In 1978, Oster and Wilson proposed a theory of adaptive caste structure in the social insects. According to this theory, the ratios of the various castes should become optimized during evolution. However, this optimalization would include a plasticity that would enable the colony to alter these ratios when the need arose. Evidence for the plasticity of the colony has come from the studies by Passera and colleagues (1996), who have shown a remarkable type of inducible defense in colonies of Pheidole pallidula.
The work-force of Pheidole pallidula is divided into two distinct castes: small-headed minor workers (who do most of the work of the hive, caring for the broods and gathering food) and big-headed major workers (soldiers) specialized for defending the colony. Passera and coworkers exposed the ants of one colony to ants of another, separating the two colonies by a wire mesh that enabled legs or antennae to pass through, but that prevented the jaws or the whole ant to get across. This contact led to the production of more soldier ants.
This is a new type of inducible defense, but it may not be widespread among ants, even in the genus Pheidole. The "job categories" of Pheidole pallidula are more specific than those of many other Pheidole colonies. In these other colonies, the major workers also mill seeds or act as food vesicles. Moreover, in many ant colonies, the removal of the major workers merely causes the minor workers to assume their roles. Perhaps there is an important temporal component, as well. It takes time to make more soldiers. Perhaps this is a "second" line of defense that can be used if the first one fails. Since neither minor or major workers of P. palludula have functional gonads (reproduction being reserved for the queen), they act similarly to body parts in this inducible defense mechanism.
Passera, L., Roncin, E., Kaufmann, B., and Keller, L. 1996. Increased soldier production in ant colonies exposed to intraspecific competition. Nature 379: 630-631.