African cichlids (Astatotilapia burtoni) inhabit the freshwater lakes of east Africa, and live in complex hierarchical societies. This social system is based on two types of males: those with a territory, and those without. Territorial males are coloured bright blue or yellow, and have prominent stripes across their faces and bodies, while non-territorial males have a drab grey appearance. Territorial males defend their territory by producing threatening displays and chasing rivals away. The males regularly engage in aggressive confrontations with each other in order to keep or gain control of a territory and the resources that come with it. The outcome of such battles has a corresponding effect on social status and reproductive success – victory, and the newly-acquired territory elevates the status of the male, so that it can court females and reproduce.
Remarkably, the male’s coloration and reproductive success are tightly regulated by social status, which has a direct effect on the structure and function of the brain. Social mobility produces a change in the size of nerve cells which produce gonadotrophin- releasing hormone (GnRH), and which are involved in the regulation of sexual development. In a male whose status has been elevated by the acquisition of a territory, the GnRH-producing neurons increase to about eight times their previous size, and begin synthesizing more GnRH; the fish becomes more aggressive, and its gonads mature and start to produce sperm. This neuro-hormonal response produces colour changes in the male – a previously subordinate male who successfully challenges another is quickly transformed from dull grey to the bright colours denoting dominance. Conversely, in a male whose status is diminished by the loss of a territory, the GnRH-producing decrease in size and produce less hormone, leading to a loss of bright coloration and a cessation of sperm production.
This beautiful example of the link between brain and behaviour was established very recently by Russell Fernald and his colleagues at Stanford University. In today’s issue of Nature, Fernald’s team shows that A. burtoni males can predict their position within the hierarchy of their group by observing fights between rival males. The experiments, which are a variation of tests used to assess the cognitive abilities of children, were conducted in a purpose-built tank consisting of one large central compartment surrounded by six smaller ones. Individual male cichlids were placed in each compartment and, over 11 consecutive days, fights were staged between pairs of fish in the outer compartments. After observing the fights between the others, the fish in the central compartment were then presented with pairs that had not fought, and their preference for one or the other of the rivals was determined.
It was found that the observing fish always preferred those that were perceived to be lower down in an implied hierarchy. This hierarchy is based on the recall of information about the fighting pairs – which of each pair won and which lost, and the disparity in strength between each of the fish in the pairs. The observers were inferring social dominance relationships and predicting their own status by a cognitive process called transitive inference. This process enables an unknown relationship between two things to be infered from known relationships. For example, if fish A always won the fight against fish B, and fish B always won against C, then the observer can infer that fish A would win a fight against fish C, even though it did not actually observe A and C fighting.
Transitive inference involves the processing of complex social information, such as the recognition of large numbers of individuals within the group and the accurate tracking of multiple relationships between individuals; it therefore requires a high degree of cognitive ability. Rats and non-human primates are able to perform this process, and, recently, researchers at the University of Nebraska’s Center for Avian Cognition demonstrated it in pinyon jays (Gymnorhinus cyanocephalus).
There is disagreement about the mechanism of transitive inference in animals. While some believe that it is evidence for logical reasoning, others suggest that it is a product of associative conditioning. This study provides some evidence that the process is indeed the result of reasoning, because the fish inferred the hierarchies not through direct experience but through observation. The findings also lend weight to the social complexity hypothesis, which states that social complexity is a major driving force in the evolution of intelligence.
Grosenick, L. et al. (2007). Fish can infer social rank by observation alone. Nature 445: 429-432.
Burmeister, S. S. et al. (2005). Rapid behavioral and genomic responses to social opportunity. PLoS Biology 3: e363 doi:10.1371/journal.pbio.0030363.
Paz y Miño, C. G., et al (2005). Pinyon jays use transitive inference to predict social dominance. Nature 430: 778-781.