Record Details
Effects of sampling design, estimators, and variability on groundfish management in Oregon
ScholarsArchive at Oregon State University
| Field | Value |
|---|---|
| Title | Effects of sampling design, estimators, and variability on groundfish management in Oregon |
| Names |
Crone, Paul R. Jr
(creator) Sampson, David B. (advisor) |
| Date Issued | 1994-12-02 (iso8601) |
| Note | Graduation date: 1995 |
| Abstract | Sampling error associated with estimates of species composition and age composition of commercial groundfish landings in Oregon from 1989 to 1991 is documented to evaluate the impact of variable landing data on fisheries management and monitoring programs. The statistical reliability and bias associated with two multistage sampling designs are investigated on the bases of the practical suitability of field procedures and accuracy and precision of the derived estimates. Additionally, the variability associated with landing estimates of age composition for five groundfish species is used to examine critically the ability of age-structured stock assessment models to describe adequately the stochastic properties of actual catch-at-age data. Finally, the spatial similarity of reproductive parameters derived from sampling data for Dover sole inhabiting marine waters off the coast of Oregon is examined in statistical and practical terms to provide management additional analyses that can be used to adopt harvest strategies that minimize detrimental effects to the exploited fish stock(s). A two-stage random sampling design combined with stratification components generated relatively reliable landing statistics. At least two-thirds of the total landings of rockfish in each port/quarter stratum consisted of estimates of species composition that had small coefficients of variation (CVs <10%). For each species sampled for age composition, at least three-fourths of the total landings included estimates for individual ages with CVs <25%. For the majority of the landings, the variation at the first stage of sampling contributed at least 63% and 90% to the variance of the landing estimates for the species and age compositions, respectively. An abbreviated version of a multistage sampling design, which incorporated a single sampling unit at the second stage, produced generally similar results to those generated by the 'complete' two-stage approach discussed above. However, results indicated that there was likely a substantial risk in using this design for species-composition sampling, given the biased variance measures associated with its application. For all practical purposes, the bias was not as problematic in age-composition sampling; however, caution and additional monitoring procedures are recommended to ensure that inferences reflect the actual statistical properties of the sampling design. An analysis of the variability in catch-at-age data clearly indicated that a multinomial probability error structure, included in models that are based on maximum likelihood estimation, more closely follows the estimated variances associated with the sampled landing data than does a lognormal error structure used in models based on least squares estimation. An analysis of maturity data from commercial landings provided some statistical evidence that spatial differences may exist in sexual maturity schedules of Dover sole inhabiting Oregon waters; however, it does not appear that the statistical findings from the maturity assessments are of magnitudes that reflect dramatic implications for management. |
| Genre | Thesis/Dissertation |
| Topic | Pleuronectidae -- Counting -- Oregon |
| Identifier | http://hdl.handle.net/1957/34868 |
