Home Archived January 13, 2017



Timothy H. Diehl and Bradley A. Bryan

Cite this report as: Diehl, T.H., and Bryan, B.A., 1993, Supply of large woody debris in a stream channel, in Shen, H.W., Su, S.T., and Wen, Feng, eds., Hydraulic engineering '93 conference, San Francisco, 1993, Proceedings: American Society of Civil Engineers, v. 1, p. 1055-1060.


The amount of large woody debris that potentially could be transported to bridge sites was assessed in the basin of the West Harpeth River in Tennessee in the fall of 1992. The assessment was based on inspections of study sites at 12 bridges and examination of channel reaches between bridges. It involved estimating the amount of woody material at least 1.5 meters long, stored in the channel, and not rooted in soil. Study of multiple sites allowed estimation of the amount, characteristics, and sources of debris stored in the channel, and identification of geomorphic features of the channel associated with debris production. Woody debris is plentiful in the channel network, and much of the debris could be transported by a large flood. Tree trunks with attached root masses are the dominant large debris type. Death of these trees is primarily the result of bank erosion. Bank instability seems to be the basin characteristic most useful in identifying basins with a high potential for abundant production of debris.


Accumulations of large woody debris at bridges have caused increased backwater and bank scour at bridges in the Harpeth River basin in middle Tennessee. Tree trunks with attached root masses play a key structural role in debris accumulations at these bridges. Previous studies of large woody debris at bridges have demonstrated that debris accumulations on bridges can be dangerous and expensive to deal with (Chang, 1973; Chang and Shen, 1979; O'Donnell, 1973). In the fall of 1992, the U.S. Geological Survey, in cooperation with the Tennessee Department of Transportation, assessed the potential supply of large woody debris in the basin of the West Harpeth River upstream from bridges at which debris has accumulated. This assessment involved estimating the amount of woody material at least 1.5 meters long, located in the channel, and not rooted in soil.

The part of the West Harpeth River basin included in the assessment has a drainage area of 267 square kilometers. Land use in the basin is mostly agricultural, with row cropping limited mostly to the valley bottoms and with forest present on most hillsides. Trees are allowed to grow to maturity on nearly all stream banks. The total relief in the basin is 195 meters.

Within the basin, channel reaches of the West Harpeth River and its two main tributaries, Leipers Fork and Murfrees Fork, were included in the debris assessment. The total length of these reaches is about 51 kilometers. Channel slopes in the reaches studied range from 0.0009 to 0.0044.


Twelve sites near bridges in the study reaches were investigated in detail. The drainage area at these sites ranged from 13 to 267 square kilometers, and the channel width ranged from 12 to 27 meters. These sites ranged in length from 100 to 290 meters, and included a total of 2 kilometers of channel, or about 4 percent of the total channel length of the reaches included in the study. Pieces of debris at least 1.5 meters long were measured at these sites. The debris concentration, defined as cubic meters of debris per kilometer of channel, was calculated at each bridge site. Characteristics and sources of woody debris, and geomorphic features of the channel associated with debris production, were identified at these sites.

Pieces of long debris (12 meters or more in length) were counted and measured in reaches of the Leipers Fork and the West Harpeth main stem totaling about 11 kilometers in length. These measurements provided additional information on the characteristics and distribution of long debris, of which only 12 pieces were found at the bridge sites. These measurements were also used to calculate the error in the estimated concentration of long debris based on the observed concentration at the bridge sites. The average number of long debris pieces per kilometer of channel was estimated by dividing the number of long pieces measured at the 12 sites by the total length of the sites. The total number of long pieces in the reaches that were not inspected was estimated by multiplying the average number of long debris pieces per kilometer of channel by the total length of these reaches.

Debris volume was estimated for the channel reaches included in the study. The volume of debris in each study reach was estimated by multiplying the length of that reach by the average debris concentration at study sites in the reach or at one end of the reach. The total estimated volume of debris in the study area was the sum of the estimated volume for all study reaches.

Site inspection process

The 12 site inspections were conducted in 100 to 290 meters of channel. These lengths were judged to be sufficient to estimate the condition of the adjacent river reach. During the inspection, debris size, condition, orientation to flow and the channel, and interaction with the channel and other debris were described (table 1). Live trees apparently in the process of becoming debris, but not yet dislodged from the bank, were not included in the investigation.

Table 1. Data noted for each piece of debris

Debris dimensions were measured with tape, calipers, and incremented range pole. Piece length was measured from end of root mass or spar butt to spar top. Pieces less than 1.5 meters long were not recorded. Diameter was measured above the root flare or any abnormally large butt swelling, and at the spar top.

Some characteristics reflect debris history and potential to accumulate. The presence or absence of branches is an indicator of length of time in the channel and the destructive forces exerted on the piece. Bark condition reflects abrasion and extent of decomposition, and may indicate how long the piece has been in the channel. Curved pieces of debris seem to be more likely to form intertwined jams.

Root-mass condition is an indicator of the origin of a piece of debris and its exposure to abrasion and fluvial erosion. A root mass containing soil indicates relatively recent introduction into the channel. A one-sided root mass probably developed on the channel bank over the lifetime of the tree. A symmetrical root mass indicates that the piece may have originally grown on or near the bank top and was undercut during channel migration or widening. Extensive wear or smoothing of the root mass indicates extensive transport within the channel. Sawn or chewed ends were documented because bank clearing and beaver activity are anecdotally cited as major sources of large woody debris.

Piece orientation relative to channel alignment, position in the channel, and type of anchorage indicate potential for transport and probable mode of transport (fully floating or dragging). Isolated pieces aligned with the flow direction, with the root mass upstream, are likely to be transported when the water rises high enough to float them. Debris anchored on stable channel features seems more likely to remain in place, particularly if the debris is aligned perpendicular to the direction of flow.


Woody debris is plentiful in the channel network. The average debris concentration at the 12 study sites was 27 cubic meters per kilometer, with a standard deviation of 26 cubic meters per kilometer. A total volume of about 1,600 cubic meters was estimated to be present in the channel reaches studied. The amount of stored debris is large compared to streams of similar size in mountainous areas (Harmon and others, 1986), but smaller than that found in low-gradient coastal-plain streams (Benke and Wallace, 1990).

Debris jams contained 23 percent of the total debris volume. Most of the rest of the debris occurred along short reaches of relatively unstable channel. The largest debris jam measured contained 4.6 cubic meters of debris in pieces at least 1.5 meters long. Two other jams observed during reconnaissance appeared to be at least twice as large. None of the large debris jams was anchored on a bridge.

Based on long debris measured at the 12 bridge sites, about 300 long pieces were estimated to be present in the channel reaches studied. The estimated number of long pieces in the 11 kilometers of inspected reaches, based on data from the 12 bridge sites, was 64. The actual number of long pieces measured in these reaches was 32. Of the total volume of debris, 21 percent was at least 12 meters long, and 12 percent was at least 15 meters long (fig. 1).

Figure 1. Graph showing relation between debris length and percent of total debris volume.
Figure 1. Relation between debris length and percent of total debris volume.

Tree trunks with attached root masses were the dominant type of long debris, and limbs and trunks separated from the stump by breaking or cutting were the dominant type of shorter debris. Of the 44 long debris pieces measured, 33 had root masses. Of the 190 measured pieces less than 12 meters long, only 62 had root masses. Only 12 of 234 measured pieces had saw cuts, and none had been felled by beavers.

The major source of debris in the study basin is trees that are undermined by fluvial erosion, fall into the stream while still alive, and finally become detached along with their basal root mass. This finding is based primarily on the prevalence of asymmetric root masses. Also, many living trees were observed to be nearly detached from the bank by erosion, and most of these trees had distinctly asymmetrical root masses. The presence of bark on 29 of the 44 long pieces measured suggests that most trees were living when they were undermined. The dominance of this debris source is consistent with previous studies (Harmon and others, 1986; Keller and Swanson, 1979; McFadden and Stallion, 1976).

Thus, debris production is associated with bank instability. Stable banks can retain trees even when the trees lean at pronounced angles and their root masses are mostly exposed. Bank instability is rarely localized at individual trees, and more often occurs throughout stretches of widening channel and along the outside bank of bends where the channel is migrating laterally.


Evaluation of debris potential based on inspection of a single site probably would have yielded a poor estimate of availability and potential for transport of debris. Debris concentration is highly variable among sites, and the likelihood of selecting a site with approximately average debris concentration would have been small.

Bank instability seems to be the channel characteristic most useful in identifying channel reaches with high potential for production of large woody debris. Lateral channel migration and widening can be detected on maps and aerial photographs. High and steep banks, erodible bank materials, and a history of channel widening or lateral migration all are useful as indicators of potential bank erosion and consequent debris production.

Appendix. - References

Benke, A.C., and Wallace, J.B., 1990, Wood dynamics in coastal plain blackwater streams: Canadian Journal of Fisheries and Aquatic Sciences, v. 47, no. 1, p. 92-99.

Chang, F.F.M., 1973, A statistical summary of the cause and cost of bridge failures: Federal Highway Administration, Office of Research and Development, Report No. FHWA-RD-75-87, 42 p.

Chang, F.F.M., and Shen, H.W., 1979, Debris problems in the river environment: Federal Highway Administration Report No. FHWA-RD-79-62, 67 p.

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

Keller, E.A., and Swanson, F.J., 1979, Effects of large organic material on channel form and fluvial processes: Earth Surface Processes, v. 4, p. 361-380.

McFadden, Terry, and Stallion, Michael, 1976, Debris of the Chena River: Hanover, U.S. Army Corps of Engineers Cold Regions Research and Engineering Laboratory Report 76-26, 17 p.

O'Donnell, C.L., 1973, Observation on the causes of bridge damage in Pennsylvania and New York due to hurricane Agnes: Highway Research Record, no. 479, p. 20-36.

Return to Publications

U.S. Department of the Interior, U.S. Geological Survey
Last update: Tuesday, 08-Jan-2013 14:46:09 EST
URL: http://tn.water.usgs.gov /publications/abstracts/debpoar2.html
USGS Privacy Policy || Disclaimer