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

1.1 Motivation for This Document

An important function of any government is monitoring the activities of its industries to gauge their productivity and health and to collect taxes. With the marine fishing industry, which exploits a publicly owned resource, government has the added responsibility of managing the industry to ensure that the natural productivity of the fish resources is not squandered and the overall integrity of the marine ecosystems is not compromised. At the least, the agencies that are responsible for monitoring and managing the fisheries need to know the magnitude of the fish landings and fishing effort. In addition, when a fish resource becomes heavily exploited, the management agencies generally require further, detailed information regarding the biological characteristics of the fish populations so that fishery scientists can assess what levels of harvest the resource can support. Many of the existing government institutions that routinely collect fishery-related data were created initially to provide basic landing statistics and then, as the fisheries expanded and fishery science matured, the institutions accepted additional responsibility for collecting the types of detailed information that now form the basis of the stock assessment and management processes.

The data collection programs for groundfish stocks commercially harvested from marine waters off Washington, Oregon, and California have been primarily designed to address two general management objectives: 1) provide estimates of the species compositions of the landings (in weight of fish), which are needed to monitor individual species that are managed by quotas or harvest guidelines, and 2) provide estimates of age compositions of the landings (in numbers of fish) and other biological characteristics, such as sex ratios, maturity stages, and length frequency distributions, which are needed to conduct stock assessments on regulated species. The data collection programs used by the individual states produce sample information that allows statistical inferences to be made about fish populations of interest. The accuracy and precision of the estimates depend on the sampling design and estimation methods that are used to derive the landing statistics.

Many of the monitoring programs used in commercial fisheries are based on the general theory and principles of sampling (e.g., Raj 1968, Sukhatme and Sukhatme 1970, Cochran 1977). Random sampling methods that have been appropriately implemented can provide valid estimates of population parameters, such as species and age compositions of the commercial landings, as well as measures of the estimates' precision, such as variances, standard errors, and coefficients of variation.

The sampling of commercial landings is a routine task and appropriate designs, field protocols, and estimation procedures have been discussed by various authors (e.g., Gulland 1955, 1966; Tomlinson 1971; Bazigos 1974; Quinn et al. 1983; Sen 1986; Pope 1988). In most cases, fishery monitoring programs are circumscribed by financial and logistical constraints, which ultimately result in estimates that reflect some level of compromise between statistical quality and practical considerations of sampling. Thus, it is imperative that commercial fishery sampling programs be well documented to ensure researchers can assess the accuracy and precision of the sample data that are to be analyzed and used to develop stock assessments and management strategies. Although most people accept that "bad" fishery data will generally not produce "good" stock assessment estimates, there have been few studies that have examined the relationship between data quality and stock assessment quality. As a consequence, there are no established standards by which to judge whether fishery statistics are adequate.

Given that fishery data collection programs in general evolve incrementally (Doubleday 1983, Pope 1988) and that fisheries management in the United States has until recently been primarily the responsibility of the individual states rather than the federal government (Knight 1977, PFMC 1990), it is understandable that many states have developed their own unique systems for collecting fishery data and deriving fishery statistics. However, when neighboring states attempt to assess and manage shared fish resources, difficulties necessarily arise if the states have not carefully coordinated their data collection activities. Currently, the Pacific Fishery Management Council (PFMC) advises the National Marine Fisheries Service (NMFS), which is responsible for the management of the groundfish fisheries off Washington, Oregon, and California. The PFMC relies on the individual states to develop and conduct independent sampling programs for groundfish landed at their respective ports. The states provide estimates of species compositions of the groundfish catch to the Pacific Fisheries Information Network (PacFIN).

The PacFIN system serves as a central database that is primarily utilized by researchers affiliated with state and federal fishery agencies that are responsible for the management of the groundfish fisheries off the U.S. Pacific coast (see section 1.2 and chapter 6 for detailed discussions that address this historical database). These data serve to monitor in-season progress towards quota attainment and as a historical record of landings for use in stock assessments and other fishery analyses.

In June 1992, the Pacific Groundfish Statistics Working Group was established, under the auspices of the Pacific States Marine Fisheries Commission, to conduct a comprehensive review of the data collection programs and data processing systems used by the individual states to monitor the commercial groundfish fisheries off the U.S. Pacific coast. Our primary objective was to produce a single document that addressed the current and past methods employed by each of the states for collecting, summarizing, and expanding the commercial groundfish landing information. We built upon a Pacific Coast Fisheries Data Committee report that outlined in general terms the data requirements and data collection activities associated with the U.S. Pacific coast groundfish fisheries (PCFDC 1990). We have endeavored to provide fishery researchers and managers with the supporting knowledge to evaluate the information associated with the groundfish fisheries of the U.S. Pacific coast, and to identify areas in which the current sampling programs and estimation methods are inadequate and can be improved. Researchers working with other fisheries, particularly those managed by multiple agencies, will undoubtedly find in this document information that is relevant and applicable to their own unique data collection programs.

We agree with Pope (1988) that "Every important data set should have someone to love and cherish it." We hope that this document will focus attention on the important, but demanding and often drab task of collecting fishery data, and that it will ultimately serve as a source of information for developing improved fishery management policies.

1.2 Historical Perspective

Since the late 1800s, commercial fishers, fish processors, and public consumers of fish products have utilized the groundfish stocks off the U.S. Pacific coast as sources of income and food (Miles et al. 1982). Three periods of growth characterize the history of the Pacific coast groundfish fishery: from the late 1800s to the early 1900s, little or no management was conducted on a disorganized and relatively small commercial fishery; from the early 1900s to the early 1980s, management on a rapidly expanding fishery was the responsibility of the individual coastal states; and currently, management on a diverse fishery and heavily exploited fish populations is coordinated by the federal government in conjunction with recommendations and support from the coastal states.

In 1976 the United States enacted the Magnuson Fishery Conservation and Management Act, which established an exclusive coastal fishing zone from three to 200 miles offshore known as the Exclusive Economic Zone (EEZ). This landmark legislation created regional fishery management councils, including the PFMC, that were responsible for developing fishery management plans. Since 1982, the domestic and foreign groundfish fisheries that operate within the EEZ off the U.S. Pacific coast have been managed in accordance with the Pacific Coast Groundfish Fishery Management Plan (FMP) and the Environmental Impact Statement for the California, Oregon, and Washington Groundfish Fishery (PFMC 1990) and subsequent amendments. Prior to the implementation of the FMP, the management of domestic groundfish fisheries was largely the responsibility of the individual states. However, many fisheries extended beyond state-imposed boundaries, which created management issues that required effective information exchange between the three state fishery agencies. To coordinate the activities of the individual states, the Pacific States Marine Fisheries Commission (PSMFC) was established in 1947. The PSMFC was not empowered with regulatory responsibilities, but was formed as a coordinating agency that would function as a data library for fishery-related information collected by the individual states.

The PSMFC recognized from its inception the necessity for coordinated collection and compilation of coastal groundfish statistics (PMFC 1948). Recognizing a need and effecting a result are two distinctly different events; not surprisingly, it was not until 1956 that the PSMFC began in earnest to develop the framework for what would become the PSMFC Data Series, a compilation of Pacific coast groundfish catch and effort data. The first published tables from the Data Series appeared in PSMFC's 14th Annual Report in 1961. The PSMFC Data Series were compiled for limited distribution, effectively to the management agencies responsible for regulation of the groundfish fishery. It begins in 1956 and ends in 1981, at which time the Data Series were replaced by PacFIN. The Data Series are contained in two volumes, one for groundfish and the other for shrimp and crab. Coincident with the development of the Data Series, the PSMFC created the PSMFC statistical reporting areas which have undergone some revisions over time.

The format and content of the Data Series were originally developed by the Research Committee of the PSMFC, an appointed group of biologists acting as technical advisors to the Commission. The Commission has also received advice from another group of technical advisors. The "Trawl Fishery Committee" (later named the International Trawl Fisheries Committee, the International Groundfish Committee, and currently the Canada/United States Groundfish Committee) was formed in 1958 to exchange information on groundfish regulations, groundfish research, and catch statistics between the United States and Canada. It has two members, one Canadian and one U.S., and a Technical Subcommittee (TSC). The U.S. member on the parent committee is the executive director of the PSMFC. The catch and effort data exchanged by the parent committee were compiled by the members of the TSC and, in fact, were the data prepared for the PSMFC Data Series.

The Data Series originally reported annual groundfish trawl landings by species and area, but this format was subsequently changed to monthly reporting at the request of the TSC. In 1970, a number of catch areas were added to the Data Series and in 1971 the Series began to report landings by gears other than groundfish otter trawl. Also in 1971, the Data Series began to be compiled under a computerized data system known as NORFISH. Early computer tapes of the compiled groundfish database have been lost; computerized data are only available since 1975 (Lynde 1986).

By the late 1970s, the data reporting system leading to the groundfish Data Series began to unravel. California stopped contributing in 1978. The United States and Canada were negotiating to exclude each other from fishing in their respective exclusive economic zones, and U.S. groundfish management was changing from exclusively state jurisdiction to state and federal jurisdiction. There was a recognized need for a revised catch data system, one which could be responsive to in-season management needs of the new federal management system. To facilitate the in-season management requirements of the FMP, the PacFIN data system was initiated in 1981, replacing the groundfish Data Series.

The PacFIN data system required several compromises when compared to the groundfish Data Series. The Data Series reported catch and effort by month and area of catch. These data were provided annually, 4-6 months after the end of the calendar year, and were effectively the contributing agency's record of final catch data. With PacFIN, the management agencies were trying to develop a catch data system with more timely reporting, something which could be used for in-season management. To achieve timeliness they compromised by using data from landing receipts that were unadjusted for area of catch, and they abandoned the reporting of trawl effort altogether. In effect, PacFIN was an approximation of the contributing agency's final data. These compromises were expected to be short-lived, with contributing states working toward in-season delivery, as soon as possible, of area-adjusted data on catch and effort. Oregon revised its catch data reporting system and began reporting area-adjusted catch data to PacFIN in 1986, while California and Washington continued to provide unadjusted data. Eventually, the contributing agencies agreed on revised formats for data transfer to PacFIN that would provide the necessary in-season adjusted data. The revised PacFIN reporting system, adopted coastwide in 1990, has become known as "redefined PacFIN."

While PacFIN data proved absolutely essential to monitor catch quotas in-season, the assessment of abundance of groundfish stocks managed by the PFMC required a historical catch data series. Stock assessment scientists were forced to compile these historical records independently from a variety of sources including the PSMFC Data Series, TSC reports, agency reports and computer files, and PacFIN data. Understandably, depending on the sources of their information, different analysts could compile different catch data for the same species. Consequently, the National Marine Fisheries Service sponsored the development of a historical landings database. This project, called HAL (Historical Annotated Landings database), contained catch data from 1956 to 1980, including information from the NORFISH system (Lynde 1986). The data were made available on computer files for convenient retrieval. Unfortunately, the database suffered from recognized data errors, some of which were captured and remedied, while others remained unresolved. As a result, the database has never received widespread use.

1.3 Overview

This document describes in detail the major features of the data systems used by the states of Washington, Oregon, and California for monitoring their groundfish fisheries. We developed this report because of a concern that the Pacific Fishery Management Council needed better data with which to manage the groundfish stocks under its jurisdiction. However, we decided that it was necessary to review and compare the three states' data collection systems before we could begin to determine how to improve data quality, either by standardization or changes in sampling levels or protocols.

The data collection systems described in this report have undergone major changes over the years and are likely to continue changing. For expediency, we have deliberately concentrated our review on the status of the systems as of 1992, but since then there have been substantial changes to portions of the systems. We have noted some of the pertinent changes in subsequent chapters. Although this document is a snapshot that was out of date before we even finished preparing it, it nevertheless provides a relevant overview of the data collection programs.

For the groundfish fisheries off the U.S. Pacific coast, there are two major categories of information routinely collected by the states of Washington, Oregon, and California: 1) weight and value of landings by species, geographic region, and season, and 2) biological characteristics of the landings (e.g., species, length, weight, and age composition). The data in the first category are used by various state and federal agencies for monitoring commercial fishing activities, collecting taxes, and for tracking landing quotas. Data in the second category are used primarily by the state fishery agencies and by members of the PFMC Groundfish Management Team, which is responsible for preparing stock assessments and developing regulations for managing the fisheries.

The document is organized as three major chapters (2-4), in which the authors describe state-specific data systems, data collection methods, and computational algorithms for the following items: landing receipts (fish tickets), trawl logbooks, species composition, and biological data. The chapter for each state follows a similar outline to facilitate comparisons between the three systems. The focus is on biological rather than economic aspects of the groundfish fishery, and little attention is given to economic information such as landed value and ex-vessel prices, although these data are often routinely collected on the landing receipts. Chapter 5 describes some special sampling programs that have been operated jointly by state and federal agencies. Chapter 6 reviews the Pacific Fisheries Information Network (PacFIN), which the state and federal agencies and the fishing industry use to share coastwide fisheries information. The concluding chapter (7) contrasts the three state systems and discusses possible methods for improving the overall quality of groundfish data. The Appendices contain examples of landing receipts used by the state fishery agencies (Appendix A) and a list of the common and scientific names for species of fish presented in this document (Appendix B).

Although some of the sampling and data systems used for the U.S. Pacific coast groundfish fishery are also used for collecting information about the commercial fisheries for salmon, herring, sardine, crab, and shrimp, we limited our review to the systems for the commercial groundfish fishery. We did not examine the program of the International Pacific Halibut Commission for monitoring the halibut fishery (Quinn et al. 1983) because this fishery is relatively small off the U.S. Pacific coast. Nor did we study the systems used for monitoring recreational landings of groundfish because the recreational landings are generally much smaller than the commercial landings, and the data collection systems are documented in various reports prepared by the NMFS (e.g., USDOC 1992). Also, we did not examine programs run exclusively by federal agencies, such as the NMFS Observer Program that monitors the offshore fisheries. Finally, although research surveys provide auxiliary information that is vitally important for stock assessments, our review does not include any sampling programs that are independent of the commercial fishery, such as the NMFS triennial bottom trawl survey (e.g., Zimmermann et al. 1994).

1.4 Acknowledgments

This chapter developed through critiques provided by all of the members of the Pacific Groundfish Statistics Working Group, including Dave Thomas, Brenda Erwin, Gerry Kobylinski, and Mark Saelens. We thank Rick Methot for reviewing drafts of this chapter and providing thoughtful comments that improved the final version.

Bazigos, G. P. 1974. The design of fisheries statistical surveys-inland waters. FAO Fish. Tech. Pap. No. 133, 122 p.

Cochran, W. G. 1977. Sampling techniques, 3rd edition. John Wiley & Sons, New York, 428 p.

Doubleday, W. G. 1983. Sampling commercial catches of marine fish and invertebrates: Workshop report (summary and conclusion). Can. Spec. Publ. Fish. and Aquat. Sci. 66:19-22.

Gulland, J. A. 1955. Estimation of growth and mortality in commercial fish populations. U.K. Min. Agric. Fish., Fish. Invest. Ser. II 18(9):1-46.

Gulland, J. A. 1966. Manual of sampling and statistical methods for fisheries biology, Part 1: Sampling methods. FAO Man. Fish. Sci. No. 3, 87 p.

Knight, H. G. 1977. Managing the sea's living resources. Lexington Books, D. C. Heath and Company, Lexington, MA, 140 p.

Lynde, M. V. H. 1986. The historical annotated landings (HAL) database: Documentation of annual harvest of groundfish from the northeastern Pacific and eastern Bering Sea from 1956 to 1980. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-NWFSC-103, 197 p.

Miles, E., S. Gibbs, D. Fluharty, C. Dawson, and D. Teeter. 1982. The management of marine regions: The North Pacific. University of California Press, Berkeley, CA, 656 p.

Pacific Coast Fisheries Data Committee (PCFDC). 1990. Groundfish data plan Pacific council Washington, Oregon, and California. Unpubl. manuscr., 24 p. (Available from Pacific States Marine Fisheries Commission-PacFIN F/AKC, 7600 Sand Point Way NE, Seattle, WA 98115.)

Pacific Fishery Management Council (PFMC). 1990. Pacific coast groundfish plan: Final amendment 4. Unpubl. manuscr., 74 p. (Available from Pacific Fishery Management Council, 2130 SW Fifth Avenue, Suite 224, Portland, OR 97201.)

Pacific Marine Fisheries Commission (PMFC). 1948. History and development of the commission: Coordinated plans for management of the fisheries of the Pacific coast. Pac. Mar. Fish. Comm. Bull. No. 1., 64 p.

Pope, J. G. 1988. Collecting fisheries assessment data. In J. A. Gulland (editor), Fish population dynamics, 2nd edition, p. 63-82. John Wiley & Sons, Chichester, England.

Quinn, T. J. II, E. A. Best, L. Bijsterveld, and I. R. McGregor. 1983. Sampling Pacific halibut landings for age composition: History, evaluation, and estimation. Int. Pac. Halibut Comm. Sci. Rep. No. 68, 56 p.

Raj, D. 1968. Sampling theory. McGraw-Hill, New York, 302 p.

Sen, A. R. 1986. Methodological problems in sampling commercial rockfish landings. Fish. Bull., U.S. 84:409-421.

Sukhatme, P. V., and B. V. Sukhatme. 1970. Sampling theory of surveys with applications, 2nd edition. Iowa State University Press, Ames, IA, 452 p.

Tomlinson, P. K. 1971. Some sampling problems in fishery work. Biometrics 27:631-641.

U.S. Department of Commerce (USDOC). 1992. Marine recreational fishery statistics survey, Pacific Coast, 1987-1989. Current Fishery Statistics. No. 9205. Washington, D.C., 367 p.

Zimmermann, M., M. E. Wilkins, R. R. Lauth, and K. L. Weinberg. 1994. The 1992 Pacific west coast bottom trawl survey of groundfish resources: Estimates of distribution, abundance, and length composition. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-42, 110 p. plus appendices. 6

Table of Contents