Functional models from limited data: A parametric and multimodal approach to anatomy and 3D kinematics of feeding in basking sharks (Cetorhinus maximus).

Abstract

Basking sharks, Cetorhinus maximus (Gunnerus, Brugden [Squalus maximus], Det Kongelige Norske Videnskabers Selskabs Skrifter, 1765, vol. 3, pp. 33-49), feed by gaping their mouths and gill slits, greatly reorienting their cranial skeletons to filter food from water. The 3D biomechanics of this behavior, however, are exceptionally difficult to study due to the size, elusiveness, and CITES status of these animals and the rarity of well-preserved specimens. To overcome these challenges, we integrated anatomical, digital design, and computer imaging approaches to reconstruct bio-realistic and poseable 3D skeletal models of feeding basking sharks. The skeleton, segmented from CT scans of intact heads, was first abstracted as a rigging for guiding skeletal positioning in 3D space. Directed by the anatomies of museum specimens and dissected beached animals, the digital scaffolding was used to virtually correct skeletal distortions (e.g., from specimen collapse), resetting the skeleton to closed-mouth symmetry. Open-mouthed feeding postures were recreated by repositioning skeletal joints to biologically relevant destination coordinates defined from videos of feeding sharks, exploiting the basking shark's steady feeding posture to build 3D photogrammetry models from successive video frames. The resultant "digital puppet" bridges diverse imaging data while capturing the coordinated motion of "hidden" cranial joints, deconstructing complex form-function relationships into computationally controllable parameters for exploring 3D skeletal movement. The input data gathered for our model provides new perspectives on basking shark cranial anatomy, while the model's biological fidelity gives insights into dynamic feeding processes impossible to observe in the laboratory. Branchial arch mechanics are comparatively poorly studied in sharks; our model can act as a platform for future kinematic modeling (e.g., of individual variation, other species), while demonstrating interdisciplinary approaches for studying large and elusive wildlife.