moderately high longline effort (see Fig. 13). Because of the com- 

 mercial longline effort in this area, a number of short-term (0-60 

 d) recoveries were obtained. Of importance in evaluating the tag 

 return data for migratory patterns is the relationship between a 

 shift of the high CPUE areas and the frequency of recoveries in 

 these areas. Recoveries about 6 mo, and 1 or more years after tag- 

 ging, are most important, if one is to assume a seasonal migratory 



pattern exists. 



Most of the tag and recovery geographical plots given in 

 Figures 5 to 10 tend to give the impression that all migration is 

 radiating outward from a geographically localized point of tag- 

 ging, and that the tagging location is the "center" of distribution; 

 this is not the case. Black marlin are tagged in an area as they 

 migrate through it at varying rates and directions. 



Emigration from the area of tagging during the first few 

 months appears, for most recoveries, to be toward the south-south- 

 east. An overall mid-point average for the first 0-60 d of release (x 

 = 30 d) was 109 nmi; for period 2, 61-120 d (x = 90 d), 547 nmi; 

 and for period 3, 121-240 d(x= 180 d), 1,386 nmi. Arc distances 

 given in Figure 12 show that, based on average distance/time (180 

 d and 1,386 nmi), the average distance of migration would be 

 from about southern New South Wales, just east of the New He- 

 brides, northeast to midway between Solomon Island and the 

 Gilbert Islands to the Equator. The longest distance recorded was 

 to the southeast of the tagging area, east of New Zealand (2,100 

 nmi, or 8.9 nmi/d). For black marlin recovered 121-240 d after 

 tagging, the average rate was 7.7 nmi/d and in the time period of 

 180 d the average distance traveled at that rate would be 1,386 

 nmi. However, this sample, having a release time of 0-240 d, 

 represents only 13.5% of the total recoveries. Based on average 

 migration rate data, black marlin tagged early in the season (Sep- 

 tember) tended to migrate away from the point of tagging at a 

 lower average rate for the first time period (0-60 d) than black 

 marlin tagged in October, November, or later in the fishing 

 season. This may be because the tagging areas are in or near the 

 spawning area, and the behavior of black marlin in this area 

 earlier in the spawning season may be different from those enter- 

 ing later in the season. 



Data obtained from this study indicate that black marlin tagged 

 in the western Coral Sea do not undergo short-term trans-Pacific 

 migrations, although some interchange over time with the eastern 

 Pacific is possible. The degree of interchange with the Indo- 

 Pacific is unclear. Emigration from the tagging area to north of 

 New Guinea was recorded; however, no recoveries were recorded 

 to the east in the Arafura, Banda, or Timor Seas or the eastern In- 

 dian Ocean. 



Some tentative estimate of the central tendency of migration 

 direction and rate can be made using the vector analysis (Fig. 1 1), 

 the graphic plots of tag and recovery' points (Figs. 4 to 9), and 

 movements of high CPUE areas over time (Fig. 14a, b, c), in rela- 

 tion to the geographical distribution of longline effort levels in the 

 southwest Pacific (Fig. 13). Figure 15 gives the approximate 

 geographical centers of high longline CPUE by month as observ- 

 ed in data presented in Figure 14a, b, c. Inspection of the longline 

 CPUE rates for the 5° areas and plots of geographical location in- 

 dicate south or southeast movement from the tagging area in the 

 summer and then a northward movement of high CPUE areas to 

 the New Guinea-Bismark Archipelago-Solomon Islands area in 

 the winter. Vector mean bearing and distance data from tag 

 results were plotted in Figure 1 1 and indicate direction and 

 distance of migration away from the tagging area for 30, 90, and 

 180 d from October, the month having the most tagging activity 

 (49%). 



Monthly average sea surface temperature isotherms are shown 

 in Figure 14a, b, c. High longline CPUE areas for black marlin 

 are located in close relation to the 26.7°C (80°F) average isotherm 

 during most months of the year. High CPUE areas are related to 

 lower temperatures and are found between the 23.9°C (75°F) and 

 26.7°C (80°F) isotherms only in February off the Queensland and 

 New South Wales coasts. Latitudinal warming and cooling as 

 reflected in sea surface temperature may be a measure of other 

 physical or biological environmental parameters that may be im- 

 portant to black marlin distribution. 



From the results of tagging (emigration rates and directions), 

 inspection of average longline effort, and CPUE, a diagramatic 

 description of black marlin migration in the southwestern Pacific 

 can be hypothesized (Fig. 16). The tagging results indicate that the 

 migratory rates and patterns of black marlin are highly variable. 

 There is, however, a central tendency of movement of tagged fish 

 not unlike that expected from observations of the movements of 

 CPUE trends. Black marlin were observed to move southward 

 from the tagging area toward southeastern Australia and New 

 Zealand in late summer, then northeast toward the Gilbert 

 Islands, and to northeast of New Guinea in the winter, returning 

 to the western Coral Sea in the spring and early summer. The in- 

 terchange rate of the population found in the Coral Sea, with the 

 population of the Indo-Pacific area, is unclear. The relationship 

 of the Coral Sea population to that in the central Pacific and 

 other areas in the western Pacific is also not defined. Though no 

 recoveries have been made in these areas, some population inter- 

 change could be expected. 



Figure 16. — A hypothetical description of black martin migration in the southwest 

 Pacific Ocean. 



18 



