9710099 Fredlund Documenting change in semi-arid grassland ecosystems through time requires better proxy data than can typically be provided by pollen, macrofossils, or other proxy data. Combining quantitative analysis of grass phytolith and carbon isotopes is a practicable approach to documenting change within semi-arid grasslands. The proposed project builds on the work begun under a small ($30k) "Proof of Concept Award" now being completed (NSF-DEB-9510065). In response to reviews of that proposal, efforts were focused on a few specific fossil records. Application of the approach to sequence of late-Pleistocene and Holocene paleosols and sediments at Wind Cave National Park, South Dakota, documents the rapid early-Holocene rise of the mixed grass prairie composition and subsequent climate-driven changes throughout the Holocene (Fredlund and Tieszen, 1997). A second study, focusing on late-Quaternary records in south Central Nebraska and adjacent Kansas (Fredlund and Tieszen, n.d.). This study demonstrates the utility of phytolith analysis in quantitatively reconstructing grassland vegetation and climate in the Great Plains of North America. The modern data presented are the typical (average) assemblages for 34 modern localities taken throughout the Great Plains. Modern assemblage variability across the region is substantial. This variability shows a coherent geographical pattern consistent with modern climate data: specifically July mean temperature. Seven fossil assemblages from three late-Pleistocene localities in the central Great Plains are presented. Two alternative phytolith-temperature regression models are applied to these fossil assemblages to estimate paleo-temperature. Resulting estimates are consistent with other efforts to calculate late Pleistocene temperature departures in the region (ca.-7.0 to -8.0( C). The fossil data also show rapid change in grassland composition and climate across the Pleistocene-Holocene transition consistent with regional hypotheses. These results demonstrate the utility of phytolith analysis as complement to, or substitute for, pollen analysis in the Great plains of North America. The proposed project has two goals: 1) To expand the modern analog data to a geographically broader and more robust set of localities that can be better used to quantitatively estimate grassland composition and climate, 2) the continued development of early Holocene fossil records from the western central Great Plains region with the end goal being a geographical synthesis showing grassland compositional and climate over the Pleistocene-Holocene transition. Methodological developments achieved by this project will be readily transferable to other semi-arid grassland biomes.
