Modeling to Support Anadromous Fish Restoration (hosted by AFS)

ROOM: Atlantis, Grand Ballroom 2
Restoration of anadromous fish populations, through measures such as flow alteration, gravel addition, floodplain and side channel creation, installation of fish screens and ladders, and dam removal, increasingly depends on computer modeling to predict the effects of the restoration actions on fish habitat, survival and passage. Modeling can be a cost-effective technique to test different alternatives, and to address permitting requirements, such as effects on floodways. At a whole-river scale, modeling can identify and rank the cost-effectiveness of floodplain restoration sites. At even larger spatial and temporal scales, modeling can be used to identify watersheds and categories of habitat restoration measures that will be most effective in increasing anadromous fish populations. This symposium will focus on recent advances in modeling to support anadromous fish restoration. The purpose of this symposium is to highlight the importance of 1) the capabilities and limitations of different models; 2) the applicability of different models at different spatial and temporal scales; and 3) model parameters and inputs.

8:00AM Using Mobile Bed Hydraulic Models to Design Solutions to Upstream Passage Barriers
  Mark Gard
The mobile bed version of the two-dimensional hydraulic model SRH2D was used to analyze 21 different alternatives to prevent the deposition of a cobble bar that blocked flow into a new fish ladder on Mill Creek. Implementation of the preferred alternative will allow unimpeded upstream migration of adult spring-run Chinook salmon to their summer rearing habitat. The presentation will focus on the strengths and limitations of the hydraulic model used, and the data requirements for the model.
8:20AM Changes in Salmon Spawning and Rearing Habitat Associated with Channel Restoration
  Mark Gard
The River2D two-dimensional hydraulic and habitat model was used to simulate fall-run chinook salmon Oncorhynchus tschawytscha spawning and fry and juvenile rearing habitat before and after restoration of stream channel sites for a range of streamflows on the Merced River and Clear Creek, California. For the Merced River, hydraulic and structural data were collected for four sites before and after restoration, representing all habitat types within the restoration reach. Habitat simulated for these sites as extrapolated to the entire restoration reach based on habitat mapping. For Clear Creek, hydraulic and structural data were collected for four sites before restoration and pre-restoration habitat was imulated. The topographic plan for the restoration was used to simulate habitat after restoration. While the restoration generally increased spawning habitat, it was less successful for rearing habitat. The results of this modeling show how they can be used in a cost-effective adaptive management framework to evaluate restoration project design prior to construction.
8:40AM Evaluation of a Multiphase Reach-Scale Restoration Project on the Lower Mokelumne River Using 2D Models and Fisheries Monitoring Data
  Robyn Bilski
Reach-scale restoration of the lower Mokelumne River began over a decade ago, in the upper one-mile reach of the river just downstream of Camanche dam. The objectives of this project include rehabilitation of the river’s longitudinal profile, raising the river bed elevation to pre-dam conditions, enhancement of salmonid spawning habitat, reconnecting floodplain habitats, and improving ecosystem health. A multiphase rehabilitation plan was implemented from 1999 through 2016. During this time 45,000 yd3 of gravel and cobble were added to the reach, two mine pits were filled, and two side-channel habitats and two floodplain habitats were excavated. 2D hydrodynamic models are used to compare the 1999 reach condition to the 2016 reach condition. Pre and post project topography, bed slope, sediment transport potential, and changes to salmonid spawning habitat, perennial rearing habitat, and floodplain rearing habitat suitability are compared. Model predictions are tested using fisheries monitoring data and limitations are discussed. As reach maintenance begins, we review early recommendations and sediment transport predictions. Newer versions of 2D models now solve added sediment transport variables including sediment concentration, erosion and deposition. We compare these predictions against empirical data collected after the flood flows of 2017 and discuss implications for maintenance planning.
9:00AM Designing a Side Channel for Retention of Juvenile Salmonids
  David L. Smith, Robert Hilldale, Tammy Threadgill, John Hannon
Habitat suitability indices are ubiquitous in the literature Their use to predict fish habitat in conjunction with 2D depth average CFD is common. HSIs have the advantage of being easily explained and because they do not include mechanistic aspects of the system in question and they avoid the uncertainty of model formulation and data development. Despite using hydraulic models to describe in detail the flow field and relying on observations of fish occurrence there is no integration that account for the physiology, cognitive capacity or internal state of the fish. Thus the mechanistic linkages between the fish sensory system, hydraulic resistance, and the ultimate habitat occupancy of the fish is not made. In this paper we propose a new habitat based procedure that does integrate between fish physiology, cognitive ability and the flow field that makes predictions of the fish distribution and can be calibrated and validated. This approach represents an increase in complexity from standard HSI driven analysis, but is less complex than a full modeling effort that include mechanistic linkages between many aspects of the ecosystem and organism in question. We believe this could help address a critical design challenge in developing new juvenile salmonid habitat.
9:20AM Chinook Salmon Physical Habitat Evolution Following River Restoration, Drought, and Flood
  Rocko Brown
River evolution has been historically governed by changing geology, climate, and biogenic processes. In the Anthropocene, humans are now playing a larger role in modulating the scale of physical change in river corridors, especially in regulated rivers in Mediterranean climates. The overall objective of this study was to evaluate the physical evolution and biological response of a highly impacted regulated river reach following restoration and subsequent drought and flood hydrology. Topographic change detection of repeat surveys, reach scale bed surface size distributions, two-dimensional hydraulic modeling of Chinook spawning and rearing habitat, and riffle persistence were analyzed for pre-project, post project and post flood conditions. Redd surveys were conducted in the project reach to evaluate utilization over time as well as estimate the relative contribution of spawning on channel change. Results show that coarse sediment augmentation, the construction of morphologic archetypes such as riffles, bars and pools, and floodplain and side channel excavation improved habitat for spawning and rearing salmonids. In addition to quantifying habitat gains this study highlights the relative role of anthropogenic, fluvial and biologically driven topographic change in regulated rivers in the Anthropocene.
09:40AM Break
1:10PM Applications of a Disease Model for Listed Coho Salmon to Evaluate Restoration Flow Regimes
  Nicholas A. Som, PhD, Russell Perry, Nicholas Hetrick, Julie Alexander
The myxosporean parasite Ceratonova shasta is known to have increased mortality rates for outmigrating populations of Klamath River juvenile salmonids for several decades. In order to evaluate how mortality risk varies according to in situ riverine characteristics, a series of fish-exposure experiments were conducted. The experimental design targeted broad ranges of water temperatures and waterborne concentrations of the parasite, which are each known to impact mortality risk, and the exposure duration to these variables, which has been hypothesized to correlate with mortality risk. These experimental exposures were conducted separately for Chinook and Coho salmon owing to observed historical variation in mortality rates among the species. In this talk, we first summarize the details and benefits of applying mixture-cure models for Coho Salmon to predict both the probability of mortality, and given susceptibility, the time to death of individuals. We then focus on an application of this model to evaluate a restoration flow regime designed to reduce C. shasta risk. More specifically, the flow regime restores scouring events, timed to be consistent with the natural flow regime, that are intended to reduce the abundance of Manayunkia speciosa, a polychaete worm known to be the intermediate host in the C. shasta lifecycle.
1:30PM Using the Stream Salmonid Simulator (S3) to Assess the Effect of Flow Management on Mortality of Juvenile Chinook Salmon Caused By the Myxosporean Parasite Ceratonova Shasta
  Russell Perry, John M. Plumb, Nicholas A. Som, PhD, Julie Alexander, Nicholas Hetrick
Flow regulation and the effects of impoundment on water temperatures of the Lower Klamath River, California, have created favorable conditions for the myxosporean parasite Ceratonova Shasta, thereby causing recurring disease outbreaks in juvenile anadromous salmonids. To reduce mortality in juvenile salmon that pass through the “infectious zone”, an 80-km reach downstream of the lower-most impassable dam, resource managers hypothesized that bed scour from annual 3-day flushing flow events would reduce polychaete worm populations, the parasite’s intermediate host. To test this hypothesis, we incorporated a disease sub-model into the Stream Salmonid Simulator (S3), a spatially-explicit population model that simulates daily growth, movement, and mortality of juvenile Chinook Salmon in the Klamath River. We ran the model over a historic 12-year period that bracketed both high- and low-disease years and simulated flushing flow events to delay and reduce C. shasta spore densities. For all populations that passed through the infectious zone, annual flushing flows significantly reduced prevalence of infection and increased survival and the abundance of fish that entered the ocean. Our analysis shows how spatially-explicit models can help understand how management actions affect fish populations that migrate through spatially and temporally variable environments.
1:50PM Increasing the Availability and Spatial Variation of Spawning Habitats through Ascending Baseflows
  Damon H. Goodman, Nicholas Som, Nicholas Hetrick
Precipitation in fall and winter is important to recharge aquifers causing the baseflow in rivers ascend during the time when Chinook Salmon construct redds. Herein, we evaluate the availability of spawning habitats under a constant streamflow common in regulated rivers against ascending baseflows patterned from free‐flowing rivers. Hydrodynamic and habitat models were leveraged together to simulate the quality, amount, and spatial distribution of spawning habitat at a series of individual flow rates, as well as the combined effect of those flow rates through a spawning season with ascending baseflows. Ascending baseflows increased the abundance of spawning habitat over individual streamflows at a site where the river channel is confined by levee‐like features. However, improvements were greater at an unconfined site that facilitated lateral connectivity and greater expansion of wetted channel area as streamflows increased. Ascending baseflows provided spatial separation in preferred habitats over a spawning season, which may reduce the risk of superimposition among runs or among species. Although replicating natural flow regimes in managed systems can be impossible or impractical, these results demonstrate that incorporating elements of the natural flow regime like ascending baseflows can benefit the restoration and conservation of riverine species.
2:10PM Mechanistic Individual-Based Models in Salmon Restoration
  Steven F. Railsback
InSTREAM and inSALMO are detailed, mechanistic, individual-based models of salmonid populations and their response to habitat alteration. The models represent how population characteristics such as abundance, persistence, spawning success, and age and size distributions emerge from channel shape, flow and temperature regimes, and the adaptive behavior of fish. Over 20 years we have applied these models to problems such as designing and evaluating the benefits of channel restoration projects, evaluating instream flow alternatives, and analyzing effects of peaking hydropower. A related model simulates how flow and temperature regimes affect breeding success of foothills yellow-legged frogs, so conflicts between salmonid and frog management can be identified and avoided. Application of these models requires field data and hydraulic simulation to create virtual stream reaches, and then calibration of the biological components. Compared to traditional habitat preference approaches, these models have similar input and calibration requirements, are more complex and challenging to understand, but provide results more directly relevant to management decisions and therefore requiring much less interpretation. Another advantage of individual-based models is that they provide a comprehensive way to think about habitat effects on populations and to design field and laboratory studies that directly improve management decision-making.
2:30PM Hydraulic Model Selection: Maximizing Model Performance for Fish and Wildlife Habitat Restoration
  Fred Meyer
Increasing two dimensional (2-D) hydraulic model platforms, improving ease of use of 2-D models, and increasing computing power have made their application in aquatic based fish and wildlife analyses commonplace. Here, 2-D hydraulic models for two different applications are considered: 1) channel restoration design, and 2) instream flow assessments. The channel restoration design application requires multiple iterations between design alternatives to meet geomorphic, riparian and ecological objectives. For this example, the RAS2D hydraulic model was used to iteratively assess a complex topographic restoration site on the Trinity River, where one of the primary objectives was too increase salmonid rearing habitat during winter and spring periods. The instream flow assessment application model’s streamflow to habitat relationships, thus the focus is on streamflow management and associated effects on fish, amphibian, and wildlife habitat (i.e., static topographic inputs). For this example, the FaSTMECH 2-D hydraulic model was developed for a steep boulder/bedrock reach of the Tuolumne River where topography was static, and the objective was to evaluate salmonids and foothill yellow-legged frog habitat due to variations in streamflow. In both cases, model output was used for WUA curve construction, individual based modeling of rainbow trout populations, and/or foothill yellow-legged frog reproduction modeling.
2:50PM Refreshment Break
3:20PM Forecasting the Downstream Migration of Adult Silver Phase American Eels
  Daniel Weaver, Douglas Sigourney, Marie-Beth Delucia, Joseph Zydlewski
American Eel (Anguilla rostrata) is a catadromous species ubiquitous in coastal waters of the eastern United States. Beginning in late summer, silver phase eels migrate from freshwater systems out to the Sargasso Sea to spawn. During downstream migration they face substantial injury and mortality passing through hydroelectric facilities. Improved predictions as to the timing and magnitude of eel migrations may inform mitigation actions (e.g., operational shutdowns) to reduce injury and mortality. We used a 30-year data set consisting of daily eel catch and environmental variables (discharge, temperature, moon phase, ordinal date) among several Northeastern US rivers. We constructed generalized additive models and generalized linear mixed models with a Bayesian hierarchical framework to describe the relationships between variables and the migration of silver eels. A Bayesian hierarchical approach effectively estimated parameters and error structure across multiple systems. This work represents a step towards the development of an auto-regressive integrated moving average (ARIMA) forecasting model to predict silver eel movement using forecasted environmental covariates.
3:40PM Understanding Sensitivity of American Shad Populations in the Northeastern United States to Climate, Dam Passage, and Fisheries
  Erin Gilligan, Dan Stich
American shad Alosa sapidissima are anadromous fish with populations on a noticeable decline throughout the entire U.S. Atlantic coast. Being highly sensitive to environmental factors, it is believed that the three primary drivers of their population and range include hydropower dams, overfishing, and global climate change. NOAA Fisheries has developed a dam passage performance model that combines cohort-based projections with individual-based models, while incorporating both temporal and spatial components. We investigated sensitivity of population abundance to life-history uncertainty and dam passage performance standards using this model for the Penobscot, Merrimack, Connecticut, and Susquehanna rivers. The relative sensitivity of these populations to passage standards at hydropower dams is not only influenced by local biology, and dam passage performance standards, but also by the configuration of the hydrosystems relative to spawning habitats. Finally, we incorporated climate and fishery scenarios into the Connecticut River model to help better understand interactions between climate change, dam passage, and fisheries in planning for the future. Concurrent consideration of all these factors, and their associated uncertainty will assist with making more robust management decisions.
4:00PM Planning for Recovery – Filling the Salmonid Rearing Habitat Deficit on the Lower American River, California
  Chris Hammersmark, Nick Southall, Michael Founds, Travis Hinkelman, Lilly Allen
Previous salmonid recovery efforts in the American River have focused on gravel augmentation to improve habitat conditions for spawning; however, in recent years the focus has shifted to include the consideration of rearing habitat enhancement efforts. But how much rearing habitat is really needed and where should it be located? This presentation will provide an overview of science-based planning efforts to quantitatively identify and prioritize potential rearing habitat enhancement projects. The need for rearing habitat is estimated with the Emigrating Salmonid Habitat Estimator (ESHE) model. The amount of existing habitat is estimated through the application of a detailed hydrodynamic model combined with habitat suitability indices to estimate habitat suitability throughout the river corridor. The difference between these two is the rearing habitat deficit which varies both spatially and temporally. A DEM of the river corridor and results of the hydrodynamic model are used to in combination to calculate the height of existing floodplain areas above typical spring flows in order to identify and develop conceptual designs for potential floodplain enhancement sites. The identified sites are compared to the estimated rearing habitat deficit combined with other factors to prioritize potential sites that will be advanced in future phases of the effort.
4:20PM Integrated Tools for Identifying Optimal Flow Regimes for Recovering Spring Chinook Salmon in the Willamette River, Oregon
  James Peterson, Jessica Pease
Sustaining the ecological integrity of lotic ecosystems, while providing needed water resources for human needs is a major challenge facing society. The Willamette River, Oregon has been altered by impoundments, flow regulation, and human land development, which has negatively affected fluvial-dependent biota such as federally threatened spring Chinook Salmon Oncorhynchus tshawytscha and winter steelhead O. mykiss. Working with an interdisciplinary team of scientists and managers, we developed a develop decision support tool that integrates flow, habitat, and temperature models with ecological models for predicting the response of spring Chinook Salmon to alternative flow regimes. The model tracks the arrival, migration, survival, and spawning of adult salmon and the emergence, growth, survival, and movement of juvenile salmon in response to changes in flows, habitat availability, and temperatures. The model is used to find the optimal allocation of streamflow through time, given constraints on water availability. Modeling results suggest that the optimal allocation of water varied by water year type and with salmon life history stage. To make these tools available to managers; we are developing user-friendly apps for evaluating potential tradeoffs associated with alternative flow management strategies.
4:40PM Decision Analysis to Identify Optimal State-Dependent Restoration Policies for Chinook Salmon in California’s Central Valley
  Adam Duarte, James Peterson
Anadromous fish are a valued resource for their economic, cultural, and ecological importance. Yet, their reliance on both freshwater and ocean environments makes them susceptible to myriad human-induced stressors throughout their life cycle, which have contributed to their decline in California’s Central Valley. We used a decision theoretic approach to identify restoration objectives and develop a probabilistic decision-support model for evaluating alternative Chinook salmon (Oncorhynchus tshawytscha) restoration actions. Using the decision-support model, we identified optimal state-dependent restoration policies that would yield the largest number of Chinook salmon adults in the future based on adult escapement, spawning and rearing habitat availability, and juvenile survival probability. Forward simulation suggested that implementing the optimal state-dependent policies resulted in a greater number of Chinook salmon adults in the future, but sensitivity analysis indicated that model predictions were strongly influenced by model uncertainties. Thus, we also discuss how monitoring data can be used to refine our belief in model uncertainties and improve future management decision making. Collectively, the results of our study revealed a management strategy that can be applied at a broad scale and may be applied to systems with similar constraints.

Organizers: Mark Gard
Supported by: AFS Fish Habitat Section, AFS Bioengineering Section

Location: Atlantis Hotel Date: October 1, 2019 Time: 8:00 am - 5:00 pm