Introduction
Globally, rivers have been severely altered over the last century due to human population growth, accelerating economic activity, land-use alteration, water development and climate change that have all interrupted fluxes of water, sediment and nutrients (Dynesius and Nilsson, 1994; Ward, Tockner, & Schiemer, 1999; Syvitski, Kettner, Correggiari, & Nelson, 2005), simplified the physical structure of habitats and floodplains (Beechie, Beamer, & Wasserman, 1994; Hohensinner, Jungwirth, Muhar, & Habersack, 2005), and degraded habitat and water quality in river systems by the loading of nutrients and pollutants (Williams, Cook, & Smerdon, 2021). These impacts to watersheds and rivers have altered riverine ecosystems dramatically and, as a result, the freshwater biodiversity of North America has become increasingly threatened. Especially severe impacts of ecosystem stressors can be observed in the dryland systems of the American Southwest where water is scarce relative to demand and there has been a growing urgency over the past few decades amongst managers and stakeholders for a more holistic approach to river management and restoration (Laub, Jiminez, & Budy, 2015).
In arid or semi-arid rivers, reduction in the frequency and/or magnitude of seasonal floods often induces a shift from a braided or anastomosing to a meandering single-thread channel, decreases channel width, causes valley alluviation, and is often accompanied by changes in riparian vegetation communities which favor domination by nonnative taxa (Vorosmarty et al., 2010; Castle et al., 2014; Udall & Overpeck, 2017). This degradation process is especially prevalent in the arid American Southwest, where four of the 14 fish species native to the Colorado River are considered threatened or endangered under the Endangered Species Act (ESA; Rinne & Minckley 1991; Laub et al., 2015). Remediating threats to the persistence of native biota in Southwest desert rivers will likely require coupled management of flow regimes and active in-channel restoration efforts in order to preserve and maintain crucial habitat (Propst, Gido, & Stefferud, 2008; Pennock, Ahrens, McKinstry, Budy, & Gido, 2021).
The San Rafael River (SRR), a tributary of the Green River in the Colorado River Basin (CRB), is representative of many desert river systems, in that an altered flow regime, fish passage barriers, degraded habitat, and nonnative fish and vegetation have combined to synergistically alter ecosystem processes and threaten the persistence of native communities (Macfarlane et al., 2017; Olden & Poff, 2005; Stromberg, Beauchamp, Dixon, Lite, & Paradzick, 2007). In an attempt to mitigate degradation, an adaptive restoration and monitoring plan was developed (Laub, Jiminez, & Budy, 2013), and there have been a few small-scale restoration projects that include the installation of beaver dam analogs (BDAs) and post-assisted log structures (PALS), as well as tamarisk removal (Keller, Laub, Birdsley, & Dean, 2015). However, mitigation goals in the SRR have tended to focus on reach-scale restoration, and the potential role of valley plugs and avulsions in restoration-conservation efforts had not been previously considered. Additionally, restoration efforts are challenged by the large spatial extent of degraded habitat (> 64 km), the lack of adequate funding for system-wide restoration and, most recently, by pervasive drought and the over-allocation of water within the watershed.
In July 2010, a severe rain event caused flash-flooding in Cottonwood Wash, a tributary of the SRR (Lyster, 2018). This rain event coincided with low flows and reduced sediment transport capacity in the SRR, resulting in the deposition of a valley plug at the confluence with Cottonwood Wash. The term ‘valley plug’ was coined by Happ, Rittenhouse and Dobson (1940) to describe the processes and deposits resulting from an occluded canal or river channel. The initial plug, stretching from 400 m downstream of Cottonwood Wash confluence to 450 m upstream of Cottonwood Wash confluence, blocked the flow of water in the main river channel and brought water levels to above floodplain elevation (Utah State University Water Research Laboratory, 2010). As a result, the SRR underwent rapid avulsion and transitioned from a single-thread channel to a multi-thread channel that began at the confluence with Cottonwood Wash and rapidly expanded upstream (Lyster, 2018). Root masses of invasive species such as tamarisk (Tamarix ramosissima ) were believed to have stabilized floodplain soils, preventing the upstream migration of channel headcuts and causing water to remain at near-floodplain levels for nearly a year. Water retention further facilitated expansion of this wide, shallow and heterogenous reach of river, which appeared to be contributing to dramatic habitat diversification within the system and, as of 2021, had continued to expand upstream.
The SRR valley plug offered a unique opportunity to examine the impacts of potential large-scale innovative restoration actions in a historically degraded desert tributary system. We predicted that the resulting rapid geomorphic change contributed to both habitat diversification and increased floodplain inundation, and was actively creating opportunities for native fish persistence. Our research objective was to determine if the valley plug had contributed to both an increased diversity of habitat types and an improved capacity for water retention against a background of otherwise extremely simplified and degraded desert river habitat. To address these aims, we calculated straightforward metrics of habitat diversity utilizing on-the-ground habitat sampling, as well as manual landscape feature digitization using Google Earth and drone-collected orthomosaic imagery.