Forecasting animal migration using SARIMAX: an efficient means of reducing silver eel mortality caused by turbines

Endangered Species Research

E. De Oliveira¹  T. Trancart¹  A. Acou¹  E. Feunteun¹ 

关键词: Forecasting models;    Migratory species;    Mitigation;    Management;    Anguillid eels;   

DOI  :  10.3354/esr00517

学科分类：动物科学

来源: Inter-Research

【 摘 要 】

ABSTRACT: Hydroelectric power plants are often considered a major cause of mortality for migratory fish. The endangered status of European eels Anguilla anguilla forces managers to make efforts to reduce this mortality. Among the mitigation measures used, turbine shutdowns appear to be an efficient method, but they involve a substantial financial loss for hydropower producers. To optimise the use of turbine shutdowns, the present study aimed to provide a precise but simple forecast of the migration peaks of migrant silver eels. We first developed a model to forecast silver eel migration using SARIMAX models and real biological data of silver eel migration from 2 fishing sites in Brittany (north-western France). This model combines exogenous covariates and past biological data to forecast future migrations. We then evaluated this model with several years of biological data from the same trap sites, based on the decision criterion that turbines should be shut down when the number of silver eels forecasted by our model represents >10% of the mean annual number of migrants. This model would have served to save 90 and 70% of eels on the Oir and Scorff Rivers, respectively, with only 3.3 to 5.6 turbine shutdowns per year on average. Initial shutdowns would have lasted 1 wk, but this study also showed that the duration of turbine shutdowns could be reduced to 2 d without a significant decrease in efficiency. SARIMAX models reduce the influence of exogenous factors in forecasting outcomes, these being the most variable factors in the current context of climate change. This model appears to be a powerful tool for ecological forecasting of endangered species in the context of global warming.

【 授权许可】

Unknown