The central complex is a conserved brain region at the core of insect navigation, integrating sensory information to guide behavior across species spanning over 400 million years of evolution. While connectomes of the fruit fly have provided unprecedented detail about its circuitry, understanding how this neural architecture varies across insects with diverse navigational capabilities requires comparative approaches. This thesis presents a multiresolution imaging, processing, and analysis pipeline designed to make comparative connectomics accessible at a reduced cost. By leveraging the predictable neuroarchitecture of the central complex, we are able to produce a global projectome with embedded local connectomes describing entire computational units.Applying this pipeline, we reconstructed head direction circuits across multiple hymenopteran species, revealing remarkable conservation in cell types and projection patterns despite over 300 million years of divergence from the fly. Circuit-level analysis, however, uncovered fundamental differences in the feedback loops underlying the ring attractor, demonstrating functionally convergent solutions to the same computational problem. Extending this analysis to downstream traveling direction neurons in the sweat bee, we identified both ancestral pathways homologous to the fly and novel circuits specific to bees, including a new cell type that may contribute to enhanced vector navigation capabilities. Together, these findings illuminate how evolution shapes neural circuits to produce behavioral diversity while maintaining core computational functions.
Valentin Gillet (Tue,) studied this question.