A quantitative framework for evaluating the mass balance of in-stream organic debris

A quantitative framework is developed for analyzing the mass budget of in-stream woody debris. Wood budgets are necessary for defining the relative importance of different recruitment processes over short and long periods, for designing spatially explicit simulation models, and for estimating the range of variability. The framework is used to analyze century-long patterns of large woody debris in streams that are governed by episodic forest death (fire and wind), forest growth and chronic mortality, bank erosion, mass wasting, decay, and stream transport. Simplified mathematical expressions are used to represent climatic, hydrologic, geomorphic, and biotic processes. Results are expressed in terms of time series and probability distributions. Predictions include that in areas of longer fire rotation (500 years) toppling of fire-killed trees comprises only 15% of the longterm wood budget yet chronic stand mortality that affect the large standing forest biomass ensures continuous large volumes of wood in streams. In contrast, toppling of fire-killed trees in forest environments with shorter fire rotations (150 years) comprise about 50% of the wood budget and indicates that field observers have a significantly higher chance of encountering low wood volumes in streams. Wood recruitment by bank erosion should increase irregularly downstream and bank erosion recruitment should exceed mortality recruitment at a bank erosion rate of approximately 5 cm per year. Recruitment from debris flows represents the single largest point source of woody debris to streams. The rarity of debris flows, in conjunction with a 3% per year annual decay rate, limits the contribution of wood from debris flows to about 12% of the long-term wood budget. Fluvial transport of wood promotes an increase in both inter-Jam spacing and jam volume downstream. The proportion of woody debris transported into a reach in comparison to lateral recruitment approaches an asymptotic maximum of 50% when tree height approaches channel width. The relationships among process rates, their spatial variance across landscapes, and the resulting probability distributions of long-term patterns of wood abundance are proposed as a set of general theoretical principles. New data on wood supply and storage at the network scale are needed to fully test the predictions made in this analysis. (C) 2002 Elsevier Science B.V. All rights reserved.