1. The biological productivity of floodplain rivers is intimately related to their flow regimes and it has been proposed that fish production should be linked to components of the flow regime in productivity models. To assess applicability of existing models of productivity in floodplain rivers, we tested predictions about growth during the early life stages of a common, short-lived fish (Australian smelt Retropinna semoni) in a non-flow-altered, temperate Australian floodplain river.
2. The morphometric condition of larval and juvenile fish measured over a five-year period was positively related to annual discharge, but the highest average seasonal growth rates occurred in two years of contrasting hydrology, one with early spring flooding and the other with predominantly low flows and a late season (within channel) flow pulse.
3. Analysis of daily growth measures indicated that timing, river height, the duration of in-channel flow events and antecedent flood events are all significant factors influencing the early growth of Australian smelt. The flexible manner in which fish growth responds to these factors appears to be an effective early life history strategy for a short-lived species occupying a highly variable environment.
4. Growth rates conformed to some predictions of the Flood Pulse Concept (in particular the Extended Flood Pulse Concept), but specific growth responses suggest that the Riverine Productivity Model and tenets of the Low Flow Recruitment Hypothesis best describe the production of Australian smelt in this system. We suggest that none of the existing conceptual models adequately describes fish productivity in temperate Australian floodplain rivers but that aspects of each are likely to be relevant under different flow conditions.