Response of European grayling, Thymallus thymallus, to multiple stressors in hydropeaking rivers

Rivers of the large Alpine valleys constitute iconic ecosystems that are highly threatened by multiple anthropogenic stressors. This stressor mix, however, makes it difficult to develop and refine conservation and restoration strategies. It is, therefore, urgent to acquire more detailed knowledge on the consequences and interactions of prevalent stressors on fish populations, in particular, on indicator species such as the European grayling Thymallus thymallus. Here, we conducted a multi-river, multi-stressor investigation to analyze the population status of grayling. Using explorative decision-tree approaches, we disentangled the main and interaction effects of four prevalent stressor groups: flow modification (i.e., hydropeaking), channelization, fragmentation, and water quality alteration. Moreover, using a modified variant of the bootstrapping method, pooled bootstrapping, we determined the optimal number of characteristics that adequately describe fish population status. In our dataset, hydropeaking had the strongest single effect on grayling populations. Grayling biomass at hydrological control sites was around eight times higher than at sites affected by hydropeaking. The primary parameters for predicting population status were downramping rate and peak amplitude, with critical ranges of 0.2-0.4 cm min^-1 and 10-25 cm. In hydropeaking rivers, river morphology and connectivity were the preceding subordinated parameters. Repeating the procedure with pooled bootstrapping datasets strengthened the hypothesis that the identified parameters are most relevant in predicting grayling population status. Hence, hydropeaking mitigation based on ecological thresholds is key to protect and restore already threatened grayling populations. In hydropeaking rivers, high river network connectivity and heterogenous habitat features can dampen the adverse effects of pulsed-flow releases by offering shelter and habitats for all life cycle stages of fish. The presented approach of explorative tree analysis followed by post-hoc tests of identified effects, as well as the pooled bootstrapping method, offers a simple framework for researchers and managers to analyze multi-factorial datasets and draw solid management conclusions.