The 2013 Port Moller Test Fishery

Transcription

1 The 2013 Port Moller Test Fishery Bristol Bay Science and Research Institute P.O. Box 1464 Dillingham, AK September 2013

2

3 The 2013 Port Moller Test Fishery Prepared by Scott W. Raborn, Michael R. Link, and Guy D. Wade LGL Alaska Research Associates, Inc West International Airport Road, Suite C-1 Anchorage, Alaska, Prepared for Bristol Bay Science and Research Institute P.O. Box 1464 Dillingham, AK September 20, 2013

4 Please cite as: Raborn, S.W., M.R. Link, and G.D. Wade The 2013 Port Moller test fishery. Final Report prepared by LGL Alaska Research Associates, Inc., Anchorage, AK, for the Bristol Bay Science and Research Institute, Dillingham, AK, 180p. ii

5 Table of Contents Executive Summary... 1 Introduction... 2 Methods... 3 Study Area... 3 Gear Description... 3 Sampling Protocol... 4 ing Based on the PMTF... 5 Additional Experiments To Test for Gillnet Saturation... 6 Reporting of PMTF Information... 7 Results and Discussion... 7 Future Work and Recommendations Acknowledgements References Table 1. Timing of information received and delivery of daily updates for the 2013 PMTF Table 2. Summary of substantive comments and predictions made in the daily interpretations of the 2013 PMTF Figure 1. Locations of fishing stations for the Port Moller Test Fishery relative to Bristol Bay.. 15 Appendix A: BBSRI s Daily Updates from the Port Moller Test Fishery, Appendix B: In-season Stock Composition Estimates for Port Moller, from ADF&G s Gene Conservation Lab Appendix C: In-season Age Composition Estimates from the Port Moller Test Fishery and Inshore Catch and Escapement, from ADF&G, Appendix D: Daily Run Summary fro Bristol Bay Salmon for 2013, from ADF&G Website iii

6 Executive Summary In 2013, the Port Moller Test Fishery (PTMF) operated from 10 June to 6 July to provide an advance indication of the run strength of sockeye salmon returning to Bristol Bay and its fishing districts. Inseason, daily catch summaries were distributed the same evening as catches were reported from the PMTF, and interpretations of these catches were usually distributed the following day. Estimates of genetic stock composition and age composition were forwarded to the distribution list soon after receipt from ADF&G. The final genetic stock composition estimates cumulative through the season were reasonably accurate given each stock s run timing and magnitude. Nominal discrepancies were the slight over-representation of the Ugashik stock (over by 3%) and under-representation of the Nushagak-Wood District (under by 5%). Combining stock composition estimates with the daily Replacement Index allowed for reasonable districtspecific forecasts of increases and decreases in catch and escapement on 25 June. Likewise, district-specific forecasts made on 4 July and 7 July utilizing stock composition estimates were fairly accurate. The aged scale samples from the PMTF yielded age composition estimates from ADF&G that were close to the yearend observed values. Timing of the 2013 Bristol Bay sockeye run proved to be aberrant, both at the PMTF and in Bristol Bay. Port Moller catches peaked on 21 June, nine days earlier than the average of 29 June. This early arrival caused the cumulative Traditional Index to be the highest on record up through 22 June before falling below average by 1 July. Likewise, total C+E in Bristol Bay peaked around June 30, which was six to eight days earlier than the average peak from 1976 to 2012 (6-7 July) making 2013 the earliest run on record. With respect to tracking the run to inshore districts, the 2013 PMTF worked reasonably well with the exception of missing catches due to bad weather. Sliding the interpolated Replacement Index forward by a travel time of nine days illustrates why we failed to detect the last movement of fish that peaked on 10 July. Clearly the interpolated values for missed fishing days 29 June-2 July were incorrect. Total run forecasts of million, 22 million, and 19 million fish were made on 28 June, 4 July, and 7 July, respectively. The resulting run through 18 July was 23 million fish. For some time we have suspected that gillnet saturation may affect PMTF catches. Initial results from this year s 20 min experimental sets versus paired 60 min sets (n=8) indicate that net saturation begins to occur beyond a Replacement Index of 50. Another anomaly this year was the pattern in sea surface temperatures throughout the season. Temperatures began about average, rose quickly, and then ended below average by season s end. Port Moller 2013, Page 1

7 Recommendations for 2014: Continue with 20 min experimental sets as opportunity allows testing for and quantifying gillnet saturation. Research and develop better models to forecast district-specific run strengths and interpolate for missed fishing days due to bad weather. Analyze the historical database for evidence of tidal influence on daily fluctuations in daily catch indices. Introduction The Port Moller Test Fishery (PMTF) has been conducted since 1967 with gillnets set at stations offshore from Port Moller, Alaska (Figure 1) (Randall 1977; Eggers and Fried 1984). The primary goal has been to predict run strength of sockeye salmon Oncorhynchus nerka, returning to natal streams in Bristol Bay approximately one week prior to their arrival in the various terminal commercial fishing districts. The PMTF operates from around 10 June through 10 July (this year 6 July was the last day) each year, and is the first check on preseason sockeye salmon forecasts. Results from the PMTF give Bristol Bay processors, fishermen, and the Alaska Department of Fish and Game (ADF&G) time to respond to suspected departures from these forecasts (Helton 1991). These responses include the shifting of haulout tenders among districts by processors and the choice by fishermen as to which districts to fish. Though the information from the PMTF is not the primary decision support upon which the fishery is prosecuted and managed, it does play an important role as managers use it to help gauge run strength. The primary objective of the PMTF was to provide information about sockeye that will arrive at the inshore fishing districts 6-9 days into the future. This information included (1) an index of abundance, (2) age composition, and (3) stock composition. The first of these run descriptors is the most tenuous owing to the variability in how many inshore fish are represented by each fish caught in the PMTF (i.e., the fish-per-index or FPI parameter), as well as variability in how long it takes fish to travel from the test fishing transect to the inshore districts. Age composition forecasts from Port Moller are more reliable and are used to help assess the validity of agespecific preseason forecasts. Likewise, stock composition estimates based on genetic samples taken from PMTF fish are perhaps the most useful information of all. These estimates reliably predict what portion of the run will manifest in each of the fishing districts approximately one week later. Combining these pieces of information is often useful. For example, high abundance indexes coupled with a large proportion of the composition belonging to a given stock adds confidence that more fish are bound for that district. A secondary objective from the PMTF was to forecast the yearend total inshore sockeye run. While this estimate is of interest to many in that it can be viewed as reducing the overall success of the Bristol Bay total run down to a single number, few inseason decisions are affected by this estimate. Moreover, forecast error around total run predictions further limit their utility. Below we describe how the test fishery works, the results from 2013, and our recommendations for 2013 research and reporting. Appended is a compilation of all results Port Moller 2013, Page 2

8 and daily updates provided to processors, fishermen, managers, and the public during the 2013 season. Daily updates in 2013 varied with the development of the salmon run, but typically summarized catches and indices (catch per effort) by station, mean body size, water temperature, fishing conditions by date, and data interpretations (Appendix A). Also appended are reports issued periodically by ADF&G throughout the season summarizing stock compositions (Appendix B) and age compositions (Appendix C) of the Port Moller catches, as well as daily run summaries of inshore catch and escapement to each fishing district (Appendix D). Study Area Methods Most Bristol Bay sockeye salmon reach the fishing districts between the end of June and the middle of July (Mundy 1979) with the peak in the fishery occurring on or around 6 July. Travel time to the fishery from Port Moller usually takes about one week, so the PMTF has generally begun on June 10 or 11. Gillnets are set at stations located along a transect from Port Moller to Cape Newenham (Figure1). Stations are 5 miles apart with Station 1 being 30 miles offshore from Port Moller and Station 12 being 85 miles offshore. Prior to 1987, odd stations were fished on the outgoing trip, the vessel anchored overnight, and even stations were fished on the return trip (Eggers and Fried 1984). Beginning in 1987, only even stations were fished (Stations 2-8 and occasionally Station 10) during both the outbound and inbound trips. In 1999, fishing at Station 10 was resumed in response to the belief that the bulk of the run may have been further out, and in 2000 fishing occurred as far out as Station 14 (95 miles offshore) (Flynn and Hilborn 2004). Usually, only five stations are fished either Stations 2-10 or Stations 4-12 depending on the previous day s offshore distribution. Gear Description The historical net configuration (from ) used for the PMTF consisted of four 50 fathom shackles (1200 ft) of multistrand monofilament dyed dark green (Momoi Shade 9), 60 meshes deep (approximately 20 ft), with 13.0 cm (5⅛ in) stretched mesh, and hung to a 2.1 to 1 ratio. This net (termed the Traditional Net) selects for ocean age-3 fish more than it does for ocean age-2 fish. Ocean age-3 fish are typically about 6 cm larger than ocean 2-fish, which affects their susceptibility to being caught in a given mesh size. This bias causes the test fishing index to change as a function of the relative abundances for each of the ocean ages comprising a given year s run. Further, it causes bias in the age composition forecasts and possibly to the stock composition estimates given that stocks differ in size structure. Based on the results from a selectivity study ( ), we developed a new, less selective net (the Replacement Net) consisting of alternating shackles of 5⅛ in mesh and 4½ in (11.4 cm) mesh. This smaller mesh selects more for 2-ocean fish, reducing the overall misrepresentation of age-specific catches in the test fishery. As a result, the various run descriptors stock composition, age composition, and abundance index provided by the PMTF are more representative of the inshore catch and escapement. We have used the Replacement Net in the PMTF exclusively since Port Moller 2013, Page 3

9 Sampling Protocol Sets lasted for approximately one hour, and deployment was perpendicular to the migratory path of the salmon on the north-south axis (Helton 1991). The vessel traveled on a course of 340 for out-going sets and a course of 160 on incoming sets. These bearings oriented the net roughly parallel to the transect bearing, which was on a line between Port Moller and the tip of Cape Newenham. It typically required 2 to 3 minutes to deploy the full net. After setting the net, the vessel moved as far away as possible from the gill net while maintaining visual contact, which varied with conditions and was sometimes restricted to a few hundred meters during times of fog to 500 or more meters in good visibility and low sea states. No attempt was made to hook or run the net to try and increase catch so as to standardize effort from year to year as skippers and vessels change. Time was recorded when the trailing buoy was deployed, when the net was fully set, when retrieval began, and when the net was fully in. Catches were converted to catch-per-unit-effort (CPUE; fish per 200 fathom hours) to adjust for small differences in fishing times among sets (larger catches take longer to pick and cause the net to fish longer). Mean fishing time in minutes for each set was calculated as: MT ( FO SO) + ( FI SI ) = SI FO + (1) 2 Where, SO=time of day the gillnet first entered water, FO=time the gillnet was fully deployed, SI=time the gillnet retrieval began, and FI=time the gillnet retrieval was completed. CPUE was then catch divided by MT and multiplied by 60 to provide fish per 200 fathom-hours. Fish were identified to species and enumerated. Sockeye salmon were sexed, measured for length (mid eye fork length MEFL), and sampled for age by placing one scale per fish on a scale card. Tissues for genetic analysis were collected for individual sockeye salmon by clipping the axillary process of the pelvic fin. Tissues were placed into trays and preserved with alcohol (ETOH). Preserved samples were stored in a cool dry location on the test fishery vessel until shipped. Fish were sampled on the test fishery vessel s deck immediately following each fishing event. Scale sampling was conducted on 100% of the sockeye catch whenever possible. This sampling goal was routinely achieved, but occasionally is not attainable because of weather, gear problems, or exceptionally large catches. In such cases, the catch was sub-sampled as randomly and as extensively as was consistent with crew safety and time constraints. Formulating the Abundance Indexes Beginning in 1985, the daily Traditional Index (TI i ) was standardized to the sum of CPUE (note that Traditional Index CPUE=fish per 100 fathom hours) for Stations 2, 4, 6, and 8 (Rogers et al. 1989, Helton 1991). TIi I2 i + I4i + I6i + I8i = (2) Port Moller 2013, Page 4

10 where, I 2i -I 8i = station and date (i) specific CPUEs (catch per 100 fathoms). Missing station points were interpolated by averaging the station specific daily indices from the two days prior to and the two days after the missing station point(s). In 1995, CPUE was highest at Station 8 causing suspicion that a substantial proportion of the run was further offshore (Flynn and Hilborn 2004). As a consequence, the Traditional Index was altered to: TI =.8( I + I + I + 2I ) (3) i 0 2i 4i 6i 8i in 1996, which gives double weight to Station 8 to account for fish passage further offshore. There was concern about causing confusion among laypersons because the magnitude of the TI i values would change from what had been released in previous years ( ). To minimize this degree of change, the scalar 0.8 was used to reduce TI i and render the output from Equation 3 more congruous with Equation 2. In 2013, missing station-date specific values were interpolated by generating normal curves (predictive models) fit across stations (one curve) and through dates (the second curve) as per Flynn and Hilborn (2004). In 2013 we used the daily abundance index formulated in 2011 the Replacement Index (RI i ) which was the average CPUE (catch per 200 fathom hours) across five stations from the entire Replacement Net (4½ in and 5⅛ in mesh combined) on a given day. We found this index easier to interpret and explain, while keeping the same statistical properties of the sum, which is what the Traditional Index is based on. The Replacement Index uses one more station than TI i and does not give double weight to Station 8. This extra station and less sensitivity to Station 8 partially reduced random fluctuation from day to day allowing the index to better track abundance through time. The five stations comprising the Replacement Index shift across the transect depending on the offshore distribution. To make catch and CPUE from the net used in 2013 (the Replacement Net) comparable to historical data (the Traditional Index), we had to remove the portions of the 2013 catch that came from the 4 ½ in mesh (because this mesh was not used in the Traditional Net). Thus, only catches from the 5⅛ in mesh shackles (two shackles) were used in Equation 3. Catches in 2013 were multiplied by a factor of 2 to make the effort correspond with previous years when catches came from the Traditional Net (four shackles of 5⅛ in mesh). ing Based on the PMTF Stock composition forecasting. Processing, analyzing, and disseminating the genetic data collected from the Port Moller test fishery must be completed in a timely manner for it to be useful for in-season fishery management. Significant logistical issues affect the availability of these data. There was a 1- to 2-day lag between the time samples were collected and arrived at Port Moller. Furthermore, there was limited air service between Port Moller and Anchorage. The processing and analysis of genetic samples at the ADF&G Anchorage genetics laboratory usually begins 3 days after the samples were collected and sometimes 4-5 days after. Time was needed to process the samples and provide a preliminary analysis of the data. Thus, in a typical year stock composition forecasts from PMTF samples generally occurred 3-5 days after sample Port Moller 2013, Page 5

11 collection; however, flight logistics and bad weather delayed this process a few times during The 2013 stock composition estimates reported by ADF&G are included in Appendix B. Age composition forecasting. Sockeye scales were aged according to European notation (Koo 1962). Thus, numerals preceding the decimal refer to the number of freshwater annuli and numerals following the decimal refer to the number of marine annuli. Total age from time of egg deposition is the sum of these two numbers plus one to account for incubation time. Age estimations were made by ADF&G personnel in King Salmon using acetate impressions of scales under low (10x) magnification using a microfiche reader. The 2013 age composition estimates reported by ADF&G are included in Appendix C. Based on the selectivity study, we developed the Replacement Net as described above. The idea was to equalize the selectivity across the four major age groups Ages 1.2, 2.2, 1.3, and 2.3. While the data collected with this net allowed for the estimation of selectivity models that can correct for residual selectivity left in the Replacement Net, we discovered that most of the selectivity across the combined raw catch was already removed. That is, 4½ in mesh selected for ocean age-2 fish while 5⅛ in mesh selected more for ocean age-3 fish. These biases seemed to cancel each other when catches from both meshes were pooled, rendering the estimate of age selectivity fairly accurate of what is passing by the test fishery. Thus, residual selectivity was nominal, our selectivity model was not used, and age composition forecasts were based only on raw catches pooled across both meshes from the Replacement Net. Total run forecasting. At the end of the 2011 PMTF, we developed a model to forecast the total run abundance based on the inseason catches only. This Daily Projection Model is based on an approach that differs from the typical forecasting method applied to Port Moller data in that it only uses information collected this season and not the historical relationship between cumulative indexes and resulting total runs from previous years. For this new method, a curve was fit to the daily Replacement Index to determine the peak day of passage and to smooth the index values. Then, parameters that define the travel time of fish from Port Moller to inshore, as well as the fish-per-index (FPI), were estimated and applied to this curve to arrive at projected year-end run abundance for the entire Bristol Bay inshore return. In 2013, we used this Daily Projection Model inseason, but modifications were made during our last few forecasts. Research and development of statistical models that forecast total run strength based on the PMTF are ongoing and will continue until an algorithm is discovered that is robust to annual variations in run entry pattern, timing, travel time between the PMTF and inshore, as well as, the FPI. Additional Experiments To Test for Gillnet Saturation In 2012, we recommended testing for gillnet saturation by setting for a shorter length of time (20-30 min instead of 60 min). As mentioned in last year s final report, we suspected gillnet saturation may inhibit the correlation between the magnitude of the PMTF catches and inshore catch and escapement. If a gillnet fished for about one hour is saturating (reducing fishing efficiency), then perhaps reducing the soak time would help correct this bias and better index larger pushes of fish. Varying soak time to deal with gillnet saturation has precedence in the Port Moller 2013, Page 6

12 literature and has allowed for corrective models (e.g., Minns and Hurley 1988; Hansen et al. 1998; Rotherham et al. 2006). This year, we decided to test this hypothesis by setting the test fishing net for 20 min at the same station following the routine 60 min set. The boat steamed for 5 min west of the station being replicated before resetting. These extra sets required more time during the day, which the crew does not have. Therefore, experimental 20 min sets were made opportunistically as time allowed. Reporting of PMTF Information The reporting of daily catches from the PMTF and our interpretation of run development were separated during the 2013 season (this was one of the recommendations made at the end of the 2012 season). This separation allowed catch summaries to be disseminated on the day of the catch, instead of waiting until the interpretations were complete. Catch summaries were sent via , using the catch table reconfigured in 2011, under the subject line PMTF Catch Update #_. The interpretation of these catches was developed in concert with inshore catch and escapement, and then disseminated the next day under the subject line PMTF Interpretation #_. We used the daily Replacement Index as an indicator of when peak abundance occurred at Port Moller, and to forecast peaks and drop offs in catch and escapement. The daily Replacement Index was reported in tabular format, as well as, in a figure to better illustrate the seasonal CPUE trend. Although we repeatedly warn about the unreliability of this approach, many stakeholders still rely on the cumulative Traditional Index to place the current year s catch trends into a historical perspective. Thus, we reported this cumulative index in tabular format along with date specific cumulative indexes and resulting total runs from previous years. We occasionally reported graphs comparing water temperatures, and stock specific daily indexes (following the release of genetic stock composition estimates by ADF&G). Various other graphs and analyses were performed that helped gauge run strength. Results and Discussion In 2013, the PMTF operated from 10 June to 6 July and caught 3,054 fish. Inseason daily catch updates were sent out the same evening that catches were reported from the test boat; interpretive reports were usually sent out the following day as planned (Table 1). Genetic stock composition estimates were forwarded to the distribution list soon after receipt from ADF&G. Notable inseason interpretations are summarized in Table 2. There were 305 addresses in the PMTF distribution list; this may represent a maximum number of recipients as some addresses may no longer be working, but the list was cleaned this year for stale addresses. Port Moller 2013, Page 7

13 Stock Composition ing The first genetic stock composition results were reported on 21 June, and there were a total of 5 reports throughout the season (Table 1). This year, stock composition results were delayed due to changes in flight logistics in and out of Port Moller, as well as, weather delays. Ideas to improve efficiency are currently being considered. The final genetic stock composition estimates cumulative through the season were reasonably accurate given each stock s run timing and magnitude (Appendix A Interpretation #16 Figure 3). Nominal discrepancies were the slight over-representation of the Ugashik stock (over by 3%) and under-representation of the Nushagak-Wood District (under by 5%). Combining stock composition estimates with the daily Replacement Index allowed for reasonable district-specific forecasts of increases and decreases in catch and escapement on 25 June (Interpretation #11 Figure 2). Likewise, district-specific forecasts made on 4 July and 7 July utilizing stock composition estimates were fairly accurate (Interpretations #14 Figure 2 and #15 Figure 2). Age Composition ing Traditionally, the test fishing net was comprised of four 50-fathom panels of 5⅛ inch mesh. In 2011, we changed the Traditional Net by replacing every other 5⅛ inch panel with 4½ inch mesh to comprise the Replacement Net in order to reduce size selectivity and render the age composition estimates at Port Moller more representative of returns to inshore fishing districts. For 2011 and 2012, the Replacement Net appears to have been successful in removing the majority of this selectivity as Port Moller catches indicated departures from the pre-season forecast of ocean age composition by June (Appendix A Interpretation #16 Figure 2). However, this year s age composition estimates from Port Moller slightly over-represented age 1.2s and 1.3s, while under-representing age 2.3s. Size selectivity creates the greatest catch discrepancy across the two ocean ages as they differ more in size than do the two freshwater ages. The direction of this year s age composition bias was not consistent within ocean ages (over for 1.3s and under for 2.3s) suggesting size selectivity was not the problem. More likely, stocks varied in their vulnerability to the test fishery, which combined with their differences in age compositions caused vulnerability across ages to vary as well. Given this explanation is correct, we can assume that ages 1.2 and 1.3 were over-represented at Port Moller, while age 2.3 was under-represented. Nevertheless, by 21 June age composition forecasts from the PMTF successfully indicated the direction of departure from the pre-season forecasts of ocean age proportions. Total Run ing The 2013 Bristol Bay sockeye run proved to be aberrant with respect to run timing both at the PMTF and inshore districts. Port Moller catches peaked on 21 June, nine days earlier than the average of 29 June. This early arrival caused the cumulative Traditional Index to be the highest on record up through 22 June before falling below average by 1 July (Appendix A Interpretation #16 Table 1). Likewise, total C+E peaked six-eight days early around 30 June Port Moller 2013, Page 8

14 ( average=6-7 July), making this year the earliest run on record (Interpretation #16 Figure 1). With respect to tracking the inshore run, the 2013 PMTF worked reasonably well with the exception of missing catches due to bad weather (Interpretation #16 Figure 4). Sliding the interpolated Replacement Index forward by a travel time of nine days illustrates why PM failed to detect the last movement of fish that peaked on 10 July. Clearly the interpolated values for missed fishing days 29 June-2 July were incorrect. Missing this many consecutive days at PM caused interpolations to be more tenuous and increased the chances that a major movement of fish passed undetected. Regardless, forecasts of million, 22 million, and 19 million fish were made on 28 June, 4 July, and 7 July, respectively (Table 2). The resulting run through 18 July was 23 million fish. Experimental Sets to Test for Gillnet Saturation For some time we have suspected that gillnet saturation may affect PMTF catches. That is, the gillnet will theoretically catch only so many fish and the rate at which fish are caught diminishes as catch accumulates. The idea is that fish caught initially spook subsequent fish, and this behavioral response occurs more and more as the net fills. Thus, a longer set results in a greater denominator (i.e., effort from catch/effort) without a concomitant increase in the numerator (catch). Some have argued that similar nets are fished inshore in the commercial fishery and catch far more fish than the PMTF net, so how could lower catches at Port Moller cause saturation? Our reasoning is that the catch environments are different between PMTF and Bristol Bay. At Port Moller, fish passing the test fishing transect are less densely distributed, in much clearer water, with no surrounding disturbance to spook them (the test fishing boat moves away from the net to minimize the effect of its presence). Fish density is greater in the fishing districts following a closure, the water visibility is zero at times, and every effort is made by the fleet to maximize catch, all of which contribute to higher catches per set as compared to the PMTF. Historical data provides us with an indication that saturation might be occurring. Over the last 23 years, annual runs have been as large as 55 million fish and daily inshore abundance (catch plus escapement) during the peak week as high as 6 million fish. During this time, only four sets yielded catches greater than 250 fish in the PMTF (n=2,066 sets; Appendix A Interpretation #8 Figure 2). Initial results from this year s experimental sets (n=8) indicate that net saturation begins to occur beyond a Replacement Index of 50 (Interpretation #11 Figure 3). These are only eight data points, and many more are needed across a range of tides and throughout the season before more definitive conclusions can be drawn with respect to saturation and at what catch levels it begins. If saturation does occur during peak passage days at Port Moller, then projections of run timing and magnitude based on the PMTF catches could be skewed substantially. At the least it would cause noise that obfuscated interpretation. Port Moller 2013, Page 9

15 Sea Surface Temperatures Another anomaly this year was the pattern in sea surface temperatures throughout the season (Interpretation #16 Figure 5). Temperatures began about average, rose quickly, and then ended below average by season s end. How temperatures affect travel time is poorly understood, but these findings are consistent with the pattern of C+E in inshore districts. Initial higher temperatures coincided with shorter travel times and earlier arrival inshore, followed by cooler temperatures lining up with longer travel times and later arrival. Nevertheless, it remains unclear how temperature affects sockeye behavior this far out in their migration with respect to run timing or susceptibility to the test fishing net. The date and location from which migration is initiated, as well as, swimming speed could both be influenced by temperature. Future Work and Recommendations As mentioned in last year s final report, and based on preliminary results from this year s experimental sets, we suspect gillnet saturation may inhibit the correlation between the magnitude of the PMTF catches and inshore C+E. We will continue to perform these experimental sets opportunistically in the coming years until enough data is collected to fully test these implications. We will continue research and development of the daily projection model. Anticipated improvements include district-specific Port Moller catch indices that utilize our recent knowledge of stock distributions across the test fishing transect based on trends from genetic stock composition estimates. As well, better methods for curve fitting and interpolations for missed fishing days due to weather are being researched. During 2012 we suggested that each daily catch index may be influenced by whether the test fishing boat was starting or ending each two day trip. Our hypothesis was that except for Station 6, the time of day at which each station was fished depends on whether the boat was inbound or outbound, which in turn determined the tidal stage occurring during each set. Because Station 6 was in the middle of the transect, it was fished around the same time (noon) each day. The daily tidal schedule changes throughout the season and certainly across years, but a quick overview of the daily Traditional Indexes patterns in previous years frequently show a daily cyclical pattern overridden by the season rise and fall in the migration. The variability removed from the catch index magnitude and timing pattern in each year may be substantial if this effect can be quantified. We may be able to demonstrate this phenomenon with the historical data going back to 1987 if a parameter that accounts for time-of-day proves significant. Beginning in 2010, the skipper of the R/V Pandalus began recording the coordinates of the gillnet buoy at the beginning and end of each set. The distance between the beginning and ending coordinates divided by the set duration should index the current velocity and direction. These data may be more informative than time-of-day, because they will account for any directional shifts in surface currents due to wind in addition to tide. Port Moller 2013, Page 10

16 Acknowledgements The R/V Pandalus was crewed by skipper Ted Jewel, engineer David Knight, and deck hand Margaret Archibald. Our fisheries technicians were Drew Stinnett and Amos Cernohouz. Guy Wade managed the Port Moller test fishery program, Matt Nemeth compiled and distributed the catch data, and Scott Raborn prepared the interpretations with assistance from Michael Link and Matt Nemeth. Fred West and Chuck Brazil (ADF&G) managed the scale aging operation in King Salmon and provided timely summaries of age composition from the Port Moller catch. Tyler Dann managed the genetic stock identification component of the project. We would like to especially thank Jeanne Boyle and Bob Murphy at the ADF&G office in Port Moller. Peter Pan Seafoods, specifically Mark Briski, George Sudar, and the dock crew were invaluable with respect to logistics. Without their help, the Port Moller test fishing program would be impossible. The 2013 Port Moller test fishing project was managed and staffed by the Bristol Bay Science and Research Institute (BBSRI). The project was funded by ADF&G, BBSRI, eight Bristol Bay processors, and by Bristol Bay driftnet fishermen (through the Bristol Bay Regional Seafood Development Association, or BB-RSDA). Processors included Canfisco (AGS/Leader Creek), E&E Foods, Icicle Seafoods, North Pacific Seafoods, Ocean Beauty, Peter Pan Seafoods, and Trident Seafoods. Although none of this is possible without the help of ADF&G personnel and funding, the interpretations contained in the updates were by no means official ADF&G interpretations of the Port Moller test fishery data or the Bristol Bay run. References Eggers, D. M., and S. M. Fried Bristol Bay salmon test fishing projects. Alaska Department of Fish and Game. Technical Data Report No pp. Flynn, L. and R. Hilborn Test fishery indices for sockeye salmon (Oncorhynchus nerka) as affected by age composition and environmental variables. Canadian Journal of Fisheries and Aquatic Resources 61: Hansen, M. J., R. G. Schorfhaar, and J. H. Selgeby Gill-net saturation by lake trout in Michigan waters of Lake Superior. North American Journal of Fisheries Management 18: Helton, D An analysis of the Port Moller offshore test fishing forecast of sockeye and chum salmon runs to Bristol Bay, Alaska. M.S. Thesis, University of Washington, Seattle. Thesis No Port Moller 2013, Page 11

17 Koo, T.S.Y Age designation in salmon. Pages in T.S.Y. Koo, editor. Studies of Alaska red salmon. University of Washington Publications in Fisheries, New Series, Volume I, Seattle. Minns, C. K., and D. A. Hurley Effects of net length and set time on fish catches in gill nets. North American Journal of Fisheries Management 8: Randall, R. C Offshore test fishing in Bristol Bay, Alaska Department of Fish and Game. Technical Data Report No pp. Rotherham, D., C. A. Gray, M. K. Broadhurst, D. D. Johnson, L. M. Barnes, and M. V. Jones Sampling estuarine fish using multi-mesh gill nets; effects of panel length and soak and setting times. Journal of Experimental Marine Biology and Ecology 331: Port Moller 2013, Page 12

18 Table 1. Timing of information received and delivery of daily updates for the 2013 PMTF. Date of test fishery sampling 10-Jun 11-Jun Date and time of test fishery catch update 10-Jun 8:00 PM 11-Jun 6:45 PM Date and time of interpretation Date and time stock composition estimates received from ADF&G (corresponds to sampling date) Date and time stock composition estimates forwarded 12-Jun 12-Jun 8:07 PM 13-Jun 12:24 PM 13-Jun 13-Jun 7:07 PM 14-Jun 3:49 PM 14-Jun 14-Jun 10:47 PM 15-Jun 10:05 AM 15-Jun NO FISHING 16-Jun 16-Jun 9:33 PM 17-Jun 12:47 PM 17-Jun 17-Jun 9:19 PM 18-Jun 1:03 PM 21-Jun 12:01 PM 21-Jun 1:23 PM 18-Jun 18-Jun 9:37 PM 19-Jun 1:45 PM 19-Jun 19-Jun 8:46 PM 20-Jun 1:43 PM 22-Jun 11:30 AM 22-Jun 11:48 PM 20-Jun 20-Jun 9:20 PM 21-Jun 3:31 PM 21-Jun 21-Jun 5:46 PM 23-Jun 10:11 AM 24-Jun 6:01 PM 24-Jun 9:32 PM 22-Jun NO FISHING 23-Jun 23-Jun 9:04 PM 24-Jun 7:47 PM 24-Jun 24-Jun 9:41 PM 25-Jun 7:12 PM 25-Jun 25-Jun 11:18 PM 26-Jun 8:44 PM 26-Jun 26-Jun 9:59 PM 26-Jun 9:44 PM 28-Jun 9:24 PM 28-Jun 10:13 PM 27-Jun 27-Jun 11:21 PM 28-Jun 7:50 PM 28-Jun 28-Jun 10:54 PM 28-Jun 7:50 PM 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul NO FISHING 3-Jul 8:46 PM 4-Jul 5:03 PM 4-Jul 4-Jul 10:26 PM 7-Jul 3:40 PM 5-Jul 5-Jul 9:51 PM 7-Jul 3:40 PM 6-Jul 6-Jul 8:16 PM 7-Jul 3:40 PM 10-Jul 5:04 PM 11-Jul 1:10 AM SEASON SUMMARY 15-Jul 5:40 PM 11-Jul 5:23 PM 12-Jul 10:00 AM Port Moller 2013, Page 13

19 Table 2. Summary of substantive comments and predictions made in the daily interpretations of the 2013 PMTF. Interpretation # Date Comment/prediction Did the prediction(s) come true? Introduced the crew and technicians. Explained the change in reporting format for catch summaries were to 10-Jun be reported the same day of sampling, but interpretations would follow the next day. 2 Noted that the catch indices were higher than usual, indicative that the run was early, larger than expected, or 14-Jun both Jun Reiterated the higher than average catches. Provided graph showing poor relationship between water temperature and the cumulative Traditional Index, and therefore, the inability of water temperature to predict run timing Jun Predicted that the yearend 3-ocean component would be 59% of the run (PREDICTION), but cautioned that sampling error could explain the discrepancy with ADF&G's forecast of 52%. Reported finding from a recent ADF&G publication on stock composition throughout the season and across PM stations. No--3-ocean component was 70% of the run, so we under estimated the deviation from pre-season forecast 5 Explained differences between calculation of the Traditional and Replacement Indexes and how they are used. 18-Jun Commented on the general pattern in catches across stations throughout the season. 7 Further commented on the patchiness of catches across stations early in the run and illustrated this pattern for 19-Jun Jun Introduced the possibility for gillnet saturation in the test fishing net with an analysis of historical catches. 10 Noted the unprecedented magnitude of cumulative C+E for this date. Speculated that the run must be larger than 24-Jun expected else have an unprecedented early run timing Jun Made general predictions about C+E in each district for the next few days based on PM catch index and stock composition estiamtes (PREDICTION). Provided initial results of 2013 net saturation experiment, which indicated considerable evidence for net saturation. More or less Estimated peak day of passage at PM was 21 June. Provided alternative projections of inshore run strength Jun timing at PM versus that for inshore. depending on travel time b/n PM and inshore. Estimated the most likely date for the inshore peak based on No--the inshore peak occurred on the historical travel times would be between 2-4 July (PREDICTION). Demonstrated the relationship between run 30 June Jun Projected the total to be millon (PREDICTION 1) and peak on 30 June (PREDICTION 2). 1. Close--Final on 18 July=23 million 2. Yes 14 4-Jul Modified the total run forecast model and changed the total run projection to 22 million (PREDICTION). Close 15 7-Jul Noted end of test fishing for 2013 on 6 July. Combining the smoothed normal curves for each district and assuming interpolated catches 29 June-2 July being accurate, the total run projection was reduced to 19 million. Final interpretation. Reviewed the performance of the test fishery with respect to forecasting age and stock compositions, as well as, total run. Provided acknowledgements. Biased low Port Moller 2013, Page 14

20 Figure 1. Locations of fishing stations for the Port Moller Test Fishery relative to Bristol Bay. Port Moller 2013, Page 15

21 Appendix A BBSRI s Daily Catch Updates and Interpolation from the Port Moller Test Fishery, 2013 Port Moller 2013, Page 16

22 From: Scott Raborn To: Scott Raborn Subject: Port Moller Test Fishery Catches--June 10, 2012 Date: Monday, June 10, :20:00 PM Attachments: PortMollerCatches_Jun10_2013.pdf Hello Everyone, The 2013 Port Moller sockeye test fishery began today. The State-owned and operated R/V Pandalus was able to fish Stations 2, 4, 6, 8, and 10. Tonight, they will anchor at Station 10 and resume fishing Stations 10-2 on the return to Port Moller. This year the Pandalus is skippered by Ted Jewel with engineer David Knight and deck hand Margaret Archibald. Our fisheries technicians are Drew Stinnett and Amos Cernohouz. Guy Wade is the project manager and is on board for this first trip. Data analyses will be conducted by Dr. Scott Raborn, who will author the daily updates along with Matt Nemeth and Michael Link. The reporting schedule will be different this year. Raw catches, their mean lengths, and station water temperatures will be reported soon after sampling is completed each afternoon/evening. However, calculation of the catch indexes (Replacement and Traditional), additional analyses concerning age composition and run strength, and an overall interpretation will be reported the following day usually before noon Alaska time. Attached is the catch summary table for today, and you can expect an updated version of this table every night for the next days (weather days notwithstanding). Interpretations will begin in the next few days once enough data are available. We will reference each catch summary and interpretation by the date to which they apply in the subject line and in the attached file name. The first genetic stock composition estimates will likely be released by ADF&G on June 20 and will cover sampling dates June As always, we will forward these estimates and the stock composition estimates to the Port Moller distribution list as soon as they are available. All for now, Scott <<...>> Port Moller 2013, Page 17

23 Port Moller Test Fishery Catches for June 10, 2013 Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 10-Jun Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Totals Weather Overcast; seas 1-2 ft; wind from NW 5-15 a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10-Jun Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Weighted means 542 c Length is measured as mid-eye-fork-length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age-2 fish=504 mm; 3-ocean=571 mm. 6/10/2013 8:00 PM Alaska Time Page 1 of 1 Port Moller 2013, Page 18

24 Port Moller Test Fishery Catches for June 11, 2013 Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 10-Jun Jun Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Totals Page 1 of 1 Port Moller 2013, Page 19 Weather Overcast; seas 1-2 ft; wind from NW 5-15 Overcast; seas 2 ft; wind from NW 5 kts 6/11/2013 6:45 PM Alaska Time a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10-Jun Jun Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Weighted means 540 c Length is measured as mid-eye-fork-length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age-2 fish=504 mm; 3-ocean=571 mm.

25 Port Moller Test Fishery Catches for June 12, /12/2013 8:07 PM Alaska Time Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 10-Jun Jun Jun Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Totals Weather Overcast; seas 1-2 ft; wind from NW 5-15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10-Jun Jun Jun Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Weighted means Page 1 of c Length is measured as mid-eye-fork-length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age-2 fish=504 mm; 3-ocean=571 mm. Port Moller 2013, Page 20

26 Interpretation of Port Moller Test Fishery Catches for June 12, /13/ :24 PM Alaska Time The daily catch indexes are reported in Table 1. Thus far, the daily Replacement Index is higher than the previous two years (Figure 1); as well, the cumulative Traditional Index is above the average for 1990 to 2012 (Table 2). Sea surface temperatures have been average and above for this time of year (Figure 2). Table 1. Estimated daily and cumulative catch indexes (Replacement and Traditional) for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun Jun Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent Traditional CPUE a by station Traditional Index Date Daily b Cumulative 10-Jun Jun Jun Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent Page 1 of 3 Port Moller 2013, Page 21

27 6/13/ :24 PM Alaska Time Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Page 2 of 3 Port Moller 2013, Page 22

28 6/13/ :24 PM Alaska Time 100 CPUE (catch per 200 fathom hours) Jun 15-Jun 20-Jun 25-Jun 30-Jun 5-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Sea surface temperatures ( ) Temperature (C) Figure 2. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for cooler years, which are represented by various colors). Page 3 of 3 Port Moller 2013, Page 23

29 Port Moller Test Fishery Catches for June 13, 2013 Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Weather 10-Jun Overcast; seas 1-2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12-Jun Overcast; seas 1 ft; wind from NE 3 kts 13-Jun % cloud cover; seas 1 ft; winds variable 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10-Jun Jun Jun Jun Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul Weighted means 545 c Length is measured as mid-eye-fork-length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age-2 fish=504 mm; 3-ocean=571 mm. Page 1 of 1 Port Moller 2013, Page 24 6/13/2013 7:07 PM Alaska Time

30 Interpretation of Port Moller Test Fishery Catches for June 13, 2013 The daily catch indexes increased substantially in the past two days (Tables 1 and 2), and the 2013 daily Replacement Index is far outpacing 2011 and 2012 (Figure 1). As well, the cumulative Traditional Index for this year is the highest on record since While these catches are encouraging, as always, we caution against weighting these results too heavily, especially since they are mainly driven by only two days of data. There are several potential explanations for the results to date: (1) the run is early, but of unknown size (2) the run is not early, but larger than expected, or (3) the Port Moller fish-per-index (FPI) is low this year; in other words, the number of fish inshore that the catch indexes at Port Moller represent is lower than average. Any one or combination of these scenarios could be happening. Sea surface temperatures are roughly in the middle of the historical range, but have climbed at an unusually high rate in the past two days (Figure 2). Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun 11-Jun Jun Jun Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Table 2. Estimated daily and cumulative Traditional Index for the 2013 Port Moller test fishery. Traditional CPUE a by station Traditional Index Date Daily b Cumulative 10-Jun Jun Jun Jun Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in the 5 1/8 inch mesh shackles and standardized to a 200 fathom net fished for 60 minutes. b Traditional daily index is based on catch/hour/100 fathoms at Stations 2, 4, 6, and 8 only (Station 8 weighted double and Station 10 not used). Note: red values were interpolated from other observed catches. Port Moller 2013, Page 25

31 Table 3. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 26

32 100 CPUE (catch per 200 fathom hours) Jun 15-Jun 20-Jun 25-Jun 30-Jun 5-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Sea surface temperatures ( ) Temperature (C) Figure 2. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for warmer years and 2012, which are represented by various colors). Port Moller 2013, Page 27

33 Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 10 Jun Jun Jun Jun Jun Jun 16 Jun 17 Jun 18 Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weather Overcast; seas 1 2 ft; wind from NW 5 15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10 Jun Jun Jun Jun Jun Jun 16 Jun 17 Jun 18 Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 547 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 28

34 PMTF Interpretation #3 for catch on June 14, 2013 The test fishing crew skipped Station 10 yesterday and started at Station 8 upon threat of bad weather to ensure fishing Stations 2-8. Plans are to fish all stations today (June 15) weather permitting. Daily catch indices continue to climb (Tables 1 and 2), with the cumulative Traditional Index for this year well above the historical maximum for this date (Table 3). High catch numbers and warmer than average sea surface temperatures cause one to wonder about how previous warm years compare. The cumulative Traditional Index appears to increase with temperature, but the relationship is noisy (Figure 1). The lack of relationship in Figure 2 illustrates why we reiterate caution when extrapolating run size from Port Moller catches this early in the season. The first genetic stock composition estimates will likely be released by ADF&G on June 20 and will cover sampling dates June As always, we will forward these estimates and the stock composition estimates to the Port Moller distribution list as soon as they are available. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun Jun Jun Jun Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Table 2. Estimated daily and cumulative Traditional Index for the 2013 Port Moller test fishery. Traditional CPUE a by station Traditional Index Date Daily b Cumulative 10-Jun Jun Jun Jun Jun Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in the 5 1/8 inch mesh shackles and standardized to a 200 fathom net fished for 60 minutes. b Traditional daily index is based on catch/hour/100 fathoms at Stations 2, 4, 6, and 8 only (Station 8 weighted double and Station 10 not used). Note: red values were interpolated from other observed catches. Port Moller 2013, Page 29

35 Table 3. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 30

36 Figure 1. Historical relationship between the Port Moller cumulative Traditional Index on 14-June and sea surface temperature for this date. Year is centered on each data point. Figure 2. Historical relationship between total run (catch + escapement) and the Port Moller cumulative Traditional Index on 14-June. Year is centered on each data point. Port Moller 2013, Page 31

37 From: To: Subject: Date: Matt Nemeth PMTF Catch Update #6 - June 15, 2013 (NO FISHING) Saturday, June 15, :43:42 PM Hello all, The boat was weathered out today, so no catches to report. Tomorrow s forecast for Port Heiden to Cape Sarichef is better than today s. PKZ PORT HEIDEN TO CAPE SARICHEF 400 PM AKDT SAT JUN TONIGHT...SE WIND 20 KT. SEAS 4 FT. GUSTS TO 35 KT OUT OF BAYS AND PASSES. PATCHY FOG. RAIN..SUN...SE WIND 15 KT. SEAS 4 FT. PATCHY FOG. RAIN..SUN NIGHT AND MON...SE WIND 20 KT. SEAS 4 FT..MON NIGHT...VARIABLE WIND 10 KT. SEAS 2 FT..TUE AND WED...NW WIND 15 KT. SEAS 4 FT..THU...VARIABLE WIND 10 KT. SEAS 5 FT. -Matt Nemeth Port Moller 2013, Page 32

38 Raw catch a of sockeye by station and mesh size Totals Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 10 Jun Jun Jun Jun Jun Jun Jun Jun 18 Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted means Water temperature ( C) by station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10 Jun Jun Jun Jun Jun Jun 16 Jun Jun 18 Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 544 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm Weather Overcast; seas 1 2 ft; wind from NW 5 15 kts Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds var. No Fishing boat weathered out 25% cloud cover; seas ft; SE5 to NE5 Port Moller 2013, Page 33

39 PMTF Interpretation #4 for catches through June 16, 2013 The test fishing crew was unable to fish on June 15 and index values for this date are interpolated. The daily Replacement Index fell yesterday (June 16; Table 1 and Figure 1), but daily ups and downs are common, and we expect PM catches to show an overall increasing trend in the coming days as the run builds. The cumulative Traditional Index continues to be the highest on record since 1990 (Table 2). Water temperatures continue to be average for this date. Having sampled 713 fish for lengths thus far in the season, we can begin to predict the age composition of the total run. In years past, PM has exhibited some predictive power with regards to age composition based on observed mean lengths (Figure 2). From this relationship, we predict 59% of the run will be 3-ocean fish, which is slightly greater than ADFG s forecast of 52%. This difference is well within the margin of error and does not necessarily indicate the run is developing out of line with the forecast. The first genetic stock composition estimates will be released by ADF&G in a few days. In anticipation of these results, we are passing along a recent publication by Dann et al. (2013) in the Canadian Journal of Fisheries and Aquatic Sciences. This paper does a nice job of illustrating how the stocks were apportioned across time and the test fishing transect for years (see Figure 5 in the attached reprint). We have taken the liberty of graphing some of their results (Table 2 in the reprint) with Figures 3 and 4 below. These results confirm our general suspicions about how the stocks move past Port Moller and line up with general catch patterns across the season/districts. Early on Egegik and Nushagak districts tend to dominate and decline throughout the 30-day test fishery while the Naknek-Kvichak and Ugashik districts start off slower, but gradually increase during the test fishery. Fish from the Naknek-Kvichak, Nushagak, and Togiak districts tend to be distributed further offshore, while those from the Egegik and Ugashik districts are closer inshore. That said, fish from every district are caught at each station and Egegik especially tends to be more evenly distributed than the rest. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun 11-Jun Jun 13-Jun Jun Jun Jun Jun 18-Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Port Moller 2013, Page 34

40 Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 35

41 100 CPUE (catch per 200 fathom hours) Jun 15-Jun 20-Jun 25-Jun 30-Jun 5-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Proportion of 3-ocean fish in yearend C+E Mean length of PM catch in 5 ⅛ mesh through 16-June Figure 2. Predicted proportion of ocean age-3 sockeye in the year-end catch + escapement based on the mean length of fish caught in the PMTF through June 16. Port Moller 2013, Page 36

42 Proportion of Port Moller catch Ugashik Naknek-Kvichak-Alagnak Togiak Egegik Nushagak-Wood-Igushik June Jun-Jul July Figure 3. Results from the top part of Table 2 in Dann et al. (2013) showing the seasonal change in stock proportions (all stations combined) for the PMTF. These estimates were based on PMTF catch samples. Port Moller 2013, Page 37

43 Proportion of Port Moller catch Proportion of Port Moller catch Ugashik Egegik Naknek-Kvichak-Alagnak Nushagak-Wood-Igushik Togiak 2 and and 12 2 and and 12 Station Early Late Figure 4. Results from the bottom part of Table 2 in Dann et al. (2013) showing the change in stock proportions across stations for the PMTF. Early (top graph) and late (bottom) were defined as before and after June 25. These estimates were based on PMTF catch samples. Port Moller 2013, Page 38

44 Raw catch a of sockeye by station and mesh size Daily Cum Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Water temperature ( C) by means station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ mean 10 Jun Jun Jun Jun Jun Jun 16 Jun Jun Jun 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 544 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 39

45 PMTF Interpretation #5 for catches through June 17, 2013 You may notice that the interpolated values for Station 8 on June 14 and for all stations on June 15 have changed a little from yesterday. This change is because with the addition of each new day s catches, the model used to generate these interpolations changes a little. As the season progresses these values will stabilize, but the overall index values remain more or less the same as it is. Catches are slightly higher inshore, but mostly centered on Station 6 this year, which produced the highest CPUE thus far in the season yesterday (Table 1). While the daily Replacement Index was not quite as high as it was on June 14, the overall trend suggests the run is building as we might expect (Figure 1). The nominal catches for stations adjacent to Station 6 suggest the run is still patchy past Port Moller. Early in the season the beginning edge of the run often produces this pattern, but as the run builds catches tend to be more consistent across stations. Catches then return to being patchy with the ending tail. The daily Traditional Index increased from the drop off on June 16 (Table 2). The daily Traditional Index fluctuates more than the daily Replacement index (Figure 2) for three reasons. First, the Traditional Index is based only on the two 50-fathom panels of the replacement net, whereas the Replacement Index is based on the entire net; thus, the Replacement Index is based on more effort. Second, the Traditional Index gives double weight to Station 8, while the Replacement Index gives equal weight to all; fluctuations in Station 8 catches will therefore cause more fluctuation in the Traditional Index. Third, the Traditional Index only uses Stations 2-8, whereas the Replacement Index uses all five stations (usually 2-10); again more effort stabilizes the ups and downs. The real utility of each index is different. For the Traditional Index, the cumulative value is useful for comparison to years past (Table 3). For the Replacement Index, the daily value is better for assessing the peak day of passage at Port Moller and correlating test fish catches with inshore catch + escapement inseason. As in previous years, we provide a brief history of the catch indices and nets in Appendix A. Water temperatures are about 1.5 degrees above average for this date (Figure 3). Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date 10-Jun Daily b 7 Cumulative 7 11-Jun Jun 13-Jun Jun Jun Jun Jun Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Port Moller 2013, Page 40

46 Table 2. Estimated daily and cumulative Traditional Index for the 2013 Port Moller test fishery. Traditional CPUE a by station Traditional Index Date Daily b Cumulative 10-Jun Jun Jun Jun Jun Jun Jun Jun Jun 19-Jun 20-Jun Totals Percent a CPUE = the number of fish caught in the 5 1/8 inch mesh shackles and standardized to a 200 fathom net fished for 60 minutes. b Traditional daily index is based on catch/hour/100 fathoms at Stations 2, 4, 6, and 8 only (Station 8 weighted double and Station 10 not used). Note: red values were interpolated from other observed catches. Table 3. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 41

47 100 CPUE (catch per 200 fathom hours) Jun 15-Jun 20-Jun 25-Jun 30-Jun 5-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Daily index scaled to a maximum value of one Traditional Index Replacement Index Jun 11-Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun Sampling date Figure 2. Comparison of PMTF daily catch indices for Each index has been scaled so that the maximum value observed thus far is one. Port Moller 2013, Page 42

48 14.0 Sea surface temperatures ( ) Temperature (C) Figure 3. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for cooler years and 2012, which are represented by various colors). Port Moller 2013, Page 43

49 APPENDIX A Description of the Traditional and Replacement Indexes The Traditional Index is the more complex of the two owing to how it evolved with the test fishery. Though it is difficult to decipher from historical documents we believe the original Port Moller Index ( ) for a given day was the average of catch-per-unit-effort (CPUE=sockeye per 100 fathom-hours) across stations (1-11) and cumulative over the course of the season through a given day s catch. Missing station points were linearly interpolated from the best available data as one report says. Beginning in 1985, the Port Moller Index was standardized to the sum of CPUE for Stations 2, 4, 6, and 8. Missing station points were interpolated by averaging the station specific daily indices from the two days prior to and the two days after the missing station point(s). In 1995, CPUE was highest at Station 8 causing suspicion that a substantial proportion of the run was further offshore. As a consequence, the cumulative index was altered in 1996, giving double weight to Station 8 to account for fish passage further offshore. There was concern about causing confusion among laypersons because the magnitude of the Traditional Index would change from what had been released in previous years ( ). To minimize this degree of change, the sum across stations 2-8 (with Station 8 counted twice) was multiplied by 0.8, which brings it back to its original scale, but allows Station 8 to have twice the influence of any other station. In years prior to 2011, the Traditional Index was based on four 50 fathom shackles of 5⅛ inch stretch mesh (a 200 fathom net we termed the Traditional Net). Beginning in 2011, we have used what we termed the Replacement Net, which is still a 200 fathom net, but the 50 fathom shackles are now alternated between 4½ inch and 5 ⅛ inch mesh (see Update #2 for a diagram of each net). This means that beginning in 2011, 100 fathoms of 5⅛ inch mesh is being fished instead of 200 fathoms as with the Traditional Net. Using catches from the 4½ inch mesh to make up the difference would bias the estimate; instead, we deemed it better to double the catches from the 5 ⅛ inch shackles for each set to formulate the Traditional Index and make it comparable to years prior to One final point we want to clarify about the Traditional Index is that in the index Table 2 above, CPUE for each station-day combination is given in fish per 200 fathoms, but the daily Traditional Index is based on CPUE for each station standardized to 100 fathoms. We continue to report the daily index this way to be consistent, but individual station CPUE s are now reported on a per 200 fathom basis because readers seem to relate to and interpret that easier. When we switched to the Replacement Net last year, we decided to simplify things with the Replacement Index. This daily index is simply the average CPUE (based on 200 fathoms) across stations from Stations So on a given day, you can interpret the daily Replacement Index as the average catch across stations for that day from a 200 fathom net standardized to a fishing time of one hour. Instead of using adjustments to the equation or statistical models to account for fish beyond Station 8, we simply include Station 10, and if need be, shift the sampling to stations further out when the run appears to be distributed further offshore. Thus, the daily Replacement Index is based on the five stations fished that day regardless of which stations they were. We also wanted to remove the double weighting of Station 8 because day-to-day variability in catch at this station can cause the Traditional Index to bounce around more than it should. We see the Traditional Index varying more than the Replacement Index in Figure 2 above. Again, part of this is undoubtedly because the Traditional Index is now based on 100 fathoms of gillnet (remember that catches from only the 5 ⅛ inch mesh is used in its calculation) across 4 stations, whereas the Replacement Index is based on 200 fathoms across 5 stations. Still, they tend to fall and rise in a similar fashion and index the same general patterns (Figure 2). Port Moller 2013, Page 44

50 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 544 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 45

51 PMTF Interpretation #6 for catches through June 18, 2013 The daily Replacement Index fell off a bit today, but remains within the range of daily fluctuations we typically see (Table 1). For example, 2011 exhibited a similar pattern in the daily Replacement Index through this date and ultimately continued to build (Figure 1). The cumulative Traditional Index remains the highest on record (Table 2). ADF&G s gene lab is still waiting for samples from the 16 th and 17 th, which are scheduled to arrive early this afternoon. If they receive the samples today (all indications are that they will), the first genetic stock composition estimates should be released sometime tomorrow. We will forward this report to the PMTF distribution list as soon as it becomes available. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun 12-Jun Jun Jun 15-Jun Jun Jun 18-Jun Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Port Moller 2013, Page 46

52 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul 2012 Figure 1. The daily Replacement Index for 2011, 2012, and Port Moller 2013, Page 47

53 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 545 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 48

54 PMTF Interpretation #7 for catches through June 19, 2013 The daily Replacement Index has been relatively constant over the past four days (Table 1), but with slight daily fluctuations similar to 2011 and The cumulative Traditional Index remains the highest on record, but this margin seems to be narrowing (Table 2). We occasionally mention how the catch indices are more variable across stations earlier in the season. We do this because variability is likely yet another indicator albeit coarse of run stage (Figure 2, below). The first few days of the run tend to exhibit low catches across all stations, presumably because the first few pushes of fish from the leading edge of the run tend to be patchy. As the run builds, catches among stations become more even (less variable) from about 20 June to the 3 July, when the bulk of the run is passing Port Moller. Afterwards, the index returns to increasing variability across stations with the ending tail. We say all of this to assist those who may live and die by the daily ups and downs of the catch indices. For instance, the daily Replacement Index on the 17 June was the second highest this year, but only because of a single high catch at Station 6. The index may very well just represent a similar or lower magnitude of fish passing the day before when the daily index was lower, but more even across the stations (Table 1). Genetics samples were delayed and will not make it to the gene lab until this afternoon. More than likely, the first stock composition estimates will be released tomorrow. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun Jun Jun 14-Jun Jun 16-Jun Jun 18-Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Port Moller 2013, Page 49

55 Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Cumulative Traditional Index Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 28-Jun Jun Sampling date 30-Jun Jul Jul Jul Jul Jul Total run (millions) Max Mean Min 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul 2012 Figure 1. The daily Replacement Index for 2011, 2012, and Port Moller 2013, Page 50

56 100% Relative variability across stations 80% 60% 40% 20% % 11-Jun 12-Jun 13-Jun 14-Jun 15-Jun 16-Jun 17-Jun 18-Jun 19-Jun 20-Jun 21-Jun 22-Jun 23-Jun 24-Jun 25-Jun 26-Jun 27-Jun 28-Jun 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul 4-Jul 5-Jul 6-Jul 7-Jul Sampling date Figure 2. Variability in the Replacement Index across stations over the course of the season for Values within each year were standardized to a maximum value of 100%. Port Moller 2013, Page 51

57 From: To: Subject: Date: Attachments: Matt Nemeth PMTF Catch Update #11 - June 20, 2013 catches Thursday, June 20, :29:25 PM PortMollerTF_CatchUpdate#11 - June 20, 2013 catch.pdf Hi All, Tonight s catch update uses actual catches at Stations 2-8, plus interpolations at Station 10. There s a small chance that I will update this later tonight to include updated numbers at Station 10, depending on what the crew does tonight/tomorrow. As for the past few days, daily and cumulative indices are included as the shaded columns. Finally, there s a gale warning for Port Heiden to Cape Sarichef on Friday night, so it s possible that some sets may be interrupted. PKZ PORT HEIDEN TO CAPE SARICHEF 400 PM AKDT THU JUN GALE WARNING FRIDAY AND FRIDAY NIGHT....TONIGHT...SE WIND 15 KT. SEAS 3 FT..FRI...SE WIND 25 KT INCREASING TO 35 KT IN THE AFTERNOON. SEAS 5 FT BUILDING TO 9 FT IN THE AFTERNOON. RAIN..FRI NIGHT...SE WIND 35 KT. GUSTS TO 50 KT OUT OF BAYS AND PASSES. SEAS 10 FT. RAIN..SAT...SE WIND 25 KT. SEAS 7 FT..SAT NIGHT...SE WIND 20 KT. SEAS 6 FT..SUN...N WIND 15 KT. SEAS 5 FT..MON AND TUE...NW WIND 15 KT. SEAS 6 FT. Good fishing to all, Matt Port Moller 2013, Page 52

58 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light NE winds 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 545 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 53

59 From: To: Subject: Date: Attachments: Matt Nemeth PMTF Catch Update #12 - June 20, 2013 catches UPDATED Friday, June 21, :06:29 AM PortMollerTF_CatchUpdate#12 - June 20, 2013 catch, version2.pdf Hi All, The Pandalus crew decided to set Station 10 tonight, and ended up with the largest Station 10 catch of the season thus far. I have attached a revised catch update; the Replacement Index rises, while the Traditional remains the same. The crew plans to fish from Station 8 in towards 2 tomorrow, hopefully avoiding the seas that are forecasted to build. As usual, Scott will follow with his analysis and interpretation tomorrow. Matt From: Matt Nemeth Sent: Thursday, June 20, :20 PM To: 'mnemeth@lgl.com' Subject: PMTF Catch Update #11 - June 20, 2013 catches Hi All, Tonight s catch update uses actual catches at Stations 2-8, plus interpolations at Station 10. There s a small chance that I will update this later tonight to include updated numbers at Station 10, depending on what the crew does tonight/tomorrow. As for the past few days, daily and cumulative indices are included as the shaded columns. Finally, there s a gale warning for Port Heiden to Cape Sarichef on Friday night, so it s possible that some sets may be interrupted. PKZ PORT HEIDEN TO CAPE SARICHEF 400 PM AKDT THU JUN GALE WARNING FRIDAY AND FRIDAY NIGHT....TONIGHT...SE WIND 15 KT. SEAS 3 FT..FRI...SE WIND 25 KT INCREASING TO 35 KT IN THE AFTERNOON. SEAS 5 FT BUILDING TO 9 FT IN THE AFTERNOON. RAIN..FRI NIGHT...SE WIND 35 KT. GUSTS TO 50 KT OUT OF BAYS AND PASSES. SEAS 10 FT. RAIN..SAT...SE WIND 25 KT. SEAS 7 FT..SAT NIGHT...SE WIND 20 KT. SEAS 6 FT..SUN...N WIND 15 KT. SEAS 5 FT..MON AND TUE...NW WIND 15 KT. SEAS 6 FT. Good fishing to all, Port Moller 2013, Page 54

60 Matt Port Moller 2013, Page 55

61 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 545 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 56

62 PMTF Interpretation #8 for catches through June 20, 2013 There were better catches in general across the fishing transect today (most notably at Station 10) increasing the daily Replacement Index (Table 1), again in line with 2011 (Figure 1). As well, the cumulative Traditional Index remains high for this date (Table 2). For some time we have suspected that gillnet saturation may affect PMTF catches. That is, the gill net will theoretically catch only so many fish and the rate at which fish are caught diminishes. The idea is that fish caught initially spook subsequent fish, and this behavioral response occurs more and more as the net fills with fish. Some would argue that similar nets are fished inshore and catch far more fish than the test fishing net, so how could the lower catches at Port Moller cause saturation? Our reasoning is that fish passing the test fishing transect are less densely distributed, in much clearer water, with no surrounding disturbance to spook them (the test fishing boat moves away from the net to minimize the effect of its presence). Fish stack up in the fishing districts following a closure, the water visibility is zero at times, and every effort is made by the fleet to maximize catch, all of which contribute to higher catches per set as compared to the test net at PM. Across 23 years worth of sets by the PMTF, (n=2,066 sets) only four sets yielded catches greater than 250 fish (Figure 2), and the years used to generate this graph included some of the largest runs on record ( ). If saturation does occur during peak passage days at Port Moller, then projections of run timing and magnitude based on PMTF catches could be skewed substantially. At the least it would cause noise that obfuscated interpretation. This year, we decided to test this hypothesis by making shorter sets immediately after the routine 60 minute sets. These extra sets require more time during the day, which the crew does not have. To offset this extra time we planned on having the crew skip Station 10 yesterday because of the low catches observed there thus far. Instead they were to double set at each station (60 min set followed by a 20 min set), which they did, and then sample Station 10 this morning (21 June). The Station 10 catch from yesterday was to be interpolated. However, due to the rough weather forecasted for sometime today the crew went ahead and fished Station 10 late yesterday evening. To ensure Stations 8-2 were fished today, they steamed back to Station 8 before anchoring for the night. Therefore, Station 10 was not fished this morning and will be interpolated for today. With only four data points, yesterday s experiment yielded some interesting results consistent with the scenario of more fish being caught at the beginning of a given set versus the end. If saturation is occurring, a longer set results in a greater denominator (i.e., effort from catch/effort) without a concomitant increase in the numerator (catch). Again, these are only four data points and many more must be collected across a range of tides and throughout the season before more definitive conclusions can be drawn with respect to saturation. We mention these preliminary results now to explain why we may opportunistically skip marginal stations in the days to come, using the time saved to repeat sets at other stations and further explore the topic of net saturation. We do not intend on skipping stations following a weather day or if catches at these stations had picked up the day before. Our goal is to maintain the current information content ordinarily provided by the PMTF while pursuing avenues that might increase our understanding of catch patterns and allow us to improve its future performance. Port Moller 2013, Page 57

63 Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun Jun Jun Jun Jun Jun 17-Jun Jun Jun 20-Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 58

64 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Number of sets Raw catch Figure 2. Frequency of raw catches from sets made during the PMTF ( and ; 23 years). Port Moller 2013, Page 59

65 From: To: Subject: Date: Attachments: Matt Nemeth PMTF Catch Update #13 - June 21, 2013 catches Friday, June 21, :56:43 PM PortMollerTF_CatchUpdate#13 - June 21, 2013 catch.pdf Hello everyone, Busy day in PMTF land. With the gale warning coming in, the Pandalus fished Stations 8, 6, and 4, before breaking off and heading in as the weather worsened. The weather is supposed to deteriorate further tomorrow, before improving on Sunday (see below). As usual, Scott will follow with his interpretation tomorrow morning -Matt PKZ PORT HEIDEN TO CAPE SARICHEF 400 PM AKDT FRI JUN GALE WARNING TONIGHT....TONIGHT...SE WIND 40 KT. GUSTS TO 50 KT OUT OF BAYS AND PASSES. SEAS 11 FT. RAIN..SAT...SE WIND 25 KT. SEAS 7 FT. RAIN..SAT NIGHT...SE WIND 20 KT. SEAS 6 FT. RAIN..SUN...NE WIND 15 KT. SEAS 6 FT..SUN NIGHT...N WIND 15 KT. SEAS 5 FT..MON AND TUE...NW WIND 15 KT. SEAS 5 FT..WED...NE WIND 15 KT. SEAS 4 FT. Port Moller 2013, Page 60

66 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 23 Jun 24 Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 545 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 61

67 PMTF Interpretation #9 for catches through June 21, 2013 Stations 10 and 2 were skipped yesterday due to inclement weather. The crew is standing down today as well, but plans are to fish tomorrow. Yesterday s catches yielded the highest Replacement Index to date for this year and were evenly distributed across Stations 4-6 indicative of the less patchy distribution we expect as the run builds. Catches should pick up further in the coming days reaching the seasonal peak sometime next week. The first set stock composition estimates show Ugashik to be stronger than expected especially given the first sample was highly influenced by catches at Station 6. Historically, Ugashik tends to have a later representation at PM and be more concentrated at Stations 2 and 4 with lower representation at Station 6 and beyond. However, this estimate had wide confidence intervals, and the stock composition estimates from 18 and 19 June are more in line with historical trends. The Nushagak and Kvichak stocks began slow during the June, but increased in the June estimate (Figure 2). Nushagak s representation in the PM genetic samples should peak soon while the Kvichak increases if the pattern in recent years is any indication. The Egegik stock dominated both sets of stock composition estimates as it usually does given its earlier run timing and size. Water temperatures continue to be a couple of degrees above average. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun 11-Jun Jun Jun Jun Jun Jun Jun 18-Jun Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Port Moller 2013, Page 62

68 Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 63

69 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul 2012 Figure 1. The daily Replacement Index for 2011, 2012, and Port Moller 2013, Page 64

70 Daily Replacement Index Ugashik Egegik Daily Replacement Index Naknek Alagnak Kvichak Daily Replacement Index Nushagak Wood Igushik Figure 2. Port Moller Replacement Index apportioned by stock based on genetic stock composition estimates. Each stock-specific index fluctuates due to changes in (1) the daily index and (2) the stock composition estimates, which combine samples from multiple dates. Different colored shadings demarcate dates for which samples were combined to form each new stock composition estimate. Port Moller 2013, Page 65

71 From: Scott Raborn To: Scott Raborn Subject: PMTF Interpretation #9 for catches through June 21, 2013 Date: Saturday, June 22, :49:09 PM No fishing today due to bad weather, but sampling should resume tomorrow. Scott Port Moller 2013, Page 66

72 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun No Fishing boat weathered out 23 Jun Overcast; seas ft; wind NE Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals #### 1988 a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 25 Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 546 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 67

73 PMTF Interpretation #10 for catches through June 23, /24/2013 7:43 PM Alaska Time The Replacement Index fell off from 21 June (Table 1; Figure 1), and given the run seems early, will likely peak in the next day or two if not already. An early run timing is consistent with the warmer than average sea surface temperatures observed thus far (Figure 2). The relatively high cumulative Traditional Index for this year is partially explained by an early run timing, but catch plus escapement (C+E) to date suggest a larger than expected run size must be making up the difference (Figure 3). Since 1967 cumulative C+E has never been this high by 23 June. The inshore run would have to have unprecedented early timing and compression to create these numbers if the run is to come in at forecast or below. The latest genetics for samples collected on June have just been released. We will include our usual graph combining the Replacement Index with the stock apportionments in tomorrow s interpretation. For now, it looks like the Kvichak is picking up steam in its representation at PM. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date 10-Jun Daily b 7 Cumulative 7 11-Jun Jun 13-Jun Jun Jun 16-Jun Jun Jun 19-Jun Jun Jun 22-Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Page 1 of 5 Port Moller 2013, Page 68

74 6/24/2013 7:43 PM Alaska Time Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Page 2 of 5 Port Moller 2013, Page 69

75 6/24/2013 7:43 PM Alaska Time 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul 2012 Figure 1. The daily (top graph) and cumulative (bottom) Replacement Index for 2011, 2012, and Page 3 of 5 Port Moller 2013, Page 70

76 6/24/2013 7:43 PM Alaska Time 14.0 Sea surface temperatures ( ) Temperature (C) Figure 2. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for warmer years, 2011, and 2012, which are represented by various colors). Page 4 of 5 Port Moller 2013, Page 71

77 6/24/2013 7:43 PM Alaska Time No. of years equal to or greater than the value ( ) Cumulative C+E through 23 June Figure 3. Historical frequency of cumulative catch and escapement (C+E) through 23 June. Page 5 of 5 Port Moller 2013, Page 72

78 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul #### 1159 Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 26 Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 546 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm Weather Overcast; seas 1 2 ft; wind from NW 5 15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds No Fishing boat weathered out 25% cloud cover; seas ft; SE5 to NE5 Clear to 25% cover; seas 1.5 ft; winds ~ 5 Clear or fog; seas calm to 4 ft; light NW Fog; ft seas; SW winds Overcast; seas ft; light wind NE & SE Overcast; seas 3 ft; winds SE/SW 5 15, No Fishing boat weathered out Overcast; seas ft; wind NE 10 Overcast; seas 1.5 ft; winds NW <= 5 kts Port Moller 2013, Page 73

79 PMTF Interpretation #11 for catches through June 24, /25/2013 7:06 PM Alaska Time The Replacement Index held steady the last two days (Table 1; Figure 1), and preliminary results from the boat indicate it will not decline today (25 June). Based on stock composition estimates combined with the daily Replacement Index through 21 June, it appears that travel time between PM and the inshore districts is about 5-6 days. If signals from the genetics and index magnitude are telling from the PMTF this year, C+E for Egegik should hold steady over the next several days, the Nushagak and Kvichak districts should increase, while the Ugashik district should decline. Readers must be aware that these projections invoke uncertainty from every aspect of the PMTF, but we will have a better idea about their information content over the next few days. Likewise, with respect to run timing and magnitude based on PM, the next several days worth of test fishing index and C+E numbers should be very telling. For now, we report some preliminary findings from our net saturation experiment this year in Figure 3. These data were generated by setting the test fishing net for 20 min at the same station following the routine 60 min set. The boat steamed for 5 min west of the station being replicated before resetting. Initial results indicate that net saturation begins to occur beyond a Replacement Index of 50, but more data points are needed to better quantify this effect and at what catch levels it begins. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date 10-Jun Daily b 7 Cumulative 7 11-Jun Jun 13-Jun Jun Jun 16-Jun Jun Jun 19-Jun Jun Jun 22-Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Page 1 of 4 Port Moller 2013, Page 74

80 6/25/2013 7:06 PM Alaska Time Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul Figure 1. The daily (top graph) and cumulative (bottom) Replacement Index for 2011, 2012, and Page 2 of 4 Port Moller 2013, Page 75

81 Daily Replacement Index Egegik C+E (millions) 6/25/2013 7:06 PM Alaska Time Daily Replacement Index Ugashik C+E (millions) Daily Replacement Index Nushagak-Wood C+E (millions) Daily Replacement Index Naknek-Kvichak C+E (millions) Figure 2. District specific daily Replacement Index (line) apportioned based on genetic stock composition estimates and catch + escapement (columns). Each district-specific index fluctuates due to changes in (1) the daily index and (2) the stock composition estimates, which combine samples from multiple dates. Page 3 of 4 Port Moller 2013, Page 76

82 6/25/2013 7:06 PM Alaska Time CPUE from 60 min set CPUE from 20 min set Figure 3. Comparison of paired sets differing in fishing time from the 2013 PMTF. Each data point shows the Replacement Index or CPUE (catch adjusted to a 60 min set) from a set duration for about 60 min compared to a set immediately following at the same station for a duration of about 20 min. Sets include Stations 2-8. Saturation would cause higher values to be below the 1:1 line; no saturation would be indicated if every point was on or around the line; the reverse of saturation would be occurring if points were above the line. Page 4 of 4 Port Moller 2013, Page 77

83 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun No Fishing boat weathered out 23 Jun Overcast; seas ft; wind NE Jun Overcast; seas 1.5 ft; winds NW <= 5 kts 25 Jun Fog; ft seas; winds W & NW, Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 27 Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 547 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 78

84 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun No Fishing boat weathered out 23 Jun Overcast; seas ft; wind NE Jun Overcast; seas 1.5 ft; winds NW <= 5 kts 25 Jun Fog; ft seas; winds W & NW, Jun Fog; seas 6 ft; winds NW Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 28 Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 547 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 79

85 PMTF Interpretation #12 for catches through June 26, /26/2013 8:43 PM Alaska Time The boat was blown around a bit today and had to finish early; therefore we have included today s catches for this interpretation. They were able to get sets in at Stations 8 and 6 before heading in. Plans are to fish all stations tomorrow, weather permitting of course. Catches at Stations 8 and 6 were down today causing the daily Replacement Index to drop (Table 1). All indications are that PM peaked on 21 June (Figure 1) making this year s run the earliest ever at PM (7 days early). If PM fails to rally, then we can begin to project expectations for the inshore run (Figure 2) and the following paragraph is based on this assumption; stronger catches at PM over the next several days shifts the PM peak to a later date and voids the scenario described below. A travel time (TT) of 6 days between PM and inshore would make the peak day inshore tomorrow (27 June). This would mean the inshore run is 9 days early; an 11 day travel time would put the inshore peak around 2 July (4 days early). Historically, early peaks at PM result in longer travel times, and our best guess from the relationship presented in Figure 3 is about 10 days making the peak day inshore somewhere between the 2-4 July. Figure 4 illustrates how changes in TT dampen the inshore deviations from average run timing as compared to the more extreme deviations at PM. The most likely curve presented in Figure 2 should begin to emerge starting on 28 June and especially by 30 June. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date 10-Jun Daily b 7 Cumulative 7 11-Jun 12-Jun Jun Jun 15-Jun Jun 17-Jun Jun Jun 20-Jun Jun 22-Jun Jun Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Page 1 of 5 Port Moller 2013, Page 80

86 6/26/2013 8:43 PM Alaska Time 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul Figure 1. The daily (top graph) and cumulative (bottom) Replacement Index for 2011, 2012, and Page 2 of 5 Port Moller 2013, Page 81

87 6/26/2013 8:43 PM Alaska Time Replacement Index (mean catch per 200 fathom hours) C+E (millions) Figure 2. Inshore catch plus escapement (C+E) of Bristol Bay sockeye salmon for The Port Moller forecasts (red lines) are based on a curve fit to the daily Replacement Index, shifted forward by various travel times (TT), and raised by corresponding fish-per-index (FPI) values. Gray bars represent observed C+E. The black solid line and dashed line represent the observed PM daily Replacement Index and its predicted curve. Total run forecast for each curve is given underneath. Page 3 of 5 Port Moller 2013, Page 82

88 6/26/2013 8:43 PM Alaska Time Travel time (TT) between PM and inshore Peak day at Port Moller Figure 3. Historical travel times between PM and inshore as a function of the peak day of passage past the PMTF. These times were estimated with the model used in Figure 2 applied to each year. Page 4 of 5 Port Moller 2013, Page 83

89 6/26/2013 8:43 PM Alaska Time Late 6.0 Inshore run timing July Early June Early PM run timing Late Figure 4. Timing of the inshore run compared to that for Port Moller. The red line depicts identical run timing between inshore and PM. Values above this line mean the inshore run was later than PM; below the line indicates the inshore run was earlier than PM. Peak days are outlined above the zero values on each axis. Page 5 of 5 Port Moller 2013, Page 84

90 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun No Fishing boat weathered out 23 Jun Overcast; seas ft; wind NE Jun Overcast; seas 1.5 ft; winds NW <= 5 kts 25 Jun Fog; ft seas; winds W & NW, Jun Fog; seas 6 ft; winds NW Jun Overcast; seas ft; winds NW Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 547 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 85

91 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10 Jun Overcast; seas 1 2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12 Jun Overcast; seas 1 ft; wind from NE 3 kts 13 Jun % cloud cover; seas 1 ft; winds variable 14 Jun % cloud cover; seas 1 3 ft; winds 15 Jun No Fishing boat weathered out 16 Jun % cloud cover; seas ft; SE5 to NE5 17 Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18 Jun Clear or fog; seas calm to 4 ft; light NW 19 Jun Fog; ft seas; SW winds 20 Jun Overcast; seas ft; light wind NE & SE 21 Jun Overcast; seas 3 ft; winds SE/SW 5 15, 22 Jun No Fishing boat weathered out 23 Jun Overcast; seas ft; wind NE Jun Overcast; seas 1.5 ft; winds NW <= 5 kts 25 Jun Fog; ft seas; winds W & NW, Jun Fog; seas 6 ft; winds NW Jun Overcast; seas ft; winds NW Jun Partly cloudy; seas 1.5 ft; winds SE < 5 kts 29 Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul 4 Jul 5 Jul Weighted means 548 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm. Port Moller 2013, Page 86

92 PMTF Interpretation #13 for catches through June 28, /28/2013 7:46 PM Alaska Time The PM daily Replacement Index stayed low yesterday, but was back up today (Table 1; Figure 1) enough to shift the estimated day of peak passage at PM a little to the right (Figure 2; gray dashed line). Stronger numbers at PM in the next few days will continue to move the estimated peak at PM to a later date and cause the estimated peak inshore to shift later as well. With the current information, an 8-9 day travel time between PM and inshore puts the run at million peaking inshore tomorrow or the next day (30 June 6 days early). Genetic stock composition estimates should be out tonight and provide estimates for June and June. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Replacement CPUE a by station Replacement Index Date Daily b Cumulative 10-Jun Jun 12-Jun Jun 14-Jun Jun 16-Jun Jun 18-Jun Jun Jun 21-Jun Jun 23-Jun Jun 25-Jun Jun Jun Jun Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Page 1 of 3 Port Moller 2013, Page 87

93 6/28/2013 7:46 PM Alaska Time 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul Figure 1. The daily (top graph) and cumulative (bottom) Replacement Index for 2011, 2012, and Page 2 of 3 Port Moller 2013, Page 88

94 6/28/2013 7:46 PM Alaska Time Replacement Index (mean catch per 200 fathom hours) C+E (millions) Figure 2. Inshore catch plus escapement (C+E) of Bristol Bay sockeye salmon for The Port Moller forecasts (red lines) are based on a curve fit to the daily Replacement Index, shifted forward by various travel times (TT) between PM and inshore, and raised by corresponding fish-per-index (FPI) values. Gray bars represent observed C+E. The black solid line and dashed line represent the observed PM daily Replacement Index and its predicted curve. Total run and TT forecast for each curve is given underneath. Page 3 of 3 Port Moller 2013, Page 89

95 From: To: Subject: Date: Scott Raborn Scott Raborn PMTF Catch Update #21 - June 29, 2013 catches Saturday, June 29, :50:55 PM Hello Everyone, The crew was weathered out today, and conditions do not look good for tomorrow. This comes at a bad time when PM seemed to be picking up as of yesterday s catches. All for now, Scott Port Moller 2013, Page 90

96 From: To: Subject: Date: Scott Raborn Scott Raborn PMTF Catch Update #22 - June 30, 2013 catches Sunday, June 30, :43:39 PM Hi All, The crew was weathered out again today, and again conditions do not look good for tomorrow. All for now, Scott PKZ BRISTOL BAY WATERS CAPE NEWENHAM TO PORT HEIDEN 400 PM AKDT SUN JUN SMALL CRAFT ADVISORY THROUGH MONDAY NIGHT....TONIGHT...SW WIND 25 KT. GUSTS TO 35 KT OUT OF BAYS AND PASSES. SEAS 9 FT. PATCHY FOG. RAIN..MON...SW WIND 25 KT. SEAS 12 FT. PATCHY FOG. RAIN..MON NIGHT...SW WIND 25 KT. SEAS 12 FT. PATCHY FOG. RAIN..TUE...SW WIND 20 KT. SEAS 9 FT..TUE NIGHT...SW WIND 20 KT. SEAS 7 FT..WED...S WIND 15 KT. SEAS 6 FT..THU AND FRI...S WIND 15 KT. SEAS 4 FT. Port Moller 2013, Page 91

97 From: To: Subject: Date: Matt Nemeth Scott Raborn PMTF Catch Update #23 - July 1, 2013 catches (no fishing today) Monday, July 01, :53:17 PM All The crew was weathered out again today. Scott will follow up an interpretation tomorrow. Earlier today, you should have received the latest ADF&G age comp document. Matt Port Moller 2013, Page 92

98 From: To: Subject: Date: Matt Nemeth Scott Raborn PMTF Catch Update #24 - July 2, 2013 catches (no fishing today) Tuesday, July 02, :54:55 PM All The PMTF has been weathered out again today. Sounds rough out there - conditions have also apparently delayed today s planned fishing opener in the Outer Port Heiden section until tomorrow. -Matt From: Matt Nemeth Sent: Monday, July 01, :45 PM To: 'Scott Raborn' Subject: PMTF Catch Update #23 - July 1, 2013 catches (no fishing today) All The crew was weathered out again today. Scott will follow up an interpretation tomorrow. Earlier today, you should have received the latest ADF&G age comp document. Matt Port Moller 2013, Page 93

99 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index Weather 10-Jun Overcast; seas 1-2 ft; wind from NW Jun Overcast; seas 2 ft; wind from NW 5 kts 12-Jun Overcast; seas 1 ft; wind from NE 3 kts 13-Jun % cloud cover; seas 1 ft; winds variable 14-Jun % cloud cover; seas 1-3 ft; winds 15-Jun No Fishing - boat weathered out 16-Jun % cloud cover; seas ft; SE5 to NE5 17-Jun Clear to 25% cover; seas 1.5 ft; winds ~ 5 18-Jun Clear or fog; seas calm to 4 ft; light NW 19-Jun Fog; ft seas; SW winds 20-Jun Overcast; seas ft; light wind NE & SE 21-Jun Overcast; seas 3 ft; winds SE/SW 5-15, 22-Jun No Fishing - boat weathered out 23-Jun Overcast; seas ft; wind NE Jun Overcast; seas 1.5 ft; winds NW <= 5 kts 25-Jun Fog; ft seas; winds W & NW, Jun Fog; seas 6 ft; winds NW Jun Overcast; seas ft; winds NW Jun Partly cloudy; seas 1.5 ft; winds SE < 5 kts 29-Jun 30-Jun 1-Jul 2-Jul 3-Jul Overcast; seas 3-4 ft; winds S kts 4-Jul 5-Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30-Jun 1-Jul 2-Jul 3-Jul Jul 5-Jul Weighted means 548 c Length is measured as mid-eye-fork-length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age-2 fish=504 mm; 3-ocean=571 mm. Port Moller 2013, Page 94

100 PMTF Interpretation #14 for catches through July 3, /4/2013 4:58 PM Alaska Time Yesterday s daily Replacement Index for 3 July was down from the upswing we saw on 28 June (Table 1; Figure 1), and the cumulative Traditional Index is now below average (Table 2). No fishing was conducted during the four days in between, and these index values had to be interpolated. Not fishing for this many consecutive days increases the potential for missing an important event at PM; however, in all likelihood PM had already peaked on 21 June. District specific forecasts are presented in Figure 2. With PM and the inshore districts peaking so early we are more certain about the travel times (TTs) between PM and each district (range=6-9 days). It appears the Ugashik and Egegik districts are close to being over, followed by the Nushagak district with the Kvichak district just now finishing its peak. Combining the four districts brings the total run forecast to about 22.2 million, but again this is based on the interpolated index values from June 29 to July 2 being reasonably accurate. Water temperatures for this year began warmer than average, but then returned to average and below (Figure 3). Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Date Daily b Cumulative 10-Jun 11-Jun Jun 13-Jun Jun Jun Jun Jun 18-Jun Jun 20-Jun Jun Jun Jun Jun 25-Jun Jun 27-Jun Jun Jun Jun Jul 2-Jul Jul Jul 5-Jul Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Replacement CPUE a by station Replacement Index Page 1 of 4 Port Moller 2013, Page 95

101 7/4/2013 4:58 PM Alaska Time Table 2. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Page 2 of 4 Port Moller 2013, Page 96

102 7/4/2013 4:58 PM Alaska Time Replacement Index (mean catch per 200 fathom hours) Naknek-Kvichak 7,754,961 TT= Egegik 6,233,506 TT= C+E (millions) Replacement Index (mean catch per 200 fathom hours) Ugashik 2,00,975 TT= Nushagak 6,251,180 TT= C+E (millions) Replacement Index corrected for saturation (mean catch per 200 fathom hours) All districts combined 22,240, C+E (millions) Figure 2. Inshore catch plus escapement (C+E) of Bristol Bay sockeye salmon in the primary fishing districts for The Port Moller forecasts (red lines) are based on a curve fit to the daily Replacement Index, shifted forward by an estimated travel time (TT) between PM and inshore, and raised by an estimated fish-per-index (FPI) value. Gray bars represent observed C+E. The black solid line and dashed line represent the observed PM daily Replacement Index and its predicted curve. Estimated total run and TT for each curve are outlined in red. The daily Replacement Index was corrected for gillnet saturation and smoothed with a three day moving average. Based on historical stock composition estimates from the genetics study (Dann et al. 2013; see PMTF Interpretation #4 on 17 June), station specific Replacement Indexes were weighted differently when forming the index time series for each district. Page 3 of 4 Port Moller 2013, Page 97

103 7/4/2013 4:58 PM Alaska Time 14.0 Sea surface temperatures ( ) Temperature (C) Figure 3. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for warmer years, 2011, and 2012, which are represented by various colors). Page 4 of 4 Port Moller 2013, Page 98

104 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Totals Weather Overcast; seas 1 2 ft; wind from NW 5 15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds No Fishing boat weathered out 25% cloud cover; seas ft; SE5 to NE5 Clear to 25% cover; seas 1.5 ft; winds ~ 5 Clear or fog; seas calm to 4 ft; light NW Fog; ft seas; SW winds Overcast; seas ft; light wind NE & SE Overcast; seas 3 ft; winds SE/SW 5 15, No Fishing boat weathered out Overcast; seas ft; wind NE 10 Overcast; seas 1.5 ft; winds NW <= 5 kts Fog; ft seas; winds W & NW, 7 15 Fog; seas 6 ft; winds NW Overcast; seas ft; winds NW 7 13 Partly cloudy; seas 1.5 ft; winds SE < 5 kts Overcast; seas 3 4 ft; winds S kts Overcast; seas 1.5 ft; winds SW ~ 5 kts a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Weighted means 548 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm Port Moller 2013, Page 99

105 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Totals a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Weighted means 548 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm Weather Overcast; seas 1 2 ft; wind from NW 5 15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds No Fishing boat weathered out 25% cloud cover; seas ft; SE5 to NE5 Clear to 25% cover; seas 1.5 ft; winds ~ 5 Clear or fog; seas calm to 4 ft; light NW Fog; ft seas; SW winds Overcast; seas ft; light wind NE & SE Overcast; seas 3 ft; winds SE/SW 5 15, No Fishing boat weathered out Overcast; seas ft; wind NE 10 Overcast; seas 1.5 ft; winds NW <= 5 kts Fog; ft seas; winds W & NW, 7 15 Fog; seas 6 ft; winds NW Overcast; seas ft; winds NW 7 13 Partly cloudy; seas 1.5 ft; winds SE < 5 kts Overcast; seas 3 4 ft; winds S kts Overcast; seas 1.5 ft; winds SW ~ 5 kts Overcast; seas 1.5 ft; winds S ~ 5 kts Port Moller 2013, Page 100

106 Raw catch a of sockeye by station and mesh size Daily Daily Totals Rep Trad Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Jul Totals Overcast; seas 3 4 ft; winds S kts Overcast; seas 1.5 ft; winds SW ~ 5 kts Overcast; seas 1.5 ft; winds S ~ 5 kts Overcast; seas 3 5 ft; wind S/SW a Raw catch = number of fish caught in 100 fathoms of each mesh size. Adding two meshes together yields actual catch at each station from the entire Replacement Net. Average sockeye length c by station and mesh size Weighted Cum Cum Water temperature ( C) by means Rep Trad station Station Date 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ 4½ 5⅛ Index Index mean 10 Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun 30 Jun 1 Jul 2 Jul 3 Jul Jul Jul Jul Weighted means 548 c Length is measured as mid eye fork length (MEFL) in mm; to put things into perspective, mean MEFL for ocean age 2 fish=504 mm; 3 ocean=571 mm Weather Overcast; seas 1 2 ft; wind from NW 5 15 Overcast; seas 2 ft; wind from NW 5 kts Overcast; seas 1 ft; wind from NE 3 kts 50% cloud cover; seas 1 ft; winds variable % cloud cover; seas 1 3 ft; winds No Fishing boat weathered out 25% cloud cover; seas ft; SE5 to NE5 Clear to 25% cover; seas 1.5 ft; winds ~ 5 Clear or fog; seas calm to 4 ft; light NW Fog; ft seas; SW winds Overcast; seas ft; light wind NE & SE Overcast; seas 3 ft; winds SE/SW 5 15, No Fishing boat weathered out Overcast; seas ft; wind NE 10 Overcast; seas 1.5 ft; winds NW <= 5 kts Fog; ft seas; winds W & NW, 7 15 Fog; seas 6 ft; winds NW Overcast; seas ft; winds NW 7 13 Partly cloudy; seas 1.5 ft; winds SE < 5 kts Port Moller 2013, Page 101

107 PMTF Interpretation #15 for catches through July 6, /7/2013 3:36 PM Alaska Time The 2013 PMTF ended yesterday with the daily Replacement Index further declining. PM peaked this year on 21 June making it about 7 days earlier than the average run timing. Catch and escapement has been less than expected the last several days dropping our total run forecast from 22.2 million (4 July estimate) to 19.0 million. Although there will be no more test fishing this year, we will send a final interpretation after a few more days of C+E. Table 1. Estimated daily and cumulative Replacement Index for the 2013 Port Moller test fishery. Date Daily b Cumulative 10-Jun Jun Jun Jun 14-Jun Jun Jun Jun Jun 19-Jun Jun Jun Jun Jun Jun 25-Jun Jun Jun Jun Jun Jun 1-Jul Jul Jul Jul Jul 6-Jul Totals Percent a CPUE = the number of fish caught in all meshes and standardized to a 200 fathom net fished for 60 minutes. b Replacement daily index is simply the average CPUE across stations. Note: red values were interpolated from other observed catches. Replacement CPUE a by station Replacement Index Page 1 of 3 Port Moller 2013, Page 102

108 7/7/2013 3:36 PM Alaska Time 100 CPUE (catch per 200 fathom hours) Jun 12-Jun 14-Jun 16-Jun 18-Jun 20-Jun 22-Jun 24-Jun 26-Jun 28-Jun 30-Jun 2-Jul 4-Jul 6-Jul 8-Jul 10-Jul 12-Jul Figure 1. The daily Replacement Index for 2011, 2012, and Page 2 of 3 Port Moller 2013, Page 103

109 7/7/2013 3:36 PM Alaska Time Replacement Index (mean catch per 200 fathom hours) Naknek-Kvichak 6,811,907 TT= Egegik 5,168,321 TT= C+E (millions) Replacement Index (mean catch per 200 fathom hours) Ugashik 2,011,086 TT= Nushagak 2,011,086 TT= C+E (millions) Replacement Index corrected for saturation (mean catch per 200 fathom hours) All districts combined 19,031, C+E (millions) Figure 2. Inshore catch plus escapement (C+E) of Bristol Bay sockeye salmon in the primary fishing districts for The Port Moller forecasts (red lines) are based on a curve fit to the daily Replacement Index, shifted forward by an estimated travel time (TT) between PM and inshore, and raised by an estimated fish-per-index (FPI) value. Gray bars represent observed C+E. The black solid line and dashed line represent the observed PM daily Replacement Index and its predicted curve. Estimated total run and TT for each curve are outlined in red. The daily Replacement Index was corrected for gillnet saturation and smoothed with a three day moving average. Based on historical stock composition estimates from the genetics study (Dann et al. 2013; see PMTF Interpretation #4 on 17 June), station specific Replacement Indexes were weighted differently when forming the index time series for each district. Page 3 of 3 Port Moller 2013, Page 104

110 PMTF Interpretation #16 Final Interpretation The 2013 Bristol Bay sockeye run proved to be aberrant with respect to run timing both at the PMTF and inshore. Port Moller catches peaked on 21 June, nine days earlier than the average of 29 June. This early arrival caused the cumulative Traditional Index to be the highest on record up through 22 June before falling below average by 1 July (Table 1). Likewise, total C+E peaked six-eight days early around 30 June ( average=6-7 July), making this year the earliest run on record (Figure 1). Traditionally, the test fishing net was comprised of four 50-fathom panels of 5⅛ inch mesh. In 2011, we changed the Traditional Net by replacing every other 5⅛ inch panel with 4½ inch mesh to comprise the Replacement Net in order to reduce size selectivity and render the age composition estimates at Port Moller more representative of inshore. For 2011 and 2012, the Replacement Net appears to have been successful in removing the majority of this selectivity as Port Moller catches indicated departures from the pre-season forecast of ocean age composition by June (Figure 2). However, this year s age composition estimates from Port Moller over-represented age 1.2s and 1.3s, while under-representing age 2.3s. Size selectivity creates the greatest catch discrepancy across the two ocean ages as they differ more in size than do the two freshwater ages. The direction of this year s age composition bias was not consistent within ocean ages (over for 1.3s and under for 2.3s) suggesting size selectivity was not the problem. More likely, stocks varied in their vulnerability to the test fishery, which combined with their differences in age compositions caused vulnerability across ages to vary as well. Given this explanation is correct, we can assume that ages 1.2 and 1.3 were over-represented at Port Moller, while age 2.3 was under-represented. The final genetic stock composition estimates cumulative through the season were reasonably accurate (Figure 3). Nominal discrepancies were the slight over-representation of the Ugashik stock (over by 3%) and under-representation of the Nushagak-Wood District (under by 5%). With respect to tracking the inshore run, the 2013 PMTF worked reasonably well with the exception of missing catches due to bad weather (Figure 4). Sliding the Replacement Index forward by a travel time of nine days illustrates why PM failed to detect the last movement of fish that peaked on 15 July. Clearly the interpolated values for 29 June-2 July were incorrect. Missing this many consecutive days at PM causes interpolations to be more tenuous and increases the chances that a major movement of fish passed undetected. Another anomaly this year was the pattern in sea surface temperatures throughout the season (Figure 5). Temperatures began about average, rose quickly, and then ended below average by season s end. How temperatures affect travel time remains poorly understood, but these findings are consistent with the pattern of C+E inshore. Initial higher temperatures coincided with shorter travel times and earlier arrival inshore, followed by cooler temperatures lining up with longer travel times and later arrival. Acknowledgements The R/V Pandalus was crewed by skipper Ted Jewel, engineer David Knight, and deck hand Margaret Archibald. Our fisheries technicians were Drew Stinnett and Amos Cernohouz. Guy Wade managed the Port Moller test fishery program, Matt Nemeth compiled and distributed the catch data, and Scott Raborn prepared the interpretations with assistance from Michael Link and Matt Nemeth. Fred West and Chuck Brazil (ADF&G) managed the scale aging operation in King Salmon and provided timely summaries of age composition from the Port Moller catch. Tyler Dann managed the genetic stock identification component of the project. We would like to especially thank Jeanne Boyle and Bob Murphy at the ADF&G office in Port Moller. Peter Pan Seafoods, specifically Mark Briski, George Sudar, and the dock crew were invaluable with respect to logistics. Without their help, the Port Moller test fishing program would be impossible. Port Moller 2013, Page 105

111 The 2013 Port Moller test fishing project was managed and staffed by the Bristol Bay Science and Research Institute (BBSRI). The project was funded by ADF&G, BBSRI, eight Bristol Bay processors, and by Bristol Bay driftnet fishermen (through the Bristol Bay Regional Seafood Development Association, or BB-RSDA). Processors included Canfisco (AGS/Leader Creek), E&E Foods, Icicle Seafoods, North Pacific Seafoods, Ocean Beauty, Peter Pan Seafoods, and Trident Seafoods. Although none of this is possible without the help of ADF&G personnel and funding, the interpretations contained in the updates were by no means official ADF&G interpretations of the Port Moller test fishery data or the Bristol Bay run. Thanks to all of you for your feedback throughout the season and please feel free to make suggestions for next year. Table 1. Historical cumulative Traditional Index by date from the Port Moller test fishery, Date Avg Min Max 11-Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jun Jul Jul Jul Jul Jul Total run (millions) Port Moller 2013, Page 106

112 8 Number of years ( ) Days early Days late Figure 1. Historical variation in peak day of return for Bristol Bay sockeye salmon compared to Port Moller 2013, Page 107

113 % of total age composition 90% 75% 60% 45% 30% 15% Pre-season Cumulative Port Moller through June Cumulative Port Moller through 6 July Total inshore run as of 14 July % 45% 30% 15% % Ocean Age 2 Ocean Age 3 0% Age 1.2 Age 2.2 Age 1.3 Age % 45% % of total age composition 60% 45% 30% 15% % 15% % Ocean Age 2 Ocean Age 3 0% Age 1.2 Age 2.2 Age 1.3 Age % % 2011 % of total age composition 60% 45% 30% 15% 30% 15% 0% Ocean Age 2 Ocean Age 3 0% Age 1.2 Age 2.2 Age 1.3 Age 2.3 Figure 2. Pre-season and Port Moller age composition estimates of Bristol Bay sockeye salmon compared to the inshore total run since we began using the Replacement Net. Port Moller 2013, Page 108

114 % of the total run 40% 35% 30% 25% 20% 15% Jun10-17 Jun18-19 Jun20-21 Jun23-24 Jun25-26 Jun27-28 Jul3-6 C+E 10% 5% 0% Ugashik Egegik Kvichak-Naknek Nushagak-Wood Togiak Figure 3. Genetic stock composition estimates from PMTF catches of Bristol Bay sockeye salmon compared to the inshore total run for Each date s estimates are based on catches cumulative across date. Port Moller 2013, Page 109

115 Replacement Index (mean catch per 200 fathom hours) day travel time C+E (millions) Figure 4. Inshore catch plus escapement (C+E; gray columns) of Bristol Bay sockeye salmon for Lines represent the PM daily Replacement Index. Missing index points for when no test fishing occurred had to be interpolated and are indicated with white markers. Slid forward by a travel time of nine days, the shaded area indicates the inshore dates most likely to have been misrepresented at Port Moller due to inaccurate interpolations for 29-June-2 July Sea surface temperatures ( ) Average Temperature (C) Figure 5. Mean sea surface temperatures averaged across the fishing transect (usually Stations 2-10) for the 2013 Port Moller test fishery. Temperatures from previous years are represented as thin gray lines (except for warmer years, 2011, and 2012, which are represented by various colors). Port Moller 2013, Page 110

116 Appendix B Inseason Port Moller Stock Composition Estimates from ADF&G s Gene Conservation Lab, 2013 Port Moller 2013, Page 111

117 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 10 17, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 10 17, A total of 877 fish were sampled and 190 were analyzed (181 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 0.8% 25.7% 31.0% 13.3% 2.2% 9.2% 7.2% 7.8% 1.9% 0.2% 0.8% 0.0% 8.8% 15.9% 6.8% 0.0% 4.3% 0.0% 0.0% 0.0% 0.0% 0.0% 4.2% 41.2% 47.9% 20.8% 8.0% 15.3% 14.9% 15.6% 8.5% 0.7% 5.2% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 10 17, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 112

118 Historical Comparison of Stock Composition Estimates / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 113

119 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 114

120 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 18 19, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 18 19, A total of 327 fish were sampled and 190 were analyzed (185 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 2.5% 13.9% 30.4% 7.9% 0.1% 20.5% 19.1% 4.2% 1.1% 0.1% 0.1% 0.0% 4.9% 21.9% 3.3% 0.0% 12.1% 8.5% 0.0% 0.0% 0.0% 0.0% 7.7% 24.1% 39.8% 13.9% 0.3% 29.7% 28.2% 11.9% 6.0% 0.1% 0.0% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 18 19, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 115

121 Historical Comparison of Stock Composition Estimates / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 116

122 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 117

123 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 20 21, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 20 21, A total of 483 fish were sampled and 190 were analyzed (183 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 9.0% 9.9% 22.6% 7.3% 0.9% 26.8% 8.8% 13.6% 0.1% 1.0% 0.1% 3.3% 0.9% 13.9% 2.4% 0.0% 18.6% 2.5% 7.9% 0.0% 0.0% 0.0% 15.5% 18.1% 32.6% 13.2% 4.6% 35.4% 15.6% 20.0% 0.1% 6.4% 0.5% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 20 21, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 118

124 Historical Comparison of Stock Composition Estimates / / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 119

125 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 120

126 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 23 24, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 23 24, A total of 407 fish were sampled and 190 were analyzed (189 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 0.3% 8.0% 22.7% 27.9% 1.9% 22.2% 6.6% 9.1% 1.3% 0.0% 0.0% 0.0% 0.0% 13.3% 18.7% 0.0% 13.9% 0.2% 2.3% 0.0% 0.0% 0.0% 2.0% 19.0% 34.3% 37.4% 9.6% 31.4% 13.7% 15.3% 7.6% 0.0% 0.0% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 23 24, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 121

127 Historical Comparison of Stock Composition Estimates / / / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 122

128 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 123

129 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 25 26, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 25 26, A total of 295 fish were sampled and 190 were analyzed (186 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 0.8% 11.5% 22.8% 11.8% 1.3% 27.8% 11.2% 10.8% 0.6% 1.3% 0.3% 0.0% 0.8% 14.5% 5.2% 0.0% 18.3% 0.8% 5.4% 0.0% 0.0% 0.0% 6.5% 22.1% 33.8% 19.1% 10.4% 37.5% 20.6% 16.6% 4.8% 5.9% 2.7% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 25 26, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 124

130 Historical Comparison of Stock Composition Estimates / / / / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 125

131 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 126

132 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 27 28, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for June 27 28, A total of 368 fish were sampled and 190 were analyzed (187 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 0.0% 7.4% 29.6% 14.0% 1.2% 28.6% 9.0% 10.1% 0.0% 0.0% 0.1% 0.0% 0.0% 16.7% 8.2% 0.0% 19.8% 3.7% 5.9% 0.0% 0.0% 0.0% 0.0% 20.0% 41.7% 20.5% 6.9% 37.8% 14.4% 15.1% 0.0% 0.0% 0.0% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, June 27 28, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 127

133 Historical Comparison of Stock Composition Estimates / / / / / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 128

134 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 129

135 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary July 3 6, 2013 All Stations Genetic stock composition estimates for sockeye salmon from the Port Moller Test Fishery for July 3 6, A total of 262 fish were sampled and 190 were analyzed (189 had adequate data to include in the analysis). Stock 90% Composition Confidence Intervals Reporting Group Estimate Lower Upper North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim 3.3% 13.8% 21.1% 18.2% 6.5% 17.0% 1.3% 13.7% 2.8% 1.9% 0.3% 0.0% 0.4% 9.1% 9.8% 2.1% 10.4% 0.0% 6.2% 0.0% 0.0% 0.0% 8.9% 26.5% 34.7% 27.0% 11.4% 24.9% 4.2% 20.8% 10.3% 4.2% 2.7% Stock Composition Estimate (%) Genetic Stock Composition Estimates for Sockeye Salmon Captured in the Port Moller Test Fishery, July 3 6, 2013 (All Stations). North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Reporting Group Wood Igushik Togiak Kuskokwim The genetic analysis was completed by the Alaska Department of Fish and Game, Division of Commercial Fisheries, Gene Conservation Laboratory. Port Moller 2013, Page 130

136 Historical Comparison of Stock Composition Estimates / / / / / / / Proportion of Catch / / /21 6/ / / /30 7/1 7/2 3 7/5 8 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / / /30 7/1 7/2 4 7/5 7 Date Port Moller 2013, Page 131

137 Historical Comparison of Stock Composition Estimates / / / / /26 6/ /30 7/1 7/2 3 7/4 5 7/ Proportion of Catch / / / / / / / /30 7/1 7/4 5 North Peninsula Ugashik Egegik Naknek Alagnak Kvichak Nushagak Wood Igushik Togiak Kuskokwim / / / / /24 6/ / /30 7/1 7/3 4 7/5 6 Date Port Moller 2013, Page 132

138 Bristol Bay Sockeye Salmon Fishery Port Moller Sockeye Salmon Stock Composition Summary June 10 July 6, 2013 Stations 2/4, 6, and 8/10 This report summarizes genetic stock composition estimates for sockeye salmon captured at different stations of the Port Moller Test Fishery in We analyzed the fish by station groups to characterize the distribution of stocks across the test fishery transect. In defining station specific groups, we balanced the goal of fine scale temporal resolution of station catches with the requirement of adequate sample sizes. As a result, we analyzed station specific catches for three time periods. Catches and corresponding samples at Stations 2 and 10 were few, so these fish were grouped with fish from adjacent stations (2 grouped with 4, 10 grouped with 8). Similarly, few fish were caught at Stations 2 and 4 so we provide a single, annual set of estimates for the inshore stations. The figure below summarizes the mean stock composition estimates for all groups while following pages provide details for each station group. The figure above depicts mean estimates for the 9 major stocks within Bristol Bay for each spatiotemporal stratum of the Port Moller Test Fishery in Time periods are along the horizontal axis while stations are along the vertical axis. The darker the red the higher the estimate, with completely red equal to 40% and white equal to 0%. See following pages for details. Port Moller 2013, Page 133