Kmgsford and Hughes: Growth, mortality, and size of Acanthochromis polyacanthus 



571 



shore waters are the warmest and outer-shelf waters 

 are the coolest (Wolanski, 2001). The opposite pattern 

 of growth to the one observed in this study would be 

 predicted by this cross-shelf gradient in water tempera- 

 ture. It is also considered unlikely that local upwelling 

 events on outer-shelf reefs could produce the observed 

 differences, but they could influence primary produc- 

 tivity and abundance of food (zooplankton) through 

 nutrient-rich waters. An increase in average annual 

 temperature correlates with maximum age in some 

 fishes (review Choat and Robertson, 2002), but we found 

 no differences in age maximum across the shelf. We 

 conclude that any differences in temperature across 

 the shelf are not persistent enough to affect cross-shelf 

 patterns of growth of A. polyacanthus. 



Differences in growth profiles can be more realisti- 

 cally attributed to cross-shelf variation in some limiting 

 resource! s). This variation in resources may influence 

 the quality and quantity of food, suitable nest sites, ref- 

 uges from predators and (or) wave exposure, and density 

 of conspecifics and (or) other species that compete with 

 A. polyacanthus for resources. Correlative studies have 

 concluded that the distribution and abundance of coral 

 reef fishes is strongly influenced (directly and indirectly) 

 by physical factors such as wave exposure, sediment 

 loads, water depth, and topographical complexity, as 

 well as by biological factors (Williams, 1982). These 

 factors also have the potential to affect growth rates. 



A combination of reduced resource levels and high 

 population densities on outer-shelf reefs strongly indi- 

 cated that growth profiles represent density dependence 

 in scarids (Gust et al., 2001, 2002). Density of con- and 

 hetero-specifics was not recorded for our study, but 

 densities of A. polyacanthus were clearly greatest on 

 the mid- and outer-shelf reefs. This observation is con- 

 trary to the pattern noted by Williams ( 1982 ) who found 

 greatest abundances of A. polyacanthus on inner- and 

 mid-shelf reefs. Thresher (1983) suggested that food 

 abundance is a limiting resource for A. polyacanthus 

 and interspecific competition for food does occur. Thus, 

 it is plausible that variation in abundance of and com- 

 petition for food across the shelf may have influenced 

 the growth rates observed in the present study. The 

 large differences in cross-shelf densities and LJs of 

 A. polyacanthus indicate that competition may be less 

 important than variation in quantity and quality of food 

 across the shelf. 



Biomass of planktivores is generally highest at mid- 

 shelf reefs on the central GBR (Williams and Hatcher, 

 1983). Although data on cross-shelf abundance and dis- 

 tribution of plankton are limited, Williams and Hatcher 

 attributed this pattern to the increased availability of 

 food (zooplankton) in mid-shelf waters. Upwelling of 

 cold, nutrient-rich water from the edge of the continental 

 shelf results in high biomasses of phytoplankton. Aging 

 of the water (time since upwelling) is accompanied by 

 a shift in dominant planktonic biomass to herbivorous 

 and then carnivorous zooplankton. This shift in biomass 

 composition occurs simultaneously with the prevail- 

 ing wind-driven passage of water across the shelf and 



ultimately leads to the greatest biomass of zooplankton 

 occurring in mid-shelf waters (Andrews and Gentien, 

 1982; Sammarco and Crenshaw, 1984; Williams et al., 

 1988). Food quality has also been previously shown to 

 limit growth and reproduction in herbivorous coral reef 

 fishes (Horn, 1989; Choat, 1991). 



Despite a high abundance of zooplankton near shore, 

 these waters also have higher turbidity than mid- and 

 outer-shelf reefs. Visual impairment caused by very tur- 

 bid waters may hinder the ability offish to feed on plank- 

 tonic organisms and this hypothesis has been suggested 

 as a factor contributing to the low relative abundances 

 of planktivorous fish on inner-shelf reefs (Williams et 

 al., 1986). It is possible that this factor may retard the 

 growth and influence the maximum size of planktivores 

 like A. polyacanthus by effectively reducing food avail- 

 ability. Interestingly, lowest L r values were found at the 

 most turbid inshore reef, Pandora. Lower visibility near 

 shore, however, did not appear to affect the mortality 

 rates of A. polyacanthus at inner-shelf reefs. 



There were clear differences in growth, size maxima, 

 and age structures for populations of A. polyacanthus 

 across the continental shelf of the central GBR. Al- 

 though Acanthochromis polyacanthus grew faster and to 

 a larger size with increasing distance from the main- 

 land, cross-shelf mortality rates and maximum ages 

 were similar. Because these populations of fish are un- 

 likely to be genetically distinct, we suggest that biotic 

 and physical processes are the most plausible cause of 

 these cross-shelf patterns. Increased abundance of zoo- 

 plankton in mid- and outer-shelf waters, coupled with 

 potential visual impairment associated with high tur- 

 bidity levels on the inner shelf, are likely mechanisms 

 that explain the observed patterns, but multifactorial 

 manipulative experiments are required to determine 

 the relative contribution of these factors to variation in 

 demographic parameters. Our study therefore cautions 

 against pooling demographic parameters over broad spa- 

 tial scales without considering cross-shelf variation. 



Acknowledgments 



We would like to thank H. Patterson, C. Bunt, W. Rob- 

 bins, and the crew of the RV Orpheus for field assistance 

 during this study. We also thank J. Ackerman for analyt- 

 ical advice and expertise and J. H. Choat for constructive 

 comments on the manuscript. We also thank John Mor- 

 rison and the staff of MARFU for assistance with the 

 maintenance of aquarium fish. The project was partly 

 funded by an ARC Grant to MJK. This is a contribution 

 from Orpheus Island Research Station. 



Literature cited 



Aldenhoven, J. M. 



1986. Local variation in mortality rates and life history 

 estimates of the coral reef fish Centropyge bicolor (Pisces: 

 Pomacanthidae). Mar. Biol. 92:237-244. 



