Ants can become an addiction, and this book is a must-read demonstration of this apparently weird fact. Ants are rather simple agents, yet their colonies and even populations show stunning behaviors in terms of coordination, organization, task allocation, reaction to environmental events, survival strategy, co-existance with neighbors, flexibility and robustness (in one, adaptivity) that make them an outstanding case study for how complex socio-ecological behavior emerge from rather basic features of agents.
The author is devoted to the topic, which is apparent from is bio and from the writing style, too - smooth, reflexive, attentive and humorous at times. The book starts with the description of the setting of the field studies and the characteristics of the ants investigated, and goes on rather concisely to face important and general questions as: how is task allocation managed? How can the behavior of a colony change in time if the average life of each ant (except the queen, who just starts the colony and lays eggs for 15-20 years underground) is very much shorter than that of the colony itself? How do colonies coexhist within a population? How do ants reactively switch tasks? What is the role of the interactions pattern (apparently more fundamental than the physical (chemo-tactile) details of the inter-ant communication)?
The author describes a host of field experiments throughout years, both in the desert and in the lab, engineered to try to provide tentative answers to such and more questions. Models of task allocation, both neaural-network based and ecological, are provided. Ignorance is often admitted, which is to be counted as a plus, either.
Some lessons are finally learned (not on populations in general, though), particularly how the behavior of an agent cannot be studied in isolation, since it also depends critically on its interactions with other agents, (homo and/or heterogeneous). The old notion that the traits of the ants were such because of genetics and so little could be done to change them is more an ultimately false excuse to avoid the study of interaction networks than a timeless law.
A fashinating book, and must-read for complexity enthusiasts.