An objective classification of largewood in streams

Abstract

Traditional, stage-based, classification systems provide a qualitative measure of decay and have been widely used to monitor and model terrestrial coarsewood and aquatic largewood dynamics. These systems are limited by subjective assignment of wood to classes, lack of measurements relating wood morphology with decay classes, and poor estimates of elapsed time-since-death within and between decay classes. To overcome these limitations, we used quantitative methods to develop a new classification system for in-stream largewood based on morphological attributes that are (1) easily measured in the field and (2) relate to the time-since-death of individual logs. Using principal components and cluster analyses, we developed a three-class system for largewood based on log length, branch order, and cover (quartile classes) of bark, vegetation, and soil. Thresholds for each attribute provided criteria that we organized in a dichotomous key that can be used to objectively and consistently assign individual pieces of largewood into mutually exclusive classes. Using time-since-death determined using dendrochronology, we verified that successive classes in our model-based decay classification represent progressive largewood decomposition through time. This model-based classification system provides an improved framework for developing management plans and modeling dynamics of in-stream largewood.