Behavioural Evolution in Cichlid Fishes – from Shared Repertoires to Explosive Divergence
Supervision: Dr. Alex Jordan (MPI-AB)
Project background
Animal behaviour is composed of multiple interacting actions and forms, built together to generate function. However, this function can be hidden from human observers and may only carry meaning for the intended recipient. In this project we aim to uncover the way behaviour has evolved in both form and function in the explosive adaptive radiation of Lake Tanganyikan cichlids. The Jordan Lab integrates field experiments, machine-learning-based video analysis, and comparative frameworks to uncover how such shared behavioural modules are redeployed and reordered across species to generate new functional forms.
Situated in the wild setting of Lake Tanganyika, this work builds on recent advances in unsupervised machine learning and computer vision that allow the decomposition of continuous behaviour into objective, reproducible components. By creating effective maps of behaviour, we can compare how different species utilise shared behavioural morphospaces and visualise how evolutionary and ecological pressures shape the organisation of behaviour across related species. We then draw from these behavioural spaces to create digital playbacks, deployed underwater with wild fish, that explore the meaning of behaviour for the animals themselves.
Project description
The PhD project will combine detailed field experiments and behavioural observations at Lake Tanganyika with quantitative analysis of behavioural sequences. The candidate will record and analyse social and territorial interactions across multiple species of lamprologine cichlids, applying automated tracking and unsupervised learning to identify shared behavioural building blocks. These data will be used to construct behavioural morphospaces, allowing evolutionary comparisons of how species assemble conserved actions into distinct functional repertoires.
Further, the project will employ digital playbacks and virtual stimuli underwater to test how animals perceive and respond to variations in behavioural form, closing the loop between quantitative description and biological meaning.
Training and environment
The student will join the Jordan Lab at the Max Planck Institute of Animal Behavior, an international group studying the evolution and function of animal behaviour in natural contexts. Training will include:
- Experimental design and data collection in the field at Lake Tanganyika
- Automated video tracking and unsupervised machine learning for behavioural analysis
- Comparative frameworks for studying behavioural evolution
- Digital playback design and underwater deployment
The student will also participate in the interdisciplinary training programme of the IMPRS-QBEE, gaining exposure to cutting-edge approaches in behaviour, ecology, and evolution.
Candidate profile
Applicants should have a strong background in behavioural ecology, evolutionary biology, or related fields, with an enthusiasm for both fieldwork and computational approaches. Prior experience with behavioural observation, programming (Python, R, or MATLAB), or image/video analysis is beneficial but not essential. Most importantly, candidates should be motivated to connect natural history, quantitative analysis, and conceptual theory to address how animal behaviour evolves.
Application
This position is fully funded for 3 years. Applications should be submitted via the IMPRS-QBEE online portal.
Start date: Flexible, Fall 2026
Application deadline: 31. October 2025
Further information: https://thejordanlab.com