PROCEEDINGS
1996/1997 STAR Grants
Ecological Assessment/
Ecosystem Indicators
Program Review
-------
Direct inquiries to:
Barbara M. Levinson
Program Manager
U.S. Environmental Protection Agency
National Center for Environmental Research and Quality Assurance (8723R)
401 M Street, SW
Washington, DC 20460
(202) 564-6911
levinson.barbara@epamail.epa.gov
-------
PROCEEDINGS
1996/1997 STAR Grants
Ecological Assessment/Ecosystem Indicators
Program Review
February 3-5,1998
Las Vegas, Nevada
Environmental Protection Agency
-------
Table of Contents
Introduction v
Section 1. Projects Initiated With Fiscal Year 1997 Support
Foliar Chemistry as an Indicator of Forest Ecosystem Status, Primary Production, and Stream
Water Chemistry 1
John Aber, Mary Martin, Charles Driscoll, Richard Hallett
Environmental Factors That Influence Amphibian Community Structure and Health as Indicators
of Ecosystem Integrity 3
Val Beasley, Luanda Johnson, Carl Richards, Rebecca Cole, Pat Schoff, Joseph Murphy,
Jack Cochran, Robert Murnane
Towards a Regional Index of Biological Integrity: The Example of Forested Riparian Ecosystems 4
Robert P. Brooks, Timothy J. O'Cornell, Denice H. Wardrop, Laura E. Jackson
Assessment of Forest Disturbance in the Mid-Atlantic Region: A Multiscale Linkage Between Terrestrial
and Aquatic Ecosystems 5
Keith N. Eshleman, Robert H. Gardner, Louis F. Pitelka, Steven W. Seagle, James N. Galloway,
James R Webb, Alan T. Herlihy
Microbial Indicators of Biological Integrity and Nutrient Stress for Aquatic Ecosystems 7
James P. Grover, Thomas H. Chrzanowski
Foraminifera as Ecosystem Indicators: Phase 1—A Marine Benthic Perturbation Index;
Phase 2—Bioassay Protocols 9
Pamela Hallock Muller
Development and Evaluation of Multiscale Mechanistic Indicators for Assessing the Integrity of Regional
Landscapes 11
Carl Richards, Lucinda B. Johnson, George E. Host
Development and Evaluation of Ecosystem Indicators for Urbanizing Midwestern Watersheds 13
Anne Spade, Jonathan M. Harbor, Midhat Hondzo, Bernard A. Engel
Characterization of Ecological Integrity of Commercially Grazed Rangelands Using Remote Sensing-Based
Ecological Indicators 14
Neil E. West, R. Douglas Ramsey
Section 2. Projects Initiated With Fiscal Year 1996 Support
Health Indicators for Salt Marsh Estuaries of the South Atlantic Bight 15
James J. Alberts, Ronald T. Kneib, Steven Y. Newell, Steven C. Pennings
Trophic Transfer of Atmospheric and Sedimentary Contaminants Into the Great Lakes Fisheries: Controls on
Ecosystem Scale Response Times 16
Joel E. Baker, Jeffery D. Jeremiason, Nathaniel E. Ostrom
Multiscaled Assessment Methods: Prototype Development Within the Interior Columbia River Basin 18
Patrick Bourgeron, Hope Humphries, Bruce Milne, Frank Davis, N. LeRoy Pqff
Multiscale Assessment of the Population Status of Tfialassia Testiidinum (Turtle Grass): A New Approach to
Ecosystem Assessment -. .-. 20
Paul R. Carlson, Michael J. Durako, James W. Fourqurean, Cynthia A. Moncrieff, Gil McRae,
Jan H. Landsberg
iii
-------
Table of Contents (continued)
Modeling Spatial and Temporal Dynamics of Montane Meadows and Biodiversity in the Greater
Yellowstone Ecosystem 22
Diane Debinski, Mark Jakubauskas, Kelly Kindscher
Modeling and Multiobjective Risk Decision Tools for Assessment and Management of Great
Lakes Ecosystems 24
Benjamin F. Hobbs, Joseph F. Koonce, Ana B. Locci
Use of Multiscale Biophysical Models for Ecological Assessment: Applications in the Southeastern
United States 26
Michael Huston
Monitoring and Restoring Hydropatterns in South Florida Ecosystems 28
Eric S. Kasischke, Edwin Romanowicz, Laura L. Bourgeau-Chavez, Curtis J. Richardson,
JeffMichalek, Kevin P. Smith
Development of Environmental Assessment, Mitigation, and Restoration Techniques for Coral Reefs 30
Robert H. Richmond
Assessment and Analysis of Ecosystem Stressors Across Scales Using Remotely Sensed Imagery: Reducing
Uncertainty in Managing the Colorado Plateau Ecosystem 32
Stephanie J. Weigel
IV
-------
Introduction
The mission of the United States Environmental Protection Agency (EPA) is to protect public health and
to safequard and improve the natural environment—air, water, and land upon which life depends. Achievement of
this mission requires the application of sound science to the assessment of environmental problems and to the
evaluation of possible solutions. The National Center for Environmental Research and Quality Assurance (NCERQA)
at EPA is committed to providing the best products in high-priority areas of scientific research through significant
support for long-term research.
One high-priority research program identified in the Office of Research and Development's (ORD)
Ecological Research Strategy is monitoring research. The monitoring research is focused on biological indicator
development at the molecular, community, and landscape levels of biological organization. These indicators will
be used for the monitoring of ecosystem condition as well as exposure evaluation. The development of new
characterization methods and technologies and improvement of multiscale monitoring designs are also high-priority
research components. This research represents the extramural component of ORD's Environmental Monitoring and
Assessment Program (EMAP).
In support of the need for monitoring research, NCERQA issued a 1996 Request for Applications (RFA)
on Ecological Assessment—Regional Ecosystem Protection and Restoration. The purpose of the solicitation was to
request proposals that led to the scientific understanding and techniques required for effective ecological risk
assessment and ecosystem protection at a regional ecosystem scale. The 1996 competition resulted in the receipt of
132 applications; 20 passed scientific peer review, and 15 were funded.
In 1997, NCERQA issued another RFA on Ecosystem Indicators. The purpose of this solicitation was to
support research that led to the development of techniques and indicators that characterize and quantify the integrity
and sustainability of ecosystems at local, regional, national, and/or global scales. The 1997 competition resulted in
the receipt of 91 applications; 13 passed scientific peer review, and 9 were funded.
Annual program reviews such as this one will allow investigators to interact with one another and to discuss
progress and findings with EPA and other interested parties. If you have any questions regarding the program,
please contact the program manager, Barbara Levinson, at 202-564-6911 or levinson.barbara@epamail.epa.gov.
-------
Section 1.
Projects Initiated With Fiscal Year 1997 Support
-------
Foliar Chemistry as an Indicator of Forest Ecosystem
Status, Primary Production, and Stream Water Chemistry
John Aber and Mary Martin
Complex Systems Research Center, University of New Hampshire, Durham, NH
Charles Driscoll
Department of Civil and Environmental Engineering, Syracuse University, Syracuse, NY
Richard Hallett
Northeastern Forest Experiment Station, USDA Forest Service, Durham, NH
The New York-New England region is the most
densely forested region in the United States. In addition,
it is one of the most densely populated, and air pollution
levels are among the highest in the country. More than
30 million people in the region require potable water,
forest products, recreation, and aesthetic renewal from
a complex patchwork of forested lands with a rich and
diverse history of human use and ownership. The sus-
tainable use of forest resources and the maintenance of
surface water quality in forested regions depends on the
integrity of biogeochemical cycles within forest eco-
systems.
Development of a successful system for mon-
itoring forest and stream biogeochemistry requires two
components: (1) a simple, integrative indicator of
current biogeochemical status and (2) a method for pre-
dicting the influence of short-term climatic and
hydrologic changes on this indicator. We hypothesize
that forest productivity, soil water chemistry, and foliar
chemistry at the whole-stand (not individual tree) level
are all tightly linked to the biogeochemical status of a
forest ecosystem (see Figure 1). We further hypothe-
size that the concentration of cations in forest canopies
will be measurable by high spectral resolution remote
sensing, as has been demonstrated for nitrogen and
lignin. Watershed-level stream chemistry, reflecting
soil water chemistry, also will be predictable from
watershed-level values of canopy chemistry derived by
remote sensing.
The White Mountain region of New Hampshire
will be used as the primary study site. At the intensive
plot scale, a long-term sampling program will be used
and augmented at the Hubbard Brook Experimental
Forest in New Hampshire, and long-term experimental
treatments will be used and augmented at the Harvard
Forest in Massachusetts to examine and attempt to
predict interannual variations in foliar and soil water
chemistry as well as woody and foliar production.
Regional subsampling scale measurements will be made
of canopy, soil and soil water chemistry, as well as
forest productivity hi a series of existing experimental
and monitoring research sites. Work at the spatially
continuous monitoring scale across the White Mountain
region will include the development of algorithms for
the prediction of canopy cation concentrations using data
from NASA's Airborne Visible-Infrared Imaging Spec-
trometer (AVIRIS).
Monitoring the biogeochemical status of forest
and stream ecosystems is a key component of assessing
environmental quality in the Northeastern United States.
Any monitoring system requiring spatially continuous
capabilities will need to use some form of remote
sensing. Forest canopies are the only portion of the sys-
tem accessible to optical reflectance remote sensing
instruments, and so they offer the most likely target sur-
face for monitoring forest health in this spatial mode.
If successful, the research proposed here will establish
the linkages between foliar chemistry and processes
controlling forest growth and element loss as well as the
methods by which remote sensing can be used to predict
canopy chemistry. This would then establish the scien-
tific basis for developing a satellite- or aircraft-based
remote sensing program for monitoring forest health and
stream water quality.
-------
Biogeochemical
Status
Site
Net Primary
Productivity
Foliar
Chemistry
Soil Water
Chemistry
Regional subsample
—• Scale •——-
Visible/IR
Reflectance
Stream Chemistry
Continuous
Figure 1. Hypothesized linkages between forest biogeochemical status and variables to be measured at three spatial scales.
-------
Environmental Factors That Influence Amphibian
Community Structure and Health as Indicators of Ecosystem Integrity
V. Beasley, L. Johnson, C. Richards, R. Cole, P. Schoff, J. Murphy, J. Cochran, and
R. Murnane
University of Illinois at Urbana-Champaign, Urbana, IL
Amphibian community structure and health may
warn of human health hazards and ecological dys-
function because of the following: (1) Amphibian fertili-
zation and development occur in water; therefore,
terrestrial juveniles can be examined and collected for
detailed study. (2) Dispersal of young results in expo-
sure to synthetic chemicals and landscape alterations.
(3) Damage to plants may severely impact amphibian
communities. Early life stages require dissolved oxygen
from plants; tadpoles feed heavily on algae; and all life
stages rely on plant cover. (4) Skin is highly permeable
and serves as a major organ of respiration and water
uptake. Skin lesions due to chemicals or pathogens may
disrupt gas, acid-base, and hydration status. (5) Juve-
nile and adult amphibians feed on invertebrates,
including insect vectors of disease. Substances toxic to
insects may harm amphibians directly or deplete their
prey. (6) In the summer, tadpole and juvenile numbers
may account for much of vertebrate biomass, serving as
essential food when reproductive demands of aquatic
predators mandate high nutrient intake. (7) All verteb-
rates share similarities in metabolism and elimination of
environmental pollutants. (8) All vertebrates share simi-
lar neuroendocrine systems and developmental pro-
cesses.
This project will examine pond sites from
Minnesota through Wisconsin to Illinois to determine
the relative influence of large scale and local factors on
amphibian community structure and health. This area
encompasses a gradient of land uses, from partially
forested areas to intensively cultivated regions. Data on
land cover, fragmentation, and connectivity of landscape
elements, position and density of roads, and water
bodies will be derived from satellite images of water-
sheds and aerial photographs of pond catchments.
Locally, the wetland ecosystem structure and function
will be assessed. Habitat parameters, water quality,
contaminants, and aquatic biota, including amphibian
communities as well as organisms that impact amphi-
bians (i.e., algae, macrophytes, snails, other inverte-
brates, and fishes), will be quantified. In addition, an
assessment will be made of amphibian health, including
malformation data, histologic lesions, and indicators of
parasitism.
Relating these findings, this project will deter-
mine if: (1) Wetland ecosystem structure and function
are correlated with amphibian diversity and community
structure. (2) Agricultural land uses are correlated with
a higher prevalence of malformations and lesions in
amphibians as well as negatively impact amphibian com-
munity structure. (3) Habitat fragmentation adversely
affects amphibian community structure and health. (4)
Aquatic herbivores and aquatic contamination by
herbicides are correlated with amphibian parasitism. (5)
Amphibian abundance is inversely related to kidney
parasitism. (6) The prevalence of deformed limbs is
correlated with encysted parasites in limb bud areas. (7)
Limb and ocular abnormalities, intersex gonads, and
changes in ratios of males to females are correlated with
waterborne contaminants.
-------
Towards a Regional Index of Biological
Integrity: The Example of Forested Riparian Ecosystems
Robert P. Brooks, Timothy J. O'Connell, and Denies H. Wardrop
Penn State Cooperative Wetlands Center, Forest Resources Laboratory, Pennsylvania State University, University
Park, PA
Laura E. Jackson
U.S. Environmental Protection Agency, Office of Research and Development, Research Triangle Park, NC
The premise of this project is that measures of
ecological indicators and habitat conditions will vary
between reference standard sites and reference sites that
are impacted, and that these measures can be applied
consistently across a regional gradient in the form of a
Regional Index of Biological Integrity (RIBI). Six prin-
ciples are being proposed to guide the development of
any RIBI: (1) biological communities with high integrity
are the desired endpoints; (2) indicators can have a bio-
logical, physical, or chemical basis; (3) indicators
should be tied to specific stressors that can be realisti-
cally managed; (4) linkages across geographic scales
and ecosystems should be provided; (5) reference stan-
dards should be used to define target conditions; and
(6) assessment protocols should be efficiently and rapid-
ly applied. Four integrative bioindicators can be
combined to develop an RIBI for forest riparian eco-
systems in the mid-Atlantic states: (1) macroinvertebrate
communities, (2) amphibian communities, (3) avian
communities, and (4) avian productivity, primarily for
the Louisiana waterthrush (Seirius motadlld). This
species depends on stream macroinvertebrates for food
and forest riparian habitats for nesting. As a common
top predator and the only obligate avian species of this
ecosystem in the Eastern United States, it is an ideal
calibrator for an index of headwater ecosystems.
Measuring the population parameters of the
Louisiana waterthrush requires a substantial investment,
but once completed, provides a means to calibrate the
other indicators, thus linking them across scales. Each
bioindicator is most strongly associated with measures
of habitat at a particular scale. Measuring productivity
for the Louisiana waterthrush relates primarily to the
quality of riparian habitat, but it is also dependent on the
availability of macroinvertebrates as food. Biomass and
composition of macroinvertebrate and amphibian com-
munities relate to instream and wetland habitat and
measures of water chemistry and sedimentation. Avian
communities relate primarily to landscape metrics.
However, by combining measures of nest productivity,
territory density, and survey abundance, attributes of the
Louisiana waterthrush span the widest range of scale.
Once tested between reference and impacted sites and
then calibrated across scales, a set of indicators could be
combined into an RIBI for the Mid-Atlantic Integrated
Assessment. By providing a reliable expression of
environmental stress or change, an RIBI can help
managers reach scientifically defensible decisions.
-------
Assessment of Forest Disturbance in the Mid-Atlantic Region:
A Multiscale Linkage Between Terrestrial and Aquatic Ecosystems
Keith N. Eshleman, Robert H. Gardner, Louis F. Pitelka, and Steven W. Seagle
Appalachian Laboratory, University of Maryland Center for Environmental Science, Frostburg, MD
James N. Galloway and James R. Webb
Department of Environmental Sciences, University of Virginia, Charlottesville, VA
Alan T. Herlihy
Department of Fisheries and Wildlife, Oregon State University, Corvallis, OR
The objective of this 3-year project is to develop,
test, validate, and demonstrate an analytical framework
for assessing regional-scale forest disturbance in the
mid-Atlantic region by establishing a multiscale linkage
between forest disturbance and forest nitrogen (N)
export to surface waters. Excessive nitrogen leakage
(export) from forested watersheds is a potentially useful,
integrative "indicator" of a negative change hi forest
function, which occurs in synchrony with changes in
forest structure and species composition. The research
focuses on forest disturbance associated with recent
watershed defoliations by the gypsy moth larva at spatial
scales ranging from small watersheds to the entire
region.
Three research hypotheses will be tested: (1) N
export from forested watersheds in the mid-Atlantic
region is primarily attributable to forest disturbances,
operating at both the local and regional scales; (2)
changes in forest species composition in the mid-Atlan-
tic region—such as declines in dominance of oak species
and increases in shade-tolerant species—have been
induced or exacerbated by gypsy moth defoliation; and
(3) if both N export and rapid forest succession are
largely disturbance-induced by gypsy moth defoliation
in the region, then broad-scale patterns of dissolved N
leakage, forest succession, and forest disturbance should
be spatially and temporally well-correlated.
There are five carefully linked tasks associated
with this project's objective that will allow testing of the
stated hypotheses. These are: (1) characterizing forest
composition, recent disturbance history, and annual N
export (see Figure 1) for intensively studied watersheds;
(2) modeling N export from intensively studied water-
sheds due to disturbance; (3) verifying N export as an
indicator of disturbance at subregional scales; (4)
scaling forest point data to landscape scales; and (5)
correlating spatial and temporal patterns of N export and
forest species composition changes with forest distur-
bance at the regional scale.
Tasks 1 and 2 will largely be accomplished using
conventional methods appropriate for small watershed
studies by cost-effectively taking advantage of a plethora
of data for small watersheds in the region that were
collected for other projects. In Task 2, data assembled
during completion of Task 1 will be used to parameteri-
ze an empirical model of N export from these water-
sheds due to disturbance—a unit N export response
function (UNERF) model (which is completely analo-
gous to the widely used linear unit hydrograph model in
watershed hydrology); an important input to the model
is a time series of gypsy moth defoliation data for each
watershed (obtained from statewide aerial mapping
programs). Task 3 will result in the subregional verifi-
cation of the UNERF model by comparing predictions
and observations of N export for a group of watersheds
not examined during Task 2. Task 4 will use a geo-
graphic information system (GIS), extrapolative tech-
niques of landscape ecology, remotely sensed imagery,
regional forest species composition data, and a limited
number of on-the-ground measurements to describe
forest composition and changes in forest species compo-
sition at the landscape scale. In Task 5, using spatial
patterns and statistical distributions of forest composi-
tion from Task 4 and time-varying gypsy moth defolia-
tion maps as inputs to a GIS-linked version of the
UNERF model, spatial and temporal patterns of N
export from forested lands hi the mid-Atlantic region
(and for important subregional units such as the Chesa-
peake Bay watershed) will be predicted.
The primary output from the project will be a
regional assessment of the effects of forest disturbance
in the form of gypsy moth defoliation on forest health
and water quality in the mid-Atlantic region based on
the establishment of a multiscale linkage between forests
and surface waters through a relatively simple, but
useful, modeling approach. The demonstration of a
viable approach to spatially extrapolating (and validating
the extrapolation) forest composition information to the
landscape or watershed scale must be recognized as a
major contribution of the project and one that will lay
the groundwork for future assessments of changes in
other types of terrestrial ecosystems. In addition, the
ability to derive statistical distributions of N export from
forested lands to the Chesapeake Bay represents a major
benefit to the existing Environmental Protection Agency
Chesapeake Bay Program.
-------
400
300
200
cr
CD
100
0
1988 1989 1990 1991 1992 1993 1994 1995
Water Year
White Oak Run O Paine Run
Staunton River D Upper Big Run
Piney River
Figure 1. Annual dissolved N export from five mid-Atlantic forested watersheds during water years 1988-1995. Note the synchrony of the
N export response to defoliation, first observed in the region in 1989.
-------
Microbial Indicators of Biological Integrity
and Nutrient Stress for Aquatic Ecosystems
James P. Graver and Thomas H. Chrzanowski
Biology Department, University of Texas at Arlington, Arlington, TX
Several chemical and biological variables
will be examined, which may provide a broadly
applicable approach to understanding the
biological consequences of nutrient loading in
aquatic systems and predictions of the resulting
community structure. This approach is based on
recent advances in aquatic microbial ecology and on
theory developed in the rapidly expanding field of
ecological stoichiometry. The indicators that will
be examined include: (1) seston C:N:P ratio; (2)
species-level responses of algae to nutrient
bioassays; (3) community-level responses of
bacteria to nutrient bioassays; (4) community
structure of algae; (5) community structure of
bacteria; and (6) the estimated ratio of algal- to
bacterial-specific growth rates. The general hy-
pothesis is that these indicators will reflect nu-
trient-related stresses, including eutrophication and
alterations of nutrient loading ratios (see Figure 1).
This general hypothesis is further elaborated by
testing the following seven specific hypotheses: (1)
indicators based on seston stoichiometry and nutrient
bioassays will agree; (2) limitation by a single nutrient
will be negatively associated with algal community
diversity and equitability; (3) colimitation by two
nutrients will be positively associated with algal com-
munity diversity and equitability; (4) bacterial com-
munity structure will shift in response to nutrient
limitation; (5) bacterial nutrient limitation will be
negatively associated with loss rate to micrograzers; (6)
degrees of algal and bacterial nutrient limitation will be
positively associated; and (7) algal nutrient limitation
will be associated with low algal productivity and
growth relative to bacteria.
A standard protocol will be employed
consisting of sampling standard water quality
parameters, chemical analysis of paniculate matter,
and algal and bacterial bioassays to identify nutrient
limitation and to estimate in situ growth rate. Over 3
years, this protocol will be applied in two warm
temperate reservoirs in Texas and in two cool temperate
lakes to enable inter-regional comparisons.
There is some evidence that algal growth in one
of the Texas Reservoirs is strongly nitrogen limited,
while growth in the other is colimited by nitrogen and
phosphorus in the warm growing season. One of the
Canadian lakes is an unmanipulated reference lake,
while the other is experimentally eutrophied by phos-
phorus additions, hi this second lake, recent research
demonstrated large shifts in algal composition and food
web dynamics associated with shifts of nitrogen and
phosphorus dynamics.
This project will demonstrate that the proposed
indicators sensitively reveal seasonal shifts in nutrient
limitation, interannual, interlake, and regional differ-
ences in nutrient stress and loading. The microbial
indicators proposed should have wide applicability in
nearly all aquatic habitats and are based on ecosystem
components with very rapid responses to environmental
changes. These indicators are short-term, and thus
feasible to repeat at larger temporal and spatial scales.
This study will reveal whether these short-term indi-
cators adequately reflect whole-lake and larger scale
responses to nutrient stresses.
-------
Experimental Lakes Area
Two regions, two lakes in each region.
North Texas
Microbiological analysis:
Algal abundance and composition
Bacterial abundance
Bacterial composition (Biolog)
Bacterial productivity
Protozoan grazing
Dilution bioassays:
Algae: N, P, trace nutrients
Bacteria: N, P, organic C
Depth Profiles
Temperature
f
OJ
Q
Samples for:
Seston C:N:P
Dissolved
nutrients
Particulate
nutrients
Figure 1. Microbial indicators of biological integrity and nutrient stress for aquatic systems.
-------
Foraminifera as Ecosystem Indicators:
Phase 1—A Marine Benthic Perturbation Index; Phase 2—Bioassay Protocols
Pamela Hallock Muller
Department of Marine Science, University of South Florida, St. Petersburg, FL
Foraminifera are by far the most useful group of
paleoenvironmental indicators used by geoscientists for
the following reasons: (1) their shells are important
sediment constituents; (2) they are small and widely
abundant; (3) different taxa have evolved to exploit most
environments, substrates, and nutritional modes in
marine systems; and (4) their shells morphologically and
geochemically record environmental conditions. This
project will develop techniques for routine use of
foraminifera as indicators of biological integrity in both
field and laboratory settings.
In Phase 1, an index is being developed and
tested for assessing perturbations of marine benthic
ecosystems, which is based on changes in key taxa of
foraminifera and that can be applied worldwide using
historical, sediment-core, and surface-sediment data
sets. This technique will provide an index for assessing
how benthic communities have changed under human
influence as compared with the natural variability in
preanthropogenic times or in unaffected areas. The
procedure requires minimal technology and can be
applied by technicians with modest training. The model
ranks relative abundances of key foraminiferal taxa,
morphogroups, and total abundances, generating an
index value when foraminiferal assemblages are com-
pared temporally or spatially. The model will be adap-
table to local and regional biotas and can incorporate
other environmental or taxonomic data that can be
scaled to the model.
In Phase 2, bioassay protocols will be developed
for foraminifera in laboratory studies on the effects of
key stressors in marine benthic environments. In ad-
dition, Phase 2 will use Amphistegina spp., which are
reef-dwelling foraminifera with algal endosymbionts.
Known stressors in culture include temperature shock,
salinity change, slight increases in UV-B radiation, algal
overgrowth, and chemical pollutants. Specific protocols
will be developed by using visual, cytological, and
biochemical responses to enable routine experimentation
with the protists. Besides being globally important in
their own right, these foraminifera can provide a model
calcifying symbiosis for testing stressors that threaten
the ecological integrity of coral-reef ecosystems. Pro-
tocols also will be adaptable to other cultivable fora-
minifera from other ecosystems.
For Phase 1, the preliminary iteration of the
index (see Table 1) has been tested on historical (1960)
and recent (1992) data sets from surface sediments
collected off Key Largo in the Florida Keys, with
results indicating a significant change in the benthic
community over that time. In addition, published data
sets from Florida Bay have been requested from the
U.S. Geological Survey; tests of those data sets using
the preliminary index will be presented.
Preliminary tests indicate the potential for
providing an index for quantifying change in marine
sediments that can be directly useful in interpreting
environmental change under anthropogenic influence.
This work has only just begun. For Phase 1, collection
and analysis of sediment cores from the Florida Keys is
planned for 1998. Data sets are being sought from
temperate locations, specifically Nova Scotia and San
Francisco Bay, to expand applicability. For Phase 2,
experiments will be initiated in early 1998.
-------
Local opportunists
Other agglutinates
Larger miliolines
Other miliolines
Large rotaliines
Other rotaliines
Other key taxa or morphogroups
Morphological anomalies
Taxanainic/IVrar|ihogniugRan{un
0 - Not recorded in region or no data
1 - Rare (< 1%)
2 - Uncommon (1% < < 5%)
3 - Common (5% < < 20%)
4 - Abundant (20% < < 40%)
5 - Very abundant (40% < < 60%)
6 - Dominant (> 60%)
# forams/gm of sediment
Other parameters
Ranked as log,0 values
Ranked as a 0-6 scale
Example: Inshore station near Key Largo, Florida
1. Known opportunists
2. Other agglutinates
3. Larger miliolines
4. Other miliolines
5. Large rotaliines
6. Other rotaliines
7. # forams/gm of sediment
8. Other catagories
1
1
6
3
1
2
0
0
1
2
3
5 x
1
3
3.9
0
0
1
3
2
0
1
0
0
0
0
2
1
0
0
No data
No data
MBPI = [T |AfB.1 - (IA:-B,l-ni = 7+3 = 1.7
N* 6
'CF = (|ArBi|-l) where |ArB,| > 1; CF =0 where |ArB,|= 0.
N* = Number of categories in which both A and B > 1 (Le., taxa occur in area and data were collected).
Preliminary tests indicate: MBPI < 1 No significant change in benthic community.
1 < MBPI < 2 Perceptible change in benthic community.
MBPI < 2 Dramatic change in benthic community.
Table 1. Proposed marine befflhic perturbation index.
10
-------
Development and Evaluation of Multiscale Mechanistic
Indicators for Assessing the Integrity of Regional Landscapes
Carl Richards, Lucinda B. Johnson, and George E. Host
Natural Resources Research Institute, University of Minnesota, Duluth, MN
In this project, suites of ecological indicators will
be developed that cross spatial scales, mechanistically
reflect ecosystem states and processes, are statistically
robust, and are applicable across regional landscapes.
In addition, these indicators will be based on readily
accessible information available in a real-time frame-
work (e.g., Landsat data). To accomplish this, the fol-
lowing objectives are proposed: (1) develop predictive
models that integrate landscape-scale factors with reach-
scale physical and chemical stream attributes to quan-
tify key compositional and structural attributes of stream
biota and derive ecosystem indicators at multiple spatial
scales; (2) evaluate the appropriate scale of terrestrial
and aquatic data necessary to resolve regional and local
aquatic resource questions; (3) improve the ability to
distinguish and quantify natural variation in indicators
from that derived from anthropogenic stressors; (4)
assess the extent to which regional and local-scale in-
dices (including standard indices of ecological integrity
such as the Index of Biological Integrity) reflect
fundamental ecosystem processes and structural pro-
perties of stream habitats and biota; and (5) quantify
confidence limits and evaluate the geographic trans-
ferability of regional and .local-scale indicators de-
veloped above.
To develop, evaluate, and integrate indicators
across multiple spatial scales, a multitiered sampling and
modeling strategy will be employed, integrating data
collected at regional scales via satellite imagery, local
scales via low-altitude imagery, and site scales via field
sampling (see Figure 1). Relationships among these
hierarchically nested scales will be quantified using an
integrated series of empirical and process models.
These data will be used to identify indicators at each
scale that reflect a critical ecosystem process or state
variables related to the integrity and sustainability of
those ecosystems. Indicators representing fundamental
driving variables and processes will be developed and
tested and then integrated into a system for identifying
positive or negative trends in the health of ecosystems
in regions that are heavily dominated by agriculture and
mixed land uses.
The results of this research will significantly
increase the ability to quantify features of terrestrial
ecosystems in strongly altered landscapes. Models of
processes that integrate features at different scales will be
developed. A sampling design that incorporates natural
and anthropogenic features of landscapes and streams at
regional, watershed, and local scales will be evaluated to
determine the transferability of these methods among
ecoregions. Extensive databases, previously developed
through EPA-funded research, will be used to validate the
use of statistical and methodological techniques.
By decomposing variance in ecosystem
parameters associated with factors such as land use
patterns and practices versus "natural" processes (e.g.,
geomorphology, topography), managers and policy-
makers can concentrate scarce resources on those
factors that have an influence on the ecological end-
points and that can be managed or manipulated. The
proposed research will assist risk assessors to prioritize
landscape and local features of terrestrial and aquatic
ecosystems that influence aquatic ecosystem integrity.
Finally, a rigoous analysis of the uncertainty associated
with indices at all spatial scales, including natural sto-
chasticity, measurement error, parameter error, and
model error, will be provided. These estimates of sta-
tistical confidence will greatly improve the utility of
ecological indicators for use in local and regional
assessments.
11
-------
Landscape Modeling Strategy
Watershed Characterization
Land Use/Land Cover
Landscape Pattern
Riparian Conditions
Hydrologic Submodel
Predict N, P, and
sediment loadings
Habitat Submodel
Predict physical conditions
(e.g., substrate, woody debris)
Reach-Scale Characterization
Physical Habitat
Water Quality
Ecosystem Prediction Submodel
Predict composition and structure
of fish and macroinvertebrate
communities
Predict ecosystem NPP and
organic matter retention
Ecological
Indicators
Community Characterization
Species Richness / Diversity
Fish IBI
Macroinvertebrate IBI
_ J
Figure 1. Conceptual model illustrating the organization of mechanistic factors operating at multiple spatial scales. Components summarize
the development of predictive models, ecosystem attributes, and fundamental ecosystem processes. Indicators are derived at all
spatial scales.
12
-------
Development and Evaluation of Ecosystem
Indicators for Urbanizing Midwestern Watersheds
Anne Spade, Jonathan M. Harbor, Midhat Hondzo, and Bernard A. Engel
Purdue University, West Lafayette, IN
Urban areas are spreading over much of the
Eastern United States, and yet our understanding of
their impact on ecosystem processes is very poor.
Understanding and managing the responses of these
systems to urbanization requires a multidisciplinary
effort ranging from site-specific to large-scale regional
studies. The use of ecosystem indicators allows for the
development of practical management approaches in the
absence of detailed monitoring of the complete system.
Although urbanization usually has negative impacts on
stream ecosystems, the causal relationships have not
been well studied. Simple empirical indicators may
actually be misleading if they are not based on specific
system characteristics. The focus of this project is on
the development of predictive indicators of urbanization
that are applicable to midwestern watersheds and stream
ecosystems and, more importantly, to illustrate an
objective methodology for developing and testing such
ecosystem indicators.
This project's objectives are to: (1) quantify the
impacts of urbanization on hydrologic regimes, water
quality, and habitat structure of stream ecosystems using
paired experimental watersheds, and to develop linked
models that accurately predict these impacts; (2) use the
linked models as a virtual laboratory within which to
generate and test indicators of urbanization and
hydrologic change in terms of responses of fish and
macroinvertebrate communities; and (3) use these
models and indicators in assessing the response of
stream communities to alternative urbanization scenarios
with extension to larger watersheds in the region.
This project will examine watersheds in transition
from rural to urban. The initial study site is a 216 km2
set of paired experimental watersheds encompassing two
adjacent third-order streams. The watersheds typify
drainages in the corn belt ecoregions; however, one is
urbanizing more rapidly than the other, with substantial
new housing developments planned in the next few
years. Stream flow, water quality, and aquatic biota
have been intensively monitored since 1991, allowing
the interpretation of temporal trends. A detailed
geographic information system (GIS) database is avail-
able for spatial analysis. The investigators hope to es-
tablish linked process models that address the impact of
land use change on the physical and chemical character-
istics of runoff, stream flow, water quality, and habitat
quality, and which then link changes in such stream
variables to changes in aquatic community structure. A
set of environmental indicators will be generated and
tested based on these relationships. The predictive rela-
tionships among indicators will then be evaluated for a
range of larger and more urbanized watersheds in the
region. Thus, indicator behavior will be examined over
a range of scales and physical settings under alternate
urbanization scenarios.
Water quality and flow regimes are currently being
evaluated for significant trends in the paired streams.
Significant differences in flow, chloride, nitrate, total
phosphorus, dissolved oxygen, and fish diversity have
been identified. In addition, the initial steps in the
modeling process focusing on temperature/oxygen levels
and land use classification also are under way.
This project will provide a sound basis for the use
of specific indicators as tools in regional planning of
watershed development. The risk analysis portion of the
work will provide a probabilistic measure as to whether
a potential urbanization scenario can achieve stream
water quality and biological targets.
13
-------
Characterization of Ecological Integrity of Commercially
Grazed Rangelands Using Remote Sensing-Based Ecological Indicators
Neil E. West and R. Douglas Ramsey
Utah State University, Logan, UT
Rangelands involve vast areas of the Western
United States with arid to semi-arid climates. Their low
biological productivity has made point-based monitoring
of the ecological status difficult to economically justify.
Several remote sensing (RS) based synoptic means of
characterizing changes in the integrity of lands will be
tested on one large ranch in northern Utah. Also, the
applicability of a transition threshold conceptual model
will be tested by using satellite remote sensing and
geographic information systems (GIS) to characterize
rangeland conditions and trends.
The proposed research will use 21 years of
Landsat satellite imagery; a GIS database of site biologi-
cal, physical, historical, and current ranch management
records; and multiple-time by nested multiple-scale
experimental design to establish causal links between
possible threshold response and human management
interventions to assess the ecological integrity of ecosys-
tems within a Western Intermountain Sagebrush Steppe-
dominated landscape subject to commercial livestock
and big game animal grazing.
The assessment will occur at multiple scales,
including: landscape, watershed, administrative (i.e.,
public versus private land), individual paddock, eco-
logical site, and piosphere (waterpoints). Watersheds
and sub-basins will be delineated by using digital
elevation models and GIS-based hydrological modeling
algorithms. The resulting landscape stratification of
geomorphological source, sink, and transfer zones will
be statistically related to satellite image-derived indices of
vegetation cover and composition, soil erosion, and
landscape configuration metrics and will result in an overall
measure of site ecological integrity and sustainability.
The proposed research will result in the
following: (1) new RS-based ecological indices for
assessing ecological integrity; (2) a synthesis of the new
transition threshold concept in rangeland ecology,
landscape ecology, and ecosystem science, which will
link spatio-temporal changes in ecosystem structure and
pattern to changes in rangeland condition and trend; (3)
a general methodology for using satellite RS and GIS
technologies with current ecological assessment concepts
for monitoring semi-arid landscapes at multiple spatial
scales; and (4) a validation of the utility of remote
sensing-based piosphere indices in North America
through the use of the developed GIS database.
14
-------
Section 2.
Projects Initiated With Fiscal Year 1996 Support
-------
Health Indicators for Salt Marsh
Estuaries of the South Atlantic Bight
James J. Alberts, Ronald T. Kneib, Steven Y. Newell, and Steven C. Pennings
University of Georgia Marine Institute, Sapelo Island, GA
This project examines simple, inexpensive, and
rapid methods for assaying and monitoring the general
health of salt marsh ecosystems hi the Southeastern
United States. It examines a suite of indigenous salt
marsh organisms to determine if biological processes
that occur on relatively short time scales (days-months)
can be used as measures of salt marsh health and
effectiveness of remediation strategies.
The research project involves laboratory and field
components. However, the project emphasizes field
studies. Paired sites at contaminated and control locales
are chosen for field sampling. In the first year of the
project, the sites included the LCP Superfund site in
Brunswick, Georgia, and a relatively pristine site on
Sapelo Island, Georgia. Years 2 and 3 will include
another paired site in Georgia and several sites in
Georgia and South Carolina, respectively. The basic
concept includes: (1) evaluating critical rates within the
macrophyte community, focusing on sublethal Impacts;
(2) determining the efficacy of reproductive potential
of three species of estuarine crustaceans as another
measure of sublethal stress; and (3) evaluating physio-
logical bioassays using marine microorganism in-
dicators.
Samplings at the LCP and Sapelo sites have been
conducted. Samples are in various states of processing,
analyses, data collection, and evaluation. Preliminary
results indicate that the bioassays tested on impacted
sites show no differences relative to reference sites.
However, the preliminary nature of these results must
be stressed, and it is inappropriate to draw conclusions
at this tune.
Samples from year 1 will continue to be analyzed.
Data will be combined with that already collected and
prepared for presentation and publication. Sites having
potential impacts of chromium (Charleston, South
Carolina), insecticides (known toxaphene site, Bruns-
wick, Georgia), and arsenic/creosote (Escambia Super-
fund Site, Brunswick, Georgia) are currently being
evaluated for extensive sampling during the coming
year. Site selection also will begin in South Carolina
and possibly Florida for year 3 project objectives.
15
-------
Trophic Transfer of Atmospheric and Sedimentary Contaminants Into
the Great Lakes Fisheries: Controls on Ecosystem Scale Response Times
Joel E. Baker and Jeffrey D. Jeremiason
Chesapeake Biological Laboratory, University of Maryland System, Solomons, MD
Nathaniel E. Ostrom
Department of Geological Sciences, Michigan State University, East Lansing, MI
The purpose of this study is to quantify the
absolute and relative magnitudes of polychlorinated
biphenyls (PCBs) transfers into the Great Lakes fisher-
ies from three exposure routes: (1) atmospheric deposi-
tion transferred through the pelagic food web; (2)
atmospheric deposition transferred, via rapidly settling
particles, through the benthic food web; and (3) transfer
from historically contaminated, in-place sediments
through the benthic food web. This will be accom-
plished by using stable isotopes and PCBs as tracers of
carbon and bioaccumulative contaminants, respectively,
through the water column and food web of Grand
Traverse Bay, an embayment of Lake Michigan. Each
of these three routes differ both in their efficiencies of
contaminant transfer and in their characteristic response
times. This study will result in a quantitative, process-
driven model of contaminant transfers in the Great
Lakes food webs that distinguishes between "new" (i.e.,
regional atmospheric deposition) and "in-place" (i.e.,
recycling from contaminated sediments) sources of con-
taminants that support the slowly changing contaminant
inventories in the highest trophic levels of the Great
Lakes.
This project's objective is to quantify the absolute
and relative flows of bioaccumulative organic con-
taminants through the pelagic, epibenthic, and benthic
food webs of the northern Great Lakes (see Figure 1).
Efficient scavenging of atmospheric-derived con-
taminants from surface waters delivers large chemical
fluxes seasonally to the epibenthic food web, and this
process "pumps" recent atmospheric loadings into the
Great Lakes fisheries. Specific objectives include: (1)
quantifying the fluxes of organic carbon on a seasonal
basis and associated contaminants from the surface
waters to near the sediment-water interface; (2) quan-
tifying trophic transfers of carbon and PCBs through the
pelagic, epibenthic, and benthic foodwebs, with
emphasis on the episodic deposition of particles to the
benthic environment in the spring and the relative
importance of infauna, amphipods, and mysids in
contaminant transfer to Great Lakes fish; and (3)
quantifying, through statistical analysis of contaminant
"fingerprints" and bioenergetics modeling, the relative
magnitudes of exposure to sedimentary and atmos-
pherically derived contaminants in the Great Lakes
fisheries.
To meet this project's objectives, an extensive
field sampling effort in Grand Traverse Bay was
successfully initiated and completed in 1997. Virtually
all major species were collected, characterizing the lake
trout food web on a seasonal basis. Sediment traps
were deployed, retrieved, and redeployed during 1997
to characterize seasonal PCB and carbon settling and
recycling dynamics. Vapor and dissolved PCB samples
were collected on a monthly basis to determine atmos-
pheric loadings. In addition, bottom sediment grabs and
bottom water (~ 2 meters off the bottom) dissolved, and
paniculate samples were collected. Preliminary PCB
concentrations in water, air, sculpin, and lake trout
suggest that Grand Traverse Bay will be an appropriate
surrogate for quantifying contaminant transfers into
northern Great Lakes fisheries.
16
-------
VaporQ
Dissolved '*—- Phytoplankton
Diatom
Aggregates"
Copepods^p|ankt.vores
e.g., alivite, smelt,
My s ids
Vertical
Migration^
bloater, herring
Fecal Pellets
Organic-Rich
Particles
Sculpin
Pelagic Transfer
Piscivores
Benthic Transfer
Amphipods
B2)
\
Ingestion
Sediment —— *- Infauna
Burial
Porewater
Figure 1. Food web structure and flows of bioaccumulative contaminants in the northern Great Lakes. Diffusive uptake of dissolved
contaminants by higher trophic levels was omitted for clarity.
-------
Multiscaled Assessment Methods: Prototype
Development Within the Interior Columbia River Basin
Patrick Bourgeron and Hope Humphries
Institute for Arctic and Alpine Research, University of Colorado, Boulder, CO
Bruce Milne
Department of Biology, University of New Mexico, Albuquerque, NM
Frank Davis
National Center for Ecological Analysis and Synthesis, University of California, Santa Barbara, CA
N. LeRoy Poff
Department of Biology, Colorado State University, Fort Collins, CO
This project's goal is to investigate multiscaled
relationships among biophysical variables and biological
features of terrestrial and aquatic systems of critical
value in the use of ecological assessment data. Exten-
sive abiotic and biotic databases compiled for the
Interior Columbia River Basin (ICRB) are employed to
address six primary objectives: (1) link biophysical and
biological patterns associated with terrestrial and aquatic
systems at different scales; (2) quantify the scaled
relations of linked biophysical and biological systems;
(3) develop methods for predicting broad and fine scale
patterns over areas of varying sizes; (4) classify land-
scapes at different scales based on biophysical and
biological characteristics and define probabilities of
response of the biotic components of landscapes; (5) test
the effectiveness of classifications based on indirect
variables (e.g., elevation, lithology, landforms) for pre-
dicting bioenvironments (e.g., groups of direct variables
such as climatic variables) and biological characteristics
of areas of varying sizes for evaluating alternative land
management strategies and conservation; and (6) de-
velop prototype multiscaled, representativeness assess-
ments for evaluating alternative land management
strategies using products from objectives 1-5.
To meet the above-mentioned objectives, nine
steps are followed: (1) identify fine scale patterns of
interest (biotic and abiotic) to be predicted by classifica-
tion, maps, and models; (2) associate indirect and direct
biophysical variables with each study ecological scale;
(3) sample fine scale patterns across coarser scale
biophysical environments; (4) determine the biophysical
variables that most influence finer scale features (biotic
and abiotic); (5) use numerical techniques to develop
ecological classifications and maps based on a reduced
number of indirect and direct biophysical variables; (6)
determine relationships between biotic components and
biophysical environments; (7) determine how accurately
classes developed using direct biophysical variables can
be predicted using indirect variables and vice versa; (8)
determine whether the hierarchical structure of biophysi-
cal environments constrains biotic distributions and, as
a consequence, constrains the hierarchical structure of
biotic communities; and (9) test the effectiveness of
classifications, maps, and models in assessing the repre-
sentativeness of areas for regional conservation planning.
The project started on March 1, 1997. Since
then, step 1 has been completed, and step 2 is near
completion. All existing relevant spatially referenced
databases covering the ICRB have been acquired and
processed, and new databases are being developed.
Thorton's DAYMET model generating biophysical
variables has been tested at a 30 meter scale in a sub-
basin of the Flathead National Forest, Montana. A total
of 35 variables representing geographic coordinates,
topography, climate/hydrology, geology, disturbance,
and management have been attributed to 27,000 vegeta-
tion plots for analyses of regional biotic distribution.
Preliminary work on steps 4, 5, and 6 is under way in
four areas: (1) testing the effectiveness of three multi-
scaled ecological classifications in predicting vegetation
patterns and disturbances across scales; (2) conducting
a regional gradient analysis of vegetation and a study of
species diversity patterns over the ICRB; (3) studying
the effect of sub-grid variation on estimation of diversity
patterns; and (4) development of preliminary predictive
statistical models to define regional probabilities of
environmental response for species (see Figure 1) and
community types.
18
-------
Interior Columbia River Basin
Vaccinium membranaceum
& V. globulare
Predicted Probability
of Occurrence
0.0-0.25
0.25-0.50
El 0.50-0.75
H 0.75-1.0
Line Legend
A1 State Boundaries
A/ Landscape
Characterization
Boundary
o 36 73
147
221
295 Miles
Figure 1. Interior Columbia River basin.
-------
Multiscale Assessment of the Population Status of Thalassia
Testudinum (Turtle Grass): A New Approach to Ecosystem Assessment
Paul R. Carlson, Gil McRae, and Jan H. Landsberg
Florida Department of Environmental Protection, St. Petersburg, FL
MichaelJ. Durako
University of North Carolina-Wilmington, Wilmington, NC
James W. Fourqurean
Department of Biological Sciences, Florida International University, Miami, FL
Cynthia A. Moncrieff
University of Southern Mississippi, Ocean Springs, MS
Seagrasses are vital components of the near-shore
ecosystem, providing food and shelter for part or all of
the life cycle of many economically important fish and
shellfish species. However, drastic declines in the dis-
tribution and abundance of estuarine seagrass and
submerged aquatic vegetation (SAV) communities have
occurred in many estuaries throughout the world during
the past 50 years, hi most cases, concurrent declines in
water quality have been blamed for seagrass loss.
Because seagrasses are sensitive to even slight
decreases in water clarity, they are ideal sentinel species
for biological monitoring of water quality. This pro-
ject's goal is to test the utility of demographic,
morphological, physiological, and chemical char-
acteristics of turtle grass (Thalassia testudinum) as
indicators of chronic stress. The investigators also are
trying to determine the most effective sampling design
and scale for assessing the impact of natural and human
impacts on seagrass beds. Turtle grass was selected for
three reasons: (1) It is abundant along the Gulf coast.
(2) Because it has a large investment in nonphoto-
synthetic tissue (roots and rhizomes), it is particularly
sensitive to stress. (3) In estuaries like Tampa Bay, it
has proven more vulnerable than other seagrass species
to human impacts.
In the first year, nine sites were sampled from the
Chandeleur Islands to the Florida Keys, using a hier-
archical sampling design based on tesselated hexagons.
At each site, 30 stations at each of three (small—100 m2,
medium—10,000 m2, and large—1,000,000 m2) scales
were sampled (see Figure 1). At each of the stations,
seagrass cover and community structure were assessed
visually in 0.25 m2 quadrats. Plant samples were
collected for demographic analysis by reconstructive aging
and for structural indices such as leaf width, length, shoot-
specific leaf area, and leaf area index. Rhizome samples
were collected for elemental (C:N:P) and stable isotope (del
C-13, del N-15, and del S-34) analyses.
The first sampling was carried out in fall 1997, so
data have not yet been analyzed. However, the
following responses to stress gradients based on our
pervious work in Florida Bay and other estuaries are
anticipated: (1) T. testudinum populations will exhibit
changes in age structure. The number of new rhizome
apices and very old shoots are both expected to decline
along stress gradients. Plastochrone intervals (blade
turnover rates) will increase. (2) Spatial patterns of
elemental (C:N:P) ratios will reflect nutrient sources for
each system. Stable nitrogen isotopes, in particular, will
reflect gradients in anthropogenic and natural (fixed
nitrogen) supplies. (3) Morphological characteristics of
T. testudinum also will reflect cronic stress effects: leaf
blades will be narrower; shoot density and leaf area
index will decline. (4) Physiological reserves, such as
rhizome starch, will be depleted by the metabolic cost of
response to chronic stress.
20
-------
Anclote, Scale 3,1997
Figure 1. Multiscale seagrass sampling grid. Tesselated hexagons (30 at each scale) distribute sampling points over small (100 m2),
medium (10,000 m2), and large (1,000,000 m2) scales at each site. The orientation and proportions of grids are adjusted to
the impact gradient at each site.
21
-------
Modeling Spatial and Temporal Dynamics of Montane
Meadows and Biodiversity in the Greater Yellowstone Ecosystem
Diane Debinski
Iowa State University, Ames, IA
Mark Jakubauskas
University of Oklahoma, Norman, OK
Kelly Kindscher
University of Kansas, Lawrence, KS
This project's goal is to examine ecological
dynamics in the Greater Yellowstone Ecosystem (GYE),
concentrating specifically upon the spatial and temporal
dynamics of montane meadow communities. The
abiotic aspects of these communities as well as the
biodiversity of plant, bird, and butterfly communities are
being examined. This involves using intensive, local
field sampling to test for relationships between species
distribution patterns and remotely sensed data (see
Figure 1). This project's long-term goal is to develop
predictive species assemblage models based on
landscape-level habitat analysis. This research involves
several steps: (1) quantifying the spatial and temporal
variability in montane meadow communities; (2) devel-
oping a spectrally based, spatially explicit model for
predicting plant and animal species diversity patterns in
montane meadows; and (3) testing the spectrally based,
spatially explicit model for predicting plant and animal
species diversity patterns in montane meadows.
A time series of satellite imagery is being used for
monitoring the extent, condition, and spatial pattern of
montane meadows on a seasonal and interannual time
scale. Spectrally based, spatially explicit models are
being developed for six meadow types using a geo-
graphic information system to stratify the study area by
topography and geology. Two regions of the ecosystem
have been sampled: the northern part of the ecosystem,
hereafter termed the Gallatin study area, which included
the Gallatin National Forest and northwestern portion of
Yellowstone National Park; and the southern part of the
ecosystem, hereafter termed Teton study area, which
included Grand Teton National Park. Twenty-five
sample sites were located in the Tetons, and 30 sample
sites were located in the Gallatins. Birds, butterflies,
and plants were surveyed at each of the sites. All spe-
cies were identified in the field or given appropriate field
names. Voucher specimens were collected. Species that
were difficult to identify are being reviewed by experts
in the field.
Data entry for the 1997 field season is nearing
completion. Quality control of the data will follow. It
is already becoming apparent that there are many species
of birds, butterflies, and plants that show strong affini-
ties for one or more spectrally defined habitat classes.
A data matrix of species as well as functional and
ecological traits is being compiled. In addition, analysis
during the next several months will focus on assessing
temporal variation in reflectance values for each sample
site and conducting statistical analyses of relationships
between spectral reflectance and field-sampled species
distribution and biophysical data. Next spring and
summer data from all of the sites will be collected
again, following the above procedures. The third year
will be spent primarily on data analysis. If there are
sufficient funds, the collection of additional field data
will occur.
-------
1. Identify clear management goals for conducting biodiversity assessment
and develop testable hypotheses
2. Establish sampling sites and choose taxa to survey
3. Collect data
by sampling
site
Organism
Group A
Organism
Group n
Physical
environmental
data
4. Conduct
multivariate
ordination using
species x site
data
Classifications of
species in
Groups A, B, ...n
5. Conduct
statistical
analysis of
abiotic data
Remotely
sensed data
Statistical treatment
by site
6a.
Compare site
classifications based
on data from different
taxa
Test for relationships
between species
distribution patterns
and patterns of variation
in environmental data
Test for relationships
between species distribution
patterns and remotely
sensed data
Validate or refute hypotheses concerning
physical environmental factors affecting
patterns of distribution of groups
of organisms
8.
Establish a new array of
sampling sites to test
hypotheses
(Go back to step 1)
Scientific understanding of
relationships between
species and environmental
parameters
Figure 1. Research design for biological diversity assessment as seen in Debinski, D.M. and P.S. Humphrey. 1997. An integrated approach
to biological diversity assessment. Natural Areas Journal 17 (4): 355-365.
23
-------
Modeling and Multiobjective Risk Decision Tools for
Assessment and Management of Great Lakes Ecosystems
Benjamin F. Hobbs
Department of Geography and Environmental Engineering, The Johns Hopkins University, Baltimore, MD
Joseph F. Koonce and Ana B. Locci
Department of Biology, Case Western Reserve University, Cleveland, OH
This project addresses restoration and protection
of the Lake Erie ecosystem. Lake Erie's successes of
the 1980s—nutrient controls, rebounding fisheries—have
not continued into the 1990s. Fish populations are once
again in decline. Contaminant body burdens are no
longer decreasing, and the invasion of exotic species
threatens to further destabilize the ecosystem. Possible
climate warming would alter habitats.
Gaps in understanding the consequences of
management actions and the lack of tested techniques
for integrating multiple objectives and risk in decision-
making hamper efforts to deal with these new chal-
lenges. The 1994 State of the Lakes Ecosystem Confer-
ence recommended that an ecosystem approach be
adopted for studying ecosystem problems and the
stresses that cause them; that well-defined ecosystem
objectives be used to measure success in restoring
ecosystem integrity; and that roundtable and interdisci-
plinary approaches to decisionmaking be taken that ami
for consensus among stakeholders. This project at-
tempts to respond to those recommendations.
In particular, the project's goal is to develop and
test an integrated ecological assessment and decision
methodology for the Lake Erie ecosystem. The purpose
of the methodology is to assist managers and stake-
holders, who are involved in the Lake Erie Lakewide
Management Plan (LaMP) and other Lake Erie manage-
ment processes, to define objectives and evaluate
tradeoffs and risks associated with future uses. The
research plan addresses the following issues: (1) the
interaction of invasions of exotic species, nutrient
reductions, and fishery harvests; (2) the influence of
near-shore and tributary habitat on the offshore commu-
nity structure and productivity; (3) the effects of alter-
ation of the offshore community on contaminant body
burdens of Lake Erie fish; and (4) the sensitivity of
emerging ecosystem objectives to climate change.
The products of the research will include the
following: (1) an expanded Lake Erie Ecosystem
Model (LEEM), with modifications to habitat, hydrol-
ogy, and climate change components; (2) the develop-
ment and application of methodologies for decision-
making under multiple objectives and uncertainty; and
(3) workshops in which Lake Erie managers, who
participate in the Lake Erie LaMP, apply and evaluate
the model and methods.
The accomplishments during the first year of the
project include enhancements to LEEM, analyses of
ecosystem response to selected issues, and framing of a
high-priority management problem for analysis by
multiobjective risk methods.
Habitat has been linked to LEEM through the use
of Habitat Suitability Indices (HSIs), which modify
recruitment and predator-prey interactions. Alternative
models have been developed for relating HSIs for wet-
lands to changes in the mean and variation of Lake Erie
levels that could result from climate change. This was
accomplished by adapting the Case Western Reserve
University Great Lakes climate change model. Pres-
ently, an inventory of habitat supply for egg and larval
stages is under way for the lake and will be linked to
LEEM. Initial analyses using the modified LEEM have
explored the relative importance of alternative stressors
(habitat loss, phosphorus, and exploitation) upon the
Lake Erie ecosystem. Changes in habitat were found to
alter total biomass and walleye populations more than
changes in P loadings.
In consultation with Region V of the U.S.
Environmental Protection Agency, the issue of phos-
phorus management has been identified as the first
management problem to be addressed by the multi-
objective and risk methodologies. In particular, there
appear to be fundamental tradeoffs between lake
productivity (especially of the recreational and com-
mercial fisheries of walleye, yellow perch, and smelt)
and the taste, odor, and visual problems resulting from
eutrophication. As a result, whether or not it is
desirable to modify phosphorus loadings in part
depends on the relative weight placed on those objec-
tives. Further, ecosystem structure changes resulting
from the zebra mussel invasion have made it more
difficult to assess the relative contribution of habitat
changes, species invasion, and decreases in phospho-
rus loadings to the recent decreases in walleye and
yellow perch populations.
LEEM will be used, together with Monte Carlo
risk-propagation methods, to characterize the effect
of alternative phosphorus targets upon the objectives
and the uncertainties surrounding those effects.
Decision trees will be used to structure the informa-
tion. Figure 1 shows the general framework of the
analysis.
24
-------
to
Land use,
restoration of
habitat
Ambient
toxics
concentrations
Lake Erie
Habitat
Models
Phosphorus
loadings
Fisheries
management
Invasions
of exotics
Available
habitat
(near-, offshore?
tributary, wetland)
Lake Erie
Ecosystem
Model
Ecological criteria
(e.g., populations,
body burdens,
health indices)
Tradeoff &
Risk Display
Model
Tradeoffs among
objectives
Sensitivities,
risk distributions
Lake levels,
tributary
flows,
water
temperatures
Harvests
Economic
criteria
(e.g., costs,
harvest value)
Economic
Impact
Models
Evaluation
Models
Great Lakes
Hydrological
Models
Lake
levels
value
judgments re:
acceptable
tradeoffs
risks
Interest
rate, economic
growth,
population
assumptions
Climate
scenarios
(transient or
steady state
Precip,
Lake leve
regulation,
water
diversions
& use
Policy costs,
shoreline pro-
tection actions
Tentative
rankings,
policy
options
Figure 1. Lake Erie ecosystem management model structure.
-------
Use of Multiscale Biophysical Models for Ecological
Assessment: Applications in the Southeastern United States
Michael Huston
The Institute for Environmental Modeling, The University of Tennessee, Knoxville, TN
This project's goal is to develop an integrated set
of landscape analysis and computer modeling tools that
can predict natural patterns of spatial and temporal
variability in the major ecological properties, which are
used for environmental assessment. The capability to
predict natural patterns of variability in these properties
is a prerequisite for determining whether these ecologi-
cal properties are actually changing, either in response
to anthropogenic impacts or to natural environmental
change.
The conceptual framework for this work is to use
the interaction of climate and geomorphology to predict
spatial and temporal variation in environmental pro-
cesses that influence the growth rates of organisms (i.e.,
net primary productivity) and the frequency and severity
of disturbances that cause mortality. The specific eco-
logical properties that are being analyzed and modeled
include: (1) spatial and temporal (interannual) variation
in primary productivity; (2) spatial and temporal (decad-
al) variation in soil carbon and hydrologic storage
capacity; (3) spatial and temporal (interannual to decad-
al) variation in the species diversity of selected func-
tional types of organisms; and (4) spatial and temporal
variation (interannual) hi the population size and dynam-
ics of selected plant and animal species.
The biophysical modeling approach being used
combines tools for visualizing and extracting landscape
scale data from digital elevation models and vegetation
maps (see Figure 1) with tools for: (1) analyzing point
samples from field data collection in the context of
landscape scale data from the geographic information
system (GIS) applications and (2) modeling hydrological
and ecological processes across landscapes using digital
elevation models and other information. The data
analysis and modeling are focused on three areas that
represent different spatial scales of pattern and processes
in the Southeastern United States: the Oak Ridge
National Environmental Research Park, 150 km2; the
Great Smoky Mountains National Park, 2,000 km2,
serving as a prototype; and the Southern Appalachian
Region, 150,000 km2. Within each of these areas,
preexisting data are being used for the analyses, with
some limited field work to address specific issues where
adequate data are not available.
Work during the first year of the project has
concentrated on the smallest spatial scale. Examples of
current projects illustrating the range of processes that
are being integrated through geomorphological modeling
include: (1) analysis of environmental monitoring data
for fish and benthic insect diversity in streams on the
Oak Ridge Research Park; (2) development of a topo-
graphically based hydrologic model for the Oak Ridge
Research Park; (3) tree-ring analysis to determine
historical patterns of spatial variation hi interannual
growth increments in tulip poplar; (4) field sampling
and computer modeling of the impact of ice storms on
forests in the Southern Appalachian region; and (5)
development of an individual-based model of black
bears for predicting bear population dynamics in the
Smokies.
26
-------
Watershed Topography Drives Both Hydrologic and
Soil Carbon and Nitrogen Dynamics
TOPOGRAPHIC
INDEX
3.5
5.5
16.5
I J
I
26
22
O
CO 18
14
56789
TOPOGRAPHIC INDEX
+ SW Facing Slopes
* Ridgetops
» NE Facing Slopes
T Valley Bottoms
Figure 1. The carbon:nitrogen ratio in soil is an important indicator of soil nitrogen dynamics, which have a strong effect on plant growth and
net primary productivity. The inset map of Walker Branch Watershed on the Oak Ridge National Environmental Research Park was
developed from a high-resolution digital elevation model (DEM) and is shaded using a hydrologic wetness index based on drainage
area and slope. High values of the index (darker colors) indicated wetter than average soil conditions. The inset graph shows the
strong correlation between soil C:N and the topographic (wetness) index, indicating that topography alone is a good predictor of this
important soil property (based on C.T. Garten, M.A. Huston, and C. Thorns. 1994. Topographic variation of soil nitrogen
availability at Walker Branch Watershed, Tennessee. Forest Science 40: 497-512).
-------
Monitoring and Restoring
Hydropatterns in South Florida Ecosystems
Eric S. Kasischke, Laura L. Bourgeau-Chavez, Jeff Michalek, and Kevin P. Smith
Earth Sciences Group, Environmental Research Institute of Michigan, Ann Arbor, MI
Edwin Romanowicz and Curtis J. Richardson
Nicholas School of the Environment, Duke University, Durham, NC
This project's goal is to develop and evaluate new
methods to monitor and predict the spatial and temporal
patterns of surface water inundation (hydropatterns) in
the different wetland ecosystems found throughout the
greater South Florida Ecosystem. The objectives of the
research are three-fold: (1) to develop approaches to use
space-borne imaging radar systems to monitor variations
in relative soil moisture and surface water inundation in
the study area; (2) to implement hydrologic models that
can be used to estimate the expected patterns of surface
water inundation (hydropatterns) in specific test sites
and the influence of anthropogenic activities (e.g.,
building of structures and management of water flow)
on these patterns; and (3) based on a variety of future
precipitation scenarios, determine the best strategy for
maintaining or improving water flow in regional wetland
ecosystems by altering structures or management
practices.
The approach will be carried out using a variety
of analysis approaches. A series of 10 test sites have
been located in the Big Cypress National Preserve and
the Everglades National Park. These field data then will
be correlated with variations in radar image intensity
measurements derived from several different satellite
systems, including ERS, Radarsat, and JERS. Ap-
proaches will be developed to use the information from
the radar imagery to map patterns of water inundation,
including the influence of man-made structures and
management activities on surface water flow patterns.
At the same time, hydrologic models will be obtained
and implemented that can be used to estimate the
expected patterns of surface water flows for the two test
sites. The model will be exercised for the time period
during which the radar imagery was collected to deter-
mine how well the models can predict actual surface
inundation patterns. The models will then be exercised
under a variety of scenarios to evaluate how future
climate change as well as changes in management
practices and restoration activities may influence natural
water flows through South Florida wetland ecosystems.
To date, efforts have been focused on establishing
field test sites and analysis of available radar imagery.
Distinct patterns have been shown to exist on radar
imagery (see Figure 1), which correlate closely with
seasonal variations in water levels. In addition, it has
been demonstrated that the influence of roads on the
interruption of the natural surface flow of water can be
detected on the radar imagery. Based on these initial
results, resource managers from the National Park
Service are now considering using radar imagery to
restore natural water flow across the Taimiami Trail,
which bisects die Big Cypress National Preserve.
Correlation of field data and radar imagery will
occur to determine the exact correlation between radar
intensity and surface water levels in different vegetation
types. In addition, hydrologic models and data sets
necessary to exercise these models are being imple-
mented.
28
-------
13 August 1996 ERS-2 SAR Image - Wet Season 15 April 1997 ERS-2 SAR Image - Dry Season
Figure 1. Two ERS radar images collected over southern Florida. Variations in image intensity are due to changes in water level and soil
moisture between the wet seasons and dry seasons.
-------
Development of Environmental Assessment,
Mitigation, and Restoration Techniques for Coral Reefs
Robert H. Richmond
University of Guam Marine Laboratory, Mangilao, Guam
The goals and objectives of this project include
the following: (1) to improve techniques used for the
assessment of coral reef health and sustainability; (2) to
develop appropriate coral reef biomonitoring protocols;
(3) to develop techniques for coral reef restoration and
guidelines for mitigation of anthropogenic disturbance;
and (4) to develop a set of criteria for EIA's and EIS'
for activities occurring on or adjacent to coral reefs, or
within watersheds that may affect coastal coral reef
ecosystems.
The approach will involve expanding present
studies on the factors affecting reproduction (fertiliza-
tion) and recruitment of corals, including tests of the
effects of water and substratum quality. Survivorship
under the conditions of decreased light, increased
nutrients, sedimentation, and contamination will be
investigated. Comparative bioassays will be performed
using EPA-approved species along with the chosen coral
and related reef species to determine if the sensitivity is
as good or better. In addition, the protocols for coral
fertilization and larval settlement bioassays will be
standardized, and the ability of transplanted (and
cultivated) corals to serve as indicators of stress from
sewer outfalls and other human disturbances will be
tested.
Preliminary findings indicate that progress has
been made on all aspects of the project. During the
coral spawning events of July and August, gametes were
collected from a variety of species, and a number of
fertilization experiments were performed. The tech-
niques for cultivating corals were refined and simplified.
Using coral larvae reared from the spawning events,
pesticide bioassays were performed using metamorphic
induction as the test of effect, with highly significant
results. Additional experiments focusing on reef res-
toration using coral larvae are under way.
The program to study effects of the sewage
outfalls and eutrophication on corals is progressing.
Cohorts of corals have been raised from larvae for
transplantation to the outfall site and two reference sites.
Due to the potential hazards involved in diving near the
sewage outfalls, specialized dry suits and full face
masks were purchased, and divers trained and certified
in their use were employed.
The grant also has been used to develop appropri-
ate policies based on scientific data. Regular meetings
have been taking place, including the principal investi-
gator and representatives from the regulatory agencies
on Guam such as the Guam EPA, the Division of
Aquatic and Wildlife Resources, the Coastal Zone
Management Program, and the Public Utilities Agency.
Also, opportunities for public participation have been
made available. During the summer, a community-
based coral reef survey was performed with support
from this grant.
The accomplishments and research results include
the following: (1) Coral Cultivation: A standardized set
of protocols has been developed for cultivating coral
larvae and coral colonies. (2) Pesticide Bioassays: A
determination has been made that different life-history
stages of corals demonstrate differential sensitivities to
pollutants. Water-soluble pollutants are particularly
problematic to egg-sperm interactions, while lipophilic
substances can interfere with recruitment (settlement and
metamorphosis). Studies of the effects of an organo-
phosphate pesticide, Chlorpyrifos, found that this
chemical can significantly affect settlement and meta-
morphosis in coral planula larvae (see Figure 1). This
chemical appears to affect both metamorphic inducers
found in the preferred settlement substrata and the larval
receptors responsible for metamorphosis. (3) Effects of
Se\vage on Corals: Several cohorts of corals have been
raised for transplantation at one outfall site and two
reference sites. (4) Criteria for Monitoring Coral Reef
Health/EIA 's and EIS': A coral reef "physical exam"
has been developed to identify the monitoring options
and provide guidance in their application. Also, a draft
of EIA and EIS procedures has been prepared as part of
a larger effort at developing legislation at both the local
and possibly the federal levels.
Procedures for coral cultivation have been
standardized to the point where it is practical to develop
coral-specific bioassays for environmental monitoring of
reefs. Using coral larvae, it was demonstrated that
standard LC50 protocols applied to adult corals can miss
substantial effects on populations and communities.
These data are being used to develop a framework for
EIA and EIS requirements for activities proposed
adjacent to coral reefs, or within watersheds affecting
coastal reef areas.
The coral transplants will be placed out hi the
field in the near future. Another series of pesticide bio-
assays is planned for later this year. The cultivated
corals are being monitored for growth rates and sur-
vivorship. Additional reproductive data are being
collected for other species with the potential for use in
bioassays. The data being generated by this research,
as well as data from other studies, are being compiled
and evaluated for application to the assessment of coral
reef health.
30
-------
40 i
settlement (numbe
CO
o
2} on-
jra 20
3
C
CO
Q.
CO
0)
o -
n -
rr1
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
s
/
/
/
t
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
f
/
/
/
/
/
\
/
/
/
/
/
/
/
/
/
',
f
/
/
'f
/^
Two-way ANOVA P«, n
F=9.37 ° °
p«0.05 ^ D
Q D
H D
Q D
T i=
life I
ay
ay
ay
ay
ay
1
2
3
4
5
nTr—
I *T- • -r-
I : : : I
I : : : I
I : : :
'/ :•:•: ' '
r/ ••;•{ '
a.
Control Control w/MeOH 20ppb BOppb
Treatment (ug/L)
101ppb
Figure 1. Effects of potential hydrophobic/lipophilic substance, Chlorpyrifos (organophosphorus pesticide), on planulae settlement
(substrate was exposed to pollutant for 24 hours).
-------
Assessment and Analysis of Ecosystem Stressors
Across Scales Using Remotely Sensed Imagery: Reducing
Uncertainty in Managing the Colorado Plateau Ecosystem
Stephanie J. Weigel
Department of Environmental Health, Colorado State University, Fort Collins, CO
This project investigates scale issues in reducing
uncertainty in ecosystem management for the Colorado
Plateau Ecosystem (CPE), by examining potential
characteristic scales at which environmental stressors
and their effects may be manifested on ecosystem
landscapes, using remotely sensed imagery. The project
objective is the development of a standardized analytical
algorithm for using multiscale, remotely sensed data in
the characterization and analysis of landscapes at the
ecosystem level.
Earth system processes and the effects of hu-
man/land interactions can be detected and monitored
over specific space and time intervals, known as charac-
teristic scales. Remotely sensed data are resampled
across a range of pixel resolutions over the spatial extent
of the CPE. Scale analysis techniques are used to
examine the landscape at different pixel resolutions,
indicating characteristic scales of ecosystem processes,
patterns, and disturbances. The Landsat MSS imagery
used is available through the North American Landscape
Characterization (NALC) program and has been initially
resampled to a 60 meter pixel resolution. NALC data
sets contain imagery from the 1970s, 1980s, and 1990s
as well as a Digital Elevation Model (DEM). Subse-
quent resampling creates a sequence of lower resolution
versions of the images (see Figure 1), corresponding in
part to the 250 meter, 500 meter, and 1,000 meter pixel
resolutions of the proposed NASA MODIS (Moderate
Resolution Imaging Spectroradiometer) sensor. Scale
effects methodologies are used to analyze each image at
each time point and resolution level as well as to analyze
difference images (e.g., 1970s/1980s). The scale ef-
fects methodologies being used include the following:
fractal analysis, local variance analysis, variogram
analysis, and multiscale variance analysis.
Based on suggestions of the proposal reviewers,
preliminary work was added to the project involving the
choice of resampling methodology for use in the project.
Eight subsets (864 x 864 pixels) representing different
landscape types across the CPE were used to examine
the responses to four different resampling methodolo-
gies: (1) averaging; (2) systematic sampling; (3) filter
weighting; and (4) filtering, using an algorithm designed
to simulate results from a lower resolution sensor
(MODIS). Results indicate that using a straight averag-
ing algorithm methodology presents some difficulties
with the way in which landscapes are represented (i.e.,
linear ground objects are weighted and represented more
heavily than their polygon counterparts in the original
image). Preliminary work with the filter weighting
algorithm indicates that it is the most appropriate
approach to effectively represent the ecosystem land-
scape at lower resolutions.
The resampling algorithm used to change the
pixel resolution of imagery for scale analyses does make
a difference on the results of subsequent analyses
performed. Choosing an appropriate resampling algo-
rithm that allows for a more realistic simulation of a
lower resolution sensor will result in more meaningful
and useful results of subsequent analyses.
Following the implementation of the filter weight-
ing algorithm to all of the images, the next steps involve
the creation of the difference images for the time steps
(e.g., 1970s/1980s; 1980s/1990s) and multiscale an-
alysis (i.e., implementation of the four scale effects
methodologies on all of the images).
32
-------
Figure 1. A 60 meter vs. 960 meter pixel resolution. Agricultural region along Gunnison River, CO.
------- |