As part of the residency programs that take place at the H.J. Andrews Experimental Forest, we ask residents to visit three "Reflections Plots," sites of significant natural and research interest. The following details about the Reflections Plots are based on notes from Fred Swanson, Spring Creek Project Senior Fellow and long-time former researcher at the Andrews Forest.

Everyone has their own views of place. The Andrews Forest scientist and manager communities have visited some of these places many times—as individuals and in groups. We have shared our views of them with one another and with thousands of visitors. We have some shared understanding of these places, but also our individual perceptions of them.

Here, we attempt to characterize the sites in several ways, beginning with the most objective and dispassionate and moving to the more subjective and interpretive. Visitors who come to reflect at the plots may wish to begin with no explicit frame of reference—to visit the sites fresh and without preconceptions. The visitor could then gradually indulge in what others think they have learned about the site and from it by reading increments of this text and then reflecting on the additional information. Visitors could then track the development of their personal engagement with the places.

A further reason for developing these notes is that these places and the stories we tell when we go there with visitors have become an important part of the oral traditions of the Andrews Forest. The Long-Term Ecological Reflections (LTEReflections) program is an effort to foster new oral and written communications and traditions about this forest and watershed, and to provide impetus to record some of the earlier Andrews Forest stories.

Let's try it. It is part of the LTEReflections experiment.







Biophysical setting, including management history

This flat bench high above Lookout Creek is covered with old-growth (ca. 500 yrs) Douglas-fir/western hemlock forest. We are in the middle of the Andrews Forest along the main valley floor, somewhat protected from storm winds and bathed in cold air drainage, which may contribute to the low fire frequency of this part of the landscape. The stand appears to have not been burned in that time frame, except for some patches of char on the bark of some of the old trees suggesting a ground fire a century or so ago. Some of the western hemlock and western redcedar trees in the middle-story of the stand may have established or been released by that light fire. The big Doug-firs exhibit the characteristic fan-shaped branching systems. Lower and middle tree strata are also populated with shade-tolerant Pacific yew, the druid of the forest with its cloak of mosses and modest stature after several centuries of growth. Logs in many stages of decay litter the forest floor. The area beneath denser patches of hemlock are very dark, forming "anti-gaps" with little understory vegetation in contrast with shrubby, canopy light gaps where overstory trees have toppled. Light salvage logging cut the original temporary road bed into the area in the ca. 1960s and research use of the area freshened up the crude road in the mid 1980s.

Research context

This is one of six sites in Mark Harmon's 200-year log decomposition study, which is the source of most of the "researcher tracks" (or trash)—plastic material for sampling various processes associated with log decay, tags for marking logs, evidence of cutting of logs for sampling purposes, netting from insect exclosures. Story has it that Jerry Franklin was the originator of the idea—one of his many monumental studies. The study began in 1985 when nearly 530 logs 6-m long and 50-cm diameter were placed on the forest floor in six locations across the environmental field of the Andrews Forest (see Harmon 1991 for establishment report). The logs came from cutting patches of forest on the Andrews and were chosen for their size and absence of defect. They are of four species representing a gradient of decay rates (from fast to slow): Pacific silver fir, western hemlock, Douglas-fir, western redcedar. Harmon rode in the log truck and interacted with the loggers in the course of setting up the study to assure that the design was achieved.

The log decay processes involving fungi, invertebrates, and other agents are tracked over time by periodically sampling a subset of logs of the four species. Logs are sectioned (cookies removed from the logs at various positions along its length) of each species are taken to the lab, photographed, and examined for bulk density, chemistry, and other properties. The 200-year aspect of the experiment refers to the planned destructive sampling of logs with sampling in geometrically increasing time steps. Harmon has attempted to carry out study documentation, data management, and marking of logs in the field so the study can proceed over that life-time, though he will not be around to complete it.

Harmon has brought many focused studies to these large-scale, long-term, experimental study sites. Effects of individual species of invertebrates and fungi are considered in terms of their roles in decomposition and nitrogen cycling. Log hydrology and biogeochemistry research examines the amount and chemistry of water landing on logs, running over their surface and thru them. All this work can then be placed in the larger context of the roles of the regions forests in carbon dynamics and sequestration. He and others have established parallel studies in stream ecosystems within the Andrews Forest.

A critical feature of any research endeavor is publication. To form a milepost zero for our knowledge about roles of dead wood in temperate forest and stream ecosystems, a team of 13 Andrews scientists published a 170-page monograph based on the global literature as of the mid-1980s with Harmon, a graduate student at the time, lead author (Harmon et al., 1986).

The feel of the place—ripple effects of learning represented by this site

When it was first established, I was very apprehensive about going to this site with forest managers and the general public, fearing negative reaction to the "waste" of good wood left to rot and the crazy behavior of making this happen. But, rather than criticize the effort, the first manager visitors asked how to do a better job of managing the dead wood resource of the forest ecosystem. They must have been impressed that scientists would think so highly of the importance of dead wood to have gone to this much trouble to study it. I believe that has been a powerful subliminal message for the study site—"Hey, this stuff is damned important in the ecosystem."

All the focus on the science of dead wood in forest ecosystems has lead to a great deal of attention to management of dead wood. Willamette National Forest and other land managers have made great changes on practices affecting dead standing and downed wood in cutting units, riparian zones, and streams, based in part on findings from studies such as the log-decomposition experiment. Harmon has coined the term "morticulture" to suggest that we need as much creativity in management of the dead wood as we give to management of the living through silviculture. Management considerations including retaining live and dead trees on cutting sites with an eye to retain the ecological roles of dead wood on the site over long periods into the future. Harmon has also modeled alternative management scenarios to examine tradeoffs between wood removal for human consumption and retention of wood on the site for carbon sequestration objectives.

Over the course of the log decomposition study, the issue of global climate warming and roles of forests in sequestration of carbon have become prominent. This study site and related modeling by Harmon and colleagues are pivotal in the argument that the dead component of forest ecosystems are critical parts of the story (Harmon et al., 1990). Some were arguing for cutting old forests and establishing vigorous plantations, but this argument ignores that the importance of carbon stored in the dead components of natural forest stands.

The year 2005 was the 10% anniversary of the log decomposition experiment. Normally we think of anniversaries as counting from day one (birth, wedding), but with a research design aimed to an endpoint, we can think ahead. But a professional lifetime is only 15-20% of the lifetime of this study, so in such long-term experiments we are: 1) banking on the future; 2) figuring on learning valuable lessons from taking the long view and seeking to learn in ways that are not possible with a short-term view; and 3) squarely addressing long-term phenomena. Other researchers will come along with new tools to address new questions using these studies.

The backdrop for this log decomposition study site, and an essential point for reflection, is a fine old-growth stand. The planned 200-year life of the decomposition study is but 40% of the lifetime of this forest stand as the study begins. The log decomposition study helps us push our thoughts into the future, while the stand in which it resides pushes us to reflect on the past half millennium that is embodied in the stand.

When we visit the site with political folk we like to reflect on when the study would have to have begun if we were completing it now—in the time of Thomas Jefferson, Benjamin Franklin, and other framers of our government. We point out the importance of commitment to processes on the time scale on which they actually operate.

An important point for reflection is the idea of starting a study you know you will not complete. Barry Lopez said of this study that it represented remarkable sense of humility on the part of the scientists. On the other hand, a monumental experiment stands as a focal point for discussion, argument, and learning at all stages of its lifetime. Now we have Robert Michael Pyle's thoughts on these matters in his field notes and poems from his Andrews Forest residency and in the short essay "The Long Haul" in Orion (fall 2004).

Literature cited

Harmon ME, Ferrell WK, Franklin JF. 1990. Effects on carbon storage of conversion of old-growth forests to young forests. Science 247:699-702.

Harmon ME. 1991. Long-term experiments on log decomposition at the H.J. Andrews Experimental Forest. Gen Tech. Rep PNW 280. Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. 28 p.

Harmon ME, Franklin JF, Swanson FJ, and others. 1986. Ecology of coarse woody debris in temperate ecosystem. In: MacFadyen A, Ford ED (eds). Advances in Ecological Research. New York. Academic Press. 15:133-302.



(Near Headquarters site. Also known as the Headquarters Reach and the River Continuum Site.)

Biophysical setting

We are standing on a gravel bar in Lookout Creek, which is draining a watershed of about 60 km2 (24 mi2) in the western Cascade Range. The adjacent forest is ca. 500-year-old Douglas-fir trees with western hemlock, western redcedar, Pacific yew, broadleaf maple, and several other tree and large shrub species. Some of the large Douglas-fir trees died in the early 1990s following bark beetle attack, which followed toppling of trees (notable on the slope above the south bank) by a wind storm. The gravel bar itself is partially covered with a developing stand of young red alder and willow established after disturbance by a major flood in February 1996. A large, channel-spanning, Douglas-fir log fell across the creek after the 1996 flood.

Management history

When considering influences of land management on a stream site, it is important to note possible effects of both local and watershed-wide activities. The main Lookout Creek road on the south side of the creek was constructed in ca. 1950 at varying distance from the stream—at this point the road is perhaps 80m away from the stream. Clearcut logging occurred in about 1960 in the area now occupied by the headquarters facilities (you can see some of the young forest plantation that grew up in the clearcut around the margins of the present opening). A small, low-quality road was constructed up Lookout Creek drainage on the Headquarters side of the valley to support the logging operations there.

Upstream of this site the watershed has been about 25% clearcut and has a rather extensive road network, nearly all put in during the 1950s and 1960s. Some evidence of this is seen in the odd chunk of asphalt and piece of culvert in the channel and on the bars.

Research context

This reach of stream has been the site of several important studies.

This is a site in the River Continuum project during the late 1970s when stream ecologists examine stream food webs in small, medium (this site), and large (the McKenzie River) channels. The crux of the River Continuum Concept (Vannote et al., 1980) is that as you move downstream from forested headwaters to the open mainstem of a river like the Willamette, the food base of the stream shifts in some predictable ways. The invertebrate communities of small streams, for example, are geared to consume forest litter and the shade of the canopy limits light that can warm the water and drive primary production in the stream. Larger channels have less forest influence and more food resources are produced by plants within the channel. This has been the fundamental concept describing how river systems work ecologically for about 25 years, and it has roots here and in three other study sites in the United States (ID, MI, PA).

At the same time that the River Continuum Concept was brewing, Andrews Forest scientists began long-term monitoring of stream channel change at various sites along the stream. Here we have a series of surveyed channel cross section (marked by fence posts and flagging on opposite sides of the channel spaced about every 30m) and we have repeated maps of the site (Faustini 2000, Swanson and Jones 2001). Repeated observations are used to track changes in sediment storage, configuration of pieces of wood, and channel habitat in this system that changes little or slowly for years and then abruptly and dramatically in floods with multi-decade recurrence intervals. During that flood we saw that debris flows occurred early in the event; log movement took place over 12 hours of so on high water of the rising limb of the floodflow; boulder movement, detectable by the deep thunking of rocks bouncing down the streambed under coffee-colored water, was audible for 48 hours or so; and suspended sediment of latte appearance persisted for quite a few days. Remarkably, many components of stream and riparian biota survived in good shape or even improved conditions, although some taxa declined dramatically, such as sculpin, small fish that reside among the boulders that moved during the flood (Swanson et al., 1998).

This site also presents a fine example of the research theme concerning forest-stream interactions—an integral part of the Andrews Forest program for three decades. Here you can see the importance of the streamside forest as sources of fine and large organic matter inputs to stream that affect food resources and habitat structure and as shade, affecting light energy available to warm water and provide energy for primary production. Another forest-stream interaction concerns the role of the hyporheic zone—the groundwater system derived from surface waters that have the opportunity to interact with riparian vegetation, including the nitrogen-fixing red alder stands common on gravel bars.

The background of knowledge of this site made it a natural spot for a stream nutrient addition experiment carried out in the late 1980s by the Oregon State University Stream Team led by Stan Gregory. Nitrogen was added to the stream during the summer and its fate in the food chain and movement downstream were tracked and compared with a "control" reach of stream above the site of nutrient addition. The basic finding was that nitrogen was taken up quickly and made its way through the food chain. The stream did not fill with algae because grazing invertebrates keep the primary producers in check.

These ecology and geomorphology studies provided great background information when the February 1996 flood hit the region. Witnessing the flood firsthand (see video taken by Gordon Grant) made it possible to sense the anatomy of a flood—to develop a watershed-wide view of sediment, wood, and boulder movement over the course of a flood. The knowledge of the history of the area, especially the 1964-1965 winter flood of similar extreme magnitude, made it possible to assess land use legacies and legacies of the earlier floods. It was very exciting to be out in the flood—as Gordon put it, "decades of boredom punctuated by moments of chaos." Essential elements of a long-term view include both the infrequent bang and the slow-change processes.

Ripple effects of learning represented by this site

This site represents several significant advances in thinking about stream and riparian ecosystems and about disturbances. The River Continuum Concept, dynamics of wood in streams of this size, and land use legacies sensed by major floods are important conceptual advances with significant roots in this study site.

Literature cited

Faustini JM. 2000. Stream channel response to peak flows in a fifth-order mountain watershed. Corvallis,OR: Oregon State University. 339 p. PhD dissertation.

Swanson FJ, Johnson SL, Gregory SV, Acker SA. 1998. Flood disturbance in a forested mountain landscape. BioScience. 48(9): 681-689.

Swanson FJ, Jones JA. 2001. Geomorphology and hydrology of the H.J. Andrews Experimental Forest, Blue River, Oregon. In: Moore GW, ed. Field guide to geological processes in Cascadia: field trips to accompany the 98th annual meeting of the Cordilleran Section, Geological Society of America: Corvallis, OR. Special Paper 36. Oregon Department of Geology and Mineral Industries. pp. 289-314.

Vannote RL, Minshall GW, Cummins KW, others. 1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences. 37(1):130-137.



(Sale Unit 5b on Blue Ridge)

Biophysical setting

This site on a northwest aspect is near the drainage divide between Lookout Creek (which equals the Andrews Forest behind us) and the upper Blue River drainage in our view to the north. Blue River is flowing from right to left (southwest). The forest in the area is a mixture of a low-density of old-growth (400-500 yrs old Douglas-fir) and mature (ca. 140-year-old) forest containing a mixture of Doug-fir, western redcedar, western white pine, and western hemlock. These two age classes appear to reflect regeneration after at least two fires in the respective time-frames. Standing at the landing first encountered along the 1508 road and looking down into the cutting unit, we see a small, headwater stream draining from left to right across the broad bench within the cutting unit.

Management history and context

The management story of this site can be interpreted at several scales—patterns within the cutting unit, relation of the cutting unit with adjacent areas, the larger landscape context of the Blue River drainage (and landscape management plan), and finally the entire area of the Northwest Forest Plan (24,000,000 acres of Federal lands in the range of the northern spotted owl extending from northern California to the Canadian border).

This site was partially cut in 2002, broadcast burned in spring 2003, and planted with several species of tree seedlings in 2004. Based on the Blue River Landscape Plan, described below and in Cissel et al. (1999), there was an attempt to leave about 45% of canopy cover in varied densities across the cut area and then to burn the site with the intent of reducing slash, maintaining roles of fire in the landscape, and killing about a third of the live trees. The objective of killing trees with the prescribed fire was to create snags in a snag-depauperate landscape (i.e., past clearcutting eliminated snag habitat for cavity-dwelling organisms and others).

The pattern of logging disturbance in the cut area reflects both the objective of having lower levels of disturbance close to the stream and riparian zone and the effects of the logging engineering system used. Logs were moved with a skyline cable system that picked up most of the log and moved it along a "yarding corridor" to the landings, where logs were collected adjacent to the road for placement on log trucks for transport to the mill. The most severe site disturbance occurred in the landing areas where all trees had been cut for safety of loggers, where yarding corridors converged, and where lots of wood accumulated (raked toward the landing during yarding and left from delimbing at the landing) and was then burned. Further away from the landings and toward the stream disturbance to the forest vegetation and soil was much less.

The cutting prescription for this site (180 yr rotation and 30% live tree retention) was set by the Blue River Landscape Plan. In other words, thinking about how the landscape works was used to direct cutting at this local level (in the old days the management focus was on management of individual sites and the landscape pattern emerged from site-level decisions). The Blue River Landscape Plan is based in part on interpretation of the fire history of the area by Pete Weisberg as part of his PhD. This specific site was selected for cutting based in part on the intent of minimizing further fragmentation of forest pattern over the landscape where some cutting is part of the management plan and policy. The local area has been greatly affected by the past dispersed patch clearcutting, and the cut area is a "leave strip" surrounded by plantations created by earlier clearcutting (note the adjacent plantations). The plantation to the right was pre-commercially thinned in ca. 2000—i.e., trees were cut to reduce tree density and the fallen saplings were left in place, except near roads where the dead trees were pulled back to reduce fuels close to the road.

This site and the Blue River drainage are parts of the Central Cascades Adaptive Management Area (AMA), one of 10 AMAs established by the Northwest Forest Plan in 1994. Work in AMAs is intended to be carried out in an adaptive management framework, meaning that the practices are designed to learn about how natural resource systems respond to management practices (adaptive management is an approach to learning first put forth over three decades ago by scientists in British Columbia and now adopted in various situations, such as the Everglades and Colorado River restoration programs). The concept is to accomplish this learning by carrying out management practices, monitoring their effects, and adjusting plans for further actions based on lessons from monitoring, operational experience, new research, and other sources. In some cases computer simulation modeling is used to depict possible outcomes of alternative management actions. Modeling results can be used to refine hypotheses and to design experimental treatments.

When we think about the Northwest Forest Plan and especially its AMA component, it is important to recognize the hierarchical framework within which each component is intended to operate. Each AMA, for example, is located in the Northwest Forest Plan regional context that includes about 80% of lands designated as reserves of various types, which are intended to provide terrestrial and aquatic habitat for many species, especially those associated with old-growth habitat. Adjacent, large Late Successional Reserves are intended to provide a buffer in the case that innovative practices within the AMAs fail for some reason.

One of the specific charges to the Central Cascades AMA is to develop innovative approaches to landscape management based in part on understanding the historic disturbance regime of the area, while carrying out some forest cutting. The Blue River Landscape Plan is a response to that charge. This work involves deep temporal perspective, both retrospectively and forward. We look back at fire history for the past 500-800 years accessible by analysis of tree rings to date past fires and then project management into the future with computer models. The futuring extends out quite a few centuries because the management scheme uses cutting rotations of up to 260 years. To "see" the future in a simulation context requires looking at least this far forward. As limnologist John Magnuson (1990) points out, the present is invisible if we don't have historical context.

Research context

Research activities relevant to conditions at this site are underway on several scales. Some studies have examined effects of partial cutting, such as response of canopy lichen communities and birds to different densities of trees left after cutting. Studies at the landscape scale concern effects of different management approaches (paradigms?) on landscape patterns and the consequences of those landscape conditions in terms of wood production (if that is part of the scheme), carbon storage, hydrology, and biodiversity. Another study theme concerns the process of learning in the course of exploring large-scale management of natural resources, like forests and watersheds. In this regard, we are learning about the adaptive management process through the practical experience of trying to do it as an AMA.

Concerning the research effort to explore different approaches to management, we can see different primary objectives and concepts about achieving them at play in the western Oregon landscape (and, indeed, globally). The wood production objective starts with intensive plantation forestry and adds measures for environmental protection (e.g., streamside buffers and snag prescriptions) as needed. The species conservation objective uses tools of conservation biology (e.g., reserve and matrix prescriptions) for one or a few species to design landscape patterns. A third approach begins with consideration of landscape dynamics and uses historical disturbance regimes to plan the timing and intensity of future patterns of cutting. This has been variously termed the historical (or natural) range of variability (HRV) or ecosystem dynamics approach. The Blue River Landscape Plan and study are intended to explore this approach on the ground and to compare it with other approaches through simulation modeling.

Ripple effects of learning represented by this site

What can the future of forestry in this area be? This broad question is addressed repeatedly through discussions on visits to this site and a few others we have used since the late 1980s as questions about the future of federal forestry have been hashed out heatedly and publicly. We do not have a term for this continuing dialog with the public—perhaps the closest term is "technology transfer," but we are talking about concepts and not technology and the information transfer is not a one-way street, but rather a continuing development of ideas shared by the research and management communities.

This question of the future of forestry can be fruitfully discussed in our unique context—a research-management partnership centered on the Andrews Forest and involving the Willamette National Forest and researchers of the Forest Service's Pacific Northwest Research Station and Oregon State University. Our unique role involves 1) a close, long-term partnership that can lead to new approaches to management that combine current management credibility and science credibility; 2) on-the-ground examples of the ideas in practice that can be used for demonstration and discussion; and 3) a research program that has basic, National Science Foundation-sponsored science at its core, so we can ask fundamental questions and explore new territory without concern for effect on the management paradigm of the day.

An important accomplishment of the partnership is the process (described above) for doing this work. The process is a product of community evolution. Visitors from elsewhere come to visit to learn about this process, which boils down to the conduct of effective, long-term working relationships.

Work on landscape management in the context of the Blue River Landscape Plan and associated efforts have influenced development of some different ideas. The notion of minimizing fragmentation ("min-frag" management) in management of patch clearcut landscapes gained attention in the late 1980s, leading to language in a Congressional appropriations bill directing the Forest Service to use that approach. The Blue River Plan has become a prime example of the HRV concept applied on the ground and, therefore, a focal point for discussion of the concept. The stand-scale practices represented here have helped stimulate more adventuresome practices elsewhere as federal foresters explore departures from the old scheme of clearcut and broadcast burning.

These ideas ripple out into the larger world via lots of field tour discussions with all kinds of folks (e.g., land owners and managers, media, elected officials, students, educators, interest group representatives), publications directed to non-scientist users of the information, and many other media.

Literature cited

Cissel JH, Swanson FJ, Weisberg PJ. 1999. Landscape management using historical fire regimes: Blue River, Oregon. Ecological Applications. 9(4):1217-1231.

Magnuson J. 1990. The invisible present. BioScience. 40(7):495-501.