Researchers Mathieu Lihoreau (Research Centre for Psychology, School of Biological and Chemical Sciences, Queen Mary University of London), Jean-Louis Deneubourg (Centre National de la Recherche Scientifique, Université de Rennes) and Colette Rivault (Service d’Ecologie Sociale, Université Libre de Bruxelles) gave hungry cockroaches (Blattella germanica) a choice between two identical food sources, and found that the animals chose whichever food source had already attracted the most other cockroaches; the effect became more marked in proportion to the number of cockroaches who had already made the more popular choice.

The researchers say that the selection of food sources 'relies uniquely on a retention effect of feeding individuals on newcomers without comparison between available opportunities', and that the behaviour 'shows similarities with the foraging dynamics of eusocial species, thus stressing the generic dimension of collective decision-making mechanisms based on social amplification rules despite fundamental differences in recruitment processes'.

Although the organization of the cockroach groups is simpler than that of eusocial animals, the researchers hypothesise that such parsimony could apply to a wide range of species.

'Eusocial' is the term used to describe species such as ants and bees in which groups are all related to one another; in the case of the cockroaches kin relationships are not a factor (something that is also true of human groups once they had evolved past the kin-group stage).

The researchers point out that the observed behaviour of B. germanica involves short-range communication between the individual cockroaches, rather than the use of pheronomes or other long-range attractants, although they do not speculate on what type of communication this might be.

The importance of collective behaviour in such life forms as cockroaches comes as something of a surprise to scientists: "What we are realising is that 50% to 60% of insects live in groups and we don't know what is happening in these groups," says Lihoreau.

Perhaps the truth of the matter will turn out to be that group behaviour emerged very early in evolution as an aid to fitness, and that thereafter solitary behaviour was the exception rather than the norm, requiring suppression of the inherent 'groupishness' of precursor species. But we don't know, because researchers haven't been looking for group behaviour in primitive life forms; perhaps now they will begin to!