![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Chapter 5 - The Implementation Process
The implementation of biological criteria requires: (1) selection of unimpaired (minimal impact) surface waters to use as the reference condition for each designated use, (2) measurement of the structure and function of aquatic communities in reference surface waters to establish biological criteria, and (3) establishment of a protocol to compare the biological criteria to biota in impacted waters to determine whether impairment has occurred. These elements serve as an interactive network that is particularly important during early development of biological criteria where rapid accumulation of information is effective for refining both designated uses and developing biological criteria values. The following chapters describe these three essential elements.
The
Reference Condition
A key step in developing values for supporting narrative and creating
numeric biological criteria is to establish reference conditions; it is
an essential feature of environmental impact evaluations (Green 1979).
Reference conditions are critical for environmental assessments because
standard experimental controls are rarely available. For most surface
waters, baseline data were not collected prior to an impact, thus impairment
must be inferred from differences between the impact site and established
references. Reference conditions describe the characteristics of waterbody
segments least impaired by human activities and are used to define attainable
biological or habitat conditions.
Wide variability among natural surface waters across the country resulting from climatic, landform, and other geographic differences prevents the development of nationwide reference conditions. Most States are also too heterogeneous for single reference conditions. Thus, each State, and when appropriate, groups of States, will be responsible for selecting and evaluating reference waters within the State to establish biological criteria for a given surface water type or category of designated use. At least seven methods for estimating attainable conditions for streams have been identified (Hughes et al. 1986). Many of these can apply to other surface waters. References may be established by defining models of attainable conditions based on historical data or unimpaired habitat (e.g., streams in old growth forest). The reference condition established as before-after comparisons or concurrent mea-sures of the reference water and impact sites can be based on empirical data (Hall et al. 1989).
Currently, two principal approaches are used for establishing the reference condition. A State may opt to (1) identify site-specific reference sites for each evaluation of impact or (2) select ecologically similar regional reference sites for comparison with impacted sites within the same region. Both approaches depend on evaluations of habitats to ensure that waters with similar habitats are compared. The designation of discrete habitat types is more fully developed for streams and rivers. Development of habitat types for lakes, wetlands, and estuaries is ongoing.
Site-Specific
Reference Condition
A site-specific reference condition, frequently used to evaluate the
impacts from a point discharge, is best for surface waters with a strong
directional flow such as in streams and rivers (the upstream-downstream
approach). However, it can also be used for other surface waters where
gradients in contaminant concentration occur based on proximity to a source
(the near field-far field approach). Establishment of a site-specific
reference condition requires the availability of comparable habitat within
the same waterbody in both the reference location and the impacted area.
A site-specific reference condition is difficult to establish if (1) diffuse nonpoint source pollution contaminates most of the water body; (2) modifications to the channel, shoreline, or bottom substrate are extensive; (3) point sources occur at multiple locations on the waterbody; or (4) habitat characteristics differ significantly between possible reference locations and the impact site (Hughes et al. 1986; Plafkin et al. 1989). In these cases, site-specific reference conditions could result in underestimates of impairment. Despite limitations, the use of site-specific reference conditions is often the method of choice for point source discharges and certain waterbodies, particularly when the relative impairments from different local impacts need to be determined.
The
Upstream-Downstream Reference Condition
The upstream-downstream reference condition is best applied to streams
and rivers where the habitat characteristics of the waterbody above the
point of discharge are similar to the habitat characteristics of the stream
below the point of discharge. One standard procedure is to characterize
the biotic condition just above the discharge point (accounting for possible
upstream circulation) to establish the reference condition. The condition
below the discharge is also measured at several sites. If significant
differences are found between these measures, impairment of the biota
from the discharge is indicated. Since measurements of resident biota
taken in any two sites are expected to differ because of natural variation,
more than one biological assessment for both upstream and downstream sites
is often needed to be confident in conclusions drawn from these data (Green,
1979). However, as more data are collected by a State, and particularly
if regional characteristics of the waterbodies are incorporated, the basis
for determining impairment from site-specific upstream-downstream assessments
may require fewer individual samples. The same measures made below the
"recovery zone" downstream from the discharge will help define
where recovery occurs.
The upstream-downstream reference condition should be used with discretion since the reference condition may be impaired from impacts upstream from the point source of interest. In these cases it is important to discriminate between individual point source impact versus overall impairment of the system. When overall impairment occurs, the resident biota may be sufficiently impaired to make it impossible to detect the effect of the target point source discharger.
The approach can be cost effective when one biological assessment of the upstream reference condition adequately reflects the attainable condition of the impacted site. However, routine comparisons may require assessments of several upstream sites to adequately describe the natural variability of reference biota. Even so, measuring a series of site-specific references will likely continue to be the method of choice for certain point source discharges, especially where the relative impairments from different local impacts need to be determined.
The
Near Field-Far Field Reference Condition
The near field-far field reference condition is effective for establishing
a reference condition in surface waters other than rivers and streams
and is particularly applicable for unique waterbodies (e.g., estuaries
such as Puget Sound may not have comparable estuaries for comparison).
To apply this method, two variables are measured (1) habitat characteristics,
and (2) gradient of impairment. For reference waters to be identified
within the same waterbody, sufficient size is necessary to separate the
reference from the impact area so that a gradient of impact exists. At
the same time, habitat characteristics must be comparable.
Although not fully developed, this approach may provide an effective way to establish biological criteria for estuaries, large lakes, or wetlands. For example, estuarine habitats could be defined and possible reference waters identified using physical and chemical variables like those selected by the Chesapeake Bay Program (U.S. EPA 1987a, e.g., substrate type, salinity, pH) to establish comparable subhabitats in an estuary. To determine those areas least impaired, a "mussel watch" program like that used in Narragansett Bay (i.e., captive mussels are used as indicators of contamination, (Phelps 1988)) could establish impairment gradients. These two measures, when combined, could form the basis for selecting specific habitat types in areas of least impairment to establish the reference condition.
Regional
Reference Conditions
Some of the limitations of site-specific reference conditions can
be overcome by using regional reference conditions that are based on the
assumption that surface waters integrate the character of the land they
drain. Waterbodies within the same watershed in the same region should
be more similar to each other than to those within watersheds in different
regions. Based on these assumptions, a distribution of aquatic regions
can be developed based on ecological features that directly or indirectly
relate to water quality and quantity, such as soil type, vegetation (land
cover), land-surface form, climate, and land use. Maps that incorporate
several of these features will provide a general purpose broad scale ecoregional
framework (Gallant et al. 1989).
Regions of ecological similarity are based on hydrologic, climatic, geologic, or other relevant geographic variables that influence the nature of biota in surface waters. To establish a regional reference condition, surface waters of similar habitat type are identified in definable ecological regions. The biological integrity of these reference waters is determined to establish the reference condition and develop biological criteria. These criteria are then used to assess impacted surface waters in the same watershed or region. There are two forms of regional reference conditions: (1) paired watersheds and (2) ecoregions.
Paired
Watershed Reference Conditions
Paired watershed reference conditions are established to evaluate
impaired waterbodies, often impacted by multiple sources. When the majority
of a waterbody is impaired, the upstream-downstream or near field-far
field reference condition does not provide an adequate representation
of the unimpaired condition of aquatic communities for the waterbody.
Paired watershed reference conditions are established by identifying unimpaired
surface waters within the same or very similar local watershed that is
of comparable type and habitat. Variables to consider when selecting the
watershed reference condition include absence of human disturbance, waterbody
size and other physical characteristics, surrounding vegetation, and others
as described in the "Regional Reference Site Selection" feature.
This method has been successfully applied (e.g., Hughes 1985) and is an approach used in Rapid Bioassessment Protocols (Plafkin et al. 1989). State use of this approach results in good reference conditions that can be used immediately in current programs. This approach has the added benefit of promoting the development of a database on high quality waters in the State that could form the foundation for establishing larger regional references (e.g., ecoregions.)
Ecoregional
Reference Conditions
Reference conditions can also be developed on a larger scale. For
these references, waterbodies of similar type are identified in regions
of ecological similarity. To establish a regional reference condition,
a set of surface waters of similar habitat type are identified in each
ecological region. These sites must represent similar habitat type and
be representative of the region. As with other reference conditions, the
biological integrity of selected reference waters is determined to establish
the reference. Biological criteria can then be developed and used to assess
impacted surface waters in the same region. Before reference conditions
may be established, regions of ecological similarity must be defined.
One frequently used method is described by Omernik (1987) who combined maps of land-surface form, soil, potential natural vegetation, and land use within the conterminous United States to generate a map of aquatic ecoregions for the country. He also developed more detailed regional maps. The ecoregions defined by Omernik have been evaluated for streams and small rivers in Arkansas (Rohm et al. 1987), Ohio (Larsen et al. 1986; Whittier et al. 1987), Oregon (Whittier et al. 1988), Colorado (Gallant et al. 1989), and Wisconsin (Lyons 1989) and for lakes in Minnesota (Heiskary et al. 1987). State ecoregion maps were developed for Colorado (Gallant et al. 1989) and Oregon (Clarke et al. mss). Maps for the national ecoregions and six multi-state maps of more detailed ecoregions are available from the U.S. EPA Environmental Research Laboratory, Corvallis, Oregon.
Ecoregions such as those defined by Omernik (1987) provide only a first step in establishing regional reference sites for development of the reference condition. Field site evaluation is required to account for the inherent variability within each ecoregion. A general method for selecting reference sites for streams has been described (Hughes et al. 1986). These are the same variables used for comparable watershed reference site selection. Regional and on-site evaluations of biological factors help determine specific sites that best represent typical but unimpaired surface water habitats within the region. Details on this approach for streams is described in the "Regional Reference Site Selection" feature. To date, the regional approach has been tested on streams, rivers, and lakes. The method appears applicable for assessing other inland ecosystems. To apply this approach to wetlands and estuaries will require additional evaluation based on the relevant ecological features of these ecosystems (e.g. Brooks and Hughes, 1988).
Ideally, ecoregional reference sites should be as little disturbed as possible, yet represent waterbodies for which they are to serve as reference waters. These sites may serve as references for a large number of similar waterbodies (e.g., several reference streams may be used to define the reference condition for numerous physically separate streams if the reference streams contain the same range of stream morphology, substrate, and flow of the other streams within the same ecological region).
An important benefit of a regional reference system is the establishment of a baseline condition for the least impacted surface waters within the dominant land use pattern of the region. In many areas a return to pristine, or presettlement, conditions is impossible, and goals for waterbodies in extensively developed regions could reflect this. Regional reference sites based on the least impacted sites within a region will help water quality programs restore and protect the environment in a way that is ecologically feasible.
This approach must be used with caution for two reasons. First, in many urban, industrial, or heavily developed agricultural regions, even the least impacted sites are seriously degraded. Basing standards or criteria on such sites will set standards too low if these high levels of environmental degradation are considered acceptable or adequate. In such degraded regions, alternative sources for the regional reference may be needed (e.g., measures taken from the same region in a less developed neighboring State or historical records from the region before serious impact occurred). Second, in some regions the minimally-impacted sites are not typical of most sites in the region and may have remained unimpaired precisely because they are unique. These two considerations emphasize the need to select reference sites very carefully, based on solid quantitative data interpreted by professionals familiar with the biota of the region.
Each State, or groups of States, can select a series of regional reference sites that represent the attainable conditions for each region. Once biological criteria are established using this approach, the cost for evaluating local impairments is often lower than a series of measures of site-specific reference sites. Using paired watershed reference conditions immediately in regulatory programs will provide the added benefit of building a database for the development of regions of ecological similarity.
Regional
Reference Site Selection
To determine specific regional reference sites for streams, candidate
watersheds are selected from the appropriate maps and evaluated to determine
if they are typical for the region. An evaluation of level of human disturbance
is made and a number of relatively undisturbed reference sites are selected
from the candidate sites. Generally, watersheds are chosen as regional
reference sites when they fall entirely within typical areas of the region.
Candidate sites are then selected by aerial and ground surveys. Identification
of candidate sites is based on: (1) absence of human disturbance, (2)
stream size, (3) type of stream channel, (4) location within a natural
or political refuge, and (5) historical records of resident biota and
possible migration barriers.
Final selection of reference sites depends on a determination of minimal disturbance derived from habitat evaluation made during site visits. For example, indicators of good quality streams in forested ecoregions include: (1) extensive, old, natural riparian vegetation; (2) relatively high het-erogeneity in channel width and depth; (3) abundant large woody debris, coarse bottom sub- strate, or extensive aquatic or overhanging vege-tation; (4) relatively high or constant discharge; (5) relatively clear waters with natural color and odor; (6) abundant diatom, insect, and fish assemblages; and (7) the presence of piscivorous birds and mammals.