FISHERY BULLETIN: VOL. 85, NO. 2 



ing the fall period of prespawning ovarian recrudes- 

 cence. Our failure to collect any partially spawned 

 females is also consistent with a single seasonal 

 spawn. 



In this study, both the GSI and the LOD proved 

 to be adequate, although not completely satisfac- 

 tory, indicators of the reproductive condition of 

 female striped mullet. However, of these two indices 

 the LOD would appear to be preferable. Determina- 

 tion of the LOD requires only the biopsy of a small 

 piece of ovary (see Shehadeh et al. 1973) which can 

 be easily accomplished without harm to the fish, 

 while determination of the GSI requires the sacrifice 

 of the fish. Furthermore, the validity of the GSI has 

 been questioned (deVlaming et al. 1982) as to its 

 accuracy in correcting for body size in a consistent 

 manner over all reproductive stages. 



On the other hand, the speed and ease of obtain- 

 ing the LOD are its only advantages over a more 

 comprehensive indicator of reproductive condition 

 —the oocyte size-frequency profile. Such a profile 

 also requires only the biopsy of a small piece of ovary 

 and is a much more accurate indicator of ovarian 

 stage, especially during the active vitellogenic 

 growth of the ovary when a developing clutch may 

 be quite spread-out in size. 



An adequate understanding of the functional rela- 

 tionship between oocyte size and stage is, of course, 

 required for correct interpretation of either LODs 

 or oocyte size-frequency profiles. Of particular in- 

 terest to us during this study were the sizes of the 

 oocyte at 1) the beginning of vitellogenesis and 2) 

 the prematuration stage of development. Our data 

 indicate that oocytes of the striped mullet are able 

 to grow to a point immediately prior to vitellogenic 

 growth, then temporarily arrest at that stage. In 

 contrast, once vitellogenic growth begins, further 

 development leading to a subsequent clutch of 

 mature eggs is apparently ensured. Thus knowledge 

 of this transition point is extremely important to in- 

 vestigators attempting to predict the future repro- 

 ductive status of these fish. Likewise, clearly iden- 

 tifying the prematuration stage of oocyte is impor- 

 tant because at this stage the oocyte is competent 

 to resume meiotic maturation culminating in the for- 

 mation of a fertilizable egg. 



We define the beginning of vitellogenesis in the 

 striped mullet oocyte by the initial appearance of 

 yolk proteins detectable by electrophoretic tech- 

 niques and the prematuration stage of development 

 by in vitro culture techniques. Our resulting pre- 

 maturation stage is in essential agreement with the 

 "functional maturity" stage of Kuo et al. (1974b), 

 as is our 0.18 mm initial vitellogenic stage with the 



initial "yolk globule" stage (0.20 mm) of these 

 authors. We did not examine smaller oocytes in 

 detail and thus did not attempt to establish specific 

 stages for previtellogenic oocytes. 



The oocyte size-frequency profiles of this study, 

 plus the additional data relating oocyte sizes and 

 stages, demonstrate that M. cephalus has a group- 

 synchronous type of oocyte development, as ori- 

 ginally defined by Marza (1938) and reiterated by 

 Wallace and Selman (1981). In such an ovary, a 

 single developing clutch of oocytes coexists with an 

 apparently asynchronous pool of previtellogenic 

 oocytes. However, it must be pointed out that in the 

 mullet this pattern is not always straightforward. 

 The movement of oocytes into vitellogenesis is quite 

 prolonged in time, so that a truly discrete clutch, 

 as characterized by the cessation of additional 

 recruitment into the clutch, does not become ap- 

 parent until the more developed oocytes within the 

 clutch are already well into vitellogenic growth. Fur- 

 thermore, even after recruitment into vitellogenesis 

 ceases, the developing clutch can be quite hetero- 

 genous in size (see oocyte size-frequency profile for 

 ovarian stage Ilia, Figure 6) and thus may not ap- 

 pear to be undergoing synchronous growth until 

 ovarian stage Illb. Such a pattern of oocyte devel- 

 opment could be very difficult to characterize with- 

 out examination of ovaries from females collected 

 throughout the period of ovarian recrudescence. 



To the best of our knowledge, the ovary staging 

 system put forth in this paper is the only such com- 

 prehensive system developed specifically for the 

 striped mullet. Based on well-defined and physio- 

 logically significant criteria, it is presented with the 

 purpose of standardizing future studies dealing with 

 reproduction in the striped mullet. 



The range (23 to 27 cm SL) we present for the 

 size at maturity of striped mullet in northeast 

 Florida is similar to the only other published report 

 for the area (25 cm SL by Stenger 1959), but much 

 lower than the values (30 to 46 cm SL) reported by 

 Jacot (1920), Thomson (1951), Ochiai and Umeda 

 (1969), Timoshek (1973), and Apekin and Vilenskaya 

 (1979). Most investigators agree that female mullet 

 reach sexual maturity at the end of their third year 

 (Thomson 1951; Timoshek 1973). We made no at- 

 tempt to age the fish because methods available 

 were not always agreed upon (see Thomson 1966). 

 However, according to the growth schedules of 

 Thomson (1966), adapted from various primary 

 sources, the age at maturity in this study apparently 

 ranged from 2V4 years (23 cm SL) to 2V2 years (27 

 cm SL). 



It thus appears that female M. cephalus attain sex- 



198 



