Prediction of water yield reductions following a bushfire in ash-mixed species eucalypt forest

Abstract

Previous work on long-term yield trends following the 1939 bushfire in mountain ash eucalypt catchments near Melbourne concluded that a relationship exists between the age of the ash forest and average annual streamflow yield. In the light of additional hydrologic and forest data, this study reassesses and extends the earlier work. A two-parameter model of the long-term yield trend following a bushfire is proposed. It is shown to satisfactorily fit rainfall-runoff data for eight catchments affected by the 1939 fires. In addition, the fits confirm earlier findings of significant yield reductions and, moreover, suggest possible recovery in yields. Despite insufficient hydrologic data to confirm these recovery trends, forest age and composition data indicate that recovery in yield should be practically complete by the time the ash forest reaches maturity (about 100–150 yr after regeneration). The yield trend model is consistent with this information. Using a generalized least squares approach, regional models for the long-term yield trend parameters L max (the maximum yield reduction) and log K (the response time) are developed. Unlike an earlier regional model only one forest parameter, namely the percentage of 1939 regrowth ash, was found to be significant in the prediction of L max. A simple simulation model based on the regional models is then developed enabling evaluation of the effect of future bushfire (and logging) on catchment yield. This was used to resolve contradictory conclusions about the significance of yield trends in the O'Shannassy catchment where the effect of fires prior to 1939 was shown to obscure the yield trends due to the 1939 fire. In addition, the simulation model was used to illustrate the potential vulnerability of Melbourne's water supply to a major bushfire.