AMINO ACID RACEMIZATION KINETICS OF GENYORNIS NEWTONI AND DROMAIUS NOVAEHOLLANDIAE USING ION-EXCHANGE AND REVERSE-PHASE LIQUID CHROMATOGRAPHY

Amino acid racemization (AAR) in eggshells of the large, flightless Australian birds Genyornis newtoni and Dromaius novaehollandiae has been a useful tool in reconstructing the paleoclimate history of Australia and for sample age estimations. Previous work using these eggshells has been restricted to D-alloisoleucine/L-isoleucine (A/I) separated by ion-exchange liquid chromatography (IELC).  New developments in reverse-phase liquid chromatography (RPLC) allow resolution of enantiomers and diastereomers from geological samples. To test the utility of RPLC, Genyornis and Dromaius eggshells were heated at 143ºC and 131.5°C for varying times to induce racemization, and then analyzed on both RPLC and IELC.  D/L ratios were calculated for glutamic acid, valine and A/I on RPLC and A/I on IELC.  A/I from the same samples measured by RPLC and by IELC is indistinguishable (r²=0.992, n=83) suggesting that RPLC provides reliable A/I as well as D/L for other amino acids. The activation energy (Ea) for racemization of glutamic acid in Dromaius eggshells is 28.5 kcal/mol, whereas valine has an Ea of 29.8 kcal/mol.  The Ea for isoleucine epimerization is statistically indistinguishable based on IELC and RPLC data (28.4 ± 0.2 kcal/mol). Previous studies showed inter-specific variations in isoleucine epimerization rate, with the rate constant in Dromaius eggshells 1.16 times that in Genyornis eggshells, a value replicated in this work.  However, glutamic acid racemization is 1.33 times faster in Dromaius, whereas valine is only 0.84 times as fast as in Genyornis.  These variations are attributed to differences in the sequence of amino acids in proteins within the calcite matrix of the two taxa.  Each amino acid was tested for its age estimation reliability for ten ¹⁴C dated Dromaius samples from central Australia. The average error for all three amino acids is auto correlated and averages 43%. The source of this error is based on variations of burial depths rather than analytical uncertainties.  Deeply buried samples have an effective diagenetic temperature (EDT) that reflects the mean annual temperature, whereas shallowly buried samples exposed to high-amplitude temperature fluctuations return an EDT up to 5ºC higher than the arithmetic mean annual temperature.