REFINING THE TIMING FOR THE MID-CENOZOIC GRASSLAND EXPANSION IN THE CONTINENTAL INTERIOR OF THE U.S (Invited Presentation)

The spread of grass-dominated plant communities during the latter half of the Cenozoic constituted a major restructuring of animal habitats and food resources, and significantly altered climate-vegetation relationships and silica cycling. It has been generally assumed that the driving forces for this ecological change was post-Eocene cooling and drying at high latitudes. However, contradictory (e.g., paleosol morphological data vs. phytolith assemblage data) or temporally poorly constrained records of when this change happened have prevented rigorous tests of potential causal factors. A rich record of phytolith assemblages from Nebraska, Wyoming, and Colorado indicate that open, grass-dominated vegetation spread in the central Great Plains during the late Oligocene-early Miocene and new phytolith data from the recently re-dated Railroad Canyon section in Idaho now suggest that grassland habitats were present in the northern Rocky Mountains from at least the early Miocene (~23 Ma), a time-frame that is more similar to the central Great Plains than previously thought.

We attempt to further constrain the timing for the expansion of grasslands in the northern Rocky Mountains by studying phytolith assemblages from two sections in Montana that have recently been dated to ~29.9-27.1 Ma (Cabbage Patch; CP) and ~26.8 Ma (Everson Creek; EC). Assemblages from CP are, with few exceptions, dominated by forest indicator phytoliths; grass phytolith assemblages consist mainly of morphotypes of an unknown grass that previous work suggested may have thrived in open, wet habitats. In contrast, the EC assemblages are characterized by relatively abundant grasses (~20-60% of phytolith assemblages as a whole), and grass communities that were dominated by C₃ pooid grasses, but also contained rare C₃/C₄ PACMAD and closed-habitat grasses. These results suggest that modern open-habitat grasses had started to become important in the northern Rocky Mountains in the late Oligocene, by 26.8 Ma, during a period of globally cooler climates and relatively low atmospheric CO₂. However, it cannot be ruled out that the observed pattern also reflects the latitudinal difference between CP and EC, or endemism in distinct depositional basins as a result of tectonic activity in the northern Rocky Mountains.