MATHEMATICAL MODELING OF DIAGENETIC AMINO ACID RACEMIZATION IN CARBONATE FOSSILS

Amino acid racemization in Quaternary fossils is often used as a qualitative geochronological tool. Measurements of D/L values in fossils are usually conducted on the total amino acid hydrolyzate, which includes both free and bound amino acids in the analyzed sample. The abundance of both D- and L-amino acids in fossils is a complex function of polypeptide breakdown and diffusional loss. Known taxonomic differences in overall racemization rates (“apparent kinetics”) indicate the complexity of these reactions.

The diagenetic modeling approach used by Collins and Riley (2000) has been applied to the analysis of quantitative amino acid data from deep-sea foraminifera (Muller, 1984). Data for free and bound D-alloisoleucine and L-isoleucine in two foram taxa ranging from ~0 to ~1000 kyr (Muller, 1984) are among the best constrained (temporally and thermally) results for any carbonate fossils. Muller’s (1984) results are modeled in terms of a three-component system, with each component having its own characteristic hydrolysis rate constant. Initial component abundances and hydrolysis rate constants are determined using least-squares curve-fitting routines, incorporating constraints using the method of Lagrange multipliers. The apparent racemization kinetics of Orbulina and Globoratalia, slow- and fast-racemizing taxa, respectively, are most influenced by significant differences in the abundance of a “refractory” component (64% and 52%, respectively) whose hydrolysis constant also varies by a factor of two (Orbulina < Globorotalia). The models developed here are appropriate for carbonate systems that are more open than fossil eggshells (Miller et al., 2000; Collins and Riley, 2000). Our current models are tested with racemization data from mollusk samples from several Quaternary coastal sequences, including some samples from arid Peruvian sites where closed-system diagenesis is more possible than in humid mid-latitude sites.