Hydrogeomorphic and biotic drivers of instream wood differ across sub-basins of the Columbia River Basin, USA

Abstract

Instream wood promotes habitat heterogeneity through its influence on flow hydraulics and channel geomorphology. Within the Columbia River Basin, USA, wood is vital for the creation and maintenance of habitat for threatened salmonids. However, our understanding of the relative roles of the climatic, geomorphic, and ecological processes that source wood to streams is limited, making it difficult to identify baseline predictions of instream wood and create targets for stream restoration. Here, we investigate how instream wood frequency and volume differ between seven sub-basins of the interior Columbia River Basin and what processes shape these differences within these sub-basins. We collected data on wood volume and frequency, discharge and stream power, and riparian and watershed forest structure for use in modelling wood volume and frequency. Using random forest models, we found that mean annual precipitation, riparian tree cover, and the individual watershed were the most important predictors of wood volume and frequency. Within sub-basins, we used linear models, finding that some basins had unique predictors of wood. Discharge, watershed area, or precipitation often combined with forest cover, riparian conifer, and/or large tree cover in models of instream large wood volume and frequency. In many subbasins, models showed at least one hydrologic variable, indicative of transport competence and one ecological variable, indicative of the reach or upstream watershed's capability to grow measurable instream wood. We conclude that basin-specific models yield important insights into the hydrologic and ecological processes that influence wood loads, creating tractable hypotheses for building predictive models of instream wood.

Natural and Anthropogenic Controls on Wood Loads in River Corridors of the Rocky, Cascade, and Olympic Mountains, USA

“…Wood likely comes more dominantly from riparian mortality (related to forest stand characteristics and hydroclimatic/disturbance regimes) and bank erosion (Benda & Bigelow, ; Piégay et al, ). Our results indicating relationships between proxies for forest stand density (elevation at an intrabasin scale and climate or logging at an interbasin scale) and wood load support the idea that land use and hydroclimatic regime determine forest characteristics and resulting wood supply (Hough‐Snee et al, ).…”

“…Unit wood volume in valley bottoms sometimes correlates inversely with channel width and drainage area (Beechie & Sibley, ; Bilby & Ward, , ; Wohl, Lininger, et al, ), although this correlation has been observed to be direct in some cases and is strongly dependent on bioclimatic region and riparian forest characteristics (Burton et al, ; Wohl, Lininger, et al, ). Wood loads have little consistent relation to channel characteristics across bioclimatic regions, but individual regions and watersheds do display significant trends, allowing wood load to be predicted by variables describing geomorphic, ecologic, and anthropogenic conditions (Hough‐Snee et al, ; Wohl, Lininger, et al, ). While mechanisms influencing wood transport and storage have been explored in flume environments (Bocchiola et al, ; Braudrick et al, ; Davidson et al, ), we still lack a good understanding of how well experimental results translate to natural conditions.…”

“…Previous broad‐scale studies of wood load spatial variability (Hough‐Snee et al, ; Wohl, Lininger, et al, ) generally conclude that wood loads can be conceptualized at either broad (interbasin) or local (intrabasin) scales by taking into account either bioclimatic or site‐specific variables (e.g., land use and channel geometry), respectively. However, a conceptual model to describe wood load spatial distribution that applies at all scales has yet to be developed.…”

Phylogenetic diversity efficiently and accurately prioritizes conservation of aquatic macroinvertebrate communities

“…This work identified important drivers of habitat degradation, but did not quantify the biotic response to such disturbances. Efforts to assess biotic response have primarily focused on the vegetation component of the PIBO dataset, examining the impact of grazing on plants (Coles‐Ritchie et al 2007), of plants on stream structure (Roper et al 2007, Hough‐Snee et al 2015, Hough‐Snee et al 2016), and the physical site factors that predict plant invasion (Al‐Chokhachy et al 2013, Menuz and Kettenring 2013). These studies are a solid foundation for understanding the health of these sites, but as a primary prey source for the imperiled fishes PIBO was designed to monitor, an understanding of the macroinvertebrates of the sites is key to achieving management goals.…”

A conceptual model of riparian forest restoration for natural flood management

“…Variations in soil type, temperature, climate and elevation will lead to variations in tree growth and mortality between regions (Liu and Malanson, ). However, it is important to note there is a great deal of variation in live tree growth rates, deadwood accumulation (Hély et al , ; Christensen et al , ; Lombardi et al , ) and decay rates (Boddy and Swift, ), at both an inter‐continental but also intra‐regional, and even intra‐basin level (Hough‐Snee et al , ). Therefore, results of any forest growth modelling exercise which are not explicitly calibrated to a given stand should not be treated as explicit quantitative predictions; rather they should be used to inform management on the trajectory and magnitude of likely forest growth.…”