DETERMINING THE TIMING OF NEOTROPICAL GRASSLAND EXPANSION USING MIOCENE SEDIMENTS FROM THE LLANOS BASIN, COLOMBIA

The environmental changes (temperature, seasonality and amount of precipitation, and increase in fire events) associated with the Late Miocene and Pliocene are thought to be drivers of the asynchronous global expansion and establishment of grassland ecosystems. The terrestrial core Saltarin-1A offers a unique opportunity to test whether the timing of the expansion of Neotropical grasslands fits with this hypothesis. Located in the Llanos Basin of Colombia, the core captures a complete Miocene section. We analyzed 138 samples for palynology and δ¹³C on bulk organic matter. Additionally, we compared the fossil pollen record with modern pollen from Holocene samples and measured δ¹³C isotopes from extant species of plants distributed in the Llanos Basin.

Our results indicate that although the Llanos Basin is currently dominated by C₄ grasses, Miocene sediments capture pollen of predominantly closed vegetation/forest types. This is supported by δ¹³C measurements of bulk organic matter indicating a dominance of C₃ vegetation. Although there is an increase of savannah vegetation types (Poaceae, Asteraceae and Cyperaceae) in the Middle to Late Miocene, our results suggest that the expansion and establishment of a modern savannah occurred after 5 Ma and that open grassland habitat is a relatively recent occurrence in the Llanos Basin. This increase of open vegetation, coincides with several changes of depositional conditions in the Llanos Basin and other Andean basins in northern South America, such as changes in sediment accumulation rate, source area, depositional environments and soil type.

Future work will include an attempt to identify whether fossil Poaceae from Saltarin-1A are C₃ or C₄. We will use a new form of superresolution microscopy (SR-SIM) to compare fine-scale morphology of fossil and modern pollen walls. We will also measure δ¹³C values from individual grass pollen grains. Our study provides a multiproxy analysis for a complete Miocene section of the Llanos Basin, and will generate new paleoecological data relevant to paleoclimatic models for the Neotropics.