CLIMATE, VEGETATION, AND STABLE ISOTOPES OF MODERN AND SEDIMENTARY N-ALKANES IN SEMI-ARID AND GRASSLAND ENVIRONMENTS OF THE NORTH AMERICAN INTERIOR

The stable carbon and hydrogen isotopic compositions of organic biomarkers are commonly used a proxy for hydrology, ecosystem and other environmental conditions. Lipid biomarkers, such as long carbon chain normal alkanes, are produced in the epicuticular wax layer of terrestrial plant leaves, and are widely preserved in the geological record with minimal isotopic degradation at temperatures below 100 to 150°C. Studies have shown that the δD and δ¹³C of long carbon chain n-alkanes are correlated with average chain length (ACL) of n-alkanes, meteoric δD, atmospheric δ¹³C_CO2, precipitation amount and vegetation type. Despite widespread application of this approach for paleoenvironmental reconstruction, many relationships between carbon and hydrogen isotope fractionation in modern vegetation and environmental parameters such as temperature, precipitation and other moisture-related parameters appear relatively weak. Furthermore, semi-arid and grassland ecosystems are poorly represented in global datasets. Thus, application of biomarker proxies to reconstruct paleoenvironmental change in these systems is problematic. To address this, we sampled soils and four plant functional types (grasses, trees, shrubs and forbs) across a large spatial gradient throughout the western United States to assess relationships between biomarker H and C isotopes and seasonal and annual climatic conditions within a semi-arid environment. These data are coupled with published data from North American sites to produce more robust and quantitative relationships between isotope values and environmental parameters and to build the North American Molecular Isoscape. Our data show predictable and statistically significant relationships between apparent fractionation and moisture-related variables. We see little correlation between precipitation amount and biomarker or bulk carbon isotopes. The nature of these relationships varies with ecosystem type. We also observe that regional differences in ecosystem type strongly bias the isotopic data. This introduces error into quantitative reconstruction of paleoenvironment in scenarios for which past ecosystem is poorly constrained. Uncertainty in quantitative paleoclimate reconstructions can be greatly improved by incorporation of paleoecosystem information.