Using a robotic fish to investigate individual differences in social responsiveness in the guppy

David Bierbach, Tim Landgraf, Pawel Romanczuk, Juliane Lukas, Hai Nguyen, Max Wolf, Jens Krause

Preprint posted on 19 April 2018

Article now published in Royal Society Open Science at

The robotic fish are coming - to socialise! Bierbach et al. use a magnet controlled, 3D-printed fish replica to study social responsiveness in guppies.

Selected by Rasmus Ern


Responding towards the actions of others is one of the most important behavioral traits whenever animals of the same species interact. Studies on the social responsiveness of fishes may improve our understanding of the spatial and temporal behaviour of this important group in the wild.

When assessing the social responsiveness of fishes towards conspecifics (members of the same species), mutual influences among interacting individuals makes it difficult to isolate the responsiveness of individual animals. The use of biomimetic robots allows researchers to circumvent this issue, but requires that the robots are accepted as conspecifics by the study species.

In the first part of this study, Bierbach et al. assessed if a magnet controlled, 3D-printed fish replica (Robofish) is accepted as conspecific by guppies (Poecilia reticulate), and if individual guppies differ in their social responsiveness.

Figure 1 Guppy replica with a group of female guppies in the test arena.

Behavioral traits may form correlated suits, so-called behavioral syndromes. An example of this is sticklebacks, where individuals with increased ‘boldness’ or ‘exploration behavior’ are less attracted by other individuals.

In the second part of this study, Bierbach et al. assessed if individual differences in social responsiveness in P. reticulate are linked to boldness and activity.


For the first part, a focal fish and a companion (Robofish or another live fish) were released in a glass tank, and their movement patterns determined via video tracking equipment. Social responsiveness was assessed via the distance between the focal fish and the companion (Inter-individual distance, IID), the strength with which focal fish adjust their own movement towards that of the companion’s movement (Time-lagged cross-correlation of velocity vectors, TLXC), and the flow of movement information between the focal fish and the companion (Transfer Entropy, TE).

For the second part, a focal fish was introduced into an opaque plastic cylinder with a small opening, sealed with a sponge. Boldness was estimated from the time it took the fish to leave the cylinder when the sponge was removed. Social responsiveness was assessed via a Robofish positioned immediately outside the cylinder. Activity was estimated from the mean velocity over a 3 min period after removal of the Robofish.


Overall, the distance between subjects was longer (IID), velocity correlations were less pronounced (TLXC), and information transfer rates were lower (TE) when focal fish were paired with Robofish. This dampened interaction pattern was attributed to the lack of mutual responsiveness present when focal fish were paired with live companions. Despite these differences, individuals exhibited consistent differences in responsiveness when interacting with a Robofish and a live companion, indicating that Robofish is accepted as conspecific by P. reticulate.

Social responsiveness was not correlated with boldness or activity, indicating that it is not part of a behavioral syndrome with these two correlating traits.


I recently conducted a study on hypoxia avoidance behavior in red drum (Sciaenops ocellatus). Experimental trials were performed on individual fish under controlled laboratory conditions. For the results to be useful from an environmental change perspective, I need to know if hypoxia avoidance behavior in the laboratory is representative of avoidance behavior in the wild. Reading this article made me think that for schooling fish species, it is also necessary to know if hypoxia avoidance behaviour of individual fish is representative of the fishes’ avoidance behaviour when they are schooling.

Future directions:

It would be interesting to test if the acceptance of Robofish as a conspecific by P. reticulate is retained in the presence of external stressors (e.g., predator odours or aquatic hypoxia). If acceptance is retained, futures studies could assess the degree to which (predator or hypoxia) avoidance behaviour is governed by the behavior of leading individuals (see related literature).

Related literature:

Bierbach et al. (preprint). Guppies prefer to follow large (robot) leaders irrespective of own size. Doi:


Posted on: 26 June 2018

Read preprint (No Ratings Yet)

Have your say

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Register here