Population models provide a powerful tool to forecast species responses to impacts and have been successfully used to inform the management of many threatened freshwater fishes. An important aspect of population model development is sensitivity analysis, used to refine the model and identify the components that most influence model outputs. We previously developed a metapopulation model for Golden Perch (Macquaria ambigua) in the Murray-Darling Basin and used this model to assess the Basin-scale impacts of environmental flows. Here, we report on a recent sensitivity analysis, which assessed model sensitivity to variation in upstream vs downstream movement, flow associations with spawning and larval drift, and low-flow thresholds.
Modelled Golden Perch populations in the southern Basin were most sensitive to changes in spawning and larval drift, whereas populations in the northern Basin were most sensitive to changes in movement rates and low-flow impacts. Importantly, the modelled effects of environmental flows were insensitive to changes in the underlying model rules. Our approach demonstrates an important part of model development, not often reported, where the sensitivity of the model can be used to identify knowledge gaps but also highlights areas in which model outputs and subsequent management decisions are robust to underlying uncertainty.