518 



Fishery Bulletin 97(3), 1999 



ographic regimes (Sedberry and Van Dolah, 1984; 

 Chiappone and Sullivan, 1994) and may show dif- 

 fering patterns of fish diversity. 



Spatial and temporal attributes offish assem- 

 blages at the three sites in the present study were 

 examined by using ordination and cluster analy- 

 sis. Visual census samples collected from March 

 through July were similar in species composition 

 and relative abundance among sites (Fig. 3). This 

 finding is in agreement with the similar plots of 

 individual and species abundances among sites 

 (Fig. 2). The relative homogeneity of these samples 

 was further reflected in the co-occurrence of many 

 taxa including haemulids (Haemulon parra, H. 

 flavolineatum, H. chrysargyreum, Anisotremus 

 virginicus,A. surinamensis), pomacentrids iSteg- 

 astes variabilis, Abudefdufsaxatilis), labrisomids 

 (Labrisomus nuchipinnis), sparids (Diplodus 

 holbrooki), labrids (Halichoeres bivittatus, Thalas- 

 sorna bifasciatum) and scarids (Sparisoma rubri- 

 pinne) (Fig. 4). With the exception of L. nuchi- 

 pinnis and S. variabilis, most taxa occurred as 

 early life stages. 



Samples from late summer (August and Septem- 

 ber) were distinct from the spring and early sum- 

 mer in both cluster analysis and ordination (group 

 2, Figs. 3 and 5). The only two samples taken in win- 

 ter (February) differed from all other samples in the 

 analyses (group 3, Figs. 3 and 5). These patterns 

 suggest that some seasonality in assemblage struc- 

 ture existed. This may reflect late spring and sum- 

 mer peaks in larval settlement in contrast to reduced 

 winter settlement and, possibly, influxes of older ju- 

 veniles from inshore lagoonal habitats. Substantial 

 numbers of many species still settled in late sum- 

 mer but were possibly subject to higher predation 

 from older individuals that settled earlier in the year. 

 Various physical disturbances (e.g. winter cold fronts, 

 summer hurricanes) and biological phenomena 

 (variation in larval recruitment) affect the composi- 

 tion of fish assemblages of nearshore hardbottom. 

 The turbidity generated by physical disturbances 

 constrains the visual surveys needed to assess their 

 immediate effects. 



Nursery habitats and nearshore hardbottom 



With increasing human modifications of coastal ar- 

 eas, detailed knowledge of habitat usage is a key com- 

 ponent of informed fishery and coastal land manage- 

 ment. Identification of essential habitats includes the 

 evaluation of spatial distributions of structural habi- 

 tats across the shelf and habitat requirements of key 

 taxa. Several lines of evidence suggest that nearshore 

 hardbottom habitats along the mainland coast of east 



25 n 



~ 20 

 o 



15 



10 



Newly settled stages 



I I I I I 



T-T 



I I I I 



AMJJASONDJFMAMJJASONDJFMAMJ 

 1994 1995 1996 



Figure 8 



Mean number of newly settled individuals per transect for all 

 species pooled at Coral Cove. 1994-1996 In =148 transects). 



Florida can serve as nursery areas for many coastal 

 fish species. Over 80'v^ of the individuals at all sites 

 were early life stages. Eight of the top ten species 

 were consistently represented by early stages. Use 

 of hardbottom habitats was recorded for newly settled 

 stages of more than 20 species. In addition, other 

 natural habitats with substantial vertical relief were 

 absent from the shallow physiographic regimes 

 where nearshore hardbottom occurred. 



Although suggestive of nursery value, these lines 

 of evidence need to be viewed in the appropriate con- 

 text. High abundances of early life stages compared 

 with adults do not guarantee that a habitat is a valu- 

 able nursery. High mortality rates in many reef fish 

 populations (Sale, 1980; Shulman and Ogden, 1987; 

 Richards and Lindeman, 1987; Jones, 1991) suggest 

 that early stages will typically be more abundant 

 than adults. If spatial distributions of all life stages 

 are homogeneous, all habitats will have more early 

 stages than adults. However, the abundances of early 

 stages on nearshore reefs probably reflect more than 

 just larger numbers of homogeneously distributed 

 recruits. Newly settled stages of eight of twelve spe- 

 cies of grunts and eight of nine species of snappers 

 of the southeast mainland Florida shelf have been 

 recorded primarily in depths less than ten m (Linde- 

 man et al., 1998). Adults of most species are typi- 

 cally uncommon or absent in shallow habitats. There 

 is considerable evidence for cross-shelf habitat seg- 



