B.01 Attachment 04 Spawning Survey-Data Report Sample.pdf

Work Completed for Compliance with the 2008 Willamette Project Biological Opinion, USACE funding: 2015 and 2016 hatchery baseline monitoring

Cameron S. Sharpe Robert L. Mapes

Brian Cannon Paul Olmsted

Michael Sinnott Bart DeBow Eric Bailey

Tracy Hoblit Thomas A. Friesen

Oregon Department of Fish and Wildlife Willamette Salmonid Research, Monitoring, and Evaluation Program

Corvallis Research Lab 28655 Highway 34

Corvallis, Oregon 97333

Prepared for the U.S. Army Corps of Engineers, Portland District Task Orders: W9127N-12-2-0004-4009 and W9127N-10-2-0008-0036

October 2017

Contents List of Tables

List of Figures

Section 1: Introduction

Section 1.1 Tasks

Section 1.2 Spring Chinook Salmon Production Program Goals

Section 1.2.1: Broodstock Collection and pNOB Goals

Section 1.2.2: Outplanting and pHOS Protocols and Goals

Section 1.2.3: Marking and Tagging of Hatchery Chinook Salmon

Section 1.3 Willamette River Escapement

Section 2: Methods

Section 2.1 Estimating Spawner Parameters: Distribution, Abundance, and Proportion of Hatchery-and Natural-Origin Chinook Salmon

Section 2.1.1: Monitoring Adult Returns

Section 2.1.2: Data Analysis

Section 2.2: Reintroduction Efforts

Section 2.3: Broodstock Sampling

2.3.1 Collection, Spawn Timing, Composition, and Disposition of Broodstock

Section 2.4: Within Hatchery Monitoring

2.4.1 Adult Monitoring

2.4.2 Juvenile Monitoring

Section 3: Results

Section 3.1: Abundance, Distribution, Spawn Timing and Composition of Naturally Spawning Adult Spring Chinook Salmon

Section 3.1.1 Adult Returns

Section 3.1.2 Redd Counts, Redd Distribution, and Spawn Timing

Section 3.1.3 Age Structure and Size Distribution on Spawning Grounds

Section 3.1.4 Spawner Abundance

Section 3.1.5 Estimates of prespawning mortality

Section 3.1.6 Origin on Spawning Grounds (pHOS)

Section 3.1.7 Straying

Section 3.1.8 Video Monitoring

Section 3.1.9 Harvest

Section 3.2: Reintroduction Efforts

Section 3.2.1 Number of Chinook Salmon Released Upstream of Dams

Section 3.2.2 Origin of Chinook Salmon Released Upstream of Dams

Section 3.3 Broodstock Sampling at Hatcheries

Section 3.3.1 Origin of Broodstock

Section 3.3.2 Broodstock Collection, Disposition, Age, and Size Distributions

Section 3.4 Juvenile Monitoring at Hatcheries

Section 3.4.1 Juvenile Production Program Goals

Section 4: Discussion

Acknowledgments

References

Appendix 1: Conservation and Mitigation Goals

Appendix 2: Spatial Scales Associated With Abundance, Spatial Distribution, and Diversity Metrics ... 129

Appendix 3: Survey reaches for upper Willamette subbasin prespawn mortality and spawner surveys .. 133

Appendix 4: Accounting of hatchery-origin Chinook salmon passing Willamette Falls

List of Tables

Table 1. Number of fish passing Willamette Falls by month, 2015 (top) and 2016 (bottom)

Table 2. Current and recent historical redd densities in comparable spawning reaches

Table 3. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the North Santiam subbasin, 2015

Table 4. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the South Santiam subbasin, 2015

Table 5. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the McKenzie subbasin, 2015

Table 6. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the Middle Fork Willamette subbasin, 2015

Table 7. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the North Santiam subbasin, 2016

Table 8. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the South Santiam subbasin, 2016

Table 9. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the McKenzie subbasin, 2016

Table 10. Peak redd counts, date peak count was recorded, and number of surveys conducted by survey section in the Middle Fork Willamette subbasin, 2016

Table 11. Age structure of natural- and hatchery-origin Chinook salmon, 2015 and 2016

Table 12. Size distribution of natural- and hatchery-origin Chinook salmon, 2015 and 2016

Table 13. Chinook salmon spawner abundance estimates, 2015

Table 14. Chinook salmon spawner abundance estimates, 2016

Table 15. Estimates of prespawning mortality of Chinook salmon in 2015

Table 16. Estimates of prespawning mortality of Chinook salmon in 2016

Table 17. Analysis results for otoliths collected from spawning ground surveys in 2015 and examined for thermal marks to verify wild status of unclipped adults

Table 18. Analysis results for otoliths collected from spawning ground surveys in 2016 and examined for thermal marks to verify wild status of unclipped adults

Table 19. Estimates of pHOS in 2015 based on counts of clipped and unclipped carcasses after adjustments following otolith analyses

Table 20. Estimates of pHOS in 2016 based on counts of clipped and unclipped carcasses after adjustments following otolith analyses

Table 21. Analysis of CWT recoveries during spawning ground surveys, at hatchery traps, and at hatcheries for run years 2015 and 2016

Table 22. Net number of marked and unmarked spring Chinook salmon and other species counted at

Upper (UB), Lower (LB) Bennett Dam and at Leaburg Dam in 2015 and 2016

Table 23. Estimated salmonid harvest in upper Willamette River subbasins from catch record card reports, 2000 - 2015

Table 24. Spring Chinook salmon outplanted, 2015

Table 25. Spring Chinook outplanted, 2016

Table 26. A comparison of estimates of live females after outplanting to peak redds observed in 2015

(top) and 2016 (bottom)

Table 27. Collection timing of Chinook salmon brood in 2015 and 2016. All fish were ad-clipped

Table 28. Age structure of Chinook salmon collected as brood at UWR hatcheries in 2015 and 2016

Table 29. Size distribution of Chinook salmon collected as brood in 2015 and 2016

Table 30. Comparison of size and age between Chinook salmon used for broodstock and NOR Chinook salmon sampled during spawning ground surveys in 2015 and 2016

Table 31. Summary of projected (Goal) and realized (Actual) Chinook salmon and steelhead trout smolt releases into the UWR in 2015 and 2016

List of Figures

Figure 1. The Willamette Basin with major dams, hatcheries, and fish collection facilities

Figure 2. Relationship between Prioritized Objectives, Reasonable and Prudent Alternatives (RPAs), Proposed

Actions (PAs), and Work Tasks conducted for spring Chinook hatchery programs in the Upper

Willamette Basin

Figure 3. Spawner abundance estimates based on redd count expansion and pHOS estimates based on carcass recoveries for reaches below dams through 2016

Figure 4. Spawner abundance estimates based on redd count expansion and pHOS estimates based on carcass recoveries for reaches above dams through 2016

Figure 5. Comparison of spawn timing in the rivers and spawn timing at the hatcheries in 2015 (left) and 2016

(right)

Figure 6. Relationship between natural-origin spawner abundance above Leaburg Dam comparing estimates from redd count expansion to estimates from dam counts

Figure 7. Recent historical mean age of natural-origin Chinook salmon in Upper Willamette subbasins

Figure 8. Comparison of broodstock collection timing to run timing of clipped and unclipped Chinook salmon in

2015 and 2016

Figure 9. Projected North Santiam CHS production goals, 2015

Figure 10. Realized North Santiam CHS production, 2015

Figure 11. Projected North Santiam CHS production goals, 2016

Figure 12. Realized North Santiam CHS production goals, 2016

Figure 13. Projected South Santiam Chinook production goals, 2015

Figure 14. Realized South Santiam Chinook production goals, 2015

Figure 15. Projected South Santiam steelhead production goals, 2015

Figure 16. Realized South Santiam steelhead production goals, 2015

Figure 17. Projected South Santiam CHS production goals, 2016

Figure 18. Realized North Santiam CHS production goals, 2016

Figure 19. Projected South Santiam STS production goals, 2016

Figure 20. Realized North Santiam STS production goals, 2016

Figure 21. Projected McKenzie Hatchery CHS production goals, 2015

Figure 22. Realized McKenzie Hatchery CHS production, 2015

Figure 23. Projected McKenzie CHS production goals, 2016

Figure 24. Realized McKenzie CHS production goals, 2016

Figure 25. Projected Leaburg Hatchery steelhead production goals, 2015

Figure 26. Realized Leaburg Hatchery steelhead production, 2015

Figure 27. Projected Leaburg STS production goals, 2016

Figure 28. Realized Leaburg STS production goals, 2016

Figure 29. Projected Middle Fork Willamette Chinook production goals, 2015

Figure 30. Realized Middle Fork Willamette Chinook production, 2015

Figure 31. Projected Middle Fork Willamette steelhead production goals, 2015

Figure 32. Realized Middle Fork Willamette steelhead production, 2015

Figure 33. Projected Middle Fork Willamette Chinook production goals, 2016

Figure 34. Realized Middle Fork Willamette Chinook production goals, 2016

Figure 35. Projected Middle Fork Willamette Chinook production goals, 2016

Figure 36. Realized Middle Fork Willamette Chinook production goals, 2016

Figure 37. Projected Middle Fork Willamette STS production goals, 2016

Figure 38. Realized Middle Fork Willamette Chinook production goals, 2016

Section 1: Introduction

The National Marine Fisheries Service (NMFS) listed spring Chinook salmon Oncorhynchus tshawytscha and winter steelhead O. mykiss in the upper Willamette River Evolutionarily

Significant Unit (ESU) as threatened under the Endangered Species Act (ESA; NMFS 1999a;

NMFS 1999b). As a result, any actions taken or funded by a federal agency in the ESU must be evaluated to assess whether they are likely to jeopardize threatened and endangered species, or result in the destruction or impairment of critical habitat. Several hatcheries produce and release hatchery salmonids in the upper Willamette Basin (Figure 1), which may impact wild populations of listed species. All hatcheries are operated by the Oregon Department of Fish and

Wildlife (ODFW) and are funded (50–100%) by the U.S. Army Corps of Engineers (USACE).

Potential risks of artificial propagation programs have been widely debated (e.g. Kostow and

Zhou 2006; Levin and Williams 2002). Risks include disease transfer, competition for food and spawning sites, increased predation, increased incidental mortality from harvest, loss of genetic variability, genetic drift, and domestication (Steward and Bjornn 1990; Hard et al. 1992; Cuenco et al. 1993; Busack and Currens 1995, and Waples 1999). Hatcheries can also bolster spawner abundance—a critical consideration for those populations on the verge of extirpation—by providing a genetic reserve, and by providing marine-derived nutrients to streams (Steward and

Bjornn 1990; Cuenco et al. 1993). Recent work, however, has shown that some hatchery fish tend to have lower reproductive success than wild fish even when broodstocks are largely comprised of wild fish (Araki et al. 2007) and productivity parameters are depressed when large numbers of hatchery salmonids mix with wild fish (Chilcote et al. 2012). However, reproductive success studies focused specifically on spring Chinook salmon yielded conflicting results with some suggesting lower reproductive success for hatchery Chinook salmon (Williamson et al.

2010), especially males (Sard et al. 2015), and others showing little difference between hatchery-and natural-origin fish (Hess et al. 2012).

The objective of this project is to conduct baseline monitoring of returning adult fish and to evaluate the potential effects of hatchery programs on naturally spawning populations of spring

Chinook salmon and winter steelhead in the upper Willamette River basin. Restoration of spring

Figure 1. The Willamette Basin with major dams, hatcheries, and fish collection facilities.

Chinook salmon under the ESA and the implementation of ODFW’s Native Fish Conservation

Policy require monitoring the number of hatchery and wild fish that comprise the spawning populations in the Willamette Basin. The Willamette Project Biological Opinion identified the need to reduce hatchery fish spawning in the wild to “the lowest extent possible (0–10%)”

(NOAA 2008).

In the Willamette Basin upstream of Willamette Falls (Figure 1), there are four distinct spring

Chinook salmon hatchery programs (North Santiam [Stock 21], South Santiam [Stock 24], McKenzie [Stock 23], and Middle Fork Willamette [Stock 22]) that are managed as integrated programs meant to provide ESA conservation benefits, and to help meet harvest objectives consistent with survival and recovery of the Upper Willamette River Evolutionary Significant

Unit (ESU).. Hatchery stocks, as well as all naturally-spawned spring Chinook salmon in the

Upper Willamette Basin, are included in the ESU.

The Upper Willamette Summer Steelhead Hatchery Program is managed to provide fish for sport fisheries and to replace loss of fisheries caused by habitat and passage loss/degradation in the

Willamette Basin and other lower Columbia River basins. The hatchery program currently includes annual smolt releases into the North Santiam, South Santiam, McKenzie, and Middle

Fork Willamette rivers. Lack of access to historical habitat and degradation of remaining habitat below the dams, especially in the North and South Santiam (the “core” populations) are the key limiting factors shared between winter steelhead and spring Chinook salmon (NOAA/NMFS

2008). In addition, summer steelhead are not native to the Willamette Basin upstream of

Willamette Falls and a third, unique, limiting factor is the potential for competition, predation and genetic introgression from out-of-ESU hatchery fish interacting with and spawning in the wild with the native winter-run. Summer steelhead were first introduced to the South Santiam

River to mitigate for lost winter steelhead production in areas inundated by Foster and Green

Peter reservoirs. The scope of work actually directed towards evaluating risks posed by summer steelhead is much smaller than that directed towards issues faced by spring Chinook.

Some work that has occurred focused upon interactions between hatchery- and natural-origin steelhead in the upper Willamette. For example, genetic analyses on steelhead describing genetic structure in the upper Willamette (Van Doornik and Teel 2012) and levels of introgression between summer- and winter-run steelhead (Johnson et al. 2015; Van Doornik et al. 2015) have occurred. McMichael et al. (2013) conducted research on ecological interactions between winter-and summer-run steelhead in the South Santiam subbasin. Jepson et al. (2015) described migration patterns of adult winter- and summer-run steelhead from Willamette Falls (Oregon

City) to their final destinations and used those data to estimate run size, spatial overlap, and temporal overlap in natal streams in the Upper Willamette.

Other steelhead work focused on the potential for reintroduction of steelhead above project dams. For example, Beeman and Adams (2015) used active tagging to study downstream passage of juvenile steelhead in the North Santiam subbasin and Hughes et al. (2016) performed similar work in the South Santiam subbasin. Johnson et al. (2016) used passive tagging to evaluate steelhead passage from releases above and below Detroit Reservoir. Romer et al. (2016) reported on juvenile steelhead migration into and out of Willamette reservoirs. Finally, Noakes et

al. (2015) have developed procedures for rearing wild-broodstock steelhead that seems likely to contribute to using the hatchery programs to reintroduce steelhead above, in particular, Detroit

Dam in the North Santiam.

This report fulfills requirements under Task Orders W9127N-12-2-0004-4009 and W9127N-10-

2-0008-0036 covering activities of May 2015–September 2017, that were implemented by

ODFW on behalf of the Corps to assist with meeting requirements of the reasonable and prudent alternatives (RPAs) and measures prescribed in the Willamette Project Biological Opinion

(BiOp) of July 2008 (NOAA/NMFS 2008). The Corps provided funding to continue ongoing monitoring activities and initiate long-term planning. The conceptual relationship between spring

Chinook salmon prioritized objectives, RPAs, and work tasks is depicted in Figure 2. The conceptual framework provided in Figure 2 combines objectives for monitoring compliance with

BiOp RPA’s that can be broadly categorized as belonging to either mitigation goals or conservation goals. Mitigation goals relate to characteristics of the hatchery programs such as species, numbers, size, location, and timing of fish released from hatcheries as well as harvest rates for returning adults. Conservation goals relate to, especially, undesirable escapement of hatchery-origin fish into natural spawning areas as well as intentional transport of natural- and hatchery-origin fish into (usually) spawning areas blocked by high-head dams. The conservation goals associated with escapement of hatchery-origin adults relate to primarily increasing natural origin abundance through outplanting, while minimizing genetic risks posed by the hatchery programs and, secondarily, potential but poorly understood ecological risks such as density-dependent disease risks to wild fish. Also, while harvest rates are clearly associated with mitigation goals, the existence of active fisheries with comingled natural- and hatchery-origin fish means some natural-origin fish will die from hook-and-release mortality, an issue better related to conservation concerns. Additional conservation issues are associated with potential ecological interactions between juvenile natural- and hatchery-origin fish. Research into ecological interactions at juvenile life history stages is not specifically part of the existing monitoring program but the issue is discussed as part of the long-term planning efforts noted above. Appendix 1 provides material on the various metrics associated with conservation and mitigation goals.

The ultimate goal of ODFW’s Hatchery Research, Monitoring and Evaluation (HRME) program is to inform decisions on operation of the USACE Willamette Valley Hatchery Mitigation

Program so that mitigation goals are met while minimizing negative impacts on naturally-produced, listed species, thereby promoting their conservation and recovery. Progress towards the ultimate goal will follow achievement of three overarching objectives:

1. Develop and maintain hatchery broodstocks to meet harvest goals and assist with implementation of the Upper Willamette Conservation and Recovery Plan for Chinook

Salmon and Steelhead, while complying with the existing genetic guidelines (Hatchery

Genetic Management Plans);

2. Rear and release high quality hatchery fish to minimize impacts on naturally-produced fish and promote conservation and recovery of listed species;

3. Manage adult returns to minimize impacts on naturally-produced populations and to aid in recovery goals.

Additional work supported by the hatchery program relates specifically to reintroduction of adult salmonids above projects. The objectives for that work are to:

1. Provide information on spawning distribution, abundance, origin (hatchery vs. wild), and pre-spawning mortality for adult Chinook salmon upstream of Detroit, Foster, Cougar, Fall Creek, Lookout Point and Hills Creek reservoirs. We will summarize adult Chinook and steelhead annual abundance for hatchery and natural origin returns for each population of concern, based on results from the carcass and redd surveys, adult trap counts, video counts, and other data as they become available (e.g. radiotelemetry, genetic sampling, etc.). This objective is intended to achieve the basic evaluation needs associated with a number of project concepts to “determine the effects of release date, outplanting site, and handling and transport protocols” on reproductive performance of outplanted Chinook above projects dams. Importantly, achieving this objective will also inform the ongoing and proposed work associated with direct measurement of reproductive fitness based on an adult to adult pedigree approach.

2. Use the data acquired under Objective 1 in combination with results from other ongoing and proposed work to assess relationships among characteristics of the spawning population and other relevant variables and to provide recommendations for conducting outplanting operations to support spawning success and use of quality habitat by spawners. This objective is intended to contribute to achieving APH-09-04 SYS-2, i.e.

“provide recommendations on management of adult UWR Chinook.”

Figure 2. Relationship between Prioritized Objectives, Reasonable and Prudent Alternatives (RPAs), Proposed Actions (PAs), and Work Tasks conducted for spring Chinook hatchery programs in the Upper Willamette Basin.

Section 1.1 Tasks

Task 1. Conduct surveys to determine the abundance, distribution and origin (hatchery or naturally produced) of spring Chinook salmon on the spawning grounds of each subbasin population (objectives addressed: SCS 4 and SCS 5).

The purpose of this task is to describe the abundance, distribution, and composition (i.e., hatchery vs. natural origin fish) of adult spring Chinook salmon returning to spawn in Upper

Willamette Basin tributaries. This task aims to describe, at varying spatial scales (Appendix 2), the population of adult returns with respect to: run size and timing, numbers of natural and hatchery origin fish collected for broodstock and outplanting, peak spawning dates, redd distribution and density, estimated natural spawner abundance, the proportion of hatchery origin fish on spawning grounds (pHOS), pre-spawning mortality (PSM) on spawning grounds, the age structure of the natural spawning population, hatchery stray rates, and harvest rates. To accomplish this, we employed a variety of data collection methods, such as monitoring the number of adipose fin-clipped and unclipped adults arriving at dams and fish collection facilities, tracking the fate and disposition of fish entering traps and transported to hatcheries, conducting redd and carcass surveys on spawning grounds, sampling carcasses that were spawned at hatcheries, and compiling fish recapture data from the Regional Mark Information System

(RMIS) and the ODFW CWT release and recovery database (CWT-Fish). Ultimately, the intent is to determine if mitigation goals have been met for harvest, broodstock, and conservation

(reintroduction/outplanting). Establishing useful numeric goals for abundance and disposition of returning hatchery adults, goals that are agreed upon by the managers and Action Agencies, is an important process that is ongoing.

The spawning ground surveys conducted as part of Task 1 are aimed at characterizing the naturally-spawning population in accessible stream reaches downstream of USACE dams.

Similar spawning ground surveys were conducted above these dams as well but are included under Task 4 as described below. This separation has been made to specifically monitor and evaluate outplanting efforts in stream reaches blocked by dams and the potential of these reaches to serve for reintroduction purposes and as sanctuaries for wild fish populations. Comparisons of estimated spawning population parameters (e.g., peak redd counts, redd densities, pHOS, and

PSM) between spawning areas downstream and upstream of USACE dams are a useful tool for identifying reaches with relatively greater habitat potential and for evaluating hatchery management practices. Such comparisons are also addressed under Task 4.

Task 2. Conduct biological monitoring of hatchery broodstock (objectives addressed: SCS 1, SCS 2, and SCS 3).

The purpose of this task is to obtain estimates of origin (hatchery, wild, strays), body size, age structure, run timing, and spawn timing of hatchery broodstock. The intent is to ensure that broodstock collected and spawned in each hatchery program adequately meet mitigation, conservation, and recovery goals, and comply with existing guidelines being developed in each

Hatchery Genetic Management Plan (HGMP).

Task 3. Conduct biological monitoring of fish rearing in hatcheries and at release (objectives addressed: SCS 6, SCS 7, and SCS 9).

This task involves monitoring of fish performance both in-hatchery (survival, growth) and post-release (migratory performance; smolt-adult return [SAR]) and includes monitoring of timing and number of juveniles released by species and stock for each hatchery.

Task 4. Estimate the relative survival of outplanted fish and abundance of outplanted fish that spawn above USACE dams (objectives addressed: SCS 4 and SCS 5).

The purpose of this task is to monitor and evaluate outplanting efforts in each of the four major

Upper Willamette River subbasins. As mentioned above, the components of this task include:

conducting spawning ground surveys in reaches where fish have been outplanted; collecting data on spawning population parameters (e.g., peak redd counts, redd densities, pHOS, and PSM);

and analysis of spawning population parameters at varying spatial scales (Appendix 2). In addition, genetic sampling of outplanted fish is conducted in support of ongoing parentage studies at several projects, and a study on the genetic diversity of the Willamette spring Chinook salmon populations (Johnson and Friesen 2014).

Section 1.2 Spring Chinook Salmon Production Program Goals

Section 1.2.1: Broodstock Collection and pNOB Goals

The intent of broodstock collection protocols at the UWR hatcheries is to sequester enough broodstock to ensure enough adults to support all mitigation and conservation requirements (e.g.

harvestable fish, broodstock for the next generation, fish for outplanting, etc.) while also ensuring that the fish taken for broodstock are phenotypically similar to naturally-produced fish

(e.g. run timing, spawn timing, age structure, etc.).

Adult collection began in May and occurred into October at all facilities. Collection protocols varied by hatchery program. In the North Santiam subbasin, broodstock were collected at the

Minto Fish Collection Facility. In the South Santiam subbasin collection occurred at the Foster

Fish Collection Facility. In the McKenzie subbasin hatchery fish volunteered to the ladder on site at the hatchery. In the Middle Fork Willamette subbasin fish were captured at the Dexter Dam trap and transported by truck to the Willamette Hatchery further upstream. At capture adults are generally anesthetized with CO2 or Aqui-S to facilitate handling.

Spawning protocols were relatively uniform across hatcheries whereby adults were crowded, anesthetized, and checked for ripeness. Unripe fish were returned to holding areas and ripe fish were killed and bled. Eggs were removed from females into spawning buckets and fertilized using a 1:1 sex ratio.

Once the subbasin HGMP’s are approved (anticipated spring 2018), incorporation of natural-origin fish into the broodstocks will occur at 5% or more per year.

Section 1.2.2: Outplanting and pHOS Protocols and Goals

Outplanting protocols varied widely throughout the subbasins. When the outplant goal is focused on disposition of hatchery-origin fish (as in the North Santiam and Middle Fork

Willamette subbasins), outplanting generally begins relatively early in the run when it becomes apparent that the run size will be adequate to provide sufficient broodstock, and ends late.

Exceptions exist at the McKenzie Hatchery and Dexter Trap when ongoing research projects require outplants at specific times either to test a particular practice (Dexter trap: early outplants) or to achieve specific escapement goals (McKenzie Hatchery: outplanting above Cougar Dam).

When outplanting is focused upon the disposition of unclipped fish (South Santiam River, Fall

Creek and, ideally, the Cougar Dam trap in the South Fork McKenzie River) then outplanting begins and ends with the capture of the first and last unclipped adult fish.

In the North Santiam River outplanted fish (hatchery-origin, predominantly) were captured at

Minto and trucked to the Breitenbush and North Santiam arms of Detroit Reservoir. On the

South Santiam River only unclipped fish captured at the Foster Dam trap were outplanted at locations ranging from near the head of reservoir to multiple locations further upstream. On the

McKenzie River outplants from the McKenzie Hatchery were exclusively adipose-clipped fish taken to the South Fork McKenzie River to complement mostly unclipped fish transported from the Cougar Dam adult trap in support of a research project evaluating productivity of hatchery-and natural-origin spawners (Banks et al. 2014). Outplanting in the Middle Fork Willamette subbasin includes releases at several locations. Adult fish from Dexter Dam trap are outplanted into the Middle Fork Willamette above Hills Creek Dam to support recovery efforts for bull trout

Salvelinus confluentus, and into Little Fall Creek, a tributary entering Fall Creek and the Middle

Fork Willamette River below Dexter Dam. Adults from both the Dexter trap and Willamette

Hatchery are also outplanted in the North Fork Middle Fork Willamette River above Lookout

Point Reservoir in various locations to support ongoing research into causes of prespawning mortality (Schreck et al. 2014; Mann et al. 2012). Finally, natural-origin adult Spring Chinook and other native fish species captured at the Fall Creek Dam trap are outplanted by USACE biologists above Fall Creek Reservoir and ODFW conducts spawning ground surveys to continue collaborative recovery efforts there.

Section 1.2.3: Marking and Tagging of Hatchery Chinook Salmon

Adult hatchery fish are identified using a combination of marks that were applied to the juveniles prior to release. All hatchery-origin Chinook salmon receive adipose fin clips and a secondary thermal otolith mark. In addition, a portion of the juvenile hatchery Chinook salmon are released with coded-wire tags (CWTs). Specific information on CWT releases from RMIS is available online at http://www.rmpc.org/. On average, 687,000 CWT spring Chinook salmon are released http://www.rmpc.org/ into the basin annually (Shaun Clements, ODFW, pers. comm.) with more than 100,000 tagged fish typically released from each hatchery.

Section 1.3 Willamette River Escapement

Escapement estimates are obtained by a combination of video counts and spawning ground surveys. Counts of clipped and unclipped Chinook salmon and steelhead trout (and other species) are obtained as fish enter the Upper Willamette River (at Willamette Falls) and at two other locations in subbasins (Bennett dams on the North Santiam River, and Leaburg Dam on the

McKenzie River).

Section 2: Methods

Section 2.1 Estimating Spawner Parameters: Distribution, Abundance, and Proportion of Hatchery- and Natural-Origin Chinook Salmon

Section 2.1.1: Monitoring Adult Returns

The majority of the spring Chinook salmon adults that pass Willamette Falls enter the North

Santiam, South Santiam, McKenzie, and Middle Fork Willamette subbasins to spawn. Returns specific to each subbasin were monitored through spawning surveys and at fish ladders or collection facilities in each of these four subbasins. Depending on management objectives for each of the subbasin hatchery programs, fish captured at collection facilities were retained for broodstock, outplanted above USACE dams, recycled downstream for additional angling opportunities, sold to offset costs of fish transport, donated to tribes, or used for stream enrichment.

2.1.1.1 Spawner Surveys: We surveyed four major eastside tributaries (North Santiam, South

Santiam, McKenzie, and Middle Fork Willamette rivers) in the Willamette Basin upstream of

Willamette Falls (Figure 1) by boat and on foot to count spring Chinook salmon carcasses and redds following established protocols (Boydstun and McDonald 2005; Schroeder et al. 2007;

Gallagher et al. 2007; Kenaston et al. 2009; Cannon et al. 2010). We counted redds from late

August through October to encompass the peak times of spawning based on data from surveys conducted in past years. Detailed maps of the subbasins are provided in the Results section and descriptions of the reaches are provided in Appendix 3.

For boat surveys we used rafts with elevated viewing towers on large river sections. On some river sections the raft stayed on one side of the river (with the other bank covered simultaneously with another raft or on subsequent surveys) over the entire length of the section to count redds, whereas on other sections the raft crossed the river to count redds on both sides. Similar techniques were used on medium-sized rivers except that we used small rafts with viewing platforms lacking elevated towers. In tributary reaches that were inaccessible to walking surveys we used inflatable kayaks. All boat surveys were conducted in a downstream direction except that a small number of reaches required paddling or rowing upstream a short distance (<100 m) when the only boat launch site was below a reach break that could not be safely passed.

For walking surveys, a stream was classified as “medium” if the surveyor had to cross the stream to observe areas on the other side, or “small” if the surveyor could observe both sides of the stream without crossing (Schroeder et al. 2005). Observers counted redds and attempted to record global positioning system (GPS) coordinates for each redd in a river section. If a GPS signal could not be obtained at a particular location, the redd was still counted. All walking surveys were conducted in a downstream direction except in a few instances when a surveyor completed a section and had the opportunity to assist a partner in a reach by surveying upstream.

2.1.1.2 Carcass Sampling: During spawning surveys all carcasses that could be recovered by hand or with long-handled gaffs were examined for adipose fin clips to determine the proportion of hatchery fish on spawning grounds. We measured carcasses (cm fork length; FL), determined sex, and estimated the proportion of remaining eggs in female fish to document pre-spawning mortality (details in section 2.1.2.5, below). Carcasses in water too deep to permit recovery or too degraded to permit inspection were recorded as unprocessable. We collected otoliths from processable carcasses without fin-clips to differentiate unclipped hatchery fish from naturally-produced fish using results from otolith analyses performed by the Washington Department of

Fish and Wildlife Otolith Laboratory (see Proportion of Hatchery Spawners, below). We used hand-held detectors manufactured by Northwest Marine Technology, Inc. (Tumwater, WA) to determine if carcasses with or without adipose fin clips had CWTs. Fish with CWTs and without fin clips might simply be mis-clipped fish, fish with regenerated adipose fins, or fish from

“double-index release groups” (intentionally released without a fin clip for fishery management purposes). We collected the snouts of tagged fish and put them in plastic bags with individually numbered labels. Tags were removed and identified at the ODFW Clackamas Fish Identification

Laboratory to establish the origin of tagged fish.

2.1.1.3 Monitoring Fish Passage at Bennett and Leaburg Dams: We used underwater video cameras to monitor net upstream movement of salmon, steelhead and other fish species through ladders at the Bennett dams on the North Santiam River and Leaburg Dam on the McKenzie

River (Figure 1). The video equipment uses software (FishTick, SalmonSoft, Inc., Portland, OR) that automatically scans and records fish movement and creates video files from these images.

The captured video images were reviewed and species, presence or absence of an adipose fin clip, direction of movement (upstream or downstream) were noted so that the net upstream movement of spring Chinook salmon and steelhead trout by presumed hatchery or wild origin could be estimated. Other fish including coho salmon O. kisutch, lamprey Entophenus and

Lampetra spp., and bull trout were also counted. Counts of clipped and unclipped Chinook salmon were later adjusted using otolith data to get estimates of actual hatchery- and natural-origin fish above the counting stations. We attempted to operate the video systems continuously throughout the migration season. On the rare occasions when a video system failed we estimated the number of fish that may have passed during these outages based on simple linear extrapolation of fish counts recorded during the time when the video equipment was operating normally, generally on the same day.

2.1.1.3.1 Video Monitoring at Bennett Dams: Passage of spring Chinook salmon (and other species) occurred at both Upper and Lower Bennett dams. The video monitoring system at upper

Bennett Dam operated continuously and, at Lower Bennett Dam, the system operated from April through December in both 2015 and 2016.

2.1.1.3.2 Video Monitoring at Leaburg Dam: Passage of spring Chinook salmon through the two fishways at Leaburg Dam was continuously monitored with video recording equipment. We recorded fish passage at both the left-bank and right-bank fish ladders.

2.1.1.4 Monitoring Harvest: Harvest estimates were obtained by summarizing ODFW online harvest reports available at http://www.dfw.state.or.us/resources/fishing/sportcatch.asp .

Section 2.1.2: Data Analysis

2.1.2.1 Peak Redd Counts and Peak Redd Densities: The peak redd count is the maximum number of redds observed in each survey section over the course of the survey season and represents an estimate of the total number of redds constructed by Chinook salmon in each section. When redd counts differed between initial surveys and resurveys conducted to evaluate variability in redd counts (described below), the resurvey counts were used to replace the initial counts. Peak redd densities were calculated by dividing the peak redd count by the length (km) of each section.

http://www.dfw.state.or.us/resources/fishing/sportcatch.asp

2.1.2.2 Spawn Timing: We compared spawn timing of naturally-spawning fish and broodstock spawned in the hatcheries. The intent of the work was to determine if the spawn timing in the hatchery differed from the average spawn timing in the river in recent years. We estimated peak spawning of naturally spawning fish by fitting a sigmoid curve to the cumulative redd counts over time for multiple years in each subbasin. The date associated with the inflection point on the fitted sigmoid curve was assumed to represent the average date of the maximum rate of redd construction in each subbasin; that is, average peak spawn timing. We then compared the average spawn timing in the rivers to the spawn timing in the hatcheries. Average spawn timing in the hatcheries was calculated as the weighted mean date of spawning, weighted by the number of broodstock spawned on each spawn date.

2.1.2.3 Spawner Abundance Estimates: We used the peak count expansion method described below to estimate total spawner abundance. We made the three following assumptions: 1) that the peak redd count in any reach of interest adequately reflected the relative abundance of fish that spawned in that reach; 2) each redd was constructed by one female; and 3) each female spawned with 1.5 males (Gallagher et al. 2007; Boydstun and McDonald 2005).

A spawner abundance estimate (A) derived from the peak count expansion method was calculated by the following equations:

A = Fspawn + Mspawn, where

Fspawn = number of spawning females = Reddpeak/Reddfemale;

Reddpeak = peak redd count, and Reddfemale = number of redds/spawning female = 1, and

Mspawn = number of spawning males = Fspawn × 1.5.

We then parsed the total spawner abundance estimate into hatchery and wild spawning cohorts by using the pHOS estimates derived from carcass sampling with adjustments based upon otolith analyses. Clearly there is a large effect that this string of assumptions has on the accuracy of the estimates of spawner abundance, and there are no estimates of precision associated with redd count expansions. The values for spawner abundance and redd count expansion should therefore be used with caution.

2.1.2.4 Proportion of Hatchery Spawners: We combined counts of clipped and unclipped fish wherever they were encountered (at video counting stations, during spawner surveys, and during monitoring of adult fish entering hatchery traps) with validation of hatchery or wild origin from otolith data to derive the proportion of hatchery spawners (pHOS) at various spatial scales. The spatial scales included basin-wide, by subbasin, above and below dams, and, in some cases, by river reach. To differentiate between hatchery and wild Chinook salmon and to implement a selective fishery, all hatchery spring Chinook salmon in the Willamette basin, beginning with the

1997 brood year, have been marked with adipose fin clips, CWTs, or both. Also, thermal marks were (and are) induced in the otoliths of all hatchery Chinook salmon released in the basin to provide an additional mark for identifying unclipped hatchery fish. Some juvenile Chinook salmon are inadvertently released without a fin clip at a rate that varies by hatchery and by brood year (Schroeder et al. 2005). However, the percentage of unclipped fish in hatchery releases has decreased in recent years with the implementation of automated fin-clipping systems. Other factors that contribute to the return of unclipped hatchery fish include the release of unclipped hatchery fish with CWTs (double-index), and natural regeneration of partially clipped adipose fins.

We estimated the proportion of natural-origin (wild) and hatchery-origin fish by adjusting counts of clipped and unclipped carcasses after examining otoliths collected from the unclipped carcasses recovered on the spawning grounds. We collected samples from adult spring Chinook salmon carcasses without fin clips on spawning grounds (North and South Santiam, McKenzie, and Middle Fork Willamette rivers). Otoliths were collected and placed into individually numbered vials. The samples were subsequently sent to the otolith laboratory operated by

Washington Department of Fish and Wildlife for analysis of thermal marks. The reach-specific proportion of hatchery origin spawners (pHOS) was derived from the counts of fin-clipped fish

(AD), unclipped thermally-marked fish (UTM) and total count of fish examined (TOT) using the equation:

pHOS = [AD + UTM]/TOT).

The reach-specific pHOS estimates were then applied to the reach-specific spawner abundance estimates and the products summed to yield subbasin-wide pHOS estimates weighted by spawner abundance (reach-specific redd counts).

In future years, after HGMPs are approved and natural-origin fish can be incorporated into brood, we will also use the otoliths to adjust estimates of the proportion of natural-origin brood

(pNOB) by using the counts of non-thermally-marked unclipped broodstock (WILDB), and the total number of broodstock (TOTB) using the following equation:

pNOB = WILDB/TOTB.

2.1.2.5 Pre-spawning Mortality: We surveyed major tributaries of the Willamette basin, both above and below project dams, by boat and on foot to estimate pre-spawning mortality (PSM) based on the proportion of unspawned female salmon carcasses observed. Female carcasses with intact or relatively intact skeins (i.e. greater than 50% eggs remaining) were considered unspawned. The 50% threshold is arbitrary but in practical terms virtually all female carcasses had either essentially no eggs remaining or completely intact skeins. The surveys were conducted in a manner identical to the spawner surveys (described above) but began in the summer prior to any spawning to permit observation of any early mortality that occurred as salmon reached spawning tributaries. Female carcasses were also checked for spawning success during the regular spawning surveys and redd counts through early October so that pre-spawning mortality could be assessed over the entire run. For every female salmon carcass that could be recovered during the pre-spawning and spawning surveys the gut cavity was cut open to visually judge the relative abundance of eggs. We then calculated PSM by dividing the number of unspawned female carcasses by the total number of female carcasses where spawning status was observed.

For the purpose of discussion in this document we arbitrarily categorize PSM as “low”, “moderate”, and “high” when estimates were <20%, 20% to 50%, and >50%, respectively.

2.1.2.6 Straying of Hatchery Fish: In the Willamette basin a “stray” is defined as any hatchery fish that does not return to its hatchery of origin and either spawns naturally or is encountered at another hatchery. In addition to estimating pHOS (described above) in each subbasin we estimated the contribution to pHOS of strays from outside the subbasin into which the juveniles were originally released.

We used handheld tag detectors to check for CWTs in carcasses recovered during surveys.

Snouts were removed from carcasses with CWTs, tagged, bagged, and frozen for later processing. Frozen snouts were delivered to ODFW’s Clackamas Fish Identification Laboratory, where the wire was extracted and decimal codes read to identify the hatchery stock and release site.

Section 2.2: Reintroduction Efforts We intercepted salmon designated for outplanting (and broodstock collection, fish sales, fish donation, and stream enrichment) at adult fish traps at the left (south) bank ladder of the Leaburg

Dam, Dexter Dam, Foster fish collection facility (FCF) and the Minto FCF. Biological data (fork length, sex, scales, presence of tags or fin clips) and specimens (otoliths [from lethally sampled fish], DNA) were collected. The count of adult fish outplanted above project dams was used as the initial basis for adult abundance above dams, modified by estimates of abundance, PSM, and distribution based on spawner surveys (described below).

We collected biological samples and data (sex, DNA sample, fin clips, date, outplant location) from all Chinook salmon that were outplanted. We subsampled outplanted fish for FL and scales at a rate intended to yield a sample size of approximately 100 fish.

Section 2.3: Broodstock Sampling

2.3.1 Collection, Spawn Timing, Composition, and Disposition of Broodstock: Traps are operated for each of the Willamette spring Chinook salmon hatcheries to collect broodstock.

Chinook salmon are also trapped at Leaburg Dam and Leaburg Hatchery and then transported to

McKenzie River Hatchery. Disposition of collected salmon is determined at each hatchery by presence or absence of an adipose fin clip and recorded. At each hatchery on each spawning date samplers recorded number of fish spawned by sex, length of broodstock, and obtained samples from fish as required (scales, otoliths, DNA, CWTs).

Section 2.4: Within-Hatchery Monitoring

2.4.1 Adult Monitoring: The bulk of within-hatchery monitoring involved tracking the fate and disposition of adult fish at each hatchery or FCF. The data were acquired by a combination of

(1) direct sampling by HRME staff at each hatchery during outplanting and spawning activities,

(2) queries of the data provided by the hatchery managers to the Hatchery Management

Information System (HMIS), and (3) interviews with the hatchery managers to verify portions of the data that were provided to HMIS.

2.4.2 Juvenile Monitoring: We obtained summaries of the number of fish released, rearing locations, release locations and size at release for both summer-run steelhead and Chinook salmon by querying HMIS for those data. We also queried RMIS to obtain information on

Chinook salmon liberation dates and release locations for CWT fish from Willamette hatcheries.

Steelhead have not been released with CWTs since the 1980s.

Other juvenile monitoring involved compiling hatchery records for size distributions and tag retention data for fish just prior to release.

Section 3: Results

Section 3.1: Abundance, Distribution, Spawn Timing and Composition of Naturally Spawning Adult Spring Chinook Salmon

Section 3.1.1 Adult Returns:

In 2015 the total count of spring Chinook salmon ascending Willamette Falls (Table 1) was

53,088 (51,046 adults and 2,042 jacks). Of the adults, 42,098 were adipose-clipped and 8,948 were unmarked. Of the jacks, 1,561 were adipose-clipped and 481 were unmarked. The run at

Willamette Falls was dominated by hatchery returns, with an estimated 82% of the 2015 run originating from Willamette hatcheries based on observed adipose mark rates (available online at http://www.dfw.state.or.us/fish/fish_counts/willamette%20falls.asp).

In 2016 the total count of spring Chinook salmon ascending Willamette Falls (Table 1) was

32,478 (30,317 adults and 2,161 jacks). Of the adults, 23,686 were adipose-clipped and 6,631 were unmarked. Of the jacks, 1,770 were adipose-clipped and 391 were unmarked. The run at

Willamette Falls was dominated by hatchery returns, with an estimated 78% of the 2016 run originating from Willamette hatcheries based on observed adipose mark rates (available online at http://www.dfw.state.or.us/fish/fish_counts/willamette%20falls.asp).

Spring Chinook salmon adults and jacks were collected at Upper Willamette Basin facilities beginning in late May or early June at all facilities, and concluding in early September through early October at the North Santiam, South Santiam, McKenzie, and Dexter facilities.

Table 1. Number of fish passing Willamette Falls by month, 2015 (top) and 2016 (bottom). CHS – spring Chinook salmon; STS = summer steelhead; STW = winter steelhead;

CHF = fall Chinook salmon; Mk = finmarked (adipose clipped); Nm = non-finmarked; JK = jack.

CHS STS STW CHF Coho Salmon

Month Adult Mk

Adult Nm

JK

Mk

JK

Nm

STS

Mk

STS

Nm

STW

Mk

STW

Nm

Adult Mk

Adult Nm

JK

Mk

JK

Nm

Adult Mk

Adult

NM

JK

Mk

JK

Nm

Jan 0 0 0 0 0 0 2 792 0 0 0 0 0 1 0 0 Feb 0 0 0 0 0 0 12 1,351 0 0 0 0 0 0 0 0 Mar 175 71 0 0 137 0 0 1,047 0 0 0 0 0 0 0 0 Apr 14,422 3,018 332 89 430 0 0 750 0 0 0 0 0 0 0 0 May 22,180 4,324 941 302 1,226 0 0 293 0 0 0 0 0 0 0 0 Jun 4,605 1,210 253 75 705 87 0 0 0 0 0 0 0 0 0 0 Jul 563 236 22 9 99 13 0 0 0 0 0 0 0 0 0 0 Aug 153 89 13 6 164 13 0 0 70 53 7 4 3 2 14 0 Sep 0 0 0 0 691 98 0 0 213…