Conference Call

March 09, 2022 Slides

Meeting Summary

The SIT’s March 9, 2022 conference call began with an overview of the Adaptive Management Update, including the updates to priorities and feedback received on the draft. The Science Coordinator reviewed the upcoming SIT timeline and asked for any pre-proposals to be submitted by the April 13 SIT call. The SIT also heard updates from subgroups. Alexander Jensen (University of Washington) provided a science talk on Modeling the distribution of Chinook salmon stocks using tagging and genetic data.

Detailed Meeting Notes


Bernard Aguilar, Mike Beakes, Tricia Bratcher, Matt Brown, Erin Cain, Felipe Carrillo, Rebekah Casey, Charles Chamberlain, Megan Cook, Flora Cordoleani, John Dealy, Jane Dolan, Adam Duarte, Laurie Earley, Brian Ellrott, Austin Galinat, Chris Hammersmark, John Hannon, Brett Harvey, Jason Hassrick, Rene Henery, Baker Holden, Josh Israel, Alex Jensen, John Kelly, Priscilla Liang, Erin Lunda, Todd Manley, Bryan Matthias, Cyril Michel, Mike Memeo, Erica Meyers, James Pearson, Jim Peterson, Corey Phyllis, Keith Marine, Michael Prowatzke, Emanuel Rodriguez, Derek Rupert, Alicia Seesholtz, Nicholas Som, Kate Spear, Erin Strange, Mark Tompkins, Annalisa Tuel, Mike Urkov, Willie Whitfield, John Wikert, Heidi Williams, Rod Wittler, Michael Wright

10:00 Welcome and Agenda Review

10:05 FY20-21 Adaptive Management Update

  • Uncertainty still exists about the best to do scope of actions
  • Jim and Adam published peer reviewed publication on SIT DSMs and processes
  • FlowWest refactored and published DSMs, available online
  • Thank you everyone for your participation, input, and perseverance, lots of work representing a lot of effort from the SIT
  • Feedback was overall positive, not much pushback from first two updates to priorities
  • Suggestion for model tools (what they are, what they’re for, how they can be used)
  • Next steps- revised version of Adaptive Management Update sent out today or tomorrow, one week window to submit addendum, process for SIT members if there are any disagreements on conclusions
  • No pushback on priorities so far
  • Brian Ellrott: NMFS had questions about the process and timeline for the sturgeon and steelhead models, and where we’re at for those, and what the next steps moving forward are with those to identify restoration priorities
  • Megan Cook: Middle of figuring out what that looks like, work out there directly related to sturgeon and steelhead needs, and what that timeline looks like. Want to understand information needs and how to move forward
  • Jim Peterson: a year out for sturgeon model, a year out from summer for steelhead model
  • Rod Wittler: On the sturgeon side, trying to coordinate sturgeon monitoring, and new monitoring IEP is proposing. Coordinating that through modeling process

10:20 Upcoming SIT Timeline

  • FInalizing latest set of model updates, make those changes with new version of model by next time we come up with new version of NTRS
  • October is deadline for model changes
  • April is deadline for submitting new pre-proposals, September is deadline for full proposal approval so SIT can decide on merits/changes

    SIT proposals

  • In progress- spawning habitat decay, in development: climate change proposal, in development: updating movement rulesets. Reach out to Megan by April 13 SIT meeting with all potential proposals so SIT can review and discuss

10:30 Subgroup Updates

Climate Change Subgroup

  • Brian Ellrott is contact for climate change subgroup
  • Everyone welcome to join
  • Kickoff meeting is tomorrow, March 10 at 11am
  • Will report out to SIT at next meeting

Salmon Demographics Subgroup

  • Met Monday, March 7, reviewed how current DSMs are pushing fish through system, how movement is related to survival
  • Reviewed updated tool to look at results from different rulesets
  • Chinook salmon movement hypotheses app:
  • Good discussion on potential new movement hypotheses to represent in DSMs
  • Organizing call in near future to confirm hypotheses to consider, what rulesets we’re going to bring back to SIT group, and updates to app

    Habitat Subgroup–activities through March 7

  • Didn’t get to full subgroup meeting in February, ready to do that now mid to late March

  • Held several working meetings with USBR sediment transport modelers- feeding into both habitat input updates, and improvement in spawning habitat decay

  • For habitat input work, modeling and data collection- provided guidance to data collection efforts, hopefully flows will allow for Sac River bathymetry work

  • Meeting to close loop on decay proposal, will hopefully be included in next round of changes to model

10:50 Science Talk: “Where do all the salmon go? Modeling the distribution of Chinook salmon stocks using tagging and genetic data” – Alexander Jensen (University of Washington)

Overview and relevance

  • Alex is a postdoc at the University of Washington, update on work done looking at survival of Chinook salmon in ocean

Data availability

  • Data is publicly available/published
  • All supplemental data come from PC Council
  • Output data preliminary, hope to finalize model soon and hope to get it out for use by others
  • Chinook salmon priority information needs- most of the modeling focuses on marine stage of fish after leave freshwater, but does touch on juvenile salmon survival, may have interesting comparisons/trends to take away

Understanding Chinook salmon marine distributions

  • Complex evolutionary structure over space and time
  • Different river basins that have unique stocks of fish- own genetic and evolutionary histories
  • Mixed-stock groupings with varying vulnerability in the marine environment

Shelton et al. 2019

  • Integrated coast-wide modeling framework
  • Modeled life history of fall-run Chinook salmon stocks
  • Relied on CWT recoveries
  • Observed spatial segregation among stocks and seasons

Shelton et al. 2021

  • Added more years of data

Next steps

  • Increase resolution of stock groupings
  • Obtain inference on untagged portion of fish
  • Characterize relationship of other life history parameters with habitat
  • Include winter- and spring-run stocks of greater conservation concern

New sources of fishery data: genetic stock identification (GSI)

  • Using genetics to assign assemblage to stock proportions, or individuals to stock probabilities, using genetic markers
  • Advantages: assign every sampled fish to group- hatchery and natural-origin both assigned
  • Disadvantages- imprecision in assignments. Lose information release year, age; assignments may be uncertain

Expected data coverage

  • CWT: recreational fisheries (CA, OR, WA, BC), commercial fisheries (CA, OR, WA, BC, AK), pelagic hake fisheries (bycatch)- CA, OR, WA, 1970s-2018
  • GSI: recreational fisheries (CA, AK), commercial fisheries (CA, OR, WA, AK), pelagic hake fisheries (bycatch)- OR, WA, net fisheries (AK), late 1990;s onward

New research objective

  • Integrate CWT and GSI information to estimate spatio-temporal dynamics of Chinook salmon distribution and abundance in the ocean
  • start with a case study of CA and S. OR stocks (Central Valley fall, California Coast, Klamath River, North California/South Oregon Coast)
  • Leverage CWT and GSI data to improve understanding of low abundance stocks and natural-origin stock components

Research Methods

Integrated model structure

  • CWT model component–modeling CWT release group abundance over time
  • # released fish with CWTs-> # fish at model age 1-> # fish at model age 2-> …-> # fish at model age n (usually until age 6)

zoid: A mixture model (and R package) for modeling proportional data with 0s and 1s in ecology

  • We developed a new method for analyzing complex proportional data
  • Models are available as R projects on GitHub and CRAN
  • Submitted to Ecology

    Time-varying juvenile mortality rate

  • Other caveats: absolute mortality rates share inverse relationship with fishery mortalities

  • Model assumes increasing fishery catchability since 1979

  • Hatchery operations are not consistent over time, among stocks, or within stocks

Future work

  • Add model functionality for spring- and winter-run life histories
  • Expand the number of modeled stocks
  • Incorporate new data sources- expand GSI to include datasets from BC and AK, obtain outmigrant estimates by stock, release year to better scale stock abundances over time
  • Expand modeling of life history parameters as a function of habitat


  • Brett Harvey: For hatchery releases, have you accounted for release location? Alex: Potentially could account for that. The way model works is it doesn’t account for differing release practices/how those practices are done
  • Corey Phillis: what to make of higher variance in survival for Central Valley fish? Is it a model artifact or is that the result of either hatchery operations or the environment or the hatchery responding to the envir (like trucking more in drought year)? Alex: good question. Would have to look into more detail to answer more fully. Looking at some years – 4-5 hatchery releases in single year, don’t see as much for other stocks which typically have 1-2 hatchery releases happening. Think more variability in practices which create that higher degree of variability. Could also in part of model artifact because Central Valley put out a lot more fish.
  • Brian Ellrott: What are juveniles defined as here? Alex: Model says juvenile mortality terms encompasses everything from point of release all the way until the next spring, which is when the model starts tracking that life stage and starts to first estimate the abundance of the life stage. So includes mainstem, time in delta, early marine environment.
  • Jim Peterson: how to peel away ocean entry survival? What can say about how CWT fish acting like wild fish? Alex: the outmigration survival estimates are coming from the more dedicated models that take into account habitat more specific to systems, which will be more valuable and more useful to this which is very broad stroke. The mortality rate coming out is an incidental output vs a focal output. Can’t help but trust the more local stuff. In terms of hatchery vs wild survival: we know hatchery vs wild distrib is expected to be similar in terms of how they behave but trying to parse out. Think survival terms can be very different – depends on how tracking at release. More to come on this.
  • Questions?

11:30 Adjourn