200735500 - Determining the Accuracy of Adult Coho Salmon Population Estimates from a Random, Spatially Balanced design using Area-Under-the-Curve
Sponsor: Washington Department of Fish and Wildlife (WDFW)
Budgets: FY07: $100,192 | FY08: $83,798 | FY09: $87,990
Short description: Compare accuracy of AUC and mark/recapture population estimates for coho salmon.
Final Council recommendation (Nov 2006)
Funding category: Expense
Recommended budgets: FY07: $100,192 | FY08: $83,798 | FY09: $87,990
ISRP final recommendation: Fundable
This project is fundable or partially fundable after integration with coho spawning habitat assessments or an explanation of why the study does not need to be integrated (e.g., no evidence that habitat is limiting). There is a strong need to fill information gaps on the abundance of adult coho salmon. The proponents did not mention habitat and any issues surrounding the habitat. Further thought should be given to the choice of sample locations for the Area Under Curve (AUC) method. Ideally, some probabilistic sampling plan would serve best - if it is practical. There is a programmatic issue larger than this specific proposal in that many of the juvenile and adult stock assessment projects proposed by WDFW should have been combined and packaged together. There is a lack of integration, many different methodologies are being used, and standardized methods or statistical review are not apparent. Additional comments and question by the reviewers are listed below. Technical and scientific background: There is clearly an identified need to improve coho escapement in the Columbia River Basin. The proposal explains the needs but it would be helpful to give reviewers a sense of the number of spawners they are dealing with and the length of the streams they are trying to enumerate. The literature review is somewhat scanty and would be improved by a description of coho spawning habitat to give the reader a sense of what they are up against in enumerating this species, e.g., spawning under stumps, flooding washing away redds, etc. Only a few references in the peer-reviewed literature are given on spawner enumeration methods and most seem to be classical and older. Map(s) of the Abernathy Creek study area and other locations described in this section would have been useful. Rationale and significance to subbasin plans and regional programs: Spawner enumeration is clearly required to enumerate adult returns, a key goal for harvest managers and this aspect is adequately demonstrated. However links to habitat restoration and ecosystem management in tributaries, also a key aspect of subbasin planning, are not highlighted. In fact there is little mention of spawning habitat at all. The tie in with juvenile assessment (to assess productivity) is only briefly mentioned on page 4, and an expansion of that linkage would give the reader a better sense of the total coho program in the subbasins. Relationships to other projects: Proponents will collaborate with other groups doing spawner enumeration but do not mention any linkage to 200736800, which is a proposal for adult salmon coho monitoring by the same proponents in some of the same watersheds. The proposal would be improved if the work were integrated with habitat assessment studies which are a key aspect of the subbasin/province plans. Perhaps there is no concern about habitat limitations and productivity, but as the proposal reads now reviewers get no sense of that aspect. Objectives: Objectives are adequately described but tie in with subbasin plans not well described, the reviewer should not have to refer to the subbasin plans to see where the work fits in. A more meaningful and measurable objective might have been: "to improve enumeration in x streams or on y stocks." No specific timelines are provided with respect to the primary objective. Tasks (work elements) and methods: The methods seem to be straightforward but adding some details would have improved the proposal, e.g.: - Map(s) showing the study area, locations of fish traps, and sections that will be surveyed (timeline), as well as a description of how appropriate (representative) the site is for addressing the objective; -What methods are being used to hold the fish while they are being tagged? -What techniques will be used for marking and tagging fish, and will handling effects be evaluated? -How will the AUC enumerators count/detect fish spawning under stumps, log jams etc? -Why is the diversity issue mentioned in methods as it is not listed as an objective? -The issue of resighting uses a reference to steelhead as justification. However steelhead spawning habitat is very different than that of coho. -Consideration of variation/confidence for estimates between reaches in Abernathy Creek would have improved the proposal. -Inclusion of a statistical power analysis if 2005 data had been available at the time that this proposal was submitted. -A greater emphasis (explanation) on cost-effectiveness of implementation and monitoring scheme. A sentence at the bottom of page 7 and top of page 8 reads, "Variance of the AUC estimates will be reduced by walking all sections believed to have suitable habitat for coho salmon spawning." Is there a potential problem here in that those areas that are identified might be in areas where the timing of spawning might differ from parts not sampled, as for example if the "suitable habitats" identified are those easily accessed by observers in the lower reaches of the stream. If something like that occurs, the AUC could underestimate the population by as much as the 27% observed. Of course, a regression equation might be developed to relate the Petersen estimates to the AUCs, but the goodness of the fit would be affected by year-to-year differences in temperature or other factors affecting choice of spawning grounds by coho. Monitoring and evaluation: Provisions for monitoring and evaluation are better than adequate. The proposal provides for assessing and comparing results of the AUC and mark recapture methods and if they do not work out, moving on to other approaches. The proponents plan to use adult age structure as a diversity metric; however, most adult coho would likely have the same age (age 1.1, or 1 winter in freshwater and 1 winter in the ocean). DNA analysis (of scales collected during this project) is also mentioned as a future diversity metric, although the proponent does not describe techniques for collecting and archiving scale samples to be used for DNA analysis. Detailed methods for the use of percentage of surveys with spawners as a spatial metric were not provided. Facilities, equipment, and personnel: Personnel are experienced in the fieldwork and inclusion of a statistician improves the credibility of the statistical methods. However it is not exactly clear what Dr. Cheng's job is in the project. All facilities and equipment are provided at no cost by WDFW (Region 5, Vancouver, WA), except for a computer. Information transfer: Plans for including data in Streamnet and other specific databases are included. Apparently inclusion in a regional database is dependent on a BPA program. The proponents should press those concerned to implement this regional database, and this could have been part of the present proposal. A plan for publishing the results of the AUC-mark-recapture method in a peer-reviewed journal should be included as the present proposal appears to generate only grey literature. Benefits to focal and non-focal species: Coho are listed and are an indicator species of subbasin and Province conditions. Knowledge of results of the comparison between AUC and mark-recapture will benefit other focal species if they are peer reviewed and published. The results of the statistical analysis are likely to benefit a set of linked projects that use similar survey methods for estimating the abundance of the Lower Columbia River coho salmon ESU, as well as other species (Chinook and chum salmon, and steelhead). However, there was no description of techniques to be used for marking and tagging adult coho salmon at the traps, or evaluation of potential harmful effects (direct mortality or delayed stress) on spawning adults or their progeny. The proposal should include consideration of how or if the traps will impact other fish species in the streams (salmonids and non-salmonids) as well as other aquatic biota if present, e.g., mammals.
State/province recommendation: MS: Recommended Action
Review group: MSRT
Recommended budgets: FY07: (n/a) | FY08: (n/a) | FY09: (n/a)
Comment: This is a comparison of accuracies for techniques. This proposal should be guided by PNAMP and CSMEP. Application of the technique should be proposed and funded under a separate proposal.