Effects of cortisol on aggression and locomotor activity in rainbow trout.

Noninvasive administration of cortisol through the diet resulted in relatively rapid (<1.5 h) and highly reproducible increases in plasma cortisol in rainbow trout, comparable to changes seen in fish subjected to substantial stress. Juvenile rainbow trout were reared in isolation for 1 week, before their daily food ration was replaced by a meal of cortisol-treated food corresponding to 6 mg cortisol kg(-1). All fish were observed for 30 min, beginning at 1 or 48 h following the introduction of cortisol-treated food. Additional cortisol (75% of the original dose on Day 2, and 50% on Day 3) was administered to the long-term cortisol-treated group. The resulting blood plasma concentrations of cortisol were similar in short- and long-term treated fish, and corresponded to those previously seen in stressed rainbow trout. Controls were fed similar food without cortisol. Half of the fish from each treatment group (controls and short- and long-term cortisol) were subjected to an intruder test (a smaller conspecific introduced into the aquarium), while half of the fish were observed in isolation. In fish challenged by a conspecific intruder, short-term cortisol treatment stimulated locomotor activity, while long-term treatment inhibited locomotion. Aggressive behavior was also inhibited by long-term cortisol treatment, but not by short-term exposure to cortisol. Cortisol treatment had no effect on locomotor activity in undisturbed fish, indicating that the behavioral effects of cortisol were mediated through interaction with other signal systems activated during the simulated territorial intrusion test. This study demonstrates for the first time that cortisol has time- and context-dependent effects on behavior in teleost fish.