NOAA-NMFS-NWFSC TM-31: Data Collection -- Groundfish (cont):

5.1 Introduction

This chapter describes federally coordinated sampling programs for lingcod, sablefish, and Pacific hake (also referred to as Pacific whiting). Sampling programs addressed here target shoreside landings of commercial fisheries, and involve coordination with two or more of the U.S. Pacific coast states. These federally coordinated programs are designed to follow the same protocols in each of the participating states and to help allocate effort to the programs in a manner consistent with coastwide needs. The sampling personnel used in each state are supervised by the states and are usually involved in other state programs.

5.2 Sablefish

Washington, Oregon, and California participate in a joint port sampling program for sablefish. As of 1995, the NMFS (Tiburon Laboratory) has the primary responsibility for coordinating this program. For example, appropriate design methods (including target sampling rates) and data summaries (including realized sampling rates) are provided by the Tiburon Laboratory.

The sablefish sampling program, which was established in 1986, is designed after the California species-composition sampling program for rockfish (see section 4.4). The coastwide sampling program continued without interruption through 1991. In 1992, sampling was discontinued in California and Oregon to allow effort to be directed towards lingcod; however, samples continued to be collected throughout the year at Washington ports. Sampling resumed in California and Oregon in 1993 and has continued in all three states through 1996. The specifics of the design were developed taking into account analyses by Hightower (1986). Sampling is done at each of the major port complexes along the U.S. Pacific coast from Morro Bay north.

The price paid for a landing depends on the distribution of the catch among "sort groups" (based on fish size), but these sort groups do not have the same status as market categories. All sort groups are not always recorded in the fish receipt databases; sometimes the original sort groups sampled are aggregated into broader categories. In California and Oregon, the sablefish sampling program differs from the California species-composition sampling program for rockfish in that the samplers are instructed to take clusters (i.e., subsamples) from each sort group on a sampled trip. In Washington, individual trips are often sorted into more groups than can be sampled in the time available. Consequently, only "ocean run" trips, which are not divided into sort groups, are sampled in Washington. These samples consist of three clusters.

Sampling personnel are requested to obtain samples from landings of both whole and dressed fish in each state. Sampling effort is distributed among gear types (fixed gear or trawl gear) and ports or port complexes. The fixed-gear season is generally quite short (several weeks), and during this period samplers are instructed to direct their sampling toward this fishery. For the trawl fishery, the current sampling goal is to obtain one complete sample (i.e., data collected from every sort group of a landing from a single boat trip) each month within each major port complex. Samplers distribute their sampling effort among different vessels when possible. Actual numbers of samples varied from 163 (9 in Washington, 50 in Oregon, and 104 in California) to 256 (44 in Washington, 99 in Oregon, and 113 in California) during the 1986-91 period.

Sablefish samples consist of two 50-lb clusters from a sort group (or three in Washington). When possible, clusters are taken from different parts of the landing (e.g., from different strap boxes, totes, or baskets). For fish landed whole, fork length and sex are recorded, and otoliths are removed. For fish landed dressed, dorsal length (from the insertion of the dorsal fin to the fork of the caudal fin) is recorded. The data are recorded on standard sablefish data sheets provided by the sablefish sampling coordinator at the Tiburon Laboratory.

Over time, sampling procedures have been modified in response to changes in the sort groups used by the fishery. In earlier years, landings tended to consist of a few general sort groups, such as ocean run (unsorted), small, medium, large, and so on. More recently there has been a tendency for the landings to be classified into more groups based on the weight of the individual fish. Beginning in 1993, the data sheet was modified so that samplers could record condition (dressed or whole) and whether the fish were landed as ocean run or sorted into weight groups. When the fish are sorted into weight groups, the sampler records the actual weight interval for the sort group sampled. The goal is still to obtain two clusters from a sort group and to sample each sort group from a trip.

Collected otoliths are cleaned, placed in plastic trays, and numbered so that later the otoliths can be matched to the measurements for that fish on the data sheet. Otoliths and data sheets are sent to the sablefish sampling coordinator at the Tiburon Laboratory approximately once a quarter.

As soon as the otoliths and data sheets are received at the Tiburon Laboratory, they are briefly examined to make sure that the otoliths can be matched to their corresponding data sheets. At the time of data entry, each sample is examined to ensure that the data sheet was correctly filled out by the sampler. The data are entered using an MS-DOS BASIC program, and the resulting data are stored in ASCII (text) form. After the data are entered, approximately 10% of the samples are randomly selected and checked against the computer files manually. If any errors are detected, then all of the samples are checked and errors are corrected. Updates to the samples from fish receipts and trawl logs are entered as they are received. These updates can include landing weight, INPFC area, minimum and maximum depth fished, and boat number. No standard reports containing derived quantities are produced. Data are made available to stock assessment scientists and others on request.

As of 1995, no ageing of sablefish is done because of a staff cutback and because low between-reader agreement in recorded ages may be contributing to biased estimates of age. Approaches are currently being explored to improve age-reading methods, which could lead to a resumption of routine age reading. Otoliths are currently stored at the Tiburon Laboratory in plastic trays and soaked in a 30% glycerin solution with a small amount of thymol added as a preservative.

5.3 Lingcod

California and Oregon participate in a federally coordinated sampling program for lingcod that began as a pilot program in 1992, while Washington continued a preexisting sampling program for this species. The federal lingcod program is coordinated from the Tiburon Laboratory by the lingcod sampling coordinator. The lingcod sampling program was modeled after the sablefish sampling program. The program was continued after 1992 originally because too few samples were collected in the first year to allow the sample data to be evaluated critically, and then because the information produced by the sampling program was viewed as a necessary component of ongoing management efforts for lingcod.

Sampling is directed towards the trawl fishery and is distributed among ports. The three major ports in Oregon and the seven major port complexes from Morro Bay north in California are covered by the sampling program. A sample consists of four 50-lb clusters, if landings are large enough. Landings of less than 50 lb are not sampled. Four clusters are selected so that roughly the same number of fish would be sampled as is collected in the Washington sampling program, where samples consist of 25 individual fish. Currently, landings of lingcod usually are not subsorted into different sort groups or market categories. The sampling design and software, however, can be adapted to accommodate multiple sort groups in the same manner as for the sablefish sampling program.

For each lingcod in the sample, the sex is determined and fork length is measured. The soft dorsal fin rays (numbers 1 through 8) are removed, placed in a labeled manila envelope, air dried, and then stored in a freezer until they are mailed quarterly to the lingcod sampling coordinator at the Tiburon Laboratory.

The target sampling goal is to collect two samples per major port or port complex every month. During the first two years of the sampling program, there was difficulty obtaining a sufficient number of samples, especially from California. Sampling effort was redistributed from other sampling programs in California to improve the quality of the estimates from the lingcod sampling program. Actual numbers of samples in 1992 were 39 in Oregon and 15 in California. Although the 39 samples in Oregon constitute less than the nominal target of 72 (two samples each month at each major port), sampling was adequate in Oregon during the months when lingcod were landed in substantial numbers. This is now also true for sampling in California. The lingcod sampling coordinator is currently considering reducing the target sampling levels to near the realized level.

Following receipt of the data sheets and fin rays at the Tiburon Laboratory, they are examined for completeness. Information from the data sheets is entered into a database program. The data entry program generates a printed copy of the information entered for each sample, which is then compared with the original data sheet. The database is updated after the fin rays have been examined to determine ages. The fin rays are embedded in plastic resin and sectioned prior to counting annual rings.

5.4 Pacific Hake

There is a federally coordinated sampling program for the shore-based Pacific hake fishery, with primary responsibility assigned to the Pacific hake port sampling coordinator at the Status of the Stocks Task, Alaska Fisheries Science Center (AFSC). Data files and archives of data sheets are maintained by the coordinator. Samples have been collected in California and Oregon since 1990. The 1990 program began as a pilot program located only in Eureka, California. In 1991, the program was expanded to include Crescent City and Newport, Oregon. Deliveries to Astoria, Oregon were also sampled in 1992. The sampling was initiated to characterize the landings of the shore-based fishery, as a supplement to data collected by onboard observers on the larger at-sea processed portion of the fishery.

Because of regulations on the catch and the migratory nature of the species, the open season for the shore-based fishery varies geographically and annually. This requires that the sampling design be modified in response to the fishery. There are two basic types of samples: "otolith" samples and "length-frequency" samples. Clusters are selected for both sample types, and both sample types are based on a fixed number of fish. For otolith samples, length, weight, and sex of individual fish are recorded and otoliths are removed for later examination. For length-frequency samples, only length and sex are determined and no otoliths are removed. Otolith samples are intended for use in constructing an age-length key, which can then be used together with the length-frequency samples to estimate catch-at-age and growth parameters. Compilations of the sample data are given in Dorn and Methot (1992) and Dorn et al. (1993). The total number of otolith samples to obtain is based on a specified total number of age determinations that can be made. This number of age determinations is contingent on the availability of personnel to read the otoliths. Within a year, the apportionment of samples among ports is roughly in proportion to expected landings, with the provision that the minimum number of samples at a port should be expected to produce 200 otoliths. This apportionment of samples among ports is made in an effort to reduce the variance of the combined (across ports) age composition, in comparison with what would be obtained with equal numbers of samples at each port. However, no formal analysis of optimal sampling allocation has been done. Starting in 1992, the recommended number of fish to collect per sample was 20, rather than the 50 fish per sample collected in 1990 and 1991. Age determinations are made from the otoliths by personnel assigned to the Age and Growth Unit at the AFSC. The number of age determinations for 1990-92 was 683, 941, and 1,070, which was based on fish collected from 15, 26, and 47 samples, respectively. The target number of otoliths for 1993 was 1,750.

A length-frequency sample consists of 100-200 fish. In 1991 and 1992, the sampling program produced 2,672 and 4,472 recorded lengths. For 1993, the initial target sampling rate was one sample every two weeks at each port, with this sampling rate subject to alteration based on changes in the expected duration of the fishery. The samplers are instructed to sample landings randomly within a port, and to distribute their sampling effort over the duration of the fishery.

After delivery of the data sheets to the Pacific hake port sampling coordinator, the data are keypunched and checked for errors by an outside contractor. Data are kept in an ASCII format at the AFSC.

5.5 Acknowledgments

Brenda Erwin (CDFG), Martin Dorn (AFSC), Don Pearson, Tom Laidig, and Peter Adams (NMFS, Tiburon Laboratory) provided information and many patient responses to repeated questions.

5.6 Citations

Dorn, M. W., and R. D. Methot. 1992. Status of the coastal Pacific whiting resource in 1992.

In Pacific Fishery Management Council, Status of the Pacific coast groundfish fishery through 1992 and recommended acceptable biological catches for 1993: Stock assessment and fishery evaluation, Appendix A, p. A1-A58. (Available from Pacific Fishery Management Council, 2130 SW Fifth Avenue, Suite 224, Portland, OR 97201.)

Dorn, M. W., E. P. Nunnallee, C. D. Wilson, and M. W. Wilkins. 1993. Status of the coastal Pacific whiting resource in 1993. In Pacific Fishery Management Council, Status of the Pacific coast groundfish fishery through 1993 and recommended acceptable biological catches for 1994: Stock assessment and fishery evaluation, Appendix A, p. A1-A82. (Available from Pacific Fishery Management Council, 2130 SW Fifth Avenue, Suite 224, Portland, OR 97201.)

Hightower, J. E. 1986. Sampling strategies for the Washington-Oregon-California sablefish fishery. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-SWFC-63, 43 p.

Table of Contents