Central American Climates of the Last Interglacial

This project provides centennial-scale multiproxy analyses for Termination II, and the last interglacial (MIS 5e, 135,000-106,000 years ago) from three recently collected sediment cores from Mexico, Guatemala and Panama. These records form the first detailed terrestrial archive of climate and vegetation change for the last interglacial from the Neotropics. The analyses include XRF, core color, and fossil pollen, charcoal and diatom analyses. A chronology for each site is established through a combination of tephrachronology and U-series dating.

Preliminary U-series dates from the Panamanian core confirms almost 5 m of deposition between ages of 130 and 115 ka. The other records appear to have similar or faster rates of deposition. Bayesian statistics are used to create transfer functions for temperature, precipitation, and number of dry days per year. These data are used to test hypotheses relating to the past migration of the inter-tropical convergence zone, varying strength of the meridional overturning circulation, the presence of quasi-cyclic climatic rhythms, and the variability of the last interglacial relative to the Holocene.

This provides the first regional estimate of whether the past interglacial was a time of forest or scrubland expansion in Central America, a finding that is of critical importance to modeling past and future carbon storage. This data provides important benchmarks against which paleoclimate models can be tested. Currently, no detailed terrestrial data are available for the last interglacial from Central America or Mexico. This data is significant to those engaged in planning for a warmer world, whether for agriculture, forestry or conservation.

The provision of highly resolved climate records from Central America allows hypotheses generated in other records to be tested. These data are also significant to ecologists studying modern forest dynamics and to evolutionary biologists investigating the origins of tropical diversity. Providing well-resolved chronologies for these cores encourages other workers using different proxies to work on them. This research leads to a significant number of new collaborations in Mexico, the USA, Germany, Costa Rica and the UK. During this project two undergraduates are trained each year, and REU and RET personnel are encouraged to publish their results. One Ph.D. student and one post-doc are supported. A female MS student in Costa Rica is engaged in the project without charge to NSF.