Nelson: Fad characteristics and associated fish assemblages 



843 



these numbers were strikingly constant across sample date 

 and, to a lesser extent, across treatments (Fig. 8). 



Recruit-enriched vs. nonenriched FADs 



Enriched FADs showed significantly higher rates of 

 recruitment than nonenriched FADs: the regression line 

 for the enriched FADs had a significant slope (f , j g|=20.76, 

 P<0.01), but the regression line for the nonenriched FADs 

 did not (F,i g)=2.29, P=0.17; Fig. 6). All additional fish were 

 juvenile sergeant major damselfish, Abudefduf troschelii . 

 These slopes are significantly different (f=3.05, 2 tailed 

 test, v=6, P=0.02; Fig. 9); enriched FADs accumulated fish 

 at a significantly higher rate (2.5 fish per hour) than did 

 nonenriched FADs that accumulated fish at a rate of 0.1 

 fish per hour Horizontal underwater visibility was 15 m at 

 the beginning of the experiment. 



Discussion 



FAD size, the presence of a fouling community, and the 

 presence of prior recruits all had positive effects on the size 

 of FAD-associated assemblages, although the latter factor 

 was assessed over a period of hours, whereas the former 

 were assessed over days. The repeated removal of an exist- 

 ing assemblage also had significant effects due at least 

 partially to treatment, but in all of these analyses sample 

 date appeared to play the largest role in determining the 

 numbers of fish(es) at these FADs. The presence of artifi- 

 cial fish or comparable-size weights did not significantly 

 affect assemblage sizes. There was little support for the 

 hypothesis that any of these factors might affect the spe- 

 cies diversity of these assemblages; only species richness 

 was significantly increased along with an increase in FAD 

 size and this result may be an effect of assemblage size 

 rather than object characteristics. Where treatment effects 

 did significantly affect the numbers of fishes, their effects 



50- 



45- 



40- 



35- 



30- 



25 



20 



15- 



10- 



5 







10- 

 9- 

 8- 

 7 

 6 

 5- 

 4- 

 3- 

 2- 

 1- 

 0-t 



A Abudefduf troschelii only 



D single 

 D triple 



26-Jul 



28-Jul 



30-Jul 



B comb ned spp. - A. troschelii 



m 



26-Jul 



28-Jul 



30-Jul 



Figure 6 



FAD-size effects (single FADs vs. triple-size FADs) on 

 aggregation size (mean no. of fishes) for Abudefduf trosche- 

 lii alone (A) and for all species combined less A. troschelii 

 (B). See Tables 3-5 for sample sizes and ANOVA results for 

 assemblage size and diversity measures. 



