Speleothem carbon isotopes in the tropics: a proxy for vegetation and what they reveal about the demise of Homo floresiensis

Most long-term speleothem climate records focus on δ18O (Cruz Jr. et al. 2005, Wang et al. 2005, Meckler et al. 2012, Carolin et al. 2013). In tropical regions, where temperature variability is relatively small, speleothem δ18O serves as a proxy for precipitation amount. On orbital timescales, changes in moisture source, sea level and global ice volume also influence speleothem δ18O. As both δ18O and δ13C are measured simultaneously, a wealth of potential paleoclimatic information from the δ13C records for numerous sites awaits publication. However, δ13C is less well understood as an environmental proxy, and a simple control and explanation may not be forthcoming. We present a new 92,000-year long speleothem δ13C record for Liang Luar cave in Flores, Indonesia. Here, δ13C acts as a proxy for soil CO2 production above the cave, which is strongly linked to vegetative changes. Since vegetation and soil CO2 production are closely linked to a climatic control, changes in speleothem δ13C can be considered an environmental response to climate change. The Liang Luar δ13C record mainly tracks the δ18O, indicating close affiliation between vegetation and climate. Peak vegetation cover and soil activity (lower δ13C) occur in the early Holocene. The Last Glacial Maximum does not have notably higher δ13C than Isotope Stages 3 and 5a. δ13C variability in the glacial is much higher than that in the Holocene. Differences between the δ13C and δ18O records indicate periods of vegetative change without a climatic cause. The largest increase in δ13C (8‰) during the last 90kyr occurred at 68kyr BP. This anomalous δ13C excursion represents a major loss of vegetation around Liang Luar. The last surviving non-human member of the Homo genus: Homo floresiensis, disappeared from the stratigraphic record in nearby Liang Bua cave between 17 and 10kyr BP (Roberts et al. 2009). The cause of the disappearance, (e.g. climate change, volcanic catastrophe or human competition, has not been established. In contrast to the environmental upheaval around 68kyr BP, the period between 17 and 10kyr BP is remarkably stable. With little change in vegetation at this time, we can rule out volcanism or climate change as a likely cause of the extinction. We are left with an intriguing puzzle as to what caused the recent extinction of Homo floresiensis. Cruz Jr. et al. 2005, Nature, v. 434, p. 63 Wang et al. 2005, Science, v. 308, p. 854 Meckler et al. 2012, Science, v. 336, p. 1301 Carolin et al. 2013, Science, v. 340, p. 1564 Roberts et al. 2009, J. Hum. Evol., v. 57, p. 484