Potential detection of illegal fishing by passive acoustic telemetry

Acoustic tagging is typically used to gather data on the spatial ecology of diverse marine taxa, informing questions about spatio-temporal attributes such as residency and home range, but detection data may also reveal unanticipated insights. Many species demonstrate predictable site fidelity, and so a sudden cessation of detections for multiple individuals may be evidence of an atypical event. During 2013 and 2014, we acoustically tagged 47 grey reef sharks (Carcharhinus amblyrhynchos) and 48 silvertip sharks (Carcharhinus albimarginatus) near reefs in the British Indian Ocean Territory (BIOT) Marine Protected Area (MPA). From March 2013 to November 2014 inclusive, tags were ‘lost’, i.e. permanently ceased to be detected within the monitoring area, at an average rate of 2.6 ± 1.0 tags per month. Between 1 and 10 December 2014, detection data suggest the near-simultaneous loss of 15 of the remaining 43 active tagged sharks, a monthly loss rate over five times higher than during the previous 21 months. Between 4 and 14 December of 2014, the BIOT patrol vessel encountered 17 vessels engaged in suspected illegal fishing in the northern BIOT MPA; such sightings averaged one per month during the previous 8 months. Two of these vessels were arrested with a total of 359 sharks on board, of which grey reef and silvertip sharks constituted 47% by number. The unusual and coincident peaks in tag loss and vessel sightings, and the catch composition of the arrested vessels, suggest illegal fishing as a plausible explanation for the unusual pattern in our detection data. A Cox proportional hazards model found that the presence of fishing vessels increased the risk of tag loss by a factor of 6.0 (95% CI 2.6–14.0, p < 0.001). Based on the number of vessels sighted and the average number of sharks on vessels arrested in BIOT during 2014, we conservatively estimate that over 2000 sharks may have been removed during the suspected fishing event. Based on average catch compositions, over 1000 would have been grey reef and silvertip sharks. Assuming a closed population mark-recapture model, over one-third of the locally resident reef sharks may have been removed from the monitoring area. The data suggest that even sporadic fishing events may have a marked impact on local reef shark populations, but also demonstrate the potential of electronic tagging a tool for detecting illegal or otherwise unreported fishing activity.