824 



Fishery Bulletin 101(4) 



80° 30' 



80° 00 



80° 30' 



80° 00' 



79° 30' 



Figure 1 



The South Carolina coast with the locations of estuaries and major river basins where striped mullet were collected 

 from 1998 to 2000. 



spawning fish, as well as previous models of reproductive 

 development (Stenger, 1959; Wallace and Selman, 1981) 

 (Table 1). This evaluation method was based on identifica- 

 tion of morphological characteristics evident in histological 

 sections. Each specimen was evaluated by two (in some 

 cases three) readers and discrepancies between readers 

 were either resolved or the specimen was excluded from 

 the analysis. 



A gonadosomatic index (GSI) was calculated for each 

 specimen following the method of Render et al. (1995) 

 where GSI was expressed as gonad weight (GW) divided 

 by body weight (BW) such that 



GSI = (GWIBW) X 100. 



The GSI values of the fecundity specimens were compared 

 among the three sampling years, as well as with GSI values 

 for female striped mullet collected during the rest of the 

 year that were mature but not reproductively active. 



Fecundity determinations were made from a total of 129 

 advanced-stage developing ovaries; 50 from 1998, 37 from 

 1999, and 42 from 2000. All of the ovaries were determined 

 by histological examinations and criteria outlined in Table 

 1 to be actively vitellogenic, having tertiary yolk-stage 



oocytes. All oocytes counted for fecundity were 400 /jm 

 or larger and in the tertiary yolk stage (Pien and Liao, 

 1975) (Fig. 2). This 400-/jm threshold has also been used 

 in other studies of oocyte development in striped mullet for 

 determining the point at which the oocytes that would be 

 spawned during that year were identifiable (Shehadeh et 

 al., 1973a). Unlike other species (particularly batch spawn- 

 ers), where fecundity counts should ideally be conducted 

 by using hydrated oocytes only, the striped mullet oocytes 

 used for fecundity counts were still presumably several 

 weeks from hydration and spawning. However, because 

 mullet are synchronous spawners, it is relatively easy to 

 distinguish the developing oocytes from the undeveloped 

 ones because of the drastic difference in size between the 

 two, as well as by the uniformity in size of the developing 

 oocytes once they reach 400 pm. 



Fecundity was estimated by using a modified gravimet- 

 ric method. The fixed whole ovary was patted dry and re- 

 weighed. The ovarian lobes were divided into four discrete 

 regions along each lobe's longitudinal axis and three sub- 

 samples (chosen at random) were taken between the two 

 lobes and prescn'ed in 50% isopropyl until oocyte counts 

 could be conducted. The subsamples ranged in weight from 

 0.025 g to 0.033 g. The subsamples from each specimen 



