382 
Fishery Bulletin 115(3) 
Table t 
Details from the electronic tagging of oceanic whitetip sharks ( Carcharhinus longimanus ) in the Atlantic and Indian oceans 
between 2011 and 2012: identification (ID) code, total length (TL), sex, location, model of pop-up satellite archival tag used, 
the period that tags were set to record data (programmed). Tagging: date when a fish was tagged, and location of tagging 
(latitude flat] and longitude [long]). Pop-up: date when a tag popped up (was released), location of released tag, and the 
number of days that the tag had remained on the fish. 
ID 
TL (cm) 
Sex 
Ocean 
Tag 
Programmed 
Date 
Tagging 
Lat. 
Long. 
Date 
Pop-up 
Lat 
Long 
Duration 
AOCS3 
167 
M 
Atlantic 
PAT-MklO** 180 d 
16/01/2011 
-0.139 
-34.218 
10/07/2011 
-3.802 
-32.466 
178 d 
AOCS4 
197* 
F 
Atlantic 
MiniPAT 
140 d 
06/12/2011 
-3.589 
-34.918 
25/04/2012 
-18.754 
-35.771 
141 d 
AOCS5 
180* 
F 
Atlantic 
MiniPAT 
140 d 
01/03/2012 
-0.501 
-37.354 
20/07/2012 
3.215 
-41.015 
141d*** 
AOCS6 
134 
F 
Atlantic 
MiniPAT 
100 d 
02/03/2012 
-0.736 
-37.534 
11/06/2012 
-0.598 
-36.235 
101 d 
AOCS7 
161 
F 
Atlantic 
MiniPAT 
100 d 
02/03/2012 
-0.435 
-37.629 
14/06/2012 
1.306 
-35.345 
104 d 
IOCS1 
183* 
F 
Indian 
MiniPAT 
100 d 
15/04/2011 
-13.119 
44.967 
24/07/2011 
-2.522 
53.554 
100 d 
*Mature individuals (size at first maturity: 180 cm TL). 
**Recovered tag. 
***Tag stopped recording data after 104 d of deployment. 
on the daily geolocation estimates from the tags and 
the NOAA daytime estimation algorithm (NOAA Solar 
Calculator, website). Day was defined as the period be- 
tween sunrise and sunset, and night was defined as 
the period between astronomical dusk and astronomi- 
cal dawn. Dusk was the hours between sunset and 
astronomical dusk, and dawn was the hours between 
astronomical dawn and sunrise. Because sharks made 
extensive horizontal movements during their monitor- 
ing periods, local sunrise and sunset times varied with 
time for all individuals. The variation, however, was 
not greater than 50 min. To facilitate graphic repre- 
sentations of aggregated data, day and night and dawn 
and dusk were depicted by their respective minimum 
and maximum estimated times. Daytime and nighttime 
depths were compared with the nonparametric Wil- 
coxon test at a 95% confidence level. For this analysis, 
depths corresponding to crepuscular hours (dawn and 
dusk) were excluded. Mean depths were grouped into 
1-h intervals to test for uniformity over the 24-h cycle. 
The uniformity was tested by using circular statistics 
(Rao’s spacing test), also at a 95% confidence level. 
A spectral analysis was carried out with the depth 
time-series data from the recovered tag of shark 
AOCS3. This analysis was not feasible for the other 
tags because of gaps caused by the data transmis- 
sion. The aim was to identify a potential periodicity in 
the vertical behavior of this shark and infer possible 
temporal patterns. A fast Fourier transform algorithm 
was used in the stats package in R, vers. 3.1.2 (R Core 
Team, 2014). The function calculates a smoothed peri- 
odogram by using Daniell windows, which are modi- 
fied moving-average filters. The raw periodogram is a 
widely fluctuating estimate of the spectrum with high 
variance, and this smoothing method provides a stable 
estimate (Bloomfield, 2004). The spectral analysis is 
particularly well suited for long-term and high-resolu- 
tion time series, including those from archival tagging 
studies (Shepard et al., 2006). 
The depth time-series data were also assessed vi- [ 
sually to examine any possible vertical patterns that j 
could have been masked when the data were grouped. 
This analysis was conducted with the help of a visual- \ 
ization tool, and the window of this software was ad- | 
justed on our computer screen to fit 2 d of data at a * 
time. In this analysis, the times of sunrise and sunset 
did not need to be estimated. Instead, the readings of f 
ambient light from the tags could be simultaneously I 
displayed with the depth readings. The light data were I 
transmitted in the form of 2 daily light curves, repre- 
senting sunrise and sunset. For the recovered tag, the 
complete time series of light readings was available. 
The simultaneous visualization of light curves and the 
depth time series was created by using the graphing 1 
and analysis software program Igor Pro, vers. 6.22A 
(WaveMetrics Inc., Portland, OR). A suite of data anal- 1 
ysis programs (WC-DAP, Wildlife Computers Inc.) was 
used to export a file formatted for use with Igor Pro, 
which facilitated the visual analysis. This pairing of ; 
light and depth data also increased the precision need- 
ed for discerning diel patterns. 
| 
Vertical movements and the environment To reconstruct • 
the thermal signature of the water column occupied 
by the oceanic whitetip sharks during the periods in 
which they were monitored, the summary data for < 
temperature at depth were used. This data product 
provides the minimum and maximum temperatures at ‘ 
selected 8-m depth intervals at a user-defined resolu- 
tion (for this study, the resolution was every 24 h). The 
average temperatures of depth intervals were interpo- j 
lated linearly to produce continuous daily temperature 
