Twenty-one percent of the freshwater tidal marsh in South Carolina and Georgia is in the boundaries of the Savannah NWR. This resource is being threatened by port development activities that are resulting in salinities moving upstream and changing the composition of the marsh and the habitat it supports. Under my direction, a team of research associates and students conducted a multi-year, multi-phased study as follows: 1) development of a geospatial database of topography, salinities, substrate types, and current vegetation cover (from Spot imagery classification), 2) a hydrologic characterization (based on integrating and synthesizing synoptic field data and long-term monitoring networks) of the deltaic network, 3) an ecological study of marsh plant community responses to environmental gradients (ordination and direct gradient analyses coupled with reciprocal transplanting ), and 4) integration of these studies in a predictive GIS based succession model for predicting change under various scenarios of port development. In addition the study included an analysis of resource partitioning by the passerine bird communities utilizing the fresh and saline wetlands that comprise the study site.
The studies documented a 2-mile upstream displacement of the salinity wedge in conjunction with the operation of a navigational feature, the tide gate (Pub. 29, 19, 33). The gradient analysis (Decorana ordination) documented a very sensitive plant community response to conditions of salinity and increased hydroperiods, with nine vegetation classes corresponding differences in salinity, elevation, and substrate type (distance from a canal). DF analyses of the 9 classes were highly significant in separating the 9 classes, although considerable overlap did occur. Percent of vegetation correctly classified by environmental variables was 87%,47%, 68%, and 79% for freshwater, oligohaline, strongly oligohaline and mesohaline sites, respectively (Pub . 36). Integrating the DF models with expected or predicted spatial salinities (along with topographic and substrate coverages) from actual tide gate manipulations as well as simulations, indicated dramatic re-organization of the marsh composition with the tide gate taken out of operation. The models indicated an 80% recovery of formerly tidal freshwater marsh sites that had undergone transformation to tidal saline marsh types (Pub.33). The results of the resource partitioning by passerine birds in the tidal fresh and brackish marshes indicated that resource partitioning tends to be more pronounced in the brackish marshes, with generalist species tending to exploit the brackish sites and specialist, the fresh sites (Pub. 38 ).
Impact: I was invited to co-organize a special session and proceedings publication of the AWRA to highlight and present the findings of this study (Pub. 33). Additionally, primarily as a result of these studies and those conducted by investigators in the Georgia Cooperative Fish and Wildlife Unit on Striped Bass impacts, the Tide Gate has been removed by the Army Corps of Engineers as an environment restoration project for past negative impacts. These studies have been used extensively by the Fish and Wildlife Service to lever and accomplish the restoration now in place. The issue and study results were presented to a joint meeting of the Congressman for the coastal Georgia District, the Assistant Secretaries for the Army and Interior, and the Director of the Georgia Ports Authority.
This study and its approach facilitated a multi agency interest in its outcome and placed the FWS and NBS in the position of scientific authority for determining the hydrological and ecological consequences of harbor development activities. As stated in the Current Position, new studies have been initiated to refine the succession modelling capability and monitor the effects of past modifications and mitigative measures. Additionally, see commendation letter from the Regional Director, USFWS.
The northernmost remnant of the intact Everglades habitat is located in the 57,235 ha (200 square mile area) of the Loxahatchee National Wildlife Refuge. Historically, this system was comprised of a spatially complex mosaic of wet prairies, saw grass stands, tree islands, alligator holes, and sloughs. However, upon completion of the extensive Army Corps of Engineers’ Central and South Florida Project, the refuge began to experience large scale habitat conversions associated with altered hydroperiods and agriculturally derived pollution (i.e., contaminants and nutrients)(Pubs. 18, 19 ). This area has been subject to a rigidly imposed regime of water deliveries and severely degraded water quality as a result of the conversion of its primary watershed to agricultural land uses.
As Principal Investigator, I directed a team of graduate students (3) and a research associate in a multi-year interdisciplinary study to resolve the issues of hydrological alterations and nutrient loading impacts on this important piece of the Everglades system. As a result of altered hydrologic regimes and excessive nutrient loading, the vegetative habitats of the refuge have responded by conversion to massive monospecific stands of cattails in areas influenced by runoff waters. There has been a tendency to drown habitats in the south of the refuge and desiccate those of the north.(Pub. 18). This study defined gradients of nutrient addition effects and hydroperiods resulting from the management of water on the refuge. The study employed a community-level investigation of vegetative associations in response to the environmental gradients with a spatial characterization of the environmental variables in the GIS. We developed a spatial hydrological simulation model, a vegetative cover database from classified satellite imagery, a spatial coverage of water column/substrate nutrient concentrations, and landscape topography, all geospatially articulated in a GIS. Simultaneous studies of wading bird and forage fish distribution in response to habitat conditions were conducted to examine the influence of hydroperiod and water quality on wading birds and their prey base.
In order to spatially portray the various water depths and hydroperiods imposed on the wetlands and examine vegetative responses, a spatially articulate hydrological model was adapted from an existing model and applied very successfully to the area. The model incorporated approximately 500 1 km cells and was verified with field data.. This capability provided a means hindcasting into the past and recreating the hydroperiod regimes for the past 16 yrs (Pub. 30). Patterns of habitat use were influenced strongly by seasonal variation in water levels as well as the decapod and fish assemblage structure varied among habitats (Pub. 36 ). The bottom line results of the study were that there were indeed impacts and vegetative change resulting form hydrological alterations, but the principal agent responsible for the conversion of approximately 8000 acres to cattail was principally phosphorus in the substrates proximal to agricultural inflows (Pubs. 18, 23, 25, 27,30, 31).
Impact: The nature of the topic and general interest in this study led to my being invited to co-convene a major symposium at the meeting of the International Wetland Conference (INTECOL 4 ) and highlight the results of the various pieces of this study (Pres. 25, 26, 27,28). Additionally, the preliminary findings were an important piece of the vast information base used by the Department of Interior and Justice Department to bring suit against the State of Florida in an attempt to seek a means of improving water delivery and quality to the refuge. The suit was settled out of court, with the government gaining major concessions from the State of Florida to create enormous water treatment facilities to scrub phosphorus from the waters entering the refuge and other protected Everglades lands. These units currently in construction are being built principally at the expense of the sugar farmers and the State. The findings of this study were no doubt influential in the enactment of the Everglades Forever Act of 1993 that established a formula for the restoration of water quality for all waters entering the Everglades and contributed significantly to the South Florida Ecosystem Restoration Initiative currently underway by the Federal Interagency Taskforce.
I was involved in a number of closely integrated studies to document the environmental impacts of diversions of a major riverine drainage from one watershed to another in an attempt to achieve the necessary head for hydropower generation. The initial project was engineered in the early 1940's and resulted in the diversion of 80% of the flow of the Santee River into the watershed of the Cooper River system and out Charleston Harbor. The project was accomplished by interconnecting adjacent impoundments, Lakes Marion and Moultrie. While successful as power generation projects, the diversion resulted in major obstruction of Charleston Harbor with sediments encouraged by the expanded flow of the Cooper River.
In an attempt to resolve the environmental complexity of this project, I managed a team of environmental scientists, geographers, and systems ecologists to synthesize existing information into a cohesive modeling effort to resolve the environmental and economic impacts of the Rediversion project. The initial effort resulted in two major accomplishments: 1) the compilation of a spatial data base of the physiographic and biologic resources of the project area, and 2) the development of a net-energy modeling analysis that translated the environmental and economic impacts of the project into common energy units that aided in identifying the major impact area and resource, the bottomland hardwood forests of the Santee River Floodplain. The spectrum of impacts included losses of anadromous fisheries, a viable shellfish industry, reduced water supplies and quality to urban and industrial consumers (Pubs. 9, 12) .
Once the focus was narrowed to the forested floodplain, another study phase was initiated that resulted in the most significant breakthroughs in the impact analysis of this project. The approach employed in this pilot effort has become the basis of my personal research program.
As Principal Investigator of this phase of the study, I coupled an individually based dynamic forest succession simulation model with the output of hydrodynamic models in a Geographical Information System (GIS) and spatially portrayed the consequences of various project alternatives in terms of forest composition and vigor over the entirety of the effected area of the floodplain. This represented an innovative and successful attempt to link a site model into a GIS to achieve spatial projections of impacts. Not only were impacts identified, but their distributions were portrayed spatially . This study made it possible to examine the extent of flooding that would occur in the forest as a result of the project.
The simulations of the GIS/succession model indicated a massive habitat degradation would occur given the design operation of the power generation facilities. The models predicted only 6000 out of 16000 acres of bottomland hardwood forest would survive the overbank flooding resulting from the project. Although cypress/tupelo would succeed in some limited portions of the area, 90% of the area would not support forested wetlands at all (Pubs.13, 12) .
The study piloted a landscape approach to impact analysis through the coupling of GIS and ecological and hydrodynamic modeling. The success of the study is evidenced in the numerous spin-off studies now underway at the Florida Unit, Louisiana State University and the Southern Biological Science Center. The project has since been implemented and the flooding patterns and negative impacts identified have been substantiated. The National Biological Service is highlighting this study as part of its public display of research in the National Biological Service in the new visitor's center at the Patuxent Wildlife Research Center.