Szedlmayer and Conti: Nursery habitat, growths rates, and seasonality of age-0 Lutianus campechanus 



633 



40 n 



20 

 



20 

 



40 



20 







60 

 40 

 20 





 60 

 40- 

 20- 







80 H 

 60 

 40 H 

 20 







"1 ^r 



11 Jul 



\ 1 80 



24 Jul 60 



9 Aug 



1 1 1 40 



20 H 

 

 1 80 



25 Aug 



FG 

 3561 



12 Sep 



22 Sep 



MN 

 I 68 



60- 

 40 

 20 H 

 -1 

 60 

 40 

 20 



r^— 1 

 80 - 

 60 

 40 

 20 



"T r 



T r 



"I I r 





12345 12345 



Stations (1995) 



Figure 7 



Mean standard lengths by station and date in 1995 for all collections that had 

 >10 fish. Numbers in bars are number offish measured. Different letters show 

 significant differences among stations and dates ( ANOVA; Waller-Duncan test; 

 P<0.05). 



due to yearly variation. Collins et al. ( 1996) sampled 

 from 1991 to 1993. Lack of evidence for August spawn- 

 ing in our study may also have resulted from low 

 survival of later spawned fish or fi"om new recruits hav- 

 ing settled to different habitats not sampled in the 

 present study 



The June and July 1994 cohorts had growth rates 

 that were significantly faster in relation to 1995 co- 

 hort growth rates. Growth rates estimated from SL 

 on varying dates were probably inaccurate owing to 

 extended settlement of new recruits, but results 



showed the same pattern as SL on age estimates, 

 i.e. 1994 growth rates were significantly faster com- 

 pared with 1995 estimates (^test, P<0.05; Fig. 5). If 

 growth rates were an indication of the probability of 

 survival (Houde, 1989; Sogard, 1997), then 1994 co- 

 horts may have had an advantage over the 1995 

 spawned cohorts. Growth rate typically decreases 

 with decreasing temperature and may have resulted 

 in the low growth rate observed for the May 1995 

 cohort. Thus, it appears that both density dependent 

 and density independent mechanisms are operating 



