Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 285 (1872)

Repeatable Group Differences in the Collective Behaviour of Stickleback Shoals Across Ecological Contexts


Repeatable Group Differences in the Collective Behaviour of Stickleback Shoals Across Ecological Contexts

Jolle W Jolles et al. Proc Biol Sci.


Establishing how collective behaviour emerges is central to our understanding of animal societies. Previous research has highlighted how universal interaction rules shape collective behaviour, and that individual differences can drive group functioning. Groups themselves may also differ considerably in their collective behaviour, but little is known about the consistency of such group variation, especially across different ecological contexts that may alter individuals' behavioural responses. Here, we test if randomly composed groups of sticklebacks differ consistently from one another in both their structure and movement dynamics across an open environment, an environment with food, and an environment with food and shelter. Based on high-resolution tracking data of the free-swimming shoals, we found large context-associated changes in the average behaviour of the groups. But despite these changes and limited social familiarity among group members, substantial and predictable behavioural differences between the groups persisted both within and across the different contexts (group-level repeatability): some groups moved consistently faster, more cohesively, showed stronger alignment and/or clearer leadership than other groups. These results suggest that among-group heterogeneity could be a widespread feature in animal societies. Future work that considers group-level variation in collective behaviour may help understand the selective pressures that shape how animal collectives form and function.

Keywords: collective behaviour; group differences; group personality; schooling; sociality; stickleback.

Conflict of interest statement

The authors declare no competing interests.


Figure 1.
Figure 1.
Schematic of the tank in which the groups of fish were tested across three different contexts: (i) the open context, an environment without food or plant cover, (ii) the foraging context, an environment with three patches of food* and (iii) the cover context, an environment with food patches* as well as plant cover*. Tracking segments are shown for one randomly selected group.
Figure 2.
Figure 2.
Heat maps depicting (a) the probability of finding neighbouring fish at a given position relative to the position of the focal fish placed at the origin pointing up, with distances expressed in units body length (BL), and (b) the relationship between group speed (mean individual speed in the group) and polarization (alignment among all group members). Panels represent the open context (i), the foraging context until all food was depleted (ii), the foraging context after all food was depleted (iii) and the cover context (iv). Plots were calculated based on a frame-by-frame basis with time steps of 1/24th s and are presented in percentages relative to the densest bin for that context. These figures show that (a) neighbouring fish could primarily be found within two body lengths side-by-side and four body lengths front-to-back, with groups being most cohesive in the open context, and (b) that there was a strong positive relationship between the speed and polarization of the groups that was not itself strongly affected by the context.
Figure 3.
Figure 3.
Line plots of group behaviour across the three contexts, with the foraging context split into time before and after food was depleted: (a) average individual median speed in the group, (b) mean group centre distance, (c) median polarization and (d) leadership structure, the variance in the proportion of time group members were in front of the group centre. Lines indicate groups, with colours indicating group rank in terms of the group's behaviour in the open context. Black diamonds indicate average group behaviour per context with error bars indicating 95% creditability intervals in terms of certainty of the mean extracted from the MCMCglmms. Non-overlapping confidence intervals indicate a significant difference.

Similar articles

See all similar articles

Cited by 7 articles

See all "Cited by" articles


    1. Sumpter DJT. 2010. Collective animal behavior. Princeton, NJ: Princeton University Press.
    1. Couzin ID, Krause J, James R, Ruxton GD, Franks NR. 2002. Collective memory and spatial sorting in animal groups. J. Theor. Biol. 218, 1–11. (10.1006/yjtbi.3065) - DOI - PubMed
    1. Couzin ID, Krause J. 2003. Self-organization and collective behavior in vertebrates. Adv. Study Behav. 32, 1–75. (10.1016/S0065-3454(03)01001-5) - DOI
    1. Gautrais J, Ginelli F, Fournier R, Blanco S, Soria M, Chaté H, Theraulaz G. 2012. Deciphering interactions in moving animal groups. PLoS Comput. Biol. 8, e1002678 (10.1371/journal.pcbi.1002678) - DOI - PMC - PubMed
    1. Herbert-Read JE, Perna A, Mann RP, Schaerf TM, Sumpter DJT, Ward AJW. 2011. Inferring the rules of interaction of shoaling fish. Proc. Natl Acad. Sci. USA 108, 18 726–18 731. (10.1073/pnas.1109355108) - DOI - PMC - PubMed

Publication types