Egalitarian cooperation linked to central oxytocin levels in communal breeding house mice

Abstract

Relationships between adult females are fundamental to understanding diversity in animal social systems. While cooperative relationships between kin are known to promote fitness benefits, the proximate mechanisms underlying this are not well understood. Here we show that when related female house mice (Mus musculus domesticus) cooperate to rear young communally, those with higher endogenous oxytocin levels have more egalitarian and successful cooperative relationships. Sisters with higher oxytocin concentrations in the paraventricular nucleus (PVN) of the hypothalamus weaned significantly more offspring, had lower reproductive skew and spent more equal proportions of time in the nest. By contrast, PVN oxytocin was unrelated to the number of weaned offspring produced in the absence of cooperation, and did not vary in response to manipulation of nest site availability or social cues of outgroup competition. By linking fitness consequences of cooperation with oxytocin, our findings have broad implications for understanding the evolution of egalitarian social relationships.

Fig. 1. Schematic overview of experiments.

Social groups of female house mice were studied in enclosures under the same controlled conditions with contrasting treatments. Social groups always consisted of two littermate sister pairs of contrasting age (older sister pairs are represented as larger than younger sister pairs). The older sisters within each group were the subjects of both experiments, and were given the opportunity to breed and rear their offspring communally. In experiment 1 (A), we manipulated the availability of protected nest sites for breeding (protected nest sites are shown as dark rectangles and unprotected nest sites as pale squares; protected nest sites were enclosed and accessed through a tunnel, unprotected nest sites were open), and the relatedness of younger non-breeding females living within the subjects’ territory (related females are represented using the same colour and unrelated females using different colours). In experiment 2 (B) we manipulated the presence or absence of neighbours (unrelated females living in a neighbouring territory), linked by connecting tunnels blocked with wire mesh (shown as dotted lines). Each enclosure contained two transponder readers that monitored the nest site attendance of subjects in occupied protected nest sites during post-natal day 0–14 (shown on the tunnel entrance to a protected nest site within each enclosure). A Upper left: older and younger sister pairs are related, and a single protected nest site is available. Upper right: older and younger sister pairs are related, and four protected nest sites are available. Lower left: older and younger sister pairs are unrelated, and a single protected nest site is available. Lower right: older and younger sister pairs are unrelated, and multiple protected nest sites are available. B Left: older and younger sister pairs are related, a single protected nest site is available, and no neighbours are present in an adjacent territory. Right: older and younger sister pairs are related, a single protected nest site is available, and neighbours are present in an adjacent territory.

Fig. 2. Relationship between PVN oxytocin concentrations of sister dyads from the same social group.

Subjects were from two experiments with independent manipulation of access to protected nest sites and outgroup competition. Subjects within each trial were classified as having a higher or lower PVN oxytocin concentration relative to one another. PVN oxytocin concentrations of sister dyads within the same social group were significantly correlated, irrespective of the experimental treatments (protected nest site availability and outgroup competition). See Supplementary Table 2 for statistical analysis.

Fig. 3. Relationship between the total numbers of offspring weaned per nest and the average PVN oxytocin concentration of sister dyads, according to whether or not the sister dyads reared young communally.

Subjects were from two experiments with independent manipulation of access to protected nest sites and outgroup competition. Higher average PVN oxytocin concentration of sister dyads was associated with a greater total number of weaned offspring when females cooperated to rear offspring in a communal nest (‘Communal nest’, red, solid line) but not when offspring were reared by a single female (‘Single nest’, blue, dashed line). See Table 2 for statistical analysis.

Fig. 4. Relationship between skew in time spent in the communal nest by communally breeding sister dyads during the active (dark) period and their average PVN oxytocin concentrations.

Subjects were from two experiments with independent manipulation of access to protected nest sites and outgroup competition. To analyse the relative time spent in the nest by communally breeding sister dyads we calculated the difference in the proportion of total time spent in the nest by each subject. Higher average PVN oxytocin concentration of sister dyads was associated with a lower skew in the relative time they spent in the communal nest, indicating that each subject spent more similar time in the nest around the time of peak maternal investment. See Table 3 for statistical analysis.