Stock Identification with the Maximum-Likelihood Mixture Model: Sensitivity Analysis and Application to Complex Problems

Wood, Chris; McKinnell, Skip; Mulligan, T. J.; Fournier, David

doi:10.1139/f87-105

Cited by 83 publications

(77 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Developing such a method is the main purpose of this paper. GSI is most accurate when there is a high degree of genetic differentiation between populations, when the baseline samples for each population are large, and when large numbers of loci are genotyped (Wood et al 1987;Kalinowski 2004), but there is no simple formula available to calculate how accurate GSI is expected to be; therefore computer simulation must be used. However, as we explain below, the simulation methods currently in use to do this are flawed in a way that leads them to consistently overestimate the expected accuracy of GSI.…”

Section: Introductionmentioning

confidence: 99%

An improved method for predicting the accuracy of genetic stock identification

Anderson

Waples

Kalinowski

2008

Can. J. Fish. Aquat. Sci.

218

277

View full text Add to dashboard Cite

Estimating the accuracy of genetic stock identification (GSI) that can be expected given a previously collected baseline requires simulation. The conventional method involves repeatedly simulating mixtures by resampling from the baseline, simulating new baselines by resampling from the baseline, and analyzing the simulated mixtures with the simulated baselines. We show that this overestimates the predicted accuracy of GSI. The bias is profound for closely related populations and increases as more genetic data (loci and (or) alleles) are added to the analysis. We develop a new method based on leave-one-out cross validation and show that it yields essentially unbiased estimates of GSI accuracy. Applying both our method and the conventional method to a coastwide baseline of 166 Chinook salmon (Oncorhynchus tshawytscha) populations shows that the conventional method provides severely biased predictions of accuracy for some individual populations. The bias for reporting units (aggregations of closely related populations) is moderate, but still present.Résumé : L'estimation de la précision de l'identification du stock génétique (« GSI ») qu'on peut espérer, étant donné une banque de données de base récoltée antérieurement, nécessite des simulations. La méthode courante comprend des simulations répétées de mélanges par ré-échantillonnage de la banque de données de base, des simulations de nouvelles banques de données de base en ré-échantillonnant la banque de données et l'analyse des mélanges ainsi simulés à l'aide des banques de données de base simulées. Nous montrons que cette méthode surestime la précision prédite de GSI. L'erreur est importante dans les populations fortement apparentées et elle augmente à mesure que de nouvelles données génétiques (locus et (ou) allèles) sont ajoutées à l'analyse. Nous mettons au point une nouvelle méthode basée sur une validation croisée de type « leave-one-out » (avec retrait d'un élément) et nous montrons qu'elle produit essentiellement des estimations non erronées de la précision de GSI. L'application de notre méthode et de la méthode courante à une banque de données de base provenant de 166 populations de saumons chinook (Oncorhynchus tshawytscha) réparties sur toute la côte montre que la méthode courante fournit des prédictions de la précision qui sont grandement faussées pour certaines populations individuelles. L'erreur dans le cas des unités d'évaluation (des rassemblements de populations fortement apparentées) est peu importante, mais réelle.[Traduit par la Rédaction]

show abstract

Section: Introductionmentioning

confidence: 99%

An improved method for predicting the accuracy of genetic stock identification

Anderson

Waples

Kalinowski

2008

Can. J. Fish. Aquat. Sci.

218

277

View full text Add to dashboard Cite

show abstract

“…3B), and not at high spawner abundances. This curvature at the lowest values of E could be the result of: (1) overestimates of R at low R due to sampling limitations and errors in CU identifi cation (Millar 1987;Wood et al 1987;Mulligan et al 1988;Brodsiak et al 1992;Gable and Cox-Rogers 1993;Waldman and Fabrizio 1994;Ricker 1997;Fabrizio 2005); (2) underestimates of E at low E, possibly associated with inadequate visual counts when spawners are uncommon, the result of focusing spawner enumeration eff ort on large abundant CUs (Grant et al 2011);and/or (3) a compensatory increase in productivity per spawner at extremely low spawner abundance. Given the limitations of compositional analysis faced with small proportions (< 1%), we do not believe this is a compensatory eff ect.…”

Section: Discussionmentioning

confidence: 99%

Stock-Recruit Analyses of Fraser River Sockeye Salmon

Akenhead¹,

Irvine²,

Hyatt³

et al. 2016

NPAFC Bull.

View full text Add to dashboard Cite

Three types of stock-recruit models (log-log, Ricker, Beverton-Holt) were applied to 57 years of adult returns (R) and eff ective female spawners (E) data from 17 biologically-based Conservation Units (CUs) of sockeye salmon from the Fraser River in British Columbia, Canada (hereafter "Fraser sockeye"). Log-log regressions of R on E showed little evidence of eff ects of density (within CUs) on survival, implying that habitat capacity does not presently limit Fraser sockeye abundance. Shared survival among CUs, the fi rst principal component of log(R/E), accounted for 46% of the variance, remarkable given the wide variety of freshwater and marine habitats experienced by the CUs. Six low-survival events in six decades accounted for much of the shared survival pattern. The r 2 values for Ricker models were low, indicating that attempts to manage and/or assess Fraser sockeye using Ricker curves fi t to individual CUs will, in general, face low explanatory power. A suite of increasingly complicated Bayesian regressions, based on the Beverton-Holt model, quantifi ed the precision of capacity estimates, but these were always imprecise and the Widely Applicable Information Criterion (WAIC) indicated overfi tting in all but the simplest models. A variance factor for the relative precision of estimates of R, based on the proportion of spawners from each CU in groups of co-migrating CUs (i.e., runs), was eff ective only in models in which WAIC indicated over-fi tting. Improving the precision of capacity estimates for Fraser sockeye salmon, using similar models, will require mobilizing biological knowledge (i.e., historical metadata) about each CU, including estimates of abundance with identifi ed precision and indicators of habitat capacity at multiple life-history stages including the abundance of potential competitors in marine habitats. Researchers analyzing stock recruitment data for salmon populations and other species are strongly encouraged to pay attention to factors aff ecting the various categories of data in their models, i.e., to examine the associated metadata.

show abstract

“…We used the maximum likelihood framework implemented in the program gsi_sim (Anderson et al, 2008;Anderson, 2010) to assign individuals to a stock. Each fish was assigned to the stock in which the probability of its genotype occurring was greatest by using the allocatesum procedure (Wood et al, 1987). We did not attempt to identify out-of-basin strays.…”

Section: Data Collectionmentioning

confidence: 99%

Abundance estimates and confidence intervals for the run composition of returning salmonids

Steinhorst¹,

Copeland²,

Ackerman³

et al. 2016

View full text Add to dashboard Cite

Abstract-In 2-stage fishery sampling, abundance is often estimated by using a primary sampling gear and total abundance is then partitioned into groups of interest by applying data on composition derived from a secondary sampling gear. However, the literature is sparse on statistical properties of estimates of run composition. We examined the accuracy and precision of estimators of composition of wild steelhead (Oncorhynchus mykiss) in the Snake River, in the Pacific Northwest. We simulated estimators, using pooled and time-stratified data. We compared confidence intervals (CIs) determined on the basis of asymptotical normality or a 2-stage bootstrap method. Stratified estimators were unbiased, except in a few cases. Joint CIs (all groups considered simultaneously) had coverages near nominal. Conversely, pooled estimators performed poorly; the proportion of biased estimates increased as the number of groups estimated increased. Using empirical data, we show that CIs met precision goals for most groups. Half-widths of CIs decreased and stabilized as the number sampled and group abundance increased. In complex scenarios, estimates of small groups will yield poor precision and some may be biased, but a stratified estimate with a conservative joint CI can be of practical use. The 2-step bootstrap approach is flexible and can incorporate other sources of variability or sampling constraints.

show abstract

Stock Identification with the Maximum-Likelihood Mixture Model: Sensitivity Analysis and Application to Complex Problems

Cited by 83 publications

References 4 publications

An improved method for predicting the accuracy of genetic stock identification

An improved method for predicting the accuracy of genetic stock identification

Stock-Recruit Analyses of Fraser River Sockeye Salmon

Abundance estimates and confidence intervals for the run composition of returning salmonids

Contact Info

Product

Resources

About