TRACKING THE SPREAD OF ANCIENT GRASSLANDS: A COMBINED PHYTOLITH-CARBON ISOTOPE PALEOVEGETATION RECONSTRUCTION FROM THE LATE MIOCENE OF SOUTHWESTERN MONTANA

Although the grassland biome is relatively young, having evolved in the early Neogene, it is critical in today’s world, providing important ecosystem services such as habitat and the majority of foods for humans. However, the exact nature of the origin and spread of C₃ and C₄ grasslands on different continents is still unclear, making studies combining paleobotanical and paleoenvironmental data necessary for understanding drivers and timing of grassland evolution. Here, we present results from the Madison Buffalo Jump locality in southwestern Montana for which we reconstructed paleovegetation using phytolith (plant opal bodies) and carbon isotopic data from paleosol organic matter. The samples are bracketed by 10.4 Ma and 9.6 Ma ashes. Previous studies using paleosol geochemistry reconstructed paleoclimate as having a mean annual precipitation of 855 mm yr^-1 and mean annual temperature of ~10°C. Preliminary phytolith analysis shows 30-65% grass phytoliths, indicating more or less grass-dominated vegetation. Grass communities appeared to have been dominated by C₃ pooids. There is small representation of tropical plants such as palms. The mean of the carbon isotope values is -24.06‰, which is consistent with pure a C₃ ecosystem. However, a few individual samples had values as heavy as -14.76‰, suggesting at least localized patches of C₄ vegetation. Both the phytolith and isotope data are consistent with previous studies that showed open-habitat grasses dominating in Montana by 17 Ma, but indicate a higher proportion of forest than other studies. These data support a scenario where the transition to grasslands involved forest-grassland mosaics, rather than a uniform opening of the landscape.