Participating Students - Anna Williams

Anna Williams
University of Michigan
Enrollment Year: 2003
Email Address: atwillia@umich.edu
Atmospheric Mentor: Mary Anne Carroll, University of Michigan
Biosphere Mentor: Milford Wolpoff, University of Michigan

Research Topics
d 18O Use as a Dietary Indicator This work is a portion of a larger project to test the hypothesis that two genera of hominid which coexisted ~2.5 Mya can be differentiated on the basis of diet. d18O analysis should discriminate between these diets if we compare specimens of each genus. However, both atmospheric and environmental components create regional and temporal variance in the d18O signal. Models and theoretical equations have suggested that atmospheric variables alone may induce as much as a 30% difference in d18O values, a difference in itself greater than dietary differences alone. The specimens we eventually want to characterize on the basis of diet are both spatially and temporally distributed; therefore we need to calibrate this potentially valuable tool for more accurate interpretation. A realistic assessment of these variables and their interactions is needed in order to develop mathematical models to assess whether dietary reconstructions using d18O can be accomplished. In order to achieve this, I will design an experiment to test the degree of influence of these external variables on the isotopic composition of an organism. The purpose of this study is to quantify the actual variation as we manipulate the atmospheric and environmental variables believed to influence the fractionation factor of oxygen, and then to examine trophic level and physiological fates of the signature. Specifically, the humidity at which plants grow, the temperature at condensation of water used to grow them with, the CO2 concentration of the atmosphere, the functional plant type (C3 vs. C4), and the part of plant used (above vs. belowground portion) will be experimentally manipulated to determine the degree to which d18O changes as a function of each independent variable. These results can then be used to develop a model that predicts variation in d18O values due to climate for organisms that are spatially and regionally distant using experimental data. Based on the output of this model, we can then ascertain the circumstances under which it is feasible to conduct the destructive d18O analysis with fossil hominid specimens. These results may then allow us to discriminate between herbivorous and omnivorous diets, as well as type of plant material processed (i.e. leafy greens vs. tubers). The results have broad implications for our models of human origins. This research may also help to refine paleoclimate variability assessments that use oxygen isotope analyses of terrestrial and non-marine substrates (i.e. otoliths, speleothems, and tooth enamel), where 18O signatures are also impacted by the factors manipulated in this study.

Back to BART Home