DEVELOPMENT OF A FORAMINIFERAL-BASED TRANSFER FUNCTION FOR COASTAL GEORGIA

Foraminiferal-based transfer functions have become widely accepted over the past decade as a method of enhancing the resolution of Quaternary sea level reconstructions. Researchers construct a “training set” using the distribution of Foraminifera in salt marshes and related environments relative to elevation, and the results are then applied downcore to assess local changes in sea level. Taxa that are susceptible to taphonomic degradation are omitted, and although some have suggested that the occurrence of infaunal taxa may prove problematic, their impact on transfer function development has not been adequately examined. Foraminiferal death assemblages and elevation data collected previously from St. Catherines Island, Georgia, were used to (a) determine whether a transfer function could be successfully developed for this region, and (b) to examine the impact of taphonomically susceptible and/or infaunal taxa. Several iterations of transfer functions were developed, and each was examined using multivariate statistical methods to determine which taxa (and processes) were controlling its development. A foraminiferal-based transfer function can indeed be developed and applied to the expansive salt marshes of coastal Georgia. However, as demonstrated elsewhere, taphonomically susceptible (both calcareous and some agglutinated) taxa must be excluded from the analysis. In addition, the inclusion of infaunal taxa, especially robust deep-dwelling taxa, can produce spurious results. Ponded-water marsh sites are also problematic because they host assemblages that do not vary with elevation in the marsh and are interpreted by the transfer function as occurring at lower elevations than they actually represent. Models therefore can be developed which have high correlation coefficients but are of no practical value. In all, taphonomy, infaunal taxa, and the inclusion of ponded-water sites can produce results that vary by about 1 m in elevation. PCA, factor analysis, and the use of iterative trials in the development of the transfer function all provide insight into the roles of individual taxa. Accurate foraminiferal-based transfer functions can be developed for this region, but they must be carefully evaluated to ensure that they track changes in elevation and not other factors.