Discovering C3 and C4 Photosynthesis Evolution Using Panicoideae and Aristidoideae Phytoliths

In the 400 million years since they first emerged on land, plants have acquired numerous adaptations to the terrestrial environment. One of these adaptations is modification of the photosynthetic pathway, from C3 to C4. The C3 pathway, the ancestral form of photosynthesis found in most plants, has limited efficiency under high temperatures and light intensities, whereas the C4 pathway offers improved productivity. C4 photosynthesis has evolved numerous times in flower- ing plants in the last 66 million years, with C4 grasses being the most diverse and ecologically dominant. Despite their current importance, we still do not understand exactly when and where the evolution of C4 photosynthesis in grasses oc- curred because the fossil record of grasses is sparse. Phytoliths (hardened silica structures precipitated within plant cells) offer a novel tool for tracking C4 evolution. We will use phytoliths as a comparative tool to examine morphological changes associated with C4 evolution. All C4 grasses are contained within the PACMAD (Panicoideae, Arundinoideae, Chloridoideae, Micrairoideae, Aristidoideae, and Danthonioideae) clade of grasses. Panicoideae, one of the largest and most diverse subfamilies within PACMAD, con- taining both C3 and C4 species, is an ideal group to study the ecological and evolutionary factors that drive the distribution of C4 photosynthesis. Through analysis of phytolith morphology, as well as overall density and distribution of phytoliths within leaf tissue using leaf clearings, we will examine a broad sampling of Panicoideae, looking for common trends amongst C4 photosynthesizing groups as compared to their C3 counterparts. To extend the scope of these conclusions beyond Panicoideae, we will seek to confirm these trends by comparison to Aristidoideae, another clade that evolved C4 photosynthesis. Preliminary data indicate that the majority of phytolith morphotypes will be bilobates, crenates, and rondels. We expect these results to correspond between observed leaf clearings and the 3D models.