Geometry, features, and panoramic views: ants in rectangular arenas

Antoine Wystrach*, Ken Cheng, Sebastian Sosa, Guy Beugnon

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    57 Citations (Scopus)

    Abstract

    When tested in rectangular arenas, the navigational behavior of the ant Gigantiops destructor can produce results similar to vertebrates. Such results are usually interpreted as supporting the ability of animals to segregate spatial geometry and features. Here, we combine a detailed analysis of ants' paths with panoramic images taken from the ant's perspective that can serve as a basis for developing view-based matching models. The corner choices observed in ants were better predicted by the use of panoramic views along with a simple matching process [rotational image difference function (rIDF)] than by models assuming segregation of geometry and features (G/F). Our view-based matching model could also explain some aspects of the ants' path (i.e., initial direction, length) resulting from the different visual conditions, suggesting that ants were using such a taxon-like strategy. Analyzed at the individual level, the results show that ants' idiosyncratic paths tend to evolve gradually from trial to trial, revealing that the ants were partially updating their route memory after each trial. This study illustrates the remarkable flexibilities that can arise from the use of taxon-like strategies and stresses the importance of considering them in vertebrates.

    Original languageEnglish
    Pages (from-to)420-435
    Number of pages16
    JournalJournal of Experimental Psychology: Animal Behavior Processes
    Volume37
    Issue number4
    DOIs
    Publication statusPublished - Oct 2011

    Keywords

    • ant navigation
    • Gigantiops destructor
    • view-based matching
    • visual compass
    • geometry

    Fingerprint

    Dive into the research topics of 'Geometry, features, and panoramic views: ants in rectangular arenas'. Together they form a unique fingerprint.

    Cite this