move southward actively in response to the southward 

 advance of cool water from late fall to early spring, the 

 fish seeking an environment more in accord with spawn- 

 ing than with feeding requirement (Blackburn and Willi- 

 ams 1975). Support for this view is seen in the northern 

 limit of larval skipjack tuna occurrence within this re- 

 gion. The distribution of skipjack tuna larvae terminates 

 sharply at lat. 10"'-12°N between long 110° and 150°W 

 (Matsumoto 1966). 



Utilizing information gained from tagging by Fink 

 and Bayliff (1970), the drift model by Seckel (1972), and 

 the longline catch data, one could postulate the routes 

 followed by the tagged fish from the eastern to the 

 central Pacific (Fig. 15). Fish A, tagged off Baja Cali- 

 fornia in September 1960 and recaptured in Hawaii in 

 June 1962, had been at liberty for 21 mo. Fish B, 

 tagged off the Revilla Gigedo Islands in April 1960 and 

 recaptured in Hawaii in August 1962, had been at 

 liberty for 28 mo. Considering that fish B had been 

 tagged 5 mo earlier than fish A, that the general 

 movement of fish in the eastern Pacific is northward in 

 the third quarter, and that both fish had been recaptur- 

 ed in the same season (only 2 mo apart) in Hawaii, it is 

 not unreasonable to assume that fish B had been tagged 

 from the same group of fish (either school or schools) as 

 fish A and that both fish, having left the eastern Pacific 

 fishery in the fourth quarter, had a) entered the Hawai- 

 ian fishery in the following spring, remaining in the area 

 a full season before being recaptured (routes Al, Bl), or 

 b) followed the longer routes A2, A3, B2, or B3. 



Fish C, tagged off Baja California in September 1961 

 and recaptured east of Christmas Island in April 1963, 

 was at liberty for 18 mo. The elasped time from release 

 to recovery and the movement across the NECC suggest 

 that the fish followed route CI or C2, or some route 

 between the two. The recovery of this fish in the SEC 

 shows that the movement involved active migration as 

 well as drifting with the current. 



Fish D, tagged off Revilla Gigedo Islands in June 

 1965 and recaptured in Hawaii in June 1967, had been 

 at liberty for 24 mo. This fish probably followed a path 

 similar to that taken by fish B. 



Fish E and F, tagged in November 1969 off Clipper- 

 ton Island and recovered in Hawaii 9 and 10 mo later, 

 respectively, lighely followed a shorter, more direct route, 

 migrating nearly due west into the Hawaiian area and 

 being caught soon after arrival. Such a movement is 

 possible simply by drifting with the current, as indicated 

 by Seckel (1972), with a minimum of active migration by 

 the fish. Current trajectories drawn from monthly sea 

 surface current charts also indicate a current path simi- 

 lar to that shown in Figure 15, with a time requirement 

 of 11 mo between the two islands (Barkley, Southwest 

 Fisheries Center, National Marine Fisheries Service, 

 NOAA, pers. commn.). 



Fish G was tagged by the Honolulu Laboratory at 

 lat. 4°11'N, long. 119°02'W on 26 October 1969 and 

 recaptured in Hawaii 21 mo later on 14 July 1971. 

 Based on the length of time at liberty, this fish likely 

 had passed through the eastern Pacific fishery during 

 the first year and had entered the Hawaiian fishery in 

 the second or third quarter of the following year. A 

 path that coincides with the assumed paths of fishes A, 

 B, and D, and with the quarterly movement of high- 

 CPUE cells of the longline fishery can be projected 



without exaggeration. The importance of this recovery, 

 as well as the one east of Christmas Island, is that it 

 showed the ability of the skipjack tuna to migrate 

 across the NECC at a time when the current was well 

 developed. 



An alternate route that these fish could have taken 

 in moving from the eastern Pacific to Hawaii is directly 

 across the California Current (Fig. 14, route A4, B4, 

 C3, D4, and G2). It seems unlikely that they did this, 

 however, because 1) if this was the normal migratory 

 route, the longline catches of skipjack tuna there should 

 have been much higher than those recorded, yet the 

 CPUE in each 5° area was not only low, but inconsi- 

 stent from year to year; and 2) fish C would have had 

 to move counter to the seasonal trend between Hawaii 

 and the equatorial region without dallying in Hawaiian 

 waters. It would seem more likely that once fish C had 

 entered the Hawaiian fishery it would have stayed there 

 for the remainder of the season and perhaps into the 

 next as well. 



In the central Pacific, the movements of skipjack 

 tuna derived from longline catch data largely represent 

 new information. In the area north of the equator, 

 between Midway and Hawaii, the movement patterns 

 indicate entry of fish into the Hawaiian fishery from 1) 

 the eastern Pacific (corroborated by tag recoveries), 2) 

 equatorial waters to the south, and possibly 3) the 

 Japanese offshore fishery. Additionally, there could be a 

 strictly local stock within the area. Captures of larval 

 skipjack tuna (Matsumoto 1958) around Hawaii indicate 

 that the species spawns there from spring through early 

 fall. Some of the larvae remain in the vicinity of the 

 islands and develop through the early juvenile stages, 

 but many are undoubtedly carried downstream in the 

 prevailing westerly current. Some of these probably get 

 caught in the North Pacific gyre and return the follow- 

 ing year or the year after as either 1- or 2-yr-old fish. 

 An indication of the presence of fish from these several 

 sources in the Hawaiian fishery was hinted at by Roth- 

 schild (1965). From length -frequency distributions of fish 

 taken by the fishery, he recognized typical modal lengths 

 of 35, 50, and 70 cm in winter and 45 and 75 cm in 

 summer. Because the positions of the modes did not 

 necessarily increase with time, Rothschild reasoned that 

 the data suggested passage of successive groups of 

 skipjack tuna through the fishery and that this passage 

 was not uniform with respect to time. 



In the South Pacific, at least two types of migratory 

 routes are indicated: a short route that takes fish from 

 an area east of the Marquesas through the Society 

 Islands and back to an area southeast of the Marquesas 

 Islands within 1 yr, and a longer route that takes fish 

 from the equator northeast of the Marquesas to as far 

 west as the Samoa-Fiji area and back east to an area 

 southeast of the Marquesas Islands in 2, or possibly 3, 



yr. 



The image one gets from these projected routes of 

 migration of the central-eastern Pacific subpopulation is 

 that the skipjack tuna or their progeny can wend their 

 way in due time from the eastern Pacific to the bound- 

 ary of the western Pacific subpopulation in both hemi- 

 spheres. 



How well this model of stock identification and mi- 

 gration reflects the actual conditions in the Pacific Ocean 

 can be tested by the more positive means of tag 



26 



