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



649 



protracted spawning season, he recorded the fecundity 

 of this species as the number of larger eggs (diameter 

 >180 fjm) estimated to be present in the ovaries of 

 mature females just prior to the commencement of the 

 spawning period. Thus, the very strong possibility that 

 some eggs with diameters <180 /jm would have been 

 destined to have become fully mature and released at 

 some stage during the protracted spawning season, i.e., 

 the species has indeterminate fecundity, was not taken 

 into account. 



In species with indeterminate fecundity, the distri- 

 bution of oocyte diameters essentially forms a contin- 

 uum, reflecting the continuous maturation of oocytes 

 throughout the spawning season and thus the progres- 

 sion through to maturity of some of the small and pre- 

 vitellogenic oocytes that were present at the beginning 

 of the spawning period. Consequently, counts of the 

 standing stock of larger oocytes found just prior to the 

 onset of spawning will result in an underestimate of the 

 potential annual fecundity of such species (Hunter et al., 

 1985. 1992; Lisovenko and Andrianov, 1991). Estimation 

 of the annual fecundity of species with indeterminate 

 fecundity thus requires a combination of data on batch 

 fecundity and spawning frequency (Hunter et al.. 19851. 

 Batch fecundity, i.e., the number of oocytes released 

 during a single spawning event, can be estimated by 

 counting the number of hydrated oocytes present in 

 ovaries immediately prior to that spawning (Hunter 

 et al., 1985). The frequency with which a fish spawns 

 during the spawning period can be determined from the 

 frequency of mature female fish possessing ovaries with 

 either hydrated oocytes or postovulatory follicles (POFs) 

 of a known age (Hunter and Macewicz, 1985). 



The spawning of many marine species of teleosts and 

 invertebrates is correlated with lunar periodicity and the 

 associated tidal cycles (e.g., Schwassmann, 1971; Taylor. 

 1984; Greeley et al., 1986; Hoque et al., 1999), with the 

 spawning of such fish species typically peaking around 

 the full or new moon (or both) (e.g., Johannes. 1978; 

 Taylor and DiMichele. 1980; Greeley et al., 1986). Many 

 fish and invertebrates with pelagic eggs spawn on high 

 or ebb tides that enable eggs and the subsequent larval 

 stages to be transported away from spawning areas, 

 in which planktivorous predators are concentrated. This 

 process thus reduces the likelihood of those early life 

 cycle stages being subjected to predation (Taylor, 1984; 

 Johnson et al., 1990; Morgan, 1990). The fact that there 

 is very little recruitment of the early 0+ individuals of 

 R. sarba into the lower Swan River Estuary, where ex- 

 tensive spawning occurs, indicates that tides transport 

 the eggs of this species from spawning areas in the 

 estuary into coastal marine waters (Hesp and Potter, 

 2003; Hesp et al., 2004). 



This investigation, which involved a detailed study of 

 the females of R. sarba in the lower Swan River Estuary, 

 had the following aims: 1) to test the hypothesis that 

 R. sarba has indeterminate fecundity; 2) to establish 

 the period during the day when the oocytes of R. sarba 

 become hydrated and when ovulation and spawning 

 occur; 3) to establish whether R. sarba spawns mainly 



when salinities are high and thus approach those of the 

 marine waters in which this species typically breeds 

 and whether spawning is correlated with the strength 

 and type (ebb vs flood) of tide in the lower reaches of 

 the Swan River Estuary; 4) to estimate the average 

 frequency of spawning for R. sarba during the spawning 

 period; 5) and to determine the relationship between 

 batch fecundity and fish length, and to use this rela- 

 tionship, in combination with the average spawning 

 frequency, to calculate the potential annual fecundity 

 of R. sarba of different sizes. 



Materials and methods 



Tide, lunar phase, and salinity 



The maximum daily tidal heights at the mouth of the 

 Swan River Estuary were calculated by using the tidal 

 prediction data of the Coastal Data Centre at the Depart- 

 ment of Planning and Infrastructure, Government of 

 Western Australia (http://www.coastaldata. transport. 

 wa.gov.au). The maximum tidal range at the mouth of 

 the Swan River Estuary is small, i.e., <0.8 m, and tides 

 can be diurnal or semidiurnal, depending on the time 

 of year (Spencer, 1956). Salinity was measured on each 

 sampling occasion by using a Yellow Springs Instru- 

 ments salinity meter (YSI model number 30, Yellow 

 Springs Instrument Co., Inc., Yellow Springs. OH). 



Sampling 



During 2001 and 2002, female Rhabdosargus sarba 

 were collected by seine netting in nearshore shallow 

 waters at distances of ca. 2.5 to 5 km from the mouth 

 of the Swan River Estuary, and by rod and line fishing 

 in water depths of 10-12 m at a distance of ca. 150 m 

 from the shore (for details of sampling region and seine 

 net, see Hesp and Potter, 2003). Sampling was under- 

 taken at least once weekly between July and November, 

 the period when R. sarba reach maturity in the lower 

 Swan River Estuary (Hesp and Potter, 2003). It was 

 restricted to the hours between dusk (ca. 18:00 h) and 

 dawn (ca 06:00 h) because extensive seine netting and 

 angling during the day in our earlier study failed to yield 

 any R. sarba. The failure to capture R. sarba by these 

 methods during daylight reflected the offshore movement 

 of this species from the shallows prior to dawn and a 

 far stronger targeting of bait by the large numbers of 

 the banded toadfish {Torquigener pleurogramma) that 

 feed in the offshore waters of the lower estuary during 

 the day. Because the lower reaches of the Swan River 

 Estuary act as a shipping harbor, alternative sampling 

 methods, such as gill netting and spearing, could not 

 be used to catch R. sarba during the day. The data for 

 2000 and 2001 were augmented by those derived from 

 fish collected from the same location by using the same 

 methods in 1998 and 1999 (Hesp and Potter, 2003). In 

 total, the results of the present study are based on an 

 examination of over 2000 R. sarba, of which 510 were 



