308 



Fishery Bulletin 92(2), 1994 



Summer 1990/91 



Photololigo sp. A 



Summer 1991/92 



19km 

 [n=171| 



24 km 

 [n=395] 



25 75 125 175 225 275 325 375 425 »45 " 75 125 175 22 5 27 5 325 375 42 5 »45 



MANTLE LENGTH (mm) 



Figure 3 



Size-frequency distribution of Photololigo sp. A and Photololigo sp. B caught at each 

 station (pooled across months) in deep (shaded) and shallow (unshaded) light-traps. Total 

 number of juveniles indicated in brackets. 



Table 5 



Results of the multiway frequency analysis to 

 examine changes in the size distribution of 

 Photololigo sp. A between time of night and depth. 



Source 



df 



Depth 



Time 



Depth x Time 



92.8 

 25.57 

 0.19 



0.00 

 0.00 

 0.66 



been validated as useful devices for monitoring rela- 

 tive abundance patterns in larval supply of pelagic 

 juvenile fish at fixed locations (Milicich et al., 1992). 



Great care needs to be exercised when interpreting 

 catch rates from different locations because changes 

 in water transparency can bias light-trap efficiency. 

 Similarly, it is not possible to quantitatively compare 

 catches from drifting and anchored light-traps 

 (Thorrold, 1992). This is because the former act as 

 lagrangian drifters and sample photopositive organ- 

 isms from within a constant light pool. In contrast, 

 the moored light-traps experience a variable water 

 flow that may greatly increase the volume of water 

 swept in an hour of sampling. Despite more inten- 

 sive sampling on the reefs, catches of Photololigo 

 were low and we conclude that spawning does not 

 occur near the reefs and that juvenile Photololigo 

 individuals are concentrated in the lagoon. In the 



