Results
- wildlife
This section of the Results addresses wildlife; it is organized as follows:
In this section, we present
analysis results on selected individual wildlife species and their habitat
performance under the planning alternatives at three time periods.
Wildlife habitat type maps are the
basis for determining habitat performance in the United States portion of the
Columbia River Basin. Two maps (Figure
IV.B.1 and Figure
IV.B.2), compiled by the Northwest Habitat Institute, illustrate
historic conditions (circa 1850) and current conditions. Habitat performance was evaluated by
comparing the amounts of various wildlife habitats for historic, current, and future
(i.e., alternative strategies) conditions. Each of the 32 wildlife habitat
types is depicted as a colored polygon with each color representing a
terrestrial, freshwater or marine habitat type. This representation of wildlife
habitat types is the first of its kind for the U. S. portion of the basin
(discussions with Canadian biologists are proceeding to continue the mapping
effort for the whole basin). This
consistent mapping effort will result in a hierarchical analysis of the
provinces, subbasins, and watersheds.
The results for habitat performance
are summarized for a few of the habitats (i.e., shrub steppe, agriculture, and
eastside mixed conifer) that illustrate great change since the 1850s. All wildlife habitats and their value for
wildlife are integrated into the Habitat Condition Indices and Functional
Analyses presented in the following sections. A summary of these data (Table
IV.B.1) is the basis for
illustrating changes between historic, current, and alternative strategies for
the basin and provinces.
The historic wildlife habitat type
map (Figure
IV.B.1) was created to illustrate the norm circa 1850 for the U.S.
portion of the Columbia Basin. This map provides an idea of what the historic
(and future) potential for an area might be.
The historic map is coarse at 1.6 miles (1 km) resolution compared to
the finer scale current map. The coarse scale map at the basin and the province
levels can under-represent some wildlife habitat types. For example, the
historic map indicates relatively large wetland areas in the Willamette and
Snake River valleys. Narrow riparian wetlands along small streams in headwater
situations are not represented at the coarse scale. The fact that these narrow
wetlands are not shown means that they are underrepresented in the historic map
and that there is less difference shown between historic and current conditions
than likely what actually occurred. Other wildlife habitats such as shrub
steppe are less likely under represented on the historic map and changes
between historic and current are likely more representative of actual changes.
The historic wildlife habitat types
serve as a reference for wildlife restoration across the basin. Management Activities or strategies that change
current conditions toward the historic condition are likely to restore wildlife
habitat types for native wildlife species and communities. The historic and
current maps also provide insight to ecosystem processes that have resulted in
ecosystem change. Insights into
ecosystem processes and functions are likely to help understand and guide the
direction (but not the detail) of what a future alternative might be.
General amounts of the three
examples of wildlife habitat types in the basin are represented by the grand
total column in Figure
IV.B.3. The shrub-steppe habitat comprised about 28
percent, east-side mixed conifer about 14 percent, and agriculture was nearly
absent in the basin under historic conditions.
The detail of the amount of these and the other wildlife habitat in the
basin will come later in the analyses for the subbasin assessments. The historic wildlife habitat types are most
interesting when compared to the map of current wildlife habitats (see below).
The province scale analysis of
wildlife habitat types was conducted for three wildlife habitat types as an
example of the type of analysis that might be conducted to assess differences
in wildlife habitat across the basin, and as a basis for assessing change in
space and time (Figure
IV.B.4).
Eastside mixed conifer habitat occurs in all provinces with the
exception of the Lower Columbia. This
habitat is more common in the Columbia Gorge, Inter Mountain, Mountain
Columbia, and Mountain Snake Provinces.
Shrub steppe habitat is also present in nine of the ten provinces. It is
most common in the Columbia Plateau, Middle Snake, and Upper Snake
Provinces. The agricultural habitat
type is close to zero in all provinces.
Wildlife habitat types mapped for
the current conditions (Figure
IV.B.2) are depicted at a minimum mapping unit
of 250 acres (100 ha). The most notable
changes from the historic map are: (1) conversion of the shrub steppe and dwarf
shrub steppe to agriculture, (2) conversion of the Willamette and Snake River
Valley wetlands and grasslands to agriculture, and (3) conversion of eastside
ponderosa pine forest to mixed conifer forest (due to fire suppression,
selective logging, and grazing).
Conversion of wetlands is detectable for large areas such as the
Willamette Valley and the Vancouver Lake area along the Lower Columbia. These large changes give managers a
perspective of the general magnitude and location of changes that have
occurred. The minimum mapping unit of the historic map precludes an accurate
representation of the relatively narrow (i.e., less than 1,000 feet wide)
historic wetlands that occurred along many of the smaller tributaries that were
likely important to beaver and salmon in historic times. Accurate analyses of wetland and riparian
changes will have to await later analyses at the subbasin and watershed scales.
The grand total percents of the
three habitats at the basin scale indicate shrub steppe is just over 20
percent, and eastside mixed conifer is just under 20 percent of the basin (Figure
IV.B.3). The most dramatic
change between historic and current conditions is the increase of 23.5 million
acres of agriculture (Figure
IV.B.5). A relatively small portion of this change
came from shrub steppe wildlife habitat type. Other wildlife habitat types such
as grassland, forest and dwarf shrub steppe have also been converted to
agriculture (Hessburg et al. 2000, Huff et al. 1995).
Changes in shrub-steppe and
eastside mixed conifer wildlife habitat types likely are better (than wetlands)
represented at the province scale. The
percent of these wildlife habitat types for current conditions (Figure
IV.B.3)
in the various provinces indicates where conversions to agriculture are the
greatest. For example, about 6 percent
(0.4 million acres) of the Mountain Snake province has been converted to
agriculture whereas almost 30 percent (3.9 million acres) of the Columbia
Plateau has been converted to agriculture.
Eastside mixed conifer forest conversion to agriculture is most
pronounced in the Mountain Columbia (about 45 percent). These changes are best illustrated in Figure
IV.B.6 where the largest increases in agricultural acreage are in the Columbia
Plateau and the Mountain Snake provinces.
Given the large conversions to agriculture in these provinces, it is not
surprising that this is where there was the largest reduction in shrub steppe
acres. The province analysis also indicates shrub steppe acres did not decrease
in all provinces and actually increased in five provinces (e.g., Mountain
Snake) along with agriculture. Subbasin analyses in these provinces should
address the reasons for these increases in shrub steppe. Eastside conifer
forest increases at the basin level can be attributed to provinces on the west
slope of the Rocky Mountains (e.g., Mountain Snake) but not all provinces had
increases in this wildlife habitat type.
A decrease in acres of this wildlife habitat type occurred in the Upper
Snake and Columbia Plateau.
Wildlife habitat types estimated by
Vail et al. 2001 (Figure
IV.B.7) clearly show a loss of over 10 million acres
of the agriculture wildlife habitat type in the future under all three
alternatives. Alternative 2, which addresses dam removal, reduces the
agriculture habitat slightly less than the other two alternatives that do not
propose dam removal. Alternative 6 reduces the agriculture habitat slightly
more than Alternative 5 (Table
IV.B.1), a slight increase in eastside conifer
forest is approximately equal for each alternative. The decrease in shrub
steppe is slightly greater in Alternative 2 than the other two alternatives.
The changes in wildlife habitat
types are similar for each alternative (generally less than 10 percent
difference between each alternative for each wildlife habitat type). Given this
similarity among alternatives, the province scale analysis focuses on one
alternative with the knowledge that trends discussed apply to all alternatives.
Alternative 2 shows that the changes in the agriculture wildlife habitat type
are quite different for each province (Figure
IV.B.8). One province, Blue
Mountain, showed a 651-acre increase and the other nine provinces showed
decreases. The decreases varied from 3 to 4.6 million acres in the Upper Snake
and Columbia Plateau to about 11 thousand acres in the Columbia Gorge.
The changes in shrub steppe also varied across the provinces with increases in
the Columbia Plateau and Upper Snake provinces. The largest decrease (1.6
million acres) in the shrub steppe habitat was in the Middle Snake. Decreases
in other provinces were less than half a million acres. Eastside conifer forest
increased in nine of the ten provinces.
The largest gain, 209 thousand acres, was in the Inter Mountain
province. One province, the Mountain Snake, had a slight (13 thousand acre)
decrease.
Biological performance for the
black bear and the bald eagle were assessed using a Habitat Condition Index
(HCI) to estimate capacity (see Methods).
As discussed in Methods, necessary fine-scale data on riparian and
aquatic habitats were not available for calculating an HCI for the American
beaver. HCI results for the black bear
and the bald eagle were calculated for each 6-HUC (in the range of the species)
and are presented here in three formats: HCI maps, cumulative integrated
capacity curves, and HCI change maps.
6-HUC information was aggregated at two levels for analysis: the basin
and province. Two types of presentation are illustrated for the species level
analysis. The black bear analysis is
very general, and utilizes HCI maps and cumulative integrated capacity curves.
The bald eagle analysis relies on HCI maps, change maps, and histograms.
The HCI maps plot an HCI value for
each 6-HUC. The highest HCI values are
represented as dark green and the shade of green lightens as the HCI values
decrease with white equivalent to zero.
White 6-HUCs represent areas outside the range of the species. Change maps have been prepared to illustrate
where the greatest and least changes are expected. Dark red shows the greatest negative changes while the pink and
white 6-HUCs illustrate the least negative change. Dark blue shows the greatest positive change while the light blue
shows the least positive change. This
analysis is especially good for alerting managers to possible problem areas for
proposed alternatives.
HCI calculations for historic
wildlife capacity are shown on the HCI maps as dark green areas (Figure
IV.B.9)
where one would expect black bear to have been abundant in the 1850s. For example, the Cascade Range from central
Oregon to Canada and the western front range of the Rocky Mountains in Idaho
and Montana show the darkest green 6-HUCs.
Areas where bears have never been abundant such as southeastern Oregon
and southern Idaho are white. The
current wildlife capacity (Figure
IV.B.10) shows less (than historic) dark
green in the above areas and noticeable absence of bears in populated and
agricultural areas (e.g., Willamette Valley.)
This is best illustrated in a comparison of the cumulative integrated
capacity curves for historic and current wildlife habitat types (Figure
IV.B.11 and Figure
IV.B.12). The historic curve shows 1000 6-HUCs with a 0.92 (or
greater) HCI whereas the current curve shows a general reduction in HCI value
with 1000 6-HUCs valued at 0.87 (or greater) HCI.
All three alternatives resulted in
HCI maps that are similar and hence are not illustrated. Little change was observed at the basin
scale between alternatives due to the relatively small amount of proposed
change in forested wildlife habitats. A
comparison between current black bear HCI and the alternatives is also slight
and difficult to detect with HCI maps at the basin level. Cumulative integrated capacity curves
illustrate the subtle differences between current (Figure
IV.B.12) and the
alternatives (Figure
IV.B.13), represented as Alternative 6. The alternatives
show a small increase in higher value 6-HUCs (probably due to an increase in
carcasses in 75 6-HUCs) but a larger increase in lower value 6-HUCs (i.e.,
below 0.4 HCI). This indicates that the
alternatives could have a positive impact on lower quality black bear habitat. The positive impact illustrated in the
cumulative integrated capacity curves is likely due to the projected reduction
in roading. The HCI assessment method
(i.e., the literature) gives considerable emphasis to the negative influence of
roads on black bear.
As one considers the results of the
black bear analysis it is important to remember the coarse scale of the
analysis and not rely on the results for decision making regarding fish and
wildlife recovery. Of greater
importance is the result that indicates little is being done to enhance black
bear habitat in forested environments and possible consequences of such an
alternative strategy for fish. For
example, this result might stimulate a fisheries biologist to ask if there will
there be adequate forests to produce large woody debris for future aquatic
habitat improvements. If as a result of
this analysis the alternative is modified to include strategies in forested
environments, the fish biologists might coordinate the location (i.e.,
landscape) for these activities to provide the most benefit for fish as well as
higher quality bear habitat. In
addition, managers at the subbasin scale of analysis should be aware that
decisions to benefit fish while beneficial for bear in some places could be
detrimental for bear in other places.
Forest structure data that may be available for subbasin analyses
(especially in stringers through shrub-steppe habitats) will likely be
important for examining potential benefits for fish as well as bear.
Fish and wildlife interaction is a
key interest of the framework analysis.
The black bear HCI analysis includes the fish carcass variable that
allows a simple but important interaction between fish distribution and black
bear habitat value. The importance of
fish carcass data can be further examined at the subbasin scale by expanding
the carcass variable to include carcass abundance (versus presence or absence
in the current HCI analysis) and seasonal use across the landscape.
The bald eagle analysis is a little
more detailed that the black bear analysis and illustrates how species-specific
data might be presented in a more quantitative format. The use of data in Table IV.B.1 and
histograms helps to identify areas where alternatives might have a negative influence
on this threatened species.
HCI calculations for the historic
wildlife capacity show few dark green 6-HUCs and relatively few 6-HUCs with low
(i.e., 0.10) HCI values (Figure
IV.B.14).
Some of this is due to the large areas of shrub steppe (poor bald eagle
habitat). Another explanation is the
general lack of fine scale information on wetlands, especially narrow riparian
wetlands, from the 1850s. For example,
much of the narrow (i.e., less than a kilometer wide) riparian stringer
wetlands that were likely present in historic conditions are under-represented
at a coarse-mapping resolution of 1 kilometer.
Areas along major rivers such as the Willamette and Snake are probably accurately
represented on the historic map but few of these areas have high HCIs. The linear nature of suitable wildlife
habitat for the bald eagle is not conducive to averaging across provinces for
either historic or current times. Instead, the analysis focuses on change in
distribution and percent of 6-HUCs increasing or decreasing in HCI value from
historic to current.
The current wildlife capacity for
the bald eagle (Figure
IV.B.15) shows more dark green 6-HUCs (than Historic)
and a wider distribution of colored 6-HUCs.
For example, the Columbia Plateau and north-central Oregon are mostly
lighter shades of green. The cumulative
integrated capacity curves for the historic (Figure
IV.B.16) and current (Figure
IV.B.17) clearly illustrate the above points. The historic curve is truncated showing less than half of the
6-HUCs occupied by bald eagles, while the current curve (Figure
IV.B.17)
extends to the right showing more 6-HUCs occupied and more HCI values between
0.25 and 0.80. Approximately 56 percent of the 6-HUCs showed in increase in HCI
from historic to current. During the same period about 30 percent of the 6-HUCs
showed a decrease.
The percent positive change in HCI
for each province from historic to current condition was above 75 percent for
the Columbia Cascade, Mountain Columbia and Upper Snake (Figure
IV.B.18). All
of the other provinces showed some moderate positive percent changes. Each
province also showed some 6-HUCs with negative changes. The three provinces
that showed the most positive percent change also showed the least negative
change. The province with the largest
negative percent change was the Columbia Gorge (77 percent) followed by the
Blue Mountain, Inter Mountain, Lower Columbia, and Mountain Snake, which were
all around -60 percent.
HCI maps produced for the three
alternatives are very similar. Figure IV.B.19 indicates all three alternatives
showed 29 percent positive and 47 percent negative change for current to
alternative conditions. In addition there is a larger percent change for
current to alternatives than there was for historic to current. Thus it appears that the alternatives could
potentially have a negative influence on bald eagles across the basin.
A closer look at the difference between Alternative 2 and the
current map using a change detection map (Figure
IV.B.20) shows where the
negative influences might occur.
The change map (Figure
IV.B.20)
illustrates changes in HCI values for each 6-HUC and only the lower 25 percent
and the upper 25 percent of change detected is plotted as red (negative) or
blue (positive). The changes illustrated
are small but indicate concentrations of red in the Columbia Plateau, the
Willamette (Lower Columbia province) and Snake (Upper Snake province)
Rivers. Much of the red color crosses
province and subbasin boundaries and as a consequence efforts to address areas
of potential concern need to be coordinated among the managers. Dark blue areas
are interspersed across the basin with slightly higher occurrence along the
Cascade Range (Mountain Columbia, Figure
IV.B.21). Light blue is also interspersed across the basin with slightly
higher occurrence along the front range of the Rocky Mountains. The reason for the possible negative
influence of alternatives on bald eagles could be due to a number of factors
such as the coarse scale of the data (i.e., wetland/riparian stringers being
under represented or the emphasis on feeding habitat versus breeding habitat
since nest site information was not available for the whole basin) or the
conversion of agriculture to shrub steppe (which is not associated with bald
eagle use.)
The most important result is that
the concept of possible negative influences on bald eagles needs to be carried
into the Multi-Species Framework analyses at the subbasin level of the
hierarchical analysis. Finer scale and
more complete data (e.g., nest locations) at the subbasin level will provide
better insight into possible negative influences for the bald eagle as well as
wildlife in general. For example, the
sandhill crane is associated with farmed fields especially during migration
through the Basin. They forage in
wetlands associated with agricultural fields and eat grains that remain in
stubble fields. Loss of these
agricultural and associated resources, without mitigative compensation, could
be particularly harsh for species such as the sandhill crane whose numbers have
been declining in recent years. Mitigation should not only consider the loss of
habitat for species such as the crane but also the interim loss of wetland and
riparian habitats that have established with contemporary management. Thus biological objectives for alternatives
should also include interim biological objectives (especially for rare wildlife
habitats) that reflect the time to implement strategies as well as the
relationship to adjacent management areas (Palik et al. 2000).
The American beaver was selected as
a species to assess because of the obvious (Schlosser and Kallemeyn 2000)
association with aquatic ecosystems as well as fish diversity and
abundance. An HCI assessment of the
American beaver included 3 components: physical condition, cover, and
food. Data collection for this species
was problematic and thus the assessment met with failure. This was because beaver is linked to small
(as well as large) streams in headwater areas of the Columbia Basin, and their
streams and associated riparian environments were not mappable at the scale of
this assessment. This is where the
cycle of dam building and dam decay influence hydraulic and fish habitat
diversity. In the headwater areas of
the basin, these processes occur at a relatively fine scale compared to the
coarse scale used for whole basin analyses.
After months of searching and trial runs, we found that there was no
consistent habitat data for the upper portions of most 6-HUCs. For example, gradient (low gradient is an
important habitat element for beaver) data was available only for the main
channels of rivers such as the John Day.
While beaver occur in these main channels their habitat is usually a
bank den rather than the dam and lodge habitat in low gradient headwater wetland
complexes. Consequently the gradient
for main stems was only good for assessing bank denning, which was not
particularly related to dam building and fish habitat diversity. We tried to develop indices for headwater
gradient such as “sinuosity index” or miles of stream per square mile of the
6-HUC. In all cases, data were either
not available or if available they were in a format significantly different
from other portions of the basin. For
this and other reasons, we concluded that the beaver HCI analyses, unlike the
bear analyses, are fine-scale dependent.
We encourage subbasin managers to retain the beaver as an evaluation species and to seek the finer-scale data necessary to evaluate beaver habitat quality. As beaver analyses are made at the subbasin and watershed scales, these analyses can be aggregated up to the subbasin and province scales. See Appendix F for the beaver habitat assessment method for HCI developed by the EWG.