Wood in rivers: interactions with channel morphology and processes

No doubt about it, wood complicates fluvial geomorphology. It messes up nice tidy streams, complicates quantitative analysis, invalidates convenient assumptions, and opens new questions about how different contemporary channels are from their pristine state. It is no coincidence that modern fluvial geomorphology developed through the study of channels lacking a substantial load of wood debris (Leopold et al., 1964) and there is little mystery as to why geomorphologists sought to study rivers or river reaches where wood does not exert a significant influence on channel morphology and processes. However, over the past several decades, recognition has grown that wood debris significantly and sometimes systematically affects channel processes in forested regions across a wide range of scales from channel roughness and bedsurface grain size (Lisle, 1995; Shields and Gippel, 1995; Buffington and Montgomery, 1999; Manga and Kirchner, 2000); to creation of in-channel features, such as pools and steps (Keller and Tally, 1979; Nakamura and Swanson, 1993; Lisle, 1995; Montgomery et al., 1995, 1996); even to large-scale controls on channel pattern (e.g., channel shifting and cut-off; Keller and Swanson, 1979; Hickin, 1984; Piégay, 1993; Piégay and Marston, 1998) and the formation of flood plains and valley-bottom landforms (Abbe and Montgomery, 1996; Piégay and Gurnell, 1997; Piégay et al., 1998; Gurnell et al., 2001). Modern forests cover almost a third of Earth’s land surface, and wood debris has been entering streams and rivers for more than 400 million years. There is little doubt that vegetation influenced ancient rivers; the first appearance of meandering stream deposits in the geologic record coincides with the evolution of land plants (Schumm, 1968; Cotter, 1978). In the less remote past, late Cenozoic changes in global vegetation patterns imparted substantial variability to the role of wood in world rivers. Whereas only a few European rivers in small isolated areas of forest refugia would have had significant wood loading during the glacial maxima, parts of eastern Australia have been continuously forested for more than 100 million years. In historic times, humans have reduced global forest cover to about half its maximum Holocene extent and eliminated all but a fraction of the world’s aboriginal forests. Although the net effects of human actions have been to homogenize and simplify river systems around the world, there are still strong regional contrasts in the geomorphological effects of wood in rivers. In most industrialized nations, the primeval character of rivers remains shrouded by time because of ancient deforestation and river clearing to improve navigation before recorded history. Human alteration of forests in Europe has been significant for at least 6000 years (Williams, 2000). Forests in southern Europe were already confined to mountainous areas by classical times (Darby, 1956), and river clearing and engineering date to the Roman era (Herget, 2000). The morphology of the riparian trees established along European stream channels reflects centuries of successive clearing campaigns, tree selection favoring root network growth, and efforts to prevent bank erosion. As the geomorphic effects of centuries of ‘‘riparian gardening’’ are revealed by progress in the understanding of both structural effects of LWD on channel forms and the natural geometry of stream channels in rural areas, the full extent of humaninduced aspects of stream geometry is becoming more and more evident in European rivers.