Direkt zum InhaltDirekt zur SucheDirekt zur Navigation
▼ Zielgruppen ▼

Humboldt-Universität zu Berlin - Institut für Biologie

Verhaltensphysiologie

Prof. Dr. Bernhard Ronacher

Prof. Dr. Bernhard Ronacher


In earlier days I compared principles of visual pattern recognition in insects (honeybees) and humans using multidimensional scaling methods.

My recent research focuses on two different species and fields:
(i) the recognition of acoustic communication signals in grasshoppers, and
(ii) navigation of desert ants


Ad (i): We aim at elucidating general principles of how acoustic signals are processed by a small nervous system. Grasshoppers are a good model system for these questions since the recognition of communication signals is vital for mate finding and thus has direct consequences for their fitness. Furthermore, the investigations can rely on a hard-wired neuronal basis since the production and the recognition of the species-specific signals are both innate behaviours. Our approach combines a tight interplay of behavioural and electrophysiological experiments (intracellular single cell recordings), complemented by theoretical methods (e.g. spike train metrics).

Ad(ii) Desert ants perform foraging excursions that extend over several ten thousand body lengths. They do not rely on scent marks to find back to their nest, rather they use path integration as a means of navigation. In collaboration with Rüdiger Wehner (Zürich) we investigated how ants estimate their walking distance, which is an essential input to perform path integration. Unexpectedly, when trained to walk over hills, the ants used not the actually walked distances but the horizontal projections of their three-dimensional itinerary – suggesting an unexpected sophistication of their path integration module. We now have expanded these experiments to the second ingredient of path integration, the compass which depends primarily on the sky’s polarization pattern. Using linear polarization filters we manipulate the information available for the ants to explore how different compass mechanisms interact and potentially complement each other. Such interactions may reduce a fundamental problem of path integration, which is inherently prone to the accumulation of errors.

Weiter zu den Webseiten der Arbeitsgruppe