A critical component of any restoration effort is to establish the natural spatial and temporal variability that exists within the ecosystem. The U.S. Geological Survey Ecosystem History of Florida Bay Project has analyzed seven cores from four sites in Florida Bay in order to establish the patterns of variability within the ecosystem. The goals of this project are to determine (1) what the ecosystem was like prior to human intervention; (2) what is the natural range of variation within the system; and (3) whether recent changes in the system can be correlated to human factors, to natural events, or to a combination of factors.

Modern samples collected at 26 monitoring sites in Florida Bay provide the necessary proxy data for interpreting the downcore historical record. The physical parameters of the environment (including salinity, temperature, substrate, clarity) and the fauna (benthic foraminifers, ostracodes, and molluscs) and flora (palynomorphs and diatoms) present at each site are recorded. (Data from the modern samples and the cores are available at http://geology.er.usgs.gov/gmapeast/fla/home.html (note: this site has moved to http://sofia.usgs.gov/flaecohist/.)) The environmental parameters controlling the spatial and temporal distribution of the biota are inferred from analysis of these modern samples. This information is extrapolated to downcore interpretations based on the biota present in the core.

Analyses of four cores located in the northern transitional, eastern, and central portions of Florida Bay reveal historical patterns of change in salinity and seagrass distribution. Changes in environmental parameters prior to 1900 are illustrated by the distribution of benthic fauna in two of the cores from central and eastern Florida Bay. Natural fluctuations occurred in salinity, but the amplitude of those fluctuations was limited to a 15 to 20 percent shift about the mean. Subtle changes occurred in the benthic fauna around 1910, but beginning around 1940, the pattern of salinity fluctuation departed substantially from the pre-1900 pattern. Post-1940, the salinity oscillates 40 to 60 percent about the mean. This pattern is seen in all indicators measured. Around 1970, a significant but short term decline occurred in salinity. A core from the northern transitional zone reflects changes in freshwater flow that have occurred during this century. The upper part of the core records a significant increase in salinity, with a slight decrease occurring in recent years. An eastern Florida Bay core represents an area of very high sedimentation rates; an increase in salinity occurs in the upper part of the core.

Changes in the interpreted salinity patterns around 1900 are consistent with the timing of the construction of the Flagler Railroad from 1905 to 1912, and the Tamiami Trail and the canal and levee systems between 1915 and 1928. Beginning around 1940, the changes in the frequency and amplitude of salinity fluctuations may be related to changes in water management practices, meteorological events (frequent hurricanes coupled with severe droughts in 1943 and 1944), or a combination of factors. These data suggest that the timing and delivery of freshwater into the Florida Bay system, and the circulation of waters in the system, whether due to human or natural factors, have changed significantly over the last two centuries, and restoration goals need to address these issues.

Natural fluctuations in the distribution of seagrass (predominantly Thalassia) are inferred from the shifts in relative abundance of epiphytal species preserved in the cores. All the cores show an increase in epiphytes and, therefore, in seagrass during the 20th century. An increase also occurs in epiphytal species that can dwell on macro-algal mats associated with Thalassia beds, which suggests that an increase in algal-mats has occurred during the 20th century. The core from central Florida Bay records an extensive period during the 1800Ės of little to no vegetative cover of the substrate based on the near absence of epiphytic species in that segment of the core. Following this period, the epiphytal species increase rapidly in abundance, implying that vegetation has the ability to disseminate rapidly. These data are significant because they indicate that seagrasses fluctuate naturally over time, and that dense seagrass coverage is not necessarily the "natural state" for Florida Bay.

A preliminary analysis of the benthic foraminifers and molluscs taken from three replicate cores at the central and eastern sites has been completed, and the salinity and seagrass patterns seen in these cores are consistent with the data from the previous analyses, indicating that the data are representative of the conditions at those sites. Similar seagrass and salinity trends have been detected in Biscayne Bay (see Ishman and others, this volume), and corresponding changes have occurred in the terrestrial Everglades (see Willard and others, this volume). The correspondence of these changes throughout the ecosystem indicates that the agent of change is regional.

Understanding the natural spatial and temporal variability that exists within an ecosystem is a critical component of efforts to restore systems to their "natural state." The paleoecologic data preserved in undisturbed cores provide insight into past conditions, natural range of variability, and response of the system to change. This historical perspective on Florida Bay allows land managers to determine cause and effect relationships in the systemsĖ responses to past changes. Additionally, the natural component of change can be filtered out of the human-induced component of change, thus allowing land managers to set realistic, cost effective, sustainable restoration goals. Goals contrary to natural patterns of change will be costly and may not be sustainable.

This work has involved collaboration and cooperation with a number of other agencies including Everglades National Park, South Florida Water Management District, National Oceanic and Atmospheric Administration, Florida Marine Research Institute, Keys Marine Laboratory, and Florida Geological Survey. We would like to thank our colleagues at those institutions for their assistance and cooperation in our research.

(This abstract was taken from the Proceedings of the South Florida Restoration Science Forum Open File Report)