638 



Fishery Bulletin 89(4), 1991 



Discussion 



Red drum larvae ranging in mean size from 1.7 to 

 5.0 mm were vertically stratified in waters of less than 

 25 m of the northcentral Gulf of Mexico. Larvae were 

 most frequently found to be concentrated higher in the 

 water column during the day than at night in vertical- 

 ly well-mixed coastal and inner shelf waters. The lack 

 of a consistent correspondance between depth of max- 

 imum larval red drum abundance and depth of max- 

 imum prey microzooplankton abundance was not un- 

 expected since, in general, high prey densities (MOO 

 organisms/L) were observed throughout the water 

 column. Brewer and Kleppel (1986) likewise found no 

 correspondance between prey densities and the day- 

 time vertical distribution of northern anchovy and 

 white croaker larvae. The time lag (2-3 hours) between 

 the collection of ichthyoplankton and zooplankton 

 samples may have affected our results if the spatial 

 coincidence of larvae and prey occurred over a shorter 

 time interval than the interval between samples. 



It is feasible to assume that red drum larvae main- 

 tain this pattern of vertical distribution by responding 

 to diel changes in illumination. In their recent review, 

 Neilson and Perry (1990) concluded that light is impor- 

 tant in mediating diel vertical migration in many fishes 

 throughout ontogeny. Diel variations in the distribu- 

 tion patterns of young fishes are associated with ac- 

 tivity, phototaxis, and brightness discrimination (Blax- 

 ter 1969). Blaxter (1973) was able to induce vertical 

 migration in herring and plaice larvae by varying am- 

 bient light levels in the laboratory or by exposing larvae 

 to natural conditions of dawn and dusk. We have 

 observed that 3-5 day-old red drum larvae in the 

 laboratory are positively phototactic and tend to con- 

 centrate at the very surface of 4 L rearing containers 

 with artifical lighting directly overhead. Larvae were 

 not observed under dark conditions. The range in ver- 

 tical movement, 5-12m, required to maintain the diel 

 distribution pattern observed in this study is theoret- 

 ically possible if the swimming capabilities of 2-5 mm 

 drum larvae are similar to those of other fish larvae, 

 i.e., with cruising speeds of 2-3 body lengths per se- 

 cond (Blaxter 1969, Theilacker and Dorsey 1980). The 

 role of the swimbladder, clearly visible in red drum 

 larvae by at least 2mm, in "depth holding" (Blaxter 

 1986) is not known. A comprehensive explanation of 

 how red drum larvae maintain vertical position in the 

 shallow depths of their early nursery grounds requires 

 a more thorough knowledge of early sensory and loco- 

 motor capabilities in this species. 



The pattern of diel vertical distribution displayed by 

 red drum larvae, implying nocturnal descent, has been 

 termed "reverse diel vertical migration" and has been 



observed by various workers (Wood 1971, Richards and 

 Kendall 1973, Ohman et al. 1983, Kuwahara and Suzuki 

 1984, Boehlert et al. 1985, Yamashita et al. 1985, Leis 

 1986, Sogard et al. 1987). From their review of the 

 literature on vertical migration among fishes, Neilson 

 and Rerry (1990) found this pattern to be less common 

 than nocturnal ascent, often associated with similiar 

 movements of prey species, and implicated in reduc- 

 ing predation pressure. Yamashita et al. (1985) at- 

 tributed this distribution pattern among Japanese sand 

 lance larvae to daytime feeding activity in the upper 

 levels of the water column. The authors suggested that 

 when feeding stops after dark, larvae become relatively 

 inactive and gradually sink to greater depths; with the 

 return of daylight, feeding resumes and larvae move 

 upward. 



Activity levels among fish larvae are associated with 

 diurnal light changes and phototaxis (Blaxter 1969). 

 The threshold light intensity found to initiate vertical 

 movement of herring and plaice larvae in the laboratory 

 was similiar to the light threshold for feeding (Blaxter 

 1986). It would be advantageous for larvae such as red 

 drum living in turbid coastal waters to move into more 

 brightly illuminated levels to feed, given the poor visual 

 acuity and limited retinomotor response capabilities 

 (light/dark adaptation) found in the early larvae of most 

 marine fishes (Blaxter 1969). 



Prior to this study, vertical distribution of red drum 

 larvae in the Gulf of Mexico had only been described 

 for estuarine/bay systems. Peters and McMichael 

 (1987) never found red drum larvae at the surface dur- 

 ing daytime hours in 1 m plankton net collections in 

 Tampa Bay, Florida, where red drum larvae were 

 found to be most abundant nearest the bay mouth. Col- 

 lections at the bay mouth, however, were taken only 

 at night during a flood tide. The most extensive data 

 on vertical distribution of red drum larvae comes from 

 Aransas Pass tidal inlet on the southern Texas coast 

 where vertical position appears to be influenced by 

 direction of tidal flow (Holt et al. 1989). Further obser- 

 vations and data indicate that high surface densities 

 are generally associated with flooding conditions and/or 

 oceanic water, while high bottom densities occur dur- 

 ing ebbing conditions and/or in bay water (S.A. Holt, 

 Mar. Sci. Inst., Univ. Texas at Austin, Port Aransas, 

 TX 78373, pers. commun. Aug. 1990). Tides along the 

 south Texas coast are principally diurnal, and during 

 the red drum spawning season, the flood tide general- 

 ly occurs at night while the ebb tide occurs during the 

 day. 



The relative proportion of the larval red drum popula- 

 tion occurring below maximum sampling depths dur- 

 ing our study is unknown, but presumed to be small 

 (Comyns et al. 1991). Observations taken during a year- 

 long survey of Mississippi Sound and adjacent coastal 



