CARBONATE ISLAND KARST AND THE QUATERNARY PALEOCLIMATE REOCRD

The oceanic setting of carbonate islands creates a karst record of Quaternary climatic change, primarily through sea-level change. Sea-level highstands are also indicated by carbonate sediment depositional cycles. Fossil coral reefs in the tectonically-stable Bahamas define the last interglacial highstand (oxygen isotope substage 5e) as approximately 12,000 years in duration by U/Th dates (131,000 to 119,000 years ago). Flank margin caves develop at the edge of the fresh-water lens, thus their magnitude and lateral extent are a measure of the duration of sea level position, and fresh-water lens position, at the time of cave formation. Sea-level lowstands result in speleothem growth in caves drained by sea-level fall. Bahamian flank margin caves formed during the last interglacial sea-level highstand time window of 12,000 years, proven by the lack of stalagmites older than 100,000 years. Blue holes, with depths over 100 m below modern sea level, are older and contain stalagmites in >350,000 years of age, indicating numerous marine fill and drain cycles caused by glacioeustasy. On tectonically uplifted carbonate islands, caves and speleothems avoid fresh or marine phreatic water overprints, and may detail aspects of climatic change by stable isotope variation. Paleomagnetic records of cave deposits provide minimum ages for caves, and show that some flank margin caves such as Cueva del Aleman on uplifted Isla de Mona, Puerto Rico, developed ~2 million years ago, before the onset of rapid glacioeustasy in the early Pleistocene. Paleosols entomb paleokarsts, and through paleomagnetic measurements of secular variation, can provide an independent chronology of karst formation. Carbonate dissolution/deposition cycles affect CO₂ flux (Ca⁺⁺ + 2HCO₃^- « CaCO₃ + H₂O + CO₂). Karst denudation of exposed carbonate platforms during glacioeustatic sea-level lowstands results in CO₂ sequestering of potentially 1.1 x 10¹³ g/y carbon, while carbonate deposition during highstands results in a potential CO₂ release of 1.2 x 10¹⁴ g/y carbon. Quaternary sea-level lowstands sufficient to expose carbonate banks are about ten times longer than highstands that flood the banks (100,000 years versus 10,000 years), creating a carbonate flux that is approximately balanced (1.2 x 10¹⁸ g carbon versus 1.1 x 10¹⁸ g carbon, respectively).