Florida Bay Research Programs and Their Relation to the Comprehensive Everglades Restoration Plan (2002)

Major changes in the Florida Bay ecosystem have been observed in the last two decades, including Thalassia die-off (Zieman et al., 1988), declining shrimp harvests (Ehrhardt and Legault, 1999), hypersalinity (McIvor et al., 1994), increased turbidity and decreased light penetration (Boyer et al., 1999), declines in the sponge population (Butler et al., 1995), and others. These changes have led to an increased focus on Florida Bay by the research community. As noted in Chapter 1, this research is being conducted by numerous governmental and academic institutions.

Some, but presumably not all, of these changes may be due to water management practices in south Florida. Since the construction and operation of the Central and South Florida Project, the amount and timing of freshwater runoff into Florida Bay has been radically different from the historical patterns, including a reduction in discharges into the Bay (Light and Dineen, 1994). Although no direct cause-effect relationship has linked south Florida water management to the sea-grass die-off, hypersalinity, declining shrimp harvests, and water-quality changes in the Bay stimulated much of the concern about the condition of the Florida Bay ecosystem reflected in the 1992 Federal Water Resources Development Act, which authorized the Restudy of the Central and South Florida Project.

This chapter summarizes some of the research that needs to be undertaken to resolve questions of the potential effects of the plan that arose out of the Restudy, i.e., the Comprehensive Everglades Restoration Plan (CERP), on Florida Bay. No implication is intended that research conducted to address other science or management goals is unimportant or unnecessary.

This report places emphasis on the potential for changes in the Bay such as increased sediment turbidity and algal blooms resulting from increased freshwater flows from the Everglades. Therefore, this chapter is organized into research needs for (1) the water budget of the Bay, (2) nutrients dissolved or suspended in this water, and (3) modeling approaches to enhance understanding of the connections between the freshwater and marine environments.

Without knowledge of the sensitivity of the Florida Bay ecosystem to alterations in freshwater and nutrient inputs, the effects of the CERP on Florida Bay are difficult to predict. The overall water budget (i.e., inflows, outflows, and storage) for the Bay appears to be dominated by rainfall and evaporation (Nuttle et al., 2000). However, these authors recognized the need for improved estimates of precipitation, evaporation, changes in runoff resulting from natural causes and water management, and exchange with the Gulf of Mexico and the Atlantic Ocean. It is even more important, however, to recognize that the overall budget of the Bay is not necessarily relevant to individual sectors of the Bay. For example, salinity and nutrient data clearly show that surface flow from the Everglades is important in the northeastern Bay (Boyer et al., 1999).

Priority research topics concerning water balance include the following:

• Accurate quantification of surface runoff into Florida Bay clearly is a high priority. Ongoing studies (Hittle et al., 2001) have been measuring freshwater discharge into northeastern Florida Bay at five creeks in the Taylor Slough and C-111 canal basins since late 1994. Continued long-term monitoring of spatial and temporal variations in surface runoff will reduce the uncertainty of the effect of the CERP on the Florida Bay water budget, and this activity is recommended. Although sheetflow over the Buttonwood Embankment (Figure 2) between the channels that breach it was not believed to be important by Hittle et al. (2001) or Davis et al., (2002), this should be confirmed by installing instrumented sites along the embankment. Likewise, significant additional freshwater may enter the Bay by diffuse seepage through the embankment, and this process also merits evaluation. Continued research on the hydrodynamic characteristics and net outflow of water from the Shark River Slough basin (e.g., Levesque and Patino, 2001) may be useful, given that some of this water appears to reach the central Bay (Lee et al., 2002).

• Aside from the shallow seepage referred to above, groundwater inputs may be important to the system in other areas. Although nearly all of this influx must be saline (Fitterman and Deszcz-Pan, 2001; Corbett et al., 1999, 2000), saline discharge that has circulated through the phosphorite-rich quartz sand deposits described in Cunningham et al. (1998) may be a significant source of phosphorus in the northwestern Bay (Brand, 2002). Furthermore, a better understanding of how the proposed hydrologic changes in the CERP may affect the position of the freshwater-saltwater contact in the subsurface should be developed. This kind of understanding may be achieved through modeling of the wetland/coastal transition zone, as described later in this chapter.

• The South Florida Water Management Model (SFWMM) predicts about the same annual runoff toward Florida Bay via Craighead Basin and Taylor Slough in the year 2050 as occurs presently (Figure 3). In contrast, flows down Shark River Slough would increase from 702,000 to 1,255,000 acre-ft per year (8.66x10⁸ to 1.55x10⁹ cubic meters per year) (Figure 4). Hence, on average, the full CERP implementation has a minimal effect on the volume of direct fresh surface water flow into the Bay but a potentially significant effect on the discharge out of Shark River Slough and Whitewater Bay, which eventually reaches the Bay. Additional definition of these flow pathways is urgently needed for modeling of impacts in the Bay. Variability must be accounted for through analysis of the full 31-year simulation period.

• After mixing with seawater, water from Shark River Slough tends to reach Florida Bay in about two to six weeks (Lee et al., 2002) and may still influence the Bay’s nutrient budget. The effects on the Bay of a Shark River Slough increase and a stable Taylor Slough/Craighead Basin freshwater flow have not been investigated, however. It is important to remember that the flows shown in Figures 3 and 4 are modeled flows, and may not accurately represent actual flows under current conditions or those after implementation of the CERP. Furthermore, the hydrologic models are subject to change (e.g., eventual conversion from the SFWMM to the South Florida Regional Simulation Model) as well as changing simulation conditions (e.g., use of a 36-year meteorological record vs. a 31-year record). Nonetheless, the implications of this potentially large shift in the magnitude of freshwater discharges down Shark River Slough as a result of the CERP need to be investigated.

• Although surface and groundwater inputs are of most direct use in understanding the effects of the CERP on the Bay, an improved water budget for the Bay as a whole, including natural variability, will help put these inputs into a larger context. Nuttle et al. (2000) estimated a water budget using evaporation rates from a rough salt balance calculated for four regions of Florida Bay, with pan evaporation rates used to distribute annual evaporation seasonally. Price et al. (2001) have been using four approaches (energy flux, vapor flux, stable isotopes, and a box model of salinity) to estimate mean rates of evaporation and its spatial and temporal variation. Both the energy and vapor flux methods depend upon measurement of net radiation, water and air temperature, relative humidity, rainfall, and wind speed and direction. Price et al.’s (2001) study is only a two-year project; continuation of such work (and expansion of the monitoring

station network) would provide the CERP with more reliable information about evaporation and precipitation over Florida Bay and about long-term climate changes.

Modification of freshwater delivery to the Bay as a result of the CERP has the potential to affect both overall nutrient loadings and the limiting nutrient status of the water by changing the nitrogen:phosphorus ratio in the source waters and ultimately in Florida Bay and the nearby coastal ocean. The consequences of changes in water quality on the Florida Bay ecosystem are a major uncertainty of the CERP. Consequently, studies are needed to address the following questions or concerns:

• Most of the dissolved nitrogen (which is mostly organic nitrogen) in Taylor Slough and the Craighead and C-111 basins passes through mangrove wetlands at the southern boundary of the Everglades. How is the nature of the dissolved nitrogen altered in this environment? Research on this and the following question is ongoing (e.g., Reyes et al., 2001a, b).

• What are the overall loadings of nitrogen and phosphorus from these wetlands through the tidal creeks that pass through the Buttonwood Embankment (Hittle et al., 2001) and into the Bay? How do these vary seasonally?

• The apparent increase in total phosphorus concentrations in Shark River Slough water as it passes through the mangrove zone, as reported by Rudnick et al. (1999), should be examined again. If confirmed, the source of the phosphorus should be determined to evaluate its potential contribution to future enhanced water flow from the CERP.

• One of the most significant uncertainties regarding nitrogen concerns the potential bioavailability of the dissolved organic nitrogen and phosphorus (DON and DOP) in Everglades water (including waters in canals that drain south from the Everglades Agricultural Area and Conservation Areas) entering Florida Bay. This includes two issues: (1) what fraction of the DON can be assimilated directly by primary producers (algae and macroscopic aquatic vegetation) in the Bay, and (2) what fraction of the DON can be transformed into nitrogen forms that can be assimilated by algae and macroscopic aquatic plants within the time frame that inflowing water resides in the Bay.

• The chemical nature of DON in Everglades surface water is not known, and such information would be useful to predict the ease and rates of its transformation to inorganic nitrogen forms via microbial metabolism and photochemical degradation. Overall, the factors and processes that control the rates of DON and DOP transformation into bioavailable forms under ambient conditions are poorly understood. What percentage of the DON and DOP pool is refractory within the time constraints imposed by water movement and circulation patterns into and out of the Bay? (Research on internal Bay phosphorus cycling is ongoing by Florida International University scientists.) Much research on Florida Bay is focused on interactions between dissolved, mineralized, sorbed, and gaseous phases of the major limiting nutrients. These transformations can be enormously complex, varying spatially and temporally depending on season, physico-chemical conditions, and a wide range of additional factors. Thus, although process-based work is resource consumptive, it is essential that the continuity of this work be maintained. The key to predicting the effects of various changes to the Everglades on the Bay systems is the use of process-based (mechanistic) models that are based on understanding transformations of the major limiting nutrients.

• Ecosystem response to nitrogen and phosphorus inputs is important in understanding the effects of the CERP on Florida Bay. This is true regardless of the source and/or form (organic vs. inorganic) of the nutrients that enter the Bay. Quantifying the magnitude of nutrient loadings by source (e.g., organic soils oxidation, urban and agricultural runoff, and regional atmospheric deposition) is relevant, however, if steps to reduce nutrient loading to the Bay become necessary. In support of this potential mitigation

effort, an effort should begin now to synthesize and evaluate present knowledge of the sources and transport of nitrogen and phosphorus within the Everglades system.

• The overall nutrient budgets for the Bay are important in order to place the findings of the earlier questions into context.

Not only is it clear that the response of Florida Bay to the CERP is uncertain, but the historical water quality and ecological conditions of the Bay are at least as uncertain. On the basis of presentations made to the CROGEE at its meetings, a restored Florida Bay appears to imply to the public and to many scientists clear water, presence of sea grass, good fishing, and ample opportunities for other recreational activities such as diving. But to what period of history do these conditions correspond? Knowledge of the Bay’s condition many decades to centuries ago is limited, and it would be useful to prepare as accurate a historical timeline as possible based on measured and anecdotal information. In this way managers, the public, and the scientific community may be made aware of the proposed future condition in relation to the recent and more distant past.

Numerical models are an essential research and management tool needed for integrating and synthesizing results from most real and hypothesized scenarios in the Everglades. Of particular interest for this report are models that may be used to link the hydrodynamic and ecological response of Florida Bay to changes in Everglades hydrology and to provide decision support and analysis tools to restoration planners.

Some success has been achieved in modeling Everglades hydrology. For example, the South Florida Water Management Model (SFWMM; http://www.sfwmd.gov/org/pld/hsm/models/sfwmm/) has been used since 1984 to simulate the hydrology and management of the water resources system of south Florida, and to evaluate hydrologic and hydraulic restoration options for the Everglades. It covers an area of 7600 square miles using a mesh of two-mile by two-mile cells, and discharges surface flows at the southern end of its area of simulation in the mangrove region of south Florida. A similarly structured model, the Natural System Model (NSM; http://www.sfwmd.gov/org/pld/hsm/models/nsm/), has been used to provide insight into a hypothesized “natural system” condition of the same domain.

However, less success has been achieved on water quality (including salinity) and ecological modeling, which will be required to evaluate the effect of CERP activities on Florida Bay. Typically, a transient, two- or three-dimensional hydrodynamic model is used to “drive” eutrophication and other water quality models that must be developed to evaluate the effects on the Bay of external forcing functions such as freshwater inflows from the Florida peninsula. Such circulation models are complex (Martin and McCutcheon, 1999), and extensive data collection activities and occasional modifications of the numerical code are required to adapt an existing model to a new location. Water quality/ecological models can be equally complex in their kinetic formulations, although their mathematics might not be quite as imposing.

Several efforts have been made to develop a circulation model for Florida Bay (e.g., Wang and Monjo, 1995; Sheng et al., 1995, 1996; Galperin et al., 1995; Roig, 1996; Davis and Sheng, 1996). While these studies were able to represent currents and salinity in the Bay with some success, their achievement was limited by the lack of long-term data necessary for development of a model that could be used operationally to evaluate management options. Characteristics of such a model would include a demonstration of its ability (through calibration with and comparison to monitoring data) to simulate circulation and salinity in response to forcing functions (e.g., tides, ocean boundary conditions, freshwater

inflows, precipitation, evaporation) over a period of years. At present, there is no model that has been adopted for management purposes, nor is there a consensus among agencies about which, if any, of the past studies might represent the most promising starting point for future model development; some of the reasons for this lack of consensus are presented by Hobbie et al. (1999).

The need for such a model is well recognized by Florida Bay and Florida Keys Feasibility Study (FBFKFS) scientists, and is emphasized by Hobbie et al. (2001), who suggest drawing from “community models” for this purpose. Community models are “…constructed by a small team of developers [and] made available to the community of users (with some reasonable restrictions) via downloading from the Internet” (Hobbie et al., 2001). Similarly, the FBFKFS might logically draw upon the “community” of those who have already performed Florida Bay modeling for advice during the model selection phase. It is important to note that some ocean models are not suitable for estuarine simulation because of their fixed orthogonal coordinate systems that cannot resolve the Bay’s complex geometry and because of their inability to simulate flooding and drying of tidal flats.

There are two requirements for successfully linking a hydrodynamic model with a water quality model. The first requirement is that the hydrodynamic model satisfy local and global mass conservation over the model grid. The two common numerical options for circulation models are finite difference models and finite element models (Martin and McCutcheon, 1999). The former are typically formulated for a rectangular x-y grid. When formulated using a control volume approach, the mass conservation requirement is guaranteed within a finite difference model. Finite element models commonly use triangular elements of varying size and are easier to adapt to complex geometries, although finite-difference methods may be adapted to curvilinear coordinates (e.g., Chau and Jin, 1995; Sheng et al., 1996) and thus mitigate this relative disadvantage. Finite element models sometimes suffer from local (individual element) problems of mass continuity, and extra effort may be required to formulate the numerical scheme in a manner to ensure local mass balance (Berger and Howington, 2002). Additional “flow projection” techniques may also be applied to finite element model output to ensure local and global mass conservation (Chippada et al., 1998; Riviere and Wheeler, 1999). Indeed, these flow projection methods appear useful for general interpolation of vector flow fields, e.g., from observational data.

The second requirement for linking such models is that the hydrodynamic model information be preserved in the water quality model. This is typically achieved by using the same model grid and time step for both the circulation model and water quality model. Both requirements lead to the conclusion that the same type of model (finite difference or finite element) should be used for both the circulation and quality objectives. In this way, constituent residence times and advective fluxes will be approximated to a higher order of accuracy, leading to concentration of constituents by evaporation, for example. Still another consideration is minimizing execution times for the combined circulation and water-quality model through means such as parallel processing so that long-term water quality simulations may be performed. The FBFKFS is expected to consider all these factors during its review of candidate models.

Notwithstanding the need for improved numerical techniques and hydrodynamic modeling, a key to successful water quality modeling in Florida Bay will be more reliable information on freshwater inflows and the nutrient balance (Berger, R.C. and Dortch, M.S., USACE Waterways Experiment Station, Personal commun., March 2002), as discussed in this report. Flow patterns are tied to salinity, which depends on freshwater inflows. The influence of freshwater nutrient inflows has been emphasized. Hence, circulation and water quality modeling will remain uncertain as long as such boundary influences are uncertain. Cerco et al. (2000) point out several other weaknesses in the knowledge of parameters and coefficients affecting water quality modeling of Florida Bay. In the event that a linked hydrodynamic-ecological model of Florida Bay proves elusive or lengthy in its development, the ability just to simulate salinity and turbidity and to track conservative constituents will still be useful for management purposes.

While waiting for an integrated hydrodynamic-water quality model to be developed, simpler techniques could prove useful. FATHOM (Flux Accounting Tidal Hydrology Ocean Model) is a coarse-grid “box model,” capable of long-term simulation (years), but dependent on simplifying assumptions about circulation (Nuttle et al., 2000; Cosby et al., 1999; Cosby et al., in prep.). Hobbie et al. (2001)

indicate that better estimates of inter-box exchange coefficients are now available that could make FATHOM more useful for initial guidance and management purposes and allow it to fill an analytical gap in the interim.

If an integrated hydrodynamic-water quality model can be developed for Florida Bay, the challenge will remain to link such a model with the SFWMM or similar model. Such an interface is vital for management purposes such as to study the effect of changes in the magnitude, timing, and spatial distribution of freshwater discharges from the Everglades on the Bay. The coarse, 2-mile by 2-mile grid, large time step, lack of tidal influence, and lack of density-driven subsurface flow features of the SFWMM present challenges to this task. However, at least two efforts are underway to interface the SFWMM (or a similar model) with a Bay model.

The first option is the USGS project “Tides and Inflows in the Mangroves of the Everglades (TIME)” (http://sofia.usgs.gov/projects/time/). This project is developing a linked surface-subsurface water flow model that responds dynamically to freshwater inflows, tides, salinities, rainfall, ET, and related forcing functions (Schaffranek et al., 2001). The goals include coupling with hydrologic modeling at the north end of the simulated region (Tamiami Trail) and quantifying the interrelation of freshwater and salt-water flows at the south end (mangrove region) of the study area. The effect of sea-level rise also can be studied with this model. The model currently is being calibrated against monitored water levels and freshwater fluxes at the Florida Bay interface. Monitoring difficulties include obtaining land elevations on mangrove islands, measuring the bathymetry of mangrove inlets, and defining tidal creek flows and salinities (Hittle et al., 2001). However, TIME has performed well in wetland regions of Everglades National Park where better monitoring data are available. This model probably will not be calibrated and tested sufficiently for use in a management setting until mid-2003, and the agencies have not agreed that this model will serve to bridge the gap between the hydrologic modeling and Florida Bay.

A precursor to the TIME model is the Southern Inland and Coastal Systems (SICS) model (http://sofia.usgs.gov/projects/sheet_flow/), developed for the Taylor Slough and C-111 drainage areas. It already has been used for management decisions regarding the effect of freshwater flows on the salinity of the northeast portion of Florida Bay. The SICS model has the same generic formulation as the TIME model and will be subsumed into the TIME model when the latter is fully operational. Interfacing of the TIME model to the SFWMM at the northern boundary depends primarily on resolution of the 2-mile by 2-mile SFWMM grid with the 500-meter TIME grid and should not be a major problem. However, the 500-meter grid and dynamic nature of the TIME model at the tidal boundary means that it runs on a short (15-minute) time step. Hence, TIME can be used only for episodic events, such as a period of drought or flooding and not for a longer portion of the 31-year SFWMM simulation. On the other hand, most estuarine circulation models are limited to simulating periods of at most several months or a few years duration.

A second modeling option for interfacing the hydrology of the CERP with the hydrodynamics of Florida Bay consists of ongoing model development by the SFWMD of its South Florida Regional Simulation Model (SFRSM) (http://glacier.sfwmd.gov/org/pld/hsm/models/sfrsm/index.html). Using a numerical approach similar to the TIME model, a linked surface-groundwater model is under development that will provide considerable additional spatial and temporal detail in the region south of the Tamiami Trail (Lal, 1998). Eventually, this model may replace the SFWMM over the whole south Florida simulation region, but according to the web site “years of development and testing will be needed before SFRSM becomes fully operational for the entire system.” Fortunately, initial implementation of the SFRSM will be in the area of Everglades National Park, the region of concern here. But the USGS, SFWMD, and Corps of Engineers have not reached consensus on how to effect the interfacing of hydrologic modeling and circulation modeling, and both the USGS and SFWMD modeling efforts are proceeding without such a mandate. Both efforts should prove useful to address many questions aside from just the interfacing issue but the three agencies should reach a consensus on the hydrologic model

most likely to serve as an interface between the overall hydrologic modeling of the CERP and circulation modeling in the Bay.

Statistical models (e.g., correlations between monitored freshwater flows and Bay salinity and algal conditions, time series analysis) may serve as a temporary bridge between CERP forcing functions and Bay response (Hobbie et al., 2001). Development of inferences from models of this type would be useful while the longer and more expensive effort to develop, calibrate and test Florida Bay circulation and water-quality models and hydrologic models linking the land and the Bay is undertaken.

Although the effects of CERP management decisions on the water quality and character of Florida Bay are the primary focus of this report, it is important to consider other influential factors, especially if model predictions prove to be highly inaccurate. For instance, the Bay has been shown to be influenced by runoff from as far away as the Mississippi River—albeit minimally and from a highly unusual hydrologic event—the major flood of 1993 (Ortner et al., 1995; Gilbert et al., 1996). Atmospheric deposition of nutrients derived from distant anthropogenic sources contributes to nutrient loads. Sea level and climate change effects may alter boundaries and bathymetry gradually in the long term while hurricanes could dramatically affect the Bay in the short term. The Florida Keys Tidal Restoration Plan (http://www.evergladesplan.org/pm/projects/proj_31.shtml), a component of the CERP, will restore tidal connections between Florida Bay and the Atlantic Ocean at four locations beneath US Route 1 and the Flagler Railroad causeway and will alter Florida Bay circulation patterns to an unknown extent. Future circulation modeling should help determine the potential effect of this action. The greatest unknowns, however, are probably those associated with the human response to the restoration and to patterns of growth in south Florida. The restoration and a reliable water supply will likely induce and facilitate population growth, thus placing additional, unexpected demands on natural resources such as the Everglades and Florida Bay. How much will such additional use degrade these resources and negate restoration efforts? Additional research and consideration of human factors are needed to illuminate these issues.

The management, coordination, and structure of ongoing research in the Bay is generally good. While it is true that some of the issues raised in this report (e.g., the need to establish restoration goals for the Bay) were raised during earlier reviews such as Boesch et al. (1996), knowledge of the Bay has increased greatly since the late 1980s. The extension of the CERP to include the Bay as well as recent activities of the Program Management Committee (PMC) should accelerate progress in addressing these important questions. As described in Chapter 1, several agencies and universities collaborate on research directed toward answers to five central questions developed by the Program Management Committee (PMC). Whether or not research is directed toward these questions and whether or not it leads to synthesis and usefulness for managers are the primary criteria affecting recommendations for new and continuing support from the PMC—an effective strategic plan for research management. All research is peer reviewed, at least through the biannual Florida Bay Science Conferences as it is prepared for peer-reviewed journals. The effectiveness of each conference is evaluated by the Florida Bay Science Oversight Panel (e.g., Hobbie et al., 2001). The oversight and coordination exhibited by these several groups are commendable.

A related issue is one of adequate time to perform the needed research. For instance, the FBFKFS is scheduled for four years, including approximately one year at the beginning devoted mainly to planning and a final year devoted mainly to preparation of documentation of the study. This leaves only about two years to perform the bulk of the research, including complex hydrodynamic and water-

quality modeling in Florida Bay. It is important that enough time be available for essential research and that project efforts not be rushed unreasonably at the expense of science.