516 



Fishery Bulletin 97(3), 1999 



cies and numbers of individuals per transect for April 

 among 1994-96 revealed no significant interannual 

 differences (ANOVA,P=0,34; ANOVA,P=0.21). Iden- 

 tical comparisons for June among the same three- 

 year period revealed no differences among mean 

 numbers of individuals (AN OVA, P=0.06), but sig- 

 nificant differences among numbers of species 

 (ANOVA, P<0.05). Pairwise comparisons using 

 Tukey's HSD showed that Coral Cove in June 1995 

 had significantly more species than in June 1996 

 (mean species numbers: 8.7 versus 5.3). 



Seasonal occurrence of only newly settled stages 

 was examined at Coral Cove (Fig. 8). The timing and 

 abundance of species occurrences suggested seasonal 

 variations with peaks of newly settled stages in late 

 spring and early summer. Sizes and numbers of 

 individuals typically increased as summer pro- 

 gressed. Abundances of early stages for most species 

 appeared to be low in February and March prior to 

 peak spawning activity for many species (Garcia- 

 Cagide et al., 1994). However, difficulties in the col- 

 lection of visual sui^veys constrained the examina- 

 tion of fall and winter patterns. 



3 -, 



2 - 



1 - 



- 



-1 - 



-2 - 



B 



5*42*° 

 51 



43 

 33 

 39 



,25 2f' 

 57 



4t 



46e ' 



56 26 



23 



U^ 



19 13 42^ 



_ - si 14 



5 40 12 



S3 





43 



sz 



"1 







Axis 1 



Figure 5 



(A) Ordination of sample scores on axes 1 and 2 from correspondence analy- 

 sis of a samples by taxa matrix of census data from Coral Cove (CC), Carlin 

 Park (CP), and Ocean Ridge (ORi. Groups 1-3 from Figure 3. (B) Ordination 

 of taxa scores from the same correspondence analysis. Numeric codes for taxa 

 are given in Figure 4. 



Ichthyofaunal characteristics 

 after habitat burial 



Prior to habitat burial, fish assemblages 

 at the two Jupiter sites were similar in 

 species composition and relative abun- 

 dance (Figs. 2, 3, and 5A). Pre- and 

 postburial numbers of individuals and 

 species at the control and impact sites 

 are plotted in Figure 9. The hypotheses 

 of no differences in total numbers of in- 

 dividuals and species before and after 

 dredge burial of hardbottom were both 

 rejected (P<0.001) in two sample ^tests 

 for equal variances following a BACIPS 

 design (Stewart-Oaten et al., 1986; 

 Osenberg and Schmitt, 1996). 



No fishes or exposed hardbottom were 

 recorded in the first postdredging sur- 

 veys at Carlin Park ( 13 April 1995). Sev- 

 eral hardbottom outcroppings ( 1 m high 

 by 3 m wide), parallel to the shore, were 

 exposed at a depth of 1.5 to 2.5 m at the 

 site during the second postdredging sur- 

 veys on 24 April. Two of ten transects 

 crossed narrow outcrops with dense 

 schools of newly settled stages of three 

 grunt species and one drum species 

 (Sciaenidae). Before burial of the reef 

 by extension of the beach width by ap- 

 proximately 60 m, such outcrops were 

 deeper and further offshore than the 

 original hardbottom sampling area. 

 Small outcrops were still present in May 

 and were occupied in two of ten tran- 

 sects by three species of grunts and one 

 damselfish (Fig. 9). Schools of newly 

 settled stages predominated. Such out- 

 crops were not encountered during the 

 20 surveys in June, and no fishes were 

 present with the exception of several 

 round scad, Decapterus punctatus. Ten 



