Stability of optimal harvest regime under stochastic variation in prawn demographic rates and fishing strategies

Although mathematical models are frequently used to assist in the management of fisheries, it is often assumed that a deterministic approximation to reality is sufficient to give reliable results. Many of the processes being modelled, however, can have a substantial stochastic component, and management often needs to be conservative to allow for this stochasticity. In the present investigation, it was found that adding stochasticity to the parameters of a model describing a prawn fishery had little effect on the optimal harvest regime. In both the deterministic and stochastic models the optimal harvest regime involved fishing a small proportion of the habitat very intensively. In particular, random variation in the fishing strategy had almost no effect on long-term harvest rates, although random variation in population parameters could increase the catch. Variation in natural mortality and fecundity influenced harvest rates under any particular fishing strategy more than variation in other parameters. It is likely that these results were obtained because density-dependent juvenile mortality damped out fluctuations in population size, and the targeting of age 2+ animals by the fishery meant that all animals had a chance to spawn before being exposed to harvesting.