to account for this variation. Therefore, the RCA calls for the use of multiple
reference sites in order to adequately account for the effects of natural environmental variation on assemblages [30–32].
We use the term reference to broadly encompass streams considered to be in
least-impaired conditions, though as discussed in Sect. 2.2, the actual level of
impairment at reference sites is highly variable among studies [32]. We use the
term impaired to refer, in general, to sites that are subject to the deleterious effects
of stressors and the term test sites to refer to those of unknown impairment status
(i.e., those for which bioassessments are needed). This chapter focuses on the RCA,
although alternative approaches for estimating reference conditions may be used
when system conditions and data availability warrant (Sect. 2.1). Application of the
RCA proceeds by first screening multiple potential reference sites to determine if
they reflect appropriate least-impaired conditions (Sect. 2.2), then classifying the
screened reference sites as to expected or quantified patterns of variability among
their biotic assemblages (Sect. 2.3).
2.1 Alternatives to the Reference Condition Approach
Though not often available, data describing past assemblages may provide valuable
information for inferring reference conditions. Investigations of sediment records,
historical accounts of landscape conditions, and museum records have been used to
infer past environmental conditions and assemblage composition in aquatic systems
[33–36]. Historical approaches, while important, lack broad applicability for
bioassessment. In lotic ecosystems, sediment deposition is generally insufficient
to provide a historical record. Moreover, data that describe assemblage composition
before anthropogenic development occurred may reflect conditions that are no
longer attainable given the effects of factors acting at large spatial and temporal
scales such as atmospheric deposition of pollutants and global climate change.
When anthropogenic impacts are spatially discernable, a paired-site approach
may be useful. For example, lotic sites impacted by point source pollution such as
mine effluent [37, 38] or municipal wastewater [39] may be paired with upstream
sites above the source of stressors. Plafkin et al. [40] referred to paired upstream
sites as controls, though this implies that confounding natural environmental factors
on the upstream–downstream comparison are being actively controlled, which is
generally not feasible. For bioassessments conducted over large spatial scales (e.g.,
ecoregions), the paired-site approach is problematic because much assemblage
variation is driven by local-scale differences in natural environmental variables
[41, 42]. Assemblage variation among reference sites, that is, variation that is not
likely caused by impairment, is typically much larger than variation among replicate samples collected at a single site. Therefore, comparisons of replicate samples
from a single reference and test site pair often do not provide a realistic representation of the effects of impairment on assemblage characteristics [43].
238
A.L. Garey and L.A. Smock
reference sites in order to adequately account for the effects of natural environmental variation on assemblages [30–32].
We use the term reference to broadly encompass streams considered to be in
least-impaired conditions, though as discussed in Sect. 2.2, the actual level of
impairment at reference sites is highly variable among studies [32]. We use the
term impaired to refer, in general, to sites that are subject to the deleterious effects
of stressors and the term test sites to refer to those of unknown impairment status
(i.e., those for which bioassessments are needed). This chapter focuses on the RCA,
although alternative approaches for estimating reference conditions may be used
when system conditions and data availability warrant (Sect. 2.1). Application of the
RCA proceeds by first screening multiple potential reference sites to determine if
they reflect appropriate least-impaired conditions (Sect. 2.2), then classifying the
screened reference sites as to expected or quantified patterns of variability among
their biotic assemblages (Sect. 2.3).
2.1 Alternatives to the Reference Condition Approach
Though not often available, data describing past assemblages may provide valuable
information for inferring reference conditions. Investigations of sediment records,
historical accounts of landscape conditions, and museum records have been used to
infer past environmental conditions and assemblage composition in aquatic systems
[33–36]. Historical approaches, while important, lack broad applicability for
bioassessment. In lotic ecosystems, sediment deposition is generally insufficient
to provide a historical record. Moreover, data that describe assemblage composition
before anthropogenic development occurred may reflect conditions that are no
longer attainable given the effects of factors acting at large spatial and temporal
scales such as atmospheric deposition of pollutants and global climate change.
When anthropogenic impacts are spatially discernable, a paired-site approach
may be useful. For example, lotic sites impacted by point source pollution such as
mine effluent [37, 38] or municipal wastewater [39] may be paired with upstream
sites above the source of stressors. Plafkin et al. [40] referred to paired upstream
sites as controls, though this implies that confounding natural environmental factors
on the upstream–downstream comparison are being actively controlled, which is
generally not feasible. For bioassessments conducted over large spatial scales (e.g.,
ecoregions), the paired-site approach is problematic because much assemblage
variation is driven by local-scale differences in natural environmental variables
[41, 42]. Assemblage variation among reference sites, that is, variation that is not
likely caused by impairment, is typically much larger than variation among replicate samples collected at a single site. Therefore, comparisons of replicate samples
from a single reference and test site pair often do not provide a realistic representation of the effects of impairment on assemblage characteristics [43].
238
A.L. Garey and L.A. Smock