Concern exists over hypersalinity – salinity exceeding marine salinity of approximately 35 ppt – in Florida Bay and its negative impact on the ecosystem. High salinity is one of several factors that may cause changes in seagrass distribution, abundance, and density (Fourqurean and Robblee, 1999). To determine the causes of historical changes in seagrass communities, as well as other ecosystem disturbances, and to improve model ability to simulate ecosystem response to future restoration, the patterns and causes of salinity variability in Florida Bay must be determined.

We describe progress towards distinguishing natural salinity variability due to climate variability from changes caused by water diversion in the Everglades during the 20^th century. Variability in salinity in Florida Bay reflects the bay’s geometry, which features the Gulf of Mexico to the southwest, the Florida Keys to the southeast, and the Everglades to the North, and its shallow bathymetry and circulation resulting from the complex of mudbanks in the bay. Florida Bay salinity varies over seasonal, interannual, and decadal timescales but the instrumental record prior to the 1990s is insufficient to attribute cause to decadal trends. Here we specifically address the application of quantitative “retrodictive” estimates of past salinity patterns derived from geochemical studies of ostracodes from sediment cores from the central part of Florida Bay. Parallel studies of molluscan shell chemistry and seasonal salinity variability using the same sediment cores will be given in a separate presentation at the meeting.

The ratio of magnesium to calcium (Mg/Ca) ions in the calcium carbonate shells of the crustacean group ostracodes is strongly influenced by the salinity and temperature in which the organism secretes its shell (see Dwyer et al., 2002). The variability in Mg/Ca ratios of Florida Bay water is “captured” in the Mg/Ca ratios of the ostracode shell when it secretes its adult shell. It is conventional to express the relationship between shell and water chemistry as follows:

where

Mg/Ca represents the atomic ratio of Mg to Ca and K _D-me is the partition coefficient for magnesium (Dwyer and Cronin 2001). Mg/Ca ratios in fossil ostracodes from sediment cores were used to construct a paleosalinity curve for central Florida Bay. The results showed that 20^th century decadal oscillations in salinity were related mainly to regional rainfall variability, which in turn was influenced by climate processes associated with interannual El Niño-Southern Oscillation and decadal Pacific North American patterns (Cronin et al., 2002).

In this study, the accuracy of the Mg/Ca salinity curve was “groundtruthed” against instrumental records of salinity from Florida Bay (Nuttle et al., 2000). Figure 1 shows the Mg/Ca paleosalinity curve plotted against the instrumental record of salinity for the last 50 years. The instrumental record was constructed from mean monthly salinity values for Whipray, Rankin, and Bob Allen Basins pooled to yield a grand mean for the central region. Table 1 summarizes the paleosalinity-instrumental salinity comparison for Russell Bank.

Both records exhibit large decadal swings in salinity in central Florida Bay ranging from > 50 ppt to the low 20s. For salinity maxima during the early 1990s, late 1970s and mid-1960s the difference between paleo and instrumental salinity was approximately -1.1, 0.9, and 3 ppt, respectively. The paleosalinity method overestimated the maxima for the 1950s by 9.7 ppt, however the instrumental record stops in 1957 and it is unlikely the two records had a similar period of record for this high salinity event. The differences for four periods of Florida Bay salinity mimima approximately 1993-1995, the 1980s, the late 1960s-early 1970s and around 1960 were 0.3, 3.6, 1.9 and 0.8 ppt, respectively. Given the temporal gaps in the instrumental record, the error bar on the sediment core age estimates, and the spatial and temporal averaging used for both records, these comparisons provide a remarkable confirmation that the Mg/Ca-based shell chemistry method yields accurate estimates of past salinity to within < 1 to 4 parts per thousand.

Figure 1 also compares paleosalinity with hydrological instrumental records from south Florida: mean monthly water height at USGS well 196a, Homestead, Florida, annual discharge at Taylor Slough since 1961, and NOAA monthly regional rainfall anomalies. The results reveal decadal patterns in rainfall and water well height since the 1930s and an inverse correlation between salinity and well height, rainfall and discharge. For example, wet periods (i.e., approximately 1960, late 1960s, early 1980s) were characterized by positive precipitation and well height anomalies and relatively low Florida Bay salinity. Periods of low salinity recorded in the Mg/Ca paleosalinity curve during the early 1950s and the early 1940s are also evident in the precipitation and well height curves. Decadal salinity oscillations over the past century are also evident from Mg/Ca analyses from Park and Bob Allen Keys, and Little Madeira Bay (Dwyer and Cronin 2001).

Figure 1. Comparison between paleosalinity and measured salinity (left), Homestead well level and Taylor discharge (center), and NOAA regional rainfall (right). [larger image]

These results demonstrate that the Mg/Ca method reconstructs salinity patterns that faithfully record salinity extremes to within a few parts per thousand and that the Mg/Ca method could be applied to new core sites to build a network of long-term salinity records in Florida Bay and adjacent bays.

References

Cronin, T. M., et al. 2002. Climate Research 19: 233-245.
Dwyer, G. S. and Cronin, T. M. 2001. Bull. of Amer. Paleont. 361: 249-276.
Dwyer, G. S., et al. 2002. AGU Monograph 131: 205-225.
Fourqurean, J. W. and Robblee, M. B. 1999. Estuaries 22: 345-357.
Nuttle, W. K. et al. 2000. Water Resources Research 36: 1805-1822.
Robblee, M. et al. 1991. Marine Ecology Progress Ser. 71: 297-299.

Contact: Cronin, Thomas M. 926A US Geological Survey, Reston, VA 20192, Phone: 703-648-6363, Fax: 703-648-6953, tcronin@usgs.gov

(This abstract was taken from the Greater Everglades Ecosystem Restoration (GEER) Open File Report 03-54)

Back to Project Homepage