Lindeman and Snyder: Nearshore hardbottom fishes of southeast Florida 



513 



variation in the data set, whereas the eigenvalue for 

 CA axis 2 was 0.124 and accounted for 9.77c of the 

 variation in the data set. Samples from August and 

 September at Coral Cove and Carlin Park separated 

 from all other samples along CA axis 1. In general, 

 samples were not spread widely along CA axis 2; 

 however, two samples. May 1995 at Ocean Ridge and 

 February 1995 at Carlin Park, did separate from the 

 other sites. 



The ordination of taxa on CA axes 1 and 2 showed 

 how the taxa were distributed in relation to the 

 hardbottom samples along these same axes (Fig. 5B). 

 The most common species (e.g. Table 1 ) clustered near 

 the origin of the ordination. Taxa with high scores 

 along CA axis 1 included infrequently occurring spe- 

 cies such as Halichoeres poeyi, Haemulon aurolin- 

 eatum , Mulloidichthys martin Icus. and Caranx ruber. 

 Low scores on CA axis 1 were Echidna catenate, 

 A.canthurus chirurgus, Chactodon occllatus, and 

 Sciaenidae sp. Species with high scores on axis 2 were 

 Sparisoma aurofrenatum, Chaetodon ocellatus, and 

 Sparisoma viride. These were most abundant at Ocean 

 Ridge in May 1995 and were responsible for the sepa- 

 ration of this sample from all others along CA axis 2. 



In comparisons of hardbottom and natural sand, 

 20 transects over natural sand plains recorded only 

 four taxa. The clupeid, Harengulajaguana, was most 

 abundant (18 juveniles in two schools total). An uni- 

 dentified Eacinostomus species, Gerres cinereus, and 

 Caranx bartholomaei were also recorded (four, one, 

 and one individuals, respectively). Hardbottom habi- 

 tats typically had over thirty times the individuals 

 per transect as natural sand habitats. Two-sample 

 ^-tests comparing hardbottom with sand habitats 

 rejected the hypothesis of no differences in mean 

 numbers of individuals per transect (P<0.005). 



Life-stage abundances 



At all sites, juveniles were the most abundant life 

 stage among the top ten species ( Table 2 ). At the two 

 intensively sampled sites in Jupiter, the numbers of 

 juveniles of all species pooled were significantly 

 greater than any other life stage (ANOVA, P<0.01, 

 Tukey's HSD). There were no significant differences 

 in abundances among newly settled, early juvenile, 

 and adult stages ( Tukey's HSD ). At both Jupiter sites, 

 at least 80% of the individuals were early life stages 

 (pooled newly settled, early juvenile, and juvenile 

 stages )( Fig. 6). Abundances of newly settled and early 

 juvenile stages at Ocean Ridge were similar to the 

 Jupiter sites (Table 2), although numbers of juve- 

 niles and adults were lower. Newly settled stages of 

 over 20 species were recorded on nearshore hard- 

 bottom structures among the three sites. 



CP_4/94a 



CP_4/94b 



CP_5/94 



CP_3/95 



CP_6/94a 



CP_4/95b 



OR_4/95 



OR 6/95 



OR_5/95 



CC_6/95a 



CC_6/95c 



CO 4/94 



CC_9/96 



CC_6/94 



CP_6/94b 



CC_4/95a . 



CO 6/96 



CP_7/94b  



CC_3/95 



004/96 



005/96  



CP_6/94c 



CC_6/95b • 



CC_7/94  



CP_7/94a 



CP_8/94 



008/94 



CP_9/94  



CC_9/94 • 



CP_2/95  



CO 2/96 . 



Dn 



Group 1 



Group 2 



Group 3 



00 



— I — I — r 



25 



-1 — I — I — I — I — I — I 



50 75 



Dissimilarity 



Figure 3 



Clustering of visual census samples (sites and times) 

 by UPGMA analysis of a normal Bray-Curtis dis- 

 similarity matrix. CC: Coral Cove: CP: Carlin Park: 

 OR: Ocean Ridge. 



Eight of the ten most abundant taxa by site were 

 represented primarily by early stages (Table 2). Co- 

 coa damselfish and hairy blenny occurred most abun- 

 dantly as adults. Adults of at least five of the top ten 

 species occurred as residents, not transients. These 

 included sergeant major, hairy blenny, cocoa dam- 

 selfish, silver porgy, and black margate. Adults of a 

 variety of less common species occurred but were of- 

 ten less abundant than early life history stages. 

 Newly settled and juvenile stages often appeared to 

 display more site-fidelity with hardbottom structure 

 than did adults. 



Six species of grunts (four Haemulon and twoA«/- 

 sotremus) and two species of damselfishes (Stegastes 



