Quantitative analysis of acoustic communication and collective behavior in animal groups
Direct supervisor: A. Strandburg-Peshkin & external collaborator Marie Roch
We are seeking a doctoral student with a quantitative background and enthusiasm for tackling biological questions to join our interdisciplinary and international team studying communication and collective behavior in animal groups. The researcher will be based at the Max Planck Institute of Animal Behavior and the University of Konstanz, located in Konstanz, Germany.
Research. The research will focus on analysis of movement and acoustic data to understand how animals use vocal signaling to coordinate collective behaviors, as part of the Communication and Coordination Across Scales Project (CCAS). CCAS is an interdisciplinary collaborative project that integrates behavioral field biology with complex systems, collective movement, bioacoustics, and machine learning.
Qualifications. We are seeking candidates with an interest in developing and employing computational approaches to the study of animal behavior and communication. This call is open to candidates from all scientific backgrounds who can articulate how their interests and training prepare them for this position. Programming skills and an enthusiasm for tackling challenging analytical problems are essential. Applicants should ideally have prior experience with the analysis of spatial, timeseries, acoustic, or network data, and/or experience with machine learning. Strong communication skills and ability to work as part of an interdisciplinary team are also essential.
Duties. The research will focus on analysis of simultaneous tracking data (movement, acoustic, and behavioral) recorded from entire social groups across three species. Candidates may work on one or more of the broadly defined topics below, or develop new directions in collaboration with other researchers on the project:
- Call detection and classification using machine learning: Develop acoustic recognition software to analyze acoustic data from tracking collars using supervised and/or unsupervised machine learning
- Behavioral state recognition using machine learning: Develop software to identify behavioral states at individual and group levels using multi-sensor tag data
- Individual decision-making: Develop and apply analytical methods to study how individuals in groups integrate spatial and acoustic information to make behavioral decisions about movement and vocal production
- Information flow through groups: Use information theoretic or other approaches to quantify how individuals influence one another and how information flows through groups
- Modeling collective behavior: Develop models linking behavioral interactions to collective outcomes across different systems
While extensive field work is not envisioned, if feasible researchers will have the opportunity to visit some of the field sites to gain insight into the study species.
Supervision and research community. Researchers will be jointly supervised by Dr. Ariana Strandburg-Peshkin (collective behavior) and external collaborator Dr. Marie Roch (bioacoustics / machine learning), and will also engage with our international team of collaborators which includes computational and behavioral researchers. The University of Konstanz and the Max Planck Institute for Animal Behavior together form a thriving research community representing a global hotspot for collective behavior and animal movement research. Researchers will be integrated into the International Max Planck Research School for Quantitative Behaviour, Ecology and Evolution (IMPRS-QBEE) also have the opportunity to join the Centre for the Advanced Study of Collective Behaviour, an interdisciplinary research community integrating biology, computer science, physics, psychology, economics, and other fields to tackle questions in collective behavior.
The Max Planck Society and the University of Konstanz strive to promote a diverse, equitable, and inclusive workplace. For details see http://www.mpg.de/equal_opportunities. We welcome applications from all backgrounds, and members of groups underrepresented in science are especially encouraged to apply.
Further project details. In the CCAS project we are leveraging innovations in tracking technology and computational modeling to determine how vocal communication influences collective behavior in animal groups. Specifically, we are recording movements and vocal signals simultaneously from all members of wild animal groups at a high resolution, and across varying degrees of spatial dispersion. Our focus is on three mammal species that solve a common set of coordination problems, but differ in spatial cohesiveness: meerkats (highly cohesive groups), coatis (moderately cohesive), and spotted hyenas (fission‑fusion). In each species, we aim to 1) fit at least one entire social group in the wild with tags that continuously record fine-scale movements and vocalizations, 2) combine supervised and unsupervised machine learning approaches to identify animal calls and movement states, 3) develop probabilistic modeling approaches to reveal how individuals integrate spatial and acoustic information, how information flows through groups, and how behavioral interactions give rise to collective outcomes, and 4) conduct targeted audio playback experiments to isolate causal relationships driving collective dynamics. Combining these approaches with long‑term data from existing field studies will allow us to shed light on both unifying features underlying coordination mechanisms across animal societies and differences imposed by distinct socio-ecological constraints.
Research Statement Instructions. Applicants should include a research statement that addresses the following points:
- Describe your scientific background and research interests, and explain how they relate to the project.
- Describe one or more specific research questions / directions you would be interested to tackle using the dataset(s) available in this project, and explain how you would go about addressing them.
Questions. For further information regarding the position, please feel free to contact Ariana Strandburg-Peshkin (astrandburg@ab.mpg.de).