19.12 Summary
Washington. Such models can be used to identify
pre settlement refugia in unsampled areas from their
topographic settings.
3. Fire group models were developed using biophysical environments to characterize the fire
regimes that drive vegetation succession in forested
landscapes of Montana (Fischer and Clayton, 1983;
Fischer and Bradley, 1987).
19.11 Methods for Aquatic
Ecosystems
Some of the previously described methods can be
used to characterize HRV for riparian or aquatic
ecosystems, such as detection of flood events using dendroecological methods or characterization
of aquatic plant species composition using pollen
analysis and plant macrofossils from sediments.
Methods such as repeat photography, maps from
historical data, simulation modeling, and characterization of biophysical environments can be applied to aquatic as well as terrestrial ecosystems.
The streamflow regime, a driving force in river
ecosystems, controls key habitat parameters such
as flow depth, velocity, and habitat volume (Richter
et aI., 1998). Long-term streamflow gauge observations provide information about the quantity,
timing, and variability of a river's flow (Kondolf
and Larson, 1995; Poff et aI., 1997). Sediment cores
extracted from lakes and wetlands characterize
temporal fluctuations in physical properties, such
as sediment size and chemical composition, in addition to presence of fish populations and aquatic
plant species composition and abundance (Uutala,
1990; Sullivan et aI., 1992; Steedman et aI., 1996).
Long-term fish surveys and commercial harvest
data describe fish species composition and abundance over time (McIntosh et aI., 1994; Kelso et
aI., 1996; Patton et aI., 1998). In reconstructing
HRV, consideration should be given to differences
in the response of aquatic and terrestrial ecosystems to particular types of disturbance, such as the
greater effect of flooding and erosion and the less
pronounced effect of fire on aquatic than terrestrial
ecosystems (Frissell and Bayles, 1996).
EXAMPLE
1. Richter et aI. (1997) described a new method
for setting streamflow-based ecosystem management targets on rivers where biodiversity conservation and protection of ecosystem function are
management objectives. The method involves determining the range of variability in 32 hydrologi285
cal parameters derived from daily streamflow values measured during a period with negligible human perturbations to the hydrological regime.
River management targets are identified based on
a statistical characterization of variability in the
suite of hydrological parameters. Predam daily
streamflow data were used to characterize the natural range of streamflow variability for the Roanoke
River, North Carolina, as a basis for recommending modifications in reservoir operations rules for
dams on the river. The method was also used to
evaluate the impacts of dam construction on hydrologic variability in two rivers in the upper Colorado River basin in Colorado and Utah (Richter
et aI., 1998).
19.12 Summary
The methods for characterizing HRV described in
this chapter produce information about the range of
variability in vegetation patterns, disturbance
regimes, or both (Table 19.1). Pollen and plant
macrofossil analyses describe changes in plant
species abundance with time and include both
canopy and understory terrestrial species; pollen
from lake sediments also contains records of
aquatic plant species. Charcoal analysis identifies
fire events. The other methods can potentially
contribute information about both vegetation and
disturbance; canopy species records are predominantly represented. Disturbance regime reconstructions for these methods can include flood events
and outbreaks of insect and disease in addition to
fire. Simulation modeling and characterization of
biophysical environments produce predicted vegetation and disturbance HRV, rather than measuring
the range of actual landscape conditions.
Methods differ in temporal and spatial extent.
Some methods, such as plant macrofossils from
middens, land surveys, repeat ground photography,
and maps from historical data, reconstruct historical conditions at one or a few time periods. Such
snapshots in time cannot characterize temporal
fluctuations. Wherever possible, these methods
should be used in conjunction with others that have
the potential to provide sequences of change over
long time periods, such as dendroecological methods, pollen analysis, macrofossils from sediment,
charcoal analyses, or simulation modeling. Reconstruction of long-term temporal variation is limited
by loss of records with time, the "fading record"
problem (Swetnam et aI., 1999). This problem is
prominent in dendroecological methods, in which
records are lost through tree death and decay, but
Washington. Such models can be used to identify
pre settlement refugia in unsampled areas from their
topographic settings.
3. Fire group models were developed using biophysical environments to characterize the fire
regimes that drive vegetation succession in forested
landscapes of Montana (Fischer and Clayton, 1983;
Fischer and Bradley, 1987).
19.11 Methods for Aquatic
Ecosystems
Some of the previously described methods can be
used to characterize HRV for riparian or aquatic
ecosystems, such as detection of flood events using dendroecological methods or characterization
of aquatic plant species composition using pollen
analysis and plant macrofossils from sediments.
Methods such as repeat photography, maps from
historical data, simulation modeling, and characterization of biophysical environments can be applied to aquatic as well as terrestrial ecosystems.
The streamflow regime, a driving force in river
ecosystems, controls key habitat parameters such
as flow depth, velocity, and habitat volume (Richter
et aI., 1998). Long-term streamflow gauge observations provide information about the quantity,
timing, and variability of a river's flow (Kondolf
and Larson, 1995; Poff et aI., 1997). Sediment cores
extracted from lakes and wetlands characterize
temporal fluctuations in physical properties, such
as sediment size and chemical composition, in addition to presence of fish populations and aquatic
plant species composition and abundance (Uutala,
1990; Sullivan et aI., 1992; Steedman et aI., 1996).
Long-term fish surveys and commercial harvest
data describe fish species composition and abundance over time (McIntosh et aI., 1994; Kelso et
aI., 1996; Patton et aI., 1998). In reconstructing
HRV, consideration should be given to differences
in the response of aquatic and terrestrial ecosystems to particular types of disturbance, such as the
greater effect of flooding and erosion and the less
pronounced effect of fire on aquatic than terrestrial
ecosystems (Frissell and Bayles, 1996).
EXAMPLE
1. Richter et aI. (1997) described a new method
for setting streamflow-based ecosystem management targets on rivers where biodiversity conservation and protection of ecosystem function are
management objectives. The method involves determining the range of variability in 32 hydrologi285
cal parameters derived from daily streamflow values measured during a period with negligible human perturbations to the hydrological regime.
River management targets are identified based on
a statistical characterization of variability in the
suite of hydrological parameters. Predam daily
streamflow data were used to characterize the natural range of streamflow variability for the Roanoke
River, North Carolina, as a basis for recommending modifications in reservoir operations rules for
dams on the river. The method was also used to
evaluate the impacts of dam construction on hydrologic variability in two rivers in the upper Colorado River basin in Colorado and Utah (Richter
et aI., 1998).
19.12 Summary
The methods for characterizing HRV described in
this chapter produce information about the range of
variability in vegetation patterns, disturbance
regimes, or both (Table 19.1). Pollen and plant
macrofossil analyses describe changes in plant
species abundance with time and include both
canopy and understory terrestrial species; pollen
from lake sediments also contains records of
aquatic plant species. Charcoal analysis identifies
fire events. The other methods can potentially
contribute information about both vegetation and
disturbance; canopy species records are predominantly represented. Disturbance regime reconstructions for these methods can include flood events
and outbreaks of insect and disease in addition to
fire. Simulation modeling and characterization of
biophysical environments produce predicted vegetation and disturbance HRV, rather than measuring
the range of actual landscape conditions.
Methods differ in temporal and spatial extent.
Some methods, such as plant macrofossils from
middens, land surveys, repeat ground photography,
and maps from historical data, reconstruct historical conditions at one or a few time periods. Such
snapshots in time cannot characterize temporal
fluctuations. Wherever possible, these methods
should be used in conjunction with others that have
the potential to provide sequences of change over
long time periods, such as dendroecological methods, pollen analysis, macrofossils from sediment,
charcoal analyses, or simulation modeling. Reconstruction of long-term temporal variation is limited
by loss of records with time, the "fading record"
problem (Swetnam et aI., 1999). This problem is
prominent in dendroecological methods, in which
records are lost through tree death and decay, but