BOEHLERT and SASAKI; PELAGIC BIOGEOGRAPHY OF THE ARMORHEAD 



surface flow is in this direction in summer months 

 (McNally et al. 1983). 



The two available studies on larval distribution of 

 armorhead are conflicting and suggest that patterns 

 of movement may differ from year to year. Komra- 

 kov (1970) conducted surveys in March-April 1969 

 and observed larvae only at northern seamounts and 

 in the open ocean several hundred kilometers north- 

 east of the SE-NHR seamounts. Borets (1979), how- 

 ever, observed 5-20 mm larvae remaining mostly 

 around the seamounts, with highest abundance 

 south of lat. 33°N in 1976. Because spawning oc- 

 curs at seamounts northwest of this area, this pat- 

 tern is consistent with southeastward drift during 

 the first few months of life (Fig. 3). If larvae remain 

 in the seamount region, they could be transported 

 to the northeast in summer months, or they may 

 actively migrate northeastward after reaching the 

 juvenile stage. A similar migration apparently 

 occurs for the pomfret, Bramajaponica, which prior 

 to spawning moves into the North Pacific current 

 region in a broad latitudinal band between the sub- 

 tropical convergence and the subarctic boundary; 

 immature fish and spent adults then move north- 

 ward into the subarctic region to feed during April- 

 May (Shimazaki and Nakamura 1981). The pomfret 

 differs from the armorhead, however, in that it 

 remains pelagic throughout its life and also is abun- 

 dant in the western North Pacific. Larval and 

 juvenile Brama sp. do occur in the same area as lar- 

 val armorhead (Borets and Sokolovsky 1978; G. W. 

 Boehlert unpubl. data). 



A possible scenario of movements after the lar- 

 val stage can be deduced from the ages of individuals 

 captured in different regions. Based upon the 

 growth of pelagic armorhead (Uchiyama and Sam- 

 paga fn. 3), most of the larger specimens indicated 

 in Figure 2 are from 1 to 2 years old, and the pelagic 

 duration for this species is most often from 1.5 to 

 2.5 years. Using daily growth increments for lar- 

 val and small juveniles and fitted growth curves for 

 larger juveniles and adults, one can convert lengths 

 to ages. The pelagic specimens of known length 

 range in estimated age from 0.04 to >2 years (Table 

 2). The youngest animals are typically found in the 

 region of the seamounts (Komrakov 1970; Fedosova 

 and Komrakov 1975; Borets and Sokolovsky 1978), 

 but intermediate-sized fish are found north and east 

 of the seamounts. Different age groups seem to be 

 distributed in different areas in the eastern North 

 Pacific (Fig. 4). Armorhead spawn in the seamount 

 region (as area A in Figure 4) during November to 

 March, and juveniles 5-25 mm long are found in the 

 SE-NHR seamounts area from February to early 



April (Fedosova and Komrakov 1975; Borets 1979). 

 Subsequent occurrences of fish older than 0.25 year 

 (Table 2) are found only in the region northeast of 

 the seamounts and in the Gulf of Alaska (area B in 

 Figure 4). Northeastward movements are contrary 

 to mean ocean currents, but Roden et al. (1982) 

 described intense northward flow along the axis of 

 the southern Emperor Seamounts that could con- 

 ceivably play a role in transport of fish at the young- 

 est stages. 



Recruitment of later stage armorhead to the sea- 

 mounts requires movement to the west. The only 

 significant westward circulation in the North Pacific 

 is found in the Alaskan Stream (Favorite 1967). This 

 narrow current occurs north of lat. 51 °N and ex- 

 tends westerly to near long. 170 °E, where it splits 

 to two branches, one turning northerly into the 

 Bering Sea and the other southerly under the influ- 

 ence of the northern Emperor Seamounts or Kor- 

 mandorskie Ridge (Favorite 1967; dashed line in 

 Figure 4). Damitsky et al. (1984) described southerly 

 movement of this water to at least lat. 40 °N and 

 suggested that it played a role in transport of 

 planktonic food for armorhead on the seamounts. 

 Unless returning armorhead are deep in the water 

 column and therefore not sampled, distributional 

 data (Fig. 2) do not support this route for westerly 

 movement. Fish near recruitment size and age are 

 instead captured in the region marked C on Figure 

 4, south of the area where younger fish are captured. 

 These fish apparently migrate back to the region of 

 the seamounts by some unknown, but probably 

 active, mechanism. 



Two less important, but plausible, movement 

 patterns may explain the rarer occurrences in south- 

 ern Japan, off Oregon and California, and in the 

 Hawaiian Archipelago (Fig. 4). Chelton (1984) dis- 

 cussed possible interannual changes in the lati- 

 tudinal position of the West Wind Drift which may 

 lead to differing magnitudes of transport in the 

 Alaska and California Currents. The rare benthic 

 occurrences of the pelagic armorhead off the west 

 coast of North America (area D in Figure 4) may 

 come from specimens that drift southward in the 

 California Current in years of greater southern 

 transport. Some of these may not settle out and re- 

 main within the subtropical gyre, possibly recruit- 

 ing to the Hawaiian chain far to the south of the 

 normal reproductive population. This route, because 

 of its distance, apparently takes considerably longer 

 than that in the subarctic gyre; based upon satellite 

 drifters, McNally et al. (1983) suggested that a full 

 circuit of the subtropical gyre takes 4.5 years. The 

 armorhead collected in the Northwestern Hawaiian 



459 



