Geography Reference
In-Depth Information
by 9-24% relative to ca. 1990. Instead of continued
decline projected for the Plan Trend 2050 and Devel-
opment 2050 scenarios, ecological endpoints for aquatic
communities recovered 20-65% of the losses observed
for these indicators since EuroAmerican settlement.
Terrestrial wildlife also responded positively under the
Conservation 2050 scenario (Figure 12.4). More species
(31% of the modeled species) gained habitat than lost
habitat relative to ca. 1990 (Schumaker et al., 2004).
The majority of wildlife species modeled for population
abundance increased by at least 10% relative to ca. 1990,
and only the mourning dove decreased by more than
10%. Like aquatic communities, most wildlife species
not only slowed their rate of decline but also recovered
a substantial amount of their historical abundance and
distribution.
The Conservation 2050 scenario consumed less water
than Plan Trend 2050 and Development 2050 with no
water availability basins having zero flow. Though con-
sumption was lower in this scenario, length of stream
channels that dry in summer increased by 70% compared
to ca. 1990 (225 km of 2nd to 4th order streams). The
region is likely to face even greater water shortages in the
future unless more stringent water conservation measures
are adopted.
values for each stream reach or pixel. However, we knew
that resource managers and decision makers faced key
resource decisions in subbasins within the larger basin.
Our overall projections of the consequences of alternative
future scenarios would not be adequate for allocating con-
servation and restoration actions at smaller spatial extents.
One alternative to reach-by-reach depictions of habitat
conditions or resource abundance is subbasin analyses
that highlight differences between major portions of a
larger landscape.
We aggregated the analyses of resource responses for
the landscape scenarios for the 12 major subbasins of
the WRB (Figure 12.5; Branscomb et al., 2002). These
12 subbasins were comprised of 77 fifth-field hydrologic
units (USGS 1997) and differed in land use, land cover,
and population density. Two measures of aquatic habitat
(habitat suitability index, large wood volume) and three
biotic responses (fish richness, cutthroat trout abun-
dance, abundance of EPT macroinvertebrate taxa) were
projected for 4,045 reaches of the WRB (see Van Sickle
et al., 2004 for model descriptions and empirical data on
which models were based).
The habitat suitability index (HSI) model for cutthroat
trout was an expert-based model derived from stream
ecosystem function and habitat use by cutthroat trout.
Separate HSI models were developed for lowland and
upland streams with watersheds predominately in the
Willamette Valley Ecoregion (lowland streams) versus
streams with watersheds predominately in the Cascade
or Coast Range Ecoregions (upland streams). HSI was
calculated from a weighted model of stream gradient,
annual mean flow, valley floor constraint, wood potential,
closed forest in riparian network, % natural vegetation
in riparian network, closed forest in the watershed,
road density in the watershed, % development land in
riparian network, and % agriculture in riparian network.
Wood volume represented the potential contribution
of large wood to the stream from the riparian area
immediately adjacent to the reach, calculated as the
weighted sum of 11 different forest vegetation classes
with older conifers weighted most heavily (Van Sickle
et al., 2004, Gregory et al., 2003). Native Fish Richness
was a function of stream order, distance along the stream
network from the sources, elevation, stream gradient,
percent agricultural land in the 120-m riparian area, and
percent developed land in the riparian area (Gregory
et al., 2002c). Cutthroat trout abundance was a function
of watershed area, percent agricultural land in the 120-m
riparian area, and percent developed land in the riparian
12.7 Informing decision makers
at subbasin extents
We developed representations of ecological resource
abundance for stream reaches with minimum mapping
units of 30-m pixels for the five scenarios of past, present,
and alternative future landscapes (Schumaker et al., 2004,
Van Sickle et al., 2004, Hulse et al., 2002). One of the
social challenges in illustrating data from remote sensing
or ecological models at the scale of river basins is the 'My
Backyard' response. When presented with spatial infor-
mation, it is human nature to examine the map to see the
characteristics of the portion of the landscape the observer
knows best, often their own home, neighbourhood, or
community. Even though overall error or uncertainty at
the landscape scale is often relatively low, the potential for
error in any single pixel or fine scale mapping unit is much
greater. Regardless of the level of accuracy at the basin or
landscape scale, most commonly the consistency between
mapped representations and the individual's personal
experience for a small subset of the basin determines their
confidence in the maps, models, and landscape analyses.
For that reason, we were reluctant to depict numerical
Search WWH ::
Custom Search