Content uploaded by Ryan R Reisinger
Author content
All content in this area was uploaded by Ryan R Reisinger on Feb 05, 2014
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
Gauging the Threat: The First Population Estimate for
White Sharks in South Africa Using Photo Identification
and Automated Software
Alison V. Towner
1,2
, Michelle A. Wcisel
1,2
, Ryan R. Reisinger
3
, David Edwards
1
, Oliver J. D. Jewell
1,3
*
1 Dyer Island Conservation Trust, Great White House, Kleinbaai, Gansbaai, South Africa, 2 Department of Zoology, University of Cape Town, Rondebosch, South Africa,
3 Mammal Research Institute, Department of Zoology and Entomology, University of Pretoria, Hatfield, South Africa
Abstract
South Africa is reputed to host the world’s largest remaining population of white sharks, yet no studies have accurately
determined a population estimate based on mark-recapture of live individuals. We used dorsal fin photographs (fin IDs) to
identify white sharks in Gansbaai, South Africa, from January 2007 – December 2011. We used the computer programme
DARWIN to catalogue and match fin IDs of individuals; this is the first study to successfully use the software for white shark
identification. The programme performed well despite a number of individual fins showing drastic changes in dorsal fin
shape over time. Of 1682 fin IDs used, 532 unique individuals were identified. We estimated population size using the open-
population POPAN parameterisation in Program MARK, which estimated the superpopulation size at 908 (95% confidence
interval 808–1008). This estimated population size is considerably larger than those described at other aggregation areas of
the species and is comparable to a previous South African population estimate conducted 16 years prior. Our assessment
suggests the species has not made a marked recovery since being nationally protected in 1991. As such, additional
international protection may prove vital for the long-term conservation of this threatened species.
Citation: Towner AV, Wcisel MA, Reisinger RR, Edwards D, Jewell OJD (2013) Gauging the Threat: The First Population Estimate for White Sharks in South Africa
Using Photo Identification and Automated Software. PLoS ONE 8(6): e66035. doi:10.1371/journal.pone.0066035
Editor: Danilo Russo, Universita
`
degli Studi di Napoli Federico II, Italy
Received February 16, 2013; Accepted May 1, 2013; Published June 12, 2013
Copyright: ß 2013 Towner et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by Marine Dynamics Shark Tours (www.sharkwatchsa.com) and the Dyer Island Conservation Trust (www.dict.org.za); these
organizations are supported by donations from tourists and VW South Africa (www.vw.co.za). Marine Dynamics provided the platform to collect data from its
shark cage diving trips and has also contributed to the study costs of AT, OJ and MW’s Masters research. These authors continue to gain financial support from
Marine Dynamics as they pursue future academic goals. The funders had no role in study design, data collection and analysis, decision to publish, or preparation
of the manuscript. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: oliverjewell@gmail.com
Introduction
White sharks Carcharodon carcharias are widely distributed apex
predators which are known to undertake extensive oceanic
migrations [1–6]. They also exhibit semi-annual site fidelity to
predictable coastal locations often associated with pinniped
colonies, effectively pooling individuals to locations, or aggregation
areas, accessible to researchers [7–9]. These aggregation areas
have provided a unique opportunity to estimate white shark
populations utilising mark-recapture methods [9,10].
Photo Identification (photo ID) has been developed as a non-
invasive method of mark and recapture in which distinctive
features of an individual can be used to recognise it against the rest
of the population during different samples, over extensive time
periods. This method is particularly appropriate when examining
vulnerable species or populations, from invertebrates [11,12] to
tigers [13], marine mammals [14–16] and sharks [17–19]. The
first dorsal fin of white sharks is often characterised by distinctive
shapes, notches, scaring and pigmentation patterns, which can be
used to recognise individuals over many years [7,19]. From such
photo ID data, mark recapture methods can be applied to estimate
population sizes, given that the model’s basic assumptions are met
adequately [14,20].
Previous white shark population estimates suggest that white
shark numbers are small relative to other apex predators [20].
Nasby-Lucas & Domeier [9] identified a minimum of 142
individual white sharks using photographic identification at
Guadalupe Island (GI) from 2001–2009. Chapple et al. [10] used
mark recapture with a closed population model and estimated the
regional population of white sharks in California to be 219
animals. However Sosa-Nishizaki et al. [20] and Nasby-Lucas &
Domeier [9] contest this estimate on the basis that the use of a
closed model was inappropriate, there was no account for inshore
recruitment [21,22], a lack of sampling from An˜o Nuevo Island (a
major aggregation site for Californian white sharks) [23] and an
insufficient sampling period for the methods to reflect appropriate
trends. Cliff et al. [24] used mark and recapture of white sharks
tagged with spaghetti tags that were captured and killed in the
KwaZulu-Natal (KZN) Sharks Board nets to estimate a population
of 1279 (CV 24%) between Struisbaai, Western Cape and
Richards Bay, KZN. The weakness of this method, however,
was that the sharks used for the estimate were killed and were then
no longer part of the population.
Few studies have attempted to estimate the South African
population using mark and recapture techniques of living
individuals [25]. The study site of Gansbaai was selected to make
a first estimate of the current population of white sharks in the
PLOS ONE | www.plosone.org 1 June 2013 | Volume 8 | Issue 6 | e66035
region. Gansbaai is a world-recognised white shark aggregation
site and is the only location in the world where cage diving trips
operate daily, weather permitting [26]. Thus, this aggregation site
provided an ideal area to collect dorsal fin images of a range of age
classes and both sexes without seasonal paucities. Additionally,
several individual white sharks from this region have been
identified at other aggregation sites, with connectivity documented
between False Bay, Mossel Bay, KwaZulu Natal, and Western
Australia [1,19,27]. This suggests that Gansbaai is a major
aggregation area for white sharks in the Southern African region
allowing for an accurate population estimate from a single site.
Methods
We encountered white sharks at two aggregation areas in
Gansbaai (Geyser Rock and Joubertsdam). Trips were permitted
by the Department of Environmental Affairs, Oceans & Coasts
(formerly operating under Marine and Coastal Management); data
collection was un-invasive and required no further ethical
clearance. We attracted sharks to one of two cage diving vessels
run by Marine Dynamics shark tours with a fish bait, seal decoy,
and a scent trail created by a mixture of fish products and sea
water. Tours were weather dependent and biased towards areas of
high white shark abundance. We obtained images of white shark
dorsal fins during 1647 trips coinciding with ca. 4120 hours of
sampling effort (,2.5 hours average) from January 2007 –
December 2011.
We imported dorsal fin images into Picasa, a photograph editing
programme (picasa.google.com). Images were organised by date,
cropped and assigned a 1–6 quality ranking defined in Gowans &
Whitehead [28]. A ranking of 6 is considered to be a fin ID of the
highest quality, i.e. the fin is entirely clear of the water, square to
the camera, with good focus, lighting and adequate zoom. Q5 is
considered high quality with only one of the previous requirements
in Q6 lacking. The Q ranking decreases as the quality of photo ID
decreases. We found that only photographs Q4 and higher
provided sufficient information to recognise individuals and allow
‘‘recapture’’ between sightings. Such photographs (n = 1683) were
then imported into DARWIN dorsal fin ID software [29]. Each
dorsal fin was traced and assigned a fixed spacing of points along
Figure 1. Sighting frequency distribution of photographically identified white sharks in the Gansbaai region (2007–2011).
doi:10.1371/journal.pone.0066035.g001
Figure 2. Discovery curve for photographically identified white sharks in the Gansbaai region (2007–2011).
doi:10.1371/journal.pone.0066035.g002
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 2 June 2013 | Volume 8 | Issue 6 | e66035
the leading and trailing edges. This trace was compared to the
entire catalogue and ranked by probability of a match. Final
matching was confirmed by eye as pigmentation, scarring, freckles
or fin changes over time can be accounted for. Fins that could not
be matched to a fin already existing in the catalogue were assigned
a unique ID code corresponding with the first date of sighting and
fin order of that trip (e.g., 20110601-2 for the second shark on 1
June 2011) and then added. Inputted fins that matched a catalogue
fin were assigned the ID code of that matched individual and also
added. Data from DARWIN was exported to MS Excel where
individual abundance and occasionality could be assessed.
We performed mark-recapture analyses of the sighting histories
of recognisable individuals using Program MARK [30], which
uses Maximum Likelihood models to estimate population param-
eters [31]. We pooled photographic sightings into quarterly
sampling intervals after comparing the results for various sampling
intervals. Population closure was not a reasonable assumption and
therefore we used the open-population POPAN parameterisation
[32,33] to estimate population parameters. In this parameterisa-
tion, N represents the size of a superpopulation; which can be
thought of as either the total number of individuals available for
observation at any time during the study or as the total number of
animals ever in the sampled area between the first and last
occasion of the study [34]. The parameter W denotes apparent
survival rate, p is the probability of observation and b represents
the probability that an animal from the superpopulation enters the
sub-population (sub-population referring to the animals occurring
in the study area). In model notation, the subscripts t and.
represent time-dependent and constant parameters, respectively,
[35] and the initial analysis is based on the fully time-dependent/
Cormack-Jolly-Seber (CJS) model {W
t
p
t
b
t
}. The first step in the
analysis involves Goodness-of-Fit (GOF) tests for the CJS model
using Program RELEASE [36] to validate model assumptions.
Models were constructed for combinations of time-dependence
and consistency for each parameter and the most appropriate
model was selected using the small sample corrected Akaike
Information Criterion (AIC
c
) [37]. Based on the GOF results of
TEST 2+ TEST 3 in RELEASE a post-hoc variance inflation
factor (c
ˆ
) may be estimated to adjust for extra-binomial variation in
the data, resulting in a quasi-Akaike Information Criterion
(QAIC
c
).
Results
We identified 532 unique individuals which were included in
the population size analyses. Figure 1 shows the sighting frequency
distribution of these animals and figure 2 shows the discovery
curve – or cumulative number of identified individuals – as the
study progressed. Of the eight models tested, two did not
converge. Both of these contained b
.
(constant probability of
entry); the two other models containing b
.
did converge, but had
very large quasi-deviances, indicating that constant probability of
entry was not a reasonable assumption. For the remaining four
models, model choice criteria as well as abundance estimates and
parameter estimates are shown in Table 1. Based on the result of
TEST 2+ TEST 3 in Program RELEASE (Table 2), a variance
inflation factor of c
ˆ
= 1.36 was estimated and applied, indicating
only slight over dispersion in the data [31]. According to the
QAIC
c
scores, model {w
.
p
t
b
t
} (constant survival, time-varying
probability of capture and probability of entry) was the most
parsimonious. This model estimated the superpopulation size at
908 individuals (95% confidence interval = 808–1008). No models
had a DQAIC
c
,2 units, which would have indicated that they
were also likely [37]. Some violation of underlying open-
population mark-recapture assumptions was evident (Table 2).
Significance of TEST 3 (p = 0.006) and one of its components
(3.SR; p = 0.0001) indicate unequal survival probabilities among
photographically captured animals.
Discussion
Unlike Chapple et al. [10], we found the computer program
DARWIN suitable for matching and cataloguing white shark
dorsal fins within a large dataset. While DARWIN had infrequent
considerable errors in ranking fins, we considered this flaw minor
when compared to the human error of matching by eye alone.
Table 1. Model choice criteria and abundance estimates (N) for four models tested in a mark-recapture analysis of individual
sighting histories of white sharks in the Gansbaai region (2007–2011), using the open-population POPAN parameterisation in
Program MARK.
Model QAIC
c
DQAIC
c
Model
Likelihood Parameters Quasi Deviance N
95% Confidence Interval
CV
Lower Upper
w
.
P
t
b
t
1 719.88 – 1 35 0 908 808 1 008 0.056
w
t
p
t
b
t
1 746.38 26.50 0 53 0 823 717 929 0.066
w
.
p
.
b
t
1 791.19 71.31 0 14 0 950 853 1 048 0.052
w
t
p
.
b
t
1 802.91 83.03 0 33 0 956 851 1 062 0.056
See text for criteria.
doi:10.1371/journal.pone.0066035.t001
Table 2. Program RELEASE goodness-of-fit results for the
fully time-dependent/Cormack-Jolly-Seber model tested in a
mark-recapture analysis of individual sighting histories of
white sharks in the Gansbaai region (2007–2011), using the
open-population POPAN parameterisation in Program MARK.
Test X
2
Df P C
ˆ
2 51.86 47 0.2900 –
3 56.75 33 0.0062 –
3.SR 48.68 18 0.0001 –
3.SM 8.07 15 0.9209 –
2+3 108.61 80 0.0184 1.36
Shaded p-values are significant at a = 0.05.
doi:10.1371/journal.pone.0066035.t002
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 3 June 2013 | Volume 8 | Issue 6 | e66035
Figure 3. Study animal ‘Darwin’, photographed numerous times from 2007–2011. This fin clearly demonstrates how small additions/changes
to notches occurring on the back of the fin can distinguish it from early photos, increasing the probability of later fin IDs being identified as multiple sharks.
doi:10.1371/journal.pone.0066035.g003
Figure 4. Nine photographs were obtained for the study shark ‘Zebra’ between 30/08/2007 and 28/06/2011. During this time, the fin
sustained significant damage to both its front and back areas, altering its profile. This affects the number of notches down the back of the fin.
doi:10.1371/journal.pone.0066035.g004
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 4 June 2013 | Volume 8 | Issue 6 | e66035
Confirming this, we found that many dorsal fin trailing edges
changed dramatically from one sighting to another, an aspect we
were only able to detect as DARWIN accounts for more than just
the notches in the trailing edge of the fin. In most cases, the
leading edge and top quarter of the fin exhibited very little change
(Fig. 3), whereas the bottom L of the fin can be unrecognisable
between sightings (Fig. S1). We also observed changes to entire fin
shapes (Fig. 4). We were able to track these changes over time due
to consistent sampling effort, allowing us to consider fresh damage
or changes to fins when cataloguing. Without consistent effort, it is
possible that individual fins which changed would not be
successfully matched to the same individual shark, thus fin IDs
from the same shark over time can be counted as multiple
individuals resulting in an over-estimation of animals. In addition,
pigmentation patterns, or ‘rosies’ [26,38], were found to change
over time in most cases (Fig. 5) which is similar to the pigmentation
changes that have been described to occur around the lower
caudal areas of white sharks in south Australia [38]. The amount
of fin degradation in the lower quarters of the fin or changes in
pigmentation patterns do not seem to relate to the size/class of the
individual upon first sighting. However, presence of copepod
parasites along the trailing edge of the fin did seem to lead to initial
notch formation. These changes in fin morphology highlight that
previously proposed methods based on counting fin notches (i.e.
‘The Rutzen Method’ in O’Connel et al. [39]; Andriotti et. al. [40])
are unreliable for long-term mark-recapture analysis.
A common bias in many mark-recapture studies is capture
heterogeneity [41]. In our study we attracted sharks by bait, thus
some individuals may have become ‘trap happy’ or ‘trap shy’ over
time [41]. This may lead to bias on estimates, but the effects of
baiting on individual sharks remains undetermined [27,42]. To
address this, future work should focus on the effects of shyness or
boldness in individual white sharks and assess whether they are
more or less likely to appear close to a baiting vessel over time as
well as incorporating such heterogeneity in behaviour in popula-
tion size estimation.
Our population estimate for white sharks in Gansbaai is
considerably higher than those obtained for other aggregation
areas [9,10,20], supporting claims that South Africa has the largest
remaining population of coastal white sharks [25]. Our estimate is
comparable to that given by Cliff et al. [24]. This is not surprising,
as most white sharks that utilise Gansbaai aggregations also move
into KZN shark netted areas [43]. Unfortunately, dorsal fin photos
of white sharks killed in KZN shark nets were not collected during
this study period, therefore we cannot compare the living fin-ID
population of Gansbaai to the culled population in KZN [24,44].
There are 11 years between the end of Cliff et al.’s [24] data
collection and the beginning of sampling in Gansbaai (this study).
This suggests white shark numbers have not shown marked
recovery from; 1) the deployment of shark nets and drum lines
along the KZN coastline in 1952, which are still in place to date
[44]; 2) the heavy fishing pressures white sharks experienced in the
1970’s and 80’s [8]; and 3) a lack of protection in neighbouring
Mozambique [43]. Despite the species being protected since 1991
[8], such a low estimate and lack of recovery rate suggests the
Southern African white shark is not receiving adequate protection
for population growth. These results highlight the need for
Figure 5. Study shark ‘Vindication’, one of the few sharks to be photographed at least once in every year from 2007–2011.
Vindication features numerous, subtle changes in harsh contrast to the minimal, significant scars on the other examples above. On Vindication’s
dorsal fin, white pigmentation ‘roise’ changes occurred slowley over time.
doi:10.1371/journal.pone.0066035.g005
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 5 June 2013 | Volume 8 | Issue 6 | e66035
effective protective measures within the entire home range of the
Southern African white shark.
Supporting Information
Figure S1 Study shark ‘Demon’, demonstrating signficant
change in the lower three quarters of the trailing edge. Despite
having a large injury to the trailing ege of the dorsal fin, the fin
identification can still be matched by using the shape of the leading
edge and the top quarter of the fin.
(TIF)
Acknowledgments
Many thanks to all staff, crew and volunteers at Marine Dynamics who
made this work possible. Special thanks to Wilfred Chivell whose passion,
drive, and threats inspired and motivated this work. Thanks also to Mike
Gibbs and others at the Dyer Island Conservation Trust and VW South
Africa for their support and sponsorship. Thanks to all that tirelessly helped
with cataloguing, cropping, ranking and inputting fins, namely: Nick Jones,
Matt Nicholson, Ellie McLennan, Tami Kaschke, Blair Ranford and Tess
Mahoney among others.
Author Contributions
Wrote the paper: OJ MW AT RR DE. Initiated the project: AT MW OJ.
Collected fin ID photographs: AT OJ MW. Determined methodology for
processing fin IDs: MW. Cataloged and cropped photos in Picasa: MW DE
OJ AT. Processed fin IDs in DARWIN: DE MW. Ran MARK analysis:
RR. Produced figures: DE RR. Wrote the paper: OJ MW AT RR DE.
References
1. Bonfil R, Mey¨er M, Schol l MC, Johnson R, O’Brien S, et al. (2005)
Transoceanic migration, spatial dynamics, and population linkages of white
sharks. Science 310: 100–103.
2. Bruce BD (2008) The biology and ecology of the White Shark, Carcharodon
carcharias. In: Sharks of the Open Ocean: Biology, Fisheries & Conservation (eds.
MD Camhi, EK Pikitch, EA Babcock), 69–81 Blackwell Publishing, Oxford,
United Kingdom.
3. Domeier ML, Nasby-Lucas N (2008) Migration patterns of white sharks
Carcharodon carcharias tagged at Guadalupe Island, Mexico, and identification of
an eastern Pacific shared offshore foraging area. Mar Ecol Prog Ser 370: 221–
237.
4. Duffy CAJ, Francis MP, Manning MJ, Bonfil R (2012) Regional population
connectivity, oceanic habitat, and return migration revealed by satellite tagging
of white sharks, Carcharodon carcharias, at New Zealand aggregation sites. In:
Global Perspectives on the Biology and Life History of the White Shark (ed. ML
Domeier), 147–158. CRC Press, Boca Raton, Florida.
5. Domeier ML (2012) A new life history hypothesis for white sharks in the North-
eastern Pacific In: Global Perspectives on the Biology and Life History of the
White Shark (ed. ML Domeier), 199–223. CRC Press, Boca Raton, Florida.
6. Weng KC, Boustany AM, Pyle P, Anderson SD, Brown A, et al. (2007)
Migration and habitat of white sharks (Carcharodon carcharias) in the eastern Pacific
Ocean. Mar Biol 152(4): 877–894.
7. Anderson SD, Chapple TK, Jorgensen SJ, Klimley AP, Block BA (2011) Long-
term identification and site fidelity of white sharks, Carcharodon carcharias, off
California using dorsal fins. Mar Biol 158: 1233–1237.
8. Compagno LJV (1991) Government protection for the great white sharks
(Carcharodon carcharias) in South Africa. S Afri J Sci 87: 2845–2846.
9. Nasby-Lucas N, Domeier ML (2012) Use of photo identification to describe a
white shark aggregation at Guadalupe Island, Mexico. In: Global Perspectives
on the Biology and Life History of the White Shark (ed. ML Domeier), 381–391.
CRC Press, Boca Raton, Florida.
10. Chapple TC, Jorgensen SJ, Anderson SD, Kanive PE, Klimley AP, et al. (2011)
A first estima te of white shark, Carcharodon carcharias, abundance off Central
California. Biol Lett 7: 581–583.
11. Caci G, Biscaccianti AB, Cistrone L, Bosso L, Garonna AP, et al. (2013)
Spotting the right spot: computer-aided individual identification of the
threatened cerambycid beetle Rosalia alpina. Journal of Insect Conservation.
doi:10.1007/s10841-013-9561-0.
12. Huffard CL, Caldwell RL, DeLoach N, Gentry DW, Humann P, et al. (2008)
Individually unique body color patterns in Octopus (Wunderpus photogenicus ) allow
for photoide ntificatio n. PLoS ONE 3(11): e3732. doi:10 .1371/journa l.
pone.0003732.
13. Hiby L, Lovell P, Patil N, Samba Kumar N, Gopalaswary AM, et al. (2009) A
Tiger cannot change its stripes: using three-dimensional model to match images
of living tigers and tiger skins. Biol lett 5: 383–386.
14. Reisinger RR, de Bruyn PJN, Bester MN (2011) Abundance estimate of killer
whales at subantarctic Marian Island. Aquat Biol 12: 177–185.
15. Langtimm CA, Beck CA, Edwards HH, Fick-Child KJ, Ackerman BB, et al.
(2004) Survival estimates of Florida manatees from photo-identification of
individuals. Mar Mamm Sci: 20: 438–463.
16. Reisinger RR, Karczmarski L (2010) Population size estimate of Indo-Pacific
bottlenose dolphins in the Algoa Bay region, South Africa. Mar Mam Sci 26(1):
86–97.
17. Bansemer CS, Bennett MB (2009) Reproductive periodicity, localised move-
ments and behavioural segregation of pregnant Carcharias taurus at Wolf Rock,
Southeast Queensland, Australia. Mar Ecol Prog Ser 374: 215–227.
18. Speed CW, Meekan MG, Bradshaw CJA (2007) Spot the match – wildlife
photoidentification using information theory. Front Zool 4: 2. doi: 10.1186/
1742-9994-4-2.
19. Jewell OJD, Wcisel MA, Gennari E, Towner AV, Bester MN, et al. (2011)
Effects of Smart Position Only (SPOT) tag deployment on white sharks
Carcharodon carcharias in South Africa. PLoS ONE 6(11): e27242.
20. Sosa-Nishizaki O, Morales-Bojorquez E, Nasby-Lucas N, Onate-Gonzalez EC,
Domeier ML (2012) Problems with photo identification as a method of
estimating abundance of white sharks, Carcharodon carcharias an example from
Guadalupe Island, Mexico. In: Global Perspectives on the Biology and Life
History of the White Shark (ed. ML Domeier), 393–404. CRC Press, Boca
Raton, Florida.
21. Weng KC, O’Sullivan JB, Lowe CG, Winkler CE, Dewar H, et al. (2007)
Movements, behaviour and habitat preferences of juvenile white sharks
Carcharodon carcharias in the eastern Pacific. Mar Ecol Prog Ser 388: 211–224.
22. Lowe CG, Blausius ME, Jarvis ET, Mason TJ, Goodmanlowe GD, et al. (2012)
Historic fisheries interaction with white sharks in the Southern California Bight.
In: Global Perspectives on the Biology and Life History of the White Shark (ed.
ML Domeier), 169–185 CRC Press, Boca Raton, Florida.
23. Klimley AP, Le Boeuf BJ, Cantara KM, Richert JE, Davis SF et al. (2001) The
hunting strategy of white sharks (Carcharodon carcharias) near a seal colony. Mar
Biol 138: 617–636.
24. Cliff G, van der Elst RP, Govender A, Witthuhn TK, Bullen EM (1996) First
estimates of mortality and population size of white sharks on the South African
coast. In: Great White Sharks: The Biology of Carcharodon carcharias (eds. AP
Klimley, DG Ainley). Academic Press, San Diego, California, 393–400.
25. Dudley SFJ (2012) A review of research on the white shark, Carcharodon carcharias
(Linnaeus), in Southern Africa. In: Global Perspectives on the Biology and Life
History of the White Shark (ed. ML Domeier), 511–532. Boca Raton: CRC
Press, Florida.
26. Towner AV (2012) Great white sharks Carcharodon carcharias in Gansbaai, South
Africa: Environmental influences over time, 2007–2011. MSc Thesis, University
of Cape Town.
27. Johnson R, Kock A (2006) South Africa’s White Shark cage-diving industry - is
their cause for concern? In Nel DC & Peschak TP (eds) Finding a balance: White
shark conservation and recreational safety in the inshore waters of Cape Town,
South Africa; proceedings of a specialist workshop. WWF South Africa Report
Series - 2006/Marine/001.
28. Gowens S, Whitehead H (2001 ) Ph otographic identification of northern
bottlenose whales (Hyperoodon ampullatus): sources of heterogeneity from natural
marks. Mar. Mam. Sci. 17(1): 76–93.
29. Stanley R (1995) DARWIN: Identifyi ng Dolphins from Dorsal Fin Images.
Senior Thesis, Eckerd College.
30. White GC, Burnham KP (1999) Program MARK: survival estimation from
populations of marked animals. Bird St Supp 46: 120–138.
31. Cooch E, White G (eds) (2009) Program MARK: a gentle introduction, 8th edn.
Available: http://www.phidot.org/software/mark/docs/book.
32. Schwarz CJ, Arnason AN (1996) A general methodology for the analysis of
capture-recapture experiments in open populations. Biomet 52: 860–873.
33. Schwarz CJ, Arnason AN (2009) Jolly-Seber models in MARK. In:
MARK: a gentle introduction (eds. E. Cooch & G. White) Program, 8th edn.
Available: http://www.phidot.org/software/mark/docs/book.
34. Nichols JD (2005) Modern open-population capture-recapture models. In:
Handbook of capture-recapture analysis (eds. SC Amstrup, TL, McDonald, BFJ
Manly), 88–123. Princeton University Press, Princeton.
35. Lebreton JD, Burnham KP, Clobert J, Anderson DR (1992) Modeling survival
and testing biological hypotheses using marked animals: A unified approach with
case studies. Ecol Monog 62: 67–118.
36. Burnham KP, Anderson DR, White GC, Brownie CR, Pollock KH (1987)
Design and analysis methods for fish survival experiments based on release-
recapture. Am Fish Soc Monog 5.
37. Burnham KP, Anderson DR (1998) Model selection and inference: a practical
information-theoretic approach. Springer-Verlag, New York.
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 6 June 2013 | Volume 8 | Issue 6 | e66035
38. Robbins R, Fox A (2013) Further evidence of pigmentation change in white
sharks, Carcharodon carcharias. Mar Freshw Res 63 (12): 1215–1217.
39. O’Connell C, Andreotti S, Rutzen M, Me?er M, He P (2012) The use of
permanent mag nets to reduce elasmobranch encounter with a simulated beach
net. 2. The great white shark (Carcharodon carcharias). J Oc Cost Manag. Available:
http://dx.doi.org/10.1016/j.ocecoaman.2012.11.006.
40. Andriotti S, Rutzen M, Mey¨er M, Oosthuizen H, Herbst B, et al. (2012) The
Rutzen Method. J Exp Mar Biol Ecol Cited in O’Connell et al. 2012 Available:
http://dx.doi.org/10.1016/j.ocecoaman.2012.11.006.
41. Pledger S, Pollock KH, Norris JL (2010) Open capture–recapture models with
heterogeneity: II. Jolly–Seber model. Biometrics 66(3): 883–890.
42. Laroche KA, Kock AA, Dill LD, Oosthuizen WH (2007) Effects of provisioning
ecotourism activity on the behaviour of white sharks Carcharodon carcharias.Mar
Ecol Prog Ser 338: 199–209.
43. Ocearch Shark Tracker. Available: sharks-ocearch.vertite.com. Accessed 2013
May 08.
44. Dudley SFJ, Simpfendorfer CA (2006) Population status of 14 shark specie s
caught in the protective gillnets off KwaZulu–Natal beaches, South Africa,
1978–2003. Mar Freshw Res 57(2): 225–240.
Population Estimate of South African White Sharks
PLOS ONE | www.plosone.org 7 June 2013 | Volume 8 | Issue 6 | e66035
