New or modified marine protected areas

Several new or modified areas were recommended to WDFW as part of the recreational regulation cycle. Five were circulated publicly and resulted in protection increasing or changing.

A new marine preserve was created at Colvos Passage, north of Gig Harbor. This area will be Closed for harvest of all species except salmon harvested by trolling. The site is 500 feet long and encompasses an underwater rock ledge and rocky reef area. WDFW has had a wolf eel study underway at this location for the past year.

An area in Hood Canal was designated as a marine preserve. The area known as Waketickeh Creek was created as a marine preserve, however, the intention of the WDFW Commission was to make this area a Conservation Area (i.e. closed to all harvest under WDFW jurisdiction) following outreach to the commercial fishers which might use the site. The site is 1700 yards long and extends 500 yards offshore. Outreach is ongoing to the commercial fishers regarding this area.

Orchard Rocks Conservation Area was enlarged to be a complete circle. Previously there was a pie-shaped wedge left out of the circular design. The area leased by the private net pen operation adjacent to the rocky habitat was not included. The recommendation to make the site round came from the dive community based on making the site boundaries easier to understand. The modification of the boundary was not adopted to change the operation of the net pen or the use of the privately leased tidelands.

Saltar’s Point Beach was adopted as a Marine Preserve. This area encompasses the tidelands owned by the City of Steilacoom. The site represents a partnership between the City and WDFW. This area is also expected to change to a Conservation Area when outreach to the commercial community is complete.

The Marine Preserve at Sund Rock was also modified as a "house-keeping" measure. The net pens, which were used in the boundary description for the site, were removed after they were damaged, so the site boundaries were changed to relate to local landmarks.

Trans-border WA-BC marine protected area activities

Islands Trust and San Juan County have had one public forum to discuss issues they have in common. In addition, they have had several planning meetings to prepare for a second forum which will take place in June on Saturna Island. At the June forum, they want to select an area of interest for study as a possible marine protected area.

A group of trans-border non-governmental organizations have undertaken a similar effort. People for Puget Sound has been mapping the resources in the transboundary area using information from other sources including the Department of Fish and Wildlife. They have an area selected for consideration based on the overlays of the information. The depth of color in the map produced is based on the number of resources known to be present. They are continuing to refine this with additional data and input from public and private sectors. The area of interest can be viewed on line at: http://www.pugetsound.org/mpa/

All seven northern counties have now established marine resource committees (MRCs) to address a variety of problems. The establishment of an MRC also qualifies the county to participate as a member of the NW Straits Commission. Part of the charge to member counties is to develop marine protected areas within their county. The NW Straits Commission includes Whatcom, Skagit, Snohomish, Island, San Juan, Clallum and Jefferson counties.

Return to Table of Contents

3. Puget Sound Ambient Monitoring Program (PSAMP) (Contributed by Sandie O'Neill and Jim West (360) 902-2843)

PSAMP is a multi-agency effort to monitor the health of Puget Sound. The Washington Department of Fish and Wildlife participates by monitoring toxic contaminants in Puget Sound fishes. To date, we have measured contaminant levels in English sole, Pacific cod, three species of demersal rockfish, Pacific herring, and chinook and coho salmon from a wide range of environments (polluted to clean). Geographic patterns of contaminants in these species was described. We have now completed focus study in Elliott Bay, Sinclair Inlet and Commencement Bay, three of Puget Sound’s most highly contaminated areas. With these studies we have documented fine-scale distribution patterns of contaminants in English sole and rockfish, as well as the effects of contaminants on reproductive competence in English sole.

Papers on English sole, salmon and rockfish were presented at the 1998 Research in Puget Sound Conference (March 1998) in Seattle, Washington. Copies are available on-line at www.wa.gov/Puget_Sound or from PSAMP researchers. Recently we have shown that English sole from urban and near urban areas were 2 to 33 times more likely to develop liver disease than fish from clean reference areas. Liver disease in English sole is strongly correlated with the presence of polycyclic aromatic hydrocarbons (PAHs) in sediments, which originate from petroleum or from the combustion of fossil fuels. In addition, we have found from our six long-term Puget Sound sampling sites that the chances of English sole developing toxicant-related liver disease have increased roughly two-fold from 1989 to 1998 in English sole from Elliott Bay (relative to clean reference sites). The chances of English sole developing liver disease have remained unchanged during this period in the five other areas (Commencement Bay, Sinclair Inlet, Everett, Hood Canal, and the Strait of Georgia).

We have completed two samplings of contaminants in Pacific herring from three to five locations in Puget Sound and Canada (1999 and 2000) and expect results from these analyses in spring, 2000. We have also continued archiving tissues from Puget Sound groundfish and forage fish for future molecular genetics studies. To date we have sampled tissues from eleven species, totaling over 2500 specimens.

4. Forage Fish

Fisheries for Coastal Pelagic Species (Contributed by Greg Bargmann (360) 902-2835)

Interest continues to grow in allowing a sardine fishery off of the Washington coast. Public meetings indicate considerable interest in a May to October fishery for sardine; there is lesser interest in fisheries for anchovy, mackerel and squid.

The department is planning a trial commercial fishery for these pelagic species. The fishery will be open May 15 to September 30. A catch quota of 4,000 metric tons of sardine will be in effect. No quota exists for the other species. Observers will be required to be present on 50% of all fishing trips. No fishing will be allowed within 3 miles of the coast to minimize interference with recreational fisheries and to reduce bycatch of groundfish.

Herring Stock Assessment (Contributed by Mark O'Toole (360) 466-4345 ext 241)

Spawning biomass estimates were made at the eighteen known herring spawning grounds in Puget Sound, along with two coastal spawning grounds, Grays Harbor and Willapa Bay. Two different direct methods of estimating abundance were utilized, spawn deposition surveys and acoustic-trawl surveys. Most spawning grounds were surveyed with either one method or the other, a few with both methods. The spawn deposition surveys generate estimates of spawning biomass only, the acoustic-trawl surveys provide estimates of spawning biomass along with age composition, size composition, etc. In general, for 1999/2000, spawning biomass estimates for the south/central Puget Sound area and the two coastal grounds are increasing (Hood Canal area estimates in particular have been up), and estimates for the north Puget Sound area and the Strait of Juan de Fuca are down.

A "new" herring spawning ground was documented this year at Wollochet Bay near Tacoma. This area was mentioned as a spawning ground in old Department reports from the 1930’s and 1940’s, but no spawning activity had been detected there in occasional spot checks over the past 25 years. In going over the old reports last fall for ESA analysis, staff noticed that references were repeatedly made to the spawning activity at Wollochet being very "early", so a spawn deposition survey was conducted in January and found that spawning was occurring. Spawning activity was documented from early January to the beginning of March, so this ground would have the "earliest" spawning season in Puget Sound.

In Grays Harbor the known herring spawning ground is in the Elk River estuary area, near Bay City. In 2000, spawn deposition surveys recorded that the spawning herring expanded their use of this ground to an area nearly double the size over the previously documented ground. In addition, herring spawning activity was discovered in the Ocean Shores area, near Pt. Damon. This has not been seen before, and it is an entirely new area.

Council Activities

The Department contributes technical support for coastal groundfish management issues via participation on the Groundfish Management Team (GMT), the Scientific and Statistical Committee (SSC), and the Habitat Steering Group (HSG) of the Pacific Fishery Management Council (PFMC). The Department is also represented on the Scientific and Statistical Committee and Groundfish Plan Team of the North Pacific Fishery Management Council (NPFMC). Landings and fishery management descriptions for PFMC-managed groundfish are summarized annually by the GMT in the Stock Assessment and Fishery Evaluation (SAFE) document.

Stock Rebuilding

The Pacific Council amended their groundfish fishery management plan this year (Amendment 12) to: 1) clarify the process for preparing and approving rebuilding programs; 2) establish rebuilding goals and objectives; 3) authorize temporary adjustments to the open access allocation for any overfished stock without a plan amendment or regulatory amendment; 4) authorize the Council and the National Marine Fisheries Service to prohibit vessels with limited entry permits from fishing in the open access fishery with the limited entry fishery is closed (that is, limited entry vessels would be prohibited from landing a groundfish species when the limited entry fishery for that species is closed); 5) revise procedures for preparing and distributing the SAFE document; and 6) declare the groundfish resource to be fully utilized by U.S. fishers and processors, eliminating foreign and joint venture fishing unless the plan is amended to reinstate such opportunities. In addition, the Council approved a draft process and schedule for preparing rebuilding plans for canary rockfish and cowcod.

Strategic Plan

The Pacific Council is in the process of developing a Strategic Plan for West Coast Groundfish which is expected to be completed by September 2000. The strategies and proposed actions in the Strategic Plan will cover a variety of management issues, including overcapacity, allocation, harvest rates, overfishing, habitat/ecosystem concerns, research and data needs, and the Council’s structure and process. The Council’s SSC Economic Subcommittee drafted a report on overcapitalization in the West Coast groundfish fishery which recommends the Council consider limited entry, buyouts, and permit stacking in the near-term given the current moratorium on Individual Fishing Quotas (IFQs); however, IFQs are best viewed as a long-term strategy for West Coast groundfish management. A draft Strategic Plan is expected to be completed for the June Council meeting. Copies of the draft plan and the SSC’s report can be obtained from the Pacific Council at (503) 326-6352.

Observer Program

The National Marine Fisheries Service drafted a package of generic regulations to set in place the process for deploying at-sea observers which the Council adopted for public review and publication in the Federal Register. This rule would establish notification requirements for vessels that may be required to carry observers and establish responsibilities and prohibited actions for vessels that are required to carry observers. When funding for a West Coast observer program becomes available, these regulations would be in place to quickly implement an at-sea observer program for groundfish fisheries that land catch shoreside. The draft regulatory package does not pertain to at-sea observers for offshore fisheries (e.g., the at-sea Pacific whiting fishery).

Harvest Rate Policy

A Groundfish Harvest Rate Policy Workshop Panel was convened to review the groundfish harvest rate and a draft report was presented to the Pacific Council in April. The draft report:

1) summarized the scientific and management background of the harvest proxy issue; 2) concisely explained some areas of common confusion; and 3) recommended default groundfish harvest rates for Pacific whiting (F_40%), Sebastes and Sebastolobus (F_50%), flatfish (F_40%), and other groundfish (F_45%). The report notes these recommendations were not developed as precautionary changes, but rather they attempt to correct previous estimates of productivity. Final action on the harvest rate policy is scheduled for the June 2000 Council meeting.

Coastal Groundfish Habitat (Contributed by Michele Robinson (360) 902-2220)

Council Activities

The Pacific Council is evaluating marine reserves as a potential fishery management tool and created an ad-hoc Marine Reserves Committee last year. The Council adopted a two-phase process in April 1999–the first phase is a conceptual evaluation of the utility of marine reserves; the second phase, if pursued, would involve the siting of specific reserve areas. The ad-hoc Committee developed management objectives for the reserves and a conceptual framework to guide the technical analysis. The management objectives (in priority order) are: 1) accelerate stock rebuilding; 2) enhance biological productivity; 3) enhance economic productivity; 4) provide insurance against management errors and/or environmental impacts; 5) conserve and protect habitat; and 6) improve opportunities for research and education. National Marine Fisheries Service scientists have analyzed the Committee’s framework and produced a draft report which will be presented to the Council in June 2000.

Department Activities

The Status of Washington’s Coastal Marine Protected Areas by Michele Robinson was published by the Washington Department of Fish and Wildlife (January 1999) and is a compilation of the existing 33 marine protected area (MPA) sites along Washington’s coast. The report identifies and classifies the coastal MPAs; describes the locations and purposes of the sites; and outlines management objectives for the establishment of future MPAs. Copies of the report can be obtained by contacting Michele Robinson at (360) 902-2220.

The Department is a member of the Olympic Coast National Marine Sanctuary Advisory Council and is participating on its Marine Conservation Working Group to evaluate the feasibility of marine zoning within the Olympic Coast Sanctuary. Specifically, the working group will make recommendations on: 1) the effectiveness of existing zoning within the sanctuary; 2) the development of an intertidal zoning strategy; and 3) the development of a public education and outreach strategy regarding zoning. At this point, the working group is focused on developing a strategy for the intertidal area and plans to have this completed by next year; following this, the working group will make recommendations regarding the status of other zoning activities within the sanctuary, perhaps for nearshore rockfish.

Pacific Fishery Management Council (PFMC) Stock Assessment (Contributed by Farron Wallace (360) 249-4628), Annette Hoffmann (360) 902-2535), and Jack Tagart (360) 902-2855)

The status of stocks for black rockfish was last determined in 1994 (Wallace and Tagart, 1994). The population was assessed using the age-structured version of the stock synthesis model. The population was regarded as healthy, stock abundance was estimated to be either increasing after passing through a low in the late 1980s or in a gentle decline showing relative stability from 1990 to 1994. The recommended allowable annual yield was 517 mt based on an F45% exploitation strategy. The current analysis reprises estimates based on the 1994 stock synthesis model, introduces a new parameterization of stock synthesis (1998 configuration) and presents a completely new model written in AD Model Builder.

Stock Structure

Tagging data collected by the Washington Department of Fish and Wildlife, suggested that Cape Flattery and Cape Falcon bound a single coastal Washington-northern Oregon black rockfish stock. This hypothesis was corroborated by a recent genetic study that evaluated a set of samples organized into three geographical clusters from California to northern Washington. The northern cluster, encompassing northern Oregon to northern Washington, was found to be significantly different from the southern clusters. We assumed that black rockfish distributed in this area represented a unit stock. All biological parameters, data analysis and yield projections presented in this assessment are intended to describe this portion of black rockfish coast-wide distribution.

Model Inputs

The stock synthesis models used catch data from 1970 to 1998 for each of three fisheries (trawl, commercial line and sport). The AD model used trawl catch data from 1986 to 1993, line catch data from 1986 to 1995, and sport catch data from 1986 to 1998. The catch data interval corresponded to the period of an active black rockfish fishery for each gear. In addition, AD model also used gear specific estimates of catch variance. For the commercial gears (trawl and line), catch variance was available for each year of catch data. Catch variance for the sport fishery was available for 1990 to 1998.

Age specific inputs included catch-at-age (by numbers or weight), proportion-at-age (by numbers), and proportion-at-age (by weight). The stock synthesis model utilized proportion-at-age in numbers for each fishery weighted by sample size. AD model incorporated proportion-at-age by number for the sport fishery, and proportion-at-age by weight for the trawl and line fishery. Additional AD model inputs were the associated variances for the estimated proportions-at-age.

Revised estimates of weight-at-age and maturity-at-age were generated for the current analysis. We determined that a single weight-at-age vector effectively represented the line and sport fishery, and a separate vector was estimated for the trawl fishery. Inspection of the 1994 stock assessment report revealed that we had inadvertently used the raw proportion mature-at-age data rather than the estimated proportion-at-age from the regression fit to the logistic. Predicted values from the regression were used in this analysis.

Auxiliary data

From 1988 to 1990 and beginning anew in 1998, black rockfish from known areas of high densities were tagged such that the tags were distributed in proportion to perceived relative abundance. Tag release data from 1988 to 1990 and in 1998 and recovery information collected between 1988 and 1994 and in 1998 where used as auxiliary data.

For the 1994 stock synthesis model configuration, two auxiliary data sets were used as black rockfish abundance indicators: tagging CPUE and recreational bottomfish effort (Wallace and Tagart, 1994). Estimates of fishing mortality derived from tag data were used as an effort index to tune the 1998 stock syntheses model configuration. Fishing mortality rates were estimated from the ratio of tags recovered to tags released and adjusted for the fraction of catch that occurred before the tagging study.

In the AD model configuration, tag recovery was modeled explicitly. Auxiliary data inputs were the annual number of tags released, the number of tags recovered stratified by year of release, annual tag loss rate, and tag reporting/recovery rates. Tag reporting rates for the 1988-94 recoveries were unknown. Analyses were conducted to evaluate the sensitivity of estimated abundance to those reporting rates.

Model Description

We used the AD model to assess current black rockfish abundance because the model explicitly included sampling uncertainty and provided the most statistically rigorous model with the fewest set of assumptions. The two stock synthesis model configurations were provided as a basis for comparison; one as a comparison to the previous assessment (1994 configuration) and one as a "parallel" to the AD model (1998 configuration). Neither stock synthesis configuration is presented in detail here.

The two key features of the AD model were (1) the parameterization of the expected catches at age and (2) the definitions of the sampling unit for the different types of data input. The parameterization chosen mostly affected parameter bias whereas the sampling unit designation mostly affected estimator variance. Both bias and variance were components of overall parameter uncertainty. The parameterization and the sampling unit definitions were designed to conform to the actual sampling protocol used, thereby propagating sampling uncertainty through to the final biomass estimates.

The first key feature, parameterization, was designed to minimize assumptions on the population dynamics. A fully parameterized model included (a) yearly fishing mortalities, (b) initial numbers at sex/age, (c) selectivity by fishery/sex/age and (d) natural mortality by sex/age. In a fully parameterized model the fishing mortalities, initial numbers and selectivities were not constrained. Natural mortality was constrained to fit a four-parameter logistic function where two of the parameters were estimated. Simulation studies conducted on fully parameterized models provided empirical evidence that the estimators and estimator variances were approximately unbiased. Additionally, they showed the estimators were approximately normally distributed.

The model for the black rockfish data included constraints on selectivity and natural mortality. When the black rockfish data were introduced into a fully parameterized model, the estimated selectivity parameters showed no discernible patterns by sex or age. Since there was not enough information in these data to estimate selectivity well, we assumed a selectivity pattern and fixed selectivity parameters at a fully selected rate for all but the youngest ages. Based on the assumptions made for the 1994 assessment (Wallace and Tagart, 1994), a constant rate of natural mortality was assumed for males and age specific rate for females. However those rates were estimated internally in the model. The remaining population’s dynamics parameters were freely estimated.

The second key feature, sampling unit definition, affected both catch age data as well as tagging data. For the catch age data, the sampling unit was defined as the "basket" or boat rather than the individual fish to mimic the port sampling procedure. The collection of "baskets" yielded empirical estimates of variance among the proportion at age vectors. Those variances were explicitly fixed into the likelihood functions describing catch at age.

For the tagging data, the sampling unit was the individual tag. This designation yielded a multinomial likelihood function where the tag recovery probabilities were calculated from the catch at age population dynamics parameters and an independently estimated tag loss rate. The recovery probabilities were also a function of a tag-reporting rate. However, the reporting rate was unknown in 1988-1994. Because of changes to the tag sampling protocol, the reporting rate in 1998 was equal to the proportion of the catch that was sampled and thus was treated as known.

Testing demonstrated that the final model converged reliably to the same solution. Initial parameter seeds were forced away from the maximum likelihood estimates by randomly generated deviations up to 10% of the original estimate. With this level of perturbation, virtually 100% of runs converged at the original maximum likelihood estimates. With a 50% perturbation approximately two-thirds of the runs converged to the original maximum likelihood estimates.

AD Model Outcomes

Initial attempts to estimate selectivity parameters using the AD model were abandoned after observing that the unconstrained estimates were highly variable from age to age. The apparent instability in these parameter estimates implied that there was too little information within the age data to obtain reliable independent selectivity estimates. Consequently, we fixed the sport and line fishery selectivity parameter values at 50% for age 6, 80% for age 7 and 100% thereafter; and, we fixed the trawl fishery selectivity at 10, 30, 65 and 100% for the age 6 through 9+ age groups. These values were consistent with catch curve estimates as well as estimates from the 1998 stock synthesis model.

Natural mortality was held constant for male black rockfish but was allowed to increase with age for females. The mortality rate was estimated by the model. However, mortality rates varied proportionately with the change in tag reporting rate. Estimated male natural mortality ranged from 0.24 at the 25% tag-reporting rate to 0.41 at the 75% reporting rate. Female mortality estimates at age 6 were lower than those for males but they were nearly double the male rate by age 16+. Female mortality rates ranged from 0.16 to 0.54 for the 25% tag reporting rate, and from 0.34 to 0.70 for the 75% rate. Model estimates of natural mortality were higher than those determined from the 1981 catch curve analysis (0.18 for males and 0.26 for females). It may be that black rockfish experienced more severe environmental conditions over the past decade than they did during the 1970s, the period affecting survival illustrated by the 1981 catch curve.

Total stock biomass varied with tag-reporting rate, with beginning (1986) biomass highly variable and ending biomass nearly constant. In general the biomass trend is shown to have declined since 1986. Minimum stock biomass ranged from 18 to 48% of maximum biomass over the 13 year interval. Beginning biomass ranged from 25 k mt at the 75% tag-reporting rate to 7 k mt at the 25% rate. Ending biomass was approximately constant at just under 6 kmt. At the higher tag-reporting rates (> 25%), 1986 biomass was noticeably greater than that estimated by the previous assessment (Wallace and Tagart, 1994).

Like natural mortality and stock biomass, recruitment varied with tag-reporting rate. Estimated geometric mean number of age 6 recruits ranged from a low 800 thousand, to a high of 2.1 million fish. Recruitment was most variable at the 25% tag-reporting rate demonstrating a 6 fold change from minimum to maximum. The greatest degree of stability among the tagging rates reported was observed at the 50% tag-reporting rate reflecting a 3 fold change from minimum to maximum. The time series of spawner-recruit data is short (7 years) and there is no indication of a definitive spawner-recruit relationship. Estimated value of the Beverton-Holt spawner recruit function shape parameter (A) is 0.76 for data taken from the 50% tag-reporting rate outputs. The non-linear regression R² value is 0.07 and clearly not significant.

Spawning biomasses supporting recruitment over the past 7 years are very high in comparison to target spawning biomass estimated by the equilibrium yield model (as much as 10 times greater). Spawning biomass supporting upcoming recruitment is 25-50% of the values in the recent past, but still approximately double the equilibrium target spawning biomass (Table 1). Nevertheless, the reduced level of spawning biomass supporting incoming recruits over the next three years may indicate that lower levels of recruitment should be anticipated.

It should be noted that the number of recovered tags was a function of the total catch used in the model, whereas the actual catch that contributed to the observed tag recoveries, was only a fraction of the total catch. However, the tag reporting rate parameters could compensate for the process error. That is, if the true tag reporting rate were known, the process error could be overcome by multiplying the tag reporting rate by the fraction of the catch that actually contributed to the tag recovery data. The year specific product of reporting rate and correct fraction of the total catch could then be entered as a substitute "tag reporting rate". The new tag-reporting rate would incorporate the necessary adjustment for the portion of the catch that contributed to tag recovery.

Tag recovery data used in the model represent catches landed in Westport. These landings account for approximately 70% of the total black rockfish catch. Available empirical data on tag reporting rate suggest a 95% reporting rate. The "adjusted tag reporting rate" required by the model to evaluate these data is 0.66 (0.7*0.95). This implies that in order to overcome the model process error, outcomes associated with the 0.5 to 0.75 reporting rate would best represent the expectation that actual reporting rate was 0.95. Model runs that used a 50% reporting rate, represent a actual tag reporting rate of 0.71 and those using a 25% reporting rate are representative of an actual tag reporting rate of 0.35.

Projected Yield

Projected yields are calculated from a simple deterministic equilibrium yield model. Data inputs included estimates of 1998 numbers-at-age; age specific schedules for natural mortality, selectivity, weight, and maturity; and, an estimate of annual recruitment. Target exploitation rates were calculated such that 45% of the estimated unfished spawning biomass was preserved.

A decision table was prepared that portrays projected yield (catch biomass), spawning biomass, and total stock biomass over the next 5 years. Lower and upper bounds to the decision table represent adjustments based on the variance associated with total stock biomass. Separate outcomes are reported for four hypothetical tag-reporting rates: 75, 50 and 25% (Table 2). As noted above, model process error affected the actual tag reporting rate associated with each of these model runs; therefore users should focus on results from the 25-75% tag-reporting rate columns to avoid incorporating model process error.

Estimated 1999 stock biomass was 9,500 to 10,100 mt dependent on tag-reporting rate. Spawning biomass in 1998 was 178 to 201% of the equilibrium spawning biomass associated with an F45% exploitation rate. Projected biomass is expected to decline over the next 5 years (Figure 1). Nevertheless, estimated spawning biomass in the year 2001 will still be 130 to 175% of the target biomass. The expected average yield over the next three years ranges from 655 to 864 mt. This is contrasted with the current sport fishing catch biomass of 254 mt.

Under the most pessimistic scenario, current catch exceeded the estimated average yield over the next three years by about 10-20%. Therefore, the black rockfish stock can be characterized as declining in abundance but healthy, i.e., displaying abundance levels in excess of those assumed to promote sustainable production.

Return to Table of Contents

Black Rockfish Tagging study (Contributed by Farron Wallace (360) 249-4628)

In 1998, WDFW began a multi-year mark-recapture survey near Westport Washington, the principal location of recreational landings of black rockfish along the Washington coast. The survey design involves five annual releases, and seven years of tag-recovery monitoring in the sport fishery. Aboard the WDFW research vessel Corliss, 2,870 and 4,041 black rockfish were captured, tagged and released during 1998 and1999 respectively. Fish were released on pinnacles distributed throughout the area fished by the Westport charter fishing fleet. Each fish was tagged with two coded wire tags (CWT) placed in the opercular musculature: one on each side of the fishes head. Double tagging of fish, and holding experiments will be used to validate assumptions about tag shedding and tag related mortality (both tags remained in all fish recovered during 1998 and 1999). The tags were marked to allow for identification of specific individuals upon subsequent recapture.

In both 1998 and 1999, approximately 43% of the total Westport recreational black rockfish catch was sampled for tags by passing fish carcasses through a CWT tube detector. A total of 48 tags have been recovered to date. Cooperation of the charter boat industry was very good and enabled us to achieve the high sample proportion of the total number of fish landed (including those filleted at sea). Mark-recapture data will be used to produce estimates of abundance, survival, and mortality for black rockfish in the Westport coastal area. Population parameter estimates will be incorporated into the 1999 black rockfish age structured model.

Data analysis show the importance of tagging as many fish as possible each year, and conducting an accurate and thorough sampling of as large a proportion of the catch as possible for tags. We hope to increase our releases and sampling rate during 2000. Study results so far are quite promising and efforts may be expanded to include the entire Washington coast in subsequent segments.

Coastal black rockfish volunteer Charter Logbook Program.

In collaboration with the sport charter boat industry, WDFW developed a new logbook to collect site specific catch and effort information during directed coastal sport bottomfish trips. This is a volunteer program that is endorsed by the Washington Charter Boat Association and in 1998. We are hopeful that more vessels will participate in 2000. This information will be crucial to ascertain the overlap of the fishery and tag releases.

Pacific Fishery Management Council (PFMC) Stock Assessment

In 1998, the PFMC adopted amendment 11 of the Groundfish Management Plan, which established a minimum stock size threshold of 25% of unfished biomass. Based on existing biomass estimates (Jagielo et al. 1997), lingcod was declared formally to be overfished, thereby requiring the development of a rebuilding plan. Tom Jagielo conducted a rebuilding analysis for northern area lingcod (Vancouver and Columbia areas) which was used by PFMC to develop a rebuilding plan for lingcod coastwide in 1999. The analysis indicated that, under conditions of no fishing mortality, the stock could rebuild within 10 years. WDFW is leading efforts to produce a new coastwide lingcod assessment for the 2000 stock assessment cycle.

Cape Flattery Lingcod Survey and New Coded-Wire-Tag (CWT) Study

The annual February-March lingcod survey with bottom troll gear at Cape Flattery was conducted for the 14th year in 2000. This survey produces estimates of lingcod survival and abundance at Cape Flattery, which have proven useful for the PFMC stock assessment, particularly as an aid to estimate recruitment. Since 1998 we have employed coded wire tags in the mark-recapture survey as internal marks, and WDFW samplers have examined as many fish as possible from the sport catch at Neah Bay with an R8-tube CWT detection system. The new survey design involves a much more labor-intensive recapture sampling effort, but eliminates the need for estimates or assumptions about tag reporting rates. The direct catch sub-sampling approach also has the potential to yield estimates of abundance with greater precision than the voluntary tag return sampling design, as estimates of the total sport catch and its variance are not required.

National Undersea Research Program (NURP) Project

A joint WDFW-NMFS project, funded by the National Undersea Research Center (NURC) began in early 1998. The project was designed as a pilot study to determine sample size requirements for estimating differences in rockfish and lingcod abundance between trawlable and un-trawlable habitats. The ultimate purpose of this work is to evaluate the "habitat" bias of trawl survey estimates of abundance. The project involved in-situ dives using the submersible Delta.

Specific objectives of the study were to 1) qualitatively assess the catchability characteristics of the submersible, and 2) determine the size of an experiment (e.g. number of submersible dives) needed to obtain sufficient statistical power to reliably compare fish densities in trawlable and untrawlable habitats.

Work in 1999 and 2000 consisted of processing videotapes to finalize estimates of area-swept and fish counts by species. Calibration of the submersible Delta's video camera was conducted with Waldo Wakefield of NMFS under controlled conditions at Sandpoint. This calibration will permit much improved quantification of area-swept measurements which are obtained from the videotape record of the dives.

The videotapes of the dive transects, which were collected using a random sampling survey design, affords additional opportunities not addressed in the original proposal. Future work of interest includes detailed habitat analysis, quantification of the amount of observable trawl impacts on bottom habitat, and potentially macroinvertebrate analysis.

Sustainable Seas Expeditions (SSE) Project

In collaboration with the Olympic Coast National Marine Sanctuary (OCNMS), WDFW participated in the first year of the Sustainable Seas Expeditions research off the coast of Washington. This project is a multi-year endeavor with support from National Geographic, the Goldman Foundation, NOAA, and the US Navy.

Annette Hoffmann and Tom Jagielo of WDFW participated in a two week cruise off Cape Flattery on the US Navy ship Discovery Bay. The WDFW work involved investigation of the utility of the new DeepWorker 2000 single person submersible for conducting quantitative fish surveys. The objective in 1999 was to conduct "proof of concept" work to evaluate an experimental approach to determine if the presence of the submersible introduces a bias in fish counts. The intention was to use two DeepWorker submersibles (one stationary and one in motion) to evaluate fish attraction or repulsion due to the moving vehicle, however, logistical constraints due to weather and equipment failure limited activities substantially. The next SSE field work in Washington is expected to occur in 2001. More information about SSE can be found at: www.sustainableseas.noaa.gov .

Return to Table of Contents

Nearshore Habitat Mapping (Contributed by Tom Jagielo (360) 902-2837)

Plans are in place to conduct detailed side-scan and muli-beam bathymetric mapping of a nearshore area off Cape Alava in June of 2000. Working in cooperation with the Olympic Coast National Marine Sanctuary, WDFW will charter a Navy vessel to conduct around the clock mapping operations. The combined efforts of WDFW and OCNMS are expected to provide for 15 days of survey effort. The detailed habitat map will be used by WDFW to develop a stratified random sampling design for submersible rockfish and lingcod surveys in 2001.

Field Evaluation of a Remotely Operated Vehicle (ROV) for Groundfish Surveys (Contributed by Marty Peoples (360) 249-4628 and Bob Pacunski (425) 379-2314).

Background

Remotely operated vehicles (ROV) have been widely used in recent years as a tool in marine environments by private industry and research groups. An ROV presents advantages over submersibles and divers in cost and safety, as well as some disadvantages including maneuverability. Marine Resources Division (MRD) staff have been investigating the utility of an ROV for assessing groundfish stocks as well as other uses. Since groundfish populations exist in a variety of habitats and depths the ROV would need to be functional in most of these habitats.

ROV’s have successfully been used by other agencies to perform similar sampling in coastal waters. However, because each application presents different variables, we performed a series of tests to assess the utility of an ROV as a sampling tool for Puget Sound and the coastal waters of Washington State. Earlier tests of the ROV in Puget Sound by MRD staff in 1998 showed some promise for this technology, however time and logistical constraints did not allow for a complete test of the ROV’s abilities. To evaluate the ROV’s ability to operate in more demanding environments, we conducted our latest testing in Neah Bay, Washington, under conditions similar to those found on the open coast of Washington.

Equipment Evaluated

The in-situ system consists of three main components; the ROV, umbilical cable (tether), and control console. The ROV used in this evaluation was a Phantom HD2+2 manufactured by Deep Ocean Engineering. The unit consists of a main body enclosed within an aluminum frame to which four (4) horizontal thrusters and one (1) vertical thruster are attached. The main body provides the flotation for the ROV and is composed of microscopic glass beads encased in a polymer resin matrix. Two fixed, 100 W lights and a color video camera with tilt control are attached to the main ROV body. The main body also encloses the control circuitry for the camera, lights, thrusters, and other optional equipment. The ROV umbilical is a 31 wire kevlar wrapped cable encased in a neutrally buoyant jacket. All video signals as well as 12 VDC and 120 VAC power are transferred through the umbilical. The control console integrates all camera, light, and thruster controls. A small video screen on the console allows the operator to constantly view the environment directly ahead of the ROV.

A Trackpoint II system provides the ROV pilot with the position of the ROV relative to the support vessel. The system utilizes a directional transceiver affixed to the vessel and a passive transponder attached to the ROV to provide this information. By interfacing the support vessel’s GPS and the Trackpoint system to a computer equipped with specialized survey software, the geographic coordinates of the ROV can be displayed to provide a track of the ROV during the deployment. Although listed as optional equipment, without the Trackpoint (or other similar equipment), it would be impossible to reference the location of the ROV to the support vessel, significantly reducing the functionality of the ROV. Hypack Survey software allowed for the display of planned survey lines, the support vessel, and the ROV via the Trackpoint II interface.

A Trimble AG122 differential GPS (DGPS) provided input to the computer for plotting and recording survey tracklines and the position of the support vessel. TriTech SeaKing dual frequency scanning was used for detecting objects around the ROV which were out of the visible range of the camera.

Evaluation Approach

The study design for these tests were not set around a rigid biological sampling regime but rather a series of tests to determine the functionality of this equipment as sampling tool in coastal and inland waters. Establishing predetermined transects in varying depths, currents and bottom types and testing the ability to follow these transects with the ROV was the primary objective. We also hoped to gauge the repeatability of these transects over time. Several types of transects were chosen for testing:

1. a straight line running with or against the current,
2. a straight line running across the current or varying angles to the current ,
3. a transect incorporating 90 degree turns, either a rectangle design or zigzag design,
4. a transect using slow turns in the shape of a figure 8 or similar design.

Test dives for this project were conducted in waters near Neah Bay. This area includes a wide variety of bottom habitats from sand to rocky reefs. Transects were selected to include the various bottom types in this area and depths. Deployments were made in waters near Bullman Reef, Sail Rock, Waadah Island and Midway Rocks in depths ranging from 10-350 ft. This time period was selected for testing due to minimal tidal exchanges, which present the weakest and most favorable currents for sampling.

Observations

Overall, this demonstration proved to be educational and enlightening despite a number of problems. The experience was particularly illustrative of the problems encountered when working with unfamiliar equipment with limited training and experience. Within the first two days it became obvious that we did not possess the necessary skills to accomplish all of our stated objectives, and that we should concentrate on learning the basics before attempting anything complicated. This being the case, we spent the first several days learning to pilot the ROV in shallow water over less demanding habitat. This turned out to be more difficult than originally anticipated since the process required a coordinated effort between the ROV pilot, vessel operator and the deck crew handling the umbilical, although the process became more efficient by the end of the week.

Operations in shallow (<100 ft) waters were accomplished with only minor problems. At times the current, wind, and tether length had an effect on the ROV’s maneuverability, however none of these factors appeared to be significant. We were able to drive the ROV over varying terrain under a variety of current and sea conditions and were successful in maintaining a relatively straight line over long distances. Operating in minimal currents (<1.5 kts) was not difficult, although we were unable to run across or against moderate to strong currents, mainly due to ROV rolling and lack of power. For the majority of the shallow water deployments, no depressor weight was used and piloting the ROV in a straight line was relatively straightforward. Umbilical drag in these depths was minimal and did not to appear to adversely affect the ROV’s performance.

Working at depths between 100' and 250' was more difficult but not impossible. As we observed in the shallow water deployments, the ROV could not be flown at oblique angles to the prevailing current since it tended to roll over onto its side. With no depressor weight attached, the increased length of the tether in the water column produced considerable drag on the vehicle, often causing it to list to one side and significantly affecting the vehicle’s maneuverability. The addition of the depressor weight to the umbilical helped to minimize this problem and improved vehicle performance. Results at these depths indicated that, in order to achieve optimal performance, the ROV should be flown either with or against the current with a depressor weight attached to the umbilical.

At depths exceeding 250', maintaining a separation distance of less than 100' was difficult while the support vessel was under power. This appeared to be due mainly to differences in current speed and direction between the surface and the bottom. When the support vessel was allowed to drift, we had no difficulty reaching the bottom or staying with the Molluscan as long as we used the depressor weight as an orientation point.

The results we experienced were encouraging for conducting shallow water (<200 ft) deployments. We did not damage the vehicle despite several hangups in the kelp and rocks, and we proved to ourselves that we could navigate the vehicle in a relatively straight line over long (500+ m) distances. We became familiar with the idiosyncrasies of the ROV system and associated electronics, and, as a result, have a much better understanding of what will be required to work in deeper, more demanding habitats. Also, given the amount of concentration required to pilot the ROV, little attention was given to the SeaKing sonar, although it was occasionally helpful in identifying reef features during several deployments.

With time and experience, it is possible that an ROV could be used to assess deepwater habitats, especially within the more protected waters of Puget Sound. However, given our experience with the ROV in this demonstration, we conclude that we would be pushing the limits of a Phantom-sized ROV by working in the demanding environment of the open coast. Significantly more experience piloting the ROV will be required before attempting any work in coastal waters.

We might have achieved greater success piloting the ROV in deeper waters had we been able to plot the position and course over ground (COG) of the ROV and support vessel. Although the Trackpoint system provides the position of the ROV relative to the support vessel, the ROV’s position is always referenced to the bow of the vessel. Thus, sudden changes in vessel heading result in the ROV "jumping" around on the display screen, confusing the ROV pilot and compromising navigational ability. This is especially problematic if the boat is constantly being steered side to side to counter the effects of wave action, making the ROV appear to "jump" back and forth to opposite sides of the vessel. By plotting the actual position of the ROV and support vessel on a computer chart, the problem of ROV "jumping" can be eliminated, making it much easier for the ROV pilot to maintain a stable course heading.

Specific Concerns

Repeatability. Most sampling designs call for predetermined transects that are repeated over time, and our results suggest that repeatable transects would be difficult or impossible to accomplish in most of the sea and current conditions encountered during our testing. The ability to perform repeated sampling appears to be greatest in shallow waters (<50') but still is problematic. Variability of wind and current conditions most likely were the primary factors affecting the ROV’s performance. In high wind conditions the support vessel became difficult to control, complicating the process of keeping the ROV within nominal separation distance (<100' in most cases). Also, the small size of the ROV restricted it to running with or against the current during most of the test dives. If the sampling design did not call for repeated transects over time, the ROV may prove useful in shallow coastal waters. In contrast to coastal waters, the more protected and predictable nature of conditions in Puget Sound increases the probability of conducting repeatable transects over time.

Umbilical dragging on bottom. The umbilical cable was often observed dragging on the bottom during test dives when using the depressor weight. This did not present a concern on flat or low relief sand or gravel bottoms where our deepwater test dives occurred, but could present problems in rockier habitats. With more experience or perhaps the addition of floats between the depressor weight and the ROV, the amount of tether dragging on the bottom could be reduced or possibly eliminated, however we were unable to test the float hypothesis during the course of the test dives. Conversely, with no depressor weight attached, the umbilical cable tended to remain suspended above the bottom, especially with the pull of the vessel, minimizing the risk of entanglement.

Ability to follow transect at 300+ feet. At depths greater than 300', it was nearly impossible to keep the ROV close to the support vessel while it was under power, preventing us from running a predetermined transect. We found that drifting with the wind and current and powering the boat only to stay in position over the ROV as it swam with the current was all we could accomplish. Hence, we could run a "transect" that the current and wind selected for us, but could not choose a direction ourselves. Consequently, under the conditions in which we conducted our testing, we were forced to conclude that the likelihood of performing repeated transects at these depths is relatively low, although reaching the bottom and viewing habitat and groundfish was certainly achievable.

Summary. Results of this demonstration suggest that the Phantom ROV in the configuration we tested would not be capable of conducting repeatable deepwater transects in an open coastal environment. It is clear however that operator experience is a critical factor in successfully piloting the ROV in demanding conditions, and, lacking unlimited access to an ROV, it is unlikely that we could gain enough experience with the ROV to improve our skills beyond the basic level. Also, the lack of geographic plotting capabilities during testing probably compromised our ability to pilot the ROV in more extreme conditions, and, without further testing, we cannot conclude to what extent this capability might have improved our results. Working in coastal current and weather conditions was extremely challenging, especially for novice pilots.

We experienced both positive and negative results during the course of our testing, and further testing of this technology should be considered. Even if an ROV proves impractical for sampling in coastal waters, the utility of an ROV for working in the more protected environment of Puget Sound should not be dismissed.

The management value of a Marine Protected Area (MPA) for fish resources is based on the concept that creating a specific geographic area where fishes, habitat, and trophic relationships are free from all consumptive exploitation, will benefit regional fish resources and, ultimately, the fisheries utilizing these resources. A successful MPA will increase the potential in the region for maintaining genetic diversity and balanced age structure in the fishes, and preserving biodiversity and vital trophic interactions. A key assumption of MPAs for fishes is that protection of reproductive age-classes will result in successful spawning, and, thus, a source of larvae for replenishment of these species in the region.

The efficacy of MPAs for marine fishes cannot be determined without validation of this key assumption. Validation of the requirement that larval and juvenile fishes produced in an MPA survive and disperse to produce recruitment to stocks in the region, requires identification and recognition of the larval and juvenile fishes.

The numerous rockfishes (Scorpaenidae) and surfperches (Embiotocidae) in nearshore temperate waters of the eastern Pacific Ocean are very important due to their ecological value in structuring temperate reef ecosystems, as well as their value in fishery harvests and non-consumptive recreational diving. These groups of semi-resident fishes provide an excellent array of species for testing the efficacy of MPAs. A major benefit of MPAs for these viviparous fishes is protection of the adults to enhance the potential for successful parturition.

This research has developed and tested the first trans-generational procedure for inducing strontium marks in larval rockfishes and juvenile surfperches. In this field-applicable method, the gestating females transfer injected elemental strontium chloride to the embryos in vivo during the normal nutrient transfer processes in these viviparous fishes. The maternal energy contribution to developing rockfish embryos can range from 11.5% to 92.1% between species (Dygert and Gunderson 1991), with the nutrient transfer occurring through absorption of maternal substances by the embryonic epidermis and hindgut epithelium (Shimizu et el. 1991). The maternal energy supplied to gestating surfperch embryos appears to be mainly from nutrient enriched ovarian fluid that is ingested and absorbed by a modified intestine (Turner 1938, 1952).

This trans-generational procedure for marking developing embryos with elemental strontium chloride was tested with brown rockfish, Sebastes auriculatus, in June 1999, and kelp perch, Brachyistius frenatus, in August 1999. Late-stage gestating females were collected in Puget Sound, Washington, and held in two treatment groups and one control group at the Point Defiance Aquarium, Tacoma. The females in the treatment groups were given intra-muscular injections of strontium chloride in either 9,000 ppm and 30,000 ppm solutions, using a saline (NaCl) solution carrier. The control group females received only saline solution injections. Detection of the strontium was through analysis of larval otoliths by Wavelength Dispersive Spectrometry (WDS) using a JEOL 733 electron microprobe (see Schroder et al. 1995).

The results from preliminary analyses of this procedure are extremely encouraging. Samples of otoliths from kelp perch juveniles in the 9,000 ppm and 30,000 ppm treatment groups showed specific bands of strontium at maximum concentrations that were about 5.0 and 9.0 times greater, respectively, than the ambient concentrations in the control group. Samples of otoliths from brown rockfish larvae in the 9,000 ppm treatment group showed concentrations of strontium that were from 1.6 to 3.3 times greater than the ambient concentrations in the control group. Each otolith was subjected to multiple SEM analyses. The inner portions of the otoliths from juvenile kelp perch in the treatment groups had ambient-level strontium concentrations prior to a distinct deposition of the injected strontium. In the extremely small otoliths from brown rockfish larvae, there was virtually no within-otolith variation in the concentrations of strontium. The between-otolith variations in the concentrations of strontium within the 9,000 ppm treatment group was possibly the result of differences between the female brown rockfish in the group, or the temporal or spatial variability of the embryos in the ovaries.

The analyses of these pilot study samples will be completed by June 2000 and the results published in a scientific journal. Funding is being sought for a more rigorous assessment of the treatment procedures.

References

Dygert, P.H. and D.R. Gunderson. 1991. Energy utilization by embryos during gestation in viviparous copper rockfish, Sebastes caurinus. Environmental Biology of Fishes 30:165-171.

Schroder, S.L., C.M. Knudsen and Eric Volk. 1995. Marking salmon fry with strontium chloride solutions. Canadian Journal of Fisheries and Aquatic Sciences 52(6):1141-1149.

Shimizu, M., M. Kusakari, M.M. Yoklavich, G.W. Boehlert and J. Yamada. 1991. Ultrastructure of the e[idermis and digestive tract in Sebastes embryos, with special reference to the uptake of exogenous nutrients. Environmental Biology of Fishes 30:155-163.

Turner, C.L. 1938. Histological and cytological changes in the ovary of Cymatogaster aggregata during gestation. Journal of Morphology 62:351-375.

Turner, C.L. 1952. An accessory respiratory device in embryos of the Embiotocid fish, Cymatogaster aggregata, during gestation. Copeia 3:146-147.

The RecFIN committee commissioned a statistical subcommittee with representatives from ODFW, CDFG, WDFW and NMFS. The primary objectives of the committee were to evaluate the different ocean sampling methods for estimating ocean recreational catch. The subcommittee met to begin their work and are in the process of conducting an assumption-based analysis.

Return to Table of Contents