Presentation Details
| Using computer vision to understand the social drivers of migration across scales Benjamin Koger1, 2, Albert Kao3, Brayden Dearden4, Andrew Berdahl5. 1University of Wyoming, School of Computing, Laramie, WY, USA.2University of Wyoming, Department of Zoology and Physiology, Laramie, WY, USA.3University of Massachusetts Boston, Department of Biology, Boston, MA, USA.4University of Wyoming, Department of Mathmatics, Laramie, WY, USA.5University of Washington, School of Aquatic and Fishery Sciences, Seatle, WA, USA |
Abstract
Social interactions have long been theorized to shape migratory decision-making, yet collecting relevant data has historically been a major challenge. Studying social migration requires recording numerous interacting individuals as they move through complex environments. Advancements in AI-driven computer vision and imaging technology, such as drones and ground-mounted cameras, can offer limited but complementary windows into these dynamics from individuals to groups to populations. We used a combination of these methods to study sockeye salmon migration near Pedro Bay, Alaska. Over 1.5 weeks of the migration, we paired daily drone-derived nesting-pond surveys, with continuous tree-mounted video recordings of the entrance of each pond, with drone derived focal follows of individual groups as they moved through the system. Across each datastream we show strong evidence for socially driven movement decisions with the effects apparent at each measured scale. The opportunities and challenges identified in this system are informative for better quantifying the dynamics of social migration and understanding the abilities of novel imaging methods to study animal behavior in the wild.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
Content Locked. Log into a registered attendee account to access this presentation.