Locascio and Burton: A passive acoustic survey of fish sound production at Riley's Hump 
105 
Riley's Hump 
B 
24032’ 
24 ° 31 ’ 
24 ° 30 ’ 
24029 1 
“ 83 ° 08 ' 
" 83 ° 07 ' 
" 83 ° 06 ' 
“ 83 ° 05 ' 
Figure 1 
(A) Location of Riley’s Hump within the Tortugas South Ecological Re- 
serve and in relation to mainland Florida, and (B) locations of acoustic 
monitoring sites on Riley’s Hump indicated on the multibeam bathy- 
metric image of this geologic feature. The scale of the image represents 
meters below sea level (mbsl). The deepwater site (RHDW) that would 
be on the left side of the image is located a bit farther west than what 
is indicated in this figure. The depth of this site was approximately 
60 m; depths at all other acoustic montioring sites located on Riley’s 
Hump were approximately 30 m. Bathymetric image is taken from fig- 
ure 7 of Mallinson et ah, 2003. 
15,094 Hz sample rate; they were moored 
to steel rebar anchored in the limestone 
substrate. All recorders were deployed 
and recovered by scuba divers, except 
for the deployment of the deepwater re- 
corder (RHDW), which was dropped in 
a weighted housing unit from the swim 
platform of the MV Spree and recovered 
by scuba divers. Visual surveys were con- 
ducted along transects of 50 m by divers 
to document the presence of grouper spe- 
cies at acoustic monitoring sites (12, 12A, 
15, RH1) during April 2010 and January 
and February 2011. All field operations 
were conducted from the MV Spree. 
Each 10-s acoustic file was analyzed in 
MATLAB, vers. R2009B (The Mathworks 
Inc., Natick, MA) with a fast Fourier 
transform to generate a power spectrum 
from which the band sound pressure level 
(SPL) in 100-Hz-wide bins was calculat- 
ed. Patterns in fish sound production on 
daily and seasonal scales were examined 
in plots of power spectra and, for grouper 
species, by direct counts of calls in a sub- 
sample of 10,000 files randomly selected 
from the entire database of acoustic re- 
cordings. The number of calls counted for 
each grouper species in the subsample 
was normalized as a ratio of the number 
of calls to number of files reviewed for 
each month at each site. Spectrograms of 
acoustic recordings were reviewed with 
Adobe Audition, vers. 2.0 (Adobe Systems 
Inc., San Jose, CA) to identify species 
present in the recordings or previously 
undescribed calls. Daily sound patterns 
were estimated by binning the number 
of calls for each species (sites combined) 
into three 8-h periods (0000-0800, 0800- 
1600, and 1600-0000, all local time) and 
comparing them by means of tests for 
analysis of variance. Acoustic time-series 
data were examined for peaks associated 
with lunar phases. 
Custom underwater audio and video 
(A/V) systems were used to verify sourc- 
es of fish sound production and to under- 
stand the behavioral context associated 
with that sound production. The record- 
ing system included a low-light, 0.001- 
lux, black-and-white flat lens board camera and 2 HTI 
96-MIN hydrophones (sensitivity -164 dB reference 
pressure [re]: IV/pPa) (High Tech Inc., Long Beach, 
MS) that recorded to a ChaseCam deck (Chase Vision, 
LLC, Cleveland, TN). Each A/V system was deployed in 
a clear, waterproof housing unit and placed on the sea- 
floor overnight at select sites where long-term acoustic 
recorders were deployed. The A/V system recorded con- 
tinuously to compact flash memory cards for approxi- 
mately 20 h. A Sony HDR-XR100 video camera (Sony 
Corp., Tokyo) fitted with an HTI 96-MIN hydrophone 
was used during visual surveys conducted by divers 
to record fish sound production and associated behav- 
ior. Audio and video data were reviewed with Corel 
Video Studio software, vers. X6 (Corel Corp., Ottawa, 
Ontario) and with Adobe Audition software to identify 
species-specific acoustic signals and associated behav- 
ioral context. 
