Search

We review recent approaches and challenges in monitoring and estimating recruitment for the main commercial tuna species managed under the auspices of the five tuna Regional Fisheries Management Organizations (tRFMOs), including: (i) the stock assessment and management context, (ii) fisheries-independent recruitment monitoring successes, failures and future options, (iii) recruitment estimation methods within statistical population models and (iv) recruitment in projections for traditional stock assessment and management strategy evaluation. Despite the diversity in tuna populations and tRFMO scientific processes, there are many common recruitment themes, including: (i) fisheries-independent recruitment monitoring is difficult and rare (the aerial surveys for juvenile southern bluefin tuna, Thunnus maccoyi, and juvenile eastern Atlantic bluefin tuna stocks, T. thynnus, are the only time series used in assessments), (ii) most statistical models estimate lognormal recruitment deviations around a stationary Beverton-Holt relationship using a penalized likelihood approach, (iii) the degree of compensation in the stock-recruitment relationship and the variance is difficult to estimate (steepness is usually fixed at 0.7 ≤ h ≤ 1.0, and the CV is most often fixed with σR = 0.6 for lognormal recruitment deviations), and (iv) estimated recruitment time series often deviate systematically from the mean relationship, but it is unclear the extent to which this reflects reality (e.g. environmental change) or estimation artefacts (e.g. incorrect model assumptions or biased data). There is a trend toward the use of model ensembles to represent recruitment (and other sources of) uncertainty, both in the context of traditional assessments and management strategy evaluation. Recommendations include: i) continued development of fisheries-independent recruitment monitoring methods, including adopting emerging technologies (e.g. genetics-based mark-recapture or abundance monitoring with a network of acoustic FADs), and (ii) the management paradigm should be designed to be robust to future recruitment uncertainties (e.g. recognize that recruitment compensation might be much lower than model-based estimates or assumptions, and be prepared to respond to substantial serially-correlated recruitment deviations, e.g. regime shifts). Following a history of over-exploitation and difficulties in reaching international scientific and management consensus, the CCSBT now provides a leading example of how tuna recruitment issues might be addressed through innovative monitoring, comprehensive modelling of uncertainty, and management with simulation-tested management procedures.

Integrated stock assessments depend on reliable catch and observational data to produce unbiased estimates of fish stock biomass and productivity. However, biases in catch data due to unreported catch are common among many of the world’s fisheries, especially those that are high value and therefore more vulnerable to illegal fishing. Some of these fisheries rely on tag-recapture data to support the estimation of biomass in integrated assessments and set precautionary catch limits intended to safeguard the fish stocks from overexploitation. Tag-recapture data is generally considered to be a more powerful indicator of stock abundance than survey data if the underlying assumptions are met. The effect of unreported catches that include tagged fish on biomass estimates in tag-based integrated assessments is unclear. We used a simulation analysis to determine the impact of under-reported catch on a hypothetical Patagonian toothfish (Dissostichus eleginoides) stock. Several catch scenarios are presented that cover different magnitudes and trend types of under-reported catch. We used the CASAL model (commonly employed to assess toothfish stocks) as both the operating model and the estimation model to evaluate how the various scenarios affected estimation performance of biomass depletion. Biomass depletion was increasingly underestimated as the magnitude of under-reporting increased, regardless of trend in catch. When unreported catch exceeds twice the amount reported combined with unreported tag returns, the estimation model will fail to detect biomass has been depleted below a target reference point of 50 % within 20 years of fishing. The lack of detection will falsely indicate current catch limits are sustainable when in reality they are not. Importantly, these estimates remained largely unchanged among different trends in under-reported catch. For assessments reliant on tag-recapture data, this suggests that it is more critical to estimate the overall quantity of unreported catch and the number of unreported tagged fish than the trend.

Indonesia is a member of three tuna Regional Fisheries Management Organisations (RFMOs), but currently has limited capacity to operationally manage its tuna fisheries to maximise benefits and minimise risks of overfishing. In 2014, the Government of Indonesia initiated discussions on the potential to develop formal harvest strategies for the management of tuna resources within its archipelagic waters. This article summarizes the development of potential empirical tuna harvest strategies for use in Indonesian archipelagic waters (Fisheries Management Areas 713, 714 and 715), using skipjack tuna (Katsuwonus pelamis) and yellowfin tuna (Thunnus albacares) as demonstration case studies. The outputs of area-specific components of Western Central Pacific Ocean stock assessment were used to condition prototype operating models for testing preliminary harvest strategies by management strategy evaluation. The case studies demonstrate the utility of scientific monitoring data to track trends in abundance of adult fish, without requiring complex stock assessment models. They also demonstrate the use of relatively simple, empirical harvest control rules to adjust the levels of fishing intensity of the Indonesian fleet to achieve specific management objectives for archipelagic waters in the context of wider regional management. The approach has a broad applicability to developing coastal nations with multi-gear fisheries and relatively limited data that harvest tuna, as well as participating in RFMOs for internationally managed stocks.

Standard Brownie mark-recapture models follow tagged cohorts of fish meaning an age-based setting is required, but there are many populations for which length-at-age cannot be characterised. This paper details the derivation of a length-based Brownie mark-recapture model, with specific application to Indian Ocean skipjack tuna. Instead of following cohorts (often tagged a number of times) the model follows the recapture history of animals tagged in a given length class and period. For Indian Ocean skipjack reporting rate uncertainty, related to mixing assumptions, is accounted for using two extreme options. Various mortality model structures were explored and length-specific natural mortality and a separable model for fishing mortality were ultimately selected. Absolute values of fishing and natural mortality depend strongly on the reporting rate assumptions, with total mortality being far more robust to assumptions about reporting rates. Estimates of fishing mortality were highest between mid 2006 and mid 2008, reducing thereafter to levels estimated in late 2005/early 2006.

Meta-analytic and multi-level stock–recruit analyses have traditionally focussed on the similar stock approach, but for a specific stock–recruit model. For six European herring stocks we embed both the stock and model levels within a fully Bayesian hierarchical framework, thus permitting the consideration of a wider class of models, specifically those that do not admit parameterisation via the steepness and unfished spawning potential. Model and parametric uncertainty is jointly characterised and the challenge of addressing model selection when the model itself is part of the hierarchy is addressed using the deviance information criterion (DIC) and posterior predictive analysis. For the six herring stocks the across-stock posterior evidence in favour of over-compensatory dynamics is fairly strong, with the Ricker and Shepherd models performing the best across the model-selection criteria. For a specific model form we perform a 20 year retrospective analysis (hierarchical and non-hierarchical) to see how temporal information flow occurs in a hierarchical framework, how this can improve our estimates of key parameters, and how this might influence management paradigms (such as Maximum Sustainable Yield) that are based on such estimates.

Short-lived fisheries stocks are subject to large fluctuations in abundance and respond rapidly to many factors including changes in oceanographic conditions, biological interactions and fishery exploitation. Management of such species requires a flexible, adaptive framework that responds rapidly to a changing environment, although such schemes are rarely operationalized. In this article, we develop a set of bioeconomic adaptive management schemes that respond to changes in economic conditions, stock abundance and catchability, using as case studies the fisheries targeting short-lived Japanese common squid (Todarodes pacificus) and Pacific saury (Cololabis saira). We suggest that such adaptive schemes have the potential to support the successful implementation of profit maximizing (MEY-based) harvest policies for borderline profitable fisheries targeting short-lived species.

The economics of enforcement is traditionally modelled using rational agents assumed to act according to their expected economic utility. In this investigation we derive this expectation by implementing a stochastic simulation model to integrate across associated levels of uncertainty using Monte Carlo. The model is developed for the South Georgia toothfish longline fishery, in which legal catches are strictly managed according to the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) conservation measures. Assuming full compliance in the legal fishery, this investigation deals with the role of a fishing patrol vessel (FPV) in excluding illegal fishing vessels. Both legal and illegal fleets are dynamic. For the legal fleet, catch is set by a harvest control rule within the model framework, with vessels leaving if fishing becomes unprofitable. The illegal catch is determined by the number of vessels operating, which predicts catch on the basis of an estimated production function. The exit of illegal vessels is determined by the probability of detection by the FPV, which is a function of days spent on patrol, and estimated profits. The entry of illegal vessels occurs at a fixed mean rate, with the model integrating over uncertainty in this parameter. An obvious trade off exists in that higher levels of illegal activity lead to lower expected returns to the legal fleet. We thus provide a cost-benefit framework for investigating the trade-offs associated with decisions on enforcement in the fishery, and confirm that current enforcement effort levels are well justified.