Williams et a\: Scales of spatial variation in demography of a large coral-reef fish 



679 



The homogeneity of demographic parameters among 

 reefs within regions presented here is not consistent 

 with a number of studies of other reef fish species on the 

 GBR (e.g. Doherty and Fowler, 1994; Hart and Russ, 1996; 

 Newman et al., 1996). These other studies demonstrated 

 significant differences in age structures, somatic growth, 

 mortality, and otolith growth among individual reefs 

 within a single geographic region for several smaller reef- 

 associated lutjanid, acanthurid and pomacentrid species. 

 The consistency of demographic parameters among reefs 

 within regions found in the present study is consistent 

 with an hypothesis that L. miniatus may move over larger 

 distances than "typical" coral-reef fish, including being 

 capable of movements among reefs within a region. There 

 are limited direct data on the movement of L. miniatus, 

 and lethrinids in general, or about the range of habitats 

 they occupy. However, L. miniatus is frequently found on 

 shoal grounds between reefs and to depths of at least 128 

 m (Newman and Williams, 1996), suggesting a strong po- 

 tential for L. miniatus to move among reefs. Movements of 

 adults among coral reefs would suggest that L. miniatus 

 does not fit the typical metapopulation model for coral-reef 

 fish, in which adults are confined to a single coral reef, 

 and the pelagic larval stage is the only means of dispersal 

 among reefs. Accordingly, differences in conditions among 

 neighboring reefs would be less likely to be manifest in 

 demographic parameters of L. miniatus than in the demo- 

 graphic parameters of more sedentary species that inhabit 

 only a single reef for their postsettlement life. 



Using microsattelite markers, van Herwerden et al. 

 (in press) examined the genetic structure of L. miniatus 

 populations on the GBR. They sampled from two reefs 

 within the Townsville (Dip and Glow) and Mackay (Bax 



and 20-137) regions in addition to two other reefs in the 

 far southern GBR (Sweetlip and Sandshoe). They found 

 no evidence of stock structure for L. miniatus populations 

 on the GBR indicating that the regional patterns in de- 

 mographic parameters of L. miniatus are not a result of 

 distinct genetic stocks. This is consistent with other genetic 

 studies that have demonstrated a lack of genetic structur- 

 ing of coral-reef fishes over large spatial scales of hundreds 

 to thousands of kilometers (Doherty et al., 1995; Shulman 

 and Bermingham, 1995; Dudgeon et al., 2000). 



The observed regional variation in demography may 

 be the result of regional differences in postsettlement 

 processes, such as competition (Jones, 1987), food and 

 habitat availability (Hart and Russ, 1996), population 

 density (Doherty, 1983), and water temperature (Conover, 

 1992). Alternatively, the regional variation in demography 

 may have resulted from regional variation in recruitment, 

 coupled with density dependent processes (Doherty and 

 Fowler, 1994), or the factors that influence larval survival 

 and settlement. Unfortunately data for these processes for 

 L. miniatus are at best limited, restricting any conclusion 

 on the causative factor(s) driving the observed regional 

 patterns. However, because demographic parameters for 

 L. miniatus do not show a linear trend with latitude, factors 

 such as water temperature, which have strong latitudinal 

 gradients on the GBR (Lough, 1994), are unlikely to inde- 

 pendently explain the observed differences. Meekan et al. 

 (2001) also found that temperature did not appear to be a 

 causal factor driving spatial differences in demography of 

 damselfishes in the tropical eastern Pacific Ocean. 



Estimates of numerical density of L. miniatus on the 

 GBR are greater in the southern regions (Mackay and 

 Storm Cay) than on reefs in the Townsville region (Wil- 



