Dapp et al.: Immediate mortality of Prionace gtauca and Galeocerdo cuvier caught by pelagic longlines 
29 
Data screening 
Data screening was conducted on the USPL data set to 
remove longline sets or individual captures that were 
deemed likely to be inaccurate. A summary of longline 
sets and individual captures removed from the data set 
is available in Supplementary Table 1. Longline sets 
and recorded captures were removed from the analysis 
in the following cases: 
• Longline sets that were made before 1992 be- 
cause records from 1982 through 1992 did not 
specify whether discards were alive or dead 
• Multiple gear types (e.g., pelagic longline and 
gillnet or pelagic longline and bottom longline) 
were used or sets did not use longline gear 
• Improbable SST (e.g., <1°C), or no SST was 
reported 
• Location data were likely to be inaccurate be- 
cause all latitude and longitude values were re- 
ported as 0°, latitude and longitude values were 
unreported, or longline set location was recorded 
outside of the designated geographic zones 
• Target catch was not reported as swordfish {Xi- 
phias gladius), species of shark, dolphinfish 
{Coryphaena hippurus), or any species of tunas 
{Thunnus spp.). If target catch was not reported 
as one of these categories, the target catch was 
specified as other and v/as considered unknown 
• Specific groupings (e.g., geographic zones, target- 
ed species, regulatory periods) with less than 100 
individuals of a given species 
• Longline sets where fishermen retained any in- 
dividuals of the species being examined. This ap- 
proach was necessary because the condition of 
whether sharks were alive or dead was not re- 
corded at the time of processing. Mortality of in- 
dividuals that were retained was unrelated to the 
environmental and operational factors examined 
in our analysis 
• Longline sets that targeted sharks, because re- 
tention was likely in these longline sets. Longline 
sets that targeted sharks were included for the 
analysis of blue shark {Prionace glauca) because 
this species was not typically retained by shark- 
targeting longline fishermen (i.e., in less than 5% 
of longline sets that reported targeting sharks be- 
fore data screening) 
• Geographic zone groupings that had less than 100 
recorded captures during each regulatory period 
when interaction effects (blue shark only) were 
examined. A table that shows excluded groupings 
and their associated sample sizes can he found 
in the supplementary materials section (Suppl. 
Table 2) 
Some variables recorded in the USPL were excluded 
from our analyses. Use of live or dead bait was not in- 
cluded as a factor because the majority of longline sets 
had either dead or unrecorded bait types (e.g., 99.8% of 
the examined longline sets that captured blue shark). 
Gangion length was not included as a factor in the 
analysis because the units of measurement (e.g., cen- 
timeters and inches) that were used were unspecified 
and varied among longline sets. Immediate mortality 
rates during demersal longline sets were not compared 
with those during pelagic longline sets because the few 
individuals remaining in the data set after data screen- 
ing was completed were recorded on demersal longline 
sets (e.g., less than 0.1% of blue sharks were caught 
during demersal longline sets). Although the number of 
hooks between floats has been used in previous studies 
as a proxy for target catch (Serafy et al., 2012; Galla- 
gher et al., 2014), this variable was not included in our 
analysis because target catch was explicitly recorded in 
the USPL, and, therefore, available. 
Species-specific immediate mortality rates (ex- 
pressed as percentages) were calculated by dividing 
the number of dead discards by the number of total 
discards. This approach was used because immediate 
mortality was not directly recorded by fishermen in the 
USPL, unwanted shark bycatch is typically released by 
commercial fishermen without bringing it onboard ves- 
sels (Moyes et al., 2006), and regulations require that 
live shark bycatch is released unharmed (NMFS^). In 
addition, we removed from our analysis longline sets 
that retained (rather than released) individuals of the 
species being examined. 
Species that were likely to be identified accurately 
by fishermen and considered in our analysis included 
blue sharks, porbeagle {Lamna nasus), oceanic whitetip 
sharks {Carcharhinus longimanus), and tiger sharks 
{Galeocerdo cuvier) (Mandelman et al., 2008; Baum 
and Blanchard, 2010). Although we report the likeli- 
hood of retention, being discarded dead, and being dis- 
carded alive for 2 occasionally retained species (oceanic 
whitetip sharks and porbeagles; NMFS^), we were un- 
able to analyze the immediate mortality rates of these 
species using the USPL data set because it could not 
be determined whether individual animals were alive 
or dead when landed and retained. Instead, we exam- 
ined immediate mortality in bycatch of blue and tiger 
sharks because these species have meat of low value 
(Vannuccini, 1999; Mandelman et al., 2008; Simpfen- 
dorfer, 2009) and are rarely retained by US. fisheries 
(NMFS^) and because a metric of dead discards in rela- 
tion to live discards would accurately reflect immediate 
mortality rates following capture. 
The effect of capture location on immediate mortal- 
ity was investigated by using the 11 zones of the US. 
pelagic longline fishery that are based on geography 
2 NMFS (National Marine Fisheries Service). 1993. Fish- 
ery management plan for sharks of the Atlantic Ocean, 287 
p. [Available from website, accessed December 2015.] 
3 NMFS (National Marine Fisheries Service). 2006. Final 
consolidated Atlantic highly migratory species fishery man- 
agement plan, 1600 p. [Available from website, accessed 
March 2016.] 
NMFS (National Marine Fisheries Service). 2015. Stock 
assessment and fishery evaluation (SAFE) report for Atlan- 
tic highly migratory species. 170 p. [Available from website, 
accessed March 2016.] 
