What do fish make of mirror images?

Julie K. Desjardins and Russell D. Fernald. 2010. Biology Letters  6: 744-747.

As early as the 1930’s researchers knew the highly territorial male Siamese fighting fish (Betta splendens) would respond aggressively and attempt to attack a mirror image of itself as if confronting a real intruder (Lissmann 1932). Tinbergen (1951) described this aggressive reaction to mirror images in another territorial fish species, the three-spined stickleback.  As with most vertebrate species, fish do not appear to exhibit mirror self-recognition (the ability to recognize an image in a mirror as themselves). Only humans, apes and (possibly) elephants, dolphins and magpies have so far been shown to be capable of self-recognition.

Consequently, researchers have long used mirrors and a fish's aggressive response to them as a proxy for fish behavior when confronted with a live intruder. This paper calls into question that assumption by showing that the neurophysiological response to a mirror image may not be identical to the response to a live opponent.

Desjardins and Fernald used a territorial African cichlid species (Astatotilapia burtoni) as their subject. Male fish (known to exhibit dominant behavior and thus, likely to aggressively defend territory) were individually confined to one half of an 8 gallon tank (divided into two compartments separated by a clear, watertight barrier). Each fish was then allowed to view either a size-matched, male Astatotilapia or a mirror image of itself (both located behind the clear barrier). A third group of fish not exposed to any stimulus was used as a control.

Each fish was observed for twenty minutes and scored for typical cichlid aggressive behavior such as biting, ramming and side body displays. After the behavioral trial, each fish was euthanized and blood collected to determine plasma levels of the hormones testosterone and 11-ketotestosterone. Brain tissue was collected from four brain regions thought to be important in the control of specific behavioral responses in fish.  These included the:

• Dorsomedial telencephalon (Dm) – a fish homologue of the amygdala (known to play a role in responses to fearful situations).
• Dorsolateral telencephalon (Dl) – the fish homologue to the hippocampus (plays a role in control of spatial learning).
• Pre-optic area (POA) – involved in reproduction.
• Cerebellum (Cce) – involved in attention.

Desjardins and Fernald measured the expression levels of two immediate-early genes (IEG), c-fos and egr-1, found in these brain regions. IEG expression in the neurons of resting animals is barely detectable but increases dramatically with the activation of certain neuronal activity patterns, including those associated with learning, memory formation, fear, stress and mating behaviors (Guzowski et al. 2010). Because IEG’s are evolutionarily conserved, they may be used as a genomic marker for neuronal activity in most invertebrate and vertebrate species. Increases in IEG expression in the brain regions examined in this study were used as an indicator of neuronal activity triggered by exposure to the specific stimulus each fish confronted.

Exposure to both the mirror and live opponent resulted in increased expression of IEGs in the brain (compared to control fish) but each stimulus type resulted in different effects in the Dm and Dl. Fish exposed to mirror images of themselves showed increased egr-1 and c-fos expression in the Dm compared to live opponent and control groups but only egr-1 was increased in the Dl in the mirror group. In contrast, the males exposed to a live opponent had higher levels of c-fos in the Dl than the mirror and control groups. Both the mirror and live opponent groups had higher egr-1 and c-fos levels in the POA than controls. Neither group was different from controls in expression of egr-1 and c-fos in the Cce. Interestingly, no differences in levels of aggressive behavior or plasma hormone were found between male fish exposed to either the mirror or a live opponent, although these groups were both higher than control fish.

So, why do IEG levels in the Dm (amygdala) and Dl (hippocampus) differ between two groups of fish showing the same behavioral and hormonal responses when exposed to the two different stimuli? Desjardins and Fernald conclude that the fish, while not recognizing themselves, may indeed recognize that the mirror image is not the same as a live opponent. The authors speculate that the mirror image may induce a fearful response in the fish, although exactly why is not known, that results in increased activation of the neuronal circuits in the Dm typical of that seen in fearful situations. Further research needs to be done to correctly tease out the meaning behind these results but the authors caution that a response toward a mirror may not be identical to a response to a live opponent.

References:

• Lissmann HW. 1932. Die Umwelt des Kampffisches (Betta splendens Regan). Zeitschrift fur Vergleichende Physiologie 18: 62-111.
• Tinbergen N. 1951. The study of instinct. London: Oxford University Press.
• Guzowski JF, Timlin JA, Roysam B, McNaughton BL, Worley PF, Barnes CA. 2005. Mapping behaviorally relevant neural circuits with immediate-early gene expression. Curr Opin Neurobiol Oct; 15(5): 599-606.