Premium
Two‐Stage Sampling for Age Distribution in the Atlantic Menhaden Fishery, with Comments on Optimal Survey Design
Author(s) -
Chester Alexander J.,
Waters James R.
Publication year - 1985
Publication title -
north american journal of fisheries management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.587
H-Index - 72
eISSN - 1548-8675
pISSN - 0275-5947
DOI - 10.1577/1548-8659(1985)5<449:tsfadi>2.0.co;2
Subject(s) - menhaden , fishery , sampling (signal processing) , fishing , sampling design , stage (stratigraphy) , fish <actinopterygii> , stratified sampling , statistics , environmental science , computer science , mathematics , biology , fish oil , paleontology , population , demography , filter (signal processing) , sociology , computer vision
Principles of probability sampling theory and optimization of survey design have been applied to the estimation of age composition in the Atlantic menhaden (Brevoortia tyrannns) purse‐seine fishery. These methods provide a framework within which stock assessment scientists can determine confidence intervals about estimated parameters and can increase the overall efficiency of sampling programs. The Atlantic menhaden survey is discussed with respect to required assumptions and serves to illustrate several problems commonly encountered in imposing theoretical requirements on the sampling of commercial catches. A two‐stage design is developed in which purse‐seine sets and fish within sets are randomly sampled. Formulas are given for estimating mean age composition with joint 95% confidence intervals for a single week and for the entire fishing season. In addition, we recommend a procedure for determining optimal survey design which admits a class of potential solutions that has been previously overlooked when several parameters (e.g., age proportions) are measured simultaneously.