428 
Fishery Bulletin 115(3) 
Predator presence 
• No predator 
4 Red grouper 
▼ Lionfish 
■ Both 
Similarity 
80 
2D stress: 0 
a 
® 
Figure 5 
Plot of nonmetric multidimensional scaling ordination, derived from the Bray- 
Curtis similarity matrix constructed by using fourth-root transformed fish abun- 
dances averaged for each predator presence or absence category: no predator; 
red grouper ( Epinephelus morio) only; species of lionfish ( Pterois spp.) only; or 
both predators. Groups of fish assemblages surveyed at grouper pits during 
2012-2015 off southwestern Florida are shown at 80% similarity. The closer 2 
groups are in space, the more similar they are to each other. The letters (a-c) 
above the group symbols represent the results of the test with the SIMPROF 
routine in PRIMER software, and statistically different fish assemblage groups 
are indicated by different letters. 
versity and evenness, with the exception of 2012, when 
those values were lower than those of other years. The 
lower species diversity in 2012 may have been observed 
because only the main ridge was sampled in that year. 
Species diversity and evenness were very similar for 
grouper pits with both predators, lionfish only, and red 
grouper only but were considerably lower for pits with 
no predators. In contrast, the highest number of spe- 
cies was observed in 2012 (again, this observation of a 
higher number of species may have occurred because 
only the main ridge was sampled in 2012), in grouper 
pits with either lionfish only or with both predators, 
inside the HAPC, and on the main ridge. 
Analyses of univariate fish abundance 
Average abundances of small fish associated with grou- 
per pits were significantly different among years and 
predator groups (Fig. 6). Abundances significantly in- 
creased from 2012 to 2015 (PcO.0001) and species were 
more abundant in grouper pits with a predator present 
(P=0.006). Average abundances of small fish were not 
significantly different among regions or HAPC groups 
(P> 0.05). In contrast, average abundances of schooling 
fish were not significantly different for any of the fac- 
tors analyzed (P>0.05). Average abundances of large 
fish associated with the grouper pits were significantly 
different only among predator groups, where higher 
abundances were observed in grouper pits with either 
red grouper only or with both predators (P<0.001; Fig. 
7). As with the multivariate analyses, we tested wheth- 
er presence of scamp had an effect on abundances of 
Table 3 
Biodiversity indices for fish communities observed dur- 
ing video surveys conducted with a remotely operated 
vehicle at the Pulley Ridge mesophotic coral ecosystem 
in the Gulf of Mexico. Values are shown for each factor: 
year; predator presence; inside the Pulley Ridge Habi- 
tat Area of Particular Concern (HAPC) (area protected 
from fishing) and outside the HAPC (unprotected area); 
and the region of Pulley Ridge. S=total number of spe- 
cies; 7T=Shannon-Wiener function of species diversity; 
J’=Pielou’s evenness. 
S 
H’ 
J’ 
Year 
2012 
50 
1.15 
0.29 
2013 
41 
2.09 
0.56 
2014 
46 
2.33 
0.61 
2015 
43 
2.26 
0.60 
Predator 
No predator 
32 
0.64 
0.18 
Red grouper only 
24 
2.10 
0.66 
Lionfish only 
64 
2.56 
0.62 
Both predators 
56 
2.48 
0.62 
HAPC 
Inside 
65 
2.38 
0.57 
Outside 
42 
2.36 
0.63 
Pulley Ridge 
Main ridge 
61 
2.05 
0.50 
Central basin 
43 
2.16 
0.57 
West ridge 
40 
2.26 
0.61 
Off main ridge 
25 
1.98 
0.61 
