Improving estimates of cryptic mortality for use in seabird risk assessments: loss of seabirds from longline hooks

Estimates of seabird bycatch within fisheries usually make no allowance for cryptic mortality
—birds killed or wounded but not brought on-board the fishing vessel. To examine loss of
seabirds we manually attached 195 seabird ‘surrogates’ —euthanized domestic ducks — to
longline hooks during a typical pelagic longline ‘tuna set’ conducted on an 18.7 m longline
vessel fishing in New Zealand waters. Hooks were either set unbaited or with a single bird
attached to the hook. Three treatments were used for hook attachments: through the lower
bill; internally hooked in the throat; or by entangling the line and hook around the wing. The
fishing operation proceeded as a normal fishing set, with the exception that most hooks were
not baited. A minimum of 750 hooks were set, with floats deployed every 12 hooks. Mean
hook soak time was 15.47 h (range 12.25 - 19.23 h).
Our results showed an almost complete retention of birds hooked during our experiment. Of
the recovered branchlines on which a duck was deployed only 1.54%, 1.56 % and 9.35% were
missing for bill, throat, and wing hooked birds, respectively. Statistical analysis showed the
difference between treatment factor levels was not significant with the difference between the
coefficients for the wing and the bill attachment levels of 1.89 (SE=1.10, P=0.084)
corresponding to a higher estimate on the logit scale and therefore a higher probability of the
carcass being missing for the wing attachment compared to a bill attachment. As expected
from the raw data the difference between the bill and throat attachments was negligible with
coefficient 0.015 (SE=1.43, P=0.99).
Replication of this study in other areas, seasons and fisheries would provide more certainty
on loss of bycaught birds, but the cost of doing so would need to be weighed up against the
gain in accuracy and precision of bycatch estimates.

RECOMMENDATIONS
1. SBWG to consider an appropriate cryptic scalar for estimating seabird bycatch in pelagic longline fisheries, noting the results of this study and
the earlier work of Brothers et al. (2010).