Brouwer and Griffiths: Reproductive biology of Argyrozona argyrozona 



267 



however, showed that seasonal temperature regulated 

 gonad development for Pagrus aratus. Based on the data 

 collected during our study, photoperiod appears to be 

 responsible for the onset of gonad maturation in carpen- 

 ter; when day length increases (but water temperature 

 is variable) in September and October, and their gonads 

 begin to develop (Fig. 10). Photoperiod was also highly 

 correlated with GSI (r = 0.86), whereas temperature 

 showed a weakly negative relationship (r=-0.16). 



Nepgen (1977) calculated spawning frequency for 

 this species with an oocyte-size-frequency analysis of 

 inactive females. Finding only one peak in the oocyte- 

 size-frequency distribution, he assumed that carpen- 

 ter spawned only once a year. In our study POFs and 

 various yolk stage oocytes were found to occur simul- 

 taneously, proving that carpenter are serial spawners. 

 Accounting for monthly trends in spawning frequency 

 and the length of the spawning season, carpenter in 

 the Tsitsikamma National Park are estimated to spawn 

 at least 30 times per year. This spawning frequency is 

 similar to other predatory reef fishes, e.g., Mycteroperca 

 microlepis (30-40 times per year) (Collins et al., 1998). 

 Nevertheless, as with other species (Danilowicz, 1995), 

 spawning fraction in carpenter during the spawning 

 season was highly correlated with water temperature 

 (r=0.931) (Fig. 9), indicating that short-term cold water 

 upwellings, a common feature of the TNP during sum- 

 mer (Schumann et al., 1982), may negatively impact 

 annual carpenter fecundity in this area. 



Although fecundity in fishes is highly variable be- 

 tween individuals (Sadovy 1996), absolute fecundity 

 increases with size (Hunter et al., 1985; Davis and 

 West, 1993; Wilson and Nieland, 1994; Collins et al., 

 1998). In our study absolute annual fecundity increased 

 markedly with fish size (Table 4) and spawning season 

 was longer for large fish (Fig. 7) (Table 3). The positive 



correlation of batch fecundity and fish size (r=0.71), 

 coupled with the increased length of the spawning sea- 

 son for the older fish, greatly increases the absolute 

 annual fecundity of larger fishes (Fig. 8). Sadovy (1996) 

 noted that for red snapper (Lutjanus compechatius) one 

 large female (601 mm FL) will produce as many eggs 

 as 212 small (420 mm FL) females. Similarly, one large 

 female carpenter of 3.3 kg will produce as many eggs as 

 72 small ones of 0.3 kg. In addition to higher fecundity, 

 the larger fish produce significantly larger eggs and 

 presumably more viable larvae (Ojanguren et al., 1996; 

 Pepin and Anderson, 1997). 



Exploited populations were traditionally managed 

 to maximize growth (Griffiths, 1997). However it is 

 imperative to maintain sufficient numbers of reproduc- 



