Predicting animal δ¹⁸O: Accounting for diet and physiological adaptation

Theoretical predictions and measured isotope variations indicate that diet and physiological adaptation have a significant impact on animals δ¹⁸O and cannot be ignored. A generalized model is therefore developed for the prediction of animal body water and phosphate δ¹⁸O to incorporate these factors quantitatively. Application of the model reproduces most published compositions and compositional trends for mammals and birds. A moderate dependence of animal δ¹⁸O on humidity is predicted for drought-tolerant animals, and the correlation between humidity and North American deer bone composition as corrected for local meteoric water is predicted within the scatter of the data. In contrast to an observed strong correlation between kangaroo δ¹⁸O and humidity (Δδ¹⁸O/Δh ∼ 2.5 ± 0.4‰/ 10% r.h.), the predicted humidity dependence is only 1.3 -1.7‰/10% r.h., and it is inferred that drinking water in hot dry areas of Australia is enriched in ¹⁸O over rainwater. Differences in physiology and water turnover readily explain the observed differences in δ¹⁸O for several herbivore genera in East Africa, excepting antelopes. Antelope models are more sensitive to biological fractionations, and adjustments to the flux of transcutaneous water vapor within experimentally measured ranges allows their δ¹⁸O values to be matched. Models of the seasonal changes of forage composition for two regions with dissimilar climates show that significant seasonal variations in animal isotope composition are expected, and that animals with different physiologies and diets track climate differently. Analysis of different genera with disparate sensitivities to surface water and humidity will allow the most accurate quantification of past climate changes.