HJA newt logo HJ Andrews
Login Donate
visit HJA Andrews on Twitter visit HJA Andrews on Facebook HJA RSS news feed

  Home > Long-Term Research Categories > Stream Ecology > Stream-Forest Interactions Summary
Long-Term Dynamics of Large Wood

Dr. Stan Gregory, Linda Ashkenas, Randall Wildman, Mark Meleason
Department of Fisheries & Wildlife, Oregon State University

Last Revised: March 7, 2002

Research Findings

Many studies have identified the characteristics of wood stored in streams (Harmon et al. 1986), but few have attempted to measure its long-term dynamics. From 1982-1985, the Long-Term Ecological Research (LTER) Program at the H.J. Andrews Experimental Forest developed a long-term study of input, storage, decomposition, and redistribution of large wood in Mack Creek (Photo 1). From 1985 to the present, we have annually surveyed a 1.1 km section of this stream (Photo2).

Mack Creek is a third-order stream flowing through a 500-year-old coniferous forest dominated by Douglas fir, western hemlock, and western red cedar. Mack Creek contains large volumes of large wood, typical of streams in old-growth forests of North America (Harmon et al. 1986). Our study section is divided into an upstream 670-m reach in an old-growth forest and a downstream 360-m reach in a 30 year-old clearcut. Within these reaches, all pieces of large wood (> 1 m long, > 10 cm diameter) are tagged and inventoried annually. Data collected include dimensions, location, geomorphic position, and decay. Standing stock, geomorphic position, input rates, and movement rates are determined from annual change measured during inventories each fall.

Density of large wood in the old-growth reach in over the past 5 years is approximately 280 pieces per 100 m (Figure 1), or 0.24 pieces/m2, and average volumes are 0.08 m3/m2. Densities and volumes in the clear-cut are significantly lower. These values include all pieces of wood touching the floodplain or active channel. Of the 303 m3/100 m of channel, 81 m3/100 m was within the active channel. Most of the number of pieces of wood was either partly or entirely in contact with the active channel, and approximately one-fifth of the wood was found on the floodplain.

Wood in old-growth streams exhibit several characteristics that reflect their long residence time and important ecosystem functions. More than 65% of the pieces of wood are less than 5 m in length, but pieces may be as long as 45 m (Figure 2). These larger pieces of wood create important geomorphic structures that serve a "backbones" for major wood accumulations.
Rootwads make wood more resistant to transport, but even in an old-growth stream, the vast majority of pieces of wood do not have rootwads. Only 6% of the pieces in Mack Creek had a rootwad connected. Most of the length of wood was located in the active channel, but the highest proportion of volume was found adjacent to the active channel on the floodplain or hillslope. Most of the wood is aggregated and does not function as single pieces. Almost 80% of the logs are in an accumulation of three pieces or more; in the old growth reach, 29% of the pieces are in full-channel spanning jams. The major mechanism stabilizing wood pieces against movement is blockage or support by other logs, although large boulders and streamside trees can also contribute to stabilization. The long residence time of wood in the channel was reflected in the state of decomposition of the wood. Most of the pieces were in decay class 3, an intermediate stage of decay. No logs were found in decay class 5, fragments of well-decayed wood, because flows within the active channel and across the floodplain destroy these well-rotted pieces.

The long-term nature of the study has allowed us to measure rates of input, fragmentation and movement. One of the least available measures of wood dynamics in streams is input rate. Many studies have documented storage of wood, but long-term studies are essential to document rates and variance of input processes. Input rates in Mack Creek were highly variable and more closely related to climatic events (windstorms, snowfall) than discharge events (Figure 3). In 1995-96, an average of almost 60 pieces of wood/100 m entered the channel, but most of this input was not related to the flood in February. A wet snowpack caused large numbers of trees to fall over throughout the H.J. Andrews in December 1995, leading to high rates of loading.

Floods are the primary disturbance event in stream ecosystems and are the most probable cause of redistribution of wood. Over the period of record from 1985-2001, less than 1% of the logs in Mack Creek moved in most years (Figure 4). DA major flood occurred in February 1996, with inundation depths of more than 1.5 m above the active channel (Figure 5).During this flood, 11% of the pieces moved more than 10 m and 89% remained in their original positions or moved only slightly. In the clearcut reach, more than 15% of the pieces moved. Probability of movement during high flows was significantly correlated to length of log, stability, and geomorphic location (see also Meleason 2001). All pieces of wood that moved during the period of record were less than the width of the active channel, and most pieces that moved more than 300 m were less than 2 m in length (Figure 6). Wood dimensions, channel morphology, and hydrologic regimes play important roles in understanding dynamics of large wood in river networks, and are critical elements of effective riparian management practices (Gregory and Ashkenas 1990, Gregory 1997). While most studies of wood in streams document amounts and sizes of wood in storage in streams and riparian areas, long-term measurements- mortality, input, breakage, storage, decay, and transport-are required to understand the processes that create and maintain patterns of wood in river networks.

Relevant Literature:

Harmon, M.E., J.F. Franklin, F.J. Swanson, P. Sollins, S.V. Gregory, J.D. Lattin, N.H. Anderson, S.P. Cline, N.G. Aumen, J.R. Sedell, G.W. Lienkaemper, K. Cromack, Jr., and K.W. Cummins.  1986.  Ecology of coarse woody debris in temperate ecosystems.  Advances in Ecological Research 15: 133-302.

Gregory, Stan; Ashkenas, Linda. 1990. Riparian management guide, Willamette National Forest. U.S. Department of Agriculture, Forest Service, Willamette National Forest. 120 p.

Gregory, Stanley V. 1997. Riparian management in the 21st century. In: Kohm, Kathryn A.; Franklin, Jerry F., eds. Creating a forestry for the 21st century: the science of ecosystem management. Washington, DC; Covelo, CA: Island Press: 69-85.

Gregory, Stanley V.; Swanson, Frederick J.; McKee, W. Arthur; Cummins, Kenneth W. 1991. An ecosystem perspective of riparian zones. BioScience. 41(8): 540-551.

Lienkaemper, G. W.; Swanson, F. J. 1987. Dynamics of large woody debris in streams in old-growth Douglas-fir forests. Canadian Journal of Forest Research. 17: 150-156.

Lienkaemper, George W.; Swanson, Frederick J. 1980. Changes in large organic debris in forested streams, western Oregon. In: Abstracts with programs: Cordilleran section of the Geological Society of America, 76th annual meeting; 1980 March 19-21; Corvallis, OR. 12(3). Boulder, CO: Geological Society of America: 116.

Swanson, Frederick J.; Franklin, Jerry F.; Sedell, James R. 1990. Landscape patterns, disturbance, and management in the Pacific Northwest, USA. In: Zonneveld, I. S.; Forman, R. T. T., eds. Trends in landscape ecology. New York: Springer-Verlag: 191-213.

Swanson, Frederick J.; Johnson, Sherri L.; Gregory, Stanley V.; Acker, Steven A. 1998. Flood disturbance in a forested mountain landscape. BioScience. 48(9): 681-689.