o s d

am ard Exe ing their own environments have received less attention. those associated with nonsocial hosts, because homeostatic OpinionThis is surprising, because it is now 40 years since it was first noted that the interior of a large ant colony represents a radically different environment from that encountered beyond its borders, a concept encapsulated in themetaphor of a colony as a ‘factory constructed inside a fortress ’ [3]. The nests that large insect societies create provide unique, environmentally buffered patches of habitat for many other organisms. The distinctiveness and quality of such patches are a direct function of the size and longevity of the colonies involved. Societies of ants and termites are remarkably long lived (Box 1). They can achieve this long-evity by recruiting new queens at regular intervals, but also the queens (and kings in termites) themselves can have life spans of decades because the colony interior is a predator-free space where selection will tend to reduce colony life will tend to reduce virulence. In the same vein, mutualists of advanced insect societies could fail to achieve maximal productivity. They are predominantly ectosym-bionts [8], which implies that they are prone to attracting their own parasites and likely to maintain independent and sexual reproductive agendas that are costly for the host. We combine these arguments to infer that advanced insect societies are bound to create biodiversity hotspots that are interesting for comparativemacroecological study, because they extend into dry and higher-latitude ecosys-tems that normally lack biodiversity hotspots. Why parasites of long-lived insect societies are expected to be nonvirulent Parasites, by definition, negatively affect hosts and this is termed virulence. Our understanding of virulence evol-Corresponding author: Hughes, D.P