May 28,  1991   Draft -  Do not quote  or cite
    POTENTIAL IMPACTS OF CLIMATE CHANGE ON PACIFIC NORTHWEST
                                   FORESTS
                                 George A. King
                   ManTech Environmental Technology Incorporated
                                      and
                                  David Tingey
                          Environmental Protection Agency
                     US EPA Environmental Research Laboratory
                               200 SW 35th Street
                               Corvallis, OR 97333
This document is a preliminary  draft.   It has not  been  formally released  by the  U.S.
Environmental Protection Agency and should not be construed to represent Agency policy. It
is being circulated for comments on its technical merit and policy implications. Do not release.
Do not quote or cite.

-------
May  28,  1991   Draft  - Do  not quote or  cite

                            TABLE OF CONTENTS

FINDINGS	  i

INTRODUCTION 	    1

CURRENT STATUS OF NORTHWEST FORESTS  	    1
      Common Forest Types	    1
      Current Land Use  	    2
      Future Resource Condition Without Climate Change	    2

SENSITIVITY OF FORESTS TO CURRENT AND FUTURE CLIMATE	    3
      Current Climate	    3
      Adaptation of Forests to Current Climate	    4
      Sensitivity to Future Climate Change	    4

METHODOLOGY 	    5
      Generation of Climate Scenarios	    5
      Vegetation Modeling - Application of the Holdridge Life-Zone Classification ....    6
      Vegetation Modeling - Local Climate-Forest Zone Correlations	    7
      Vegetation Modeling - Forest Gap Models	    7
      Carbon Model Results 	    9

RESULTS  	    9
      Holdridge Vegetation Scenarios	    9
            Simulation of Current Vegetation  	    9
            Estimated Vegetation Change	    9
            Assumptions/Limitations  	   10
      Local Climate-Forest Zone Correlations	   10
      Forest Gap Models - CLIMACS Results  	   11
            Simulation of Current Forests	   11
            Estimated Changes in Forest Compositions	   11
            Assumptions/Limitations  	   11
      Forest Gap Models - ZELIG Results  	   12
            Simulations of Current Vegetation	   12
            Estimated Changes in Forest Composition  	   13
            Assumptions/Limitations  	   14
      Potential Changes in Carbon Storage	   14
      Expert Judgement - Effects of Climate Change on Forest Disturbances  	   15
      Major Uncertainties	   16
            Transient Dynamics	   16
            Direct Effects of C02  	   17

CONCLUSIONS  	.t	   19

-------
May  28,  1991   Draft - Do  not quote  or cite

      Climate Induced Changes in Forest Composition and Distribution	   19
       	   19
      Implications for Forest Management 	   19

ACKNOWLEDGEMENTS  	   21

LITERATURE CITED	   22

-------
May  28,  1991   Draft  -  Do  not  quote  or  cite
               FINDINGS

Trace  gas induced climate  change could
significantly alter the climate of the Pacific
Northwest. Mean annual temperatures could
increase 2° to 5° C.   The seasonality of
precipitation will likely remain the same, but
with annual totals remaining unchanged or
increasing 20%.

       The potential effects of these climate
changes on Northwest  forests  have been
estimated  using a  variety  of  modeling
approaches and climate scenarios.  Overall,
26  to  90% of  the area in the Northwest
would change from  one general vegetation
type  to  another.    Forest  area  in  the
Northwest could  decrease   5  to  25%.
Remaining  forest  land would  differ in
species composition,  and likely  be  less
productive than current forests.  In Oregon,
Douglas-fir dominated forests would increase
in area, whereas the more productive western
hemlock  forests  would decrease.   Forest
vegetation zones would increase in elevation
from 500 to  1000m. Alpine and subalpine
forests  could  disappear from  all but the
highest elevations in the region.

       Forest disturbances  such as fire and
pest/pathogen    outbreaks   would   likely
increase in frequency, speeding vegetation
change in response to climate change.

       There  are two key limitations to the
data presented here.    First, the transient
(time-dependent) dynamics of change have
not been adequately addressed.  How forest
dynamics respond to a rapidly changing but
variable climate is uncertain.  Second, the
direct effects of enhance«jCO2 concentrations
on  forest  species  growth  has  not  been
considered   in  any   of  the  modeling
simulations. Laboratory experiments suggest
the potential for increased drought tolerance
by  individual  plants  under  higher  C02
concentrations.  The landscape scale impacts
of higher C02 concentrations  are uncertain.
       Forest managers are  thus presented
with a difficult problem.  Specifically, how
should current forests be managed given 1)
our uncertainty  as  to  the  magnitude  and
direction of future climate change and 2) the
potential  for   large   changes   in   forest
composition and distribution if  the climate
does change as currently simulated by state-
of-the-art climate models?

-------
May 28,  1991   Draft  - Do not quote  or cite
                                  INTRODUCTION

       The forests of the northwestern United States are dominated by evergreen conifers that
are long-lived and grow to sizes unmatched in other parts of the world (Waring and Franklin
1979, Franklin 1988).  Biomass storage is higher than any other vegetation type in the world
(Waring and Franklin 1979, Franklin 1988, Harmon et al. 1990). In sum, these forests are unique
among north temperate forest types, in which deciduous trees are usually the dominant life-form.
Besides their intrinsic value as unique ecological systems, Northwest forests serve as an important
source of timber and are a focus of the tourism industry.  The potential for significant changes
in the regional climate caused by global increases in radiatively important  trace gases (IPCC
1990) has raised concern over the future of Northwest forests.  The purpose of this chapter is to
summarize our current understanding of the potential effects of climate change on this important
regional resource.  We  will focus on the forests in Washington and Oregon, with some results
and analysis applicable in  the state of Idaho.

                  CURRENT STATUS OF NORTHWEST FORESTS

Common Forest Types

       Forests occupy 72 million acres of land in Washington, Oregon, and Idaho, or about 46%
of the region (Figure 1, Table la,b).  Forests occupy higher elevations east of the Cascades, and
dominate the land west of the Cascades.  Douglas-fir (see Table 2 for  scientific nomenclature)
and western hemlock are the most important trees in forests ranging from sea-level to 700-1000
m west of the Cascades (Franklin 1988).  Douglas-fir typically dominates younger forest stands,
with hemlock and western  red cedar increasing in importance after 400-600 year*  Grand fir and
western white pine ef&ebe minor components of forests west of the Cascades.  In central Oregon,
incense cedar, sugar pine,  and ponderosa pine may occur. Hardwoods  are rare in these forests,
occupying recently disturbed sites or riparian zones.  Big leaf maple and red alder are the most
common  hardwood species. On drier sites, Pacific madrone, golden chinkapin and Oregon white
oak are common (Franklin 1988).  Understory composition varies  widely in this forest zone,
controlled generally by moisture availability (Franklin 1988).

       Along the Pacific coast, a narrow band of Sitka spruce - western hemlock forest prevails.
Sitka spruce is only found within a few tens of kilometers of the coast, forming pure stands along
the coast (or with lodgepole pine)  because of its  salt tolerance.  Hemlock is usually more
abundant, however (Franklin 1988).

       At higher elevations in the Cascades and  Olympic Mountains, mixed conifer forests occur
(Franklin 1988). Common tree species include silver fir, western hemlock, noble fir, Douglas-fir,
western red cedar, and western white pine.  At higher elevations and towards the interior,

-------
May 28,  1991   Draft  -  Do  not  quote  or  cite

lodgepole pine, subalpine fir, Englemann spruce, western larch, mountain hemlock, and Port-
Orford-Cedar can be present. Western hemlock often forms pure stands at the highest elevations
in the Cascades and Olympic mountains that can support forests.

       Complex geology and steep gradients in moisture and temperature make the Klamath
Mountains in southwestern Oregon and northern California the most diverse forested region in
the study area (Whittaker 1960, Franklin 1988).  Mixed conifer and evergreen hardwood forests
are typical here.  Douglas-fir is usually the dominant conifer, and the evergreen trees tanoak,
canyon live oak, Pacific madrone, and golden chinkapin can form a hardwood understory.

       Carbon storage in old growth Douglas-fir - western hemlock forests averages over 600
Mg/ha, while second growth 60 year old forests average 250 - 275 Mg/ha (Harmon et al. 1990).
Productivity in eastern Oregon forests (east of the Cascade crest) is significantly lower.

       In contrast to conifer forests elsewhere in North America, disturbances are infrequent in
Pacific Northwest forests. The natural fire rotation in these forests is about 400-500 years and
tends to be shorter in the southern part of the range (Franklin  1988). Wind storms can produce
widespread tree mortality, but  are of lesser importance in the Cascades than along the coastal
mountains. Pest and pathogen outbreaks are  not as frequent as in other western conifer  forests
(Franklin  1988).
Current Land Use

       About 75% of the forest land in the Northwest are managed for timber production (Table
1, USDA 1989).  Over the region, about 63% of the timberland is publically owned, although
in Washington State timberland is equally divided between public and private ownership. Forest
industry owns about half of the private forest lands in the Northwest
                                                     
-------
May 28,  1991   Draft  - Do  not quote  or  cite

timber supplies in the future, while maintaining large tracts of heretofore unprotected old growth
forest.

       Another management challenge is to assess the environmental effects of the conversion
of natural stands into managed plantations (Swanson et al. in press).  A key issue is how to
disperse  cutting patterns on a forested landscape.  Franklin and Foreman (1987) analyze  the
effects of various cutting strategies on  landscape patterns.  An important result was  that  the
widely used practice  of dispersed 10-20 ha  cuts through a virgin landscape  will eventually
fragment the remaining forests into smaller  and smaller contiguous  tracts.  This effectively
reduces the amount of interior forest available as habitat for old growth dependent  species.
Te&hniques at modelng landscape patterns have been developed to help evaluate alternative
cuJfmg patterns (Hemstrom 1990, Cissel 1990, Swanson et al. in press).

       SENSITIVITY OF FORESTS TO CURRENT AND FUTURE CLIMATE

Current Climate

       "Climatically the [Pacific Northwest] region experiences wet mild winters and warm dry
summers. The dormant season, when  shoot  growth  is inactive,  is  characterized  by heavy
precipitation with daytime  temperatures usually above freezing. Away from the coast,  the
growing  season is characterized by warm temperatures, clear days, and little precipitation. Water
storage in snowpack, soils and vegetation - as well as pulses of fog, clouds, or cool maritime air
which reduce  evapotranspiration - obviously are more important during a summer drought"
(Waring  and Franklin, 1979). The forests, in this region, grow under a wide range of temperature
and moisture regimes. Mean annual temperature range is about 5° C, while precipitation varies
about an order of magnitude (approximately 300 cm) (Franklin et al., in press).

       In general, both summer and winter temperatures decrease from south to north (Fig.  2).
The north to south temperature gradient is steeper in the winter than the summer.  Focusing again
on the region west of the Cascade creset, at Gold Beach, OR the mean January temperature is
8.3° C while at Forks, WA  it is 3.8° C.  In  the summer, mean July temperatures are  18.0° C at
Andrews, and  15.7° C at Forks, WA.  Similarly, there is a large decrease in  the duration of  the
growing  season which ranges from 365 days at Gold Beach, OR to 281 days at Forks, WA.
Similarly, the  growing season becomes shorter and  the temperatures lower as  the elevation
increases. Summer (mean monthly) temperatures only reach 13.2° and  12.4° C at Rainier/Paradise,
WA and Crater Lake, OR with 5 to 6 months displaying mean monthly temperatures below  0°
C.
                                                                          t>
      There is a distinct seasonality in precipitation with the winters tending to be wet and  the
summers dry (Figure 3). Depending on the  temperature and elevation, the precipitation falls as
either rain or snow; significant snow pack occurs at higher elevations.  During the winter, monthly
precipitation ranges from 40 to 80 cm, while during the summer months precipitation can be less

-------
May 28,  1991   Draft  - Do not  quote  or  cite

than 10 cm/month. There are stronger orographic and elevational influences on precipitation than
a south to north influence. Rain  shadow effects are particularly evident east of the crest of the
Olympic Mountains and Cascade Mountains.

Adaptation of Forests to Current  Climate

       The strong seasonality of climate in the Northwest is a strong determinant of regional
vegetation patterns. Evergreen conifers are well-adapted to this climatic regime, in contrast to
hardwood species.  These adaptations  are summarized by Waring and Franklin (1979) and 1)
include the ability to photosynthesize under the cool winter temperatures, 2) needle-shaped leaves
which  reduce leaf temperature and respiration during the warm dry summer, 3) a large volume
of sapwood that stores water which can be utilized during dry summer days,  and 4) the ability
to reestablish their vascular water column if it's  broken during evaporative stress (whereas many
hardwood species can not; Sucoff 1969, Burke  et al. 1976, Sakai 1983).

       Within the region soil moisture  and temperature are important controls of ecotones
between forest types (Daubenmire 1943,  Zobel  et al. 1986, Franklin 1988).  Lower elevational
zones are controlled primarily by soil moisture, whereas upper elevational zones are controlled
by temperature.  Snowpack and wind can be important controls at uppermost elevations in the
Cascades and Olympic Mountains (Scott 1980). Also, Leverenz  and Lev (1987) point out the
seed chilling  requirement needed by Douglas-fir for successful regeneration in coastal regions.
Sensitivity to Future Climate Change

       A  logical extension of the hypothesis  that climate controls  the dominant vegetation
patterns in the region is simply that future climate change will affect regional vegetation patterns.
The key question is: how much can climate change before major vegetation redistribution takes
place? One way to answer this is to look at the response of vegetation to past climate variation.
Within  the historical record, short-term droughts had little impact on vegetation (Graumlich
1987).  Over the past few centuries, droughts similar to those  of the 1920s  and 1930s occurred
at least once every century since 1675, as determined by tree ring analysis (Graumlich  1987).
This variation obviously affected tree growth, but did not change regional  vegetation patterns.
       In the longer term, temperatures were about 2° C warmer during the early  Holocene
(10,000 - 7,000 years before present) (Brubaker  1988) during which time the vegetation was
much different from today's. Douglas-fir was more important than today in western Washington,
and oak savanna extended north of its present limit in the Williamette Valley in  Oregon to the
southern part of the Puget Trough (Barnosky et al. 1987). Thus it appears that climate changes
of greater than  2° C will likely have significant impact on regional vegetation patterns.

-------
May 28,  1991   Draft  - Do not quote or cite

       A complicating factor in assessing forest sensitivity to climate change is the long life
spans of trees in the Northwest (500 - 1000+ years; Franklin 1988; Brubaker 1986, Franklin et
al. in press).  Adverse climate conditions may eliminate seedling establishment  and sexual
reproduction, while mature trees  survive for  several centuries.  Trees which  can reproduce
asexually may persist "out of equilibrium" with climate for thousands of years (e.g. Neilson and
Wullstein 1983). The key definition that needs to be  made here is "adverse".  It most certainly
has an  upper and lower bound in this context, since mature trees can be killed by severe climate
conditions as well.  Unfortunately,  the autoecological databases for making such a determination
are too incomplete at  this time for defining environmental tolerances  of seedlings and mature
trees of many species  in the Northwest  However, modeling exercises using forest gap models
(discussed below) may help define the term for certain species.

                                  METHODOLOGY

       The sensitivity of Northwest forests to climate  change has been analyzed using a variety
of modeling techniques as well as  expert judgement based on an understanding of how climate
controls the current distribution of  forest types in the region.  As will be described, the modeling
analyses are complimentary, as no single vegetation model can provide the specific detail and
geographic context needed in an impacts assessment.  The specific limitations and assumptions
of the specific modeling analyses will be described in the results sections. Also, expert opinions
on climate change impacts (e.g. Franklin et al. in press) are also discussed in the results section.

Generation of Climate Scenarios

       Four low resolution general circulation models (GCMs),  OSU (Schlesinger and  Zhao
1989),  GISS (Hansen  et al.  1983), GFDL (Manabe and Wetherald 1987), and UKMO (Wilson
and Mitchell 1987) were used to project future climate conditions assuming  a doubling of
"current" ambient CO2 concentrations. Because the grid sizes of each model (typically about 4°
latitude by 5° longitude) is large relative to the  area of the  Pacific Northwest, we can not expect
these models to exactly reproduce  the current or projected climate but rather to provide insight
into changes over  broad  regions when  the atmospheric CO2 is  increased (Jenne,  1988).  To
convert the low resolution GCM output to specific locations, the ratios of 2X: IX CO2 model runs
were used to^multiple^current mean monthly temperature and precipitation for  each location.
Temperatures were first convened to °K before calculating the ratios.  Current climate data for
selected locations in Washington and Oregon was provided by Charles B. Halpern, Department
of Forest Science, Oregon State University from long-tem NOAA summaries. The length of the
time-series used to create the mean values varies with location, and ranged between 30 and 60
years.

       The overall  effect of climate change in the Pacific Northwest |s that forests will become
significantly warmer.  The GCMs  project a 2° to 5°  C increase  in  annual  mean temperatures
(Table  3) with projected temperatures decreasing from south to north.  The UKMO predicts the

-------
May  28,  1991   Draft  - Do not quote or cite

greatest warming, with January temperatures increasing 5-6° C, and July temperatures increasing
9-11° C.  The temperature seasonality is projected to persist under climate warming with all
months showing an increase (Figure 2). Growing degree days, a measure of the cumulative
increase in temperature over the growing season, is projected to increase significantly (Table 3).
The OSU GCM projects a 37 to 73% increase in growing degree/days while the GISS model
projects significantly more warming (83 to  133%).

       Under the projected climate change, there are no seasonal shifts in precipitation;  the
pattern  of relatively dry summers and wet  winters will persist (Figure  3).  Although  the
seasonality is  maintained, the proportions of rain and snow may change from the current
conditions. Two of the GCMS (OSU and GFDL) predict unchanged annual precipitation, while
the other two (GISS and UKMO) predict about 20% greater annual precipitation. In terms of the
impact on vegetation, changes in soil moisture  is of greatest interest. This depends in a large part
on potential evapotranspiration (PET), estimates of which are listed in Table 4.  The projected
changes in PET are dependent on the method  of calculation and must be viewed as preliminary
estimates at this time (Marks 1990). Overall, the  UKMO scenarios present^ the largest change
from current conditions, particularly the large increases in summer temperature.

       The  climates projected  from the  GCMs represent  a significant shift from  present
conditions.  When viewed in a south to north transect, the climatic change is equivalent to
shifting current climates from 200 to 500 km  north from the present locations, i.e., moving the
climate of northern California into northern Oregon  and the climate of northern Oregon into
northern Washington (Franklin et al., in press). Similarly, when these changes are viewed from
an elevational perspective, it is equivalent to moving  current climatic conditions at the base of
mountain 500 to 1000 m upward (Franklin  et  al., in press).

Vegetation Modeling -  Application of the Holdridge Life-Zone Classification

       The Holdridge Life-Zone classification system has been used to simulate the effect of
climate change on global vegetation patterns (Emanuel et al.  1985a,b; Prentice and Fung 1990,
Smith et al. submitted). We will summarize  the results of Smith and Leemans (1990) for the
Pacific Northwest. Although the model was applied at a relatively coarse resolution (0.5° x 0.5°)
for application  to a regional case study, the Holdridge results  are the only published data
available that provide estimates of regional changes in the distribution of forests in the Northwest.
       The Holdridge Life-Zone classification system relates major vegetation formations with
mean annual biotemperature, precipitation, and the ratio of potential evapotranspiration (PET) to
mean annual temperature. Biotemperature is basically an index of the growing season. PET as
defined by Holdridge (1967)  is  a linear function of biotemperature,  and thus is not an
independent variable in this model. Using these climate variables, Holdridge created a triangular
axis system relating climate and vegetation (Figure 4).   Smith and Leemans (1990) applied the

-------
May 28,  1991   Draft  - Do not  quote  or  cite

Holdridge classification system to a gridded, global database of mean monthly temperature and
mean annual precipitation (Leemans 1990) for current climate, and simulated future climate using
GCM output described above. The climate data base has a resolution of 0.5°, which consequently
is the resolution of the vegetation scenarios.

Vegetation Modeling - Local Climate-Forest Zone Correlations

       Franklin et al. (in press) used correlations of climate and forest zonation in the central
Oregon Cascades (Franklin 1988) to estimate changes in the areal extent of forest communities
under two climate change scenarios.   In particular, the current temperature for  a particular
elevation was assumed to increase by either 2.5° and 5° C.  The current relationship  between
temperature and forest types were used to define the new elevational bands the forest  types would
occupy under  these two scenarios.   Then the areal extent of  the forest  zones at their new
elevations was determined using  a elevation model relating area with  a given elevational band.
These results are more detailed than the Holdridge simulations, but are limited to the central
Oregon Cascades.

Vegetation Modeling - Forest Gap Models

       Models  that simulate forest dynamics on small plots (usually 1000 m2)  have been
developed for a variety of forest communities worldwide (Shugart and West 1980).  The models
were initially devised in order to simulate forest succession in a gap formed in a closed forest
produced by the death of a overstory tree.  In recent  years,  these models have been used to
estimate changes in species composition in response to  climate change (Solomon 1986, Botkin
et al. 1989, Dale and Franklin 1989, Urban and Shugart 1989, Bonan et al.  1990).  The models
simulate tree growth for every individual on a plot, as  well as seedling establishment  and tree
mortality. Mathematical functions of key demographic  processes (e.g. annual diameter growth)
are  derived for the  maximum potential  behavior of each  tree species included in  the model.
Growth rates or other processes are then reduced  according to environmental constraints such as
shading,  soil moisture,  and temperature  (Urban and Shugart 1989).  Thus, environmental
feedbacks are an important component of these models. As a forest gap closes, available light
decreases  at the  forest  floor, shifting the  probability  of sapling establishment  from shade
intolerant species to shade tolerant species.  Since the models incorporate stochastic processes,
a large number of plot simulations are run and the average results presented as the model output.

       Forest gap models were initially developed for  eastern deciduous forests (Botkin et al.
1972, Shugart and West  1977), but several gap models have been developed for western forests
(Dale and Hemstrom 1984, Kercher and Axelrod 1984, Urban et al. 1990).  Two models have
been used to simulate climate change effects in  the Pacific Northwest, CLIMACS (Dale  and
Hemstrom 1984, Dale and Franklin 1989) and ZELIG (Urban  1990, Urban et al. 1990).

-------
May  28,  1991   Draft  - Do not quote or  cite
       CLIMACS was modified from FORET for application to the Northwest, specifically for
western Washington and Oregon.  The major modifications include introducing a moisture factor
in the growth equations, calculating regressions of height to diameter for use in the annual growth
equations, making mortality dependent on successional class of the species, and incorporating
natural disturbances and clear-cut logging (Dale and Hemstrom 1984, Dale et al. 1986).  Twenty-
one tree species are included in the model. The model was run assuming a 5°  C warming in
northwest Oregon,  using data from Cascade Head on the Oregon coast. The climate change
scenario was applied to a typical stand structure of a forest 140 years old at Cascade Head.

       ZELIG is an updated version of the FORET model as well (Urban 1990) and was used
to simulate climate change impacts at three sites in Washington and Oregon  (Urban et  al. 1990,
unpublished report).   The major effort in implementing the  model  in the Northwest  was
developing a set of species parameters (e.g. growing degree day limits) describing life history and
growth characteristics for each Northwest tree species.  In contrast to eastern United States where
relief is substantially less, tree species in the mountains of the Northwest often have upper and
lower elevational range limits. Correlations of climatic variables and these range limits were thus
established at  a number of latitudinal bands  from northern California into British Columbia.
Because of sparse  climate data, regional  regression models  were developed to predict mean
monthly temperature  (Urban et al. 1990). Substantial latitudinal variation occurs in the climatic
limits of some of the major species in the  region (to be discussed  more later).

       ZELIG was tested initially at the H.J. Andrews LTER site (44.2° N, 122.2° W) in  the
central_OregQiL_Cascades because of large amount of forest stand data available for model
verification. The model was applied at three elevations, 500m, 1000m, and 1500m.  The climate
parameters defining the range limits of species at the Andrews latitude (44° N) were used in  the
model run.  Model simulations started  from bare  ground (in  contrast  to the CLIMACS
simulations), ran for 500 years, and were replicated for 20 plots. Two climate change scenarios
were used, those generated from the OSU GCM (Schlesinger and Zhao 1988) and GISS (Hansen
et al. 1983). The  sensitivity of forest dynamics  at Mount Rainier, WA (46.8° N, 121.7°  W,
1654m) and Gold Beach, OR (42.4° N, 124.4° W) to climate change was also tested, using  the
species parameters set for 44° N. These results must be viewed as more tenuous than the results
for the Andrews, because of differences in species characteristics at these two different latitudes.

       ZELIG was also used to begin studying the transient response of Northwest forests to
climate change.  Transient climate changes were applied to the  model as  linear increases in
growing degree days (GDD) starting at a specific year and ending  at specific years.  The
increase was started at year 0, 200, and 500 in separate simulations, and either a 600 GDD or
1300 GDD warming was applied.  A 600 GDD  corresponds to the OSU 2xC02 equilibrium
scenario, and the 1300 GDD increase corresponds to the GISS 2xC02 equilibrium scenario.

-------
May 28,  1991   Draft -  Do  not  quote  or  cite


Carbon Model Results

      Potential changes in carbon storage in the central Oregon Cascades was analyzed using
the DFC model (Harmon et al.  1990). This model tracks carbon stored onsite (e.g. changes in
ecosystem storage)  and offsite (e.g. the  fate of harvested wood).  A simple 5° C warming
scenario was applied at three elevations in the Oregon Cascades (Urban et al. 1990).

                                      RESULTS

Holdridge Vegetation Scenarios

                           Simulation  of Current Vegetation

      Before presenting  the climate change results it is important to consider how well the
vegetation model simulates today's vegetation patterns. The Holdridge classification presents a
much coarser taxonomic  resolution of vegetation in the Northwest (Table 5, Rgures 5-8).
Generally, the Holdridge model does a good job of separating forest from non-forested land. The
main forest types in the Holdridge model under current conditions are temperate forest and boreal
forest, whereas the  SAP  classification has eight major forest types delineated based on the
dominant tree (e.g. Douglas-fir forests). All the forests west of the Cascade crest are considered
temperate forests, while subalpine  forests are predicted for parts of the Cascade crest, and
portions of the Northern Rocky Mountains in Idaho.  The Holdridge model does not depict the
heterogeneity of forest vegetation in central Idaho. The Holdridge estimate of forested land in
the region reasonably approximates the Forest Service estimates of forested land (Holdridge
318,000 km2 vs.  290,000 km2 USDA 1989).

                             Estimated Vegetation Change

      Major shifts  in vegetation patterns occur in the Northwest under four climate change
scenarios, according to the Holdridge system (Smith et al. submitted).  Overall, 26% (OSU) to
90% (UKMO) of the tri-state region will change from one Holdridge vegetation class to another
(Table 5, Figures 5-8). Most vegetation change in predicted under the UKMO climate scenario,
with least change predicted under the OSU  scenario.  Total forested area decreases under all
scenarios from 5 to 25%.  Boreal (subalpine) forests decrease by at least 50%, while temperate
forests generally  contract (except for the OSU scenario in which these forests increase by 20%).
Temperate forests remain  or expand at  higher elevations  in  the region.  In contrast, warm
temperate forest expand many fold.

      Interpreting what these  changes in broad vegetation classes means  in  terms of more
specific forest types  identified in the introduction is somewhat problematic. If one compares the
current distribution of forest types and the Holdridge classes, temperate forests in the Northwest

-------
May  28,  1991   Draft  -  Do  not  quote  or  cite

correspond to the entire suite of conifer forests in the region. Warm temperate forests are only
found in the  southeastern United States, corresponding to oak dominated forests in the middle
of the eastern United States (Holdridge 1967, Kuchler 1964).  Thus it appears that the Douglas-fir
- hemlock forests decrease in regional extent, being replaced by more drought tolerant oaks.

                                Assumptions/Limitations

       The limitations of this analysis are numerous and should be kept in mind when analyzing
the output.  First, the climate  change analysis assumes that the Holdridge system adequately
defines current vegetation patterns in the region, and that current climate-vegetation correlations
are unchanged in the future.  Although the Holdridge system only correctly classifies 40% of the
globe's vegetation, the model does a reasonable job of separating forest and non-forest vegetation
in the Northwest.  It  is uncertain whether current climate-vegetation correlations will remain
unchanged in the future. New broad-scale vegetation models are being developed (Neilson et al.
in press, Prentice) which are based  on  a mechanistic understanding  of how climate  controls
vegetation  patterns.   These models should  be  more robust under  different  climates.   All
vegetation models used in climate change analysis should be extensively tested, either in different
parts of the world, or at different times in the past (e.g. last 15,000 years).

       The model also does not simulate the dynamics of vegetation change. The climate change
results must be viewed  as a snapshot of future vegetation patterns after a double CO2 climate
change has occurred and the vegetation has come back into equilibrium with the regional climate.
In other words, forest dynamics (tree establishment and migration) are not simulated by the
model.

       The model only simulates potential natural vegetation; land use is not considered in the
model.  The influence of soils  on vegetation  is not factored into the analysis, nor are the direct
effects of higher CO2 concentrations on plant growth.  Finally, the life zone classifications are
very broad and applied on a coarse  resolution. Interpreting the vegetation change scenarios in
terms of actual species distribution is problematic.

Local Climate-Forest  Zone Correlations

       Mean annual temperatures of major plant zones in the region differ by about 1.5° to 2.0°
C at Mount Rainier, WA (Franklin 1988) and 2.5° C in southwestern Oregon (Atzet and Wheeler
1984, Franklin et al. in press).  Thus, a 2° degree wanning would completely shift a forest type
one  zone upward; a 4°  degree  warming would shift them two zones upward  (Franklin et al.
1990).

       The potential impact  of climate change on the elevation and area! extent of forest zones
in the central  Oregon Cascades was estimated using current correlations between forest zones and
mean annual temperature. Two simple climate warming scenarios (+2.5°C and +5°C)  were used
                                           10

-------
May 28,  1991  Draft -  Do  not quote  or  cite

to drive the correlational model (Franklin et al. in press). Douglas-fir dominated forests would
increase from 8% of current forested land to 39 or 27% (Table 6, Figure 9). The moister western
hemlock forests would decrease in areal extent.  Subalpine and alpine forests would decrease in
areal extent.  Decreases in area of upper elevational forests is simply a function of less area
occupying a given elevational band as elevation up a mountain range increases. Total forested
area would decrease east and west of the Cascade crest (Table 6).

       This correlational approach has the same limitations as those discussed for the Holdridge
approach, except the taxonomic resolution is much finer. Also, the approach assumes that current
forest zones will not change in composition, all the species in them will move an equivalent
distance upslope. However, paleoecological data clearly indicate that this assumption is incorrect.
Species moved independently in the past in response to warming after deglaciation (Davis 1981,
Webb ).  Each species has its own unique  set of climate limits.  Thus vegetation change in the
future will certainly be more complex than suggested here.

Forest Gap Models -  CLIMACS Results

       Simulation of Current Forests

       Before future simulations of forest composition as simulated by CLIMACS are discussed
it is important to evaluate how well the model simulates current forest dynamics.  CLIMACS has
been verified against data from HJ. Andrews (Hemstrom and Adams 1982) and validated for the
western Olympic peninsula. For xeric and mesic sites at Andrews the  model correctly predicts
Douglas-fir as the dominant tree and modeled  and  measured leaf area and basal area compare
favorably. For the Olympic Peninsula site, the model correctly projects Douglas-fir as dominant
with western hemlock and silver fir in the understory (after 500 years). The  size distribution is
similar for both modeled and measured stands (Dale et al. 1986).

                        Estimated Changes in Forest Compositions

       CLIMACS was run under a 5° C warming scenario for Cascade Head on the Oregon coast
(Dale and Franklin 1989).  Soil moisture status was not changed in the scenario. The warming
scenario was imposed on a stand  140 years old.   Under current  climate  conditions, the
composition of the stand would change from one dominated by Douglas-fir to one dominated by
western hemlock (Figure 10). Silver fir would be a co-dominant.  Under the  warming scenario,
western hemlock also succeeds Douglas-fir as a the dominant tree. However, grand fir replaces
silver fir as a codominant (Figure 10).

                               Assumptions/Limitations

       The limitations of gap  models  will be discussed fully, after the ZELIG results are
summarized.  One limitation unique  to the CLIMACS analysis jije that soil moisture was not
                                          11

-------
May 28,  1991   Draft  - Do not  quote  or cite

altered so the drought effects of warming were not analyzed. These could be quite pronounced
in the Northwest under the GFDL and UKMO scenarios.

Forest Gap Models - ZELIG Results

                            Simulations of Current Vegetation

       Much of the work in adapting ZELIG to run in the Northwest was spent developing a list
of species parameters (e.g. growing degree limits) for inclusion in the model (Urban et al. 1990).
Climatic correlates with tree species range limits were established for a number of latitudinal
bands in the Northwest, ranging from a species southern geographic limit to its northern limit.
A complicating result of this analysis in terms of modeling forests throughout the  Northwest is
the fact that growing degree day (GDD) limits for a species vary with latitude. For some species,
the maximum at higher latitudes  is greater than the minimum at lower latitudes.  There are
several explanations for this, ranging from genetic variation within a species, to the possibility
that some other factor (soil moisture, snow pack, wind) besides GDD limits species distributions
in some parts of their range.

       This complicates the forest gap modeling, because a decision needs to be made as to what
speciesparameters should be input into the model when the model is run at a particular site.
EitheTIhe local parameter  suite, &f the regional limits are the two  choices.  In the ZELIG
analysis, parameters for the latitude of the HJ. Andrews site were used in the model runs.  For
climate change analysis, this assumes that seed sources will be restricted to the area surrounding
the site (within about 100 km).

       Despite the problems  the  genetic  variability of  a species presents to modeling  forest
response to climate change, the variability will provide a species the ability to better respond to
climate change.  The probability  of a genotype being present in a species adaptive to a future
climate is increased with greater genetic variability within a species.

       The results  from the ZELIG simulations for current and future conditions at Gold Beach,
OR, HJ. Andrews, OR, and Mount Rainier, WA are summarized in Figure 11. The model does
a poor job of simulating current vegetation patterns at HJ. Andrews.  At 500 m, Douglas-fir and
western hemlock are the dominant trees in the current forests. In the model simulation, chinkapin
and ponderosa pine are dominant.  Douglas-fir and western hemlock are only minor components
of the modeled stand.  In essence  the model predicts vegetation  adapted to much drier sites than
occur at  the Andrews. The model performs better at 1000m,  where  Douglas-fir and western
hemlock  dominate  natural stands and modeled stands. Douglas-fir is still underestimated by the
model, however.

       A priori one would expect the model not to perform as well at the two other  sites because
of the apparent latitudinal differences in species adaptations to climate. At Gold Beach (where
                                          12

-------
May  28,  1991   Draft  - Do not  quote or cite

annual precipitation is greater than at the Andrews, and temperatures are 1.3° C warmer), current
forests are dominated by western hemlock and Sitka spruce (Franklin and Dyrness 1988). At the
end of the 500 year run, the model predicts chinkapin and tan oak as dominants.  Hemlock is a
very minor component (sitka spruce was inadvertently left out of the species set, will be rerun
but won't change  outcome).   Again the  model is biased towards predicting a drier type  of
vegetation (even more so than  at Andrews) than what actually occurs at the site.

       At 1600m elevation at Mount Rainier, mountain hemlock and silver fir are dominant with
subalpine fir present (Franklin and Dyrness 1988, Franklin et al. 1988).  Silver fir, mountain
hemlock, western white pine, subalpine fir are simulated as dominant species by the model.  In
terms of the dominant trees, the  model does  a good job (better than at  the other two sites)  of
simulating the forest stand composition. In sum, ZELIG appears to be biased towards simulating
drier types of vegetation at warm sites, but does a better job of forest simulations at higher sites.
It does not simulate the dynamics of Douglas-fir well (Hansen et al. unpublished).

                        Estimated Changes in Forest Composition

       ZELIG predicts major changes in species composition under the OSU and GISS climate
scenarios (Figure 11). Under the OSU scenario, the dominant trees simulated under current
conditions are replaced by tanoak and red alder at Andrews.  At Gold Beach, dominants under
current conditions are replaced by red alder. At Rainier, vegetation change is less, with the major
change being a replacement of mountain  hemlock with  western hemlock.  The driving factor
behind these changes is the increase in GDD at each site. Even under the moderate wanning of
the OSU scenario, only a few species have maximum GDD limits less than that of the predicted
climate.  At Gold Beach, only  red alder and tanoak will  grow.

       The  increase in temperature under the GISS  scenario is  so  great that GDD  limits  at
Andrews and Gold Beach exceed the maximum values for any species included in the Northwest
modeling effort.  Thus, the model simulates no tree growth at these sites.   At Rainier, silver fir
becomes a minor component with western hemlock and Douglas-fir becoming more important.
In essence the model  predicts an  upward  movement of species distribution in  the Washington
Cascades.

       The general response of mid-elevation forests at Andrews to the transient climate scenario
varies with the age at which the warming applied (Urban et al. 1990). If the warming is applied
at a stand age of 200  or 500 years, a short episode of high tree mortality occurs and the  total
biomass on  the plot decreases. The total stand biomass doesn't return to prewarming levels
because of the growing degree constraints on  tree growth as discussed above.  If the warming
began from the beginning of the simulations,  succession  to more heat tolerant species proceeds
without significant forest dieback.
                                          13

-------
May 28,  1991   Draft  - Do not  quote  or cite
       These results are distinctly different from the CLIMACS results, in which total biomass
did not drop after a climate change was imposed on a young stand.

                                Assumptions/Limitations

       A number of assumptions are made when the two forest gap models are applied in the
Northwest. In developing correlations between climate and species distribution, it is assumed that
their current distribution reflects their environmental tolerance, and that the species grow best at
the center of its range (Urban et al. 1990).  Temperature (through growing degree day totals) is
assumed to be the primary control of species range limits.  Genetic variability with a species
parameter set is assumed to be non-existent. Seeds of all the species included in the simulation
are assumed to be^ instantaneously available  if the climate  is favorable for their germination.
Other assumptions are discussed by Urban et al. (1991).   The form of the species response
functions to climate is assumed to be correct  and similar for all species.

       The limitations of Northwest forest stand models are  substantial at this point in their
development.  The ZELIG model has problems simulating low elevation (warm) forests  in the
Northwest, and consequently the 2xC02 scenarios can only be interpreted in terms of relative
change in dominance.  Specific changes in  species dominance are  likely  inaccurate, except
perhaps for the Rainier site.  Thus problems probably exist both in the species parameterization
and form of the model algorithms.

       More importantly, the models do not adequately account for the  importance of soil
moisture in controlling species distributions at lower elevations (Daubenmire 1968,  Zobel et al.
1976, Waring and Franklin 1979, Neilson et al. 1989). Moreover, the simulation of soil moisture
is biased  towards simulating wetter conditions than what  actually  occurs  at a site,  because
vegetation cover does not affect available soil moisture in the  model. Also the contribution of
snowmelt to soil  moisture  during  the spring  and early summer in the  growing season is
underestimated.  Finally, other factors besides temperature and soil moisture may limit species
distributions in particular regions, such  as chilling requirements along the coast (Leverenz and
Lev 1987) and snowpack at higher elevations (Scott 1980).

       The model simulations do not consider the direct effects of enhanced CO2 concentrations
on water use efficiency or changes in disturbance frequencies (Urban et al. 1991). Model output
also depends in part on the species assemblages input as potential occupants of a stand.  Many
of the species specific parameters are expressed relative  to the other species in the assemblage.
These could change as assemblages change, they remain constant in these model runs.

Potential Changes in Carbon Storage

       A climate warming of 5° C could decrease detrital carbon stores by 30% across a diverse
                                          14

-------
May  28,  1991   Draft  - Do not  quote  or cite

array of sites (Urban et al. 1990).  At low, middle, and high elevations in the Oregon Cascades
this could result in fluxes of 40, 55, and 75 Mg C per hectare to the atmosphere. Production
would have to increase by about 10% to offset  these losses.  The increase in production is
unlikely except at middle elevations. Low elevations could be constrained by low soil moisture
and replacement of Douglas-fir by slower growing species.  Shallower soils at higher elevation
may limit production increases there. Thus, assuming no change in production, and weighting
the detrital carbon loss by the area! extent of each elevation zone as done by Franklin et al. (in
press) suggests that there could be a net loss of 60 Mg carbon per hectare  from the Oregon
Cascades.

       It  should be emphasized that  this  analysis does not  factor in changes in species
composition, nor the potential for catastrophic forest dieback (Neilson personnal communication)
to inject a pulse of carbon into the atmosphere.  Predicted changes to less-productive forests
adapted to warmer and drier conditions could substantially reduce the amount of carbon  stored
in the Northwest.

       To put the potential carbon loss in perspective to recent land use changes, conversion of
old growth forests to younger plantations in western Oregon and Washington has resulted in
declines in carbon storage of 305 - 370 Mg per hectare (Harmon et al. 1990).


Expert Judgement - Effects of Climate Change on Forest Disturbances

       Although as yet not quantitatively modeled in the Northwest, the effects of climate change
on forest composition and structure could be felt initially and most extensively through altered
disturbance regimes (Overpeck et al. 1990, Franklin  et al. in press).   Disturbances destroy the
resilience of the existing forests and coupled with climate change provide conditions  for forest
reestablishment  that  may be more  severe than previously  existed.   As noted  by  Brubaker
(1986)"[disturbances]  should also mitigate  the  lagging  effects of  long  tree  lifespans by
accelerating rates of population decline  when climate change makes conditions unfavorable for
seedling establishment."

       Fire frequencies  are likely to  increase  in the region  given  increased  temperatures,
unchanged precipitation and higher potential evapotranspiration.  In southern Oregon  increased
fires could increase the dominance of some hardwood species that sprout after fires, (e.g. tanoak;
Franklin  et al. in press). The primary effect of  fire would probably be  one of speeding the
conversion of a forest to one favored under warmer conditions.

       New or more severe insect problems are probable given projected climate change. The
altered climate may provide a more favorable environment for the insect or the trees may become
more susceptible to the insect pest as a  consequence of climate-change ino-easesln tree  stress.
For example, the balsam woolly aphid (Adelges piceae)  is an introduced p^st.can be a serious


                                          15

-------
May  28,  1991   Draft -  Do  not quote  or  cite

problem in the Pacific Northwest in Pacific silver fir and low-elevation occurrences of subalpine
fir (Mitchell, 1966; Franklin and Mitchell, 1967, Franklin, 1991). Currently, the aphid has been
restricted to low and middle elevations by temperature limitations. During the summer the second
generation of the  aphicUmust reach the first instar  stage to survive the winter.   The higher
subalpine  zones of the (CQ^al and Cascade  mountains rarely experience sufficient heat for the
second generation to develop sufficiently to overwinter* fyence, to^few aphids attain the critical
stage to produce dense populations (Mitchell, 1966; Franklin et al., 1991).

       However, given the large increase in growing degree/days (Table 3) and mean temperature
increases of 2°  to 5° C, it would be possible for the aphid to successfully reproduce and spread
at the higher elevations where subalpine fir is a major component (Franklin et al., 1991). Mature
subalpine  fir are susceptible to the aphid; consequently high levels of subalpine fir mortality are
likely (Franklin et al., 1991).

       Increased frequency  and severity of pest outbreaks are likely in forest stands subjected
to increased physiological stresses associated with climate change (Mattson and Haack, 1987).
Even under conditions of  stable  climate, the majority  of pest outbreaks  are associated  with
increased  host-plant stress (Franklin et al., 1991).

       Disturbances can be viewed  as events that hasten the adjustment of forest vegetation to
the current climate conditions. But this could be disruptive as climate change may create more
severe conditions  for forest reestablishment which can cause significant changes in the forest
composition  and function  (Franklin et al.,  1991).  Transitions in vegetation types will  be  a
problem as forest loss may occur faster than forest reestablishment, especially at the lower and
upper tree lines. There will  likely be a shift in area from forest to non-forest vegetation (Franklin
et al., 1991).  "In general,  natural forest ecosystems,  with  their greater  compositional and
functional redundancy, are  expected  to show greater resistance to change and recover more
rapidly following disturbance than are intensively managed forests." (Franklin et al., 1991).

Major Uncertainties
                                   Transient Dynamics

       The model results presented  here assume an instantaneous climate change, and simulate
the equilibrium response of vegetation to that climate change.  That is,  the models simulate
vegetation patterns in the region after species have responded to the climate change.  One reason
for this assumption is the lack of transient climate change scenarios (two GISS transient scenarios
are  available however, Hansen et al.  (1983)).   Simulating  transient climate change  using
atmospheric  models  is  a   difficult task,  necessitating better ocean-atmosphere  feedbacks.
Incorporating a more realistic ocean circulation model in a climate model greatly increases the
computer  power required to operate the model, and taxes the fastest computers now available.
Other technical and conceptual issues  need to be resolved as well before transient scenarios can
be produced.


                                           16

-------
 May 28,  1991   Draft  -  Do  not quote or cite

        Ecosystem models also are limited in their ability to simulate transient dynamics. Broad-
 scale vegetation models that simulate the equilibrium response of vegetation to climate change
 are  still being refined; development of regional scale models to  depict transient behavior of
 vegetation is still a ways off. Local scale models, such as forest gap models,  are available for
 driving by a transient climate scenario.  Several scenarios of the transient response of forests to
 climate change have produced in other parts of the U.S. (Urban et al. 1989, Botkin et al. 1989?).
 However, forest gap models in the Northwest still need refinement in their simulation of current
 forests before their simulations of future climate impacts can be considered to be more accurate.

        Interestingly,  the  simple  transient scenario used  in the ZELIG approach  suggests that
 biological inertia (e.g. the long lifespan of the forest trees in the Northwest) may not significantly
 delay changes in forest composition.  The magnitude of the climate warming is so great that
 forest dieback occurs even in mature trees.  This result is driven by the temperature limitations
 imposed  on the species in the modeling runs which  may not be relevant in lower elevational
 forests.  Still, changes in soil moisture could be substantial enough that mature trees rapidly
 dieback to be replaced by more drought adapted species.

        One difference  between  the eastern and  western U.S. in terms of transient responses of
 vegetation is  the differences  in dispersal distances required for a species to track its favored
 climate.  The generally flat terrain  in the eastern U.S. means a 1° C temperature change will
 move a climate zone a considerable distance latitudinally. In the west, the mountainous terrain
 means that a 1°  C change will move  up only a few hundred meters of elevation and perhaps only
 a few km in actual distance.  Thus, western tree species will have shorter distances to migrate
 to track their favored environment (as long as the temperature change does not raise the preferred
 elevational  band above the highest elevation in the local region).  Thus mountain vegetation will
 more likely stay in equilibrium  with the changing climate than eastern forests.

        Genetic  variation in  certain western tree species is relatively high (e.g.  Douglas-fir).
 Thus, within a  small region, the probability of a genotype being  present that is adapted to a
 potential  future climate is relatively high.   However, current forest management techniques
 limiting the range of genotypes being replanted on cutover sites may decrease the ability of local
^forests to successfully adapt to future climate change.
>
                                   Direct Effects of CO2

        The other major uncertainty in predicting forest response to climate change is the potential
 effect of  higher CO2 concentrations on forested ecosystems. Many assessments of the impacts
 of climate change on forests have focused on the potentially negative direct effects of increased
 temperature and drought on the  trees, but have not considered  the potential impacts of elevated
 CO2 on forests.

        There is a substantial body  of  literature that  photosynthesis is increased  and stomatal
                                            17

-------
May  28,  1991   Draft -  Do  not quote  or  cite

conductance  decreased by  supra ambient levels of CO2 (Agren, et al., 1991; Hollinger, 1987;
Surano et al., 1986).  However, the magnitude of the changes, especially for mature trees is
uncertain.  Also it is largejyonknown if the changes  in photosynthesis and conductance will
persist as the tree acclimati^^o the elevated CO2 or if the increases can be supported by current
nutrient levels in the forest soils.

       The extent to which possible nutrient limitations in forest soils will limit the direct effects
of CO2 on productivity is also unknown. In a review of a series of experiments of the response
of seedlings and young trees to CO2, Jarvis (1989) concluded that tree growth was stimulated by
added C02 even under severe nutrient limitations. However, it is not known if these stimulations
will persist in mature trees in nutrient-poor soils.

       Studies with loblolly pine (Pinus taeda L.) seedlings found that drought stress did not
preclude a growth response to elevated CO2^and that the effects of drought stress were largely
mitigated by elevated CO2 (Tschaplinski and Norby, 1991).  The reported decreases in stomatal
conductance and increases  in photosynthesis  yieldtf an increase in water-use efficiency by trees
(Conroy et al., 1988).

       A key issue in the CO2 debate is how(wifl>egetation  over a landscaperespond to higher
     concentrations.  The scientific problem  is one of developing the techniques to extrapolate
    experimental work to a forest stand. One approach is to use simulation  models.  Solomon
/and West Q investigated the impact of greater water-use efficiency of forest population dynamics
with a simulation model. They concluded that forest response to elevated CO2 was dampened by
the processes of tree regeneration  and death. The increase in water-use efficiency was  not
sufficient to compensate for the probable increase in drought severity expected from a warmer
and possible drier climate.

       In another simulation study of forest stand pfocpssies^ Shugart (1984)  found that the
primary effect of elevated CO2 was to increase the rate^atgapsin the forest closed. Consequently,
forests recovered  faster from disturbance^uid forests with  more disturbances  were more
responsive to additional CO2. "Although...slm/ulation studies of possible responses of forests to
enhanced levels of CO2 are based on simple  assumptions and formulations, the indications that
population dynamics  may  override  the consequences of CO2 fertilization  are interesting. The
direct effects of CO2 must therefore be considered within the larger scale consequences of
population dynamics and responses of different species to CO2-induced climate change" (Agren
et al., 1991).
                                           18

-------
May 28,  1991   Draft  - Do  not quote  or  cite



                                   CONCLUSIONS

Climate Induced Changes in  Forest  Composition and Distribution

       Despite  the  uncertainties and concerns on  model precision addressed above in the
discussion of climate change impacts on Northwest forests, some overall conclusions can be made
concerning climate  change impacts on Northwest forests. The foremost of these  is that the
distribution and composition  of forests in Washington and Oregon could change  substantially,
given the GCM  scenarios of warmer temperatures and generally unchanged precipitation and
ignoring the  potential  mitigating effects of increased CO2 concentrations.  The  Holdridge,
Elevation-zone, and forest gap models (except for the CLIMACS results) all forecast shifts to
forests better adapted  to warmer and drier conditions.  Temperate  forests in the  Holdridge
scenarios are generally restricted to upper elevations and total forest acreage decreases 5 to 25%.
In central Oregon,  total forested area is projected  to decrease  by almost half under a  5° C
warming.  Forest zones could move up one complete elevation band under the same degree of
warming.  Oak woodlands and Douglas-fir dominated forests are likely to increase in areal extent,
while the more productive hemlock forests will undergo significant contraction. Subalpine and
alpine vegetation is  likely to  disappear.

       The forest modeling work, although potentially more precise, is still in need of additional
refinement. ZELIG  results do show an upward movement of forest type^a)id a general decrease
in forest biomass. The model does not simulate low elevation or dry siWs'well.  Upper elevation
simulations perform better, perhaps because temperature actually does  becomes the limiting
constraint on forest  growth as assumed by the model.

       Expert judgement suggests that disturbance frequencies (fires and pest/pathogen outbreaks)
could increase under warmer conditions. Increased disturbance rates would speed the conversion
of forest types and decrease the biological inertia represented by long-lived trees.  Even without
disturbance, the ZELIG results suggests that mature trees would dieback under the 2xCO2 climate
conditions simulated by GCMs.

Implications for Forest Management

       Climatically  induced changes of the magnitude predicted here raise a number of forest
management issues.
       1).  Given current rotation lengths and the possibility of substantial climate change over
the period of that rotation, what species should be planted now?
       2).  How much must current uncertainties in climate predictions and vegetation response
be  reduced before  irreverisble  management  practices   predicated  on climate change are
implemented?


                                         19

-------
May 28,  1991   Draft  - Do not quote  or  cite

       3). Are their "no regrets" management practices for adapting to climate change that can
be implemented without a substantial cost if climate change does not occur as expected?
       4). How should current strategies at protecting endangered species (e.g. the spotted owl)
be altered considering the prospect of large changes in  future climate?
       5).  How can forest  management practices be  altered to promote the sequestering of
carbon and  slow atmospheric concentrations of CO2?
                                       20

-------
May 28,  1991   Draft  - Do  not  quote  or cite

                             ACKNOWLEDGEMENTS

Dean Urban from the University of Virginia and colleagues at the Pacific Northwest Lab of the
U.S. Forest Service supplied the ZELIG model and the initial 2xCO2 climate simulations through
Interagency Agreement DW12934129.  Charles Halpern of the Department of Forest Sciences
at Oregon  State University provided climate data for specific  localities in western Oregon and
Washington. Terry Droessler at the Envrironmental Research  Laboratory, Corvallis, performed
the  2xC02 simulations for Gold Beach and Rainier.  The authors thank Don Phillips and Ron
Neilson for reading an earlier version of the manuscript.
                                         21

-------
                                LITERATURE CITED

Agren, G.I., R.E. McMurtrie, W.J. Parton, J. Pastor and H.H. Shugart. 1991. State-of-the-art
       of models of production-decomposition linkages in conifer and grassland ecosystems.
       Ecological Applications 1:118-138.

Atzet, T., and D.L. Wheeler.  1984. Preliminary plant associations of the Siskiyou Mountain
       Province. USDA Forest Service, Pacific Northwest Region, Portland, OR.

Bamosky, C.W., P.M. Anderson, and P.J. Bartlein. 1987.  The northwestern U.S. during
       deglaciation; Vegetational history and paleoclimatic implications, in W.F. Ruddiman
       and H.E. Wright, Jr. editors, North America and Adjacent Oceans during the Last
       Deglaciation.  The Geology of North America, Volume K-3. Geological Society of
       America, Boulder, CO.

Bonan, G.B.  1989. A computer model of the solar radiation, soil moisture, and soil thermal
       regimes in boreal forests. Ecological Modelling 45:275-306.

Bonan, G.B., Shugart, H.H., and D.L. Urban. 1990.  The sensitivity of some high-latitude
       boreal forests to climatic parameters. Climatic Change 16:9-30.

Botkin, D.B., J.F. Janak, and J.R. Wallis.  1972.  Some ecological consequences of a
       computer model of forest growth.  Journal of Ecology. 60:849-873.

Botkin, D.B., R.A. Nisbet, and T.E. Reynales.  1989. Effects of climate change on forests of
       the Great Lake States.   In J.B.  Smith and D.A. Tirpak, editors.  The potential effects
       of global climate change on the U.S.:  Appendix D-Forests. Office of Policy,
       Planning and Evaluation, U.S. EPA, Washington, DC.

Brubaker, L.B.  1986. Responses of tree populations to climatic change. Vegetatio 67:119-130.

Brubaker, L.B.  1988.  Vegetation history and anticipating future  vegetation change, in Agee,
       J., and D. Johnson, editors, Ecosystem management for parks and wilderness.  Seattle,
       WA, University of Washington Press.

Burke, M. J., L. V. Gusta, H.  A. Quamme, C. J.  Weiser, and P. H. Li.  1976.  Freezing injury
       in plants. Annual Review of Plant Physiology 27:  507-528.

Cissel, J. 1990. An approach  for evaluating stand significance and designing forest
       landscapes. COPE Report (Oregon State  University). 3:8-11.

Conroy, J.P., M. Kiippers, B. Kiippers J. Virgona, and E.W.R. Barlow.  1988. The influence of
       CO2 enrichment, phosphorus deficiency and water stress on the growth, conductance
       and water use of Finns radiata D. Don. Plant, Cell and Environment 11:91-98.

Dale, V.H., and R.H. Gardner.  1987.   Assessing regional  impacts of growth declines  using a
       forest succession model.  Journal of Environmental Management. 24:83-93.

Dale, V.H., and J.F. Franklin.    1989.  Potential effects of climate change on stand
       development in the Pacific Northwest.  Can. J. For. Res.  19:1581-1590.
                                          22

-------
Dale, V.H., and M.A. Hemstrom.  1984.  CLIMACS: A computer model of forest stand
       development for western Oregon and Washington.  USDA Forest Service Research
       Paper PNW-327.

Dale, V.H., M. Hemstrom, and J. Franklin.  1986.  Modeling the long-term effects of
       disturbances on forest succession, Olympic Peninsula, Washington.  Canadian Journal
       of Forest Research.  16:56-67.

Daubenmire,  R.F. 1943. Vegetational zonation in the Rocky Mountain.  Botanical Review 9:
       326-393.

Daubenmire,  R.  1968. Soil moisture in relation to vegetation distribution in the mountains of
       Northern Idaho. Ecology 49:431-438.

Emanuel, W.R. 1985.  Response to comment: Climatic change and the broad-scale
       distribution of terrestrial ecosystem complexes.  Climatic Change  7:457-460.

Emanuel, W.R., H.H. Shugart,  and M.P. Stevenson.  1985. Climatic change and the broad-
       scale distribution of terrestrial ecosystem complexes.  Climatic Change 7:29-43.

Franklin, J.F. and R.G. Mitchell. 1967. Successional status of subalpine fir in the Cascade
       Range. USDA Forest Service Research Paper PNW-46. Pacific Northwest Forest and
       Range Experiment Station, Portland, OR.

Franklin, J.F., F.J. Swanson, M.E. Harmon, D.A. Perry, T.A. Spies,  V.H. Dale, A. McKee,
       W.K.  Ferrell,  S.V. Gregory,  J.D. Lattin, T.D. Schowalter, D. Larsen, I.E. Means. In
       press.  Effects of global climate change on forests  in northwestern North America,  in
       R.L. Peters and T.E. Lovejoy, editors, Global Warming and Biological Diversity.  Yale
       University Press.

Franklin, J.F. 1988. Pacific Northwest forests,  in M. Barbour and W.D.  Billings, Jr, editors,
       North American Terrestrial Vegetation. Cambridge University Press, New York. 103-
       130.

Franklin, J.F., and C.T. Dyrness.  1988. Natural Vegetation of Oregon and Washington.
       Oregon State  University Press, Corvallis,  OR.

Harmon, M.E., J.N. Long, and  J.F. Franklin.  1982.  Relations between vegetation and
       environment.  Pages 28-44 in R.L. Edmonds, editor.  Analysis of Coniferous Forest
       Ecosystems in the Western United States. US/IBP Synthesis Series  14. Hutchinson
       Ross,  Stroudsburg, PA.

Franklin, J.F, F.J. Swanson, M.E. Harmon, D.A.  Perry, T.A. Spies, V.H. Dale, A. McKee,
       W.K.  Ferrell,  S.V. Gregory,  J.D. Lattin, T.D. Schowalter, D. Larsen, J.E. Means. 1991.
       Effects of  global climatic change on forests in the northwestern North American, in
       The Consequences of the Greenhouse Effect for Biological Diversity. Yale University
       Press  [In Press]

Franklin, J.F. and R.T.T. Forman.  1987. Creating landscape patterns by cutting: ecological
       consequences  and principles. Landscape Ecology 1:5-18.
                                          23

-------
Graumlich, L.J.  1987.  Precipitation variation in the Pacific Northwest (1675-1975) as
       reconstructed from tree rings.  Annals of the Association of American Geographers
       77:19-29.

Hansen, J., G. Russell, D. Rind, P. Stone, A. Lacis, S. Lebedeff, R. Ruedy, and L. Travis.
       1983.  Efficient three-dimensional global models for climate studies: Models I and II.
       Monthly Weather Review 3:609-662.

Hansen, J., I.  Fung, A. Lacis, D. Rind, S. Lebedeff, R. Ruedy and G. Russell. 1988. Global
       climate changes as forecast by Goddard Institute for Space Studies three dimensional
       model. Journal of Geophysical Research. 93:9341-9364.

Harmon, ME., W.K. Ferrell, J.F. Franklin.  1990.  Effects on carbon storage of conversion of
       old-growth forests to young forests.  Science 247:699-702.

Hemstrom, M. 1990.  Alternative timber harvest patterns for landscape diversity. COPE
       Report (Oregon State University). 3:8-11.

Holdridge, L.R.  1947.  Determination of world formulations  from simple climatic data.
       Science 105:367-368.

Holdridge, L.R.  1967.  Life Zone Ecology.  Tropical Science Center,  San Jose, Costa Rica.

Hollinger, D.Y. 1987. Gas exchange and dry matter allocation responses to elevation of
       atmospheric CO2 concentrations in seedlings of three tree species. Tree Physiology
       3:193-202.

Intergovernmental Panel on  Climatic Change (IPCC). 1990. Scientific Assessment of Climate
       Change. Report for WG1 Plenary Meeting. Cambridge Unviersity Press, Cambridge,
       UK

Jarvis, P.G. 1989. Atmospheric carbon dioxide and forests. Philosophical Transactions of the
       Royal  Society of London 324:369-392.

Jenne, Rl. 1988.  Data From  Climate Models, the CO2 Warming. National Center for
       Atmospheric Research, Boulder, CO.

Kalkstein, L.S., editor.  1991.  Global Comparisons of Selected GCM Control Runs and
       Observed Climate Data.  Report prepared for US EPA, Office of Policy, Planning, and
       Evaluation, Washington, DC.

Kercher, J.R., and M.CJC. Axelrod. 1984.  A process model of fire ecology and succession in
       a mixed-conifer forest.  Ecology 65:1725-1742.

Leemans, R.  1990.  Possible changes in natural vegetation patterns due  to a global wanning.
       Publication No. 108 of the Biospheric Dynamics Project, International Institute for
       Applied Systems Analysis A-2361, Laxenburg, Austria.

Leverenz, J.W., and D.J. Lev.  1987.  Effects of carbon dioxide-induced climate changes on
       the natural ranges of six major commercial tree species in the Western United States.
       Pages  123-155 in W.E. Shands and J.S. Hoffman, editors. The Greenhouse Effect,


                                          24

-------
       Climate Change, and U.S. Forests. The Conservation Foundation.  Washington, DC,
       USA.

Manabe, S., and R.T. Wetherald. 1987.  Large-scale changes in soil wetness induced by an
       increase in carbon dioxide. Journal of the Atmospheric  Sciences.  44:1211-1235.

Marks, D.  1990. A continental-scale simulation of potential evapotranspiration for historical
       and projected doubled-CO2 climate conditions, in H. Gucinski, D. Marks, and D.P.
       Turner, editors.  Biospheric Feedbacks to Climate Change: The Sensitivity of Regional
       Trace Gas Emissions, Evapotranspiration, and Energy Balance to Vegetation
       Redistribution.  Status of Ongoing Research.  U.S. Environmental Protection Agency.
       EPA/600/3-90/078.

Mattson, W.J.  and R.A. Haack. 1987. The role of drought in  outbreaks of plant-eating insects.
       BioScience 37:110-118.

Mitchell, R.G. 1966. Infestation characteristics of the balsam wooly aphid in the Pacific
       Northwest USDA Forest Service  Research Paper PNW-35. Pacific Northwest
       Research Station, Portland, OR.

Neilson, R.P. and L.H.  Wullstein. 1983.  Biogeography of two southwest American oaks in
       relation to atmospheric dynamics.  Journal of Biogeography 10:275-297.

Neilson, R.P.,  G.A. King  and G. Koerper. Towards a rule-based biome model. Submitted to
       Landscape Ecology.

Neilson, R.P.,  and G.A. King. Submitted. Continental scale biome responses to climatic
       change. Proceedings of EMAP Conference,  Oct. 16-19, 1990.  Elsevier Science
       Publishers, Ltd., Dordrecht, The Netherlands.

Neilson, R.P.,  G.A. King, R.L.  DeVelice, J. Lenihan, D. Marks, J. Dolph, B. Campbell, and
       G. Glick. 1989.  Sensitivity of ecological landscapes  and regions to global climate
       change. U.S. Environmental Protection Agency, Environmental Research Laboratory,
       Corvallis, OR. EPA/600/3-89/073.

Overpeck, J.T., D. Rind, and R. Goldberg.  1990. Climate-induced changes in forest
       disturbance and  vegetation. Nature 343:51-53.

Parry, M.L., and T.R. Carter.  1989.  An  assessment of the effects of climatic change on
       agriculture. Climatic Change 15:95-116.

Parry, M.T., T.R. Carter, N. Konijin, and J. Lockwood.  1987. The impact of climatic
       variations on agriculture,  in Introduction to the HASA/UNEP Case Studies in
       Semi-Arid Regions. International Institute for Applied Systems Analysis, Laxenburg,
       Austria.

Prentice, K.C., and I.Y. Fung.  In press.  Bioclimatic simulations test the sensitivity of
       terrestrial carbon storage to perturbed climates. Nature.

Sakai, A. 1983. Comparative studies on  freezing resistance of conifers with special reference
       to cold adaptation .and its evolutive aspects.   Canadian Journal of Botany 61:2323-


                                          25

-------
       2332.

Schlesinger, M.E., and Z.C. Zhao.  1989. Seasonal climatic change introduced by doubled
       CO2 as simulated by the OSU atmospheric GCM/mixed-layer ocean model.  Journal of
       Climate 2:429-495.

Scott, W.F. 1980. Geology of the Columia Basin.  Journal of Forestry. 9:537-541.

Shugart, H.H.  1984. A Theory of Forest Dynamics. Springer-Verlag, New York, NY.

Shugart, H.H., and D.C. West.  1977.  Development of an Appalachian deciduous forest
       succession model and its application to assessment of the impact of the chestnut
       blight.  Journal of Environmental Management. 5:161-170

Shugart, H.H., and D.C. West  1980.  Forest succession models. BioScience  30:308-313.

Smith, T.M., R. Leemans, W. Cramer, I.C. Prentice, A.M. Solomon, and H.H. Shugart.
       Submitted.  Simulating changes in the distribution of global terrestrial vegetation under
       CO2-induced climate change: Comparison among scenarios based on general
       circulation models. Climate Change.

Solomon, A.M.  1986. Transient response of forests to CO2-induced climate change:
       simulation modeling experiments in eastern North America.  Oecologia 68:567-579.

Solomon, A.M. and D.C.  West.  1986. Atmospheric carbon dioxide change: Agent of future
       forest or decline? in J.G. Titus, editor, Effect of Changes in Stratospheric Ozone and
       Global  Climate. Vol 3:23-38. Environmental Protection Agency, Washington, DC.

Sucoff, E. 1969. Freezing of conifer xylem and the cohesion-tension theory.  Physiologia
       Plantarum 22:424-431.

Surano, K.A., P.P. Daley, J.L.J.  Houpis, J.H. Shinn, J.A. Helms, R.J. Palassou and M.P.
       Costella. 1986. Growth and physiological responses of Pinus ponderosa Dougl ex P.
       Laws,  to long-term elevated CO2 concentrations. Tree Physiology 2:243-259.

Swanson, F.J., R.P. Neilson, G.E.Grant. In press.  Watershed management--a broad
       perspective,  in RJ. Naiman and J.R. Sedell, editors, New perspective for watershed
       management: Balancing long-term sustainability with cumulative environmental
       change.  Springer-Verlag.

Thomas, J.W., E.D. Foreman, J.B.  Lint, E.G. Meslow, B.R. Noon, and  J. Verner. 1990. A
       conservation strategy for the northern spotted owl.  U.S.  Department of Agriculture,
       Forest Service, Portland,  OR.

Thomthwaite, C.W.  1948. An approach toward a rational classification of  climate. Geographic
       Review 38:55-89.

Thomthwaite, C.W.  and J.R. Mather. 1955.  The Water Balance. Publications in Climatology
       Vol 8, Laboratory  of Climatology.

Tschaplinski, T.J. and R.J. Norby.  1991. Osmotic adjustment in  loblolly pine  seedlings in


                                         26

-------
       response to water stress and elevated C02. (Draft manuscript).

U.S. Department of Agriculture. Forest Service.  1970.  In National Atlans of the United
       States.  United States Geological Survey, Washington D.C.  Plates 154-155.

U.S. Department of Agriculture. Forest Service.  1988.  Final supplement to the EIS for an
       amendment to the PNW regional guide. Vol. 4. Spotted owl guidelines.

U.S. Department of Agriculture. Forest Service.  1990.  An analysis of the timber situation in
       the United States: 1989-2040. Appendix C. General Technical Report RM-199.
       Department of Agriculture,  Forest  Service. Rocky Mountain Forest and Range
       Experiment Station, Ft. Collins, CO.

Urban, D.L. 1990. A versatile model to  simulate forest pattern: A user's guide to ZELIG
       Version 1.0. Environmental Sciences, University of Virginia, Charlottesville, VA.

Urban, D.L., and H.H. Shugart.  1989.  Forest response to climatic change: a simulation
       study for southeastern forests,  in J.B. Smith and D. Tirpak, editors, The Potential
       Effects of Global Climate Change  on the United States.  EPA-230-05-89-050, U.S.
       Environmental Protection Agency, Washington, DC.

Urban, D.L., M.E. Harmon, and C.B. Halpern.  1991. Modeling forest response to climatic
       change: A case study and stocktaking. (Draft manuscript).

Urban, D.L., M.E. Harmon, C.B. Halpern, and F.J. Swanson.  1990.  Potential response of
       Pacific Northwest forest ecosystems to climate change.  Report to the US EPA,
       Environmental Research Laboratory, Corvallis, OR

Waring, R.H. and J.F. Franklin.  1979.  Evergreen coniferous forests of the Pacific Northwest
       Science 204:1380-1386.

Whittaker, R.H.  1960.  Vegetation of the Siskiyou Mountains, Oregon and California.  Ecol.
       Monogram. 30:279-338.

Wilson, C.A., and J.F.B. Mitchell.  1987.  A doubled CO2 climate sensitivity experiment with
       a GCM including a simple  ocean.  Journal of Geophysical Research 92:13315-13343.

Zobel, D.B., A. McKee, G.M. Hawk, and C.T. Dyrness.   1976.  Relationships of environment
       to composition, structure, and diversity of forest communities of  the central western
       Cascades of Oregon. Ecological Monographs  46:135-156.
                                         27

-------
Table la. Total forest land in the Pacific Northwest (USDA 1989).  Data is for 1987.
      State
      Idaho
      Oregon
      Washington

      Total
  Total  Total  Timberland   Reserved     Other
  Land  Forest              Timberland   Forest
  Area  Area                           Land
              (thousands of acres)
  52692 21818  14533
  61546 28057  22084
  42483 21856  16848
156721
71731 53465
(290285 km2)
3051
1777
2765

7593
4234
4196
2244

10674
Table Ib.   Timberland by ownership in the Pacific Northwest (USDA 1989).  Data is for 1987.

      State   All          Total  Total         Total  Total         Total
             Ownerships  Public Federal       State  Private       Forest
                                                                Industry
                         (thousands of acres)
      Idaho  14534
      (percent)
      OR    22085
      (percent)
      WA   16849
      (percent)
      Total  53468
      (percent)
         11435 10310
         79     90
         13706 12462
         62     91
         8652  5026
         51     58
         33934 27979
         63     82
                   1036   3099
                   6      21
                   827    8379
                   6      38
                   2025   8197
                   23     49
                   3903   19734
                   12     37
                   1198
                   39
                   5114
                   61
                   4588
                   56
                   11000
                   56

-------
Table 2.  Common tree species in the Pacific Northwest
Abies amabilis
Abies procera
Abies lasiocarpa
Abies grandis
Acer macrophyllum
Alnus rubra
Arbutus menziesii
Calocedrus decurrens
Castanopsis chrvsophylla
Chamaecvparis nootkansis
Larix occidentalis
Lithocarpus densiflorus
Picea sitchensis
Picea englemannia
Pinus monticola
Pinus lambertina
Pinus ponderosa
Pseudotsuga menziessii
Quercus chrvsolepis
Quercus garryanna
Thuja plicata
Tsuga heterophylla
Tsuga mertensiana
Silver fir
Noble fir
Subalpine fir
Grand fir
Big leaf maple
Red alder
Pacific madrone
Incense cedar
Golden  chinkapin
Port-Orford-Cedar
Western larch
Tanoak
Sitka spruce
Englemann spruce
Western white pine
Sugar pine
Ponderosa pine
DouglaS-fir
Canyon live oak
Oregon white oak
Western red cedar
Western hemlock
Mountain hemlock

-------
Table 3. The comparison of the impact of climate change on mean annual temperature and growing degree/days.
The temperature is from historical records or calculated from the OSU  and GISS GCMs while growing
degree/days is calculated from the Weather routine in the  ZELIG simulation model.
Location
Forks, WA
Rainier, WA
Longview, WA
Astoria, OR
HJ Andrews, OR
Crater Lake, OR
Gold Beach, OR
Ashland, OR
Current
Temp. C
9.7
3.1
10.8
10.4
10.3
32
11.6
11.1
OSU
Temp. C
12.0
5.4
13.1
12.7
12.6
5.5
13.9
13.4
GISS
Temp. C
14.7
8.0
15.9
15.4
15.3
8.1
16.7
16.2
Current
Degree/Day
1584
511
2007
1773
1928
625
2225
2191
OSU
Degree/Day
2367
883
2794
2595
2645
984
NA
2933
GISS
Degree/Day
3340
1360
3774
3601
3663
1458
4066
3882

-------
Table  4.    The  comparison of  the  impacts  of  climate  change  on  annual
precipitation and  potential evaporation  (PET).  The precipitation  is  from
historical records  or  calculated from the OSU and GISS GCMs  while  PET is
calculated from  the Weather routine  in  the  ZELIG  simulation  model  (Urban
1990,  Urban et al.  1990)  using the Thornthwaite relationship. The ratios for
precipitation and PET are GCM/Current condition.
Location
Forks, WA
Rainier, WA
Longview, WA
Astoria, OR
HJ Andrews, OR
Crater Lake, OR
Gold Beach, OR
Ashland, OR
Current
Prec. cm
302.3
289.3
1173
176.8
135.9
469.8
210.0
48.0
OSU
Ratio
1.45
1.45
1.45
1.46
1.46
1.46
1.47
1.46
GISS
Ratio
1.27
1.27
1.27
1.27
127
1.27
1.26
1.26
Current
PET cm
62.8
42.5
66.0
63.9
64.0
41.4
65.4
67.0
OSU
Ratio
1.10
1.16
1.10
1.09
1.10
1.18
1.09
1.12
GISS
Ratio
1.21
1.34
1.24
1.21
1.24
1.34
1.22
1.27

-------
Table 5.   Areal extent of Holdridge life form classes before and after a double CO2 climate change.  Area (thousands of km2) and
percent change (in parantheses) from current conditions are given for each life  form catgory.
Life Form Category
Cold Desert
Hot Desert
Steppe
Chapparral
Boreal (Subapline)
Forest
Temperate Forest
Warm Temperate
Forest
Tropical Semi-Arid
Tropical Seasonal
Forest
Tropical Dry Forest
Total Forest Area
Current Area
77
0
215
4
130
186
2
0
0
0
318
OSU
69
(-10)
4
225
(5)
37
(757)
63
(-51)
223
(20)
15
(571)
0
0
0
301
-5
GISS
9
(-88)
6
185
(-14)
144
(3192)
26
(-80)
171
(-8)
95
(4310)
2
0
0
292
-8
GFDL
56
(-27)
23
170
(-21)
148
(3301)
15
(-89)
142
(-24)
78
(3521)
2
0
2
239
-25
UKMO
2
(-97)
15
111
(-49)
255
(5760)
0
(-100)
93
(-50)
110
(4986)
30
18
4
254
-20

-------
Table 6.  Percent of area in various vegetation zones in the central Oregon Cascade Range (latitude 44°30' north) under current climate
and with  increases of 2.5 and 5 degrees C (Franklin et al. in press).
                                         Climate
Zone
Western Slopes
Nonforested
Quercus woodland
Pseudotsuga
Tsuga heterophylla
Abies amabilis
Tsuga mertensiana
Alpine
Total Forested
Cold Snow Zone1
Eastern Slopes
Artemisia steppe
Juniperus occidentalis
Pinus ponderosa
Abies grandis
Tsuga mertensiana
Alpine
Total Forested
Cold Snow Zone2
1 includes halt ot Abies amabilis zone and all of the
2 includes Tsuga mertensiana and Alpine zones.
Current

0
0
8
56
24
9
3
97
24

0
29
22
19
17
13
58
30
Tsuga mertensiana and

+ 2.5° C

0
8
39
38
13
2
0
92
9

51
14
12
14
6
3
32
9
Alpine zones.

+ 5.0° C

0
39
27
24
2
0
0
53
1

77
11
7
4
1
0
12
1



-------
Figure Captions

Fig. 1.  Forest types in the Pacific Northwest (USDA 1970).

Fig. 2.  Current and predicted mean monthly temperatures for three sites in the Northwest

Fig. 3.  Current and predicted mean monthly precipitation for three sites in the Northwest

Fig. 4.  Holdridge life-zone classification system (Holdridge 1967).

Fig. 5.  Vegetation redistribution in the Pacific Northwest under the OSU climate scenario using the Holdridge life-zone classification
systems (Smith et al. submitted). In the map on the lower right, red depicts the pixels that change from one vegetation type to another.

Fig. 6.  Vegetation redistribution in the Pacific Northwest under the GISS climate scenario using the Holdridge life-zone classification
systems (Smith et al. submitted). In the map on the lower right, red depicts the pixels that change from one vegetation type to another.

Fig. 7.  Vegetation redistribution in the Pacific Northwest under the GFDL climate scenario using the Holdridge life-zone classification
systems (Smith et al. submitted). In the map on the lower right, red depicts the pixels that change from one vegetation type to another.

Fig.  8.   Vegetation redistribution in the  Pacific Northwest  under the UKMO climate scenario using the Holdridge life-zone
classification systems (Smith et al. submitted). In the map on the lower right, red depicts the pixels that change from one vegetation
type to another.

Figure  9. Potential shifts in forest zones under two different climate warming scenarios (Franklin et al. in press).

Figure  10. A. Simulations of current forests at Cascade Head using CLIMACS. B. Potential changes in forest vegetation at Cascade
Head as simulated by the CLIMACS model (Dale and Franklin 1989).  In both simulations, the model was initiated at year 140 using
the stand structure of plots at Cascade Head was used to  initiate the model at year 140.

Figure  11. Potential changes in forest tree species abundance at Rainier, WA, HJ. Andrews, OR, and Gold Beach, OR as simulated
by ZELIG (Urban 1990. Urban  et  aL  1990, 1991). ABAM = Abies  amabilis, ALRU = Alnus rubra, LTDE = Lithocarpus densiflora,
PSME  = Pseudotsuga menziesii, ABGR = Abies grandis, CACH = Castanopsis chrysophylla, THPL = Thuja plicata, ABLA = Abies
lasiocarpa, CHNO = Chamaecyparis  nootkatensis, PIMO = Pinus monticola, TSHE = Tsuga heterophylla, ABPR = Abies procera,
LBDE  = Libocedrus decurrens, PIPO = Pinus ponderosa, TSME = Tsuga mertensiana.

-------
   Douglas-fir
   Bortock-Sitka spr
f:| Ponferosa pine
   Hhile pine
    Lodgepole pine
    Lard)
    Fir-spruce
    Hanhoods
    PiDyon-jmiper
    Non-forest
                                               FIGURE  1

-------
        Rainier/Paradise, WA
          Lat. 46.8, Lon. 121.7, Elev. 1654m
      Mean Monthly Temperature - °C
30


25


20


15


10


 5


 0


-5
   1   234887891011  12

                  Months
     H.J. Andrews Forest,  OR
          Lat. 44.2. Lon. 122.2, Elev. 500m
      Mean Monthly Temperature - °C
30


25


20


15


10


 5


 0


-5
           Gold  Beach, OR
          Lat. 42.4, Lon. 124.4, Elev. 15m
      Mean Monthly Temperature - °C
   1   234587991011  12

                  Months
                Figure 2

-------
        Rainier/Paradise,  WA
         Lat. 46.78. Lon. 12173. Elev. 1654m
      Mean Monthly Precipitation - cm
                 8789

                   Months
80



SO



40-



30



20



10
      H.J.  Andrews Forest, OR
          Lat. 44.2, Lon. 122.2. Elev. 500m
      Mean Monthly Precipitation - cm
    12348878910t112

                   Months
           Gold Beach,  OR
           Lat. 42.4. Lon. 124.4. Elev. 15m
       Mean Monthly Precipitation - cm
                Figure  3

-------
                 Latitudinal
                 beli:
0    1.5  r--
J •/ .„,•   \ /  -r-  \     T-  Vrf *O     O
 «-
_ -\- -.	W I -.	>• f r	\.  <^f - 
       line
                                                    FIGTOE 4

-------
FIGURE 5

-------
FIGURE 6

-------
FIGURE 7

-------
.i 1   -.!--• Ul -.(I' 1
.)!   <\r\   !'->f^
                       FIGURE  0

-------
   POTENTIAL VEGETATION CHANGE
       IN THE OREGON CASCADES
      CURRENT
      +2.5UC
 7-
 .H
LJ
UJ
 1-
   OREGON
   WHITE OAK

   MOUNTAIN
    HEMLOCK
DOUGLAS
FIR

ALPINE
WESTERN
HEMLOCK
         +5.0 C
PACIFIC
SILVER FIR
                                    MANTECH / ERL-C =
              FIGURE 9

-------
               WESTERN HEMLOCK
                 AND OTHERS
   140    240   340    440    540
            STAND AGE (y«ar»)
640
100
                WESTERN HEMLOCK
                   AND OTHERS
  140    240    340    440   540
           STAND AGE (y«ari)

       FIGURE 10 A & B
640

-------
          H.J. Andrews, Gold Beach & Mt.  Rainier
         Species Relative Importance (%), Year 500
SPECIES
               ALRU

          [777] PSME
                       BOG

                         GBeach
OTHER
THPL
          BOG

             Rainier
rrrr^n

I////J PIMO
r jr A TSHE
MODEL

LOCATICN

 ABPR
 LBDE
 PIPO
 TSME
                         FIGURE  11

-------