SOME STATISTICAL CONSIDERATIONS OF THE DESIGN OF 

 TRAWL SURVEYS FOR ROCKFISH (SCORPAENIDAE) 



William H. Lenarz and Peter B. Adams* 



ABSTRACT 



This study is in two parts. The first part reviews statistical theory for choosing among random, 

 stratified random, and systematic sample survey schemes when strata are of equal size and receive 

 equal sampling effort. The theory is applied to data collected during a pilot trawl survey for rockfish in 

 Queen Charlotte Sound, British Columbia, and a full scale survey along the coasts of Washington, 

 Oregon, and California. The results indicate that on a scale of about 80 km, a systematic survey scheme 

 provides more precise estimates than the other schemes. However, the differences in precision are 

 slight and probably should not outweigh other factors such as logistical constraints in the design of 

 trawl surveys. The second part of the study reviews statistical theory for sampling from negative 

 binomial distributions. Results of the Queen Charlotte Sound pilot survey indicate that except for fish 

 with very low densities, numerous tows of short distances are relatively more precise than fewer tows of 

 longer distances for trawl surveys for rockfish. 



The Fisheries Conservation and Management Act 

 of 1976 requires the development of fishery man- 

 agement plans for each marine fishery under 

 jurisdiction of the United States. The requirement 

 emphasizes the need to assess the status of U.S. 

 fisheries. Estimates of stock abundance are essen- 

 tial for fishery assessment, and trawl surveys 

 often are used to estimate absolute or relative 

 stock abundance where suitable data from a 

 fishery itself are lacking. 



Very little data are available from the com- 

 plex fisheries for rockfish (genus Sebastes) of the 

 Pacific coast of North America. The fisheries 

 are complex because many species and types of 

 gear are involved. Often landing statistics do 

 not specify species, and catches are seldom suf- 

 ficiently sampled for age, length, and sex com- 

 position. Catch per effort data also are not 

 reported by species and are difficult to interpret 

 because of temporal changes in fishing power and 

 target species. 



Because of this lack of needed data, the North- 

 west and Alaska Fisheries Center of the National 

 Marine Fisheries Service initiated a large scale 

 trawl survey of rockfish stocks from southern 

 California to the Aleutian Islands. The first stage 

 of the 4-yr survey was to conduct pilot surveys in 

 the Monterey Bay area, Calif., and Queen Char- 



Southwest Fisheries Center Tiburon Laboratory, National 

 Marine Fisheries Service, NOAA, 3150 Paradise Drive, Tiburon, 

 CA 94920. 



lotte Sound, British Columbia, during 1976. The 

 overall goal of the pilot surveys was to provide 

 information for design of the full scale survey. 

 Gunderson and Nelson describe the pilot survey 

 and present preliminary results of the effort. A 

 full scale survey was conducted in 1977 off the 

 coasts of Washington, Oregon, and California. 

 Design of the 1977 survey was partly based on 

 results of the pilot surveys. Gunderson and Sam- 

 ple (1980) discussed the 1977 survey. 



While trawl surveys have proven to be a useful 

 tool for assessing fish stocks, problems still remain 

 in their design and analysis. This paper presents 

 analyses of results of the Queen Charlotte Sound 

 pilot survey and 1977 survey. The analyses were 

 aimed at answering three questions: 1) Should the 

 full scale survey design be based on a random, 

 stratified random, or systematic scheme? 2) Do 

 results of the 1977 survey indicate that aspects of 

 the design based on the pilot survey were correct? 

 3) What are the trade offs in precision between 

 distance trawled and number of tows? Ancillary to 

 question one are the questions: 1) Are there 

 significant benefits to be gained by choosing one 

 or a combination of the three sampling schemes? 

 2) Are there significant biases in estimates of 



Manuscript accepted December 1979. 

 FISHERY BULLETIN: VOL. 78, NO. 3, 1980. 



^Gunderson, D. R., and M. O. Nelson. 1977. Preliminary 

 report on an experimental rockfish survey conducted off 

 Monterey, California and in Queen Charlotte Sound, British 

 Columbia during August-September, 1976. Prepared for Feb- 

 ruary 15-16, 1977 Interagency Rockfish Survey Coordinating 

 Committee Meeting, Northwest and Alaska Fisheries Center, 

 Seattle, Wash., 82 p. 



659 



