6
Building the Foundation for Adaptive Management

When the U.S. Congress approved the Comprehensive Everglades Restoration Plan (CERP) in the Water Resources Development Act of 2000 (WRDA 2000), there was clear recognition that the Central and South Florida Comprehensive Review Study (or Yellow Book; USACE and SFWMD, 1999) provided only a general outline for restoration of the Everglades and not a detailed restoration plan. Considering the need to move forward in the face of some uncertainties, the Yellow Book proposed an adaptive management plan founded on a comprehensive ecosystem monitoring program. This approach provides a mechanism for emerging scientific information to be incorporated into the plan and for unforeseen consequences of the restoration project to be addressed. Congress subsequently approved funding for an Adaptive Management and Monitoring program in WRDA 2000 and the Programmatic Regulations (33 CFR §385.31) directed the U.S. Army Corps of Engineers (USACE) to adopt an adaptive management approach.

In this chapter, progress toward building the necessary foundations for adaptive management is described, and major recent monitoring and assessment reports and associated issues are reviewed, building on previous National Research Council (NRC) reports (NRC, 2003b, 2007). The committee is specifically charged to review monitoring and assessment protocols and progress (see Chapter 1) because monitoring and assessment are essential for evaluating CERP progress. The chapter begins with an overview of the concept of adaptive management, and progress in creating and applying an adaptive management framework for the CERP is discussed. Next, progress in developing the monitoring and assessment plan is reviewed as a critical component of the adaptive management process. Specifically, the CERP System-wide Performance Measures report (RECOVER, 2007b) and the first full analysis of the status of the South Florida ecosystem (the 2007 System Status Report; RECOVER, 2007c) are discussed in detail. Previous NRC reports (NRC, 2003b, 2007) reviewed the monitoring plan titled Monitoring and Supporting Research (RECOVER, 2004).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 189
6 Building the Foundation for Adaptive Management When the U.S. Congress approved the Comprehensive Everglades Resto- ration Plan (CERP) in the Water Resources Development Act of 2000 (WRDA 2000), there was clear recognition that the Central and South Florida Compre- hensive Review Study (or Yellow Book; USACE and SFWMD, 1999) provided only a general outline for restoration of the Everglades and not a detailed restoration plan. Considering the need to move forward in the face of some uncertainties, the Yellow Book proposed an adaptive management plan founded on a comprehensive ecosystem monitoring program. This approach provides a mechanism for emerging scientific information to be incorporated into the plan and for unforeseen consequences of the restoration project to be addressed. Congress subsequently approved funding for an Adaptive Management and Monitoring program in WRDA 2000 and the Programmatic Regulations (33 CFR §385.31) directed the U.S. Army Corps of Engineers (USACE) to adopt an adaptive management approach. In this chapter, progress toward building the necessary foundations for adaptive management is described, and major recent monitoring and assess- ment reports and associated issues are reviewed, building on previous National Research Council (NRC) reports (NRC, 2003b, 2007). The committee is spe- cifically charged to review monitoring and assessment protocols and progress (see Chapter 1) because monitoring and assessment are essential for evaluating CERP progress. The chapter begins with an overview of the concept of adaptive management, and progress in creating and applying an adaptive management framework for the CERP is discussed. Next, progress in developing the monitor- ing and assessment plan is reviewed as a critical component of the adaptive management process. Specifically, the CERP System-wide Performance Mea- sures report (RECOVER, 2007b) and the first full analysis of the status of the South Florida ecosystem (the 2007 System Status Report; RECOVER, 2007c) are discussed in detail. Previous NRC reports (NRC, 2003b, 2007) reviewed the monitoring plan titled Monitoring and Supporting Research (RECOVER, 2004). 189

OCR for page 189
190 Progress Toward Restoring the Everglades Information management and the status of hydrologic and ecological models, which are essential to a well-functioning adaptive management process, are also discussed in this chapter. ADAPTIVE MANAGEMENT Adaptive management facilitates natural resource management or environ- mental restoration activities when uncertainty about the potential outcomes of management actions is present (NRC, 2007). It offers a means to proceed with- out a fixed design and to reduce uncertainty through the iterative refinement of management actions, ideally based on experimentation (Lee, 1999; Walters and Holling, 1990). Of the many applications of adaptive management, the most- effective ones have well-structured processes that include: 1. management objectives that are regularly revisited and accordingly revised, 2. a model or models of the managed system, 3. the monitoring and evaluation of outcomes, 4. mechanisms for incorporating what is learned into models guiding future decisions, and 5. a collaborative process for stakeholder participation and learning (NRC, 2004). Through effective application of such a process, decision making moves from a trial-and-error process to one of experimentation based on continuous monitor- ing, assessment, and reevaluation. Since its inception, the CERP has taken an adaptive management approach to address uncertainty in the restoration. This approach is mainly passive, entail- ing detailed scientific analysis, planning, monitoring, and assessment, combined with feedback to restoration design and operation (see Figure 6-1). In some cases, however, active (i.e., experimental) rather than passive adaptive management may better assist in achieving restoration goals, because substantial uncertainties remain about the degree to which a resilient, self-sustaining ecosystem can be restored under the dramatically changed environment of South Florida. Oppor- tunities for active adaptive management experiments are numerous (NRC, 2007), and the CERP’s Decompartmentalization (Decomp) project in particular is an example of a project that would likely benefit greatly from the active adaptive management approach described in Box 6-1. Opportunities also exist for incre- mental adaptive restoration, which was conceived by the previous committee (NRC, 2007) as a way to advance restoration in the face of contentious uncer-

OCR for page 189
Building the Foundation for Adaptive Management 191 FIGURE 6-1 The CERP Adaptive Management Strategy. Figure 6-1.eps SOURCE: Adapted from RECOVER (2006a). bitmap tainties and sequencing constraints, while utilizing active adaptive management to resolve critical uncertainties and improve future project planning (see also Chapter 3). The Restoration, Coordination, and Verification (RECOVER) team has fleshed out many dimensions of CERP adaptive management, including extensive work to create monitoring and assessment plans and protocols that are discussed later in this chapter. In 2006, RECOVER published the CERP adaptive management strategy, which outlines a framework for linking monitoring and assessment activities to management decisions and to plan updates and revisions at both system-wide and project levels (Figure 6-1; RECOVER, 2006a). As required by the Programmatic Regulations, the USACE and South Florida Water Management District (SFWMD) also have drafted a Guidance Memorandum for Assessment and Adaptive Management that recapitulates and elaborates upon the CERP adaptive management strategy (USACE and SFWMD, 2007a). The Guidance Memorandum lays out a framework for preparing CERP technical reports and for recommending changes or adaptive management actions as needed at

OCR for page 189
192 Progress Toward Restoring the Everglades BOX 6-1 Decomp Adaptive Management Plan The Water Conservation Area 3 Decompartmentalization and Sheet Flow Enhance- ment project (Decomp) has been described as “the heart of Everglades restoration” because of the tremendous ecological benefits the project provides (USACE and SFWMD, 2002). The objective of Decomp is to reestablish hydrologic connectivity between WCA-3A, WCA-3B, and Everglades National Park such that hydropatterns in these regions better approximate those experienced in these areas historically. Decomp project planning is well behind schedule due to ongoing stakeholder con- flicts (e.g., disagreements over the need to completely fill canals and thereby eliminate bass fishing habitat) and constraints in the project planning process (Light, 2006; NRC, 2007). To expedite progress the U.S. Army Corps of Engineers (USACE) and the South Florida Water Management District (SFWMD) are developing the Decomp project in three separate project implementation reports and a “Decomp physical model” (DPM), which combines features of a pilot study with an active adaptive management experi- ment. USACE preferred the term physical model over experiment because the latter implies research, which is not in the USACE purview. Because the specific plans are still under development, this section focuses on the general direction being taken with the Decomp adaptive management plan. The plan to decompartmentalize the water conservation areas has raised a number of thorny scientific questions: What is the relationship between the extent of sheet-flow restoration and the rate and degree of ecosystem recovery? What are the effects of partial versus extensive backfilling of canals? How will levee degradation impact adja- cent areas? How will tree islands respond to different water depths and hydroperiods? To date this scientific uncertainty has delayed restoration because interest groups have challenged the credibility of proposed modifications. Decomp adaptive management plan has been designed to reduce scientific uncertainty and to accommodate stake- holder concerns. The program combines gathering of baseline information, field trials of alternative decompartmentalization approaches (through the DPM), monitoring and assessment, and feedback to the agencies and stakeholders. Currently the DPM consists of two proposed experiments that will be installed along the L-67 levee connecting WCA-3A to WCA-3B: the before-after control impact (BACI) flow way and the repeated measure flow way. The BACI flow way experiment would compare geomorphic and ecological effects associated with three canal-backfilling designs (complete, partial, and no backfill) installed in three contiguous 3,000-ft gaps in L-67C (see Figure 4-3). Water would be supplied by a 200-ft (61-m) gap located 2 miles upstream in L-67A. Although the experiment involves no replication, the BACI approach is intended to provide statistical validity to the results: outcomes will be compared to project, subregional (or module), and system-wide levels of the restoration. The memorandum also provides guidance on reporting standards and expectations for peer review. The draft CERP Adaptive Management Implementation Guidance Manual (RECOVER, 2007e) provides clear technical guidance to project delivery teams

OCR for page 189
Building the Foundation for Adaptive Management 193 trends in adjacent areas without levee breaches. Field studies are scheduled for 3 years. The repeated measure flow way consists of three 3,000-ft partially backfilled gaps in L-67C intended to measure local variation in the ecological effects of this restoration alternative (Sklar, 2007b). The DPM designs entail field-scale engineering projects that may significantly reduce uncertainty while advancing Decomp design and implementation. Large-scale changes in the water management system supported by information from the DMP experiment provide opportunities for learning as well as providing some, albeit fairly small, eco- logical restoration benefits. This approach is consistent with the incremental adaptive restoration approach as suggested by a previous NRC committee (NRC, 2007). The experimental design reflects and responds to operational, fiscal, and political realities. Concerns about the ability of the DPM to fully address scientific uncertainties associated with Decomp include: 1. The current design does not address the relationship between the extent of sheet-flow restoration and ecological restoration; 2. Management constraints may prevent moving enough water through the system to allow fair comparison of the ecological implications of complete versus partial canal backfilling; 3. The proposed design may not accommodate infrequent high flows that could have an important structuring role in the Ridge and Slough ecosystem; and 4. The 3-year assessment period may not allow sufficient time to distinguish treat- ment effects. These concerns are well recognized by RECOVER scientists, and final plans for the DPM are evolving. CERP scientists hope that the scientific uncertainties associated with Decomp can be greatly reduced by the DPM such that the contentious issues regarding the project design will be lessened or even resolved. Of much graver concern to implementing Decomp are the political uncertainties associated with completion of the Tamiami Trail component of the Modified Water Deliveries to Everglades National Park project (Mod Waters) (Chapter 4). Currently, it is not clear if Mod Waters ever will be completed as described in the authorizing legislation, or how the recommended Tamiami Trail modifi- cation plan will impact the ecological benefits that can be achieved by Decomp. on whether and how to apply adaptive management at the project level. The manual also provides a brief discussion of system-wide implementation of adap- tive management, promising a more detailed treatment of this topic in future versions. The manual sets out conditions for employing passive versus active management and advocates incremental adaptive management (NRC, 2007)

OCR for page 189
194 Progress Toward Restoring the Everglades as the preferred active adaptive management approach when there is stake- holder gridlock or political pressure for quick restoration benefits (RECOVER, 2007e). Previous NRC reviews of CERP adaptive management concluded that the MAP is generally scientifically strong and that the adaptive management strategy provides a sound organizational basis for passive adaptive management (NRC, 2003b, 2007). Nevertheless, some remaining scientific and institutional chal- lenges need to be squarely addressed, including improved analytical methods and ecological models for upscaling local monitoring data to subregional and system-wide evaluation, stronger coupling of hydrologic and ecological monitor- ing, and the need for a mechanism to document how science has been used in support of decision making. These are difficult scientific and technical challenges that face all large-scale wetland restoration efforts. Social and institutional issues affecting the effectiveness of CERP adaptive management may ultimately prove less tractable than scientific issues. To date, the pace and scope of restoration have been driven more by special interest groups and the need to resolve stakeholder conflicts (for examples, see Box 4-1) than by the desire to learn (Light, 2006), and traditional command-and-control plan- ning and governance structures have prevailed over more adaptive approaches (Gunderson and Light, 2006). Stakeholder and collaborative relationships appear to be fraying as the CERP suffers continued delays (NRC, 2007). Agency com- mitment to continued long-term funding of the monitoring and assessment plan is not assured. Moreover, as discussed in detail in Chapter 3, project-by-project implementation of the plan is laden with planning and review processes that, although well-intended, not only slow restoration but also may operate against effective adaptive management. For instance, analyses of various project alter- natives as part of the project implementation report development process (see Chapter 3, Box 3-1) rely on traditional performance-based assessments that do not account for uncertainty or the potential for learning under alternative designs. This conventional “management for optimality” operates against adaptively managing for resilience and surprise (Gunderson and Light, 2006). CERP MONITORING AND ASSESSMENT The monitoring and assessment plan is the foundation of the CERP adap- tive management program. The CERP monitoring and assessment plan provides the framework that the RECOVER teams use to measure and understand the ecosystem’s responses to the CERP and to help determine how well the CERP is meeting its goals and objectives (RECOVER, 2004). Using a “performance assess- ment” process, the information generated from the monitoring and assessment

OCR for page 189
Building the Foundation for Adaptive Management 195 plan provides information required to make informed decisions about the need to alter restoration plans through the adaptive management process (Figure 6-1). Thus, without effective, long-term monitoring and assessment, adaptive manage- ment cannot occur (NRC, 2004). Monitoring and assessment also can support CERP project planning, design, implementation, and operation. Performance assessment (Box 2 in the adaptive management strategy; Figure 6-1) involves determining how the ecosystem is responding to restora- tion activities. Performance assessment consists of three activities: evaluation, monitoring, and assessment. Evaluation uses a variety of modeling approaches (discussed later in this chapter) and existing data and known relationships among variables to project potential ecosystem changes due to restoration activities. Monitoring is the gathering of data for one or more variables that can be used to generate quantitative indicators (or performance measures, described in more detail in the next section) that compress information about complex, related phenomena. Assessment in an adaptive management framework is not simply a report of the status of individual performance measures; it involves comparing the values of groups of linked performance indicators with known and expected values that are based on baseline data, modeling, or extensive literature review to determine the actual response of the system to manipulations, including CERP restoration actions. The CERP monitoring and assessment plan is founded on conceptual eco- logical models that are an assembly of logical hypotheses that describe the relationships among societal actions, environmental stressors, and ecosystem characteristics and the linkages among the physical, chemical, and biological elements within the natural system. They identify key driving factors, processes, stressors, and functional relationships based on extensive reviews of scientific data (Ogden et al., 2005). The conceptual models cover 11 physiographic regions and the entire South Florida ecosystem (i.e., the total system conceptual model) (Figure 6-2a). For assessment purposes, these have been combined into four geographic modules (see Box 6-2 and Figure 6-2b) and two modules that relate to the total system (the South Florida Hydrology Module and the South Florida Mercury Module). These conceptual models are used to inform research, plan- ning, operational management, and the performance assessment aspects of the adaptive management plan. The monitoring aspects of performance assessment are described by three RECOVER documents (see Table 6-1): 1. Monitoring and Supporting Research Plan (MAP I) (RECOVER, 2004), which establishes the monitoring and assessment plan; 2. Development and Application of Comprehensive Everglades Restora-

OCR for page 189
196 Progress Toward Restoring the Everglades A B FIGURE 6-2 Boundaries of the 11 conceptual ecological models (A) and the four monitoring and assess- ment modules (B). Figure 6-2.eps 2 bitmaps (vector A & B) SOURCE: RECOVER (2004). tion Plan System-wide Performance Measures (Performance Measures Report) (RECOVER, 2007b), an analysis and justification of monitored performance measures; and 3. Draft Quality Assurance Systems Requirements (QASR) (RECOVER, 2007d) containing quality assurance and data management protocols (see Box 6-3). All aspects of monitoring including what is monitored and why (MAP I and Performance Measures Report), the spatial and temporal extent of the monitoring (MAP I), and the methods used to collect the data and assure data quality (QASR) are described in these documents. A framework for analyzing monitoring data and evaluating the restoration progress is described in a fourth

OCR for page 189
Building the Foundation for Adaptive Management 197 BOX 6-2 Monitoring and Assessment Plan Module Groups The monitoring and assessment plan performance measures are grouped into four geographic modules (Figure 6-2b) and two additional modules for hydrology and mercury bioaccumulation monitor- ing. The modules are the organizing elements and research units of the monitoring and assessment plan. Each of the modules includes one or more of the conceptual ecological models briefly described in the text and in RECOVER documents (RECOVER, 2004, 2006f, 2007c) and in previous NRC reports (2004, 2007). The modules are designed to test the conceptual model working hypotheses as the CERP is implemented. The module groups are teams of scientists with technical expertise in ecology, hydrology, water quality, and human systems. The module groups oversee the monitoring and research programs in the four major regions of the Everglades and carry out the performance of assessment. • Northern Estuaries (NE) Module Group includes St. Lucie/Indian River Lagoon-South, Caloosahatchee Estuary, Lake Worth Lagoon, and Loxahatchee River Estuary. • Greater Everglades Wetlands (GE) Module Group includes the ridge and slough, southern marl prairies, mangrove estuaries, and Big Cypress swamp regions. • Southern Estuaries (SE) Module Group includes Florida and Biscayne Bays, and the South- west Florida Coast. • Lake Okeechobee (LO) Module Group. • South Florida Hydrology Monitoring module is designed to assist in evaluating water supply and protection for urban and agricultural areas. • Mercury Bioaccumulation. SOURCES: RECOVER (2004, 2006f). TABLE 6-1 Activities and Components of the Monitoring and Assessment Plan Component Purpose Status Monitoring and Supporting Research Describes monitoring plan and Final/January 2004; (MAP I) research justifying plan. Performance Revision of MAP I is (RECOVER, 2004) measures are identified. Focus is the planned for 2008 natural system. Implementation plan is included. Assessing the Response of the Describes the guidelines and Final/December 2006 Everglades Ecosystem to Implementation procedures used to synthesize of the CERP (Assessment Report) monitoring data to assess ecosystem (RECOVER, 2006e) response to the CERP. Development and Application of Describes performance measures Final/October 2007 Comprehensive Everglades Restoration monitored; baseline information. Plan System-wide Performance Measures (Performance Measures Report) (RECOVER, 2007b) Draft Quality Assurance Systems Describes quality assurance protocols Final Draft (peer-review Requirements (QASR) for all performance measures. Includes comments incorporated)/ (RECOVER, 2007d) some information on data validation, October 2007 management, and data archiving.

OCR for page 189
198 Progress Toward Restoring the Everglades BOX 6-3 Quality Assurance Systems Requirements (QASR) Manual The QASR document (RECOVER, 2007d) is an extensive manual developed and periodically updated by the CERP to ensure the quality of the data collected during both research and monitoring activities. It represents the mechanism by which the CERP can ensure that all data collected through the multiple agencies, projects, evaluation and assessment activities all adhere to the same standards and requirements of planning, performance, analysis and data archiving, and that data are comparable among compo- nents of the CERP. Separate sections of the report are devoted to detailed discussions of methods and quality assurance procedures for monitoring water quality, water quan- tity, hydraulics, soils and sediments, hydrometeorology, and biological entities. Other chapters specify quality assurance procedures and recommended analytic methods for laboratory chemical determinations and remotely sensed data. The document also addresses research and monitoring administration and lines of responsibility, require- ments for data management and archiving, and data quality evaluation and assessment. Extensive appendices list specific EPA and other official sources of methods, and methods established in the peer-reviewed literature for a large number of analytes in water, soils, and sediments, for animal and plant groups (e.g., fish, amphibian, birds, macroinvertebrates, plants), for use of remote sensing data, and for methods of data analysis such as methods of measurement and calculations of flux rates. The set of documents constitutes perhaps the most comprehensive, thorough and wide-ranging summary of specific methods in existence, as it integrates over hydrology, meteorology, soil science, environmental chemistry, animal, plant, and microbial ecology, and remote sensing, as well as data management and assessment protocols. The QASR also includes a very detailed protocol for documenting sample acquisition, analysis methods, and QA/QC analysis. Data custody is assigned to both the agency collecting the data and the Information and Data Management Program of the CERP, which has created the CERPZone Infrastructure to foster data availability and sharing among CERP par- ticipants. If the standards and methods for data acquisition, verification and quality assurance, and data archiving and sharing can be fully implemented as described in this document, it will itself be a striking achievement of the CERP. document—Assessing the Response of the Everglades Ecosystem to Implemen- tation of the CERP (Assessment Report) (RECOVER, 2006e). Collectively, these four documents (see Table 6-1) constitute the monitoring and assessment plan. The first full implementation of the assessment strategy is described in the 2007 System Status Report (RECOVER, 2007c). The following sections reflect the committee’s evaluation of two major monitoring and assessment reports released between 2006 and 2008: the Perfor- mance Measures Report (RECOVER, 2007b) and the 2007 System Status Report (RECOVER, 2007c). The recently finalized Quality Assurance Systems Require- ment Manual (Recover, 2007d) is also discussed in Box 6-3.

OCR for page 189
Building the Foundation for Adaptive Management 199 Performance Measures Because of the complexity of ecosystems, there is no single, simple variable whose response can be used to answer questions about the status of ecosystems, their changes over time, and their responses to management. Instead, indicators are developed that have a suite of properties that support their use for manage- ment (Niemi and McDonald, 2004; NRC, 2000). The Performance Measures Report (RECOVER, 2007b) supports the monitoring plan by describing the set of indicators (referred to as performance measures; see Appendix F) used to deter- mine the effects of CERP implementation. A performance measure may involve a single variable or, more commonly, is developed as a construct of the combined response of several readily measurable environmental attributes (e.g., population size, water nutrient concentration, water flow rates and durations). The choice of directly measured variables is justified by a review of the scientific literature demonstrating a causal relationship between the measured variables and the desired environmental outcome(s) of interest (restoration target). For example, mangrove forest production and soil accretion is an important performance mea- sure of the food webs critical to fisheries production in Florida Bay and wading bird populations in the Everglades. To assess the condition of this indicator it is necessary to measure variables such as mangrove species composition, canopy density, root growth, decomposition of leaves and roots, sediment deposition, and wetland surface elevations. Performance measures have been developed for both indicators of eco- system condition and for critical stressors on the ecosystem (e.g., estuarine salinity, soil and water phosphorus concentrations, hydropatterns). This allows the evaluation and assessment processes to focus on the current understanding of cause-and-effect relationships. This is a great strength of the performance measure system, because the cause-and-effect view of ecosystem dynamics is crucial for implementing an adaptive management approach. Performance measures play a key role in the restoration because they are used in the planning phase to evaluate project benefits at local and regional scales (termed “evaluation” in CERP documents), and they also are intended to be used to assess the effectiveness of projects once they are implemented (termed “assessment” in CERP documents). In project evaluation, hydrologic and ecologi- cal models are used to project the effects of various alternative project plans on performance measures. CERP performance assessment using the performance measures is described in more detail later in this chapter. Some performance measures might also serve as potential “report card” indicators. Report card indi- cators are key indicators used to inform the public about how the natural system is responding to restoration efforts (e.g., the abundance and nesting success of

OCR for page 189
212 Progress Toward Restoring the Everglades monitoring budget is challenging. If monitoring data are collected for too many performance measures, resources (money and manpower) are wasted for little gain in information. However, if the monitoring budget is cut too severely, critical information that is needed to guide adaptive management will be lost. Moni- toring that is transferred from the individual projects to RECOVER will stretch an already thin monitoring budget even more. This will inevitably reduce the total number of performance measures, the frequency of measurement, and/or the spatial density of the measurements. If cuts to the monitoring program are sufficiently extensive, the information provided will be severely compromised, making it impossible to detect how the ecosystem is responding to restoration activities. If monitoring and assessment information is going to be available to sup- port adaptive management of the CERP, monitoring of ecosystem response to projects should be a priority. CERP monitoring responsibilities at all levels of project management across and within agencies should be clearly established. While monitoring in and of itself does not ensure restoration progress, without monitoring plans tailored to improve understanding of ecosystem response and the outcomes of project implementation from local to whole ecosystem scales, uninformed management decisions will be made with potentially undesirable ecosystem consequences. Investments in monitoring at the outset of restoration and during the entire restoration process are as important to the CERP as invest- ments in construction projects. INFORMATION AND DATA MANAGEMENT Effective performance assessment in support of adaptive management relies on a well-designed information and data management system, which ensures that investments in monitoring are maximized and data are widely available to be utilized to the fullest extent. Primary responsibility for managing CERP data and information falls to the Information and Data Management (IDM) program, housed with the USACE in Jacksonville and the SFWMD in Ft. Lauderdale. As described in the Program Management Plan: Information and Data Management (USACE and SFWMD, 2007b), the CERP informatics strategy is based on a com- mon information system (CERPZone) that enables Web-based sharing of informa- tion. CERPZone is not a centralized data repository for all CERP data. Instead, it is designed to facilitate search and retrieval of CERP data and documents over an “extranet” of participating agencies, universities, and other organizations. As of March 2008, 26 agencies and organizations were registered CERPZone users. The CERPZone architecture is consistent with the principles laid out in the CERP Master Program Management Plan (USACE and SFWMD, 2000), which

OCR for page 189
Building the Foundation for Adaptive Management 213 specifies general requirements such as tools for data and information sharing, networked servers for sharing documents, schedules, financial, scientific and geospatial data information, security, and content standards. The 2000 Master Program Management Plan also stipulates development of a document man- agement and control system for tracking and documenting decisions affecting restoration design and implementation. The IDM program currently operates with an annual budget of $6–7 mil- lion and is jointly funded by the SFWMD and the USACE. IDM activities are now closely coupled to those of the Interagency Modeling Center. In addition to developing and maintaining CERPZone, the IDM program, upon request and approval, provides tools and technical support to CERP projects. Examples of sup- port activities include database administration, Web application development, GIS analysis and map production, and computing infrastructure support (e.g., server backups and data archiving). Thus, CERP activities such as RECOVER participate in CERPZone but are also “customers” that can request special services on a cost recovery basis (e.g., preparation of maps for the 2007 System Status Report). CERPZone provides access to an impressive amount of information. Roughly 200,000 online documents and 1 Terabyte of data are catalogued in an electronic data catalogue (EDCat). Documents are stored and tracked using Documentum, an electronic document management system. A data access, storage, and retrieval (DASR) system and a GIS portal enable access to the data. A Model Manage- ment System has been developed to access model code and input and output data. Metadata management tools (e.g., Morpho4) are being used to standardize data documentation, and tools are being developed to support spatial queries (EDCat v. 2.0). Administrative tools are in place to manage CERPZone access, data storage requests, and technology project requests. A public Web site (www. evergladesplan.org) is maintained to share information with the general public. Future development plans through FY 2010 are listed in the April 2007 Program Management Plan (USACE and SFWMD, 2007b). The CERP IDM program provides a sound structure and good functionality for collaborative CERP data and information management. Its effectiveness depends on user awareness of and competence in using available tools and applications, timely provision of research and monitoring data from primary data stewards to CERPZone, and community adherence to data standards and guidelines (these are articulated in a series of guidance memoranda). Experiences of CERP staff to date indicate that the CERP information and data management system is working well for the documentation and retrieval of monitoring data. Most monitoring data and associated metadata collected under http://knb.ecoinformatics.org/morphoportal.jsp. 4

OCR for page 189
214 Progress Toward Restoring the Everglades contracts executed with RECOVER are archived, discoverable, and accessible, and data gaps are mainly attributable to some individual principal investigators who have been slower to share their data. However, based on experience, assembling the 2007 System Status Report, integration of CERP monitoring and data, and data synthesis for ecosystem assessment are challenging and tools are needed to facilitate the production of standard assessments. Spatial query tools currently under development will help. CERP IDM program staff should work closely with RECOVER scientists to identify and prioritize products that should be routinely derived from CERP monitoring data or tools to help scientists pro- duce such products. MODELING IMPROVEMENTS IN SUPPORT OF ADAPTIVE MANAGEMENT As discussed by NRC (2007), models are critical tools used in adaptive man- agement to test the understanding of and to predict the ecological and hydrologic consequences of management alternatives and ecosystem drivers (e.g., rainfall, sea-level rise, climate change). The CERP was developed using simulation mod- els to evaluate expected outcomes of various restoration scenarios. Both moni- toring and modeling support the adaptive management process by providing information to allow informed alterations to the CERP during its implementation. The monitoring program will measure ecosystem response to restoration, and the modeling program provides a system-level context for integrating the responses. As abstract representations and simplifications of the complex real world, models are useful tools for integrating and updating current knowledge of a system and for identifying and prioritizing critical uncertainties. South Florida restoration activities are supported by an enormous and impres- sive multiagency modeling effort. Numerous models have been or are being developed by researchers from agencies such as the SFWMD, USACE, other federal agencies, independent consultants, and academic institutions in the United States and elsewhere. The models vary in stage of development and application; some have been widely applied for evaluation and planning of CERP projects, while others are still being developed, calibrated, verified, or reviewed. The following sections review the current state of restoration modeling of hydrologic and ecological systems and evaluate the status against modeling needs for effective adaptive management. Hydrologic Modeling The current system model for the Everglades restoration continues to be the SFWMM, developed and maintained by the SFWMD. Operating on a

OCR for page 189
Building the Foundation for Adaptive Management 215 2-mile-by-2-mile grid (hence “2 × 2 model”), the SFWMM simulates the verti- cal water balance of rain, evapotranspiration, and groundwater seepage as well as overland flow and flow at hydraulic structures. Generally operating on a daily time step, the model simulated a 31-year historic meteorological period (1965 to 1995) for the original CERP planning in the 1990s, whereas recent runs have extended the period of record by 5 years (1965 to 2000), with ongoing efforts to extend the record further forward in time. Extending the simulation period to a starting date prior to 1965 is complicated by fewer rain gages and other moni- tors, although conceivably this could be done through extrapolation and time series techniques. The physics of the 2 × 2 model are well documented, and the SFWMM has undergone extensive peer review (e.g., Bras et al., 2005). However, the SFWMM consists of more than just a simulation of rainfall, runoff, and flow; the operations of the massive and complex SFWMD system are also incorporated into the model, as are operating rules for all CERP components. These operational algorithms are embedded in the Fortran code and are difficult to change and understand by all but the most familiar users. That is, if changes in operations of CERP components are to be simulated, the process involves hard-coding the changes into the SFWMM, rather than in the form of external model input. The SFWMD is well aware that this restriction limits the flexibility of the model with regard to straightforward evaluation of alternative plans and is working for a more flexible means to input operating rules in the next genera- tion of CERP water simulation, namely, the Regional Simulation Model, or RSM (J. Obeysekera, SFWMD, personal communication, 2008). The issue is exacerbated in that system response is more sensitive to operat- ing rule changes than to most changes in the physics of the model. Because of the intricate way in which operations are coded into the SFWMM, it is difficult to tell what change in rules causes what change in output, as well as the rela- tive impact of different operating rules within the system and the relative impact of operating rules versus, say, climate change. This again points to the urgency of differentiating policy from physics in the crucial CERP modeling tools, as is under way in the development of the RSM. The RSM is intended to address this issue by incorporating two simulation engines: one for hydrology and hydraulics (the Hydrologic Simulation Engine, or HSE) that itself has substantial improvements in the physics, and one for management (the Management Simulation Engine, or MSE), for easy variation of operations. Two external panels (Bales et al., 2007; Chin et al., 2005) have recently reviewed the latest version of RSM (and its application to the natural system, the NSRSM). The RSM is approximately 2 or 3 years away from replacing the SFWMM for routine CERP analysis (e.g., first use for CERP as a whole may be in 2010), but it is being tested now in more self-contained basins north of

OCR for page 189
216 Progress Toward Restoring the Everglades Lake Okeechobee (Van Zee, 2007) and is being applied successfully to a variety of local problems (e.g., the water conservation areas, Everglades National Park, and lower east coast [the so-called “Glades model”]; C-111; and Biscayne Bay coastal wetlands). Further improvements in numerical modeling have been identified in a num- ber of earlier reports (e.g., the strategic modeling plan of 2003 [Plato Consulting, Inc., 2003]; NRC, 2007; RECOVER 2005c). Included in these reports are state- ments that the need to model the “fate and transport of nutrients, sediments and nonpoint source pollution is imperative” (RECOVER, 2005c, p. 40). The SFWMD intends to eventually add nutrient and sediment transport models linked to the RSM. Another limitation of the RSM is that it is a two-dimensional model, where only the average flow in the vertical direction is considered. An eventual linkage of the RSM with local project-scale models to analyze fine-scale problems (e.g., sediment transport for tree island or seepage management) is in the planning stage. Limited staff and funding resources mean that these additional envisioned improvements to the RSM will not be accomplished for many years. Hydrologic and hydraulic applications for site-specific CERP project analysis will continue to be handled by local-scale models (e.g., hydraulic models for channels), often with the SFWMM or the RSM used to provide boundary conditions. As recommended in the Strategic Modeling Plan (Plato Consulting, Inc., 2003), RSM development and application are using software engineering and project management processes based upon the Capability Maturity Model Inte- grated (CMMI)5 developed by the Software Engineering Institute at Carnegie Mellon. CMMI is a process improvement model consisting of industry best practices that organizations use to improve their business, project management, and software and engineering processes in order to increase product quality, work efficiencies, and customer satisfaction. SFWMD modelers report that the implementation of CMMI processes within the Hydrologic and Environmental Systems Modeling Department of the SFWMD has brought tremendous value and improvements to the RSM, and future progress depends upon continuing these efforts. Similarly, the Interagency Modeling Center has been instrumental in achieving consistency of modeling efforts across agency boundaries. The design coordination team, consisting of SFWMD and USACE personnel, set the model- ing priorities. The priority modeling tasks are then executed by the Interagency Modeling Center (J. Obeysekera, SFWMD, personal communication, 2008). The pace of model development is limited by the staff resources available—a problem typical of all agencies in the CERP. Staff time for model development is also impacted by the need for production runs of the SFWMM and/or the RSM. Additional information on CMMI can be found online at http://www.sei.cmu.edu/cmmi/. 5

OCR for page 189
Building the Foundation for Adaptive Management 217 Possible reductions in SFWMD funding will not help this situation. One way in which agencies have tried to adapt to limitations on resources is to use the simplest model that will address specific needs. If the SFWMM is not needed to resolve a local question, then a simpler or more site-specific hydrologic or hydraulic model may suffice. One example of “simpler” is to use the SFWMM to develop regional boundary conditions surrounding a smaller application area, within which the more sophisticated RSM is used to provide small-scale, detailed modeling (J. Obeysekera, SFWMD, personal communication, March 2008). Another example of “simpler” is the use of a regional spreadsheet model in the analysis of Tamiami Trail modifications for the Modified Water Deliveries to Everglades National Park project (USACE and SFWMD, 2008a). The model uses measured flows as inputs to a 25-year simulation of flows entering Shark River Slough. The Natural System Regional Simulation Model (NSRSM), the application of RSM to the predevelopment Everglades but with adjustments for topography and landscape constraints, is hampered (just as it has been for the NSM) by lack of predevelopment data with which to calibrate and verify simulation results. Ongoing SFWMD efforts to resolve predevelopment topography will aid in this effort. Regarding application of models such as RSM to climate change scenarios, natural variability in rainfall patterns is partially accounted for by the current 36-year input rainfall time series. Current efforts to develop an input time series dating back to 1914 will certainly be an improvement in developing an under- standing of long-term climate fluctuations and trends. In addition, for climate- change analysis, scenarios of future rainfall time series are needed, as well as changing sea-level boundary conditions. The CERP could not have been developed without models like the SFWMM and cannot continue to progress without models like the RSM. It is clear that agency staff and managers recognize the value of modeling, and modeling is performed in every case in which it is needed. The principal impact of limited resources is to slow model development. Use of the SFWMM instead of the RSM because the RSM is not yet ready for application to the full CERP leads to evaluations that are likely less accurate and certainly more cumbersome. Tech- nical challenges in model development continue to be addressed by a capable model development team. Ecological Modeling Ecological models linking hydrology and water quality to species and eco- system dynamics have the potential to significantly improve the effectiveness of

OCR for page 189
218 Progress Toward Restoring the Everglades CERP monitoring and assessment and to accelerate learning through adaptive management (DeAngelis et al., 2003; NRC, 2003b, 2007). A spectrum of models, ranging from qualitative conceptual models to complex dynamic simulation models, has been developed to link water management to ecological outcomes. While each has value, spatially explicit models that are logically coupled to the water management models and support landscape-level planning and decision making have been especially important to CERP (Sklar et al., 2001) Two ecologi- cal models in particular, the Across Trophic Level System Simulation (ATLSS) model and Everglades Landscape Model (ELM), were developed in large part to support restoration design, monitoring, and evaluation (DeAngelis et al., 1998; Sklar et al., 2001). ATLSS is a set of spatially structured models operating at 0.01—1 km2 resolution. Coarse (2 × 2) output from the SFWMM is downscaled to finer- scale hydrologic grids (currently the rate-limiting step in model operation) and then used in process models for lower trophic levels such as zooplankton and phytoplankton, structured population models for fish and macroinvertebrates, and individual-based models for large vertebrates such as Cape Sable seaside sparrows (CSSSs), wood storks, alligators, and white-tailed deer. ELM is a spa- tially structured “patch” model that couples hydrology, energy, and nutrients to species-specific plant growth and vegetation pattern across the landscape (Sklar et al., 2001). Between 1997 and 2001 ATLSS was used intensively for evaluating CERP alternatives (http://www.atlss.org/). Researchers have continued to extend ATLSS capabilities, adding fire and vegetation succession modules and refining and testing spatially explicit species index models and population viability analyses for target species such as the snail kite, CSSS, and American alligator. Software has been developed to support model visualization and decision making, and ATLSS code has been revised to allow the model to be operated using parallel computing to improve scalability and reduce run time (Wang et al., 2006). Despite the large investment in ATLSS models for restoration planning, they have not been used in CERP implementation and are not integrated with MAP monitoring efforts for model validation and refinement, integrated assessment, or for forecasting alternative management scenarios. This lack of integration may be due in part to technical complexity and computing requirements of the ATLSS models. Until recently, the expertise and infrastructure to operate ATLSS models existed only at the University of Tennessee, hindering integration with the water management models. Several modules (e.g., periphyton, crayfish, snail kite, and alligator modules) could be readily applied at the 2-mile-by-2-mile scales of current water management models (although the variable grid sizes of the RSM [0.1 to 2 miles] now under development would be much better for linking to

OCR for page 189
Building the Foundation for Adaptive Management 219 the ecological models). ATLSS has recently been installed and is being operated at Everglades National Park, and the Department of the Interior (DOI) intends to locate one or two ecological modelers at the Interagency Modeling Center to operate ATLSS and other ecological models. The latter would be a positive step, as limited DOI involvement has precluded coupled hydrologic-ecological modeling at the Interagency Modeling Center. Given DOI’s relatively low level of investment in modeling staff, it will be probably at least 2 years before ATLSS will be operational at the Interagency Modeling Center. Recent development of ELM (version 2.5) has focused on linking water management scenarios to regional water quality outcomes, specifically total phosphorus concentrations in surface water and net phosphorus accumulation in the Greater Everglades ecosystem. An external peer review of ELM 2.5 concluded that the model formulation was fundamentally sound and relatively unbiased. The peer-review panel argued for close integration of ELM with CERP monitoring and adaptive management and stated that ELM “may be the primary method that SFWMD should use to guide any monitoring changes in the future, not just for ELM itself, but to monitor progress in restoring the Everglades” (Mitsch et al., 2007). Nevertheless, ELM was not integrated into system status monitoring and reporting for 2006 and 2007. In summary, ecological models have seen diminishing use in CERP since the development of the CERP during the mid-1990s (the Restudy). The trend is not due to technical problems with the models and is not a hiatus while next- generation models are developed; rather, it reflects reduced staffing for model application and development. The committee can only echo previous NRC reports in stating that integrated hydro-ecological modeling has an important role in project planning, monitoring, assessment, and adaptive management. To improve the application of ecological models for the CERP planning and management, the DOI needs to invest more attention and resources in eco- logical modeling and data management activities at the Interagency Modeling Center. CONCLUSIONS AND RECOMMENDATIONS To facilitate restoration progress despite some scientific and engineering uncertainty, Congress mandated an adaptive management approach for the CERP. Adaptive management requires the support of effective monitoring and assessment protocols and adequate hydrologic and ecological models. In this chapter, recent progress and major issues with respect to CERP science to sup- port the adaptive management process were reviewed. The major findings are highlighted below.

OCR for page 189
220 Progress Toward Restoring the Everglades The RECOvER team has now produced nearly all of the elements needed to implement a decision-making framework using adaptive management to assess scientific uncertainty. Documents describing the adaptive management process (RECOVER, 2006a; 2007e), and all aspects of performance assessment (i.e., the monitoring plan [RECOVER, 2004], an assessment plan [RECOVER, 2006f], performance measures [RECOVER, 2007b], and quality assurance requirements [RECOVER, 2007d]) are completed. Conceptual ecological models that are the foundation of the monitoring and assessment documents have been peer- reviewed and published. The information management and data management system and the Interagency Modeling Center are actively developing tools to support the assessment and planning aspects of decision making and assisted in production of the 2007 System Status Report, the first in a series of assessment reports that documents the ecosystem response to implementation of CERP projects. The System Status Reports are a critical component of the adaptive management strategy; they are the vehicle used to transmit new scientific infor- mation to restoration managers. These are significant accomplishments, and their importance should not be underestimated. However, the CERP adaptive management scheme could be improved by addressing several major issues, which are summarized below. In order for monitoring and assessment information to adequately support CERP adaptive management, a robust program of ecological monitoring should remain a priority. While monitoring in and of itself does not ensure restora- tion progress, without monitoring to understand ecosystem response to project implementation from local to whole ecosystem scales, uninformed management decisions will be made with potentially undesirable ecosystem consequences. A well-justified and documented set of performance measures has been developed, and a scientifically robust process for updating, refining, and adding to the set of performance measures is in place. RECOVER should continue to move forward to fully develop those performance measures that are currently monitored and to reconcile performance measure standards for evaluation and assessment. The periodic review of performance measures should consider ways to make sure that the total number of variables monitored is appropriate to their purpose for informing decisions and to the funding available for monitoring efforts. It also is important to match the frequency of monitoring with the speed of change of the variables that are monitored and to increase reliance on remotely sensed data collection methods. Revisions of the monitoring and assessment system should be firmly grounded in the use of the data for planning and management decision making. The 2007 System Status Report achieved its stated objectives to test the monitoring and assessment plan and establish as long a baseline as possible to

OCR for page 189
Building the Foundation for Adaptive Management 221 capture the natural variance of CERP performance measures. In most cases, the System Status Report also provided some discussion of the ability to detect per- formance measure change or trends. In doing so, the first System Status Report serves as the reference that will be used to gauge system response as CERP projects are implemented, and it is extremely valuable. Insights learned during the production of the report should be incorporated into the revision of the Monitoring and Assessment Plan (MAP I) and the conceptual ecological models, as needed, and for reprioritization of the performance measures. To maximize the usefulness of System Status Reports for adaptive man- agement, RECOVER should develop succinct summaries in future reports that clearly address whether the interim and longer-term goals are being met; if not, why not; and what CERP operations or design changes are most likely to move ecosystem response closer to the interim goals. RECOVER should also aim for a more coordinated approach in future assessments that allows for additional consideration of restoration effects across modules. The CERP Information and Data Management program provides a sound structure and good functionality for collaborative CERP data and information management. Performance assessment depends on a well-designed information and data management system so that monitoring data can be widely available and utilized to the fullest extent. The effectiveness of the Information and Data Management program depends on user awareness of and competence in using available tools and applications, timely provision of research and monitoring data, and community adherence to the articulated data standards and guidelines. Based on experience assembling the 2007 System Status Report, it appears that timely integration of CERP monitoring data remains challenging, and efforts to develop tools to facilitate the production of standard assessments should continue. Integrated hydrologic, ecological, and water quality modeling tools are needed for science to have a fully developed role in CERP decision making and ecosystem management. CERP planning and assessment of performance indicators are dependent on the modeling tools; as model development and implementation lag, so does access to more accurate and functional tools. Models are needed for each ecological indicator (performance measures) to compare predicted and monitored indicator responses to the CERP for effective adaptive management decision making. This will only occur when: • ecological modeling and data management activities are fully incorpo- rated and funded in the CERP’s Interagency Modeling Center, • Water quality and sediment transport models become routinely available and integrated with the new Regional Simulation Model (RSM), and

OCR for page 189
222 Progress Toward Restoring the Everglades • These physical-chemical models can be readily linked to ecological models. Linking of models (the third point above) continues to be a particularly slow endeavor. Impediments to achieving the synergy between planning, monitoring, and assessment needed to support effective adaptive management are related to shrinking resources, loss of staff, and time. Shrinking CERP resources mean that the trade-off between use of staff for model development versus for model production runs for CERP planning favors the latter. Moreover, if staff numbers are reduced, the knowledge and training of departing professionals go with them. This committee recognizes that resources are limited but notes that model development is a long-term proposition and should continue with as much sup- port as possible so the tools required to restore and manage the ecosystem are available in the future.