190 
Fishery Bulletin 117(3) 
Figure 1 
Map of sampling locations where bottom longline gear was set in the northern 
Gulf of Mexico during March-April 2017. Crosses indicate longline sets baited 
with northern shortfin squid (Ilex illecebrosus), and circles represent long- 
line sets baited with Atlantic mackerel (Scomber scombrus ). Gear was set in 
waters off Louisiana (LA), Mississippi (MS), Alabama (AL), and Florida (FL). 
Gray lines indicate the 10-, 20-, 30-, 40-, 50-, and 100-m isobaths. 
squid ( Illex illecebrosus ). Conversely, tele- 
osts commonly targeted with bottom long- 
line gear in the northern Gulf of Mexico, 
such as red snapper ( Lutjanus campecha- 
nus), yellowedge grouper ( Hyporthodus 
flauolimbatus), and tilefish ( Lopholatilus 
chamaeleonticeps), showed no preference. 
The authors acknowledged that the use of 
alternating bait types could have biased 
results because of potential interaction 
effects of baits, a possibility suggested 
by Watson et al. (2005) and Foster et al. 
(2012). To address this issue and fur¬ 
ther investigate the efficacy of bait type 
in reducing shark bycatch, the objective 
of this study was to assess if the prefer¬ 
ence of sharks for Atlantic mackerel over 
northern shortfin squid baits reported by 
Driggers et al. (2017) would be evident in 
a single-bait experimental design. 
Materials and methods 
Bottom longline gear was deployed from the NOAA Ship 
Oregon II in the northern Gulf of Mexico from 26 March 
through 19 April 2017. All sets were made within an 
area bounded by 87.50°W to 93.00°W longitude at depths 
between 15 and 85 m (Fig. 1). Stations were preselected 
on the basis of a random starting point and subsequent 
stations being spaced 18.5 km apart in a grid to maxi¬ 
mize sampling effort and minimize effects of sampling 
sites being in close proximity to one another. On several 
occasions, preselected stations were moved to avoid haz¬ 
ards to navigation (e.g., vessels, petroleum platforms, 
and shipping lanes) or excessively fast currents (>1 m/s). 
Sampling gear consisted of 1.8 km of 4-mm diameter 
monofilament mainline and 100 gangions. Each gangion 
was constructed of a size 148-8/0 snap, 3.2 m of 3-mm 
diameter monofilament, 0.5 m of 2.4-mm diameter fish¬ 
ing wire, and a hook (#15/0 non-offset, circle, Mustad 
#399600*, O. Mustad & Sqn A.S., Gjpvik, Norway) and 
was identical to that used in Driggers et al. (2017). Each 
hook was baited with Atlantic mackerel or northern 
shortfin squid cut to fit the circle hooks. Bait consisted 
either entirely of Atlantic mackerel or northern shortfin 
squid for each longline set with bait randomly assigned 
at the beginning of each set. 
Bottom temperature (in degrees Celsius) and salinity 
were measured at each sampling location, while the long- 
line gear was in the water, by using an SBE 911plus CTD 
water profiler (Sea-Bird Scientific, Bellevue, WA). Soak 
time, defined as the time elapsed from completing deploy¬ 
ment of the gear to beginning retrieval of the gear, was 
limited to 1 h in an effort to minimize mortality rates of 
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. 
captured organisms and to maximize the number of long- 
line sets conducted during the allotted sampling period. 
Further, the relatively short soak time reduced data loss 
associated with shark depredation and potential bias asso¬ 
ciated with capturing large sharks feeding upon smaller 
captured fish, a behavior that would not reflect prefer¬ 
ence for Atlantic mackerel or northern shortfin squid. 
Upon capture, each fish was identified to species and its 
fork length was measured to the nearest millimeter on a 
straight line along the axis of the body from the tip of the 
rostrum to the caudal notch. 
Depth, temperature, salinity, soak-time, and fish-length 
data associated with sampling locations where each bait 
was used were assessed for normality and homoscedas- 
ticity by calculating kurtosis, skewness, and E-ratio. Data 
were considered normally distributed when kurtosis and 
skewness were between -2 and 2. For data that adhered 
to the assumptions of parametric statistics, /-tests were 
used to compare mean values; for data that did not meet 
assumptions, median values were compared by using a 
Mann-Whitney W test. To determine if the number of indi¬ 
viduals captured on each bait was significantly different 
from 1:1, a chi-square test with Yates correction for con¬ 
tinuity was used on a species-specific basis. Sharks that 
were captured on hooks occupied by a smaller captured 
fish were removed from all analyses (2 blacktip sharks, 
Carcharhinus limbatus), and fish not measured because of 
shark-inflicted damage were excluded from comparisons 
of body length. 
To account for the possibility that catch data were 
biased because of schooling behavior or areas of high 
abundance (i.e., high catch in limited cases because of 
habitat specificity or aggregations), catch data for each 
species were examined to determine if an unusually high 
number of individuals were caught on any specific set. 
