Stress levels of dominants reflect underlying conflicts with subordinates in a cooperatively breeding species

Maintaining dominance status had long been considered to be less stressful than subordination. However, no consistency in stress levels of dominant and subordinate individuals has been demonstrated. Tactics used to achieve and maintain dominance could be determinant. In cooperatively breeding species, conflicts between dominants and subordinates are expected since dominant individuals tend to monopolize reproduction while subordinates seldom reproduce. Reproductive skew models predict that subordinates' reproductive opportunities are either allotted or subject to competition with dominants. In the former case, no policing of subordinates by dominants is expected. In the latter, dominant should exert a control over the subordinates possibly leading to higher stress levels in dominants than in subordinates, which could be further elevated as the number of potential competitors in the group increases. In the present study, we aimed to test these hypotheses by assessing individual's stress level using the neutrophils to lymphocytes ratio (N:L) in a wild cooperatively breeding rodent, the Alpinemarmot (Marmota marmota). We found that dominants exhibit higher N:L ratio than subordinates and that dominants' N:L ratio increases with the number of unrelated same-sex subordinates in the group. We conclude that controlling unrelated subordinates is stressful for dominants, as expected under tug-of-war models. These stress patterns reveal conflicting relationships between dominants and subordinates over the reproduction and social status acquisition. This study highlights the influence of the nature, strength, and direction of conflicts on stress levels. Significance Statement In cooperatively breeding species, reproductive skew models predict that subordinates' reproductive opportunities are either allotted or subject to competition with dominants and, thus, can modulate the relative stress level between dominants and subordinates. In the first case, no policing of subordinates by dominants is expected, while in the second, dominant should exert a control over the subordinates which should lead to higher stress level in dominants than in subordinates, and particularly when subordinates are unrelated to the dominants. In Alpine marmots, we found that dominants exhibit higher stress level than subordinates and dominants' stress level increased with the number of unrelated same-sex subordinates. These patterns are in agreement with the predictions of the tug-of war models of reproductive skew and indicate that controlling subordinates is costly for dominants.