FISHERY BULLETIN: VOL. 86, NO. 3 



ber. Based on all the information, the major recruit- 

 ment may occur in spring with some additional 

 recruitment in August and September (Fig. 5); data 

 are not sufficient to determine the recruitment 

 pattern in detail. 



A broad geographic separation exists between the 

 location of pelagic captures of larger armorhead and 

 the location of the spawning populations (Figs. 2, 

 4). If recruitment to the seamounts occurs predom- 

 inantly in spring (as suggested for 1973 in Figure 

 5), then temporal sampling patterns may have 

 missed these fish, although JAMARC pomfret 

 surveys covered this area in some seasons (Table 1). 

 Locating the seamounts, which have small (2-5 km) 

 summits must be a formidable task given the wide 

 ocean areas over which armorhead are distributed. 

 A similar situation exists for rock lobster, Jasus 

 tristani, in the South Atlantic, that recruits from 

 an upstream population some 2,000 km away and 

 in sufficient numbers to support a fishery on Vema 

 Seamount in some years (Lutjeharms and Heydorn 

 1981). 



Open-ocean migrations of fishes may depend 

 upon many potential cues, including electric fields 

 (McCleave and Power 1978), magnetic fields 

 (Walker 1984), gyres (Williams 1972), and phero- 

 mones (Nordeng 1977). Certain characteristics of 

 these isolated, open-ocean seamounts may promote 

 their detection by armorhead. First, current-topog- 

 raphy interactions may create significant signals in 

 physical and biological features. The region of the 

 SE-NHR seamounts is active in front development 

 (Roden and Paskausky 1978); upwelling, eddies, and 

 other aspects of flow complexity also occur around 

 these seamounts (Roden et al. 1982) and down- 

 stream from them (Royer 1978). The biological sig- 

 nals may include increased chlorophyll in response 

 to upwelling or doming of isotherms (Genin and 

 Boehlert 1985), or aggregations of various organ- 

 isms and the larger animals which prey upon them 

 around seamounts (see review in Boehlert and Genin 

 1987). Gravity anomalies associated with seamounts 

 may also play a role; positive gravity anomalies exist 

 at the summit and slopes, and negative anomalies 

 are seen in the surrounding "moat" regions (Wedge- 

 worth and Kellogg 1987). Seamounts often have 

 strong magnetic dipoles associated with them, and 

 the dipole might serve as a landmark for magnetic 

 orientation by fish (Klimley^). While fish have been 

 shown to have magnetoreceptors (Walker et al. 

 1985), their use of magnetic maps remains specula- 



T. Klimley, Scripps Institution of Oceanography, La JoUa, CA 

 92038, pers. commun. 30 June 1987. 



tive but possible (Gould 1985). Although we cannot 

 postulate the mechanism that armorhead use for 

 recruitment, it is clear that the effects of seamounts 

 may be detected at distances greater than their area 

 alone would suggest. 



Interannual Variations in 

 Recruitment Strength 



The year-class strength of armorhead recruiting 

 to seamounts appears to be independent of the 

 parent stock size (Wetherall and Yong 1986). As an 

 example, Borets (1975) estimated that on the SE- 

 NHR seamounts from 1968 to 1973, the stock size 

 varied by a factor of <1.8 while recruitment varied 

 by >5.5 times. The relative abundance of armorhead 

 at the SE-NHR seamounts area increased in 1986 

 after a long period at a very low level. This increase 

 probably corresponds to the high abundance of 

 pelagic specimens captured in the northeastern 

 Pacific in 1985 (Fig. 2B; Table 2) that consisted of 

 two age groups. The increased recruitment at the 

 SE-NHR seamounts in 1986 suggests that environ- 

 mental conditions were favorable to the survival of 

 young armorhead in the 1984 and 1985 winter 

 seasons. 



A wide variety of factors, both biotic and physical, 

 can affect survival and ultimate year-class strength 

 in fishes (Lasker 1978). During the 2 years between 

 spawning and recruitment for armorhead, an ex- 

 tended migration through varied pelagic environ- 

 ments occurs (Fig. 4). Feeding conditions for larval 

 and juvenile stages are characterized by interannual 

 variability; Fedosova (1980) suggested that warm 

 years were more productive for zooplankton prey 

 and, thus, favorable to the survival of young armor- 

 head. Interannual variation in atmospheric systems 

 (Seckel 1988) or large-scale ocean currents of the 

 kind described by Mysak et al. (1982) may also play 

 a role in armorhead recruitment strength. Changes 

 in the position of the Alaska gyre by up to 700 km 

 southwest of its normal position may have occurred 

 from 1981 to 1985, with an associated increase in 

 seawater temperature (Royer and Emery 1987). 

 Large-scale atmospheric phenomena, such as the 

 longitudinal position of the Aleutian Low, may 

 create definite interannual variations in winter wind 

 systems that may be seen in surface current pat- 

 terns (Seckel 1988). These patterns may, in turn, 

 be related to the latitudinal position of the subtrop- 

 ical front, which varies interannually between lat. 

 28° and 32°N (Roden 1970). Variability in these 

 features influences surface drift (McNally 1981), 

 which in turn affects the neustonic young of armor- 



462 



