McKinzie and Szedimayer: Mortality estimates for Balistes capriscus based on acoustic telemetry 177 
natural mortality [M] of 0.28; i.e., Z=F+M) determined 
from the 2015 stock assessment (SEDAR, 2015). From 
this 2015 assessment, the estimated F for 2013 was 0.12, 
but previous estimates indicate that F exceeded sus- 
tainable levels and ranged from 0.35 to 0.67 in that year 
(SEDAR, 2011; Burton et al., 2015). Estimates of M have 
been relatively constant over time, ranging from 0.27 to 
0.28 (SEDAR, 2011; Burton et al., 2015; SEDAR, 2015; 
GMFMC, 2017b). 
Estimates of F, M, and Z for gray triggerfish are criti- 
cal for management but are highly dependent on fishery- 
dependent data from the Marine Recreational Information 
Program, which is known to have difficulties (NASEM, 
2017). Also, it is relatively easy for management to esti- 
mate Z from age—frequency analyses, but it is more difficult 
to separate Z into F and M for an exploited stock (Ricker, 
1975). Some authors have proposed methods to estimate 
M with equations that attempt to relate WM to life history 
characteristics, such as maximum length, maximum age, 
or growth rates (Hewitt and Hoenig, 2005; Charnov et al., 
2013). Mortality can also be estimated through mark- 
recapture studies on the basis of the number of tagged fish 
that were caught and reported by fishermen (Pine et al., 
2012). However, there are inherent caveats with conven- 
tional mark-recapture studies; for example, fishermen 
not reporting catch of tagged fish, fish shedding tags, and 
fish emigrating from study areas (Schwartz, 2000; Pollock 
et al., 2001; Denson et al., 2002; Pine et al., 2003, 2012). 
Telemetry methods have been used in terrestrial ani- 
mals for some time (Trent and Rongstad, 1974; Pollock 
et al., 1995) and more recently have been applied to 
aquatic organisms (Hightower et al., 2001; Heupel and 
Simpfendorfer, 2002; Pollock et al., 2004; Starr et al., 2005; 
Knip et al., 2012). In the northern Gulf of Mexico, telem- 
etry methods have been successfully used to estimate F 
and M of red snapper residing on natural and artificial 
reefs, independent of tag returns by fishermen (Topping 
and Szedlmayer, 2013; Williams-Grove and Szedlmayer, 
2016a; Mudrak and Szedlmayer, 2020). Telemetry can 
also be used to estimate non-reporting by fishermen and 
tagging mortality (Hightower et al., 2001; Williams-Grove 
and Szedlmayer, 2016a; Mudrak and Szedlmayer, 2020). 
In previous studies, telemetry methods have been applied 
to gray triggerfish to estimate movement patterns (Herbig 
and Szedlmayer, 2016; McKinzie, 2018; Bacheler et al., 
2019a, 2019b), and responses to baited fish traps (Bacheler 
et al., 2018), but have not been reported for mortality esti- 
mations. In our study, we applied telemetry methods to 
estimate mortality of gray triggerfish from artificial reef 
habitats in the northern Gulf of Mexico from 2013 through 
2017. We estimated M, F, and Z with a known-fate model 
(Kaplan and Meier, 1958; Pollock et al., 1989) on the basis 
of fine-scale positions obtained from Vemco Positioning 
Systems! (VPS; Innovasea Systems, Boston, MA) com- 
posed of acoustic receivers and transmitters, and these 
1 Mention of trade names or commercial companies is for identi- 
fication purposes only and does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 
estimates can potentially enhance management efforts for 
this species. 
Materials and methods 
Study area and array design 
Covering 64 km”, the study area was located 23-35 km 
south of Dauphin Island, Alabama, in the northern Gulf of 
Mexico. The study area contained 26 steel-cage artificial 
reefs (2.5 x 1.3 x 2.4 m), each positioned 1.4—1.6 km apart 
at depths of 18-35 m (Fig. 1). The artificial reefs were 
deployed at unpublished locations from 2006 through 2010 
within a designated reef building zone (Hugh Swingle 
General Permit Area; Syc and Szedlmayer, 2012). A VPS 
array of acoustic receivers (VR2W, Innovasea Systems; 
69 Hz) and synchronization transmitters (V16-6x, Inno- 
vasea Systems; 69 kHz, transmission delay of 540-720 s) 
was deployed at each of 5 reef sites to monitor patterns of 
fine-scale movements (in meters) and to estimate mortal- 
ity of gray triggerfish tagged with acoustic transmitters 
(V13-1L or V13P-1L, Innovasea Systems; 69 kHz, trans- 
mission delay of 40-80 s, battery life of 566-991 d). The 
depths at the 5 reef sites were 18, 23, 25, 26, and 35 m. 
In addition, a single VR2W acoustic receiver was placed 
at each of 21 surrounding reef sites to detect large-scale 
movements (in kilometers) and emigrations away from 
VPS sites (Fig. 1). 
Each of the 5 VPS sites consisted of 5 acoustic receivers: 
a center receiver placed 10—20 m north of the reef site and 
4 receivers placed 300 m north, south, east, and west of 
the center receiver (Piraino and Szedlmayer, 2014). The 
V16-6x synchronization transmitters were attached 1 m 
above each receiver to standardize the internal receiver 
clocks (Piraino and Szedlmayer, 2014). The design of the 
receiver arrays at VPS sites permitted high detection effi- 
ciency of tagged gray triggerfish (>88%; McKinzie, 2018). 
Every 4-6 months, data were downloaded and all receiv- 
ers were exchanged. Data from the VPS receivers were 
sent to Innovasea Systems for post-processing after every 
download. A stationary control transmitter (V13-1L) was 
placed at a known location within each VPS array to val- 
idate the accuracy of Innovasea-derived positions and to 
ensure continuous data collection throughout the study 
period. 
Each of the 21 surrounding reef sites contained a sin- 
gle VR2W receiver that detected presence and absence of 
tagged fish. Every 6-12 months, data were downloaded 
and these single receivers were exchanged. The maximum 
radius for transmitter detection around each receiver 
was estimated to be 770 m (McKinzie, 2018). Therefore, 
tagged gray triggerfish were detected in 82% (53 km”) of 
the area within the array of 21 receivers used to monitor 
large-scale movements, leaving 18% (11 km?) of the area 
with little coverage. The areas of low detection occurred 
over paths that were 100-300 m wide between the cov- 
erage areas of single receivers at surrounding reef sites 
(McKinzie, 2018). 
