Tolotti et al.: Vertical movements of oceanic whitetip sharks (Carcharhinus longimanus) 
387 
□Type I BType II □ Type III DNA 
0% 20% 40% 60% 80% 100% 
AOCS3 
AOCS4 
AOCS5 
AOCS6 
AOCS7 
IOCS1 
Figure 4 
Proportion of the 3 different behavior types of vertical movement 
during day and night observed on the depth time series of oceanic 
whitetip sharks ( Carcharhinus longimanus ) tagged in the Atlantic 
and Indian oceans between 2011 and 2012. The identity codes (e.g., 
AOCS3) identify individual sharks. 
in the model with data from multiple sharks (Fig. 7). 
The set of explanatory variables in the best models for 
individual sharks varied slightly between sharks, and 
horizontal positions, unlike MLD and SST, were always 
significant (Fig. 7). Summary tables with model results 
are presented in Supplementary Table 2 and Supple- 
mentary Figure 5 (online only). 
To assess in detail the effect of environmental fac- 
tors on variations of vertical movement, the cumula- 
tive sums of MLD, SST, and depth SD were plotted for 
each individual. The cumulative sums were subtracted 
from the mean and rescaled to highlight periods when 
values were above or below the mean (Fig. 8). These 
plots reflect and facilitate understanding of the model 
results. The results for shark IOCS1 provide an ex- 
ample of the overall observed trend. For this individ- 
ual, a perfectly aligned correlation between MLD and 
vertical activity was observed. When MLD was above 
average and, therefore, the species preferred environ- 
ment was extended, the SD was also above average. 
The opposite trend was also true. It is not surprising 
that the MLD was a highly significant factor in the 
GAM for this individual; the GAM had a high degree 
of deviance explained (64.7%). Another interesting case 
is shark AOCS3. In the beginning of the monitoring 
period for this shark, the same relationship with MLD 
and vertical activity (SD), as that found for IOCS1, was 
observed. Around May, however, a major shift occurred. 
Instead of decreasing with MLD, vertical activity of 
this shark started to increase regardless of the reduc- 
tion of preferred environment. At this time, the SST 
started to exceed average values; hence, its addition as 
a significant factor in the model of AOSC3. This same 
trend was observed for shark AOCS5. 
Discussion 
Recent studies have shown that the oceanic 
whitetip shark is an epipelagic predator 
largely confined to the mixed layer (Musyl 
et al., 2011; Howey-Jordan et al., 2013; To- 
lotti et al., 2015b) — a finding that was also 
confirmed with this study. To date, however, 
detailed information on how this species oc- 
cupies the epipelagic environment has been 
lacking. This study revealed that oceanic 
whitetip sharks had complex vertical move- 
ment patterns, pronounced diel changes 
and behavioral shifts, and that environ- 
mental factors influenced vertical activity. 
Several statistical approaches were com- 
bined to identify these vertical movement 
patterns. This research represents the first 
description of satellite archival data from a 
tag deployed on an oceanic whitetip shark 
in the Indian Ocean. 
Diel patterns 
The occurrence of diel behavior has not been 
reported from previous research on vertical 
movements of oceanic whitetip sharks (Howey-Jordan et 
al., 2013). Conversely, diel patterns were observed for 
all 6 individuals analyzed in our study. Despite some 
variability, diel patterns occurred at least once during 
the monitoring period of each tagged shark. Diel vertical 
movements are very common for fish species with verti- 
cal ranges that exceed that of oceanic whitetip sharks, 
such as the blue shark (Prionace glauca), bigeye thresh- 
er ( Alopias superciliosus ), swordfish, and bigeye tuna 
( Thunnus obesus) (Musyl et al., 2003, 2011; Abecassis 
et al., 2012; Lam et al., 2014; Coelho et al., 2015). These 
species usually occupy deep waters during the day and 
remain in the mixed layer during the night. They are 
believed to follow the vertical migration of mesopelagic 
prey species within the deep scattering layer (Dagorn 
et al., 2000; Bernal et al., 2009). 
Several epipelagic predators have also been reported 
to display diel patterns in their vertical behavior. For 
the silky shark (C. falciformis ) and dolphinfish ( Cory - 
phaena hippurus) (Merten and Appeldoorn, 2014; Fil- 
malter et al., 2015), for example, vertical movements 
similar to the type-I behavior described previously for 
the oceanic whitetip shark have been observed. Al- 
though opposite to the general diel migration pattern 
described for pelagic species with wide vertical ranges, 
type-I behavior could also be linked to feeding on me- 
sopelagic prey from the deep scattering layer. Filmal- 
ter et al. (2015) reported that silky sharks during the 
night were more vertically active and were observed 
at depths similar to the nocturnal depth range of pe- 
lagic species known to follow the migration of the deep 
scattering layer. The authors hypothesized that these 
increased vertical oscillations were associated with 
feeding activity. 
