NOTE Chen et al.: Feeding and growth of larval Etrumeus teres and Brevoortia patronus 



185 



ing reared in the laboratory at 24-26°C (Miller et al. 

 1979), as well as for gulf menhaden reared at 18-22°C 

 (Hettler 1984). Further, we used alternative empirical 

 methods to support our assumption that otolith growth 

 increment formation occurs daOy (Hales 1987). By com- 

 paring the width of marginal increments with the width 

 of the proximal completely-formed increment, we 

 determined the percentage of larvae with partially- 

 formed or completely-formed marginal increments over 

 a 24-hour period (8-10 December 1981; Table 1). The 

 frequency of increment formation was inferred from 

 these percentages and from the relationship of otolith 

 radius and larval length. 



The Laird version of the Gompertz growth model was 

 used to describe growth from the logarithm, of length 

 and the estimated age of larvae (Zweifel and Lasker 

 1976). Growth curves of round herring and gulf men- 

 haden larvae were compared by using the predictive, 

 resampling method described by Kappenman (1981). 

 Data for gulf menhaden growth were taken from 

 Warlen (1988) for comparisons with the growth of 

 round herring. 



Results 



Distribution and co-occurrence 



In all, 419 round herring larvae were identified in the 

 present collections, four fewer than gulf menhaden 

 (Sogard et al. 1987). Collections of the larvae of both 

 species indicate that they co-occur infrequently. Round 

 herring and gulf menhaden larvae occurred together 

 at 15 of 45 locations where collections produced either 

 species. Larval round herring were collected most fre- 

 quently throughout the water column at the offshore 

 stations in water 91m deep, although one of the largest 

 single collections was made at 18 m (Table 1). Larval 

 gulf menhaden were collected mostly inshore at the 

 18 m stations along each transect. The larvae of these 

 species co-occurred mainly at the 91m stations along 

 each transect. 



Diet comparisons 



Only 56 round herring larvae had food in their guts. 

 Larval round herring had eaten primarily copepod 

 nauplii, copepodites, and adults, with pteropods (mainly 

 Limacina trochiformis), tintinnids, invertebrate eggs, 

 and Eucalanus spp. nauplii contributing lesser percent- 

 ages (Table 2). Eucalanus nauplii were considered a 

 discrete food organism separate from other copepod 

 nauplii, because its form and size differed markedly; 

 Eucalanus spp. nauplii have long, paddle-like appen- 

 dages and are more than three times larger than the 

 other copepod nauplii observed in the guts of larvae. 



The width of food organisms ranged from 40 to 280 f^m, 

 a width range comparable to that found for gulf men- 

 haden (Govoni et al. 1983). 



Of the 88 round herring larvae collected simultan- 

 eously with gulf menhaden, 26 had food in their guts. 

 There were differences in the gut contents of these 26 

 round herring and 26 randomly selected gulf menhaden 

 larvae collected simultaneously (Table 3). Larval round 

 herring had eaten cylopoid copepods {Oncaea spp. and 

 Corycaeus spp.) and pteropods more frequently, but 

 calanoid copepodites and adult copepods less frequent- 

 ly, than had larval gulf menhaden. Percent similarity 

 of the diets of these larvae was 52.2, a value that in- 

 dicates marginal overlap in diet (Schoener 1970). 



