USGS - science for a changing world

South Florida Information Access (SOFIA)

publications > paper > application of a weighted-averaging method for determining paleosalinity: a tool for restoration of south florida's estuaries > introduction

Application of a Weighted-Averaging Method for Determining Paleosalinity: A Tool for Restoration of South Florida's Estuaries


Figures & Tables
Paleoecologic analyses of biotic assemblages are currently being applied to a number of societal issues, including global change, land use, and ecosystem restoration (for example see, Brush and Hilgartner 2000; Byrne et al. 2001; Cronin et al. 2005; Dowsett 2007; Oswald et al. 2003; Vandergoes and Fitzsimons 2003; Willard et al. 2005; Willard and Cronin 2007). The common element in these studies is the application of a modern analogue dataset to the interpretation of fossil biotic assemblages preserved in sediment cores, using either the analytical transfer function (Imbrie and Kipp 1971) or the modern analogue method (Hutson 1979). These studies generally use pollen or microfossil assemblages because the broad geographic distribution of these groups is consistent with examining large-scale environmental changes. For smaller-scale ecosystem or watershed-based studies, however, macrofossils can play a significant role in paleoecologic interpretations.

Mollusks can be particularly useful environmental indicators since they are found in terrestrial, freshwater, estuarine, and marine ecosystems, and they represent several levels of heterotrophic consumers (grazers, deposit feeders, suspension feeders, and carnivores). Generalized environmental determinations have been made using mollusks since Lamarck investigated the Paris Basin (Lamarck 1802) and Conrad explored the Atlantic Coastal Plain (Conrad 1838) in the early 1800s. Detailed paleoecologic analyses of mollusks have been conducted using qualitative comparisons to living fauna, for example to determine paleo-depths (Brett et al. 1993), marine cycles (Dominici and Kowalke 2007; Kauffman 1969), marine environments (Allmon 1993; F├╝rsich and Kauffman 1983), and salinity (Hudson 1963). It is rare, however, to see the statistical application of analogue datasets to molluscan paleoecologic analyses. Rousseau's (1991) development of a climatic transfer function for terrestrial mollusks is an exception. A more common application of mollusks to environmental studies is the analyses of stable isotopes of carbon and oxygen in the shells to determine water temperature and salinity (for example, Andreasson et al. 1999; Arthur et al. 1983; Byrne et al. 2001; Cornu et al. 1993; Jones and Allmon 1995; Krantz 1990; Surge et al. 2001). Stanton and Dodd (1970) demonstrated a close correspondence between salinity derived from molluscan assemblage analysis and from oxygen isotopic methods.

A primary goal of the Comprehensive Everglades Restoration Plan (CERP) is to restore the flow of freshwater through the terrestrial ecosystem and into the estuaries to a more natural state (U. S. Army Corps of Engineers 1999). By setting performance measures and target salinities for the estuaries that reflect the natural pre-anthropogenic system, restoration managers hope to achieve that goal. A number of previous studies have examined the salinity history of Florida Bay using paleoecologic assemblages (for example, Alvarez Zarikian et al. 2001; Brewster-Wingard et al. 2001; Nelsen et al. 2002), stable isotopes of corals (for example, Swart et al. 1996, 1999), elemental analyses of ostracode shells (Dwyer and Cronin 2001), and stable isotopes of mollusks (Halley and Roulier 1999). These previous studies are summarized in Wingard et al. (2007a); however, the CERP groups responsible for setting restoration target salinities for the estuaries prefer historical salinity data that are amenable to modeling and statistical analysis (Browder et al. 2008).

A number of cores have been collected by the USGS in Biscayne Bay, Florida Bay, and the southwest coastal area of south Florida to establish pre-1900 temporal and spatial salinity patterns within the estuaries. The purpose of this paper is to determine the reliability of molluscan assemblage data from sediment cores in the reconstruction of historical salinities. We use a simplified version of the modern analogue technique (Hutson 1979), closely allied to weighted-averaging techniques (ter Braak and Juggins 1993; ter Braak and Looman 1986; ter Braak and van Dam 1989; Yuan 2005); all of these methods utilize data on living organisms to quantify biotic changes in terms of some ecologic variable of interest, in this case salinity. Like all paleoecologic studies, we are operating under the assumption that the "present is the key to the past" and that the species we are studying have the same ecological requirements today that they did in the past. In addition, we use assemblages rather than individual or selected species, in order to address the issue of multiple variables controlling the distribution and/or changes in environmental preferences of the fauna that may have occurred over time. As Bosence and Allison (1995, p. 1) explain in a discussion on the importance of using assemblages in paleoenvironmental analyses, "it is unlikely that all [species] will have changed their ecological requirements synchronously." We also discuss where the data and methods need to be improved to provide higher precision of the paleosalinity record. We believe these methods can provide data that are applicable to the management goals of CERP.

Before applying a modern analogue weighted-averaging method to the interpretation of faunal assemblages from core samples, however, we wanted to test the method using a modern dataset collected under known environmental conditions. We are making the assumption that if the method works reasonably well at estimating current conditions, than it will work reasonably well in cores collected from the same estuary and containing extant species. Testing the modern dataset allows us to answer the following questions. How good is the modern analogue dataset at predicting the known salinity? What temporal resolution--days, months, or years? What combination of the data subsets and what statistical measures (mean vs. median) provide the best correlation to the known salinity? Answering these questions will allow for more accurate interpretation of the historical salinity record preserved in the cores, and thus, will provide restoration managers with the ability to set performance measures and targets that accurately reflect the natural system. An example of the application of the cumulative weighted percent (CWP) method to one core, collected in the Northern Transition zone of Florida Bay is provided.

< Abstract | Methods >

| Disclaimer | Privacy Statement | Accessibility |

U.S. Department of the Interior, U.S. Geological Survey
This page is:
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 04 September, 2013 @ 02:04 PM (KP)