FIELD AND LABORATORY ASSESSMENT OF PATTERNS IN 



FECUNDITY OF A MULTIPLE SPAWNING FISH: THE 



ATLANTIC SILVERSIDE MENIDIA MENIDIA' 



David 0. Conover^ 



ABSTRACT 



Patterns in fecunditj' (i.e., spawning frequency, batch fecunditj', annual egg production) of a multiple 

 spawning fish, the Atlantic silverside, Menidia menidia, were assessed by 1) gonadal analysis of field 

 specimens, 2) visual observations of spawning periodicity in the field, and 3) laboratory experiments. The 

 gonadal analysis assumed that the difference between total number of eggs (recruitment + mature) per 

 female just prior to the beginning of spawning, and recruitment egg retention per female at the end of the 

 spawning season, represented annual egg production. Annual egg production estimated in this manner was 

 893 ±197 eggs/g ovary- free body weight (±95% C.L.). Batch fecundity (no. eggs in the most advanced size 

 class/g ovar\'-free body weight, ±95% C.L.) varied significantly during the breeding season, being lowest 

 near the beginning (179 ± 21) and end (181 ± 28), and highest during the middle (266 ± 34 and 267 ± 23) of 

 the breeding season. Batch fecunditj' averaged over the entire breeding season, was about 225 eggs/g 

 ovary-free body weight, indicating that each female must spawn about four times. The gonadal estimate was 

 tested by inferring the actual spawning frequency from daily, visual observations of spawning in the field. 

 These observations showed that spawning occurred on a fortnightly cycle coincidental with new and full 

 moons, and that each female spawned at most once per semilunar period. There were about four semilunar 

 spawning phases during the breeding season, indicating close correspondence with the results of gonadal 

 analysis. 



Laboratory observations demonstrated that female Af. menidia are physiologically capable of spawning 

 more frequently than in nature. Total egg output in the laboratory was about twice that in the field. 



Accurate estimates of fecundity are important in 

 describing the dynamics of fish populations. In some 

 fishes, all eggs mature synchronously and are shed in 

 a single batch over a relatively brief period of time 

 each year (Bagenal 1967). Estimating fecundity in 

 such species is a simple process of enumerating the 

 number of ripening eggs per female. However, in 

 many other fishes, ova mature in multiple batches 

 that are spawned successively within one spawning 

 season. These species have been termed multiple 

 spawners, batch spawners, serial spawners, or frac- 

 tional spawners by various authors (Bagenal and 

 Braum 1971; Hempel 1979; DeMartini and Fountain 

 1981; Gale 1983; Snyder 1983). Little is known 

 about the patterns in fecundity of multiple spawners, 

 even though many marine and freshwater fishes 

 from diverse taxonomic groups in both temperate 

 and tropical regions produce eggs in this manner. 

 Determination of annual egg production in multiple 

 spawners is difficult and recent studies have in- 



Contribution No. 89 of the Massachusetts Cooperative Fishery 

 Research Unit, University of Massachusetts, Amherst, MA, and 

 Contribution No. 437 of the Marine Sciences Research Center, 

 State Universitj' of New York, Stony Brook, NY. 



^Marine Sciences Research Center, State University of New 

 York, Stony Brook, Long Island, NY 1 1794. 



dicated that in some species, previous assessments 

 may be in error by as much as an order of magnitude 

 (Hunter and Goldberg 1980; Hunter and Leong 

 1981; DeMartini and Fountain 1981). 



Frequency distributions of egg size (diameter) 

 within ovaries of multiple spawning fishes are 

 characteristically multimodal (Hempel 1979). In 

 most multiple spawners, a synchronously maturing 

 batch of eggs that is accumulating yolk sequentially 

 arises from a much larger group of previtellogenic 

 immature eggs, termed "recruitment" ova (Clark 

 1925; Bagenal and Braum 1971; Jones 1978; Hunter 

 and Leong 1981). The problem in estimating fecun- 

 dity has been to determine how many modes or 

 batches of eggs are spawned annually. The conven- 

 tional approach has been to count only the largest 

 eggs or those above an arbitrary size (e.g., all yolked 

 eggs) under the assumption that smaller eggs would 

 be resorbed or spawned in later years. Whenever 

 this assumption is incorrect, fecundity can be grossly 

 underestimated. A second approach has been to infer 

 the spawning frequency of adults from the propor- 

 tion of field-collected females in ready- to- spawn con- 

 dition (i.e., those containing hydrated eggs, e.g., 

 Demartini and Fountain 1981) or from females 

 showing evidence of having just spawned (i.e.. 



Manuscript accepted October 1984. 



FISHERY BULLETIN: VOL. 83, NO. 3, 1985. 



331 



