~The Marine Conservation Society, Seychelles~  

 
   

 

 

 

Seychelles whale shark monitoring newsletter 

  Jun 2007  Vol 5, No. 2
   
 

 

NOTICE BOARD

Whale Shark Sightings 
 Jan - Jun 2007


In-water Sightings
1 reported Sighting
1 in-water encounters

0 shark identified
0 shark sexed

Aerial Survey

Aerial surveys will restart as from August 2007

 

 

 

 

 

 

 

Please send us any comments or news for inclusion by clicking

www.mcss.

 Photo Identification Reveals Some Surprises!

The introduction of the Interactive Individual Identification System, IRIS as it was known, photo-identification software to the Marine Conservations Society Seychelles Whale Shark Programme during the last season has already brought to light one or two surprises!

Prior to 2005, images were scanned from photographs or slides and as such were limited in number. Since 2005, with the advent of affordable digital photography, the photo ID project was viable and almost every whale shark encountered by the programme was photographed.

The IRIS visual comparison of S208: original sighting in 2001 compared to an image taken on 23rd October 2006 showing the matched spot patterns on each image.  

Project leader, whale shark researcher David Rowat, has now compiled all the identification photos of individual whale shark encounters from 2001 through to the end of last season, i.e. 2006. A total of 9,168 images!

These were then sorted to select images of the left and right sides, showing the area just behind the last gill slit, which is regarded as having the most individual spot patterning.

From the selected images, digital fingerprints were created for each image using IRIS and stored in a database for comparison. 632 files were in the left-side database and 415 in the right-side series. One by one, each fingerprint was compared to all the others of the same side and matches were combined within each year group. For each comparison the IRIS programme compares the fingerprints and give a list of the top 50 best ranked matches. The researcher then uses the ranking score (the lower the better), a visual comparison and a tool called the spot cloud, which shows the deviation of the matched spots from each other, to confirm or decline the best matched images.

After these comparisons were completed, the final count left a total of 300 images in the left-side database and 215 in the right.

Final tallies for each of the years in ongoing but in the 2006 season 195 individual animals have been identified with over 220 matches being made within the season. The surprises have been in the number of animals that were recorded in 2006 that were also matched to previous years.

The The IRIS ‘spot cloud’ pattern showing the difference in the position of marked spots from the two images of whale shark S208, with 19 pairs of spots being matched giving a score of 0.05, which indicates an almost perfect match.

Of the 195 individuals 36 were also recorded in 2005, five from 2004, two from 2002 and one from 2001. The 2001 shark was in fact a tagged shark, S208, recorded on the 28th August 2001 and was matched to three sightings in 2006 on the 2nd, 23rd and 25th of October 2006. Similarly, in 2005 four sharks from 2001 were identified, two had been tagged previously, S201 and S210, while the other two had not been tagged but were photographed on the 27th of October 2001

While photo identification is extremely time consuming, it does have the potential to provide some interesting insights in the study of whale shark populations.

To MCSS Home Page

To Newsletter Home Page