Physiological effects of dominance hierarchies: laboratory artefacts or natural phenomena?