Pop-up Archival Tags Used to Model Daytime Vertical Habitat of Swordfish in the North Pacific Ocean

PIFSC Research Assistant Melanie Abecassis, an employee of the University of Hawaii Joint Institute for Marine and Atmospheric Research, recently analyzed data recorded by Wildlife Computers pop-up archival transmitting tags attached to 28 swordfish caught and released in the North Pacific Ocean as part of an 8-year study to learn about swordfish movements in horizontal and vertical dimensions of their pelagic habitat. She was joined in the research by Heidi Dewar of the Southwest Fisheries Science Center and PIFSC scientist Jeff Polovina.

Four of the swordfish were tagged by longline vessels in the central Pacific. The other 24 fish were tagged by harpoon boats off San Diego, CA. Of the latter group, 13 fish radiated away from California, while 5 remained near San Diego, one of them for all 180 days it was tracked.

Tracks of swordfish outfitted with pop-up archival transmitting tags.  Triangles indicate release and 
            recovery locations. Dots mark estimated positions. Dashes denote that position data could not be obtained.
Tracks of swordfish outfitted with pop-up archival transmitting tags. Triangles indicate release and recovery locations. Dots mark estimated positions. Dashes denote that position data could not be obtained.

Track durations ranged from 10 to 180 days. Classic diel movement (shallow depths at night, 0–100 m; below 200 m during the day) was observed with the exception of occasional daytime basking at or near the sea surface in the general area of the California Current in the fall and winter. Five fish dove as deep as 1200 m. The overall temperature range recorded was 3.2 to 28.8 °C.

Overall, tagged swordfish spent 60% of the time between 0 and 100m, 22% of the time between 200 and 400 m and 7% of the time below 500 m. The overall distribution of residence time vs. temperature exhibits 2 modes, one between 7.5 and 10 °C and another between 16 and 18 °C.

Proportion of time spent by tagged swordfish in various intervals of depth and temperature.
Proportion of time spent by tagged swordfish in various intervals of depth and temperature.

During the daytime, the tagged swordfish spent 66% of the time at depths between 200 and 750 m and 24% of the time above 50 m (basking). They spent 65% of the daytime in waters between 5 and 10 °C. During the night, they spent 97% of the time above 100 m and 37% at the surface (<5 m). They spent 91% of the night in waters warmer than 16 °C, with a mode between 16 and 18 °C (31%). Swordfish feed near the surface during the night and at depth during the day when their prey is more densely distributed and less active.

Abecassis, Dewar and Polovina developed generalized additive models (GAMs) to assess the separate and combined effects of oxygen, temperature and light on daytime foraging depth. The predictors retained in the model are the concentration of chlorophyll-a (used as a proxy for light), the concentration of dissolved oxygen at 400 m, and the temperature of the water column at 400 m (400 m being the overall average daytime mean depth occupied by the tagged swordfish when basking events were neglected). Chlorophyll-a concentration by itself explained 57 % of the variation in daytime mean depth registered by swordfish; dissolved oxygen concentration at 400 m explained 51% and temperature at 400 m explained 47%. According to the model, decreases in daytime mean depth are associated with increases in chlorophyll-a concentration, whereas increases in daytime mean depth are associated with increased oxygen concentration and temperature at 400 m. The best model combining all three factors explained 78% of the variability in observed daytime foraging depth.

The model performs very well in predicting daytime depth for individual tags. Monthly maps of predicted daytime foraging depth can be produced using chlorophyll-a input data from MODIS-Aqua and oxygen and temperature data from the World Ocean Atlas. Such maps could potentially be used to improve targeting of swordfish in deep daytime longline sets. Deep daytime sets would entail less risk of unwanted bycatch than the shallow night-time sets typically used in the swordfish longline fishery.

This research was partially funded by the Pelagic Fisheries Research Program (PFRP).