From low- to high-value fisheries: Is it possible to quantify the trade-off between management cost, risk and catch?

Abstract

Fishery managers are faced with the challenge of maintaining sustainable fisheries at the lowest possible cost while conforming to international and national obligations. Given that fisheries range from low to high value, there is a real need to understand how to trade ecological and economic risks, and the various costs associated with their management, against the benefits from catch. Key to this is an understanding of (a) the costs corresponding to a given level of acceptable risk, or conversely, (b) the change in risk given a change in cost investment. This paper first defines biological, economic and ecosystem risk at a whole-of-fishery level, and then develops a simple model to quantify the trade-offs between risk, cost and catch. Using as case studies Australia’s federally managed fisheries that range from data-rich to data-poor, risk was quantified for target species in terms of both their limit and target reference points (defined as ‘‘biological risk’’ and ‘‘economic risk’’, respectively), and for ecosystems in terms of overall ecological impact (defined as ‘‘ecosystem risk’’). A statistical linear model was used to quantify the risk–cost–catch frontier for each of the three forms of risk. The most parsimonious models were statistically significant for each. However, the management and research costs were mostly positively correlated with risk, indicating that these tended to be reactive to risk, as opposed to risk decreasing in response to increased costs. The only model where this was not the case was for the ecosystem risk, which is probably because these risks have only recently been assessed and the management response to these risks across all the fisheries has so far been limited. For target species risks, it was not possible to develop a model for proactive use. However, the method itself has merit and, if the costs were defined to a greater level of resolution, and/or a time-dynamic modelling approach considered, these issues could potentially be addressed.