Advection is clearly shown by the fact that the heat 

 exchange equations for Koko Head from February to 

 September 1964 disclosed a net loss during this period. Yet 

 there was a substantial increase in the temperature of the 

 surface waters. This increase was due to the northward 

 transport of water by currents. 



Seckel has presented evidence that the appearance of 

 the season tuna in the island area is closely correlated with 

 changes in oceanographic conditions, and particularly the 

 presence of warm water of relatively low salinity that flows 

 into the area from the south and southeast. 



Wakes in the Lee of Islands 



Skipjack tuna of the eastern and central Pacific are 

 found over a range of about 6,000 miles. As a rule they 

 are caught, as is obvious from figure 12, no more than a 

 few miles from the islands in Hawaii, and in the eastern 

 Pacific as far as k few hundred miles offshore. If these 

 very small needles in a very wet haystack are to be located, 

 our knowledge of the mechanisms that concentrate them, 

 i.e., oceanographic conditions, needs expansion. 



The very first oceanographic cruise conducted by the 

 Laboratory in Honolulu in 1949 located a natural phenom- 

 enon the implications of which are only now being ex- 

 plored. To the we.st of the great island of Hawaii a wide 

 counterclockwise eddy was found. On a subsequent fishery 

 exploration cruise in the area, skipjack tuna schools were 

 found to be more plentiful near the eddy (350 miles from 

 land) than near the shores of the Hawaiian Islands. Later, 

 in 1962, E. C. Jones, a Laboratory scientist, studied the 

 "island effect" on the zooplankton (which make up part of 

 the tuna's food) near the Marquesas Islands in the central 

 South Pacific Ocean. He found inshore forms of zooplank- 

 ton organisms much farther to sea (80-100 miles) than 

 could be explained by the operation of pure chance. 



Thus the presence of islands in the ocean might alter 

 the productivity of the waters considerably beyond the 

 immediate shores. 



A mechanism that could explain the presence of near- 

 shore copepods far to sea and the presence of tuna con- 

 centrations hundreds of miles from the nearest islands has 

 now been advanced by Richard A. Barkley, an oceano- 

 grapher at the Laboratory in Honolulu. He has postulated 

 that an island standing in the path of an ocean current 

 would set up what is known as a von Karman wake, a 

 disturbance in the stream that results in the regular for- 

 mation of eddies that move downstream some distance from 

 the island (the distance depending on the island's size 

 and the strength of the current) . To test this theory, the 

 Toxvnsend Cromwell and Charles H. Gilbert went to sea 

 in January and February 1968 to study the structure of 

 the current downstream from the large island of Hawaii 

 (diameter about 80 miles) and tiny Johnston Island (dia- 

 meter about 10 miles) . At Johnston they found a wake that 

 neatly fitted the theory (fig. 19). Eddies rotating clock- 

 wise and counterclockwise formed alternately in the lee of 

 the island. They took about 2 days to form and grow to full 

 size. When they reached maturity, they began to migrate 

 downstream (northwest) at a rate of about 2.2 miles a day. 

 The eddies were about 50 miles across. The picture off Ha- 

 waii was more complex, owing to interference from other 

 nearby islands, but eddies of the expected size and speed 

 were found. 



Since the counterclockwise eddies are divergent, bringing 

 cool, enriched water near the surface, they should be more 

 productive than adjacent waters and consequently provide 

 more forage for fishes, including tunas. The eddies also 

 would explain the non-random dispersion of nearshore 

 forms far to sea observed bv Jones. 



28 



