Development of an operating model for evaluation of harvest strategies for the Eastern Tuna and Billfish Fishery
At a recent meeting of the Fisheries Assessment Group (FAG) for the ET&BF (held June 1998), the rapid increase in the catches of bigeye tuna and broadbill swordfish taken in recent years were noted. It was also noted, that given the present levels of investment being made in the fishery, the increase in catches was likely to continue.
The FAG was informed of concern expressed at the meeting of the Standing Committee on Tunas and Billfish (held June 1998) over the current level of exploitation of bigeye tunas in the Pacific Ocean. Concern was also expressed over the susceptibility of broadbill swordfish to over-exploitation in light of the long-lived nature of this fish. This species is considered to be over-exploited in the Atlantic Ocean.
In light of these developments the FAG concluded that there was an urgent need to identify and evaluate appropriate performance indicators and harvest strategies to allow for the controlled and sustainable development of this fishery. In particular, Eastern Tuna MAC needs to avoid a situation where there is an unsustainable level of investment in the fishery. This need is reflected in the 1998 research priority list for Eastern Tuna MAC where the identification and evaluation of harvest strategies for the key ET&BF species using a management strategy evaluation approach was ranked sixth (ranked fifth in the 1999 priority list).
At present, the size of the catches of tunas and billfish that can be taken on a sustainable basis within the eastern AFZ remains unknown. This is due to uncertainties in and/or absence of the necessary information on which this advice can be based. Nevertheless, there is a need for Eastern Tuna MAC to identify appropriate performance indicators and management (or harvest) strategies for the continued and sustained development of this fishery.
The Management Strategy Evaluation approach, mentioned in the previous section, has been identified as the most appropriate method to adopt in achieving these outcomes. In particular, this method allows evaluation and selection of appropriate performance indicators and harvest strategies across a range of possible stock scenarios based on the comparison of a range of performance measures which would typically include the risk to the stock, rewards in the form of catches and the medium to long-term stability of these rewards. By allowing for the evaluation of harvest strategies across a range of possible stock hypotheses, the uncertainties in our knowledge concerning the biology of the target species is explicitly incorporated into the process.
Evaluation and selection of appropriate harvest strategies and performance indicators is udertaken using Monte Carlo simulation based on an operational model of the fishery. Some preliminary work on developing such a model for the ET&BF was undertaken during the study recently completed by CSIRO. The objective of this study (entitled “Evaluation of performance indicators in the Eastern Tuna and Billfish Fishery – a preliminary study”) was to illustrate the procedure used to evaluate the potential usefulness within the ET&BF of several performance indicators. These performance indicators would be used to assist in the management of this fishery. In particular, the study focused on the ability of the indicators based on catch-rates and size (length or weight) statistics associated with the catch of broadbill swordfish to correctly identify situations when the biomass drops below a certain level (40% of initial biomass was used in the study). The results demonstrated that while catch-rates can provide a very misleading impression regarding trends in abundance, performance indicators based on the length (or weight) frequency of the catch appear to be more satisfactory. However, while the relationships between these indicators and biomass are approximately linear, there is not a direct relationship between the decreases observed in the indicator variables and the underlying biomass. It is also clear that they can be quite imprecise (so that the performance indicator is frequently “triggered” too early or too late).
It is likely that “better” performance indicators could be developed (for example, based on smoothed averages or on fits of population models to the data). However, this remained beyond the scope of the preliminary study. Furthermore, refinement / further development of the operating model used in evaluation of such performance needs to be undertaken before this work can be carried out (see below).
It also needs to be noted that the main reason for calculating performance indicators is so that they can used in the management of the fishery (e.g. the effort should be reduced / stabilised if some performance indicator is triggered). It therefore seems sensible that rather than evaluating the performance of different performance indicators it is necessary to evaluate the performance of the combination of a performance indicator and the rules that specify what will happen if the performance indicator is triggered – that is, a management procedure. Work has already commenced to evaluate harvest strategies for the eastern stock of gemfish (Punt and Smith). It should be noted that results from previous evaluations of harvest strategies indicate that it is the performance of the whole harvest strategy that is important and not necessarily that of its components (such as any performance indicators). It is quite conceivable that a harvest strategy based on a performance indicator that is relatively poor will outperform one based on a performance indicator that is quite satisfactory if the rules used to deal with the case in which the performance indicator is triggered are inappropriate. This application addresses the need identified by Eastern Tuna MAC to undertaken such an evaluation of harvest strategies within the ET&BF.
The analyses undertaken in the preliminary study were based on a very simple operating model of the fishery (and in this sense are illustrative only) and considerable further work is necessary before decisions regarding the future management of the ET&BF can take place. Some of the major issues that are ignored in the current framework are discussed below.
a) The present operating model considered the dynamics of a single species only. However, there is no conceptual reason why the operating model could not be (and should not be) extended to mimic the technical interactions among the fisheries for the various species targeted in the ET&BF and the impact of changes in targeting practices.
b) The fishery was assumed to consist of a single homogeneous fleet. In reality, there are several distinct components to the fishery. Further work needs to identify each component and how it’s targeting / selectivity practices may differ.
c) The movement dynamics were assumed to be the same each year (i.e. there is no stochasticity in movement) and not to depend on density. In reality, the fraction of a stock in each area will vary considerable from one year to the next due to environmental fluctuations and movement may depend on density if some areas are more preferable to others.
d) The spatial structure of the fishery in the model was quite simple and should be considered in more detail.
e) The work needed to evaluate harvest strategies requires extensive input about models, performance measures, and harvest strategies from the stakeholders. Further work must emphasize the link between the technicians / modellers and the stakeholders. This may necessitate several workshops to specify hypotheses and harvest strategies, and to discuss the results.
The need for further development of the model along the lines just listed, together with the need to extend the work on the evaluation of performance indicators and undertake the evaluation of harvest strategies for the ET&BF is addressed by this application.
