Mojica et al.: Recruitment of Albula vulpes 



669 



Results 



Winter environmental conditions 



Winds in Exuma Sound during the four winters gen- 

 erally blew from the east or southeast; there is a con- 

 siderable cross-shelf component of motion across the 

 shelf edge that runs northwest-southeast (Fig. 2). The 

 predominant along-shore component of the wind was 

 towards the northwest. Passage of occasional storm 

 fronts or cold fronts with winds from the northeast was 

 characterized by cross-shelf winds at velocities exceed- 

 ing 5 m-sec -1 and by along-shore flow to the southeast. 



Currents on the outer edge of the shelf flowed pre- 

 dominantly along the shelf toward the northwest 

 with only a weak cross-shelf component of motion 

 (Fig. 2). Flow onto the shelf, and occasional current 

 reversals along the shelf to the southeast were typi- 

 cally associated with the passage of storm fronts 

 (Shenkeretal., 1993). 



The hours of dark flood tide cycled throughout the 

 study, ranging from to 7 hours per night (Fig. 3). 

 The nights with hours of dark flood tide occurred 

 when the full moon was visible during the entire 

 night flood tide. During the week after each full moon, 

 moonrise became progressively later each night, and 



the amount of flood tide occurring between sunset 

 and moonrise increased rapidly. 



Temporal patterns of recruitment during 

 winter 



A total of 3,079 A. vulpes leptocephali were collected 

 during the 277 nights of sampling over four winters 

 (Fig. 3). Recruitment levels varied greatly among 

 years; a low of 316 leptocephali were taken the first 

 winter and 1,421 during the third winter. Several 

 large peaks in recruitment, reaching a maximum of 

 190 fish/night, were detected during the third and 

 fourth winters. Over all four years, 90.2% of the lep- 

 tocephali were captured by the nets fishing the up- 

 per 1 meter of the water column. 



Periodogram analysis of A. vulpes recruitment in- 

 dicated a very strong cycle with a period of 30.7 days, 

 which suggests a lunar or tidally influenced cycle 

 (Fig. 4). Autocorrelation of recruitment data (Fig. 4) 

 also identified cycling centered around 30 days. 

 Cross-correlations between recruitment and hours 

 of dark flood tide (Fig. 4) showed a strong positive 

 association on nights with a high degree of flood tide 

 occurring during moonless portions of the night and 

 a negative relationship centered around the full moon. 



Cross-shelf wind 



Along-shore wind 



199091 1991-92 1992-93 



Cross-shelf current 



1990-91 1991-92 1992-93 1993-94 



Along-shore current 



' Y W 



30 



20 



10 







-10 

 -20 

 -30 



Figure 2 



Cross-shelf (above line=movement offshore, below line=movement onshore land along-shore (above 

 line=to northwest, below line=to southeast) components of motion for wind and currents at Lee 

 Stocking Island. Dotted vertical lines indicate where years are joined by lunar date. Solid bars on 

 abscissae indicate when no data were available. 



