Estimates of salmon and steelhead harvest rates in

Estimates of salmon and steelhead harvest rates in Columbia River fisheries based on the recovery of PIT tags Thomas Buehrens & Dan Rawding Photo: R. Ehlke Thomas Buehrens • thomas. buehrens@dfw. wa. gov • 2015 PIT Tag Workshop • Skamania Lodge • 1/27/2015

Photo: A. Stephenso Acknowledgements • • • Alan Byrne Doug Case Quinten Daugherty Roger Dick Jr. Robin Ehlke Michelle Groesbeck Lisa Harlan Rick Heitz Jay Hesse Joe Hymer • • • Ken Keller Kevleen Melcher Bret Morgan Bryant Spellman Steve Vander. Ploeg Ben Warren Jimmy Watts Bob Woodard Numerous field samplers! Columbia River TAC Funding: • BPA • Washington Columbia River Salmon and Steelhead Endorsement • NOAA-Mitchell Act Thomas Buehrens • thomas. buehrens@dfw. wa. gov • 2015 PIT Tag Workshop • Skamania Lodge • 1/27/2015

Historic Harvest Estimation in the Columbia River • Historic Harvest Estimation • Stock assignment based on CWTs, timing, size • Hatchery stocks substituted for wild stocks • Harvest reporting groups not based on biological populations • Need for new methods • ESA listed wild populations constrain fisheries • Population dynamic models require accurate harvest impacts Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Photo: Conclusions A. Stephenson

Project History • WDFW, ODFW, PSMFC have historically sampled fisheries catches for biological data and CWT to make harvest estimates • In 2010 BPA began funding WDFW to sample mainstem Columbia River fisheries for PIT tags • Sampling program includes mainstem sport, commercial, and treaty fisheries • New methods include Parentage Based Tagging (PBT), Genetic Stock ID, and PIT tags, but unlike genetic and Coded Wire Tag (CWT) programs, PIT tagging does not occur for harvest purposes Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Goal: Determine feasibility of using PIT tags to estimate harvest rates in Columbia R. fisheries • Sample PIT tags in fisheries catch • Expand PIT tag recoveries by speciesand week-specific fishery sample rates • Identify biological harvest reporting groups based on PIT tags • Estimate harvest rates for reporting groups Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Fishery Areas Also sport fishery Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Fisheries & Sampling • Sampling Data: – Species, weight, CWT, PIT tags, gutted/intact, number sampled – Total catch reported separately • Treaty – Above BON (Zone 6) Summer and Fall fisheries • Only fish sold to commercial buyers • No C&S or overbank sales • Commercial – Below BON (Zones 1 -5) Spring, Summer, Fall • Sport – Below BON Feb-Oct Thomas Buehrens • thomas. buehrens@dfw. wa. gov Photo: A. Stephenson • Introduction • Methods • Results • Conclusions

PIT Harvest Reporting Groups • Harvest rate based on susceptibility to harvest – Run timing confirmed with graphical analysis – Body size tested for ocean age/size selectivity – Migratory Behavior not yet! • Based on population structure – NOAA-designated hierarchical population structure • We developed hierarchical groups based on timing and biology Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

PIT Tag Harvest Model Components • Harvest Rate = Tagged Catch / Tagged Run-size • Tagged Catch = # of harvested fish with PIT tags – Number of tags sampled – # of harvested fish sampled for PIT tags (week by species strata) – Total number of harvested fish (week by species strata) • Tagged Run-size = # of fish passing BON with PIT tags • Organized by harvest reporting groups Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Catch and Run-size Depend on PIT Reader Efficiency Catch: Handheld Readers Allflex RS 601 -3 • Two best types >> 99% for all species • Robust to minor protocol violations Destron Fearing FS 2001 F-ISO • Other readers infeasible or lower detection rates Run-size: Bonneville Dam • Not many fish escape detection at BON • Allows for BON count of tags to be denominator of harvest rate with little error • 2011 data shown but consistent among years Species Ck Co Sth So Rate 0. 986 0. 979 0. 987 0. 992 Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Example: Steelhead Harvest in the Zone 6 Fall Fishery Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

PIT tag Recoveries (2011): Steelhead Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Cumulative Timing of Steelhead at BON: Support for reporting groups? Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Cumulative Passage at BON 5/1 Cumulative Timing of Steelhead at BON: Support for reporting groups? 6/1 7/1 8/1 9/1 10/1 11/1 Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Cumulative Timing of Steelhead at BON: Support for reporting groups? Cumulative Passage at BON Lower Columbia (LCR) Mid-Col. River (MCR) Snake A-run (SNA) Upper. Col. River (UCR) Snake B-run (SNB) 5/1 6/1 7/1 8/1 9/1 10/1 11/1 Thomas Buehrens 5/1 6/1 7/1 8/1 9/1 10/1 11/1 • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Harvest Rates of Steelhead in Zone 6 Fall Fishery Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Can we use PIT tags to incorporate age structure into harvest estimates? Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Ocean Age based on PIT tag travel days 0 1 3 2 1 • Based on days between release and detection as adult at BON • Initially used graphical analysis 0 1 3 2 4 • Confirmed by examining individual tag histories 1 • Currently collecting scales for further validation Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods 2 3 • Results • Conclusions

Age Selectivity Species • Suggests Age selectivity Coho in fisheries • Net fisheries catch is older than run • Mostly not managed beyond jack/adult • Can cause evolutionary change • Suggests need to move to adult-only or agestructured harvest estimates PIT Ocean ages at return Sockeye Sport 0, 1 NA NA 1, 2, 3 NA NA Sp. Chinook 0, 1, 2, 3, 4 Su. Chinook 0, 1, 2, 3, 4 Fa. Chinook 0, 1, 2, 3, 4 Steelhead Treaty Commercial 1, 2, 3 NA NA Size-selective Weakly Size-selective Not Size-selective Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Summary • WDFW successfully built on CWT sampling infrastructure to add other sampling (genetic, PIT tags, etc. ). • Detection efficiency of handheld and BON readers was very high permitting harvest estimation • PIT tag run timing data support grouping by geography/population designation • Tag recoveries often sufficient to make biologically-based ESUlevel harvest estimates; improve upon historical methods • PIT tags allowed assignment of ocean age and showed age selectivity in fisheries Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Future • Age-structured/adult only PIT tag harvest estimates • Hierarchical models for population-specific estimates • Increase sample sizes? – Incorporate PIT tag fishery recoveries in other Col. River sampling programs – Expand juvenile PIT tagging to cover poorly represented populations • Work with IDFG and CRITFC to combine genetic, PIT, CWT methods Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Thanks! Photo: R. Ehlke Thomas Buehrens • thomas. buehrens@dfw. wa. gov • 2015 PIT Tag Workshop • Skamania Lodge • 1/27/2015

Extra

Harvest Model Assumptions • Harvest susceptibility of PIT tagged fish is representative of untagged fish • Sampled catch is representative of entire catch • Total catch reported by TAC accurate • Sport Creel, Commercial Fish Tickets, Tribal Creel (estimates total of catch sold to buyers + C&S + over the bank) • For treaty fisheries, assumes commercial fish sales are representative of non sampled catch • Model currently treats all tags as adults • Lots of precocious life histories for Chinook, • Less of an issue for coho, steelhead, sockeye Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

• Introduction • • • Project goals Methods • • • Harvest rate data and model Detection study Fishery sampling Reporting Groups Harvest Reporting Groups Results • • Background/Need for new methods to estimate harvest Sampling stats BON detection rate PIT Harvest Groups Harvest Rates Coho Mixture Model Sockeye Subtraction Method More work to be done Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Detection Study (what % of tags are detected when tagged adults scanned? ) • Conducted an experiment at hatcheries under replicated field conditions • Tested efficiency under a variety of conditions: – – – Reader type Species Size Sex Simulated sampling protocol violations • Lots of replicates (100 fish * 4 samplers * 4 units per reader * protocols w/ + w/o deviations) Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

A Destron Fearing FS 2001 F-ISO B G “Sword” Reader All Flex RS 601 -3 (2 units) (Not Pictured) C Biomark Flat Plate D Biomark 24” square H Destron Fearing HPR (2 units) (Not Pictured) E Psion Teklogix data logger with RFID – LF module F Biomark Pocket Reader Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Detection Study Cont. : effect of protocol violations • Best units (AF and DF) only • Protocol = oval method, touching fish • Protocol violations • Single Swipe • Not touching fish AND single swipe Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Detection Results • >> 99% for all species with: Allflex RS 601 -3 and Destron Fearing FS 2001 F-ISO • when protocols followed • ~ 60% for Biomark pocket reader • ~ 95% but highly variable for Destron HPR • ~ 85% for “Sword” Reader • Flat Plate, Psion, Square not suitable for field sampling (slow, or unwieldy) Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Results: both readers robust to minor violations! • Likely due to detection distances >> fish body depth Photo: A. Stephenson Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Sockeye Subtraction Method The proportion of the tagged population is estimated by The number of Wenatchee and Snake PIT tags at BON is summed by statistical week and divided by the proportion of each population that is PIT tagged and is estimated by The weekly estimate of Okanogan sockeye passing BON is estimated by The total number of Wenatchee and Snake River fish harvested is the harvest rate of these groups calculated from PIT tags times the estimated number of these fish passing BON The number of Okanogan sockeye salmon harvested in the Zone 6 fishery is calculated by subtracting the number of Wenatchee and Okanogan sockeye from Thomas • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions the total. Buehrens Zone 6 catch

Sockeye Subtraction Method • Allowed estimation of weekly run timing of 3 populations • Demonstrated different run timing • Could allow fisheries to target abundant Okanogan & Wenatchee stocks while minimizing impacts to listed Snake stock Weekly Percentage 45% Snake Wenatchee Okanogan 30% 15% 0% 25 26 27 28 29 30 31 32 Statistical Week Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Coho Mixture Model • Use Pit tag detections of “early” stock as prior for bi-model mixture model of two normal distributions • Estimate early and late proportions for total counts for at BON Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Sampling Rates • Only non-dressed fish • Goal of 20% Commercial Season Week Spring 14 Spring 15 Spring 20 Spring 21 Summer 25 Summer 26 Chinook 2010 2011 Treaty Coho 2010 2011 Coho Sockeye Steelhead 2010 2011 0% 3% 5% 8% 8% 2% 29 Summer 30 18% 0% Summer 31 7% 1% 14% Fall 32 33 34 35 19% 8% 12% 18% 11% Fall 36 31% 12% Fall 37 23% Fall 38 Fall 39 36% Fall 40 Fall 41 Fall 42 Fall 43+ 25 Summer 26 Summer 27 Summer 28 Summer 14% 15% 32 34 Fall 35 Fall 36 Fall 39 Fall 40 23% 17% Fall 41 28% 26% 41% Fall 42 20% 34% 21% Fall 43 33% Chinook 2010 2011 10% 15% 25% 16% Summer Fall Thomas Buehrens Season Week 37% 64% 23% 50% 19% 47% 18% 17% 11% 26% 14% • Bold are periods w > 10% of annual catch • Red is <20% sample rate Sport 10% 21% 24% 33% 40% 16% 20% 5% 7% 18% 12% 16% 11% 22% 26% 13% 25% 17% 25% 29% 13% 16% 19% 14% 18% 20% 33% 12% 24% 9% 24% 28% 44% 30% • thomas. buehrens@dfw. wa. gov 12% 0% 11% 24% 6% • Introduction Month Chinook Steelhead 23% 2% 27% 11% 14% 10% 14% 23% 22% 8% 5% 16% 14% 2 3 4 5 8% 17% 15% 8% 6% 6% 5% 4% 6 10% 7 13% 8 9 10% 9% 11% 19% 16% 10 4% 10% 12% 20% 12% • Methods • Results • Conclusions

• Introduction Outline • Background/Need for new methods to estimate harvest • • • Project goals Fishery sampling for PIT tags Detection Study Harvest rate data and model Harvest Reporting Groups Results • • • Sampling stats BON detection rate PIT Harvest Groups Harvest Rates Coho Mixture Model Sockeye Subtraction Method Thomas Buehrens • thomas. buehrens@dfw. wa. gov Photo: A. Stephenson • Introduction • Methods • Results • Conclusions

Age Selectivity Species Coho • Suggests Age selectivity in fisheries Sockeye • Net fisheries catch is older than run BON Treaty Sport Ocean Age Run % Catch % 0 25 1 0 0 NA NA 1 2003 99 50 100 NA NA Sp. Chinook • Mostly not managed beyond jack/adult • Can cause evolutionary change Su. Chinook • Preliminary: • catch is unexpanded • could be biased by age-specific run timing Fa. Chinook • 2012 move to age. Steelhead structured harvest Thomas Buehrens • thomas. buehrens@dfw. wa. gov estimates 1 2 3 0 59 5 1040 89 73 6 505 0 0 6 100 0 0 NA NA NA Commercial Catch % 0 0 9 100 NA NA NA 8 0 0 0 1 2 3 4 1827 30 3013 50 654 11 1 0 3 12 5 0 15 60 25 0 10 7 4 0 48 33 19 0 6 21 1 0 21 75 4 0 0 532 24 0 0 0 1 2 3 4 890 40 610 27 189 9 1 0 3 14 4 0 14 67 19 0 0 0 4 100 0 0 7 1 0 0 88 13 0 0 1 2 3 4 477 9 2036 40 1021 20 1597 31 11 0 2 12 32 73 0 2 10 27 61 0 0 2 1 1 8 23 1 3 3 24 68 3 1 842 53 2 714 45 3 • Introduction 21 1 • 11 39 17 61 Methods 0 0 0 40 20 40 0 4 50 • 0 Results 0 • NA NA Conclusions NA NA

PIT tag Recoveries (2011): Coho Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Cumulative Passage at BON Coho Harvest Reporting Groups Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

PIT tag Recoveries (2011): Sockeye Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

Cumulative Passage at BON Sockeye Harvest Reporting Groups Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

PIT tag Recoveries (2011): Chinook Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions

CK “ESU” and “Rivers” reporting groups

PIT Tag Harvest Model Direct Acyclic Graph for Estimation of Harvest Rates Estimate Harvest Rate by Group Thomas Buehrens #estimate number of fish based on tickets by period #Estimate expanded PIT Tags by sampling period • thomas. buehrens@dfw. wa. gov • Introduction #estimate detection rate Of PIT tag readers • Methods • Results • Conclusions

Fishery Periods Thomas Buehrens • thomas. buehrens@dfw. wa. gov • Introduction • Methods • Results • Conclusions