Linking models across scales to assess the viability and restoration potential of a threatened population of steelhead (Oncorhynchus mykiss) in the Middle Fork John Day River, Oregon, USA
Introduction
Efforts to recover threatened and endangered species often rely on models to identify management strategies that will improve the status of at-risk populations (Good et al., 2007, Sweka and Wainwright, 2014). Fundamental to this work is the task of building models that are structurally or numerically consistent with the spatial and temporal scales of relevant biological processes and management actions (Green et al., 2005). For species like Pacific salmon and steelhead (Oncorhynchus spp.), this task is especially challenging given that they respond to factors operating at widely varying spatial and temporal scales. For example, the foraging success, growth, and survival of individuals can be influenced by conditions within small (10–100 m2) patches of stream habitat (e.g., Rosenfeld and Taylor, 2009), yet a river’s overall capacity to produce juvenile fish may be set by habitat limitation for another life stage that is governed by broader-scale geomorphic controls (e.g., spawning gravel availability, Buffington et al., 2004, Palm et al., 2007). Similarly, global climatic phenomena (e.g., El Niño Southern Oscillation) can affect the survival of salmon during marine life stages (e.g., Sharma et al., 2013) to an extent that makes it difficult to detect the effects of factors operating during freshwater stages (e.g., Schaller et al., 2013). Accordingly, there is a need for modelling approaches that can accurately depict complex life histories and simultaneously maximize the realism of species–environment relationships for life stages targeted by management and restoration.
The need for such tools is particularly acute in the Columbia River Basin, home to a large-scale, long-term, and costly stream restoration program that was developed to recover salmon and steelhead populations through spawning and rearing habitat improvements and barrier removal (Leonard et al., 2015). Although restoration comes in many forms, the general goal of most projects is to increase the quality or quantity of habitats that are believed to limit particular life stages and overall population abundance (Barnas et al., 2015). In our study region, ongoing restoration aims to improve in-stream habitat complexity and water quality (e.g., cooler summer temperatures), among other conditions, based on a combination of riparian restoration and woody structure placement projects (Rine et al., 2016). To support prioritization of specific restoration activities, regional panels have identified restoration projects that are expected to confer the greatest demographic benefits to populations (e.g., Booth et al., 2016). However, at least two key uncertainties must be addressed before this question can be reliably answered. Firstly, candidate restoration projects must be translated into an expected habitat change (e.g., reduced fine sediment levels or cooler summer temperatures, Bartz et al., 2006, Jorgensen et al., 2009) and associated fish response (e.g., an increase in freshwater survival, Honea et al., 2009). Secondly, because the benefits of specific restoration projects are often localized in nature, these reach-scale expectations must be upscaled to the basin/population level so that their overall value can be assessed (Wheaton et al., In Press). If such uncertainties can be addressed quantitatively, then the long-term population-level benefits of restoration alternatives can be assessed using a life-cycle modelling approach.
Despite a rich history of life-cycle model (LCM) applications to salmon recovery-related questions (see Good et al., 2007 for a recent review), previous attempts to link habitat restoration scenarios with future population performance have met with limited success. For example, studies have modelled population responses to hypothetical survival increases assumed to be achievable through habitat restoration, but without explicit consideration of current conditions or restoration feasibility (e.g., Kareiva et al., 2000). More recent assessments have integrated habitat–survival relationships and basin-specific habitat condition into modelling (e.g., Honea et al., 2009, McHugh et al., 2004, Scheuerell et al., 2006), assessed restoration potential based on relationships between watershed condition and in-stream habitat metrics (e.g., Bartz et al., 2006, Jorgensen et al., 2009, Sharma et al., 2005), or increased the spatial resolution of LCMs (Scheuerell et al., 2006). The realism of LCMs has thus improved greatly, yet these tools continue to inform restoration practice at a coarse level and in the absence of mechanistic insight on fish–habitat associations. Thus, approaches that integrate LCMs with reach-level habitat observations and restoration plans are needed.
Recent advances in fish-habitat monitoring and modelling within the Columbia River Basin have made it possible to integrate biological responses and habitat condition at scales relevant to individual restoration projects. Specifically, the Columbia Habitat Monitoring Program, initiated in 2011, has collected the data necessary to parameterize hydraulic, drift-foraging, and habitat suitability models at several hundred survey sites (i.e., ca. 100–600 m long stream reaches) across the region (Wheaton et al., In Press). Using the Middle Fork John Day River (MFJD) as a test case, here we demonstrate a modelling framework that (1) leverages these data to estimate reach-level juvenile rearing and adult spawning capacity as a function of physical habitat, (2) upscales these reach predictions using larger scale basin data (e.g. GIS, remote sensing) to provide an estimate of population or basin total carrying capacity for key life stages, and (3) uses these values, along with other demographic data for the target population, in a LCM context to quantify the current status of the basin’s threatened steelhead (O. mykiss) population. We apply this framework retrospectively for validation purposes, then prospectively to assess the potential benefits of two commonly pursued restoration approaches; one that aims to enhance rearing capacity and survival for juveniles by providing cooler summer temperatures (ODEQ, 2010) and another that aims to increase juvenile carrying capacity through increased structural/hydraulic complexity of select reaches (via large wood addition; Li et al., 2016). Our objectives are two-fold: (1) to describe the modelling approach, which is sufficiently flexible for other applications in which similar data exist (note, our description here is brief, but we provide complete details in a companion supplement); (2) to apply it to a specific case and interpret our findings relative to the population’s conservation status. Additionally, in pursuit of these objectives, we demonstrate a practical approach for upscaling reach-level mechanistic models to inform population-level assessments.
Section snippets
Study system and model population
The Middle Fork John Day River drains the highlands of north central Oregon and is a primary tributary to the John Day River (Fig. 1). Its basin is moderate in size (ca. 2100 km2; average base flow 7.2 m3 s−1 [gauging station at rkm 24]) and spans habitats ranging from near-alpine at its crest (ca. 2200 m) to sagebrush steppe near its confluence with the North Fork John Day River (ca. 650 m) (O'Brien et al., In Review). The climate of the MFJD basin is semi-arid, characterized by cool, wet winters
Baseline scenario validation and results
The base (SQ) LCM parameterization, informed by a combination of published and unpublished data synthesis, original analysis, and network extrapolation results (Fig. 3, Appendices A & B in Supplementary Material), produced abundance dynamics consistent with recent sampling observations for MFJD steelhead. Escapement from baseline simulations ranged from 489 to 5482 whereas sampling estimates ranged from 811 to 5859 (Fig. 5). Total life cycle productivity (spawner-to-spawner) values computed
Discussion
Our analysis offers three insights of relevance to the management of habitat for steelhead within the Middle Fork John Day River basin. Firstly, based on an independent analysis and using a different set of criteria (i.e., IUCN, 2016), our status quo simulation results corroborate recent assessments (e.g., Ford, 2011) concluding that the MFJD steelhead population continues to warrant conservation attention. For example, although simulated escapements often met recovery targets, quasi-extinction
Acknowledgements
This work was funded by the Bonneville Power Administration (BPA) as part of the Integrated Status and Effectiveness Monitoring Program (BPA project # 2013-017-00) and the Columbia Habitat Monitoring Program (BPA project # 2011-006-00) and is the culmination of efforts from many affiliated with these programs, as well as other entities working within the Middle Fork John Day River Basin, including: Chris Beasley, Jody White, and Kevin See (Quantitative Consultants, Inc.); David Byrnes and Joe
References (112)
- et al.
Quantifying the role of woody debris in providing bioenergetically favorable habitat for juvenile salmon
Ecol. Model.
(2014) - et al.
Can stream and riparian restoration offset climate change impacts to salmon populations?
J. Environ. Manage.
(2017) - et al.
A quantitative framework for the analysis of habitat and hatchery practices on Pacific salmon
Ecol. Model.
(2005) - et al.
A review of statistical methods for the evaluation of aquatic habitat suitability for instream flow assessment
River Res. Appl.
(2006) - et al.
Escapement and Productivity Monitoring of Summer Steelhead and Spring Chinook Salmon in the John Day River
(2015) - et al.
Is habitat restoration targeting relevant ecological needs for endangered species? Using Pacific Salmon as a case study
Ecosphere
(2015) - et al.
Translating restoration scenarios into habitat conditions: an initial step in evaluating recovery strategies for Chinook salmon (Oncorhynchus tshawytscha)
Can. J. Fish. Aquat. Sci.
(2006) - et al.
Comparative thermal requirements of westslope cutthroat trout and rainbow trout: implications for species interactions and development of thermal protection standards
Trans. Am. Fish. Soc.
(2007) - et al.
Predicting natural channel patterns based on landscape and geomorphic controls in the Columbia River basin, USA
Water Resour. Res.
(2014) - et al.
Process-based principles for restoring river ecosystems
Bioscience
(2010)