HALES: ROUND SCAD IN THE SOUTH ATLANTIC BIGHT 



ity for structure (Klima and Wickham 1971; Feder 

 et al. 1974; Hastings et al. 1976). 



Although fish sizes within each depth zone over- 

 lapped, the observed pattern was fairly consistent 

 and distinct for all cruises; adults composed the 

 catch in 9-18 m depths, and fish in deeper waters 

 (mostly juveniles) showed a positive size-depth corre- 

 lation. Adults migrate inshore in spring to feed and 

 spawn, and offshore in winter to avoid cold waters. 

 However, the apparent movement of juveniles to 

 deeper waters is not understood. Correlations be- 

 tween fish size and depth are numerous in aquatic 

 habitats (Helfman 1978), but explanations based on 

 changing physiological tolerances (Bullis and Struh- 

 saker 1970), foraging strategies (Polloni et al. 1979), 

 and predation responses (Hobson 1972) have been 

 difficult to demonstrate in most fishes. 



Reproductive Biology 



Although results (Fig. 3) indicate that D. punc- 

 tatiis spawn primarily from March through August, 

 spawning probably occurs through September and 

 to a lesser extent throughout the year. Collections 

 were not made during September and October of 

 1980, but water temperatures in August and Sep- 

 tember in the South Atlantic Bight are generally 

 similar (Atkinson et al. 1983). Round scad larvae 

 have been collected in winter in the South Atlantic 

 Bight (Fahay 1975; Powles and Stender 1976) and 

 the eastern Gulf of Mexico (Leak 1981). Larval oc- 

 currence has been correlated with water temper- 

 ature in the eastern Gulf of Mexico (Leak 1981), and 

 sufficient water temperatures (>20°C) occur 

 throughout the year in parts of the South Atlantic 

 Bight. 



The pattern of oocyte development is generally 

 quite variable and complex in serial spawners: 

 oocytes develop asynchronously or synchronously 

 in groups, and ova are released in batches. Three 

 observations suggested that round scad are serial 

 spawners: 1) the occurrence of three distinct modes 

 in the frequency distributions of oocyte diameters 

 from ripe ovaries collected in spring; 2) two modes 

 in those distributions from ripe fish collected in late 

 summer; and 3) evidence of spawning in ovaries hav- 

 ing a frequency distribution of oocyte diameters 

 similar to developing ovaries. Spawning was in- 

 dicated by disorganized ovarian septa with con- 

 spicuous spaces, debris in the ovarian lumen, 

 residual atretic oocytes, and brown bodies. 



Although estimates of fecundity in round scad are 

 comparable to those of D. pinnulatiis (Yamaguchi 

 1953), D. macrosoma, and D. rtisselli of similar size 



(Tiews et al. 1970), the conventional method applied 

 here probably underestimated fecundity. Because 

 gonads used in fecundity estimation were not ex- 

 amined histologically, it was not possible to deter- 

 mine if spawning had occurred recently in specimens 

 used for fecundity measures. In addition, sufficient 

 numbers of specimens were not examined to deter- 

 mine spawning frequency from running-ripe 

 (DeMartini and Fountain 1981) or postovulatory 

 (Hunter and Goldberg 1979) females. Previous 

 studies on other serial spawning fishes (Hunter and 

 Goldberg 1979; Hunter and Leong 1981; DeMartini 

 and Fountain 1981; Conover 1985) have shown that 

 estimates of annual fecundity (total number of ova 

 spawned in 1 year) can differ from conventional 

 fecundity estimates (which ignore multiple spawn- 

 ing) by an order of magnitude. Serial spawning 

 fishes generally have low relative ovary weights 

 (Martinez and Houde 1975; Smith and Lasker 1978; 

 DeMartini and Fountain 1981), but can expend over 

 100% of their body weight per year in eggs (Hubbs 

 1976; DeMartini and Fountain 1981). If observed 

 fecundity in round scad (6,200-51,000) is extra- 

 polated from the 4.3% relative ovary weight (Table 

 7) to total body weight, then fecundity estimates of 

 142,000-1,173,000 would result. If observed fecun- 

 dity is divided by the proportion of oocytes in the 

 most advanced developmental mode (32%, from 

 Figure 3), then batch fecundities of approximately 

 2,000-16,000 per female (130-230 eggs/g body 

 weight) and annual fecundity estimates (based on 

 10 d spawning cycle for 6 months) of 36,000-288,000 

 would result. Both estimates are entirely specula- 

 tive, but support the contention that the conven- 

 tional method underestimated fecundity, and 

 emphasize the need for additional studies on the 

 fecundity of round scad. 



Round scad mature at a smaller size than reported 

 for other species of Decapteriis, which reach matur- 

 ity at 18-20 cm (Yamaguchi 1953; Tiews et al. 1970). 

 The small size at which round scad become sexual- 

 ly mature suggests that they are under strong selec- 

 tion pressure to mature rapidly. The natural mor- 

 tality of round scad in the eastern Gulf of Mexico 

 is high (Houde et al. 1983). Compared with tem- 

 perate and boreal species, many tropical clupeoids 

 also mature at small sizes, seldom attain large size 

 and have high adult mortality rates (Blaxter and 

 Hunter 1982; Houde et al. 1983). 



Feeding Habits 



Zooplanktivores feed during the day or at night, 

 but seldom during both periods (de Silva 1973; Hob- 



263 



