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

Studies of fish behaviour have demonstrated the existence of social interactions that result in dominance hierarchies. In environments in which resources, such as food, shelter and mates, are limited, social competition results in some fish becoming dominant and occupying the most profitable positions. This behaviour has been observed in natural environments and also in many laboratory‐based experiments. When two fish have been confined in a small tank, one of them usually has exhibited behaviour that suggests it is dominant over the other submissive animal. Physiological consequences of social interaction can be seen in both dominants and subordinates but are more extreme in the subordinate. However, this scenario is without doubt an artificial situation. Fewer experiments have been conducted using laboratory experiments that are more socially and physically complex than those experienced by dyads in tanks. In simple fluvial tanks, through which water is recirculated, the physiological responses of fish to social competition have generally been qualitatively similar to those recorded among dyads. However, when environmental disturbances, complex resource distributions, increase in water flushing, presence of predators and competing species of fish have been included in experimen‐tal designs, there have been fewer, diminished or no physiological dierences between dominant and subordinate fish. There have been very few studies of physiology in relation to dominance in natural habitats, and those that have been conducted suggest that under some circumstances hierarchies may cause less intense physiological responses than have been suggested based on results of laboratory studies in simple environments. Possible reasons for these variations are discussed. The need is identified for a well structured experimental approach to the investi‐gation of the causes and consequences of hierarchies if the ecology of wild fish is to be modelled eectively based on physiological processes. It is also suggested that the further development and application of techniques for monitoring physiologies of fish in the wild is important.