Harter et al.: Fish assemblages associated with grouper pits in the Gulf of Mexico 
427 
18 
1 
0 
f 16 
P<0.001 
CO 
% 14 
> 
TD 12 
C 
8 »• 
1 
\ c 
! * 
0 
O) A . 
2 4 
0 
2 • 
a / 
a 
2012 2013 2014 2015 
Figure 3 
Average abundance of lionfish ( Pterois volitans or P. miles) mea- 
sured as number of individuals by year from surveys conducted 
from 2012 through 2015 at Pulley Ridge off southwestern Flor- 
ida. The P-value is given for the results from one-way ANOVA. 
Statistically significant differences are noted with different let- 
ters (a— c); for years with the same letter, the difference is not 
significant. Error bars indicate standard errors of the mean. 
SIMPER routine indicated that the primary 
species responsible for the groups clustering 
in this way were cardinalfish and 3 damselfish 
species, including the sunshinefish ( Chromis 
insolata ), the purple reeffish ( Chromis scotti), 
and the yellowtail reeffish ( Chromis enchrys- 
ura). These species, as well as several others, 
including the scamp, striped grunt, and school 
bass, had higher abundances in grouper pits 
with either both predators or lionfish only 
than in grouper pits with no predators or red 
grouper only. Because the scamp was the most 
abundant economically important species ob- 
served (with the exception of red grouper), we 
needed to test the potential effect this species, 
as a predator, could have on fish assemblages 
in grouper pits through predation. The results 
of tests with the ANOSIM routine support the 
assertion that the presence of scamp did not 
affect fish assemblages of the grouper pits 
(i?-statistic=0.169). 
Species diversity and evenness did not dif- 
fer considerably by HAPC or region but were 
different by year and predator presence (Table 
3). All sampling years had similar species di- 
per pits in 2012, 73.2% in 2013, 91.4% in 2014, and 
86.5% in 2015. Although the maximum number of li- 
onfish observed in a single pit was 74, the average 
abundance was 6.10 individuals/pit (SE 0.60), and the 
average abundance increased significantly over time 
(P<0.0001) (Fig. 3). Abundance of lionfish throughout 
the sampling area and presence of red grouper are 
displayed in Figure 4. Both species were distributed 
throughout the region, but the highest abundances of 
lionfish were located primarily outside the HAPC on 
the west ridge as well as a few places along the main 
ridge inside the HAPC. 
Analyses of multivariate fish communities 
Fish species composition associated with grouper pits 
was not significantly different for year, region, or HAPC 
factors. It did, however, differ depending on the preda- 
tor species present CR-statistic=0.402, P<0.01). Three 
significantly different groups resulted from the SIM- 
PROF test (P< 0.05), indicated by the letters on the 
MDS plot (Fig. 5). Grouper pits with lionfish only and 
with both predators formed one group, which meant 
that their fish assemblages were not significantly dif- 
ferent from one another. Grouper pits with red grou- 
per only formed their own group, as did those with no 
predators, which meant that their fish assemblages 
were significantly different from all other groups. The 
groups clustered together in this fashion at 80% sim- 
ilarity — a result that meant that the species composi- 
tion of grouper pits with lionfish only and with both 
predators was 80% similar. Pairwise tests with the 
Figure 4 
Average abundance of lionfish ( Pterois volitans 
or P. miles) measured as number of individuals 
per grouper pit, for each pit surveyed during 
2012-2015 off southwestern Florida. The size 
of the white circles indicates the level of abun- 
dance. Small black circles indicate the presence 
of a red grouper ( Epinephelus morio). The poly- 
gon outlined in black represents the Habitat 
Area of Particular Concern at Pulley Ridge. 
