Atlantic Guitarfish Photo ID

Hi everyone,

We’re starting an Atlantic Guitarfish photo-ID study in South Florida, USA. They have complex spot patterns, here are some examples:



For photo ID so far, we have tried using a program “13S” and have focused on mapping the spots from the tip of their rostrum to their spiracles. As Guitarfish often will bury themselves in the sand, we often run into the issue that sand gets stuck on their rostrum and obscures part of the spot pattern. As the area between their spiracles is more consistently visible and free of sand, we are curious about focusing on that smaller area between the spiracles as a region for ID? Additionally, the program we have been using is very labor intensive, with the user having to manually select up to 150 spots to compare between individuals.

Could anyone advise if any of the algorithms that have been integrated into Wildbooks would potentially be suitable for distinguishing individual guitarfish with minimal user input? We have logged over 500 encounters since starting the project!

Any suggestions or pointers would be greatly appreciated!

Thank you!

Olivia

What in interesting species! In the data you have so far, do you have sets of verified individuals?

Yes! We are also working with a few aquariums to get photos of captured guitarfish, as verified individuals and to track whether the spot patterns remain stable over time.

Awesome. I’m going to send you an email so we can get the ball rolling on next steps.