Comparison of virtual population analysis and statistical kill-at-age analysis for a recreational, kill-dominated fishery

2005 ◽  
Vol 62 (2) ◽  
pp. 436-452 ◽  
Author(s):  
Paul Radomski ◽  
James R Bence ◽  
Terrance J Quinn II
Keyword(s):  
Measurement Errors ◽  
Population Analysis ◽  
Careful Consideration ◽  
Estimation Errors ◽  
Virtual Population Analysis ◽  
Virtual Population ◽  
Substantial Error ◽  
Age Structured ◽  
Forward Calculation ◽  
Over Time

We used simulations to compare the distributions of estimation errors for virtual population analysis using forward calculation (FVPA) and three variants of statistical kill-at-age analysis (KAA). The KAA variants assumed constant, time-blocked, and nonadditive selectivity. Simulations were based on a recreational walleye (Sander vitreus) fishery in Lake Mille Lacs, Minnesota. The focus of our experiments was on how model mis-specification (incorrect assumptions about selectivity for KAA or that kill had no error for FVPA) interacted with the magnitude of measurement errors and fishing mortality. We found that KAA models outperformed FVPA when they assumed the correct selectivity pattern, even when kill was measured without error. Of particular concern was a strong tendency by FVPA to overestimate stock size when kill was measured with substantial error. When KAA was based on an incorrect assumption regarding fishery selectivity and kill was measured with little error, wide distributions of errors and substantial biases sometimes resulted. KAA models that allowed fishery selectivity to change over time performed about as well as a constant-selectivity KAA model when selectivity was constant, and they performed much better when selectivity changed over time. Careful consideration of alternative fishery selectivity models should be a fundamental part of any age-structured assessment.

Variable Natural Mortality Rates Inflate Variance of Recruitments Estimated from Virtual Population Analysis (VPA)

1992 ◽  
Vol 49 (10) ◽  
pp. 2020-2027 ◽  
Author(s):  
Michael F. Lapointe ◽  
Randall M. Peterman ◽  
Brian J. Rothschild
Keyword(s):  
Mortality Rate ◽  
Environmental Factor ◽  
Statistical Power ◽  
Population Analysis ◽  
Sensitivity Analyses ◽  
Natural Mortality ◽  
Environmental Correlates ◽  
Virtual Population Analysis ◽  
Virtual Population ◽  
Over Time

We used a simulation model to determine whether estimates offish recruitment obtained from virtual population analysis (VPA) (1) have the correct interannual variability and (2) yield high statistical power (>0.8) when correlated with an environmental factor, given that the "true" instantaneous adult natural mortality rate (M) likely varies over time but a constant M is used in VPA (MVPA). Under such circumstances, VPA exaggerates variability in recruitment, which reduces the probability of correctly detecting environmental correlates with recruitment. The magnitude of these effects increases with increases in (1) the absolute value of the true mean M, (2) the variation in M over time, (3) the relative error in MVPA, and (4) the magnitude of MVPA relative to FL (instantaneous terminal fishing mortality rate) and decreases with increases in magnitude of the true variation of recruitment or the true correlation with the environmental factor (all else being equal). This bias is not large under most conditions, but it is likely to be more important in short-lived, high-M species than the similar but counteracting bias caused by aging errors. Sensitivity analyses can demonstrate how various MVPA values affect conclusions about environmental correlates with recruitment.

Virtual Population Analysis (VPA) Equations for Nonhomogeneous Populations, and a Family of Approximations Including Improvements on Pope's Cohort Analysis

1986 ◽  
Vol 43 (12) ◽  
pp. 2406-2409 ◽  
Author(s):  
Alec D. MacCall
Keyword(s):  
Mortality Rate ◽  
Population Analysis ◽  
Seasonal Pattern ◽  
Cohort Analysis ◽  
Natural Mortality ◽  
Close Approximation ◽  
Virtual Population Analysis ◽  
Virtual Population ◽  
Special Case

A set of "backward" virtual population analysis (VPA) equations relates catch (Ct) from continuous fishing between times t and t + 1 to population n size (Nt, Nt+1) when a portion of the stock is unavailable to fishing. The usual VPA equations become a special case where the entire stock is available (i.e. the stock is homogeneous). A close approximation to the VPA equations is Nt = Nt+1 exp(M) + CtM/(1 − exp(−M)), which has properties similar to Pope's "cohort analysis" and is somewhat more accurate in the case of a continuous fishery, especially if the natural mortality rate (M) is large. Much closer simple approximations are possible if the seasonal pattern of catches is known.

scholarly journals Evaluating the effect of predation mortality on forage species population dynamics in the Northeast US continental shelf ecosystem using multispecies virtual population analysis

2008 ◽  
Vol 65 (9) ◽  
pp. 1689-1700 ◽  
Author(s):  
Megan C. Tyrrell ◽  
Jason S. Link ◽  
Hassan Moustahfid ◽  
William J. Overholtz
Keyword(s):  
Population Dynamics ◽  
Continental Shelf ◽  
Population Analysis ◽  
Northwest Atlantic ◽  
Virtual Population Analysis ◽  
Virtual Population ◽  
Predation Mortality ◽  
Forage Species ◽  
Species Population ◽  
Northeast Us

AbstractTyrrell, M. C., Link, J. S., Moustahfid, H., and Overholtz, W. J. 2008. Evaluating the effect of predation mortality on forage species population dynamics in the Northeast US continental shelf ecosystem using multispecies virtual population analysis. – ICES Journal of Marine Science, 65: 1689–1700. An expanded version of multispecies virtual population analysis (MSVPA) is used to analyse the effects of predation by 14 key predators on Atlantic herring and Atlantic mackerel in the Northwest Atlantic ecosystem for the period 1982–2002. For herring, MSVPA produced greater abundance estimates than single-species assessments, especially for the youngest age classes. The average rate of predation mortality for herring aged 0 and 1 was also higher than the standard total natural mortality rate (0.2) for the 21-year time frame (0.84–3.2). The same was true for mackerel in this MSVPA (0.37–1.6). Consumptive removals of herring and mackerel generally increased over time. From 1999 to 2001, the biomass removed by predators exceeded each species' commercial landings. The sum of consumption and landings notably exceeded the multispecies maximum sustainable yield for herring for the years 1995–2002 and for mackerel for the period 1999–2002. We highlight the importance of accounting for predation on forage species in the context of changes to the fish community that have taken place in the Northwest Atlantic over the past few decades.

scholarly journals Testing the stability of the suitability coefficients from an eastern Bering Sea multispecies virtual population analysis

2005 ◽  
Vol 62 (5) ◽  
pp. 915-924 ◽  
Author(s):  
Jesús Jurado-Molina ◽  
Patricia A. Livingston ◽  
Vincent F. Gallucci
Keyword(s):  
Stomach Content ◽  
Bering Sea ◽  
Population Analysis ◽  
Data Set ◽  
Predator Species ◽  
Eastern Bering Sea ◽  
Virtual Population Analysis ◽  
Virtual Population ◽  
Total Data ◽  
The Stability

Abstract Suitability coefficients are important for the estimation of predation mortality in a multispecies virtual population analysis (MSVPA) and subsequent use in the multispecies forecasting model (MSFOR). Testing the assumption of the stability of the suitability coefficients is important in assessing the robustness of the predictions made with MSFOR. We used different statistical methods to partially test this assumption for the eastern Bering Sea MSVPA model with eight species, using stomach content data for the years 1985–1989. Comparison of the estimates from two different sets of stomach content data (set one with all data and set two mainly with data from 1985) suggested that the differences between the two types of estimates were much reduced when the number of predator stomachs sampled increased. In a second approach, we contrasted the residual variances of partial data sets with the results from the fit of the total data set. Results suggested a small increase (∼10.8%) in the variation of the suitability coefficients. Comparison of the means of the suitability coefficients associated with each predator species suggests that only 13 of the 50 possible pairwise contrasts were significantly different (α = 0.05). In general, results suggested that the predator preferences and prey vulnerabilities remained stable over the time period studied. Therefore, MSFOR could be considered as a tool to advise fisheries managers within a multispecies context.

Comparison of the performance of age-structured models with few survey indices

2018 ◽  
Vol 75 (6) ◽  
pp. 2016-2024
Author(s):  
Hiroshi Okamura ◽  
Yuuho Yamashita ◽  
Momoko Ichinokawa ◽  
Shota Nishijima
Keyword(s):  
Stock Assessment ◽  
Population Analysis ◽  
Assessment Model ◽  
Sufficient Information ◽  
Abundance Estimates ◽  
Virtual Population ◽  
Stock Biomass ◽  
Abundance Indices ◽  
Spawning Stock ◽  
Age Structured

Abstract Age-structured models have played an important role in fisheries stock assessment. Although virtual population analysis (VPA) was once the most widely used stock assessment model for when catch-at-age information is available, (hierarchical) statistical catch-at-age analysis (SCAA) is about to take that position. However, the estimation performance of different age-structured models has not been evaluated sufficiently, especially in cases where there are few available abundance indices. We examined the performance of VPA and SCAA using simulation data in which only the abundance indices of spawning stock biomass and recruitment were available. The simulation demonstrated that VPA with the ridge penalty selected by minimizing retrospective bias provided near-unbiased abundance estimates without catch-at-age error and moderately biased estimates with catch-at-age error, whereas SCAA with random-walk selectivity suffered from problems in estimating parameters and population states. Without sufficient information on abundance trends, naïvely using SCAA with many random effects should be done cautiously, and comparing results from various age-structured models via simulation tests will be informative in selecting an appropriate stock assessment model.

Sign in / Sign up

Export Citation Format

Share Document