PROCESSING PHOTOGRAPHIC IDENTIFICATIONS OF WHALE SHARKS USING THE INTERACTIVE INDIVIDUAL IDENTIFICATION SYSTEM (I3S). DRAFT DATA COLLECTION PROTOCOL, VERSION 1.1 (12/3/07)

Simon J. Pierce Lead Scientist - Whale Sharks Manta Ray & Whale Shark Research Centre Tofo Beach, Mozambique Ph: +258 2935 6254 Email: [email protected] Website: http://mozmarinescience.googlepages.com/

INTRODUCTION Whale sharks (Rhincodon typus) are the largest fish in the world. The iconic status and placid nature of the species has made these sharks the subject of popular marine tourism industries. However, it has become increasingly evident in recent years that regional whale shark populations have suffered major declines. These declines are linked to directed fisheries targeting whale sharks for their fins and meat, and the continuing nature of this threat has led to whale sharks being listed as ‘Vulnerable’ internationally on the IUCN Red List of Threatened Species. Though broadly distributed through tropical and subtropical waters, whale sharks are rare or seasonal throughout most of their range. Whale sharks are a highly migratory species and individual sharks are likely to regularly travel through multiple political jurisdictions, potentially varying from fisheries target to protected species as they do so. Thus, to effectively conserve the species it is important to assess its population status over broad geographical scales. Researchers need to estimate the size of remaining populations and whether the global whale shark population is subdivided into different stocks, which would then necessitate a series of independent regional population estimates. Though data is difficult to gather as individual scientists and research teams, these aims can be achieved relatively quickly and inexpensively through collaborative studies. PHOTOGRAPHIC IDENTIFICATION Whale sharks are easily distinguishable from other species due to their large size and distinctive white-spotted dorsal colouration. These spot patterns are individually unique and appear to be consistent over time, enabling the long-term resighting of individuals and the application of standard sighting-resighting based population estimation methods (Meekan et al. 2006).

Photographic identification has several advantages over conventional tagging, particularly for a large threatened species like the whale shark. Photo-ID is non-invasive, reducing both the potential for detrimental impacts from the tags or a behavioural response to tagging, which may bias future resighting. Given that underwater digital cameras are now ubiquitous within the diving and marine tourism industry, it is easy to obtain ID shots from tourism operators or interested clients themselves as well as from within the scientific community. The tedious task of comparing these photographs is reduced by automating the process using the Interactive Individual Identification System (I3S) (available from http://www.reijns.com/i3s/). This free software package allows researchers to either process their own sight-resight data or to send ID photographs to a centralised location for analysis and reporting. The use of a common method between research teams allows single-location databases to be combined to look for matches across a wide geographical range. The aim of this document is to present a user-friendly proven method for analysing within-population data that also allows collaborative studies without modification. The use of I3S (and this manual) will require a basic familiarity with the program which can be gained by completing the tutorial exercises in the I3S user manual (available from http://www.reijns.com/i3s/download/I3S_download.html). DATA COLLECTION PROTOCOL

The flanks of the shark are the areas used for identification, based on other successful studies. Photographs should be taken from right angles to the shark. The important areas to include in the photograph are the upper and lower fifth gill slit and the inner trailing edge of the pectoral fin. These flank areas approximate a two-dimensional surface, contain large distinctive spots and include suitable reference points for comparison between images. Identification photographs can be taken using either a digital or film still camera. Lowerquality, but usable, still images can also be taken from underwater video footage.

Obviously, all images must be in a digital format before use with I3S. Select images of highest clarity for matching, but if hard-drive space for your I3S library is limited then consider your computer screen size and resolution and re-size your images appropriately. I re-size my I3S library images (taken with a 7.2 MB Sony digital camera) to 25% of their original size with no loss of clarity. Some pre-processing of images can often be useful before using I3S. Adobe Photoshop (or equivalent software) is a useful tool for maximising image clarity: try adjusting brightness and contrast, and/or using ‘auto-level’ or ‘levels’ on your photographs, to make spots vivid and obvious. I generally auto-level all my shots to reduce the bluish tinge common to underwater photographs taken without using a strobe. I do not use the camera’s internal flash when taking photographs to reduce backscatter, the reflection of light from suspended particles. Ideally, for each shark sighted the following data would be obtained: 1. 2. 3. 4. 5.

Both right- and left-side ID shots (see below) Location of sighting (GPS coordinates) Photographs of any scars or distinctive features The sex of the shark A measured or estimated length

Ultimately though, even just a single ID shot and location remains useful. The presence of scars or distinctive features is a useful means of validating identifications, while also being interesting data in its own right. Similarly, a location is necessary to clarify the geographical range of individual or populations of sharks. The sex of sharks can be distinguished externally by the presence or absence of claspers, the male reproductive organs, on the pelvic fins. Claspers are relatively small in immature male whale sharks, and small males may be difficult to distinguish from females. For non-specialists, it may be necessary to photograph the pelvic fins to confirm sex. Length can either be estimated by comparison to objects of known size, such as swimmers, or measured directly using a tape measure or string. USING THE I3S SOFTWARE PACKAGE The I3S website and user manual has detailed documentation on the program’s development and use. This section assumes you have been through the tutorial examples in the user manual and are now a confident I3S user. I3S was originally developed for ragged-tooth shark (Carcharias taurus) research. Whale shark use is similar aside from the use of different reference points.

The photo above is a left-side ID shot in I3S. The reference points for whale sharks are the top of the fifth gill slit (‘dorsal 1’), the trailing edge of the pectoral fin (‘dorsal 2’) and the bottom of the fifth gill slit (‘pelvic’). Dots are placed on all the natural spots to the rear of the fifth gill slit and above the pectoral until you run out of either spots or shark. Place as many spots as possible. The photo below is a right-hand ID shot – it’s effectively just a mirror image of the left-hand ID. The spot pattern differs on each side, so it’s useful to have both in the database.

Clear photographs taken from right angles to the shark should have few problems in matching images. See Speed et al. (2007) for technical details of image-matching whale shark photographs with I3S. I generally do visual checks on the first three to five topranked photo matches. I use only the best image from each flank of individual sharks to minimise the database size and maximise the matching-speed of the program, but in ragged-tooth sharks it proved beneficial to have up to three images included from each shark (van Tienhoven et al. 2007), and if you’re not constrained by hard-drive space then I recommend you follow their recommendations. For convenience I use a separate folder for each whale shark within the ‘sharks’ folder in I3S, and replace images when resightings produce a higher-clarity photograph. FURTHER READING AND RECOMMENDED LINKS That’s it. Go forth and use I3S on every species you wish. If you have ideas for further directions for the software, please contact the developers, Jurgen den Hartog and Renate Reijens at [email protected]. If you are interested contributing or collaborating with the large-scale whale shark study now in progress, encompassing the Indian Ocean and elsewhere, please get in touch with me ([email protected]) for more information. Good luck, and have fun! The I3S website (www.reijns.com/i3s) has full information on downloading, installing and using the program. It has some links to relevant papers and information on current research. For information on whale shark research in Mozambique, the Manta Ray & Whale Shark Research Centre has a website at http://mozmarinescience.googlepages.com/ See the following references for further details on whale shark photo-identification and using I3S: Meekan MG, Bradshaw CJA, Press M, McLean C, Richards A, Quasnichka S, Taylor JG (2006) Population size and structure of whale sharks Rhincodon typus at Ningaloo Reef, Western Australia. Marine Ecology Progress Series 319:275-285. van Tienhoven AM, den Hartog JE, Reijns RA, Peddemors VM (2007) A computer-aided program for pattern-matching of natural marks on the spotted raggedtooth shark Carcharias taurus. Journal of Applied Ecology. Speed CW, Meekan MG, Bradshaw CJA (2007) Spot the match – wildlife photoidentification using information theory. Frontiers in Zoology 4:2. Available online from http://www.frontiersinzoology.com/content/4/1/2.