Warlen et al ; Recruitment of larval Bievoortia tyrannus to North Carolina and New Jersey estuaries 



611 



Figure 1 



Location of sampling sites for larval Atlantic menhaden (.Brevoortia tyrannus) collected at Little 

 Sheepshead Creek, New Jersey, and Fivers Island, North Carolina, during 1989-90, 1990-91, and 

 1992-93. 



and approximated those measured in North Carolina. As in 

 North Carolina, samples were preserved in 95% ethanol. 



We assumed that larval fish densities within and among 

 years at a site were comparable (see Witting et al., 1999): 

 therefore year-to-year relative abundances could be esti- 

 mated at that site. However, differences between sites are 

 not easily compared because the catch efficiency of the two 

 types of passive nets is not known and there is no indepen- 

 dent estimate of absolute abundance at either site. 



Age, growth, and birthdate determination 



Larvae were randomly subsampled from individual weekly 

 net sets at both locations in proportion to their contribu- 

 tion to the total nightly catch. All larvae in catches up to 20 

 fish were used. In catches of >20 fish, subsample sizes were 

 proportional to catch but generally no more than 50 fish 

 were aged per week from any one site. Experienced otolith 

 readers at the Beaufort Laboratory determined the ages of 

 1435 larvae from North Carolina and 444 from New Jersey 

 for our study. Larvae were measured to the nearest 0. 1 mm 

 standard length (SL) with an ocular micrometer Estimated 

 age was the number of sagittal otolith growth increments 

 (Maillet and Checkley, 1990) plus an empirically derived 

 value for the number of days (five) from spawning to first 

 increment formation (Warlen, 1992). Additionally, late- 

 larval to early-juvenile Atlantic menhaden reared in the 



laboratory through the winter at ambient water tempera- 

 tures, formed daily otolith growth increments even when 

 temperatures declined to 3°C (Ahrenholz et al., 2000). 

 We assumed that the age at initial increment deposition 

 in larval otoliths did not vary and that otolith increment 

 deposition rate was constant within and between sampling 

 seasons and the two sampling sites. 



To reduce the chance of underestimating ages of lar- 

 vae, we conducted a detailed examination of the otolith 

 microstructure of larvae collected in the 1990-91 season. 

 Otoliths from fish with our assigned January-February 

 hatching dates were re-examined and their growth incre- 

 ments measured on an image analysis system. One reader 

 made all counts and measurements in close consultation 

 with two other experienced otolith readers. In contrast 

 to the North Carolina larvae. New Jersey larval otoliths 

 had areas in the middle portion of the counting path with 

 narrower increments. However, New Jersey larvae did not 

 have checks on their otoliths, which indicated that otolith 

 growth did not stop. The otolith radius, standard length, 

 and age relations suggested that New Jersey fish grew 

 more slowly than North Carolina fish but did not indicate 

 a systematic underestimation of the age of New Jersey 

 fish due to narrow increments. 



Average growth of larvae was described by the Laird 

 version (Laird et al., 1965) of the Gompertz growth equa- 

 tion (Zweifel and Lasker, 1976). The model was fitted to 



