ABSTRACT
This chapter explores alternative ways available for characterizing at least a proportion of the uncertainty inherent in any modelling situation. In the chapter, the authors have already optimally fitted a four-parameter model and examined ways of characterizing the uncertainty around those parameter estimates. They describe a specialized function negLLO() to deal with likelihood profiles around model outputs. The prawn catch data are relatively noisy, which is not unexpected with prawn catches. That endeavour and tiger prawn catches are correlated should also not be surprising. The endeavour prawns are generally taken as bycatch in the more valuable tiger prawn fishery, so one would expect the total tiger prawn catch to have some relationship with the total catch of endeavour prawns. The characterization of uncertainty is important because it provides some idea of how confident one can be when providing management advice.
