CENOZOIC VEGETATION CHANGES IN THE NORTHERN ROCKY MOUNTAINS: ADDING THE PHYTOLITH RECORD

Based on macrofossils and palynofloras, Cenozoic plant community evolution in the Northern Rocky Mountains has traditionally been reconstructed as a progressive shift from subtropical evergreen or semi-deciduous forests in the early Eocene to drier and more montane habitats in the late Eocene. Middle-late Eocene floras in the central Rocky Mountains have been thought to reflect open savanna woodlands which, judging by the scarcity of grass fossils, consisted exclusively of small trees and shrubs. During the Oligocene, this vegetation type presumably spread, culminating with the development of conifer forests and steppe vegetation in the middle to late Miocene. In contrast, palynofloras from south-western Montana have pointed to the presence of grass-dominated habitats already in the early Oligocene. Thus, the role of grasses in Cenozoic Northern Rocky Mountain ecosystems has been incompletely known; in particular, the timing for the spread of steppe elements such as dry-adapted, open-habitat grasses was uncertain.

A recently described record of plant silica (phytolith) assemblages helps shed light on vegetation change in the Northern Rocky Mountains. The majority of the Eocene and Oligocene phytolith assemblages are suggestive of closed forests with abundant dicotyledons, relatively rare palms, and little grass. Both typical closed-habitat grasses (e.g., bamboos) and open-habitat grasses were very uncommon. On the other hand, phytoliths of unknown grasses are frequent or dominant in a few cases, potentially supporting the suggestion that savanna woodlands existed in Montana as early as the late Eocene. However, because the autecology of these grasses is presently unclear, this hypothesis remains untested. In the early Miocene, open-habitat grasses expanded, resulting in grass-dominated savanna woodlands, similar to contemporaneous vegetation types in the Central Great Plains.

The study also points to differences between interpretation of Cenozoic vegetation based on, respectively, phytoliths and other paleobotanical evidence. For example, previous work has failed to show the persistence of palms in the Northern Rocky Mountains through the middle Miocene. Conversely, the phytolith record does not clearly reflect the strong coniferous element evident in macrofossil and palynofloral data.