The results in the present study demonstrate that rearing rainbow
trout at a density of 25 kg m−3 is associated with increased energy
expenditures, as evidenced by higher routinemetabolic rates compared
to groups reared at intermediate and high densities. Although behaviour
was not quantified directly, these findings are in agreement with
a previous study, in which we found that levels of aggression between
individuals held at 25 kgm−3, quantified by scale loss,were constrained
at high densities (Laursen et al., 2013). Interestingly, in the latter study,
a higher rearing density (140 kg m−3) led to chronic stress, as evidenced
by neuroendocrine parameters (elevated central signalling of
serotonin and low plasma levels of cortisol), while no such indications
were shown at low density. However, due to the practical issues of
performing quantitative behavioural observations on individual fish at
such high rearing densities, the indirect measures of aggression obtained
by Laursen et al. (2013) serve as an indication of increased aggressive
behavioural interactions at lowdensities. Here, it is therefore speculated
that fish reared at density of 25 kg m−3 were more engaged in aggressive
social interactions, and that this behaviour incurs the additional
energetic cost. In small experimental populations, Salmonid fishes can
be highly aggressive and form dominance hierarchies