Factors related to fish kill events in Mediterranean reservoirs

Fish kills are widespread visible events perceived by the civil society as a major cause for concern about water quality and ecosystem health. Investigations conducted so far have linked fish kills to multiple factors, but extensive studies examining quantitatively the likelihood of fish kills are missing. In the present study, factors related to fish kills in 67 Mediterranean reservoirs located in the Iberian Peninsula were investigated. Moreover, the variation in the likelihood of fish kills with the foreseen consequences of anthropogenically driven climate change upon temperature and precipitation was assessed. During the 23-year period studied (1995-2017) the number of fish kills per reservoir varied from 0 to 3 and happened mostly in Southern reservoirs and during the warmest months, particularly in June. Twelve explanatory variables showed significant differences between reservoirs with and without fish die-offs. Reservoir depth, surface area oxygen and chlorophyll a concentrations were the variables that, together, best discriminated between reservoirs with and without fish kills, explaining 67.2% of the total variation in the occurrence of fish die-offs. The variables retained in the mortality model explained unique parts of that variation, but a considerable amount of the explained variation was shared by all the variables. The number of fish kill events appears to be rising during the examined period and the projection made indicated an increase in the likelihood of fish kills towards the end of the 21st century, particularly for the RCP8.5 emission scenario, with the prevalence of fish kills reaching 0.865 in Southern reservoirs. In the future, oligotrophication is likely the only strategy to consistently reduce fish kills in Mediterranean reservoirs.