Tree islands are considered key indicators of the health of the Everglades ecosystem because of their sensitivity to both flooding and drought conditions. Tree islands also act as a sink for nutrients in the ecosystem and may play an important role in regulating nutrient dynamics. Although management strategies to restore and even create tree islands are being formulated, the published data on their age, developmental history, and geochemistry is limited. To determine underlying controls on the formation and development of Everglades tree islands, this project will integrate floral and geochemical data with geologic and vegetational mapping activities to establish the timing and environmental controls on development of both healthy and degraded tree islands throughout the Everglades. We also will study the role of tree islands in the geochemical budget of nutrients in the Everglades and investigate the use of sediment phosphorous as a tracer of historic bird populations in the Everglades.

Statement of the problem:

Although maintainence of healthy tree-islands is a critical component of most Everglades restoration plans, little scientific data exists on their age, patterns of development, geochemistry, and response to environmental alterations. Loss of tree islands over the last few decades has been documented clearly in parts of the Water Conservation Areas, but the conditions needed to create and maintain tree islands are not clearly understood. Preliminary results of research on temporal patterns in vegetation and geochemistry from two tree islands in WCA 3B (Willard, et al, in press; Orem et al, in press) indicate that initial development of tree-island vegetation began on those islands as long as 3,200 years ago, and the present vegetation was in place by 1,700 years ago. These studies also indicate different hydrologic conditions and much higher phosphorus levels on tree islands than in surrounding sloughs even before development of tree-island vegetation. Earlier work by Loveless (1950) reviewed existing hypotheses for tree-island formation and suggested that tree islands overlie topographic highs in underlying bedrock. However, the hypothesis has not been tested. Floating or "pop-up" islands, which are prevalent in Loxahatchee National Wildlife Refuge (NWR), were studied by Gleason et al (1980), who analyzed peat petrography and pollen from two islands and determined that the islands formed after floating battery peats were stranded upon sawgrass plains and ultimately were colonized by trees to form small tree islands.

In light of the relatively sparse data generated on Everglades tree islands to date, a number of questions remain about their formation, including: 1) Is there a common pattern of tree island formation throughout the Everglades? 2) Did most tree islands develop at about the same time (i.e. did certain climatic events control tree island development)?; 3) Did tree island heads and tails develop concomittantly? If not, what factors controlled tail development (hydrology vs. nutrients)? 4) Did underlying bedrock topography and lithology control location of tree island development? 5) Are high levels of phosphorus observed on tree-island heads in preliminary work tied to groundwater discharge, presence of wading bird populations, or some other factor? 6) What environmental factors must be in place for development of tree islands, and what factors may lead to degradation and loss of tree islands?

Project objectives and strategy:

This project aims to develop a vegetational and geochemical history of tree islands to be integrated with mapping of current vegetation, topography and lithology of underlying limestone, and hydrologic studies to understand the environmental parameters that have controlled past tree-island formation. Such an interdisciplinary effort is necessary to determine how geologic and environmental factors interacted to form Everglades tree islands. Specific goals of this project include: to document the timing of tree-island formation across the region; to establish patterns of vegetational development and geochemical changes on the islands; to compare development of different types of tree islands; and to develop a model of tree-island formation that may be used in restoration of degraded islands and, possibly, creation of new islands. The project will use several paleoenvironmental proxy methods and geochronological techniques on sediment cores to establish the vegetational and geochemical history of tree islands in the Water Conservation Areas (WCA), Loxahatchee NWR, and Everglades National Park (ENP). Selection of tree islands for study will be made in consultation with collaborators at South Florida Water Management District (SFWMD), the Florida Game and Fresh-Water Fish Commission (FGFC), Loxahatchee NWR, and ENP. These islands will be selected to provide regional coverage and ultimately will include several different types of islands, including fixed, tear-drop shaped tree islands, pop-up islands, drowned islands, willow heads, and cypress heads. Paleontological and geochemical data from the islands will be integrated with data generated by collaborators on bedrock surface topography and composition, vegetation, surface and groundwater hydrology, and remote sensing work to produce models for formation of each island and, ultimately, to develop models of tree-island formation on a regional basis.

Paleoenvironmental reconstructions will be based on analysis of sediment core samples collected in transects along the length and breadth of selected tree islands. Integration of geochronologic and vegetational data from each island will identify phases in development of tree-island vegetation from the original marsh/slough communities as well as the timing of tree-island formation. Comparison of these patterns among different types of tree islands across the region will establish whether common regional patterns of development exist. Through analysis of transects of cores collected along the length of the island, we will document the timing of tree-island head and tail formation and determine whether they formed synchronously or whether tree-island heads formed before tails. Such information is critical to evaluate hypotheses that tails formed due to accumulation of debris behind the head (hydrodynamic hypothesis), due to chemical/nutrient enrichment from the head (chemo-hydrodynamic hypothesis), or due to hydrologic controls tied to climatic and environmental changes. An understanding of processes governing head and tail formation is necessary for development of plans to restore and create tree islands.

Transects of sediment cores also will be collected across the breadth of tree-island heads from marsh to marsh to compare long-term vegetational and geochemical patterns on and off the island. These will include cores for solid phase paleoecology, geochemistry, and dating as well as cores for pore-water geochemical studies. Wells are being installed on heads and adjacent marshes of two to three tree islands by SFWMD for monitoring of groundwater discharge on and around tree-island heads. During the drilling of each well, thirty-foot long cores of the underlying limestone will be obtained. In collaboration with personnel at SFWMD, we will describe the lithology of these cores and determine wheter the limestone composition is different under the heads compared to the surrounding marsh. Also, SFWMD and FGFC are collecting data across these transects on sediment depth to bedrock to determine whether topographic highs underlie the islands.

Initial geochemical and biological data from tree-island cores in WCA 3B indicates correlations between specific vegetational assemblages and elevated phosphorus levels in the tail and marsh. Regional sampling of tree islands will establish whether this is a common pattern on Everglades tree islands or a local phenomenon. The source of phosphorus is unclear, but bird guano is a likely source (Powell, et al, 1991). We will examine the sources of phosphorus to the tree islands through tracer and isotope studies of sediment cores collected on the islands and from guano and sediment cores collected from extant bird rookeries in the Everglades. Another possible source is upwelling ground water; analysis of samples from wells installed on key islands by SFWMD will clarify the relative roles of wading birds and upwelling ground water in phosphorus enrichment of tree-island sediments.

Sustained high water levels during the last few decades have resulted in degradation and loss of tree islands in much of WCA 2A. Determination of vegetational patterns on drowned and healthy islands should help establish the sequence of vegetational changes during tree-island drowning and provide useful data to aid restoration efforts on specific tree islands. Integration of these long-term patterns with ongoing mapping of bedrock topography and veetation will facilitate development of models of tree-island formation and evolution. An understaning the role of various environmental parameters in controlling plant community composition and geochemistry,n both healthy and degraded islands, is critical in determine\ing appropriate management strategies to sustain tree islands, restore degraded islands, establish appropriate monitoring strategies, and evaluate the feasibility of creating tree islands.

Potential impacts and major products:

Because tree-island health is influenced directly by water-management practices, SFWMD and FGFC have developed tree-island research plans to establish protocols to monitor tree-island health as a guide for adjustments of Everglades hydrology. Programs in both agencies need data on patterns of tree-island development to aid implementation of hydrologic regimes based on the Natural Systems Model and for assessment of the relative ecological impacts of alternative plans being considered by the C&SF Project Comprehensive Review Study (the "restudy").

Because little research has been devoted to evaluating the different models of tree-island formation that have been proposed, basic information and data on tree-island formation and geochemistry is a critical need of resource managers. A commonly accepted model for formation of fixed tree islands assumes that they are associated with topographic highs in the bedrock underlying the Everglades, with the head of the island overlying the high. We will evaluate the validity of this hypothesis by integrating floral and geochemical records with data on bedrock topography and composition.

The development of tree-island tails also is poorly understood, and several hypotheses for tail formation have been proposed. These include: a "hydrodynamic" hypothesis that the tail develops due to litter from the head being deposited downstream by water currents; a "chemo-hydrodynamic" hypothesis that the tail develops due to nutrient release from decomposing plant material on the head; and a hydrologic hypothesis that the bedrock under the tail also is elevated relative to the surrounding marsh, and formation of both the head and tail are due to hydrologic changes. Comparison of geochemical and floral data from well-dated cores should help evaluate the validity of these hypotheses; this is needed to develop workable, sustainable plans for restoration of degraded islands.

Tree islands also may play an important role as nutrient sinks and sources within the tree island/slough/marsh complex that dominates large parts of the Everglades. The role of tree islands in nutrient cycling within the ecosystem, however, is virtually unknown. Basic data on the nutrient geochemistry of tree islands within the context of the surrounding slough and marsh system will provide managers with the information necessary for integrating tree islands into a nutrient model for the ecosystem. In addition, profiles of phosphorus in dated cores from tree islands may provide a useful proxy for historical wading bird populations in the ecosystem. Although it generally is accepted that wading bird populations in the Everglades have declined dramatically over the last 100 years, the reasons for this decline still are unclear. An understanding of historical trends in wading bird populations will provide information on what environmental factors are most important. It also will provide a basis for managers to manipulate the conditions within the ecosystem to maximize the potential for recovery of wading bird populations.