A NEW PROXY FOR ESTIMATING LEAF AREA INDEX (LAI) IN THE FOSSIL RECORD

Vegetation structure, a fundamental aspect of ecosystems, controls everything from primary productivity, hydrological cycling, fluxes in energy, carbon reservoirs and atmospheric gases, to soil moisture, fire regime, erosion, and the evolution and composition of faunal communities. Yet, in paleoecological studies, we do not have a way to reliably measure it.

This study presents a proxy to quantify habitat structure by estimating Leaf Area Index (LAI), the area of foliage per area of ground, from the fossil record using phytoliths. It uses the well-known relationship between sunlight and epidermal plant cell morphology whereby cells of leaves growing in shade are larger and more undulated than those growing in sun. Silica filled epidermal cells are casts of the original cells that faithfully preserve cell size and shape, and are often preserved as phytoliths in soils and microfossils in paleosols. We hypothesized that anticlinal epidermal cells and hence phytoliths contained in soils under closed habitats would be larger and more undulated due to diffuse levels of irradiance under a forest canopy. To test this, we collected assemblages of modern phytoliths from soils under various canopy light environments in Costa Rica. From each site, LAI was measured using hemispherical photography. Anticlinal epidermal cells, presumably from dicots, ferns and non-herbaceous monocots were photographed and measured. Consistent with expectations from theory, we found significant correlations between LAI and both mean anticlinal phytolith area and mean undulation. We developed a multiple linear regression model that correlates cell area and cell undulation with LAI.

Using the model created from the modern dataset, we present a record of significant change in LAI linked to climate change during the middle Cenozoic (49-11 Ma) in Patagonia. This new method adds an important untapped tool to paleoecological studies filling a major gap in paleoenvironmental reconstruction. The method has the advantage of being taxon-independent, therefore it can be applied to strata of any age so long as phytoliths are preserved. LAI is an important variable used in ecological studies and in climate modeling. With this new proxy, there is the potential to create global paleo-LAI maps to include vegetation feedbacks in deep-time paleoclimate models.