Stock assessments for many U.S. Pacific coast groundfish stocks are developed using
the catch-at-age method known as Stock Synthesis. In this work a simulation package was
developed and used to evaluate the sensitivity of the Stock Synthesis program. More
specifically, the evaluation focused on the impacts of input data errors and stock
characteristics on the accuracy and precision of Synthesis estimates. Factors examined
included the length of the time series of data, the rate of natural mortality, the shape of the
fishery and survey selectivity curves, the trend in the rate of fishing mortality, the recruitment
pattern, and errors in the observed data for annual catch, fishing effort, fishery and survey age
composition, and survey biomass indices. First, the study evaluated the sensitivity of the Stock
Synthesis program applied to populations with simple multinomial age compositions. The
length of the data series and sample size were the two most influential factors. Second, the
study focused on populations with compound multinomial age composition, in which the age
composition data were over-dispersed relative to simple multinomial samples. When the
fishery age composition actually followed a compound multinomial distribution, the estimates
produced by the Stock Synthesis program, which assumed simple multinomial distributions
with maximum sample sizes of 400 fish, were moderately more biased and more variable.
When applying Synthesis to populations whose age compositions follow compound
multinomial distributions, the results from the experiments indicated that a common
configuration, in which age sample sizes in the likelihood specification are limited to 400 fish
per sample, probably gives age composition data too much emphasis. The experiments
indicated that using 200 as the upper limit provided more accurate results than using 400.
Third, the actual stock assessment of yellowfin sole (Limanda Aspera) was taken as a case
study and it was found that more accurate assessment results could be achieved from a better
balance in the amount of sampling effort allocated to age composition data versus survey
biomass estimates. / Graduation date: 2002
| Identifer | oai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/28564 |
| Date | 28 September 2001 |
| Creators | Yin, Yanshui |
| Contributors | Sampson, David B. |
| Source Sets | Oregon State University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0018 seconds