2.2 Reference Site Screening
Reference site screening is the process of estimating the degree of anthropogenic
impairment at study sites, such that those with acceptably low stress levels may be
designated as reference sites. Screening criteria vary among studies based on both
data availability and on the opinions of investigators as to which criteria are most
appropriate. Some have advocated for the use of professional judgment regarding
whether observed assemblages represent reference conditions [44, 45]. Professional
judgment may also be applied to the environmental conditions at the sites, providing a more independent, and potentially less biased, means of estimating impairment status [46]. Advocates of professional judgment often favor consensus
opinions among groups of experts [45, 47]. Potential criticisms of such approaches
include a lack of empirical support for decisions regarding reference designations
and circular logic when sites are designated as reference based on the assemblage
that is also used for bioassessment. However, Davies and Jackson [45] recently
showed that the opinions of trained experts were highly consistent when rating
ecological integrity based on assemblage data, leading the authors to assert that
well-informed professional opinions provide reliable, ecologically relevant benchmarks for bioassessment.
Independent stressor variables (i.e., those not related to the assemblage used for
bioassessment) are often used to estimate impairment status. Screening approaches
that employ both professional judgment and independent stressor variables produce
reference sites that are of higher ecological integrity than those selected using either
approach alone [48]. Variables used to estimate anthropogenic stress include
physicochemical water quality parameters, land-cover features derived using a
geographic information system (GIS) that indicate development, and physical
habitat quality assessments based on site observations. Studies often employ a
filtering approach, whereby threshold levels for each measured stressor variable
are set to designate sites as reference or impaired [21, 49]. The filtering criteria used
to select reference sites are generally evaluated with an all approach for reference
sites and any approach for non-reference sites. For example, Blocksom et al. [49]
required that sites meet all of the reference criteria listed in Table 2 to be designated
as reference, but considered sites impaired if any one of the impaired criteria was
met.
The level of anthropogenic impairment considered acceptable varies greatly
among studies, and many authors fail to provide clear descriptions of how reference
conditions are defined. Recognizing this problem, Stoddard et al. [32] advocated for
the use of the terms minimally disturbed condition (MDC) to describe expected
conditions in the absence of substantive anthropogenic disturbance, least-disturbed
condition (LDC) to describe the best available conditions present under current
disturbance regimes, and best attainable conditions (BAC) to describe the expected
conditions when all avoidable sources of anthropogenic influence are removed
(BAC is generally intermediate between LDC and MDC). MDC sites, within
virtually unimpacted, near pristine watersheds, are rarely available, and most
Principles for the Development of Contemporary Bioassessment Indices for. . .
239
Reference site screening is the process of estimating the degree of anthropogenic
impairment at study sites, such that those with acceptably low stress levels may be
designated as reference sites. Screening criteria vary among studies based on both
data availability and on the opinions of investigators as to which criteria are most
appropriate. Some have advocated for the use of professional judgment regarding
whether observed assemblages represent reference conditions [44, 45]. Professional
judgment may also be applied to the environmental conditions at the sites, providing a more independent, and potentially less biased, means of estimating impairment status [46]. Advocates of professional judgment often favor consensus
opinions among groups of experts [45, 47]. Potential criticisms of such approaches
include a lack of empirical support for decisions regarding reference designations
and circular logic when sites are designated as reference based on the assemblage
that is also used for bioassessment. However, Davies and Jackson [45] recently
showed that the opinions of trained experts were highly consistent when rating
ecological integrity based on assemblage data, leading the authors to assert that
well-informed professional opinions provide reliable, ecologically relevant benchmarks for bioassessment.
Independent stressor variables (i.e., those not related to the assemblage used for
bioassessment) are often used to estimate impairment status. Screening approaches
that employ both professional judgment and independent stressor variables produce
reference sites that are of higher ecological integrity than those selected using either
approach alone [48]. Variables used to estimate anthropogenic stress include
physicochemical water quality parameters, land-cover features derived using a
geographic information system (GIS) that indicate development, and physical
habitat quality assessments based on site observations. Studies often employ a
filtering approach, whereby threshold levels for each measured stressor variable
are set to designate sites as reference or impaired [21, 49]. The filtering criteria used
to select reference sites are generally evaluated with an all approach for reference
sites and any approach for non-reference sites. For example, Blocksom et al. [49]
required that sites meet all of the reference criteria listed in Table 2 to be designated
as reference, but considered sites impaired if any one of the impaired criteria was
met.
The level of anthropogenic impairment considered acceptable varies greatly
among studies, and many authors fail to provide clear descriptions of how reference
conditions are defined. Recognizing this problem, Stoddard et al. [32] advocated for
the use of the terms minimally disturbed condition (MDC) to describe expected
conditions in the absence of substantive anthropogenic disturbance, least-disturbed
condition (LDC) to describe the best available conditions present under current
disturbance regimes, and best attainable conditions (BAC) to describe the expected
conditions when all avoidable sources of anthropogenic influence are removed
(BAC is generally intermediate between LDC and MDC). MDC sites, within
virtually unimpacted, near pristine watersheds, are rarely available, and most
Principles for the Development of Contemporary Bioassessment Indices for. . .
239