Fish In Silico with Hydrodynamic and Social Interactions Forces – FISHSIF
The role of hydrodynamics in fish communication
How do fish and other aquatic species communicate to form coherent schools? Numerous social models, mainly based on vision, have succeeded in reproducing collective behavior and movement. Very few studies, however, involve hydrodynamics in this communication. Yet we know that fish are sensitive to flow via their lateral line, which enables them to measure the movement of other fish and obstacles.
Collective motions of fish in complex media.
From planktonic species (a few micrometers) to large mammals (up to several meters), the movements of aquatic organisms cover a wide range of length and time scales. Their individual and collective movements, as well as their mutual intra- and interspecific interactions, have significant ecological and environmental impacts. Several social models have been successful in predicting the behavior of schools of fish based on their mutual cognitive interactions (i.e. perception/action), modeled by social forces. While these social approaches apprehend fish school structures as the result of an internal social life or interactions with their environment (conspecifics, predators, prey, obstacles, food), another idea has been regularly put forward by hydrodynamicists, claiming that fish could benefit from schooling behavior in a perspective of locomotion efficiency. The question that remains open in biology is to be able to estimate the relative importance of cognitive positioning - compared to hydrodynamic positioning - within the shoal. Indeed, it has been demonstrated experimentally that hydrodynamic drag is drastically reduced within a shoal, enabling the fish to save energy. Hydrodynamic simulation goes in the same direction, but is generally applied to configurations with given fixed geometries, e.g. tandems, phalanges or diamonds. These static hydrodynamic approaches are therefore unable to compare the relative importance of social and hydrodynamic forces: compound models describing the dynamics of fish school formation are still lacking. The aim of this project is to couple dynamic models of cognitive behavior with comprehensive hydrodynamic calculations. The resulting numerical simulations will help us to understand how hydrodynamics influence the organization within a school of several individuals, and how a complex environment can shape this organization. This coupling between hydrodynamic interactions and cognitive models is new and has never been proposed in the scientific literature before. It will represent a breakthrough in physical models and serve as an effective tool for behavioral biology. The construction of the physical model will benefit from direct comparison with controlled live experiments and ethological data in order to optimize the balance between physical and behavioral forces.
On the one hand, we have developed experiments to understand how collective movements are perturbed by a complex medium: i) evacuation by a constriction, ii) movement in a network of pillars. On the other hand, we have developed a minimal model of a swimmer without body deformation, based on force and torque dipoles, which enables precise Navier-Stokes calculations in 3D.
As the complex medium disturbs the collective motion, this allows us to probe the important interactions that “weld” a shoal together, and thus to deduce the most important and robust parameters in existing social models.
In the evacuation experiment i) we change the size of the opening through which small fish of a few centimetres (Paracheirodon Axelrodi), better known as Cardinalis, escape. The fish are pushed towards the opening by a regular movement of a landing net. The experiment is repeated around twenty times to derive reliable statistics.
In the pillar experiment (ii), we use Danio rerio, better known as zebrafish. We vary pillar density to understand how the shoal is disturbed by pillars as their density increases.
Our minimal numerical model can reproduce the propulsion of a swimmer for a wide range of Reynolds numbers, and generate wake vortices in the inertial regime, reminiscent of the flow generated by the flapping tails of real fish. The idea is to use this simplified model to simulate a large number of fish in 3D hydrodynamic interaction, combined with proven models for social interaction.
The study i) between theoretical partners, experimentalists and ethologists focused on the collective social behavior of fish having to evacuate through a constriction. Published in 2023 [1]. This study demonstrated that, unlike most animals and humans, fish respect a social bubble and avoid contact even in a state of panic when evacuating an area of the aquarium. This work has been very well received in the mainstream press (BBC Focus, Telegraph of India, Mediapart, etc.) [2].
A collaborative study (ii) between theoretical/numerical and experimental partners was published [3,6] on collective movement in a complex environment (network of pillars). This study demonstrated that when the obstacles (the pillars) are at a distance less than the social distance between fish, an abrupt break in collective movement occurs and the shoal is diluted in the entire aquarium.
A third study, a numerical one, concerns the development of a minimal model of an aquatic organism's individual swim and its hydrodynamic signature. This study is currently being submitted [4,6].
Unlike most models based on the deformations of the swimmer's body, our original numerical model is based on the forces produced by an organism on the surrounding fluid as it swims. This model is minimal, but it enables us to accurately and rapidly resolve the flow produced by the swimmer, and we have been able to extract a universal behavior from the Stokes regime (the swimming regime of micro-organisms) to turbulent scales (large fish or marine mammals). This law relates the Reynolds number, which depends on the swimmer's speed, to a new dimensionless number, the thrust number, linked to the amplitude of the forces exerted. Thanks to its numerical lightness, this model will enable us to study the collective movements of several dozen or even hundreds of swimmers.
To popularize our work, we have already had two videos made by a professional concerning the first two studies in our project [2] and participated in a conference for the general public [5].
[1] R. Larrieu, P. Moreau, C. Graff, P. Peyla & A. Dupont, Fish ecacuate respecting a social bubble, Scientific Reports, 13, 10414 (2023)
[2] Mainstream press and videos: liphy-annuaire.univ-grenoble-alpes.fr/pages_personnelles/philippe_peyla/communications.html
[3] B. Ventéjou, I. Magniez- -Papillon, E. Bertin, P. Peyla & A . Dupont, Behavioral transition of a fish school in a crowded environment, Phys. Rev. E 109, 064403
[4] B. Ventéjou, T. Métivet, A. Dupont & P. Peyla, Universal Scaling Laws for a Generic Swimmer Model, arxiv.org/abs/2407.04511 (2024)
[5] Conference for the general public:
pintofscience.fr/event/de-lindividu-au-banc-une-modelisation-de-poisson
[6] Data base: perscido.univ-grenoble-alpes.fr/datasets/DS406
Our aim is now to use our numerical model to couple hydrodynamic interactions with a 3D social model currently under study. Our innovative numerical method makes it possible to take into account dozens or even hundreds of fish without altering calculation time. This represents a major step forward, paving the way for numerical simulations that can be compared with the experiments described above.
The integration of external elements such as obstacles or external flow is easily implemented in our code. Our FEEL++ [1] based code executable with input and output files and user manual will be made freely available.
More broadly, we believe that our work opens the way to the study of active matter at high Reynolds numbers, which is currently confined to the microscopic scale and Stokes regimes.
Translated with DeepL.com (free version)
Forcing a fish school through a bottleneck: a smooth evacuation
Renaud Larrieu, Philippe Moreau, Christian Graff, Philippe Peyla & Aurélie Dupont, Scientific Reports 13, 10414 (2023)
Collective movement in living organisms is a phenomenon of self-organization of a large number of individuals observed in nature from the micrometer scale (bacteria, plankton) to meters (birds) or even kilometers (school of sardines). This organization at a scale much larger than each individual is the result of interactions with the social (conspecifics), ecological (predators) and physical (obstacles, current) environment. The goal of this project is to understand how the coupling between ethodynamic and hydrodynamic interactions influence the organization of small fish schools, and how a complex environment can modify this organization. Thus, we hope to understand how the relative balance between social and hydrodynamic forces evolves with environmental conditions. To do so, our numerical simulations will combine a cognitive model with the direct resolution of 3D hydrodynamics. The construction of the physical model, in particular the balance between physical and behavioral constraints, will benefit from a constant dialogue between in vivo experiments in controlled environments and theory.
Project coordination
Philippe Peyla (Laboratoire Interdisciplinaire de Physique)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partnership
LIPHY Laboratoire Interdisciplinaire de Physique
LPNC LABORATOIRE DE PSYCHOLOGIE ET NEUROCOGNITION
INRIA GRA Inria Grenoble Rhône-Alpes
Help of the ANR 230,200 euros
Beginning and duration of the scientific project:
March 2022
- 36 Months
