ÁñÁ«ÊÓƵ

Ants are arguably the world's most successful insect group. There may be as many as 200 ants for every square metre of land. Their nests are found in the most inhospitable of deserts and even in the frozen tundra of the far north. Ants have been part of the terrestrial community for a long time, having diverged from the wasp lineage 60-100 million years ago. During that time, large numbers of plants and animals evolved to coexist with ants, which on their own scale of existence, have had as big an impact on the natural world as people.

How have they achieved this success? Ask anyone anywhere, and it is almost certain that you will be told it is because, like us, ants live together in societies or colonies. The parallels with humanity were recognised by the earliest philosophers. The Bible exhorts us to consider the ways of the ant because with prudent industry she gathers up grain in time of plenty to feed her young in winter. Incidentally, all queen ants and their sterile workers are female; males are merely ephemeral fly-like creatures produced in certain seasons solely for the task of fertilising new queens.

For the past 400 years naturalists have unravelled the intricacies of ant sociality, sometimes in the hope of drawing lessons for human society but more often just from the sheer curiosity engendered by these insects. They have uncovered a bewildering array of alternative life styles analogous to those of humans. Many species have small colonies, which, like primitive tribes, depend on the individual prowess of hunter-scavengers. Other larger colonies are formed by species evolved to be farmers, some husband aphids and bugs like cattle, others grow crops of nutrient-rich fungi on compost made from freshly cut leaves.

The lifestyle of the tropical army ants is the most awesome. Like the marauding hordes of Ghengis Khan, a million or more fearsome workers descend on a territory, pillage it and move on. Despite the scare stories of them devouring people and cattle, it is actually social insects, particularly other ants, that suffer most from their raids. Like old city states, they are overrun by the hoard and left devastated and abandoned. It may come as a surprise to some to discover that this devastation actually benefits the long-term diversity of tropical ecosystems. This is just one of the many titbits of army ant biology recorded by William Gotwald Jr, an American who has studied these amazing creatures for more than 25 years. Gotwald summarises a lifetime's research in his new book, Army Ants: The Biology of Social Predation.

The book is well illustrated with many figures and more than 40 excellent colour photographs. It makes an entertaining read for social insect specialists, but it is particularly aimed at naturalists who are fascinated either by complex and highly evolved life styles or by tropical ecosystems. Gotwald first reviews army ant distribution and then describes how colonies undergo cycles of growth and reproduction, and how the nomadic marauding life-style relates to this. He shows how morphology and behaviour, particularly the response to chemical messages, is important to social cohesion. Most of the second half of the book is concerned with the secondary biodiversity generated by army ant sociality; they support an enormous number of "guests" - symbionts and parasites - as well as specialist predators. This section will be of great interest to natural historians.

Early in the history of writings on ants, great play was made of their selfless, altruistic behaviour. They were often credited with individual intelligence that they used to "decide" to help each other. While this notion was particularly attractive to some moralists it was strongly resisted by others, such as the Rev Wasmann (Psychology of Ants and of Higher Animals, c.1900) who considered ants "the climax in development of instinctive life throughout the animal kingdom" and appealed, "Do away with all books, pamphlets and periodicals, whose only purpose is to raise brute to the level of man". In other words, he believed that while man's social behaviour is motivated by higher ideals, that of ants is fixed by inheritance.

However, inherited altruism posed great difficulties for believers in Darwinian evolution. If an individual abandons its chance to reproduce in favour of helping another (that is, it becomes a worker), how can this be an inherited trait? Altruists should always become extinct. Darwin himself recognised this problem and suggested that sterile workers might evolve if the community or colony benefited.

It became popular to call an ant colony a superorganism. The queen was seen as combining the roles of gonad and brain while groups of workers behave like specialist tissue and organs; each bit striving for the good of the whole. In the preface to his book, Gotwald reminds us that the superorganism concept fell into disrepute among scientists as the developing science of genetics highlighted its inadequacies and contradictions. Yet he speaks for many field ecologists when he confesses to a temptation to compare army ant colonies with higher-order predators.

He can relax. Apparently the superorganism is making something of a comeback in modern scientific thinking. Andrew Bourke and Nigel Franks (the second of whose work is freely cited by Gotwald) in their new book Social Evolution in Ants conclude: "The superorganism cannot be dismissed from the study of social insects. It may at times be desirable. But it is also not antithetical to gene selection."

Bourke and Franks are to be congratulated on producing a remarkably clear synthesis of the complex ideas that surround the evolution of sociality. I cannot praise their book too highly. For some years now, I have struggled to maintain a grasp on the theoretical debate as it fragmented around topics such as kin conflict and multiple mating. Bourke and Franks have now drawn all these together in a single logical sequence. The book may not tell theoreticians anything new but it will be indispensable for field ecologists such as Gotwald and myself.

The book is organised into 12 self-contained chapters. The first three are devoted to the underlying theory. Starting from the idea of kin selection, which has dominated thinking on social insects since it was formulated by William Hamilton in 1963, the authors debate the levels at which selection operates. Full discussion is given to the role of haplo-diploid reproduction in the evolution of sociality in ants: in the ants, bees and wasps, unfertilised eggs become males whereas fertilised eggs become females. This results in full sisters being related by a factor of three-quarters rather than by the half that is normal in diplo-diploid systems such as our own.

The next chapters are devoted to consideration of the specialist topics of sex ratio theory, kin conflict and polygyny (multiple queens in colonies). The last four consider life-histories, mating biology and evolution of division of labour. Complex theories that underpin the text but are not directly relevant to the argument are considered separately in a series of self-contained "boxed" topics.

Bourke and Franks are an impressive combination. They leave me in no doubt that all the curious ramifications of ant social behaviour that I have observed can be explained by natural selection acting on genes.

While Gotwald's book will be at home on any naturalist's bookshelf, or in any senior school or college library, Bourke and Franks's will have more specialist academic appeal. Social Evolution in Ants will surely become essential reading for any student interested in the evolution of insect sociality and will be a standard reference source for teachers and researchers alike.

Graham W. Elmes is a Natural Environment Research Council senior scientist attached to the Institute of Terrestrial Ecology, Furzebrook, Dorset.