442 



Fishery Bulletin 103(2) 



did not appear to peak at some intermediate current 

 level. These results, however, represented the total 

 catch of all fishes, and the relationship between cur- 

 rent speed and light trap catches may be more taxon 

 specific (Doherty, 1987). When analyzed at the family 

 level, a bell-shaped relationship may have occurred for 

 Clupeidae, Engraulidae, and Blenniidae; however, the 

 pattern was indistinct and there was generally little 

 difference among families. 



The lack of any strong differences in the relationship 

 between light trap CPUEs and current speed among 

 the dominant families was unexpected, considering 

 the potential differences in swimming abilities. Be- 

 cause larvae and juveniles of demersal fishes are gener- 

 ally believed to have lower swimming speeds (Blaxter, 

 1986), it was anticipated that catches of synodontids 

 and blenniids would have been more negatively affected 

 by increasing current speed than relatively stronger- 

 swimming pelagic taxa (e.g., scombrids and carangids). 

 Perhaps larvae of demersal taxa have greater swim- 

 ming capabilities than previously considered, as has 

 been recently found for certain settlement-stage larval 

 reef fishes (sustained swimming speeds of 20-60 cm/ 

 sec; Stobutzki and Bellwood, 1994; Leis and Carson-Ew- 

 art, 1997). However, despite possible strong swimming 

 abilities, few larval and juvenile demersal or pelagic 

 fishes were collected at current speeds >40 cm/sec, and 

 of these the majority were preflexion larvae that were 

 undoubtedly passively entrained in the light trap. It is 

 possible that the larvae and juveniles of taxa collected 

 at platforms were unable to maintain the metabolic 

 power required to swim against the stronger currents 

 over extended distances from the light trap (Fisher and 

 Bellwood, 2002). 



Currents may have interfered with the functioning of 

 the light traps. Assuming that larval and juvenile fishes 

 were able to swim against the stronger currents, their 

 ingress into the light trap may have been impeded by 

 turbulence created by the current flow around the trap. 

 If turbulence occurred after some critical current speed, 

 then this may explain the lower CPUEs beginning at 

 around 30 cm/sec observed for each of the dominant 

 families. 



Higher turbidity also appeared to have a negative ef- 

 fect on light trap catches at platforms. Light trap catch 

 efficiency should be greatly impaired by highly turbid 

 waters because greater light attenuation would reduce 

 the effective sampling radius of the trap. In addition, 

 the phototactic response of larval and juvenile fishes 

 may be lower at lower light intensities (Gehrke, 1994; 

 Stearns et al., 1994). However, it is uncertain whether 

 the relatively small range of turbidities (0.1-2.6 NTU) 

 sampled during this study would result in a significant 

 decrease in light trap catch efficiency, particularly given 

 the intensity of the light source used (250,000 candle- 

 power). The observed patterns may have been a reflec- 

 tion of intrusions of turbid coastal and Mississippi River 

 plume water at the platforms, during which light trap 

 catches comprised large numbers of coastal clupeids and 

 relatively few other taxa (Fig. 5). 



Although they were treated separately for the purpos- 

 es of this study, the effects of current speed and turbid- 

 ity also may have been interrelated. A positive relation- 

 ship between turbidity and current speed was found for 

 a limited data set where both variables were available 

 (r-=0.28, P<0.0001). It is unlikely that this relationship 

 was caused by the resuspension of benthic sediments, 

 given the water depth at the platforms (20-230 m), but 



