One key event in insect evolution was the development of mandibles with two joints, which allowed powerful biting, but restricted their movement to a single degree of freedom. These mandibles define the Dicondylia, which constitute over 99 percent of all extant insect species. It was common doctrine that the dicondylic articulation of chewing mandibles remained unaltered for more than 400 million years. We report highly modified mandibles overcoming the restrictions of a single degree of freedom and hypothesize their major role in insect diversification. These mandibles are defining features of parasitoid chalcid wasps, one of the most species-rich lineages of insects. The shift from powerful chewing to precise cutting likely facilitated adaptations to parasitize hosts hidden in hard substrates, which pose challenges to the emerging wasps. We reveal a crucial step in insect evolution and highlight the importance of comprehensive studies even of putatively well-known systems.

Tomographic scans of ethanol-preserved insect heads were performed at the UFO-I station of the Imaging Cluster at the KIT light source using a parallel polychromatic X-ray beam produced by a 1.5 T bending magnet. The beam was spectrally filtered by 0.5 mm aluminum and the resulting spectrum had a peak at about 15 keV, with a full-width at half maximum bandwidth of about 10 keV. A fast indirect detector system was employed, consisting of a 12 µm LSO:Tb scintillator and a diffraction limited optical microscope (Optique Peter) coupled with a 12bit pco.dimax high speed camera with 2016 x 2016 pixels. Scans were done by taking 3,000 projections at 70 fps and an optical magnification of 10x, resulting in an effective pixel size of 1.22 µm. We used the control system concert for automated data acquisition and online reconstruction of tomographic slices for data quality assurance. Online and final data processing included flat field correction and phase retrieval of the projections based on the transport of intensity equation. X-ray beam parameters for algorithms in the data processing pipeline were computed by syris and the execution of the pipelines, including tomographic reconstruction, was performed by the UFO framework.