Abstract

Bycatch of harbour porpoises (Phocoena phocoena) occurs in most gillnet fisheries throughout the continental shelf area in the Northern
Hemisphere. A few behavioural studies have attempted to ascertain the detection of a barrier consisting of passive echo enhancing
structures. Past experiments with active sound added to fishing nets often did not achieve statistically significant bycatch reduction due
to small sample sizes (reflecting a low catch per unit effort of nets).
Using inoffensive experimental ‘nets’, behavioural data obtained by theodolite tracking can yield larger sample sizes. Three different
device types mounted on a floatline were tested for their potential rule in reducing bycatch. For this purpose behavioural responses
(avoidance and closest approach) of porpoises to these devices were compared to responses to a control which consisted of a floatline
only. The devices were two different passive reflectors (target strength ranging from -38dB to -24dB) and a 2.9kHz pinger with a source
level of 115dB (re 1 μPa at 1 m).
A total of 335 porpoise groups (distributed in almost equal proportions for all treatments) were recorded. 92.4% of porpoise groups
avoided the pinger equipped floatline whereas only about half of the groups avoided the other stimuli (reflectors: 48.6% and 58.9%.
floatline only: 51.8%). The difference between pingers and all the other stimuli was significant. Closest observed approach distances
were 34m (SE = 3.57) for the floatline only. 33m (SE= 3.65) and 30m (SE = 3.38) for the reflectors and 133m (SE = 6.60) for the
pingers, the difference between pingers and all the other stimuli being significant.
Schooling was observed in 36% of the groups with more than two individuals near the pingers. Since porpoise density in
the area did not decrease after permanent use of pingers for six days there is no indication for long-term displacement from the
area. The porpoises may have habituated to the sound of the pingers since frequency of avoidance and minimum distance to pingers
decreased towards the end of the six week study period. This trend was not significant, however. Suggestions for future research are
given.