Surface distributions of salt-marsh foraminifera from Connecticut, USA: modem analogues for high-resolution sea level studies

Salt-marsh foraminifera are routinely used as sea-level indicators since their vertical distribution is closely linked with elevation relative to the tidal frame. The precise nature of these relationships is variable in time and space, and the accuracy of sea-level reconstructions depends upon the selection of appropriate modem analogues that reliably reflect past fauna-environment associations. The marshes of Connecticut, USA, are sites of ongoing research seeking to produce high-resolution records of sea-level change, yet, little published data regarding their modem foraminiferal distributions exist. This paper presents new surface foraminiferal data from three Connecticut salt-marshes and evaluates their suitability as modem analogues for past sea-level changes. The results indicate that significant intra- and inter-site variability between these marshes and those of neighbouring states exists. As a consequence of this, the extrapolation of fauna-environment relationships developed from marshes with different hydrographic, physiographic, vegetative or climatic characteristics may produce erroneous reconstructions. even when adjusted for variations in tidal range. These errors are potentially greatest if single 'indicator' species are used since the relative abundance of individual taxa does not vary consistently with elevation, even in high marsh environments from the same site. Whilst cluster analysis demonstrates that foraminiferal assemblages from Connecticut are vertically zoned with respect to mean high water (MHW), the composition, elevation and height range of these zones is variable between sites. This spatial heterogeneity results in reconstructions of relatively low and variable precision that restricts their utility in high-resolution sea-level research. New studies seeking to distil decimetre-scale changes in relative sea level will need to employ quantitative methods capable of combining multi-site information to develop fauna-environment relationships that capture this spatial variability. Reconstructions will be most precise when employing local foraminiferal distributions, but may require the collection of additional modem analogue samples from other regions when the extant foraminiferal population of a site differs in composition from its sub-fossil counterpart. (C) 2003 Elsevier B.V. All rights reserved.