Behavioural response of farmed Atlantic salmon (Salmo salar L.) to artificial underwater lights: Wavelet analysis of acoustic telemetry data

J.B Ulvund,S Engebretsen,J.A Alfredsen,T Kristensen,H.A Urke,P.A Jansen

doi:10.1016/j.aquaeng.2021.102196

Abstract

Biological, environmental, economic and ethical issues become increasingly pertinent as the scale of the aquaculture industry expands. This study used acoustic telemetry data and wavelet analysis to investigate behavioural patterns of Atlantic salmon when exposed to artificial underwater lights in fully stocked production cages located on the Norwegian coast. Using acoustic depth sensor tags, time series of depth registrations were gathered from 21 individual salmon distributed over three cages during a five-month experimental period. Underwater lights, normally used to suppress pre-harvest sexual maturation amongst Atlantic salmon, were installed at eight-metre depth and switched on in the middle of the experimental period. Swimming depth registrations initially showed a typical diurnal swimming behaviour, manifested by registrations generally in deeper waters during day-time than during night-time. The diurnal swimming behaviour abruptly ceased after the onset of lights. The change in swimming behaviour was detected by wavelet analysis and coincided with the introduction of underwater lights. Results from this study demonstrate the utility of wavelet analysis as a timely surveillance tool when investigating behavioural patterns of a periodic nature in fish, and specifically the individual response of farmed salmon to artificial lighting in a genuine industrial setting.

Highlights

Observing fish and their movements in the sea cages forms a vital part of daily salmon farming husbandry
783,436 depth registrations were collected from 1st November 2016 to 2nd February 2017 (Cage A: 264,440, Cage B: 284,395 and Cage C: 234,601) with total detections per individual fish in the range 33,074–41,458
After aggre gating to an hourly resolution, we have a total of 46,977 hourly depth registrations, out of which 44 were interpolated to substitute missing values occurring due to signal-collision or during battery changes and other maintenance operations

Summary

Introduction

Observing fish and their movements in the sea cages forms a vital part of daily salmon farming husbandry. This allows us to detect deviations from normal behaviour, which can be used as an early warning of potentially adverse conditions in the fish farm, allowing time to initiate closer investigations and implement mitigating measures (Oppedal et al, 2011; Hvas et al., 2020) Such changes can be challenging to detect in full-scale production cages given the size of the sea cages and the number of fish in modern-day aquaculture. In Atlantic salmon (Salmo salar L.) farming, typical tasks include manual sightings and sampling of fish for health and welfare assessment through monitoring of surface activity and manually counting sea lice. These surveillance tools are based on direct visual inspection, relying on in situ observation of the fish. Artificial lights have been shown to enhance appetite and growth rate (Endal et al, 2000) and are adopted as a common farming practice throughout the industry

Objectives

Methods

Results

Discussion

Conclusion