were not as well repi>esented in the catches as they were in the population since the 

 largest fish hooked were most frequently lost. 



Trolling surveys in the Phoenix Islands indicate an even higher catch rate than 

 in the Line Islands (table 4), but the sample is somewhat smaller and may not be as 

 reliable. The Phoenix Island catches closely resembled in species composition those 

 from the Line Islands, with a catch of 43.0 tuna per 100 line-hours, all of them 

 yellowfln. 



The amount of primary trolling at other Island areas in the central Pacific 

 (table 4) was Inadequate for an evaluation of abundance, but very good catch rates 

 for yellowfln were experienced during the few hours of trolling performed at Maiden, 

 Starbuck, and Jarvis Islands. 



The secondary trolling (table 5) Is less reliable than the primary trolling as 

 an index of relative abundance among the several Island groups because tuna trolling 

 in all the island areas, possibly excluding the Hawaiian Islands, Is particularly 

 productive in very specialized locations, usually near reefs. Inasmuch as these lo- 

 cations were only occasionally and unsystematlcally traversed during secondary trol- 

 ling the catches do not afford unbiased estimates for the several island areas. The 

 records will, however, be useful later in the discussion. 



Oceanic and Semioceanlc Trolling 



The 5,525 line-hours of secondary oceanic and semioceanlc trolling resulted in 

 a very low tuna catch. Only 70 fish were landed: 7 skipjack, 8 yellowfln, 40 

 dolphin, 12 wahoo, 1 little txuiny, 1 black marlln, and 1 barracuda (table 5). This 

 Is an average of 1.3 fish per 100 line-hours, 0.3 of which were tuna. 



Table 5. — Results of semioceanlc and oceanic 

 trolling 



It is of particular Interest to note that the nearly equal catches in the 

 oceanic and semioceanlc trolling (table 5) are at variance with the number of sur- 

 face schools, which were seen 5 times as frequently in the semioceanlc zone (page 

 7). Thus it appears likely that the greater relative number of sightings in the 

 semioceanlc areas is mainly a reflection of the greater abundance of birds within 

 180 miles of Islands. This confirms the surmise in the previous section that the 

 fish school sighting!* decreased when passing from the semioceanlc to the oceanic 

 provinces because the bird population decreased. 



The secondary trolling in the semioceanlc zone resulted in smaller tuna catches 

 (0.3 per 100 line-hours) than the secondary trolling In the combined island zones 

 (2.3 per 100 line-hours) (tables 4 and 5), suggesting differences in abundance that 

 were not evident from the sightings. Actually, the higher troll catch in the 



14 



