Technical Approaches Toward Achieving No Net Loss
This chapter describes lessons that can be drawn from the preceding chapters to increase the likelihood that mitigation requirements will in fact move forward and serve water quality and other functions in the nation's watersheds. The committee offers practical guidelines for designing and constructing sustainable compensation wetlands, for assessing the functional endpoints of those wetlands, and for creating the institutional reforms that will identify and secure the desired results.
OPERATIONAL GUIDELINES FOR CREATING OR RESTORING WETLANDS THAT ARE ECOLOGICALLY SELF-SUSTAINING
The compensatory mitigation process has been weakened by insufficient scientific knowledge as well as limitations of wetland regulatory institutions. The committee agreed on 10 guidelines applicable to compensatory mitigation projects. The guidelines suggest that in most cases wetland restoration is preferred to wetland creation. Also, the guidelines address mitigation project setting, site design, landscape setting, the importance of reference wetlands, and long-term management. Special attention must be paid to hydrological and topographical variability, subsurface characteristics, and the hydrogeomorphic and ecological landscape and climate of a site. Systems that are designed to incorporate natural processes will be more likely to ensure long-term sustainability.
1. Consider the hydrogeomorphic and ecological landscape and climate. Whenever possible, locate the mitigation site in a setting of comparable
landscape position and hydrogeomorphic class. Do not generate atypical “hydrogeomorphic hybrids”; instead, duplicate the features of reference wetlands or enhance connectivity with natural upland landscape elements (Gwin et al. 1999).
Regulatory agency personnel should provide a landscape setting characterization of both the wetland to be developed and, using comparable descriptors, the proposed mitigation site. Consider conducting a cumulative impact analysis at the landscape level based on templates for wetland development (Bedford 1999). Landscapes have natural patterns that maximize the value and function of individual habitats. For example, isolated wetlands function in ways that are quite different from wetlands adjacent to rivers. A forested wetland island, created in an otherwise grassy or agricultural landscape, will support species that are different from those in a forested wetland in a large forest tract. For wildlife and fisheries enhancement, determine if the wetland site is along ecological corridors such as migratory flyways or spawning runs. Constraints also include landscape factors. Shoreline and coastal wetlands adjacent to heavy wave action have historically high erosion rates or highly erodible soils, and often heavy boat wakes. Placement of wetlands in these locations may require shoreline armoring and other protective engineered structures that are contrary to the mitigation goals and at cross-purposes to the desired functions.
Even though catastrophic events cannot be prevented, a fundamental factor in mitigation plan design should be how well the site will respond to natural disturbances that are likely to occur. Floods, droughts, muskrats, geese, and storms are expected natural disturbances and should be accommodated in mitigation designs rather than feared. Natural ecosystems generally recover rapidly from natural disturbances to which they are adapted. The design should aim to restore a series of natural processes at the mitigation sites to ensure that resilience will have been achieved.
As described in other chapters, regulatory agency personnel often do not have either the time or the education and training to consider important, broader issues such as landscape setting. It is imperative that the Corps, EPA, and other advisory agency personnel receive additional training in landscape ecology and other considerations that are poorly represented in the present mitigation process.
2. Adopt a dynamic landscape perspective. Consider both current and future watershed hydrology and wetland location. Take into account surrounding land use and future plans for the land. Select sites that are, and will continue to be, resistant to disturbance from the surrounding landscape, such as preserving large buffers and connectivity to other wetlands. Build on existing wetland and upland systems. If possible, locate
the mitigation site to take advantage of refuges, buffers, green spaces, and other preserved elements of the landscape. Design a system that utilizes natural processes and energies, such as the potential energy of streams as natural subsidies to the system. Flooding rivers and tides transport great quantities of water, nutrients, and organic matter in relatively short time periods, subsidizing the wetlands open to these flows as well as the adjacent rivers, lakes, and estuaries.
3. Restore or develop naturally variable hydrological conditions. Promote naturally variable hydrology, with emphasis on enabling fluctuations in water flow and level, and duration and frequency of change, representative of other comparable wetlands in the same landscape setting. Preferably, natural hydrology should be allowed to become reestablished rather than finessed through active engineering devices to mimic a natural hydroperiod. When restoration is not an option, favor the use of passive devices that have a higher likelihood to sustain the desired hydroperiod over the long term. Try to avoid designing a system dependent on water-control structures or other artificial infrastructure that must be maintained in perpetuity in order for wetland hydrology to meet the specified design. In situations where direct (in-kind) replacement is desired, candidate mitigation sites should have the same basic hydrological attributes as the impacted site.
Hydrology should be inspected during flood seasons and heavy rains, and the annual and extreme-event flooding histories of the site should be reviewed as closely as possible. A detailed hydrological study of the site should be undertaken, including a determination of the potential interaction of groundwater with the proposed wetland. Without flooding or saturated soils for at least part of the growing season, a wetland will not develop. Similarly, a site that is too wet will not support the desired biodiversity. The tidal cycle and stages are important to the hydrology of coastal wetlands.
4. Whenever possible, choose wetland restoration over creation. Select sites where wetlands previously existed or where nearby wetlands still exist. Restoration of wetlands has been observed to be more feasible and sustainable than creation of wetlands. In restored sites the proper substrate may be present, seed sources may be on-site or nearby, and the appropriate hydrological conditions may exist or may be more easily restored.
The U.S. Army Corps of Engineers (Corps) and Environmental Protection Agency (EPA) Mitigation Memorandum of Agreement states that, “because the likelihood of success is greater and the impacts to potentially valuable uplands are reduced, restoration should be the first option considered” (Fed. Regist. 60(Nov. 28):58605). The Florida Department of
Environmental Regulation (FDER 1991a) recommends an emphasis on restoration first, then enhancement, and, finally, creation as a last resort. Morgan and Roberts (1999) recommend encouraging the use of more restoration and less creation.
5. Avoid overengineered structures in the wetland's design. Design the system for minimal maintenance. Set initial conditions and let the system develop. Natural systems should be planned to accommodate biological systems. The system of plants, animals, microbes, substrate, and water flows should be developed for self-maintenance and self-design. Whenever possible, avoid manipulating wetland processes using approaches that require continual maintenance. Avoid hydraulic control structures and other engineered structures that are vulnerable to chronic failure and require maintenance and replacement. If necessary to design in structures, such as to prevent erosion until the wetland has developed soil stability, do so using natural features, such as large woody debris. Be aware that more specific habitat designs and planting will be required where rare and endangered species are among the specific restoration targets.
Whenever feasible, use natural recruitment sources for more resilient vegetation establishment. Some systems, especially estuarine wetlands, are rapidly colonized, and natural recruitment is often equivalent or superior to plantings (Dawe et al. 2000). Try to take advantage of native seed banks, and use soil and plant material salvage whenever possible. Consider planting mature plants as supplemental rather than required, with the decision depending on early results from natural recruitment and invasive species occurrence. Evaluate on-site and nearby seed banks to ascertain their viability and response to hydrological conditions. When plant introduction is necessary to promote soil stability and prevent invasive species, the vegetation selected must be appropriate to the site rather than forced to fit external pressures for an ancillary purpose (e.g., preferred wildlife food source or habitat).
6. Pay particular attention to appropriate planting elevation, depth, soil type, and seasonal timing. When the introduction of species is necessary, select appropriate genotypes. Genetic differences within species can affect wetland restoration outcomes, as found by Seliskar (1995), who planted cordgrass (Spartina alterniflora) from Georgia, Delaware, and Massachusetts into a tidal wetland restoration site in Delaware. Different genotypes displayed differences in stem density, stem height, belowground biomass, rooting depth, decomposition rate, and carbohydrate allocation. Beneath the plantings, there were differences in edaphic chlorophyll and invertebrates.
Many sites are deemed compliant once the vegetation community becomes established, as described in the Coyote Creek case study (Appendix B). If a site is still being irrigated or recently stopped being irrigated, the vegetation might not survive. In other cases, plants that are dependent on surface-water input might not have developed deep root systems. When the surface-water input is stopped, the plants decline and eventually die, leaving the mitigation site in poor condition after the Corps has certified the project as compliant.
7. Provide appropriately heterogeneous topography. The need to promote specific hydroperiods to support specific wetland plants and animals means that appropriate elevations and topographic variations must be present in restoration and creation sites. Slight differences in topography (e.g., micro- and mesoscale variations and presence and absence of drainage connections) can alter the timing, frequency, amplitude, and duration of inundation. In the case of some less-studied, restored wetland types, there is little scientific or technical information on natural microtopography (e.g., what causes strings and flarks in patterned fens or how hummocks in fens control local nutrient dynamics and species assemblages and subsurface hydrology are poorly known). In all cases, but especially those with minimal scientific and technical background, the proposed development wetland or appropriate example(s) of the target wetland type should provide a model template for incorporating microtopography.
Plan for elevations that are appropriate to plant and animal communities that are reflected in adjacent or close-by natural systems. In tidal systems, be aware of local variations in tidal flooding regime (e.g., due to freshwater flow and local controls on circulation) that might affect flooding duration and frequency.
8. Pay attention to subsurface conditions, including soil and sediment geochemistry and physics, groundwater quantity and quality, and infaunal communities. Inspect and characterize the soils in some detail to determine their permeability, texture, and stratigraphy. Highly permeable soils are not likely to support a wetland unless water inflow rates or water tables are high. Characterize the general chemical structure and variability of soils, surface water, groundwater, and tides. Even if the wetland is being created or restored primarily for wildlife enhancement, chemicals in the soil and water may be significant, either for wetland productivity or bioaccumulation of toxic materials. At a minimum, these should include chemical attributes that control critical geochemical or biological processes, such as pH, redox, nutrients (nitrogen and phosphorus species), organic content, and suspended matter.
9. Consider complications associated with creation or restoration in seriously degraded or disturbed sites. A seriously degraded wetland, surrounded by an extensively developed landscape, may achieve its maximal function only as an impaired system that requires active management to support natural processes and native species (NRC 1992). It should be recognized, however, that the functional performance of some degraded sites may be optimized by mitigation, and these considerations should be included if the goal of the mitigation is water- or sediment-quality improvement, promotion of rare or endangered species, or other objectives best served by locating a wetland in a disturbed landscape position. Disturbance that is intense, unnatural, or rare can promote extensive invasion by exotic species or at least delay the natural rates of redevelopment. Reintroducing natural hydrology with minimal excavation of soils often promotes alternative pathways of wetland development. It is often advantageous to preserve the integrity of native soils and to avoid deep grading of substrates that may destroy natural below-ground processes and facilitate exotic species colonization (Zedler 1996a,b).
10. Conduct early monitoring as part of adaptive management. Develop a thorough monitoring plan as part of an adaptive management program that provides early indication of potential problems and direction for correction actions. The monitoring of wetland structure, processes, and function from the onset of wetland restoration or creation can indicate potential problems. Process monitoring (e.g., water-level fluctuations, sediment accretion and erosion, plant flowering, and bird nesting) is particularly important because it will likely identify the source of a problem and how it can be remedied. Monitoring and control of nonindigenous species should be a part of any effective adaptive management program. Assessment of wetland performance must be integrated with adaptive management. Both require understanding the processes that drive the structure and characteristics of a developing wetland. Simply documenting the structure (vegetation, sediments, fauna, and nutrients) will not provide the knowledge and guidance required to make adaptive “corrections” when adverse conditions are discovered. Although wetland development may take years to decades, process-based monitoring might provide more sensitive early indicators of whether a mitigation site is proceeding along an appropriate trajectory.
WETLAND FUNCTIONAL ASSESSMENT
The goal of no net loss refers to both wetland acres and wetland function, as the functions contribute to the watershed where the wetland is located. Therefore, when setting compensatory mitigation goals, the functions of a wetland proposed for fill need to be precisely characterized
and, if possible, quantified, as should the functions of the proposed compensatory mitigation project. Even if the mitigation goal does not seek in-kind replacement of functions, functional assessment provides a foundation for considering the watershed consequences of out-of-kind mitigation. Functional assessment helps determine whether the location and design of a compensation wetland will secure the functions that are emphasized for the watershed.
In practice, mitigation attention often is focused on relatively few of the numerous functions that wetlands can provide—for example, habitat, water-quality improvement, and various hydrological functions (groundwater recharge and floodwater desynchronization). The committee does not believe that a science-based functional assessment should be used to assess or rank all the societal values of a wetland. In some cases, technical assessment and the social values of each function have been merged into one assessment procedure. It is recognized that these functions have human value by the societal, economic, and other services they provide and that the values emphasized should be reflected in the location and design of the compensatory wetland. However, the committee believes there are other points in the process of mitigation planning to consider tradeoff among functions where, based on a systematic functional assessment that evaluates all functions objectively, weighting factors can be introduced into the mitigation planning process to consider the broader perspectives about their relative importance (see Chapter 8 for a discussion of this point).
Complete characterization of a compensatory mitigation site requires an assessment of the level of performance attainable for each wetland function under different site designs. This would include consideration of various natural hydrological, geochemical, and ecological attributes and processes. In addition, functional assessment of prospective compensation sites will help establish the design and the monitoring and assessment procedures for the wetland to be created or restored.
Most wetland scientists argue that science-based, regionally standardized procedures are preferable to best professional judgment in comprehensively evaluating wetland function for both impacted and mitigation sites. As a result, the general absence of a uniform approach to assessing wetlands as multifunctional ecosystems have likely encouraged less complex wetland mitigation designs and rudimentary measures of achieving mitigation goals.
THE FLORISTIC APPROACH
In early wetland mitigation efforts, functional assessment was usually confined to lists or qualitative descriptions. Furthermore, although
permit requirements often suggest the need to consider area and function, structural characteristics (usually the amount of vegetation cover) may be used as a criterion to judge whether functional replacement is achieved (e.g., Kentula et al. 1992b). Vegetation structure (e.g., percent cover) is a pervasive example of one structural attribute that is often the default indicator of wetland function. One example of such a singular criterion is the Floristic Quality Assessment developed by Swink and Wilhelm (1979, 1994) for wetlands in the Chicago region and several Midwest states1 (Andreas and Lichvar 1995; Taft et al. 1997; Herman et al. 1996; Mack et al. 2000). The floristic and similar approaches basically characterize a mitigation site solely on the vegetation present. The assumption underlying this approach is that wetland vegetation is a comprehensive indicator of the hydrological and ecological status of the site, and specific vegetation parameters can be used to indicate the functions of the mitigation site. The reasoning behind this approach is that if the vegetation community is healthy and has “natural” species diversity, the ecological components (e.g., physical, biological, and biochemical) that support the vegetation must be present.
In Swink and Wilhelm's (1979, 1994) application of Floristic Quality Assessment, indicators are based on the site in question; thus, a riparian system would have completely different vegetation parameters than a coastal salt marsh. In the Floristic Quality Assessment, each plant species was assigned a coefficient of conservatism (C) ranging from 0 (ubiquitous species) to 10 (species having narrow habitat tolerances), based on the authors' knowledge of the flora of the Chicago region (Swink and Wilhelm 1979, 1994). Other indicators used in past evaluations include percent canopy and/or ground cover, percent survival of specific indicator species, tree height, and species diversity. In many areas, floristic assessment has been the method of choice because vegetation parameters are easy to measure, provide a dramatic visual indicator of compliance (full canopy, tall trees), and allow resource agencies to write well-defined performance criteria for the mitigation sites.
However, the assumptions and premises of the floristic approach are often unclear or incompletely specified when examining the regional spectrum of wetland types. Low plant diversity is not always characteristic of “inferior” hydrogeological and geochemical settings, and high plant diversity is not necessarily a de facto indicator of the multitude of wetland functions (e.g., NRC 1995). Systematic assessment of more than just floristic quality indicators reduces dependence on such speculative assumptions.
D.M.Ladd. The Missouri Floristic Quality Assessment system. The Nature Conservancy, St. Louis, MO, in preparation.
HABITAT EVALUATION PROCEDURES AND THE HYDROGEOMORPHIC APPROACH
The possible array of procedures now available for functional wetland assessment has grown to the point that there is considerable confusion about what is acceptable or preferable and by which regulator or scientist (e.g., 40 procedures are recognized by Bartoldus (1999); see Appendix H). Most procedures are site-specific, with only a few providing assessments at the wetland system or landscape scale. Many are specifically designed to assess one or a few wetland functions, such as fish and wildlife habitat, and lack any procedures to assess other functions or a comprehensive assessment of all functions. Many limit consideration to wetland functions with societal value. Some were developed to generate scores that are scaled to wetland area, such that functions are explicitly assumed to be multiplicative (which is not always the case). Although most use systematic models, many are based on qualitative and often subjective interpretations rather than measurement of discrete variables or parameters. Some procedures, such as habitat evaluation procedures (HEPs) (USFWS 1980, 1981; Sousa 1985), have become operationally codified in regulatory procedures as either required or recommended elements of wetland assessment. HEP was one of the two functional assessment procedures that Bartoldus (1999) considered applicable in all 50 states. Meanwhile, the lack of a broadly accepted, generalized functional assessment procedure as a universal screening tool has led to hybrids that are designed to meet perceived unique needs, such as that for wetland banking (Stein et al. 2000). Among the 40 procedures evaluated by Bartoldus (1999), only seven have been applied or are being considered to establish credits in mitigation banks.
In the mid-1990s, the Corps and Natural Resources Conservation Service (NRCS) agreed to the formal adoption of the hydrogeomorphic (HGM) approach (see Box 7–1) as a uniform procedure for functional assessment in the Clean Water Act's Section 404 program and the U.S. Department of Agriculture programs (Smith et al. 1995). Because it is exclusively based on wetlands and not social processes and has applicability at both the watershed and the landscape scales, HGM was attractive to wetland scientists. It was seen as particularly applicable to wetland mitigation because target hydrology could be based on the influence of water sources, wetland type, and the relative ease or difficulty of establishing certain hydrological regimes. Another of the recognized strengths of HGM is the assessment of functional performance based on a domain of reference systems that capture the presumed optimum natural function. Reference sites are essential for the precise identification of specific wetland attributes and processes for the mitigation site (e.g., hydrology
The Hydrogeomorphic Method
HGM classification (Brinson 1993) is a functional classification and as such differs from other wetland classification systems, such as the Cowardin system, which was designed for use in national wetland inventory mapping (Cowardin et al. 1979). The HGM approach classifies a wetland based on its setting in the landscape, its source of water for the wetland, and the dynamics of the water on-site. Setting in the landscape, in the context of HGM, refers to distinctions among, for example, wetlands that occupy depressions, river flood plains, or estuary fringes. Similarly, by water source, the committee means to distinguish among wetlands that receive surface water, as opposed to those that receive primarily precipitation or groundwater. By dynamics of the water on-site, or flow, the committee is interested in distinguishing wetlands that have unidirectional horizontal flow from those that have vertical flow (upwelling) and those that have horizontally bi-directional flow. HGM classification groups wetlands with similar structure and function and emphasizes features of wetlands that are relatively independent of the biogeographical distribution of species and requires recognition of factors external to the wetland (Brinson 1993). Wetlands within a class are assumed to be hydrogeomorphically and functionally similar and to have functional attributes different from wetlands in other classes. According to the HGM perspective —for example, a slope wetland (i.e., a groundwater-driven wetland) dominated by emergent vegetation is functionally different from a riverine wetland (i.e., a wetland of the active flood plain) with emergent vegetation.
functions in terms of saturation duration, depth, and frequency not only seasonally but also annually). The enhancement of functions (such as control of water levels or flows to enhance vegetation, water quality, or waterfowl habitat) was to be considered outside the domain of reference sites. In addition, fundamental incorporation of reference wetlands meant that assessments were sensitive to regional variations in the functional performance of hydrogeomorphic subclasses. However, in one respect, HGM and similar assessment procedures are still deficient at assessing the effect of wetland mitigation at the landscape scale. Although they may effectively assess the functions of a wetland site in a hydrogeomorphic, landscape setting, these procedures will not necessarily examine whether the development of a wetland will reduce the functional value of adjacent wetlands or put at risk significant other areas.
HGM AS A FUNCTIONAL ASSESSMENT PROCEDURE
The original Corps-sponsored HGM functional assessment procedures have been modified to meet different, often project specific, needs,
but most versions have not been accepted by the Corps (Bartoldus 1999). Although many include elements of the HGM concepts and/or terminology, most deviate substantially from the intent, premises, and design of HGM (Brinson 1995; 1996). Most of these do not include complete data sets, particularly, appropriate reference site data. Examples of these derivative procedures include the Washington State Wetland Function Assessment Project (Hruby et al. 1998), Minnesota Routine Assessment Method for Evaluating Wetland Functions (Minnesota Board of Water and Soil Resources 1998), EPA (Bartoldus et al. 1994), Method for Assessment of Wetland Function (MDE method) (Fugro East Inc. 1995), and the Rapid Assessment Procedure (Magee and Hollands 1998).
However, HGM has many useful applications in functional assessment and the mitigation process in general. For instance, the advantages of evaluating mitigation performance using an assessment of hydrological equivalence (Bedford 1996) based on hydrogeomorphic classification (Brinson 1993) are effectively demonstrated by an EPA Wetlands Research Program evaluation of mitigation projects in the Portland, Oregon, metropolitan area (Gwin et al. 1999; Shaffer et al. 1999; Shaffer and Ernst 1999; Magee et al. 1999) (see Box 7–2). HGM classification analysis revealed seven HGM regional classes in the Portland sample wetlands that were defined using the Cowardin system. Analysis showed that the vegetation, soil, and hydrological variables that were measured differed significantly among HGM classes.
Almost all of the mitigation wetlands belonged to HGM classes that were atypical of the region. There were no naturally occurring analogs for the hydrogeomorphic types they represented (Gwin et al. 1999). This explains why the results presented above comparing features of mitigation wetlands and naturally occurring wetlands pointed to differences between the two groups. One would expect differences in hydrogeomorphic features to be reflected in the soils, vegetation, and, especially, the hydrology. These results demonstrate that the diversity in the hydrogeomorphic characteristics of wetland exemplified in the regional HGM classes is related to the diversity of their extant hydrology (Shaffer et al. 1999). Because hydrology is a critical forcing function for other wetland attributes, changes in hydrology can be assumed to have significant effects on a variety of wetland functions.
However, development and testing of HGM at the regional level are inconsistent, uncoordinated, and dependent on needs and funding. Perhaps more important, as exemplified by the Corps's recent development of HGM regional guidebooks for several wetland classes (Ainslie et al. 1999; Brinson et al. 1995), most variables incorporated into the assessment models remain measures of wetland structure rather than processes. This may be the inevitable result of tension between the costs of doing func-
tional assessment in terms of staff time and funds and the available staff and budgets. Compromises may be necessary. Although HGM as a specific functional assessment procedure may not be meeting expectations and may be too costly to implement in all cases, it has put a focus on the need for assessing wetland function at the landscape scale (see Box 7–2).
Functional Assessment of Hydrological Equivalence Using HGM in the Portland, Oregon, Metropolitan Region
The Portland metropolitan area, located In northwestern Oregon at the north end of the Willamette Valley, was chosen for study of the functionality of mitigation sites because rapid urbanization and development have placed wetlands in the area at high risk for modification and destruction (Holland et al. 1995). The sites sampled were small (~2 ha) palustrine wetlands ranging from those dominated by emergent marsh to those dominated by open water (Cowardin et al. 1979), that is, the wetland types historically most common in the Willamette Valley (Davis 1995; Guard 1995) and most frequently required as mitigation for permitted losses of freshwater wetlands in the Portland area and the State of Oregon (Kentula et al. 1992a,b). The study wetlands were located in a variety of land-use conditions, including urban, agricultural, and undeveloped. Ninety-six sites (45 naturally occurring and 51 mitigation wetlands) were assessed in terms of morphology, hydrology, soils, and vegetation in the summer of 1993. The mitigation wetlands ranged in age from 1 to 9 years, averaging 5. In addition, the hydrological characteristics of approximately half the sites were monitored through January 1997.
EVALUATION USING STRUCTURAL INDICATORS
Characteristics considered desirable in naturally occurring wetlands are commonly used as permit conditions and design criteria for mitigation wetlands. Therefore, a comparison of naturally occurring wetlands and mitigation wetlands, using variables that have or could be used as permit conditions, illustrates the types of results produced by such an approach. For example, one might conclude that the mitigation wetlands in the Portland study could be called compliant based on the characteristics of the plant community. Within 5 years after construction, most mitigation wetlands would have met a criterion of 80% cover per square meter where emergent vegetation occurred on the site. However, only a small portion of the site was vegetated on many of the mitigation wetlands, because most of the sites were occupied by deep open water. The naturally occurring wetlands and mitigation wetlands both had plant communities composed of about 50% native species. On average, a slightly higher percentage of the species per site was native to the mitigation wetlands (mitigation wetlands = 47%; naturally occurring wetlands = 43%). However, the wetland flora in the area, in general, was degraded by the predominance of exotic species. The species composition of naturally occurring wetlands and mitigation wetlands is different (p < 0001) with species richness per
Furthermore, it should be recognized that, as it is currently developed as an assessment tool, HGM is principally a diagnostic method, not a prescriptive “cookbook.” In this respect, the HGM models do not specifically lay out design parameters that guarantee the likelihood that hydrology, desired wetland vegetation, and desired animals will be reestab
site higher on mitigation wetlands (p=.0006). However, the species composition of mitigation wetlands less than 3 years in age differed from that of mitigation wetlands more than 3 years in age due to the influx of introduced species, averaging 11 additional species per site as the site aged. So the mitigation wetlands may not maintain native plant species over time, especially in the face of changes occurring with urbanization of the landscape.
In the case of the organic matter content of the soils, naturally occurring wetlands and mitigation wetlands are significantly different. There is less organic matter in the top 5 cm of the soil of mitigation wetlands (p=.0001) and at 15 to 20 cm (p= 0.0551) than in naturally occurring wetlands. There was no substantive relationship between soil organic matter concentration and the age of mitigation wetlands (r2=.0232, p=.6003). This suggests that development of a soil organic matter content similar to that of naturally occurring wetlands may not be achieved for a very long time, if ever. Finally, the hydrological characteristics of the mitigation wetlands differed from the naturally occurring wetlands. As mentioned above, mitigation wetlands had more open water than naturally occurring wetlands. On average, 57% of the area of the mitigation wetlands was flooded, while 28% of the area of the naturally occurring wetlands was flooded during the year (p < 001). The predominance of deep open water on mitigation wetlands was indicated by higher mean annual water levels (0.85 m) on mitigation wetlands than on naturally occurring wetlands (0.25 m, p < 001). Hydrological variability also differed between naturally occurring wetlands and mitigation wetlands. The mean difference between the 10th and the 90th percentiles of water levels was 0.60 m for naturally occurring wetlands and 0.32 m for mitigation wetlands (p < 01). The difference between the 10th and the 90th percentiles was used to represent conditions commonly found in the wetlands as it minimizes the effects of extreme storm events.
Given the above analysis, conclusions about the performance of the mitigation wetlands would depend on whether only vegetation characteristics were considered and on how one viewed the predominance of alien plant species. In addition, there is evidence that the conclusion might change with time, especially with time periods longer than the 5-year monitoring requirement often associated with permits. Regardless, these kinds of analyses do not overcome the inherent problems of using structural similarity to infer functional equivalence, let alone determining the effects of permit decisions on the resource as a whole. Faced with this dilemma, it was found that the concept of hydrological equivalence as exemplified in HGM classification brought important insights to the evaluation of mitigation wetlands.
lished or the likelihood that exotics will not invade. HGM provides no analytical structure (e.g., a decision matrix) for inserting information about factors that influence the likelihood that hydrology, desired wetland vegetation, and desired animals will be reestablished or that exotics will not invade. This can only come with explicit application and monitoring using HGM in the design of wetland mitigation projects, and the resulting feedback on the correspondence between HGM indicators and monitored performance. This gap between HGM, and most other functional assessment procedures for that matter and the need for specific scientific and technical guidance for self-sustaining, functional mitigation wetlands remains a major hindrance to effective wetland mitigation.
It is possible that there is no single “best” wetland assessment procedure, because the specific needs vary with the situation, especially if a quick screening technique is needed (Smith 1993). However, in the mitigation process it is essential that there be an ability to relate the structural characteristics of a site to the resulting functions. Only in that way can the compensation site be designed to secure certain functions. The level of the function is calculated relative to levels in reference sites in the same subclass of wetlands within the same watershed or ecoregion. Perhaps functional assessments will evolve to meet this goal. The functional assessment procedure has the following desirable attributes:
It includes reliable indicators of the important wetland processes (hydrology, sedimentation, and primary production) or a scientifically established structural surrogate of those processes.
It assesses function over a broad range of performance conditions, such that differences in wetlands can be relatively easily distinguished.
It is integrative over space and time, and its indicators are not vulnerable to seasonal or other fine-scale temporal or spatial variability.
It results in a continuous, parametric scale that has not been reduced to a relative rank.
It assesses all recognized functions so that the assessment encompasses all goals for the mitigation.
1. Dependence on subjective, best professional judgment in assessing wetland function should be replaced by science-based, rapid assessment procedures that incorporate at least the following characteristics:
Effectively assess goals of wetland mitigation projects.
Assess all recognized functions.
Incorporate effects of position in landscape.
Reliably indicate important wetland processes, or at least scientifically established structural surrogates of those processes.
Scale assessment results to results from reference sites.
Are sensitive to changes in performance over a dynamic range.
Are integrative over space and time.
Generate parametric and dimensioned units, rather than nonparametric rank.
2. Impact sites should be evaluated using the same functional assessment tools used for the mitigation site.