Preparation of data on observed protected species captures, 2002–03 to 2014–15

Incidental captures of protected species in New Zealand waters are recorded by fisheries observers when they are on-board commercial fishing vessels. As observer coverage varies across fisheries, statistical models are used to estimate the total number of incidental captures across New Zealand fisheries. These bycatch assessments rely on the linking of observer records to fisher-reported effort data. Preparation of the data also requires corrections to errors in the capture records (e.g., misidentifications of species), and the inclusion of capture records from other sources, such as seabird necropsies and examination of photographs taken by observers.

This report documents the data preparation process used for observer records of protected species captures for the period between the 2002–03 and 2014–15 fishing years. The data preparation process included the linking of observer and fishing effort data, the expanding of multiple capture events that were reported as single captures, and updates to capture records and species identifications, following expert reviews through necropsies and photograph examinations. Other updates to the dataset included the omission of records such as data that were not consistent with the definition of incidental captures or were from waters outside New Zealand’s Exclusive Economic Zone. Some seabird captures that were only identified by fisheries observers also had their identifications modified through an imputation process. The imputation used information from other captures that were formally identified, from necropsy or photograph, to infer the likely identification.

The final dataset consisted of 8921 protected species captures for the fishing years between 2002–03 and 2014–15. Included in this dataset were 6621 captures of seabirds.

Three substantive issues were found with the data used for estimating marine mammal and seabird bycatch to the end of the 2013–14 fishing year. These issues were missing observer trip records, inclusion of seabird deck captures, and the incorrect matching of photograph identifications to the observer capture data. Addressing these issues with resulting changes to the data will lead to changes in the number of estimated captures, especially of seabirds. When the seabird bycatch estimation is updated to the end of the 2014–15 fishing year, estimates for all previous years will also be updated.

This report highlights several areas where improvement in the processes would help with more timely (and potentially more accurate) records of bycatch data. Multiple organisations work with the observer capture data, and depend on them for their reporting. Developing collaborative working methods would help to develop a consistent, reliable, and timely view of protected species bycatch information.