Aggressiveness predicts dominance rank in greylag geese: mirror tests and agonistic interactions

Abstract

Individual differences in aggressiveness, if consistent across time and contexts, may contribute to the long-term maintenance of social hierarchies in complex animal societies. Although agonistic interactions have previously been used to calculate individuals' positions within a dominance hierarchy, to date the repeatability of agonistic behaviour has not been tested when calculating social rank. Here, we examined the consistency and social relevance of aggressiveness as a personality trait in a free-flying population of greylag geese (Anser anser). For each individual, we quantified (i) aggressiveness using a standardized mirror stimulation test and (ii) dominance ranking based on the number of agonistic interactions won and lost in a feeding context. We found that individual differences in aggressiveness were significantly repeatable and that individuals' aggressiveness predicted their dominance rank position. The flock showed a robust and intermediately steep dominance hierarchy. Social rank was higher in paired birds, males and older birds, and most agonistic interactions occurred between individuals with moderate rank differences. We suggest that selection favours aggressiveness as a personality trait associated with resource acquisition and social rank, whereby a dominance hierarchy may increase the benefits of group living and reduce costs over conflict within dyads.

Keywords: anatidae; dominance hierarchy; individual differences; mirror stimulation test; personality; social rank.

Figure 1.

The relationship between aggressiveness (latency to approach mirror, seconds) and the number of agonistic interactions (log-transformed) that an individual (a) donated and (b) received. Raw data are presented as circles. Full model output in table 1. Shaded area in panel (a) represents 95% CI around the marginal effect regression line.

Figure 2.

Predicted values of the effect of the number of geese on the number of agonistic interactions (log-transformed). Feeding and non-feeding contexts are shown separately. Full model output in table 1. Shaded area represents 95% CI around the marginal effect regression line.

Figure 3.

Individual dominance rank with 95% CI (1000 randomizations) for each goose in July 2023, calculated from 1411 interactions. Dominance ranks are ordered from top (upper left) to bottom (lower right). Heterosexual paired males = deep blue, homosocial paired males = black, unpaired males = cyan, paired females = red, unpaired females = pink and birds of unknown sex = grey. Three geese (Eastwood, Joe and Rosenrot, shown at bottom right) were not represented in any interactions.

Figure 4.

Shape of the dominance hierarchy of greylag geese in July 2023, plotting the probability for a dominant goose to win a conflict with respect to the rank difference with its opponent, with 95% CI (1000 randomizations). Point size represents the number of interactions in the dataset for each rank difference.

Figure 5.

The relationship between goose dominance rank and (a) sex, (b) hatch year and (c) pair type. Raw data are presented as circles. Lower numbers indicate higher rank within the dominance hierarchy. Full model output in table 2a . Error bars in (a) and (c) represent 95% CI around the estimated marginal means, and the shaded area in (b) represents 95% CI around the marginal effect regression line.