Methods Report: Monitoring Mark-Selective Recreational Chinook Fisheries In the Marine Catch Areas of Puget Sound (Areas 5 through 13)

In recent years, abundant runs of hatchery Chinook salmon (Oncorhynchus tshawytscha) have been mixed with depressed runs of wild Chinook salmon in the marine environments of the Puget Sound and Strait of Juan de Fuca. Providing recreational anglers with opportunities to harvest abundant hatchery stocks while simultaneously protecting weaker, wild stocks, such as Endangered Species Act (ESA)-listed Puget Sound Chinook salmon, has proven to be a significant conservation and management challenge. The combination of large-scale hatchery marking (i.e., adipose fin clipping) programs and mark-selective harvest regulations makes it possible for anglers to pursue and harvest hatchery Chinook salmon while minimally impacting wild salmon populations. In such "mark-selective fisheries" (MSFs), anglers are generally allowed to retain adipose-fin clipped ("marked") hatchery fish and are required to release unharmed any unclipped ("unmarked," predominantly wild) salmon encountered.

The Washington Department of Fish and Wildlife (WDFW) implemented the first recreational pilot mark-selective Chinook fishery in the marine waters of Washington State within Areas 5 and 6 (Strait of Juan de Fuca) during summer 2003, based on agreements between the State of Washington and Northwest Treaty Tribes during the annual North of Falcon salmon season-setting process (WDFW 2008a). The pilot fishery purpose, as stated in state-tribal agreement documents (e.g., Northwest Treaty Tribes and the Washington Department of Fish and Wildlife, 2007), is defined as follows:

"The purpose of the 'pilot' fishery is to collect information necessary to enable evaluation and planning of potential future mark-selective fisheries. The 'pilot' fishery provides a basis for determining if the data needed to estimate critical parameters can be collected and if the sample sizes needed to produce these estimates with agreed levels of precision can be realistically obtained."

Over the past eight years (since 2003), in addition to the mark-selective Chinook fishery in Areas 5 and 6, WDFW has implemented additional pilot-level mark-selective Chinook salmon fisheries in several Puget Sound Marine Catch Areas (Areas 5 through 13) during both the summer and winter seasons (Appendix A; Figure 1). The first wintertime Chinook MSF was established on a pilot basis in Areas 8-1 and 8-2, from October 2005 through April 2006, and has continued each winter season ever since (with varying fishery season length; see Appendix A). Additionally, beginning in 2007, summer selective Chinook fisheries were established in Areas 9, 10, 11, and 13 and winter selective Chinook fisheries in Areas 7, 9 and 10. Also, as a result of the 2009 North of Falcon process, Chinook MSFs were established for the first time in Areas 11 and 12 during the winter season (February through April) of both 2010 and 2011. Thus, as of the close of summer 2010 fishing season, pilot summer selective Chinook seasons have occurred in Areas 5 and 6 for eight years (2003-2010; Thiesfeld and Hagen-Breaux 2005a, Thiesfeld and Hagen-Breaux 2005b, WDFW 2008a, WDFW 2009a, WDFW 2010g, and WDFW 2011) and in Areas 9, 10, 11, and 13 for four years (2007-2010; WDFW 2007a and 2007b, WDFW 2009b and 2009c, WDFW 2010e and 2010f, and WDFW 2011). Pilot winter selective Chinook fisheries have occurred in Areas 8-1 and 8-2 for five complete seasons (2005-06, 2006-07, 2007-08, 2009, and 2009-10; WDFW 2008b, WDFW 2009d, WDFW 2010b), Areas 9 and 10 for three winter seasons (2008, 2008-09, and 2009-10; WDFW 2010c, WDFW 2010d), Area 7 for three winter seasons (2008, 2009, and 2009-10; WDFW 2009e, WDFW 2010a), and in Areas 11 and 12 for one winter season from February 1 through April 30, 2010. Consistent with the 2004 Puget Sound Chinook Harvest Management Plan (Puget Sound Indian Tribes and WDFW 2004), a key goal of implementing each of these mark-selective Chinook fisheries has been to provide meaningful opportunity to the recreational angling public while minimally impacting ESA-listed Puget Sound Chinook salmon.

The State of Washington and Northwest Treaty Tribes have planned the pilot mark-selective Chinook fisheries in Puget Sound based on assumptions about the performance of each fishery and how the fishery was predicted to affect wild (unmarked) and hatchery (marked) Chinook salmon. For example, the total number of marked and unmarked Chinook salmon encountered in these fisheries was estimated during the pre-season planning process using the Chinook Fishery Regulation Assessment Model (FRAM) and assumptions about fish abundance and angler effort levels. The sampling and monitoring programs in place for the pilot selective fisheries have provided a means of verifying these pre-season assumptions. More fundamentally, results of the sampling programs have been used to determine if the data needed to provide usable estimates of critical parameters can be collected, and if the sample sizes needed to produce these estimates with agreed levels of precision can be realistically obtained (WDFW 2008a and 2008b).

Comprehensive Sampling and Monitoring Program

Given the pilot nature of the mark-selective Chinook fisheries in Puget Sound, WDFW's Puget Sound Sampling Unit (PSSU) has been tasked with implementing a comprehensive sampling and monitoring program to collect the data needed to evaluate each pilot mark-selective Chinook fishery and its impact on unmarked salmon. As per state-tribal agreement (e.g., WDFW and NWIFC 2009), we have developed area-specific sampling plans consisting of several comprehensive and complementary sampling components, including dockside creel sampling, test fishing, on-water or aerial effort surveys, and angler-completed voluntary trip reports (VTRs). We have tailored area-specific sampling plans so that we could reliably estimate the following critical parameters needed for evaluating mark-selective fisheries: i) the mark rate of the targeted Chinook population, ii) the total number of Chinook salmon harvested (by size [legal or sublegal] and mark-status [marked or unmarked] group), iii) the total number of Chinook salmon released (by size and mark-status group), iv) the coded-wire tag- (CWT) and/or DNA-based stock composition of marked and unmarked Chinook mortalities¹, and v) the total mortality of marked and unmarked double index tag (DIT) CWT stocks. In addition, we have acquired and analyzed relevant data characterizing other aspects of the pilot fisheries, including descriptors of fishing effort, fishing success (catch [landed Chinook] per unit effort), the length and age composition of encountered Chinook, and the overall intensity of our sampling efforts.

History of By-Area Monitoring Plans

Area-specific sampling and monitoring plans for mark-selective Chinook fisheries in Puget Sound have evolved over the past eight years, ever since the first mark-selective Chinook fishery was established in Areas 5 and 6 in 2003, in response to state-tribal negotiations at the annual North of Falcon and Pacific Fishery Management Council (PFMC) salmon season-setting processes. Factors such as the need for timely in-season creel estimates of salmon encounters (i.e., retained and released salmon) to assess pilot Chinook MSFs in-season, costs of shore-based and on-water monitoring, feasibility of on-water sampling of Chinook encounters by size/mark status, bias and precision considerations for dockside and on-water based estimates, expansion of the WDFW voluntary trip report (VTR) program in recent years, and other factors specific to each area have been influential in affecting the development and evolution of annual selective fishery monitoring plans. In addition, reaching state-tribal agreement on area-specific monitoring plans has hinged on whether or not certain sampling components are included or not in each area's sampling design, and at what defined intensity level of sampling --e.g., components such as dockside creel sampling, test fishing, on-water effort surveys or aerial surveys, test fishing, and/or angler-completed voluntary trip reports. In the next section of this report (Marine Catch Areas of Puget Sound), we describe the evolution of our agreed-to sampling and monitoring plans for each Puget Sound Marine Catch Area (Areas 5-13) in which mark-selective Chinook fisheries have been implemented and evaluated.

Methods Development to Estimate Total Chinook Encounters

In addition to developing and implementing comprehensive monitoring designs, a key facet of WDFW's selective fishery evaluation program over the years has included working with our tribal counterparts to develop and refine methods for estimating Chinook encounters and associated mortalities in mark-selective Chinook fisheries. To analyze and report on selective Chinook fishery data from the 2003 through 2007 seasons, we applied two methods to estimate total Chinook encounters in marine recreational mark-selective fisheries in Puget Sound. Consequently, within our annual post-season selective fishery reports for the 2003-2007 seasons, we presented two different Chinook encounters estimates, resulting from the two separate methods, and compared them. As Conrad and McHugh (2008) discuss in their thorough evaluation of the two methods, both the Method 1 (M1) and Method 2 (M2) approaches were designed to estimate the same quantity (total Chinook encounters), yet they often yielded differing results, which made it difficult to interpret post-season estimates of fishery impacts. The two approaches are best described (from Conrad and McHugh's [2008] publication) as follows:

• Method 1 (M1) -- M1 estimates of total Chinook encounters are derived from the combination of dockside observations of landed catch and angler interview responses about salmon releases; thus, as Conrad and McHugh point out, the accuracy of Method 1 estimates depends heavily on the ability of anglers to correctly recall and report the number of Chinook they actually encountered and released. M1 relies on creel survey data to estimate the total number of Chinook harvested and the total number of Chinook released and then apportions the total encounters (estimated number harvested plus estimated number released) to four size/mark status categories (legal-size and marked [LM], legal-size and unmarked [LU], sublegal-size and marked [SM], and sublegal-size and unmarked [SU]) using test fishery or angler-completed voluntary trip report (VTR) data.
• Method 2 (M2) -- M2 estimates of Chinook encounters are obtained using the creel survey estimates of the total number of legal-size, marked Chinook harvested in combination with the test fishery or VTR data to estimate both the total number of Chinook encounters and to apportion the encounters to four size/mark status categories (LM, LU, SM, SU). As Conrad and McHugh (2008) discuss, the M2 estimator was derived assuming that anglers retain all LM Chinook encountered; therefore, its accuracy depends on the extent to which angler behavior deviates from this idealized case.

To identify a single, reliable estimate of total Chinook encounters in mark-selective Chinook fisheries, Conrad and McHugh (2008) quantitatively evaluated sources of bias in the M1 and M2 approaches and considered possibilities for correcting the bias within each method. For Method 1, they reviewed evidence that suggested a combination of digit bias² and prestige bias³ contributed to M1 over-estimating the true number of Chinook encounters, especially when encounter rates were high in a selective fishery. For Method 2, they evaluated evidence indicating that LM Chinook release occurs on both an intentional and unintentional basis. In combination, the authors found that intentional and unintentional releases likely contributed to a 12-13% underestimate of actual (true but unknown) encounters by M2.

Based on their analyses and practical considerations regarding the most feasible bias correction approaches, Conrad and McHugh ultimately recommended using Method 2 with a correction for the release of legal-size marked Chinook as the preferred method for estimating total Chinook encounters in mark-selective Chinook fisheries. In particular, they determined that an "unbiased" estimate of total Chinook encounters could be obtained under Method 2 using:

Bias-Corrected M2 = Original M2 Estimate / (0.87)

In August 2008, WDFW and tribal representatives conducted a thorough technical review of Conrad and McHugh's recommended bias-corrected (i.e "M2-adjusted") method for estimating total Chinook encounters in mark-selective Chinook fisheries. The state-tribal technical group agreed that the recommended M2-adjusted approach would enable the most reliable, single estimate of total Chinook encounters in mark-selective Chinook fisheries. Consequently, starting with our selective fishery data analyses conducted in Summer/Fall 2008 and thereafter, WDFW applied Conrad and McHugh's M2-adjusted estimation approach to report the estimated Chinook encounters and associated mortalities in Chinook MSFs (see Appendix B for complete computational details). From this point forward, our post-season reports contained one "best estimate" of Chinook encounters; we no longer presented two different, sometimes conflicting, estimates of Chinook encounters based on two separate approaches with inherent biases. Further, based on Conrad and McHugh's (2008) analysis and recommendations, we determined that we could apply the M2-adjusted method in cases where an estimate of the total number of LM Chinook harvested is obtained through less intensive survey approaches (e.g., estimates of total Chinook harvest resulting from the WDFW Catch Record Card [CRC] system, coupled with field estimates of LM Chinook relative abundance from voluntary trip reports).

Selective Fishery Reporting--Need for Centralized Methods Report

Since pilot mark-selective Chinook fisheries were first implemented in Puget Sound in 2003, WDFW has produced a detailed post-season data report evaluating each mark-selective Chinook fishery implemented in Puget Sound, as well as two multi-year reports (WDFW 2008a, 2008b). We have generated a separate post-season data report for each area and season (winter or summer), containing estimates of each of the critical selective fishery parameters listed above along with comparisons to FRAM pre-season predictions of the key parameters.

In July 2010, technical staffs from the WDFW Puget Sound Sampling Unit, Northwest Indian Fisheries Commission (NWIFC), and Puget Sound Treaty Tribes met to discuss potential reporting efficiencies in WDFW's mark-selective Chinook fishery post-season reports. NWIFC and tribal representatives had initiated the idea for such a meeting, considering that we at WDFW had been submitting a separate post-season report for each area and season (since 2003) to the co-managers, resulting in redundancies between individual reports, particularly in the Methods section. Also, over the years we kept adding sections to the selective fishery annual reports, in response to individual tribal co-manager requests, and sustained those additions in each future report, resulting in ever-lengthening post-season reports. From both the WDFW and tribal technical perspectives, we needed to prioritize the most essential reporting elements and achieve efficiencies to streamline the selective fishery reporting work load.

Thus, at the July 2010 meeting the WDFW and tribal staffs worked on prioritizing the most essential elements (i.e., tables, figures, and appendices) needed in WDFW's annual post-season selective fishery reports in an effort to define reporting efficiencies. Based on these decisions (details available in a WDFW memo dated August 16, 2010, summarizing the July 2010 meeting), WDFW would begin implementing the agreed-to reporting efficiencies within our draft 2009-10 winter selective fisheries report.

Further, we agreed that a key efficiency in the annual reporting process would be for WDFW staff to produce a centralized Methods Report. The Methods Report would be a stand-alone document that includes the details of each area's Chinook MSF study design (for both winter and summer fisheries), sampling procedures, data analysis methods, and all equations used to generate estimates and variances. We determined that the Methods Report would be cited in future annual data reports (starting with the draft winter 2009-10 post-season report), enabling us to eliminate methods-related redundancies contained in the previous by-area annual reports. Thus, the Methods Report presented herein is a key outcome of state-tribal collective ideas for creating selective fishery reporting efficiencies.

Methods Report Objectives

The purpose of this Methods Report is to provide detailed documentation of WDFW Puget Sound Sampling Unit's sampling designs and procedures used to monitor and evaluate the critical selective fishery data parameters (as defined above) for pilot mark-selective Chinook fisheries implemented in Puget Sound since 2003. In particular, we focus on documenting details of the most current selective fishery monitoring designs, sampling procedures, data analysis methods, and all equations used to generate estimates and variances. As such, this Methods Report is considered a companion document accompanying each of WDFW's annual post-season data reports (i.e., serves as a Methods Section for each report), beginning with the draft 2009-10 winter selective fisheries report and continuing with each report thereafter. Further, we will update this Methods Report if any of the methods described herein are modified in the future (i.e., after 2010) based on state-tribal technical agreement.

In the following pages, we first provide a full description of each Marine Catch Area in Puget Sound (Areas 5 through 13) with accompanying maps. Within each area's section, we also present an overview of the recreational mark-selective Chinook fishery seasons that have occurred in the area to date. Additionally, we describe the evolution since 2003 of our agreed-to sampling and monitoring plans for each Puget Sound Marine Catch Area in which mark-selective Chinook fisheries have been implemented and evaluated using sampling and monitoring programs. In the subsequent Methods section of the report, we provide a detailed account of our sampling procedures and post-season estimation methods within the context of the Puget Sound Sampling Unit's four primary sampling designs, which we present in a sequence of numbered "Sections" as follows: 1) Full Murthy Estimate Design; 2) Reduced Murthy Estimate Design; 3) Aerial-Access Design; and 4) Baseline Design. If specific sampling protocols from Section 1 of this report are identical to methods used within the sampling designs presented in Sections 2 through 4, we cite the methods description in Section 1 to avoid redundancies in later sections. Also, we refer the reader to WDFW's "2010 Puget Sound Sampling Manual" (available at www.recfin.org) to review examples of specific data collection forms that are used when implementing each sampling design. Finally, in a series of appendices to this report we provide: A) the detailed history of intensive (i.e., creel estimates) versus baseline sampling in Puget Sound/Strait of Juan de Fuca Marine Catch Areas with selective fishery seasons and sampling plan components (years 2003-2010); B) a detailed description of our encounters and mortalities estimation scheme; C) aerial-access design estimators; D) statistical week tables by calendar year (corresponding to statistical weeks referenced in WDFW's annual selective fishery data reports); and E) a table of sampled sites and sampling designs per area and season (winter or summer) for the Puget Sound recreational mark-selective fishery sampling program.

¹ Though the necessary tissue samples have been collected, DNA-based estimates of stock composition are presently unavailable for Puget Sound/Strait of Juan de Fuca mark-selective fisheries. In the present report, methods for producing CWT-based (unexpanded) estimates of the stock composition of marked Chinook harvest are provided.

² A type of angler recall error in which anglers report the number of salmon they encounter as a rounded approximation of what they actually release (e.g., reporting that 10 salmon were released when the true number was actually 9 or 11). See Conrad and McHugh (2008) for further detail.

³ When anglers exaggerate the number of fish that they caught to be perceived as a better angler. See Pollock et al. (1994) and Conrad and McHugh (2008) for further detail.