Monday, 6 October 2008
George R. Brown Convention Center, Exhibit Hall E
Recent work on the geochemistry of exquisitely preserved Eocene foraminiferans suggests that early Eocene tropical paleotemperatures may have been warmer and more stable through time than previously suspected, and that previous foram-derived paleotemperatures may have been compromised by the growth of early cements. The composition of fossil mollusks can help to resolve this discrepancy because well-preserved shell carbonate can be selected for sampling to the exclusion of any potentially altered or secondary material. The drawback is that most fossil mollusks are collected from shelf settings where salinity variation can affect stable oxygen isotope values. We sampled two individuals of the bivalve Venericardia hatcheplata from each of seven stratigraphic horizons within the Hatchetigbee Formation of the U.S. Gulf Coastal Plain in southwestern Alabama and examined intra-annual δ18O trajectories across growth bands. Data from all 14 shells reveal similar mean values and ranges of seasonal variation that, when converted directly to paleotemperature, yield mean values of ~28°C with a seasonality of ~10°C. Early work on a single shell from this formation produced similar results, but the unexpectedly warm and seasonal temperatures led workers to presume the influence of seasonally lowered salinity. However, this demonstrated consistency of values suggests that these data may reflect true shallow-shelf paleotemperatures independent of paleosalinity influence, for one might expect contributions of fresh water in an estuarine setting to be more variable over time seasonally, inter-annually, and on the scale of time-averaging. Such values are consistent with those derived from well-preserved tropical planktonic forams, suggesting that early Eocene tropical and sub-tropical paleotemperatures may have been comparable and quite warm. The potential significance for paleoclimate work of these and other data from shelf mollusks makes it all the more important that reliable proxies for paleosalinity be developed so as to more accurately interpret stable isotope data from these settings.