Figure 1. Location of tree islands sampled between 1998 and 2000. The sampling strategy for core transects along tree islands is shown in the inset; modified versions of this sampling were used on each island. [larger image]

Tree-island formation and development are influenced by a number of environmental factors, including hydrologic regime, nutrient influx, and underlying geologic structure. Design of management strategies to maximize the health of tree islands and the Everglades wetland as a whole requires an understanding of how changes in these parameters affect the diverse plant and animal communities on these features. To determine how tree-island communities have responded to past climatic and environmental changes, transects of sediment cores have been collected from tree islands in the Everglades to determine: the timing of tree-island formation throughout the region; underlying controls on tree-island formation; patterns of vegetational development on tree islands;nutrient trends on and around tree islands; and the use of sediment phosphorus as a tracer of historic bird populations in the Everglades.

To date, sediment cores have been collected on sixteen tree islands: two in Arthur R. Marshall Loxahatchee National Wildlife Refuge, one in WCA 2A, eleven in WCA 3A, and two in Water Conservation Area (WCA) 3B (fig. 1). Sediments in these cores are described lithologically, dated using radiometric techniques (¹⁴ C and ²¹⁰ Pb), and analyzed palynologically and geochemically. Integration of these data with field evidence for peat/sediment thickness and elevation, bedrock elevation, and existing vegetation are being used to develop models of tree island formation and development.

Palynological and geochemical analyses of cores collected on the head, near tail, far tail, and adjacent wetland have been completed for two tree islands in WCA 3B, Gumbo Limbo island and Nuthouse island (fig. 1). Both are elongate, teardrop-shaped tree islands, with trees clustered at the north end of the island on the head and a tapering tail to the south with shrubs and thick marsh vegetation. Sediment cores collected in transects across the head and along the length of both islands contain records of the last 4.3 thousand years (ka) and predate tree island formation.

Three stages of tree-island development have been identified from both cores (fig. 2). Prior to tree-island formation, sites on the tree island were covered by sawgrass marsh vegetation with common weedy annuals, indicating moderate water depths and hydroperiods with periodic intervals of drought. In the wetlands adjacent to the head, however, slough vegetation, characteristic of deeper water depths and longer hydroperiods grew concomittantly. In the early stage of tree-island development (beginning at 1200 BC on Gumbo Limbo and 300 AD on Nuthouse), ferns, trees, and shrubs became more abundant; this stage persisted for 1,000 to 2,000 years. Mature tree-island vegetation is indicated in the pollen record by strong dominance of fern spores and increased abundance of tree and shrub pollen. Such vegetation has been in place on Gumbo Limbo island since 800 A.D. and on Nuthouse Island since 1500 A.D. Thus, these features are geologically old, with formation occurring as long ago as 1200 BC, early tree-island vegetation lasting 1,000 to 2,000 years, and mature tree-island vegetation persisting for the last 500-1,200 years.

Vegetational reconstructions from transects along the length of these tree islands indicate that sites on the island were drier than the surrounding wetlands throughout their histories, with shallowest water depths on the heads and progressive deepening of water depths with increasing distance along the tail. Thus, relatively localized hydrologic conditions appear to have played a major role in determining where tree islands formed; ongoing research on topography and lithology of the underlying limestone should clarify the roles of climate and geologic framework in influencing tree-island formation. The impact of artifically altered water depths on tree-islands is illustrated by preliminary results from a “drowned” tree island in WCA 2A. Vegetational reconstructions indicate an abrupt change from vegetation characteristic of a tree-island head to that similar to a far tail site, with much deeper water (fig. 3). Based on the current age model, this change occurred within the last fifty years, after high water levels were maintained for more than a decade. Thus, a tree island that had existed for about 2,000 years was altered to an unprecedented degree over a span of a few decades, highlighting the sensitivity of tree islands to extreme hydrologic change.

Sediments collected on these tree island heads are characterized by extremely high phosphorus content, both before and after tree-island formation. Sediments from the surrounding wetlands, in contrast, have relatively low phosphorus content, similar to marshes elsewhere in the Everglades. In tree-island tails, phosphorus levels were low prior to tree-island formation and increased after formation, when water depths decreased. We hypothesize that these phosphorus records may be a tracer of historic bird populations in the Everglades and are analyzing additional cores from rookeries to establish the validity of this hypothesis.

Percent Abundance of Pollen of Major Plant Taxa, Near Tail, Treece's Island, WCA 2A
Figure 3. Percent abundance of pollen of major plant taxa in core from head of Treece's Island, Water Conservation Area 2A. [larger image]

Back to Project Homepage