Dr. André Punt¹
1US SAFS
October 13, 2015 9:00 (PST): FSH 203

Some Insights into Data Weighting in Integrated Stock Assessments

The results of fishery stock assessments based on the integrated analysis paradigm can be sensitive to the values for the factors used to weight each of the data types included in the objective function minimized to obtain the estimates of parameters. These assessments generally include index data, length-composition information and conditional age-at-length data, and algorithms have been developed to select weighting factors for each of these data types. This paper introduces methods for weighting conditional age-at-length data that extend an approach developed by Francis (2011) to weight age- and length-composition data. Simulation based on single-zone and two-zone operating models are used to compare five tuning algorithms that are constructed as combinations of methods to weight each of these data types. The single-zone operating models allow evaluation of the algorithms in terms of the ability to provide unbiased estimates of management-related quantities and the correct data weights in the absence of model mis-specification while the two-zone operating models allow the impacts of model mis-specification on the performance of tuning algorithms to be explored.

Presentation

Dr. Elizabeth Councill^1,2
1NWFSC
²US SAFS
September 29, 2015 9:00 (PST): FSH 203

Autocorrelated recrutiment deviations

Abstract: To be posted …

Dr. Kotaro Ono^1
1UW SAFS

June 02, 2015 9:00 (PST): FSH 203

Harvest strategies for multispecies fisheries under catch constraints

Current groundfish species in the Bering Sea, Alaska, are managed under strict annual quotas. However, there are strong bycatch regulations in place that compromise the ability of the fleet to catch the entire quota. As a consequence, single species population forecasting might not be very accurate if the multispecies nature of the fishery is ignored. In addition, uncertainty in environmental fluctuation and fishermen behavior can affect the accuracy of population forecasting and bias the effect of management actions. Management strategy evaluation (MSE) is a computer simulation method that consists in modelling both the fish population and the management cycle (stock assessment and decision making) of a fishery to evaluate the performance of the management actions. This procedure has been shown to be very important as it can caution the user about the potential undesired effect of their proposed management actions (e.g. unwanted reduction in quotas). In this study, a MSE was developed to test 1) how the inclusion of a multispecies catch tradeoff affect the short- to long-term performance of the model (e.g. catches, discounted revenue) compared to traditional single species model, and 2) how these models perform in the presence of environmental and fishermen behavioral uncertainty.

Ian Stewart^1
1International Pacific Halibut Commission

May 05, 2015 9:00 (PST): FSH 203

It is standard practice to conduct fisheries stock assessments generating a point estimate of stock size, which is then translated into one or more future fishery catch targets and limits. The general approach often relies on the output of a single stock assessment model, and therefore does not thoroughly account for alternative hypotheses or major sources of uncertainty such as fixed parameter values, model approach (e.g., statistical catch-at-age, VPA), model structure (e.g., treatment of spatial dynamics, delineation of fishing fleets, time-series length) and data weighting. In 2012, the International Pacific Halibut Commission made the transition from point estimates to risk-assessment, based on a decision table produced for the annual management process. The decision table represents a composite of probability-weighted results from alternative models, allowing a comparison of potential benefits (fishery yields) with the probabilities of various risk metrics, including stock and fishery trends and status. In 2013, the approach was extended to include the use of an ‘ensemble’ of models, following methods used in climate and hurricane forecasting. These changes have led to increased transparency about perceived risk, the availability of more information for the decision makers, and a clear delineation between scientific and policy considerations. The ensemble approach to stock assessment also provides a conceptual link between single model analyses and fully developed Management Strategy Evaluation of harvest policy and management procedures. Potential extensions and improvements to the approach, as well as a brief summary of decisions based on this information will also be discussed.

Dave Fournier¹, James Ianelli^1,2 and Steve Martell^1,3
1ADMB Foundation; ²Alaska Fisheries Science Center; ³International Pacific Halibut Commission

April 21, 2015 9:00 (PST): FSH 203

A proposal for rehabilitating the multinomial likelihood for compositional data

The multinomial likelihood is widely used in many stock assessment models for evaluating compositional data. However, this simple model does not account for over-dispersion and does not allow for correlation among adjacent categories. In response to recent publications proposing alternatives for the multinomial, we propose to extend and modify the ideas put forward by Hranfnkelsson and Stefánsson (2004) to better accommodate over-dispersion and correlation in compositional data. Specifically, to develop a self-scaling multinomial type estimation procedure and optionally incorporate positively autocorrelated errors. Models that incorporate composition data need to take into account over-dispersion and correlation if uncertainty in model parameters, and the risks associated with management actions, is desired.