Roumillat and Brouwer: Reproductive dynamics of Cynosc/on nebu/osus 



483 



ren, 2001; Nieland et al., 2002; Lowerre-Barbieri et al. 8 ) 

 have used the gonadosomatic index (GSI) to delineate 

 the spawning season in spotted seatrout. Even though 

 the GSI provided a good approximation of the spawning 

 season, histological data alone provided more precise 

 evidence. Spotted seatrout in South Carolina began 

 spawning near the end of April of each year and ceased 

 by early September. Similarly, Lowerre-Barbieri et al. 8 

 reported that the spawning season for spotted seatrout 

 in Georgia extended from late April to mid-September. 

 We found histological evidence of initial spawning in 

 specimens captured in 20°C water, although approxi- 

 mately 75% of spawning occurred when ambient water 

 temperatures were greater than 25°C. In laboratory 

 experiments, Brown-Peterson et al. (1988) found no suc- 

 cessful spawning in water below 23°C but pointed out 

 that others (McMichael and Peters, 1989) found eggs 

 and larvae in 20.4°C water. 



We found that females became mature approximately 

 one full year after their birth. A female born in May of 

 one year would be reproductively active in May of the 

 following year. Females born later in the season would 

 not be mature as the same successive season began; 

 therefore, not all one-year-old females were mature 

 when the spawning season began in May, but became 

 mature before that season ended. This maturity sched- 

 ule has also been reported for spotted seatrout in Loui- 

 siana (Nieland et al., 2002). However, Lowerre-Barbieri 

 et al. 8 found that all one-year-old females were mature 

 in coastal Georgia. A limited sample size or habitat 

 segregation of mature and immature trout (Lowerre- 

 Barbieri et al. 8 ) may have contributed to their result. 



The size at first maturity for spotted seatrout in this 

 study was 248 mm TL. This size is comparable to what 

 others have reported in other areas of the species' range 

 (Brown-Peterson et al., 1988; Brown Peterson and War- 

 ren, 2001; Nieland et al., 2002; Mercer 1 and references 

 therein; Lowerre-Barbieri et al. 8 ). Our estimate of size 

 at 50% maturity (268 mm TL) was larger than what 

 Nieland et al. (2002) reported for 100% mature trout 

 in Louisiana (250 mm TL). However, Nieland et al.'s 

 (2002) statement that animals are 100% mature at 250 

 mm TL, does not agree with the growth equation they 

 report for female trout when age = 1. Because we found 

 size at 100% maturity among female spotted seatrout 

 in South Carolina to be about 300 mm TL, we wonder 

 whether Nieland et al.'s (2002) growth equation for 

 female TL was meant to represent SL. Were this the 

 case, they might have offered a different rationale for 

 size at maturity among trout in Louisiana. 



Brown-Peterson et al. (1988) and Brown-Peterson 

 and Warren (2001) reported size at 100% maturity of 

 356 mm and 309 mm TL (using the SL-TL conversion 

 found in our "Methods: section) for spotted seatrout in 

 Texas and Mississippi, respectively. Brown-Peterson et 

 al. (1988), however, chose a combination of gears that 

 may not have sampled the trout population in Texas 

 representatively for size-at-maturity estimation. In Mis- 

 sissippi, Brown-Peterson and Warren (2001) used a 

 more appropriate gear for capture of late juvenile and 



early adult fish. Our estimate of size at lOO 1 * maturity 

 was quite similar to theirs. 



Spawning frequency 



Determining the number of multiple spawning events 

 during a single season for individual fish has been prob- 

 lematic. Initially, there was little understanding of the 

 reproductive dynamics of spotted seatrout, and BF esti- 

 mates were reported to represent the output for a whole 

 season (Pearson, 1929; Sundararaj and Suttkus, 1962; 

 Overstreet, 1983). Hunter et al. (1985) and Hunter and 

 Macewicz (1985) developed techniques to overcome these 

 limitations by providing protocols for the use of hydrated 

 oocytes in determining BF and SF among group-syn- 

 chronous species. 



To use the techniques of Hunter (1985) and Hunter 

 and Macewicz (1985) appropriately, it is critical to obtain 

 a representative sample of the spawning population. 

 DeMartini and Fountain (1981) and Lisovenko and Adri- 

 anov (1991) maintained that the relative occurrence of 

 hydrated oocytes (as determined macroscopically) was an 

 effective measurement of SF when the spawning popula- 

 tion was sampled representatively. However, when sam- 

 pling a species that spawns in aggregations at specific 

 geographic locations, as do many of the sciaenids, it is 

 inherently impossible to obtain a statistically representa- 

 tive sample of the spawning population for SF estimation 

 based on FOM. Because the window of opportunity is 

 temporally and spatially constrained, obtaining a sample 

 that includes all sizes and ages involved is not feasible; 

 the only choice in this situation is to sample in a directed 

 fashion. This was the sampling strategy used to target 

 females for BF counts; the majority of the animals cap- 

 tured whose oocytes evidenced FOM were obtained in a 

 nonrandom fashion. Additionally, we assumed that fishes 

 demonstrating FOM were moving toward deeper water 

 spawning aggregations and away from our capture gear. 

 For these reasons, we felt that our SF estimates based 

 on the proportion of females with oocytes in FOM were 

 biased and we excluded them from AF estimation. This 

 is an important matter to keep in mind when comparing 

 frequencies of spawning based on different methods. 



Because obtaining representative numbers of ani- 

 mals with late-maturing oocytes is not often feasible, 

 researchers have relied on the relative abundance of 

 postovulatory follicles (POFs) to calculate SF (Brown- 

 Peterson et al., 1988; Brown-Peterson and Warren, 2001; 

 Nieland et al., 2002; Lowerre-Barbieri et al. 8 ). The POF 

 method lacks the limitations (described above) of the 

 FOM method. Because the method we chose allowed 

 us to sample all sizes and ages of fish in the estuary, 

 obtaining representative numbers of animals with POFs 

 was accomplished effectively. Therefore, we felt that our 

 estimates of SF based on the POF method were more 

 precise and we chose to use them in deriving AF. 



The POF method depends on the ability to assess the 

 disappearance of these structures. Hunter and Macewicz 

 (1985) systematically sampled captive spawning ancho- 

 vies to develop histological criteria for POF atrophy in 



