Underlying trends confound estimates of fish population responses to river discharge.
Conservation management of freshwater ecosystems often focuses on mitigating or reversing the negative effects of altered patterns of river discharge. Assessments of management interventions frequently focus on direct, short-term responses to discharge without consideration of underlying population trends that span multiple years. We sought to determine the effects on fish populations of annual variation in river discharge and water temperature after accounting for underlying population trends. We used data on five native and one non-native fish species, collected over 7-20 years in seven rivers in the Murray-Darling Basin, south-eastern Australia. Population trends explained 3.4%-24.6% of the total variation in abundance and biomass of our six study species, while discharge and water temperature explained a further 1.2%-11.4% of this variation. However, population trends confounded the effects of discharge and water temperature, which suggests that the effects of annual discharge conditions may be contingent on past conditions and factors intrinsic to populations (e.g. age structure). Failing to account for population trends led to a combination of plausible and implausible associations with discharge and water temperature. Plausible associations included positive associations with the magnitude of spring discharge and negative associations with the number of days where discharge was below the long-term 10th percentile. Determining whether estimated associations are real or artefactual requires a greater focus on the processes that underpin multiyear population trends. Our results highlight the potential for underlying trends in populations to confound the short-term effects of discharge. Potential confounding of short- and long-term changes in populations underscores the need to assess responses to river discharge in the context of overarching environmental conditions, including factors other than discharge.