Global Change Research Program
Environmental Research Laboratory
Corvallis, OR
FY-94 & 95

-------
U.S. Environmental Protection Agency
Office of Research and Development
GLOBAL CHANGE RESEARCH
TERRESTRIAL BIOSPHERE INTERACTIONS
Project Summaries
FY-94 & 95
Environmental Research Laboratory
Corvallis, Oregon
Dr. Thomas A. Murphy, Director
Dr. Peter A. Beedlow
GCRP Technical Director
(503) 754-4634
Dr. David T. Tingey
Program Leader:
(503) 754-4621
U.S. Environmental Protection Agency
Environmental Research Laboratory
200 SW 35th Street
Corvallis, Oregon 97333
USA

-------
Tcn-cstnal Biosphere Interactions
The ERL-C Research Program:
Terrestrial Biosphere Interactions
The global change research at (he Corvallis
Environmental Research Laboratory (ERL>C)
is a component of the Office of Research and
Development's (ORD) national program. Re-
search is conducted both on site, by EPA staff,
university cooperators, and contractor scien-
lists, and off site, by scientists at a number of
universities and other government laboratories.
Program Goals
The Corvallis research program is focusing on:
•	Predicting the response of terrestrial
vegeia tion to climate change, and sub-
sequent feedbacks to the climate system,
and
•	Assessing the extent to which terrestrial
ecosystems, particularly forests, affect
the global carbon cycle.
Overview of Ecosystem Research
The ecological uncertainties concerning global
change arc substantial, but the potential risks are
most serious. A balanced policy demands that
research be carried out now to:
*	Reduce or resolve the significant scien-
tific uncertainties regarding the effects
of global climate change on natural and
managed ecosystems, and
*	Develop the predictive capability of pro
ducing the scientific basis for cost-effec-
tive mitigation and adaptation options.
The U.S. Environmental Protection Agency has
identified global change as one of its highest
research priorities. The EPA Global Change
Research Program (GCRP) is managed from the
Office of Research and Development (ORD),
and is implemented through eight EPA research
laboratories.
Priority Scientific Questions
1.	How much will the response of the terrestrial
and near ooastal biosphere amplify or dampen
global climate change associated with green-
house gases?
Currently, it is estimated that on average about
50 percent of the carbon emissions to the atmo-
sphere is removed by sinks. For many years it
was assumed that the ocean was the major sink
and that the annual flux of carbon in the terres-
trial biosphere was in equilibrium with the atmo-
sphere if tropical deforestation was not consid-
ered. Today we know that global forest systems
capture over 110 Gt. C annually, with decompo-
sition and respiration contributing approximately
100 Gt. of CO2: to the atmosphere. Net emis-
sions from tropica] deforestation and burning
has been estimated to range between 1 and 2 Gt.
of carbon per year.
However, uncertainties regarding the size of
carbon pools and flux exist. Evidence suggests
that this sink could increase or decrease in effec-
tiveness and importance under a changing cli-
mate. A major objective of the EPA research is
to provide the scientific basis for closing the
carbon budget with respect to carbon flux to and
from the terrestrial biosphere. Early results of
the program will provide better estimates of
current fluxes. Future results will allow us to
account for terrestrial biofeedback in predic-
tions of global climate change.
2.	How will Climate Change Effect Important
Terrestrial Ecosystems?
The rate and magnitude of change could exceed
the capacity of natural systems to adapt without
dramatic disruptions, and tax our ability to man-
age forest and agriculture systems. At the present
time, our ability to predict the effects of climate
change on natural ecosystems is limited in rwo
ways. First, current climate models do not
ERL-Corvjllis

-------
Terrestrial Biosphere Interactions
provide credible estimates of change at the re-
gional scales necessary to link change to eco-
logical effects. Second, we are limited in our
ability to estimate ecological effects at regional
and continental scales.
The EPA effects program has been designed to
improve our ability to provide policy-makers
with information about potential effects of cli-
mate change on the natural environment and
major agricultural systems. The approach is to
develop a process-based understanding of glo-
bal climate change impacts and use this under-
standing to develop process-based models that
can predict steady-state shifts in regional scale
vegetation owing to climate change. Later the
program will expand to address transient re-
sponse of vegetation and associated ecological
resources. Of particular concern are systems,
including managed systems, that have been
judged as vulnerable to climate change. This
research will provide the scientific basis to esti-
mate potential effects of global climate change
on terrestrial ecosystems.
Research and Project Areas
The GCRP at ERL-Corvallis is organized into
three research areas:
•	Process and effect studies;
•	Modeling; and
•	Assessments.
Process and Effect Studies -This research area
comprises research activity to improve our
knowledge of how the global climate influences
the terrestrial biosphere and how the terrestrial
biosphere influences the global climate. Current
and planned projects include:
•	Effects of climate change on rice ecosys
terns;
•	Ecophysiological effects on forest tree
species;
•	Field studies;
•	Effects on terrestrial habitat and biodi-
versity; and
•	Interactive effects on ecological com -
plexity and ecosystem functioning.
Modeling • Modeling is part of the interpreta-
tion, synthesis, and extrapolation of process
studies, modeling is conceived, developed, and
exercised to further our understanding and test
hypotheses. The modeling area at ERL-C in-
cludes two projects:
Terrestrial systems response, vegetation redis-
tribution, and the related water balance; and
Terrestrial systems model evaluation.
Assessments - Periodic assessments are under-
taken in order to adjust the direction and focus of
our efforts and to provide policy support. A
series of phased ecological assessments have
been developed or are planned for:
•	Carbon pools and fluxes inventories for
managed ecosystems including forests
and agricultural systems;
•	Carbon pools and fluxes inventories for
managed ecosystems including forests
and agricultural systems;
•	Ecological effects of global climate
changes; and
•	International assessments.
Facilities and Expertise
The ERL-C program focus is a result of the
laboratory's unique facilities and scientific ex-
pertise in the area of vegetation ecology and
ecophysiology.
Experimental Facility - The Terrestrial Eco-
physiological Research Area (TERA) at ERL-
Corvallis is uniquely suited for the long-term
study of both the above and belowground physi-
ological responses of plants to climate change.
The core of the TERA facility is an array of 12
field chambers that provide the complete cli-
mate control of an enclosed plant/soil system.
Each chamber is independently regulated using
individual programmable process controllers,
with a central host computer system providing
communication linkage with a central database
manager and a site meteorology tower. Ambient
weather condition and CO2 levels are continu-
ously monitored by the host system, and these
conditions are use to derive the desired experi-
ERL-Corvallii
PACE 2

-------
Tefrejinal Biosphere Interaciioni
mental treatments. The belowground chamber
provides a large isolated soil reservoir which is
outfitted with an array of sensors for continuous
monitoring of temperature and soil moisture.
Sample ports providing for the detailed study of
emitted gases, soil water chemistry, and visual
root observation are also built into each system.
The TERA facility is ideally suited to test
hypotheses concerning the response of forest
trees to increased C02» increased temperature,
and decreased soil moisture. The experiments
being conducted at TERA will provide a better
understanding of the effects of climate stress on
above and belowground carbon dynamics, nu-
incnt dynamics, and tree waier use efficiency
and cvapotranspiraiion rates.
Spatial Analysis, Modeling and Simulation •
The Environmental Research Laboratory -
Corvallis has unique capabilities for performing
large scale ecological modelling by virtue of its
extensive Spatial Analysis, Modelling,and Simu-
lation (SAMS) computer facility. This network
provides a computing facility unique within
EPA and uniquely capable for the large scale
spatial modelling work outlined here. The net-
work currently serves 75 scientists, and includes
approximately 30 workstations and 40
Macintoshes/PCs. There is in excess of 50
gigabues of on-line storage, and a variety of
ljscr primers, plotters, CD-ROM readers, tape
drives for 3 different formats, optical read/write
units, and color scanners. A fiber optics link
provides direct supercomputer access. A wide
variety of software including various Geographi-
cal Information Systems and image processing
systems are available for spatial analysis and
modelling.
Scientific Expertise • Some 25 agreements with
universities and other Federal agencies are, or
have been funded through the Corvallis
laboratory to broaden the scope of the research
and to access the best scientists throughout the
world. In addition, the ERL-C program is
cooperating with the International Geosphere-
Biosphere Programme (IGBP) Global Change
and Terrestrial Ecosystems Project (Get t). The
1GBP/GCTE provides access and coordination
with research activities world wide, and provides
a forum for scientific review and planing. The
Corvallis, Oregon research community is among
the largest cadre of forest ecology researchers in
the world. The USDA Forest Service, National
Park Service, Bureau of Land Management, Soil
Conservation Service and EPA employ over 200
forest scientists and technicians in the Corvallis
community. Many of these federal research
programs focus on the role of forest systems in
the global carbon cycle in support of CEES /
GCRP. In addition, the College of Forestry at
Oregon State University is among the most
highly acclaimed forest research institutions in
the world. Unique research capabilities include
data bases, experimental forestsandplots ranging
across a gradient of precipitation, topography
and soils.
Pip 3
ERL-Corvtllis

-------
Terresirial Biosphere Imeraciions
ERL-Corvallij
PACE 4

-------
Tcrrcsinal Biosphere Interactions
Effects of Climate Change and UV-B
on the Rice Ecosystem
Problem/Goal
To determine the effects of stratospheric ozone
depletion and a changing global climate on the
irrigated rice ecosystem.
Background
As the stratospheric ozone layer is depleted
surface fluxes of ultraviolet-B (UV-B) radiation
to the earth's surface are expected to increase.
Concurrently, a build-up of trace gases (C02,
methane, and others) in the atmosphere is caus-
ing atmospheric wanning and associated cli-
matic change. The potential effects of increased
UV-B and climate change on the world's agri-
culture are of great immediate concern since
crop productivity must continue to increase to
meet the demands of a growing human popula-
tion. In addition, agricultural crops can emit
greenhouse gases to the atmosphere, contribut-
ing to climate change.
Recent research has focusscd on the direct im-
pacts of climate change on potential yield from
crops. However, more critical and realistic
impacts of climate change will occur on com-
plex agricultural systems including several crops
as well as insects, diseases, and weeds which can
substantially reduce yields.
Rice is (he most important food crop in the
world, supplying 21% of all calories consumed
by human beings. Rice is particularly important
in Asia where it supplies as much as 75% of the
daily calories in countries such as Bangladesh
and Myanmar. Our knowledge of the response
of the rice plant to these environmental changes
is quite meager. Furthermore, the net impact of
the changes will depend on a balance between
potential increases in yields with increased CO2
for a C3 species, such as rice, vs. unpredictable
effects of temperature and precipitation changes.
For example, crop areas could expand north-
ward with increased temperatures, but suffer
from high temperatures in other areas.
In addition, rice is unique among crops in that it
is a major contributor to global emissions of
methane, a critical greenhouse gas. Thus, any
efforts to assess the risk to rice from climate
change must also consider the impacts of cli-
mate change and mitigation strategies on meth-
ane emissions from rice fields.
Science Objectives
There will be 4 foci for specific research in this
area:
1.	To determine the responses of rice geno-
types and important components of the
rice system (diseases, insects) to UV-B
radiation, increased C02, and increased
temperatures.
2.	To evaluate the effects of UV-B radia-
tion, C02, and changing precipitation
and temperature on crop yield on a re-
gional basis in south and east Asia.
3.	To determine methane emissions from
rice fields and the impacts of climaic
change on the emissions.
4.	To evaluate adaptation options can be
identified to minimize negative effects
of UV-B and climate on rice yields and
methane emissions from rice fields.
Approach
The research is being conducted at the Interna-
tional Rice Research Institute (IRRI) in the Phil-
ippines, the world's center for rice research,
with further collaboration with leading rice re-
search groups in Asia, the United States and
Europe. Experimental studies focus on field
conditions with supporting laboratory and
phytotron research. Physiological rice simula-
tion models are being used to integrate and
extrapolate experimental results to the regional
level using Geographic Information Systems
(CIS). This research will characterize at re-
ERL-Corvallu

-------
Terrestrial Biosphere Interactions
gional scales the risks to rice productivity from
UV-B and climate change.
This program is part of a continuing research
effort initiated in FY 90 and continuing for five
years. Its general direction and focus responds
to program-level peer reviews, and the indi-
vidual projects within it are subjected to periodic
peer reviews. The scientific teams are in place.
Data acquisition and analyses are ongoing. Spe-
cific activities at IRRI and it's collaborators are:
UV-B EFFECTS ON RICE - Field studies on
effects of UV-B on rice genotypes (tolerant and
sensitive) are being carried out emphasizing
phenology, morphology, growth, and yield of
rice plants. Studies are being conducted during
both the wet and dry cropping seasons. Phytotron
(temperature and humidity controlled green-
houses) studies are focussing on the completion
of cultivar screening studies and the mecha-
nisms for differences in UV-B sensitivity among
cultivars. Collaborative studies are underway
with the People's Republic of China.
UV-B AND CLIMATE CHANGE EFFECTS
ON DISEASE - Research models the effects of
UV-B and climate change on the rice - rice blast
disease system. The results from the modeling
are being evaluated on a spatial basis for Asia
ureas using a GIS. Experiments at Washington
State University are determining the direct ef-
fects of UV-B on the blast fungi itself and on
fungi-rice interactions.
C02 AND TEMPERATURE EFFECTS ON
RICE - Experiments in the IRRI Phytotron
focus is on screening for the range sensitivity to
increased C02 and temperature among domes-
tic cultivars and wild genotypes of rice. Experi-
ments in open-top field chambers focus on the
interactive effects of CO2, temperature, and
nitrogen fertilization on rice. The emphasis is on
photosynthetic responses of the rice genotypes,
phenological, morphological, growth, and yield
component responses
EMISSIONS OF METHANE FROM WET-
LAND RICE FIELDS- Studies at IRRI use an
automatic sampling and analyzing system for
continuous measurement of methane emissions
from wetland rice fields.based on a collabora-
tive effort with the Fraunhofer Institute for At-
mospheric Research in Germany. Studies are
being conducted on the effects of effects of
source and manner of fertilization, and rice cul-
tivar on methane emissions in rice fields. The
effects of soil properties (pH, redox potential,
electrical conductivity, texture organic matter,
total nitrogen) on methane production are being
studied in the laboratory for rice soils. Collabo-
rative studies at Louisiana State University
(LSU), focus on the influence of soil redoxpH
conditions on methane production and methane
oxidation rates using soil microcosms, the ef-
fects of soil properties on methane production
rates on soils in the laboratory, and effects of
organic matter, urea, and main vs. second rice
crop on methane emission rates in for rice fields.
MODELING OF IMPACT OF CLIMATE
CHANGE ON RICE YIELDS - Modeling ac-
tivities are being carried out at IRRI and in
Japan, the Republic of Korea, Peoples Republic
of China, Malaysia, India, and the United States.
The rice model ORYZA1 and a model from
Japan are being used to estimate rice yield.
Researchers are carrying out computer simula-
tion and spatial analysis using a GIS to produce
estimates of the impacts of climate change on
rice yields in different agroclimatic regions of
Asia.
IMPACTS OFCLIMAT1CCHANGE ON RICE
ECOSYSTEM - Experiments are being con-
ducted on the impacts of climate change (in-
creased temperature) on rice-insect interactions.
Experiments evaluate impacts on insect popula-
tion demographics, feeding rates, predator-prey
interactions. The experiments are being con-
ducted in the Phytotron and laboratories at IRRI
with collaborative research with the People's
Republic of China and Thailand. Models of
insect/rice interactions are being evaluated for
use in simulations of impacts of climate change
on pest outbreaks for key agroclimatic regions
of Asia, especially through collaborative re-
search with the Republic of Korea.
ERL-Corvallis
PACE 6

-------
Tcrresinal Biosphere Interactions
MAJOR PRODUCTS
FY94- Annual Report on Effects of UV-B
and Climate Change on Irrigated Rice
Ecosystem
FY95 Assessment of impacts of UV-B ra-
diation and climate change on the
wetland rice cropping system (Pro-
ceedingsoflnternational Symposium
on Rice and Climate Change in
March, 1994).
FY95 Annual Report on Effects of UV-B
and Clim.uc Changeon Irrigated Rice
Ecosystem
FY96 Final Report on Effects of UV-B and
Climate Change on Irrigated Rice
Ecos>sicm
BUDGET (SK)
FY 90
FY91
FY92
1300
1735
1550
FY9?
FY94
FY9S
1663
1338
500
Page 7	ERL-Corvillii

-------
Terrestrial Biosphere Interactions
ERL-Corvallii	FACES

-------
Terrestrial Biosphere Inieraciioni
Ecophysloloeical Effects of CO2
and Climate Change
Problem/Goals
The research is focused on three major goals
including:
1.	What are the effects of elevated COj and
climate change on the growth and pro-
ductivity of forest trees and soil pro-
cesses?
2.	Will elevated COj and climate change
alter the carbon sequestration potential
of forest trees and forest soils?
3.	What is the magnitude of elevated CO2
and climate change impacts on forest
trees and soil, and will the impacts be
widely distributed?
Background
Concentrations of carbon dioxide (CO2) and
other trace ga^es are increasing in the Earth's
atmosphere due to human activities. Evidence
suggests that the increased gas levels will pro-
duce increased global temperatures and result-
ant changes in precipitation, cloudiness and other
atmospheric factors (known collectively as "cli-
mate change"). The atmospheric changes are
likely to affect all Earth processes either directly
or indirectly. Of interest to future integrity of
terrestrial systems and human security are the
responses of forests. Understanding forest re-
sponses is crucial for resource management plan-
ning because of the long-lived nature of forests
and the long lead time to realize many of the
societal benefits derived from forests.
Science Objectives
Existing data do not provide scientifically de-
fensible answers to these policy issues at the
individual tree or forest stand levels. An experi-
mental study was undertaken to contribute to the
major project goals and policy issues. The ex-
periment, which utilizes the Terrestrial Eco-
physiology Research Area (TERA) of the
Corvallis* Global Processes and Effects Pro-
gram (GPEP), is designed to provide:
1.	Determine of how CO2 and climate
change will affect forest seedling eco-
systems as represented by North Ameri-
can Pacific Northwest forests, and
2.	Assess how forest ecosystems will feed
back to atmospheric trace gas chemistry.
Approach
The TERA experiment is a complex, multi-
faceted, highly integrated empirical effort fo-
cusing on both numerous abiotic and biotic
aspects of the above- and below-ground portions
of a Douglas fir seedling "ecosystem". The
project is comprised of four separate but inter-
acting research activities:
•	Scoping Studies,
•	Experimental Tasks,
•	Modeling Tasks, and
•	Integration and Inference Activities.
Current work is focused on the experimental
(integrated plant and soil, and integrated above-
and below-ground studies) and the iterative
modeling activities. The experimental work is
organized around seven tasks:
•	Shoot Carbon and Water Flux,
•	Shoot Growth and Phenology,
•	"Ecosystem" Nutrients,
•	"Ecosystem" Water,
•	Litter Layer,
•	Root Growth and Phenology, and
•	Soil Biology.
Work will begin to include the inference and
integration activities asmorephenomenological
data are collected.

ERL-Corvallis

-------
Terrestrial Biosphere Interactions
The TERA facility is a state-of-lhe-science re-
search center with the capability to investigate
the effects of elevated CO^ and climate change
on plants and soils. This facility is unique
compared with others because it is designed to
measure and track accurately ambient CO2, tern-
perature and dew point while operating continu-
ously for several years. The central component
of the facility is 12 SPAR (Soil-Plant-Atmo-
sphere Research) units, called terracosms, which
utilize solar radiance and are capable of provid-
ing complete climate control of an enclosed
plant/soil system.
The experimental design is a 2 x 2 factorial [two
levels of CO2 (ambient & ambient +200 ppm)
and two levels of temperature (ambient & ambi-
ent +4 C)j with three replications of each treat-
ment. The experimental treatments are defined
in relation to current ambient conditions; addi-
tions to the ambient conditions are made to
create the elevated treatments. The use of ambi-
ent and ambient-plus treatments insures that the
experimental treatments maintain the climatic
linkages.
In addition, to the treatment variables, the ex-
periment includes Controlled Variables (levels
controlled during the experiment) and Initial-
ized Variables (levels set at the beginning of the
experiment, and then allowed to vary according
to treatment effects). Controlled Variables in-
clude vapor pressure deficit and soil moisture.
Initialized Variables include soil nutrient pools,
and soil flora and fauna.
M^jor Products
FY95 Root longevity and turnover under
conditions of elevated CO2 and tem-
perature.
FY96 Responseofsoilbiologicalprocesses,
populations and diversity to elevated
CO2 and temperature.
FY97 Parameterized and calibrated tree
growth model to describe effects of
elevated CO2 and temperature
FY98 "Ecosystem" Budgets for Carbon,
Water & Nitrogen.
Budget ($ K)
Em
615
FY95
1200
Em
1000
FY 96-98
1300
£124
1200
The experiment also consists of a field study in
the Oregon Cascade Mountains. Three field
sites ranging in elevation from approximately
500 to 1400 m were planted to Douglas fir at the
same density as in the terracosms. Selected
environmental conditions are being monitored.
A subset of the seedlings will be harvested
annually to make comparisons with seedling
growth and soil processes in the terracosms, and
to calibrate the process model under develop-
ment to project treatment effects on tree growth.
F.RI.-Cotvallis
PACE 10

-------
Terrestrial Biosphere Interactions
Habitat Sensitivity to Climate
Change
Problem/Goal
The problem is to determine the potential effects
of habitat modification through land use change
and global climate change on animal habitats
and the consequent impacts on animal popula-
tions.
Background
EPA's Science Advisory Board listed habitat
modification, global climate change, and loss of
biodiversity as three of the highest risk environ-
mental problems. Habitat modification from land
use changes is currently adversely affecting bio-
di\ ersity and other ecological services. Climate
change has the potential to compound the prob-
lem by producing additional climatic habitat
modification, as well as further affecting land
use. The SAB further recommended the devel-
opment of tools for comparative ecological risk
assessments for multiple anthropogenic stres-
sors.
Science Objectives
1.	Determine the comparative risk to a se-
lected fauna] group at (he regional to
continental scale from land use change
habitat modification, climate change, and
combined effects.
2.	Determine the comparative risk to ani-
mal habitats and biodiversity for a se-
lected landscape from land use change
habitat modification, climate change,and
combined effects.
Approach
A competitive Request for Proposals was issued
for regional (Science Objective 1) and landscape
level (Science Objective 2) comparative risk
assessment studies. The choice of fauna! groups,
geographic areas, and methodology was left up
to the investigators to propose and justify. One
three year Cooperative Agreement (10/92 • 9/
95) was established for each of the two types of
studies.
LANDSCAPE STUDY - The University of
Georgia (Ronald Pulliam, PI) was awarded the
Cooperative Agreement on "Comparative Risk
Assessment of Climate Change and Other An-
thropogenic Stressors: Habitat and Biological
Diversity on the Savannah River Site, South
Carolina." The approach is to model population
and community dynamics in response to climate
and land use changes via their influence on
habitat suitability and availability. CIS land-
scape mapping will be combined with a spatially
explicit population dynamics simulation model
that is sensitive to the distribution of habitat
patches across the landscape. An existing ver-
sion of this model has already been used to
model the response of one particular bird species
to land use changes at the Savannah River Site.
The model will be generalized to allow applica-
tion to a variety of vertebrate species. This
generalized model will be used in a comparative
risk assessment to estimate the relative impacts
of climate and land use changes on a variety of
vertebrate groups. The goal is to develop a
genera] approach which could be applied to a
variety of landscapes, animal species, and large-
scale stressors.
REGIONAL STUDY • South Dakota Slate Uni-
versity (Carter Johnson, PI) was awarded the
Cooperative Agreement on "Global Warming
and Prairie Wetlands: Potential Consequences
for Waterfowl". The approach is to:
•	Develop hydrologic and vegetation mod-
els for temporary, seasonal, and semi-
permanent wetlands;
•	Assemble them into a landscape-level
wetlands model;
•	Project the effects of climate change and
land use change on wetlands area and
habitat quality by wetlands type;
•	Develop remote sensing protocols to
P»ge11
ERL-Cbrvallit

-------
Terrestrial Biosphere Interactions
delect changes in wetland area and con-
dition which might indicate climate
change responses; and
Relate projected changes in wetlands
habitat to wate rfowl population size, pro-
ductivity, and species diversity.
Major Products
FY94 Manuscript on comparative risk as-
sessment for Savannah River Site
region, comparing climate change,
habitat fragmentation, and land-use
change as stressors on the bioticcom-
munity.
FY94 Report on comparative risk assess-
ment for effects of land use change
and climate change on waterfowl in
prairie wetlands.
Budget ($ K)
FY92
440
FY93
230
FY 9 4
358.4
ERL-Corvillis
PACE 12

-------
Terminal Biosphere Interactions
Field Validation - Processes/Response
Problem/Goal
The field data required to parameterize simula-
tion models and validate model projections as
well as extrapolate the results of experimental
studies to larger spatial and temporal scales are
lacking. For example, most forest carbon bud-
gets have been made from forest inventories that
were designed to evaluate productivity, mer-
chantable lumber, silvjcultural practices,while
informaiion on soil organic matter (SOM) gen-
erally came from broad soil surveys or were
estimated. Although valuable for original as-
sessments of carbon pools and fluxes, there is a
critical need to evaluate many of those assump-
tions, validate the conversion concepts (tree
diameter at breast height, dbh, to total forest
carbon), and look carefully at the soil carbon
pools and flux rates
Background
The development of field calibration and valida-
tion sck requires extensive measures of biologi-
cal processes and rates. For example, the U.S.
Forest Service (USFS) is involved in various
studies to understand the effect of different for-
est nunagement strategies on forest growth,
health and wildlife habitat. One of the most
ambitious is called the Long Term Ecosystem
Productivity Program (LTEP). It was initiated
by the USFS in the Pacific North West several
years ago with the investment of over 30,000
acres dedicated for an experimental period of
200 years. This activity involves a large team of
scientist who are engaged in measuring various
ecological parameters and correlating them with
forest growth and stability. Much of the data
needed to make carbon budgets and understand
the impacts of forest management on carbon
pools and fluxes is contained in the data of these
studies.
Objectives
The initial objectives of the research are:
1.	Determine carbon pools and fluxes in
forests managed to attain a variety of
serai conditions. Compare the rates of
carbon sequestration and retention on
these sites. Particular attention to below
ground processes will be part of these
budgets.
2.	Compare temporal patterns of carbon
flux and sequestration pools under dif-
ferent management practices to make
specific recommendations regardingfor-
est management and harvest cycles.
3.	Determine scaling factors for carbon and
water fluxes in coniferous forests for
extrapolating the results of experimental
chamber studies to larger spatial and
temporal scales.
Approach
Measured carbon fluxes will be compared with
that inferred from ecological measurements and
models. The USFS will use data from the LTEP
to create site specific carbon budgets. Since
these data were not collected specifically for this
purpose, additional sampling and analysis by the
USFS will be required. We will use the results
to compare management practices for their im-
pact on carbon pools and fluxes. The site
specific carbon budgets will be authenticated
data sets of carbon in all compartments of vari-
ous forest types. Alt sites will be measured and
evaluated in the same manner and all aspects of
the measurements, sampl ing procedures and sta-
tistical treatment will be subjected to quality
assurance evaluation. These data sets will form
a basic resource for model validation. Carbon
budgets will provide the information to evaluate
the impact of various forest management prac-
tices on carbon sequestration and flux. This will
P»ge 13
ERL-CorvalUs

-------
Terrestrial Biosphere Interaction*
be an important aspect of a national mitigation
strategy. Corollary information regarding at-
mospheric, and soil conditions along with forest
type will be the basis for making recommenda-
tions in support of forest management practices
which foster increased carbon sequestration.
Ecophysiological measures (e.g. photosynthe-
sis, water flux, carbon allocation, soil carbon
dynamics) of plant and soil processes will be
measured on forest trees of various age structure
and stand conditions to develop the scaling fac-
tors required to extend experimental data from
small tree studies to mature stands distributed
across a landscape.
In support of AREAL's GCRP program estab-
lish gas flux measurements in a representative
forest system nearthe ERL-C. Apply traditional
ecological measurements and models to this
site, and compare the carbon fluxes to the di-
rectly measured flux. Carbon flux values for
Western coniferous forests will yield under-
standing of the effects of important meteorologi-
cal parameters. Especially important are tem-
perature and frequency and timing of precipita-
tion. These values will allow validation of forest
carbon dynamic models and the yearly patterns
will allow validation of carbon budget models
and also aid in the extension of experimental
results to natural conditions.
This project would also support field data col-
lection in conjunction with the EMSL-LV re-
mote sensing program, specifically, the pro-
posed Forest Pilot project.
Deliverables:
FY96 Report on carbon dynamics in a
mature Douglas fir forest,
FY97 Report of Forest management prac-
tices on Carbon cycling.
FY98 Development of scaling factors for
extending the resultsof EPA'sTERA
experiment to larger temporal and
spatial scales.
Budget: <$ K)
£124 H21	FY96-98
600 600	1000
ERL-Corvallit
PACE 14

-------
Terrestrial Biosphere Interactions
Global Change and Ecological Complexity
Problem/Goal
Determine the impact of global change on eco-
logical complexity in natural and managed eco-
systems, and the accompanying effects on eco-
system functioning. The research will not sim-
ply focus on responses of dominant plants (e.g.
trees or crops) to environmental changes, but
rather, it will address the relationships among
multiple ecosystem biotic components, the ef-
fects of environmental changes on these biotic
components, and how the interacting responses
affect ecosystem processes such as productivity
and carbon dynamics.
Background
Ecological complexity can be measured on many
scales.These include the genetic diversity within
species, species diversity within ecosystems,
and ecosystem diversity across the landscape.
Human activities have already caused signifi-
cant reductions in diversity at all of these levels.
Future global change, including continued hu-
man population growth, land use change, and
the likelihood of global warmingwill exacerbate
all these problems, and will likely accelerate the
loss of ecological complexity. Ecological com-
plexity is related to ecosystem functioning in
ways that are only partly understood. Further
understanding of these relationships at multiple
scales and the ways in which rapid environmen-
tal change may affect them is critical at this stage
of human development and the history of life on
Earth.
The research is designed to determine the impact
of global climate change on two fundamental
ecological questions:
1. Does the number of species (biodiver-
sity) "count" in system processes (e.g.,
nutrient retention, decomposition, pro-
duction, etc) over short- and long-term
time spans, and in the face of global
change?
2. How is system stability and resistance
affected by species diversity and how
will global change affect these relation-
ships?
SCIENCE OBJECTIVES
To assess the impact of CO2 and climate change
on ecosystem function and complexity, it is
necessary to consider several scales of study.
Consequently, there will be 3 foci for specific
research projects in this area ranging from small
scale (plot) soil/plant ecosystems to the land-
scape level:
1.	SOIL BIOTIC COMPLEXITY - Deter-
mine the effects of altering the diversity
of soil organisms on plant and soil car-
bon & nutrient dynamics, and the effects
of CO2, temperature, and moisture on
these processes.
2.	COMPLEX CROPPING SYSTEMS -
Determine the effects of climate change
(COji temperature, moisture) on major
biotic plant stresses (e.g., insects, dis-
eases, and weeds) and the joint influence
of these biotic stresses and climate change
on the productivity of complex cropping
systems.
3.	LANDSCAPE SCALE COMPLEXITY
Examine potential landscape scale re-
sponses in ecological complexity result-
ing from interacting effects of climate
change and land use change.
APPROACH
The work will build upon current EPA research
in the GCRP and related programs including
work on rice ecosystems, habitat sensitivity, the
TERAexperimental facility, vegetation redistri-
bution modeling, and the Biodiversity Research
Consortium. Data base analyses and simulation
modeling will utilize the spatial data bases and
distributed computing facilities assembled in
P»gei5
ERL-Corvillis

-------
Terminal Biosphere Fncersctiom
the ERL-C Spatial Analysis, Simulation, and
Modeling (SAMS) facility. The proposed ap-
proaches were selected so that the EPA research
wouldbe closely linked with andsupportGCTE's
Focus 4 "Global Change and Ecological Com-
plexity" and Focus 3 "Global Change Impacts
on Agriculture and Forestry". Experimental stud-
ies, analysis of existing geographic data bases,
and simulation modeling will be used in the
research.
Experimental Studies
SOIL BIOTIC COMPLEXITY-This work will
build upon the existing ecophysiological re-
search in the TERA experimental facility. An
experimental approach will be taken to examine
the effects of increased C02 and temperature on
plant and soil carbon & nutrient dynamics and
the relationships among soil organism diversity
on these processes. Three different scales of
experimental systems will be utilized, repre-
senting a gradient from open canopy, open nutri-
ent cycling tree seedling mesocosm systems, to
closed canopy, closed nutrient cycling forests.
COMPLEX CROPPING SYSTEMS - This work
will build upon the existing rice ecosystem re-
search project. Experiments will be conducted
on different scales of complexity ranging from a
single species to multi-species systems. Single
species experiments will be conducted to estab-
lish fundamental relationships among crop and
pest population characteristics and climate
change parameters (CO2, temperature, mois-
ture). Dual species experiments (e.g. insect/crop,
disease/crop, and weed/crop) will obtain data
for simple general models of insect feeding,
disease severity, and plant competition. Com-
plex, multi-species experiments will examine
climate change impacts on crop yield, disease
prevalence, and insect and weed populations as
endpoints. A range of environments including
controlled environment chambers, open or closed
field chambers, and open FACE (Free Air CO2
Enrichment) systems may be utilized.
Analysis of Existing Geographic Data Bases
COMPLEX CROPPING SYSTEMS - Analysis
of existing biological, agricultural, and geo-
graphic data bases will be used to explore the
potential for significant impacts on the geo-
graphic distribution of cropping systems and
their major pests.
LANDSCAPE SCALE COMPLEXITY - The
research planning process will develop specific
plans for analyses of spatial data bases to exam-
ine the potential for climate and land use changes
impacts on complexity at the landscape scale.
For example, the Biodiversity Research Consor-
tium (EPA, USFS, USFWS, USGS, The Nature
Conservancy, and several universities) is in the
process of producing integrated data bases on
animal distributions and spatial patterns of suit-
able habitat for 6 pilot study regions, and even-
tually for the contiguous United States. These
data bases may be analyzed using a Geographic
Information System to examine the connectivity
of existing suitable habitat for selected verte-
brates as possible dispersal/migration corridors
in response to environmental changes. Analyses
may be done at several different spatial scales.
Simulation Modeling
SOIL BIOTIC COMPLEXITY - Acurren t ERL-
C research program includes the use of single
tree growth models to simulate responses to
climate change. The proposed work will build
on this foundation and will link in soil process
models and some components of the soil biota,
particularly the structure and function of soil
food webs.
COMPLEX CROPPING SYSTEMS - Model-
ing will integrate experimental results and spa-
tial data bases to predict crop yield and pest
population responses on field, landscape, and
regional scales. This approach will include iden-
tification, improvement, testing, and running of
models.
LANDSCAPE SCALE COMPLEXITY - Simu-
lation modeling activities will build upon cur-
rent ERL-C work examining redistribution of
ERL-Corvallit
PACE 16

-------
Terrestrial Biosphere Interactions
natural vegetation and agroecological zones in
response to climate change. Possible areas for
further work inc! ude the use of forest gap models
to project changes in forest structure and suit-
ability as habitat for forest animals; projecting
potential range shifts of forest and agricultural
pest populations in response to global change;
etc.
MAJOR PRODUCTS
FY98 Changes in production as affected
by changes in insects, weed inci-
dence and disease severity in con-
junction with climate change.
FY99 Interacting effects of climate change
and soil biotic diversity on carbon
and nutrient cycling.
FY99 Animal habitat structural changes
including connectivity of migration
corridors in response to climate
change.
BUDGET ($ K)
FY9S FY96-98
1200 1200
Pap 17
ERL-Corvsllis

-------
Terminal Biosphere Interaction!
ERL-Corvjlht	PACE 18

-------
Terrestrial Biosphere Interactions
Processes of Ecosystem Response to
Global Environmental Change
Problem/Goal
Current interest is increasing among CEES, IPCC
and governmental agencies in the definition and
measurement of ecological responses to global
environmental change. Political action on abate-
ment of greenhouse gas emissions is not likely to
occur until scientific evidence is presented to
indicate that the biosphere is responding in some
functional manner to global environmental
change. The development and signing of the
Montreal Convention on ozone indicates that the
evidence should be both conclusive and scien-
tifically obvious. While no one ecological re-
sponse will be definitive, the usual ambiguity of
scientific research can be circumvented in the
formJtion of policy, if theory can predict the
emergence of an interrelated multitude of eco-
logical changes which subsequently are mea-
sured in the field.
Background
Although no one ecological response will be
definitive for change detection, this limitation
can be circumvented, if ecological theory can
predict the emergence of number interrelated
measures of ecological changes which subse-
quently are measured in the field. Thus, defini-
tive evidence of changes in ecosystem structure
and functioning must be obtained at the land-
scape to global scale. Considering that the
singularity of the earth precludes rigorous test-
ing and scientific proof, and considering the
complexity of ecosystems in combination with
our sketchy knowledge of how they function, the
task may seem impossible. Yet, such evidence
could be obtained in a few years through (a) the
careful choice for study of simple, widespread
ecosystems which are themselves important in
the global carbon cycle, (b)definition of a unique
"fingerprint" of expected changes, and (c) judi-
cious use of environmental monitoring which
was aimed at measuring the characteristics of
the fingerprint.
Detection of a fingerprint implies that the natu-
ral variation in ecosystems from one time to
another has been documented well enough to
distinguish it from chemical and climate
changerelated variations. Hence, the challenge
is to provide the definitive evidence of signifi-
cant change in ecosystem functioning at a global
scale, to document the difference between the
changes and those resulting from natural ecosys-
tem dynamics, and to demonstrate that the only
logical source of change has been shifting atmo-
spheric chemistry and climate.
Scientific Objectives
The primary objectives of a program designed
for early or first detection of significant ecosys-
tem response to global environmental change
are to:
1.	provide definitive evidence of signifi-
cant, unidirectional and lasting changes
in ecosystem functioning and to demon-
strate that these changes are occurring
globally;
2.	document the differences between the
measured changes and those resulting
from natural ecological dynamics, or
from other external forcing; and
3.	demonstrate that shifts in atmospheric
chemistry and/or climate provide the
most complete and logical source of the
changes.
Approach
The data to be collected in support of these
investigations must extend throughout the globe
to avoid false research trails created by region-
ally-unique changes. They must involve long-
lived trees and forest communities in order to
assure directionality and ecological significance
Pap 19
ERL-Corvallis

-------
Terrestrial Biosphere Inicracnonj
of changes. Only perennial vegetation pos-
sesses age structures capable of predicting veg-
etation composition and density into future cen-
turies. Finally, the data collection must focus on
identifiable tree and forest borders where cli-
matic gradients and thresholds control ecologi-
cal dynamics, to select the least ambiguous and
largest magnitude responses.
The most suitable subject of first-detection in-
vestigations is the circumpolar belt and related
montane areas in which low temperatures limit
tree growth. This isone of the most temperature-
sensitive ecotones in the world. Absent or mini-
mal are the intensive land uses that mask climate
response and the influence of other gaseous
pollutants The physiognomic contrast of forest
and tundra, or montane forest and alpine vegeta-
tion, is easily recognized in remotely sensed
data.
The strategy must be to hypothesize, a priori, a
set of related ecological responses, from a set of
measured environmental changes, correspond-
ing to the criteria stated above, then to provide
rigorous statistically-relevant tests of their pres-
ence. In the absence of rigorous scientific proof,
the documentation of a large set of non-defini-
tive changes (corresponding to an ecological
fingerprint) can permit intelligent policy formu-
lation based on otherwise ambiguous science.
The considerations discussed above suggest that
the long-term project should be focussed on six
primary subject areas:
1.	Fossil pollen data (mostly extant) can be
used to describe the global geography of
the ecotone during the past 10,000 years.
2.	Tree-ring data (also mostly extant) can
be used to describe the growth patterns
of the trees before and during the indus-
trial age.
3 Remotely sensed data from aerial pho-
tography and satellite imagery should be
applied to describe the detailed geogra-
phy of the ecotone recently, and today.
4.	Ecosystem process studies should be
aimed at quantifying the spatial and tem-
poral relationships between the ecotone
and the factors which control it.
5.	Mathematical modeling, based prima-
rily on the evidence collected in parts 1-
4, should be used to describe likely fu-
ture responses of the ecotone, some of
which will be measurable now as defini-
tive hypothesis tests.
6.	General Circulation models of the atmo-
sphere, and available monitoring and
research data, should be used to provide
estimates of current and future climate to
drive ecological models, and of past cli-
mate to validate ecological models.
Mqjor Products
FY95 Major scientific meeting to define
ecological fingerprint inboreal wood-
lands and montane treeline 94
FY99 Characterization of symptomatic eco
system response to global environ-
mental change
Budget ($ K)
FY 94	FY95	FY 96-98
400	400	400
ERL-Corvillis
PACE 30

-------
Terrejinil Biosphere tmeraeiions
Transient Response of Vegetation to
Environmental Change
Problem/Goal
The temporal response of the terrestrial bio-
sphere to changing climate is unlikely to keep up
with the rapid temporal pattemof climate change,
inducing gradually increasing lags in vegetation
dynamics. The lagged ecosystem properties
(slowed growth rates of individuals; loss and
gain in species; declines in community density)
are likely to generate irregularities in slowly
changing rates of terrestrial carbon storage and
release, such that the earth will behave as a
source of carbon during several decades at one
time, and as a sink for carbon during several
decades at other times. In addition to carbon
cycle irregularities, currently undocumented
potential declines in biodiversity are possible as
species die out in certain areas and their replace-
ments appear (here much later. The objective of
the research is to understand these transient
responses of vegetation communities from local
to global scales and to devise, test and apply a
framework for their prediction under changing
climate and atmospheric chemistry.
Background
Accurate prediction of transient responses of
long-lived vegetation to rapidly changing cli-
mate is critical for addressing questions of veg-
etation redistribution; in fact, that is the only
place where our central issue of "vegetation
redistribution" is addressed; the rest of the mod-
els we and others produce are either aimed at
predicting "environmental changes which must
force a redistribution of vegetation" or else,
"vegetation which eventuallycould appear some-
time following its redistribution". These latter
approaches can generate valid estimates of the
vegetation which must change or disappear with
climate change, a prediction which constitutes
half the concern about vegetation response to
environmental change. However, they cannot
predict the vegetation which will replace "cli-
matically outmoded" communities, which con-
stitutes the other half of the concern. Here,
replacement vegetation depends on lags in mi-
gration of tree species to appropriate growth
sites, and slow growth of trees to reproductive
maturity. These lags, in turn, control variations
in the flux of carbon between atmosphere and
vegetation, generating pulses of atmospheric
carbon when niches remain empty, and pulses of
carbon sequestration when appropriate species
become established in emptied regions.
Success in predicting transient responses of spe-
cies composition, plant density, establishment,
growth and mortality, can be based on processes
which define transient responses in vegetation:
migration of species across landscapes dissected
by human uses, plant succession and delayed
reproductive maturity of species, growth de-
clines from chronic stress of continuous warm-
ing and shifts in disturbance regimes in the form
of wildfire and pest/disease epidemics.
The potential for environmental change rates so
rapid that species cannot migrate fast enough to
fill newly-available niches indicates the need to
predict migration potential. Indeed, climate
change rate may be too rapid for completion of
life-cycles of slower growing species, generat-
ing local extinctions of species. Predicting the
magnitude of these problems demands consider-
ation of processes defining effects of external
forcing on separate life stages and understand-
ing of the implications of interspecific competi-
tion during each life stage.
Scientific Objectives
The capability to predict carbon cycle character-
istics from simultaneous temporal and spatial
vegetation dynamics requires a landscape-level
approach to processes which both impede and
facilitate propagule transport and establishment,
and which generate multi-dccadal lags between
establishment of seedlings and tree growth to
reproductive maturity. The goal of this research
is to generate a framework for assessing effects
Page 21
ERL-Corvillu

-------
Terrestrial Biosphere Interaction!
of forest succession and tree migration, to test
the quantitative aspects of that framework on
field data (fossil pollen evidence; current age
distributions), and to apply the framework to
predicting future vegetation and carbon cycle
dynamics which must result from global envi-
ronmental changes if the hypotheses on which
the framework is based are correct. This re-
quires, in turn, spatially distributed global data
sets on current soil properties, topography, land
use, climate and vegetation distributions. Spa-
tially distributed global data sets from earlier
periods of different climate are also required on
soils, topography, climate and vegetation. The
soils properties must be documented to define
local characteristics of agricultural and wildland
vegetation productivity, along with the relation-
ships and feedbacks be tween vegetation produc-
tivity and hydrologic cycle elements.
Approach
The research will include field experiments and
data analyses, tests of predictive frameworks
and climate impact assessments from the Co-
lumbia River Watershed, the North American
continent, and the global terrestrial biosphere.
The hypothesized nature of transient processes
will be tested by comparison of predicted tree
migrations and succession with measured tree
migration in the midwestem U.S. and western
Europe. There, data networks describing veg-
etation dynamics of the past 15,000 years are
dense enough to support such analyses. Addi-
tional hypotheses on current importance of tran-
sient processes will be examined by remotely-
sensed and ground data collections from high
latitudes under the title of "Early detection of
ecological response to global environmental
change." Relationships between ecological pro-
cesses and climate and land use on the current
landscape will be examined to determine the
resulting carbon fluxes and sequestration char-
acteristics, and, their relationship to hydrology.
Statistically-defined relationships tmong car-
bon stock and flux variables will be tested by
comparison of predictions to specific data col-
lected at individual research sites.
M^jor Products
FY 95 Global analyses of the potential to
modify current terrestrial carbon flux
by changing land use pattern
FY 96 Global framework for integrating tree
migration, forest succession, and
landscape-level disturbances, tested
at landscape scales in the midwest
and Europe and linked to a global
integrated assessment model
FY 98 Estimates of global carbon cycle ir-
regularities (pulses) during the next
century from transient processes, land
use and climate change predictions
FY 98 Measurements of current transient
response of ecosystems to global
change from high latitudes
Annual Budget ($ K)
EL24 FY 95 FY 96-98
451.2	600	900
ERL-Corvalhs
FACE22

-------
Terrestrial Biosphere Interaction*
Global Terrestrial Model Validation
Problem/Goal
Global models of biospheric change and stabil-
ity are being constructed and compared to cur-
rent biospheric properties to define their valid-
ity. Models which can accurately mimic today's
landscapes are expected to be reliable predictors
of future landscape and regional characteristics.
Yet, many of the predictions diverge from one
another when the several models of a given kind
are constrained by conditions of a "different
world" of the future. Because no more than one
of a set of comparable predictions can be correct,
we must assume that at least the remaining
predictions (and perhaps all predictions) are
incorrect One means to determine the degree of
inaccuracy of separate terrestrial biosphere mod-
els is to compare model behavior under differ-
ent-world conditions in which real properties of
the diffcrent-uorld can be documented, thai is,
in paleoecological data of the past several thou-
sand years. This research program isdesigned to
generate the paleoecological data needed in
model testing through such hindcasts and to test
the tcrrcsirial biosphere models being devel-
oped tn EPA.
Background
Earth systems models, global integrated assess-
ment models, and biosphere models, among
others, are composed of multiple components
each designed to replicate the dynamics of a
specific process. Frequently, the means for
testing the accuracy and validity of the indi-
vidual components is obvious; the means for
testing the accuracy and validity of the global
model as a whole is not frequently obvious
because there is only one world, the world from
which the models were constructed. Like the
castles and water projects drawn by the Dutch
artist, M. Escher, global models can be built in
which the individual model components are
correctly related to one another but the model as
a whole is fundamentally flawed. Such flaws
may be the source of differences in model be-
havior already detected in ad hoc comparisons
of biome geography, leaf area distributions and
net primary or ecosystem production. Hence,
wholesystem tests are required.
Most models in fact are tested on the whole
system, by comparison to modem field data
which has been excluded during the model de-
sign. For example, Neilson proposed develop-
ing his MAPSS leaf area model on vegetation
data from the western hemisphere, and testing
the model on excluded vegetation in the eastern
hemisphere. Prentice and others constructed
their BIOME vegetation geography model from
individual, geography-free plant physiological
thresholds to vegetation classes documented by
Olson and others. Both the MAPSS and BIOME
models generate comparably-accurate vegeta-
tion geography when compared to today's veg-
etation but each gives significantly different
predictions of vegetation geography when con-
strained by the same future climate scenarios.
The only appropriate source of whole-globe data
describing conditions different from those of
today are in the historic and prehistoric past. The
Dutch research team developing the 1MAGE2
global integrated assessme nt model, for example,
is assembling data setsonhistoricglobal changes
of the past 100 years for comparison to simu-
lated changes during that period. However, the
past 100 years contains a very small amount of
environmental variance compared to that ex-
pected in the natural biosphere under, e.g.,
doubled greenhouse gas (GHG) concentrations.
The period from the Last Glacial Maximum
(LGM) until the present, in contrast, contains at
least the magnitude of climate change expected
under doubled GHG concentrations, including
an increase in CXfo itself. The climate change
since the LGM, although an order of magnitude
slower than expected from GHG-induced cli-
mate change, included some 5 to 9°C of warm-
ing with a considerable intensification of the
hydrological cycle. The warmth of the mid-
Postglacial warm period (Hypsithermal in North

ERL-Corvalhs

-------
Terrestrial Biosphere Interactions
America: Climatic Optimum in Europe) was
some 1 to 2°C above cunent global tempera-
tures. This research project will focus on the
period of about 21,000 years since the LGM.
Scientific Objectives
The validation of global terrestrial biosphere
models with Quaternary (ice age) paleoecologi-
cal data will concentrate on documenting veg-
etation reconstructed primarily from the ubiqui-
tous fossil pollen evidence. Data from the more
irregularly-available plant macrofossil evidence
will be used only for specific research tasks,
such as identifying local plant genera and spe-
cies represented among otherwise indistinguish-
able pollen. The "proxy" data needs this project
will fill include assessing the validity of static
vegetation geography models, particularly in
their ability to reproduce the ecosystem geogra-
phy, taxonomic composition, and density of
global LGM and Hypsithermal landscapes. The
proxy data will also be assembled to test EPA
transient vegetation models, especially in their
ability to replicate the temporal sequences of
species migration during the past 10,000 years
documented in the Michigan peninsulas, in
Northwest Europe, and elsewhere.
It is important to note that there are no direct
measurements of the ecosystem composition,
leaf area density or productivity data which
models replicate. Instead, (here are only tempo-
ral sequences of e.g., fossil pollen accumulation
and/or composition, combined with empirical
relationships developed on the modern land-
scape between fossil pollen and the biotic vari-
ables. Hence, a considerable portion of the
research effort will be aimed at evaluating the
accuracy of the empirical relation ships on which
the proxy data are based.
Approach
The EPA will use a combination of in-house data
compilations, cooperative agreements with uni-
versities, and Interagency Agreements (NOAA
& USGS) to assemble globallycomprehensive
vegetation reconstructions which are compared
to specific outputs of static and transient models
cited above. Much of the work will comprise
close communication with other groups which
have similar goals. The data compilations car-
ried out by EPA and those sponsored by EPA
among other researchers will be carefully coor-
dinated with regional efforts now underway,
including those coordinated for North America
by Overpeck at NOAA, Boulder, Colorado, the
European and African PMIPS Project of Guoit
and others, and the more broadly-based research
sponsored by the PAGES Project of IGBP, coor-
dinated by Jack Eddy in Geneva, Switzerland.
M^Jor Products
The project will initially generate globally-com-
prehensive maps of vegetation properties at the
LGM, the Hypsithermal, and locally concen-
trated maps for the Holocene based on the larg-
est and most available literature data sets. Suc-
cessive iterations will be increasingly detailed in
terms of vegetation variables generated, and of
geographic specificity.
FY97 Initial map of ecosystem geography
for the globe at 21,000 and 6,000
years before present, and initial tests
of then-current static vegetation
models.
FY97 Application of local species mapped
abundances in Michigan Peninsulas,
and Northwest Europe, to testing tran-
sient models of vegetation.
FY98 Second iteration map of ecosystem
geography, leaf-area and net primary
production variables for the globe at
21,000 and 6,000 years before
present.
Annual Budget ($ K)
FY96-?8
500
ERL-Corvillis
PACE 24

-------
Tenesmi! Biosphere Interaction*
Assessment of Forest Ecosystem Management:
Effects on National and Global Carbon Dynamics
Problem/Goal
Forest systems can sequester carbon at signifi-
cant rates while providing needed goods and
services, consequently there is a growing con-
sensus in the international community that the
influence of various forest management prac-
tices on carbon sequestration should be deter-
mined. Under the Framework Convention on
Climate change, the U.S. is committed to stabi-
lizing CC>2 emissions at the 1990 level by the
year 2000. Furthermore, as a signatory, the U.S.
is committed to developing a national inventory
of greenhouse gases within 6 months of imple-
mentation of the Convention. Terrestrial eco-
systems exchange large quantities of carbon
with the atmosphere each year, they are impor-
tant components of inventories and stabilization
plans for several countries, including the U.S.
Because forest systems can sequester carbon at
significant rates, research is needed to quantify
the availability of land for more intensive man-
agement practices in both developed and devel-
oping nations.
Background
Relatively few national technical assessments
have been completed for forest management
impacts on COi emissions. This project will
conduct research specific to managed forest and
agroforest systems as input to the processes of
negotiating international treaties on climate
change and forestry. A number of preliminary
analyses indicate a promising potential for man-
aged forest and agroforest systems to offset
atmospheric CO? increases on a global scale. In
anticipation of U.S. commitments under the
Framework Convention, the project entered into
research on several pertinent tasks.
Science Objectives
The Forest Systems Project seeks to assess for-
est and agroforest management practices and
technologies for their effects on global carbon
dynamics.
1.	Assess the potential of forest manage-
ment practices to influence carbon dy-
namics in the boreal, temperate, and
tropical forest regions of the world.
2.	Assess land use trends, carbon dynamics
and forest management options for na-
tions, including the former Russia, Bra-
zil and Mexico to inventory CO2 emis-
sions, particularly from changes in for-
est ecosystem land-use.
Approach
Objective 1:
Published data from a number of sources and
countries have been compiled to develop an
extensive global database on forest carbon se-
questration and conservation. A consistent ap-
proach for relating forest land use to national
carbon budgets for four large forest nations in
boreal, tropical and temperate regions is under
development. A comprehensive economic analy-
sis of carbon storage opportunities through im-
proved forest management is being completed.
A framework for conducting a risk assessment
approach to forest adaptation to climate change
will be completed. Remote sensing/CIS analy-
ses of land availability for carbon storage in
Latin America under various climate and policy
scenarios is underway.
Pigc2S
ERL-Corvallis

-------
Tenesinal Biosphere Interactions
Objective 2:
Past and present land cover is stratified using
remote sensing technology, carbon density for
each land cover class is characterized using field
research and existing data. Carbon dynamics in
Brazil, Russia, U.S. and Mexico are being as-
sessed. GIS is being used to develop
spatiallydistributed models to estimate the car-
bon budgets during the past twenty years. A
field-based carbon budget for southern Mexico
will be completed and used as the basis for
recommending forest management policy op-
tions to impact greenhouse gas accumulation.
Major Products
FY94 The contribution of forest land use to
national carbon flux: case studies in
the United States, Russia, Brazil and
Mexico
FY95 Global analysis of the feasibility to
conserve and sequester carbon in
boreal, temperate, and tropical forest
systems- impact of land-use patterns
Annual Budget ($ K)
FY94	FY95
600	200
ERL-Corvsllii
PAGE 26

-------
Terrestrial Biosphere Interaciioni
Global Integrated Assessment Models
(GIAM)
Problems/Goals
The overall goal is to create a relatively simple
model of the complete global system in order to
assess effects of economic and ecological deci-
sion-making on atmospheric greenhouse gas
(CHG) concentrations. Of necessity, the model
must be simple enough to allow anomalous or
surprising results to be traced to their origins
within the model, for evaluation of their impor-
tance. Yet the model must be comprehensive
enough to allow assessing the impact of GHGs
of changes in trade policy, carbon tax schemes,
and similar actions which work through linkages
among the human and natural dimensions of the
global environmental system. At the very least,
a modeling framework is needed in which to
evaluate the many policies being considered in
the National Action Plan.
The fundamental problem is to predict the
changes in the concentrations of greenhouse
gases in the atmosphere in the future from knowl-
edge of dynamics in, and relationships among
•	Atmospheric chemistry and physics,
•	Ocean biology, chemistry and physics,
•	Terrestrial ecology, biophysics and bio-
geochemistry,
•	Human population dynamics and
demography,
•	Resource use, supply and demand in
industry, energy, forestry and agricul-
ture, and
•	World trade in these resources.
The goal is to assess the impact on this world
system generated by various proposed adaptive
and mitigative actions, and by no action at all.
Background
The Presidential Science advisor, John Gibbons,
has requested all federal agencies doing global
change research to examine the means to con-
duct global integrated assessments that would
include estimates of future changes in climate
and atmospheric chemistry, the roles of chang-
ing human populations, land use, energy and
resource use, and so on. Initial work has begun
on such global integrated assessment models,
and at least one (the Dutch IMAGE 2.0 Model)
is currently available and running on SPARC
Workstations. Some models which also include
all the global social and natural systems are just
being initiated, such as the GCAM Model of
Battelle Pacific Northwest Laboratories. Others
concentrate primarily on the social systems (ES-
CAPE Model, University of East Anglia; PAGE
Model, Carnegie-Mellon; DICE Model, Massa-
chusetts Institute of Technology), while natural
scientists are developing much more detailed
global biosphere models, but in the absence of
modeled control by the human systems (e.g.,
ESM, EPA-Athens; BATS, University of Ari-
zona, Tucson; EVE/GENESIS, National Center
for Atmospheric Research, Boulder).
The most comprehensive of the available inte-
grated models so far includes an atmosphere
which responds to GHGs in terms of cl imate and
chemistry; oceans which take up and release
carbon via surface circulation, deepwater for-
mation and surface temperature (climate); in-
dustrial and energy-use GHG sources related to
production output and demand which are framed
in terms of changing population and resource
utilization by differing levels of technology; and
a biosphere in which geographically- realistic
land use and vegetation store and release carbon,
as determined by climate and agricultural de-
mand, the latter a response to population change
and the availability of arable land.
Scientific Objectives
The IMAGE2 model and in the future, the GCAM
model, explicitly simulate the quantitative cou-
pling among human activities and natural pro-
cesses which normally are not considered by
modelers. The result is a model which can be
Page 27
ERL-Corvillit

-------
Terrestrial Biosphere Interaction!
interrogated by changing the emissions of spe-
cific industrial sectors, by planting trees in geo-
graph ical I y realistic local ions, by reducing popu-
lation in specified countries, etc. Either model
will be appropriate for assessing the ecosystem
and GHG implications of human activities, es-
pecially of international programs to mitigate
and adapt to environmental change.
Our objectives include inserting within one of
these global integrated assessment models, a
much more detailed version of North America,
specifically aimed at defining the role of North
American energy and resource use in global
GHG concentrations, and the role of the rest of
the world in North American climate, atmo-
spheric chemistry and economy. In addition, we
will add global change processes operating in
North America which are likely to be important
in the future, but which are not now present in
global-scale models in which the lowest com-
mon denominator in data quality determine the
quality for the whole model. These U.S. or
North American models will allow assessment
of the impact on global GHG concentrations by
proposed mitigation policies to be taken by the
U.S., and of impacts on the U.S. of international
environmental and economic policies. The im-
plications of specific policies being suggested
for inclusion in the National Action Plan will be
relatively easy to assess with the "embedded"
North American/Global IMAGE or GCAM
model.
Approach
The North American component of a G1AM
version would take its annual atmospheric and
oceanic circulation estimates and CO? exchanges
from a GIAM (IMAGE 2.0, GCAM, etc.), but
would calculate new North American values for
annual changes in vegetation, land use, popula-
tion, energy use, industrial activity, and so on. It
would also include the processes currently under
study in our transient vegetation modeling work,
such as the lagged responses of ecosystems to
rapid climate change, such as shifting mortality
and growth by trees, migration of species as a
function of land use and terrain, and so on. The
natural systems would be modeled by in bouse
personnel and the social system models would
be developed by external personnel through
competitive RFPs.
Mqjor Products
FY97 North American assessment model
developed and hardwired into a
GIAM.
FY98 An assessment of effects of chang-
ing human population and resource
use upon atmospheric chemistry, cli-
mate and the biosphere.
FY99 An evaluation of proposed mitiga-
tion and adaptation strategies in the
U.S. to reduce global GHGs and to
simultaneously increase U.S. pro-
ductivity.
Annual Budget ($ K)
mi	FY96-98
400	400
ERL-Corvallis
face 28

-------
Tenejinal Biosphere Interactions
Integrated Risk Assessment for
Rice Cropping System
Problem/Goal
A changing global climate will put the produc-
tivity and sustainability of major crops at risk.
To define the extent and intensity of that risk,
integrated assessments are needed for key inter*
national crops so that strategies can be devel-
oped to minimize losses. Rice is the most
important crop for direct human consumption,
therefore, the integrated assessment will focus
on rice.
Background
Global climate change due to increases in green-
house gas emissions will have substantial im-
pacts on terrestrial ecosystems. These changes
may also impact the ability of ecosystems to
further modify atmospheric concentrations of
these gases. Impacts of climate change on crop-
ping systems can have dramatic costs in terms of
human food supplies with economic and social
impacts. The impacts will be more severe in
lesser developing countries undergoing rapid
changes in urbanization and increased intensity
of agriculture.
Recent research has focussed on the direct im-
pacts of climate change on potential yield from
crops. However, more critical and realistic
impacts of climate change will occur on com-
plex agricultural systems including several crops
as wel I as insects, diseases, and weeds which can
substantially reduce yields.
Rice is the most important food crop in the
world, yet our knowledge of the response of the
rice plant to these environmental changes is
quite meager. Furthermore, the net impact of the
changes will depend on a balance between po-
tential increases in yields with increased CO2
for a C3 species, such as rice, vs. unpredictable
effects of temperature and precipitation changes.
For example, crop areas could expand north-
ward with increased temperatures, but suffer
from high temperatures in other areas.
In addition, rice is unique among crops in that it
is a major contributor to global emissions of
methane, a critical greenhouse gas. Thus, any
efforts to assess the risk to rice from climate
change must also consider the impacts of cli-
mate change and mitigation strategies on meth-
ane emissions from rice fields.
Research is being conducted at the International
Rice Research Institute and collaborating groups
in Asia, the United States, and Europe, on the
effects of global climate change on the irrigated
rice cropping system. Outputs from that re-
search include experimental and modeling data
on the effects of increases in CO2 and tempera-
ture on rice, diseases, insects, weeds; and on
methane emissions from rice fields. The re-
search includes modeling of the impacts of cli-
mate change on potential rice yield. However,
current research but does not include modeling
of interactions among rice and other cropping
system components which affect yield and meth-
ane emissions. Therefore an integrated assess-
ment of the effects of climate change on the rice
cropping system is needed to determine effec-
tive strategies to maintain and sustain rice yields
while reducing methane emissions from rice
fields.
Science Objectives
There will be 4 foci for specific research in this
area:
1.	to characterize current and predicted lev-
els of CO3 .temperature, and precipita-
tion in critical rice growing areas;
2.	to develop a cropping system model to
integrate responses of rice plants and
other components of the rice system to
climate change and accompanying land
use change;
Page 29
ERL-Corvillu

-------
Terrestrial Biosphere Interactions
3.	to assess the risk to rice ecosystems from
climate change through the use of simu-
lation models and a geographic informa-
tion system (CIS); and
4.	to assess the mitigation and adaptation
options for the rice cropping system in
response to climate changes.
Approach
This assessment wit! be carried out according to
the Ecological Risk Assessment Framework pro-
posed by the EPA's Risk Assessment Forum.
The study will fall under four areas:
PROBLEM DEFINITION/SCOPING - A work-
shop will identify a conceptual model for the
irrigated rice cropping system and evaluate avail-
able information regarding its key components.
A research program will be developed which
will be integrated with, and a core project of
Activity 3.1 "Effects of Global Change on Key
Agronomic Species" of the Global Change and
Terrestrial Ecosystems (GCTE) program of the
International Geosphcre-Biosphere Programme
(1GBP).
CHARACTERIZATION OF STRESS - Gen-
eral Circulation Model (GCM) outputs for key
rice producing areas of Asia will be evaluated
assuming different atmospheric CO2concentra-
tions. The GCM projections will be used with
historic and current climate data to produce
estimates of temperature, precipitation, cloudi-
ness, and wind patterns for key agroclimatic
areas. These outputs will be inputs for regional
analysis during risk characterization.
CHARACTERIZATION OF ECOSYSTEM
EFFECTS • Data from controlled experiments
(field, controlled environment, laboratory) will
be evaluated to determine the range and inten-
sity of response of key species is the irrigated
rice cropping system. Crop yield, methane
emissions, and insect, disease, and weed popula-
tions will be specific endpoints, but other param-
eters such as system carbon and nitrogen bal-
ances will also be studied.
RISK CHARACTERIZATION - Risk charac-
terization using the outputs from stress and eco-
logical effects characterization. Rice yields and
methane emissions will be calculated for irri-
gated rice producing areas of Asia using a rice
cropping system model. The outputs from the
cropping system model will serve as inputs to
GIS spatial databases. These databases will then
be used to produce rice yield change estimates,
methane emission, and/or other parameters on a
regional basis. Tlie risk characterization will
include an evaluation of the economic costs of
different strategies to maintain rice yields while
reducing methane emissions. It will also include
an analysis of potential shifts in land use re-
quired to maintain rice production with climate
change, and the impacts of those shifts on
unmanaged systems. Ultimately, assessments
based on these estimates will provide options for
policy makers and rice producers when they are
called upon to make recommendations to miti-
gate the effects of global climate changes on
rice.
M^jor Products
FY96 Cropping System Model to de-
termine impacts of climate
change on rice yields and meth-
ane emissions.
FY97 Assessment of impacts of cli-
mate change on rice yields and
methane emissions.
FY98 Assessmentofimpactsofglobal
climate change and land use on
Asian rice production and meth-
ane emissions.
FY99 Recommendations for maintain-
ing rice yields and reducing meth-
ane.
Budge! ($K)
FY94	FY95	FY96-98
175.8	200	400
ERL-Corvillii
PACE 30

-------