Paper No. 4
Presentation Time: 2:30 PM

SEEING THE FOREST THROUGH THE TREES: RECONSTRUCTING VEGETATION STRUCTURE FROM THE CELLULAR LEVEL USING PHYTOLITHS


DUNN, Regan1, STRÖMBERG, Caroline A.E.2, LE, Thien-Y3 and LOESER, Aidan3, (1)Department of Biology, University of Washington, Kincaid Hall 24, Seattle, WA 98195, (2)Department of Biology & Burke Museum, University of Washington, Box 351800, 24 Kincaid Hall, Seattle, WA 98195-1800, (3)Department of Biology, University of Washington, Box 351800, Kincaid Hall 24, Seattle, WA 98195, dunnr@uw.edu

In many cases, traditional paleobotanical studies using fossil leaves or pollen lack the ability to accurately reconstruct the structure of ancient vegetation. Phytoliths extracted from paleosols represent highly localized fossil plant assemblages and have been used to reconstruct habitats in studies involving grassland evolution. However, the existing method for reconstructing vegetative structure and habitat type (open vs. closed) using phytoliths depends on taxonomic occurrences of presumed open-habitat grass types. This method may lack resolution to reliably reconstruct vegetation when open habitat grass phytoliths are absent, and it depends on the assumption that open-habitat grasses have always been so. Therefore, a taxon-free, morphological approach may be more informative when considering fossil phytolith assemblages. This study presents a novel proxy for estimating Leaf Area Index (LAI), a measure of habitat openness, from the fossil record. It uses the known relationship between sunlight and epidermal plant cell shape, whereby leaves growing in shade have more undulated, or wavy cells than sun leaves. Two experiments were conducted: 1) a greenhouse study where 5 species of grass were grown under differing light quantities; and 2) a study of phytolith assemblages extracted from modern soils collected from variable light environments in Costa Rica (savanna, shrubland, deciduous tropical forest, flooded savanna and multi-tiered rainforest). Results from the first experiment indicate that with the exception of Chusquea (bamboo), epidermal long cells in grasses are more undulated when grown in shaded environments. In the second experiment, Leaf Area Index (LAI) for each soil collection was estimated using hemispherical photography. Phytoliths were extracted from the soils using standard methods and silicified epidermal cells from all taxa (including those from ferns and angiosperms) were photographed and traced. Phytolith undulation was standardized using the Phytolith Undulation Index (PUI). Preliminary results from this experiment indicate a highly positive correlation between cell undulation and LAI where undulation increases with increasing LAI. This finding potentially adds a new tool to paleoecological and evolutionary studies where data about habitat structure are desired.