This is an exemplary, innovative project. We thank the proponents for submitting a revised proposal as well as a point-by-point response addressing all of the topics identified in our preliminary comments. In our preliminary comments, we requested responses on the following topics:

1. Compatibility with NOAA status assessment
2. MPG and ESU identification
3. Alternative approaches

The proponents added three appendix tables (A1-A3) in response to our request for more detail about how data generated by this project contribute to the abundance, productivity, or diversity measures used by NOAA for hierarchical assessment of ESU viability.

Table A1 highlights the extent to which this project supports assessments in a variety of other projects by genetically distinguishing sex and the hatchery origin of unmarked fish (that would otherwise be assumed to be natural origin). These and other data contributed by the project are clearly important for improving estimates of pHOS, natural spawner abundance, and indices of genetic and life history diversity.

Tables A2 and A3 support the ISRP’s preliminary comments that the genetic stock groupings to which fish can be assigned by this project often do not exactly match the TRT populations of steelhead and Chinook salmon, and in some cases even the MPGs of Chinook salmon.

Moreover, the proponents’ response does not indicate how estimates of abundance or productivity measured at Lower Granite Dam for genetic stocks comprising mixtures of TRT populations are (or might be) adjusted to provide the spawner-to-spawner data typically required for NOAA's VSP and ESU status assessments. We therefore conclude that the proposal has somewhat overemphasized the utility of this project for the NOAA viability assessments of TRT populations. That said, this project also supports a basinwide effort to implement a variety of approaches for monitoring status and trends at multiple hierarchical levels. To maximize efficiency and effectiveness of that overall effort, it remains important to periodically review and refine the approaches being developed in this and other projects. The ISRP urges the proponents to consider, in future work plans, whether a greater focus on SNPs associated with adaptive traits might improve the capability to resolve TRT populations and MPGs.

Preliminary ISRP report comments: response requested

Response request comment:

We commend the proponents for a clearly written, well-organized and well-justified proposal. The standardization and application of accurate methods of parentage-based tagging (PBT) and genetic stock identification (GSI) are remarkable achievements by this project (in collaboration with project 200890700). The resulting single nucleotide polymorphism (SNP) baselines for Snake River steelhead and Chinook salmon have enabled (1) cost-effective and routine monitoring of spatial and temporal trends in diversity and genetic structure of natural-origin Snake River populations; (2) estimation of stock composition of harvests in mainstem fisheries and escapements past Lower Granite Dam; and (3) evaluation of proportionate natural influence (PNI) of integrated hatchery programs in Idaho.

The proposal states that the project contributes significantly to NOAA and other investigators by providing data for viability assessments of independent populations defined by the Technical Recovery Teams (“TRT populations”), major population groups (MPG), and evolutionarily significant units (ESU). However, the ISRP was unable to understand how, or the extent to which, this project informs these status assessments. The ISRP requests the proponents to address the following points in a revised proposal, and to provide a brief point-by-point response to explain how and where each issue is addressed in the revised proposal:

1. Compatibility with NOAA status assessment. Clarify how estimates from this project are used in the NOAA status assessments. In many cases, it seems that data generated by this project could only be used to estimate abundance and productivity of multi-population groupings returning to Lower Granite Dam. It seems that such estimates would not be adequate for viability assessments that typically rely on estimates of spawner abundance and spawner-to-spawner productivity for TRT populations.
2. MPG and ESU identification. List the steelhead and Chinook salmon populations by MPG and ESU for which this project provides data that are used by NOAA for viability status assessments.
3. Alternative approaches. If the project enables an alternative approach to viability assessment of abundance and productivity, then explain the alternative approach and compare it with the NOAA approach to demonstrate its utility.

Q1: Clearly defined objectives and outcomes

This project addresses management issues and uncertainties that are highly relevant to the Council's 2014 Fish and Wildlife Program, 2017 Research Plan, and High-Level Indicators as well as numerous subbasin plans. However, it is less clear how the project supports and contributes to ESA Recovery Plan objectives.

The current proposal combines objectives from two previous BPA projects to test the feasibility of using PBT and GSI to help manage hatchery and wild Snake River steelhead and Chinook salmon. The eight revised objectives are clearly specified, and most meet SMART criteria. An exception is Objective 2, which is not quantitative. Instead of just saying “discover new SNPs,” we suggest indicating more precisely the number of new SNPs, or the attributes of new SNPs, that need to be discovered for the project to achieve this objective.

In future proposals, it would be advantageous for the proponents to include additional objectives related to adaptive management and reporting, as this would serve to highlight other very successful outcomes from the project.

The project now comprises a well-established annual cycle of activities that is expected to continue for the foreseeable future.

Q2: Methods

The proposal includes a succinct but comprehensive overview of methods, which is organized appropriately by objectives, clearly explains the rationale for various approaches, and provides convenient links to details elsewhere (e.g., monitoringresources.org and references to associated projects and the primary literature). Sampling and analytical protocols are documented in more detail in the annual report for 2019. Standard methods are being used for statistical analyses, and confidence intervals or significance test probabilities are provided to support most conclusions. The proponents also use state-of-the-art techniques that were peer-reviewed in their primary publications, and hence, are considered scientifically appropriate.

A notable weakness in the methods section is the lack of detail about how data generated by the project contribute to the abundance, productivity, or diversity measures (i.e., viable salmonid population (VSP)) parameters used by NOAA for hierarchical assessment of ESU viability. VSP parameters used to assess the viability of TRT populations are typically based on multi-generational adult spawner-to-spawner data. This proposal does not describe how estimates of abundance or productivity measured at the Lower Granite Dam for stocks comprising mixtures of TRT populations are adjusted to provide the spawner-to-spawner data typically required for NOAA's VSP and ESU status assessments. It seems that in many cases, the scale of the abundance and productivity estimates provided from this project would not align sufficiently with that needed for assessments at the TRT population, MPG, or ESU level. In summary, the methods section should indicate more precisely which data generated by this project are actually used by NOAA for viability assessment and describe more clearly how these data are adjusted to meet (or circumvent) the requirement for spawner-to-spawner abundance and productivity estimates for TRT populations and the specific populations for which data are generated.

Q3: Provisions for M&E

A brief paragraph refers to a history of periodic adjustments to genetic marker panels, genotyping platforms, and statistical tools through regular meetings with genetic collaborators to evaluate results and discuss new proposals. Significant adjustments to improve cost-effectiveness include transitioning to:

• Absorptive chromatography paper that can hold 50–100 samples per sheet and occupies much less space than an equivalent number of ethanol-filled vials;
• 300-400 GTseq SNP genotyping panels for both Chinook and steelhead that provide near zero false-positive and false-negative rates at less cost than the original 96 samples X 96 SNP loci system;
• A 5-year rotating schedule for resampling populations to update GSI baselines;
• Whole-genome sequencing of pools of individuals (Pool-seq) to more cost-effectively estimate allele frequencies across the genome at the population scale; and 
• The (proposed) addition of microhaplotypes (multiple, tightly linked SNPs that exhibit contrasting allele frequencies across populations) to the existing GTseq SNP panels in hope of further improving GSI accuracy for steelhead.

The proposal states that project results are regularly evaluated and discussed during meetings with collaborators. Although this project adjustment process appears to be working well, the ISRP would like to see more explanation, in future proposals or the next annual report, of the decision process by which the proponents allocate effort and resources among objectives, such as finding new SNPs and updating and expanding baselines.

The project has excelled at sharing information and providing information to support management decision processes. Monitoring and research results are presented in annual IDFG and BPA reports, at various meetings (e.g., LSRCP, IDFG Anadromous Meeting, Steelhead Workshop, Coastwide Salmonid Genetic Conference), and in the primary scientific literature. Impressively, the proponents have authored or co-authored over 20 papers on work undertaken in this project. The Gantt chart (Fig. 19) clearly indicates the annual cycle of activities, and the table in section 8 (Relationships to other projects) helps to clarify roles and responsibilities in collaborations with six other projects.

Q4: Results - benefits to fish and wildlife

The proponents have worked collaboratively with CRITFC staff (project 200890700) to develop and standardize SNP panels for steelhead and Chinook salmon that cost-effectively integrate application to both PBT and GSI, and identify the genetic sex of each species. The ISRP noted an apparent inconsistency between pages 15 and 27 of the proposal which state the current Columbia-basin-wide panel contains “390 SNPs for steelhead and 299 SNPs for Chinook Salmon” versus “368 SNPs for steelhead and 343 SNPs for Chinook Salmon,” respectively, with Hess et al. (2020) cited in both instances.

On average, 4,900 steelhead and 12,000 Chinook salmon broodstock are sampled each year to create PBT baselines comprising all steelhead and Chinook salmon broodstock used in hatcheries throughout the Snake River basin. This effort allows the Snake River PBT program to genetically “tag” about 95% of 20 million steelhead and Chinook salmon smolts released annually.

Since 2018, the steelhead GSI baseline has represented 23 TRT populations and all 6 MPG. These steelhead collections are pooled to create 45 “GSI populations” for stock composition analysis and 10 “genetic stock” groups for reporting mixture proportions. The Chinook salmon baseline represents 31 of 41 TRT populations and all 5 MPG. The Chinook salmon collections are pooled to create 30 GSI populations for analysis and 6 genetic stocks for reporting.

PBT has now superseded coded-wire tagging (CWT) to estimate the harvest of hatchery Chinook salmon in the Snake River basin, although CWTs are still used to monitor ocean and downriver harvests, and to assess and compare alternative hatchery rearing and release strategies. Multiple year results have shown that PBT and CWT methods provide similar accuracy, but PBT can provide greater precision because of the larger number of “tags” available.

The proponents have used GSI and PBT in combination (working collaboratively with projects 199005500, 199107300, 198335003, and 201800200) to estimate abundance and stock composition of wild steelhead, spring/summer Chinook salmon, and fall Chinook salmon passing Lower Granite Dam. Abundance, productivity, and measures of genetic diversity are provided to NOAA as part of requirements to review the listing classification of Snake River steelhead and Chinook salmon at least once every five years.

Incorporating PBT with GSI has significantly improved the accuracy of wild escapement estimates at Lower Granite Dam by detecting untagged hatchery-origin fish that would otherwise be mistaken for wild fish, resulting in a significant overestimation of natural abundance. PBT analysis identified that, on average from 2014-2018, 19.6% of Chinook salmon and 8.3% of steelhead adults passing Lower Granite Dam were hatchery-origin, despite having no physical or mechanical marks. Similarly, a comparison of stock-specific abundance estimates for hatchery Chinook salmon returning in 2016 to 2019 revealed that the in-season PIT-tag method accounted for only 65% (averaged across all release groups) of the total detected by PBT.

The project also demonstrated that PBT should be used as part of long-term monitoring of proportionate natural influence (PNI) for integrated hatchery programs in Idaho. Analyses based on physical marks consistently overestimated PNI by overestimating the proportion of natural fish in both the natural spawning and broodstock components of the hatchery program compared to analyses based on PBT.

The proponents have fully addressed previous ISRP recommendations:

• Genetic data from PBT and GSI projects in the Columbia River basin (and throughout the Pacific Coast of North America) are now stored in FIshGen (McCane et al. 2018) and available to any lab running PBT projects;
• PBT data for PIT-tagged fish are now being linked to life history data from scale sampling at Lower Granite Dam (i.e., age at maturity) and migration behavior from PIT-tag detections at in-stream arrays (IPTDS). This step allows VSP metrics to be assessed at the scale of some TRT populations. However, because in-stream array coverage is not complete across the landscape, the method cannot be applied to all TRT populations.