U.S. Department of the Interior
U.S. Geological Survey
OFR 2007-1203

Methods

The cores discussed in this report were collected over a nine year period, from 1994 to 2003. Core locations are shown on Figure 1 and listed in Table 1. Three methods were utilized to develop the age models for the cores: (1) ²¹⁰Pb analyses of the sediments; (2) first stratigraphic appearance of pollen of Casuarina (Australian pine); and (3) radiocarbon dates on shells or wood (Table 1). The age models for the last century of deposition were based on ²¹⁰Pb and pollen biostratigraphy. Radiocarbon dates of individual shells or wood fragments were used to establish a chronologic framework for the lower portion of the cores.

Figure 1. Satellite image showing core locations. Numbers correspond to listing in Table 1. Red dots represent locations of cores with synthesized age information discussed in this paper. Yellow dots mark locations of cores that have had only 210Pb analyses done. 210Pb data for all cores (yellow and red) are in Appendix 1. [larger image]

Table 1: Location of all cores discussed in report or listed in Appendix. Collection information and analyses conducted are listed. [click on the thumbnail image above to view Table 1]

²¹⁰Pb Analyses

Decay of unstable ²¹⁰Pb isotopes into its daughter products is a well-documented method for dating 20^th century sediments (Appleby and Oldfield, 1978; Abril and others, 1992; Ducat and Kuehl, 1995; Appleby, 1997; Panayotou, 2002; Walling, 2003). ²¹⁰Pb (²¹⁰Pb) activity was measured by alpha spectroscopy using the method outlined in Flynn (1968) in which ²¹⁰Pb and its progeny, ²¹⁰Po, are assumed to be in secular equilibrium. Supported ²¹⁰Pb activity was determined by continuing measurements until activity became constant with depth. Excess ²¹⁰Pb activity was calculated by subtracting the supported ²¹⁰Pb activity from the total ²¹⁰Pb activity.

Age models have been used to quantify sediment accumulation rates based on ²¹⁰Pb decay profiles that take into account the variables of atmospheric flux, sediment supply, and mixing. Robbins (1978), Oldfield and Appleby (1984), Carroll and others (1995), Robbins and others (2000), and Nie and others (2001) have reviewed the various methods, applications and approaches to ²¹⁰Pb age modeling. The chronologic model used in this study, developed by Robbins and others (2000), is a simple first-order model. In this model, the atmospheric ²¹⁰Pb flux and sediment accumulation rate are assumed to be constant and any variability in ²¹⁰Pb concentrations, with the exception of decay, is averaged by sedimentological processes. The surface activity is assumed to be constant and equal to the flux of ²¹⁰Pb divided by the sediment accumulation rate in terms of grams per square centimeter.

²¹⁰Pb data for 50 cores from Florida Bay and Biscayne Bay are provided in Appendix 1, and the depth where ²¹⁰Pb values reach background levels is indicated in Table 2. Excess (unsupported ²¹⁰Pb) and total ²¹⁰Pb activity are shown in Figures 2-9. A component of the ²¹⁰Pb analysis is a sedimentological analysis of grain size and loss on ignition (LOI) - a general indicator of organic material. These data also are presented in Appendix 1. To obtain LOI, sample is placed in a weighed crucible, weighed, placed in a muffle furnace for six hours at 450°, cooled, reweighed and percent loss calculated. Grains size data are obtained by sieving a wet sample with deionized water through a 63 micron sieve, drying, and weighing each size fraction. For a detailed discussion of the collection and analytical procedures, geochemistry, and correlation to radium-226 and coral bands, see Holmes and others, 2001.

Table 2: Summary of data used to determine beginning of 20th century deposition in cores: first occurrence datum for Casuarina equisetifolia and depth to background levels of lead-210. Data are illustrated in Figure 2 and 3. Total lead-210 is shown in Figures 4-9. [click on the thumbnail image above to view Table 2]

Figure 2. Excess (unsupported) 210Pb activity in decays per minute per gram (dpm/g) plotted against depth in centimeters for Biscayne Bay cores. Data in blue, occurrence depth of Casuarina in green, post-modern carbon-14 in red. Black indicates approximate depth where 210Pb activity reaches background values. In some cores excess 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). [click on the plots above to view a larger version]

Figure 3. Excess (unsupported) 210Pb activity in decays per minute per gram (dpm/g) plotted against depth in centimeters for Florida Bay cores. Data in blue, occurrence depth of Casuarina in green. Black indicates approximate depth where 210Pb activity reaches background values. In some cores excess 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). [click on the plots above to view a larger version]

Figure 4. Total 210Pb activity in decays per minute per gram in five cores from the Northern Margin of Florida Bay in Little Madeira Bay and at Trout Creek. Red mark indicates approximate depth where 210Pb activity reaches background values (Table 2). In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Figure 5. Total 210Pb activity in decays per minute per gram in cores from Russell Bank, Florida Bay. Red mark indicates approximate depth where 210Pb activity reaches background values (Table 2). In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Figure 6. Total 210Pb activity in decays per minute per gram in five cores from Bob Allen mudbank in Florida Bay. Red mark indicates approximate depth where 210Pb acitivity reaches background values (Table 2). In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Figure 7. Total 210Pb activity in decays per minute per gram in cores from central and eastern Florida Bay and northern Biscayne Bay. Red mark indicates approximate depth where 210Pb activity reaches background values (Table 2). In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Figure 8. Total 210Pb activity in decays per minute per gram in cores from western Florida Bay. Red mark indicates approximate depth where 210Pb acitivity reaches background values (Table 2). In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Figure 9. Total 210Pb activity in decays per minute per gram in cores from Pass Key, Florida Bay. Red mark indicates approximate depth where 210Pb activity reaches background values (Table 2) In some cores total 210Pb values appear to decrease further below the point where they reach background values; this is due to decreased radium values with increased depth in core as discussed in Holmes and others (2001). Data are in Appendix 1. [larger image]

Pollen Biostratigraphic Analyses

An excellent biostratigraphic marker for post-1900 AD sediments is the occurrence of Casuarina equisetifolia pollen (Australian pine) in Florida sediments. Casuarina equisetifolia is an exotic species introduced into south Florida in the late 19^th century (Langeland, 1990). Calibration of Casuarina pollen abundance with ²¹⁰Pb geochronologies indicates that C. equisetifolia pollen first occurred in south Florida sediments at ~1910+/- 15 years, becoming common after 1940 (Duever and others 1986, Wingard and others 2003). For this study, pollen typically was analyzed at 10 cm increments, increasing the error range of the first occurrence of Casuarina. It should be noted that peak pollen production was only obtained when stands of trees reached maturity, typically a few decades after germination. Therefore, the first consistent appearance of Casuarina pollen often is 10-30 cm higher than the point at which ²¹⁰Pb values reach background levels. When the presence of Casuarina is the only stratigraphic evidence available to identify sediments deposited during the 20^th century, we assign it an age of AD 1910 +/- 15 years. Otherwise, we rely more strongly on ²¹⁰Pb data to develop age models for the 20^th century (see Table 2).

Pollen data were obtained on ten of the thirteen cores discussed below; no residual material was available for analysis in the three remaining cores. Full pollen data sets for nine of these cores have been previously published (Brewster-Wingard and others, 1997; Ishman and others, 1996 (data revised); 1998; Wingard and others, 1995 (data revised); 2003; 2004). Methods for processing followed procedures described in Willard and others (2001).

¹⁴C Analyses

Radiocarbon dates provide an estimated age and a basis of comparison of the cores for the pre-20^th century deposition. Stuiver (1986, p. 106) discusses the problems associated with ¹⁴C for samples less than a thousand years old - “the relative age uncertainty of ¹⁴C dates caused by counting statistics is about one percent or more. It is large because the statistical uncertainty in the measured sample activity becomes large in relation to the decrease in ¹⁴C activity caused by decay over such a short period.” A number of factors may affect the accuracy of ages based on ¹⁴C, including species-level fractionation of carbon, up-take of “old carbon”, global marine and local reservoir effects, variations in ¹⁴C production in the atmosphere over time, circulation of marine carbon in the open ocean and in ground water, and dissolved inorganics (Lowe and Walker, 1997).

Samples were dated with accelerator mass spectrometry (AMS) by Beta Analytical Inc. (33 samples) and the USGS Radiocarbon Lab (15 samples) (Table 3). Material was selected for analysis based on general position within the individual core and quality of preservation of the material. Radiocarbon 2 σ age ranges were calibrated to calendar years using Calib 5.0 (Stuiver and Reimer 1993; 2005) and either a marine or terrestrial carbon correction factor as indicated on Table 3. All calibrated dates herein are presented as calendar years before the date of core collection and designated yrBP. Ages on young shell material deposited after 1950 are expressed in PMC (percent modern carbon).

Table 3: Results of carbon-14 analyses on core samples from Biscayne Bay and Florida Bay. Data are listed alphabetically by core location within each bay. This table includes all data on carbon-14 analyses available for the estuaries in south Florida. Age models have not been developed for all cores. [click on the thumbnail image above to view Table 3]

Development of Age Models

Age information, including ¹⁴C, ²¹⁰Pb and pollen where available, were compiled and examined using a mixed effect regression model provided by the CAgeDepth.fun function within the Windows-version of the statistical software R (see Heegaard, 2003 for the function and http://cran.r-project.org/ for the R software; last accessed 11/8/2006). The CAgeDepth function takes into consideration the variance between samples (depth and time) and the variance within each sample (sample thickness and probable age range). The output produces an estimated age-depth curve with a 95% confidence interval (Heegaard and others, 2005) (Appendices 2-3).