Harter et al. : Assessment of fish populations and habitat on Oculina Bank 
201 
Table t (continued) 
open 
OHAPC 
OECA 
PAY 
RUB 
OUT 
PAV RUB 
OUT 
SD 
LO 
PAV 
RUB 
OUT 
SD 
LO 
Sphyraenidae 
Sphyraena barracuda 
— 
— 
— — 
— 
— 
— 
— 
0.07 
— 
— 
— 
Bothidae 
Cyclopsetta fmbriata 
0.14 
Undetermined 
— 
0.14 
0.98 — 
— 
0.83 
— 
2.44 
— 
0.10 
— 
0.29 
Balistidae 
Balistes capriscus 
— 
— 
— — 
— 
— 
— 
— 
— 
0.10 
— 
— 
Monacanthidae 
Aluterus monoceros 
— 
— 
— — 
— 
— 
— 
— 
0.07 
— 
— 
— 
Stephanolepis hispidus 
— 
— 
— — 
0.07 
— 
— 
— 
— 
0.10 
— 
— 
Monacanthus spp. 
— 
— 
— — 
0.19 
— 
— 
— 
0.07 
— 
— 
— 
Ostraciidae 
Lactophrys quadricornis — 
— 
— 
— — 
- 
- 
— 
— 
0.21 
— 
— 
— 
Lactophrys spp. 
— 
— 
— — 
— 
— 
— 
— 
— 
0.17 
— 
— 
Tetraodontidae 
Sphoeroides spengleri 
2.88 
0.51 
3.62 0.41 
0.42 
— 
— 
2.72 
0.83 
0.16 
— 
0.44 
Sphoeroides spp. 
— 
— 
— — 
0.34 
— 
— 
— 
0.07 
— 
— 
0.57 
Diodontidae 
Chilomycterus spp. 
0.20 
found in every habitat and management 
area. Rock hind ( Epinephelus adscensio- 
nis ), speckled hind ( E . drummondhayi), 
grey triggerfish ( Balistes capriscus ), and 
grunts (family Haemulidae) were only 
observed in the OECA. 
Multivariate analyses based on 39 fish 
species across 473 transects indicated no 
differences in fish assemblages among 
hardbottom habitat types or management 
areas. MDS ordination portrayed a poten- 
tially useful representation of relation- 
ships among ROV transects in two-dimen- 
sional space (stress=0.2; see Clarke and 
Warwick, 2001) and showed no distinct 
groupings (Fig. 2). ANOSIM results con- 
firmed these conclusions, fish assemblages 
were not significantly different among 
hardbottom habitat types (ANOSIM, Glob- 
al R=0.128, P= 0.001 ) or management ar- 
eas (ANOSIM, global #=0.061, P=0.002). 
For ANOSIM, the P value is highly sensi- 
tive to sample number and, therefore, the 
likelihood of committing a type-I error is 
high. For that reason, the R value is more 
important than the P value. R equals 0 
when groups are the same and R equals 
1 when groups are different (Clarke and 
Warwick, 2001). 
Among habitat types, species richness 
was highest on rock outcrops and low- 
est for standing dead O. varicosa (Fig. 3). 
Average taxonomic distinctness (4 + ) was 
Figure 2 
Multidimensional scaling (MDS) ordination of habitats (A) and manage- 
ment areas (B) based on the Bray-Curtis similarity matrix calculated 
from square root transformed fish densities (39 species). Data were 
collected from remotely operated vehicle (ROV) transects conducted on 
the Oculina Bank during April-May 2003 and October 2005. 
