25.6 Variability and Ecosystem Dynamics in Assessments
regime of high-frequency, low-magnitude events
(e.g., 2-year return period) that forms the active
channel width also drives formation of pools by
scour through secondary flow processes (Thompson, 1986). The hydrological scour that forms patterns of meander formation and pool scour is considered equilibrium dissipation of energy in
planform and vertical dimensions, respectively
(Langbein and Leopold, 1966) and are regularly
spaced. The optimal spacing of pools is considered
5 to 7 channel widths, and this distributional pattern is considered universal in a wide variety of
bedrock and alluvial streams (Keller and Melhorn,
1978).
Factors that cause variability and deviation from
a normal rhythmic distribution are important for interpretation of stream processes and features and
development of restoration strategies. Channel
planform morphology and pool spacing are altered
by contact with high terraces or valley hillslopes,
variable floodplain, and sediment sizes (Milne,
1982). Features such as bedrock outcrops, large
wood, large boulders, and rootwad-protected banks
are considered large roughness elements and cause
local scour that forms variably sized stationary
pools (Lisle, 1986). The size of pools formed by
these structural features is related to the size of the
feature scaled to the width of the stream (Lisle,
1986). Pool spacing and size relations can be
skewed by large-scale paleohydrological events,
which create large bedforms and pools that are
spaced relative to the magnitude of the event
(O'Conner et al., 1986). Size and density of pools
vary in geomorphic river segments of different
lithology and channel morphology, such as in the
Colorado River, Grand Canyon section (Howard
and Dolan, 1981). Also in this setting, large-pool
distribution is indirectly influenced by coarse-scale
patterns of structural properties such as jointingfaulting, because debris torrents and tributary formation follow fault patterns, and the larger pools
are associated with debris torrent deposits.
The processes and patterns related to habitat formation and distribution differ in steeper, more confined upland streams (Grant et aI., 1990; Montgomery and Buffington, 1997). This is partly
because the predominant bankfull discharge may
not be competent to effectively transport and sort
the larger substrate sizes encountered in smaller upland streams, and thus pools may be poorly or irregularly formed (Miller, 1958; Milne, 1982). In
steeper upland streams with large substrate sizes,
bed entrainment that forms channel units occurs at
higher flows of relatively large magnitude and low
frequency (e.g., 25 year + return period; Grant et
379
al., 1990). Also, hillslope processes and valley features often directly impinge on upland channels
(Grant and Swanson, 1995), which affects habitat
patterns. Channel bedforms and structural features
such as boulders may have a clustered distribution
and are associated with mass wasting processes that
are more prevalent in certain valley reach types
(Benda, 1990; Grant et al., 1990; Grant and Swanson, 1995). The coarse-scale distribution of structural pool types in the Oregon Coast Range reflect
these upland controls and are related to stream
power, essentially channel slope, and drainage area
(Stack and Beschta, 1989). This relation with habitat formation and stream size is demonstrated in
forested streams, where pool spacing is inversely
related to wood loading, and wood influence on
pool spacing decreases with increasing stream
width (watershed size) (Montgomery et aI., 1995).
25.6 Variability and Ecosystem
Dynamics in Assessments
In ecological assessment and restoration, we need
to recognize that aquatic systems are not static entities, but are naturally dynamic. This requires a
framework that incorporates the properties and
processes that both cause and constrain ecosystem
dynamic variability (such as channel dynamics).
Thus we must manage and restore aquatic systems
in context of these dynamics, not necessarily for
some stable or optimum state. Four general types
of variability need to be accounted for in ecological assessments. The first two types are spatial
variability due to change in geoclimatic or biophysical characteristics. Horizontal variability reflects changes across a single spatial scale (e.g.,
changes in lithology, climate). Vertical variability
results from change in scale (e.g., from regional to
local scale). A third type of scale change is change
in size, which in lakes, and especially for streams,
fundamentally affects physical processes. A fourth
general type is temporal variability as a result of
process dynamics. The first two types of spatial
variability are well accounted for in classifications
that emphasize process, form, and composition of
aquatic systems at various spatial scales (Maxwell
et aI., 1995; also see Chapter 24). Although the spatial context of controlling and driving variables are
incorporated in these classifications, the actual
processes and dynamics are usually only implicitly
implied. This process-driven dynamic variability is
emphasized in this section, particularly channel dynamics and climatic fluctuations.
regime of high-frequency, low-magnitude events
(e.g., 2-year return period) that forms the active
channel width also drives formation of pools by
scour through secondary flow processes (Thompson, 1986). The hydrological scour that forms patterns of meander formation and pool scour is considered equilibrium dissipation of energy in
planform and vertical dimensions, respectively
(Langbein and Leopold, 1966) and are regularly
spaced. The optimal spacing of pools is considered
5 to 7 channel widths, and this distributional pattern is considered universal in a wide variety of
bedrock and alluvial streams (Keller and Melhorn,
1978).
Factors that cause variability and deviation from
a normal rhythmic distribution are important for interpretation of stream processes and features and
development of restoration strategies. Channel
planform morphology and pool spacing are altered
by contact with high terraces or valley hillslopes,
variable floodplain, and sediment sizes (Milne,
1982). Features such as bedrock outcrops, large
wood, large boulders, and rootwad-protected banks
are considered large roughness elements and cause
local scour that forms variably sized stationary
pools (Lisle, 1986). The size of pools formed by
these structural features is related to the size of the
feature scaled to the width of the stream (Lisle,
1986). Pool spacing and size relations can be
skewed by large-scale paleohydrological events,
which create large bedforms and pools that are
spaced relative to the magnitude of the event
(O'Conner et al., 1986). Size and density of pools
vary in geomorphic river segments of different
lithology and channel morphology, such as in the
Colorado River, Grand Canyon section (Howard
and Dolan, 1981). Also in this setting, large-pool
distribution is indirectly influenced by coarse-scale
patterns of structural properties such as jointingfaulting, because debris torrents and tributary formation follow fault patterns, and the larger pools
are associated with debris torrent deposits.
The processes and patterns related to habitat formation and distribution differ in steeper, more confined upland streams (Grant et aI., 1990; Montgomery and Buffington, 1997). This is partly
because the predominant bankfull discharge may
not be competent to effectively transport and sort
the larger substrate sizes encountered in smaller upland streams, and thus pools may be poorly or irregularly formed (Miller, 1958; Milne, 1982). In
steeper upland streams with large substrate sizes,
bed entrainment that forms channel units occurs at
higher flows of relatively large magnitude and low
frequency (e.g., 25 year + return period; Grant et
379
al., 1990). Also, hillslope processes and valley features often directly impinge on upland channels
(Grant and Swanson, 1995), which affects habitat
patterns. Channel bedforms and structural features
such as boulders may have a clustered distribution
and are associated with mass wasting processes that
are more prevalent in certain valley reach types
(Benda, 1990; Grant et al., 1990; Grant and Swanson, 1995). The coarse-scale distribution of structural pool types in the Oregon Coast Range reflect
these upland controls and are related to stream
power, essentially channel slope, and drainage area
(Stack and Beschta, 1989). This relation with habitat formation and stream size is demonstrated in
forested streams, where pool spacing is inversely
related to wood loading, and wood influence on
pool spacing decreases with increasing stream
width (watershed size) (Montgomery et aI., 1995).
25.6 Variability and Ecosystem
Dynamics in Assessments
In ecological assessment and restoration, we need
to recognize that aquatic systems are not static entities, but are naturally dynamic. This requires a
framework that incorporates the properties and
processes that both cause and constrain ecosystem
dynamic variability (such as channel dynamics).
Thus we must manage and restore aquatic systems
in context of these dynamics, not necessarily for
some stable or optimum state. Four general types
of variability need to be accounted for in ecological assessments. The first two types are spatial
variability due to change in geoclimatic or biophysical characteristics. Horizontal variability reflects changes across a single spatial scale (e.g.,
changes in lithology, climate). Vertical variability
results from change in scale (e.g., from regional to
local scale). A third type of scale change is change
in size, which in lakes, and especially for streams,
fundamentally affects physical processes. A fourth
general type is temporal variability as a result of
process dynamics. The first two types of spatial
variability are well accounted for in classifications
that emphasize process, form, and composition of
aquatic systems at various spatial scales (Maxwell
et aI., 1995; also see Chapter 24). Although the spatial context of controlling and driving variables are
incorporated in these classifications, the actual
processes and dynamics are usually only implicitly
implied. This process-driven dynamic variability is
emphasized in this section, particularly channel dynamics and climatic fluctuations.