Irschick, D. | Search bioRxiv

bioRxiv2026

Dynamic dorsal body morphology encodes engineering design principles of fish propulsion and hydrodynamics

Understanding how biological shape and movement interact with surrounding fluids represents a fundamental challenge at the intersection of biology, physics, and engineering. Fish locomotion exemplifies this challenge: body morphology and swimming kinematics together determine the hydrodynamic forces and flow structures that enable efficient propulsion and maneuverability. Whereas biologists have long sought to connect morphological variation to swimming performance, traditional morphometric approaches provide limited insight into the fluid mechanical consequences of shape differences. Similarly, although computational fluid dynamics can reveal detailed flow physics, simulating hydrodynamics across diverse and dynamic morphologies remains prohibitively expensive for systematic investigation. To bridge this gap, we introduce a data-driven framework that connects fish body shape dynamics to hydro-dynamic performance through compact morphospace parameterization and reduced-order modeling. Using CFD simulations of 15 fish species from the Digital Life Project database (www.digitallife3d.org/3d-model), we generate hydrodynamic datasets capturing the shape-flow relationship. Principal Component Analysis (PCA) extracts four dominant shape parameters from dorsal body profiles, which are then integrated into an Inverse-Design with Dynamic Mode Decomposition (ID-DMD) framework to model the resulting fluid dynamics. The resulting modal analysis suggests that locomotion strategies emerge from specific shape-flow interactions. We further demonstrate the frameworks utility through single- and multi-objective shape optimization, showing how it enables efficient exploration of the morphology-hydrodynamics relationship. This approach offers a novel analysis and design tool for understanding how biological form and motion interact with fluid mechanics, with applications ranging from bio-inspired vehicle development to evolutionary biomechanics.

biophysics↗

bioRxiv2024

PicoCam: High-resolution 3D imaging of live animals and preserved specimens

1. The PicoCam system is a multi-camera photogrammetry rig used for generating high-resolution 3D models of animals in the field, or preserved specimens in laboratory or museum settings. The digital measurement of 3D models is increasingly useful for studying body shape. However, methods that capture sub-millimetric detail often do so at the cost of portability and versatility; this system aims to bridge this gap. 2. The PicoCam system employs 3D digital photogrammetry, a process that generates accurate, full-color, 3D models from sequences of photographs. By using six cameras and a rotating base, the system is able to capture multiple angles in rapid succession - a key advantage for both 3D-imaging live specimens and efficiently scanning museum specimens. Through the use of macro lenses and high-resolution camera sensors, this system can capture sub-millimetric detail without sacrificing portability. 3. In this study, we 3D imaged the bills of 19 species of hummingbirds using the PicoCam system and measured length, height, width, surface area, and volume of their bills. We examined eight species in the field and 11 from the Burke Museum in Seattle. We chose Hummingbird bills as a model system, as their fine-scale interspecific differences in 3D shape can have significant functional and behavioral implications, and could tell us more about how these traits predict habitat and resource use. 4. The prospect of a common 3D-imaging method for both museum and field use is compelling when documenting structures shape within and among species. The PicoCam system is also valuable for quantifying the 3D shape of fine-scale phenotypes (like hummingbird bills) that benefit from digital measurement, preservation, and improved accessibility. Lastly, the PicoCam system allows "digital 3D collection", by which the shape of biological structures in the field can be recorded and stored in a public database without the need to collect a specimen. This opens the door to studies in which multiple 3D image captures of the same individual across different time scales permit 3D shape/color comparisons (e.g., seasonal, ontogenetic changes).

evolutionary biology↗