Elsevier

Fisheries Research

Volumes 129–130, October 2012, Pages 82-93
Fisheries Research

Population-specific escapement of Columbia River fall Chinook salmon: Tradeoffs among estimation techniques

https://doi.org/10.1016/j.fishres.2012.06.013Get rights and content

Abstract

In the multi-stock Columbia River system, managers estimate fall Chinook salmon, Oncorhynchus tshawytscha (Walbaum), escapements using various combinations of spawning ground surveys, harvest data and fish counts at dams and hatcheries. Our objectives were to improve upon the traditional methods, and to evaluate trade-offs among methods. Using data from radio-tagged (n = 4421) and PIT-tagged (n = 1950) adult salmon over eight years, we applied a mark-recapture method to estimate population-specific escapements, both aggregating data within year and stratifying them by week. Mark-recapture estimates differed between estimation techniques and from estimates generated using traditional methods. Stratifying data by week measured escapement estimate uncertainty more reasonably than aggregating data within year. Radiotelemetry provided better spatial resolution among populations for tributary spawners whereas PIT tags provided low-cost, easily replicated estimates using an existing detection system. Mark-recapture techniques had several advantages over current practices: quantifying uncertainty, transparent methods and reduced sensitivity to survey biases.

Highlights

► Improved techniques for spawning escapement estimation are needed for fall Chinook salmon in the Columbia River. ► A mark-recapture method was applied to data on radiotelemetry and PIT tagged salmon collected during eight years. ► Because of the variability in salmon migration timing, and sampling restricted by water temperature, we stratified data by week. ► Stratified analyses measured population-specific salmon escapement and associated uncertainty more reasonably than aggregated methods.

Introduction

Adult escapement estimates for anadromous Pacific salmonids (Oncorhynchus spp.) are used to assess a variety of management and conservation objectives. These include monitoring population status and trends (Baker et al., 1996), developing life-cycle survival models (Kareiva et al., 2000, McClure et al., 2003), evaluating restoration and recovery efforts (e.g., Peters et al., 2001), and forecasting adult returns (e.g., Hyun et al., 2005, Hyun et al., 2012b). Escapement data are perhaps most frequently used to manage fisheries, especially in mixed-stock river systems where robust populations are intermingled with populations of conservation concern (e.g., Knudsen, 2000, Good et al., 2007).

The accuracy and precision of salmonid escapement estimates differ widely among estimation techniques and it is important to understand the strengths, biases, and sources of uncertainty in each approach. Most methods are constrained by some combination of available resources (e.g., personnel, time, equipment), characteristics of the enumeration site (e.g., depth, turbidity, accessibility), salmon behaviors (e.g., immigration and emigration timing, residence time, prespawn mortality), and study design (e.g., sampling intervals and tag recovery likelihood). The importance of these factors varies among methods and information requirements. For example, widely used visual surveys across a geographic sampling frame like redd counts (hereafter “surveys”) can be cost effective, but are sensitive to weather, river conditions, and among-observer differences (Chapman et al., 1986, Jones et al., 1998, Groves and Chandler, 1999, Dunham et al., 2001, Holt and Cox, 2008). Direct enumeration (hereafter “counts”), such as at passage constrictions like weirs, fences, or dam fishways is generally considered more reliable than visual methods. However, installation and maintenance can be cost-prohibitive and count data are also subject to errors. These include unmonitored upstream or downstream passage events (Dauble and Mueller, 2000, Boggs et al., 2004) and the counting biases associated with fish density, separation among populations (Murdoch et al., 2010), timing effects (Hilborn et al., 1999, Parken et al., 2003), and fallback and reascension at count stations (Boggs et al., 2004). Mark-recapture studies have important statistical advantages over many survey and enumeration approaches, including the ability to quantify uncertainty in escapement estimates (Lebreton et al., 1992, Schwarz and Seber, 1999). However, mark-recapture estimates can be sensitive to temporal and spatial sampling biases, the number of fish tagged and released, and capture and recapture probabilities (Schwarz et al., 1993, Arnason et al., 1996, Korman et al., 2002, Hyun et al., 2012a).

In this study, we applied a mark-recapture method to estimation of population-specific Columbia River fall Chinook salmon (ocean-type Oncorhynchus tshawytscha [Walbaum]) escapement, using radiotelemetry and Passive Integrated Transponder (PIT) tag data. When mark-recapture data are used, the definition of escapement depends on the location(s) where ‘recapture’ (i.e., recovery, or detection of marked fish) is made. In this paper, we define escapement as to a spawning site (i.e., potential spawners) because pre-spawn survival, harvest mortality, spawning activity and spawning success were not monitored in either the radiotelemetry- or PIT-tagged salmon (with very minor exceptions). This is common in studies of this type. We further compared the mark-recapture escapement estimates with salmon counts at dams (count-based estimates) and agency point estimates (survey-based estimates).

The fall Chinook salmon run includes the abundant Hanford Reach population, the threatened Snake River population (U.S. Endangered Species Act, NMFS, 1992), and additional populations in the upper Columbia River, Yakima River, and several smaller tributaries and satellite spawning groups (Dauble and Watson, 1997, Myers et al., 1998, Dauble et al., 1999). The run is a major contributor to Alaskan and Canadian ocean fisheries and to Columbia River commercial, tribal, and sport fisheries (Waples et al., 1991, Myers et al., 1998). Two groups are indicator stocks for the Pacific Salmon Commission (PSC) and are used to set international and regional harvest quotas and spawner escapement targets (PSC, 2000). These are the Deschutes River population and the “upriver bright” aggregate that includes Hanford Reach, Snake, Yakima, upper Columbia, and several less abundant populations (Fig. 1).

Our primary objective was to improve current methods of estimating population-specific escapement of upriver bright and Deschutes River fall Chinook salmon. Our approach was to estimate escapement using mark-recapture data from recent radiotelemetry and PIT-tag studies, and then to compare those estimates with independent estimates from management agencies and from adult Chinook salmon counts at dams. We applied analytical estimators of the escapement and its uncertainty to mark-recapture data, where data were both pooled (i.e., seasonal) and stratified by week to assess effects of seasonal fall Chinook salmon run composition changes (e.g., Jepson et al., 2010). A second objective was to evaluate tradeoffs associated with radiotelemetry versus PIT-tag monitoring systems, including differences in inferential power, spatial resolution, and ability to discriminate among populations. For the comparison, we simulated PIT-tag detection histories from a larger radiotelemetry dataset to help describe escapement variability among populations and years, and to help evaluate the current capability of the adult PIT tag detection system to estimate escapement. The final objective was to use the mark-recapture results to evaluate the relationship between population-specific escapement precision and sample size requirements for future escapement studies.

Section snippets

Salmon collection and tagging

In the 2000–2005 radiotelemetry and 2008–2009 PIT-tag studies, adult fall Chinook salmon were trapped at Bonneville Dam (Columbia River kilometer 235) in the adult fish facility adjacent to the Washington-shore fish ladder (Fig. 1). A weir in the main ladder was lowered to direct salmon into the facility where they were either diverted into anesthetic tanks (25 mg/L clove oil solution) for tagging or returned to the main ladder without handling. Salmon were intended to be tagged in proportion to

Tagging, detection and distribution

In total, 4421 radio-tagged and 1950 PIT-tagged Chinook salmon were used for distribution and mark-recapture analyses (Table 1, Table 2). Relative under-sampling occurred during run peaks in all years and tagging was restricted in some weeks in August 2004, 2005, and 2009 (Fig. 2). Relative over-sampling (2000, 2008) and under-sampling (2005) also occurred late in some migrations. Overall, tagging was qualitatively most proportional to the run in 2003 and was least proportional in 2005 (Fig. 2).

Comparisons among escapement estimation methods

There were clear tradeoffs among estimation techniques. First, an important advantage of the mark-recapture methods was the ability to calculate escapement variance estimates (and thus confidence intervals) for each population. Quantifying uncertainty is critically important, especially for managers attempting to simultaneously address optimal harvest and conservation objectives. Second, the mark-recapture method was relatively transparent and required few data sources (i.e., only Bonneville

Acknowledgments

We are grateful to the many people who provided their time and assistance during the course of this study. We especially acknowledge T. Bjornn (UI), C. Peery (USFWS), and L. Stuehrenberg (NMFS) who were responsible for the early radiotelemetry studies. Field operations were conducted by B. Begay, W. Perez, C. Torbeck, J. Mainord, R. McConville, and M. Whitman (CRITFC), and R. Ringe, K. Tolotti, S. Lee, C. Boggs, W. Daigle, D. Queampts, P. Keniry, M. Morasch, M. Heinrich, T. Dick, and C. Morat

References (47)

  • S. Hyun et al.

    Integrated forecasts of fall Chinook salmon returns to the Pacific Northwest

    Fish. Res.

    (2012)
  • Arnason, A.N., Kirby, C.W., Schwarz, C.J., Irvine, J.R., 1996. Computer analysis of marking data from stratified...
  • T.B. Baker et al.

    Status of Pacific salmon and steelhead escapements in southeastern Alaska

    Fisheries

    (1996)
  • C.T. Boggs et al.

    Fallback, reascension and adjusted fishway escapement estimates for adult chinook salmon and steelhead at Columbia and Snake River dams

    Trans. Am. Fish. Soc.

    (2004)
  • E.L. Brannon et al.

    Population structure of Columbia River basin chinook salmon and steelhead trout

    Rev. Fish. Sci.

    (2004)
  • B.J. Burke et al.

    Performance of passive integrated transponder tags and radio tags in determining dam passage behavior of adult Chinook salmon and steelhead

    N. Am. J. Fish. Manage.

    (2006)
  • D.W. Chapman et al.

    Effects of river flow on the distribution of Chinook salmon redds

    Trans. Am. Fish. Soc.

    (1986)
  • M.R. Dare

    Mortality and long-term retention of passive integrated transponder tags by spring Chinook salmon

    N. Am. J. Fish. Manage.

    (2003)
  • D.D. Dauble et al.

    Status of fall chinook salmon populations in the Mid-Columbia River, 1948–1992

    N. Am. J. Fish. Manage.

    (1997)
  • D.D. Dauble et al.

    Fall chinook salmon spawning in the tailraces of lower Snake River hydroelectric projects

    Trans. Am. Fish. Soc.

    (1999)
  • D.D. Dauble et al.

    Difficulties in estimating survival for adult chinook salmon in the Columbia and Snake rivers

    Fisheries

    (2000)
  • J. Dunham et al.

    Sources and magnitude of sampling error in redd counts for bull trout

    N. Am. J. Fish. Manage.

    (2001)
  • J.W. Gibbons et al.

    PIT tagging: simple technology at its best

    Bioscience

    (2004)
  • Fryer, J.K., Mainord, J., Whiteaker, J., Kelsey, D., 2009. Upstream migration timing of Columbia Basin Chinook and...
  • T.M. Goniea et al.

    Behavioral thermoregulation and slowed migration by adult fall Chinook salmon in response to high Columbia River water temperatures

    Trans. Am. Fish. Soc.

    (2006)
  • T.P. Good et al.

    Recovery planning for Endangered Species Act-listed Pacific salmon: using science to inform goals and strategies

    Fisheries

    (2007)
  • P.A. Groves et al.

    Spawning habitat used by fall chinook salmon in the Snake River

    N. Am. J. Fish. Manage.

    (1999)
  • R. Hilborn et al.

    Estimating spawning escapements from periodic counts: a comparison of methods

    Can. J. Fish. Aquat. Sci.

    (1999)
  • K.R. Holt et al.

    Evaluation of visual survey methods for monitoring Pacific salmon (Oncorhynchus spp.) escapement in relation to conservation guidelines

    Can. J. Fish. Aquat. Sci.

    (2008)
  • S. Hyun et al.

    A statistical model for in-season forecasts of sockeye salmon (Oncorhynchus nerka) returns to the Bristol Bay districts of Alaska

    Can. J. Fish. Aquat. Sci.

    (2005)
  • S. Hyun et al.

    Accounting for tag loss and its uncertainty in a mark-recapture study with a mixture of single- and double tags

    Trans. Am. Fish. Soc.

    (2012)
  • Jepson, M.A., Keefer, M.L., Naughton, G.P., Peery, C.A., Burke, B.J., 2006. Run-timing, escapement, and harvest of...
  • M.A. Jepson et al.

    Stock composition, migration timing, and harvest of Columbia River Chinook salmon in late summer and fall

    N. Am. J. Fish. Manage.

    (2010)
  • Cited by (0)

    View full text