Prioritizing Wetland and Stream Restoration and Protection Using Landscape Analysis Tools

Transcription

1 Prioritizing Wetland and Stream Restoration and Protection Using Landscape Analysis Tools The Environmental Law Institute recently completed A Handbook for Prioritizing Wetland and Stream Restoration and Protection Using Landscape Analysis Tools. This Article provides a synopsis of that guide, focusing on the processes used to develop such tools and how those tools are applied to evaluate and prioritize sites for preservation and restoration. By Eric Sweeney, Philip Womble, and Rebecca Kihslinger A substantial increase in infrastructure investment over the coming decade is likely to have a significant impact on the nation s wetlands and streams. Likely investments in surface transportation improvements, water development projects, and renewable and traditional energy development, all threaten to harm these critical aquatic habitats. The potentially deleterious effects of these development projects can be minimized, however, through strategic planning and science-based cumulative impacts analysis, evaluation of avoidance and minimization options, and the selection of high-value compensation sites. The strategic identification of high-value conservation and restoration sites can help to efficiently prioritize opportunities to offset impacts to wetland and stream resources. Landscapelevel approaches to the prioritization of wetland and stream restoration and protection sites can also provide a platform for integrating both regulatory and nonregulatory aquatic resource conservation efforts in a more holistic manner. The Environmental Law Institute (ELI) recently completed A Handbook for Prioritizing Wetland and Stream Restoration and Protection Using Landscape Analysis Tools. 1 The handbook is designed to provide valuable information to guide the development, establishment, and refinement of approaches for identifying restoration and protection priorities. The handbook defines landscape prioritization tools in terms of inputs, outputs, and objectives, and compiles an inventory of data and methods used by prioritization programs. This Article focuses on the processes used by the various prioritization programs to develop these tools and how the tools are, or will be, applied to evaluate and prioritize sites for preservation and restoration. In this Article, we: Identify the variety of component processes used by prioritization program developers upstream (e.g., determining prioritization objectives and defining input factors and weightings) and downstream (e.g., calibrating tools and validating model outputs) of the application of landscape prioritization analysis tools. Describe how landscape prioritization tools are applied to evaluate potential wetland and stream restoration and protection sites across a range of objectives and guide regulatory and nonregulatory program decision-making. Discuss the transferability of prioritization approaches and barriers to the development and implementation of landscape prioritization tools. Landscape Prioritization Programs and Tools To develop the handbook, ELI examined 30 prioritization programs and 115 associated landscape prioritization tools. These programs and tools represent a broad continuum Figure 1. Single-objective tools integrate multiple input factors to produce a prioritization output that evaluates a single prioritization objective. Figure 2. Multi-objective tools integrate multiple input factors to produce an output that evaluates multiple prioritization objectives. Input factor 1 E.g., percent forested Input factor 2 E.g., in priority conservation area? Input factor 3 E.g., percent impervious Prioritization Output Prioritization objective E.g., habitat quality Input factor 1 E.g., percent forested Input factor 2 E.g., in priority conservation area? Input factor 3 E.g., percent impervious Prioritization Output Prioritization objective 1 E.g., habitat quality Prioritization objective 2 E.g., water quality Prioritization objective 3 E.g., feasibility of restoration 21

2 Determination of prioritization objectives Stakeholder feedback A Prioritization objectives Data analysis Determination of input factors/weightings Stakeholder feedback Analysis of field data B Input factors/ weightings C Input data QA/QC Field verification Desktop review Application of landscape prioritization tools Aquatic resource condition Habitat quality Carbon storage Historic functional change Cost-effectiveness Social values Feasibility of restoration Suitability for preservation Flood mitigation Surface water supply Future impacts Sustainability of restoration Groundwater supply Water quality Tool validation Correlation analysis Systematic field confirmation D F Landscape prioritization analysis Prioritization Output E G Tool calibration Stakeholder feedback Analysis of field data Refinement of identified priorities Field methods Expert/stakeholder input Figure 3. General process applied by the 30 landscape prioritization programs evaluated in this study. Methods applied by programs included (A) determination of prioritization objectives, (B) determination of input factors/weightings, (C) input data QA/QC, (D) application of landscape prioritization tools, (E) calibration of landscape prioritization tools, (F) refinement of identified priorities, and (G) validation of landscape prioritization tools. of geospatial methods and data currently used in the United States to identify wetland and stream restoration or protection priorities at watershed or landscape scales. We defined a prioritization program to include any agency, organization, or researcher who developed one or more landscape prioritization tool. A prioritization tool is a landscape metric or index that integrates a variety of data inputs to rank or score sites in terms of their value for aquatic resource restoration or protection for one or more prioritization objectives (e.g., habitat quality). Of the 115 tools we examined, we found that they fell into two broad categories single-objective tools and multi-objective tools. Single-objective tools integrate a set of input factors to derive an output representing a single prioritization objective, such as habitat quality (see Figure 1). Multi-objective tools integrate multiple input factors to obtain an output that represents multiple prioritization objectives, such as habitat quality, water quality, and feasibility of restoration (see Figure 2). The Components: Upstream and Downstream Processes All prioritization programs analyzed for the handbook are defined by a core geospatial analysis tool, as illustrated in Figure 1 and Figure 2 and Box D in Figure 3. However, there are a number of steps that prioritization programs applied both upstream and downstream of this analysis in order to determine objectives, weight input factors, and validate the tool. Figure 3 illustrates the generic sequence of these processes. Upstream (i.e., prior to conducting the prioritization analysis), the prioritization programs we studied applied a variety of approaches to determine prioritization objectives, determine input factors/weightings, and ensure input data Quality Assurance/Quality Control (QA/QC) (A-C in Figure 3). Then, following the application of the geospatial analysis, programs applied a variety of approaches to calibrate their prioritization tools, validate the tools, and refine the identified priorities (E-G in Figure 3). The upstream and down- 22 national wetlands newsletter

3 stream approaches applied by the programs we studied are described in more detail below. Upstream Processes Determination of Prioritization Objectives (Fig. 3 (A)) In most cases, the program developers we studied predetermined prioritizing objectives without applying any formal methods. However, seven programs applied a specific method for identifying prioritization objectives. Five of the 30 programs identified prioritization objectives by soliciting stakeholder input on watershed/landscape priorities. For instance, to determine what habitat types to target, the developers of the NOAA Habitat Priority Planner Mississippi-Alabama Habitats Tool solicited input from a stakeholder group that included over 60 state and local representatives involved with habitat management in coastal Alabama. 2 Two programs integrated watershed/landscape data analysis as part of their process for identifying prioritization objectives. For example, as part of its Standard GIS Methodology for Wetland Analysis, the Arkansas Multi-Agency Wetland Planning Team identified objectives specific to each of its Watershed Planning Areas by evaluating readily available watershed-scale GIS data sets that capture basic wetland characteristics. 3 Determination of Input Factors/Weightings (Fig. 3 (B)) Sunrise River Mitigation Site Selection Survey 12. When looking for potential mitigation sites within the Sunrise River Watershed, which is more important? Protection from Urban Sprawl (Criterion 5)? 13. When looking for potential mitigation sites within the Sunrise River Watershed, which is more important? Connectivity with Existing Public Lands (Criterion 8)? 14. When looking for potential mitigation sites within the Sunrise River Watershed, which is more important? High Biodiversity (Criterion 7)? 15. When looking for potential mitigation sites within the Sunrise River Watershed, which is more important? Conditions Reasonably Distanced/Removed from Human Disturbance (Away from Roads and City Centers) (Criterion 8)? 16. When looking for potential mitigation sites within the Sunrise River Watershed, which is more important? Mitigation Inside the Floodplain (Criterion 8)? Figure 4. The Corps used a web-based survey to solicit input from a stakeholder team to help determine weightings for each prioritization criterion to apply in the Spatial Decision Support System prioritization model. Ten of the prioritization programs that we evaluated applied methods for selecting input factors and weightings. Seven programs determined input factors/weightings using expert/ stakeholder feedback. For example, in a series of workshops, the U.S. Army Corps of Engineers Sunrise River Watershed-Based Mitigation Pilot Program developers engaged a stakeholder team that identified criteria (e.g., hydrologic connection to tributaries) that it considered to be most important for targeting wetland compensation mitigation efforts within each of the relevant sub-watersheds. 4 Stakeholders individually completed a web-based survey in which they ranked selected criteria against one another in a series of pairwise comparisons (see Figure 4). The Corps then applied the Analytic Hierarchy Process (a type of Multi-Criteria Decision Analysis) to determine weightings for each criterion as part of its Spatial Decision Support System prioritization tool. Three of the programs we studied determined input factors and weightings using statistical analysis. These programs assembled field-based data sets (e.g., rapid assessment data) for wetland sites and determined the extent to which these field data correlate with available landscape data sets (e.g., percent tree ). Each program then built landscape analysis models by combining the landscape data sets that were most strongly correlated with the field-based measures. Weller et al. 2007, for example, used this approach to develop landscape assessment models that predicted five different Functional Condition Index scores (e.g., hydrology, habitat, etc.) for flat and riverine wetlands in the Nanticoke Watershed. 5 Input Data Quality Assurance/Quality Control (Fig. 3 (C)) Several prioritization programs applied Quality Assurance/ Quality Control (QA/QC) methods to ensure that data were valid prior to using them as input factors in their landscape prioritization analyses. These methods included a variety of approaches, including field verification and desktop review. Three of the studied methods applied rapid or intensive onthe-ground assessment methods to confirm the accuracy of input spatial data sets. For example, the Arkansas Multi- Agency Planning Team ground-truthed, or collected on location, the input data sets it used as part of its Standard GIS Methodology for Wetland Analysis using windshield surveys, field visits, and local knowledge. 6 23

4 Programs employed various methods for desktop review, including the application of predefined QA/QC guidelines (e.g., only using GIS data of known origin), comparison of input data to other data sources, and examination of data input integration. For instance, the Maryland Watershed Resource Registry is in the process of developing a method for field-verifying its various input data sources to ensure the quality of model outputs. 7 Downstream Processes Tool Calibration (Fig. 3 (E)) Tool calibration based on field data can be used to improve the accuracy and effectiveness of landscape prioritization tools. Programs reported a variety of approaches to tool calibration, including stakeholder feedback and analysis of field data. Two of the prioritization programs examined calibrated their landscape prioritization tools through stakeholder evaluation of outputs. The developers of the NOAA Habitat Priority Planner Mississippi-Alabama Habitats Tool, for example, presented output maps from its models to the stakeholder group that had been involved in the original model development. 8 This stakeholder group drew upon its collective expertise to provide feedback that was used to refine model parameters and improve results. Two other programs established weightings for input factors used in their models by analyzing outputs against field data. The Virginia Institute of Marine Science (VIMS) Wetland Condition Assessment Tool, for example, correlated counts of frequently observed stressors near randomly sampled wetland sites with model outputs for wetland condition for those sites. 9 VIMS then used changepoint analysis a method used to detect threshold changes in data to identify nonlinear thresholds not originally accounted for by its model, using what it learns from this analysis to refine model parameters. Tool Validation (Fig. 3 (F)) Figure 5. The Colorado Natural Heritage Program found a strong correlation between Ecological Integrity Assessment results and Landscape Integrity Model results. Tool validation methods are applied to determine the accuracy of the results generated by the landscape prioritization tools. Five of the programs we studied applied systematic field confirmation to compare results of their tools against field data. For instance, the Michigan Tech Research Institute compared site suitability rankings produced using the Wetland Mitigation Site Suitability Tool with rankings obtained from field monitoring data. The tool correctly assessed wetland suitability for 19 of the 20 sites identified by the model. 10 Two of the programs we studied applied statistical methods to validate tool accuracy by correlating tool outputs against rapidly or intensively obtained data. The Colorado Natural Heritage Program, for example, found strong correlations between outputs from its Landscape Integrity Model and data from three rapid assessments (e.g., Ecological Integrity Assessment) and one intensive assessment (Vegetation Index of Biotic Integrity) (Figure 5). 11 Refinement of Identified Priorities (Fig. 3 (G)) Twelve of the prioritization programs reviewed for our studied applied methods for refining the results of their landscape prioritization analysis. Seven programs used field data to further refine their selection of sites based on model outputs. Strager et al. s method, for example, scored potential wetland restoration sites, initially identified using a landscape prioritization analysis, based on rapid assessment criteria collected in the field by wetland specialists (Strager et al. 2012). 12 Strager et al. then evaluated the highest scoring sites using intensive assessments to identify which sites are most feasible for wetland and stream mitigation banking. Five programs refined outputs of landscape prioritization tools using expert/stakeholder input. For example, The Nature Conservancy s (TNC s) Willamette Basin Synthesis Project first produced a Union Portfolio of potential conser- 24 national wetlands newsletter

5 Figure 6. Using the WVSP online mapping tool, members of the public can draw (e.g., the polygon in light gray above) recommended additions or changes to Conservation Opportunity Areas identified in TNC s Union Portfolio. Members of the public can then print changes as a PDF for submission to TNC for incorporation into the Union Portfolio. vation sites by combining five existing Willamette conservation assessment maps (e.g., state parks vegetation). Members of the public refined site boundaries of this portfolio by using a nomination form and online mapping site to draw features over the Union Portfolio indicating which areas should be added or changed (Figure 6). 13 Application of Landscape Prioritization Tools to Prioritize Specific Objectives The 115 tools analyzed in this study targeted a wide variety of prioritization objectives, including biophysical functions (e.g., water quality improvement), social values (e.g., nature-based tourism), opportunity metrics (e.g., feasibility of mitigation), and condition metrics (e.g., based on stressors inferred from surrounding land use). The tools often rank or score hydrologic units (e.g., HUC-12s), wetland polygons, or pixels (e.g., 30m 2 raster cells) in terms of one or more prioritization objectives. Overall, we categorized prioritization objectives assessed by the tools into 14 groups: Aquatic resource condition Carbon storage Cost-effectiveness Feasibility of restoration Flood mitigation Future impacts Groundwater supply Habitat quality Historic functional change Social values Suitability for preservation Surface water supply Sustainability of restoration Water quality For example, The Duck-Pensaukee Watershed Approach Pilot team, led by TNC and ELI, applied a variety of tools for identifying wetland sites suitable for wetland restoration and preservation across a range of objectives or wetland services. 14 These objectives included the provision of wildlife habitat, flood abatement, surface water supply, water quality protec- 25

6 Clean Water Act wetland mitigation Clean Water Act water quality programs NRCS Wetland Reserve Program State Wildlife Action Plans Endangered Species Act 10 mitigation National Environmental Policy Act effects analysis State/local wetland mitigation State water quality programs Nonregulatory markets for ecosystem services Other nonregulatory restoration/protection Figure 7. In its 2012 report, TNC-ELI provided output maps from its assessments of watershed needs (i.e., areas of historic functional loss) and tools for identifying site-specific priorities. For example, its assessment of flood abatement needs (left) identifies HUC-12s in which site-specific restoration and preservation priorities for flood abatement (right) might be targeted to promote a watershed approach to regulatory and nonregulatory wetland conservation. tion, carbon storage, shoreline protection, and provision of fish habitat. To assess watershed needs for each service, the tool applied a unique method for quantifying historical losses of each service across sub-watersheds. For example, Figure 7 compares the Pilot s output map for historic flood abatement service losses with its output map showing priority wetland sites for restoring and protecting flood abatement services. Application of Landscape Prioritization Tools to Regulatory and Nonregulatory Programs We identified 10 categories of regulatory and nonregulatory applications for the results of the prioritization tools examined in our study. The application categories include the following: In terms of regulatory applications, program developers reported that the landscape prioritization tools are (or will be) most commonly applied to support site selection for 401/404 wetland compensatory mitigation, including general site selection, in-lieu fee program site selection, bank site selection, watershed approach, and determination of permit requirements. The Arkansas Multi-Agency Wetland Planning Team s GIS Methodology for Wetland Analysis, for example, is commonly used to inform mitigation site selection by the State Wetland Mitigation Banking Program (operated by the Arkansas State Highway and Transportation Department and the Arkansas Natural Resources Commission). On the nonregulatory side, program developers most often reported that the outputs of their prioritization tools are (or will be) used to aid in the selection of Wetland Reserve Program (WRP) sites by the Natural Resource Conservation Service (NRCS). As part of its process for allocating WRP funding, NRCS may award more points to proposed sites that fall within priority areas identified by prioritization programs. For example, multiple states throughout the Mississippi Alluvial Valley use prioritization results from the USGS Forested Breeding Bird Decision Support Tool to award WRP points. Transferability of Landscape Prioritization Tools Many of the prioritization programs in our study indicated that their approaches would be easily transferable to other programs seeking to develop their own landscape prioritization tools, due to their ease of use, minimal funding limitations, use of readily available data, and/or readily adaptable frameworks. For example, several programs characterized their tools as easy to use because they relied on basic raster calculation methods in ArcGIS. Kramer et al. (2012) explained that because its raster models are based on data sets that can be readily interchanged, the approach can be easily adopted by many potential users. 26 national wetlands newsletter

7 Other prioritization programs described their tools as highly transferable because their tools use national and readily available data sets as data inputs. Many of the data sets used by the Virginia Department of Conservation and Recreation GIS Model, for instance, are national data sets (e.g. SSURGO data) that can simply be reapplied if the model is transferred to other states. In another example, Weller et al. (2007) note: Our method could be applied wherever a large group of field assessments (say 50 or more) can be matched with appropriate digital geographic data. 15 Barriers to Development and Implementation Finally, ELI documented at least eight types of barriers encountered by the prioritization programs in the development of their tools. These barriers include the following: Data limitations Technical capacity Funding and staff time Property rights concerns Promoting use of the tool Bureaucratic obstacles Stakeholder collaboration Maintaining updated data The programs we studied identified 16 different types of data gaps that limited the functionality of their tools, including a lack of data on wetlands, soils, agriculture, and urbanization; inadequate resolution of elevation data; and a lack of information on local impacts, among others. Several programs identified property rights concerns as a barrier to the implementation of their tools. For example, TNC is cautious in applying results from its Aquatic Ecoregional Assessment because many landowners in Virginia are sensitive to TNC identifying specific locations on a map for restoration/conservation. Finally, other programs characterized the maintenance of updated input data to be a significant obstacle. The Kramer et al. (2012) tool developers, for example, expected that the most significant data-related concern going forward would be their ability to continuously update the tools inputs with new data sets so that outputs would remain as relevant as possible. Benefits of Landscape Prioritization Methods As highlighted in this article, ELI s new handbook analyzes the objectives and components of existing landscape prioritization tools and summarizes the applications, transferability, and limitations of these tools. As such, the handbook provides useful information and practical models for practitioners seeking to develop or refine prioritization methodologies. The programs examined for our handbook also highlight the wide variety of ways in which landscape prioritization tools benefit wetland restoration and protection, including efficient identification of restoration and protection sites that address multiple conservation objectives, streamlined permitting processes for transportation and natural resource agencies undertaking compensatory mitigation, effective cost-benefit analysis with respect to functional return on investment, and reduced costs associated with field monitoring. Our goal is that our handbook will provide models and information for all states, tribes, and local governments and other organizations involved in the siting of wetland and stream restoration and protection projects. By improving the ability of wetland programs to site projects on a landscape basis, we hope that this handbook will contribute to an overall improvement in watershed and human health. Endnotes 1. The handbook is available for free download at The handbook is accompanied by an interactive website ( that includes information on how research was carried out, the project s findings, and factsheets on the programs studied. 2. NOAA Habitat Priority Planner Mississippi-Alabama Habitats Tool. Accessible from: 3. The Multi-Agency Wetland Planning Team, The Standard GIS Methodology for Wetland Analysis, 4. U.S. Army Corps of Engineers St. Paul District Sunrise River Watershed-Based Mitigation Pilot Study Spatial Decision Support System. Accessible from: Weller et al. (2007) Wetland Condition Assessment Tools. Accessible from: from Jennifer Sheehan, Arkansas Multi-Agency Wetland Planning Team Coordination Office (Oct. 13, 2011). 7. Maryland Watershed Resources Registry. Accessible from: sites/default/files/docs/md_wrr_factsheet.pdf. 8. NOAA Habitat Priority Planner Mississippi-Alabama Habitats Tool. Accessible from: 9. Virginia Institute of Marine Science Wetland Condition Assessment Tool. Ac cessible from: pdf. 10. Michigan Tech Research Institute, Validation Report: Wetland Mitigation Site Suitability Tool (2009), PageLibrary B004F2A59.nsf/h_Toc/BA CC B60045 FF3C/?OpenDocument. 11. Feedback provided by Joanna Lemly, Wetland Ecologist, Colorado Natural Heritage Program (May 16, 2012). 12. Strager et al. (2011) Banking Site Selection Model. Accessible from: eli.org/sites/default/files/docs/strager_bsst_factsheet.pdf. 13. Interview with Dan Bell, Willamette Valley Consevation Directory, The Nature Conservancy (May 22, 2012). 14. The Nature Conservancy & Environmental Law Institute Duck-Pensaukee Watershed Approach Pilot. Accessible from: tnc_eli_dpwap_factsheet.pdf. 15. Weller et al. (2007) Wetland Condition Assessment Tools. Accessible from: 27