286 



Fishery Bulletin 98(2) 



three years, and an analysis of covariance to log-trans- 

 formed data was applied (Draper and Smith, 1981). 



Results 



Length distributions for Brazilian menhaden ranged 

 from 26 to 36 cm FL and did not differ among 

 the years sampled (P>20). All individuals collected 

 were adult mature females with yolked or hydrated 

 oocytes; no immature ovaries were observed. 



The oocyte diameter distribution of gravid B. aiirea 

 females (with hydrated oocytes) showed four groups 

 of oocytes (Fig. 3). The smallest group was com- 

 posed mainly of primary growth oocytes (smaller 

 than 120 pm). The next larger group was composed 

 of cortical alveolus stage and partially yolked oocytes 

 (primary yolk stage) ranging from 120 to 500 pm. A 

 third group included advanced yolked oocytes (sec- 

 ondary yolk stage) ranging from 500 to 700 pm. The 

 fourth group was the largest and corresponded to 

 the hydrated oocytes measuring 1100 to 1500 pm. 

 The continuous distribution from primary growth 

 oocytes to advanced yolked oocytes is a characteris- 

 tic pattern in multiple spawning fishes. 



Spawning frequency was estimated by examining 

 ovarian tissue from 169 mature females sampled 

 between 21 and 23 of November 1995 during the 

 main spawning peak. Of all specimens examined, 



41% 



^ 



25 



Ze- 



is - 



Primary growth stage 



Cortical alveolus and primary yolk stage 



Secondary yolk stage 



Hydrated oocytes 



«=749 



^ 



200 400 600 800 1000 1200 1400 



Oocyte diamclcr (urn) 



Figure 3 



Oocyte diameter distribution for an ovary of Brazilian menhaden containing 

 hydrated oocytes. The different shadings represent the four groups of oocytes 

 identified. 



about 14% had new postovulatory follicles (day-0) 

 and 10% had day-1 postovulatory follicles (Table 1), 

 the average was 11.98% (SD=7.53%). 



Batch fecundity estimates ranged from 20,000 

 hydrated oocytes for a 27-cm-FL female to 130,000 

 hydrated oocytes for a 34-cm-FL female. A power 

 model and a linear model were fitted to the rela- 

 tionships BF versus FL and BF versus TW, respec- 

 tively (Fig. 4). Analysis of covariance indicated that 

 the slope of the regression of fecundity on length 

 did not differ between years, but the intercepts were 

 significantly different ( 1994-95, F, j 74,^6.68, P<0.05; 

 1994-97, F, 1 gg, ^ 26.02, P<0.01; 1995-97, F, j gj ,=4.53, 

 P<0.05). Relative fecundity ranged from 60 to 212 

 hydrated oocytes per gram of female (ovary free). 

 These values were different (P<0.05) between the 

 years 1994 ( 107 ±29 hydrated oocytes) and 1995 ( 135 

 ±34 hydrated oocytes), and 1994 versus 1997 ( 149 ±30 

 hydrated oocytes), but not between 1995 and 1997. 



Discussion 



Brevoortia aurea is a multiple spawner with inde- 

 terminate annual fecundity, according to our obser- 

 vations of maturing ovaries with postovulatory 

 follicles and yolked oocytes (partially spent stage). 

 These observations suggest that after one batch 

 of eggs is spawned, a new batch develops and is 

 released ( Hunter etal., 1992). Further, 

 the oocyte size-frequency distribu- 

 tion of gravid females shows different 

 batches of growing oocytes, including 

 hydrated eggs. Fractional spawning 

 has been suggested for other con- 

 geners, such as Atlantic menhaden 

 and gulf menhaden (Brevoortia patro- 

 nus), on the basis of oocyte diameter 

 distributions (Lewis and Roithmayr, 

 1980; Lewis et al., 1987). These spe- 

 cies have a long reproductive season, 

 6-8 mo. (Powell, 1994), similar to 

 that of B. aurea. Atlantic and gulf 

 menhaden spawn during the fall and 

 winter (Powell, 1994), whereas Bra- 

 zilian menhaden spawn in spring- 

 summer, as do most of the fishes 

 inhabiting the Rio de la Plata estu- 

 ary (Macchi and Acha, 1998). 



During November the temperature 

 ranges from 19° to 2rC in the 

 spawning area of Brazilian menha- 

 den. Because duration of the POF 

 stage varies at different tempera- 

 tures, disappearance of the lumen 



