CWD is also important for carbon storage (see carbon cycling) and as a major factor in the susceptrbility of to fire (see Forest fire mod- Coarse woody debris (CWD), synonymous with large els). Carbon dioxide released from storage in CWD woody debris, is a term used to describe from by or burning may be a substantial dead . Although an exact definition is not avail- contributor to global warming. death and sub- able, many researchers include all pieces of wood that sequent decomposition also are important processes are at least 10 cm in diameter and 1 m in length. CWD for the cycling of nutrients in forests (see Ecosystem can consist of any coarse part of a dead tree, such as element cycling). trunks or .parts of trunks, branches, stumps, and root Challenges to the measurement and quantification wads. A distinction is often made between standing of CWD range from sampling issues to multivariate dead trees, snags, and dead trees on the ground, logs. statistics for studying occurrences and com- Dead roots in the soil are generally not considered as munities on dead wood. A brief overview of some CWD although there is no principal reason why they specific environmetrical challenges follows. should not be. Researchers often distinguish between CWD and fine woody debris (FWD), where the latter consists Sampling Methods for CWD of the parts of a dead tree smaller than CWD, exclu- CWD can be assessed using many different tech- sive of needles and leaves. The diameter threshold niques. Remote sensing may be used, but will gen- value distinguishing CWD from FWD varies between erally result in coarse estimates. With most kinds of different studies. satellite data the inferences will be indirect, linking Interest in CWD has increased dramatically dur- identifiable features of a forest stand with the normal ing the recent decades, primarily because of increased amounts of CWD present in such stands. With aerial understanding of the important role CWD plays in photograph interpretation, direct measurements of the forest and aquatic ecosystems. CWD provides essen- amount of standing and fallen dead trees (such as is tial for a wide range of terrestrial and aquatic used to assess fuel loading for forest fire risk) can species. Fallen trees provide habitat for a variety of often be made. However, to obtain accurate estimates small vertebrates including rodents, snakes and sala- manders, and many species of , , fungi of CWD, field assessment is generally required. Some important probability sampling methods for assess- and depend on dead wood for their survival. ing CWD are plot based sampling, strip surveying, Fallen trees in moist environments also serve as nurse and line intercept sampling (see Sampling, environ- logs for seedling establishment. In streams, CWD mental). The latter method is applicable only to affects channel morphology by deflecting flow and downed logs. Data from ground-based surveys often encouraging patterns of scour and fill. Debris dams is expressed as volume (e.g. cubic meters) or mass slow the flow of water, dissipate energy, trap sed- per unit area, typically kilograms per square meter iment and , and create microhabitats for and benthic macroinvertebrates. kw-2). Both on land and in water, habitat simplification resulting from timber harvesting and the subsequent Studies of Decomposition Rates of CWD decreases in CWD input has been linked to long-term changes in species and community composition. It is The rate of decomposition varies with factors such widely acknowledged that practices need to as tree species, dimension, resin content, and macro- be modified to reestablish suitable habitat conditions and microclimatic conditions. Two major approaches for species depending on CWD. to modeling decomposition can be distinguished. For an individual species, the value of CWD Firstly, regression techniques are often used to deter- depends on factors such as tree species, size, degree mine and predict the decrease in dry weight over of decomposition, as well as site and local micro- time. In this context, the exponential model is com- climatic conditions. Classification systems have been monly used. Secondly, probabilities of transition of developed to categorize CWD by the degree of CWD units between different decomposition classes decomposition. can be estimated. Thus, matrix models can be applied 2 Coarse woody debris

to study the amount of CWD present in different Managemeni: Foreshy and Fishery Interactions, E.O. decay classes over time. Tree mortality functions, Salo & T.W. Cundy, eds, University of Washington, commonly derived using logistic regression, may Seattle, pp. 143-190. be used to provide the input of new CWD to the Harmon, M.E., Franklin, J.F., Swanson, F.J., Sollins, P., Gregory, S.V., Lattin, J.D., Anderson, N.H., Cl&e, models. S.P., Aumen, N.G., &dell, J.R, Lienkaemper, G.W., Cromack, K. & Cummins, K.W. (1986). Ecology of coarse woody debris in temperate ecosystems, Advances Species Occurrences and Communities on in Ecological Research 15, 133-302. CWD Jonsson, B.G. 62 ms, N., eds (2001). Ecology of woody debris in boreal forests, Ecological Bulletirn 49. To quantify and predict CWD substrate quality for Maser, C. & &dell, J. (1994). From the Forest to the Sea: different species, methods such as logistic regres- the EcoIogy of Wood in Streams, Rivers, Estuaries. and sion are commonly used. Thus, estimates of species Oceans, St Lucie Press, Delay Beach. preferences between different kinds of CWD can be Maser, C. & Trappe, J.M., eds (1984). The Seen and Unseen World of the Fallen Tree. US Department of Agricul- obtained. With multivariate statistical methods (e.g. ture, Forest Service General Technical Report PNW-164. ordination) patterns of species communities can be Pacific Northwest Forest and Range Experiment Station, mapped versus substrate quality and other relevant Portland. factors. Samuelsson, J., Gus&son, L. & Inge%g, T. (1994). Dying and Dead Tree. A Review of Their Importance for Biodi- wrsity, SLU, Uppsala. Further Reading

Bisson, P.A., Bilby, RE., Bryant, M.D., Dolloff, C.A., Grette, (See abo ; Tree morphology) G.B., House, RA., Murphy, M.L., Koski, K.V. & Sedell, J.R. (1987). in forested streams in the pacific northwest: past, present, and futme, in Streamside G&AN STAHL & C. ANDREW DOLLOFF Coarse woody debris

Gijran Stahl & C. Andrew Dolloff Volume 1, pp 361-363

in

Encyclopedia of Environmetrics (ISBN 0471 899976) Edited by

Abdel H. El-Shaarawi and Walter W. Piegorsch

0 John Wiley & Sons, Ltd, Chichester, 2002