1 December 1998

Mr. Stuart Appelbaum
Project Manager, C&SF Project Restudy
U.S. Army Corps of Engineers
P.O. Box 4970
Jacksonville, Florida 32232-0019

Dear Mr. Appelbaum;

We commend the efforts of the Corps of Engineers and the South Florida Water Management District in the Central and South Florida Project Comprehensive Review Study (Restudy). The results establish a strategy for restoring the altered hydrology of south Florida toward its natural state. This sets the stage for restoration and conservation of the ecosystems in south Florida, particularly the Everglades and Florida Bay.

By this letter, the Florida Bay Program Management Committee (PMC) provides its comments on the draft report (USACE, 1998) as part of the public consultation phase of the Restudy. The PMC is comprised of members from 14 federal and state agencies, and its role is to oversee and direct research in Florida Bay and adjacent marine systems. These comments address, in part, the anticipated benefits for Florida Bay of the hydrological restoration described in the draft report. Critical gaps remain in our understanding of marine ecosystems in south Florida and their links to hydrologic conditions and water quality in the southern Everglades. The comments also identify where gaps limit our ability to predict with confidence the ecological benefits that will accrue from hydrological restoration. These are areas in which work is most needed as the Restudy process moves towards implementation.

The comments of the PMC are summarized below, with details provided in the attached report:

• In general, the proposed restoration plan promises to benefit the ecological health of Florida Bay. However, the effects of the plan are less clear for Biscayne Bay and the Florida Keys National Marine Sanctuary.
• Given the uncertainties of the method used to assess the impacts on Florida Bay, the report may overstate the differences among the various restoration alternatives with respect to their expected effects on Florida Bay.
• The regression equations relating salinity at coastal locations to water levels at P33 are neither robust nor reliable measures of ecological conditions in Florida Bay. Therefore, managers should not rely on these in the detailed planning leading to implementation of restoration.
• The PMC will develop a detailed set of ecological performance measures and restoration targets for Florida Bay in the next 12 months.
• Existing hydrologic models must be refined so that they provide more reliable predictions of volume and location of flows into Florida Bay, Biscayne Bay and across the southwest Florida coast.
• The proposed restoration plan must be sufficiently flexible to accommodate the remaining uncertainty in the freshwater flow and water quality requirements of the marine ecosystems of south Florida.

The high priority given to the Florida Bay Feasibility Study in the Project Implementation Schedule demonstrates that the need to close critical gaps in our understanding of Florida Bay is generally recognized. Results from the research directed by the PMC will help to establish the freshwater flow and water quality requirements of Florida Bay and adjacent marine ecosystems. The PMC looks forward to working more closely with the Restudy team in the coming months.

Sincerely,

cc:
Working Group of the South Florida Restoration Task Force
Governor's Commission for a Sustainable South Florida
Southern Everglades Restoration Alliance
Florida Bay Program Management Committee
Florida Bay Science Oversight Panel

In summary, the proposed hydrologic restoration plan described in the draft report, i.e. Alternative D13R, promises ecological benefits to Florida Bay. This conclusion is based entirely on the anticipated increase in water levels at station P33 in central Shark Slough. The statistical relationships between historical water levels here and salinity in Florida Bay are consistent with the hypothesis that water flows, and levels, in Shark Slough influence hydrological conditions in the entire region, including the near-shore areas of Florida Bay. On the basis of this hypothesis, an increase in water levels in Shark Slough implies that flows of freshwater to Florida Bay will increase, in general. This increased flow will offset, to some unknown degree, the decreases in freshwater supplied to the bay in recent decades that have contributed to its ecological decline. However, other factors, such as local rainfall, also influence salinity in Florida Bay, and this introduces uncertainty into the assessment of the effects of hydrologic restoration.

Without an analysis of the effects of this uncertainty, it is not possible to judge whether distinctions made between various restoration alternatives are real or simply an artifact of the uncertainty and the weighting scheme used to integrate the results of different performance measures. For example, the comparison of alternatives illustrated in Table 7-8 and 7-9, and described in Appendix D, takes no account of the uncertainties in the salinity predictions. Similarly, there is doubt about the improvement in Florida Bay that is claimed for alternative D13R compared to alternative D13, Tables 7-12 and 7-13.

Managers must recognize that these relationships are not sufficiently robust to use in the detailed planning during the implementation phase. Use of statistical relationships to predict the future response of salinity in Florida Bay can be done only so long as the regional hydrology remains unchanged from the historical period for which the relationships were calibrated. Changes planned as part of the hydrological restoration raise doubts about the continued validity of these relationships. For example, we note that several elements of the proposed plan would change critical aspects of water flow in the area between Shark Slough and Florida Bay. The distribution of flows into Shark Slough across Tamiami Trail will be shifted from west to east. Changes are planned in the canals and levees to reduce the loss of water by seepage along the eastern boundary of Everglades National Park. Further, work has recently been completed to alter the levees along the C111 canal, which is in addition to the changes in the proposed plan.

Limitations of the salinity relationships point directly to the need to develop a more refined set of assessment tools for Florida Bay. In addition, assessment tools are needed for Biscayne Bay and the Florida Keys National Marine Sanctuary. These tools should be flexible enough to accommodate anticipated changes in water flows and water quality provided to the natural systems in south Florida and precise enough to meet managers’ planning needs. The comments that follow identify the source of limitation in the current assessment tools.

Lack of confidence in the hydrological model imposed a bottleneck on the assessment of the effects of restoration alternatives on Florida Bay. In the planning and assessment process, the South Florida Water Management Model (SFWMM) was used as the link between the description of restoration alternatives and assessment of their ecological consequences. Output from hydrologic simulations for each alternative, i.e. time series of simulated water levels and flows, provided input into the ecological assessments. With respect to Florida Bay, the modelers expressed a general lack of confidence in the simulated water levels and flows close to the edge of the model domain, adjacent to Florida Bay. Ecologists on the Alternative Evaluation Team (AET) responsible for Florida Bay were discouraged from developing performance measures and targets based on the simulated hydrologic response at these coastal locations. Instead, the final assessment of the implications on Florida Bay is based on water levels from one location upstream in Shark Slough, P33, and regression equations that relate the variation in salinity at various coastal locations to these water levels. Other heuristic measures were used by the AET from time to time, but the draft report relies solely on the quantitative assessment based on the salinity regression equations.

Other sources of information about the hydrologic models reinforce the modelers’ caution in the use of simulated hydrology close to the coastal boundary. First, comparison of observed hydrologic conditions near the headwaters of Taylor Slough with conditions simulated by the model indicates that the model provides a poor description of this hydrologic feature. This finding is particularly significant because Taylor Slough is a principal source of surface water discharge into Florida Bay. Second, problems with predictions of the hydrologic model near Florida Bay is evident in results of the SFWMM calibration completed just prior to the latest phase of the Restudy (South Florida Water Management District 1997). Very large uncertainties in water levels are reported for the calibration point located at S197, near the southeast boundary of the model.

With respect to predicting the response of Florida Bay, the southern boundary of the hydrologic simulation model effectively occurs at P33, a point that is about 30 miles inland from the bay. Overall, the regression-based salinity models make perhaps the best use of the information available within the overall constraints imposed by the accuracy of the hydrology models and the time available for the assessment of alternatives. However, the result is that the assessment of restoration alternatives essentially ignored all differences among alternatives in the resulting hydrology of the areas south and east, except to the degree to which these are reflected in the simulated water levels for P33.

Criticism of the hydrologic models does not ignore the present deficiencies in our understanding of ecology of marine systems in south Florida. Over the next 12 months, the PMC will develop an improved set of performance measures and restoration targets for Florida Bay based on the results from recent research. Specifically, the initial development of circulation and mixing models of the bay will extend our capacity to assess the effect of augmented flows on salinity throughout the bay. Synthesis of paleoecological studies will provide information on the salinity regime in the bay prior to development. We also expect progress in understanding the factors contributing to seagrass die-off and the occurrence of plankton blooms. However, improvement in the performance measures for Florida Bay depends on continued refinement of hydrological models in coastal portions of the Everglades.

Whatever the specific nature of the performance measures may be, we anticipate that they will depend on 1) the volume of freshwater flow entering the bay, 2) the spatial and temporal distribution of freshwater flows along the coast, and 3) their impact on water quality in the marine ecosystems. The objective set out by the Governor’s Commission Conceptual Plan for the Restudy should guide development and/or refinement of hydrologic models in the coastal areas of south Florida. That is, the models should "be able to predict volume and location of flows across the mangrove zone" (page 6-35). We are aware of at least two projects working in this area, i.e. the development of the FEMWATER model by the Corps of Engineers and development by the USGS/WRD of SWIFT2D, a detailed hydrology model for the area of the Everglades north of Florida Bay. These modeling efforts must be coordinated with the efforts to model hydrodynamics and water quality in Florida Bay and adjacent marine systems. In addition, they must integrate improved topography data, and calibration and validation of the models should make use of long-term data on water levels along the southwest coast that are being collected by the USGS/BRD project on mangrove ecology.

1) Throughout the draft report, reference is made to operational flexibility, which is built-in to key components, as a strategy to accommodate the unexpected in the final design, implementation and operation of the restoration plan. It should be evident that a great deal of uncertainty remains about the effects of the proposed restoration plan on Florida Bay. However it is not clear whether or not the benefits of the operational flexibility built into components far upstream in the watershed can propagate downstream and provide the flexibility in flows needed to offset the uncertainties in the effects on Florida Bay. Is it possible to quantify the amount of adjustment in flows to Florida Bay and Biscayne Bay that can be achieved through flexible operation of the altered C&SF Project?

2) We take issue with the response by the Restudy team to DOI’s recommendation on flow to Florida Bay (Recommendation #13a of the U.S. Fish and Wildlife Service draft Fish and Wildlife Coordination Act Report Recommendations, page 12-6). The response in the draft report argues that a variety of different hydrological targets in planning for restoration, such as "hydroperiod duration and stages," should be used "rather than a single parameter such as flow volume." The ecological restoration of the Florida Bay estuary depends most directly on restoring the volume of freshwater flow, its seasonal timing and spatial distribution along the boundaries of the bay. Estuaries differ from wetlands in this respect. Hydroperiod and stages at inland locations are only indirectly, and imperfectly, related to the flow into Florida Bay. Attention to predicting flow crossing the coast is needed in order to improve the tools for assessing the effects of hydrologic restoration on Florida Bay, Biscayne Bay and the Florida Keys National Marine Sanctuary. For example, a given volume of freshwater will have a different effect on Florida Bay if it is discharged through Shark Slough than if it is discharged directly into the bay through Taylor Slough. Allocation of flow to the southwest coast, rather than to Taylor Slough or the C111 basin, also has implications for freshwater flows reaching Biscayne Bay.

3) We reiterate the recommendation by DOI (Recommendation #15 of the U.S. Fish and Wildlife Service draft Fish and Wildlife Coordination Act Report Recommendations, page 12-9) for improvements to the hydrologic models, particularly as related to Florida Bay. This will require improved information on water balance fluxes, particularly evapotranspiration in the C111 basin, and data on surface and groundwater levels for model calibration, in addition to the better topographic data identified in DOI’s recommendation. The response by the Restudy team should describe current efforts to develop refined hydrology models adjacent to Florida Bay, i.e. FEMWATER and SWIFT2D, the integration of new data with these models, and coordination with complementary modeling efforts in Florida Bay itself.

4) The draft report incorrectly uses the units of "parts per trillion" when referring to salinity for Florida Bay on pages 3-11 (first paragraph) and 3-23 (second full paragraph). The correct units are "parts per thousand" for the magnitude of the values stated.

South Florida Water Management District, 1997. Draft Documentation for the South Florida Water Management Model, pp 239.

U.S. Army Corps of Engineers, 1998. Draft Integrated Feasibility Report and Programmatic Environmental Impact Statement, pp 3511.