Hesp et al.: Timing and frequency of spawning and fecundity of Rhabdosargus sarba 



651 



the ovarian duct and thus whether or not ovulation had 

 commenced at the time of capture of that fish. 



For all histological studies of the gonads, part of the 

 mid region of one of the ovarian lobes was placed in 

 Bouin's fixative for ca. 48 hours, dehydrated in a series 

 of ethanols, embedded in paraffin wax, cut into 6-um 

 sections, and stained with Mallory's trichrome. The ova- 

 ries were fixed within 1-3 hours of capture of the fish. 



To test the hypothesis that R. sarba has indetermi- 

 nate fecundity, the diameters of 100 oocytes in histo- 

 logical sections of stage-VI ovaries of two fish caught 

 during the spawning period were measured to the near- 

 est 10 /jm by using an eyepiece graticule in a compound 

 microscope and the stage of each of those oocytes was 

 recorded. Measurements were restricted to oocytes in 

 which a nucleus was visible in their center to ensure 

 that the oocytes had been sectioned through their center 

 and that the diameters were thus measured accurately. 

 This approach could not be used to measure the oocyte 

 diameters of hydrated oocytes in histological sections 

 because the nucleus of these oocytes undergoes germi- 

 nal vesicle breakdown. 



Histological sections of numerous ovaries were used 

 to determine the timing of the formation and degen- 

 eration of postovulatory follicles (POFs). An age was 

 assigned to the POFs found in ovaries of fish caught 

 at different times of the day, based on the timing of 

 ovulation and the degree to which those POFs had 

 degenerated (Hunter and Goldberg, 1980; Hunter and 

 Macewicz, 1985). Histological sections were also used 

 to determine the relative abundance of the different 

 stages of atresia in ovaries at different times during 

 the spawning period. 



The jars containing the ovarian lobes that had been 

 preserved in formalin at the different time intervals on 

 1. 2, and 13 September 2001 (see earlier) were shaken 

 until the oocytes of each ovary had become evenly sus- 

 pended in the solution. The resultant solution from 

 each ovary was then passed through a 125-/jm sieve 

 to remove the smallest oocytes, and we were able thus 

 to focus our study more specifically on the vitellogenic 

 oocytes. Comparisons of the appearance of the larger 

 oocytes under a dissecting microscope with those of the 

 different oocyte stages in histological sections of the 

 other ovarian lobe of the same fish were used to allocate 

 the oocytes observed under the dissecting microscope 

 to a specific stage in oocyte development. Each oocyte 

 in a representative subsample of 100 oocytes from each 

 formalin-preserved ovarian lobe was measured under a 

 dissecting microscope with an eyepiece graticule. This 

 approach enabled the diameters of hydrated oocytes to 

 be measured accurately, which was not possible with 

 histological sections (see earlier). 



Categorization of stages in atresia, fecundity estimates, 

 and spawning frequency 



On the basis of their histological characteristics, atretic 

 oocytes were allocated to either the a or /3 stages, by 

 using the criteria of Hunter and Macewicz (1985). Mature 



ovaries were categorized according to the proportions 

 of their a and /3 atretic oocytes (Hunter and Macewicz, 

 1985). Thus, atretic state = ovaries with yolked oocytes 

 but no a atretic oocytes; atretic state 1 = ovaries in which 

 less than 509c of the yolked oocytes are in the a stage 

 of atresia; atretic state 2 = ovaries in which less than 

 509c of the yolked oocytes are a atretic and atretic state 

 3 = ovaries which contain no yolked oocytes but do pos- 

 sess |3 atretic oocytes. During the present study, atretic 

 state 1 ovaries were further divided into three categories 

 on the basis of the percentage of a atretic yolk granule 

 oocytes in histological sections, namely early (<10%), mid 

 (10-359? I and late (36-50%) atretic state 1, an approach 

 similar to that adopted by Farley and Davis (1998). 



The batch fecundities of 31 R. sarba were estimated 

 from the number of hydrated oocytes in one of the ovar- 

 ian lobes of fish that had been preserved in 109c neu- 

 trally buffered formalin. These fish were chosen because 

 histological examination of their other ovarian lobe dem- 

 onstrated that the ovaries were in atretic state or early 

 state 1, i.e., less than 10% of their yolk granule oocytes 

 were atretic and newly formed POFs were not present 

 (Hunter et al., 1992; Nichol and Acuna, 2001). The for- 

 malin-preserved ovarian lobe was dried with blotting pa- 

 per and ca. 180-200 mg of tissue was removed from each 

 of its anterior, middle, and posterior regions and weighed 

 to the nearest 1 mg. These pieces of tissue were placed on 

 separate slides, covered with 309c glycerol and examined 

 under a dissecting microscope. The oocytes were then 

 teased apart and the number of hydrated oocytes record- 

 ed. The number of hydrated oocytes in each of the three 

 pieces of ovarian tissue of known weight were then used, 

 in conjunction with the weight of both ovarian lobes, to 

 estimate the total number of hydrated oocytes (=batch 

 fecundity) that would have been present in the pair of 

 ovarian lobes of each fish. The prevalence of spawning on 

 any given night is expressed as the percentage of female 

 fish with hydrated eggs (ovarian stage VD among all fe- 

 male fish with stage-V I mature) and stage-VI (spawning) 

 ovaries. These estimates were based on an examination 

 of samples collected between 22:00 and 01:30 h, when it 

 was possible to determine which female fish were going 

 to spawn in the ensuing few hours (see Hunter et al., 

 1985, for further details of this method). 



Results 



Although mean monthly salinities in the lower Swan 

 River Estuary in late spring to early winter were close 

 to that of full strength sea water (359cc), they fell pre- 

 cipitously to a minimum of 23%c (minimum individual 

 value=14'?c) in August, and then rose sharply in early 

 to mid-spring (Fig. 1). 



Staging of the ovaries and confirmation of 

 indeterminate fecundity 



The characteristics of each macroscopic stage of the 

 ovaries of R. sarba and the corresponding histologi- 



