Modeling and interpreting triple oxygen isotope variations in vertebrates, with implications for paleoclimate and paleoecology

Abstract

The 18O/16O ratios of biominerals have been widely used for reconstructing ecophysiology and climatic settings of modern and extinct animals. However, the 18O/16O ratios of body water, which largely determine the 18O/16O ratios of biominerals, are influenced by a host of competing factors. Regional climate and local hydrology are dominant controls on water isotopic composition before water is consumed by an animal. Behavioral and physiological factors, modified by local climate, also have a strong influence on body water compositions. The addition of a third isotope, 17O (expressed as Δ’17O) potentially allows for further resolution of these factors. Here we construct a generalized triple oxygen isotope mass balance model based on the 18O model of (Kohn 1996) (Geochim. Cosmochim. Acta 60, 4811–4829) and examine the influence on vertebrate animal body water Δ’17O of numerous climatic, ecological, and isotopic variables. We evaluate the model against new and previously published triple oxygen isotope data from modern and fossil animals. The model predicts that animals from arid environments will have wider ranges and lower minimum values of body water Δ’17O than animals living in humid environments. Leaf water consumers are more sensitive to variations in relative humidity and have lower Δ’17O than surface water consumers, which more closely track meteoric water compositions. In this model, factors such as body mass and relative proportions of evaporative versus nonevaporative effluxes from the animal have a lesser influence on animal Δ’17O. If δ18O of meteoric water is invariant, body water isotopic compositions will form approximately linear arrays in Δ’17O versus δ18O space with slopes of ∼0.52. Study of Δ’17O becomes most useful when δ18O of meteoric water is variable or unknown (as is generally the case for fossil animals); in this case Δ’17O of body water responds more strongly to changes in relative humidity, evaporated water inputs, and animal water use efficiency. These predictions are generally supported by observations of Δ’17O for modern animals. This agreement suggests that Δ’17O analysis of animal tissues has great potential as a paleo-aridity proxy in continental environments and as a proxy for learning about the ecology of modern and extinct animals.