Despite their relatively small brains with comparatively low neuron counts, insects show complex navigation behavior such as seasonal long-range migration, path integration, and precise straight-line movement. Spatial navigation requires a sense of current heading, which must be tethered to prominent external cues and updated by internal cues that result from movement.
Global external cues such as the position of the sun may provide a reference frame for orientation. Sunlight is polarized by scattering in the atmosphere, which results in a sky-spanning polarization pattern that directly depends on the current solar position and makes polarization information, like the sun itself, useful as an external reference cue. Internally, moving through the environment generates optic flow---the motion of the viewed scenery on the retina---, which may inform about turning maneuvers, movement speed, and covered distance. Many insects use these external and internal cues for orientation, and the neuronal center for spatial navigation likely is the central complex, a higher-order brain structure where sensory information is integrated to form an internal compass representation of the current heading.
This thesis addresses the question how celestial compass cues, specifically the polarization pattern, and optic flow are processed in the central complex of the desert locust, a long-range migratory insect. All chapters except the last one are electrophysiological studies in which single central-complex neurons were intracellularly recorded while presenting visual stimuli. The neurons' anatomy was histologically determined by dye injection in order to infer their role in the neural network.
The studies in Chapters 1 and 2 show that the central complex contains a neuronal compass that robustly signals the sun direction based on direct sunlight and the integration of the whole solar polarization pattern. This shows that the locust brain uses all available skylight cues in order to form a unified compass signal, enabling robust navigation under different environmental conditions.
The study in Chapter 3 further examines how neurons at the input stage of the central complex process skylight cues. Already at this stage, single neurons integrate visual information from large areas of the sky and have receptive fields suitable to build the skylight compass.
Chapter 4 sheds light on the detection sensitivity for the angle of polarization, finding that central-complex neurons are highly sensitive in this regard, adapted to analyze the skylight polarization pattern almost in its entirety and under unfavorable environmental conditions.
In Chapter 5 the locust central complex was scanned for neurons that receive optic flow information. Neurons at virtually all network stages are sensitive to optic flow, mainly uncoupled from skylight-cue sensitivity. This highlights that sensory information is flexibly processed in the central complex, presumably depending on the animal's current behavioral demands. Further, the study hypothesizes how horizontal turning motion is processed in order to update the internal heading representation, backed up by a computational model that adheres to brain anatomy and physiological data.
Altogether, these studies advance the understanding of how external and internal cues are processed in the central-complex network in order to establish a sense of orientation in the insect brain.
Finally, I contributed with data sets and programming code to the development of the InsectBrainDatabase (www.insectbraindb.org), a free online database tool designed to manage, share and publish anatomical and functional research data (Chapter 6).