Article alert: Hydrogeomorphic and Biotic Drivers of Instream Wood Differ Across Sub-basins of the Columbia River Basin, USA

Full article:
Please email me or link to Researchgate for PDF of final version.
My ET-AL teammates Alan Kasprak and Becca Rossi, Fluvial Habitat Center bosses, Joe Wheaton and Nick Bouwes, and USFS fish biologist, Brett Roper and I recently finished up an effort to model the abundance of instream large wood across the Columbia River Basin, USA. Instream wood, also known as large wood, large woody debris, etc., is an important component of stream and river evolution, and is monitored in many habitat programs across North America. As water flows over and around wood jams and pieces, local velocity changes, causing heterogeneous areas of erosion and deposition around the wood. For example, as water is forced around a piece of wood, velocity increases around the edges and porous root networks. Immediately downstream of a large piece of wood, this increase in velocity may cause sediment scour that forces pools to form. Similarly, as a flood over the wood recedes, velocity over wood slows allowing sediment to deposit and forming bars. Because of how wood causes heterogeneous velocity across a stream, it is an important driver of stream planform and channel complexity that provide diverse and dynamic aquatic and riparian habitats.
Site photos from each of the seven sub-basins.

This project, a part of the Columbia Habitat Monitoring Program (CHaMP), looked at wood loads, and the hydrogeomorphic and ecological factors that correspond to them in seven CHaMP-monitored watersheds. Historically, models of wood have been site specific, providing detailed inference into the processes that shape wood recruitment, transport, and storage at a single reach or a few reaches. In contrast, watershed managers, including habitat restoration biologists and stream restoration practitioners, often need to know what processes or pattern correspond to a given habitat attribute at broad scales. In the Pacific Northwest of the continental United States, numerous sub-basins are monitored for trends in their condition over time. Because wood may be the primary element shaping aquatic habitats in small streams, reaches or entire watersheds, areas that lack wood should be identified for restoration. More importantly, areas that are inherently wood-limited by the processes that grow trees, recruit them to the channel, and move them throughout a stream network should be identified. By identifying these patterns, basin-specific hypotheses (and models) can be constructed
Study reaches within the Wenatchee (A), Entiat (B), Tucannon (C), John Day (D), Grande Ronde (E), Lemhi (F), and South Fork Salmon (G) basins, USA. 
In this research, we identified that wood loads differ between sub-basins of the Columbia River Basin, and that the processes responsible for growing and transporting wood also differ between these sub-basins. Accordingly, a sub-basin without sufficient climate to grow forest vegetation or the hydrology and channel form to accumulate, rather than transport, wood will be intrinsically wood-limited. If wood is part of a habitat restoration or management plan in these wood limited streams, then restoration that introduces wood to channels (sensu Camp's HDLWD), restores forests that grow trees, or retains wood that is already in channels, may be necessary. In unconfined, valley bottoms where wood typically arrives from upstream forested areas, actions that encourage wood retention could include reintroducing beaver, maintaining keystone large wood, or using post lines to collect wood. In many cases, targets for wood should shift away form homogeneous, one size fits all approaches, instead targeting the mechanisms or processes that limit wood, and it's influence on habitat dynamism.
Instream wood differed in volume and frequency across the seven monitored sub-basins. Note that the John Day and Lemhi have much less wood than the other sub-basins.
For example, the South Fork Salmon is a wet, forested sub-basin with abundant wood while the Lemhi and John Day (and even the Wenatchee and Upper Grande Ronde) have less discharge, fewer high magnitude floods, and less precipitation and forest cover that result in lower wood growth and reduced wood movement. Accordingly, in areas where disturbance (e.g. grazing, logging, wildfire) has removed riparian forest, channels have been levied or straightened (e.g. near roads), and forest vegetation is unlikely to naturally recover, restoration of wood may be difficult and rely on forest restoration to increase local wood loads.

For more information, check out the full article at River Research and Applications or look at the PeerJ pre-print. Please email me for a pdf of the final version, or download via Researchgate if you're unable to access it directly via RRA.

I thank Dr. Simon Dixon for a particularly thoughtful and generous review that greatly improved the manuscript.
Non-metric multidimensional scaling ordination (first two of three axes) shows that hydrology and forest growth, etc. differed (right side) across the seven basins (left side)