Proceedings: Views from the Ridge—Considerations for Planning at the Landscape Scale

Home , Ait

United States Department of Proceedings: Views From the Agriculture Forest Service Ridge—Considerations for Pacific Northwest Research Station Planning at the Landscape Scale General Technical Report PNW-GTR-596 January 2004 Photo credits Top photo taken by Roger Clark, bottom left to right Kevin Burke, Roger Clark

Compilers Hermann Gucinski (retired) was a physical scientist, Forestry Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR 97331; Cynthia Miner is a general biologist and Becky Bittner is a forester, Pacific Northwest Research Station, 333 SW First Avenue, Portland, OR 97208-3890. Papers were presented at a conference by the authors, who are therefore responsible for the content and accuracy. Opinions expressed may not necessarily reflect the position of the U.S. Department of Agriculture. The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement or approval by the U.S. Department of Agriculture of any product or service to the exclusion of others that may be suitable. Proceedings: Views From the Ridge—Considerations for Planning at the Landscape Scale

Hermann Gucinski, Cynthia Miner, and Becky Bittner Editors

Views From the Ridge—Considerations for Planning at the Landscape Scale, Vancouver, WA, November 2-4, 1999

Sponsored by: Pacific Northwest Research Station, USDA Forest Service Western Forestry and Conservation Association U.S. Department of Agriculture Forest Service Pacific Northwest Research Station Portland, Oregon General Technical Report PNW-GTR-596 January 2004 Abstract Gucinski, Hermann; Miner, Cynthia; Bittner, Becky, eds. 2004. Proceedings: views from the ridge— considerations for planning at the landscape scale. Gen. Tech. Rep. PNW-GTR-596. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 133 p. When resource managers, researchers, and policymakers approach landscape management, they bring perspectives that reflect their disciplines, the decisions they make, and their objectives. In working at a landscape level, they need to begin developing some common scales of perspective across the variety of forest ownerships and usages. This proceedings is a compilation of 22 papers presented at a conference that addressed divergent views on landscape management. The conference was a forum for exchanging concepts and knowledge from research and management experiences about managing landscapes. The program addressed the issues of managing landscapes when everyone has a different perspective; approaching landscape management from aquatic, terrestrial, and socioeconomic viewpoints; and characterizing landscape management. Keywords: Landscape management, forest policy, forest management, aquatic, terrestrial, socioeconomic. Contents 1 Introduction Hermann Gucinski

Conference Overview 5 From Red Queens to Mad Hatters—A Wonderland of Natural Resource Management Lance Gunderson 11 What Is a Landscape and How Is One Studied? Virginia H. Dale and Barry R. Noon

How Do We Determine Scale? 17 Landscapes: The Dilemma of Scale and the Role of Theory Lee Benda and Daniel Miller 22 Scale Considerations for Linking Hillslopes to Aquatic Habitats Robert R. Ziemer 33 Landscape Assessment in a Multiownership Province Thomas A. Spies 38 Implications of Scale on Viability Assessments of Terrestrial Wildlife Species Martin G. Raphael 42 Projecting Potential Landscape Dynamics: Issues and Challenges Ronald P. Neilson 48 Scaling and Landscape Dynamics in Northern Ecosystems Marilyn Walker 57 Landscape Perspectives Roger N. Clark, Linda E. Kruger, Stephen F. McCool, and George H. Stankey 61 Landscape-Level Management, It’s All About Context Bruce Shindler 62 Human Adaptation to Environmental Change: Implications at Multiple Scales Ellen M. Donoghue, Terry L. Raettig, and Harriet H. Christensen 67 Land Use and Land Cover Dynamics: Changes in Landscapes Across Space and Time Ralph J. Alig

Worldviews Panel 75 Private Landowner Perspective on Landscape Management Carl F. Ehlen 78 Back to the Future: The Value of History in Understanding and Managing Dynamic Landscapes Penelope Morgan Case Studies 87 Northwest Forest Plan Nancy M. Diaz 93 Recent Changes (1930s-1990s) in Spatial Patterns of Interior Northwest Forests, USA P.F. Hessburg, B.G. Smith, R.B. Salter, R.D. Ottmar, and E. Alvarado

Landscape Methodology 95 Analysis and Communication of Monitoring Data With Knowledge-Based Systems Keith M. Reynolds and Gordon H. Reeves 105 Seeing the Forest and the Trees: Visualizing Stand and Landscape Conditions Robert J. McGaughey 110 A Method to Simulate the Volume and Quality of Wood Produced Under an Ecologically Sustainable Landscape Management Plan Glenn Christensen, R. James Barbour, and Stuart Johnston

Transforming Information From Many Perspectives to Action on the Ground 119 A Social Science Perspective on the Importance of Scale Linda E. Kruger

Conclusion 127 Summary Remarks on Views From the Ridge T.F.H. Allen 130 The Future of Landscape-Level Research at the Pacific Northwest Research Station Thomas J. Mills Introduction

Hermann Gucinski1

Fred Swanson, a scientist who has devoted much provide a yet larger prospect (“Vogelschau” of his career to the study of the processes that [bird’s-eye view] may be the most apropos Ger- affect the character of a landscape, was asked man word for this), whereas “a view from the val- while leading a tour of the H.J. Andrews Experi- ley” would help discern the delivery of results from mental Forest in Oregon with the group poised upland processes when taking an aquatic or ripar- atop a commanding ridge, “Well, just what is a ian perspective. good landscape?” These processes, the delivery of mass and en- As a scientist committed to expanding knowledge ergy, among others, bring us to the scientific part rather than to just citing known facts, he took this of the inquiry—the appropriate subject of “land- question as a serious challenge, admitted to find- scape ecology.” Richard T.T. Forman and Michael ing it perplexing, and used it in turn to engender Godron (1986) used this definition in their seminal deep discussion, and some introspection, around work: the campfire that evening. Landscape ecology is the study of struc- Before we can use the same question as a frame- ture, function and change in a heteroge- work for the contributions in this volume, it might neous land area composed of interacting be good to ask: “What is a landscape?” and, be- ecosystems. It therefore focuses on: cause this is a scientific endeavor, “What is land- • structure, the spatial patterns of landscape ecology, and how might it contribute toward scape elements and ecological objects the larger question?” For it was indeed the objec- (such as animals, biomass and mineral tive of the symposium, and the proceedings that nutrients); resulted, to take the “View From the Ridge” as a scientific challenge, rather than a subjective one, • function, the flows of objects between and to try to learn if a scientifically framed answer landscape elements; and can also help with the subjective part of the • change, alterations in the mosaic through needed answer. time. Webster’s unabridged dictionary and the Oxford Monica Turner (1989) defines it thus: English Dictionary have similar definitions: “a portion of land or territory, which the eye can compre- Landscape ecology emphasizes the inter- hend in a single view, including all the objects so action between spatial pattern and eco- seen” for the former, and “a view or prospect of logical process—that is, the causes and natural in-land scenery, such as can be taken in consequences of spatial heterogeneity at a glance from one point of view; a piece of across a range of scales. Two important country scenery” for the latter. Both offer some aspects of landscape ecology distinguish it additional definitions: “a tract of land with its dis- from other subdisciplines within ecology. tinguishing characteristics and features, esp. First, landscape ecology explicitly ad- considered as a product of modifying or shaping dresses the importance of spatial configu- processes and agents (usually natural)”—“a view, ration for ecological processes. Not only is prospect of something”—“a distant prospect: a landscape ecology concerned with how vista,”—“a bird’s-eye view; a plan, sketch, map.” much there is of a particular component Thus, our title, “Views From the Ridge” is appro- but also with how it is arranged. Second, priate, although one contributor, Bob Ziemer, landscape ecology often focuses upon astutely notes that a “view from space” would spatial extents that are much larger than those traditionally studied in ecology. 1Physical scientist (retired) Forestry Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR 97331.

1 The following papers show the interest focused on the view from the ridge to a quite specific ridge, the importance of spatial configuration for an eco- namely Oak Ridge, site of the Oak Ridge National logical process and will demonstrate the astonish- Laboratory and its rich history of contributions to ing variety of both viewpoints and scales used to landscape ecology. Today’s efforts are enhanced address problems. This shows the science to be by better tools and methods, as Spies notes in evolving and is a sign of the value placed on prob- his contribution, and they also require looking lems in this area. A brief look at the history of for- beyond political boundaries such as ownerships. estry science shows its earliest endeavors, silvi- McGaughey introduces us to valuable visualiza- culture and entomology, to be largely driven by tion tools now available to permit examination of stand-scale processes. Problems were consid- landscape “treatment” scenarios. Raphael argues ered solvable at that scale. Not only that, but the that being able to hold onto all scales is a neces- nature of the problem and suggested solutions sity when considering the problem of “viability” of were seen as largely separable. You isolated each a species that is threatened. Clark et al. wrestle variable, and treatments of that variable were with the subjective components of a landscape as based on the discipline within which it fell. they relate to the values that people assign to it. Although this is oversimplified, earlier scientific And, as Shindler reminds us, right up to the endeavors engendered the separation of the dis- present, our decisions have almost always been ciplines and slowed understanding of the interde- made at the stand level. Morgan expands our pendence of ecological processes. “If all you have vista to not only view a landscape in space, but is a hammer, every problem will begin to re- become aware of it over time such as a knowl- semble a nail.” This approach worked for a long edge of history permits. Here the concept of “his- time. torical range of variation” can be a vital tool. Benda and Miller take landscape processes under Many phenomena contributed to the expanding the lens, focusing on the differences between view of ecological processes. We recognized that stochastic and deterministic processes. Confus- the accumulation and translocation of long-lived ing these can have disastrous consequences. pollutants required a comprehensive look at bio- Neilson reminds us of the nature of “emergent logical, hydrological, and atmospheric processes. properties” when going to a larger scale, and uses So did the need to develop concepts of “carrying climate response as a useful integrator for these capacity” as the pressures of nonsustainable re- processes. source use—hastened by population growth— brought conditions that could no longer be The intent here is not to give away the principal described or understood at stand levels. This theses and conclusions of each contributor. In- development challenged resource managers to stead, I want to whet the appetite of the reader, abandon linear approaches to management— convey the breadth and depth that are contained ecological processes are cyclical in nature, and in this volume, reflect on the variability of the ap- evolutionary development has allowed highly inte- proaches outlined, and weigh their implications. grated system functioning that produces no Finally, I hope the information given and the chal- waste. Superimposing solutions based on linear lenges raised will encourage the reader to rethink thinking caused new problems to pop up in places the question, “What is a good landscape?” in light outside the scientific discipline where the original of available science as it illuminates the subjective problem first appeared. Addressing the conse- needs we have. quences resulted in yet other problems, leading to a vicious circle of “stimulus-response.” Literature Cited That dynamic can be broken by moving up in Forman, R.T.T.; Godron, M. 1986. Landscape scale from the domain of the initial problem, seek- ecology. New York: John Wiley and Sons. ing to understand the connections while standing on the ridge—or, alternatively, in the valley. This Turner, M.G. 1989. Landscape ecology: the effect can help us regain perspective, and find pathways of pattern on process. Annual Review of Ecol- to solutions. Dale and Noon tackle this by relating ogy and Systematics. 20: 171-191.

2 3 Conference Overview

4 From Red Queens to Mad Hatters—A Wonderland of Natural Resource Management Lance Gunderson1

Lewis Carroll (1899) created a number of memo- cesses that support that stability domain, conse- rable characters who appear and reappear quences of those assumptions, and explanations throughout the adventures of Alice in Wonderland. for policy and action. At least five such caricatures A few of these characters highlight and resonate are described in the following paragraphs (modi- with my sense and experience of the kinds of fied from Holling and Gunderson 2002). issues faced by resource managers. Just as Alice is bewildered by the complexities and uncertain- Nature Flat ties of Wonderland, managers often face similar Although this perception is perhaps so simplistic issues of sensemaking and coping with the inevi- as to be silly, it is included here as a starting point table surprises of complex systems of humans from which subsequent contrasts with other views and nature. In the following sections, I use these can be made. In this view, the word “flat” is used rich characters from Wonderland as an entree to describe a system in which there is little or no into a set of topics that will hopefully stimulate stability of structure over time. The hands of some ideas related to the paradigms that underlie humans, given sufficient resources can change our management of natural resources at the land- nature, and there are no feedbacks or conse- scape scale, beginning with the Red Queen. quences of those actions—it is much like rolling a ball around on a cookie sheet. The processes that Red Queens and Other Caricatures affect the state of nature are random or stochas- “Well, in our country,” said Alice, still pant- tic. In such a view of nature, policies and politics ing a little, “you’d generally get to some- are random as well, often described as garbage where else—if you ran very fast for a long can dynamics (March and Olsen 1989). It is a time, as we’ve been doing.” nature that is infinitely malleable and amenable to human domination. As such, the issues of re- “A slow sort of country!” said the Queen. source use, development, or control are identified “Now, here, you see, it takes all the run- as issues of people and resolved by activism or ning you can do, to keep in the same community organization. place. If you want to get somewhere else, you must run at least twice as fast as Nature Balanced that!” ……..Lewis Carroll The second is a view of nature at or near an equilibrium condition, which can be static or dynamic. The Red Queen’s action depicted in this scene is Hence if nature is disturbed, it will return to an often used as a metaphor by evolutionary biolo- equilibrium through (in systems terms) negative gists to describe whether species can adapt to feedback. As such, nature seems to be infinitely rapidly changing environments. I see similar “Red forgiving. This is the view of nature that underpins Queen dilemmas” as practitioners work as fast logistic growth where the issue is how to navigate as they can to cope with rapidly and unpredictably a looming and turbulent transition—demographic, changing systems. This metaphor can be ex- economic, social, and environmental—to a sus- tended to include a wider set of assumptions or tained plateau. This is the view of several institu- myths that help to explain how nature operates. tions with a mandate for reforming global re- Each of these myths has a different supposition of source and environmental policy: the Brundtland how stable systems are (in the sense of respond- Commission (Brundtland and Khalid 1987), the ing to perturbations over time), the types of pro- World Resources Institute, the International Insti- tute of Applied Systems Analysis (Clark and Munn 1Associate professor and Chair, Department of Environmental 1986), and the United Nations (Munasinghe and Studies, Emory University, Atlanta, GA 30322; Tel: 404-727- McNeely 1995) for example, who are contributing 2429; e-mail: [email protected]

5 skillful scholarship and policy innovation. They are and management approaches that are adaptive. among some of the most effective forces for But both of these presume a stationary land- change. scape. In this case, our cookie sheet has been molded and curved in three dimensions, but it is Nature Anarchic fixed over time. If the previous myth is one where the system sta- Nature Evolving bility is defined as a ball at the bottom of a cup, this myth is one of a ball at the top of a hill. It is The emerging fifth view is evolutionary and adap- globally unstable. It is a view dominated by hyper- tive. It has been given recent impetus by the para- bolic processes of growth and collapse, or where doxes that have emerged in successfully applying increase is inevitably followed by decrease. It is a the previous more limited views. Complex sys- view of fundamental instability where persistence tems behavior, discontinuous change, chaos and is only possible in a decentralized system with order, self-organization, nonlinear system behav- minimal demands on nature. It is the view of ior, and adaptive evolving systems are all the some extreme environmentalists. If the previous present code words characterizing the more review assumes that infinitely ingenious humans do cent activities. Such thinking is leading to integra- not need to learn anything different, this view as- tive studies that combine insights and people from sumes that humans are incapable of learning. developmental biology and genetics, evolutionary This is implicit in the writings of Tenner (1996), biology, physics, economics, ecology, and com- where he argues that all technology that is un- puter science. It is a view of an actively shifting leashed will eventually “bite back.” This view pre- landscape with self-organization (where the stabil- sumes that small is beautiful because of the ity of the landscape affects behavior of the vari- inevitable catastrophe of any policy. It is a view ables, and the variables, plus exogenous events, where the precautionary principle dominates affect the stability of the landscape). It is a view of policy, and social activity is focussed on mainte- cross-scale interactions of processes–dubbed nance of the status quo. panarchy in previous writings (Gunderson et. al. 1995, Holling and Gunderson 2002). It is a view Nature Resilient that requires a focus on active policy probes of a shifting domain and a focus on institutional and The fourth view is of nested cycles organized by political flexibility for learning. fundamentally discontinuous events and processes. That is, there are periods of exponential Comparing and contrasting these underlying cari- change, periods of growing stasis and brittleness, catures or worldviews, provides some insight into periods of readjustment or collapse, and periods how we create our wonderlands in order to make of reorganization for renewal. Instabilities organize prescriptions for action. It is also useful to note the behaviors as much as do stabilities. This has how these partial myths are adopted and rein- recently been the focus of fruitful scholarship in a forced and prescribed by a variety of disciplines, wide range of fields—ecological, social, eco- as described in the next section, organized nomic, and technical. These dynamics have simi- around Mad Hatters of various disciplines. larities in Harvey Brook’s view of technology (1986), Brian Arthur’s and Kenneth Arrow’s recent Disciplinarily Mad Hatters views of the economics of innovation and competition (Waldrop 1992), Mary Douglas’ (1978) and Alice had been looking over his shoulder Mike Thompson’s (1983) views of cultures, Don with some curiosity. “What a funny watch!” Michael’s view of human psychology (1984), and she remarked. “It tells the day of the Barbara Tuchman’s (1978) and William McNeill’s month, and doesn’t tell what o’clock it is!” (1979) views of history. The “nature resilient” view “Why should it?” muttered the Hatter. is a view of multiple stable states in ecosystems, “Does YOUR watch tell you what year it is?”

6 “Of course not,” Alice replied very readily, equally narrow community development ones. “but that’s because it stays the same year Each one builds their efforts on theory, although for such a long time together.” many would deny anything but the most pragmatic “Which is just the case with MINE,” said and nontheoretical foundations. The conservation- the Hatter. ists depend on theories of ecology and evolution, the developers on variants of free market models, Alice felt dreadfully puzzled. The Hatter’s the community activists on theories of community remark seemed to have no sort of mean- and social organization. All these theories are ing in it, and yet it was certainly English. ”I correct in the sense of being partially tested and don’t quite understand you,” she said, as credible representations of one part of reality. The politely as she could. problem is that they are partial. That partiality of ……..L. Carroll concepts leads to the search for theories of The Mad Hatter utters seemingly nonsensical change that bridge disciplines in rich ways. questions and answers in his dialogues with Alice One such integrated theory is Holling’s (1992) at the tea party. He reminds me of those of us adaptive cycle. The heuristic combines ecological who are technically or scholarly oriented, the edu- theories of succession and ecosystem develop- cated skeptics who pay attention to different indi- ment with other concepts of stability and resil- cators, have different frames of reference and ience. It is a rich framework to indicate how different interpretations and understandings. As natural capital and connectivity of systems in- with the Mad Hatter, in attempting to communi- crease slowly over time. But those properties lead cate our technical understanding to policymakers to an increasing vulnerability and inevitable peri- or decisionmakers, more often than not, we are ods of destruction and reorganization. Authors politely misunderstood. And all of us are like the have used this framework to explain co-evolution Hatter in that we wear our disciplinary hats of resources and management through time proudly—whether as ecologist, biologist, political (Light et al. 1995), business and organizational scientist, economist, or any other scholarly cat- dynamics (Westley 1995), and political systems egorization. Yet there is a growing sense that (Holling and Sanderson 1996). these discipline-based organizations of inquiry and understanding are problematic. The above excerpt from the tea party, in which Alice and the Mad Hatter discuss differences in Management of global and regional resources is their watches, also suggests to me that one of the not an ecological problem, nor an economic one, key challenges that resource managers face is nor a social one. It is a combination of all three. overcoming obstacles of scale. Those obstacles And yet actions to integrate all three inevitably are both theoretical and practical. How managers shortchange one or more. Sustainable designs attempt to analyze and learn from their actions driven by conservation interests ignore the needs are both related to issues of scale. Most models for an adaptive form of economic development and modes of inquiry are scale bound and depen- that emphasizes enterprise and flexibility. Those dent. Walters (1997) cites the cross-scale prob- driven by economic and industrial interests act as lem as a severe obstacle in most assessment/ if the uncertainty of nature can be replaced with modeling activities. Development of new theories human engineering and management controls, or is needed to help address ecosystem and natural ignored altogether. Those driven by social inter- resource dynamics across space and time scales. ests act as if community development and em- Over the last 40 years, time and space have been powerment of individuals hold the key and there separated for analytical purposes. Most field eco- are no limits to the imagination and initiative of logic investigations either freeze space and ex- local groups. As investments fail, the policies of periment over time or freeze time and look at government, private foundations, international spatial patterns (witness the explosion and ubiq- agencies, and nongovernmental organizations flip uity of geographic information system technology from emphasizing one kind of myopic solution to in resource management agencies). Perhaps another. Over the last three decades, such poli- there are practical reasons for this pattern, but it cies have flipped from large investment schemes, also can be explained in part because of underly- to narrow conservation ones to, at present, ing theoretical frameworks. There is a growing

7 sense that this separable-dimension framework rather on more enduring system properties such results in assessments with different outputs, sug- as resilience, adaptive capacity, and renewal ca- gesting the need for integration and reconciliation. pability. Indirectly, these system properties have been explored in large complex ecosystems Through The Looking Glass and What such as the Grand Canyon (Walters 1997), the Alice Found There Columbia River basin (Volkmann and McConnaha 1993), and the Everglades (Gunderson 1999). A For some minutes Alice stood without resilience or adaptive capacity framework involves speaking, looking out in all directions over both the human components of the system (op- the country—and a most curious country it erations, rules, policies, and laws) and the bio- was. There were a number of little brooks physical components of the landscape and its running across from side to side, and the ecosystems. The shift of focus to a learning basis ground between was divided up into is likely to require flexible linkages with a broader squares by a number of hedges, that set of actors, or network. Another way of saying reached from brook to brook. this bluntly is, until management institutions are “I declare it’s marked out just like a large able and willing to embrace uncertainty and sys- chessboard!” Alice said at last. “There tematically learn from their actions and respond to ought to be some men moving about that learning, adaptive management will not con- somewhere—and so there are!” she tinue in its original context, but rather be redefined added in a tone of delight, and her heart in a weak context of “flexibility in decisionmaking” began to beat quick with excitement as (Gunderson 1999). she went on. “It’s a great game of chess But what does it take to be hopeful in a world that that’s being played all over the world—if is perhaps becoming much more unforgiving? As this is the world at all, you know. Oh, what the degree of human impact continues to increase fun it is! in scale, a key unanswered question is whether ……..L. Carroll the adaptive capacity of both ecologic and social One of the enduring and endearing aspects of the systems can keep pace with this expanding hu- Alice in Wonderland stories is that she faces an man footprint. Under those conditions, the pre- uncertain world (from the nightmarish to the sub- scription for facilitating constructive change lime) with hope and wonder. For some who deal appears to be: with issues of natural resource management, • Identify and reduce destructive constraints and there is only doubt and gloom. However, I tend to inhibitions on ecological change, such as per- agree with Alice, and I suggest that there is rea- verse subsidies (e.g., sugar farming in the son to hope. But that hope is founded on develop- Everglades). ing and creating new ways to think about and manage issues of the environment. • Protect and preserve the accumulated experience on which change will be based (such as Perhaps it is time to rethink the paradigms or managers in land management agencies with foundations of resource management institutions, multiple decades of experience). and place more emphasis on development of sustaining foundations for dealing with complex • Stimulate innovation in a variety of fail-safe resource issues. Learning is a long-term proposi- experiments that probe possible directions in tion that requires a ballast against short-term poli- ways that are low in costs for people’s careers tics and objectives. Another shift likely will require and organizations’ budgets (such as adaptive a change in the focus of actions away from man- policies in the Grand Canyon). agement by objectives and determination of • Encourage new foundations for renewal that optimum policies toward new ways to define, un- build and sustain the capacity of people, derstand, and manage these systems in an ever- economies, and nature for dealing with changing world. That focus should not be solely change. on variables of the moment (water levels, population numbers) and their correlative rates, but

8 These suggestions are founded on the premise Gunderson, L.; Holling, C.S.; Light, S.S., eds. that we must learn our way into an uncertain fu- 1995. Barriers and bridges for the renewal of ture. That learning should help guide deliberations ecosystems and institutions. New York: Colum- toward workable and sustainable futures. Those bia University Press. deliberations will not solve all social or distribu- Holling, C.S. 1992. Cross-scale morphology, tional issues, but rather might help frame ways to geometry and dynamics of ecosystems. Eco- work through this wonderland of resource man- logical Monographs. 62(4): 447-502. agement—we don’t have the luxury of awakening and realizing that it may have been a dream. Holling, C.S.; Gunderson, L.H. 2002. Panarchy, understanding transformations in systems of Acknowledgments humans and nature. Washington, DC: Island Press. This is a contribution from the Resilience Net- work, funded by a grant from the John D. and Holling, C.S.; Sanderson, S. 1996. Dynamics of Catherine T. MacArthur Foundation. Buzz Holling (dis)harmony in ecological and social systems. continues to inspire creative and novel ap- In: Hanna, S., ed. Rights to nature. Washing- proaches to resource management. ton DC: Island Press: 57-85. Light, S.S.; Gunderson, L.H.; Holling, C.S. Literature Cited 1995. The Everglades: evolution of manage- Brooks, H. 1986. The typology of surprises in ment in a turbulent ecosystem. In: Light, S.S.; technology, institutions and development. In: Gunderson, L.H; Holling, C.S. eds. Barriers Clark, W.C.; Munn, R.E., eds. Sustainable and bridges to the renewal of ecosystems and development of the biosphere. Cambridge, institutions. New York: Columbia University United Kingdom: Cambridge University Press: Press: 103-168. 325-348. March, J.G.; Olsen, J.P. 1989. Rediscovering Brundtland, G.H.; Khalid, M., eds. 1987. Our institutions. New York: Free Press. common future. Oxford, United Kingdom: Ox- McNeill, W.H. 1979. The human condition: an ford University Press. ecological and historical view. Princeton, NJ: Carroll, L. 1899. Through the looking glass and Princeton University Press. what Alice found there. New York; London: Michael, D.N. 1984. Reason’s shadow: notes on Macmillan and Co., Ltd. the psychodynamics of obstruction. Technical Clark, W.C.; Munn, R.E., eds. 1986. Sustainable Forecasting and Social Change. 26(2): 149- development of the biosphere. Cambridge, 153. United Kingdom: Cambridge University Press. Munasinghe, M.; McNeely, J. 1995. Key con- Douglas, M. 1978. Cultural bias. Royal Anthropo- cepts and terminology of sustainable develop- logical Institute. [Publisher and location un- ment. In: Munasinghe, M; Shearer, W., eds. known.] Defining and measuring sustainability: the biogeophysical foundations. Washington DC: Gunderson, L. 1999. Resilience, flexibility and World Bank: 19-56. adaptive management—antidotes for spurious certitude? Conservation Ecology. 3(1): 7. Tenner, E. 1996. Why things bite back: technol- http://www.consecol.org/vol3/iss1/art7. ogy and the revenge of unintended conse- (10 November 2003). quences. New York: Knopf.

9 Thompson, M. 1983. A cultural bias for compari- Waldrop, M. 1992. Complexity: the emerging son. In: Kunreuther, H.C.; Linnerooth, J. Risk science at the edge of order and chaos. New analysis and decison processes: the siting of York: Simon and Schuster. liquid energy facilities in four countries. Berlin: Walters, C. 1997. Challenges in adaptive man- Springer: 232-262. agement of riparian and coastal ecosystems. Tuchman, B.W. 1978. A distant mirror. New York: Conservation Ecology [online]. 1(2): 1. http:// Ballantine. www.consecol.org/vol1/iss2/art1. (10 November 2003). Volkman, J.M.; McConnaha, W.E. 1993. Through a glass, darkly: Columbia River Westley, F. 1995. Governing design. In: Light, salmon, the Endangered Species Act, and S.S.; Gunderson, L.H.; Holling, C.S., eds. adapative management. Environmental Law. Barriers and bridges to the renewal of ecosys- 23: 1249-1272. tems and institutions. New York: Columbia University Press: 391-427.

10 What Is a Landscape and How Is One Studied?

Virginia H. Dale1 and Barry R. Noon2

Introduction systems. To deal with this complexity, first hierarchy theory (O’Neill et al. 1986) and then risk Because this workshop is entitled “Views From analysis (Bartell et al. 1994, Suter 1992) arose. the Ridge,” I interpret my task in reviewing contri- However, neither of these advances recognized butions to resource management from landscape the spatial relationships inherent to ecological ecology to be to provide you with the view from interactions. Landscape ecology has recently Oak Ridge. The Environmental Sciences Division come to be a new field of study that explicitly fo- at Oak Ridge National Laboratory (ORNL) was cuses on spatial interactions (Turner 1989, Turner formed in 1955 as a response to the concern of and Gardner 1991). how radiation was affecting the environment. The Division’s objectives rapidly expanded to include Need for a Landscape Perspective analysis of environmental effects of all aspects of energy use. This topic clearly encompasses all From an ecological viewpoint, a landscape is a aspects of environmental sciences, and now our spatial extent over which ecological processes Division houses 96 scientists with degrees from take place (King 1997). More simply, it refers to a 18 fields of study. spatially heterogenous area that has a similar geomorphology and disturbance regime (Turner The history of advancements in environmental and Gardner 1991). sciences at ORNL parallels developments in land management and ultimately leads to a landscape A landscape perspective is needed to address perspective. Therefore, I thought it would be use- today’s land management problems for several ful to highlight those scientific achievements as a reasons. It is now recognized that the spatial precursor to discussing landscapes. Over the scale of environmental problems is large and that years, ORNL scientists have participated in the all ecological processes (and management ac- development of key research areas that formed tions) occur in a spatial context and are con- the basis for landscape ecology (Hagen 1992). strained by spatial location. As an example, Supported by the International Biological Pro- Fraser fir (Abies fraseri (Pursh) Poir.) are dying in gram, the field of systems ecology was created. the southern Appalachians as a result of herbivory Systems ecology views ecological interactions and population dynamics of the introduced woolly from a holistic perspective and seeks to quantify adelgid (Adelges piceae Ratzeburg), but the in- interactions of varied components. Global change sects’ distribution, and thus fir mortality, is influ- research was initiated shortly after views of the enced by the topographic conditions that restrict Earth from space led to a realization that the eco- the fir to the highest peaks (Dale et al. 1991). logical system was global. But this view was Furthermore, ecological systems can be viewed tainted with clouds of pollution. Recognition that as spatially and temporally hierarchical. That is, human impacts occur on a global as well as local processes observed at one level of organization scale led to remediation science as a way to use arise from lower level behaviors and are con- the scientific process to learn about ways to rem- strained by higher level processes. Therefore, edy pollution problems. Yet this application of solutions for contemporary environmental prob- ecological science to real-world problems was lems need to be provided within a spatial context. difficult because of the intricacies of ecological For example, natural areas that provide essential ecological services are limited in extent, and their 1 Corporate fellow, Environmental Sciences Division, Oak contributions must be interpreted within the land- Ridge National Laboratory, Oak Ridge, TN 37831-6036; Tel: scape matrix in which they reside and with the 865-576-8043; e-mail: [email protected] understanding that environmental conditions may 2 Department of Fishery and Wildlife Biology, Colorado State change over space as well as time (as with global University, Fort Collins, CO 80523-1474; Tel: 970-491-7905; warming). Thus spatially optimal solutions to land e-mail: [email protected] management options should be considered.

11 A Landscape Approach to Land presence of appropriate species, populations, and Management communities as well as the occurrence of processes at appropriate scales (Angermeier and Karr The defining attributes of an ecological landscape 1994, Karr 1991). To maintain integrity, it is neces- are structure, composition, function, and change sary to perpetuate the “characteristic” structure, (fig. 1). Structure deals with the physical relation- composition, and processes of ecological sys- ships of landscape elements to each other: their tems, preserve those key elements of landscape shape, patchiness, juxtaposition, etc. Composition geometry that facilitate essential processes, retain refers to the variety of elements that make up a the productive capabilities of the land, and main- landscape (e.g., cover types, land forms, etc.). tain the evolutionary capabilities of ecological Function indicates the ecological processes that systems. Often resource extraction or use com- occur on the landscape (e.g., persistence of promises these features of integrity, and thus patches, rates of nutrients and energy flow, ero- management actions seek ways to reinstate these sion, etc.). These attributes lead to the key focus features across the landscape. of landscape ecology: estimating the reciprocal relationships between landscape structure, func- The Land Use Committee of the Ecological Soci- tion, and composition and how they might change ety of America recommends several guidelines to over time. assist managers in decisions about the use of land (Dale et al. 2000). The guidelines are pre- Linking landscape ecology to application takes sented in full awareness that all of these rules of several steps. First, analyses must move beyond thumb cannot be implemented in every (or even a description of the attributes of structure, func- most) situations. These guidelines suggest that, tion, and composition to analyze interactions when possible, land managers should: among these attributes. Second, structure, function, and composition should be considered at • Examine impacts of local decisions in a re- multiple spatial scales (see fig. 1). For example, it gional context. is critical to know how structure, function, and • Plan for long-term change and unexpected composition of the landscape are reflected in events. population or ecosystem features. Finally, knowledge of temporal dynamics of landscape change • Preserve rare landscape elements and associ- is essential. Natural disturbance regimes can be ated species. characterized by their frequency, spatial extent, • Avoid land uses that deplete natural re- intensity, and duration. System resilience has sources. evolved in the context of these disturbances, and thus it is important to compare human-caused • Retain large contiguous or connected areas disturbance regimes to natural disturbance re- that contain critical habitats. gimes to determine if human impacts lie within the • Minimize the introduction and spread of non- bounds of system resilience. The concept of his- native species. torical range of variability has value as a bench- mark for human-induced changes. It is important • Avoid or compensate for the effects of devel- to understand reciprocal relationships between opment on ecological processes. disturbance and landscape pattern (e.g., distur- • Implement land use and management prac- bances both respond to and create landscape tices that are compatible with the natural po- pattern). tential of the area. Key goals of responsible land management are to These guidelines are based on ecological prin- provide for societal needs without usurping the ciples such as the idea that the size, shape, and resources of future human generations and to spatial relationships of habitat patches on the maintain ecological integrity and thus sustainable landscape affect the structure and function of ecological systems. The concept of ecological ecosystems (Dale et al. 2000). This landscape integrity refers to system wholeness, including the principle has several corollaries:

12 Figure 1—The structural, compositional, and functional components of ecological systems as viewed at various levels of ecological organization: the landscape, ecosystem, and population (from Dale and Beyeler 2001).

• Landscape elements vary in spatial distribu- Improved procedures are needed for selecting tion and quality in time and space. ecological indicators to assess the status and trend of ecological systems (allowing interpreta- • The structure and attributes of landscape ele- tion of indicators for large spatial and temporal ments and patches affect movement (flows of scales). matter and energy), which in turn affects spatial distributions. Acknowledgments • The nature of boundaries between patches and matrix elements controls horizontal flows Comments on the abstract by Tony King and across landscapes. two anonymous reviews are greatly appreciated. Some of the ideas reported here were developed • The spatial context (neighborhood) of a patch on a project funded by the Conservation Program affects its properties and dynamics. of the Strategic Environmental Research and Several new insights to management arise from a Development Program through the U.S. Depart- landscape perspective. The concept of ecological ment of Defense Military Interagency Purchase integrity needs further interpretation within the Requisition No. 74RDV53549127 and the Office landscape perspective so that it becomes meas- of Biological and Environmental Research, U.S. urable and thus a practical management tool at all Department of Energy (DOE). Oak Ridge National scales. Methods are needed to reliably estimate Laboratory is managed by Lockheed Martin En- an expected range of natural variation for specific ergy Research Corporation for DOE under con- ecological systems (see Parsons et al. 1999). tract DE-AC05-96OR22464.

13 Literature Cited Karr, J.R. 1991. Biological integrity: a long- neglected aspect of water resource manage- Angermeier, P.L.; Karr, J.R. 1994. Biological ment. Ecological Applications. 1: 66-84. integrity versus biological diversity as policy directives: protecting biotic resources. King, A.W. 1997. Hierarchy theory: a guide to BioScience. 44: 690-697. system structure for wildlife biologists. In: Bissonette, J.A., ed. Wildlife and landscape Bartell, S.M.; Gardner, R.H.; O’Neill, R.V. 1994. ecology. New York: Springer-Verlag: 185-212. Forecasting toxicological risk in aquatic ecosystems. New York: Columbia University O’Neill, R.V.; DeAngelis, D.L.; Waide, J.B.; Press. Allen, T.F.H. 1986. A hierarchical concept of ecosystems. Princeton, NJ: Princeton Univer- Dale, V.H.; Beyeler, S.C. 2001. Challenges in the sity Press. 254 p. development and use of ecological indicators. Ecological Indicators. 1: 3-10. Parsons, D.J.; Swetnam, T.W.; Christensen, N.L. 1999. Uses and limitations of historical Dale, V.H.; Brown, S.; Haeuber, R.A. [et al.]. variability concepts in managing ecosystems. 2000. Ecological principles and guidelines for Ecological Applications. 9: 1177-1178. managing the use of land. Ecological Applica- tions. 10: 639-670. Suter, G.W., II. 1992. Ecological risk assessment. Chelsea, MI: Lewis Publishers. Dale, V.H.; Gardner, R.H.; DeAngelis, D.L. [et al.]. 1991. Elevation-mediated effects of bal- Turner, M.G. 1989. Landscape ecology: the effect sam woolly adelgid on southern Appalachian of pattern on process. Annual Review of Ecol- spruce-fir forests. Canadian Journal of Forest ogy and Systematics. 20: 171-197. Research. 21: 1639-1648. Turner, M.G.; Gardner, R.H., eds. 1991. Quanti- Hagen, J.B. 1992. An entangled bank: the origins tative methods in landscape ecology: the of ecosystems ecology. New Brunswick, NJ: analysis and interpretation of landscape het- Rutgers University Press. erogeneity. New York: Springer-Verlag. 536 p.

14 15 How Do We Determine Scale?

16 Landscapes: The Dilemma of Scale and the Role of Theory

Lee Benda and Daniel Miller1

Introduction: The Dilemma of Scale a range of scales (Naiman and Bilby 1998, Swanson et al. 1988), it has proven difficult to Landscapes fall into a class of objects that fit the develop general principles on how stochastic and adage: “you know it when you see it, but it’s hard deterministic landscape factors, in combination, to define.” As one looks from a ridge into a for- govern habitat development. One example of this ested valley, words such as ecosystems, interac- limitation is the continuing inability to define natu- tions, disturbance, and connectivity come readily ral disturbance regimes, including the range of to mind. But, it has proven to be another matter variability in aquatic and riparian environments entirely to scientifically define a landscape as a (Benda et al. 1998, Naiman et al. 1995). In prac- whole system of interacting parts that can be tice, this has often led to a preference for single readily integrated into management of natural environmental states and single-value regulatory resources, environmental assessments, water- thresholds by many scientists, managers, and shed restoration, and regulation. This is the di- regulators. lemma of scale. Landscapes comprise thousands of components and processes that are difficult to The lack of quantitative and predictive theory on characterize at any one time. Furthermore, inter- landscape-scale processes has created a depen- actions among landscape components occur over dence on case studies that often have focused on decades to centuries, making it difficult to analyze unique and detailed aspects of landscapes, and them over the short duration of research studies. on classification systems that emphasized spatial The study of landscapes, therefore, poses a com- determinism (i.e., stochastic effects ignored). plicated problem. Moreover, theory absence discourages hypothesis testing and commensurable research efforts Major components of riverine habitats depend on (similar things measured in similar ways). These the supply, routing, and storage of inorganic and problems in combination increase the perceived organic materials that originate from terrestrial difficulty of landscape-scale and environmental sources. A stochastic climate exerts a degree of problems. In sum, the lack of theory hinders the randomness in the supply of sediment and or- scientific analysis of landscapes and therefore ganic debris to channel networks over 10 to 100 management planning at the landscape scale. years; topography and channel-network geometry impose a spatially determined organization in the Role of Theory in the Study of routing and storage of those materials. Hence, aquatic and riparian habitats have both stochas- Landscapes tic and deterministic origins. Studies that have “Theory” refers to an explicit set of rules and pa- incorporated stochastic effects have been re- rameters that are used to describe observed phe- ferred to as disturbance ecology (Pickett and nomena in a quantitative manner and that accord White 1985), temporal hierarchies (Frissel et al. with the empirical record (Gell-Mann 1994, Pop- 1986), pulses (Junk et al. 1989), and landscape per 1972). Theories should make testable predic- dynamics (Benda et al. 1998). Studies focusing tions and hence be in a continual state of evalua- on deterministic aspects are described in terms tion, rejection, and modification (Popper 1972). In of continuums (Vannote et al. 1980), spatial hier- the study of landscapes, theories are generally archies (Frissel et al. 1986), ecotones (Naiman et applied at small spatial and temporal scales (i.e., al. 1988), and classification systems (Montgomery slope stability and sediment transport theories). and Buffington 1997, Rosgen 1995). Despite a The term “concept” in the aquatic sciences gener- sustained interest in ecological processes over ally refers to new and innovative ideas, and it is that class of knowledge where the greatest strides 1 Earth Systems Institute, 1314 NE 43rd St., Suite 206-207, have been made in articulating the multivariate Seattle, WA 98105-5832; Tel: 425-787-1960/206-633-1792; attributes of landscapes. Concepts, however, do e-mail: [email protected]/[email protected] not make testable predictions in the same way

17 theories do because they contain components The Study of Large Numbers of that are not fully or explicitly defined. Concepts Interacting Landscape Processes may play a key role in the development of a discipline because they act as verbal precursors to the Clues for developing inductive or deductive theo- development of quantifiable and testable theories ries at large scales pertinent to landscapes are (Haines-Young and Petch 1986). A third class of found in the scientific disciplines that have already knowledge, “classification systems,” provides an tackled problems involving large numbers of de- organizing framework for grouping items into simi- terministic and stochastic elements. One suc- lar categories. Classification is a powerful tech- cessful strategy has been to represent the inter- nique for developing a common vocabulary and actions of many small-scale processes by larger for arraying physical and biological properties of scale parameters. For example, application of certain watershed elements. Typically, classifica- analytical mechanics to the problem of landslide tion is a precursor to development of theories and prediction views soil as a “continuum,” even laws (Hempel 1966). though soil is composed of many individual grains. For soil, continuum mechanics represents Developing theoretical understanding pertinent to the multitude of millimeter-scale grain-to-grain large scales is a critical step in the coordinated interactions by meter-scale parameters such as and organized study of landscapes. First, land- soil cohesion, bulk density, and soil friction angle. scape-scale theory would encourage the meas- (A similar approach has been applied to problems urements of landscape attributes in similar ways, involving turbulent fluid flow [e.g., fluid mechan- thereby contributing to a regional pursuit of gen- ics]). Analytical mechanics is most effective when eral principles, similar to theories at smaller dealing with problems at relatively small spatial scales. This would tend to counter the notion that and temporal scales, and it runs into difficulty every place is unique and has to be studied when applied at larger scales. For landsliding, the uniquely on its own merits. Second, because land- use of a one-dimensional, infinite-slope model is scape study is fundamentally interdisciplinary, an example of simplifying the interactions of mul- theory would encourage and guide how different tiple three-dimensional unit volumes of soil in- scientific disciplines would need to converge or volved with failures. merge in studies of various phenomena. Third, theory makes it easier and more defensible to Statistical mechanics provides another tech- extrapolate findings from one landscape to an- nique for predicting the behavior of exceedingly other, thereby obviating the need for re-creating large numbers of randomly behaving elements. A the wheel at every location. Fourth, theory allows purely statistical approach can describe the be- “bridges” to be built among incomplete data (ei- havior of gases that contain vast numbers of ran- ther temporally or spatially), a strategy that could domly colliding molecules (James Clark Maxwell be cast in terms of “hypotheses,” but that would [1831-1879] and Ludwig Boltzmann [1844-1906]). allow for more comprehensive understanding. To calculate the macroenergy state of a gas in This would make explicit the gaps in data and response to applied temperature and pressure, understanding and would aid in targeting future molecules are parameterized by probability distri- research priorities. Fifth, general theoretical prin- butions of energy states. As pointed out by Dooge ciples would create a hierarchy of scientific under- (1986), however, there are large differences be- standing in which case studies of processes or tween the statistical mechanical approach that conditions obtained over small spatial and tempo- depends on concepts of energy equilibrium and ral scales would be evaluated in the context of the average conditions, and the nonequilibrium and larger scales that characterize landscapes. These transient conditions manifest in hydrologic and advantages would apply to research in the veg- geomorphic processes that are of interest to sci- etative, geomorphic, hydrologic, and biotic scientists and resource managers. ences; in natural resource management; and in restoration and conservation biology.

18 In the context of these methods, landscapes con- distributions and spatial patterns of sediment and tain too many components to be treated strictly wood flux to streams. Included is how temporal deterministically and too few components to be distributions of material fluxes evolve along a treated purely statistically. Environmental systems channel network owing to asynchronous material that fall between the end members of determin- supply, network geometry, attrition, spatial scale ism and stochastism but that exhibit both charac- (drainage area), and transport and storage re- teristics have been referred to as “intermediate gimes. The derived long-term probability distribu- number systems,” or systems of “organized com- tions of material flux and storage indicate the plexity” (Weinberg 1975). This characterization stochastic and deterministic origins of aquatic and also has been extended to ecological and geo- riparian landforms. Probability distributions also morphological systems (Allen and Starr 1982, define the space-time structure of variability or the Graf 1988). To deal with systems of organized natural disturbance regime. complexity, “systems theory” has been developed The general theory is a work in progress, and (Von Bertalanffy 1968). Application of systems hence there is need for testing and refining gen- theory typically relies on building comprehensive eral principles pertaining to climatic and vegeta- mathematical models that are used to scale up tion disturbances, erosion regimes, sediment analytical descriptions of processes at small routing, and wood recruitment. In addition, new scales to predict the macrobehavior of a system theoretical principles covering riparian vegetation of such processes over larger space and time and the formation of aquatic and valley floor habi- scales. This so-called “upwards approach” has tats are needed. Many of the overarching interac- been applied to the study of certain hydrological tions between stochastic and deterministic problems (Rodriguez-Iturbe and Valdes 1979, landscape factors can be sketched on the back of Roth et al. 1989, Smith and Bretherton 1972). a napkin. Simulation modeling and field studies are needed to make more quantitative and land- Pursuit of General Landscape Theory scape-specific predictions (Benda and Dunne The concept of “organized complexity” is pro- 1997a, 1997b; Benda and Sias 1998). Refer to posed as a framework for unifying random and General Landscape Theory of Organized Com- organized attributes of landscapes that govern the plexity (Benda et al. 1999) for a more thorough flux, storage, and routing of mass between and discussion. within terrestrial and aquatic systems. Random- ness refers to behavior that is not predictable in Potential Applications of Landscape detail (such as climate), but that can be described Theories: in probabilistic terms. Organization refers to spatial regularities or patterns in a landscape and can 1. Guide field studies of landscape-scale pro- include laws of stream ordering and bifurcation cesses. (Horton 1945, Strahler 1952), and systematic 2. Provide context for studies conducted at variations in network geometry, such as decreas- small spatial and temporal scales. ing channel gradient and increasing channel width with increasing drainage area (Leopold et al. 3. Define natural disturbance regimes through 1964). probability and frequency distributions. Pursuit of landscape theory requires landscape- 4. Evaluate environmental change through shifts scale parameters. These include temporal distri- in distribution form (in time or space). butions of the frequency-magnitude characteris- 5. Promote risk assessments that use a probabi- tics of climatic, hydrologic, and geomorphic listic approach. events, and spatial distributions that characterize the attributes of large numbers of landscape ele- 6. Base environmental analyses, resource man- ments (Benda et al. 1998). A general landscape agement planning, environmental regulation, theory is proposed that predicts how mixtures of watershed restoration, and conservation biol- climate, topography, lithology, and vegetation im- ogy on a theoretical foundation. pose overarching constraints on the probability

19 Literature Citations Horton, R.E. 1945. Erosional development of streams and their drainage basins: hydro- Allen, T.F.H.; Starr, T.B. 1982. Hierarchy: per- physical approach to quantitative morphology. spectives for ecological complexity. Chicago, Bulletin of the Geological Society of America. IL: University of Chicago Press. 310 p. 56: 275-370. Benda, L.; Dunne, T. 1997a. Stochastic forcing of Junk, W.J.; Bayley, P.B.; Sparks, R.E. 1989. The sediment routing and storage in channel net- flood-pulse concept in river-floodplain systems. works. Water Resources Research. 33(12): Canadian Special Publication of Fisheries and 2865-2880. Aquatic Sciences. 106: 110-127. Benda, L.; Dunne, T. 1997b. Stochastic forcing Leopold, L.B.; Wolman, M.G.; Miller, J.P. 1964. of sediment supply to the channel network Fluvial processes in geomorphology. San Fran- from landsliding and debris flow. Water Re- cisco: Freeman. 522 p. sources Research. 33(12): 2849-2863. Montgomery, D.R.; Buffington, J.M. 1997. Benda, L.; Miller, D.; Dunne, T. [et al.]. 1998. Channel reach morphology in mountain drain- Dynamic landscape systems. In: Naiman, R.; age basins. Geological Society of America Bilby, R., eds. River ecology and management: Bulletin. 109: 595-611. lessons from the Pacific Coastal Ecoregion. Washington, DC: Springer-Verlag: 261-288. Naiman, R.J.; Bilby R.E., eds. 1998. River ecol- Chapter 12. ogy and management: lessons from the Pacific Coastal Ecoregion. New York: Springer-Verlag. Benda, L.; Miller, D.; Sias, J. [et al.]. 1999. General landscape theory of organized com- Naiman, R.J.; DeCamps, H.; Pastor, J.; plexity. Seattle, WA: Earth Systems Institute. Johnston, C.A. 1988. The potential importance of boundaries to fluvial ecosystems. Benda, L.; Sias, J. 1998. Wood budgeting: land- Journal of the North American Benthological scape controls on wood abundance in Society. 7: 289-306. streams. Seattle, WA: Earth Systems Institute. 70 p. Naiman, R.J.; Magnuson, J.J.; McKnight, D.M. 1995. The freshwater imperative: a research Dooge, J.C. 1986. Looking for hydrologic laws. agenda. Washington, DC: Island Press. 165 p. Water Resources Research. 22: 465-585. Pickett, S.T.A.; White, P.S., eds. 1985. The ecol- Frissell, C.A.; Liss, W.J.; Warren, C.E.; Hurley, ogy of natural disturbance and patch dynam- M.D. 1986. A hierarchical framework for ics. New York: Academic Press. stream habitat classification: viewing streams in a watershed context. Environmental Man- Popper, K.R. 1972. The logic of scientific discov- agement. 10(2): 199-214. ery. 6th ed., rev. London, England: Hutchinson. Gell-Man, M. 1994. The quark and the jaguar: Rodriguez-Iturbe, I.; Valdes, J.B. 1979. The adventures in the simple and complex. New geomorphic structure of hydrologic response. York: W.H. Freeman and Company. 329 p. Water Resources Research. 15(6): 1409– 1420. Graf, W.L. 1988. Fluvial processes in dryland rivers. Berlin, West Germany: Springer. 346 p. Rosgen, D.L. 1995. A stream classification system. In: Johnson, R.R.; Zeibell, C.D.; Patton, Haines-Young, R.; Petch, J. 1986. Physical ge- D.R. [et al.], eds. Riparian ecosystems and ography: its nature and methods. London, their management: reconciling conflicting uses. England: Harper and Row. 230 p. Gen. Tech. Rep. RM-120. Fort Collins, CO: Hempel, C.G. 1966. Philosophy of natural sci- U.S. Department of Agriculture, Forest Service, ence. Upper Saddle River, NJ: Prentice Hall. Rocky Mountain Forest and Range Experiment Station: 91-95.

20 Roth, G.; Siccardi, F.; Rosso, R. 1989. Hydrody- Swanson, F.J.; Kratz, T.K.; Caine, N.; namic description of the erosional develop- Woodmansee, R.G. 1988. Landform effects ment of drainage patterns. Water Resources on ecosystem patterns and processes. Research. 25: 319-332. BioScience. 38(2): 92-98. Smith, T.R.; Bretherton, F.P. 1972. Stability and Vannote, R.L.; Minshall, G.W.; Cummins, K.W. the conservation of mass in drainage basin [et al.]. 1980. The river continuum concept. evolution. Water Resources Research. 8(6): Canadian Journal of Fisheries and Aquatic 1506-1529. Sciences. 37(1): 130-137. Strahler, A.N. 1952. Hypsometric (area-altitude) Von Bertalanffy, L. 1968. General systems analysis of erosional topography. Bulletin of the theory. New York: Braziller. Geological Society of America. 63: 1117-1142. Weinberg, G.M. 1975. An introduction to general systems thinking. New York: Wiley- Interscience.

21 Scale Considerations for Linking Hillslopes to Aquatic Habitats

Robert R. Ziemer1

The title of this conference, “Views From the are we successful in dealing with the full complex- Ridge,” suggests a particular scalar view of is- ity of the issue across disciplines. Traditional ways sues. From the ridge, one obtains a somewhat of looking at problems are either from the top broad but restricted view of the landscape. Cer- down, or from the bottom up. tainly, “Views From Space” would provide a larger spatial overview in which landscape pattern be- Top Down comes a dominant theme. For an aquatic or ripar- The top-down approach (fig. 1) starts with some ian theme, “Views From the Valley” would suggest land use activity, such as logging, grazing, or ur- looking upward to the hillslope and ridges, in con- banization. The next step is to identify the onsite trast to looking down from the ridge. Issues con- changes produced by that activity; that is, how cerning appropriate scale have been prevalent in does that land use activity modify the site—soil, most of the recent landscape assessments, in- vegetation, terrain, slope, and so forth. Then, how cluding the Northwest Forest Plan (FEMAT 1993), are these onsite changes translated into altered the PACFISH (1994) strategy, and, most recently, watershed products? Primary products of altered the Forest Service Roads Analysis (USDA Forest watersheds are water, sediment, organics, chemi- Service 1999) procedure. In all of these efforts, cals, and heat. And finally, how are these products three struggles were common: (1) issue identifica- transported away from the site of disturbance to tion and integration of information across multiple cause some offsite impact? For example, sup- disciplines, (2) appropriate spatial scales, and (3) pose there are logging and associated roads in a appropriate temporal scales. particular watershed. These activities compact the soil, modify the vegetation, and alter the topogra- Issue Identification and Integration of phy by making the slope steeper at road cuts and Information Across Multiple fills. These physical changes can modify runoff Disciplines timing and volume, wood input to streams, surface erosion, and landslides. The result can pro- Several decades ago, some of us thought that it duce changes in peak flow, base flow, water would be a good idea to get a bunch of fishery temperature, channel condition, and sediment. researchers and watershed researchers together Society is more concerned about the conse- for a joint meeting. The joint meeting lasted about quences of these changes offsite: increased 4 hours until someone voiced the opinion that we flooding, increased sedimentation, fewer salmon, had nothing in common to speak about and the and so forth. By looking at the full set of potential meeting broke up into two different rooms: one influences of a land-disturbing activity, a broader room for the biologists and another for the physi- range of potential concerns can be identified than cal scientists. Since that time, interdisciplinary if we simply focused on our favorite impact. work has improved. At least now we occasionally can identify common issues. But still, people con- Bottom Up tinue to struggle with understanding the crosscut- ting complexity within a common issue. For Another equally useful approach (fig. 2), which is example, foresters tend to identify forestry issues a common engineering exercise, is to identify as centered around trees; hydrologists see for- some offsite impact and trace the way back up to estry issues as related to water quality or quantity; find the activity that caused that offsite impact. For biologists see the same forestry issues as revolv- example, if a bridge was washed out, there could ing around birds, salamanders, or fish. Seldom be many potential reasons including increased peak flow, channel erosion, water diversion, bat- 1 Chief research hydrologist, U.S. Department of Agriculture, tering by debris, and so forth. Identification of the Forest Service, Pacific Southwest Research Station, 1700 correct process and successive linkages is impor- Bayview Dr., Arcata, CA 95521; Tel: 707-825-2936; e-mail: tant in order to be successful in preventing future [email protected] failures or to identify the guilty party.

22 Figure 1—The top-down approach starts with a land-disturbing activity, then describes the onsite changes, the subsequent effects of these changes, and finally the consequences (from Ziemer and Reid 1997).

Figure 2—The bottom-up approach starts with an identified consequence (bridge washed out), then describes the important conditions and linkages that could have produced the problem, then the processes that caused the conditions, and finally links to the land-disturbing activities (from Ziemer and Reid 1997).

23 To be most successful, we should analyze the and ocean conditions. The land use and terrestrial issues simultaneously from the top down and conditions include the traditional issues and link- bottom up by linking the land use activity to poten- ages: logging, grazing, agriculture, urbanization, tial offsite impacts, and also by linking identified dams, and so forth, with their associated effects. offsite impacts to potential land use activities The human predation component addresses (fig. 3). sport, commercial, and subsistence fishing. The ocean conditions influence a major portion of the Putting It Together salmon’s life cycle. As an example, let us take “disappearing salmon” The traditional view of the problem (fig. 5) is to as an issue for consideration (fig. 4). If we were to ignore all of this complexity and other influences simply focus on the number of salmon at a par- and focus on the parts that we particularly care ticular point as the appropriate metric of success, about. We select a land use of interest and evalu- we may develop some programs of salmon resto- ate the linkages and pathways between that land ration that are rather silly when the problem is use (logging) and the target concern (disappear- considered within its broader context. For ex- ing salmon). Commonly, we further narrow the ample, we might try to restore habitat above a scope to a specific component, for example, to dam or a culvert where the fish are unable to woody debris. We want to demonstrate that a reach. Or, perhaps the reason the fish numbers change in woody debris has some effect on disap- are low is because they were caught downstream. pearing salmon. So this becomes our top-down By producing diagrams similar to figure 4, we can approach. We only think about how woody debris begin to visualize and understand the complexity is affecting the salmon and we ignore all of the and interactions within the issue of concern. The other influences. process of developing the diagram is more impor- It is common to find that an agency only considers tant that the final diagram itself. In building the those components for which they are directly re- diagram, individuals with different backgrounds sponsible and ignores the potential effects of and focus can identify where their knowledge con- other land uses. For example, a forestry agency tributes to the solution of a single issue. In figure becomes only concerned with the effects of forest 4, there are three major components potentially land management on salmon, while the influence affecting salmon: land use, human predation,

Figure 3—The top-down and bottom-up approaches can be merged into a single- analysis approach by linking the land use activity to potential offsite impacts, and also by linking identified offsite impacts to potential land-use activities.

24 Figure 4—A generalized diagram of some possible important interactions affecting “disappearing salmon” (from Ziemer and Reid 1997).

Figure 5—Example of a typical shortcut that ignores many of the important components in the generalized diagram to link a particular component to “disappearing salmon” (from Ziemer and Reid 1997).

25 of agriculture, urbanization, dams, fishing, and so supply for a small community, evaluating the forth are ignored because the forestry agency is subwatershed directly above the water supply only authorized to regulate logging or manage intake is the appropriate geography and scale. timberland. This focus is appropriate at a later Areas beyond that direct influence are not rel- time when the agency decides upon a program of evant to the problem. If, however, we are dealing action. Unfortunately, such a myopic view often with the effect of a project on anadromous fish, misses the context of the agency’s program within then we are dealing with a much different geo- the larger issue and can lead to uneven regulation graphical and spatial arena. For each of the boxes or to ineffective management actions. and linkages in the disappearing salmon diagram (fig. 4), there are sets of scales that are appropri- The end point of problem simplification is to select ate to that box. For the fish population, the scales some index that directly links the activity to the range from the individual stream reach to the target issue without regard to other influences Pacific Northwest, including the ocean. Salmon (fig. 6). For example, a group working to restore stocks from the Columbia River may compete in salmon runs in the South Fork Trinity River in the ocean for food and resources with salmon northwestern California assumed that their favor- from northwestern California. Anything that ite variable, changes in the volume of large pools changes the competitive advantage of one stock in the mainstem river, was related to the number is important to consider. Further, there may be of returning salmon. The group decided to meas- migratory wandering of fish from one river system ure annual changes in the volume of these large to another. A depleted stock from one river may pools and then to correlate these annual pool result in success of stocks from another river volume changes to logging and road building, because of reduced competition, or vice versa. It which were assumed to produce decreased pool is important to recognize such external forces that volume, and to the amount of future watershed are operating at the large scale outside of the im- rehabilitation, which was assumed to result in mediate frame of reference. Similarly, within a increased pool volume. In other words, pool vol- given river system, it is not possible to evaluate ume was the index that was to tie changes in land the value of improving fish habitat quality at the use to fish. None of the other components or influ- small watershed scale without some understand- ences upon fish numbers were to be evaluated or ing of how habitat along the migratory route influ- considered. The problem was that the group had ences the population. In the extreme example, no information about what was happening to fish improving salmon habitat above a migration bar- downstream and no independent indication that rier will have no effect, because the fish will never there was any relation between fish numbers and be able to use that habitat. pool volume, let alone between land use and pool volume. The appropriate scale or geography depends on the issue to be addressed. Some issues remain Spatial Scale fixed in one location (trees, soil fertility), whereas others are mobile (animals, water, sediment). Individuals who design projects, such as timber Products associated with aquatic issues (water, sales, roads, grazing permits, recreation facilities, heat, chemicals, wood, sediment) tend to move and so forth, are quite accustomed to and com- downslope or downstream and are constrained fortable in dealing with the project or subwater- within defined topographic boundaries. Fish shed scale (fig. 7) that ranges from 10 to a few move upstream and downstream, so for them, thousand acres. Project designers are less accus- watershed boundaries are useful geographic tomed to evaluating the context of that project limits. Terrestrial animals (deer, birds) are not within larger scales. The appropriate size of that constrained by watershed boundaries, and the larger scale depends strongly on the issue being watershed concept is not particularly useful. For considered. If, for example, there is a concern these animals, movement range is a more useful about the effect of a project on the drinking water scale than topographic boundaries.

26 The Index Approach

Altered riparian Vegetation Logging, grazing, vegetation change urbanization, etc.

Less woody More erosion debris ShallowECA, Industrialization TMDL, water etc. Aggradation Higher Altered circulation High water More predation peak and temperature velocities flows

Ocean Habitat Freshwater High water change and estuary temperature Dams Migration Disappearing blockage Sport Salmon Commercial Altered spawning Fishing gravels Subsistence

Figure 6—Occasionally, an index is used to link a land-disturbing activity to the target, ignoring all of the complexity and interactions that may also influence the target (from Ziemer and Reid 1997). (ECA = equivalent clearcut area, TMDL = total maximum daily load.)

Figure 7—A hierarchy of spatial scales.

27 A survey of any geographic area will result in a much shorter than 75 years, often 10 or fewer high variance for most parameters. For example, years. Any 10-year period in figure 9 could find consider a hypothetical survey of 30 streams to conditions ranging from no severe storms to mul- evaluate the risk of mortality of some species of tiple storms. In other words, the temporal scale fish (fig. 8). Some streams have good habitat and needed to adequately represent the significant a low risk of mortality resulting from some action, geomorphic or ecologic drivers is often orders of whereas others will have a high risk. The level of magnitude longer than our monitoring database. “acceptable” risk has two components, biological How does this relate to the level of regulation and and social. If the species is abundant, it may be risk of mortality? Suppose that the average of the biologically and socially acceptable to adopt a streams depicted in figure 8 had a monitoring level of regulation that would overprotect some record of 30 years (fig. 10). The maximum risk streams and underprotect others. As the species of mortality, and perhaps the appropriate level becomes rarer, a higher level of regulation may be of regulation, could differ substantially based on appropriate, depending on the consequences of which period is monitored: for example, years making a judgment error. The problem with regu- 1 through 10, years 11 through 20, or the entire lations that produce or require a generic “designer 30-year record. stream” is that stream systems are dynamic and may require a wide range of evolving habitat con- What is the appropriate time scale to consider? ditions to be productive. The stream systems The answer depends strongly upon the issue. described by Reeves (this volume) require a sub- Different folks or the same folks considering stantial amount of perturbation and resulting pro- different issues look at the problem differently ductivity changes over time. Designing for the (table 1). For those in the corporate world of prof- perceived “ideal” condition in all places all of the its and losses, a quarter (of a year) is an impor- time may lead to a poor stream condition in the tant scale. Corporate well-being 150 years from future. Further, a poor condition today may con- now is often not an important consideration to the tain exactly the components needed for the best board of directors. Politicians like to see programs habitat in the future. that they sponsor put into effect and have some result during their time in office. For politicians, Temporal Scale the election cycle (2, 4, or 6 years) is an important time scale. The length of a human life is an impor- It is well known that “significant” hydrologic or tant time scale for people, and sometimes plan- meteorologic events occur rarely, and the tempo- ning includes several generations, that is, ral distribution of these events is not uniform. This planning cycles ranging from 10 to perhaps 100 presents a problem because most monitoring years. For most people, something that happened activities represent only a short snapshot of the 20 years ago was a long time in the past. With temporal distribution of events. If the long-term some exceptions, such as planning for infrequent distribution was uniform and well behaved, the but catastrophic events such as earthquakes and snapshot may be an adequate representation of floods, something that happens once every 20 the expected population of future events. How- years or so is beyond the immediate concern of ever, if the events are not uniformly distributed most people. However, a 20-year time scale is (fig. 9), then any short period of monitoring can extremely long for an insect species having sev- produce flawed information. For example, as- eral life cycles per year, or extremely short for a sume habitat conditions are monitored on a redwood or bristlecone pine having a life cycle of stream continuously for 75 years, considered by 1,000 years or longer. An individual storm be- most to be an exceedingly long record. If the comes very important for the domestic water user monitoring period ran from year 1 to 75 (fig. 9), who turns on the water tap and finds the water to the conditions represented would be greatly differ- be turbid. Geomorphic events that shape the ent than if the period was from year 75 to 150. stream channel may occur only once a decade, More realistically, most monitoring activities are century, or millennium.

28 Endangered species

Underprotected

Overprotected Abundant species Risk of mortality mortality of Risk Levels of regulation Levels of

Stream identifier

Figure 8—Hypothetical survey of streams for risk of fish mortality and the level of regulation needed to protect fish at two levels of abundance (from Ziemer 1994).

Figure 9—Distribution and magnitude of severe storms during a single 300-year simulation (from Ziemer 1991).

29 Figure 10—Hypothetical risk of fish mortality based on monitoring streams for different periods of time and the effect of cycles or unusual events on the perceived level of regulation needed to protect fish (from Ziemer 1994).

Table 1—Appropriate time scales Entity Period Years

Corporations Quarterly profits and losses 0.25 Politicians Election cycles 2, 4, or 6 Humans Memory of significant events 1 to 20 Humans Lifespan 50 to 100

Insects Life cycle 0.2 to 1 Anadromous fish Life cycle 2 to 4 Humans Life cycle 50 to 100 Trees Life cycle 100 to 1,500

Domestic water user Individual storm 0.1 to 5 Channel adjustments Large storm 1 to 1,000

30 Finally, how one views the world depends strongly The second model (fig. 11, curve b) suggests that on conceptual models about how things operate. a small amount of disturbance in watersheds hav- For example, our belief about how the level of ing the best habitat can result in a rapid decline in watershed disturbance is related to salmon habi- habitat quality. Once the habitat quality is low, tat (fig. 11) has a strong influence on land man- additional disturbance has little incremental effect agement and restoration strategies. The initial on habitat quality. Conversely, curve b suggests assumption for both curves a and b is that the that recovery of habitat quality in heavily disturbed best habitat represents that area having the least watersheds will require a huge effort before any watershed disturbance. In one model (fig. 11, improvement will result. Many past land manage- curve a), a watershed can be increasingly dis- ment plans followed assumptions of curve a. The turbed with little effect on habitat quality until a Northwest Forest Plan (FEMAT 1993) aquatic threshold is reached, beyond which there is a conservation strategy follows the assumptions of precipitous decline in habitat quality. The manage- curve b, that is, to identify and protect those wa- ment objective would be to allow disturbance ac- tersheds that have the best remaining habitat (key tivities to continue until just before the point is watersheds), and to concentrate continued har- reached where habitat quality begins to drop rap- vesting in those areas having the poorest habitat idly. Conversely, curve a suggests that a severely (matrix). It is important to determine which of degraded habitat can be restored with a small these models best represents the relationship reduction in the amount of watershed disturbance. between watershed disturbance and habitat quality.

Figure 11—Two conceptual models of the relation between watershed disturbance and salmonid habitat: (a) habitat quality is not degraded until substantial watershed disturbance is reached; (b) habitat quality is degraded most quickly during initial stages of watershed disturbance (from Ziemer 1997).

31 Conclusion Ziemer, R.R. 1991. An approach to evaluating the long-term effects of land use on landslides, Management and policy strategies to sustain a erosion, and stream channels. In: Proceed- resource depend on physical and biological hy- ings, Japan-U.S. workshop on snow ava- potheses that often are untested. The success or lanche, landslide, debris flow prediction and failure of a particular strategy will depend strongly control. Tsukuba, Japan: Organizing Commit- on how the resource actually responds once that tee of the Japan-U.S. Workshop on snow ava- strategy is applied. Understanding the response lanche, landslide, debris flow prediction and of the resource, in turn, will depend critically on control: 533-542. viewing that resource from the appropriate scale in time and space. Ziemer, R.R. 1994. Cumulative effects assessment impact thresholds: myths and realities. Literature Cited In: Kennedy, A.J., ed. Cumulative effects assessments in Canada: from concept to prac- Forest Ecosystem Management Assessment tice. Edmonton, AB: Alberta Associa- Team [FEMAT]. 1993. Forest ecosystem man- tion of Professional Biologists: 319-326. agement: an ecological, economic, and social assessment. Portland, OR: U.S. Department Ziemer, R.R. 1997. Temporal and spatial scales. of Agriculture; U.S. Department of the Interior In: Williams, J.E.; Wood, C.A.; Dombeck. M.P., [et al.]. eds. Watershed restoration: principles and practices. Bethesda, MD: American Fisheries PACFISH. 1994. Environmental assessment for Society: 80-95. Chapter 6. the implementation of interim strategies for managing anadromous fish-producing water- Ziemer, R.R.; Reid, L.M. 1997. What have we sheds in eastern Oregon and Washington, learned, and what is new in watershed sci- Idaho, and portions of California. Washington, ence? In: Sommarstrom, S., ed. What is wa- DC: U.S. Department of Agriculture, Forest tershed stability? Proceedings, sixth biennial Service; U.S. Department of the Interior, watershed management conference. Report Bureau of Land Management. 68 p. + attach- No. 92. Davis, CA: University of California, ments. Water Resources Center: 43-56. U.S. Department of Agriculture, Forest Ser- vice. 1999. Roads analysis: informing decisions about managing the national forest transportation system. Misc. Rep. FS-643. Washington, DC. 222 p.

32 Landscape Assessment in a Multiownership Province

Thomas A. Spies1

Controversies about forest sustainability, ad- Modeling Study (CLAMS) is a large interdiscipli- vances in the sciences of landscape ecology and nary effort designed to evaluate aggregate ef- ecosystem management, and new tools such as fects of different forest policies on the ecological geographic information systems (GIS) and remote and socioeconomic conditions of the Coast sensing have led natural resource policymakers, Range province as a whole (Spies et al. 2002; planners, and managers from the stand, to the www.fsl.orst.edu/clams). Here, I briefly describe landscape, and to regional scales. As scientists our general approach and present an example of and managers have expanded to these broader a simulation of changing forest landscape condi- scales, they have typically encountered multi- tions over time. I discuss the potential ecological ownership landscapes. The management and consequences of the mosaic of different owner- scientific challenges posed by multiownership ship policies and conclude by identifying some of landscapes are especially complex. Species and the challenges associated with building integrated ecosystems do not recognize legal boundaries regional models. between ownerships, and the landscape dynam- The goal of CLAMS is to develop and evaluate ics of individual ownerships are controlled by a concepts and tools to help understand patterns complex of economic, social, political, and bio- and dynamics of ecosystems at province scales physical forces. The aggregate ecological condi- and to analyze the aggregate ecological and so- tions of landscapes are controlled by the spatial cioeconomic consequences of forest policies for pattern and dynamics of individual owners and different owners. Our approach is based on the ecological interactions among those ownerships. assumption that by knowing landscape structure The dominant disturbance regimes in many land- and dynamics of vegetation we can project conse- scapes are now directly or indirectly controlled by quences of different forest policies for biological human activities. Consequently, to understand and social responses. The major steps in our and predict these anthropogenic disturbances, we approach are: must also understand their linkages to economics, policy, and sociology. Solutions to problems of 1. Build high-resolution spatial models (grain conservation policy and practices for multiowner- size of 0.1 to 10 ha) of current biophysical ship landscapes do not lie in isolated owner-by- conditions (e.g., vegetation, ownership pat- owner planning and management. Broader scale terns, topography, streams) across all owners approaches are needed. Work in multiownership by using Landsat satellite imagery, forest in- landscapes also reveals the need for increased ventory plots, and GIS layers. integration among ecological and social sciences. 2. Conduct surveys and interviews of forest The challenges to conducting integrated regional landowners to determine their expected man- assessments are numerous and frequently not agement intentions (e.g., rotation ages, thin- appreciated by scientists who typically have little ning regimes, riparian management intensity) experience in these types of efforts. under current policies and to develop spatial A group of Pacific Northwest Research Station land use change models based on retrospec- and university scientists is currently involved in tive studies. a research program that is designed to test and 3. Project expected successional changes in evaluate multiownership issues at province forest structure and composition under differ- scales. The Coastal Landscape Analysis and ent management regimes by using stand dynamics models. 1 Research ecologist, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry 4. Build a landscape change simulation system Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR based on forest management intentions and 97331; Tel: 541-750-7354; e-mail: [email protected] forest stand models to project potential landscape structure for 100 to 200 years.

33 5. Develop biophysical response models for agement. For example, a large, rare fire event habitat quality for selected terrestrial and could influence ecological and socioeconomic aquatic vertebrate species, viability of se- systems for decades and centuries. Relatively lected vertebrates species, coarse-filter fine-scale processes such as disease or land- measures of community and landscape con- slides could affect biophysical potential across ditions, historical range of natural variation of large areas. We do not mean to imply that these forest successional stages, and landslide and other processes are not important but our initial debris flow potential. interest is in isolating the effects of management actions. The model should be viewed not as a 6. Develop socioeconomic response models for predictive tool but rather as a computer-based measures of employment and income by eco- experiment to provide insights into the relative nomic sector, timber value and production, effects of different forest policies. recreational opportunities, and contingent value of biological diversity to the public. We developed a prototype of our landscape simulator for the Coast Range province and ran 7. Estimate potential ecological and socioeco- it for a 100-year scenario under current policies nomic consequences of current forest policies (fig. 1). Patterns of current forest condition are by using the landscape simulator and various nonuniformly distributed across ownerships. Cur- response models. rent vegetation patterns in the province are char- 8. Evaluate, test, and revise overall simulator acterized by a predominance of early and mid- system and submodels. sized conifer forests. Forests dominated by trees of the largest size classes (large and very large 9. Provide policymakers, landowners, and the conifers) are rare and restricted primarily to public public with results of spatial projections of lands. Broadleaf forests are less common than consequences and interact with those groups coniferous forests and tend to be concentrated in of people to help inform debate and facilitate riparian areas. Old-growth forest condition (ap- collaborative learning. proximately equivalent to the very large conifer At this point in the project we are simulating only class) is currently a small percentage of the total forest management-related disturbances (e.g., area, and what is remaining is concentrated on clearcutting, partial cutting, thinning) and landslide Bureau of Land Management (BLM) and Forest and debris flow disturbances. We focus on these Service lands in the southwestern portion of the because they are among the most frequent in the province. Little old growth occurs on private land, region, potentially have large impact on measures but some small remnant patches do occur and of biological diversity, and are of great interest in form the basis of Habitat Conservation Plans for current policy debates. We are not simulating the northern spotted owl (Strix occidentalis stochastic disturbances such as wildfire, wind, caurina). Conversely, open (pasturelands, mead- insects, and disease. Studies in the region indi- ows, agricultural lands, and recent clearcuts) and cate that wildfire occurs infrequently (150 to 400 early-successional stages of forest (typically for- years) and its spatial pattern is only weakly con- ests less than 15 to 20 years old) occupy almost trolled by topography, especially for large fire 40 percent of the province and are concentrated events (Impara 1997). Smaller wind and pathogen on private lands. disturbances are quite frequent, but they are diffi- By 50 years into the simulation of future condi- cult to predict and typically occur at patch sizes tions, the pattern of vegetation classes has below our level of spatial resolution for this provin- changed dramatically. Amounts of large-diameter cial study. Climate change is another process that classes have increased, especially on federal we do have in the current model. These other lands, and the spatial pattern of vegetation has ecological processes could be incorporated into begun to resemble the underlying ownership pat- future modeling efforts, either directly in the simu- tern. Young plantations (10 to 30 years old) on lation model or as scenarios (e.g., effects of a federal lands have matured and are beginning to large fire) for comparative analysis. These pro- blend into the matrix of large conifer size classes. cesses could have profound implications to man- On private lands, intensive forest management

34 Figure 1—Current conditions (as of 1995) and simulated changes in forest types of the Coast Range at 50 and 100 years into the future under current policies.

(40- to 50-year rotations) keep these landscapes extremes is a wide range of mixtures of succes- cycling between early successional stages and sional dominance and dispersed or blocked spa- harvest-age timber plantations. By 100 years, the tial patterns. Consequently, we hypothesize that contrasting patterns of vegetation across owner- a new landscape pattern is emerging in this prov- ships are even stronger. ince in which ownership patterns, management Although total amounts of late-successional forest strategies, and boundaries will control patterns of have increased dramatically in the Coast Range biophysical processes more than in the past. The in this simulation, the spatial pattern of these for- ecological and socioeconomic consequences of ests creates considerable potential edge effects changing diversity and spatial pattern are the pri- and spatial pattern interactions, especially on BLM mary focus of our ongoing research efforts. checkerboard lands. The simulations suggest that To summarize, we have learned that recently en- large watersheds of the Coast Range will develop acted forest policies in the Oregon Coast Range into a mosaic of very different landscape types have the potential to create novel landscape pat- based on the amount and spatial pattern of forest terns of vegetation. We hypothesize that in this conditions. These landscapes range from water- dynamic landscape, the combination of complex sheds dominated by late-succesional forest to ownership patterns, contrasting management watersheds dominated by early-successional regimes, and ecological processes create the and mature-forest plantations. Between these spatial interactions that could not be predicted based on information from individual ownerships

35 in isolation from each other. Although simple • The difficulty of developing spatial infor- analysis of ownership patterns indicates a strong mation about landscapes and regions. Spa- potential for aggregate effects in this province, tial information about provinces and regions is more detailed analyses are needed to test the inadequate and will always be flawed. The degree and distribution of these effects. The spa- challenge is to determine when data quality is tial interactions that we expect to have the great- good enough to provide a first approximation est impact on the ecological systems of this at large scales. province are the following: • The value of landscape projections. Spatial • Imbalances and gaps in seral stage distribu- projections of possible future landscapes are tions across environmental strata, including a powerful way to engage policymakers and subecoregions, watersheds, and topographic stakeholders in joint learning efforts. positions. • The challenge of measuring ecological • Gaps in distribution of habitat of relatively effects. We lack quantitative measures of wide-ranging species (such as the northern ecological response. The challenge is to blend spotted owl or salmonids) whose movement empirical, modeling, and expert judgment patterns are at similar scales to ownership approaches to provide working hypotheses for tracts and management allocations within use in model projections and to direct future ownerships. research. • Decline in aquatic habitat quality in stream • The challenge and importance of scale. reaches and watersheds as amounts of large The spatial and temporal scales of ecological, wood are lost because of forest and agricul- policy, and socioeconomic processes and tural management practices along streams measures are typically not the same. Continu- and upslope in landslide and debris flow ous attention to scale is needed to ensure that areas that are sources of large wood for linkages can be made among components. many streams. • The diversity of ways that integration ei- The building of integrated regional models to as- ther happens or not. Integration across disci- sess forest policies is a relatively new endeavor, plines is central to the effort. Although not all and many scientists have little experience with scientists in the team have the time and inter- this type of integrated research, which also may est to attend to integration of the project as a be conducted in an unfamiliar policy and public whole, one or a few leaders must pay close environment. We have learned much about the attention to this process. process of building integrated regional models to • The challenge of conducting science in assess ecological and socioeconomic effects. Our public policy and private landowner envi- lessons learned include: ronment. Applying landscape ecology to large • The importance of problem definition and multiownership areas cannot be done entirely the conceptual model. Without adequate within the walls of a research institution. Sci- problem definition and conceptual framework, entists must interact with policymakers and the process can degenerate into separate stakeholders in new and sometimes uncom- studies that may not meet project goals. fortable ways. These interactions can be time consuming and disruptive to the “normal” pro- • The importance of involving policymakers cess of research. Without them, however, the and identifying policy questions. Without relevance of the effort can be seriously jeopar- incorporating policymakers and specific policy dized. questions at the beginning, the potential relevance of the project will be diminished.

36 These lessons on the process are important for logical and socioeconomic values than the current scientists, policymakers, funding agencies, and policies. Finally, we expect to discover how much management agencies. These types of efforts fine-scale information is needed to answer our require, above all, patience, leadership, long-term questions and understand the dynamics of land- funding, and flexibility to deal with different per- scapes and ecosystems at broad spatial scales. spectives and changing goals. Beyond these challenges and the lessons from Literature Cited the research process, we expect this effort to Impara, P.C. 1997. Spatial and temporal patterns make significant contributions to policymaking of fire in the forests of the central Oregon and the science of landscape assessments. We Coast Range. Corvallis, OR: Oregon State expect that the detailed ecological and spatial University. 354 p. Ph.D. dissertation. structure of the model will help us understand the relative importance of fine-scale management Spies, T.A.; Reeves, G.H.; Burnett, K.M. decisions at broad scales and the importance of [et. al.]. 2002. Assessing the ecological conse- broad-scale policy decisions at fine scales. We quences of forest policies in a multi-ownership expect to determine the locations in a region province in Oregon. In: Liu, J.; Taylor, W.W., that can provide the greatest contribution to eds. Integrating landscape ecology into natural biodiversity goals. We expect to learn if a different resource management. Cambridge, MA: Cam- mix of policies could provide greater overall eco- bridge University Press: 179-277.

37 Implications of Scale on Viability Assessments of Terrestrial Wildlife Species

Martin G. Raphael1

Introduction Key points in this definition include considerations of numbers, distribution, interactions A prudent manager desiring to meet objectives of (i.e., among individuals within populations), habi- ecological sustainability and society’s economic tat, and boundary (i.e., planning area). New draft needs and social desires must consider a variety regulations, as proposed following the report of of scales, both temporal and spatial. This is per- the Committee of Scientists (Federal Register haps an obvious point, but the techniques and 2000 65: 67514-67581), define species viability operational requirements to manage at multiple as: scales are not always so obvious. Large-scale assessments and conservation strategies, such A species consisting of self-sustaining and as the Northwest Forest Plan and the ongoing interacting populations that are well distributed Interior Columbia Basin Ecosystem Management through the species’ range. Self-sustaining Project, illustrate the challenges and frustrations populations are those that are sufficiently of managing at multiple scales. In particular, abundant and have sufficient diversity to dis- meeting the legal requirements for species viabil- play the array of life history strategies and ity, as embodied in the Endangered Species Act forms to provide for their long-term persis- and the National Forest Management Act, has tence and adaptability over time. forced a clear and explicit recognition of the need New considerations introduced in these regula- for a broader view and has crystalized the difficul- tions include adaptability (a concept that includes ties managers face. In this paper, I introduce genetic diversity), and time. some of these challenges and describe recent attempts to meet them. All Scales Count The viability requirement of the National Forest To evaluate whether management will contribute Management Act illustrates the many facets of to or detract from species viability, one must con- species conservation. Viability is defined in the sider a variety of scales, both spatial and tempo- current regulations (36 CFR 219.19) as: ral. The reason for this is simple: animals re- ...a viable population shall be regarded as one spond to their environment at a variety of scales. which has the estimated numbers and distri- For example, consider the marbled murrelet bution of reproductive individuals to insure its (Brachyrhamphus marmoratus). This bird nests continued existence in the planning area. In on large-diameter branches of large-diameter order to insure that viable populations will be coniferous trees, usually within patches of older maintained, habitat must be provided to sup- forest. At the microsite level, a tree must have a port, at least, a minimum number of reproduc- sufficiently large limb to support the bird’s nest. At tive individuals and that habitat must be well a slightly larger scale, the important elements are distributed so that those individuals can inter- the nest tree itself and the structure of the canopy act with others in the planning area. surrounding that tree. A nesting murrelet feeds in coastal waters and makes daily trips to the nest to 1 Research wildlife biologist, U.S.Department of Agriculture care for its young. The canopy must allow space Forest Service, Pacific Northwest Research Station, Forestry for a flight path into the nest. At the same time, Sciences Laboratory, 3625 93rd Ave. SW, Olympia, WA cover must be sufficient to protect the nest from 98512-9193; Tel: 360-753-7662; e-mail: [email protected] potential predators. Therefore, the structure of the nest stand is also important as the size and shape of that stand may influence the abundance of predators and the susceptibility of the nest to predation. Over a broader scale, one must consider

38 the total extent of suitable nesting habitat, as this used to illustrate the importance of consideration determines the size of the murrelet population and of many scales. As evidenced in figure 1, spotted its overall distribution over a landscape. owls choose nest sites in concentrated patches of nesting habitat. This is suggested by the greater As illustrated in this example, several levels of difference between nest sites and random sites at habitat selection exist. The smallest scale is the smallest circle size. As circles around nests where individual animals obtain resources at grow larger, the differences between amounts of any point in time. It is where a nest is located or habitat in nest sites and random sites grow where a prey item is captured–a specific microsite smaller. At very large sizes, both nesting circles or habitat element. An intermediate scale is an and random circles would reflect average condi- animal’s home range, the geographic area an tions in the broader landscape. This example animal traverses to obtain resources during a shows that smaller scales reveal important bio- specified period. A description of habitat condi- logical relationships. In turn, larger scales are tions within a home range might include the ex- necessary to describe population phenomena, so tent of various cover types, their arrangement both scales are important in assessing habitat and pattern, and the composition of each of the conditions. types of cover. A larger scale encompasses the set of conditions that support a population of ani- Another important scale consideration is the effect mals over time. The size of the area to consider of study area boundary on measures of landscape is determined by the extent of the population or pattern. Using the example in figure 1, note that may be an arbitrary definition covering a specified the smallest circular area contains only one patch number of individuals from within a larger popula- of habitat and that the boundary defined by the tion. At an even broader scale, patterns of bio- circle cuts off part of that patch. A measure of diversity might be described. This scale covers patch size is biased because the circle is too the geographic area needed to support a commu- small to contain the whole patch. Similarly, meas- nity of species, each of which is represented ures of edge can be biased because artificial by populations. In this case, the area is large edges are created around the boundary of the enough to contain a variety of cover types and study area. This suggests that calculation of structures, each of sufficient extent to support measures of landscape pattern should be done these populations. on a landscape that is much larger than the average patch size in that landscape, perhaps on the Research on habitat relationships of the northern order of 10 to 50 times as large as the average spotted owl (Strix occidentalis caurina) can be patch size.

50 45 40 OwlOwl sites Sites RandomRandom sites Sites 35 30 25 20 Percent Habitat Percent 15 10 Habitat (percentage of area) of area) (percentage Habitat 00 200200 400400 600600 800 10001,000 12001,200 CircularCircular area Area (hectares) (ha)

Figure 1—Amount of spotted owl nesting habitat in circular areas of varying size. The figure on the left is a comparison of mean (+ standard error) percentage of habitat in a sample of 89 nest sites and 100 randomly located sites, Coast Range, Oregon. The figure on the right is a sample of one site showing habitat (dark) and a nested set of circular areas.

39 The temporal dimension is also important. Spe- Species Viability in the Interior cies viability cannot be assessed at a single point Columbia River Basin in time. At any particular time, a species (or population) exists or it does not. Land managers are The interior Columbia River basin presents an- generally more interested in the future. A typical other good illustration of the importance of scale question is, If I manage the land in some manner, in evaluating viability of terrestrial wildlife (fig. 2). is it likely the species will persist into the future? In this case, we2 assessed habitat conditions for Time can be assessed relative to generation time a set of individual species at two geographic of the species in question. Short-lived species scales, the subwatershed and the entire basin. (e.g., most insects or annual plants) experience At the subwatershed scale, we estimated the total many generations over a decade; longer lived extent of primary habitat within each of about species may take a century or more for multiple 7,000 subwatersheds that make up the basin. generations to pass. Viability assessments of Quality of habitat was adjusted by using models vertebrates are often considered over hundreds that accounted for more specific habitat elements of years. that might affect populations in that watershed.

2 Raphael, M.G.; Wisdom, M.J.; Rowland, M.M. [et al.]. 2001. Status and trend of habitats of terrestrial vertebrates in relation to land management in the interior Columbia River basin. In: Forest ecology and management. Elsevier Science B.V. 153(1-3): 63-87.

Figure 2—Trend in habitat of the pygmy nuthatch from historical to current to hypothetical future in the interior Columbia River basin.

40 We then displayed the geographic distribution of As shown in this example, trend in any one habitat quality over a species’ range at three peri- subwatershed is not particularly informative in ods in time (fig. 2): historically (circa 100 years evaluating viability. It is necessary to take a much ago), currently (average conditions over the past broader view encompassing the species range to decade), and future (100 years into the future gain such insight. But conditions within each under a supplemental draft environmental impact subwatershed are critical for building an aggre- statement alternative). We used three general gate picture of habitat suitability for a given spe- indices to evaluate, at the broad scale, condition cies over an extensive landscape. of habitat as it might influence species viability at each period. We summarized the total amount Conclusion of habitat (actually a weighted average of habitat quality relative to historical conditions), percent- Scale is a vitally important consideration in as- age of historical range that is occupied (extent), sessing species viability. No one scale is suffi- and connectivity of habitat. By using Bayesian cient. We need to consider all scales, from local methods and conditional probability tables, these to regional, and each is important in answering three indices were combined (see footnote 2 for different questions. detailed methods) to summarize the likelihood of conditions that would support the species at each time.

41 Projecting Potential Landscape Dynamics: Issues and Challenges

Ronald P. Neilson1

Natural resource science and management seem infiltration. Others are intermediate, or landscape to have entered the “age of the landscape.” His- scale, such as metapopulation dynamics or seed torically, resources have been studied and man- dispersal. Few processes are truly large scale. aged from a stand, or site, perspective and have Most large-scale patterns are recognized as generally been approached as single-issue prob- “emergent properties,” that is, they arise from lems. Forests, water, fish, and biodiversity were small-scale processes and interactions yet reveal treated largely as separate topics. The recent shift coherent large-scale patterns. The distributions of to a landscape perspective is an explicit recogni- species and ecological zones and the regional tion that the world is far more complex than can patterns of species diversity are examples of be comprehended from strictly site-based, single- emergent properties. Such large-scale emergent issue studies. Thus, scientists and managers are properties often are correlated with large-scale challenged to address resource issues at multiple climate patterns. However, the regional climate scales: site, landscape, regional, national, interna- patterns are themselves emergent properties of tional, and integrated across multiple resources. the small-scale fluid dynamics of the atmosphere, Examples of the latter include vegetation and playing out on a rotating sphere and interacting wildlife habitat interactions; upland vegetation with oceans, continents, mountains, and other dynamics affecting water quantity and quality; surface features, including vegetation. climate, vegetation, pest, disease, and fire inter- Correlations between large-scale resource pat- actions; ecosystem productivity and biodiversity; terns and climate are useful in calling attention to and many others. Natural resource management the relationship and point to likely causation. If we has become so pervasive that it is comparable in could view the climate as a very slowly changing scale to most large-scale natural patterns of re- phenomenon, for example, glacial-interglacial source structure and function. cycles, then we could largely ignore mechanisms My intent in this discussion is to point out some of of interaction between resources and climate and the prevailing issues and challenges facing re- simply focus on the resource processes. How- source managers and scientists alike in predicting ever, we now know that the world is heating up or “forecasting” possible trajectories of such com- and that perhaps half of the current rate of heating plex landscapes. Under the assumption that the is human caused (Trenberth 2001). We cannot climate and land use patterns of the future will be say with certainty what the future climate holds, different from the past, I presume that a process- only that we cannot assume the status quo. Thus, based modeling approach is required, that is, an we must strive to understand and predict resource approach that is robust to changing environmental interactions and future trajectories from first prin- conditions and that can comfortably incorporate ciples, which take us to the small scale of dy- increasing levels of complexity. The discussion namic processes. Yet, we must acquire predictive reflects my personal views and is not intended to capabilities of future resource trajectories over be a complete review of the issues and technolo- large spatial extents. It is ultimately the regional gies available. emergent properties of resources for which we must manage. The challenge of the day, there- Natural resource issues occur across a range of fore, is to develop predictive capability over very spatial and temporal scales. Most processes are large spatial extents from small-scale processes. small scale, for example leaf physiology or water

1 Bioclimatologist, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR 97331; Tel: 541-750-7303; e-mail: [email protected]

42 A Framework for Science and scales while relying on a core modeling paradigm Management that carries through all scales. Thus, the core to all of the scales is the site-based model, which Scientists appear to be converging on a three- must incorporate such things as leaf physiology; scale structure for developing model-based pre- competition among life forms for light, water, and dictive capabilities of resource trajectories and nutrients; soil hydrology; and other processes. multiresource interactions. These are the site, or At each unique scale, scale-specific phenomena stand scale, spanning a few tens of meters; the likely must be included, but the approach may landscape scale, spanning a few to tens of kilo- vary depending on whether it is top down or meters; and the regional or continental scale, bottom up. Many forest simulation and manage- spanning hundreds to thousands of kilometers. ment models do not require the physiological, Approaches to these three scales are either top biogeochemical, and hydrological processes down (empirically based) or bottom up (process (Crookston and Stage 1999). These are empiri- based), being distinguished primarily by the reso- cally based models that cannot accept a chang- lution of the simulation grid. Stands may be con- ing climate as input. I thus restrict my comments sidered as spatially homogeneous at a scale to to the process-based approach to ecosystem about 10 meters and may be treated as point modeling. simulations, e.g., traditional stand or “gap” models (Shugart and Smith 1996). Stands are also often The unique aspect of global ecosystem modeling simulated as spatially explicit with a grid resolution in comparison to more traditional ecological mod- of perhaps 1 meter, but this is less common eling is that the primary points of interest are the (Miller and Urban 1999, Peters 2000). Each grid emergent, large-scale spatial patterns and their cell contains a full model simulation and may or dynamics. Large-scale biogeographic modeling may not include cell-to-cell interactions. The spa- can be simulated from small-scale processes tially explicit approach is bottom up; the spatially (leaf to landscape), but calibrated to large-scale aggregated approach is top down. The stand or biogeographic and hydrologic patterns (Neilson point model can be placed on a grid for spatially 1995). The challenge is to find the simplest model explicit simulations of heterogeneous landscapes structure that is sufficient to capture the neces- with a grid resolution of perhaps tens of meters. sary processes at all the appropriate scales. One Regional-scale models are of necessity coarse may find that a particular mathematical represen- grid, ranging in grid resolution from about 1 km up tation of a process works fine in most locations, to about 50 km or more. Clearly, even the highest but needs to be recalibrated at each. Presumably, resolution regional model must absorb consider- however, if one is appropriately representing the able spatial heterogeneity within each grid cell. processes, a single calibration will suffice in all locations. For example, I used large-scale biogeo- Independent models have been developed at all graphic and hydrologic patterns as simultaneous of these scales; however, there is little coherency constraints to help determine both model struc- among those differently scaled models. For ex- ture and calibration (Neilson 1995). ample, traditional “gap” models, which simulate succession in a small opening, such as a single- In constructing the Mapped Atmosphere-Plant- tree blowdown, use empirical curves based on Soil System (MAPSS) vegetation distribution degree-day sums (temperature sums over time) model, I coupled the simulation of leaf area index (Botkin et al. 1972), whereas other models might (LAI, the area of leaves per unit area of ground) to use elaborate physiologically based processes for the simulation of site hydrology (Neilson 1995). the same purpose (Running and Hunt 1994). In The LAI was used, in part, to define boundaries or the extreme, such models can produce opposite ecotones between different vegetation types, such responses to external forcing, such as climate as that between forest and savanna, thus provid- change, even though both may be equally well ing the biogeographic constraint at the regional calibrated to current conditions (Loehle and and continental scales. The simulation of LAI was LeBlanc 1996). dependent upon the accurate simulation of sea- sonal soil moisture and its distribution through the One approach to building a verifiably accurate soil profile, as transpiration is a function of leaf regional-scale model is to take advantage of the area, stomatal conductance, and vertical root top-down and bottom-up approaches across all distribution. The hydrologic module was calibrated

43 against locally observed runoff in contrasting re- finest scale of interest. This is the inverse of the gions across the United States and was initially natural landscape, where the large-scale patterns structured quite simply. However, the simple are emergent from below and constrained from structure required unique calibrations for each above. Management emerges from the top and is region of the country. I incrementally increased implemented below. However, top-down manage- the complexity of the hydrology model until I found ment also clearly is constrained by the bottom-up the minimum complexity that would accurately spatial resource availability and condition. Nested- simulate runoff in all the diverse regions of the scale management modeling can be vertically country with a single calibration, and which also integrated with the nested-scale ecological model- accurately simulated the LAI to provide proper ing. Some of the issues and challenges in con- mapping of vegetation boundaries. Thus, the structing the ecological side of this dual hierarchy model was simultaneously constrained by both are addressed below. vegetation and hydrologic patterns at the large scale of emergent properties, but was based on Subcell Heterogeneity: Representing the simulation of small-scale processes. the Landscape in Coarse Grids A robust framework of science can be constructed Much of the challenge of large-scale resource by combining the top-down and bottom-up ap- management revolves around the opposing con- proaches. Each spatial scale can be represented cepts of spatial heterogeneity and homogeneity. from both perspectives. At the top of the hierar- For example, as one moves from wet to dry along chy, grid cells are sufficiently coarse to simulate an aridity gradient, the density of a forest will thin wall-to-wall coverage but with uncertainties owing to a point where a grassy understory just begins to subcell heterogeneity, because each grid cell is to be supportable because of enhanced under- usually treated as being spatially homogeneous. story light. This is an unstable threshold condition Higher resolution grids can be used; however, the because new processes suddenly enter the sys- finer scales would only be feasible over smaller tem forcing it away from the point of transition as spatial extents, that is within a sampling frame- follows. Introduction of an understory entrains work over specific watersheds with subplots and positive feedbacks through competition for water, so on. Each scale can inform the other; process which further thins the canopy overstory. Addi- constraints can be imposed through all levels tional feedbacks through fire can thin the over- from a common small-scale theory, and large- story even more, allowing yet more grass and scale constraints, such as biogeographic patterns, more fire until equilibrium is reached. Thus, the can be imposed from above. Thus, each level has system undergoes a transition from a spatially the ability to peer up or down one level, and the homogeneous entity to a heterogeneous one. structure can be moved up and down levels for as Along this hypothetical aridity gradient, with no many levels as are of interest. Yet, the core pro- topographic complexity, there is an endogenous cesses are applied at all spatial scales, thus en- shift from a homogeneous system (forest) at the suring that the patterns and sensitivities of the wet end to a heterogeneous system (savanna) applications at different scales will be fully inter- with increasing aridity and back to a homoge- nally consistent. neous system (grassland) with further increases Historical management practices can be similarly in aridity. The system has also gone through scale approached, spatially explicit at a scale of focus, transitions, from coarse-scale forest, to fine-scale statistical below, and deterministic above. For tree-grass patches, to coarse-scale grassland. example, at the national scale, targets might be Transitions between these physiognomic shifts in set for harvest or fuel reduction practices, but heterogeneity are generally termed ecotones and their locations are not determined. These targets occur at the point of finest scale interfingering might be further dissected to specific regions, pattern between the adjacent homogeneous providing a spatially determined component, but types. within each domain, the cutting or thinning repre- If we interject topographic complexity to the above sents a statistical target and is not spatially deter- moisture gradient, the spatial disposition of eco- mined. This process can continue iteratively, until tones both horizontally and with elevation can final operations are spatially determined at the become quite complex along both elevational and

44 horizontal temperature and moisture gradients. A standing of the landscape patterns. The empirical transect along the west slope of the Rocky Moun- analysis that led to the description of this wedge tains from southern Idaho to the Mexico border of ecotones was based on a set of nested-scale illustrates the complex shifts in elevational eco- experimental seedling transplants along environ- tones along latitudinal temperature and moisture mental gradients at scales of meters (shrub to gradients (Neilson 2003). Winter precipitation intershrub), tens of meters (landscape unique decreases from north to south, but summer pre- geomorphic positions), hundreds of meters (el- cipitation increases from north to south in the form evation), and hundreds of kilometers (regional) of the Arizona monsoon. In combination, the win- (Neilson and Wullstein 1983). Yet, simulations at ter temperature and summer moisture gradients the relatively coarse scale of 10-km resolution create a latitudinal “wedge” of ecotones, upper (Neilson 1995) were able to elicit the same re- elevational ecotones increasing with decreasing gional gradients in ecotones, providing inferences latitude and lower elevational ecotones decreas- to spatial patterns and processes at much smaller ing with decreasing latitude. In the southern part than grid-cell resolution. of the transect, the wide elevational separation of Such convergence of ecotones may tend to occur ecotones creates the classic ecosystem zonation where steep airmass gradients converge. I pro- patterns described by Whittaker and Niering pose that these “nodes” of airmass convergence (1965) on the Santa Catalina Mountains, Arizona, drive a rescaling of ecological gradients, which is and earlier by Merriam (1890) on the San Fran- most apparent at the landscape scale. At a dis- cisco Mountains, Arizona. At the northern part of tance from these nodes, homogeneous vegeta- the transect, however, the elevational ecotones tion occurs in large patches, but near the nodes, converge, creating a tremendous amount of spa- smaller patches of vegetation intermix on micro- tial diversity across microhabitats within compara- sites dictated by topography and soil (Neilson et tively small landscapes. The result is a spatial al. 1992). But, perhaps most interesting, is the pattern of complexity through the region that con- possibility of inferring landscape-scale patterns tains both vertical and horizontal gradients of di- from the coarse-scale patterns simulated by the versity. Peet (1978) described a similar gradient models. Such top-down inference of landscape- along the east slope of the Rocky Mountains. scale pattern must, of course, be validated with It is well recognized that diversity tends to in- direct simulation of spatially explicit landscape crease at ecotones as the dominant organisms patterns across regional gradients. from neighboring domains intermingle into com- Much of the art in landscape simulation arises plex combinations (Hansen et al. 1992). Trees in the challenge to validate innovative top-down and grass, for example, interdigitate at the prairie- modeling techniques with direct, bottom-up simu- forest ecotone, enhancing diversity. The same lations and ground- or satellite-based data. The type of interdigitation and spatial diversity gradi- question becomes one of how to represent and ents occur at elevational ecotones at the southern simulate subcell patterns and dynamics accu- end of the Rocky Mountain transect, for example rately, but without having to do the direct simula- in the Santa Catalina Mountains. At the northern tion at all locations. A topographically induced end of the transect, however, with the spatial con- mosaic of forests and grasslands would appear vergence of ecotones, the different vegetation as a savanna in a large grid cell. The savanna zones sort out on unique topographic slopes, as- may provide sufficient accuracy for some issues, pects, and soils, compressing the interdigitation of but for others, it clearly will not. The spatial het- vegetation from the macroscale to the microscale erogeneity can be handled through explicit bot- and creating a wholly new elevational zonation tom-up simulation of the complex terrain, or as pattern. unique simulations of each patch type as an Thus, attempts to understand the patterns of al- aggregated homogeneous entity. There are also pha, beta, and gamma diversity (local, gradient, innovative ways to explicitly recognize heteroge- and regional) at, for example, only one end of the neity in the aboveground components at the land- Rocky Mountain transect, would be only partially scape scale, while preserving a more homogene- revealing and would provide little general under- ous belowground competitive environment.

45 Fires and other disturbances in the landscape global. Most large-scale modelers are attempting produce significant problems for large-scale simu- to incorporate the important processes that occur lations with coarse grids. Fires create a mosaic of at all three scales: patch (competition, gas ex- uneven-aged patches, with new patches being change); landscape (fire, dynamic heterogeneity); created as often as each year in some cases. and global (emergent, spatial pattern). It will be There are numerous differences in the vegetation very important for practitioners working within any of an area 1 year after fire and 15 years after fire. one of these three modeling communities to coor- In contrast, the differences in vegetation 100 dinate closely with those working at the other years and 115 years after fire are minimal if the scales. Patch models built around one type of climate and substrate are the same. Thus, one ecosystem or in one region may not be well struc- approach is to allow creation of new patches each tured for working in other systems or regions. year and track them individually, but as they be- Consistency of process should be maintained come increasingly similar with age, merge them across scales. If models are to be nested or back together. Initially in the course of model de- linked across scales, then their processes should velopment, these patches, in an otherwise homo- be based upon the same theoretical underpin- geneous grid cell, might be noninteracting and nings, or they may not translate well across only be represented uniquely by their areas and scales. An area of research that I believe may ages. In grid cells with complex terrain, these have some potential, but that remains largely patches could be maintained on unique soils and untapped, is the possibility of downscaling from with unique climates, but again nonspatially. Even- regional to landscape patterns by using coarse- tually, some level of interaction among patches scale information, as in the relation between spa- could be implemented, but still without spatially tial ecotone convergence and landscape diversity explicit representation within the cell. Spatially patterns. explicit simulations on fine-scale grids, coupled The key points of this discussion serve to empha- with ground and satellite data would assist in the size the importance of accurate simulation of eco- validation of these top-down approaches. system constraints at all relevant scales. Under a Topographically induced spatial heterogeneity rapidly changing climate and with changing physi- presents additional difficulties when integrating ology under elevated CO2, constraints normally across resources, such as between vegetation assumed to be stationary must now be assumed and hydrologic processes. Hydrologic processes to be dynamic and must be explicitly simulated. are strongly coupled to vegetation processes and Heterogeneous landscapes are among the most span scales from local infiltration processes to complex, yet globally among the most common regional river routing, yet most of the physics oc- types of ecosystems. Accurate simulation of land- curs at very fine scales. A bottom-up modeling scape patterns and processes under global approach will couple the vegetation and hydro- change requires attention to organism-level and logic point-scale processes and also account for lower processes within the constraints of biome- the horizontal flows of surface and subsurface level dynamic biogeography. A carefully crafted, water. Thus, accurate representation of riparian hierarchical framework with dual-scale represen- zones and stream hydrology is possible. Repre- tation at each level can provide the structure for sentation of these interactions in coarse grids also integration among natural resources and between may be possible along the lines described above, management and natural processes in a rapidly but would require direct linkage of subgrid patch changing environment. types for routing of water between upland and lowland parcels. English Equivalents Conclusions 1 meter (m) = 3.28 feet 1 kilometer (km) = 0.62 miles Current modeling approaches for natural resource management often are organized around three different scales: patch, landscape, and regional to

46 Literature Cited Neilson, R.P.; King, G.A.; DeVelice, R.L.; Lenihan, J.M. 1992. Regional and local veg- Botkin, D.B.; Janak, J.F.; Wallis, J.R. 1972. etation patterns: the responses of vegetation Rationale, limitation, and assumptions of a diversity to subcontinental air masses. In: Northeastern forest growth simulator. IBM Hansen, A.J.; di Castri, F., eds. Landscape Journal of Research and Development. 16(2): boundaries. New York: Springer Verlag: 129- 101-116. 149. Crookston, N.L.; Stage, A.R. 1999. Percent Neilson, R.P.; Wullstein, L.H. 1983. Biogeogra- canopy cover and stand structure statistics phy of two southwest American oaks in relation from the Forest Vegetation Simulator. Gen. to atmospheric dynamics. Journal of Biogeog- Tech. Rep. RMRS-GTR-24. Fort Collins, CO: raphy. 10: 275-297. U.S Department of Agriculture, Forest Service, Rocky Mountain Research Station. Peet, R.K. 1978. Latitudinal variation in southern Rocky Mountain forests. Journal of Biogeogra- Hansen, A.J.; Risser, P.G.; di Castri, F. 1992. phy. 5: 275-289. Epilogue: biodiversity and ecological flows across ecotones. In: Hansen, A.J.; di Castri, F., Peters, D. 2000. Climatic variation and simulated eds. Landscape boundaries: consequences for patterns in seedling establishment of two domi- biotic diversity and ecological flows. New York: nant grasses at semi-arid—arid grassland Springer-Verlag: 423-438. ecotone. Journal of Vegetation Science. 11: 493-504. Loehle, C.; LeBlanc, D. 1996. Model-based assessments of climate change effects on for- Running, S.W.; Hunt, E.R. 1994. Generalization ests: a critical review. Ecological Modelling. 90: of a forest ecosystem process model for other 1-31. biomes, biome-BGC, and an application for global-scale models. In: Ehleringer, J.R.; Field, Merriam, C.H. 1890. Results of a biological sur- C., eds. Scaling processes between leaf and vey of the San Francisco Mountains region and landscape levels. San Diego: Academic Press: desert of the Little Colorado, Arizona. North 141-158. American Fauna. 3: 1-136. Shugart, H.H.; Smith, T.M. 1996. A review of Miller, C.; Urban, D.L. 1999. A model of surface forest patch models and their application to fire, climate and forest pattern in the Sierra global change research. Climatic Change. 34: Nevada, California. Ecological Modelling. 114: 131-153. 113-135. Trenberth, K.E. 2001. Stronger evidence of hu- Neilson, R. 2003. The importance of precipitation man influences on climate: the 2001 IPCC seasonality in controlling vegetation distribu- assessment. Environment. 43: 8-19. tion. In: Weltzin, J.; McPherson, G., eds. Pre- cipitation and terrestrial ecosystems. Tucson, Whittaker, R.H.; Niering, W.A. 1965. Vegetation AZ: University of Arizona Press: 47-71. of the Santa Catalina Mountains, Arizona. (II) A gradient analysis of the south slope. Ecology. Neilson, R.P. 1995. A model for predicting conti- 46: 429-452. nental-scale vegetation distribution and water balance. Ecological Applications. 5: 362-385.

47 Scaling and Landscape Dynamics in Northern Ecosystems

Marilyn Walker1

This manuscript overviews issues of scale in rela- B.S.P.) is also present in open, sparse stands. tion to the core research of the USDA Forest Ser- There is a gradient between full-scale wetlands vice, Pacific Northwest Research Station (PNW) with few trees to fully closed forests without Boreal Ecology Cooperative Research Unit lo- permafrost. cated in Fairbanks, Alaska. The unit’s main re- The need to explicitly recognize and account for search focus is the long-term ecology of boreal scale in ecosystem studies has become increas- forests in Alaska’s interior region. Other ongoing ingly appreciated in the past two decades. Studies and related projects focus on pattern and change that examine pattern or process in space must in tundra and on effects of silvicultural practices consider explicitly the sources of variation that are on white spruce (Picea glauca (Moench) Voss; either secondary or primary to the research. This the major economically important tree species) is not a new idea; it is the foundation of biogeog- regrowth following various cutting and clearing raphy and classical plant ecology, both of which treatments. Alaska’s interior region and North look at patterns of organisms in space and at- Slope have experienced a warming trend over tempt to determine the causal mechanisms be- the past two decades that has been unprec- hind those patterns. What has become increas- edented in the last century (Barber et al. 1998, ingly recognized is how closely together temporal Hammond and Yarie 1996, Juday et al. 1998, and spatial patterns are tied in most systems, and Serreze et al. 2000). Understanding how key pat- also how traditional ecological hierarchies do not terns and processes scale in time and space is map directly to more fundamental hierarchies absolutely critical to understanding how this sys- (Allen and Hoekstra 1992, Allen and Starr 1982, tem may respond to changing temperatures. O’Neill et al. 1986). Because all landscapes have An alternative title for this paper, “Boreal forest– some degree of dynamism, the spatial patterns at Tundra with trees?”, makes reference to the simi- any point in time are the result of multiple pro- larities between tundra vegetation and taiga forest cesses operating at multiple time scales (Delcourt understory. The understory vegetation in flat, et al. 1983). The National Science Foundation’s low-lying areas in northern boreal forests is re- (NSF) Long-Term Ecological Research (LTER) markably similar in both composition and struc- Program addresses the problem of temporal ture to tussock tundra vegetation (Jorgenson et al. scales directly by focusing research resources at 1999, 2001; Walker et al. 1994). This type is selected sites over periods sufficient for explicit codominated by sedges (Eriophorum vaginatum study of decadal-scale pattern (Franklin 1987). L., Carex lugens Holm) and shrubs (Betula nana The Bonanza Creek and Caribou-Poker Creeks L. or B. glandulosa Michx., Ledum palustre L. ssp. LTER site in the Fairbanks region is one of six decumbens (Ait.) Hulten or L. groenlandicum LTER sites nationally that are cofunded by NSF Oeder, Salix L. spp.) and underlain by complete and USDA, and is one of two such sites within coverage of mosses and lichens (Sphagnum L. PNW. spp., Hylocomium splendens (Hedw.) Schimp. Northern ecosystems represent opportunities in B.S.G., Pleurozium schreberi (Brid.) Mitt.). In to examine natural processes at their limits. By the taiga, black spruce (Picea mariana (P. Mill.) focusing at these limits, we can better understand these processes and their roles in complex situa- 1 Unit leader, U.S. Department of Agriculture, Forest Service, tions where many factors are operating simulta- Pacific Northwest Research Station, Boreal Ecology neously. What is most distinctive about these Cooperative Research Unit, P.O. Box 756780, University of Alaska Fairbanks, Fairbanks, AK 99775-6780; Tel: 907-474- systems as a whole is that a single factor, tem- 2424; e-mail: [email protected] perature, controls essentially all other processes (Chernov and Matveyeva 1997). Therefore changes in this single factor will affect almost all

48 processes and patterns. I examine the interac- gradient do not follow a classic life-zone pattern tions between climate, species abundance and of movement up and down an altitudinal gradient. distribution, and disturbance regimes along a Instead, trees and shrubs first come into the land- gradient from the northern limit of plant growth to scape at midslope positions (fig. 2) (Van Cleve the most northern forests (fig. 1). This gradient and Viereck 1981). The presence of shallow covers key thresholds of biodiversity, including the permafrost results in a sealed hydrologic system first introduction of woody-dominated ecosystems and drainage conditions unsuitable for tree growth in the low arctic to the first dominance of trees in at lower slope positions. The exception is black the northern boreal forests. The introduction of spruce, Picea mariana, which grows in open new functional types into the system, first shrubs stands in permafrosted lowlands or on north-fac- and then trees, has dramatic impacts on temporal ing slopes. and spatial patterns, and on how disturbance and The introduction of trees into the landscape has vegetation interact. an overridingly important influence on the spatial Vegetation zonation within the arctic regions has and temporal domains of the systems (fig. 3). The been described by many authors, and various main differences between the two systems are: approaches have been used to describe these 1. Insects and herbivores have a more consis- geobotanical zones (Walker 2000). Although tent and extensive role in the boreal system. slightly different criteria can be used resulting in different outcomes, within Alaska it is useful to 2. The roles of fire and flooding are out of scale recognize three broad zones defined by summer relative to other disturbances in the boreal temperature and dominant plant structural types. system, with fire being the more significant. Table 1 lists characteristics of the three major 3. The formation and melting of ice wedges is arctic zones along with transitional and fully boreal a significant disturbance in the Arctic, zones. There are multiple processes and charac- whereas in the current boreal climate, the teristics associated and interacting with these main effect is surface melting of permafrost structures. Species richness is very tightly corre- (thermokarst), with less recovery. Ice wedge lated with temperature at multiple scales in these formation is on a scale small enough to be systems (Chapin and Danell 2001, Walker 1995, considered insignificant as a landscape- Walker et al. 2001). The decrease in species rich- forming process. ness from south to north is mainly due to loss of species rather than turnover in species, so that 4. The current timber harvest regime is quite the most northerly floras represent depauperate small relative to other stand-replacing distur- versions of the southern ones (that is not wholly bances. However, new technologies and true, but it is a reasonable representation). There- economies in fiber harvest may result in a fore, the latitudinal similarities among the vegeta- very different regime in the near future. tion and flora increase from south to north, 5. Patterns of eolian loess distribution in the because there is a smaller pool of species avail- systems differ because of an interaction with able. Until treeline is encountered, the dominant topography, vegetation, and wind regime. species are the same or closely related through- Arctic rivers follow broad, often anastomosed out the zone; however, the species representing pathways with large expanses of barren or the northern treeline vary throughout the region nearly barren silt and sand bars. Nearly con- (Alexandrova 1980). stant wind redistributes this loess across the Because this is such a major gradient in both landscape, with major influence on patterns climate and structure, many other variables are of vegetation composition and productivity simultaneously changing and correlated along the (Walker and Everett 1991, Walker et al. gradient including available moisture, energy bal- 1998). The Alaska boreal forest has little wind ance and radiative transfer, distribution and depth compared to the Arctic, and although there of permafrost (permanently frozen ground), bio- are active sand and silt bars associated with mass and production, and others. Changes in medium and large rivers, much of what is vegetation structure and composition along this blown off is effectively trapped by the riparian and flood-plain vegetation.

49 Figure 1—Major ecosystems of Alaska. Source: Markon, C. 1991. US Geological Survey digital map of the major ecosystems of Alaska. Based on the map unit boundaries delineated by the Joint Federal-State Land Use Planning Commission for Alaska in 1973. Scale 1: 2,500,000.

50 Table 1–Structural characteristics of major arctic-boreal vegetation and ecosystem zones Mean July Zone Vegetation structure temperature Polar desert Mainly herbaceous or nonvascular, discontinuous scattered plants. 1-3 °C Mosses, when present, limited to acrocarpous species or small hummocks of individual clones. High arctic Mainly herbaceous (Graminae and Cyperaceae); woody species in 4-6 °C genera Salix and Dryas with woody sections limited to below ground or 5 cm above ground, areas of semicontinuous cover in protected areas. Increasing diversity of pleurocarpous mosses in small clusters of plants. Low arctic Codominated by herbaceous (mainly graminoids of Cyperaceae) and 7-9 °C woody species in the Ericaceae family and in genera Salix, Dryas, and Betula. Woody sections to as high as 40 cm above ground. Continuous cover. Development of full moss and lichen carpet underlying vascular species with high nonvascular species diversity. Sphagnum spp. dominate moss carpets in certain regions. Treeline or Tall shrubs from 1 to 2 m tall representing four to five families, with 10-12 °C woody tundra graminoids of Cyperaceae subdominant. Scattered islands of coniferous trees, occasional stands of poplar (Populus spp.) or tree birches (Betula spp.). Increased moss biomass and production. Boreal forest Uplands completely dominated by fully developed forest, mainly of Picea, 12-15 °C Betula, Populus, and Larix. Graminoid-dominated stands of Cyperaceae limited to areas of permafrost in lowlands. Moss biomass and production extremely variable depending on forest type and condition, from none in deciduous upland forests to thickest development of carpets in open, permafrost lowlands.

51 Figure 2—Arrangement of major structural-dominance vegetation types along gradients of megascale climate and mesotopography (catenas) in Alaska and western Canada. (Latitude is shown for each location.) Symbols: P–polar desert, see table 1; H–heath, ericaceous and rosaceous prostrate and dwarf shrubs with lichens; W–fen (wet meadow) dominated by species of Carex (s) or grasses (g), mainly Dupontia fisheri R.Br.; T–tussock tundra, Eriophorum vaginatum L., Carex bigelowii Torr. ex Schwein. with low and prostrate shrubs; S–shrub tundra, species of erect Salix L., Betula L., and sometimes Alnus P. Mill. with an understory of prostrate shrubs, graminoids, and mosses; O–open woodland/muskeg, combination of sparse black spruce Picea mariana (P. Mill.) B.S.P. with tussock tundra; F–forest, mainly Picea glauca (Moench) Voss as well as Populus tremuloides Michx. and Betula papyrifera Marsh.

52 Figure 3—Spatial and temporal domains of the major landscape-forming natural disturbances in boreal and low arctic ecosystems, respectively (adapted in part from Walker and Walker 1991). Timber harvest is overlain on the boreal system to show its extent relative to other disturbances.

The increased importance of fire and animals is a function can be clearly linked to a series of mostly direct result of the presence of trees but also is a abiotic controls (Walker et al. 1994, Walker et al. function of climate. Climate sets an ultimate limit 1998). Although the arctic biome in Alaska has on tree growth and in some cases animal distribu- varied significantly in spatial extent and appear- tion (particularly insects; Chapin and Danell ance over the past million years, most of its flora 2001), but the presence of trees also drives both formed in major glacial refugia and has since insect and fire regimes. In the Arctic, patterns of been redistributed to form existing patterns vegetation and major differences in ecosystem (Hultén 1937). The Alaska boreal ecosystem, in

53 contrast, had its dominant species invade only Forest harvest in the Alaskan interior represents a within the past 10,000 years, in a series of scale issue mainly in the sense that current forest “waves” that represented almost instantaneous harvest is on a scale small enough that it can be appearance over large areas (Anderson and considered ecologically insignificant. A closer ex- Brubaker 1994, Hu et al. 1993). Existing patterns amination indicates this view to be somewhat of a within the Alaska boreal zone can be explained falsehood, however. Most harvest is on private almost completely by a combination of mesotopo- lands owned by regional corporations. Forest graphy and succession following disturbance industries on these lands are not economically (Van Cleve et al. 1983, 1996). viable from a market perspective but are subsi- dized as a way of getting local timber and fire- The ability of a few species to play a dominant wood and providing employment in small villages.4 role in ecosystems processes has major implica- Harvest on public lands is primarily within the tions for many management issues as well as for Tanana Valley State Forest and to a lesser degree our ability to comprehend the potential for change on Bureau of Land Management lands. Current in the system under changing climate. Because harvest rates on these lands are at about 800 arctic systems are controlled by fairly fixed sets acres per year, and many offered sales are not of regimes, developing comprehensible models purchased (Crimp et al. 1997). Forest harvest is a of landscape change related to climate change public concern, however, for the following rea- should be possible (Epstein et al. 2000). In the sons: (1) Sustainability of the harvest resource, boreal system, however, the strong interaction mature white spruce stands, is little understood between biota, climate, and disturbance regime and highly variable in time and space. There has makes for a much more chaotic system (Rupp et been no systematic analysis of forest practices al. 2000, Starfield and Chapin 1996). Getting fire over the last 30 to 50 years to analyze changes in management “right” is perhaps the single most composition following harvest, but there is some important management question in Alaska’s inte- evidence that certain harvest practices may result rior region today. Fire has more than major eco- in a long-term dominance of deciduous species, logical consequence in Alaska; fire fighting is a an undesirable management outcome if white major source of cash economy in many small spruce harvest is to be sustained. Natural reseed- villages.2 The Alaska Fire Service’s current policy ing and successful establishment of white spruce is to control human-caused fires and to let natural are dependent on timing of cone crops and herbi- fires (started by lightning) burn unless they vore population levels, both of which vary in time, threaten life or property.3 Their goal is to support a creating an uncertainty that is beyond manage- natural regime, with the idea that ultimately such a ment possibility. (2) Most harvests are in highly regime will maximize all major economic land- visible areas near roads and along rivers, and scape uses: wildlife habitat, recreation, and timber thus engender public scrutiny and debate. Al- harvest. Because fire regime is so closely tied to though the level of harvest is essentially invisible both climate and vegetation, it seems unlikely that relative to large timber industries, it is of great any policy other than the current one could be concern locally. (3) The Alaska economy is almost justified by scientific evidence. A natural fire re- completely dependent on natural resources, both gime will forever be a moving target, however, from extraction and tourism viewpoints. Changing and thus management goals need to follow the global markets for fiber and new technologies that regime rather than dictate it. make fiber use more practical could potentially lead to a regional fiber industry that would not be 2 Chapin, S.F., III. 2001. Personal communication. Professor, so dependent upon certain vegetation types. University of Alaska, Fairbanks, P.O. Box 757520, Fairbanks, AK 99775. Thus, although the current timber industry effect is masked by the effects of fire, fiber extraction on 3 Jandt, R. 2001. Personal communication. Ecologist, Alaska large scales could change this. A firm ecological Fire Service, Bureau of Land Management, P.O. Box 35005, Fort Wainwright, AK 99703. 4 Wurtz, T. 2001. Personal communication. Research ecologist, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Boreal Ecology Coopera- tive Research Unit, P.O. Box 756780, University of Alaska Fairbanks, Fairbanks, AK 99775-6780.

54 understanding of how disturbances operate in this Chernov, Yu.I.; Matveyeva, N.V. 1997. Arctic system in time and space is therefore essential to ecosystems in Russia. In: Wielgolaski, F.E., our ability to provide input on sustainable man- ed. Polar and alpine tundra. Amsterdam, The agement practices. Netherlands: Elsevier: 361-507. In conclusion, Alaska is considerably different Crimp, P.M.; Phillips, S.J.; Worum, G.T. 1997. from other states in terms of having extensive, Timber resources on state forestry lands in the remote landscapes that are mostly undeveloped. Tanana Valley. Fairbanks, AK: Alaska Depart- Management of Alaska landscapes must consider ment of Natural Resources, Division of For- the interactions between the spatial and temporal estry. 80 p. scales of various disturbances. Both long-term Delcourt, H.R.; Delcourt, P.A.; Webb, T., III. data collection and retrospective analysis are 1983. Dynamic plant ecology: the spectrum of important tools for understanding the frequency vegetational change in space and time. Qua- and importance of certain types of events. Manag- ternary Science Reviews. 1: 153-175. ing for sustainability must consider that the range of variability in key disturbances will rarely match Epstein, H.A.; Walker, M.D.; Chapin, F.S., III; with the affected spaces and timeframes required Starfield, A.M. 2000. A transient, nutrient- for most management decisions. based competition model of arctic vegetation. Ecological Applications. 10(3): 824-941. Metric Equivalents Franklin, J.F. 1987. Importance and justification 1 acre = 0.405 hectares of long-term studies in ecology. In: Likens, G.E., ed. Long-term studies in ecology: ap- Degrees Fahrenheit = 1.8 °Celsius + 32 proaches and alternatives. New York: Springer- Verlag: 3-19. Literature Cited Hammond, T.; Yarie, J. 1996. Spatial prediction Alexandrova, V.D. 1980. The Arctic and Antarctic: of climatic state factor regions in Alaska. their division into geobotanical areas. Cam- Ecoscience. 3: 490-501. bridge, United Kingdom: Cambridge University Press. 247 p. Hu, F.S.; Brubaker, L.B.; Anderson, P.M. 1993. A 12,000-year record of vegetation change and Allen, T.F.H.; Hoekstra, T.W. 1992. Toward a soil development from Wien Lake, central unified ecology. New York: Columbia University Alaska. Canadian Journal of Botany. 71: 1133- Press. 384 p. 1142. Allen, T.F.H.; Starr, T.B. 1982. Hierarchy: per- Hultén, E. 1937. Outline of the history of arctic spectives for ecological complexity. Chicago, and boreal biota during the Quaternary period: IL: University of Chicago Press. 394 p. their evolution during and after the glacial pe- Anderson, P.M.; Brubaker, L.B. 1994. Vegeta- riod as indicated by the equiformal progressive tion history of northcentral Alaska: a mapped areas of present plant species. Lehre, Ger- summary of late-Quaternary pollen data. Qua- many: J. Cramer Publishing. 168 p. ternary Science Reviews. 13: 71-92. Jorgenson, M.T.; Roth, J.; Raynolds, M. [et al.]. Barber, V.A.; Juday, G.P.; Finney, B.P. 1998. 1999. An ecological land survey for Fort Wain- Tree-ring data for climate reconstruction and wright, Alaska. Report 99-9. Hanover, NH: U.S. indication of twentieth century warming and Army Cold Regions Research and Engineering summer drying. Paleotimes. 6: 20-22. Lab. 83 p. Chapin, F.S.; Danell, K. 2001. Scenarios of Jorgenson, M.T.; Roth, J.; Smith, M.D. [et al.]. biodiversity change in the boreal forest. In: 2001. An ecological land survey for Fort Chapin, F.S.; Sala, O.; Huber-Sannwald, E., Greely, Alaska. Rep. TR-01-4. Hanover, NH: eds. Global biodiversity in a changing environ- U.S. Army Cold Regions Research and Engi- ment: scenarios for the 21st century. New York: neering Lab. 85 p. Springer-Verlag: 101-120.

55 Juday, G.P.; Ott, R.A.; Valentine, D.W.; Barber, Walker, D.A. 2000. Hierarchical subdivision of V.A. 1998. Forests, climate stress, insects, and arctic tundra based on vegetation response to fire. In: Weller, G.; Anderson, P., eds. Implica- climate, parent material, and topography. Glo- tions of global change in Alaska and the Bering bal Change Biology. 6: 19-34. Sea region. Fairbanks, AK: Center for Global Walker, D.A.; Auerbach, N.A.; Bockheim, J.G. Change and Arctic System Research: 23-49. [et al.]. 1998. Energy and trace-gas fluxes O’Neill, R.V.; Deangelis, D.L.; Waide, J.B.; across a soil pH boundary in the Arctic. Nature. Allen, T.F.H. 1986. A hierarchical concept of 394: 469-472. ecosystems. Monographs in Population Biol- Walker, D.A.; Everett, K.R. 1991. Loess ecosys- ogy 23. Princeton, NJ: Princeton University tems of northern Alaska: regional gradient and Press. 262 p. toposequence at Prudhoe Bay. Ecological Rupp, T.S.; Starfield, A.M.; Chapin, F.S., III. Monographs. 61: 437-464. 2000. A frame-based spatially explicit model Walker, D.A.; Walker, M.D. 1991. History and of subarctic vegetation response to climatic pattern of disturbance in Alaskan arctic terres- change: comparison with a point model. Land- trial ecosystems: a hierarchical approach to scape Ecology. 15: 383-400. analyzing landscape change. Journal of Ap- Serreze, M.C.; Walsh, J.E.; Chapin, F.S., III plied Ecology. 28: 244-276. [et al.]. 2000. Observational evidence of re- Walker, M.D. 1995. Patterns of arctic plant com- cent change in the northern high-latitude envi- munity diversity. In: Chapin, F.S., III; Korner, ronment. Climatic Change. 46(1/2): 159-206. C., eds. Arctic and alpine biodiversity: patterns, Starfield, A.M.; Chapin, F.S., III. 1996. Model of causes, and ecosystem consequences. Eco- transient changes in arctic and boreal vegeta- logical Studies. 113: 1-18. tion in response to climate and land use Walker, M.D.; Gould, W.A.; Chapin, F.S., III. change. Ecological Applications. 6: 842-864. 2001. Scenarios of biodiversity changes in Van Cleve, K.; Dyrness, C.T.; Viereck, L.A. arctic and alpine tundra. In: Chapin, F.S., III; [et al.]. 1983. Taiga ecosystems in interior Sala, O.; Huber-Sannwald, E., eds. Global Alaska. Bioscience. 33(1): 39-44. biodiversity in a changing environment: scenarios for the 21st century. New York: Springer- Van Cleve, K.; Viereck, L.A. 1981. Forest suc- Verlag: 83-100. cession in relation to nutrient cycling in the boreal forest of Alaska. In: West, D.; Shugart, Walker, M.D.; Walker, D.A.; Auerbach, N.A. H.; Botkin, D., eds. Forest succession: con- 1994. Plant communities of a tussock tundra cepts and application. New York: Springer- landscape in the Brooks Range foothills, Verlag: 185-211. Alaska. Journal of Vegetation Science. 5: 843-866. Van Cleve, K.; Viereck, L.A.; Dyrness, C.T. 1996. State factor control of soils and forest succession along the Tanana River in interior Alaska, USA. Arctic and Alpine Research. 28(3): 388-400.

56 Landscape Perspectives

Roger N. Clark,1 Linda E. Kruger,2 Stephen F. McCool,3 and George H. Stankey4

In this presentation, we describe perspectives times by citizens. Technical definitions are impor- about landscapes and scales of analysis appropri- tant for technical analyses but not necessarily ate to understanding the nature of relationships important to everyone; they are a means to often- between people and the environment of which unclear ends. The challenge is to understand the they are a part. We also discuss how to improve needs and questions to be addressed before lines integration of human and biophysical consider- are drawn on maps and data collection started. ations in the study and management of land- Such problem framing is the most important, yet scapes. Given the purpose of this conference, the least well-done step, particularly if all interested paper is not in the format of a formal review or parties are not included up front. synthesis of the literature. This paper reflects the In a sense, the meanings that landscapes hold experience of the authors whose work has collec- are constructed by those viewing the landscape or tively spanned several decades and a variety of interacting with it in other ways. Each meaning is topics related to people and natural resource in- different, not better or worse. Full understanding teractions. Our observations might best be con- of the values and meanings landscapes produce sidered as propositions to aid discussion rather requires that analyses be inclusive of the people than scientific conclusions or testable hypotheses. who interact with the landscape. The implication of different types of meanings is that conflict is Landscapes Are in the Eye of the likely to occur where such meanings collide. Beholder Landscapes come in many sizes. There is no one People Think and Act at Multiple right definition for what a landscape is (or what it Scales for Many Reasons means to different people) or the scale(s) appro- No one way to divide time and space will account priate for understanding relationships between for the multiple values, concerns, and uses that humans and natural resources. Various needs people bring to the understanding of natural re- and questions will define the appropriateness of sources. Some ways to think about how land- landscape meanings and scales of analysis. scapes can be considered from a social science Sometimes these needs are defined by scientists, perspective are summarized below. other times by forest managers, and yet other Many values are important to people. Many 1 Research social scientist, U.S. Department Agriculture, values are important to people as they think about Forest Service, Pacific Northwest Research Station, Forestry and use forests and other landscapes. These th Sciences Laboratory, 400 N 34 Street, Suite 201, Seattle,WA include a variety of commodity, public use, amen- 98103; Tel: 206-732-7833; e-mail: [email protected]. ity, environmental quality, spiritual, and health 2 Research social scientist, U.S. Department Agriculture, values. Such meanings are attached to land- Forest Service, Pacific Northwest Research Station, Forestry scapes by different types of people and at differ- Sciences Laboratory, 400 N 34th Street, Suite 201, Seattle,WA 98103; Tel: 206-732-7832; e-mail: [email protected]. ent scales. For example, recreation can be thought of as people using microsites such as 3 Professor, School of Forestry, University of Montana, campsites or as driving for pleasure across larger Missoula, MT 59812; Tel: 406-243-5406; e-mail: [email protected]. landscapes. And the array of values often blends biophysical, economic, and social domains in 4 Research social scientist, U.S. Department Agriculture, different combinations across space, people, and Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, Corvallis, OR 97331; Tel: 541-737- time. This means that to understand the meaning 1496; e-mail: [email protected]. and importance of these values, expertise beyond the biophysical sciences is required.

57 Photo by Roger Clark

People organize in many ways. There are a entific needs. Meanings cannot be aggregated variety of ways to think about how people (indi- upward; people may define an entire watershed viduals) are combined at different scales and how as a suitable place for timber harvesting, yet hold a social organizational hierarchy can be de- claims to spiritual, aesthetic, and recreational scribed. These include individuals, family and meanings at the site level. household groups, neighborhoods, communities, Human lives and activities consider multiple counties/boroughs, states/provinces, nations, and temporal scales. Ways of defining time include ultimately, the globe. The interests (political, eco- the past, today, tomorrow, weeks, seasons, years, nomic, cultural) people hold in the landscape at decades, and generations. These may or may not different scales and the decisions they make coincide with how time is considered by special- about how they interact with it may cut across ists concerned with biophysical phenomena. Dif- these different levels. Each level or scale is char- ferences between biological and social scales of acterized by different emergent properties, such significance are frequently at the root of conflict— that the next higher scale is not simply an aggre- such as when forest plans are considered over a gation of units. There are often mismatches be- 50-year timeframe but budgets are appropriated tween these organizational units and biophysical annually. Considerations of time often influence scales that need to be reconciled before any public acceptance of forest management prac- analysis begins if an integrated solution is de- tices. sired. Putting the pieces together later has proven to be difficult. We must beware of the ecological fallacy when drawing conclusions about people. People act at multiple spatial scales. These What may be true at a higher scale, such as the include microsites, areas (e.g., a grove of trees, county level, may not be so at lower scales, such meadows), drainages, watersheds, landscapes as the communities in the county; the attributes of (e.g., “the valley”), regions (e.g., Puget Sound), a transportation system may not apply to the indi- continents, and the globe. It is important to con- vidual roads within; qualities of a dispersed recre- sider such ways of defining scales because differ- ation area may differ when one looks at specific ent social, cultural, and institutional properties sites; and the distribution of meanings across a may emerge at each scale. Appropriate scales landscape cannot necessarily be summed to ar- may be defined by the processes at work, interac- rive at an overall assignment of landscape mean- tions within and between components of complex ings. Looking up and down this sort of scale is not biophysical and social systems, and policy or sci-

58 Photo by Linda Kruger simply an aggregation or disaggregation issue. It specific places, and how they define their critical is likely in many cases that different processes habitat needs, often are not correlated with certain work at different scales. The perspectives people types of biogeographical features mapped by have when they think at different scales influence biologists. These concepts can help us think judgments about the appropriateness and accept- about how people relate to landscapes at multiple ability of change. And what may be acceptable at scales and improve our ability to understand the one scale may not be so at another. effects of policy options and management practices on existing and potential public values and Human habitats are definable. Concepts that uses. help explain why some wildlife species behave as they do also help explain the behavior of people. Corridors channel energy. Natural topography For example, we have found in studies of and features and human-created corridors chan- recreationists that home ranges exist for both nel air, water, critters, and people. The intersec- residents and tourists (a migratory species!). tions of corridors (water crossings, power Edges, particularly along water, affect use. Critical corridors, dams) or flows within them often reveal habitats are definable and include attributes that conflicts and compatibilities between public values relate to many temporal and spatial scales. Hu- and uses and other resource values. man habitats—for recreation, spiritual, and aes- Acceptance of change varies both in time and thetic purposes—are dynamic in time and space. space. Judgments about the acceptability of Travel corridors and how they are accessed influ- changes depend on their nature, extent, cause, ence patterns of use. Place attributes and mean- and location with respect to the meanings and ings (at multiple scales) influence choices people values people attach to specific places and land- make. However, the meanings people attach to scapes.

59 Problem Framing Is the Key to Ideologies and beliefs (worldviews) condition how Success we think. The common focus on getting answers and solutions before clarifying the questions and It is important not to “start” until one gets the problems often leads us in the wrong direction. issues and questions right. This usually means The language of science and expertise makes it we need to step back from individual, disciplinary difficult for interested citizens to easily engage in definitions and join with other “–ologies” and in- processes and activities that affect them and to terests. We need to be sure we are not solving contribute the knowledge they have about the “solutions” or solving the “wrong” problem. We complex systems we all value. Technology can be must learn from one another about how we “see” both a means and an end but often drives us in and define landscapes so that we can jointly con- directions that later prove to be less than effec- struct opportunities and redefine problems and tive. For example, more and better geographic issues and then develop explicit questions to information system technology can be a hin- drive joint actions. Action in today’s society re- drance if the wrong questions are under study. quires multiple actors, and joint agreement on problems and opportunities is needed before Conclusions effective action can be implemented. It is important to determine where questions require “broad All landscape definitions are human conventions– and integrated” as well as “narrow and deep” ap- they do not “exist” in any objective sense. We proaches. Picking the wrong boundaries and carry expectations and definitions with us. And scales is inevitable unless driven by questions. these expectations and definitions are dynamic. Because integrated approaches require a full What often divides diverse interests and needs range of interests, types of knowledge, and points are assertions that there is one best way to de- of view, such approaches will challenge the power scribe landscapes and scales. and authority normally accorded to experts and Integrated multiscale approaches will link humans specialists. with biophysical landscapes. But social and biophysical considerations often are not in sync at Problem Framing and Resolution Must least initially. Improved problem framing before Include Various Value Systems any action to collect data will improve the integration of human consideration, values, and uses into Because landscape values and meanings differ landscape analyses. by time, scale, and individual or group, there is no “correct” definition. Although this suggests that People form strong opinions about places and the diversity may be an obstacle, it may be an oppor- characteristics of places at multiple scales. They tunity as well. What can unite us is recognition of also are concerned about the appropriateness of the power of both individual and collective per- resource management uses (in time and space). spectives. Although we “look at” the same places Legacies on the land affect judgments in a differ- we “see” them differently, so to really understand ent way. It is hard for some people to relate to them we need to look at them together and de- large landscapes when they are concerned about scribe what we see and why. But this is not an favorite places. easy task. Planning processes that are inclusive So, landscapes are socially defined in multiple raise the possibility of increased understanding, ways for multiple purposes from multiple perspec- improved representativeness in public participa- tives. They are complex—socially and biophysi- tion, an opportunity to learn, and eventually identi- cally. These definitions are dynamic. Landscapes fying ways to get to a desired future. at multiple scales can be used as a means to unite and understand interactions among bio- What’s Getting in the Way? physical and social systems. As we think about and conduct analyses of land- We need to take a more holistic view when think- scapes from multiple perspectives, it becomes ing about landscapes. Boundaries matter—they evident that to develop a more holistic approach is define time and space for many relevant pur- difficult. A number of things make designing and poses. The trick is to look across boundaries on implementing integrated approaches challenging. the map and in the mind.

60 Landscape-Level Management: It’s All About Context1

Bruce Shindler2

Abstract Forest managers and researchers are now working at a landscape level, but is our public with us? For most people, landscape-level management is not a clear concept. To better understand how citizens view landscapes, we need to go beyond attempts to “educate the public” and instead interact with citizens to promote learning, find appropriate outreach activities and simulation techniques, and directly address questions about risk and uncertainty. Various scales of analysis will be necessary, including the geographic, temporal, and normative contexts and their relevance to citizens.

1 Shindler, B. 2000. Landscape-level management: it’s all about context. Journal of Forestry. 98 (12): 10-14. Reprinted with permission from the Society of American Foresters.

2 Associate professor, Department of Forest Resources, Oregon State University, Corvallis, OR 97331; Tel: 541-737- 3299; e-mail: [email protected].

61 Human Adaptation to Environmental Change: Implications at Multiple Scales

Ellen M. Donoghue,1 Terry L. Raettig,2 and Harriet H. Christensen3

Introduction government. Physical and biological effects often are relevant only indirectly as they have shaped The purpose of this paper is to describe recent the boundaries of government entities such as and ongoing work that contributes to understand- cities, counties, and states. Understanding the ing social and economic change at broad scales. relevant secondary data available is a prerequisite In particular, the paper focuses on a recently re- to primary data collection efforts. leased atlas on human adaptation to environmental change at the county level. The paper also Winnowing this mass of information on social and discusses issues pertaining to data availability economic variables and selecting the appropriate and geographic and temporal scales. data to address the research or policy issues at hand is a formidable task. Also, there is the need Issues of scale are important to understanding to provide meaningful tabular, graphical, and geo- social and economic change. Two important con- graphical displays of this information as an aid to siderations affect social and economic analyses understanding and interpretation. It is in this arena at the broad scale. First, there is the wealth of that the work of the Rural Communities and secondary data collected by an array of federal, Economies Team of the USDA Forest Service’s state, and, in some cases, local government Pacific Northwest Research Station has been agencies (Salant and Waller 1995). Frequently concentrated. A related issue of interest is how these data sets provide information for specific different scales of the social and economic vari- variables across multiple scales and on a rela- ables are hierarchically related and how these tively consistent basis for extended periods of relationships impact the interpretation usefulness time. Examples include the decennial census of the variables. Team work has focused on these information that is available from individual cen- three areas: (1) the collection of data on key so- sus “blocks” or “block groups” and additive to the cial and economic variables for the Pacific North- county, state, and national levels and the employ- west; (2) the development of geographical dis- ment information collected at the firm level by plays (atlases) that include interpretations of employment security departments and provided at these variables (Christensen et al. 2000, Raettig a range of geographic and industrial sector scales et al. 2001); and (3) the exploration of theoretical and levels of detail. The second consideration is issues concerning social and economic variables that most of this secondary information is avail- and the linkages between different spatial and able only at scales defined by political units of temporal scales.4

1 Research social scientist, U.S. Department of Agriculture, Recently in the West, several interagency, Forest Service, Pacific Northwest Research Station, Forestry multidisciplinary, broad-scale assessments were Sciences Laboratory, P.O. Box 3890, Portland, OR 97208- conducted in response to, or in anticipation of, 3890; Tel: 503-808-2018; e-mail: [email protected]. changes in resource management. These as- 2 Economist, U.S. Department of Agriculture, Forest Service, sessments include the Forest Ecosystem Man- Olympic National Forest, 1835 Black Lake Blvd. SW, agement Assessment Team (FEMAT 1993), the Olympia, WA 98512-5623; Tel: 360-956-2449; e-mail: Sierra Nevada Ecosystem Project (SNEP 1996), [email protected].

3 Research social scientist (retired), U.S. Department of 4 Raettig, T.L. [N.d]. Spatial and temporal scale issues in Agriculture, Forest Service, Pacific Northwest Research Northwest Forest Plan monitoring and evaluation activities. th Station, Forestry Sciences Laboratory, 400 N 34 , Suite 201, Manuscript under review. On file with: USDA Forest Service, Seattle, WA 98103; Tel: 206-732-7824; e-mail: Pacific Northwest Research Station, Forestry Sciences [email protected]. Laboratory, P.O. Box 3890, Portland, OR 97208.

62 and the Interior Columbia Basin Ecosystem An important consideration in collecting second- Management Project (Quigley and Arbelbide ary data is the specific definition for the variable of 1997). Social assessments at the county and interest. Employment information, for instance, is community level were conducted as part of these based on place of employment in the employment bioregional assessments (Doak and Kusel 1997, security and regional economic information sys- Harris et al. 2000, Horne and Haynes 1999). tem, but is based on place of employee residence Many dimensions of human life were reported on for the decennial census and CPS. Temporal including culture, the history of human-natural scale also is a consideration. Employment Secu- resource interactions, socioeconomic conditions rity Department employment and wage informa- and trends, and measures of community resil- tion is often available on a monthly as well as iency and community capacity. This work expands annual basis. The decennial census collects infor- on these efforts, by depicting other variables over mation only once per decade with estimates pro- different geographic and spatial scales. It also vided for key census variables in intervening contributes to meeting the requirements for social years. and economic monitoring and forest planning. For The key economic variables depicted in the atlas instance, in the Northwest Forest Plan (NWFP) include employment, income, unemployment, and region, the record of decision for the NWFP di- economic diversity. The key social variables in- rects agencies to conduct three types of monitor- clude population (and characteristics of the popu- ing—implementation, effectiveness, and lation, such as race and ethnicity), migration, validation—to detect desirable and undesirable education, poverty, and crime. These variables changes as a result of ecosystem management reflect key indicators that are used in social and (USDA and USDI 1994). Planning rules for Na- economic assessments and studies on socioeco- tional Forest System land also have requirements nomic well-being. In addition to spatial and tempo- for social and economic sustainability (see for ex- ral scale issues, the atlases focus on absolute, ample, USDA Forest Service 1999). Finally, the trend, and relative change in the values of the work reflects appropriate roles for scientists in variables. An example has been the documenta- assessments, analyses, and monitoring activities tion and interpretation of how different sectors of associated with forest management. the wood products industry have been impacted by changes in timber harvest in the Pacific North- Data Collection west (Raettig and McGinnis 1996). The atlases focus on county-level information. For For the Pacific Northwest, work also has begun many of the social and economic indicators avail- on the collection of community-level data incorpo- able from secondary sources (except for the de- rating the secondary data that are available and cennial census), the county is the smallest unit for may ultimately include the collection of selected which information is disseminated or published. In primary data items (Donoghue 2003). The social, the case of the commonly used economic vari- economic, and resource data, such as timber ables such as employment and income, informa- harvests, assembled by the team have provided tion cannot be disclosed or published that can be the basic information in an analysis of the North- traced to an individual firm or employee (Wash- west Economic Adjustment Initiative (Raettig and ington Employment Security Department 1997). Christensen 1999). The data and analytical work For other variables, such as those collected in the also have been resources for continuing efforts Current Population Survey (CPS), small sample related to Northwest Forest Plan social and eco- sizes may preclude the calculation of data for nomic monitoring. smaller geographical areas. County-level data can be easily aggregated into larger scale areas such Atlases as planning regions, states, and even multistate regions. They also provide useful context for so- The Atlas Of Human Adaptation to Environmental cioeconomic assessments at smaller scales, al- Change, Challenge, and Opportunity for the Pa- though in and of themselves may not be adequate cific Northwest (Christensen et al. 2000), and the proxies for conditions and trends at smaller scales Atlas of Social and Economic Conditions and (Beckley 1998). Change In Southern California (Raettig et al. 2001) provide geographical expression of change

63 in social, economic, and timber-related variables Specific maps displayed in these sections: at the county level. The Pacific Northwest atlas • Section 2: Change in timber harvest and wood was designed to provide Northwest Forest Plan products employment managers and planners with a comprehensive reference depicting the dimensions, location, Change in public timber harvest magnitude, and direction of social and economic Change in public and private timber change in the Northwest Forest Plan region. In- harvest cluded are displays that show patterns of the so- Change in wood products employment cial and economic diversity and health of the Federal lands-related payments to region. This atlas focuses on change during the counties period of timber harvest reduction and initial economic restructuring in the region. Most maps are • Section 3: Change in economic conditions: based on changes in variables between 1989 and economic performance 1994. Unemployment rate compared to The first section of the Pacific Northwest atlas region (Christensen et al. 2000) contains five base maps Change in total employment that depict the region, the Northwest Forest Plan Wage trends provinces, counties, population distribution, major Wage level transportation routes, and the public land ownerships. Six sections follow that address: • Section 4: Structural economic change 1. What kinds of social and economic changes Economic diversity have taken place in the face of reduced tim- Fastest growing nonfarm industries ber harvest? Are Pacific Northwest communi- Slowest growing nonfarm industries ties changing? If so, how? • Section 5: Characteristics of the population 2. What changes have occurred in the timber industry since 1990? Has timber employment Change in county population changed? Is private harvest increasing? Change in unincorporated and incorporated population growth 3. Have changes in federal harvest had a signifi- Migration status and trends cant effect on county revenues? Change in ethnicity 4. Are western Oregon and Washington and Educational attainment northern California singularly dependent on natural resources? • Section 6: Social issues 5. What federal assistance has aided cities and Change in income maintenance rural areas? Change in poverty rate Changes in violent crime 6. How have the population characteristics Changes in property crime changed? What are the trends in migration, Alcohol-related incidences educational attainment, and changes in ethnicity? • Section 7: Federal assistance 7. What have been the changes in selected Northwest Economic Adjustment social issues such as rates of poverty, prop- Initiative erty and violent crimes, and alcohol-related incidences?

64 Accompanying the maps are text, supporting Literature Cited graphics, and tabular displays that interpret the information and patterns on the maps, explore Beckley, T.M. 1998. The nestedness of forest the relevant issues in more depth, and provide dependence: a conceptual framework and a working tool for managers and planners. empirical exploration. Society and Natural Re- sources. 11: 101-120. A similar atlas was prepared as part of the socioeconomic assessment for the Southern California Christensen, H.H.; McGinnis, W.J.; Raettig, Conservation Strategy, a broad-scale planning T.L.; Donoghue, E. 2000. Atlas of human ad- effort for the four national forests in southern Cali- aptation to environmental change, challenge, fornia (Raettig et al. 2001). This atlas covers 26 and opportunity: northern California, western counties in southern California and includes maps Oregon, and western Washington. Gen. Tech. for the same variables displayed for the Northwest Rep. PNW-GTR-478. Portland, OR: U.S. De- Forest Plan region, except timber harvest infor- partment of Agriculture, Forest Service, Pacific mation was not displayed owing to the relative Northwest Research Station. 66 p. insignificance of timber harvest on the four na- Doak, S.; Kusel, J. 1997. Well-being assessment tional forests of that region. Both atlases are avail- of communities in the Klamath region. able on CD-ROMs that include some interactive Taylorsville, CA: Forest Community Research; features. contract 43-91W8-6-7077. Report prepared for the USDA Forest Service, Klamath National Theoretical Issues and the Linkages Forest. Between Different Spatial and Donoghue, E.M. 2003. Delimiting communities in Temporal Scales the Pacific Northwest. Gen. Tech. Rep. PNW- If social and economic variables are to be useful GTR-570. Portland, OR: U.S. Department of for managers and planners, a thorough under- Agriculture, Forest Service, Pacific Northwest standing is necessary of the implications of the Research Station. 51 p. information displayed and the relationship be- Forest Ecosystem Management Assessment tween decision processes and different scales of Team [FEMAT]. 1993. Forest ecosystem man- the variables. Work has been done exploring the agement: an ecological, economic, and social relationships between individual and aggregate assessment. Portland, OR: U.S. Department data (the “ecological fallacy”) (King 1997, McCool of Agriculture; U.S. Department of the Interior and Troy 1998) in political science and other so- [et al.]. [Irregular pagination]. cial sciences. Still, other work has explored the issues of spatial data, unit definition, and scale in Fotheringham, A.S.; Wong, D.W.S. 1991. The geographic and geographic information systems- modifiable areal unit problem in multivariate related problems (Fotheringham and Wong 1991, statistical analysis. Environment and Planning Longley and Batty 1997). The Rural Economies A. 23: 1025-1044. and Community Team has work in progress that Harris, C.C.; McLaughlin, W.; Brown, G.; will produce a synthesis of the theoretical work Becker, D. 2000. Rural communities in the that has been done in these areas (see footnote inland Northwest: an assessment of small 4). Ultimately this theoretical work will provide the communities in the interior and upper Colum- basis for understanding scale issues and how bia River basins. Gen. Tech. Rep. PNW-GTR- these issues impact the use and interpretation of 477. Portland, OR: U.S. Department of Agri- social and economic variables in the planning, culture, Forest Service, Pacific Northwest monitoring, and management settings that are Research Station. 120 p. part of the ecosystem management processes now being implemented.

65 Horne, A.L.; Haynes, R.W. 1999. Developing Raettig, T.L.; McGinnis, W.J. 1996. Wood prod- measures of socioeconomic resiliency in the ucts employment in the Pacific Northwest: interior Columbia basin. Gen. Tech. Rep. emerging trends. In: Proceedings of the 30th PNW-GTR-453. Portland, OR: U.S. Depart- annual Pacific Northwest regional economic ment of Agriculture, Forest Service, Pacific conference. Portland OR: Pacific Northwest Northwest Research Station. 41 p. (Quigley, Regional Economic Conference and Center for T.M., ed.; Interior Columbia Basin Ecosystem Urban Studies, Portland State University: 45- Management Project: scientific assessment). 50. King, G. 1997. A solution to the ecological infer- Salant, P.; Waller, A.J. 1995. Guide to rural data. ence problem: reconstructing individual behav- Revised ed. Washington, DC: Island Press. ior from aggregate data. Princeton, NJ: 140 p. Princeton University Press. 342 p. Sierra Nevada Ecosystem Project [SNEP]. Longley, P.; Batty, M. 1997. Spatial analysis: 1996. SNEP final report to Congress: assess- modeling in a GIS environment. New York: ments and scientific basis for management John Wiley and Sons. 392 p. options. Davis, CA: University of California, Centers for Water and Wildland Resources. McCool, S.F.; Troy, L.R. 1998. Estimating effects 1528 p. Vol. 2. of natural resource policy: the impact of the ecological fallacy. RJVA 97-8016. Missoula, U.S. Department of Agriculture, Forest Serv- MT: U.S. Department of Agriculture, Forest ice. 1999. National Forest System land and Service, Rocky Mountain Research Station. resource management planning. Federal Reg- 34 p. ister. 64(192): 54074-54112. Quigley, T.M.; Arbelbide, S.J., tech. eds. 1997. U.S. Department of Agriculture, Forest Serv- An assessment of ecosystem components in ice; U.S. Department of the Interior, Bureau the interior Columbia basin and portions of the of Land Management. 1994. Record Klamath and Great Basins. Gen. Tech. Rep. of decision for amendments to Forest Service PNW-GTR-405. Portland, OR: U.S. Depart- and Bureau of Land Management planning ment of Agriculture, Forest Service, Pacific documents within the range of the northern Northwest Research Station. 2066 p. spotted owl. [Place of publication unknown]. 74 p. [plus attachment A: standards and guide- Raettig, T.L.; Christensen, H.H. 1999. Timber lines]. harvesting, processing and employment in the Northwest Economic Adjustment Initiative re- Washington Employment Security Depart- gion; changes and economic assistance. Gen ment. 1997. User’s guide to labor market Tech. Rep. PNW-GTR-465. Portland, OR: U.S. information. Olympia, WA: Labor Market and Department of Agriculture, Forest Service, Economic Analysis Branch. 32 p. Pacific Northwest Research Station. 16 p. Raettig, T.L.; Elmer, D.M.; Christensen, H.H. 2001. Atlas of social and economic conditions and change in southern California. Gen. Tech. Rep. PNW-GTR-516. Portland, OR: U.S. De- partment of Agriculture, Forest Service, Pacific Northwest Research Station. 66 p.

66 Land Use and Land Cover Dynamics: Changes in Landscapes Across Space and Time

Ralph J. Alig1

Introduction strategies (e.g., Franklin 1989), warranting a more explicit treatment of the human system in con- Given interactions among terrestrial, aquatic, junction with ecosystem properties. Integrating the and socioeconomic systems, planning at the land- human dimension more directly into analysis sys- scape level could benefit from including informa- tems is critical, as choices made by resource and tion from different scales. This paper offers ex- land managers become increasingly important for amples of broader systems views that may inform the long-term sustainable use of our natural re- natural resource managers and policymakers sources (e.g., biodiversity, wildlife habitat, carbon dealing with restoring and maintaining healthy sequestration) within the context of suitable hu- ecosystems while addressing the demands of man welfare. people. World growth in human populations and income has resulted in social, economic, and Examining the state of the land is a key part of technological changes that profoundly affect the sustainability and livability analyses. Land use global management and use of natural resources. changes can provide opportunities to help society The world population will continue to grow, possi- adjust to changing demands for and supplies of bly from 5.9 billion in 1998 to 8.9 billion by 2050. renewable resources from the Nation’s forest and The U.S. population also will continue to grow, aquatic ecosystems. Interfaces between land particularly in the southern and western regions, uses, such as riparian zones on forest and agri- with a projected increase of more than 120 million cultural land, should be considered alongside people by 2050. Projected increases in population growing urban areas and increased fragmentation and income will, in turn, increase demands for of some forests. In this paper, I examine drivers of renewable resources. Demographic shifts, such changes in the large private forest land base, and as the aging of the population and increasing eth- discuss changes in land use and land cover by nic and racial diversity, will affect the patterns of major categories. demand for natural resources. Land Base Changes Over Time Growing human populations and associated use and consumption of the Earth’s natural resources From 1800 to 1930, forest land area in the United are placing more pressure on the chemical, bio- States declined by some 300 to 350 million acres logical, and ecological systems of the planet. (Clawson 1979), primarily in the East. This was Human activities are significant drivers of environ- about a one-third loss in forest cover. Some of mental changes on the mosaic of land uses and this land was employed for urban and infrastruc- land ownerships (e.g., Cohen et al. 2002). Analy- tural developments, but the largest portion was sis of the ecosystem response to changes in cleared and converted to agriculture. These land these human-derived environmental factors is use changes reflected federal policies of the time warranted at multiple scales, depending on the to transfer the original public domain to private policy-relevant questions. Changes in societal hands and to expand agricultural production. With values are part of the dynamics affecting forest the closing of the public domain, establishment of ecosystems and influencing land management permanent federal forest reserves, conversion of most suitable nongovernment forest lands to 1 Team leader and research forester, U.S. Department of some form of cropping or pasture, and dramatic Agriculture, Forest Service, Pacific Northwest Research improvements in agricultural productivity, the net Station, Forestry Sciences Laboratory, 3200 SW Jefferson movement of land between forestry and agricul- Way, Corvallis, OR 97331; Tel: 541-750-7267; e-mail: [email protected]; [email protected]. ture has become far less marked. Since 1950,

67 U.S. forest land has declined about 4 percent, need to be factored into landscape analyses, with the largest losses being to developed uses given implications for fish habitat, riparian areas, (Alig et al. 1999b). wetlands, and recreational opportunities. Although the large amount of forest land clearing In addition to an expanding population, the U.S. for agriculture of the last century is probably a population is aging. Early in the 20th century, 1 in phenomenon of the past, continued expansion of 25 Americans was over age 65. In 2000, 1 in 8 urban and developed areas is likely (e.g., Alig and was over 65 (USDC Bureau of the Census 2002). Healy 1987). In 1994, forests covered 65 percent Other demographic changes include continuing of nonfederal land in the Pacific Northwest, and shifts in ethnic and racial compositions. The these forests are under increasing pressure of changing U.S. demographic composition likely conversion. Oregon’s and Washington’s popula- will impact demands for forest and range re- tions have increased faster than the national aver- sources and how resources are managed (USDA age over the last decade. Most of the increase Forest Service 2001). was in the western part of the region, especially Americans’ incomes have grown substantially (in along the I-5 corridor. For example, Portland real terms or constant dollars) in recent decades, was one of the fastest growing cities in the United and average per capita personal incomes in Or- States over the last decade. Projections are that egon have increased more than 75 percent over the Pacific Northwest’s share of total population the last 30 years. Projected increases in discre- will increase, as the Nation’s population also tionary income will increase demands for renew- expands. able resources, but may also lead to further con- Urban sprawl and livability are issues receiving version of forests for developed uses. increased attention around the Nation, as the The United States in the last decade of the 20th Nation becomes more urban, with more than 75 century was in a period of economic growth, as percent of U.S. residents now living in cities stock markets experienced relatively large gains. (USDC Bureau of the Census 2002). In the Pa- In the new millennium, wealthier nations are likely cific Northwest (PNW), approximately 400,000 to have a greater ability to accommodate restora- acres of forest were converted to urban and de- tion and maintenance of ecosystems than are veloped uses between 1982 and 1997 (USDA those nations with economies under serious NRCS 2001). Such changes can interact with stress. However, some development patterns may other environmental alterations, such as habitat increase vulnerability of land-based systems in fragmentation, wetland loss, loss of biodiversity, spite of greater wealth. With more assets and and water pollution, and the cumulative impacts infrastructure located in coastal areas, this may on any particular ecosystem need to be consid- complicate rehabilitation of ecosystems home to ered carefully. Pacific salmon. Such trends also may increase vulnerability of coastal areas to pollution and to Socioeconomic Trends extreme events that are under study for any link- A number of socioeconomic trends affect options ages to global climate change (e.g., Easterling for restoring and maintaining ecosystems, and et al. 2000). affect the supply and demand of land-based The United States is also currently in the midst of goods and services. Along with the move to cities, rapid changes and advancements in technology more and more people are moving to the Nation’s capabilities in many economic sectors. The new coasts. Some 53 percent of the U.S. population information technology has affected agricultural lives in the 17 percent of the land in the coastal production and the forest sector to a lesser de- zone (within 50 miles of a coast). Further, the gree. Socioeconomic conditions will influence largest population increases over the next several how the technology is disseminated, and this will decades are projected to continue to be in coastal affect how well our region and country can effect areas. Increases in population and personal in- improvements in the state of the land. With grow- come tend to lead to reductions in aggregate for- ing and wealthier populations, development pres- est area (e.g., Alig and Healy 1987). Such trends sures and recreational use stresses on forests, wetlands, and riparian areas are likely to increase.

68 Types of Land Use and Land Cover expected or inadvertent, on another (e.g., forest Dynamics sector vs. agriculture sector). In recent years policies have increasingly focused on riparian zones Five categories of change involving the land base and incentives for promoting more favorable are (1) afforestation, (2) deforestation, (3) owner- aquatic habitat for salmonid stocks. Many past ship, (4) forest type transitions, and (5) land man- studies have examined policy impacts by either agement intensity (Alig and Butler 2002, Alig et al. (1) ignoring spillovers in other sectors or (2) sim- 2002). Such changes arise from an interaction of ply “adding up” impacts across the two sectors, biophysical, ecological, and socioeconomic pro- ignoring feedbacks or interactions through the cesses and forces, often operating at a variety markets for land. At times, this also has included of scales. For example, market forces tend to ignoring linkages between resource conditions on operate at much larger scales than biophysical public and private lands, and associated intercon- processes commonly studied at microlevels such nections via markets. as stands or reaches of a stream. I will provide a brief overview of the types of land base changes, Deforestation with illustrations of temporal and spatial considerations drawn from empirically based studies. The Deforestation is the conversion from forest to example studies range in geographic scope from nonforest use. Between 1982 and 1992, about the Oregon Coast Range province (Spies, this 250,000 acres were deforested on nonfederal volume) to the global level of analyses for current land in the region. The annual rate of conversion investigations of climate change. of forests to urban/developed uses increased in the 1990s to 32,000 acres, from the approxi- Afforestation mately 25,000 acres in the 1980s (USDA NRCS 2001), with stronger population growth and more Afforestation is the forestation, either by human economic activity. In total, between 1982 and or natural forces, of nonforest land. Two examples 1997, the net outcome of land use shifts was are planting conifers on former cropland and a regional loss in forest area of approximately former Christmas tree plantings that are allowed 400,000 acres. The destination of most converted to revert to “wild” forest. In the Pacific Northwest forested acres was to urban and developed uses. (Oregon and Washington), most afforested acres were formerly in pasture and range use. Tree Ownership Change planting is a major component of afforestation. Approximately 10 percent of the area of U.S. The main ownership change between traditional private tree planting in 1994 was in the Pacific ownership categories has been between forest Northwest in contrast to more than 80 percent in industry (FI) and nonindustrial private forest the South. Most tree planting in the South was on (NIPF) owner classes. The FI land is owned by nonindustrial private lands, whereas in the PNW companies for the purpose of timber production region, the majority was on forest industry lands. (see for example, Zheng and Alig 1999). The Nonfederal forest area in both Washington and NIPF lands compose the remainder of the private Oregon has steadily declined between 1982 and forest land base. Forest industry owns most of the 1997 surveys by the USDA Natural Resources private timberland in the PNW region, and gener- Conservation Service (2001), indicating that con- ally manages forest land intensively for timber versions of forest to other uses, primarily devel- production. Land exchanges among private own- oped uses, have in net been larger than the ers have been more common than exchanges amount afforested. The regional net loss has between private and public owners (Zheng and been more than 300,000 acres of forest over that Alig 1999). For example, Forest Inventory and 15-year period. Analysis (FIA) surveys in western Oregon indicate a net shift of 750,000 acres from NIPF to FI own- Tree planting has been a target of various poli- ers for 1961-94. Such changes in ownership can cies. The United States has a long history of for- result in notably different land management, par- est policies that jointly pursue both economic and ticularly as some private owners respond to mar- ecological objectives, often involving policy instru- ket changes prompted by reductions in federal ments designed to affect land use or cover. Public timber supply. policies aimed at one sector have impacts, either

69 With mergers, acquisitions, and other institu- harvest. Even for those stands not disturbed, tional changes, there also have been significant 12 percent of Douglas-fir stands on FI lands and changes within the industrial ownership category 23 percent on NIPF lands changed forest type. (e.g., combining Willamette and Weyerhaeuser Such findings suggest that recognition of owner- companies in 2002). The diverse NIPF owner ship differences can be important, beyond the category also continues to change in composi- public vs. private classifications sometimes used. tion. Along with this, forest fragmentation and parcelization and other factors may result in more Land Management Intensity noncorporate individual owners (e.g., Sampson Within a given forest type, investments in forest 2000), but they are likely to have smaller tract management can significantly influence options sizes on average. Accompanying changes in tract for different mixes of goods and services. Man- size may be changes in the values and percep- agement intensity includes the timing and type of tions of forest landowners. New owners some- harvests, which have differed notably over time by times bring different land management attitudes ownership in this region. Investments in improved compared to traditional owners, with changing planting stock on private land have allowed higher values often the result of exurbanization, or urban forest production per acre in some areas, lessen- residents moving to rural settings. ing pressures on other areas looked upon for nontimber goods and services. From a broad Forest Type Transitions perspective, demands for different mixes of land- Within the land base retained in forests over time, based goods and services are likely to shift over changes in forest cover types are caused by a time with a growing and wealthier population. combination of natural and human-related forces. Because some forest-based processes are de- Earlier literature dealing with forests focused on cades or centuries in length, planning such invest- forest succession and natural disturbances, with ments with adequate lead time is critical. emphasis on natural forces prompting forest type changes. However, human-caused disturbances Assessments of Land Base Changes such as timber harvests now dominate in most PNW forests (Alig et al. 2000, Cohen et al. 2002). The 2000 Resources Planning Act (RPA) assess- Over the last FIA survey cycle, human-caused ment by the USDA Forest Service (2001) is de- disturbances were more frequent, by at least an signed to examine the current situation of the order of magnitude, than recorded natural distur- Nation’s forest and rangeland ecosystems and bances. The probability of harvesting on private assess the prospective situation over the next five lands was about 2 to 10 times more likely than decades. In support of that national assessment, full- or partial-replacement disturbances from wild- research pertaining to drivers of land use change fire, the primary natural disturbance agent of the at macroscales indicates that major determinants past. In the approximatly 10 years between FIA of land use changes are population, personal measurements, timber harvests affected about income, and potential profits from land enterprises 20 percent of private timberland, whereas fire such as forestry or agriculture (Alig et al. 2002). affected less than 1 percent. For example, the Forest and Agricultural Sector Model (Adams et al. 1996a, Alig et al. 1998) has Given this disturbance backdrop, clearcutting endogenous determination of land use changes. Douglas-fir (Pseudotsuga menziesii (Mirb.) The linked model of the U.S. agriculture and for- Franco) stands on FI lands resulted in 15 percent est sectors has land demands and supplies in being replaced by hardwood (primarily red alder both sectors that are endogenous and determined [Alnus rubra Bong.]) stands, and another 3 per- simultaneously with demands and supplies for cent went to other types. Clearcutting red alder products from those lands. One possible objective stands on FI lands resulted in 72 percent being function in the model is economic maximization replaced by other forest types. In contrast, only (Adams et al. 1996b). Examination of land use 42 percent of red alder stands on NIPF lands changes within a multisector context is important changed forest types after clearcutting, reflecting for assessing alternative futures of “sustainable fewer intentional type-conversion efforts after forestry,” “sustainable agriculture,’’ “sustainable communities,’’ or some composite.

70 Owner Behavior ture, and urban/developed uses, are becoming more evident. Natural resource policy issues now A long-standing issue in forest resource supply often involve forest and nonforest components in studies is the likely behavior of private owners the overall system. For example, the condition of and how behavior may be affected by incentives riverine systems is affected by a broad range of and institutional factors. In addition to changing land uses, and opportunities for improvements characteristics of the owner population over time, exist on both wildlands and other lands (e.g., we also need to consider how changes in owner- pastureland in riparian zones). Efforts to better ship objectives may impact the resources. The understand how systems function, especially ter- larger, publicly owned companies are affected by restrial, aquatic, and socioeconomic systems, shareholder expectations, as well as pursuing should bear in mind policy-relevant questions and acceptable rates of return on investment within a consider the marginal costs versus marginal ben- changing institutional environment of incentives efits of undertakings. and regulations. The NIPF owners, in particular, possess multiple objectives, causing them to re- From a long-term conservation perspective, pur- spond to socioeconomic forces and policies in suit of new options for improvements in natural complex and sometimes unpredictable ways. resource conditions is increasingly faced with Nontimber services likely could be enhanced ef- compatibility issues among competing interests fectively by targeting incentive programs toward in the land and incentives for owners. Efforts to select groups of NIPF owners (Kline et al. 2000). better align commercial uses of the forests with Other research has found that financial incentives other conservation objectives has led to increased are not as critical to owners’ decisions in foresting interest in what is being called sustainable for- riparian zones as some have suggested, and estry, although there are similar efforts tied to other factors include “neighborly concerns’’ and other major competing interests in the land, concerns about restrictions on land management such as sustainable agriculture or sustainable and loss of flexibility (Kingsbury 1999). Behavior communities. by the diverse NIPF owner group also is affected When one looks at managing forested systems by whether owners live on their forested property for multiple uses within ecological limits, then in contrast to absentee owners. Another factor is consideration of capital markets and other inter- increasing density of people per forested unit, connections between sectors of the economy especially in areas such as the Puget Sound area. become more important. Policy issues associated with sustainability and planning information to aid Future Directions in collective decisions about managing natural The landscape comprises a mosaic of the major resources are affected by changing product mar- land uses, which reflects and warrants a wide ket and institutional conditions, as in recent de- view of forces at work in shaping the landscape, cades with changes tied to marked reallocation in both temporal and spatial dimensions. Projec- of rural land. Improvements in U.S. agricultural tions indicate that demographic and other forces production technology allowed expanded agricul- will continue to affect that mosaic of land uses. tural production on a fairly constant land base. If the U.S. population increases by 126 million Less pronounced and in a lagged fashion, forest people over the next 50 years, along with a rise volumes have increased on a timberland base in personal income, that could result in a net loss that is 4 percent smaller than the one in 1952. of 15 million acres of timberland (Alig et al. 2002, Demand for wood products is expected to keep 2003). Future improvements in policy could be growing, driven by the same population increases facilitated by acknowledging different ownership and economic development that affect demands characteristics and a broader representation in for other major land uses. However, intensified analyses and policy deliberations of those with timber management on the supply side repre- competing interests in the land, including public sents some of the largest projected changes in- information about actions and possible conse- volving private timberland (Alig et al. 1999a, quences. With increasing populations and pres- 2002), and could act to moderate or temper up- sures and conflicts on the land, relationships ward pressure on future wood prices. Private tim- among major land uses, such as forestry, agricul- berlands have the biological potential to provide

71 larger quantities of timber in an environmentally Alig, R.; Benford, F.; Moulton, R.; Lee, L. sound manner than they do today. Many of the 1999b. Long-term projections of urban and opportunities for intensified forest management developed land area in the United States. In: can be undertaken with positive economic returns, Otte, K., ed. Balancing working lands and de- but changing social and institutional aspects may velopment [CD-ROM]. Philadelphia, PA: Ameri- affect actualization. Most of the timber manage- can Farmland Trust and Land Trust Alliance: ment intensification opportunities in the United [Pages unknown] States are in the South and the Pacific Northwest Alig, R.; Butler, B. 2002. Land use and forest west side. Further, more forests around the world cover projections for the United States: 1997- are managed today compared to earlier decades, 2050. In: Proceedings of the 2001 national and this includes an expanding area of relatively Society of American Foresters (SAF) conven- productive plantations. Even if projected intensifi- tion. Bethesda, MD: Society of American For- cation were implemented, most of the NIPF esters: 93-115. timberland would still be concentrated in low management-intensity classes that involve naturally Alig, R.; Healy, R. 1987. Urban and built-up land regenerated stands. Sustainability analyses can area changes in the United States: an empiri- be enhanced if both land use and land investment cal investigation of determinants. Land Eco- options are examined. Analyses should be explicit nomics. 63: 215-226. with respect to timing of tradeoffs in addition to Alig, R.; Mills, J.; Butler, B. 2002. Private timber- the growing attention given to spatial details. This lands: growing demands, shrinking land base. would increase the usefulness of such tools in Journal of Forestry. 100 (2): 32-37. policy analyses, such as in the 2000 RPA assessment and forest carbon analyses. Alig, R.; Zheng, D.; Spies, T.; Butler, B. 2000. Forest cover dynamics in the Pacific Northwest Metric Equivalents west side: regional trends and projections. Res. Pap. PNW-RP-522. Portland, OR. U.S. 1 acre = 0.405 hectares Department of Agriculture, Forest Service, Pacific Northwest Research Station. 22 p. Literature Cited Alig, R.J.; Plantinga, A.J.; Kline, J.D. 2003. Adams, D.; Alig, R.; McCarl, B. [et al.]. 1996a. Land use changes involving forestry in the The forest and agricultural sector optimization United States: 1952 to 1997, with projections model (FASOM): model structure and policy to 2050. Gen. Tech. Rep. PNW-GTR-587. applications. Res. Pap. PNW-RP-495. Port- Portland, OR: U.S. Department of Agriculture, land, OR: U.S. Department of Agriculture, For- Forest Service, Pacific Northwest Research est Service, Pacific Northwest Research Station. 92 p. Station. 60 p. Clawson, M. 1979. Forests in the long sweep of Adams, D.; Alig, R.; McCarl, B.A. [et al.]. American history. Science. 204: 1168-1174. 1996b. An analysis of the impacts of public timber harvest policies on private forest man- Cohen, W.; Spies, T.; Alig, R. [et al.]. 2002. agement in the U.S. Forest Science. 42(3): Characterizing 23 years (1972-1995) of stand 343-358. replacement disturbance in western Oregon forests with Landsat imagery. Ecosystems. 5: Alig, R.; Adams, D.; Chmelik, J.; Bettinger, P. 122-137. 1999a. Private forest investment and long-run sustainable harvest volumes. New Forests. 17: Easterling, D.; Evans, J.; Groisman, P. [et al.]. 307-327. 2000. Observed variability and trends in extreme climate events: a brief review. Bulletin Alig, R.; Adams, D.; McCarl, B. 1998. Impacts of of the American Meteorological Society. 81: land exchanges between the forest and agri- 417-425. cultural sectors. Journal of Agriculture and Applied Economics. 90(5): 31-37.

72 Franklin, J.F. 1989. The “new forestry.” Journal of U.S. Department of Agriculture, Natural Re- Soil and Water Conservation. 44(6): 549. sources Conservation Service. [NRCS]. 2001. The national resources inventory, 1997. Kingsbury, L. 1999. Oregon’s conservation re- Revised. Washington, DC. serve enhancement program: likely participa- www.nhq.nrcs.usda.gov/NRI/1997. [Date last tion and recommendations for implementation. accessed unknown]. Corvallis, OR: Oregon State University. M.S. thesis. U.S. Department of Commerce, Bureau of the Census. 2002. Gateway to census 2000. Kline, J.; Alig, R.; Johnson, R. 2000. Fostering Washington, DC. www.census.gov. the production of nontimber services among (10 November 2003). forest owners with heterogenous objectives. Ecological Economics. 33: 29-43. Zheng, D.; Alig, R. 1999. Changes in the nonfederal land base involving forestry in western Sampson, N. 2000. People, forests, and forestry: Oregon, 1961-1994. Res. Pap. PNW-RP-518. new dimensions in the 21st century. In: Portland, OR: U.S. Department of Agriculture, Sampson, N.; DeCoster, L., eds. Proceedings, Forest Service, Pacific Northwest Research forest fragmentation 2000. Washington, DC: Station. 22 p. American Forests: 53-59. U.S. Department of Agriculture, Forest Serv- ice. 2001. 2000 RPA assessment of forest and range lands. Report FS-687. Washington, DC. 78 p.

73 Worldviews Panel

74 Private Landowner Perspective on Landscape Management

Carl F. Ehlen1

Presentation Objectives These reasons for ownership will shape the landowners’ goals and objectives and thus how they Mason, Bruce & Girard has been asked to give will proceed with operations on their property. you our impressions of the private landowner per- Ultimately, these reasons also will shape how they spective on landscape management. To do this will form any perspectives on landscape manage- subject justice, we need to discuss private land- ment. owner goals and objectives, state and federal regulations, and forest land management issues State and Federal Regulations as they relate to landscape management. As I prepared these remarks, I have drawn on my The three Western states (Washington, Oregon, nearly 30 years of experience within the forest and California) have very comprehensive laws products industry while with Georgia-Pacific Cor- governing timberland operations. These laws poration and through the contacts I have made primarily regulate forest management operations while with Mason, Bruce & Girard. I have tried to that have the potential to impact public resources capture questions we hear being asked and to such as fish and wildlife habitat and clean water. hopefully give you an understanding of why these They may be modified somewhat, depending on questions are important to the private forest land- the size of a given ownership, although generally owners within our region. This region needs diver- speaking most will apply to all landowners if man- sity, flexibility, room for dissension, and more agement activities occur. nonbureaucratic processes. Any perceived ben- Landowners, regardless of size of holdings, must efits of landscape management, at this point in determine how best to fit their goals and objec- time, are being viewed as far outweighing the tives with how these laws impact their individual costs, which will be borne, in a large part, by the ownership. For example, a landowner with small individual forest-land owner. acreage with fish-bearing streams flowing through the property may have few management options Private Landowners’ Goals and as compared to a landowner whose property con- Objectives tains only non-fish-bearing streams and, who The first question we need to ask is, Why do therefore, may have several management op- people own forest land? What is their driving moti- tions. Corporate ownerships generally are large vation? Is it profit, is it enjoyment of being able to enough to work around the regulations and still manage a forest based on their own interests, is it meet corporate mandates. for recreational purposes, or is it ownership for As an example of how regulations can impact the the beauty of it? management of our private forest lands, Mason, Although there are some who own timberland to Bruce & Girard completed a study for the Oregon enjoy it, most private landowners manage their Forest Industries Council in the early 1990s to test lands with the objective of profiting from it in the requirements of Oregon Senate Bill 1125. This some manner, whether on an annual basis or bill, which was backed by forest landowners, set over some other timeframe. There are many dif- new requirements designed to spread timber har- ferent ideas on what forest land management vest over the landscape by restricting clearcut should look like, from very intensive to a hands-off areas to 120 acres in size and requiring regenera- approach. tion to be 4 feet high, 4 years old, and “free to grow” before harvesting an adjacent timber stand. “Free to grow” is defined as “a tree or stand of 1 Forestry consultant, Mason, Bruce & Girard, Inc., 707 SW Washington Street, Suite 1300, Portland, OR 97205; Tel: 503- well-distributed trees, of acceptable species and 224-3445; e-mail: [email protected] good form, with a high probability of remaining or

75 becoming vigorous, healthy, and dominant over Landscape Management Perspectives undesired competing vegetation.” For the purpose and Questions of this definition, trees are considered well distributed if 80 percent or more of the area contains at Just as management goals and objectives differ least 200 trees per acre and not more than 10 among private landowners, so do their perspec- percent contains fewer than 100 trees per acre. tives, and therefore their definitions, of landscape The study determined that the “free to grow” re- management. To our knowledge, there is no ob- quirement, when combined with leave areas be- jective, all-encompassing definition of landscape tween harvest units and maximum regeneration management or what it means to the individual harvest unit size, created the following: landowner in terms of impacting their timber resource. 1. Significant impacts on the size, number, distribution, and timing of regeneration harvests. Is “landscape management” providing for all public resources on every acre, or is it providing for all 2. A slowing of the rate of harvest on individual public resources through a mosaic of age and tracts. size classes of forests across the landscape? 3. The potential for significantly increasing the Is it important that all landowners provide this miles of roads, which not only adds to a land- mosaic, regardless of size of ownership? owner’s cost but also adds to the probability for increased environmental damage. Does landscape management include all landowners—private, state, and federal alike—in a 4. Land being taken out of production. given watershed or viewshed or larger area? 5. Increased stand fragmentation as the maxi- Are all landowners to be treated equally? mum regeneration harvest size is further reduced, thus potentially jeopardizing the ability Will a landowner be able to factor in individual or of a landowner to fully harvest the annual corporate goals and objectives? growth on the property. Given the above, we believe that the following, at 6. A significant reduction in the size of regenera- a minimum, must be studied and analyzed before tion harvest units with the average always private landowners would even consider buying being less than the maximum allowed. into any landscape management scheme. 7. More broken ownership, which makes it more Economic Analysis difficult to efficiently schedule harvesting activities. “Land” or “soil” expectation is an indicator of the value of a piece of bare land as a forestry invest- 8. More dispersion of regeneration units over the ment. As additional regulations are imposed, soil landscape. expectation values on low-site lands begin to be- Private landowners have long supported forest come marginal, particularly east of the Cascade regulations backed by scientific evidence. Now, crest, but the same is true on the west side as however, as regulatory proposals pass from sci- well. This concern begins to beg the issue of dis- entific to political, private landowners are becom- investment of lower site lands. ing more and more leery of what might be ex- Current state and federal regulations are costly to pected of them as they continue to manage the private landowner. To date we know of no their property and risk the potential loss of their analysis to determine the additional costs and investments. benefits to the private landowner operating under Will more regulations, particularly if seen as politi- any landscape management framework. An eco- cal, drive more disinvestment of forest land? nomic analysis is imperative, as any regulations

76 in addition to what we have today would be ex- tection attributes as opposed to keeping them tremely costly. As environmental regulation in- productive and profitable while continuing to pro- creases, we arrive at the point of getting small tect our public resources? marginal benefits at an enormous cost, most of Is there a balance that can be reached between which will be borne by the landowner. private timber production and habitat require- Has anyone taken the time to do an analysis of ments? this sort?

Rotation Length Issues Administration of Landscape Landowners must be able to determine the requi- Management Requirements site rotation length for their property based on the A great deal of thought must go into how land- productive capacity of the land and their goals scape management issues are to be adminis- and objectives, while at the same time being able tered. These issues must be examined by each to harvest their timber in response to market landowner rather than on a watershed or land- changes in accordance with current regulations. scape basis. Problems will erupt as soon as one How will landscape management schemes allow landowner is restricted or impacted to a greater for this within the required timing? degree than other landowners. How will final rotations be impacted? Who will be the regulator? Will landowners be allowed to set their own rota- Who determines if landowners are complying with tions and harvest ages depending on silvicultural the requirements? objectives and market fluctuations? How will anyone be able to make a determination in an objective and legal manner? Silvicultural Issues With the serious decline in federal timber sale Summary Comments volume, combined with the continuing demand for Our principal issue at this point is that we do not wood products, many private landowners are en- know what landscape management looks like nor deavoring to harvest timber equal to the growth do we know what, if anything, a private landowner rates of their property. Many private landowners will have to give up. Is this another attempt to are investing heavily in various silvicultural activi- further regulate the private landowner? Will land- ties to assure that their forest land is producing scape management become a form of “central- timber at its productive capacity. These activities ized planning?” include, among others, planting genetically improved seedlings, brush control, early and mid- We do not know of any private landowner in the rotation fertilization, and midrotation thinning. We Northwest who does not embrace the need for the are seeing a quality of forest management activi- protection and the sustainability of public re- ties that is unprecedented. And yet, landowners sources. The current regulatory framework is continue to improve on these efforts. inherently designed to provide for both. Will these efforts and investments go for naught? How one landowner’s goals fit with an adjacent In other words, will the financial investments in landowner’s goals and how a given landowner’s planting, vegetation control, precommercial thin- operations fit within the definition of managing on ning, etc. made early in the life of a rotation be a landscape scale, we assume, are still to be de- lost if a landowner is not allowed to harvest his termined. In the end, however, private landowners crop? must be included in any discussions of landscape management, particularly if regulatory proposals How does even-age management fit with land- are being considered. There is too much value at scape management goals and objectives? stake for discussions on this issue to proceed Are we as a society only interested in how our otherwise. private forest lands look and their resource pro-

77 Back to the Future: The Value of History in Understanding and Managing Dynamic Landscapes

Penelope Morgan1

Describing past conditions is part of a coarse-filter Defining Historical Range of Variability management strategy for sustaining biological diversity (Cissel et al. 1994, Haufler 1994, Haufler Historical range of variability characterizes fluc- et al. 1996, Hunter et al. 1989, Swanson et al. tuations in ecosystem conditions or processes 1994). A description of past ecosystem structure over time (Landres et al. 1999, Manley et al. 1995, and its variability is useful in exploring the causes Morgan et al. 1994, Swanson et al. 1994), and and consequences of change in ecosystem char- thus provides researchers and managers with a acteristics over time, and thus for understanding reference against which to evaluate recent and the set of conditions and processes that sustained potential ecosystem change. Landres et al. (1999) ecosystems prior to their recent alterations by defined natural variability as “the ecological condi- humans. It provides a context for interpreting tions and their variability over space and time natural processes, especially disturbance, and it relatively unaffected by people.” Thus, HRV de- allows variability in patterns and processes to be fines bounds of system behavior that remain rela- understood in terms of a dynamic system. Natural tively consistent over time. It is described for a resource managers can use such information to timeframe relevant to understanding the behavior evaluate the magnitude, direction, and causes of of contemporary ecosystems and the implications landscape change, especially change induced by for management, usually prior to the intensive people. An assessment of current conditions rela- influence of European-Americans and with relative to the historical range of variability (HRV), tively consistent climate. Many people use a also called natural variability, also can be used to timeframe of 100 to 700 years before present to identify management goals, particularly where interpret successional dynamics and to character- those goals encompass sustainability, ecological ize past variation in forest structure and composi- restoration, conservation of biological diversity, tion, but that depends on the available historical and maintenance of natural processes (Hann and information and the ecosystem of interest. Histori- Bunnell 2001, Landres et al. 1999, Morgan et al. cal data on disturbance occurrence and effects 1994, Swanson et al. 1994, White and Walker are increasingly limited the further back in time we 1997). The concept is less useful when manage- look. Using recent centuries includes burning and ment objectives are narrowly focused on maximiz- other actions of American Indians, and encom- ing commodity production or sustaining the passes climatic variability, such as the Little Ice population of an endangered plant or animal Age. Hann et al. (1997) used the last 2,000 years (Landres et al. 1999). as the timeframe for HRV for the interior Colum- bia River basin. In this paper, I give a brief overview of the HRV concept, which is also called natural variability Utility of This Concept and reference variability, and its use in understanding and managing dynamic landscapes. I The utility of this concept is based on two con- briefly discuss the value and limitations of history, cepts: (1) past conditions and processes provide and then summarize what we have learned from context and guidance for managing ecological historical ecology that is of use to managers of systems today, and (2) spatial and temporal vari- dynamic landscapes. ability caused by disturbance is vital to ecosystem integrity (Landres et al. 1999). However, the utility 1 Professor, College of Natural Resources, University of will also depend on the management issues and Idaho, Moscow, ID 83844-1133; Tel: 208-885-7507; e-mail: the social and ecological context for decisions [email protected]. (Landres et al. 1999). The HRV often is used in

78 combination with societal values and other infor- points to large areas, and from short time inter- mation to identify desired future conditions vals (years and decades) to long time intervals (Haufler et al. 1996, Landres et al. 1999). Histori- (centuries) is problematic and fraught with error cal information has been used to guide manage- (Morgan et al. 2001). We can more readily recon- ment of waterflow on the Colorado River (Poff et struct past vegetation structure than past pro- al. 1997) and in the Everglades (Harwell 1997), cesses, and when we reconstruct past forest and in management of fire (Brown et al. 1994, structure, we are more confident about the num- Skinner and Chang 1996) and forest structure ber of large trees than we are about how many (Baker 1992, Camp et al. 1997, Cissel et al. 1999, small trees there were (Allen et al. 2002). Hierar- Gauthier et al. 1996, Hann and Bunnell 2001, chy theory can be helpful in interpreting data from Keddy and Drummond 1996, Ripple 1994). It also different spatial and temporal scales. has been used in ecological restoration (Allen et al. 2002, Covington and Moore 1997, White and Historical Range of Variability and Walker 1997), and to deepen our understanding Desired Future Conditions of the processes driving forest change (Lertzman et al. 1997; Lesica 1996; Mladenoff and Pastor Comparing current and desired future conditions 1993; Swetnam and Betancourt 1990, 1998; (DFCs) to HRV is useful for identifying manage- Wimberley et al. 2000). Ecosystem management ment strategies and priorities (fig. 1). I sometimes often relies heavily on a description of past vari- demonstrate this with a ball (existing conditions) ability in defining desired future conditions rolling around on a large plate (HRV). When con- (Christensen et al. 1996, Kaufmann et al. 1994, ditions are within the HRV, the ball rolls around on Manley et al. 1995). the plate. The edges of the plate keep the ball on the plate (the system is bounded), until something The greatest values for characterizing HRV are pushes it over the edge (outside of HRV). Then understanding (1) how the processes and factors the ball drops (change is rapid, and though it may driving ecosystem change differ from one place be somewhat predictable, the ball might not return and time to another, (2) the influence of driving to the plate). Further, I can hold the ball steady, variables in the past, and (3) how those driving but doing so takes energy on my part (an external variables might influence ecosystems today and subsidy to the system in the form of human ac- in the future (Landres et al. 1999). Managers de- tion). Now imagine a small plate representing veloped the HRV concept in a search for legally DFCs. This plate is smaller because DFCs are defensible strategies for sustaining biological di- typically more limited than HRV. It is possible to versity and ecological integrity. have all three stacked up: the ball on the small Data for describing HRV differ in temporal and plate and both on the large plate. We do have spatial scale (Morgan et al. 2001, Swetnam et al. forest ecosystems where current conditions are 1999). Both natural (fire-scarred trees, pollen, within the DFCs and those are within HRV. This charcoal, etc.) and human archives (old photo- includes, for example, some subalpine forests. In graphs, early survey data, etc.) are sources of such cases (row 1 of fig. 1), our management historical data (Swetnam et al. 1999). Although decision might be to maintain this desirable condi- we know more about fire history than about other tion, which we view as sustainable. We can, how- disturbances, none of our fire history data are ever, imagine other combinations. Many urban comprehensive across both time and space, and and agricultural areas, for instance, would be most are better at describing frequency than se- characterized with the ball on a small plate (cur- verity, size, or spatial pattern of disturbances. Few rent conditions within DFCs) but both separate adequately inform us about spatial pattern and from the large plate (both outside of HRV). dynamics at the landscape scale (Morgan et al. Whether or not such conditions are sustainable 2001, Swetnam et al. 1999). Also, we have more depends on whether we as a society are willing to data for points and small areas than we have for assume the external subsidies (e.g., financial landscapes and watersheds (Landres et al. 1999, subsidies to farmers, and fertilizer and herbicide Morgan et al. 2001, Swetnam et al. 1999). Thus, made from fossil fuels), effects (e.g., quality of life extrapolation is required, but extrapolation from and environment), and risks (e.g., climate changes).

79 Status Graphical representation Management action

HRV ≥ DFC ≥ CC |…xxx...| Maintain |______| HRV ≥ DFC ≠ CC |…..…...| xxx Restore |______| HRV ≠ DFC ≥ CC |______| |…xxx...| Evaluate carefully to assess risks, sustainability, and the ≥ ≠ HRV CC DFC |______xxx______| |…..…...| external subsidies of DFC; reevaluate DFC HRV ≠ DFC ≠ CC |______| |……...| xxx

Figure 1—Comparing both current ecological conditions (CC, xxx) and desired future conditions (DFC |———|), to historical range of variability (HRV, |_____|) is useful for identifying management actions. (Redrawn from Landres et al. 1999).

Thus, current conditions, DFCs and HRV may the use of HRV description, modeling succession, or may not overlap. All three or any two might be and identifying areas of departures from HRV as similar, or all might be different. Depending on a means for prioritizing for restoration in Colorado. social, economic, and political issues, restoration Their approach is similar to one applied in the is suggested where desired conditions are within Sierras by Caprio and Graber (2000). Cissel HRV, but current conditions are different (row 2 et al. (1999) contrast two approaches to land- of fig. 1). When DFCs differ from HRV or current scape management in the Blue River watershed conditions, desired future conditions may need to in Oregon. An approach based on using logging to be reevaluated (rows 3, 4 and 5 of fig. 1). External approximate past disturbance regimes while also subsidies will be required to maintain such a con- maintaining stream buffers created less forest dition, but people can choose to bear both the fragmentation than an alternative based on matrix subsidies and the risk of change to the system. and corridors under the Northwest Forest Plan. Ideally, scientists and managers can quantify the implications of different choices as the basis for What Have We Learned? informed decisions by the stakeholders. Haufler et al. (1996) suggest combining knowledge of HRV Historical ecology provides many lessons. For with knowledge of species needs to identify “ad- instance, we now understand that disturbance equate representation” of elements of ecosystem was and is ubiquitous. At any given time, most diversity. ecosystems are in some state of recovery from disturbance. Directly and indirectly, disturbances Examples of Applications of Historical strongly influence the structure, composition, diversity, and function of the ecological systems Range of Variability in Landscape we manage. Further, disturbances interact, and Management the relative importance of disturbances varies with Hann and Bunnell (2001) describe the develop- scale. For instance, it is likely that at some broad ment of landscape management prescriptions spatial scales, regional and global climate based upon assessment of changes from HRV. supercedes local fuel and topography in determin- They give examples at several spatial scales, ing fire patterns (Swetnam and Betancourt 1998). including the mapping of historical fire regimes, It is likely, for instance, that El Niño drives fire current fire regimes, and condition classes reflect- patterns across the Southwestern United States ing the degree of change for the conterminous (Swetnam and Betancourt 1990) and in the United States done by Hardy et al. (2001) and Patagonian region of South America (Kitzberger available on the Web (http://www.fs.fed.us/fire/ and Veblen 1997). The results of historical analy- fuelman). Hann and Bunnell (2001) also describe sis can sometimes be surprising. For instance,

80 Swetnam and Lynch (1993) found that spruce Implications and Challenges for budworm (Choristoneura occidentalis Freeman), Managers a defoliating insect in conifers, is more prevalent during wet than during dry years, counter to the We need new approaches to landscape planning commonly held notion that defoliation occurs in that effectively integrate the three very different drought-stressed trees. approaches currently applied.2 Traditional approaches to evening out the flow of commodities Scale matters. Wimberley et al. (2000) modeled have taught us a lot about modeling complex and the historical range of variability of old forests in interacting systems, but it is often difficult to incor- the Oregon Coast Range. The historical range of porate ecological uncertainty and change into this variability was progressively narrower as they approach. Conservation planning based upon increased the analysis area by orders of magni- corridors and matrices accommodates the needs tude from a small watershed to a large watershed of many species but doesn’t recognize that land- and then to a region. scapes are dynamic. Basing landscape manage- Changes in forests at the landscape scale are ment prescriptions on disturbance regimes builds great and variable (Hann et al. 1997). Forests upon our knowledge of how and why landscapes at high elevation have changed in different ways change (Cissel et al. 1994, 1999) but may not than forests at low elevations over the last 60 provide for the needs of species and people. years, and both biophysical and social factors Thus, we need to develop approaches to land- are correlated with different types and degrees scape planning that accommodate multiple spe- of change (Black 1998). Although changes in cies and processes in ecosystems subject to fire regimes are extensive, the changes are not introduced species, processes, and human the same everywhere (Hardy et al. 2001, http:// population pressure. For whitebark pine (Pinus www.fs.fed.us/fire/fuelman). albicaulis Engelm.) ecosystems, for instance, landscape planning must incorporate the needs Rigorous testing of hypotheses by using historical and dynamics of grizzly bears, Clark’s nutcrack- data has even greater potential for rapidly advanc- ers, squirrels, fire, the introduced blister rust, and ing ecological understanding. Time series analy- climate change, along with scenic values and sis, comparative case studies, and analyzing watershed protection. Historical information can synchronous events across broad spatial scales help us to understand this system and to build have taught us a great deal about the relative models, which are then used to compare alterna- influences of climate, land use, topography and tive futures for this ecosystem (Keane and Arno vegetation on fire regimes (Rollins et al. 2000a, 2001). 2000b, 2001; Swetnam 1993; Swetnam et al. 1999; Swetnam and Betancourt 1998), and about Human-induced changes to our global, regional, the interactions between insects, forests, and and local environments pose great challenges to climate (Swetnam and Lynch 1993). natural resource managers. Our climate is changing (IPCC 2001). Introduced species have altered Through modeling (e.g., Miller and Urban 1999, our environment as well (Vitousek et al. 1996). Wimberley et al. 2000), comparison of managed Some will find HRV less useful in addressing and unmanaged landscapes, and examination of management of ecosystems that are clearly historical data, we are coming to understand how undergoing changes that are to some degree and why landscapes change. That understanding unprecedented when coupled with human popula- is very important to managers of dynamic land- tion pressures. However, effective management scapes, even when those landscapes are subject to introduced species, structures, and processes. 2 Swanson, F. 1998. Personal communication. Research geologist, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR 97331; Tel: 541-750-7355; e-mail: [email protected].

81 will continue to depend on our understanding of Black, A.E. 1998. Forest patterning in the interior ecosystems. The ecological implications of cli- Columbia River basin: 60 years of social and mate change and the dynamics of ecosystems biophysical interaction. Moscow, ID: University reinforce the need for a broader focus on restor- of Idaho. 137 p. Ph.D. disseration. ing ecological integrity, resilience, and sustain- Brown, J.K.; Arno, S.F.; Barrett, S.W.; Menakis, ability, rather than on restoring some “vignette” J.P. 1994. Comparing the prescribed natural or past condition (Landres et al. 1999, White and fire program with presettlement fires in the Walker 1997). Selway-Bitterroot Wilderness. International The potential consequences to biological diversity Journal of Wildland Fire. 4: 157-168. are great if current forest conditions are outside of Camp, A.; Oliver, C.; Hessburg, P.; Everett, R. and remain outside of the historical range of vari- 1997. Predicting late-successional fire refugia ability. Consequences include increased vulner- pre-dating European settlement in the ability to stand-replacing fires where few ever Wenatchee Mountains. Forest Ecology and occurred historically, which can put soils, streams, Management. 95: 63-77. and fish at risk (Hann et al. 1997). This does not mean, however, that we must return our forests Capiro, A.C.; Graber, D.M. 2000. Returning fire completely to historical conditions to sustain bio- to the mountain: Can we successfully restore logical diversity. Thoughtful evaluation of change the ecological role of pre-Euroamerican fire from historical conditions; evaluation of social regimes to the Sierra Nevada? In: Cole, D.M.; needs; presence of exotic species, structures, McCool, S.F.; Borie, W., O’Laughlin, J., comps. and processes; and future climate all should play Wilderness science in a time of change confer- a role in identifying DFCs and the management ence. Vol. 5: Wilderness ecosystems, threats, alternatives to achieve them (Landres et al. 1999). and management Proc. RMRS-P-15-VOL-5. Wilderness ecosystems, threats, and manage- Acknowledgments ment. Ogden, UT: U.S. Department of Agricul- ture, Forest Service, Rocky Mountain I am deeply indebted to the many creative and Research Station: 233-241. thoughtful managers and scientists who helped me develop my ideas. Wendel Hann was one of Christensen, N.L.; Baruska, A.M.; Brown, J.H. the managers who originated the HRV concept. [et al.]. 1996. The report of the Ecological So- Discussions with him and with Tom Swetnam, ciety of America committee on the scientific Peter Landres, Dave Parsons, Bob Keane, Art basis for ecosystem management. Ecological Zack, Steve Bunting, and many others have Applications. 6(3): 665-691. helped me shape my ideas. Lee Harry, a silvicul- Cissel, J.H.; Swanson, F.J.; McKee, W.A.; turist on the Beaverhead National Forest, origi- Burditt, A.L. 1994. Using the past to plan the nated the bowl, ball, and plate demonstration, future in the Pacific Northwest. Journal of For- and thus his ideas were the genesis for figure 1. estry. 92: 30-31, 46. Literature Cited Cissel, J.H.; Swanson, F.J.; Weisberg, P.J. 1999. Landscape management using historical Allen, C.D.; Savage, M.; Falk, D.A. [et al.]. 2002. fire regimes: Blue River, Oregon. Ecological Ecological restoration of Southwestern ponde- Applications. 9(4): 1217-1231. rosa pine ecosystems: a broad perspective. Ecological Applications. 12(5): 1418-1433. Covington, W.W.; Moore, M.M. 1997. Restoring ecosystem health in ponderosa pine forests of Baker, W.L. 1992. Effects of settlement and fire the Southwest. Journal of Forestry. 95: 23-29. suppression on landscape structure. Ecology. 73: 1879-1887. Gauthier, S.; Leduc, A.; Bergeron, Y. 1996. For- est dynamics modelling under natural fire cycles: a tool to define natural mosaic diversity for forest management. Environmental Moni- toring and Assessment. 39: 417-434.

82 Hann, W.J.; Bunnell, D.L. 2001. Fire and land Keddy, P.A.; Drummond, C.G. 1996. Ecological management planning and implementation properties for the evaluation, management, across multiple scales. International Journal of and restoration of temperate deciduous for- Wildland Fire. 10: 389-403. est ecosystems. Ecological Applications. 6: 748-762. Hann, W.J.; Jones, J.L.; Karl, M.G. [et al.]. 1997. Landscape dynamics of the basin. In: Kitzberger, T.; Veblen, T.T. 1997. Influences Quigley, T.M.; Arbelbide, S.J., tech. eds. An of humans and ENSO on fire history of assessment of ecosystem components in the Austrocedrus chilensis woodlands in northern interior Columbia basin and portions of the Patagonia, Argentina. Ecoscience. 4: 508-520. Klamath and Great Basins, Gen. Tech. Rep. Landres, P.B.; Morgan, P.; Swanson, F.J. 1999. PNW-GTR-405. Portland, OR: U.S. Depart- Overview of the use of natural variability con- ment of Agriculture, Forest Service, Pacific cepts in managing ecological systems. Eco- Northwest Research Station: 363-1055. Vol. 2 logical Applications. 9(4): 1179-1188. Hardy, C.C.; Schmidt, K.M.; Menakis, J.P.; Lertzman, K.; Spies, T.; Swanson, F. 1997. Sampson, R.N. 2001. Spatial data for national From ecosystem dynamics to ecosystem man- fire planning and fuel management. Interna- agement. In: Schoonmaker, P.K.; von Hagen, tional Journal of Wildland Fire. 10: 353-372. B.; Wolf, E.C., eds. The rain forests of home: Harwell, M.A. 1997. Ecosystem management profile of a North American bioregion. Wash- of south Florida. BioScience. 47: 499-512. ington, DC: Island Press: 361-382. Haufler, J.B. 1994. An ecological framework for Lesica, P. 1996. Using fire history models to planning for forest health. Journal of Sustain- estimate proportions of old growth forest in able Forestry. 2(3/4): 307-316. northwestern Montana, USA. Biological Con- servation. 77: 33-39. Haufler, J.B.; Mehl, C.A.; Roloff, G.J. 1996. Using a coarse-filter approach with a species Manley, P.N.; Brogan, G.E.; Cook, C. [et al.]. assessment for ecosystem management. 1995. Sustaining ecosystems: a conceptual Wildlife Society Bulletin. 24(2): 200-208. framework. R5-EM-TP-001. Albany, CA: U.S. Department of Agriculture, Forest Service, Hunter, M.L.; Jacobson, G.L.; Webb, T. 1989. Pacific Southwest Region and Pacific South- Paleoecology and the coarse-filter approach to west Research Station. maintaining biological diversity. Conservation Biology. 2: 375-385. Millar, C.I.; Woolfenden, W.B. 1999. The role of climate change in interpreting historical vari- Intergovernmental Panel on Climate Change ability. Ecological Applications. 9: 1207-1216. [IPCC]. 2001. Climate change 2001: impacts, adaptation and vulnerability. http://www.ipcc.ch/ Miller, C.; Urban, D.L. 1999. Forest pattern, fire (1 May). and climate change in the Sierra Nevada. Eco- systems. 2: 76-87. Kaufmann, M.R.; Graham, R.T.; Boyce, D.A. [et al.]. 1994. An ecological basis for ecosys- Mladenoff, D.J.; Pastor, J. 1993. Sustainable tem management. Gen. Tech. Rep. RM-GTR- forest ecosystems in the northern hardwood 246. Fort Collins, CO: U.S. Department of and conifer region: concepts and manage- Agriculture, Forest Service, Rocky Mountain ment. In: Aplet, G.H.; Olsen, J.T.; Johnson, N.; Forest and Range Experiment Station. 22 p. Sample, V.A., eds. Defining sustainable forestry. Washington, DC: Island Press: 145-180. Keane, R.E.; Arno, S.F. 2001. Restoration concepts and techniques. In: Tomback, D.F.; Arno, Morgan, P.; Aplet, G.H.; Haufler, J.B. [et al.]. S.F.; Keane, R.E., eds. Whitebark pine com- 1994. Historical range of variability: a useful munities: ecology and restoration. Washington, tool for evaluating ecosystem change. Journal DC: Island Press: 367-401. of Sustainable Forestry. 2(1/2): 87-111.

83 Morgan, P.; Hardy, C.; Swetnam, T.W. [et al.]. Swanson, F.J.; Jones, J.A.; Wallin, D.O.; 2001. Mapping fire regimes across time and Cissel, J.H. 1994. Natural variability—implica- space: understanding coarse- and fine-scale tions for ecosystem management. In: Jensen, patterns. International Journal of Wildland Fire. M.E.; Bourgeron, P.S., tech. coords. Ecosys- 10: 1-14. tem management: principles and applications, Volume II. Eastside forest ecosystem health Poff, N.L.; Allan, J.D.; Bain, M.B. [et al.]. 1997. assessment. Gen. Tech. Rep. PNW-GTR-318. The natural flow regime. BioScience. 47: 769- Portland, OR: U.S. Department of Agriculture, 784. Forest Service, Pacific Northwest Research Ripple, W.J. 1994. Historic spatial patterns of old Station: 80-94. forests in western Oregon. Journal of Forestry. Swetnam, T.W.1993. Fire history and climate 92: 45-49. change in giant sequoia groves. Science. Rollins, M.; Swetnam, T.W.; Morgan, P. 2000a. 262: 885-889. Twentieth century fire patterns in the Gila/Aldo Swetnam, T.W.; Allen, C.D.; Betancourt, J.L. Leopold Wilderness Complex in New Mexico 1999. Applied historical ecology: using the past and the Selway-Bitterroot Wilderness Area to manage for the future. Ecological Applica- Idaho/Montana. In: Crossing the millennium: tions. 9(4): 1189-1206. integrating spatial technologies and ecological principles for a new age in fire management: Swetnam, T.W.; Betancourt, J.L. 1990. Fire- Proceedings from the Joint Fire Science con- southern oscillation relations in the Southwest- ference and workshop. Moscow, ID: University ern United States. Science. 249: 1017-1020. of Idaho: 161-169. Swetnam, T.W.; Betancourt, J.L. 1998. Meso- Rollins, M.; Swetnam, T.W.; Morgan, P. 2000b. scale disturbance and ecological response to Twentieth-century fire patterns in the Selway- decadal climatic variability in the American Bitterroot Wilderness Area, Idaho/Montana Southwest. Journal of Climate. 11: 3128-3147. and the Gila/Aldo Leopold Wilderness Com- Swetnam, T.W.; Lynch, A.M. 1993. Multicentury, plex. In: Cole, D.N.; McCool, S.F.; Borrie, W.T.; regional-scale patterns of western spruce bud- O’Loughlin, J., comps. Wilderness science in a worm outbreaks. Ecological Monographs. time of change conference: Vol. 5. Wilderness 63(4): 339-424. ecosystems threats and management. Proc. RMRS-P-15. Fort Collins, CO: U.S. Depart- Vitousek, P.M.; D’Antonio, C.M.; Loope, L.L.; ment of Agriculture, Forest Service, Rocky Westbrooks, R. 1996. Biological invasions as Mountain Research Station: 283-287. global environmental change. American Scien- tist. 84(5): 468-478. Rollins, M.G.; Swetnam, T.W.; Morgan, P. 2001. Evaluating a century of fire patterns in two White, P.S.; Walker, J.L. 1997. Approximating Rocky Mountain wilderness areas using digital nature’s variation: selecting and using refer- fire atlases. Canadian Journal of Forest Re- ence information in restoration ecology. Con- search. 31(12): 2107-2123. servation Biology. 5(4): 338-349. Skinner, C.N.; Chang, C. 1996. Fire regimes, Wimberley, M.D.; Spies, T.A.; Long, C.J.; past and present. In: Status of the Sierra Ne- Whitlock, C. 2000. Simulating historical vari- vada, Volume II: assessments and scientific ability in the amount of old forests in the Or- basis for management options. Wildland Re- egon Coast Range. Conservation Biology. sources Center Report 37. Davis, CA: Univer- 14(1): 167-180. sity of California, Centers for Water and Wild- land Resources: 1041-1070. Sierra Nevada Ecosystem Project final report to Congress.

84 85 Case Studies

86 Northwest Forest Plan

Nancy M. Diaz1

In the early 1990s, the forest industry in the Pa- agency officials; the selected alternative was cific Northwest was in a condition that has been adopted in 1994 (USDA USDI 1994) and became described as “gridlock” (FEMAT 1993). Because what is now known as the Northwest Forest Plan of mounting evidence that old-growth forest eco- (NFP) (fig.1). The NFP guides management of 24 systems were declining in health and integrity, million acres of federal forest lands (including and because of the listing or potential listing of national forest and BLM lands, as well as national species (such as the northern spotted owl (Strix parks) within the range of the northern spotted occidentalis caurina) and various salmon stocks) owl. by the U.S. Fish and Wildlife Service as endangered or threatened, forest management prac- The Northwest Forest Plan–An tices were under intense criticism. Research Ecosystem Management Approach suggested that a large proportion of the historical acreage of old growth had been lost to timber The Northwest Forest Plan was one of the earliest harvest, and that the remaining stands were and largest attempts to implement what was at heavily fragmented, causing risks to old-growth- that time a relatively new concept in public land dependent species (FEMAT 1993). In addition, use planning–that of managing whole systems, habitat loss from roads, culverts, and degradation including broad-scale vegetation patterns, bio- of riparian areas was contributing to problems physical processes, and whole suites of species. with fish populations. On federal forest lands This approach has become widely known as “eco- (those lands managed by the USDA Forest Ser- system management.” vice and USDI Bureau of Land Management The NFP treats the land base as a system of in- [BLM]) within the range of the northern spotted terconnected parts, providing a single conserva- owl (western Washington and Oregon, and north- tion biology strategy for wildlife, plants, fish, and ern California), a series of court injunctions fungi associated with late-successional and old- against logging and delays from appeals of timber growth forests. The old-growth orientation of the sales had resulted in a dramatic decline of timber NFP distinguishes it from other biodiversity con- harvest operations, with major economic, social, servation efforts. It weaves together efforts tar- and political consequences. geted at individual species potentially at risk (the In 1993, President Clinton convened a large “survey and manage” component, described in group (ultimately over 100 contributors) of fed- the next section), with broad-scale strategies that eral agency and university scientists, the Forest sweep in groups of species and ecological pro- Ecosystem Management Assessment Team cesses (reserves, special guidelines for actively (FEMAT), and gave them the task of producing a managed areas). It specifically provides for active plan that would both provide for viability of old- management for both commodity production and growth ecosystems and their associated species restoration. And finally, it is a long-term plan (100- and provide for a sustainable level of economic year timeframe in which target conditions become benefits from forest products (Johnson et al. fully implemented) with checkpoints along the way 1999). Led by Jack Ward Thomas (later Chief of to determine if adjustments are needed in order to the USDA Forest Service), the FEMAT group de- meet the goals. veloped several options for consideration by A primary feature of ecosystem management is that it is science based. To achieve sustainability, 1Chief, Branch of Physical Sciences, OR/WA State Office, plans must be built with an understanding of the U.S. Department of the Interior (USDI) Bureau of Land Management, P.O. Box 2965, Portland, OR 97208-2965; Tel: 503-808-6036; e-mail: [email protected]

87 as partners in formulating policies and major implementation steps. The participating agencies are: o USDA Forest Service—Pacific Northwest and Pacific Southwest Regions, and Pa- cific Northwest and Pacific Southwest Research Stations o USDI Bureau of Land Management o USDI National Park Service o USDI Fish and Wildlife Service o USDI Geological Survey o U.S. Environmental Protection Agency o USDC National Marine Fisheries Service o U.S. Army Corps of Engineers o USDI Bureau of Indian Affairs o USDA Natural Resources Conservation Service

Northwest Forest Plan Land Alloca- tions, Standards, and Guidelines The heart of the NFP is in the system of reserves, Figure 1–Federal land allocations within the range of the actively managed lands (matrix), and operational northern spotted owl. guidelines it lays out. The key features are: • Matrix lands–Making up about 16 percent of interactions and possibilities among the compo- the total NFP acreage, matrix lands are where nents of forest ecosystems and their associated most active timber harvest is expected to human communities. The NFP not only incorpo- occur. Within managed areas, special stan- rated science, it was developed largely by scien- dards and guidelines protect elements of late- tists. This mode of plan development helped gain successional and old-growth forests to provide the scientific credibility needed to reverse the connectivity for old-forest-dependent organ- court injunction against timber harvest, and as isms across the landscape, and provide refu- the plan proceeded to be implemented, resulted gia for them during the time that managed in significant growth in scientist-manager collabo- areas are in earlier successional stages. rations. These standards include leaving 15 percent The NFP includes three major facets: of mature green trees in managed areas and specific requirements for protection of snags • A strategy (composed of land allocations and and large down wood. standards and guidelines) for managing 24 million acres of federal forest land. • Late successional reserves (LSRs)–About 7.5 million acres (31 percent of the total) of federal • The Northwest Economic Adjustment Initia- forest lands within the NFP area are in large tive, a $1.2 billion program (spread over reserves within which late-successional and 5 years) that provides support to timber-de- old-growth forest structures and processes pendent communities, tribes, and businesses. are maintained and protected. The purpose of • An interagency model of policymaking, LSRs is to provide large blocks of habitat for wherein the participating federal agencies act species associated with old forests. The LSRs

88 are scattered throughout the NFP area such o Watershed analysis–A “systematic proce- that organisms can disperse among them dure to characterize the aquatic, riparian, across the entire area (standards for interven- and terrestrial features within a water- ing matrix areas help provide connectivity shed,” used to “refine riparian reserve across managed lands). Significant acreage in boundaries, prescribe land management LSRs is currently in younger stands (former activities, including watershed restoration, plantations); up to age 80, silvicultural treat- and develop monitoring programs” (USDA ments may be used to foster old-growth stand USDI 1994). structures and habitats. Once a stand reaches o Restoration–Special emphasis has been age 80, no further management typically oc- placed on proactively identifying and re- curs.2 All activities (for example, recreation) in storing degraded aquatic and riparian LSRs must be neutral or beneficial with regard sites (e.g., road crossings with inad- to late-successional and old-growth qualities. equate culverts or stream reaches with When LSRs are combined with other types of insufficient large instream wood. reserves (congressionally and administratively designated areas, such as wilderness areas • Monitoring and adaptive management–A and national parks), the resulting acreage is prime goal of the NFP is to learn from experi- approximately 78 percent of the total NFP ence, and adjust the plan when appropriate. area. This cycle of doing, learning, and adjusting is termed “adaptive management.” Much of • Aquatic conservation strategy (ACS)–The adaptive management application occurs in- NFP provides for the viability of native fish formally and opportunistically. Two “institution- stocks and overall riparian values through the alized” facets of adaptive management in the ACS. There are four components to this ele- NFP are: ment of the NFP: o Monitoring–The NFP calls for monitoring o Riparian reserves–Similar in policy and at multiple scales for an array of factors. intent to LSRs, riparian reserves consist The most formalized monitoring occurs of buffers of varying widths (generally 100 at the NFP area-wide scale, in ongoing, to 300 feet) adjacent to all streams and systematic assessments of northern spot- wetlands in the matrix. About 11 percent ted owl populations, marbled murrelets of the total NFP acreage is in riparian (Brachyramphus marmoratus), aquatic reserves. Besides providing benefits to and riparian conditions, late-successional both riparian and aquatic habitats and and old-growth habitats, and adherence processes, riparian reserves also provide to NFP guidelines in implementation of late-successional habitat connectivity land management activities. Monitoring across the matrix between LSRs. for biological diversity, socioeconomic o Key watersheds–Selected watersheds factors, and tribal relations is being within the NFP area are designated as planned as well. “key watersheds,” with special require- o Adaptive management areas (AMAs)– ments for the protection of at-risk fish Ten areas were designated AMAs in the stocks and water quality. NFP, constituting about 6 percent of the total acreage. The AMAs were geographi- cally located to capitalize on opportunities 2 There may be exceptions to this rule on a case-by-case basis, if needed, to reduce risk to old-growth characteristics. to develop partnerships (“collaborative For example, without thinning or other mechanical treatments stewardship”) with local communities, and for fuels reduction and enhancement of stand structure, some were intended to be places where experi- stands on the eastern flank of the Cascades with a high- mentation outside the matrix standards frequency, low-intensity natural fire regime may be at risk of and guidelines would be allowed (to test stagnation, insect and disease outbreaks, or destruction by wildfire. Conceivably, treatments after age 80 could be the adequacy of the standards and guide- allowed in these stands; however, such exceptions have not lines) and where new management ap- been widely sought. proaches could be developed and tested.

89 • Survey and manage (S&M) mitigation–When • “Institutionalized” collaboration (standing com- the consequences of implementing all the mittees and chartered groups with specific components listed thus far were evaluated, it mandates) among agencies and the public. was determined that there were still approxi- • Provision for direct participation by nonfederal mately 400 species of old-growth-associated governments including tribes. plants, animals, and fungi whose persistence was still uncertain. For many of the species, • Consistent data collection and mapping lack of information on distribution and abun- across ownerships. dance was a major factor in their being in- • Specific inclusion of research and monitoring cluded. This is not surprising, as the group for decisionmaking and adaptive manage- contains many rather obscure nonvascular ment. plants, fungi, invertebrates, and arthropods, in addition to the better known flora and fauna. The NFP sets out a special set of rules man- Challenges dating surveys for S&M species prior to imple- The NFP was the first plan of its kind to attempt menting management activities, and various such a comprehensive, ecologically based effort degrees of protection, depending on the with diverse agencies on a large land base. As needs. Provision was made for S&M species plan implementation has proceeded, questions to be evaluated periodically as new informa- and issues have emerged that present significant tion is obtained, with the goal of eventually challenges to the interagency partners. either determining that species should not have been included in the first place, or pro- At the head of the list of compelling questions is viding an adequate plan of protection. that of meshing ecosystem-based and single- species approaches to conserving late-succes- Has the NFP Succeeded? sional and old-growth biological diversity. Except for the survey and manage component, the NFP For the most part, it is too early to tell whether the assumes that providing for adequate habitat, con- NFP has succeeded in meeting its goals. Cer- nectivity, and ecological processes across the tainly, the harvest levels envisioned have not been landscape will protect biological diversity. The achieved, largely because of recent lawsuits, lack FEMAT group viewed this assumption as a hy- of political support for harvest of old-growth for- pothesis that would be tested through monitoring, ests on matrix lands, and other factors. Results of with subsequent plan adjustments if necessary.3 monitoring efforts for species and aquatic/riparian However, an experimental approach to land man- systems are still several years away. Agencies are agement presumes that some degree of risk is still struggling with adaptive management, both as acceptable, a point around which there clearly is a concept, and in the context of the AMAs. On the lack of consensus. On the other hand, attempting other hand, the injunction against timber harvest to manage a large set of cryptic species at a very was lifted, the agencies involved are working to- low level of risk is far beyond the resources of any gether proactively and collaboratively, and com- of the agencies; in fact, it is probably an unattain- munities are increasingly involved in management able goal to begin with. Strategies that minimize decisions. risk to individual species by focusing on providing habitat may be the only practical solution to In 1998, the Interagency Steering Committee biodiversity conservation concerns. How to create (ISC) commissioned a review of the NFP (Pipkin such strategies that are scientifically sound and 1998). The report identified several elements of legally defensible remains a productive avenue for the NFP as “useful and proven models” worthy of further exploration. inclusion in future assessments:

• A common regional “vision” for forests and the 3 To date, a framework for testing this hypothesis has not timber economy that all participants can work been developed. toward.

90 The need to work at multiple scales was explic- differing points of view among the agencies re- itly stated as a goal for the NFP (FEMAT 1993). garding how much and what type of experimenta- However, technological tools for sliding up and tion is appropriate within AMAs. down the continuum of scales of space and time Another example is fire-prone late-successional have been slow in coming. Such tools would have reserves on the east flank of the Cascade crest, to provide the ability to both scale up (for ex- where managers have hesitated, again because ample, assemble the cumulative effects of finer of the risk of appeals and lawsuits, to seek ex- scale phenomena) and scale down (disaggregate emptions from LSR guidelines in order to imple- larger scale phenomena into smaller temporally or ment prescriptions that may reduce the risk of spatially explicit fragments). Currently, the NFP is stand damage from fire and insects. Both of these being implemented at several scales (stand, wa- examples illustrate that application of a static set tershed, landscape, province, region), but it is of rules to an inherently dynamic situation may often impossible to translate data, findings, or actually increase risk in the long run (through lost effects to larger or smaller scales because of opportunities for learning and improved manage- inconsistencies in methodologies, assumptions, ment in the case of AMAs, and through increased and so on. threat to old-growth values in LSRs). There is an apparent tension between what I will Finally, because the NFP applies exclusively to term static versus dynamic approaches to de- federal lands, the role of nonfederal lands, es- fining target conditions for a variety of ecosystem pecially private lands, in the overall ecological, elements within the NFP. The static approach is social, economic, and political context of the re- manifested in the creation of rules, standards, and gion cannot be fully assessed. It is especially diffi- guidelines that set a threshold or narrow range of cult to understand how private lands contribute to conditions that must be achieved and maintained. the overall whole when large elements are un- This approach often has the terms “regulatory” or regulated and landowner behavior is difficult to “statutory” applied to it. In contrast, the dynamic predict. approach is evidenced in the notion of adaptive management (experimental management with a The federal agencies involved in the NFP have specific goal of learning and potentially adjusting set in motion numerous efforts to resolve these practices), and in the idea that because ecosys- and other issues. The NFP has not proved to be tems are themselves dynamic, a wider range of the ultimate solution to conflicts over the use and conditions will naturally (and should be allowed to) management of federal forest lands in the Pacific exist through time. Both approaches have a place Northwest. Although the agencies have achieved in the NFP, but the extent to which each applies to major successes, small and large, the challenges the various components of the NFP has yet to be listed above, combined with the continuing need resolved. The main difference between the ap- to build public trust, have constituted real barriers. proaches is the perception of risk associated with The NFP’s strength lies in the strong science un- variation. derpinnings on which it is founded, in the scope of the ecological, social, and political issues it at- One of the most problematic examples of the tempted to address, and in its built-in flexibility to tension between the two approaches has been in adapt and improve. It remains to be seen whether adaptive management areas. Although AMAs the agencies and public together can capitalize on were delineated with the express purpose of these strengths to overcome the difficulties cited evaluating the assumptions of the NFP through here and realize the NFP’s promise of protection testing a wide range of practices (some of which for old-growth ecosystems while contributing to may fall outside standards and guidelines), land regional economic health. managers have been understandably reluctant to venture outside the prescriptions of the NFP because of the very real threat of appeals, lawsuits, Metric Equivalents and other forms of resistance from outside enti- 1 foot = 0.304 meters ties that may lack confidence in managers’ abili- ties to meet broad NFP goals. There are also 1 acre = 0.405 hectares

91 Literature Cited Pipkin, J. 1998. The Northwest Forest Plan revis- ited. Washington, DC: U.S. Department of the Forest Ecosystem Management Assessment Interior, Office of Policy Analysis. 117 p. Team [FEMAT]. 1993. Forest ecosystem management: an ecological, economic, and social U.S. Department of Agriculture, Forest Serv- assessment. Portland, OR: U.S. Department ice; U.S. Department of the Interior, Bureau of Agriculture; U.S. Department of the Interior of Land Management. 1994. Record of deci- [et al.]. [Irregular pagination]. sion for amendments to Forest Service and Bureau of Land Management planning docu- Johnson, K.N.; Swanson, F.J.; Herring, M.; ments within the range of the northern spotted Greene, S. 1999. FEMAT assessment–case owl. [Place of publication unknown.] 74 p. [Plus study. In: Johnson, K.N.; Swanson, F.J.; Her- attachment A: standards and guidelines]. ring, M.; Greene, S., eds. Bioregional assessments. Washington, DC: Island Press: 87-116.

92 Recent Changes (1930s–1990s) in Spatial Patterns of Interior Northwest Forests, USA1

P.F. Hessburg,2 B.G. Smith,3 R.B. Salter,2 R.D. Ottmar,4 and E. Alvarado5

Abstract We characterized recent historical and current vegetation composition and structure of a representative sample of subwatersheds on all ownerships within the interior Columbia River basin and portions of the Klamath and Great Basins. For each selected subwatershed, we constructed historical and current vegetation maps from 1932 to 1966 and 1981 to 1993 aerial photos, respectively. Using the raw vegetation attributes, we classified and attributed cover types, structural classes, and potential vegetation types to individual patches within subwatersheds. We characterized change in vegetation spatial patterns using a suite of class and landscape metrics, and a spatial pattern analysis program. We then translated change in vegetation patterns to change in patterns of vulnerability to wildfires, smoke production, and 21 major forest pathogen and insect disturbances. Results of change analyses were reported for province-scale ecological reporting units (ERUs). Here, we highlight significant findings and discuss management implications. Twentieth-century management activities significantly altered spatial patterns of physiognomies, cover types and structural conditions, and vulnerabilities to fire, insect, and pathogen disturbances. Forest land cover expanded in several ERUs, and woodland area expanded in most. Of all physiognomic conditions, shrubland area declined most due to cropland expansion, conversion to semi- and non-native herblands, and expansion of forests and woodlands. Shifts from early to late seral conifer species were evident in forests of most ERUs; patch sizes of forest cover types are now smaller, and current land cover is more fragmented. Landscape area in old multistory, old single story, and stand initiation forest structures declined with compensating increases in area and connectivity of dense, multilayered, intermediate forest structures. Patches with medium and large trees, regardless of their structural affiliation are currently less abundant on the landscape. Finally, basin forests are now dominated by shade-tolerant conifers, and exhibit elevated fuel loads and severe fire behavior attributes indicating expanded future roles of certain defoliators, bark beetles, root diseases, and stand replacement fires. Although well intentioned, 20th-century management practices did not account for landscape-scale patterns of living and dead vegetation that enable forest ecosystems to maintain their structure and organization through time, or for the disturbances that create and maintain them. Improved understanding of change in vegetation spatial patterns, causative factors, and links with disturbance processes will assist managers and policymakers in making informed decisions about how to address important ecosystem health issues.

1 Hessburg, P.F.; Smith, B.G.; Salter, R.B.; Ottmar, R.D.; Alvarado, E. 2000. Recent changes (1930s–1990s) in spatial patterns of interior Northwest forests, USA. Forest Ecology and Management. 136: 53-83. Reprinted with permission from Elsevier. 2 Research ecologist, (Hessburg) and geographer (Salter), U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, 1133 N Western Ave., Wenatchee, WA 98801; Tel: 509-662-4315; e-mail: [email protected], [email protected]. 3 Quantitative ecologist, U.S. Department of Agriculture, Forest Service, 1230 NE 3rd St., Suite 262, Bend, OR 97701; Tel: 541-383-4023; e-mail: [email protected]. 4 Research forester, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, 400 N 34th, Suite 201, Seattle, WA 98103; Tel: 206-732-7828; e-mail: [email protected].

5Research forester, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry Sciences Laboratory, 400 N 34th, Suite 201, Seattle, WA 98103; Tel: 206-732-7842; e-mail: [email protected].

93 Landscape Methodology

94 Analysis and Communication of Monitoring Data With Knowledge-Based Systems

Keith M. Reynolds1 and Gordon H. Reeves2

Introduction Institute, Redlands, CA) geographic information system (GIS) to provide knowledge-based rea- Several broad-scale ecoregional assessments soning for landscape-level ecological analyses have been conducted in the United States in re- (fig. 1). Major components of the EMDS system cent years (Everett et al. 1994, FEMAT 1993, include the NetWeaver knowledge-base system, SNEP 1997, USDA FS 1996), and several more the EMDS ArcView application extension, and the are in progress. Each assessment has used a assessment system (Reynolds 1999a, 1999b; now well-standardized approach to define the Reynolds et al. 1996, 1997a, 1997b). analytical problem. A scoping process was used to identify and evaluate critical issues deserving A knowledge base is a formal logical specification consideration. A needs assessment was per- for interpreting information and is therefore a form formed to identify data requirements and analyti- of meta database in the strict sense (Jackson cal methods needed to respond to the issues. A 1990, Waterman 1986). Interpretation of data by wide array of statistical, simulation, and optimiza- a knowledge-base engine (a logic processor) tion procedures has been used to address various provides an assessment of system states and components of the overall assessment problem. processes represented in the knowledge base as To assert that the suite of analyses used in these topics. Use of logical representation for assessing assessments was conducted ad hoc would be a the state of systems frequently is desirable or disservice. However, it is fair to say that, although necessary. Often, the current state of knowledge assessment teams may have carefully coordi- about a problem domain is too imprecise for sta- nated the conduct of analyses with integration in tistical or simulation models or optimization, each mind, there is little evidence in the reports that of which presumes precise knowledge about rel- effectively integrated analysis was achieved. evant mathematical relations. In contrast, knowledge-based reasoning provides solutions for Landscape Analysis With Ecosystem evaluating this more imprecise information, and Management Decision Support some knowledge-based systems can provide useful analyses even in circumstances in which The USDA Forest Service Pacific Northwest Re- data are incomplete. search Station released the first production ver- Knowledge-based solutions are particularly rel- sion of the ecosystem management decision evant to ecosystem management because the support (EMDS) system in February 1997. The topic is conceptually broad and complex, involving EMDS integrates a knowledge-base engine into numerous, often abstract, concepts (e.g., health, the ArcView3 (Environmental Systems Research sustainability, ecosystem resilience, ecosystem stability, etc.) for which assessment depends on 1 Research forester, Pacific Northwest Research Station, numerous interdependent states and processes. Forestry Sciences Laboratory, 3200 SW Jefferson Way, Corvallis, OR 97331; Tel: 541-750-7434; e-mail: Logical constructs are useful in this context be- [email protected]. cause the problem can be evaluated as long as the entities and their logical relations are under- 2 Research fisheries biologist, Pacific Northwest Research Station, Forestry Sciences Laboratory, 3200 SW Jefferson stood in a general way and can be expressed by Way, Corvallis, OR 97331; Tel: 541-750-7314; e-mail: subject matter authorities. [email protected] 3 The use of trade of firm names in this publication is for reader information and does not imply endorsement by the U.S. Department of Agriculture of any product or service.

95 Figure 1—Architecture of EMDS system. The acronym, GUI, indicates a graphic user interface through which a user interacts with components of the EMDS application. The assessment system uses a custom GUI to manage communication between the NetWeaver logic engine and code internal to the EMDS ArcView extension. ESRI = Environ- mental Systems Research Institute.

Nestucca Basin Example: Evaluation quired to implement the Aquatic Conservation of Salmon Habitat Suitability Strategy of the Northwest Forest Plan, the authors designed a prototype knowledge base suitable for The Nestucca basin is a 5th-code hydrologic unit application to true 6th-code watersheds in the Or- located in the northern portion of the Oregon egon Coast Range province (fig. 2). Starting with Coast Range province (fig. 2) and contains 45 basic requirements identified in the NMFS matrix, true and composite 6th-code watersheds. In 1998, Reeves summarized logical requirements for the a watershed assessment team from the Siuslaw knowledge base. The authors then worked to- National Forest rated each true 6th-code water- gether to implement the knowledge-base design shed in the basin with respect to its suitability for in NetWeaver, and then reviewed the initial design providing salmon habitat, following the rating sys- with the assessment team to make refinements, tem of the decision matrix of the National Marine corrections, and additions to the initial version. Fisheries Service (NMFS). In subsequent discussions with the authors, assessment team mem- The most basic outputs from an EMDS analysis bers expressed a number of reservations about are maps of evaluated indices for topics included the approach that had been used for rating water- in an analysis. In our example, the primary map of shed condition. interest (fig. 3) displays the evaluated index for salmon habitat suitability for each 6th-code water- To improve on the basic approach to watershed shed in the basin. Composite watersheds, for evaluation for salmon habitat suitability as re- which the knowledge base was not designed, are

96 Figure 2—The Nestucca basin is in the central Oregon Coast Range.

Figure 3— Salmon habitat suitability in 6th-code watersheds in the Nestucca basin. Patterned features are composite watersheds for which the knowledge base is not appropriate. The scale expresses strength of support for the proposition that the biophysical condition of a watershed provides suitable salmon habitat. The extremes of the scale, -1 and 1, indicate no support and full support, respectively. Intermediate values express some degree of support.

97 displayed with a lined pattern. All information that ment times. Similarly, more detailed summaries of is logically antecedent to topics selected for analy- conditions in an assessment area (figs. 8 and 9) sis also must be evaluated by the logic engine provide useful background information for design and so is displayable either in maps, graphs, or in of restoration programs as well as a basis for tabular output generated by the logic engine. more detailed statistical inferences about the efficacy of these programs. Maps characterizing landscape condition are powerful communication tools, but it is at least as important to be able to explain the basis for ob- Identifying Restoration Objectives and served results in a clear, intuitive manner. One of Criteria the most significant features of knowledge-base Problem specification for ecological assessment systems is their ability to display a logic trace that may well deserve to be classified as a wicked explains the derivation of conclusions. In this ex- problem (Allen and Gould 1986). Evaluation moni- ample, the top portion of the browser window toring must consider potentially numerous states displays an expandable outline of knowledge- and processes of biophysical, social, and eco- base structure, and the bottom portion displays nomic components of an ecosystem. Many enti- the logic structure of the topic currently selected in ties may have both deep and broad networks of the outline (fig. 4). logical dependencies as well as complex intercon- The NetWeaver logic engine reasons with incom- nections. Although constructing such complex plete information, when needed, by using fuzzy representations in NetWeaver is not trivial, it is at math (Zadeh 1965, 1968), and the engine evalu- least rendered feasible by the precision and com- ates the influence of any missing information with pactness of fuzzy logic relations, and by the respect to its contribution to completeness of an graphic, object-based representation of logic net- analysis. Input data for the Netsucca basin analy- works in the system interface. No subject-matter sis contained a few missing observations on data authority, nor for that matter any group of authori- fields (fig. 5). However, the engine implements ties, is capable of holding a comprehensive cogni- evidence-based reasoning, producing partial tive map of such a complex problem domain in evaluations of topics with incomplete information their consciousness. On the other hand, the (figs. 3 and 4). The knowledge base is, in effect, a graphic, object-based form of knowledge repre- mental map of the problem. Its structure contains sentation in NetWeaver is highly conducive to the information that can be used to compute the influ- incremental evolution of knowledge-base design ence of missing information (fig. 6). The computa- from simple to complex forms. tion of influence is based on the level at which Two essential aspects of the knowledge-based information enters the logic structure, how many approach to evaluation monitoring that we have references there are to it, and how frequently the described and illustrated are its goal orientation associated database field contains missing obser- and use of formal logic in problem specification. vations. The assessment system in EMDS also The implications for realizing truly integrated includes tools to optionally synthesize information analysis are significant. Knowledge-base design about data influence and the logistics of acquiring begins with identification of the questions (formu- data to prioritize data needs (Reynolds 1999a). lated as propositions) that the analysis ultimately Finally, the logic engine generates tabular output is designed to address. Specification proceeds by from which an analyst can summarize additional identifying lines of reasoning that decompose the information such as the frequency distribution of original propositions into progressively more con- index values (fig. 7) or the frequency with which crete ones until links to data are identified. The data substantially contribute to a particular conclu- result, effectively, is a mental map of the problem, sion (figs. 8 and 9). Frequency distributions of including not only identification of all topics perti- indices provide a simple synoptic view of condi- nent to the problem, but their logical interdepen- tions over an assessment area at a point in time dencies. The final specification thus not only (fig. 7), and useful statistical inferences about the provides logical links between data requirements efficacy of restoration programs, for example, and the original motivating questions, but provides could be drawn from comparisons of such fre- a specification for how the data are to be inter- quency distributions over consecutive assess- preted, taking into account interrelations among pieces of information.

98 Figure 4—The knowledge base with the EMDS hotlink browser tool displays the state of the knowledge base for features (e.g., watersheds) selected in the EMDS Assessment View.

Figure 5—Frequency of missing data fields in the input data table for the Nestucca basin analysis.

99 Figure 6—Influence of missing information in the Nestucca basin analysis calculated by the NetWeaver logic engine.

Figure 7—Frequency distribution of the salmon habitat suitability index for the Nestucca basin analysis. Truth value interval is a measure of the degree of support that data provide for a proposition. This index also can be interpreted as a measure of suitability.

100 Figure 8—Frequency with which data in high-gradient drainages contributed to the conclusion of a compromised or degraded biophysical condition. Drainages with an average reach gradient greater than 4 percent were classified as high gradient; in-channel conditions were not available.

Figure 9—Frequency with which data in low-gradient drainages contributed to the conclusion of a compromised or degraded biophysical condition. Drainages with an average reach gradient less than or equal to 4 percent were classified as low gradient; data on in-channel conditions were available.

101 Performing Evaluations With trusted. Powerful and sophisticated linear pro- Incomplete Information gramming systems such as FORPLAN largely failed on this account. It is probably fair to say that almost all significant monitoring and assessment programs start out Our experience to date with EMDS, and with its with significant data gaps. The NetWeaver logic knowledge-base browser interface in particular, engine is particularly well suited for use in applica- has been that the presentation of results from tions such as evaluation monitoring in particular logic models is easy to grasp at an intuitive level and ecosystem analysis in general because it is so the results are accessible to broad audiences able to reason with incomplete information and with only limited explanation. An analyst can easily still provide meaningful results, and because it run the browser interactively in front of an audi- can evaluate the influence of missing information. ence, navigating the structure of the evaluated knowledge base, and explaining the derivation of We noted earlier that the number of topics in any the logic as he or she goes. The earlier logic dia- reasonably realistic problem and the number of gram (fig. 4) provides a good example. Composite interdependencies among them can easily over- watersheds (WS type = 1) are screened out of the whelm the capacity of the human brain to reason analysis with a logic switch and remain in an un- effectively about such questions as, What is the determined logic state. For true watersheds, the most useful information that I could acquire at this right-hand path under the WS type logic switch is point in time to improve the completeness of my followed. Biophysical condition of a true water- analysis? It is important to appreciate that the shed evaluates as suitable to the degree that influence of information is highly dynamic: the there is suitable riparian vegetation and a suit- influence of what is missing depends both on the able upland environment. A second logic switch observations that are currently available, and on checks whether the data for the watershed in- the values in those observations. Consequently, clude an evaluation of in-channel conditions. use of iterative analysis of data influence in EMDS If a low-gradient stream survey was performed following successive, incremental additions of (lowGradSurvey > 0), then the evaluation includes new data could conceivably reduce resources the upland, in-channel, and riparianVeg topics; devoted to data collection on the order of 50 per- otherwise the evaluation of biophysical condition cent in typical applications. only depends on the upland and riparianVeg top- In our very small example, the distributions of ics. Each of the three topics, upland, in-channel, missing data (fig. 5) and data influence (fig. 6) and riparianVeg, has its own logic specification mirror each other. More generally, and particularly (not shown). for larger and more complex logic models, the two distributions may be very different owing to inter- Evaluation Within Larger Contexts dependencies among data elements. Even in this Almost all landscape analyses are performed simple case, however, it is clear that missing data within the context of some broader scale. Our on water temperature are the largest source of knowledge base for evaluation of salmon habitat uncertainty in the current analysis (fig. 6). It also is suitability is a prototype constructed for the spe- quite possible that if some of the missing water cific context of the Oregon Coast Range. Conse- temperature observations were acquired, pool quently, it is unlikely that the current version is quality, fines, and farm use might no longer have sufficiently general for application throughout the nonzero influence. planning region of the Northwest Forest Plan, which encompasses the range of the northern Explanation of Monitoring Results spotted owl (Strix occidentalis caurina). The ability to explain monitoring results to col- Fortunately, it is not difficult to generalize proto- leagues, line officers, interested publics, etc. is at types such as ours for broader applicability. In least as important as the ability to perform an some cases, for example, it may be sufficient to analysis in the first place. If the logic of underlying replace hard-wired arguments with a calculated models cannot be presented in clear, unambigu- one that contains a more general fuzzy member- ous terms to such audiences, results will not be ship function whose parameters vary with additional context information. For more dramatic

102 Figure 10—An example of knowledge-based integration across spatial scales. variations in logic structure, it may be necessary cally implemented in a spreadsheet or database to add new logic switches that similarly select application) to synthesize information for input to alternative logic pathways based on additional the next coarser scale. A second knowledge base context information. In a technical sense, knowl- processes the synthesized information to provide edge-base generalization is a relatively minor an assessment of landscape-level attributes at issue. On the other hand, it leads naturally to dis- the top of the figure. Finally, knowledge-base out- cussion of the more interesting question of puts at the broader landscape scale may feed multiscale implementations. back to the fine scale as context information that influences evaluations at the fine scale. This Extending Ecosystem Management simple conceptual model provides the basis for a Decision-Support Applications to formal logical specification of analyses that are Multiple Spatial Scales consistent across scales. Hierarchies, or even networks, of knowledge-based analyses as sug- It is relatively easy in principle to extend integrated gested would be highly consonant with ecosystem analysis via knowledge-based reasoning over theories concerning the hierarchical organization multiple spatial scales (fig. 10). Data from fine- ecosystems. scale landscape features such as 6th-code watersheds are first processed by a knowledge base English Equivalents designed for that scale. Knowledge-base output, shown as evaluated states in the middle of the 1 kilometer = 0.62 miles figure, then go through an intermediate filter (typi-

103 Literature Cited Reynolds, K.M.; Saunders, M.; Miller, B. [et al.]. 1997a. An application framework for deci- Allen, G.M.; Gould, E.M., Jr. 1986. Complexity, sion support in environmental assessment. In: wickedness, and public forests. Journal of GIS 97 Proceedings of the eleventh annual Forestry. 84: 20-23. symposium on geographic information sys- Everett, R.; Hessburg, P.; Jensen, M.; tems. Washington, DC: GIS World: 333-337. Bormann, B. 1994. Eastside forest ecosystem Reynolds, K.M.; Saunders, M.; Miller, B. [et health assessment. Volume I: executive sum- al.]. 1997b. Knowledge-based decision sup- mary. Gen. Tech. Rep. PNW-GTR-317. Port- port in environmental assessment. In: Pro- land, OR: U.S. Department of Agriculture, ceedings 1997 ACSM/ASPRS annual con- Forest Service, Pacific Northwest Research vention and exposition. Bethesda, MD: Ameri- Station. 61 p. can Society for Photogrammetry and Remote Forest Ecosystem Management Assessment Sensing: 344-352. Vol. 4 Team [FEMAT]. 1993. Forest ecosystem man- Sierra Nevada Ecosystem Project [SNEP]. agement: an ecological, economic, and social 1997. Final report to Congress, addendum. assessment. Portland, OR: U.S. Department Davis, CA: University of California, Centers for of Agriculture, Forest Service; U.S. Department Water and Wildland Resources. 328 p. of the Interior, Bureau of Land Management; [et al.]. [Irregular pagination]. U.S. Department of Agriculture, Forest Ser- vice. 1996. The Southern Appalachian assess- Jackson, P. 1990. Introduction to expert systems. ment summary report. Tech. Pap. R8-TP-25. Reading, MA: Addison-Wesley Publishers. Atlanta, GA: Southern Region [Region 8]. 526 p. 118 p. Reynolds, K.; Cunningham, P.; Bednar, L. Waterman, D.A. 1986. A guide to expert systems. [et al.]. 1996. A design framework for a knowl- Reading, MA: Addison-Wesley Publishers. edge-based information management system 419 p. for watershed analysis in the Pacific Northwest U.S. AI Applications. 10: 9-22. Zadeh, L.A. 1965. Fuzzy sets. Information and Control. 8: 338-353. Reynolds, K.M. 1999a. EMDS users guide (version 2.0): knowledge-based decision support Zadeh, L.A. 1968. Probability measures of fuzzy for ecological assessment. Gen. Tech. Rep. events. Journal of Mathematical Analysis and PNW-GTR-470. Portland, OR: U.S. Depart- Applications. 23: 421-427. ment of Agriculture, Forest Service, Pacific Northwest Research Station. 63 p. Reynolds, K.M. 1999b. NetWeaver for EMDS user guide (version 1.1): a knowledge base development system. Gen. Tech. Rep. PNW- GTR-471. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 75 p.

104 Seeing the Forest and the Trees: Visualizing Stand and Landscape Conditions

Robert J. McGaughey1

Introduction spatial and temporal interactions that occur across landscapes. Traditional work methods The appearance of forested landscapes and indi- involving fieldwork, maps, and aerial photographs vidual stands after forest management activities provide enough information to assess individual is critical to public acceptance of these activities. treatments but may not provide adequate informa- Even with thorough planning, detailed site-specific tion to fully evaluate the cumulative impact of analysis, and careful monitoring, many manage- treatments or the impact of treatments implement activities will not be truly successful if the mented over time. Computer-based landscape public views the resulting landscape as an eye- simulations are a recognized tool for assessing sore. Unfortunately, many people judge the suc- the potential visual impact of land use decisions cess or failure of management activities based on and management activities. Visualization tools the visual impact of the activity to an otherwise that include a stand projection component or pro- “natural” landscape. vide linkages to such models can simulate and Forestry professionals have long used visualiza- depict stand and landscape changes over time. tion techniques to address a variety of forest Such presentations help to communicate stand management problems. Prior to the advent of and landscape conditions and how these condi- computerized methods, they used “artists’ renditions change as a result of management activities, tions” to communicate the effects of land man- natural disturbances, and growth over time. Dur- agement activities. Perspective sketches and ing the treatment design process, visualizations scale models continue to help communicate the depicting treatments within their landscape con- spatial arrangement and extent of management text provide important feedback. Such feedback activities to the lay public. However, current man- can help resource specialists develop and imple- agement practices use more detailed silvicultural ment better landscape management plans. Fur- prescriptions involving small treatment areas scat- thermore, visualizations help communicate tered over larger landscapes and the removal or management activities to other resource special- modification of specific stand components. In ists and public stakeholders. addition, treatment regimes are designed to achieve a desired condition over long timespans Overview of Visualization Techniques by using a series of treatments. Given the broad range of treatments considered and changes to Computerized visualization methods range from stand conditions over time, it is difficult to produce simple diagrams to complete virtual realities. Four traditional artists’ renditions that reflect the ex- methods, described in detail and compared in pected condition accurately and with a minimum McGaughey (1998), are commonly used to pro- of artist-contributed bias. In addition, the time duce visual representations of forest operations: required to produce artists’ renditions limits their • Geometric modeling use when depicting several alternative treatment strategies. • Video imaging Foresters charged with selecting stands for treat- • Geometric video imaging ment and designing silvicultural prescriptions of- • Image draping ten find it difficult to comprehend the complex Geometric modeling methods build geometric models of individual components (ground surface, 1 Research forester, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, University of trees, other plants, and structures) and then as- Washington, P.O. Box 352100, Seattle, WA 98195-2100; Tel: semble the component models to create a scene (206) 543-4713; e-mail: [email protected], that represents a forest stand or landscape. [email protected]

105 Video imaging uses computer programs to modify attributes as a function of stand age, to present scanned full-color video or photographic images information describing forest stands and stand to represent changes to stand and landscape conditions. These communication tools, suitable conditions. To portray a wide range of landscape for conveying simple data relationships to scien- conditions, video-imaging techniques use a library tific and professional audiences, do not always of images that represent different forest condi- provide an easily understood representation of a tions to replace portions of an original image. forest stand or the changes in stand structure that can occur after disturbances or natural processes. A hybrid approach, called geometric video imag- The Stand Visualization System (SVS) was de- ing by this author, combines geometric modeling signed to help foresters understand and commu- and video-imaging techniques to produce realistic nicate stand conditions and changes in stand images that accurately represent data describing conditions resulting from management activities the effects of forest management activities. Op- or natural processes (McGaughey 1997). erators use geometric modeling to produce images that specify the location, arrangement, and The SVS uses geometric modeling techniques to scale of stand or landscape features. These im- create images depicting stand conditions. As part ages then guide video-imaging manipulations that of the stand construction process, SVS converts modify a photographic image to reflect proposed tables describing the population of standing trees, stand or landscape changes. By combining the down material, and understory shrubs into a list of precise spatial location capabilities of geometric individual stand components, e.g., trees, down modeling and the photo-manipulation capabilities logs, and shrubs. Standing trees and down mate- of video imaging, hybrid methods result in spa- rial are defined by using the number of trees or tially accurate, photo-quality images. However, logs per hectare. Understory shrubs can be de- hybrid techniques, like video imaging, require fined by using the number of plants per hectare extensive libraries of tree and stand images to but are more commonly defined by using a per- represent an appropriate range of species, tree centage of cover. The SVS generates enough sizes, growth forms, landscape positions, and plants to achieve the desired cover. Plant loca- treatment options. tions are generated by using a variety of spatial patterns including patterns that mimic natural and Image draping mathematically “drapes” an image planted stands. The SVS also reads output from over a digital terrain model to create a textured the Forest Vegetation Simulator (Wykoff et al. surface. The draped image is typically a satellite 1982) and ORGANON (Hester et al. 1989). The scene, aerial photograph, orthophoto, or scanned images produced by SVS, although abstract, pro- map sheet. vide a readily understood representation of stand Of the four methods, geometric modeling pro- conditions and help communicate silvicultural vides the most consistent link between data de- treatments and forest management alternatives to scribing stand and landscape conditions and a variety of audiences. The SVS provides the features in a generated image. In general, there is following specific capabilities: a one-to-one relationship between data elements • Display overhead, profile, and perspective and objects in the final image. Geometric model- views of a stand. ing produces images that are generally less realistic than images produced from photographs. • Differentiate between stand components by However, if photographic icons are used to repre- using different plant forms, colors, or other sent individual trees and other objects, geometric types of marking as specified by users. modeling can produce images that contain many • Provide tabular and graphical summaries of of the details normally associated with photo- information represented in a stand image. graphs. • Facilitate the design of silvicultural treatments Visualization of Stand/Plot-Scale by allowing users to select individual stand Areas components and specify treatments. Foresters use stand tables, showing the relative • Display information describing individual stand abundance of plant species and size classes, and components as they are selected by the user. simple graphs, showing single or multiple stand

106 Once an SVS tree list has been created, users such as roads, streams, and points of interest can can simulate silvicultural treatments by selecting be represented as points, lines, or polygons on individual trees or groups of trees and specifying the ground surface or as solid walls or point mark- a treatment action such as removal or pruning. ers sitting on the ground surface. To represent As an alternative, treatment rules such as “re- ground surface texture, EnVision can fill polygon move 75 percent of the stand basal area from features with textures derived from aerial photo- below” can be applied to the entire stand. Figure 1 graphs. This “synthetic aerial photo” capability can shows a 1-hectare stand before and after a simple be used to provide ground texture representing thinning treatment that removes all of the alder vegetation under a canopy of larger trees that are (Alnus Hill.) trees and a portion of the Douglas- drawn as geometric models. fir (Pseudotsuga menziesii (Mirb.) Franco) trees. EnVision can use a variety of data to describe Figure 2 shows the same stand after a treat- conditions within individual stands. The simplest ment that removed all of the alder and six small form of stand data consists of the distribution of patches containing 30 percent of the Douglas-fir tree sizes for each species present in a stand. basal area. More complex data include individual tree records Visualization of Landscape-Scale from inventory plots that include species, diameter Areas at breast height, total tree height, live crown ratio, crown width, and the number of trees per unit A wide array of techniques is available for pre- area represented by the record. Stand data are senting landscape-scale information ranging in merged with data describing the growth form and complexity from simple tables and graphs to general crown shape for each tree species to color-coded maps to photo-realistic computer construct a geometric model for each tree in a visualizations. Computer visualization techniques stand. are especially useful given the complexity of stand treatments, the spatial distribution of treatment Multiscale Visualization units, and the desire to view landscape conditions The ability to visualize individual stands by using over time. Landscape visualization provides the SVS has significantly enhanced a forester’s ability viewer with an easily identified image of a land- to understand and communicate stand conditions scape complete with topographic features and and the effects of silvicultural treatments on a stand conditions. A new visualization system be- stand. Through linkages with the Forest Vegeta- ing developed by the author, EnVision, is de- tion Simulator (Teck et al. 1996, Wykoff et al. signed to portray landscape conditions by using 1982), SVS provides visualizations of stand data describing individual stands. EnVision pro- growth and development. However, understanding vides several capabilities not previously available the dynamics within an individual stand does not or not available in the same application. For ex- necessarily lead to a better understanding of the ample, EnVision can fill individual polygons with a dynamics of an entire landscape. The complexity texture derived from an aerial photograph while of landscape management plans, variety of stand rendering individual trees in other polygons. treatments, and time period for which most plans An EnVision scene is based on a gridded digital are developed make it difficult to understand and terrain model2 that defines the ground surface. communicate the plan’s impact on the landscape. EnVision provides a variety of methods to repre- A system that links the design and visualization of sent the ground including gridlines and contour individual stand treatments with the ability to dis- lines, with and without hidden area removal, and play the stands within their landscape context is as a shaded, lighted surface. Surface features needed. EnVision provides this linkage by allowing the use of an SVS tree list to represent each 2 Gridded digital terrain models that correspond to the U.S. stand on a landscape. In operation, EnVision rep- Geologic Survey’s 7.5-minute quadrangle series are available licates the SVS tree list to fill the stand polygon by for most areas in the United States. Other terrain model types, e.g., TIN models, can be used to generate a gridded model by using geograpic information system (GIS) software.

107 Figure 1—SVS visualization of a 1-hectare stand (a) before and (b) after a uniform thinning treatment. The thinning removed all of the alder and 25 percent of the Douglas-fir basal area from below.

Figure 3—Landscape visualization produced by EnVision Figure 2—SVS visualization showing an over- showing a stand polygon filled by using the SVS stand shown head view of the stand shown in figure 1 but in figure 2. Individual trees have been rendered only in the after a treatment that removed all of the alder polygon of interest. Other stand polygons have been filled and six small patches containing 30 percent of with textures that represent the stand structure. the Douglas-fir basal area. using a simple tiling algorithm. This linkage be- Conclusions tween SVS and EnVision makes it possible to This paper has discussed a variety of techniques design spatially explicit treatments within SVS and that can help land managers visualize, under- visualize that treatment within the landscape con- stand, and communicate stand- and landscape- text. Figure 3 shows a landscape where the 1- scale management actions. In particular, this hectare stand in figure 2 has been used to fill a paper has highlighted the SVS and EnVision ap- 21-hectare stand polygon. In this example, SVS plications developed by the author. tile reflection has been used to minimize any pattern effects that might result when filling the stand polygon.

108 Experience with SVS, UTOOLS/UVIEW land- Hester, A.S.; Hann, D.W.; Larsen, D.R. 1989. scape analysis and visualization system (Ager ORGANON: southwest Oregon growth and and McGaughey 1997), and Vantage Point proto- yield model user manual, Version 2.0. type (Bergen et al. 1998) indicates that such visu- Corvallis, OR: Oregon State University, Forest alization tools help resource specialists and land Research Laboratory. 59 p. managers make better decisions. However, such McGaughey, R.J. 1997. Visualizing forest stand tools also can produce photo-realistic images that dynamics using the Stand Vsualization Sys- can mislead viewers. Practitioners must be care- tem. In: Proceedings of the 1997 ACSM/ ful when using such tools that they do not inten- ASPRS annual convention and exposition. tionally or unintentionally misrepresent existing or Bethesda, MD: American Society of Photo- expected conditions (Wilson and McGaughey grammetry and Remote Sensing: 248-257. 2000). Failure to do so will seriously limit the use- Vol. 4. fulness of these visualization techniques for future projects. McGaughey, R.J. 1998. Techniques for visualizing the appearance of forestry operations. English Equivalents Journal of Forestry. 96(6): 9-14. 1 hectare (ha) = 2.47 acres Teck, R.; Moeur, M.; Eav, B. 1996. Forecasting ecosystems with the Forest Vegetation Simula- Literature Cited tor. Journal of Forestry. 94(12): 7-10. Ager, A.A.; McGaughey, R.J. 1997. UTOOLS: Wilson, J.S.; McGaughey, R.J. 2000. Presenting microcomputer software for spatial analysis landscape-scale forest information: What is and landscape visualization. Gen. Tech. Rep. sufficient and what is appropriate? Journal of PNW-GTR-397. Portland, OR: U.S. Depart- Forestry. 98(12): 21-28. ment of Agriculture, Forest Service, Pacific Wykoff, W.R.; Crookston, N.L.; Stage, A.R. Northwest Research Station. 15 p. 1982. User’s guide to the stand prognosis Bergen, S.D.; McGaughey, R.J.; Fridley, J.L. model. Gen. Tech. Rep. INT-133. Ogden, UT: 1998. Data-driven simulation, dimensional U.S. Department of Agriculture, Forest Service, accuracy and realism in a landscape visualiza- Intermountain Forest and Range Experiment tion tool. Landscape and Urban Planning. 40: Station. 112 p. 283-293.

109 A Method to Simulate the Volume and Quality of Wood Produced Under an Ecologically Sustainable Landscape Management Plan

Glenn Christensen,1 R. James Barbour,2 and Stuart Johnston3

Abstract Adopting the Northwest Forest Plan into federal land management brought with it many new challenges to forest planning. One challenge is to determine what effect using new and untested silvicultural treatments to meet ecologically based objectives will have on the production of wood or other products that society values. Determining the quantity and quality of potential outputs as a consequence of ecological restoration activities is important to helping understand the cost to society of implementing such a plan. The difficulty comes from the scale of the problem. No longer are short-term stand-level analyses adequate; evaluation of forest outputs has to be at meaningful spatial and temporal scales. To be successful, new methods will have to be developed. The objective of this paper is to present a method used to evaluate wood removals from a landscape management plan that was developed to meet ecological objectives by mimicking past disturbance cycles. This required extending stand-level techniques for simulating quality and quantity of wood to 980 stands in the Blue River watershed of western Oregon. Silvicultural prescriptions include a range of thinning intensities and frequencies combined with extended rotation ages (100 to 260 years). Study results include the range of products that might be manufactured under different silvicultural alternatives and evaluate the volume as well as the quality of primary and secondary products. Primary product quality estimates include log attributes such as diameter, branch size, juvenile wood, and growth ring count. Secondary product estimates include lumber or veneer volume recovered by grade and likely end use.

Introduction planning on development of a road network to aid in wildfire suppression and dispersion of harvest Interest in the management of forests at the land- activities. Stand treatments were primarily clear- scape level has increased dramatically over the cuts designed to meet several objectives including past several decades. An area where this has rapid stand regeneration and the creation of edge recently received much attention is on lands man- and early seral wildlife habitat. Fueled by change aged by the USDA Forest Service in the Pacific in public perception of clearcuts and a growing Northwest. Prior approaches to forest manage- concern over the effect of forest fragmentation ment on these lands focused landscape-level on the loss of old-growth forest habitat on wildlife, especially the northern spotted owl (Strix 1 Forester, U.S. Department of Agriculture, Forest Service, occidentalis caurina), and water quality during the Pacific Northwest Research Station, Forestry Sciences late 1980s, all Forest Service timber sales were Laboratory, P.O. Box 3890, Portland, OR 97208; Tel: 503-808- temporarily suspended. To end the suspension, 2064; e-mail: [email protected] the Northwest Forest Plan (USDA and USDI 2 Research scientist, U.S. Department of Agriculture, Forest 1994) was developed and is the guiding plan for Service, Pacific Northwest Research Station, Forestry 9.7 million ha of federally managed forest land. Sciences Laboratory, P.O. Box 3890, Portland, OR 97208; Tel: 503-808-2542; e-mail: [email protected] The Northwest Forest Plan (NWFP) is based on a system of static reserves, riparian corridors, and 3 Forester, U.S. Department of Agriculture, Forest Service, Siuslaw National Forest, Mapleton Ranger District, 4480 general forest land called matrix lands. A criticism Highway 101, Building G, Florence, OR 97439; Tel: 541-902- of the NWFP is the use of static reserves on a 6958; e-mail: [email protected]

110 dynamic landscape. Recently, however, consider- Study Area able interest has emerged as a different approach to landscape-level planning, and two studies have The 23 900-ha Blue River watershed is almost been initiated in the central Willamette National entirely managed by the USDA Forest Service Forest of Oregon. Both the Augusta Creek study and includes the H.J. Andrews Experimental For- (Cissel et al. 1998) and the Blue River landscape est, an area with an extensive history of ecosys- plan (Cissel et al. 1999) have focused upon inte- tem research. Of the 298 million ha of forest land grating historical disturbance regimes into land- within the United States, 19 percent (56 million scape- and watershed-level management plans. ha) is managed by the USDA Forest Service These approaches use information on historical (Smith et al. 1994). The Blue River watershed is and current landscape conditions, disturbance located in the Willamette National Forest within history, and social goals to set objectives for fu- the McKenzie River watershed, a tributary of the ture landscape structures that provide desired Willamette River in western Oregon (fig. 1). habitat, watershed, timber supply, and other func- The landscape is steep, highly dissected volcanic tions (Cissel et al. 1999). The intent is not to terrain of the Cascade Range. Annual precipita- mimic historical conditions but rather to use them tion exceeds 2500 mm falling mostly in October as a reference in developing and evaluating man- through April as rain at lower elevations and snow agement alternatives to meet these goals. in higher areas. The landscape ranges from 317 The Blue River landscape plan is an integrated to 1639 m in elevation and is covered largely by landscape management strategy that was devel- coniferous forests dominated by Douglas-fir oped to achieve both ecological and social objec- (Pseudotsuga menziesii (Mirb.) Franco), western tives. As stated in the plan, the “primary goal is to hemlock (Tsuga heterophylla (Raf.) Sarg.), and sustain native habitats, species, and ecological Pacific silver fir (Abies amabilis Dougl. ex Forbes) processes while providing sustained flow of wood (Cissel et al. 1999). fiber for conversion to wood products” (Cissel et al. 1999). The key underlying concept is that by Approach simulating certain aspects of the historical fire The approach for this analysis is to use the Blue regime through forest management activities, we River landscape plan as the basis to determine can sustain the historical range of variability nec- scheduling of stand treatments and silvicultural essary to preserve these native habitats, species, prescriptions. The analysis focuses on wood pro- and ecological processes while still providing a duction as trees are harvested to meet broader predictable flow of timber. ecological objectives. The main tool used for Regardless of which approach is used, method- simulation is the LMS, a computerized system ologies and protocols are needed that describe that integrates landscape-level spatial information, the quantity, quality, and value of the wood pro- stand-level inventory data, and distance-indepen- duced under landscape-level plans. Managers dent individual tree growth models to project also need to know when wood will be removed changes though time across forested landscapes and from what locations on the landscape. In ad- (McCarter 1997). Once current conditions are dition to difficulties from planning at large spatial determined for each stand as accurately as pos- and temporal scales, most modern forest man- sible, silvicultural treatments are scheduled and agement requires the use of complex silvicultural growth projections made. As stands are grown prescriptions in the form of thinnings, group selec- and treatments made, information is collected on tions, and individual tree selections to meet spe- individual tree volume and quality characteristics. cific stand structure objectives. To address these complexities, the Landscape Management Sys- tem (LMS) was developed at the University of Washington (McCarter 1997).

111 Figure 1—Location of Blue River watershed.

This analysis uses methods similar to those de- Modeling Wood Production by Using veloped by Christensen (1997) and Barbour et al. the Landscape Management System (1996) to evaluate the volume and quality of wood from a range of silvicultural regimes. Tree volume The LMS model was chosen for its landscape- and quality information is analyzed with the level analysis capabilities, as well as its integra- TREEVAL (Briggs 1989, Sachet et al. 1989) tion of various component programs to provide model to estimate potential lumber volume and detailed tree volume and wood quality information. grade recovery. The TREEVAL model was devel- Of particular interest is the ability to choose oped by the USDA Forest Service, Pacific North- among several different growth and yield models west Research Station to provide financial in- within the program to allow maximum flexibility in formation and analysis of product recovery to the number of species, kinds of silvicultural treat- support silvicultural decisions in coast Douglas-fir ments, and output produced. Included in LMS are (P. menziesii var. menziesii). Analyses conducted stand- and landscape-level visualization programs by Christensen (1997) and Barbour et al. (1996) that permit, in addition to traditional quantitative are temporally explicit, evaluating wood produced analysis, generalized visual representations of from individual stands through time. They are different management options. spatially implicit in terms of where the wood came A detailed description of the methodology for each from. The current analysis adds the spatial dimen- part of the project can be found in Christensen et sion to this process, providing not just wood pro- al.4 For this paper, a brief overview is provided duction information as a stand develops over giving the basic approach used during each time, but also providing information on the source phase of the study. of wood from across many stands within a watershed. This gives planners the ability to understand 4 Christensen, G.A.; Stuart, J.; Malinick, T.E. [In preparation]. how stand treatments will interact to provide a Simulating the volume and quality of wood produced under predictable flow of wood in terms of volume and an ecologically sustainable landscape management plan in grade recovery from a defined landscape. the Oregon Cascade Range: results from the Blue River Landscape Plan. Res. Pap. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station.

112 Current Stand Conditions For this study, reserve areas were assumed to receive no timber harvest and are not included in As part of the analysis for the Blue River Land- the analysis. Landscape areas were divided into scape Plan, much basic information had already three types based on interpretation of historical been assembled. The most important of this infor- fire-return intervals and intensity. The three land- mation was geographic information system (GIS) scape areas are (Cissel et al. 1999): data files. From the GIS data, we knew the stand type, vegetation association, and how the stand is • Landscape area 1–Frequent fire-return inter- to be managed. Now we needed an inventory for val and moderate severity (40 to 60 percent each stand with specific tree measurements. The mortality). LMS software requires specific stand inventory • Landscape area 2–Moderate fire-return inter- data including stand or polygon number, species, val and moderate-to-high severity (60 to 80 diameter at breast height, height, crown ratio, and percent mortality). expansion factor, on a per-acre basis. The spatial data requirements vary depending upon the • Landscape area 3–Infrequent fire-return inter- growth-and-yield model selected but can include, val and high severity (>80 percent mortality). by stand or polygon number, number of plots, Cissel et al. (1999) developed silvicultural pre- regional location, site index, habitat type, age, scriptions for each landscape area to closely slope, aspect, and elevation. Among the available mimic natural stand development following a fire data sets, it was determined that the inventory (table 1). Treatments were designed to produce plots maintained by the current vegetation survey stand structures and wildlife habitat closely re- (CVS) in the Willamette National Forest had the sembling historical conditions within the water- most applicable stand inventories in addition to shed. These prescriptions were used in the LMS providing the most widely distributed sample plots. model for this analysis. It is assumed that for each By using CVS data for stand inventories we were prescription, reforestation (through replanting) will able to determine current vegetation condition for succeed under some level of overstory. It is also many stands. However, a significant number of assumed that, as after a natural fire, overstory stands remained that had no available tree-level retention will be nonuniform and distributed as data. To overcome this, we used all available in- patches and scattered individual trees throughout formation to assign inventory data to these the stand. Patchy overstory retention also will stands. By using a combination of vegetation as- stimulate understory growth in the gaps aiding sociation, stand type (tree species and structure), replanting success and development of vertical age, elevation, slope, aspect, and site index we structural diversity. Further structural diversity, were able to select the closest match from stands both horizontal and vertical, will be developed that had inventory data. Once this process was through several thinning entries prescribed for complete, we had an intact inventory for 980 each landscape area. stands giving us the current vegetation condition Simulation of stand growth within LMS used the of the entire watershed. ORGANON growth and yield model (Hann et al. 1994). ORGANON has the advantage of allow- Silvicultural Prescriptions, Stand ing detailed silvicultural treatments and provides Treatment Scheduling, and Growth quantitative estimates of important stand attri- Projections butes such as percentage of crown cover following a thinning. Another advantage of ORGANON As part of the Blue River Landscape Plan, the is that it provides estimates of tree quality charac- watershed was divided into two basic manage- teristics as an optional output. Tree characteristics ment categories, reserves and landscape areas. estimated by ORGANON include height to each Reserves were established to protect special in- branch, largest branch diameter, diameter of the terest areas such as late-successional spotted juvenile wood core at each branch whorl, and owl habitat and riparian areas along fish-bearing diameter inside bark of the stem at each branch streams. Landscape areas are the areas de- whorl. A similar output from LMS has been devel- signed to meet a variety of ecological and social oped but has not been fully tested. A limitation of objectives where some level of timber harvest will occur (fig. 2).

113 Figure 2—Blue River Landscape Plan treatment areas (Cissel et al. 1999).

114 Table 1—Blue River silvicultural prescriptions Prescription elements Landscape area 1 Landscape area 2 Landscape area 3 Rotation age (years)/% regeneration harvested annually 100/1.0 180/0.56 260/0.38 Landscape block sizes <40 ha (% of area) 60 20 20 40-80 ha (% of area) 20 40 40 80-160 ha (% of area) 20 20 40 Retention level (% of existing overstory crown closure) 50 30 15 Retention mixture Shade intolerant – Shade intolerant – Shade intolerant – (% of species dependent 65 80 95 on plant association) Shade tolerant– Shade tolerant– Shade tolerant– 35 20 5 Reforestation density (trees per hectare) 500 750 1000 Reforestation mixture Shade intolerant– Shade intolerant– Shade intolerant– (% of species dependent 40 60 75 on plant association) Shade tolerant– Shade tolerant– Shade intolerant– 60 40 25 First thinning (tph)a 500 at year 35 500 at year 20 to 25 500 at year 12 to 15 Second thinning (tph) 250 at year 35 275 at year 40 275 at year 40 Third thinning (tph) 150 at year 65 200 at year 70 200 at year 60 Fourth thinning (tph) Not planned 125 at year 100 125 at year 100 Low-severity fire (in Not planned Once between years 100 Twice between years addition to fuel and 180 100 and 260 treatments) aTph = trees per hectare. Source: Cissel et al. 1999. growth models is the inability to simulate patches as landscape blocks. An individual block ranges and gaps within a stand. To accurately estimate in size from about 40 to 160 ha and may contain the effect of these treatments on tree growth and several stands. Harvest entries were scheduled quality, models will need to have this failing cor- with an area-control approach by using the land- rected. Currently the best we can do is to use scape block as the basic unit. Blocks were desig- spatially uniform stand treatments and assume nated for entry in 20-year periods over 200 years. the effect on overall stand growth and quality is Scheduling of harvest entries in LMS is accom- negligible. plished through a scenario file that contains the year of treatment, stand number, and treatment A key part of the Blue River Landscape Pan is description. By using this file, silvicultural treat- when and where stand treatments occur. To move ments can be programmed for multiple stands away from further forest fragmentation, Cissel et across the entire landscape through time. De- al. (1999) applied treatments to areas designated

115 spite being able to simulate the growth of many ber production comes from tree removals as part stands for a large area, the current model cannot of stand treatments used to develop specific simulate interactions among stands. structural attributes through time. All approaches have an eventual rotation age that is longer than that used in traditional management for timber Analysis of Wood Volume and Quality production. The combination of using unique Analysis of wood production is completed after stand treatments with long rotations will have an stands have been grown and silvicultural treat- effect on the volume and quality of wood available ments applied by using the LMS model. LMS can for commodity production. It is unknown what the be configured to simulate bucking of trees into magnitude of this effect will be, and simulation user-defined log lengths. Currently, logs can be models give us an opportunity to look at potential defined by only diameter and length, but work is trends. being completed that incorporates additional quality characteristics such as number of branches, English Equivalents branch diameter, growth ring count, and diameter 1 hectare (ha) = 2.47 acres of juvenile wood core. For this study we focused on lumber production from each regime and dif- 1 millimeter (mm) = 0.039 inches ferences in tree characteristics. Tree-quality char- 1 centimeter (cm) = 0.39 inches acteristics are summarized for each regime from information in the ORGANON wood quality output 1 meter (m) = 3.28 feet file. To simulate lumber production, the TREEVAL model is used. TREEVAL is designed to accept Literature Cited ORGANON wood-quality output files as input. TREEVAL calculates financial information allow- Barbour, R.J.; Johnston, S.; Hayes, J.P.; ing comparisons of different silvicultural regimes. Tucker, G.F. 1996. Simulated stand character- For this study we were interested primarily in the istics and wood product yields from Douglas-fir volume and quality of lumber recovered for each plantations managed for ecosystem objectives. regime. Lumber recovery equations from Fahey et Forest Ecology and Management. 91(1997): al. (1991) have been programmed into the model 205-219. estimating yields of veneer and lumber. Lumber Briggs, D.G. 1989. Tree value system: descrip- recovery can be combined for visually graded and tion and assumptions. Gen. Tech. Rep. PNW- machine-stress rated (MSR) lumber. TREEVAL GTR-239. Portland, OR: U.S. Department of simulates the bucking of trees to a nominal log Agriculture, Forest Service, Pacific Northwest length of 4.9 m and a minimum small-end diam- Research Station. 24 p. eter of 13 cm. Volumetric lumber grade recoveries were estimated for each log by using the Christensen, G.A. 1997. Development of Doug- TREEVAL model. las-fir log quality, product yield, and stand value after repeated thinnings in western Conclusion Oregon. Corvallis, OR: Oregon State University. 176 p. M.S. thesis. For this paper, we wanted to demonstrate that by Cissel, J.H.; Swanson, F.J.; Grant, G.E. [et al.]. linking existing simulation models, a detailed 1998. A landscape plan based on historical fire analysis can be made from a landscape-scale regimes for a managed forest ecosystem: the perspective of wood volume and quality from Augusta Creek study. Gen. Tech. Rep. PNW- many individual stands. This information can be GTR-422. Portland, OR: U.S. Department of used to evaluate wood removals based on differ- Agriculture, Forest Service, Pacific Northwest ent approaches to planning on a large scale. Tim- Research Station. 82 p.

116 Cissel, J.H.; Swanson, F.J.; Weisberg, P.J. Sachet, J.K.; Briggs, D.G.; Fight, R.D. 1989. 1999. Landscape management using historical Tree value system: users guide. Gen. Tech. fire regimes: Blue River, Oregon. Ecological Rep. PNW-GTR-234. Portland, OR: U.S. De- Applications. 9(4): 1217-1231. partment of Agriculture, Forest Service, Pacific Northwest Research Station. 45 p. Fahey, T.D.; Cahill, J.M.; Snellgrove, T.A.; Heath, L.S. 1991. Lumber and veneer recov- Smith, W.B.; Faulkner, J.L.; Powell, D.S. 1994. ery from intensively managed young-growth Forest statistics of the United States, 1992, Douglas-fir. Res. Pap. PNW-RP-437. Portland, metric units. Gen. Tech. Rep. NC-GTR-168. OR: U.S. Department of Agriculture, Forest St. Paul, MN: U.S. Department of Agriculture, Service, Pacific Northwest Research Station. Forest Service, North Central Forest Experi- 25 p. ment Station. 147 p. Hann, D.W.; Olsen, C.L.; Hester, A.S. 1994. U.S. Department of Agriculture, Forest Serv- ORGANON users manual. Corvallis, OR: De- ice; U.S. Department of the Interior, Bureau partment of Forest Resources, Oregon State of Land Management. 1994. Record University. 113 p. of decision for amendments to Forest Service and Bureau of Land Management planning McCarter, J.B. 1997. Integrating forest inventory, documents within the range of the northern growth and yield, and computer visualization spotted owl. [Place of publication unknown]. into a landscape management system. In: 74 p. [Plus attachment A: standards and Teck, R.M.; Moeur, M.; Adams, J., eds. Pro- guidelines]. ceedings of the Forest Vegetation Simulator conference. Gen. Tech. Rep. INT-GTR-373. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 159-167.

117 Transforming Information From Many Perspectives to Action on the Ground

118 A Social Science Perspective on the Importance of Scale

Linda E. Kruger1

All ecological processes and types of ecological structure are multiscaled. Each particular structure relates to a particular scale used to observe it such that, at the scale of perception, the entity appears most cohesive, explicable, and predictable. The scale of a process becomes fixed only once the associated scaled structures are prescribed and set in their scaled context. Scaling is done by the observer; it is not a matter of nature independent of observation. (Allen and Hoekstra 1992: 11, emphasis added.) This paper offers a very brief introduction to some scale-related considerations from the perspective of a social scientist. The paper does not provide an indepth discussion, but my hope is that enough information is provided to provoke thought and additional exploration in these areas.

Scale and the Human Frame of places, usually at small scales. However, when Reference asked to comment on a forest plan at the scale of a national forest, many of us (as citizens) may be As Allen and Hoekstra (1992), among others, at a loss as to how to express how we feel about have noted, scale is a social construction. This those special places. Thus, both agency planning means that scale is not something that is out processes and research questions need to recog- there on the ground that exists apart from humans nize the interconnectedness across multiple but rather it is something we create by assigning scales and scale down to smaller scales so that defining attributes to it. A particular scale does not citizens can better understand the implications of exist in reality and has no meaning until we give it a decision or finding on “their place.” Often the meaning (Tuan 1974). Thus, those who select a boundaries identified for purposes such as plan- particular scale have a responsibility to clarify its ning and research are not socially significant or meaning to others. meaningful to citizens. In addition, scale only has meaning relative to the Nassauer (1997) suggests that landscape scale human frame of reference. Recall the relation does not have a universal definition. She identifies Gulliver had to the lands and people he visited. two common uses of the concept: The differences in scale that he experienced were all relative to his own size and personal past ex- (1) a heterogeneous combination of eco- perience. Because each of us brings a different systems, which affect each other across past experience with us, we will see the same space and time, and (2) a “middle” scale, things differently. This means that we need to look within a hierarchy, of ecological processes at places and projects together so we can share that affects smaller-scale processes and with each other what each of us “sees” rather is affected by larger-scale processes. Both than assuming we are all seeing and experiencing the concepts of heterogeneity and hierar- a place in the same way. chy emphasize the connectedness of the landscape across space and time. Con- Most of us (as humans) experience places at the nectedness, then, is an essential property scale of a site–a place we go fishing, camping, or of the landscape scale. (Nassauer hiking, for example. Some of us may experience 1997: 73) a sense of place or special attachment to certain Nassauer (1997: 73) goes on to suggest that “at the scale of human beings, connectedness refers 1 Research social scientist, U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, Forestry to landscape structure that allows flows of water, Sciences Laboratory, 400 N 34th Street, Suite 201, Seattle, nutrients, energy or species that people have WA 98103; Tel: 206-732-7800; e-mail: [email protected] noticed and believe to have ecological value.”

119 From this perspective, landscape scale is a cul- Determining appropriate planning or study bound- tural concept based on what we determine to aries is important. Just as important is keeping in have value. Because we as humans manage mind that boundaries are always porous and open landscapes, we often arrive at a definition of land- across time and space (Massey 1995). The identi- scape scale by identifying what we perceive as ties of places are products of links to other times management units (Nassauer 1997). Based on and places. Although each place is unique, its this perspective, landscape scale could mean interconnections to other times and places are someone’s backyard, a watershed, ranger district, important in defining its uniqueness. or national forest. “In each case, the ecological functions of individual patches of lakes, streams, Appropriate Scale turf, fields, forests, or even pavement are connected to one another” (Nassauer 1997: 73). Scale is an important consideration for both man- Thus, landscape as a scale of analysis is similar agers and scientists. Every resource issue and to the social concept of community in that land- question has a spatial and temporal component scapes, like communities, are nested and embed- or context. In choosing the appropriate scale, we ded in each other at various scales, rather than must consider at what scale our knowledge is representing one specific scale. reliable and at what scale we are most comfort- able making predictions (Haskell et al. 1992). It is this connectedness of ecological functions What the appropriate scale is depends on the across space that makes working across land- question(s) being asked and can range from a scapes essential. However, working across land- small personal space (How will this plant look on scapes is often easier said than done. Patchwork my desk?) to the entire earth (How might this ownership patterns put landscape-level decisions activity influence global change?). For social in- in conflict with our culture’s belief in private prop- quiry the most common scales are community, erty rights. More work is needed to develop tools county, state, and region. Counties are frequently and incentives for working across boundaries in used for analysis because they are a major unit ways that minimize the threat to people’s sense of used by the USDC Bureau of the Census. Machlis private property rights. (See Kalinowski’s [1996] et al. (1995) participated in the social assessment discussion of Leopold’s land ethic, specifically his process portion of the Interior Columbia Basin differentiation between land as property and land Ecosystem Management Project (ICBEMP). They as territory.) suggest the county as the best scale for social The ability to take a cross-boundary perspective— analysis, noting that data are readily available and looking across management jurisdictions, owner- that counties are increasingly involved in resource ships, and administrative and legal boundaries— policy through planning and zoning activities. They expands the possibility of considering questions at also suggest that the county scale fits best with the appropriate scale (Knight and Landres 1998). landscape-level analysis (Machlis et al. 1995: Like scale, boundaries are created by society for 13-16). a variety of social purposes. Boundaries organize Although there are advantages to the county scale space. They often cut across other social relations and major studies such as ICBEMP and Forest that define space. For example, the city I live in Ecosystem Management Assessment Team lies within a school district that includes two other (FEMAT) do not support higher resolution com- cities. It is bisected by a county line. Boundaries munity-level studies, there also are limitations to define to whom we pay taxes, for whom we can working at the county scale. The effects of re- vote, and who will represent us in government, source decisions are often felt at the local level, where our kids go to school, where our water and communities within a county can differ dra- comes from, and in some parts of the world even matically, particularly in a county that has a mix of where people can and cannot travel. Creating urban and rural areas. Working at the larger, boundaries is a process of place-making or con- county-level scale also tends to disenfranchise structing an identity that differentiates what and people whose interests are more local. These who are inside the boundary from what and who are often the people who will feel a study “has are outside the boundary. been done to them.” It also can be hard to make county-level data meaningful to a community. As

120 a society, we need to explore methodologies that Considerations in Choosing a enable people to learn about themselves and their Particular Scale places at a variety of scales that are meaningful to them. In choosing a particular scale to use when designing a study or planning area, asking the following The ability to consider a question at the appropri- questions may be helpful. ate scale may be constrained by a variety of factors including the scale at which data are avail- • What assumptions underlie the choice of this able, time or funding, political considerations, scale? Are these reasonable assumptions? logistical problems, and project location (Commit- • What is the justification for choosing this scale tee on the Applications of Ecological Theory to over others? Environmental Problems 1986). In many cases, ownership, management jurisdiction, or another • What is being overlooking or ruled out by administrative or legal constraint instead of the choosing this scale over others? Are we miss- question or issue itself determines the spatial area ing any critical questions? that is considered. • Are we choosing a particular scale because Working across disciplines or multiple types of we have data available or because measure- information may require synthesizing finer scale ment is easier than at other scales? information to a coarser scale common with oth- • What are the implications of working at this ers. For example, socioeconomic data may be scale? available only at the county scale, whereas biophysical data may be available at a combination • Does our choice of scale match our of scales including the watershed, stream reach, objectives? stand, or province. The identification of a cross- How does the choice of a particular scale affect cutting question may be a helpful first step in iden- ecosystem management efforts? In an attempt to tifying an appropriate scale at which to integrate begin to address this issue, Norton (1992) identi- across disciplines. fies four ways in which scale affects ecosystem The Committee on the Applications of Ecological management. Theory to Environmental Problems (1986: 97) 1. Spatial boundary issues. Jurisdictional bound- determined that a “fundamental problem in pre- aries usually do not correspond to natural dicting and controlling cumulative effects is the features or biogeography. frequently large mismatch between the scales or jurisdictional boundaries of management authority 2. Temporal perspective. Biological (and social) and the scales of the ecological phenomena in- systems are dynamic. Thus managers may volved or their effects.” The committee notes that have to decide whether to manage for current an affected environment often crosses several social values, production of “essential ser- jurisdictions. In addition the impacts of an action vices,” or features that are critical to maintain- may be felt in a different jurisdiction far from ing natural processes. where the offending action took place. 3. Organizational structure and scale. This as- Thus, the scale that is chosen must be large pect of scale has to do with substructure. enough to encompass interactions, larger scale Once a whole system is defined, subunits, processes, and cumulative effects (over time and each with their own boundaries, must be iden- space) without losing sight of important details, tified. Taking an example from social science, smaller scale processes, and interactions. At a nation may be made of states, which might broader scales, important variability can be be subdivided into counties, which have within masked; at narrower scales important effects, them cities and towns and neighborhoods levels, and nature of impacts can be missed; and where families and individuals live. at scales that are mismatched, it can be impossible to develop linkages between human communities and forests.

121 4. Scale and degree of impact. The degree of The Inclusion of Citizens in Planning impact an action may have on a system de- and Decisionmaking pends both on the scale of the activity and the scale of the system. Activities that might de- Transforming information into implementation on grade a small site might have negligible effect the ground requires both recognition that humans at a larger scale. As the scale of the activity are part of ecosystems (rather than an outside increases, it might have higher degrees of force) and the inclusion of citizens in planning impact at ever increasing system scales. Of- and implementation. This requires us to shift our ten the effects of our actions have impacts at thinking from “unfortunately we have to work with a variety of scales. people” to “citizens are critical to an informed, effective process.” Collaborative planning, when So what happens if our choice of scale is less used in Forest Service planning processes, re- than optimal? There are several possible out- quires participation of researchers, managers, comes of analysis or action at the wrong scale. and citizens. It recognizes that citizens have The following are drawn from the Committee on knowledge and real world experiences to contrib- the Applications of Ecological Theory to Environ- ute to the planning process. Frequently citizens mental Problems report Ecological Knowledge also bring a passion for action, and citizen partici- and Environmental Problem-Solving (1986). pation can result in better decisions and an im- 1. Well-intended efforts can have minimal or proved process. Land managers can share re- even adverse effects if planned for too short sponsibility for environmental quality and health a timeframe or too small an area. with those outside the agency only if they provide opportunities that include citizens as meaningful 2. Key processes might be overlooked at larger participants in planning and decisionmaking pro- temporal or spatial scales. cesses. 3. Averaging over larger areas can mask the It also requires an understanding of the various importance of processes at smaller scales scales at which people identify a landscape as a because it is more difficult to detect subtle place that they relate to. Geographer Gillian Rose differences. (1995) suggests that people develop a sense of In addition to these concerns, a focus on single belonging or connection to places at multiple actions at whatever scale can result in obscuring scales including a local scale, a regional scale, a cumulative effects over time and space. Cumula- national scale, a supranational scale, like Europe, tive effects can take several forms ranging from and even at a global scale, as our social and eco- activities that add materials to the environment, nomic systems become more globalized. This such as discharging effluents, to those that re- means that as managers consider who should be move materials from the environment, such as involved in local planning processes, they must harvesting timber. Cumulative effects can take the think beyond the local geographic community. form of changes over large areas for long periods “What kinds of evidence, scientific or otherwise, of time that result from management actions, are necessary to justify a given choice regarding such as when a forest is managed based on the scale on which environmental problems are single stands. Cumulative effects also occur when addressed?” Norton (1992: 35) answers his own several actions compound each other; for ex- question by suggesting an interdisciplinary ap- ample, when logging roads are built, logging, rec- proach that acknowledges that questions of scale reation, hunting, poaching, and other activities involve value judgments. “The correct scale on increase and affect a variety of forest species. which to address a management problem is de- The importance of understanding and considering termined by what society wants to accomplish the implications of cumulative effects makes a with that system” (Norton 1992: 36). Norton joins strong case for the consideration of multiple spa- others in suggesting that citizen involvement in tial and temporal scales. dialogue and debate is necessary in order to facilitate consideration of alternative actions and desired conditions.

122 Scale may influence citizen involvement in man- personal experiences both with the specific place agement and planning. It can be much more diffi- and the agency proposing the activity and their cult for citizens to comment on large-scale level of trust in the agency often play a role. A proposals that are not linked to specific small- sense that “too much has already been cut” or scale places they can personally identify with. general concerns over wildlife, recreation, or sce- Forest plans are at a scale many citizens find nic values also may play important roles. Finally, difficult to respond to. We might speculate that how the decisionmaking process is carried out is watersheds, especially at the local community important to some people. If people feel they level, are a more meaningful scale as evidenced didn’t have opportunities for meaningful participa- by the number of grassroots watershed groups. tion or if they didn’t feel heard, they may object to an activity regardless of how the project looks on Frequently, federal land management agencies the ground. I think a quote from the child’s story, A conduct planning processes at the scale of a na- Little Prince (Saint-Exupery 1943: 87) aptly de- tional forest. Ecosystem management is based on scribes this situation: “What is essential is invis- planning at a large scale, often including areas ible to the eye.” What is essential to understand- larger than a national forest. The FEMAT and the ing how acceptable a forest practice is may have ICBEMP each covered several states. At these less to do with how the forest appears and more broad scales it is easy for people and their inter- to do with aspects of how and why it looks the ests and concerns to get lost. way it does and where it is located relative to How meaningful public participation can be is places people care about. often a function of the scale at which a study is initiated. The scale or scope of a place being Literature Cited studied is important in the sense that it has to be a scale that people can relate to. What is impor- Allen, T.F.H.; Hoekstra, T.W. 1992. Toward a tant is that the scale is appropriate to the question unified ecology. New York: Columbia University or issue being addressed and that it does not Press. 384 p. obscure what is meaningful and what people can Brunson, M. 1993. “Socially acceptable” forestry: relate to. What does it imply for ecosystem manage- We must also take care that the tools and tech- ment? Western Journal of Applied Forestry. niques that we use do not obscure what it is that 8(4): 1-4. is meaningful and important to people. Technol- Committee on the Applications of Ecological ogy has provided us with tools that enable us to Theory to Environmental Problems. 1986. create pictures of what a stand or landscape will Ecological knowledge and environmental prob- look like immediately after harvest and at intervals lem-solving. Washington, DC: National Acad- thereafter. In addition to its use in presenting infor- emy Press. 388 p. mation, this technology is often used to solicit public response to harvest practices (Shindler et Haskell, B.D.; Norton, B.G.; Costanza, R. 1992. al. 1995). Although the technology can serve a What is ecosystem health and why should we useful purpose in helping citizens visualize what worry about it? In: Costanza, R.; Norton, B.G.; the land will look like over time as the forest Haskell, B.D., eds. Ecosystem health: new greens up, some researchers have concerns with goals for environmental management. Wash- its use in measuring public judgments of accept- ington, DC: Island Press: 3-20. ability. Kalinowski, F.A. 1996. Aldo Leopold as hunter Studies have found that judgments of acceptability and communitarian. In: Vitek, W.; Jackson, J., of management practices are based on more than eds. Rooted in the land: essays on community just looking at a scene and responding to what is and place. New Haven, CT: Yale University there (Brunson 1993, Shindler and Cramer 1999). Press: 140-149. Research done by some of our collaborators has Knight, R.L.; Landres, P.B., eds. 1998. Steward- identified a number of factors that play a role in ship across boundaries. Washington, DC: Is- the acceptability of resource management deci- land Press. 371 p. sions and actions. For example, the respondents’

123 Machlis, G.E.; Force, J.E.; Dalton, S.E. 1995. Rose, G. 1995. Place and identity: a sense of Monitoring social indicators for ecosystem place. In: Massey, D.; Jess, P., eds. A place in management. Walla Walla, WA: Interior Co- the world? Places, cultures and globalization. lumbia Basin Ecosystem Management Project. Oxford, United Kingdom: Oxford University Unpublished report. On file with: Pacific North- Press: 88-132. west Research Station, Human and Natural Saint-Exupery, A. 1943. The little prince. New Resources Interactions Program, Forestry York: Harcourt, Brace and World, Inc. 113 p. Sciences Laboratory, 400 N 34th Street, Suite 201, Seattle, WA 98103. Shindler, B.; Cramer, L.A. 1999. Shifting public values for forest management: making sense Massey, D. 1995. The conceptualization of place. of wicked problems. Western Journal of Ap- In: Massey, D.; Jess, P., eds. A place in the plied Forestry. 14(1): 1-7. world? Places, cultures and globalization. Ox- ford, United Kingdom: Oxford University Press: Shindler, B.; Peters, J.; Kruger, L. 1995. Social 45-85. values and acceptability of alternative harvest practices on the Tongass National Forest. Nassauer, J.I. 1997. Cultural sustainability: align- 101 p. Unpublished report. On file with: Pacific ing aesthetics and ecology. In: Nassauer, J.I., Northwest Research Station, Human and Natu- ed. Placing nature: culture and landscape ecol- ral Resources Interactions Program, Forestry ogy. Washington, DC: Island Press: 67-83. Sciences Laboratory, 400 N 34th Street, Suite Norton, B.G. 1992. A new paradigm for environ- 201, Seattle, WA 98103. mental management. In: Costanza, R.; Norton, Tuan, Yi-Fu. 1974. Space and place: humanistic B.G.; Haskell, B.D., eds. Ecosystem health: perspective. Progress in Human Geography. new goals for environmental management. 6: 213-252. Washington, DC: Island Press: 23-41.

124 125 Conclusion

126 Summary Remarks on Views From the Ridge

T.F.H. Allen1

No. Ecosystem management will not work unless I was asked not to prepare remarks ahead of the it involves explicitly giving people options that they workshop, but rather to respond to what I saw are prepared to use, and a sound basis to select and heard. Let me say at the outset, I have been among them. much impressed with the general thrust of this workshop. Many of the papers were technical and What I have described above is what Silvio represented sound research, but the real strength Funtowicz and Jerry Ravetz call postnormal sci- of the presentations as a whole was in the genu- ence (Ravetz 1999). Although normal, merely ine eagerness in tackling big problems. The big modern science is meticulous as it tries to ap- problem, as I see it, is that we have many large- proach reality with its models; the postnormal scale problems with regard to renewable re- scientist cannot afford to be so idealistic. Post- sources, and that there is a huge divide between normal situations are characterized by (1) high how science prefers to work and what needs to stakes, (2) insufficient data, and (3) a short time be done. We should be impressed with the high fuse. The data from normal science will not be level of appreciation for the difficulties, and the coming in time, for the issue will have passed, win vigor of the response in so many of the presenta- or lose, before meticulous experiments and confi- tions. I saw courage in the presenters I heard. dence limits can be achieved. Even so, scientists must offer their best advice anyway, and the man- As an example, we are really quite good at quanti- ager must act on that best advice. Clearly finding fying arcane models about cutting regimes. But out what is the reality of the situation is irrelevant, what is the use of this to the end consumer of but, if it is any comfort, reality was never achiev- these results? The models mean nothing until we able anyway. visualize the landscape in images that anyone can see and understand. I saw exactly that at this Beyond postnormal science is postmodern sci- workshop. The new gigabytes and gigahertz of ence. In the classical world and modern world the computer capability are powerful, because they novice became dexterous, and the apprentice let us translate abstractions, such as a 10-percent became a craftsman, while the master went be- cut, into an image that the public can see and yond all to create something in a new framework. evaluate. New technology really does change the The novice and apprentice achieve quality by game, as we fly down streams and summarize being meticulous, delivering exactly what is their condition with thermal imagery. The prob- needed every time. That is called structural qual- lems are large, but the new technology moves ity. The master introduces a different quality, a human achievements upscale to meet many of dynamical quality of good change, which itself them. Even the papers that used no technological undermines the premises that underlie structural “fancyware,” were generally very cognizant of the quality. Quality classical work was approved by need to serve the interface between humans on common standards, where the external consen- the landscape and the biogeophysical landscape sus was that everyone liked the picture of the itself. When we manage an ecosystem, we do not Patron. The difference between classical and manage the material ecological system itself; modern is that in the modern world, external real- rather we manage the presence of the people ity becomes the external reference. Elites, be they who impact the system. And that does not mean cubist painters or 20th-century scientists, assert we manipulate the public covertly or dishonestly. that by looking at the world their special way, ultimate reality may be approached. But in a 1 Professor, Botany Department, University of Wisconsin, postmodern world there is no external reference, Madison, WI 53706-1381; Tel: 608-262-2692; e-mail: be it consensus or reality. In postmodernity, it is [email protected] not that it is all in our heads and anything goes, it is rather that we only deal with data, not with ob-

127 jective reality. Without an external reference, the quality. Leaf-cutting ants have evolved further to postmodern scientist must make the process of use leaves instead of guano, and so overcome science justify itself, but how? The answer is that the scarcity issue. But leaves are a low-quality we need to keep science a high-quality activity resource, and so demand very organized pro- in both structural and dynamic features. Even cessing. Diminishing returns on a high-quality though we cannot now pretend that science is resource, such as guano, end in collapse when objective, it is quality that is endogenous to doing the supply runs out. Diminishing returns on a low- science that continues to give science its de- quality resource, such as leaves, end in collapse served privileged position (Funtowicz and Ravetz when the huge system grows into excessive de- 1992). Science is meticulous so it can repeat mand. Counterintuitively, more capital is built on anything it does, and that amounts to structural diffuse, low-quality resources than on high-quality quality. Science also challenges its products at resources. every level from the null hypothesis all the way to It appears that human society often switches from fighting over new paradigms. That is dynamical high-quality resources (HQR) to low-quality re- quality. sources (LQR) in cycles of elaboration. The ener- Structural and dynamical quality, that is, honest getically easy entry to a general type of resource carefulness and creativity, make science the best is on HQR, and the elaboration upscale is on game in town. And we saw lots of both sorts of LQR. Thus (1) hunting (HQR) leads to (2) agricul- quality at this workshop. Some papers were tech- ture (LQR). In an agricultural setting, (3) looting nical and were splendid displays of structural neighbors (HQR) gives way to (4) imperial taxa- quality. Other papers were full of dynamical qual- tion of peasants (LQR). In the First World we live ity, as managers, politicians, scientists, and the on (5) industrialization (HQR). Without meaning to lay public made visionary statements of what to describe any sort of grand historical narrative, we do next, as we reach out to a world full of people note that some political arrangements collapsed with values living in systems that are the purview at various stages, e.g., Mongols in the West at of the Forest Service. step 3, or the Romans at step 4. It is good to see such courage and commitment We look as if we are about to come off the back to dealing with the values of people in the bio- side of step 5, fossil carbon, and need a new LQR sphere, because the problems are huge. Joseph phase. Because renewable energy sources are all A. Tainter (1988), of the Rocky Mountain Re- low quality, they will have to be extensively cap- search Station, points out that societies are prob- tured, rather like sun on crops. Much more eco- lem-solving units that elaborate processes or logical damage has been done by LQR agriculture approaches to deal with resource issues. It ap- compared to HQR fossil fuels. The change to the pears inevitable that problem solving proceeds to LQR hydrogen economy using wave, wind, bio- greater cost and less benefit with vicious dimin- mass, and solar energy may be wrenching, as ishing returns. Eventually societies lose either the human settlement decentralizes. The diffuse na- political will or the ability to deal with the last prob- ture of high-technology information systems al- lem, whatever form it may take, and the people lows for energy self-sufficient production systems just walk away from civilization. scattered across the landscape. There are alternative scenarios of the move to renewable energy Some societies on the edge of collapse find a new where decentralization is delayed or circum- lease on life by switching to some fundamentally vented, but these have their downside. They in- new resource, such as coal when Britain ran out volve consumption of the mountains of dirty coal, of wood. Looking at attoid ants, we see that the as a society builds massive LQR infrastructure primitive species farms fungi on grasshopper where forests of windmills sit on sculptured coast- droppings, which amounts to jet fuel for growing lines (Allen et al. 2001). Avoiding renewable en- fungi. The problem is that insect droppings are a ergy appears to invite the four Horsemen of the highly processed, scarce material, albeit of high Apocalypse, so we cannot afford to fail. And fail

128 we will unless we can encourage the populace at Literature Cited large that the transition is worth it. The human focus and the emphasis on advice from lay opin- Allen, T.F.H.; Tainter, J.A.; Pires, J.C.; ion at this workshop are most encouraging at a Hoekstra, T.W. 2001. Dragnet ecology, “Just time when deep concern for public opinion is not the facts ma’am”: the privilege of science in a misplaced. The coming LQR society is set to dev- post-modern world. BioScience. 51: 475-485. astate coastlines, deserts, forests, and wildlife. It Funtowicz, S.O.; Ravetz, J.R. 1992. The good, is for this reason that the visionary tone of this the true and the post-modern. Futures. 24(10): workshop is so important. Even under the happi- 963-974. est outcome, ecological stress will be great. Pre- cisely because of impending ecological disaster in Ravetz, J.R. 1999. What is post-normal science? the move to extensive use of renewable energy, Futures. 31: 647-653. only the best stewardship will do. The problem is Tainter, J.A. 1988. The collapse of complex soci- large, so the view of the stewards of natural re- eties. Cambridge, MA: Cambridge University sources such as forests, wildlife, and fisheries Press. must be at least as visionary as the most expan- sive view from the ridge.

129 The Future of Landscape-Level Research at the Pacific Northwest Research Station

Thomas J. Mills1

Key Drivers of Landscape-Level What put us on the landscape trek? In part, it was Research the result of our fine-scale and single-discipline research. Although often initiated from a single Station scientists at the Pacific Northwest Re- disciplinary perspective, this research has pro- search Station (PNW) work to advance the basic duced important scientific findings. At the same understanding of how systems function, especially time, it has shown that some of those fine-scale terrestrial, aquatic, and socioeconomic systems. phenomena also were components of large sys- This provides the foundation of scientific under- tems. We have found that some relations exist standing to better understand the consequences and manifest themselves only at a large scale. and risks of choices made in managing the land. As we examined outbreaks of insects and disease It also provides the foundation on which to help during the past several decades, for example, we design new land management options. began to understand the need to look beyond the In the 1990s, in response to key issues, Station stand. Also our research on fire disturbance pro- scientists increasingly focused their research on cesses and smoke management led to a broader the landscape scale. This workshop reflects a perspective. The ecological effects of forest frag- body of work designed by PNW to consciously mentation and the interactions between fragmen- examine this scale. In many ways, we have only tation and habitat quality for some wildlife species started to understand systems at the landscape begged a landscape-scale view. Increasingly we scale in the Pacific Northwest, but it is an impor- understand that pattern and patches across the tant part of our research program that is already landscape affect relations such as those among demonstrating that it holds considerable promise. the nesting, feeding, and migrating of some wild- We appreciate being able to share information life species. with those of you who make land management In addition to our fine-scale research, some defin- decisions or affect those decisions and the scien- ing land management issues helped us to look at tists here who have validated and built on our broader scaled relations. Issues such as manage- research results. ment of old-growth forests, recovery of threatened Much of our landscape-scale research has fo- and endangered species, and forest health are cused on national forest land issues because they examples. These types of issues have been driv- offer landscape-scale laboratories. Our researchers of our research program, just as our scientific ers often work side by side with land managers as understanding has helped shape the character of we conduct studies and assessments on the very these issues. large scales of the national forests. We also con- Land management issues such as these, rein- duct landscape-scale research on other lands forced by our past research, have led to the fol- such as the state lands of Washington. Much of lowing overarching land management questions what we are learning is applicable to all owner- that demand broader scale, functionally integrated ships. As we study forests at the landscape scale, research: How do the effects of land management I believe we can find some additional options that actions accumulate over space and time? What may create compatibility among interests and are the effects of interactions among different across ownerships. land units, and how do those interactions affect overall outcomes? Another related question, given 1 Deputy Chief, Business Operations, U.S. Department of the inability to treat all parcels, is, How can we Agriculture, Forest Service, 201 14th Street SW, Washington, DC 20250; Tel: 202-205-1707; e-mail: [email protected] prioritize among the numerous parcels to maximize our objectives?

130 These questions require a landscape-scale per- Policymakers have tried to solve this challenge by spective and integration among different scientific implementing land allocations akin to land use disciplines. Moving to broader scale is usually zoning. Land allocations in national forest plans coupled with the need to study the interactions are a simple version of this. Our research chal- among different components of the land, as well lenge is to sufficiently understand the broad-scale as the interaction among the different land com- phenomena so that creative solutions may be ponents within the landscape. Integration among developed to provide an array of values that blend physical and biological disciplines is needed, but uses across the landscape, and even shift uses integration between the biophysical and social across time in a way that increases benefits and sciences also is essential if we are to understand reduces negative effects and risks. the full array of consequences of land management actions and to create new land management Building on the Foundation of Past options. Research The Station has a deep history of long-term re- Many PNW scientists and their colleagues have search studies and an emerging body of land- presented examples of landscape-scale research scape-scale research, but research integrated at this workshop. The PNW Station is making a across disciplines at the landscape scale is only concerted effort to address the landscape scale. now emerging. Although working across disci- Broad-scale studies like that of land cover dynam- plines can be complicated and time consuming, it ics and land use changes described by Ralph Alig is in the resulting comprehensive picture that and the study of the implications of scale for the policymakers likely will find solutions to complex study of terrestrial wildlife populations by Martin problems. Raphael are only two examples of the work presented here. The poster session showed a variety Land Management Opportunities of projects related to landscapes that PNW has Among the Challenges underway. And yet during this workshop, although we presented much, it was not all work conducted Not only does the landscape-scale perspective at the landscape scale by the Station. Some land- add significant complexity beyond what is appar- scape research spans decades for issues such ent through a finer scale look at the land, it also as insects and disease. Some of our research holds the promise of land management opportuni- has just begun, and information is yet to be ties not apparent at the finer scale. For example, developed. if the focus is solely on the fine scale, a manager or the public might conclude that only the area As with the rest of our research, much of this being studied can provide some desired value. work on the landscape scale is conducted with They might also conclude that all such sites must colleagues at universities and through partner- provide the same set of values. Our research has ships with land managers. John Cissel’s presenta- shown, however, considerable variability across tion of the Augusta Creek study was an example. the landscape. The land is not now uniform nor Critical in our success is the ability to work with was it ever, even when nature was managing it. the National Forest System and other land man- For example, not all sites on the west side of Or- agers on large tracts of land. The contribution we egon and Washington were prime salmonid habi- have made so far, our continued cooperation with tat at any one time even in the past. And some of universities, and our partnerships with land man- the most constraining laws, such as the Endan- agers create a strong foundation from which to gered Species Act, require only that values be continue studies at a broad scale. provided on enough sites to meet an overall goal, not that those values need to be provided from Challenges to Landscape-Level every site. A landscape view helps us understand Research how a mosaic of uses might be scattered across the landscape. Each parcel might be devoted to In accomplishing landscape-level research, espe- delivery of some set of values with the designated cially with research integrated across many disci- uses moving across the landscape over time, as plines, we face significant challenges. To make nature moved them. meaningful progress, we need to face and overcome these challenges.

131 We need to fully develop concepts, models, and emphasis to define objectives and hold managers especially testable hypotheses. For the most part, accountable for performance results. We have these do not yet exist. Questions abound, but overcome this challenge in some places, such as rigorous scientific methods to answer the ques- the large-scale studies in the Willamette National tions need to be developed further. Without the Forest, but we need to build on those few suc- benefit of stronger methods, there is always the cesses, including working across ownership risk that the personal values of scientists will be- boundaries. come embedded in the framing of questions or Even when we are successful with large-scale, the interpretation of results. designed experiments, we invariably have few Landscape-scale research often but not always, replications. It is difficult to commit extensive requires integration of several scientific disci- acreages, and it is hard to find enough compa- plines. Success depends on our ability to under- rable sites. We need to somehow develop means stand interactions among biophysical components of testing hypotheses that are not so dependent and between biophysical and social systems. But on large sample sizes. We will never have the the science perspective of the different disciplines sample sizes typical of fine-scale studies, but that often focuses at widely divergent scales, and typi- cannot stop us from drawing defensible infer- cally with very different boundaries. For example, ences. most social science is conducted at broad scales We need to expand our capability to draw infer- and creates behavioral models developed at the ences from other than designed experiments. broader spatial scales but still useful at smaller Modeling has long been a tool to learn about rela- scales. The biophysical sciences, on the other tions. Models will play an even more important hand, are challenged to “scale up” with both data role in large-scale research, but we need to de- and models. For scientists to address landscape velop better means to measure the confidence we issues, synthetic and interdisciplinary science can place in their results. We need to develop must overcome these technical and conceptual new methods to reveal the confidence, or in turn, difficulties. To date, we have mostly accomplished the uncertainty, of our scientific information that integrated research by synthesis of knowledge describes large-scale relations. already gained. Instead, we need to plan for integrated approaches as we develop hypotheses, Landscape research requires vast amounts of concepts, models, and methods. We must gain data. Other opportunities for collection of impor- the ability to determine early in our research how tant landscape data may emerge as monitoring we want to address the pieces to understand the data are generated, such as is proposed for the whole. federal lands in the range of the northern spotted owl. We will have to develop more rigorous data Through its ability to bring together staff scientists accumulation and data management systems and those from universities, the Station has con- than customarily are needed for finer scale re- tributed to and will continue to work toward such search. This technological challenge emerged integrated research. The Station has an advan- with the study of landscape relations in the interior tage in being able to bring scientists together, and Columbia basin as in the work described by Paul we need to exercise that advantage. Hessburg. Just as important as the technology, Landscape science requires at least some large- however, is the discipline to use it. scale experiments. Large-scale experiments are Successfully addressing landscape-scale issues challenges from several perspectives. One is the requires complex decisionmaking, often involving difficulty of convincing land managers and the multiple agencies and landowners. A willingness public that the search for knowledge should re- to research transboundary effects is necessary, quire testing ideas, and we are testing them be- and success in both science and implementation cause we do not know what the outcome will be, will require public-private cooperation. These even as we are guided by a hypothesis. Even if transboundary challenges highlight the impor- land managers are willing to accept and partici- tance of socioeconomic landscape research. pate in these experiments, management with Biophysical research alone simply will not be suc- uncertain outcomes is not consistent with the cessful in providing the scientific information to deal with larger scale issues.

132 Landscape science may be burdened with the Conclusion a priori expectation that it will produce more than scientific information. Scientists, managers, and We have a strong foundation of research on the public may develop exaggerated expectations which to build future landscape-scale research, about the nature of results of studies about land- and the PNW Station and others in the science scapes, perhaps even more than about finer community have already begun to do landscape- scaled research studies. Sound scientific informa- scale research. There are compelling scientific tion is an essential ingredient to sound decisions, and policy reasons for conducting landscape- but information alone does not make decisions. scale research, especially research integrated Any decision invariably requires the integration across scientific disciplines. Working with our of multiple components that can only be accom- research collaborators and land management plished by value-based weighing of tradeoffs partners, we hope to build on this foundation and among those components. That is the stuff of make meaningful advances—advances that may decisionmaking, not science. While insisting that not only help us better estimate the conse- the available science must be fully and faithfully quences and risks of current management op- considered in the decision process, those respon- tions but also help us all create new options that sible for research results must be careful to avoid might embody more compatibility among the the misperception that science “makes” decisions. many values society seeks from the land.

133 This page has been left blank intentionally. Document continues on next page. This page has been left blank intentionally. Document continues on next page. This page has been left blank intentionally. Document continues on next page. The Forest Service of the U.S. Department of Agriculture is dedicated to the principle of multiple use management of the Nation’s forest resources for sustained yields of wood, water, forage, wildlife, and recreation. Through forestry research, cooperation with the States and private forest owners, and management of the National Forests and National Grasslands, it strives—as directed by Congress—to provide increasingly greater service to a growing Nation.

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, gender, religion, age, disability, political beliefs, sexual orientation, or marital or family status. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD).

To file a complaint of discrimination, write USDA, Director, Office of Civil Rights, Room 326-W, Whitten Building, 14th and Independence Avenue, SW, Washington, DC 20250-9410 or call (202) 720-5964 (voice and TDD). USDA is an equal opportunity provider and employer.

Pacific Northwest Research Station, 333 S.W. First Avenue, P.O. Box 3890, Portland, Oregon 97208- 3890.

Pacific Northwest Research Station

Web site http://www.fs.fed.us/pnw Telephone (503) 808-2592 Publication requests (503) 808-2138 FAX (503) 808-2130 E-mail [email protected] Mailing address Publications Distribution Pacific Northwest Research Station P.O. Box 3890 Portland, OR 97208-3890 U.S. Department of Agriculture Pacific Northwest Research Station 333 S.W. First Avenue P.O. Box 3890 Portland, OR 97208-3890 Official Business Penalty for Private Use, $300