Aggressive encounters lead to negative affective state in fish

Abstract

Animals show various behavioural, neural and physiological changes in response to losing aggressive encounters. Here, we investigated affective state, which are emotion-like processes influenced by positive or negative experiences, in a territorial fish following aggressive encounters and explore links to bold/shy behavioural traits. Eighteen 15-month old Murray cod (Maccullochella peelii) received three tests in order to determine bold/shy behavioural traits then underwent a typical go/no-go judgement bias (JB) test. The JB apparatus had five adjacent chambers with access provided by a sliding door and fish underwent a training procedure to enter a chamber at one end of the apparatus to receive a food reward but were chased using a net if they entered the chamber at the opposite end. Only one third (N = 6) of fish successfully completed the training procedure (trained fish), and the remaining 12 fish failed to reach the learning criterion (untrained fish). Trained fish housed with a larger aggressive Murray cod for 24 h were significantly less likely to enter intermediate chambers during probe tests compared to control fish, demonstrating a pessimistic response. Trained fish showed "bolder" responses in emergence and conspecific inspection tests than untrained fish, suggesting that shyer individuals were less able to apply a learned behaviour in a novel environment. Our limited sample was biased towards bold individuals but supports the hypothesis that losing an aggressive encounter leads to pessimistic decision-making.

Fig 1

Plan of the aquariums used for (a) emergence, (b) exploration, (c) conspecific inspection and (d) judgement bias testing. Chambers in the JB testing aquarium are referred to as positive, POS; intermediate positive, IP; intermediate, I; intermediate negative, IN and negative, NEG.

Fig 2

Visual representation of the judgement bias set-up for various tests with Murray Cod: (a) discrimination tests (only positive, POS, and negative, NEG, chambers were opened); (b) probe tests paradigm (in addition to the POS and NEG chambers, IP, intermediate positive; I, intermediate; and IN, intermediate negative chambers were also opened) and (c) Mean predicted probability of fish entering each chamber in probe tests. Aggression treatment fish (white squares, N = 3) were less likely to enter the three intermediate chambers than control fish (black squares, N = 3). 95% Wald confidence intervals are shown.

Fig 3

(a) Time taken to emerge from the isolation chamber in the emergence test and (b) time spent in the near zone in the social inspection test by trained and untrained fish. Different letters indicate statistical significance at P<0.05. Range, interquartile range and medians are shown.