Respiratory metabolism of different developmental stages (larvae, pseudergates, nymphs, soldiers, neotenic reproductives; 0.6-4.5 mg body mass) of Prorhinotermes simplex was individually monitored by scanning respirographic method sensitive to subnanoliter amounts of O(2) consumption or CO(2) output per minute. Specimens exposed to dry air after removal from the colony performed enormously large, discontinuous bursts of CO(2) lasting usually 2 min. The volume of CO(2) produced during the burst often surpassed the volume of the whole body and it was 10- to 20-fold in excess of the air-filled endogenous tracheal volume. The initial velocity of CO(2) production during the burst was more than 90-fold faster in comparison to O(2) consumption. In the presence of enough moisture within the respiratory vessel, the termites breathed continuously without any larger outburst of CO(2). This fact fully corroborates validity of the so-called water retention theory in discontinuous CO(2) release. The highest rates of O(2) consumption were found in the second instar larvae (0.9 mg, 1000-2000 microl O(2)/g/h), the soldier caste was intermediate (700 microl O(2)/g/h) while pseudergates and neotenic reproductives consumed between 300 and 600 microl O(2)/g/h, at 25 degrees C. All developmental stages feeding on a cellulose diet had CO(2)/O(2) values (RQ) over 1 (1.2-1.4, i.e. carbohydrate metabolism), pigmented soldiers fed by the workers had RQ around 0.75 (predominating lipid or protein metabolism). The unusually large, sudden eruptions of CO(2) in specimens exposed to dry air allow us to make the following conclusions: (1) the bursts were due to special chemical processes, such as by enzymatic hydration of carbonic acid by carbonic anhydrase and; (2) the bulk of chemically evolved gaseous CO(2) escaped from the body by a mass flow supported by active ventilation, not by a passive diffusion. These results demonstrated that the periodic emissions of CO(2) and the associated homeostatic regulation of the respiratory acidaemia were under perfect physiological control. The termites could thus actively select the type of CO(2) release best suited to the extant environmental or internal physiological conditions, i.e. from a completely continuous respiration to occasionally cyclic or completely discontinuous CO(2) release.