Forward and inverse methods for extracting climate and diet information from stable isotope profiles in proboscidean molars

Intratooth stable isotope profiles in enamel provide time series of dietary and environmental information that if correctly interpreted, serve as archives of seasonal variability in past environments. A major challenge in interpreting these profiles arises from time averaging imparted by enamel mineralization and developmental geometry, whereby the primary (delta C-13 or delta O-18) input signal is attenuated and shifted, which can potentially lead to incorrect interpretations of the magnitude or frequency of seasonal variability. Several forward and inverse models have been developed to reconstruct the primary input signal from intratooth profiles in continuously growing teeth. Here the models developed by Passey and Cerling (2002) and Passey et al. (2005) are extended to molars of Elephantinae, which grow over a long but finite interval of time. Proboscidean molars are particularly attractive for intratooth profiles because they may contain a decade or more of information and they are often well preserved in the fossil record because of their thick enamel and large size. Forward model parameters are established using histological analysis of molar thin sections of extant African elephants (Loxodonta africana) and a mammoth (Mammuthus columbi) and by micro-CT analysis of L. africana molar plates. The density of immature enamel is about 65% of the final density of mature enamel. The appositional length varies from approximately 35 to 55 mm, and the maturation length is about 70 mm. Histological methods are used to determine crown formation time (CFT) in elephant and mammoth molar plates. CFT for the elephant and mammoth molar plates studied in thin section are about 5-6 years and 11 years, which translate to mean growth rates of about 21 mm/year and 16 mm/year, respectively. Coeval molar and tusk profiles from a zoo elephant are compared. The tusk isotope profile serves as a proxy for the primary input signal, and thus provides an opportunity to evaluate the forward and inverse models. The results from the zoo elephant profiles demonstrate that the inverse model accurately reconstructs the amplitude and overall structure of the primary input signal. Inverse model results of mammoth molar profiles show double the range of delta C-13 in measured enamel profiles. Inversion model results illustrate that improved reconstruction of the primary input signal can lead to more accurate interpretations of the seasonal variability of diet and body water and by extension, vegetation and precipitation in past environments.