The relationships among sprint performance, voluntary swimming activity, and social dominance in juvenile rainbow trout.

The specific objectives of this study were to determine whether sprint performance in juvenile rainbow trout is correlated with either voluntary swimming activity or aggressive behaviors and to determine the reciprocal: the effect of swimming activity and aggression on sprint performance. Sprint performance was assessed by rapidly accelerating trout (5-7-cm fork length) to a fixed velocity (40, 42, or 45 cm s(-1)) and then holding them at that velocity until fatigue. There was considerable interindividual variation in sprint performance not explained by variations in body size, but intraindividual performance was highly repeatable over at least 2 mo. Voluntary swimming was measured as the frequency of transits (voluntary transit activity, VTA) between two identical tanks via a connecting channel with two different flow regimes: zero or minimum velocity (0 or 2.5 cm s(-1)) and high velocity (84 cm s(-1)). There was a strong correlation between sprint performance and VTA in minimal current but no correlation in high current. Furthermore, sprint performance did not predict the outcome of dominance encounters. Experience with rapid acceleration, especially when voluntary, led to a pronounced improvement in sprint performance in proportion to the number of acceleration events. Social dominance encounters had a more complex effect: a significant reduction in sprint performance in previously high-performance sprinters and the reverse for low performers. We propose that there are four independent axes of interindividual variation in juvenile rainbow trout: spontaneous and rheotaxis-stimulated locomotor activity, aggressive activity, and the trainability of sprint performance. The independence of these axes has the potential to produce a much larger diversity in behavioral and ultimately physiological phenotypes than would be produced if the axes were linked.