Sources of carbon dioxide in penguin air sacs.

CO2 tensions in the caudal air sacs of birds cannot be quantitatively predicted by current models of avian respiration, mainly because the contribution of neopulmonic parabronchial gas exchange has not been determined. To overcome this problem we studied penguins that have purely paleopulmonic lungs. Three penguins were anesthetized, intubated, and ventilated at a constant respiratory rate and different tidal volumes (VT). PO2 and PCO2 were measured in arterial blood and end-expired, mixed-expired, interclavicular air sac, and caudal thoracic air sac gas. Interclavicular air sac and end-expired gas had similar compositions. Caudal thoracic air sac gas was intermediate in composition to end-expired and inspired gas, and its PCO2 was 1.5-3.5 times greater than the value predicted from reinhaled dead space. This difference between measured and predicted caudal thoracic air sac PCO2 increased with VT but showed no relationship to changes in dead space-to-VT ratio. The difference is not explained by stratification or diffusive gas exchange across air sac walls. The results can be explained by postulating that inspired gas passes over exchange surfaces on its path to caudal air sacs. This is unexpected in the purely paleopulmonic lungs of penguins and suggests that airflow may not be caudocranial in all paleopulmonic parabronchi.