Mojica et al.: Recruitment of Albula vulpes 



671 



der moonless conditions (Fig. 5). The 

 single peak in recruitment was not 

 associated with specific wind or cur- 

 rent patterns (data not shown); the 

 very low numbers of fishes collected 

 during following months precluded 

 use of correlation analysis to exam- 

 ine the relationship between recruit- 

 ment and environmental conditions. 



Larval ages and spawning 

 patterns 



Though the formation of daily otolith 

 increments for A. vulpes has yet to 

 be confirmed, the daily formation of 

 increments in Anguilla japonica lep- 

 tocephali has been verified (Ume- 

 zawa et al., 1989). We thus assumed 

 that increments are produced daily 

 in bonefish and that deposition of 

 increments begins at hatching. Given 

 these assumptions, the average lar- 

 val duration was 56 days, a range of 



Table 1 



Results of cross-correlation analyses between recruitment of A. vulpes and 

 along-shore and cross-shelf wind and current vectors for four winters. The 

 table shows all significant correlation coefficients between larval recruitment 

 and wind and current data, with a lag corresponding to the number of days 

 indicated. Positive correlations indicate signficiant relationships with trans- 

 port to the northwest (along-shore component) or offshore (cross-shelf compo- 

 nent!. Numbers below years in parenthesis indicate the number of days used 

 in each analysis. NS = not significant (a=0. 05 ). ID = insufficient current meter 

 data for analysis. 



Year 



1990-1991 



(72) 



1991-1992 

 (62) 



1992-1993 



(55) 



1993-1994 



(58) 



Wind 



Current 



Along-shore Cross-shelf Along-shore Cross-shelf 



NS 



0.28/0 days 



-0.22/3 days 



NS 



NS 



NS 

 NS 

 NS 



ID 



NS 

 ID 



NS 



ID 

 NS 

 ID 



NS 



D 



E 



''  '"'■ '"■■ ■■■'■■■  litfll J 



July August 



Summer 1992 



Figure 5 



Abundance of bonefish, Albula vulpes, larvae 

 collected each night from 24 June to 4 Septem- 

 ber 1992 (top graph) and hours of dark flood tide 

 during the sampling period (bottom graph). 



41-71 days (Fig. 6). Regression analysis of back-cal- 

 culated hatching date and recruitment date revealed 

 a strong relationship. Backcalculation of hatching 

 dates from otolith data indicated continuous spawn- 

 ing from mid-October through early January (Fig. 

 6). The maximum number of otoliths examined per 

 day was 7 for fish recruiting on 30 January. The back- 

 calculated spawning dates of these fish ranged from 

 16 November to 27 December. 



Discussion 



Variation in the recruitment of tropical marine fishes 

 is considered to be a dominant influence on the size 

 and distribution of adult populations. Doherty and 

 Fowler (1994) have shown that variations in the re- 

 cruitment of a reef-dwelling territorial pomacentrid 

 over a ten-year period could explain 909c of the varia- 

 tion in adult populations. Dramatic variability in 

 larval recruitment has been detected by light trap 

 surveys (e.g. Doherty, 1987; Milicich et al., 1992), 

 visual surveys (e.g. Sale, 1980; Robertson et al., 1988; 

 Robertson, 1992), and calculation of settlement pat- 

 terns from otoliths (e.g. Thresher et al., 1989; 

 Wellington and Victor, 1989). Recent work in the 

 Bahamas (Shenker et al., 1993;Thorroldet al., 1994, 

 a and b) and French Polynesia (Dufour and Gazlin, 

 1993) have shown that moored channel nets can be 

 used to monitor onshore larval movement directly 



