In contrast, studies testing the environmental intelligence hypothesis focused on diet-related aspects such as home range size, general habitat complexity, and extractive foraging. Comparative studies on the social brain hypothesis (SBH) identified either group size, mating systems, or social bonds as proxies for the selective social pressure. Social and environmental complexity emerged as two competing hypotheses to explain how increased cognitive demands drive brain size evolution. In addition, the artificial selection on brain enlargement is known to result in higher cognitive abilities in laboratory tests, such as discrimination learning in mice, and reversal-learning and numerical abilities in female guppies. There is substantial variation in brain size between vertebrate species, with absolute and/or relatively larger brains possessing ‘better’ cognitive abilities. Selective enlargement of brain parts, that is mosaic brain adjustment, appears to be driven by population density in cleaner fish. In contrast, the size of the mesencephalon, rhombencephalon, and brain stem, assumed more basal in function, were independent of local fish densities. Here, we report a positive relationship between the local density of cleaners, as a proxy of both intra- and interspecific sociality, and the size of the cleaner's brain parts suggested to be associated with cognitive functions, such as the diencephalon and telencephalon (that together form the forebrain). A suitable study model is the mutualist ‘cleaner’ fish Labroides dimidiatus, a species that removes ectoparasites from a variety of ‘client’ fishes in iterative social interactions.

While previous work has focused on comparisons between species, how social complexity affects plasticity in brain morphology at the intraspecific level remains mostly unexplored. The ‘social brain hypothesis' proposes a causal link between social complexity and either brain size or the size of key brain parts known to be involved in cognitive processing and decision-making.

Osirix lite 9.5.2