434 



Fishery Bulletin 98(2) 



Figure 4 



Sea-surface-temperature .satellite imager>' dunng (A) 12 and, 

 (Bi 15 and, (C) 16 September 1994 off North Carolina. Station 

 locations are referenced. See Figure 1 for station numbers. 



rence of these larvae appeared to be similar on both 

 days as was the presence of Gulf Stream water. 



On 16 September 1994 (day 2), hydrographic con- 

 ditions had changed markedly offshore (Fig. 5). At 

 the surface of station 2, warm, low-salinity water 

 was present, indicating water of coastal origin. But 

 the temperature was higher than observed over the 

 Carolina shelf in the SST imagery (Fig. 4), indicat- 

 ing a more southern coastal origin for this water. 

 Accompanying these changes in outer-shelf waters, 

 the length-frequency distributions of the two most 

 abundant taxa, Bothiis sp(p). and S. papillosum, dif- 

 fered significantly between day 1 and day 2 (Knaskal- 

 Wallace Test; P<0.01 for both taxa). Very small larvae 

 of both middle-shelf taxa were abundant in outer- 

 shelf waters on day 2 but absent on day 1 (Fig. 6). 



Discussion 



Gear comparison 



A series of pelagic larvae and juveniles was collected 

 with complementary gear for Aca «//!«//•« s sp(p).,An^;- 

 gonia sp(p)., Bothits sp(p)., Scaridae type A, and 

 Selar crumenophthalmus. The Methot frame trawl 

 was not successful in collecting large larvae or early 

 juveniles of economically important reef fish species 

 (snappers and groupers), probably due to the low 

 abundance of adult populations. In other studies, the 

 frame trawl has successfully collected large larvae 

 and juveniles of those species whose adults are 

 extremely abundant (i.e. Theragra chalcogramma 

 [Shima and Bailey 1994; Brodeur et al., 1995]) and 

 Engraulis mordax, [Methot, 1986]). Economically 

 important reef fish species are relatively rare and 

 larger size pelagic larvae and juveniles are rarer 

 than young larvae. Longer frame-trawl tow times 

 (one hour) will be required, but this would limit 

 the extent of the area to be sampled and decrease 

 the quality of the specimens. The bongo sampler is 

 generally limited in its ability to collect late-stage 

 larvae or early pelagic juveniles owing to net avoid- 

 ance (Shima and Bailey, 1994); however, it is a more 

 effective sampler than either the neuston net or 

 frame trawl in collecting early stage larvae, and it is 

 needed to obtain information on spatial and tempo- 

 ral spawning. 



The frame trawl is effective in capturing lepto- 

 cephali, and because the leptocephalus is a true oce- 

 anic form regardless of habitat as an adult (Smith, 

 1989), it is a valuable indicator species in studies 

 dealing with cross-shelf transport. Similarly, the 

 frame trawl is effective in capturing Trichopsetta 

 vcutralis, a resident of the Gulf of Mexico (Robins 



