Santos et al.; Effects of leader material on catches of longline fisheries in the Indian Ocean 
223 
for catches with greater than 30 individuals. The re¬ 
sponse variable of the models was the proportion of the 
catches (or dead specimens in the mortality models) in 
each longline set, calculated as the number of animals 
(or dead specimens in the mortality models) given the 
number of hooks used in each set. The odds-ratios of 
the parameters with their respective 95% confidence 
intervals (CIs) were calculated and used for interpre¬ 
tation. For this purpose, the monofilament leader was 
considered the baseline (control) configuration, and the 
odds-ratios were calculated for changing to the alterna¬ 
tive wire (experimental) leader. 
Data analysis for this study was carried out with 
R statistical software, vers. 3.2.0 (R Core Team, 2015). 
Most analyses were performed with functions available 
in the core R program. Additional libraries were used 
for the Levene tests to compare homogeneity of vari¬ 
ances (library car; Fox and Weisberg, 2011) and for the 
permutation tests (library perm; Fay and Shaw, 2010). 
Plots were built by using ggplot2 (Wickham, 2009), and 
the map was built by using mapplots (Gerritsen, 2014), 
shapefiles (Stabler, 2013), and a function for the north 
arrow created by Tanimura et al. (2007). 
Results 
Power tests 
Overall, a total of 82,656 hooks were used during the 
experimental fishing sets (82 sets), corresponding to 
41,328 hooks of each leader type. According to the con¬ 
ducted power analysis, the number of hooks deployed 
is larger than that necessary to detect a 50% (5904 
hooks) or 25% (23,613 hooks) change in the catch 
rates (in number) of the blue shark, the most cap¬ 
tured shark species in Indian Ocean pelagic longline 
fisheries. The same was also true for detecting such 
a change in swordfish catch rates (5088 and 20,346 
hooks, respectively). 
Catch composition 
A total of 33 taxa were caught during this study, spe¬ 
cifically 11 species of sharks (2 species belonging to 
Mobulidae), 6 species of billfish, 3 species of tuna, 12 
species of other bony fishes and a single species of ma¬ 
rine turtle (Table 1). For the species composition of the 
catch, the highest number of species was recorded for 
wire leaders (31 out of the 32), compared with 24 taxa 
caught with monofilament leaders. For the different 
groups of species, monofilament leaders in contrast to 
wire leaders resulted in the catch of an equal number of 
billfish and tuna species, but a lower number of sharks 
(7 us. 9) and other bony fishes (8 us. 12, respectively) 
A total of 2385 specimens were caught during this 
study, of which 329 were discarded either because they 
were noncommercial species, belonged to species whose 
retention is prohibited by the lOTC, or they were dam¬ 
aged by depredation. Among the overall catch there 
were 1077 billfishes, 845 sharks, 412 other bony fishes, 
50 tuna, and a single marine turtle, as summarized in 
Table 1. A total of 15 taxa were retained in the experi¬ 
ments, the swordfish and blue shark being the most 
abundant species, representing 50.1% and 37.3% of 
the retained catches in number, respectively. All the 
remaining species represented less than 1.5% of the 
total retained catch in number, the exceptions being 2 
other bony species, the dolphinfish {Coryphaena hip- 
purus) and the escolar {Lepidocybium flauobrunneum), 
which represented 6.7% and 3.0%, respectively (Table 
1). In addition, 18 taxa were systematically discarded, 
specifically 9 sharks, 8 other bony fishes and a marine 
turtle, the loggerhead turtle {Caretta caretta). Amongst 
the discarded groups of species, other bony fishes and 
sharks accounted for 64.7% and 15.8% of the discarded 
specimens, whereas commercial species such as bill- 
fishes (10.3%) and tunas (0.9%) were also discarded 
owing to predation or the small size of specimens. The 
most commonly discarded bony fishes were the long- 
nose lancetfish {Alepisaurus ferox) and snake mackerel 
(Gempylus serpens), accounting for 28.6% and 26.7% 
of discards, respectively. In the case of elasmobranchs, 
the pelagic stingray (Pteroplatytrygon uiolacea) was the 
most commonly discarded species (7.6%), followed by 
the 2 shark species prohibited to be fished: the bigeye 
thresher (3.0%) and the oceanic whitetip shark (2.4%). 
The frequency of occurrence varied greatly among 
species, with the swordfish being the most frequently 
caught species (present in 100% of the sets), followed 
by the blue shark (91.5%), dolphinfish (37.8%), long- 
nose lancetfish (36.6%), and snake mackerel (29.3%) 
(see details in Table 1). 
Catch rates 
The effects of the leader type on the CPUEn were group 
and taxon specific. The values of CPUEn were higher 
when wire leaders were used for other bony fishes and 
sharks, as well as for the overall catch (on average 
32%, 30% and 13%, with 95% CIs of 8-60%, 13-49% 
and 5-23%, respectively). At the species level, only 
CPUEn for blue sharks showed a significant increase 
on wire leaders, which was on the order of 31%, with 
95% CIs varying between 14% and 52% (Fig. 3). Sword¬ 
fish and billfishes showed lower CPUEn on wire lead¬ 
ers, but the detected differences were not significant. 
The effects of the leader type on the CPUEb of retained 
species were again group and taxon specific. However, 
in the CPUEb, higher catch rates on wire were noted 
only for the shark group, blue shark in particular and 
the overall retained catch (on average 53%, 56% and 
15%, respectively). In the case of the target species 
(swordfish), although the wire leaders had a negative 
effect (decrease of 11%) on the mean CPUEb, this effect 
was not statistically significant (Tables 1 and 2). 
In terms of the VPUE, no significant differences were 
detected in changes from monofilament to wire lead¬ 
ers (permutation test: difference in means= -$44.99 
(-€32.62), P=0.874) because the decrease of swordfish 
