Ferreira and Russ: Population structure of Plectropomus leopardus 



639 



great differences in the size and age structure were 

 not likely to be detected. A fourth assumption is that 

 effects are attributable to fishing and not some other 

 factor. The failure to detect a significant difference 

 between either mean size or age of leopard coral- 

 grouper on open and closed reefs, however, was due 

 largely to variability between replicate reefs. If, in 

 the present study only Glow (closed) and Grub (open) 

 had been compared, the result would reveal a classic 

 effect-of-fishing scenario, with a larger range of sizes 

 and ages and significantly larger mean sizes and ages 

 observed on the reef closed to fishing. In contrast, if 

 only Yankee (closed) and Hopkinson (open) had been 

 compared, no effect of fishing would have been de- 

 tected on the population structure. These results 

 emphasize the importance of replicate reefs in analy- 

 ses of the effects of fishing on coral reef fish popula- 

 tions. More replicates (i.e. more reefs per treatment 

 group) would increase the degrees of freedom and 

 thus the power (i.e. likelihood of detecting a given 

 effect size) of the nested AN OVA. 



One possible reason for the differences between 

 the two open reefs is the fact that they are appar- 

 ently not subject to the same fishing pressure. Grub 

 is renowned for its excellent anchorage, and there- 

 fore is a preferred site for recreational and commer- 

 cial fishing vessels. Aerial surveys conducted by the 

 Great Barrier Reef Marine Park Authority between 

 1989 and 1992 (GBRMPA 7 ), indicated that Grub is 

 frequented by boats 2.2 times more frequently than 

 Hopkinson and that fishing vessels are sighted 3 

 times more often at Grub than at Hopkinson. Such 

 factors should be taken into account in designing 

 future sampling and experimental programs on the 

 effects of fishing on the Great Barrier Reef. 



Nevertheless, there was a major and consistent 

 difference between the open and closed reefs that 

 were analyzed. For the two closed reefs, the popula- 

 tion structure was dominated by the presence of a 

 strong year class which settled in early 1984. A simi- 

 lar pattern was not obvious on the open reefs, and a 

 corresponding strong mode was not evident at Grub 

 or at Hopkinson. Occurrence of strong year classes 

 is a well-documented phenomenon in commercial 

 catches of temperate species (Hjort, 1914; Sissen- 

 wine, 1984; Rothschild, 1986). For temperate spe- 

 cies, year-class strength has been linked to early life 

 history processes since the beginning of this century 

 (Hjort, 1914). However, for populations of coral reef 

 fish, the importance of recruitment as a major driv- 

 ing force in the temporal variability of abundance 

 has been recognized only recently (Williams, 1980; 



7 GBRMPA data base. 1992. Great Barrier Reef Marine Park 

 Authority. P.O. Box 1379, Townsville, Q48 10, Australia. 



Doherty and Williams, 1988, a and b; Doherty, 1991; 

 Doherty and Fowler, 1994). 



There is evidence for the possibility of strong 

 recruitment pulses of reef fishes occurring concur- 

 rently on midshelf reefs off Townsville which are 

 separated by distances of up to 10-30 km (Doherty 

 and Williams, 1988, a and b; Williams, 1991). The 

 age-structure data for the two closed reefs provides 

 circumstantial evidence in support of pulses of re- 

 cruitment being synchronous on reefs at least 10 km 

 apart ( Fig. 1 ). With the assumption that the four reefs 

 received a similar pulse of recruitment in 1984, it is 

 apparent that fishing mortality has operated to 

 largely decrease the abundance of this year class. 

 On the closed reefs, this strong year class was pro- 

 tected from fishing for almost its entire life and as a 

 result its dominance was maintained. In contrast, 

 on the open reefs, the same year class probably sup- 

 ported the fisheries disproportionately in relation to 

 the other age classes, and consequently abundance 

 was reduced. An alternative hypothesis is that the 

 settlement pulse occurred only on the two closed reefs 

 owing to some process independent of fishing. 



A common question regarding the effects of fish- 

 ing on protogynous hermaphroditic fishes is how the 

 sex structure of the population would respond to fish- 

 ing mortality. If sex change is determined by age and 

 size and selective removal of larger and older indi- 

 viduals occurs, the result would be a decrease in the 

 proportion of males in the population. However, if 

 sex change is behaviorally induced, the population 

 would be expected to compensate to some extent for 

 the selective removal of males by female-to-male sex 

 change, i.e. by changing sex at smaller ages and sizes. 



The mean size and age observed for each stage 

 seemed to follow the size and age structure of each 

 population. Mature females and young males were 

 larger and older at Glow than those at Grub. Age 

 and size of transitional-stage fish did not differ sig- 

 nificantly between reefs, but this is not surprising 

 given the high variability in the age and size of sex 

 transition characteristic of leopard coralgrouper and 

 the small numbers of transitional individuals. Ma- 

 ture males from Yankee were smaller than those at 

 Hopkinson. This is possibly a consequence of the age 

 distribution and consequent size distribution. Yan- 

 kee had proportionally more 6- and 7-year-old fish 

 and not many in the older age classes; therefore most 

 males would be 6- or 7-year-olds, resulting in a small 

 overall mean size. At Hopkinson, the age frequency 

 was more evenly distributed, without strong modal 

 classes for years 6 and 7 and having a wider range of 

 age classes. 



The comparison of frequency of developmental 

 stages between reefs showed significant variation. 



