YOSHIDA: AMERICAN SAMOA LONGLINE FISHERY 



for space on the gear. Furthermore, there is 

 another compHcating variable: the fishermen ap- 

 parently seek out yellowfin tuna when the catches 

 of albacore are poor. They do this by modifying the 

 longlines to fish shallower and by fishing closer to 

 the equator where yellowfin tuna catches are 

 known to be better. The fact that the price of 

 yellowfin tuna increased from an average of $270 a 

 ton in 1965 to $394 a ton in 1970 may also have been 

 a factor. 



As noted earlier, the CPUE of albacore in rela- 

 tion to increased fishing effort has been declining, 

 especiallv in the years subsequent to 1967. During 

 1968 to 1971, tne yellowfin tuna CPUE was rela- 

 tively good. It is apparent that as albacore fishing 

 deteriorated, the vessels expended more effort to 

 catch yellowfin tuna. The relation between al- 

 bacore and yellowfin tuna CPUE in the Samoan 

 longline fishery from 1963 to 1971 is shown in 

 Figure 9. It appears that an inverse relation 

 existed between albacore and yellowfin tuna 

 CPUE: When yellowfin tuna CPUE was high, al- 

 bacore CPUE was low, and vice versa. The 

 correlation coefficient (r = -0.6636; df = 7), 

 however, was not significant. 



< 



z 



3 



s 



o 

 o 



X 



o 

 o 



2.00 300 400 



CATCH / 100 HOOKS - ALBACORE 



Figure 9.— Relation between albacore and yellowfin tuna CPUE. 



SPATIAL AND 

 TEMPORAL CONSIDERATIONS 



Observations on the American Samoa fishery 

 during the period from 1954 to 1965 indicated that 

 the longline vessels shifted fishing grounds with 

 season (Otsu and Sumida 1968). Seasonal and 

 geographical differences in CPUE, however, were 



not readily apparent in the early years. In the 

 following sections I will examine the spatial and 

 temporal distribution of effort and CPUE in the 

 fishery from 1966 through 1971. 



Effort 



In the early years of the fishery, the longliners 

 confined their fishing largely to the vicinity of the 

 Samoa Islands. Over the years, the vessels 

 gradually extended their operations to more dis- 

 tant waters, and by 1965 the fishing grounds 

 reached from about long. 175°E to about long. 

 120°W between the equator and lat. 30°S (Otsu 

 and Sumida 1968). From 1966 to 1971 there was a 

 further extension of the fishing grounds; the ves- 

 sels fished from off the east coast of Australia to 

 long. 100°W and from about lat. 10°N to about lat. 

 40°S. The fishing effort, however, was not 

 uniformly distributed throughout the 

 geographical limits of the fishery. Rather, there 

 were distinct seasonal patterns in the spatial dis- 

 tribution of fishing effort. 



A composite geographic distribution of the 

 fishing effort on a quarterly basis for 1966 to 1971 

 summarized by 2° squares between the equator 

 and lat. 40°S and east of long. 150°E is shown in 

 Figure 10. As composite charts they can reflect 

 only "average" conditions. 



In the first and fourth quarters, the vessels 

 generally fished to the north of lat. 20°S, and areas 

 of concentrated fishing effort developed in about 

 the same area each year. In the second and third 

 quarters, the vessels expanded their operations to 

 the south of lat. 20°S and, in addition to the usual 

 area of concentrated effort to the north, an area of 

 high fishing effort also developed to the south of 

 lat. 20°S. However, there apparently has been a 

 change from earlier years in the operations of the 

 fleet because prior to 1966 the vessels fished in the 

 north during both the first and second quarters 

 (Otsu and Sumida 1968). These figures indicate 

 that the vessels are moving south in the second 

 quarter, earlier than previously. In any event, 

 these seasonal changes in the concentration of 

 fishing effort have been interpreted to reflect the 

 movement of albacore in the South Pacific Ocean 

 (Suzuki 1961; Otsu and Sumida 1968). 



Catch Per Unit of Effort 



The mean quarterly CPUE for 1966-71 plotted 

 by 2° square areas is shown in Figure 11. In a 



753 



