Abstract
One of the practical problems with quantifying the amount of energy used by fish implanted with electromyogram (EMG) radio transmitters is that the signals emitted by the transmitters provide only a relative index of activity unless they are calibrated to the swimming speed of the fish. Ideally, calibration would be conducted for each fish before it is released, but this is often not possible. Consequently, calibration curves derived from more than one fish are used to interpret EMG signals from individuals that have not been calibrated. We tested the validity of this approach by comparing EMG data within three groups of three wild juvenile white sturgeon Acipenser transmontanus implanted with the same EMG radio transmitter. We also tested an additional six fish that were implanted with separate EMG transmitters. Within each group, a single EMG radio transmitter usually did not produce similar results in different fish. Grouping EMG signals among fish produced less accurate results than having individual EMG-swim speed relationships for each fish. It is unknown whether these differences resulted from different swimming performances among individual fish or from inconsistencies in the placement or function of the EMG transmitters. In either case, our results suggest that caution should be used when applying calibration curves from one group of fish to another group of uncalibrated fish.