Kazuhiko Hiramatsu

Developing a management procedure robust to uncertainty for southern bluefin tuna: a somewhat frustrating struggle to bridge the gap between ideals and reality

Fisheries management is conducted to achieve sustainable use of fishery resources, mainly through regulation of fishing activities. For almost a decade, the Commission for the Conservation of Southern Bluefin Tuna (CCSBT) struggled to reach agreement on a total allowable catch (TAC) for southern bluefin tuna (SBT) because of stock assessment uncertainties. To address this, in 2002 the CCSBT commenced development of a management procedure (MP), a pre-agreed set of rules to determine how the TAC will be adjusted as new monitoring data become available. The CCSBT Scientific Committee tested various candidate MPs using operating models which simulate fish population and fishery dynamics as well as incorporate process, observation, and model uncertainties. Candidate MPs were evaluated using performance measures related to the following management objectives: maximize catches, avoid stock collapse, and minimize interannual catch variation. Of the MPs explored, some relied solely on empirical data [i.e., adjusted TAC based on catch per unit effort (CPUE) trends], whereas others were more complicated, based on population models. In 2005, the CCSBT adopted a model-based MP that realized a moderate catch with low variability and avoided stock collapse. This MP struck a compromise between the risk-prone and risk-averse standpoints of the different stakeholders. However, despite this concerted scientific effort, the MP was not implemented because, shortly after its adoption, it became evident that historical catches may have been substantially underreported. This complication necessitates returning to near the beginning of the development process. MP approaches have various advantages and challenges to be explored further. However, it is essential to lessen human-introduced uncertainty (such as catch misreporting) by enhanced enforcement, and to increase management robustness to biological uncertainties by implementing MPs.

Abundance estimation of diving animals by the double-platform line transect method

In conventional line transect theory, it is assumed that all animals on the line are detected. This article introduces an extended and generalized hazard probability model without the need for such an assumption. The proposed method needs a survey design with independent observers having the same visual region and assumes an explicit distinction of simultaneous and delayed duplicates. It can take account of random heterogeneity caused by surfacing behavior as well as systematic heterogeneity by covariate effects. Furthermore, it can be easily extended to cases in which data from incompletely independent observers are available. The abundance estimate is based on the Horvitz-Thompson estimator in unequal detectability sampling scheme. Simulation studies suggest that the proposed method has good performance. The method is applied to a real data set on Antarctic minke whales in the illustration.

Impacts of the nonlinear relationship between abundance and its index in a tuned virtual population analysis

The abundance index used in a tuned virtual population analysis (VPA) is usually assumed to be proportional to actual abundance. However, the actual abundance and abundance index do not always have a linear relationship. Such nonlinearity can cause biases in abundance estimates as well as retrospective biases arising from systematic differences in abundance estimates when more data are successively added. Severe retrospective biases can damage the reliability of stock assessments. In this study, we use an approach to estimate an additional parameter that controls the nonlinearity in a tuned VPA. A performance test using simulated data revealed that the tuned VPA was able to accurately estimate the nonlinearity parameter and thus yielded less biased abundance estimates and smaller retrospective biases. We also found that estimating the nonlinearity parameters was effective even under other model misspecification scenarios, such as disregarding historical increases in catchability and time-varying natural mortality. Moreover, we applied this approach to some Japanese fish stocks and evaluated its validity. We found that estimating the nonlinearity parameters in the tuned VPA enhances the reliability of fisheries stock assessments.

Fisheries Stock Assessment

Fisheries stock assessment methods are categorized into two types: survey-based methods, such as acoustic surveys and swept area and line transect methods, and model-based methods, such as virtual population analysis (VPA), surplus production models, DeLury method, and age-structured production models. The advantages and disadvantages of each are briefly reviewed, followed by a more extensive review and discussion of VPA and several extensions of it. Tuned VPA and statistical catch-at-age analysis have been used globally for many different stocks. An example of tuned VPA and its reliability and validity is presented based on experience in Japanese fisheries stock assessment. Stock assessment methods have been developing in two directions—simple and complex models. The former relates to the recent development of management strategy evaluation. An example of the latter approach is integrated analysis. Complex models are used for data-rich stock assessments and simple assessment models for management purposes and data-poor stock assessments.

Estimation of Fish Mortality Rates with a Change-point from Tag Recoveries.

This study deals with the problem of a change-point alternative in the estimation of fishing and natural mortality rates from tag recoveries. We consider some models of mortality rates, and estimate parameters by a maximum likelihood procedure. Taking the overdispersion of the recovery data into account, we employ the likelihood function of a normal approximation of Dirichlet compound multinomial distribution. In a case study of red sea bream Pagrus major in the Seto Inland Sea, we could not detect the change-point of the natural mortality rate. It was suggested that fluctuation in the recovery data substantially influenced the estimation of the changepoint mortality rates.

A method for estimating mortality coefficients of Chum salmon Oncorhynchus keta from tagging experiments

Mature chum salmon are caught by coastal and river fisheries during their spawning migra-tion. In the tagging experiments of chum salmon in coastal waters of Japan, most early recaptures occurred in coastal fisheries, and the number of recaptures in the river fisheries gradually increased later. Therefore both fishing intensities are time dependent. In this paper, we consider a method for estimating the non-fishing mortality coefficient (M, day-1) and the time dependent mortality coefficients by coastal fisheries (F, day-1) and by river fisheries (R, day-1) from the tag recoveries using a method by Hearn et al. and the well known catch equation. We also apply this method to an example of actual recapture data. In this example, M was estimated to be 0.141, F varied from 0.000 to 0.100 and R fluctuated between 0.000 and 0.545.