A tool to predict the impact of anthropogenic noise on fish

Abstract

Anthropogenic (man-made) noise is a global problem in aquatic and terrestrial environments. In the shallow seas around many countries, including the UK, large windfarms are being constructed using pile driving to create a solid base for the turbines. Offshore pile driving creates pulsating noises and vibrations of very high intensities, which has been shown to be deleterious to a variety of aquatic species. Using a hydrodynamic model that predicts the propagation of underwater noise while taking into account bathymetry, tidal movements and currents, we integrated a numerical behavioural tool that models fish behaviour in response to noise. Using agent based modelling, scientifically published data and parameters obtained from carefully controlled experiments, we modelled the impact of noise on European sea bass (Dicentrarchus labrax) as they encountered pile driving during migration from the ocean to a spawning site close to the shore. Taking our empirical experiments into account - which showed a negative impact of noise on feeding behaviour and increased oxygen consumption - the model predicts that the fish took significantly longer to arrive at the spawning site. This effect could have important implications at a population level, as fish would use more energy to reach the site and might desynchronize spawning behaviour, which in turn would influence larval survival and life history processes that reduce fitness. This tool not only shows the value of using numerical models to predict animal behaviour in a complex environment, but also highlights the merit of using such models to predict anthropogenic impacts that would otherwise be difficult or too costly to obtain.