Login / Signup

Reorientation and propulsion in fast-starting zebrafish larvae: an inverse dynamics analysis.

Cees J VoesenekRemco P M PietersFlorian Titus MuijresJohan L van Leeuwen
Published in: The Journal of experimental biology (2019)
Most fish species use fast starts to escape from predators. Zebrafish larvae perform effective fast starts immediately after hatching. They use a C-start, where the body curls into a C-shape, and then unfolds to accelerate. These escape responses need to fulfil a number of functional demands, under the constraints of the fluid environment and the larva's body shape. Primarily, the larvae need to generate sufficient escape speed in a wide range of possible directions, in a short-enough time. In this study, we examined how the larvae meet these demands. We filmed fast starts of zebrafish larvae with a unique five-camera setup with high spatiotemporal resolution. From these videos, we reconstructed the 3D swimming motion with an automated method and from these data calculated resultant hydrodynamic forces and, for the first time, 3D torques. We show that zebrafish larvae reorient mostly in the first stage of the start by producing a strong yaw torque, often without using the pectoral fins. This reorientation is expressed as the body angle, a measure that represents the rotation of the complete body, rather than the commonly used head angle. The fish accelerates its centre of mass mostly in stage 2 by generating a considerable force peak while the fish 'unfolds'. The escape direction of the fish correlates strongly with the amount of body curvature in stage 1, while the escape speed correlates strongly with the duration of the start. This may allow the fish to independently control the direction and speed of the escape.
Keyphrases
  • aedes aegypti
  • drosophila melanogaster
  • high resolution
  • zika virus
  • machine learning
  • high speed
  • artificial intelligence
  • convolutional neural network