Example of a complex fishery model/Courtesy A.T.Charles
The fishery resources being essentially invisible, the fishery scientist constructs conceotual and mathematical models of fisheries, using them both for stock assessments and elaboration of management advice. Assuming his model is close enough to reality, the fishery scientist projects, with a variable degree of precision and confidence what might happen to a fishery under various conditions. For example, the scientist will calculate how particular levels of exploitation will effect the stock and influence catches, spawning stock size, or revenues. Various scenarios can be elaborated that guide the decisions of the fishery manager facing complex choices.
As experimentation is usually not possible, the model can also be used to assess, ex ante, the effectiveness of proposed actions to regulate the fishery, such as setting catch quotas, increasing the mesh sizes of nets, banning certain types of fishing, or closing protected areas. It can also be used to set the requirements for shore-based handling and processing facilities on a year-to-year basis. At it highest degree of complexity, it can be used to simulate the reactions of the stocks, the economy of the sector, and the behaviour of the fishers (e.g. in terms of compliance) to changes in climatic or economic drivers, in probabilistic terms.
Models used in fisheries reflect a wide range of complexity. They may represent the dynamics of one stock of a single species under average conditions of fishing. That may also include more than one species (e.g. predators and preys) and various fishing fleets (e.g. industrial and artisanal) with their complex interactions. They may also represent with significant detail:
The more complex models have higher data requirements, may be more realistic, and also more difficult to interpret.
Models designed to mimic the reaction of fish populations to fishing, generally start from the premises that the recruitment of young fish to an unexploited stock and the average growth of the individuals in the stock approximately balances the loss through natural causes such as predation. In that situation a stock varies naturally under changing environmental conditions. When fishing starts, recruitment and average growth will change, compensating to some extent loses to the population due to fishing. Nevertheless, the additional (fishing) mortality will drive the stock to new and lower levels of equilibrium (still affected by environmentsl changes) with lower biomass and younger fish on average. Accounting for natural variations from year too year, many conventional models assume that the "equilibrium" (or sustainable) catch is maximum somewhere around the level at which biomass is about half of the biomass in the unexploited stock.
Conventional population dynamics models also indicate that since fish grow very fast until they reach the age of sexual maturity, it pays to avoid catching them before this age. From these points, it follows that the pattern of fishing featured in the model must strive to maintain the mature breeding stock at a level that provides adequate recruits to it. Excessive removal of either the mature breeding stock or of fish about to be recruited to it, leads to biological overfishing of the stock.
Conventional models aim to set the level of fishing effort required to meet predetermined biological, economic or social objectives, such as Maximum Sustainable Yield (MSY), maximum net economic return or some level of employment. They usually represent based on single species fisheries. They are still be applied in many areas (e.g. in support of single species quota management strategies) but the multispecies fisheries commonly found in tropical waters require models that take a more holistic approach making far greater allowances for interactions between species and their place in aquatic ecosystems.
Multispecies and multifleet models for instance, are used when the data are available and integrate the relationships between species and fisheries. New developments in ecosystem modelling of fisheries (e.g. Ecopath, Ecosim, Ecospace) represent a new generation of models, the practical impact of which on fisheries science are still to be fully appreciated and, by lack of practical alternative also assume ecosystem "equilibrium". The last generation of Multi-Agent models (MAMs) allow, for the first time, a complete representation of a fishery, with its natural and human components, their local and global environment, with great detail, and across multiple scale, combining qualitative and quantitative information, from numerous disciplines, in complex representation that assume neither equilibrium nor reversibility.
It is important to always remember that models are human constructions constrained by human understanding and computing power and as such remain theories to be confronted to the perceptions of stakeholders (in participatory modelling) and tested and progressively improved through adaptive approaches.