Combining acoustic and trawl data for estimating fish abundance

In this study, different scenarios of uncertainties in a multispecies fishery are presented. Uncertainties have been considered in the estimation of the species composition of the tropical tuna catch as well as in the total catch. Although three species (yellowfin, skipjack and bigeye) are involved in the tropical tuna fishery, only uncertainties in the bigeye vs. skipjack have been taken into account, in order to simplify the analysis. Regarding the total catch, bigeye has been considered as the species with higher uncertainty due to the IUU catch component. Analyses have been focused in evaluating the sensitivity of the current assessment methods to these uncertainties. Taking into account that the two sources of uncertainty affect the age structure of the catch as well as the total catch, surplus production methods have been applied in the assessment.

Operational management procedures (OMPs) provide an alternative to the conventional stock assessment treadmill, where increasingly complex models are run each year. These models may, on occasion, be modified significantly by scientists and produce estimates of stock status that, depending on circumstances, may influence the management action. In contrast, OMPs allow for simpler assessment/management models to be developed, simulation-tested and validated against more complex models used to describe alternative plausible states of nature. A large variety of candidate management procedures may be tested, but once one is agreed upon by the stakeholders of the management system, standard practice is for it to be implemented for several years with periodic evaluations to see that it is working as intended. OMPs represent a powerful tool that is firmly rooted in the existing knowledge base, and recognises the uncertainties inherent in this knowledge base by developing management strategies that are robust to these uncertainties. They also allow a systematic analysis of the knowledge and control requirements of alternative management tools.

Spectacular long-term fluctuations (of the order of 110 years) have been seen in catches of Atlantic bluefin tuna (Thunnus thynnus) independent of human exploitation. The actual causes of such changes could have been generated by contrasting but equally plausible dynamic processes, such as changes in carrying capacity or migration. A simulation framework was therefore used to evaluate the consequences for the perception of the stock obtained via virtual population analysis (VPA) as used by ICCAT. It was shown that the performances of VPA depend on the underlying processes that generated the long-term fluctuations in catch and on the current phase of the dynamics, both of which will be extremely difficult if not impossible to determine from current stock assessment methods. Further, while interpretation of variation in catch resulting from changes in the carrying capacity can be estimated correctly by VPA, it is more difficult to estimate trends and absolute values of fishing mortality and spawning stock biomass when variations in catch result from changes in migration. There was also considerable confounding between the dynamics and increasing effort, as currently seen, that makes it difficult to infer the actual dynamics on the basis of commercial catch data. The main outcome of the study is that knowledge of the underlying process that generates long-term variations in catch is of key importance, because distinct processes lead to contrasting population dynamics and to different perceptions of stock status.

Fromentin and Kell (eTIP result: ) showed that identifying the underlying process that generates long-term variations in bluefin tuna catch is of key importance, because distinct processes lead to different perceptions of stock status. In this study, the performances of stock assessment methods with respect to: - Their ability to provide good estimates of MSY, FMSY and BMSY, and - Assessing stock status and exploitation level relative to these MSY targets were evaluated. In addition, the robustness of the current ICCAT management strategy (based upon MSY) to uncertainty about the true population dynamics and historical exploitation levels was evaluated and contrasted with a simpler management strategy. Reference points appear to be more robust to dynamic uncertainty than the estimates of absolute values and trends in F and SSB, but they perform better when fluctuations in catch result from changes in carrying capacity than in migration. F-based reference points were less biased and more precise than biomass- and/or yield-based reference points and F0.1 appeared to be the best proxy for FMSY. However, F0.1 cannot indicate past and current levels of exploitation relative to FMSY when there is uncertainty about the dynamics. The F0.1 management strategy based on ADAPT-VPA stock size estimates finally only performed slightly better than a strategy based on a simple size limit in term of SSB. However, the former leads to much lower catch levels than the latter. We conclude that the performances and robustness of distinct management strategies strongly depend on biological processes (i.e. the underlying dynamics) and when they are implemented relative to the stock size cycle, but also on concrete objectives, such as fleet composition, gear selectivity, economic constraints. Such an investigation can only be resolved through simulation studies.

Operational management procedures (OMPs) provide an alternative to the conventional stock assessment treadmill, where increasingly complex models are run each year. These models may, on occasion, be modified significantly by scientists and produce estimates of stock status that, depending on circumstances, may influence the management action. In contrast, OMPs allow for simpler assessment/management models to be developed, simulation-tested and validated against more complex models used to describe alternative plausible states of nature. A large variety of candidate management procedures may be tested, but once one is agreed upon by the stakeholders of the management system, standard practice is for it to be implemented for several years with periodic evaluations to see that it is working as intended. OMPs represent a powerful tool that is firmly rooted in the existing knowledge base, and recognises the uncertainties inherent in this knowledge base by developing management strategies that are robust to these uncertainties. They also allow a systematic analysis of the knowledge and control requirements of alternative management tools.

Atlantic Bigeye is a highly migratory species widely distributed through the whole Atlantic. Recent assessments conducted by the ICCAT SCRS are highly uncertain due to the large uncertainties in biological parameters, such as M, as well as in fishery statistics (e.g. IUU fleets, juvenile catch) or abundance indexes. Similar uncertainties can be found in the Indian or Pacific stocks. The development of an operational model has been recommended by different regional fishery organizations as a good approach to investigate the sensitivity of the assessment results to the uncertainties in the different inputs. This document presents an operational model for the Atlantic Bigeye as part of a simulation framework that will allow the performance of management strategies based on current and improved biological knowledge and management procedures to be evaluated.

Plaice (Pleuronectes platessa) and sole (Solea solea) are the primary targets of a mixed flatfish beam-trawl fishery in the North Sea. The fishery is currently managed through single-species Total Allowable Catches and trawl mesh size regulations, with fleets fishing in the northern North Sea required to use a larger mesh size than in the south. We modelled interactions between these fish stocks and fishing fleets, taking into account the impact of changes in fish distribution, stock productivity, and resilience on fishery sustainability and profitability. Over time, the plaice population appears to have shifted northwards and the productivity of both plaice and sole has varied. Biological changes had a considerable impact on the limit management reference points currently used by ICES. Biomass reference points shifted with changes in productivity, whilst fishing mortality reference points remained constant (although the resulting impact on the stock changed). Single-species reference points currently used are often conflicting and encourage high-grading and discarding of smaller fish. As fishing activity is largely driven by economic considerations (in the light of management controls), alternative economic management targets were considered. In order to achieve maximum economic yield from the fishery, a considerable decrease in fishing effort of the southern fleet is required, owing to the smaller mesh size and increased vulnerability of young plaice before they move north with age. Sustaining maximum employment within the fishery could, in contrast, be achieved when fishing effort was close to the break-even level. The analysis suggested that at 2002 effort levels, at equilibrium fleets would break even in only one scenario of productivity and resilience. Common alternative biological target reference levels (e.g. maximum sustainable yield and its proxy F0.1) were also examined. They resulted in considerably different effort levels in the north and south, and different profit levels. While we examined a range of possible management objectives in a multifleet, multispecies fishery, selection of the most appropriate depends on priorities assigned by managers and users. Distribution of resulting effort between fleets in the fishery must then be considered, which as shown in this study may not conform to the European aim of relative stability. The choice of alternative reference points should be evaluated through simulation, considering the system in place and the uncertainties inherent within it.

FLR is a generic software framework for fisheries modelling and the evaluation of management strategies. The framework allows the creation of simulation models, which can be conditioned on a range of stock and fleet assumptions, through which current and alternative management strategies can be evaluated. The framework is able to explicitly include a variety of process and model uncertainty with respect to knowledge of the dynamics of fisheries systems, their response to management and our ability to monitor, assess and control them. This enables the development of advice that is robust to these sources of uncertainty. Currently the framework is being used to develop bio-economic models, multi-annual management plans and fishery independent assessment methods within a variety of EU Projects. FLR has been developed using R, a free, open source, multi-platform, software environment for statistical computing and graphics. R provides many useful software tools and can be extended to include packages (i.e. libraries) which are of interest to particular groups. FLR is open source and all users have access to the code. This allows many developers from different projects, bodies or institutes with a variety of objectives to collaborate on coding. This helps in transparency, enables users to learn from each other and can speed up the process of validation and bug fixing. The development of additional FLR packages by users is strongly encouraged by the FLR development team. The simultaneous collaboration of multiple developers on the same files within an open source environment is achieved using Concurrent Versions System (CVS). All changes are stored on a central server, the CVS repository, from which previous or current versions can be extracted. The FLR CVS project is currently located at: http://sourceforge.net/cvs/?group_id=108356, where all modules are currently available and anyone can browse through the CVS tree. Details on how to checkout packages from the CVS repository can be found at: http://www.flr-project.org/doku.php?id=doc:devmanual.

Past studies have indicated that numbers-at-age 1 of albacore may be correlated with the winter NAO index in the previous year. Two main hypotheses have been stressed to explain such observation, i.e. an impact of the NAO on the recruitment or on migration patterns and consequently in the availability and catchability of specific age classes to the surface fishery. The objective of the present paper is to test whether such processes can be distinguished using VPA outputs. We first computed simple cross-correlations between the NAO at various lags, catches-at-age, Numbers-at-age and F-at-age estimated in the last assessment. Since results were inconclusive, simulations were undertaken. The results showed that the probability of detecting spurious correlations between a given environmental variable, such as the NAO, and VPA outputs appears to be strongly related to the level of errors due to the observation process and modelling assumptions made within the assessment procedure. We concluded that correlations based solely on VPA outputs should be interpreted with great care.

Robustness of both short-term stock biomass recovery and longer-term sustainable management strategies to different plausible climatic change scenarios were evaluated for North Sea cod (Gadus morhua), where climate was assumed to impact growth and recruitment. In the short term, climate change had little effect on stock recovery, which depends instead upon reducing fishing effort to allow existing year classes to survive to maturity. In the longer term, climate change has greater effects on stock status, but higher yields and biomass can be expected if fishing mortality is reduced. Incorporating environmental covariates in stock assessment predictions will not achieve sustainable resource use. The implications of climate change for biological reference points depend upon the mechanism through which temperature acts on recruitment, i.e. on juvenile survival or carrying capacity. It is not possible to distinguish between these processes with stock assessment data sets alone. However, this study indicates that reference points based on fishing mortality appear more robust to uncertainty than those based on biomass. Ideally, simpler management procedures are required that meet pre-agreed management objectives and are robust to uncertainty about the true dynamics.

Deliverables

Share this page

Download