Island zones probably does not represent a difference In general tuna abundance but 

 In the main reflects the higher relative populations near land of easily captured 

 yellowfln. In addition, the reef areas, known to harbor concentrations of yellow- 

 fin, are occasionally traversed during secondary trolling, erratically increasing 

 the results of Insular secondary trolling. 



DISCUSSION 



The sightings smd trolling catches suggest several generalizations on the 

 distribution of surface tunas in the central Pacific. First, close to the reefs 

 there appear to be concentrations of tunas dominated by the yellowfin. Beyond the 

 immediate vicinity of the reefs the presence of land does not seem to affect the 

 abundance of tuna but does affect the species composition. Yellowfln schools are 

 sighted more frequently in the island than in the semioceanic and oceanic zones, 

 where skipjack dominate. This may indicate that the presence of land alters the 

 ecology in such a way that yellowfin can exist in larger numbers in the adjacent 

 waters. Skipjack, on the other hand, dominate the semioceanic sind oceanic zones, 

 being almost completely replaced by yellowfin in the island zones of the Line and 

 Phoenix islands, and to a limited degree of the Hawaiian Islands. 



Since it appears that the mere presence of land does not alter the density of 

 the tuna population, except in limited areas, we should consider the possibility 

 that the different levels of abundance (disregarding species) in different island 

 and oceanic areas are a function of different levels of general productivity of the 

 oceanic waters. To examine this hypothesis, it is necessary to suppress variation 

 in the tuna population associated with different seasons, either by comparing dif- 

 ferent areas in the same season or by making comparisons between areas having 

 representation throughout the year. The latter course appears the sounder, because 

 we do not have the information on homologous seasons, and if comparisons are limited 

 to segments of the year there is danger of comparing one area at its peak with 

 another area at sui "off season." Finally, comparisons must be restricted to areas 

 or zones equally accessible to aquatic birds, because most schools are overlooked 

 unless working birds advertise their presence. 



In order to compare the relative abundance of fish schools with productivity 

 we shall use the standing crop of zooplankton, the level of which has been shown to 

 be related to the oceanic circulation (Cromwell 1954, King and Demond 1953, Sette 

 MS, etc.), as an index of relative oceanic productivity. Considering the three 

 island areas. King and Demond (1953) provide data indicating that the relative order 

 of island groups with respect to zooplankton abundance is Hawaiian Islands, Phoenix 

 Islands and Line Islands in ascending order. If these same groups of islands are 

 listed according to the apparent abundance of fish schools, using the average of 

 the unweighted average quarterly sighting rates from table 1 (1.4, 1.9, 2.1), their 

 order remains unchanged. A similar comparison of fish-school abundance using the 

 semioceanic sightings (table 2) gives the same order, though the comparison is less 

 precise because of poor stratification in time. 



In the oceanic area the abundance of surface fish and birds does not conform to 

 the abundance of plankton as well as in the insular areas. King (MS) indicates 

 that north of 10°N. the abundance of zooplankton is relatively low as is also the 

 abundance of birds and fish schools. In the Counter current he reports considerably 

 more zooplankton than to the north, and here surface schools are also relatively 

 abundant. In the South Equatorial Current between the Equator and 5°N. latitude he 

 reports even more zooplankton than in the Counter current, but fish schools are only 

 about as abundant as in the zooplankton-poor region north of the Countercurrent. 

 South of the Equator (0° to 10°S.) there is somewhat less zooplankton than in the 

 Countercxirrent, but fish schools are slightly more abundant than in any of the zones 

 under consideration, though there is reason to suspect that the sampling south of 

 the Equator has not been adequate. 



The discrepancies between fish schools and plankton in the oceanic area may 

 not be in serious conflict with the general idea that the size of the population of 

 animals (fish and birds) at one trophic level is related to the size of the popula- 

 tion at lower levels (plankton), for we are here considering only one element of 

 the fish population (surface schools that are mostly skipjack). It was pointed out 

 that the distribution of surface schools does not correspond with the distribution 

 of subsurface or deep-swimming tunas (mostly yellowfin). The fact that deep- 

 swimming yellowfin tuna are most abtindant in regions of few surface tuna schools 



15 



