SS1.01 Assessing Potential Environmental Impacts of Aquaculture
Date: Thursday, June 13, 2002
Time: 2:45:00 PM
Location: Lecture Theatre
 
McFarlaneWJ, The University of British Columbia, Centre for Aquaculture and the Environment, Vancouver, Canada, wmcfarla@interchange.ubc.ca
Williams, H, , Memorial University of Newfoundland, C-CORE , St. John's, Canada, holly.williams@c-core.ca
Rowsell, D, , Memorial University of Newfoundland, C-CORE, St. John's, Canada, dean.rowsell@c-core.ca
Moccia, R, , University of Guelph, Aquaculture Centre, Guelph, Canada, moccia@uoguelph.ca
Gosine, R, , Memorial University of Newfoundland, C-CORE, St. John's, Canada, rgosine@engr.mun.ca
McKinley, R, S, The University of British Columbia, Centre for Aquaculture and the Environment, Vancouver, Canada, mckin@interchange.ubc.ca
 
Development of an automated reasoning system by monitoring activity levels in rainbow trout
image
This study exploited swimming behaviour of adult rainbow trout (0.8-1 kg) to predict the energetic needs and overall well being of fish reared under aquaculture conditions. Through the use of state-of-the-art electromyogram (EMG) transmitters implanted into lateral red muscle, we analyzed patterns of EMG activity (representing muscle activation) in fish during states of fasting and satiation. Individual activity levels allowed us to determine pattern signatures for prediction of the energetic status and needs of fish. This pattern analysis was accomplished by examining distinct features of the collected EMG data, including maximum signal, frequency analysis, shape parameters and time series, using techniques such as Fast Fourier Analysis (FFT), Wavelets and filters. Initial validation trials illustrate that EMG signals correlate with red muscle activation, and EMG procedures have no effect on swimming performance. Based on the pattern analysis, it appears that feeding state can be predicted from activity levels. Ultimately, the development of an automated reasoning system, to control automatic feeders and monitor environmental parameters from the perspective of the fish, is anticipated. This study represents the first of its kind, involving a collaboration of fisheries science and engineering, to create an automated reasoning system that allows free ranging fish to directly communicate their physiological needs.