672 



Fishery Bulletin 93(4), 1995 





E 



41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 



Larval duration 



Recruitment date 

 Figure 6 



Frequency distribution of estimated larval duration (top 

 graph) and relationship between estimated hatching date 

 and recruitment date (bottom graph) of bonefish, Albula 

 vulpes, collected during winter 1993-94. 



through tidal passes and can help elucidate the pat- 

 terns of recruitment and mechanisms driving trans- 

 port. Although adult population sizes of highly mo- 

 bile fishes such as Albula vulpes have yet to be mea- 

 sured, it is likely that whatever variation does occur 

 in the abundance of adults is at least partially af- 

 fected by recruitment variability. 



A virtually universal pattern observed by studies 

 on daily recruitment of tropical marine fishes is the 

 association of peaks of recruitment with dark phases 

 of the lunar cycle (Pfeiler, 1984; Robertson et al., 

 1988; Robertson, 1992; Dufour and Gazlin, 1993; 

 Shenker et al., 1993; Thorrold et al., 1994, a and b). 

 This pattern may be a function of a lunar spawning 

 cycle, followed by a fixed larval duration. Alterna- 

 tively, it may be an active response of fishes, enabling 

 them to remain in the plankton until dark conditions 

 permit them to move onshore, thus enabling them to 

 avoid the "wall of mouths" of visual predators along 

 reef edges (Hamner et al., 1988) and at settlement sites. 



Our data suggest that A. vulpes follows the latter 

 strategy. Analysis of larval otoliths of winter recruits 

 (assuming that otolith increments are deposited 

 daily, beginning at hatching) indicates that these fish 

 spawn continuously from late October through De- 

 cember (Fig. 6). However, owing to preservation prob- 

 lems with some of the large recruitment pulses (>100 

 animals/night) in winter 1993-94, when otolith 

 analyses were performed, we cannot exclude the pos- 

 sibility that the level of spawning activity varies over 

 time. Significant spawning activity in the Bahamas 

 probably extends until spring, with the large pulse 

 of recruitment in late June 1992 presumably result- 

 ing from spawning in April and May. 



After hatching, leptocephali remained in the pe- 

 lagic environment of Exuma Sound for 41-71 days 

 (with a mean of 56 d) in winter 1993-94. Despite 

 this relatively broad range of larval duration, the 

 metamorphosing larvae exhibited a very strong cy- 

 clical recruitment pattern (Fig. 4) that was not con- 

 sistently related to meteorological conditions or cur- 

 rents measured along the shelf-edge seaward of the 

 sampling stations (Table 1). This response supports 

 the contention of Thorrold et al. (1994, a and b) that 

 various species can actively control their onshore 

 movements in certain environments, perhaps by de- 

 laying their metamorphosis until suitable environ- 

 mental conditions or opportunities develop (Welling- 

 ton and Victor, 1989). However, variable shrinkage 

 of leptocephali, due to different times between cap- 

 ture, death in the nets, and preservation, prevented 

 backcalculation of growth rates and a test of the abil- 

 ity of A. vulpes to delay metamorphosis. 



In series of ichthyoplankton surveys from Janu- 

 ary through February 1991, A. vulpes leptocephali 

 were found to be widely dispersed at night over a 

 transect extending from the shelf edge near LSI to 

 24 km offshore (Drass, 1992), indicating that some 

 larvae were always close enough to the shore so that 

 they might become entrained in the flood tides 

 crossing the narrow shelf. Despite the continuous 

 presence of leptocephali close to the coast, however, 

 their onshore movement was temporally restricted. 

 The occurrence of favorable low nocturnal illumina- 

 tion levels may be the parameter that limits cross- 

 shelf movement of larvae to specific "windows of op- 

 portunity." These windows of opportunity were de- 

 fined by the relationship between lunar and tidal 

 conditions. During bright, moonlit nights, recruit- 

 ment levels were very low (Figs. 3 and 5); recruit- 

 ment increased as larger amounts of night-time flood 

 tide occurred prior to the progressively later moon- 

 rise. Extremely little recruitment occurred on nights 

 when there were less than two hours of nocturnal 

 flood tide under moonless conditions, whereas the 



