An interaction mechanism for the maintenance of fission–fusion dynamics under different individual densities

• DOI: 10.7717/peerj.8974
• Publication Date: 2020-05-14T09:17:56Z
• Authors: David Bierbach; Stefan Krause; Pawel Romanczuk; Juliane Lukas; Lenin Arias-Rodri...; Jens Krause
• Keywords: 0106 biological sciences; Poecilia; Animal Behavior; QH301-705.5; Markov chain; R; 01 natural sciences; Social network analysis; Fission–fusion society; Cave fish; Medicine; 14. Life underwater; Biology (General)

Animals often show high consistency in their social organisation despite facing changing environmental conditions. Especially shoaling fish, fission–fusion dynamics that describe for which periods individuals are solitary or have been found to remain unaltered even when density changed. This compensatory ability is assumed be an adaptation towards constant predation pressure, but the mechanism through can actively compensate changes yet unknown. The aim of current study identify behavioural patterns enable this active compensation. We compared two populations live-bearing Atlantic molly ( Poecilia mexicana ) live adjacent habitats with very different predator regimes: cave mollies inhabit a low-predation environment inside sulfidic low predatory water bugs Belostoma sp.), and directly outside (henceforth called “surface” mollies) high-predation environment. analysed under fish densities 12 6 per 0.36 m 2 . As expected, surface spent more time being than mollies, difference was result less likely discontinue contact (once they had partner) resume alone) mollies. Interestingly, were also switch among partners A random walk simulation predicted each population reduced encounters treatment. While largely followed prediction, maintained interaction probabilities at density. Surface achieved by reduction size convex polygon formed group as decreased. may allow them maintain while still able visit large parts available area group. slight (21%) visited observed insufficient explain how dynamics. Finally, we discuss potential movement rules could account test performance.