FISHERY BULLETIN: VOL. 83, NO. 3 



presence of 1-d-old postovulatory follicles, observed 

 histologically, Hunter and Goldberg 1980). Spawn- 

 ing frequency is then multiplied by batch fecundity 

 (number of eggs in the largest size class) to arrive at 

 annual fecundity. Excellent examples of this method- 

 ology can be found in DeMartini and Fountain 

 (1981), Hunter and Goldberg (1980), and Hunter and 

 Macewicz (1980). A third technique for estimating 

 fecundity in multiple spawners has been laboratory 

 experiments where females are confined and allowed 

 to spawn repeatedly (Gale and Gale 1977; Gale and 

 Buynack 1978, 1982; Hislop et al. 1978; Gale 1983). 

 The problem here is that spawning frequency and 

 fecundity are dependent on the food supply (Wootton 

 1973, 1977, 1979; Townshend and Wootton 1984), so 

 that it may be difficult to interpret laboratory data 

 unless detailed studies of feeding rate and/or fecun- 

 dity of fish in nature have been previously conducted 

 (e.g.. Hunter and Leong 1981). 



The ovarian cycle of many fishes that breed during 

 a restricted season involves two major, alternating 

 phases of oocyte production and growth: 1) a pre- 

 vitellogenic phase during which new oocytes are pro- 

 duced, cell organelles form, and cytoplasmic growth 

 occurs; and 2) a vitellogenic phase during which 

 growth is faster and yolk accumulates in the ovum 

 (Ball 1960; Jones 1978; Tokarz 1978; Baggerman 

 1980). The greatest production of new oocytes and 

 phase one growth occurs during the postspawning 

 season, with vitellogenic growth and maturation of 

 eggs occurring just prior to and during the spawning 

 season (this may not be true in tropical or other 

 fishes that breed throughout most of the year, see 

 reviews by Ball 1960; Jones 1978; Baggerman 1980). 

 Hence, in multiple spawners having a restricted 

 breeding season, the reservoir of recruitment eggs 

 may be largely formed prior to the breeding season. 

 If true, then the number of recruitment eggs per 

 female should consistently decline as the breeding 

 season progresses. The rate of decline in number of 

 eggs per female would provide an estimate of 

 seasonal egg production, and spawning frequency 

 per female could be estimated from the total number 

 of eggs shed divided by batch fecundity. The method 

 could be tested by comparing the estimated spawn- 

 ing frequency with the actual spawning frequency 

 determined independently in some other manner. 



The purpose of this paper is to describe patterns of 

 batch fecundity and annual egg production in the 

 Atlantic silverside, Menidia menidia. The analysis 

 employs the method described above: I show how the 

 total number of eggs per female (recruitment -i- 

 maturing) present at the beginning of the spawning 

 season minus recruitment eggs per female retained 



at the end of breeding can be used as an estimate of 

 total egg production. The method is tested by show- 

 ing that predicted spawning frequency is identical to 

 the spawning frequency inferred from direct visual 

 observations of spawning periodicity in a field popu- 

 lation. Observations of spawning frequency and egg 

 production under laboratory conditions are used to 

 demonstrate that individual females are physiolo- 

 gically capable of maintaining the egg production 

 rates and spawning frequency estimated from field 

 populations. 



The Atlantic silverside, Menidia menidia, (Pisces: 

 Atherinidae) is a multiple spawning marine fish that 

 inhabits coastal waters of eastern North America 

 (Middaugh 1981; Conover and Kynard 1984). 

 Menidia menidia is an annual fish: all individuals 

 mature at age 1 and < 1% of breeding populations 

 are 2 yr old. The ecological importance of M. ynenidia 

 in terms of biomass transport from salt marsh to off- 

 shore communities and as forage for piscivorous 

 fishes has been previously documented (Bayliff 1950; 

 Conover and Ross 1982; Conover and Murawski 

 1982). The Atlantic silverside is an excellent species 

 for studying patterns in fecundity because it is 

 numerous and can be easily collected, spawning is 

 easy to visually observe in the field, and it readily 

 breeds in the laboratory. 



METHODS 



Gonadal Analysis 



Field Sampling 



Fish were sampled from the salt marsh region of 

 Essex Bay, MA, with beach seines. Specimens were 

 collected every 2 wk during 1977 as part of a general 

 study of the population ecology of M. Tnenidia. Addi- 

 tional specimens were collected intermittently in the 

 spring during the breeding season. All collections 

 were made in daylight within 1-2 h of low tide, and 

 all specimens were immediately preserved in 10% 

 buffered Formalin^. Detailed descriptions of Essex 

 Bay and sampling methodology are available in Con- 

 over and Ross (1982). 



Fecundity 



Gonads were excised and weighed (nearest 0.01 g) 

 from fish captured on 1 1 dates from October 1976 to 

 July 1977. All fish were measured for total length 



'Reference to trade names does not imply endorsement by the Na- 

 tional Marine Fisheries Service, NOAA. 



332 



