Buttonwood Embankment: the historical perspective on its role in northeastern Florida Bay sedimentary dynamics and hydrology

Charles W. Holmes, Debra Willard, Lynn Brewster-Wingard, Lisa Wiemer and M.E. Marot

(Click on any of the images below for a larger version.)

Figure 1

The peninsula of Florida is separated from Florida Bay by a coastal ridge called the Buttonwood Embankment.  The Buttonwood Embankment, averaging 0.5 m (1.5 ft) in height, was characterized by Craighead (1964) as the "embankment that impounds the freshwater of the lower three counties of Florida".  However, the creeks, lakes, and ponds adjacent to the embankment are presently brackish to marine.  The change from fresh to brackish waters has been detrimental to the marsh. Studies have been launched to determine the best methods to reestablish and sustain a more fresh/estuarine environment. To judiciously plan for this restoration two questions have been posed to clarify the history of the area. (1) Have the environmental changes over the years resulted from rising sea level, hydrologic management practices, or both? and (2) What is the role of the "Coastal Levee" in the hydrologic (surface and groundwater) regime of  southern Florida and Florida Bay?

Short-term Developments During the past century, significant environmental changes have occurred in northeastern Florida Bay.  Although minor vegetative changes occurred in the early part of the 20th century, of most concern are the changes that have occurred between 1950-1960.   In the bay, there was the infilling of the pass between Pass Key and Lake Key (fig3), the extension of the mudbank south of Porjoe Key, and changes in sedimentary accumulation rates within the banks. Along the bay fringes within the Everglades, at least one pond began to close (Fig 3). Further inland, along Taylor Creek, the saw-grass plain was encroached by mangrove forest. These events are attributed to the advancing marine environment.

1950

1994

Figure 2 - Filling of Pass Key Pass

1940

1994

Figure 3 - Development of Lake One

Figure 4

Figure 5

Sea level data from Key West shows a stepwise rise over the past century.  Figure 4 is a plot of the monthly means from 1910 to the present. The best fit line for each decade is shown in white.  The best fit lines suggest that between 1910 and 1930 sea level remained essentially constant. Between 1930 and 1948, sea level rose at a relatively rapid rate of approximately 5 mm/yr.  From 1940 to 1970, sea level was constant and then began to rise at about 2 mm/yr to the present level. Superimposed are two periods of sea level jumps, 1948 and 1971.  During these "events" sea level remained high throughout the year (Figure 5 details the 1947-1949 jump). Correspondingly, analysis of rate of sediment accumulation within the Bay, suggest that changes in sedimentary dynamics coincides with these jumps (Figure 6a-Pass Key and Figure 6b -Bob Allen Key) . Figure 7a is a Pb-210 plot of a core from Russell Bank.  This diagram shows a change in sediment rate beginning about 1949 with the formation of a grass bed.  Figure 7b is from Halley and Roulier (1999) which demonstrates the changing isotopic composition of Brachiodontes exustes with time.

Figure 6a

Figure 6b

Figure 7a

Figure 7b

Cores from the Mangrove Fringe Cores were collected recently from the northern parts of Joe Bay, Little Madeira Bay, Taylor Creek Lake One, and Lake Monroe.  Previous cores were taken in the mangrove peat and chronicle long term change in the environment, but these cores were taken within the basins and appear to reveal more recent changes. The cores contain a basal freshwater carbonate overlain by a marine/estuarine sequence.  The carbon dating of the freshwater material ranges from 3000 BP in Lake Monroe, to 1400 BP in Little Madeira Bay. The Pb-210 age model of the marine sequence, however, suggests that the change in environment occurred within the last 50 years.

Joe Bay

Little Madeira Bay

Taylor Creek Lake 1

Monroe Lake

Figure 8

Long term development During the past 2000 years, the Medieval Warm period (1200 to 700 BP) and the Little Ice Age (500 to 200 BP) have left a record in the sediments of South Florida. Concurrent with this climate variability, or because of it, the sea level has changed.  Although sea level in the Florida Bay area, as measured at Key West and Miami, has been rising an average of ~3 mm per year for the past 150 years, there is a dichotomy of opinion on the nature of sea level rise prior to this time.  Many investigators present evidence of a slow and continuous rise in sea level (Scholl and others, 1966; Robbin, 1984) while others present evidence of step-type changes in sea level (Wanless and others, 1995). Some evidence also suggests that, during the Medieval Warm Period, the sea stood 0.5 m higher than today between 1200 -700 years BP (Fairbridge, 1974; Stapor and others, 1991). Our interpretation of the core data seems to be consistent with the latter interpretation. For example, brackish invertebrate taxa and a negative sulfur isotopic signature, occur in a core landward of the embankment at about 570 BP (Willard and Holmes, 1997).

Figure 9

Figure 10

The determination of sea level changes on a millennium scale has been the target of many investigators.  Our interpretation of the published data and with the addition of thirty more analyses from this investigation suggested that dates from shells and non-peat organic material increases the scatter of the data.  Spackman and Dolsen (1962) determined that the best material for measuring sea level was that basal mangrove peat (Figure 10). Figure 9 was constructed using mangrove peat (red dots).  The green stars are data from Davis (1979) and are basal freshwater peats.  Statistical analysis of the data (white line) indicates that sea level has risen less than one meter over the last 2000 years. The white best fit line intersects time zero about 20 cm below sea level as predicted by the Spackman and Dolsen model.

Taylor Creek Transect	Crocodile Point Transect
Figure 11	Figure 12

	Stage 1 Fresh water marl overlying basal peat. Modern Analog- Freshwater coastal ponds.
	Stage 2 Sea level rise causes a breach in coastal levee.  Deposition of estuarine carbonate marl is the result of marine and fresh water mixing.  Modern Analog- Western section of Lake One, Taylor Creek.
	Stage 3 Mangrove encroachment and peat deposits cover estuarine marl as sea level rise slows.  Modern Analog- Eastern section of Lake One, Taylor Creek.
	Stage 4 Development of a central dry pond with a loss of mangroves.  The sediment deposited in this pond is filtered by the fringing mangroves with deposition keeping up with sea level rise. Modern Analog- Crocodile Point.
	Stage 5 Continual rise in sea level results in a thick deposit of fine carbonate mud with terrestrial fauna.  Modern Analog- Crocodile Point.
	Stage 6 Fall in sea level causes erosion of the marine shoreline producing a ridge.  Modern Analog- Shoreline of Little Madeira Bay.

Figure 13

A conceptual model of bank formation was developed through examination of the embankment sediment record: the model consists of six phases that are tied to sea level changes during the past 2,000 years.  During phase 1, the region underlain by the present Buttonwood Embankment was a series of freshwater lakes and ponds, preserved as freshwater peats admixed with freshwater marl.  These basal sediments are overlain by estuarine carbonate mud (Phase 2) capped by a mangrove peat (Phase 3).  14C dates of this peat range in age from 1400 to 1700 BP.   The mangrove layer is overlain by sediments Cottrell (1989) called a supratidal carbonate.  On the Buttonwood Ridge, the supratidal sediment is very fine and devoid of any internal sedimentary structures.  Scattered fossils within this layer are terrestrial, having lived on leaf litter and on the underbrush. `C dates of these fossils and organic material picked from the cores give ages between 1200 and 900 BP, falling within the Medieval Warm Period.  The process of deposition of this unit is unclear, but it is thought to resemble the processes, that are currently adding sediment to the area of Crocodile Point.  On Crocodile Point the fringe mangrove forest filters out sediment during very high water that invades the central low area.  Dating of this accumulation indicates that it is accreting at the same rate as sea level is rising in the Bay, so the supratidal carbonate on the Buttonwood Embankment probably represents a rise in sea level.  A core, collected 10 km inland from the coastline, has a zone with a brackish suite of pollen that dates at the end of the Medieval Warm Period.  This would be consistent with a higher sea level.  During the Little Ice Age that followed, sea level may have dropped, exposing the supratidal muds and leading to erosion creating the Embankment as seen today.

Paleogeography of the Southeastern Everglades

Figure 14

Figure 15

Figure 16

The three cores with fauna and flora records in the southernmost Everglades suggest that the climate events of the last 2000 years had a significant effect on the paleogeography of the region.  It is also suggested that the site furthest east (Mud Creek) had a slightly different history than those sites within the Taylor Slough delta.   There is evidence that during the Medieval Warm Period (about 1100 to 1400 AD) a marine influence reached some 10 km inland from the present shoreline. This is consistent with the conceptual model of the development of the "Buttonwood Ridge".