde Community in the center, and the Confederated Tribes of the Siletz Indians of Oregon in the southern part of the AMA. Major private industrial forest land owners include Boise Cascade, Cavenham, Georgia Pacific, Hampton, Simpson, Stimson, and Willamette Industries. Approximately 60,000 people live near the federal lands in the AMA, with 77% living in rural communities. The population is growing significantly and aging in coastal counties as retirees and second-home buyers acquire homes on the coast. Employment in the retail and service industries has increased while employment in manufacturing and the forest industry has decreased over the last 35 years, leading to a decline in wages for many people. Economic growth is associated with tourism and residential development (Leonard 1997).