The equilibrium concentration of oxygen-18 in body water: Implications for the accuracy of the doubly-labelled water technique and a potential new method of measuring RQ in free-living animals

A model is developed which predicts the equilibrium enrichment of ¹⁸O in body water. By setting limiting values to parameters in the model, the maximum and minimum enrichments of ¹⁸O in body water are predicted to be 0·99 and 1·02 × the enrichment of the environmental source of drinking water. Empirical tests of the model were attempted using Long-eared bats (Plecotus auritus) and Pipistrelle bats (Pipistrellus pipistrellus). In Long-eared bats the enrichment of body water relative to the local water pool always fell within the predicted limits. In Pipistrelle bats the difference in body water enrichment between pregnant and lactating bats was qualitatively in accord with the model predictions. The implications of variability in the equilibrium enrichment of ¹⁸O in body water, for the accuracy of the doubly-labelled water technique, are explored. Four levels of accuracy in evaluation of equilibrium enrichment are identified. The error in estimates of carbon dioxide production due to inaccuracy of the equilibrium enrichment rapidly becomes limiting to the potential duration of doubly-labelled water experiments. In small animals with rapid turnover different methods of evaluating equilibrium enrichment have little effect on the potential experimental duration. In large animals, however, the method of measuring equilibrium enrichment has a much greater influence on the potential duration of the measurement. It is suggested all studies on larger animals should take a blood sample, prior to injection of labels, to assess the equilibrium enrichment. Measurement of the ratio of body to environmental enrichment of ¹⁸O may provide an estimate of RQ for some animals.