CAREY and ROBISON: DAILY PATTERNS IN ACTIVITIES OF SWORDFISH 



threshold for vision in deep-sea fish (Denton and 

 Warren 1957; Clarke and Denton 1962). The 

 swordfish should be able to locate its prey visually 

 in its dimly lit environment both day and night. 



Our depth records for swordfish show many 

 variations from the U-shaped pattern expected if 

 they are following isolumes. Much of this varia- 

 tion can be explained as modification of a response 

 to light by other environmental factors. On 10 

 November the depth record for no. 7 was skewed, 

 with the greatest depth reached late in the after- 

 noon (Figure 8). The deepening isotherms during 

 this day indicate that the fish was moving from 

 shelf and slope water into the warm, clear water 

 of the Gulf Stream. Light attenuation is much 

 greater inshore than in the Gulf Stream ( Jerlov 

 1968) where the same level of illumination occurs 

 at greater depths. The gradual increase in depth 

 during the fading daylight on this afternoon 

 coincided with the movement from slope to Gulf 

 Stream water and can be interpreted as the fish 

 maneuvering deeper to maintain a constant light 

 intensity in the clearer water. 



On 9 November, no. 7, which had reached 400 m 

 by midmorning, returned to 100 m at noon. This 

 unexpected midday rise from depth occurred when 

 the fish left a region of clear blue slope water and 

 passed under a streamer of dark grey-green shelf 

 water ( indicated as a marked thermal inversion in 

 Figure 8 and as a light-colored region in the 

 satellite infrared image, Figure 7). Swimming 

 under this dark shelf water, the swordfish entered 

 a shadowed area. By rising toward the surface it 

 would have returned to a light level which pre- 

 vailed at greater depth in the blue water. A 

 change of attenuation coefficient from 0.035 in 

 blue water to 0.140 in shelf water would result in 

 the same light intensity at 400 m as at 100 m 

 depths, respectively. These are reasonable atten- 

 uation values and it is possible that the swordfish 

 was maintaining a constant light level during this 

 vertical movement. 



Oxygen 



A vertical movement in response to sunrise and 

 sunset occurred in all of the Baja California 

 experiments. An interesting feature is apparent 

 when these depth records are aligned vertically 

 (Figure 9). The rapid descent which began about 

 an hour before dawn each morning reversed at 

 sunrise when the fish came back up to about 

 100 m. A well-developed oxygen-minimum layer 



exists in this area of the eastern tropical Pacific 

 and the oxycline at the top of the low oxygen 

 region is parallel to the thermocline, but at some- 

 what greater depth (Griffiths 1968, fig. 11 and 27). 

 In its normal movement away from the surface 

 with increasing light at dawn, the swordfish 

 penetrated deeply into this low oxygen layer, then 

 returned to spend most of the day at shallower 

 depths with higher oxygen concentration. The 

 presence of the oxygen-minimum layer caused the 

 swordfish to spend the day at a shallower depth 

 and higher light level than normal. 



In Baja California swordfish frequently came up 

 during the day to bask on the surface with the tips 

 of their dorsal and caudal fins out of the water for 

 periods of 15-80 min. This basking behavior, at a 

 time of day when we would expect them to be at 

 their greatest depth, may be related to the low 

 oxygen concentration at depth. The depth record 

 for no. 6 superimposed on an isotherm pattern in 

 Figure 5 shows that the fish was well below the 

 thermocline during much of the day. We did not 

 measure oxygen, but using the data in Griffiths 

 (1968) we can infer concentrations from water 

 temperature and depth. When swordfish no. 5 and 

 6 were below the thermocline, they were in an 

 environment with an oxygen concentration only 

 10 to 20% that of air-saturated water. This is 

 a much lower concentration than the 60*%^ sat- 

 uration suggested as the lower limit for skipjack 

 tuna in the vicinity of an oxygen-minimum layer 

 (Ingham et al. 1977). The less active swordfish, 

 with its large mass of white muscle, might be more 

 resistant to anoxia and able to accumulate an 

 oxygen debt. If so, the periodic excursions to well- 

 aerated surface waters would allow it to recover 

 from anoxia. There appears to be a rough correla- 

 tion between time spent on the surface and the 

 length of the preceding period below the ther- 

 mocline, as would be expected if this were a 

 recovery process: Ts = 0.2 Tl + 16 ( r = 0.6; 

 n = 10) where Ts is time on surface in minutes and 

 Tl is time at depth in minutes. 



In the Baja California area, there was a marked 

 difference between frequencies at which the in- 

 shore and offshore swordfish came to the surface 

 during the day. The inshore fish surfaced six times 

 in 7 d and spent 2.8'7f of the daylight hours on the 

 surface. The offshore fish surfaced 10 times in 2.5 

 d, spending 25.7% of the daylight hours on the 

 surface. This difference may be related to oxygen 

 concentrations, for the inshore fish were near the 

 mouth of the Gulf of California where the oxygen- 



289 



