#OpenAccess Article Alert: Direct and indirect drivers of instream wood in the interior Pacific Northwest, USA

While I briefly touched on the instream wood research that Alan Kasprak, Brett Roper, Christy, Meredith and I completed earlier this year, I wanted to take a moment to discuss both the rationale for our research and the rationale for where we published it in a longer article alert. 

A big question in federal and state aquatic monitoring programs is how riparian vegetation data can be used to make inference about aquatic ecosystems. This is partly because many programs have access to classified NAIP or LANDFIRE imagery that can be used to assess forest and rangeland changes following disturbances like wildfire, management actions like grazing retirement, or road building. Many organizations also measure riparian forest structure or composition as indicators of the integrity of the matrix surrounding a given stream network. This data can be used to assess individual streams over time or to compare the riparian composition and structure of groups of streams across successional or environmental gradients. Riparian vegetation is a bit political however, and fisheries and watershed managers often want to know, "what does riparian vegetation tell us about stream habitat quality, geomorphic change, or management impacts to the channel or fish population?" Because riparian vegetation is the keystone element of many stream networks, conferring shade to channel, stabilizing banks, contributing allochthonous energy to aquatic foodwebs, and forming habitat units like undercut banks, the popular thought has been that vegetation should be an indicator of aquatic processes and/or health. The relationships between riparian vegetation and aquatic ecosystem status have been well elucidated in the macroinvertebrate and fish literature.
Undercut banks are an important fish habitat type that is formed when flows carve out refugia along the active channel under riparian meadow and forest vegetation. http://water.epa.gov/type/rsl/monitoring/vms41.cfm

Accordingly, the project that would become our manuscript was undertaken as an early stab at identifying how large-scale monitoring data can be used to make inference about the drivers of aquatic ecosystem services. In the case of small streams, we know that instream wood (also often called large woody debris) is an important driver of channel change in small streams. For example, trees fall into channels where they shape local hydraulics that cause heterogeneity in shear stress that sorts sediment, forces scour and creates complex step-pool riffles (also referred to as plane-bed morphology). These geomorphic units are often tied to fish life cycles as fish use pools to forage or as refugia from predation and riffles for spawning. But what does it take to get trees into the channel as instream wood? Well, intuitively, trees or large shrubs that grow large enough to influence small channels. What changes the composition of a stream's riparian vegetation and will this influence instream wood loading? 

We started by assessing how much wood has been found in 720 low-order streams of the interior Columbia and upper Missouri River basins, USA. We looked at trends between environmental gradients, including climate, management, and matrix forest cover. In short, the coolest, wettest areas exhibited more large, long-lived and persistent tree species like Pseudotsuga menziesii, Thuja plicata, and Picea engelmannii and were unlikely to have recently burned or be heavily grazed. In short, forested areas have different climates than shrubland or grassland ecotones. We also saw that these species that occurred at reaches with heavy instream wood loads had distinct channel morphologies: they were wider, steeper and more likely to connect to steep hillslopes in the surrounding watershed.

A visual on the different riparian ecosystems based on their wood loading. Reaches are shown from first quartile to fourth quartile (least to most wood).
Given the correlations between riparian vegetation composition and instream wood loads and environmental gradients and riparian vegetation, we built a path model to look at how environmental change in climate, channel form and disturbance shape vegetation and how vegetation shapes instream wood. The big takeaway was visualized in a massive spaghetti-monster of a structural model: 
Nobody should ever forgive me for the complexity of this graph. Ever.
In short, wood volume and frequency were positively correlated to the forested vegetation types within the NMDS ordination. Woody vegetation was negatively correlated to grazing and wildfire and positively correlated to precipitation and catchment elevation, which also corresponded to wider channels with lower gradients, a product of numerous wide, low-elevation channels in Idaho's northern panhandle forests. We can infer from these relationships that anything that shapes either the ability of a landscape to grow and maintain forest vegetation or any factor that changes the channel dynamics responsible for transporting large wood will likely change instream wood loads. Accordingly, any area where woody vegetation might be predicted to decline due to climate change or disturbance will likely show less instream wood. In many cases, the relationships could be more complex, such as when a fire or insect outbreak kills trees that are quickly contributed to the active channel, but understory recruitment maintains riparian wood production. These same disturbances might also change channel forms to types that evacuate rather than store wood as water and sediment are rapidly evacuated from hillslopes, reshaping channels in large flow events. While not a conclusive predictive model of future scenarios, we outlined that direct vegetative and indirect climatic and disturbance effects are responsible for how wood loading varies across landscapes. Accordingly, managers can visualize the different gradients across which instream wood targets should be set, modifying goals in areas where wood recruitment is unlikely due to potential riparian vegetation composition.

We chose to publish this article in the open-access riparian ecology journal, Riparian Ecology and Conservation. I learned of the journal by chance in early 2013 when looking around for open access journals that catered to the applied water resources and ecology communities. I was pleased by the quality and thoroughness of the reviews that we received, even if southern hemisphere field work led to a less rapid review process than we had initially anticipated. With a great editorial board that includes Christer Nilsson, John Stella, Henri DeCamps, Lee Brown, and Editor, Yixin Zhang, I think that eventually the journal will rise to mid-level sub-disciplinary prominence amid Wetlands, Hydrobiologia, and River Research and Applications. This may happen quickly as all articles are distributed under a Creative Commons attribution license (CC BY-NC-ND 3.0).