PROCEEDINGS
1998 WATER AND WATERSHEDS
PROGRAM REVIEW
28-29 January 1998
Corvallis, Oregon
Sponsored by the EPA/NSF Partnership for Environmental Research
\ NATIONAL SCIENCE FOUNDATION
^60 ST4t.
ENVIRONMENTAL PROTECTION AGENCY
-------
Cover Photo: Aerial photograph of the White Clay Creek watershed near the Stroud Water
Research Center, Chester County, Pennsylvania. The photo, taken after a snowstorm, shows patches
of forest (dark areas) interspersed among deforested meadow or cropland (white) areas. This region
of southeastern Pennsylvania contains many of the study streams for Sweeney et al.'s study entitled
"Streamside Reforestation: An Analysis of Ecological Benefits and Societal Perceptions" (see page
38). Photograph from Stroud Center archives; R.L. Vannote.
-------
PROCEEDINGS
1998 WATER AND WATERSHEDS
PROGRAM REVIEW
28-29 January 1998
Corvallis, Oregon
Sponsored by the EPA/NSF Partnership for Environmental Research
ENVIRONMENTAL PROTECTION AGENCY
NATIONAL SCIENCE FOUNDATION
-------
Table of Contents
Introduction .... ... . . . . . vii
Section 1. Projects Initiated With Fiscal Year 1997 Support
Community Values and the Long-Term Ecological Integrity of Rapidly Urbanizing Watersheds . 2
M. Bruce Beck, T.C. Rasmussen, B.C. Patten, K.G. Porter, B.C. Norton, A. Shepherd
Connecting Ecological and Social Systems: Watershed Research Relating Ecosystem Structure
and Function to Human Values and Socioeconomic Behaviors . ... 3
Gaboury Benoit, S. Kellert, M. Ashton, P. Barten, L. Bennett, D. Skelly
Social and Ecological Transferabiliry of Integrated Ecological Assessment Models ... .4
Linda A. Deegan, James Kremer, Thomas Webler
From Landscapes to Waterscapes: Integrating Framework for Urbanizing Watersheds . 6
PanosDiplas, W.E. Cox, D.F. Kibler, V.K. Lohani, R.G. Greene, D.J. Bosch, L.A. Shabman,
K. Stephenson, E.F. Benfield, P.S. Nagarkatti, S. Mostaghimi, D.J. Orth
An Integrated Ecological and Socioeconomic Approach To Evaluating and Reducing Agricultural
Impacts on Upper Mississippi River Watersheds . 8
Prasanna H. Gowda, R.J. Haroa, A.D. Ward, T.L. Napier
Integrated Ecological-Economic Modeling of Watersheds and Estuaries at Multiple Scales 10
Charles Hopkinson, Edward Rastetter, Joseph Vallino
Linking Watershed-Scale Indicators of Changes in Atmospheric Deposition to Regional Response Patterns 11
Jeffrey S. Kahl, I. Fernandez, D. Mageean, S. Ballard, S. Norton, J. Cosby, P. Ludwig, L. Rustad
Comprehensive Watershed Management: A Spatial Water Quality Assessment System 12
C. Gregory Knight, Robert P. Brooks, Barry M. Evans, James M. Hamlett, Archie J. McDonnell,
Gary W. Petersen, Todor N. Hristov
Development and Implementation of Decision Support Systems for Predicting Economic and Ecologic
Impacts of Alternative Land and Water Management Policies in Urbanizing Regions 14
Daniel P. Loucks, Tammo S. Steenhuis, Mark B. Bain, Warren Brown, Walter R. Lynn
Effects of Natural and Anthropogenic Processes on Tillamook Bay and Its Watershed: An Integrated
Process Study and Land Use Perspective 16
James McManus, Paul Komar, M.S. Jesse Ford, Courtland Smith, Debbie L. Colbert, Greg Bostrom
Ecological Risks, Stakeholder Values, and River Basins: Testing Management Alternatives for the
Illinois River . . . . 17
Mark Meo, Baxter Vieux, James Sipes, Edward T. Sankowski, Robert Lynch, Will Focht,
Keith Willett, Lowell Caneday
Risk-Based Urban Watershed Management: Integration of Water Quality and Flood Control Objectives 19
Vladimir Novotny, Robert Griffin, David Clark, Douglas Booth
Impact of Social Systems on Ecology and Hydrology in Urban-Rural Watersheds: Integration for
Restoration 20
Steward T.A. Pickett
Section 2. Projects Initiated With Fiscal Year 1996 Support
Strategic Renewal of Large Floodplain Rivers: A Preliminary Status Report ... . .25
John B. Braden, Misganaw Demissie, Eric DeVuyst, Paromita Mitra, Daniel Schneider,
Richard E. Sparks, David C. White, Renjie Xia
in
-------
Table of Contents (continued)
Integrating Modeling and Management of Agriculturally Impacted Watersheds: Issues of Spatial
and Temporal Scale . . . . . . . . 26
Patrick L. Brezonik, K. William Easter, David Mulla, James A. Perry
Urban Stream Rehabilitation in the Pacific Northwest: Physical, Biological, and Social Considerations ... 28
Stephen J. Surges, Derek B. Booth, Sally Schauman, James R. Karr
Influences of Forest Fragmentation on Watershed Functions in Northern Vietnam—Preliminary
Field Results . . 29
Jeffrey Fox, T. Giambelluca, A.T. Rambo
Geochemical, Biological, and Economic Effects of Arsenic and Other Oxyanions on a Mining
Impacted Watershed . . .... . 31
Glenn C. Miller, Watkins W. Miller, Scott Tyler, Douglass Shaw, Ron Hershey, Lambis Papellis,
Susan Anderson
Effectiveness of Regulatory Incentives for Sediment Pollution Prevention: Evaluation Through Policy
Analysis and Biomonitoring ... . . 32
Seth R. Reice, Richard N. Andrews
Watershed Protection in Agricultural Environments: Integrated Social, Geomorphological, and
Ecological Research To Support Ecosystem-Based Stream Management . . 33
Bruce L. Rhoads, Edwin E. Herricks, David Wilson
Towards an Integrated Regional Model of River Basins of the Western Pacific Rim . ... 34
Jeffrey E. Richey
An Integrated Approach To Assessing Water Management Options in a Major Watershed: Extending
a Hydrodynamic-Water Quality Model To Include Biological and Politico-Economic Components 35
Paul Sabatier, Loo Botsford, Mike Johnson, Jay Lund, Peter Moyle, Gerald Orlob, James Quinn,
Peter Richerson, Tom Suchanek, Marca Weinberg
Modeling Effects of Alternative Landscape Design and Management on Water Quality and Biodiversity
in Midwest Agricultural Watersheds . . . . . 36
Mary Santelmann, K. Freemark, D. White, S. Polasky, G. Matzke, J. Eilers, J. Bemert,
B. Danielson, R. Cruse, J. Nassauer, S. Galatowitsch
Streamside Reforestation: An Analysis of Ecological Benefits and Societal Perceptions . . 38
Bernard Sweeney, Thomas Bott, John Jackson, Louis Kaplan, J. Denis Newbold, Laurel Standley,
Richard Horwitz, W. Cully Hession, Janet Johnson, James Finley, Caren Glotfelty, Cecilia Ferreri
Integrated Urban Watershed Analysis: The Los Angeles Basin and Coastal Environment ... .40
Richard P. Turco
Section 3. Projects Initiated With Fiscal Year 1995 Support
Development and Application of Spectroscopic Probes for Measurement of Microbial Activity in
Aquatic Ecosystems .... . . ... . . . . . . 45
Carol Arnosti, Neil V. Blough
Watersheds and Wetlands: Large-Scale Disturbances and Small-Scale Responses . . .47
Charles Andrew Cole, Robert P. Brooks, Denice Heller Wardrop
Integrated Ecological Economic Modeling and Valuation of Watersheds .... ... . .49
Robert Costanza, RoelofBoumans, Tom Maxwell, Ferdinando Villa, Alexey Voinov,
Helena Voinov, Lisa Wainger
IV
-------
Table of Contents (continued)
Oyster Reefs as Structural and Functional Components of Tidal Creeks: An Ongoing Ecosystem
Experiment ... . . ... 51
Richard F. Dame, E. Koepfler, L. Gregory, T. Prins, D. Allen, D. Bushek, C. Corbett, D. Edwards,
B. Kjerfve, A. Lewitus, J. Schubauer-Berigan
Tracing the Fate of Nitrogen Inputs From Watersheds to Estuaries . 53
Linda A. Deegan, Bruce J. Peterson
Probing the Relationship Between Fulvic Acid Aggregation, Metal Ion Complexation, and the Binding
of Organic Compounds ... 55
A. Dixon, W.R. Carper, C.K. Larive
Diffusional Rate Limitations hi Heterogeneous Porous Media: Model Structure, Scale, and Geologic
Characterization 56
David L. Freyberg, Paul V. Roberts
Integrating Planning, Forecasting, and Watershed Level Ecological Risk Assessment Techniques:
A Test in the Eastern Cornbelt Plains Ecoregion 57
Steven I. Gordon, Andy Ward, Dale White
Development of Geomorphological Artificial Neural Networks (GANNs) for Modeling Watershed Runoff ... 59
Rao S. Govindaraju
Physicochemical Mechanisms Governing Virus Filtration ... 60
Stanley B. Grant, Terese M. Olson, Mary K. Estes
Watershed Impacts on Sediment Pollution History and the Viability of the Zooplankton Egg Bank 61
Nelson G. Hairston, Jr., Colleen M. Kearns, Charles T. Driscoll
The Role of Colloidal Particles in the Transport of Chemicals Through an Agricultural Watershed . . 62
George M. Hornberger, Janet S. Herman, James E. Saiers
Geomorphic, Hydrologic, and Ecological Connectivity in Columbia River Watersheds: Implications for
Endangered Salmonids . ... .... . 64
Hiram W. Li, Bruce A. Mclntosh, J. Boone Kauffman, Judith L. Li, Robert L. Beschta,
Patricia McDowell
Resistance of Communities to Chronic Haloaromatic Contamination From Biogenic and Anthropogenic
Sources . . .... . . . . 65
David E. Lincoln, Sarah A. Woodin, Charles R. Lovell, V. Pernell Lewis
Influences of Watershed Land Use on Stream Ecosystem Structure and Function 66
Judith L. Meyer, E.A. Kramer, M.J. Paul, W.K. Taulbee, C.A. Couch
The Role of Hg (II) Reduction and Chemical Speciation in Controlling the Concentration of Mercury
and Its Methylation in Natural Waters . . . . 68
Franfois MM. Morel
Formation and Propagation of Large-Scale Sediment Waves in Periodically Disturbed Mountain
Watersheds ... .... . 69
Gary Parker
Multiscale Statistical Approach to Critical-Area Analysis and Modeling of Watersheds and Landscapes ... 70
G.P Patil, W.L. Myers
-------
Table of Contents (continued)
Contemporary Water and Constituent Balances for the Pan-Arctic Drainage System: Continent to
Coastal Ocean Fluxes . . . . . 72
Bruce Peterson, Charles Vorosmarty, Richard Lammers
Modeling Temporal Rainfall via a Fractal Geometric Approach . . .74
Carlos E. Puente
Effects of Food Web Structure and Nutrient Loading on Lake Productivity and Gas Exchange With
the Atmosphere . • • - • • • • • • • . . 75
Daniel E. Schindler, Stephen R. Carpenter, James F. Kitchell, Jonathan J. Cole, Michael L. Pace
A Comparative Institutional Analysis of Conjunctive Management Practices Among Three Southwestern
States . ...... 76
Edella Schlager
Water and Sustainable Development in the Binational Lower Rio Grande/Rio Bravo Basin . 77
Jurgen Schmandt
Environmental Change and Adaptive Resource Markets: Computer-Assisted Markets for Resource
Allocation . . .79
Vernon Smith, S. Rassenti, E. Hoffman, R. Howitt, A. Dinar
Detecting Fecal Contamination and Its Sources in Water and Watersheds . . .80
MarkD. Sobsey
Ecoregion-Specific Comparison of Stream Community Responses to Nutrient Gradients Using Both
Survey and Experimental Approaches . . .81
R. Jan Stevenson, Mike Wiley, Joe Holomuzki
Holocene Floodplain Development as a Function of Climate Change and Human Activities:
The Arroux and Loire Rivers, Burgundy, France . 82
Eric C. Straffin, Michael D. Blum
Response and Compensation to a Bivalve Invasion by an Aquatic Ecosystem 83
David L. Strayer, Nina Caraco, Jonathan J. Cole, Stuart Findlay, Michael L. Pace
Scaling Up Spatially Distributed Hydrologic Models of Semi-Arid Watersheds . . 84
David G. Tarboton, Christopher M. U. Neale, Keith R. Cooley, Gerald N. Flerchinger, Clayton L. Hanson,
Charles W. Slaughter, Mark S. Seyfried, Rajiv Prasad, Charlie Luce, Greg Crosby, Changyi Sun
Traveling Wave Behavior During Subsurface Transport of Biologically Reactive Contaminants:
Implications for In Situ Bioremediation . . 85
Albert J. Valocchi
Carbon Exchange Dynamics in a Temperate Forested Watershed (Northern Michigan): A Laboratory
and Field Multidisciplinary Study . . . .87
LynnM. Walter, L.M. Abriola, J.M. Budai, G.W. Kling, P.A. Meyers, J.A. Teeri, D.R. Zak
A Comparison of Agricultural vs. Forested Basins: Carbon and Nutrient Cycling Within the
Hyporheic Ecotone of Streams 89
David S. White, Susan P. Hendricks, Timothy C. Johnston, George Kipphut, William E. Spencer
In Situ Assessment of the Transport and Microbial Consumption of Oxygen in Groundwater . 99
Tadashi Yoshinari, R.L. Smith, J.K. Bohlke, K. Revest
Index of Authors . . 91
VI
-------
Introduction
The U.S. Environmental Protection Agency / National Science Foundation (EPA/NSF) Water and
Watersheds competition is one of three special extramural awards competitions supported by the
EPA and the NSF under a partnership for environmental research initiated in 1994.
The competition emphasizes interdisciplinary research taking a systems approach to issues of water
and watersheds. Its goal is to 1) develop an improved understanding of the natural and anthropogenic
processes that govern the quantity, quality, and availability of water resources in natural and human-
dominated systems, and 2) improve understanding of the structure, function, and dynamics of the
terrestrial and aquatic ecosystems that comprise watersheds.
The 1995 Water and Watersheds competition reviewed 656 proposals requesting ~$335M and made
36 awards. In 1996, the competition announcement was narrowed to focus more on interdisciplinary
research—as well as to limit the number of proposals. The 1996 competition reviewed 249 proposals
requesting ~$168M and 12 awards were made. The agencies further narrowed the announcement in
1997 primarily in response to concerns that the competition had too low a success rate. Proposals
were required to integrate physical, ecological and social science research and, for the first time, the
investigators were encouraged to take a community-based approach. The 1997 competition, with an
emphasis on urban/suburban research, reviewed 128 proposals requesting ~$77M and made 13 awards.
The abstracts in this volume are organized alphabetically, within three sections that correspond to
the year of award. The most recent awards (FY97) are first. Many of these projects were just getting
underway as this publication went to press and the abstracts indicate goals and plans. The FY96
cohort of projects appear next. These abstracts report early findings and plans for future years. The
projects that were initiated with FY95 support are in the third section. These abstracts report results
based on several years of research.
The competition is currently entering its fourth year, and the U.S. Department of Agriculture
(USDA) lias joined the two original agencies in supporting the effort. The theme for the fourth year
competition is rehabilitation of watersheds. Program reviews such as this one will allow investigators
to interact with one another, and to discuss progress and findings with EPA, NSF, USDA and other
federal officials who are interested in the program.
Any opinions, findings, conclusions, or recommendations expressed in this report are those of the
investigators that participated in the research and the Program Review meeting, and not necessarily
those of the NSF or the EPA. For further information on the EPA/NSF Water and Watersheds
competition please contact the Program Officers: Ms. Barbara Levinson, EPA,
levinson.barbara@epamail.epa.gov or Dr. Penelope Firth, NSF, pfirth@nsf.gov.
vn
-------
Section 1.
Projects Initiated With Fiscal Year 1997 Support
-------
Community Values and the Long-Term
Ecological Integrity of Rapidly Urbanizing Watersheds
M. Bruce Beck and T. C. Rasmussen
Daniel B. Warnell School of Forest Resources, University of Georgia, Athens, GA
B.C. Patten and K.G. Porter
Institute of Ecology, University of Georgia, Athens, GA
B.G. Norton
School of Public Policy, Georgia Institute of Technology, Atlanta, GA
A. Shepherd
City Planning Program, College of Architecture, Georgia Institute of Technology, Atlanta, GA
Untoward Disturbances
Controls/Policy
THE LAKE/WATERSHED
Natural Science Model
(Hydrology; Ecology)
Short-Term Response
Long-Term Response
Desired Short-
Term Response
Potential Changes of
Model Structure
THE COMMUNITY
Belief Networks and
Plurality of Mental Models
(Social/Policy Sciences)
.Mismatch & Dissonance
Desired Long
Term Response
Figure 1. Schematic representation of the interaction between natural and social/policy science models in which the tasks are to identify: (a) on
which critical, key unknowns in the natural science base may hinge the reachability of feared/desired futures; and (b) how community
and stakeholder concerns for the future change in the light of changing information about the hydrological and ecological status of the
lake and its watershed.
Watersheds surrounding metropolitan Atlanta are
expected to experience substantial development during
the next two decades. Indeed, Atlanta's continued rapid
economic development could be constrained by pro-
blems of access to sufficient water supplies from the
relatively small headwater catchments to its north.
Within these watersheds lies Lake Lanier, which is used
for hydroelectric power generation, water supply, flood
protection, and most significantly, recreation.
The investigators are developing a new approach
in which both community interests and a complex
(mathematical) representation of the lake's ecosystem
can be engaged in exploring how shorter term individual
preferences can be reconciled with longer term
community values regarding the maintenance of the
integrity of an environmental system. The primary
hypothesis is that citizens' value commitments are
dependent upon scale (both in time and place) and that
these scales have their counterparts in the range of time-
constants typifying the behavior of the physical system:
the short term, spanning 0-5 years; and the long term,
more than 20-100 years. The secondary hypotheses are
that: (1) radical shifts of behavior, as gauged by
qualitatively different patterns of model outputs (such as
populations of fish), may be a function of slowly
evolving changes in the values of the model's co-
efficients (i.e., growth, mortality, and predation rates);
and (2) longer term change, and hence maintenance of
the integrity of a system, is primarily a function of such
coefficient variations.
Adaptive community learning will be facilitated
by working with small groups of stakeholders. The
procedure will comprise several iterations around the
cycle of eliciting community values, allowing their
modification, encoding stakeholder concerns about the
long-term future (as target features to be matched by the
model's simulation results), identifying consensus on
policy options and/or reconciling discord between
longer and shorter term aspirations, and generating in-
sight into the attainability of the target futures through
the scientific model (see Figure 1). The model will be
developed from past studies of Lanier and its watershed
and will incorporate new material for description of the
microbial food web and sediment biochemistry, the
latter being supported by modest field and laboratory
work. The model will be used principally within the
computational framework of a sensitivity analysis
employing Monte Carlo simulation.
In addition to developing and testing a prototype
approach to facilitating adaptive community learning,
with special reference to preserving the attainability of
goals on an intergenerational time scale, this project will
develop a new model for the dynamics of an ecosystem
in a southeastern impoundment and policy guidance on
options for maintaining the longer term integrity of such
a system.
-------
Connecting Ecological and Social Systems:
Watershed Research Relating Ecosystem Structure
and Function to Human Values and Socioeconomic Behaviors
Gaboury Benoit, S. Kellert, M. Ashton, P. Barten, L. Bennett, and D. Shelly
Yale Environmental Studies, New Haven, CT
Ecosystem
Health
Primary Productivity
Organic Matter Cycling
Biogeochemical Cycling
Hydrologic Regulation
Energy Processing
Biodiversity
Human
Environmental Values
Aesthetic Dominionistic ^
Humanistic Naturalistic
Negativistic Moralistic I
v Scientific Symbolic /
Utilitarian
Socioeconomic Benefits
Property Values
Material Production
Consumption Patterns
Outdoor Recreation
Water and Land Use
Intellectual Skills
Sense of Place
Community Stability
Figure 1. Hypothesized feedback loop linking ecosystem structure and function, environmental values, and Socioeconomic benefits.
The goal of this project is to examine how eco-
logical and social systems influence each other through
either positive or negative feedbacks within watersheds.
The central hypothesis is that the relative health and in-
tegrity of ecosystems cause, and in turn, are caused by
enhanced human performance and productivity. Con-
versely, damaged and degraded ecosystems cause, and
in turn, are caused by diminished human performance
and productivity.
The current research will use observational
techniques integrated by standard statistical methods to
measure the quantifiable linkages between biophysical
and social systems. This investigation also will lay the
groundwork for a field experiment to test whether re-
storing a degraded watershed can significantly enhance
social interactions in associated human communities.
State-of-the-art analytical methods will be used to
characterize approximately 30 sub-watersheds in terms
of key hydrological, chemical, biological, economic,
and social parameters. This large number of biophysi-
cal and social measures will be integrated within a
quantitative context that allows us to test our pre-
dictions and to generate an overall index of watershed
quality.
The first step has been to select study sites from
among an initial set of more than 60 possible candidates.
The investigators see objective selection of appropriate
sites as crucial to the long-term success of this project.
The final set of sub-watersheds need to fulfill several
criteria: (1) relatively homogeneous land use or land
cover in each; (2) all large enough to include a sufficient
number of residents to yield robust statistics on social
science surveys, even in areas having the minimum
regional population density; (3) adequate in size to
ensure year-round streamflow; (4) all sites within +/-
50 percent of the mean area; (5) representative of the
entire range of development from rural to urban; and (6)
covering the full spectrum of environmental quality
from nearly pristine to badly degraded. Because these
last two characteristics are not entirely correlated, a set
of sites are being sought that includes representatives
having: (1) a high level of development with good
environmental quality and (2) low development with
poor environmental quality as well as the more common
converses of these two.
Every sub-watershed within our three river basins
fulfilling criteria 1 through 4 have been delineated on
24,000 scale USGS topographic maps and all ambiguous
boundaries clarified through field visits. Preliminary
land cover and land use have been calculated for each
sub-watershed, allowing them to be ranked as a test of
criterion 5. Final land use characteristics will be de-
rived from an existing detailed geographic information
systems' database. Rapid assessments of water quality
and biological characteristics are being conducted to
evaluate criterion 6.
Also, we are in the process of designing and
testing surveys to evaluate human attitudes, values,
knowledge, and behavior towards nature and economic
conditions within the watershed. The process of select-
ing sub-watershed study sites also has required us to
begin examining in detail how to reconcile differences
between physical watershed boundaries and political or
other human social boundaries.
-------
Social and Ecological Transferability
of Integrated Ecological Assessment Models
Linda A. Deegan
The Ecosystems Center, The Marine Biological Laboratory, Woods Hole, MA
James Kremer
University of Connecticut, Avery Point, CT
Thomas Webler
Social and Environmental Research Institute, Leverett, MA
The coastal zone, near-shore waters and adjacent
uplands, is heavily used for a variety of purposes—not
all of which are compatible. In past years, urban and
suburban development in the uplands have been ac-
companied by profound changes in ecosystems with
declines in fish and shellfish harvests. However, people
are reluctant to believe mat what occurs in their back-
yards can affect the health of the sea. This project will
develop a general integrated ecological assessment
model for coastal watersheds. This model will be used
to determine how the level of involvement with the
model by managers and planners affects their acceptance
of it as a useful management tool. Our project will
determine how best to make the model relevant and
useful to local citizens in their decisions regarding land
use management in the coastal zone. Two interrelated
issues will be addressed: (1) the ecological transfer-
ability of the model, the model's ability to capture the
dynamics of a previously unconsidered ecosystem; and
(2) the social transferability of the model, the
willingness of local managers and planners to use the
integrated planning model.
An existing watershed-loading model will be used,
and an estuary model will be extended to include a new
and socially important management endpoint—fish and
shellfish. This new integrated assessment model will be
used in a series of experiments designed to test
simultaneously the ecological and social transferability of
the model in a variety of new settings. Empirical
relationships will be developed between predicted N-load
and endpoint responses, using new field data and
literature information. This project's research involves
using replicated experiments to determine the effect of
familiarity with and involvement in the model on its
acceptance and usefulness by the planning community.
The effect of four treatment conditions on model
acceptance will be tested (see Figure 1): (1) explanation
of the model by an "expert," (2) participation in a
learning workshop ("Wkshp I"); (3) a learning workshop
supplemented with intensive dialogue ("Wkshp II"); and
(4) a learning workshop supplemented with a citizen
field-data collection program ("Wkshp + Field").
There is an urgent need for an estuarine
ecological model of broad application, incorporating
components that are of direct concern to the public. It
is not possible to build "from scratch" an integrated
ecological model for each community using local re-
searchers and local data. Most communities do not
have the local researchers nor the data needed, and
society cannot afford the money nor the time to fund
de novo model development for each local case.
Therefore, planners must rely on models developed
elsewhere to be applied to their local situation. It is
common knowledge, however, that people distrust
policy recommendations for their community when
they are based on data gathered elsewhere. One of the
most interesting results will be to explain how much
gain in model acceptance is achieved by differing
levels of training and familiarity building. Conclusions
will be drawn about whether the existing models
provide information that community policymakers find
useful and relevant. This research project is designed
to benefit urban/suburban coastal communities interest-
ed in protecting estuarine ecosystems from nitrogen
loading. Integrated ecological assessment models have
the potential to improve the competence of local
communities in making policy decisions. This research
will result in an integrated ecological model of the
consequences of coastal land use change on estuarine
systems and, perhaps more importantly, better infor-
mation on how to apply that model to new environ-
mental and social settings.
-------
CD
0
c
CL
0)
O
O
•u
o
Expert Wkshp I Wkshp II Wkshp + Field
increasing interaction
Treatment
Figure 1. Hypothetical response of subjects to increasing intensity of interaction with the model. Dependent variable is tine
degree of acceptance of the model as a useful planning tool.
-------
From Landscapes to Waterscapes:
An Integrating Framework for Urbanizing Watersheds
Panos Diplas, W.E. Cox, D.F. Kibler, V.K. Lohani, and R.G. Greene'; D.J. Bosch, L.A. Shabman, and K.
Stephenson 2; E.F. Benfield, and P.S. Nagarkatti3; S. Mostaghimi *; and D.J. Orth 5
' Department of Civil Engineering;2 Department of Agricultural and Applied Economics;3 Department of
Biology;4 Department of Biological Systems Engineering;5 Department of Fisheries and Wildlife Sciences,
Virginia Polytechnic Institute and State University, Blacksburg, VA
A generally adverse relationship between urban-
ization and water resources conditions is normally
assumed. However, the variability of impacts from
alternative landscape scenarios on water supply, flood-
ing, and the ecological status of waters within a water-
shed subject to urbanization/suburbanization is not well
understood. Approaches and methods for analyzing
cause-and-effect relationships from a comprehensive
perspective are not widely available.
Land settlements within urbanizing areas take
place in a socioeconomic framework within which
landowners attempt to maximize returns associated with
land consistent with public policies on such issues as
taxation, density of development, wastewater disposal,
and drainage. These constraints have increasingly been
applied on the basis of hydrologic units such as
watersheds. However, watershed management has tend-
ed to consist of a collection of disparate measures that
focus on specific aspects of development in relative
isolation and sometimes without adequate evaluation of
unintended side effects. For example, measures to
protect surface waters from erosion and sedimentation
may result in additional aquifer recharge and an increase
in chemical content of runoff, which ultimately may
affect not only use of groundwater but also use of the
surface water into which the groundwater discharges.
This situation creates the potential for ineffective
management actions and wasted resources as individual
control measures are implemented without evaluation of
their interactions or cumulative effects. The lack of
methodology for a comprehensive assessment of water-
shed management strategies is a major obstacle to
assessment of policy effectiveness and adoption of a
more holistic approach to watershed management.
The goal of this project is to develop pro-
cedures for integrated assessment of the hydrologic,
ecological, and economic consequences of alternative
landscape scenarios that occur during the urban-
ization/suburbanization processes. More specific
objectives include: (1) development of an integrated
hydrologic framework for assessing impacts of
alternative land-scapes on surface and subsurface
water flows and movement of sediments and pol-
lutants; (2) development of procedures to predict the
response of fish and macroinvertebrate communities
to urbanization-induced changes in water quantity,
water quality, and other biological conditions; and (3)
identification and assessment of policy and economic
conditions consistent with alternative landscape
scenarios as well as estimation of the effects of
alternative landscapes on land values and fiscal
consequences for local governments.
A case study focusing on the Upper Roanoke
River Watershed will be employed to test the operation
of the system of linked models to be developed. This
study area is shown in Figure 1. This project will use
a stakeholders' panel from the case study area to
ensure relevancy to actual community watershed
management.
The project will produce a set of linked models
incorporated into a general methodology for holistic
watershed assessment. Figure 2 illustrates the
methodology that will be used as the basis for this
assessment. The creation of this analytical capability
will facilitate the development of policy likely to be
effective in achieving social objectives by allowing an
evaluation of consequences during policy devel-
opment.
-------
Washington, D C.
Figure 1. Upper Roanoke River watershed study area.
Initial Conditions in Watershed Study Area:
Alternative Landscape Scenarios Modeled
Land Use Change
Urban Rural
Change In Population Change in Land Cover
Change in Settlement Patterns Change in BMP Practices
Change in BMP Practices Change in Animal Density
Policy & Economic
Conditions Consistent With
Land Use Change
Policy, Preferences, &
Changes in Hydrologic Parameters
Hydrology & Hydraulic Submodels
Output Indicators Display: User Friendly Computer Interface
Aquatic Conditions
Biolocial Impact Assessments
Biomarker Species
Flood Hazards
Change in Flood Risk
Stream Quality
Economic Conditions
Land Price Changes
Net Fiscal Returns
Figure 2. Study components and integration plan.
-------
An Integrated Ecological and Socioeconomic Approach To Evaluating
and Reducing Agricultural Impacts on Upper Mississippi River Watersheds
Prasanna H. Gowda and R.J. Haroa
University of Wisconsin-La Crosse, La Crosse, WI
A.D. Ward and T.L. Napier
The Ohio State University, Columbus, OH
Nutrient-enriched waters from the Mississippi
River are one of the main causes of hypoxic zones in the
Gulf of Mexico and adversely impact aquatic
ecosystems within the basin. Nutrient loadings in the
Mississippi River have been associated with the
production of a substantial portion of the Nation's corn,
soybean, and wheat, but the specifics of these links
remain unclear. Our research will: (1) use a process
model to predict agricultural discharges from two
watersheds in the Upper Mississippi River basin (see
Figure 1); (2) use observed and predicted runoff, sed-
iment, and nutrient loadings to estimate the contribution
of urban areas, wastewater treatment plants, and point
sources to the total loading; (3) evaluate potential water
quality benefits associated with the adoption of
alternative management strategies on these watersheds;
(4) develop regional-scale predictive models of
ecosystem health, biodiversity, and sustainability by
relating information on biota and ecosystem processes
to current and potential landscape composition and
structure; and (5) identify factors that affect the adoption
of conservation production systems among landowners
in the two watersheds. Our research focuses on the
Lower Minnesota River watershed in eastern Minnesota
and the Maquoketa River watershed in northeastern
Iowa (see Figure 2). A spatial-process watershed
modeling approach that incorporates a field scale model
will be used for predicting daily hydrologic and water
quality responses. Landsat Thematic Mapper data,
Natural Resources Conservation Service's soils data-
bases, published farming system information,
topographic data, and historic climatic data will be used
in conjunction with Geographic Information Systems
(GIS) software to delineate watershed and hydrologic
units to be used as inputs to the model. Available water
quality data will be used to determine flow attributes
associated with point discharges and nonagricultural
activities.
Potential alternative farming systems will be
developed based on interactions with stakeholders in
each watershed. Process-model estimates will be made
of potential reductions in environmental effects
associated with the adoption of alternative farming
systems. Data on macroinvertebrate communities in the
study watersheds will be collected and analyzed in the
laboratory. Statistical relationships between the habitat,
biotic community structure, and land use activities will
be developed and used to predict expected regional
attainment for management plans. Finally, the project
will evaluate socioeconomic factors that have influenced
the success of past soil and water protection initiatives
within the watersheds. Socioeconomic characteristics
of the farm operator and the farm enterprise will be
used to develop statistical models to predict the
following: (1) adoption of specific soil and water
conservation production systems; (2) agricultural pro-
ductivity; and (3) propensities to adopt innovative
conservation farming systems. A survey will interview
750 landowner-operators in each watershed. Socio-
economic, ecological, and water quality data will be
analyzed by using a combination of parametric and
nonparametric statistics. It is anticipated that factors
affecting the adoption of conservation behaviors at the
farm level will be identified. This will make it possible
to implement more effective technology transfer
programs within the region and provide a tool to assess
the environmental benefits of adopting alternative
management practices.
-------
Figure 1. An infrared aerial photograph (acquired on September 12, 1989) showing movement of sediments from Maquoketa
River watershed into Pool 13 of the Upper Mississippi River.
b. Lower Minnesota River watershed in Minnesota
a. Upper Mississippi River Basin c. Maquoketa River watershed in Iowa
Figure 2. A map showing the location of Lower Minnesota River and Maquoketa River watersheds in the Upper Mississippi
River basin.
-------
Integrated Ecological-Economic Modeling
of Watersheds and Estuaries at Multiple Scales
Charles Hopkinson, Edward Rastetter, and Joseph Vallino
The Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA
C.J. Cleveland
Center for Energy and Environmental Studies, Boston University, Boston, MA
IPSWICH RIVER SYSTEM
& ASSOCIATED WETLANDS
IMMOBILIZATION
SEDIMENTATION
DENITRIFICATION
RIPARIAN FOREST
Figure 1. Annual budget of N loading, uptake, and export for the Ipswich River drainage basin (metric tonnes). Objectives of the research
program are to develop spatially explicit models of water and nutrient transport and fate in the Ipswich River watershed.
The consequences of large-scale, long-term
changes within watersheds on estuaries are poorly
understood. Integrated assessments of ecological im-
pacts of economic development in coastal watersheds
will require interdisciplinary approaches. Ecological-
economic models must be integrated into biogeo-
chemical and landscape models to realistically assess
human impacts on watersheds and estuaries. The
objectives of this project are to: (1) integrate socio-
economic models with spatially explicit land use, bio-
geochemistry, and watershed hydrology models to
provide comprehensive assessments of human activities
on nutrient, sediment, and water fluxes in a -500 km2
coastal watershed; and (2) to develop aggregation
schemes that will allow the linked models to be applied
to other watersheds.
The Ipswich River watershed in the Boston
metropolitan area will serve as the experimental water-
shed. Rapid economic and land use changes in this
region, plus ongoing research and an extensive data-
base, make the Ipswich River drainage basin particularly
appropriate for this research project (see Figure 1).
To assess human impacts on watershed and
estuarine dynamics, four types of models will be used:
(1) socioeconomic models, (2) spatial land use change
models, (3) terrestrial hydrologic models, and (4)
stream hydrologic/biogeochemical models. A goal of
our research is to develop a means of assessing the
impacts of urbanization and land use change on
watersheds and estuaries in general, not just a specific
assessment for the Ipswich River. Pending successful
completion of Ipswich River watershed modeling,
aggregation schemes will be developed that will allow
for applying our linked models to larger watersheds or
watersheds with less intensive input data than ours.
Aggregation procedures can ensure that our approach is
transferable. Our approach will enable us to better
assess human impacts on watersheds and to evaluate
management strategies for rninirnizing or reversing
watershed, riverine, and estuarine degradation.
The following scenarios will be evaluated
following the development and linkage of models: (1)
How do the magnitudes and temporal patterns of water,
sediment, organic matter, and nutrient loadings differ
among subcatchments with different land use patterns,
and how are loadings processed within the river before
being discharged to the coastal zone? (2) What effects
do the spatial patterns of land use and river habitats
have on magnitudes and temporal patterns of loadings
to the mouth? How did historical land use and economic
activities affect the magnitudes and temporal patterns of
estuarine loadings? (3) What changes in watershed land
use can be expected based on projected economic
activity, and what will be the impact on water yield and
material loadings?
Stakeholder involvement is an integral part of our
research. Our research will be implemented with the
involvement of three stakeholder groups in the Ipswich
River watershed. Collaboration with these groups will
facilitate the collection of data critical to parameterizing
hydrologic and biogeochemical models of the
watershed.
10
-------
Linking Watershed-Scale Indicators of Changes in
Atmospheric Deposition to Regional Response Patterns
Jeffrey S. Kahl \ 1. Fernandez 2, D. Mageean 3, S. Ballard 3, S. Norton 4, J. Cosby 5, P. Ludwig 6, and
L. Rustad 2 7
1 Water Research Institute; 2 Department of Applied Ecology and Environmental Sciences; 3 Smith Center
for Public Policy; 4 Department of Geological Sciences, University of Maine, Orono, ME
5 Department of Environmental Sciences, University of Virginia, Charlottesville, VA
6 Champion International, Inc., Bucksport, ME
7 U. S. Forest Service, Durham, NH
o
z
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
Figure 1. Bear Brook Watershed in Maine (BBWM) stream time series.
The possibility of negative impacts from acidic
deposition on aquatic ecosystems was widely recognized
in the United States less than 20 years ago. Methods to
assess the status of aquatic systems have been available
for only about 10 years, and the models to evaluate this
assessment are still being modified and tested. In this
project, a specific ecosystem response to experimental
watershed acidification is being determined, and this
knowledge is being scaled to a regional level to
determine the extent of acidification and N-saturation in
a sensitive subpopulation of high-elevation lakes.
Concurrently, mechanisms are being developed to
involve state, federal, and industry resource managers
in the process of using these results in management and
policy.
The project activities are hierarchical from site
specific to regional. First, the indicators of, and model
predictions for, acidification and N-saturation at the
experimental site will be examined at the Bear Brook
Watershed in Maine (BBWM). This site has had an 8-
year experimental treatment of dry ammonium sulfate.
The model of acidification of groundwater in catch-
ments, and its nitrogen sequels, will predict acidification
and responses as well as determine where BBWM is
located on the response continuum. From new and on-
going data collection on soils, stream chemistry, and
forest growth, the response indicators will be assessed
at this site. Second, the site-intensive information will
be scaled to the region using data from high elevation
lakes in Maine and their watersheds. These lakes had
the highest concentrations of nitrate of any known lake
population in the northeastern United States in the
1980's. Their acidification status was similar to lakes
in the Adirondacks. Resampling of these lakes will
occur in 1997 and 1998 to determine lake-chemistry
changes during the past decade. Using mechanisms
defined at BBWM, parallels will be determined between
Bear Brook and the high elevation lakes. Then the
focus will be on transferring new information about
changes in water quality and forest productivity to the
state, federal, and industry contacts. A key component
is the early and ongoing involvement of these
stakeholders in a two-way communication process.
At BBWM, dry ammonium sulfate has been
applied bimonthly since 1989 to the 10.2 ha. experi-
mental catchment, with an adjacent 10.9 ha. catchment
serving as an untreated reference. This experimental
loading is 1,800 eq/ha/yr, vs. 600 eq/ha/yr ambient wet
+ dry N deposition. In the experimental watershed,
nitrate concentrations increased from a volume-weighted
mean of 26 ^eq/L in 1989 to 78 ^eq/L by 1996 (see
Figure 1). The yearly maximum increased from 40 to
106 jueq/L. Nitrate concentrations in the reference
stream decreased from an average of 23 to 3 /ieq/L, and
the yearly maximum decreased from 43 to 11 fteq/L,
reflecting a regional decrease in nitrate concentrations
and flux.
This information will be used by the EPA to meet
the congressional mandate in the Clean Air Act
Amendments of ascertaining trends in ecological re-
sponse and determining the effectiveness of the Act in
influencing these trends. Site-specific data from Bear
Brook scaled to the regional high elevation lakes also
will provide a template for the recognition and
understanding of possible N-saturation and base cation
depletion. This information will be evaluated for use in
management and policy decisions by industry, and at the
local, state, and federal levels.
11
-------
Comprehensive Watershed Management:
A Spatial Water Quality Assessment System
C. Gregory Knight
Department of Geography and Center for Integrated Regional Assessment, Pennsylvania State University, University
Park, PA
Robert P. Brooks, Barry M. Evans, James M. Hamlett, Archie J. McDonnell, and Gary W. Petersen
Environmental Resources Research Institute, Pennsylvania State University, University Park, PA
Todor N. Hristov
Institute of Water Problems, Bulgarian Academy of Sciences, Sofia, Bulgaria
The geography of natural resources, land use,
and human economic activity influence the nature and
spatial distribution of water quality conditions and the
health of many aquatic ecosystems. The goal of this
project is to create an approach to achieve multiple goals
for water quality improvement from both physical and
biological viewpoints in river reaches that are subject to
a variety of geographical, social, and economic factors.
Basic to achieving this complex goal is the ability to
answer two kinds of questions: (1) if we invest in pollu-
tion control at one location, where and to what degree
will the stream quality goals be achieved? and (2)
vice-versa, to attain particular quality standards for a
given reach for a basin, what alternative strategies could
be implemented under various stream-flow scenarios?
Thus, integrated approaches to water quality improve-
ment must be linked to available technological interven-
tions, to spatial and temporal patterns of implementation
and to social and economic dimensions of community
decisionmaking.
This project is a collaboration between the
Institute of Water Problems in the Bulgarian Academy
of Sciences and the Pennsylvania State University,
Environmental Resource Research Institute and Center
for Integrated Regional Assessment, using the Yantra
River Basin (see Figure 1) as a study site. This basin,
the location of Bulgaria's first river basin planning
council, is among the major Bulgarian contributors to
Danube pollution as a result of its long history of
industry and its farming and livestock production.
This work will bring together geographic infor-
mation systems (GIS), water quality models, and
decision-support software for simulating the impacts of
community decisions about improving water quality.
GIS will assist in the development of model input for
estimating nonpoint source pollution where directly
measured data are not available and will bring together
such key variables contributing to water pollution and
remediation as population centers, settlements, industry,
soil types, land use, water use, stream hydrology, treat-
ment facilities, and biochemical and aquatic indicator
measures. GIS also will help decisionmakers visualize
the impacts of alternative community decisions about
priorities for pollution abatement.
This project's approach has been to link new and
existing water quality model components in a holistic, and
transferable, GIS-based water quality assessment system.
Bulgarian colleagues will contribute important components
of the research beyond empirical data for model
calibration, including work with aquatic indicators,
experience in modeling multireservoir hydrologic systems,
multicriteria decision modeling, and interaction with local
resource managers. The initial review of decision support
software suggests that a basin managment system
developed at the International Institute for Applied Systems
Analysis (Austria) can be coupled with existing models
used in the United States and other new approaches. This
project will develop scenarios based on a finite combination
of economic and climatic conditions, which are
representative of past and future conditions in the basin.
12
-------
/\,Vratsa O <-^- }•
S s—fy I.ovech
YANTRA RIVER BASIN
BULGARIA
Figure 1. The Yantra River Basin, a Danube sub-basin in Bulgaria
13
-------
Development and Implementation of Decision Support
Systems for Predicting Economic and Ecologic Impacts of
Alternative Land and Water Management Policies in Urbanizing Regions
Daniel P. Loucks ', Tammo S. Steenhuis 2, Mark B. Bain 3, Warren Brown 4, and Walter R. Lynn 5
'Civil and Environmental Engineering; 2Agricultural and Biological Engineering;3Natural Resources;
4'Institute for Social and Economic Research; sCenterfor the Environment, Cornell University, Ithaca, NY
Community planning and environmental boards
often evaluate proposed changes in land and water use.
The aim of this research is to develop and then provide
government agencies with an interactive computer-based
technology for predicting the economic and ecological
impacts of decisions on water and land use policies. To
do this, this project must identify and quantify how
various land use policies impact surface and ground-
water quantity and quality, and how flows and pollutant
concentrations affect aquatic and adjacent terrestrial
systems. Also, the investigators must attempt to predict
how changes in land and water use policies influence
socioeconomic activities, local economic development,
and tax revenues.
Our immediate goals are to improve the ability to
predict the interdependent socioeconomic, terrestrial,
hydrological, and ecological processes that define the
dynamics of land use changes and their effects on water
and associated biota. This capability will be incor-
porated into a data-driven, interactive, microcom-
puter-based decision support system (see Figure 1).
This research is being conducted by a multidisciplinary
team that includes a regional economist, three agri-
cultural-environmental engineers, and an ecologist.
Work is being conducted closely with town and county
planners, and the methods will be tested on a local
watershed. The approach and models being developed
will be applicable to a variety of watersheds similar to
those found in the Northeastern United States.
In the few weeks since this project began, work
on developing the required surface groundwater
simulation models for water quantity and quality has
been conducted. This will be an expanded Interactive
River-Aquifer Simulation (IRAS) model that has been
developed steadily. To assess economic impacts of land
and water use policies and to predict likely scenarios for
future land use changes, this land use and land cover
information is being gathered for the five-county
economic region in which this study watershed is
located. Data from the Census Bureau's Survey of
Population and Housing for 1970, 1980, and 1990; the
New York State Office of Real Property Services' tax
assessment for every land parcel; and USGS' GIRAS
files on land use and land cover are being obtained.
The initial objectives include building a model that
accounts for land converted to residential uses since
1970, and then using it to identify areas in the watershed
having a high probability for future residential develop-
ment under various economic growth scenarios. The
structure of the ecological components of the simulation
model are being developed by expanding a simple
system relating macroinvertebrate and fish species to
predicted habitat and water quality status. New quan-
titative information from the surface groundwater sim-
ulation model will permit much more sophisticated
predictions on aquatic communities than were previously
available.
This project's goal is to develop a model system
that will simulate decision options, policy ramifications,
and environmental enhancement strategies. The model
should be valuable for uses such as designing the spatial
distribution of conservation incentives, selecting re-
quirements for development mitigation measures, com-
paring zoning and siting approaches, and planning
reserves, greenways, and other forms of natural area
designations.
14
-------
MODULE 1
Social &
Economic
MODULE 2
Rainfall -
Runoff
MODULE 5:
Decision
\Support
-------
Effects of Natural and Anthropogenic Processes on Tillamook Bay
and Its Watershed: An Integrated Process Study and Land Use Perspective
James McManus ', Paul Komar', M.S. Jesse Ford2, Courtland Smith ', Debbie L. Colbert', and
Greg Bostrom 3
1 College of Oceanic and Atmospheric Sciences; 2 Department of Fisheries and Wildlife;3 Department of
Anthropology, Oregon State University, Corvallis, OR
Figure 1. Tillamook Bay and its watershed.
Human activities can significantly alter the
physical, chemical, and biological nature of aquatic
environments. Furthermore, community perceptions of
environmental hazards can lead to mitigative actions that
may themselves have their own detrimental impacts on
ecosystems. There is an interplay between social per-
ceptions and societal decisions that can influence the
physical environment, chemical cycles, and biological
communities. An understanding of the problems as-
sociated with the interactions between social and natural
systems requires a multidisciplinary approach.
This project is employing such techniques in
addressing research problems associated with the inter-
play between natural and anthropogenic processes within
the Tillamook Bay watershed. Tillamook Bay is a
drowned-river estuary, formed initially about 9,000
years ago when rising sea level at the end of the Quater-
nary ice age inundated the lower reaches of the Trask,
Wilson, Tillamook, Kilchis, and Miami Rivers, which
currently drain into the bay (see Figure 1). The Bay is
located in the Pacific Northwest, roughly 100 km west
of Portland, Oregon, and approximately 80 km south of
the Columbia River.
Land use practices vary significantly among the
different riverine watersheds. The Trask runs through
the urban center of Tillamook, whereas the Tillamook
River to the south is heavily influenced by dairy farming
activities. The northern-most three rivers experience
progressively less urban and agricultural activities
within their lower watersheds and increase in the
relative proportion of forested lands. Within the frame-
work of examining the relative influence of different
watershed activities on processes occurring within the
Tillamook basin, four specific hypotheses will be tested:
(1) Land use practices have significantly altered the
sedimentary budget of the Tillamook Bay watershed.
(2) Carbon, nutrient, and trace metal cycles are sig-
nificantly different among the five major river systems
feeding the Tillamook estuary because of the different
land use practices. (3) Differences in land use practices
between the Tillamook and Kilchis River sub-basins
have led to quantifiable differences in aquatic ecosystem
health and biotic integrity in the riverine and stream
environments. (4) Local knowledge of watershed act-
ivities and processes are influenced by many social
factors that may result in perceptions quite different
from scientific measures of sedimentation, nutrient
cycles, and aquatic health.
The above hypotheses will be approached with
field and analytical programs designed to yield infor-
mation on spatial and temporal patterns within the
watersheds. This project will attempt to explain the fac-
tors influencing watershed quality in such a way that
plans can be developed to enable residents of the five
river basins to improve the environmental quality of
Tillamook Bay and the basin in which they live.
16
-------
Ecological Risks, Stakeholder Values, and River
Basins: Testing Management Alternatives for the Illinois River
Mark Meo
Science and Public Policy Program, University of Oklahoma, Norman, OK
Baxter Vieux
Civil Engineering and Environmental Science, College of Engineering, University of Oklahoma, Norman, OK
James Sipes
Division of Landscape Architecture, College of Architecture, University of Oklahoma, Norman, OK
Edward T. Sankowski
Department of Philosophy, College of Arts and Sciences, University of Oklahoma, Norman, OK
Robert Lynch
Department of Occupational and Environmental Health, College of Public Health, University of Oklahoma Health
Sciences Center, Oklahoma City, OK
Will Focht
Department of Political Science, College of Arts and Sciences, Oklahoma State University, Stillwater, OK
Keith Willett
Department of Economics, College of Business Administration, Oklahoma State University, Stillwater, OK
Lowell Caneday
College of Education, Oklahoma State University, Stillwater, OK
The Illinois River, one of the most scenic rivers
in Oklahoma, has been the center of political contro-
versy about private property rights and environmental
protection for more than 25 years. The Illinois River
has provided multiple social benefits to the citizens of
Oklahoma through its use for recreation, water and
power supply, flood control, and nutrient removal. Yet,
the inability of different interests to reach agreement on
how to protect the Illinois River watershed has placed its
hydrologic resources at increased risk of long-term
degradation. With the absence of a unique environmen-
tal issue or feature to catalyze political support for
policy change, the Illinois River basin exemplifies the
challenge to sustainable river basin management.
This 3-year interdisciplinary research project
demonstrates how different environmental and social
values held by river basin stakeholders can be identified
and compared so that more effective environmental
protection strategies can be determined and adopted by
local land and water use interests and state agencies.
The investigators will develop and test an integrated
impact assessment management protocol for the Illinois
River watershed by linking together the ecological,
economic, hydrological, social, and political aspects of
the watershed in an interdisciplinary approach that
provides a more realistic framework for calculating,
communicating, and negotiating environmental risks and
competing social values (see Figure 1). In the first 2
years of the project, the research team will: (1) identify
stakeholder views on the Illinois River basin; (2)
determine the economic effects of alternative land and
water uses for three study sites in the river basin; (3)
calculate the ecological risks associated with different
intensities of resource use; (4) develop hydrologic
models using Geographic Information Systems (GIS)
that incorporate water quality aspects of alternative land
use practices; and (5) develop computer-generated im-
agery of each of the three sites that will enable stake-
holders to visualize more easily the implications of
different management options for the river basin's re-
sources. At the same time, members of the research
team will be investigating stakeholders' perspectives of
natural, economic, and sociopolitical impacts through
interviews and focus group sessions. These groups will
include technical experts, lay stakeholders, and policy-
makers. Background data will be drawn from the
investigators' prior studies of the Illinois River water-
shed, its carrying capacity, and comprehensive land use
plans.
In the third year of the project, stakeholder groups
will be engaged in a policy dialogue and a test of the
effectiveness of integrated computer models to facilitate
the risk communication of complex environmental
management issues. Visual simulations developed from
CIS-based hydrological models will be shown to stake-
holders in conjunction with focus group sessions to
ascertain management preferences and the overall legiti-
macy of negotiated agreements. Negotiation workshops
will be held to develop a consensus about land use prac-
tices that afford an adequate level of protection to the
basin. The entire process will be tested to determine the
degree to which the process is viewed by experts and lay
stakeholders as efficient, effective, and legitimate, and
therefore acceptable.
17
-------
Year 1: Baseline Assessment
Year 2: Alternatives Evaluation
Modeling N1A
regional policy
Modeling EIA
regional policy
alternatives
1
\
Year 3: Policy Recommendation
xU
Prepare
visualizations
\
/
Meeting of lay
stakeholders and
researchers
\
/
Refine policv
alternatives
alternatives
1
\
^
Modeling of EIA
basin wide
policy alternatives
Phone survey for
larger evaluation
Figure 1. Illinois River Project: Research Plan. Ecological risks, stakeholder values, and river basins: testing management
alternatives for the Illinois River.
18
-------
Risk-Based Urban Watershed
Management: Integration of Water Quality and Flood Control Objectives
Vladimir Novotny, Robert Griffin, David Clark, and Douglas Booth
Marquette University, Milwaukee, WI
Figure 1. The Milwaukee River is being restored. The picture shows a site that was previously an urban reservoir with very poor water quality.
The dam that has created the reservoir is being removed, contaminated exposed sediments are either removed or capped and vegetated,
stream banks are being reinforced by bioengineering, and the channel is restored to provide a natural habitat to fish, including
migrating salmon from Lake Michigan.
Urbanization is known to increase flooding
potential and to impair the ecological integrity of water
bodies. In the past, however, most urban drainage pro-
jects focused almost solely on flood protection and were
conveyance oriented. This project will develop and test
methodology for watershed management in urban and
urbanizing watersheds. Management is based on quan-
titative risk assessment of urbanization to the watershed
ecosystem, including people residing in the watershed
and affected by the receiving water body and biota
residing in receiving water bodies. An evaluation of the
flood and ecological risks will help to set priorities for
management, find an optimum solution considering both
risks, and provide necessary information to agencies for
funding decisions.
The objectives of this project are to: (1) develop
statistical flow, loading, and water quality models
applicable to risk assessment; (2) develop objective and
quantitative risk assessment procedures for estimating
ecological risks of stormwater and subsurface discharges
from urban and suburban watersheds; (3) develop a
methodology for the assessment of flood control and
water quality benefits and resolving conflicts between
flood control and ecological preservation-restoration
objectives; (4) develop benefit/cost models for urban
watershed management to optimize both flood control
and receiving system ecological integrity; (5) research
innovative financing of urban watershed management,
identify key players, and assess the willingness to pay
for different types of benefits; and (6) examine
homeowners' risk/benefit perceptions, values, effective
responses to the risk, subjective norms, sociocultural
backgrounds, and use of communication in the willing-
ness to pay for these different types of benefits.
Figure 1 shows the concept of optimization of
flooding and ecological risks in watershed management.
Ecological risk assessment quantitatively enumerates the
hazards of pollution discharges and physical alteration
of urban water bodies and their hydrology/hydraulics to
the biota residing in the water body and riparian eco-
systems. A similar notion is applied to flood risk that
denotes a probability of damage to the water body itself,
to properties in the flood plain, and hazards to human
health and well being.
Presently, no ecological standards are available
for urban drainage and water body restoration projects,
and most projects are driven by flood protection ob-
jectives, leading to conveyance-oriented modification of
the urban (suburban) water bodies. Conveyance solu-
tions (such as channel lining and straightening) have
adverse ecological consequences and commonly are
economically inefficient. Today, such projects and
solutions may not be acceptable. Incorporating eco-
logical integrity objectives may lead to storage-oriented
approaches and restoration of ecological integrity of
urban receiving water bodies.
The methodologies developed in this project will
be tested on two pilot watersheds in Milwaukee County.
One watershed is fully urbanized and is found in the
City of Milwaukee. The other pilot watershed is in a
suburban part of the County that is undergoing rapid
urbanization.
19
-------
Impact of Social Systems on Ecology and Hydrology
in Urban-Rural Watersheds; Integration for Restoration
Steward T.A. Pickett
Institute for Ecosystem Studies, Millbrook, NY
Policymakers, planners, and managers need to
understand the biophysical and social processes con-
trolling water yield and quality as well as aquatic
biodiversity in human-dominated landscapes. This
requires an integrated watershed model, which can
address three types of social interventions for ameliorat-
ing water quality: Type 1— reducing point sources of
pollution at the "end of the pipe;" Type 2—regulating
land use to reduce nonpoint inputs; and Type 3—
modifying sociocultural processes within different types
of land uses to minimize the polluting behavior of
individuals, groups, and institutions. Types 2 and 3 are
underutilized.
This research, to be initiated in late 1997, will
develop a model to test whether the sociocultural factors
involved in Types 2 and 3 control water quality,
watershed dynamics, and stream biota of watersheds
that extend across urban through rural land uses.
Specifically, the model will test to what degree
sociocultural factors affect water quality, watershed
dynamics, and stream biota indirectly by modifying
ecological processes, or directly by modifying people's
actions. Indirect sociocultural effects include clearing,
planting, or managing forest patches that can affect the
extent, distribution, and structure of vegetation cover.
Direct sociocultural effects on ecological processes
include inputs of solids and nutrients from dumping or
runoff from fertilizers and pesticides. The research will
be conducted in the 17,150 ha Gywnns Falls watershed
in the Baltimore region of the Chesapeake Bay drain-
age.
An integrated approach is needed for two
reasons. First, the U.S. EPA has mandated a 40 percent
reduction in nitrogen loading to the Chesapeake Bay.
Controlling point sources of pollution (Type 1) has not
met this mandate. Second, forest conversion and sub-
urbanization are proceeding rapidly (Type 2), leaving
few options to mitigate water quality. Thus, a model of
the interaction between the ecological and physical
patterns and processes in the landscape with socio-
cultural patterns and processes (Type 3) is a potential
decisionmaking tool to identify and evaluate social and
ecological interventions for improving water quality in
the Chesapeake region.
20
-------
LAND COVER-ECOLOGICAL
HIERARCHY
Land Cover Type
Patch Configuration
Soil
Permeability Class
SOCIO-POLITICAL
HIERARCHY
Political Units
Land Use
Neighborhood
Household
WATERSHED
HYDRO-ECOLOGICAL
HIERARCHY
Watershed
Sub-Watershed
Sub-Catchment
HUlslope
Floodplain
Channel
Figure 1. The three hierarchies of patch or land units to be integrated in the proposed research. Sociopolitical units are those defined by
political or administrative entities. Watershed hydroecological units are those delimited by decomposing large watersheds into smaller
units that contribute runoff. The land cover-ecological hierarchy describes the specific surface features that govern the flow rate and
pathway of runoff and the degree of saturation.
21
-------
Section 2.
Projects Initiated With Fiscal Year 1996 Support
-------
Strategic Renewal of Large
Floodplain Rivers; A Preliminary Status Report
John B. Braden ', Misganaw Demissie 2, Eric DeVuyst', Paromita Mitra, Daniel Schneider ''3,
Richard E. Sparks '-3, David C. White ', and Renjie Xia 2
'University of Illinois at Urbana-Champaign, Urbana, IL
2 Illinois State Water Survey, Champaign, IL
3 Illinois State Natural History Survey, Champaign, IL
136
O
_Q
rt 134
Pre-dam (1887)
natural dry
season
100 200
Day of Ycnr
Figure 1. Effects of Illinois River dams.
300
This project is investigating the potential for
selective restoration of the hydrology and ecology of
floodplains. Interrelated ecological, hydrologic, and
economic models are being developed to compare the
effects of current floodplain management versus
ecosystem-based alternatives. Specific objectives are:
(1) to estimate thresholds at which floodplain restoration
measures can have significant impacts; (2) to identify
combinations of structural changes and floodplain
restoration efforts that enhance sustainability; and (3) to
identify institutional obstacles to policies that might
otherwise balance the costs and benefits of floodplain
restoration. Our project uses the LaGrange Reach of
the Illinois River as a study site. Figure 1 illustrates the
impact that current river management has had on river
and floodplain hydrology.
The hydrology of the Illinois River varies on time
scales of days, seasons, and decades. The role of varia-
bility is being examined at each of these time scales on
the ecological processes in the floodplain and the
resulting economic activities. Ways to increase the
overall economic value of the river using ecological and
hydrologic restoration are being sought.
Project hydrologists are combining one- and
two-dimensional hydraulic models to compute flow
fluctuations in three critical time scales: (1) short-
term changes related to navigation needs, (2) seasonal
patterns, and (3) long-term extremes (both floods and
droughts). For selected restoration sites, intensive
modeling using the two-dimensional model will allow
prediction of lateral-flow velocities and sedimentation
processes that influence biotic regimes. To date, the
one-dimensional model has been run under selected
high flow conditions to test its sensitivity and to
simulate the removal of levees. The analysis will be
extended to low flow conditions. The model results
appear highly sensitive to Manning's roughness co-
efficient.
Project ecologists are creating dynamic models of
floodplain vegetation, for both forested and nonforested
regions. A critical feature of the models has been to
incorporate daily hydrological variation rather than the
average annual conditions used in existing floodplain
forest models. These dynamic ecosystem simulators
interact with complex water regimes provided by the
hydraulic model and, in turn, generate input for static
habitat suitability models for floodplain animal species.
More than a century of river-stage records, floodplain
forest descriptions from early 19th century surveyor
notes, and information from present-day vegetation
surveys have been used to calibrate the forest model.
The model has demonstrated that hydrological variation
at long time scales is necessary to maintain floodplain
forest diversity. The next steps include incorporating
the nonforest vegetation model and linking the dynamic
models to habitat suitability.
Project economists are developing a model to
estimate the impacts of restoration on the local econ-
omy. This includes collecting, organizing, and assess-
ing data of the various economic elements (e.g., farm
production budgets, crop insurance, recreation values,
flood damage records and estimates, transportation
values, institutional influences, etc.). A limited land use
allocation model has been developed and used to assess
how agricultural land use might change as levee heights
are reduced. Agricultural land use values will be
supplemented with survey data on recreational values to
permit a more complete analysis of social values
associated with floodplain land uses. This project has
emphasized improving the information flows between
disciplines and model components, perhaps through
expanded use of geographic information systems.
25
-------
Integrating Modeling and Management of Agriculturally
Impacted Watersheds: Issues of Spatial and Temporal Scale
Patrick L. Brezonik', K. William Easter2, David Mulla3, and James A. Perry 4
1 Water Resources Center and Department of Civil Engineering;2 Department of Agriculture and Applied
Economics; 3 Department of Soil, Water and Climate;4 Department of Forest Resources, University of
Minnesota, St. Paul, MN
Scientists and managers concerned with water-
shed management are faced with a dilemma: Scientists
are developing models that explicitly address cumulative
effects, large spatial scales, and longer timeframes, but
societal policy shifts are forcing managers to make
decisions on an increasingly more local basis. Effective
implementation of watershed management principles
requires models that resolve this dilemma (e.g., models
that incorporate knowledge about effects at longer
temporal scales and larger spatial scales into more local-
ized decisionmaking). The goals of this project are to
develop a more quantitative understanding of the
implications that this dilemma poses to aquatic science
and watershed management and to develop solutions for
resolving the dilemma.
This project focuses on the Minnesota River
basin, which occupies much of southern Minnesota. The
basin presents great opportunities for integrative
research to advance watershed science as well as
contribute to solving a major environmental problem.
The basin drains almost 40,000 km2, including 34,000
km2 of intensively fanned land, and it is considered one
of the most polluted rivers in the country. High
concentrations of nitrogen, phosphorus, pesticides,
bacteria, and sediment, mostly from diffuse sources,
cause degraded conditions throughout the river and
many of its tributaries. In turn, the river acts as a major
source for these pollutants to the Mississippi River, into
which it flows at the Twin Cities.
A state-initiated program, the Minnesota River
Assessment Program, studied water quality in the basin
and set improvement objectives during the period of
1989-1993. This program led to a phase in which
comprehensive basin plans are being developed and
implemented. These plans include Best Management
Practices (BMPs) for farm management, wildlife and
habitat protection, sewage treatment facilities, riparian
zone and wetland restoration, and pesticide/nutrient
management. However, these plans rest on an in-
complete knowledge base. Barriers to implementation
include incomplete knowledge about sources and the
effectiveness of proposed BMPs. This project is ad-
dressing these needs with one overarching objective: to
improve the understanding of how biophysical and
socioeconomic variables interact in agricultural water-
sheds of varying scales, landscape conditions, and land
use management practices to affect the export of nu-
trients and their impacts on instream biological com-
munities, and in turn, to assess the role of knowledge
about those effects in decisionmaking processes on local
level land use.
We are conducting the research in three coor-
dinated phases, respectively dealing with landscape
processes, aquatic processes, and socioeconomic issues.
Biophysical phases of the project encompass spatial
scales ranging from the whole drainage basin to several
scales of watersheds. In particular, we are evaluating:
(1) ways to account for scale effects in hydrologic and
nutrient transport modeling within watersheds; (2) the
effectiveness of agro-ecosystems as complementary
landscape units (to watersheds) for analysis of nutrient
export from the land (see Figure 1); (3) the scales at
which aquatic ecosystems process and retain nutrients
and how they influence rates of stream recovery when
land management practices are improved; and (4) ways
to identify priorities for controlling diffuse sources of
sediments and phosphorus based on technical, social,
and economic feasibility. All of these are critical for the
protection and enhancement of aquatic ecosystems by
integrated watershed management.
26
-------
Steeper Cropped Moraine
Steeper Urban Moraine
Flatter Moraine
,J QuUtnli
Olivia Till
Diyer Blue Earth Till
Wetlei Blue Earth Till
Poorly Drained Wetter Silts and Clays
Poorly Drained Dryer Silts and Clays
Figure 1. Map of Minnesota River Basin delineating its 12 major watersheds and 10 agro-ecoregions.
27
-------
Urban Stream Rehabilitation in the Pacific
Northwest: Physical, Biological, and Social Considerations
Stephen J. Surges, Derek B. Booth, Sally Schauman, and James R. Kan-
University of Washington, Seattle, WA
Figure 1. McAllister Creek, Thurston County, WA.
Society is eager to rehabilitate urban streams and
to limit future damage as new areas are urbanized. Any
framework for rehabilitation or protection must be
grounded in understanding the specific factors respon-
sible for degradation, their consequences, and the causal
pathways linking them to human activities. With an im-
proved understanding of what determines stream condi-
tion in urban areas, we can better evaluate prospective
rehabilitation candidates; define realistic rehabilitation
goals; guide the design of successful rehabilitation pro-
jects; and limit damage to high-quality aquatic systems
through improved planning, design, construction, and
management in urban areas.
One of the goals of this project is to document the
consequences of urban development on the physical and
biological condition of urban streams and to demonstrate
specific rehabilitation strategies likely to restore valued
properties of those systems (see Figure 1). Another
goal is to develop successful rehabilitation methods for
urban areas and to test public attitudes and preferences
for rehabilitation measures. The questions for this
project are framed in a process-based, watershed
context: How do landscape processes translate to pat-
terns in channel form and process, and how does
urbanization affect spatial and temporal pattern of
landscape processes? Since April 1997, we have ac-
complished several important tasks: (1) selected sites for
examining processes, elemental changes, and cultural
contexts of stream degradation, benthic macroin-
vertebrates, channel forms, bed sediments, and cultural
landscapes—we have considered 18 sites that span a
gradient from low to high urbanization; (2) evaluated
visual preferences for rehabilitation design and human
behaviors exhibited toward urban riparian areas and
rehabilitation designs; and (3) assessed rehabilitation
project costs and outcomes. These efforts are being
expanded as we continue the development of our
project.
28
-------
Influences of Forest Fragmentation on Watershed
Functions in Northern Vietnam—Preliminary Field Results
Jeffrey Fox,1 T. Giambelluca,2 and A.T. Rambo1
1 Program on Environment, East-West Center, Honolulu, HI
2 Department of Geography, University of Hawaii, Honolulu, HI
It is likely that the atmospheric and hydrologic
effects of deforestation depend on the degree of
fragmentation of the remaining forest. As fragment size
is reduced, a greater fraction of the remaining forest is
in close proximity to a forest edge, where humidity, air
temperature, and wind speed may enhance the
transpiration rate. Forest clearing and subsequent land
use can have major effects on soil hydrologic properties,
including hydraulic conductivity. A fragmented land-
scape will translate into high spatial variability in the
saturated hydraulic conductivity of the soil known as
Ks. As a result, overland flow may be less likely to
reach the stream channel or to attain velocities capable
of initiating erosion or transporting sediment as far as
the stream channel. Together, fragmentation effects on
evaporation and overland flow can alter the hydrologic
process differently than monolithic clearing. Thus far,
one of several planned intensive field experiments has
been conducted to investigate edge effects on trans-
piration. During the first experiment, a transect was
instrumented, crossing both upwind and downwind
forest fragment edges. Along the transect, four re-
cording microclimate stations were created. Also, two
arrays of Granier-type sap flow sensors were installed,
monitoring 10 trees near the forest edge and 5 trees in
the interior of the patch. Preliminary analysis of the sap
flow measurements indicate significantly higher
transpiration in near-edge trees as compared with trees
of the same species located in the patch interior.
To scale up the hydrologic effects from a single
forest fragment to a larger area, knowledge of the larger
area's land cover is needed. Preliminary analysis of the
research area's landscape suggested that the relatively
homogeneous land cover, dominated by forest, present
in 1952 has become much more heterogeneous and
fragmented over time. To test this hypothesis, 1995
Landsat TM imagery was acquired, classified, and
compared with a 1952 air photo interpretation. Because
of the different resolutions of the two data sets,
intensive field work was carried out to adequately
ground truth the Landsat TM image. More than 200
ground truth points were collected as randomly as
possible. A method was devised for using differential
global positioning systems or GPS ( + /- 7 meter
accuracy), and the satellite imagery was registered to the
field data with a resulting root mean square or RMS
error of 0.74 pixels (22.2 meters). A "supervised
classification" (whereby the image analyst "supervises"
the process by choosing the information categories on
classes desired and then selecting training areas that
represent each category) of the imagery resulted in
seven land cover classes. Two accuracy assessments of
the classification were done. The first showed 79 per-
cent of 155 check points correctly classified without
attempting to account for GPS surveying and regis-
tration error. The second attempted to account for GPS
surveying and registration error and resulted in 95
percent of the check points being correctly classified.
A comparison of the Landsat TM classification with the
1952 air photo interpretation suggests that the research
area's land cover is more fragmented and heterogeneous
than in 1952. Today, trees still make up a relatively
large part of the land cover (27 percent), but it is
significantly more fragmented than in 1952 and the
average tree cover fragment size is two-thirds of a
hectare as compared with the air photo in 1952.
Preliminary socioeconomic research results
indicate that the local community manages the
fragmented landscape for different types of bamboo.
The land cover analysis also corroborates this. The
largest land cover type is dominated by bamboo (32
percent), and the average fragment size is 2 hectares.
Future efforts will include continued field work
to investigate edge effects on transpiration in dry and
wet seasons, further ecological surveys to determine the
major tree species that compose the vegetative cover,
and additional socioeconomic studies to determine
factors that cause and maintain forest fragmentation.
29
-------
Figure 1. Forest fragment where we established four recording microclimate stations, Ban Tat, Vietnam.
30
-------
Geochemical, Biological and Economic Effects of
Arsenic and Other Oxyanions on a Mining Impacted Watershed
Glenn C. Miller, Watkins W. Miller, Scott Tyler, and Douglass Shaw
University of Nevada, Reno, NV
Ron Hershey and Lambis Papellis
Desert Research Institute, Reno, NV
Susan Anderson
University of California at Davis-Bodega Marine Laboratory, Davis, CA
Rain/Snow
Evaporation
~
Ground Water
\
Flow In
Turnover
Wall Rock ,
Dissolution
Pit Wall
Sloughing
Chemical
Precipitation
Pit Lake Water
'. \ Flow Out
Newly Oxidized Rock
(from air penetration of ground
water cone of depression)
Figure 1. Processes affecting pit lake chemistry.
More than one-half of the annual production of
gold is mined in the Humboldt River Basin, between
Elko and Winnemucca, Nevada. More than 15 major
precious metals mines are located in this watershed,
including the Nation's two largest mines, which together
produce more than 5 million ounces of gold per year.
The expansion of precious metals mining in the past 15
years has been made possible, in part, by relatively new
techniques of cyanide extraction that have allowed
profitable extraction of gold at very low concentrations.
This expansion of mining has introduced environmental
problems that are poorly understood, both in terms of
their immediate environmental impact and in our ability
to predict long-term changes that may occur as mines
begin to close. Many of these large mines penetrate the
groundwater table. The required pumping of large
amounts of water significantly impacts the groundwater
systems. Arsenic (As) is often associated with gold
deposits, and release of As in soils and surface waters
has become an increasing concern.
This research is focusing on the geochemical,
biological, and economic impacts of As released during
mining in the Humboldt River watershed in northern
Nevada. The objectives are to examine four aspects of
precious metals mining: (1) studies on the evolution of
precious metals pit lakes, (2) long-term drainage from
cyanidization heaps, (3) biological effects of arsenic and
antimony, and (4) economic effects of long-term alter-
ation of the watershed.
The pit lake studies are using a variety of
geochemical models (see Figure 1) to develop a pre-
dictive understanding of an existing 106 meter deep pit
lake in Yerington. The water quality in the pit lake
appears to be significantly affected by dissolution of
wall rock because the equilibrium models underpredict
the contaminant concentrations in the lake. Trace met-
als are particularly difficult to model. Initial laboratory
studies on leached heap material show a large array of
constituents being released from the rock. As expected,
As concentrations in the drainage water are controlled
by sorption to the heap material. Development of math-
ematical models are considering both equilibrium dis-
solution effects for heap drainage water quality and
chemical reactions on the heap material.
Biological research on the effects of arsenite on
water flea (Ceriodaphnia) indicate that brood sizes are
decreased from 9.4 +/- 1.8 to 1.5 +/- 1.1 individuals
when two generations are exposed to arsenite at 1.5
mg/L. These multigenerational studies indicate that ar-
senite may cause long-term genetic damage at sublethal
doses.
Studies on the economic impacts of mining on
the Humboldt River system have been initiated using
five focus groups with interests in the Humboldt
River watershed. Groups consisting of represen-
tatives of the mining industry, the agricultural in-
dustry, the conservation community, and the govern-
ment have provided views on the potential impacts.
Water quantity issues, rather than water quality,
appear to be a dominant concern, and the focus
groups have reflected strongly held views of water
usage in the watershed.
31
-------
Effectiveness of Regulatory Incentives for Sediment Pollution
Prevention: Evaluation Through Policy Analysis and Biomonitoring
Seth R. Reice and Richard N. Andrews
University of North Carolina at Chapel Hill, Chapel Hill, NC
Minimum Disturbed
Area Requiring
Erosion Plan
# Field
Staff
Total Area
(Miles 2)
# Active
Projects
Orange County
Wake County
District 4*
(16 Counties)
0.5 Acres
1 .0 Acres
1 .0 Acres
3
13
5
400
858
8,116
-60
-500
-1000
* District 4 oversees all construction projects in all 16 counties without a Local Erosion and Sediment Control
Program. It covers all governmental construction in the District 4 area, including Orange and Wake Counties. So,
a single stream can have adjacent construction sites administered by different regulatory agencies.
Figure 1. Characteristics of the regulatory jurisdictions studied.
A critical problem in American rivers and
streams is sedimentation. Sedimentation degrades water
quality, alters habitat for fish and macro invertebrates,
limits the ecosystem functions and services, and reduces
the aesthetic and economic value of rivers and streams.
Many regulations and policy incentives have been
devised to control sediment pollution of our rivers. Yet,
there has rarely been an attempt to reconnect the
policies with the ecology of the rivers. That is the goal
of this research. The overall objective is to study me
effectiveness of different policy incentives in the
reduction of die ecological risks and consequences of
sedimentation. The aim is to create more effective
management strategies to provide environmentally
sustainable social and economic development in U.S.
watersheds. This work will integrate the social and
regulatory theory behind sediment ordinances and
policies as well as the resultant ecological impacts of
sedimentation on the rivers and streams. This project's
goal is to discover which policies and regulations really
work to enhance stream biota and ecosystem health and
the reasons why they are or are not effective.
A comparative watershed approach is being used
to contrast the ecological effects of different intensities
of sediment control standards and enforcement. Similar
watersheds are being selected from three political
jurisdictions that differ in the stringency of their
sediment and erosion control requirements, their
staffing, and enforcement (see Figure 1). Field bio-
monitoring data, water chemistry, and leaf litter
decomposition rates are being used to document changes
in stream ecosystems above and below construction
projects under these different regulatory regimes.
North Carolina has one of the strongest state laws
in the Nation for erosion and sedimentation control, and
at least 38 of its county and municipal governments have
enacted even more stringent control ordinances. The
differences among these regulatory regimes, applied to
similar types of socioeconomic activities (e.g.,
commercial and housing construction, road building) on
otherwise similar watersheds, provide opportunities to
draw broader inferences on a comparative basis. An
unusually detailed evaluation of the North Carolina
erosion and sedimentation program and a close working
relationship with the North Carolina Division of Land
Quality provide substantial baseline and supporting
information for this research. This project focuses on
the regulations governing controls on sedimentation into
rivers and streams to determine their effectiveness and
ecological consequences. A direct comparison of dif-
ferences in the programs' regulatory stringency with
their ecological outcomes in the streams will expose
what actually results in the protection of stream eco-
systems. This will determine the actual effectiveness of
these regulations in maintaining stream ecosystem
health, functioning, and biodiversity. In the end, the
answer to the question, "What really works?" will be-
come clear.
32
-------
Watershed Protection in Agricultural Environments:
Integrated Social, Geomorphological, and Ecological
Research To Support Ecosystem-Based Stream Management
Bruce L. Rhoads, Edwin E. Herricks, and David Wilson
University of Illinois at Urbana-Champaign, Urbana, IL
Historical Social/Cultural Context
T T
Values/Attitudes Towards the Environment I
t
/' Predominant
/ Ethic
Legal
/ Economic
/ Available i
\
Context
Context
r Technology
/ Interaction with Environment 1-
/ Scientific
/ Research
Perceptual
Assessment
/ /' Environmental Response I ~~
Scientific Knowledge |
Nonscientists
Local Knowledge I
Figure 1. Conceptual model of interaction between scientists and nonscientists m the social negotiations of community-based resource
management.
The overall goals of this project are to provide an
improved understanding of the roles of scientists and
scientific information in community-based environ-
mental decisionmaking (see Figure 1) and to advance
knowledge of the connections between stream geo-
morphology and aquatic ecology in human-modified
agricultural stream systems. The proposed project has
a four-pronged research design: (1) historical analysis of
the attitudes and values of rural stakeholders toward
water and watersheds in the agricultural Midwest,
(2) social analysis of the mechanisms that facilitate and
impede infusion of new scientific knowledge into local
decisions about stream management in agricultural
watersheds, (3) GIS- and field-based analysis of inter-
relations between geomorphological and ecological
dynamics of human-modified agricultural stream
systems at the watershed and reach scales, and (4) dis-
semination of scientific information from the geo-
morphological and ecological research to farmers and
other local stakeholders.
First-year data collection activities have con-
centrated mainly on the social component of the project.
Fifty-seven in-depth, open-ended interviews have been
conducted with drainage district commissioners,
farmers, and associated stakeholders in the rural
communities of East Central Illinois. The interviews
have focused on three issues: (1) farmer and stakeholder
perception of watersheds as an agricultural, cultural,
and aesthetic resource; (2) current usage and main-
tenance of watersheds; and (3) perceived best and rel-
evant strategies for future usage and maintenance. Pre-
liminary findings indicate that farmers are a surprisingly
heterogeneous group in outlook and perspective and that
important commonalities can be identified. Farmers
value streams not only for the purpose of land drainage,
but also perceive themselves as fundamental stewards of
the land. Future work will explore the interplay of
these contrasting values. Additional open-ended inter-
views will be conducted after scientific/technical infor-
mation is imparted to farmers and stakeholders to gauge
their responses.
Work on the geomorphological component of the
project has emphasized the development of a geo-
referenced grid of control points for digitizing stream-
channel traces and for analyzing changes in these traces
through time. Preliminary analysis indicates that over
the last several decades, humans have been the dominant
agents of geomorphic change in the fluvial systems of
East Central Illinois. Field activities have involved the
selection, preparation, and mapping of seven study sites
for detailed investigations of the fluvial dynamics of
human-modified agricultural streams. During the next
year, geomorphological research will focus on GIS
analysis of stream-channel change and on data collection
at the field sites.
A fish-species data matrix containing critical
habitat and life history information for Illinois fish
species is under development. The matrix will be used
to assist in interpreting habitat fisheries relationships
from field data and will provide a means of specifying
critical habitat conditions when management programs
are being developed. A systematic fisheries collection
program has been initiated at the field sites in co-
ordination with geomorphic data collection activities.
Preliminary sampling suggests that fish-species diversity
is tied strongly to physical habitat diversity.
33
-------
Towards an Integrated Regional Model
of River Basins of the Western Pacific Rim
Jeffrey E. Rickey
University of Washington, Seattle, WA
Evapotranspiration
Soil Texture
Vegetation
Precipitation
Temperature
Population
Runoff
Figure 1. Schematic of integrated regional model for the Western Pacific Rim.
The extraordinary pace of development and
population growth along the Western Pacific Rim has
placed dramatically increasing pressure on river basins
and their downstream coastal ecosystems. Although
decisions about the usage and allocation of water
resources are generally made according to economic and
political criteria, the sound management and optimiza-
tion, and hence sustainable use of these resources, will
require increasingly sophisticated information on the
functioning of the biophysical systems and how they are
affected by socioeconomic and political institutions. To
address these issues, the investigators are building an
"integrated regional model" of the river basins of the
Pacific Rim, with a focus on Southeast Asia. The
model has the objective of coupling hydrological
pathways and biogeochemical indicators with the
interactions and effects of human impact to describe
how materials are mobilized and transported from the
land surface to the coastal zone. The project is ad-
dressing two interrelated questions: (1) The science
question is, "What are the effects of changing land use
(and climate) on the mobilization of water and its
dissolved and paniculate load from the land surface into
the coastal zone at local to regional scales?" (2) The
practical question is, "How is information and training
in a geographically and politically diverse region
mobilized and focused on the problem?"
The project approach is to link large scale
attributes of drainage basins to fundamental bio-
geochemical processes and fluxes through rivers via a
regional-scale routing model (see Figure 1). This ap-
proach is derived from the extensive experience of this
group on the Amazon River. To address the inherent
space and time scaling assumptions that must be made,
the investigators have developed a "drainage basin
element" model (DBE), expressed within the modeling
environment of a global biogeochemical model. Each
DBE is georeferenced within a drainage network and
linked to databases of basin characteristics (e.g.,
elevation, soil texture, and vegetation types). The
immediate requirements are to translate precipitation
into runoff with a water balance model for each DBE in
a manner that allows estimates of residence time and
flow through each element (to "mobilize" the chem-
istry). Selection of the size and temporal resolution of
each DBE is a "scaling" problem, where tradeoffs must
be made between expectations and data availability.
The current expression of the model for the
region is 1 degree (about 100 km), based on readily
available data sets and summed to a monthly composite.
At this scale, the hydrographs of such rivers as the
Mekong and Yangtze are readily reproduced. However,
it is not possible to resolve the dynamics of, for
example, the myriad small rivers of the coastal areas so
typical of the region, nor to address the typical scales of
land use and cover change. Hence, the next phase of the
work will be to establish the model for 1 km elements,
taking advantage of newly available data. An integral
part of this project will be to establish a network of
cooperating institutions across the region, including
science, policy, and economic organizations. Based on
preliminary feedback, this project will contribute to
useable assessments of water resources along the Pacific
Rim.
34
-------
An Integrated Approach To Assessing Water Management
Options in a Major Watershed: Extending a Hydrodynamic-Water
Quality Model To Include Biological and Politico-Economic Components
Paul Sabatier, Loo Botsford, Mike Johnson, Jay Lund, Peter Moyle, Gerald Orlob,
James Quinn, Peter Richerson, Tom Suchanek, and Marca Weinberg
University of California at Davis, Davis, CA
The primary objective of this project is to develop
and demonstrate an integrated methodology for as-
sessing management alternatives for watersheds that
support aquatic species at risk. Such a methodology
requires an interrelated set of deterministic and sta-
tistical models designed to characterize quantitatively the
responses of sensitive species to hydrodynamic, water
quality, and ecological influences within riverine and
estuarine reaches of the watershed under various mana-
gement alternatives. Alternatives may include restor-
ation of riparian habitat, regulation of stream flows,
curtailment or rescheduling of diversions, and control of
point and nonpoint sources of pollution. The project
links not only hydrologic, water quality, sedimentation,
and fishery models, but also the impacts of various
management alternatives on urban and agricultural water
supplies. Finally, the conditions under which such
modeling capability can produce consensus on the
perceived impacts and feasibility of various management
alternatives in situations of political conflict will be
assessed. The models are being developed for the
Sacramento River from Shasta Dam through the
Sacramento-San Joaquin Delta, with the principal
fisheries being the winter run chinook salmon, striped
bass, and Delta smelt.
In the 12 months since funding began, the
project has focused on: (1) developing the hydrologic
and water quality (i.e., temperature, salinity, and toxic
particle) models for 1984, 1992, and 1993; (2)
integrating models of particle and fish movement into
the hydro model, and then estimating winter run and
striped bass survival; (3) continuing long-standing
work on the Clear Lake and Cache Creek watersheds,
particularly with respect to sedimentation and mercury;
(4) gathering the data to build the models relating
water quantity and quality to urban and agricultural
water use; and (5) conducting a mail survey of
approximately 1,300 policy participants in Bay-Delta-
Central Valley water policy.
The number of water years was increased from
three to six to include a wider variety of flow and tempera-
ture conditions. Project members also have obtained ad-
ditional support to study mercury contamination and urban
water management in the context of this research.
35
-------
Modeling Effects of Alternative Landscape Design and Management
on Water Quality and Biodiversity in Midwest Agricultural Watersheds
Mary Santelmann, K. Freemark, D. White, S. Polasky, and G. Matzke
Oregon State University, Corvallis, OR
J. Filers, and J. Bernert
E & S Environmental Chemistry, Inc., Corvallis, OR
B. Danielson, and R. Cruse
Iowa State University, Ames, IA
J. Nassauer
University of Michigan, Ann Arbor, MI
S. Galatowitsch
University of Minnesota, St. Paul, MN
Serious concerns exist about ecological degrad-
ation from modern agriculture. An integrated, multi-
disciplinary approach linking multiple purposes of
watershed could inform and improve our management
of agricultural watersheds and water resources. The
goal of this project is to develop alternative future
landscape design and management scenarios and to
evaluate their effects on agricultural production, water
quality, and native biodiversity in watersheds charac-
teristic of the midwestern United States.
This project involves the following: (1) iterative
design of alternative future landscape and management
scenarios; (2) development of computer models to
estimate effects of landscape change on agricultural
production, water quality, and biodiversity (i.e., aquatic
organisms, terrestrial vertebrates, plants); and (3) a
farm planning exercise to incorporate input from local
farmers and decisionmakers and to explore how human
attitudes as well as practical and economic constraints
are translated into land use and management decisions
and the spatial implications of these decisions at the
watershed scale. This project is focusing on two
agricultural watersheds (5,130 and 8,790 ha) in central
Iowa. Databases exist on land cover, avian bio-
diversity, and aquatic biodiversity for these watersheds
as a result of the USEPA/USDA Midwest Agrichemical
Surface/Subsurface Transport and Effects Research
(MASTER) program.
For each watershed, 3 alternative scenarios for 25
years into the future are being developed using a
normative expert informed design process (see Figure
1). The scenarios reflect a set of alternative national
priorities with an emphasis on: (1) production of agri-
cultural commodities such as corn, soybeans, and
livestock; (2) conservation of native biodiversity and
associated ecosystem functions; (3) improvements in
water quality; and (4) a compromise among all three.
Digital maps of each scenario may incorporate the
effects of heterogeneous implementation of each
watershed design to reflect situations in which farmers
and/or landowners may decide not to participate in
programs designed to achieve improvements in water
quality and biodiversity. Therefore, the manner in
which patterns of implementation may influence the
success of these efforts can then be evaluated. Digital
coverages among scenarios can also be recombined to
evaluate effects on ecological and reorganized model
results.
Existing spatially explicit computer models for
terrestrial vertebrates (multiple species and individual
species populations) are being adapted for the present
watersheds. Parallel models for terrestrial vertebrates
will crossvalidate results. New models for plant bio-
diversity (particularly in wetlands) are being developed
from existing databases in central Iowa and southern
Minnesota. Currently, only rudimentary accounting
models are being developed for agricultural production
and water quality.
This project is pioneering a landscape design
approach for agricultural watersheds based on a
normative process using experts. Evaluation efforts will
contribute to the development and comparison of
spatially explicit computer models for individual versus
multiple species for a variety of taxa. The longer term
significance lies in the ability of the project to inform
landowners and policymakers (for example, those
crafting Farm Bill 2000) of effects of land use and
management choices on water resources, ecosystem
function, and human social systems in the Western Corn
Belt Region.
The next steps for this project will involve the
completion of digital maps of future scenarios as well as
the construction and parameterization of computer
models.
36
-------
\
\—
Design
Scenarios ,-,•
«•
/
X
Design
Scenarios
2
Examine
CON
Design
Scenarios
3
C E P T U A L
Define
5 P E C 1
Design
Scenarios
4
F 1 C 1 TY
Design
Scenarios
U^ for Modeling and
Field Testing
^^^^^—^^^^^^^•k
^•^^•t
PORTRAY COMPARE PORTRAY
COMPARE
PORTRAY
, COMPARE
PORTRAY
. COMPARE
PORTRAY
Figure 1. Agricultural watersheds—A normative, expert design process.
-------
Streamside Reforestation: An Analysis of
Ecological Benefits and Societal Perceptions
Bernard Sweeney, Thomas Bott, John Jackson, Louis Kaplan, J. Denis Newbold, and Laurel Standley
Stroud Water Research Center, Academy of Natural Sciences, Philadelphia, PA
Richard Horwitz and W. Cully Hession
Patrick Center for Environmental Research, Academy of Natural Sciences, Philadelphia, PA
Janet Johnson
University of Delaware, Newark, DE
James Finley, Caren Glotfelty, and Cecilia Ferreri
Penn State University, University Park, PA
Streamside forests are recommended as a land use
practice to protect aquatic ecosystems from nonpoint
source pollution. To date, the scientific basis for the
benefits of these forest buffers has stressed their role in
removing watershed inputs of nutrients, sediments, and
toxic contaminants. In this project, forests along small
streams in the Piedmont region of Eastern North
America are viewed as the primary regulators of stream
width. Preliminary measurements indicate that small
forested streams are consistently wider than contiguous
meadow reaches. The alteration of stream width that
results from deforestation profoundly influences the
stream ecosystem, both locally and in downstream
rivers and estuaries, through effects on habitat and
water quality.
Public policies to restore Streamside forests will
require documentation of ecological benefits and an un-
derstanding of societal factors affecting their implemen-
tation. The principal goals of this project are to: (1) un-
derstand the relationship between forest buffers, stream
width, and the condition of stream ecosystems; (2) de-
velop databases concerning stream organisms and their
activities that can guide the development of policies
concerning restoration of Streamside forests; and
(3) identify and quantify social and economic issues
affecting Streamside forest restoration.
This project includes four interrelated com-
ponents: (1) field measurements in 15-18 different
streams to ensure general applicability of results;
(2) supplemental experiments in controlled model
streams; (3) mathematical modeling to assess watershed-
wide water quality implications; and (4) interviews and
surveys to assess landowner attitudes and land use
practices. The ultimate goal is to bring both the natural
and social science components into focus for
policymakers and regulatory agencies.
Field work began in March 1997. Longitudinal
profiles and detailed cross-sectional measurements
have been performed on 18 stream reaches (i.e.,
paired meadow-woodland reaches on 9 streams) to
estimate the geomorphological characteristics of each
stream reach (e.g., slope, cross-sectional area, width,
depth, sinuosity, pool and riffle characteristics,
floodplain height and width, and habitat
heterogeneity). Preliminary results indicate that
woodland reaches are 1.6 to 3.4 times wider than
meadow reaches. Community metabolism studies have
been conducted on 5 streams (10 reaches) using open-
system measurements of dissolved oxygen change.
Macroinvertebrate communities in the riffle and pool
sections of 9 streams (18 reaches) were sampled
quantitatively in March, May, July, and November of
1997. Fish community analysis was performed by
electroshocking studies of 6 streams (12 reaches; 3-4
passes per reach) in the fall, including length and
weight analysis of individual fish of each species.
Degradation/decomposition studies of natural organic
matter (i.e., humic substances, amino acids,
carbohydrates, and total dissolved organic carbon) and
pesticides (i.e., atrazine, dursban, linuron, and meth-
oxychlor) in meadow and forest reaches was initiated
in 1997. Preliminary studies involving bromide and
chloride releases in 5 streams (10 reaches) have been
completed in 1997 in preparation for nitrogen and
phosphorus spiraling work.
In the anthropological work, investigators de-
veloped, pretested, and refined an interview protocol
and objective statements describing Streamside forests.
Target watersheds and streams in central and southeast-
ern Pennsylvania, and northern Delaware and Maryland
were randomly selected, and the interview process with
landowners was started in 1997 (see Figure 1).
38
-------
f\/ Watersheds
A/R°ads
/\ / Rivers and Streams
Land Use/ Land Cover
Water
Low Intensity Development
High Intensity Residential
High Intensity Industrial/Coi
Hay/Pasture
Row Crops
Lawns/Parks/Golf Courses
:.vV: Forest
I Wetlands
N
Figure 1. Contemporary land use map near rural Chester County, Pennsylvania, showing the location of study streams (blue
lines) and their watersheds (red boundaries) where stream ecosystem structure and function are being evaluated in
response to the presence/absence of a streamside forest. Each dot represents either the upstream or downstream limit
of a given reach of stream; each study stream has two experimental reaches—one forested, one deforested. Map was
derived from EPA Region Ill's 30 m Landsat Thematic Mapper Land Cover Data Set Version 3alb by Cully Hession,
Linda Misiura, Maeve McBnde, Academy of Natural Sciences of Philadelphia.
-------
Integrated Urban Watershed Analysis:
The Los Angeles Basin and Coastal Environment
Richard P. Turco
University of California, Los Angeles, CA
This project is investigating the processes that
control water availability and quality in a major urban
watershed—the Los Angeles basin in southern California
(see Figure 1). The study consists of a synthesis of data
and models that bear on numerous aspects of the water
resource problem, including: regional meteorology and
climatology; basin hydrology, vegetation and land use;
human water consumption and disposition; runoff
sources of sediments, toxics and nutrients; air pollutant
transport, transformation and surface deposition; down-
stream wetlands ecology and impacts; and coastal water
circulation, biogeochemistry, and sediments. This pro-
ject consists of a broad synthesis of observational
data—in situ measurements, remotely sensed data, and
local geographical information—and model simulations
of material flow, deposition, transformation, and
bioassimilation.
Regional simulations of storms over the Los
Angeles basin were carried out with the MM5 meso-
scale dynamics model. This project focuses on a record
rainfall event that occurred in 1993. Model calculations
of precipitation compare favorably with meteorological
observations. The corollary aspects of extreme regional
drought are under investigation through a tree ring
analysis, which correlates moisture availability with
seasonal growth rates and wood density. A 500-year
reconstruction indicates an extended period of severe
drought in southern California in the late 1500's, within
the epoch of the "Little Ice Age."
A coastal ocean model for the Los Angeles
watershed is close to completion. To provide realistic
large scale forcing of the circulation, the model
incorporates a nesting of spatial scales from that of the
Pacific basin, to the California Bight, to Santa Monica
Bay. Moreover, initial results were obtained from a bio-
geochemistry code that is being integrated with the
coastal dynamics model to study the "red tides." The
coastal modeling also is being connected with the runoff
component of the project, for which a GIS-based
analysis has been developed to quantify the chemical
composition (nutrients, toxics) in key regional Los Angeles
outflows. The runoff model is being calibrated using
streams in the region, particularly those heavily impacted
by urbanization. The runoff component uses a com-
prehensive hydrological database and modeling system
developed for several of the Los Angeles watersheds.
New research was conducted on the response of
vegetation in a major southland estuary to anthropogenic
nitrogen and phosphorus. Measurements taken along an
upstream-to-downstream transect through the Newport
Bay estuary indicate that macroalgae efficiently
scavenge nutrients from the freshwater inflow. The
levels of nitrogen in the salt marsh sediments were quite
low, however, and the nitrogen/phosphorus ratios in
these sediments were unusually low.
An analysis of the contribution of airborne
particulates to the southern California coastal ocean
surface microlayer showed that the concentrations of a
suite of trace metals (i.e., Cd, Cu, Ni, Pb, Zn, Cr, Mn,
Fe, and Ag) are more strongly correlated with a
planktonic source than with either an aerosol or bulk
water source. The plankton appear to bioconcentrate
these metals effectively, although the mechanism by
which this occurs remains uncertain. Hence, the
situation of southern California is quite different from
that in the Great Lakes, where the trace metal
composition of the surface layer, bulk water, and
aerosols are strongly correlated. In other work, a novel
analysis of DDT in offshore sediments succeeded in
separating the contributions due to sewage outflows and
to industrial dumping. The sewage source is identified
through a close relationship that exists between DDT
and another class of stable organic compounds used in
detergents. Thus, coastal regions with heavy industrial
DDT pollution have been identified, which will help in
formulating remediation schemes.
40
-------
:MalibuufjF
CreelifW
II
Santa
Monica Balloha
Bay Wetlands
III
Los Angeles
Basin-
I
Newport
Bay
Figure 1. Los Angeles Regional Watershed.
41
-------
Section 3.
Projects Initiated With Fiscal Year 1995 Support
-------
Development and Application of Spectroscopic Probes
for Measurement of Microbial Activity in Aquatic Ecosystems
Carol Arnosti
University of North Carolina-Chapel Hill, Chapel Hill, NC
Neil V. Blough
University of Maryland-College Park, College Park, MD
The microbial degradation of organic matter is a
key part of carbon cycling in aquatic systems. Bacteria
are responsible for a high organic-carbon turnover in
water and sediments, hydrolyzing organic macro-
molecules to smaller pieces that can be remineralized
and interconverting organic structures through a variety
of reactions. Measuring the net degradative capabilities
of a complex community of microorganisms is a major
challenge in understanding the dynamics of carbon
cycling because we lack the means to measure ac-
curately the rates at which bacteria hydrolyze large
macromolecules to smaller pieces, which can be further
transformed or remineralized. In this project, we are
developing a new generation of spectroscopic probes
that can be used to measure extracellular enzymatic
hydrolysis rates of organic macromolecules in the water
column and sediments. This approach is based on
efficient intramolecular energy transfer between donor
and acceptor fluorophores covalently attached to single
macromolecules, so hydrolysis rates can be measured
using simple fluorescence techniques. For this study, the
fluorophores are covalently linked to polysaccharides,
which comprise a significant proportion of total organic
matter. The sensitivity of these probes should permit
measurement of hydrolysis rates at picomolar levels;
ease of analysis means that a large number of samples
can be surveyed rapidly.
Two different polysaccharides as well as lower
molecular weight oligo-saccharides with different sets of
fluorophores have been covalently labeled to determine
optimal probe characteristics and coupling strategies.
Based on initial laboratory investigations, we are focus-
ing on two fluorophore pairs. Laboratory investigations
demonstrate that the double-labeled probes exhibit
fluorescence energy transfer. Donor quantum yield is
significantly higher in single-labeled as compared to
double-labeled probes. Field tests of these probes are
being conducted; the addition of fluorescein/Texas Red-
xylan to porewater from Delaware Bay sediments
demonstrated an increase in the ratios of donor to
acceptor emission with incubation time (see left side of
Figure 1). Concurrent measurements using gel perme-
ation chromatography showed changes in substrate
molecular weight (hydrolysis of xylan; see right side of
Figure 1) consistent with the fluorescence measure-
ments. Changes in fluorescence characteristics were
observable on much shorter time scales than changes in
substrate chromatography; the fluorescence analyses
also required much less time than chromatographic
analyses (< 4 min. vs. > 1 hr). Our technique already
allows us to screen rapidly large numbers of samples to
evaluate enzymatic activity and select samples for fur-
ther investigation.
This project focuses on using purified enzymes to
calibrate the fluorescence responses of the double-
labeled probes in the laboratory. The synthesis of
several new double-labeled probes of different chemical
structures are being planned to test for a wider range of
enzyme activities. This expanded suite of probes will
then be used to make larger scale field measurements of
enzyme activities in the water column and sediments.
Preliminary studies have indicated that we may find
distinct differences in relative activities of specific
enzymes in sediment pore waters as compared to the
water column, and an expanded suite of probes will
enable us to investigate this possibility in detail. These
fluorescent-labeled probes will provide the means of
making rapid, high-resolution measurements of specific
enzymatic activities with minimal disturbance to the
naturally complex microbial community, providing a
more realistic picture of the transformations that actually
take place.
45
-------
-------
Watersheds and Wetlands: Large
Scale Disturbances and Small Scale Responses
Charles Andrew Cole, Robert P. Brooks, and Denies Heller Wardrop
Penn State Cooperative Wetlands Center, University Park, PA
As pressures for development increase and more
wetlands are permitted for modification or destruction,
our understanding of the watershed-level impacts of
these activities remains poor. In addition, although
wetland restorations usually try to provide a replacement
that looks and operates like a natural wetland, the
complexity of these systems may limit success. The
goals of this project are to: (1) assess characteristics of
natural reference wetlands and created wetlands, by
watershed, disturbance, and hydrogeomorphic (HGM)
category, to determine improved design characteristics
for created wetlands implemented for mitigation
purposes; (2) assess succession in reference wetlands
using seed banks and soil dating; (3) assess character-
istics of created wetlands of different ages to begin
determining successional pathways; (4) compare created
wetlands with the reference sites to see if the created
wetlands are successful in any sense; and (5) use
disturbance theory to evaluate the impacts of the
surrounding landscape on both reference and created
wetlands and their successional trajectories.
The Penn State Cooperative Wetlands Center
(CWC) has been assessing wetland structure and
function by classifying a wetlands' position within the
watershed (e.g., headwater) and by its source of water
(e.g., groundwater). Research indicates that sur-
rounding disturbance impacts wetland function so that
events occurring at the watershed scale will likely be
revealed by changes in function at the wetland scale.
The CWC is studying 63 reference wetlands by
measuring soil characteristics, sedimentation rates,
hydrology, basin morphometry, plant community
composition, biomass and production, seed banks,
macroinvertebrates, and wildlife habitat. The investi-
gators have found that soil characteristics, sedimentation
rates, hydrology, and plant communities all are
predictable (to some degree) based on position and
sourcewater for the wetland.
This project involves studying 12 created
wetlands and looking at the same parameters. Our
research found that created wetlands are not at all like
natural wetlands in Pennsylvania. They have different
physical structure, different soils, markedly different
hydrology, and different plant communities. A pre-
liminary time series analysis shows that created sites are
not converging in structure or function towards natural
wetlands.
In addition, this project is considering the historic
use of each of several watersheds to determine if past
land use practices resulted in any discernable changes in
wetland plant communities. A subset of the reference
wetlands is being examined by sampling the soil column
and dating soil layers using Cs137. Once the soil layers
are dated, the seeds in the seed bank will then be
germinated or otherwise identified to develop a plant
community history. A comparison of land use changes
since the 1930's for a 1 km radius around each site will
be made to see if land use has been reflected in the seed
bank (i.e., the plant community). To date, successful
dating of the soil layers has been accomplished, and the
process of collecting seed bank samples is under way.
Our findings are helping to develop data on
wetland function by HGM subclass. HGM will soon be
the primary wetland classification and assessment
protocol within the United States, and we will be able to
provide real data towards assessing functions. Also, we
are adding to basic ecological knowledge on succession
by looking at historic seed banks in reference wetlands
and at chronological sequences in created wetlands. This
will help to define permit expectations and requirements
as well as aid in streamlining an already contentious
mitigation process. We are just now beginning to ex-
plore the impacts of watershed-level activities on
individual wetland function. This data will allow for a
better understanding of large scale decisions on site-
specific processes.
47
-------
Water Authority "AS": 1994-1996
McCall Darn "AS": 1995-1997
•dr
H'H«! H« J|«Hs
Bald Eagle Creek "C720": 1996-1997
100
75
1 I
I -
G -50
-75
BESPPEM"AS": 1994-1996
Figure 1. Examples of wetland hydrographs by HGM subclass.
-------
Integrated Ecological Economic
Modeling and Valuation of Watersheds
Robert Costanza, Roelof Boumans, Tom Maxwell, Ferdinando Villa,
Alexey Voinov, Helena Voinov, and Lisa Wainger
Institute/or Ecological Economics, University of Maryland, Solomons, MD
This integrated modeling effort is intended to
provide tools to manage water quality and ecosystem
responses by synthesizing data from intensive experi-
mental studies on small scales and using dynamic
models to extrapolate that information to large drainage
basins. A grid-cell, process-based model developed for
the 2,500 km2 Patuxent River watershed in Maryland
integrates information over areas of spatial, temporal,
and complexity scales. A unit model calculates stocks
and flows for water, nutrients, plants, and animal
populations within each cell, and cells are linked to
simulate water and material fluxes across the landscape
(see Figure 1). An economic submodel estimates land
development patterns and effects on human decisions
from site characteristics, ecosystem properties, and
regulatory paradigms. Patterns of development are used
in the ecological model to estimate effects of human
activities. The software we have developed links geo-
referenced and other databases to the model for input
and calibration. We have developed methods and tools
to simplify and partially automate model calibration and
result analysis.
Preliminary model runs show good agreement
with measured data for components of the model at
several scales and for several years. Streamflow in the
hydrologic submodule was compared with 2 years of
gage data after the initial calibration year. The model
predicted overall surface and groundwater flow across
the watershed generally within 10 percent of daily
values, although some large flood peaks deviated to a
larger extent from measured values. Several sub-
watersheds were used to test model behavior at smaller
spatial extents (23 km2 to 940 km2). This modular
approach to calibration worked well for the hydrologic
component, with little fine tuning required when switch-
ing scales. Forest biomass calculated by the model
matched boundary conditions established from calibra-
tion data. Temporal and spatial calibration data on
annual increments to forest biomass were developed
using species-specific tree ring records and spatial
distributions in the watershed. Seasonal and long-
itudinal dynamic records for phosphorus and nitrogen
concentrations in the river were compiled from numer-
ous sources for calibration. Unit model estimates were
constrained to boundary conditions for nutrients and
deposited matter.
The investigators developed methods to examine
the effect of spatial pattern of land use on ecological
indicators developed from model variables. Preliminary
results from an adaptation of the hydrologic model for
a suburban watershed suggest that sensitivity to land use
proportions and spatial pattern decrease sharply above
60 percent high-density residential and commercial uses.
Stream buffers in particular lose the ability to mitigate
storm peak flows at this level of development. These
effects will be tested in the full landscape model.
Scenarios are being developed to investigate the
implications of changing land management policies on both
the ecological and economic components of the system.
These include: (1) clustered vs. sprawl for agricultural and
urban areas development; (2) implementation of nutrient
and water management practices for agricultural and
urban areas; (3) development with varying proportions
of forest, agricultural, and urban uses; and (4) spatial
arrangement of land uses. Dynamic models examine the
numerous competing effects of human and naturally
imposed variation to suggest potential long-term impacts
to system function and resiliency. Model results can
help guide land use policy and help to develop goals for
ecosystem quality indicators.
49
-------
Figure 1. Elevation of the Patuxent Watershed. Grid cells shown are 3 km. Model grid cells range from 0.2-1 km.
50
-------
Oyster Reefs as Structural and Functional
Components of Tidal Creeks; An Ongoing Ecosystem Experiment
Richard F. Dame, E. Koepfler, L. Gregory, T. Prins
Coastal Carolina University, Conway, SC
D. Allen, D. Bushek, C. Corbett, D. Edwards, B. Kjerfve, A. Lewitus, and J. Schubauer-Berigan
Baruch Marine Laboratory, Georgetown, SC
ATMOSPHERE
WATER
PARTICULATE
FILTRATION
SEDIMENTATION
DISSOLVED
METABOLITES
MINERALIZATION
RESUSPENSION
RELEASE
AEROBIC SEDIMENTS
DON
DEITRIFICATION
MOBILE PHQSpHATBpO
METHANOGENESISCH
Figure 1. Ecological processes coupling oyster reefs to tidal creek waters.
Shallow tidal estuaries are highly productive and
ecologically diverse systems because the complexity of
the physical environment generates a multitude of
habitats. These estuaries are threatened by environ-
mental changes such as pollution, overharvesting of
fisheries, and sea level rise. The investigators in this
project are conducting an ecosystem-level experiment to
examine the structural and functional role of oyster reefs
in tidal creeks (see Figure 1). Eight similar tidal creeks
were chosen to standardize for oyster dry body biomass
to creek water volume at full bank conditions. Each
creek has been topographically and bathymetrically
surveyed, and these data have been used to determine
area and volume relationships for each creek basin.
Preliminary findings from the first, premanipu-
lation year include: (1) less mature creeks with upland
interfaces have lower nutrient concentrations than more
mature creeks nearer the ocean; (2) oyster growth
appears to be negatively correlated with distance from
the creek mouth; (3) grazing by bivalves and micro-
zooplankton is important in controlling phytoplankton
year round; (4) nutrient limitation effects are only found
during the late fall to early spring period (see Figure 2);
(5) tide pools within creeks accumulate nutrients and serve
as refuges for nekton; and (6) differences in nekton
assemblages among creeks appear to be determined by
creek geomorphological and flow pattern variations.
In the manipulation year, which begins in March
1998, oyster reefs will be removed from four creeks,
and their structure and function will be compared to
four manipulated creeks for 1 year. The replicated
experimental design will further the basic understanding
of the role of oyster reefs in sustaining the ecosystems
that they dominate.
51
-------
Growth response to
grazer reduction
Growth response
toNH
Spring
Summer Fall
Season
Winter
Figure 2. Predicted phytoplankton responses in creeks without oyster reefs compared to creeks with oyster reefs.
52
-------
Tracing the Fate of Nitrogen
Inputs From Watersheds to Estuaries
Linda A. Deegan and Bruce J. Peterson
The Ecosystems Center, The Marine Biological Laboratory, Woods Hole, MA
This project's goal is to illuminate the trophic
pathways by which river-borne dissolved inorganic
nitrogen is incorporated into and processed by food
webs in the upper, tidal freshwater-oligohaline reaches
of coastal plain estuaries. Specific research goals
include determining: (1) the key species in pelagic and
benthic food webs; (2) the strength of linkages between
benthic and pelagic food webs; and (3) the importance
of watershed nitrogen inputs to consumer organisms.
To meet these goals, we conducted the first whole-
ecosystem stable isotope tracer amendment to be carried
out in an estuary. First, 15N-enriched nitrate was added
to the upper reach of the Parker River, Massachusetts,
during the summer of 1996 and, through intensive
sampling, the progress of the isotopic label through the
water column, organisms, and sediments was tracked
during the amendment and for several months there-
after.
Three major sources of organic matter fuel the
consumers of the upper Parker River estuary: (1) most-
ly microscopic pelagic (water column) producers, (2)
benthic (sedimentary) primary producers, and (3)
detritus derived predominantly from the tidal freshwater
marsh (see Figure 1). The pelagic diatom, Actinocydus
normanii, was identified as a critical determinant of the
fate of watershed-derived nitrogen in the Parker River
estuary during the summer period when biological
activity reaches its maximum. A bloom of this single-
celled alga rapidly assimilated watershed-derived nitrate
entering the upper estuary and was the primary vector
of this nitrogen to benthic and pelagic food webs.
Among those species feeding directly on the diatom, or
on an herbivorous intermediary, up to one-half of their
assimilated nitrogen was derived from this source, and
thus from new nitrogen entering from the watershed.
On the basis of 15N tracer incorporation, 25-50 percent
of the dietary nitrogen assimilated by planktivorous
fishes, such as juvenile alewife and white perch, was
derived originally from A. normanii by way of their
copepod food. Surprisingly, we found that the tracer
was rapidly assimilated by major crustacean benthos
(i.e., mud crabs, amphipods, and grass shrimp), indicat-
ing an unexpectedly strong linkage of pelagic primary
production with benthic secondary production. It is
estimated that 20-40 percent of their assimilated nitro-
gen derived from the pelagic diatom. A. normanii was
incorporated into surficial sediments, providing both a
food reserve for benthos, and a probable nitrogen
source for locally important benthic primary producers
that fed small, productive intertidal fauna. Other spe-
cies were largely uncoupled from A. normanii and
depended more on unlabeled food sources, derived most
likely from freshwater marsh detritus. Abundant
benthic-feeding fishes (e.g., white sucker, mumrnichog)
assimilated the tracer minimally by feeding primarily in
a detritally based food web.
This project is among the first to measure directly
the relative trophic importance of in situ primary pro-
ducers and detrital organic matter in an intact ecosys-
tem, and it is the first to do so by using a stable isotope
tracer in an estuary. The research has made a major
contribution to understanding the fate of watershed-
derived nitrogen in the critical zone where rivers inter-
face with estuaries, a region that has received relatively
little study despite its potential to control nutrient trans-
formation and transport from uplands to coastal waters.
53
-------
Trophic Pathways of Nitrogen Flow in the Upper Parker River Estuary
§
I
v>
grass shrimp
mud crab
amphipods
Nereis
pelagic copepods
E. affinis
tidal freshwater
marsh detritus
mummichog
white sucker
deposited detritus
benthivores
planktivores
alewife
white perch (Juv-)
^^Tnjcrophytobenthos, bacteria
N
remifieralfeation
sediments
striped bass
Figure 1. Conceptual model of nitrogen flow through biota of the upper Parker River estuary, showing trophic linkages supported
by isotopic evidence, feeding ecology, and habitat use.
-------
Probing the Relationship Between Fulvic Acid Aggregation,
Metal Ion Complexation, and the Binding of Organic Compounds
A. Dixon ', W.R. Carper2, and C.K. Larive '
' Department of Chemistry, University of Kansas, Lawrence, KS
2 Department of Chemistry, Wichita State University, Wichita, KS
D Region 1
13 Region2
H Aromatic
• Average
Sm Angle BPPLED Spin Echo- Relaxation-
X-Ray BPPLED BPPLED
Figure 1. This figure shows the results of NMR experiments for the analysis of the average molecular size of fulvic acid solutions. These results
are compared with the reported radius of gyration for this sample using small-angle X-ray scattering. Region 1 refers to the portion
of the proton NMR spectrum containing CH3, CH2, and CH groups adjacent to aromatic, carbonyl and carboxyl groups. The standard
NMR experiment, the Bipolar Pulse Pair Longitudinal Encode-decode pulse sequence (BPPLED), produces a radius of gyration that
is the average over all fulvic acid molecules and is directly comparable to the small-angle X-ray scattering results. We have used two
spectral editing methods based on spin-spin coupling (spin echo) and T, relaxation to modify the standard BPPLED experiment to detect
fractions of die sample that are either lower or higher in molecular weight than the average. All three NMR experiments are being
used to see how the average size of each fration changes as a result of cadmium complexation.
Soluble organic acids (humic substances) leach
into ground and surface waters from soil and plant
organic matter. Humics play a key role in water quality
because they are the major constituents of dissolved
organic carbon and they are important in controlling the
bioavailability and transport of toxic metals and organic
pollution. The goal of this project is to develop a better
understanding of the physical and chemical properties of
natural aquatic organic materials such as fulvic acids.
Our approach makes use of the high specificity of nuclear
magnetic resonance spectroscopy (NMR). NMR pro-
vides information about the local chemical environment
of a molecule or chemical group. For example, binding
constants of NMR active metal ions (such as cadmium)
can be measured by monitoring the concentration-
dependence of the NMR chemical shift, hi addition, the
NMR chemical shift also yields information about the
local chemical environment of the metal. In this way,
it is possible to determine which functional groups (such
as carboxylate or amine) form the complex.
NMR relaxation measurements and detailed
molecular mechanic calculations also are being used to
further elucidate the nature of the metal-fulvic acid
complex. One interesting facet of this project is the
relationship between metal complexation and fulvic acid
aggregation. NMR experiments have been modified for
the measurement of diffusion coefficients to make them
more specific for different classes of compounds. In this
way, the extent to which metal ion complexation induces
aggregation in a fulvic acid sample can be examined.
The high specificity of NMR can provide additional
information about the fulvic acid aggregates when
compared with that provided by other analytical
methods, which could be used for this purpose, such as
small angle X-ray scattering and dynamic light
scattering. The ultimate goal of this project is to use
NMR and specially labeled compounds to examine the
extent of metal binding to a fulvic acid solution, how
metal ion binding affects the aggregation of the fulvic
acid sample, and how metal-induced aggregation affects
the binding of organic pollutants by the fulvic acid
aggregates. The molecular level specificity of our an-
alytical methods will increase our knowledge of the
molecular-level behavior of aquatic fulvic acids. This
information will be useful to develop a better under-
standing of how the chemistry of fulvic acids contributes
to the availability and transport of pollutants in aquatic
ecosystems.
55
-------
Diffusional Rate Limitations in Heterogeneous
Porous Media: Model Structure, Scale, and Geologic Characterization
David L. Freyberg and Paul V. Roberts
Department of Civil Engineering, Stanford University, Stanford, CA
(B)
80'^
60
20
80
40 60
[cm]
Figure 1. Regions of low groundwater velocity within a point bar deposit. (A) Natural log of the hydraulic conductivity (cm/s) for a vertical
section. (B) Regions of low velocity (dominant flow direction is left to right): mean velocity magnitude=4.TxlO^cm/s; dark regions
are those with velocity magnitudes less than 4.8xlO"5cm/s.
The goal of this project is to enhance our ability
to predict the transport and fate of hazardous chemicals
in subsurface waters by increasing our understanding of
the nature and significance of diffusion during transport
in saturated, heterogeneous porous media. The objec-
tive is to develop the understanding and tools necessary
to evaluate and predict the relative importance of
diffusion, or diffusion-like processes, over a range of
spatial scales (from grains to regionals). The approach
rests on a program of numerical experimentation and
analysis examining: (1) the mathematics of multiscale
diffusional rate limitations on sorption; (2) the inter-
actions between geologic structure, heterogeneity,
upscaling, and real and apparent transport process rep-
resentation in mathematical models; and (3) physically
based aggregation schemes that are consistent with
different model formulations of transport processes and
multiscale diffusional rate limitations. The investigators
are using both random-field models and a very
fine-scale, quasideterministic, numerical model to re-
present spatial heterogeneity and uncertainty.
The researchers have developed and applied a
model to analyze a set of laboratory column experiments
conducted to determine the rate at which trichloro-
ethylene (TCE) can be removed from natural soils. A
wide range of diffusional time scales for the TCE
sorption-desorption process, spanning up to 15 orders of
magnitude, has been found. To investigate the effects
of a distribution of contaminant sorption time scales in
macroscopically homogeneous aquifers, the temporal
moments of a contaminant pulse were examined. A
distribution of sorption time scales was found to not
affect the first two temporal moments, but can affect
strongly and even dominate the third moment. In ex-
tending the analysis to heterogeneous aquifers, the
statistics of the arrival times of a contaminant are being
examined after transport through a hydraulic con-
ductivity field. Preliminary results suggest that we will
be able to distinguish among scenarios in which: (1) the
heterogeneity of the aquifer controls the spread of
arrival times, while sorption can be treated as an
equilibrium process; (2) the kinetic sorption controls the
spread of arrival times, and the aquifer can be treated as
macroscopically homogeneous; and (3) both hetero-
geneity and kinetic sorption contribute significantly to
the spread of arrival times.
Pure advection through spatially heterogeneous
groundwater velocity fields itself can lead to skewed
breakthrough curves. In addition, velocities in some
regions of a heterogeneous velocity field may be so low
that solute transport is dominated by diffusion, even
though the mean velocity is relatively high. To inves-
tigate the nature of low-velocity regions, the investi-
gators began by exploring simple geometries such as
spherical inclusions. Subsequently, this project has
examined simulated flow through a highly resolved
representation of a point-bar deposit. Preliminary results
show that regions of very low velocity do exist, and that
the spatial structure of these regions is complex and a
strong function of the conductivity heterogeneity as well
as of the mean head gradient (see Figure 1).
The results from this research will contribute to
a fundamental understanding of the impact of dif-
fusionally rate-limited sorption and diffusionally
controlled slow release from zones of low velocity on
transport, with applications to contaminant remediation
design, site characterization, and the interpretation of
field monitoring and laboratory experimental data.
Continuing work is focused on exploring
heterogeneity-kinetic sorption interactions in different
heterogeneous settings; characterizing low-velocity re-
gions and their relationship to geologic structure and
flow as well as transport boundary conditions; and
exploring the roles of aggregation, upscaling, model
formulation, and spatial structure in predictive modeling
of transport.
56
-------
Integrating Planning, Forecasting, and Watershed Level Ecological Risk
Assessment Techniques: A Test in the Eastern Cornbelt Plains Ecoregion
Steven I. Gordon and Andy Ward
The Ohio State University, Columbus, OH
Dale White
Ohio Environmental Protection Agency, Columbus, OH
The objectives of this research are: (1) to test the
relationships between biological conditions of streams
and the nature and distribution of human activities on
the watershed; (2) to demonstrate methods for linking
physical models of urban and agricultural impacts on
runoff volume and runoff quality; (3) to define the
relationships between physical model forecasts and the
biological conditions of streams; and (4) to integrate all
of the findings into an expert system to be used by
planners.
During the past year, we have been working
simultaneously in three areas: (1) assembly of a com-
prehensive regional database and testing of watershed
scale relationships between biological measures of water
quality and watershed characteristics; (2) testing the
impacts of different scale soils data on our predictive
models of hydrologic response; and (3) finalizing
methodologies and data for preparing and analyzing
riparian zone and detailed watershed data aimed at
testing models at larger scales.
The USEPA RF3 stream Geographic
Information System (GIS) files were used as the basis
to divide the 4 Eastern Cornbelt Plains region into two
distinct sets of watersheds—25 coarse scale watersheds
and 137 fine scale watersheds. For the region as a
whole, Landsat imagery was used to define the major
land use categories. A number of other data were
compiled using the same geographic base, including
the IBI (Index of Biologic Integrity), the major
dependent variable in our regional scale analyses. The
data were collected at 522 locations throughout the
watersheds.
At this regional scale, a number of empirical
relationships between the IBI as a measure of biological
quality and some proxy variables for point and nonpoint
water pollution stresseshave been tested. More detailed
modeling at the watershed scale has proceeded by
testing the impact of soils data at two scales on the
results of our sediment model. A test of the model
behavior using two soils data at two different scales has
been completed. Finally, work at the riparian zone level
has led to a strategy for extracting relatively detailed
vegetation-type assemblages and patterns in the riparian
zone using Landsat imagery. A new set of techniques
will be used in the coming year to derive the most
detailed data sets and to test empirical models at this
scale against our regional scale and watershed scale
empirical and physical models.
A set of empirical models for the Eastern
Cornbelt Plains Ecoregion has been defined, which
explains significant amount of the variation in IBI levels
across watersheds. Table 1 compares the results for the
larger scale (more detailed) and smaller scale
watersheds.
The significant variables are the percentages of
dense urban settlement, and several measures of
watershed condition and location, including stream
order, substrate type, and pool and riffle run quality.
Adding an estimate of chemical stressors using
STORET data produced significant improvement in
model results. IBI levels also are strongly influenced
by upstream and downstream IBI.
Model runs made for the Big Darby Creek
basin with the two alternative sources of soils data
lead to different conclusions concerning scale changes.
The predicted annual flow, sediment, and nitrate
losses for the analysis with conventional tillage and no
subsurface drainage are presented in Table 2. There
is no significant difference between the mean annual
flow and sediment values predicted by using data of
finer vs. coarser resolution. However, the mean
annual nitrate loss prediction based on the two soils
databases are significantly different due to the
approximately 10 percent difference in the loss for
1992. There are considerable differences in the flow,
sediment, and nitrate prediction for individual map
units of fine vs. coarse resolution (not presented). The
results suggest that coarser resolution data can be used
to model hydrological responses at a watershed scale
but should not be used at a field scale, or if the goal
is to predict responses for short time periods. It
should be noted that in this study, the model was not
calibrated and the results only provide an indicator of
the relative responses associated with different
agricultural practices and soil characteristics. It is
probable that sediment losses are underpredicted and
some of the model input variables need to be
adjusted.
57
-------
Table 1. Regression models for mean IBI at the watershed scale.
Coarse Watersheds (25)
R2 Values
R2 Values
R2 Values
.384
.494
.694
Fine Watersheds (137)
.409
.525
.703
Significant
Independent Variables
Dense urban; substrate;
stream order;
Pool/Riffle
Adding chemical
conditions based on
ranking of segment by
STORET readings
Adding upstream and
downstream IBI to
account for spatial
autocorrelation
Table 2. Comparison of annual flow depths, sediment losses, and nitrate losses for conventional tillage
without subsurface drainage.
Precipitation
Flow Depth (mm) I Sediment Loss (ton/ha) I Nitrate Loss (kg/ha)
STATSGO MUIR STATSGO MUIR STATSGO MUIR
1991
1992
1993
1994
1995
Mean
718
792
931
696
937
815
197
211
264
155
317
229
198
226
270
160
318
235
3.0
3.7
3.7
3.9
6.8
4.2
2.8
3.9
3.8
3.9
6.9
4.3
9.3
16.2
9.9
6.9
10.7
10.6
9.3
17.8
10.2
6.8
11.0
11.0
58
-------
Development of Geomorphological Artificial
Neural Networks (GANNs) for Modeling Watershed Runoff
Rao S. Govindaraju
School of Civil Engineering, Purdue University, West Lafayette, IN
Input
Weighted
Summation
Output
Figure 1. A schematic of modular artificial neural network.
Predicting streamflow is a fundamental com-
ponent of a wide range of watershed-modeling efforts.
In this project, geomorphological artificial neural
network techniques have been applied to predict
watershed response to precipitation and temperature.
These networks can be used to resolve watersheds at
different spatial and temporal scales through a massively
interconnected network of processing units. System
architecture is determined by the topology of the
watershed. Network development is characterized
through an iterative training procedure. Model
calibration involves the selection of the optimal values
of connection strengths and processing-unit thresholds,
so that the network best reproduces historical stream-
flow records for the watershed. This requires a large
number of training examples, in terms of historical
precipitation and streamflow records for a watershed.
Once network training has been accomplished, it can be
used to forecast streamflows for the watershed based on
rainfall and temperature inputs. The model preserves
some geomorphological and probabilistic characteristics
of watershed response as represented by historical
records. Data from several Kansas watersheds are
being used to test the performance of the model. To
avoid problems associated with over-training, an
additional cross-validation data set has been used to
determine optimal training and network architecture.
The data set is partitioned into three portions: (1) a
training set, (2) a cross-validation data set, and (3) a
testing data set. Historic records have been used to
determine the time lags for watershed response to
precipitation and have been incorporated in our
modeling effort. A cross-correlation analysis between
the average monthly precipitation and watershed runoff
revealed that the past 3 months of rainfall data would
have a significant impact on the runoff for some Kansas
watersheds.
Singular neural networks predict average
streamflow characteristics satisfactorily, but they have
not been successful in predicting high and low runoff
events. A number of singular networks were arranged
in a layered structure to form a modular network, and
stream discharges were partitioned into three classes,
each using a singular network (see Figure 1). Each
singular network or module covered a particular range
of the dataset, and a final module assigned weights to
the prediction of the individual modules. Each module
was developed based on the data in a class, but the final
network was based on the entire training dataset. This
kind of network resulted in an improved prediction of
streamflows for three Kansas watersheds. Inclusion of
rainfall and temperature from previous months allows
the model to better represent watershed conditions.
The preliminary success of modular networks
suggests that different rules are required to represent
extreme and average events. The architecture of the
modular network can be adapted to represent a stream
network and includes the geomorphology of a
watershed. Future efforts are planned to extend our
applications of geomorphological artificial neural
networks and consider the influence of spatial and
temporal resolution on watershed-runoff computations.
59
-------
Physicochemical
Mechanisms Governing Virus Filtration
Stanley B. Grant and Terese M. Olson
Department of Civil and Environmental Engineering, University of California, Irvine, CA
Mary K. Estes
Division of Molecular Virology, Baylor College of Medicine, Houston, TX
3.00
2.00
1.00
,23.4 r»M
11.7 HM
"0.0 nw
NanoScope
Scan size
Setpoint
Scan rate
Contact AFM
3.000 JJM
-2.025 U
1.001 Hz
Number of sanples
256
Figure 1. Atomic force microscope image of Norwalk virus particles absorbed to a mica surface. This adsorption experiment was conducted
in an aqueous solution at pH 5 and lOOmM NaCl. Under these conditions, the virus is positively charged and the mica is negatively
charged. Consequently, there is no electrostatic barrier to virus adsorption.
The use of reclaimed sewage for groundwater
recharge raises a number of important human health
issues. Foremost among these is the possibility that
human enteric viruses present in the reclaimed sewage
may be transported to production wells and, ultimately,
end up in a water distribution system. Laboratory-scale
experiments are being conducted that are aimed at
identifying the geochemical features of a recharge basin
influencing the removal of viruses from groundwater by
physicochemical filtration (see Figure 1). Initial experi-
ments of this project examined the influence of water
pH on the surface charge of recombinant Norwalk virus-
like particles and their filtration rates in packed beds of
quartz sand. Many outbreaks of gastroenteritis in the
United States are caused by waterborne or foodborne
transmission of Norwalk virus. The initial results from
this project suggested that pore water pH may be the
most important factor in deterrnining the capacity of
groundwater systems to provide natural disinfection by
physicochemical filtration.
Further experiments have used a total of four
different viruses, including three coliphage (a somatic
phage, and two different male-specific phages) and
Norwalk virus particles generated using a baculovirus
expression system. In all cases, we found that pore
water chemistry dramatically influences the removal of
these viruses in packed beds of quartz sand. When the
filtration results are combined with microelectro-
phoresis measurements of the viruses' surface electrical
potential, the results are generally consistent with the
predictions of classical theory—that is, electrostatic
interactions between the viruses and quartz sand ul-
timately determine a filter's ability to remove viruses
from water.
60
-------
Watershed Impacts on Sediment Pollution
History and the Viability of the Zooplankton Egg Bank
Nelson G. Hairston, Jr., and Colleen M. Kearns
Section of Ecology and Systematics, Cornell University, Ithaca, NY
Charles T. Driscoll
Department of Civil and Environmental Engineering, Syracuse University, Syracuse, NY
log Qig Hg/g Sediment)
10 -10
E
u
Q.
»
Q
O
E
•o
0
V)
1950
1930
1890
1850
1810
1770
1730.
Onondaga 1690'
Lake 1650
10
100 0
j %Hatch
i
f
5
AH.
'' Oneida
> Lake
1990-
1970-
1950-
1930-
1890
1850
1810-
1770-
1730-
20
40
60
80
100
% Hatch of Diapausing Eggs
Figure 1. Comparison of the mercury concentrations (logarithmic scale) and hatching success of zooplankton diapausing eggs in the sediments
of two lakes in central New York State. Oneida Lake has only received metals from atmospheric deposition on the watershed.
Onondaga Lake has experienced direct inputs from industry effluent. Sediment dates determined by 210-Pb analysis. Note that
analysis of hatching success in Oneida Lake is only partially complete.
The zooplankton community of lakes and its
response to environmental change may be substantially
impacted by the diversity, viability, and population
dynamics of long-lived diapausing eggs buried in lake
sediments. The history of environmental change in
lakes is manifested in the pollutants that enter their
watersheds and are transported to the lake, where they
ultimately accumulate in the sediments.
This project is investigating the impact of this
pollution on diapausing egg viability. The two major
sources of heavy metals in lake sediments in the
Northeastern United States are: (1) direct inputs of
industrial effluent to the watershed or lake and (2)
atmospheric inputs that are deposited within the wate-
rshed and wash into the lake. Replicate sediment cores
taken at two lakes in central New York State reflect
their distinct histories of industrial impacts. In both
lakes, cores were sectioned, dated by 210-Pb, and
analyzed for concentrations of metals, including lead
(Pb) and mercury (Hg). In both lakes, fluxes of these
metals to the sediments increased from 1850 to the
mid-to-late 1900's. The increase was initially due to at-
mospheric deposition to the watersheds, but in
Onondaga Lake, major inputs from chemical and steel
industries on the shore occurred during the 1960's,
1970's, and 1980's.
Sediment enrichment factors (modern metal
deposition/metal deposition in 1850) were on the order
of 2 to 3 for Pb and Hg in Oneida Lake, but between 30
and 120 for these metals in Onondaga Lake. These
deposition patterns will be compared with those for
remote lakes in the Adirondack region of New York
State. Diapausing eggs of crustacean and rotifer zoo-
plankton are found in abundance in the sediments of
both lakes (see Figure 1). Both show substantial hatch-
ing for eggs deposited 40 to 100 years in the past.
There is no evidence for a reduction in egg viability in
Onondaga Lake, as compared with Oneida Lake, or
within Onondaga Lake as a function of sediment metal
concentration. Apparently, the '"ametabolic" state of
zooplankton eggs in diapause protects them from metal
toxicity in the sediments. These eggs may thus repre-
sent a source for the recovery of lake zooplankton
communities from past environmental pollution.
61
-------
The Role of Colloidal Particles in the
Transport of Chemicals Through an Agricultural Watershed
George M. Hornberger and Janet S. Herman
University of Virginia, Charlottesville, VA
James E. Saiers
Florida International University, Miami, FL
Organic contaminants exhibit complex chemical
patterns as they are transported through watersheds to
streams. This project has examined the chemistry of the
widely used herbicide, atrazine, as it passes through
hydrologic systems—from its initial occurrence in the
unsaturated soil zone in agricultural areas through the
groundwater of a bedrock water-supply aquifer to its
appearance in a stream draining the watershed. The
goal has been to expand our understanding of herbicide
movement through the entire soil-bedrock-stream system
in an agricultural watershed. This information will
permit informed management decisions about land use
and its impact on a sustainable water supply.
Work is being conducted on Muddy Creek,
which drains an agricultural catchment in the Shenan-
doah Valley, Virginia. The research addresses three
primary questions: (1) What factors control the varia-
bility in the concentrations of colloids and agricultural
chemicals in the aquifer-stream system at Muddy Creek?
(2) How do aquifer materials and colloidal particles
affect the movement of agricultural chemicals such as
atrazine and hydroxyatrazine? (3) Do laboratory-based
observations of colloid-facilitated transport of agri-
cultural chemicals provide a reliable estimate of trans-
port at the field scale?
In this project, the forms and transport of
herbicides in the groundwater flow system will be
reported. To evaluate in situ soil-water composition,
samplers were installed to collect water held under
different tensions in the unsaturated zone at three depths
at three sites in the catchment. Tension lysimeters were
installed in 8 cm diameter holes at depths of 15, 45, and
90 cm. Zero-tension (pan) lysimeters, which sample
water flowing through macropores in the soil, were
installed by excavating a trench and fitting three pans at
the same depths as the tension lysimeters. Soil-water
samples were collected, and water volume was recorded
from both types of lysimeters approximately every 4-6
weeks from April 1996 through June 1997. Samples
were analyzed for atrazine, nitrate, dissolved organic
carbon, and colloid concentrations. In the clay-silt loam
at our field site, significantly higher concentrations of
atrazine were identified in zero-tension lysimeters at
three depths (15, 45, and 90 cm) than in corresponding
tension lysimeters. Atrazine concentrations in the zero-
tension lysimeter were as high as 21 ppb, whereas the
corresponding concentrations in tension lysimeters were
only ~2 ppb (see Figure 1). Colloid concentrations in
the zero-tension lysimeters were as high as 650 mg/L.
Although the differences between the particle-associated
atrazine and the dissolved atrazine in the pan
lysimeters are small, they are significant statistically
(p = 0.027). Concentrations of nitrate are highly
variable and not significantly higher in the pan
lysimeters than the tension lysimeters, indicating that
different processes may control the transport of
nitrate relative to atrazine.
Our results indicate that: (1) transport of atrazine
is largely via macropores; (2) colloids are mobilized and
transported over distances of at least 1 meter in the
vadose zone at our site; and (3) nitrate is not a good
surrogate for judging atrazine transport. Macropore
sampling is required to estimate transport of herbicides
in the field, and colloid-facilitated transport of strongly
sorbing constituents is of potential importance at
agricultural sites such as Muddy Creek.
62
-------
a.
20
§15-
03
c
a
£10-
5 -
n
•i rii ll
JL
1
i
i
apr may jun aug
1996
20 -
f 15 •
^
0)
c
a
£10-
5 -
n -
. . 1
|-i
-.
-
i
i
[
1 pan
rrt
tension
Lnn
oct
feb
mar
fl.
_, »n n
ll
may jun
1997
f
|
^H raw
rlT
n
filtered
nti
ll
tfinn
i
T
f
b.
apr may
jun aug oct
1Q96
feb mar may jun
1QQ7
Figure 1. Atrazine concentrations in (a) pan and tension lysimeters and (b) raw and filtered fractions from pan lysimeters.
63
-------
Geomorphic, Hydrologic, and Ecological Connectivity in
Columbia River Watersheds: Implications for Endangered Salmonids
Hiram W. Li, Bruce A. Mclntosh, J.Boone Kauffman, Judith L. Li, and Robert L. Beschta
Oregon State University, Corvallis, OR
Patricia McDowell
University of Oregon, Eugene, OR
100n
Fishes
-o-Warmwater spp.
redside shiner
speckled dace
northern squawfish
largescale sucker
-iV'Coolwater spp.
mt whitefish
torrent sculpin
bridgelip sucker
-a- Coldwater spp.
chinook salmon
rainbow trout
ABODE FGHI JKLM
Warm Cool Cool Warm Cool
Geomorphic reach
Figure 1. Patterns of fish assemblage structure corresponding to temperature patches among stream reaches.
Coldwater fish communities throughout the
Columbia River basin face high risks of extinction due
to degraded freshwater habitats. The survival and
persistence of these fish communities will depend on
our understanding of the processes controlling stream
temperatures and fish habitats. This project is investi-
gating the interplay among geomorphic, hydrologic, and
riparian-community features and the dynamics of
groundwater input and hyporheic flow in pristine and
badly disturbed streams in the high desert of the
Columbia River. Images captured using Forward Look-
ing Infrared (FLIR) videography revealed that the
temperature profile in a watershed disturbed by
intensive livestock grazing, the Middle Fork John Day
(MFJD), varies by as much as 50° C. However, a
stream draining the wilderness basin, the North Fork
John Day (NFJD), exhibits the normative temperature
patterns. Patterns are similar throughout the year, with
consistent magnitudes between stream differences in all
seasons.
Presently, there is no clear, single explanation for
the temperature patches. Ground truthing has estab-
lished that it is not an artifact. Intensive stream gaging
at selected locations, dye release tests, and measures of
vertical hydraulic gradient do not indicate that they are
caused by hyporheic groundwater exchange, which is
relatively minor in comparison to that documented in
streams of the northern Rocky Mountains. Likewise,
analyses of bed materials suggest that hydraulic con-
ductivity may be low, although this will be decided with
further analysis by ground penetrating radar to map
potential pathways of hyporheic flow by measuring
depth to bedrock and defining alluvial stratigraphy.
These studies were conducted in the MFJD and adjacent
drainages with similar geomorphologies. Biogeochem-
ical relationships between subsurface flows, sols, and
vegetation in the riparian zone were characterized by
measuring annual dynamics of water-table depth, soil
redox potential, and chemistry of subsurface and stream
waters. In contrast to instream measurements, pre-
liminary results show some degree of lateral hyporheic
exchange between the stream and riparian communities
as documented by water temperature, redox potential,
and dissolved organic carbon concentrations. Livestock
grazing appears to affect groundwater exchange. Un-
grazed riparian reaches were found to have greater root
bio-mass, soil organic matter, water infiltration rates,
pore space, and lower soil bulk densities.
Fish surveys in the two watersheds reveal that
longitudinal fish community patterns strongly reflect the
temperature profiles. There is a gradual shift from cold
water to warm water communities in the wilderness
stream system (NFJD); whereas, the relative composi-
tion of community types in the disturbed stream (MFJD)
follows the distribution of temperature patches (see
Figure 1). Subsurface flow within reaches was found to
be very slow and poorly oxygenated, suggesting that
high stream temperatures may force organisms to
choose between microhabitats of cooler temperatures,
but with low dissolved oxygen versus mainstream
habitats that are warmer, but higher in dissolved
oxygen. Water chemistry, primary production, macro-
phytic and algal biomasses, macro invertebrate abun-
dances, and fish diets at 12 sites along the MFJD were
to examine the effects of water temperature and possible
hyporheic up-welling/down-welling on the trophic net-
work. This project's analyses do not support the hy-
porheic corridor concept as an explanation for the
distribution of organisms, but an evaluation of the
factors dictating temperature conditions will be made.
64
-------
Resistance of Communities to Chronic Haloaromatic
Contamination From Biogenic and Anthropogenic Sources
David E. Lincoln, Sarah A. Woodin, Charles R. Lovell, and V. Pernell Lewis
University of South Carolina, Columbia, SC
Halogenated aromatic compounds are important
pollutants characterized by their toxicity, persistence,
and accumulation in the environment, and abundant
production and frequent use in a variety of industrial
processes. This project examines the extent to which the
capacity to degrade naturally occurring halogenated
aromatic compounds determines the biological impact of
anthropogenic halophenol pollutants on marine benthic
communities. Dehalogenase enzyme activities and ge-
netic probes are being used to assess the potential for
dehalogenation among the polychaete worms, which
dominate these sedimentary ecosystems, and the
worm-associated bacteria, as well as other principal
community members. The enzyme assays and DNA
probes plus induction experiments will enable us to
determine if dehalogenation potential is an important
determinant of organism survival and persistence in
biogenically and anthropogenically contaminated lo-
cations.
Our findings to date highlight the abundance and
broad distribution of benthic macro- and micro-
organisms capable of production and/or degradation of
halometabolites. A large proportion of invertebrate
taxa, which are numerically dominant in coastal marine
sediments, are halometabolite producers and have
enzymatic dehalogenation activity. The native worm
species—the most abundant intertidal benthic macro-
fauna—appear to be segregated among sites with regard
to their dehalogenation capacity, and this may be related
to the presence of halophenols in their environment.
Twelve reductively dechlorinating anaerobic bacterial
strains have been isolated from the burrows of a
bromphenol-producing and a nonproducing species.
These strains reductively dechlorinate 2,4,6-tri-
chlorophenol.
A survey of polychaetes at an undisturbed
estuarine site showed that 40 percent of the organisms
contained halogenated metabolites. Additional surveys
indicated that native species appear to be segregated
among sites with regard to their halgenation and
dehalogenation capacity, and this may be related to the
presence of halophenols in their environment.
Twelve reductively dechlorinating anaerobic
bacterial strains have been isolated from polychaete and
hemichordate burrow lining materials. Dechlorination
of 2,4,6-trichlorophenol, with phenol as the apparent
terminal product, was initially rapid but declined after
successive subculturing of all strains. A new method
for analysis of phospholipid fatty acid data, allowing a
more sensitive detection of changes in community
composition, also was developed as part of this project.
The investigators have initiated a series of species
transplant and halophenol contamination experiments in
the field to assess the ability of worms that do not
produce halogenated aromatic compounds to survive in
their presence and to determine whether dehalogenation
activity can be induced. Experiments creating refuge
from predators also are in progress to determine the role
of predators in controlling the representation of fauna
with halometabolites.
Additional near-term activities include the
completion of purification and characterization of the
bacterial dehalogenase enzymes and the development of
probes to characterize the genetic potential for dehalo-
genation by bacteria associated with the macrofauna.
Dehalogenation activities against brominated (common
biogenic) and chlorinated (common anthropogenic) halo-
aromatics will be tested. We have developed a new
method, based on neural networks, for analyzing phos-
pholipid fatty acid profiles from sediment bacterial
communities. The new method is more sensitive than
conventional cluster and principle components analyses
and yields greater resolution of community composition
than previous approaches.
65
-------
Influences of Watershed Land Use on
Stream Ecosystem Structure and Function
Judith L. Meyer, E.A. Kramer, M.J. Paul, W.K. Taulbee
Institute of Ecology, University of Georgia, Athens, GA
C.A. Couch
U. S. Geological Survey, Norcross, GA
Because natural resource management is more
frequently conducted at the landscape and ecosystem
levels, it must be determined whether the species and
habitat-based approaches currently used by water-quality
monitoring programs provide data relevant to assessing
interventions at the scale of the landscape or ecosystem.
This project has two objectives: (1) to determine how
differing patterns of land cover and land use in a
watershed alter stream ecosystem structure and function;
and (2) to compare measures of stream ecosystem
function (rates of community metabolism and transport
and cycling of carbon and nutrients) with traditional
chemistry-based, water-quality assessment and organ-
ism-based bioassessment. The project approach has
been to measure ecosystem structure and function in
streams draining eight watersheds in the Chattahoochee
River basin; two streams drain watersheds in each of
four land use categories (i.e., forest, agriculture, urban,
and suburban). Urban and suburban watersheds are
within metropolitan Atlanta. Poultry is the primary
commodity produced in the two agricultural watersheds.
The eight watersheds studied range in size from 47
200 km2; human population density ranges from 190
8,200 persons/km2. The progress in four areas of re-
search will be reported, including: (1) watershed
characterization, (2) patterns of material transport, (3)
organic matter dynamics, and (4) nutrient spiraling.
Land cover and land use has been characterized
in the watersheds, and impervious surface area has been
estimated in different land use classes. Percent im-
pervious surface increases with distance from the stream
in urban and suburban watersheds. An Index of Biotic
Integrity (ffll) based on fishes was lowest in watersheds
with much of the impervious surface area in com-
mercial, industrial, and transportation land uses. There
apparently is a threshold at about 30 percent impervious
surface area; when percent impervious surface rises
above that, IBI scores decrease sharply. Correlations
between IBI scores and forest cover in riparian areas
(50-meter buffer) were similar to correlations between
IBI scores and forest cover in whole watersheds, even
though riparian zones had 20 percent greater forest
coverage. This analysis suggests that in urban water-
sheds, the success of riparian restoration projects may
be limited by the influence of land use elsewhere in the
watershed.
Export of sediments, nitrate, total phosphorus,
and paniculate organic carbon were highest in agri-
cultural watersheds (see Figure 1). In contrast, dis-
solved organic carbon export was greatest in urban
watersheds, presumably a result of wastewater inputs
and runoff from impervious surfaces. These data, when
combined with standing crop and respiration data, will
provide valuable comparative measures of organic
carbon turnover in urban, agricultural, and forested
watersheds in the same physiographic area.
Organic matter inputs, retention, storage, and
metabolism were measured seasonally in the eight
streams. The amount of leaf litter falling directly into
streams was greatest in agricultural streams and lowest
in urban streams. However, lateral input of litter was
greatest in urban streams. Retention of organic matter,
measured as uptake lengths of fine particles (fluore-
scently labeled yeast), and standing crop of benthic
organic matter was lowest in urban streams. Leaf de-
composition rates were highest in urban and agricultural
streams and lowest in forested streams. All streams
were heterotrophic, and variance in net ecosystem
metabolism was best explained by the standing crop of
benthic organic matter.
Uptake lengths of nutrients were shortest
(200-400 meters) in forested streams and longest
(> 800 meters) in the urban streams. Uptake is best
explained by hydrologic properties of each channel,
especially the size of the transient storage zone and
its exchange rate. If human activities decrease tran-
sient storage zone size, then nutrient uptake length
will increase, thus indicating a reduction in the ability
of the stream to remove nutrients from the water
column.
66
-------
Sediment
Total P
Nitrate + nitrite
Forest Agriculture Suburban Urban
Watershed Type
Forest Agriculture Suburban
Watershed Type
DOC
POC
Urban
Figure 1. Average daily export of suspended sediment (metric tons km2 d"'), total phosphorus (kg km"2 d"1), nitrate and nitrite
(kg km2 d"1), dissolved organic carbon (kg km"2 d"1), and particulate organic carbon (kg km"2 d"1). Values shown are
standard deviations for the first year of the study.
means
67
-------
The Role of Hg (II) Reduction and
Chemical Speciation in Controlling the
Concentration of Mercury and Its Methylation hi Natural Waters
Franfois M.M. Morel
Department of Geosciences, Princeton University, Princeton, NJ
(A) Filtered river water
c
.0
"2
H—'
c
o
o
c
o
» Hg(0)
i Hg total
• Hg(ll)
(B) Unfiltered river water with KCI added
9 T :
M chloride added
Time (min)
Figure 1. Time series of dissolved gaseous Hg°, total Hg and Hg(II) concentrations in (A) filtered river water and (B) unfiltered river water
with 0.5 M chloride. At the onset, 8.3 nM of Hg° was added. Hg(II) was obtained by the difference between Hg° and total Hg.
High concentrations of mercury (Hg) that bioac-
cumulate in fish are a major environmental and human
health concern. Research has shown that there is a poor
correlation between the extent of Hg contamination and
the Hg content of fish in the environment. The
objective of this project is to elucidate the conditions
and processes that lead to a high rate of formation of
methyl mercury, the organic mercuric species that
accumulates in the food chain. This project's research,
which is based on experimental process studies in the
laboratory and in the field, has followed two principal
axes: (1) the determination of the chemical conditions
that lead to a high rate of bacterial uptake and
methylation of Hg; and (2) the quantification of the
transformation processes—dissolution, reduction, oxida-
tion—that control the concentration of the various forms
of Hg in aquatic systems.
In natural waters, Hg is methylated chiefly by
sulfate-reducing bacteria that live in anoxic waters or
sediments and respire by reducing sulfate to sulfide.
This project has tested the hypothesis that the presence
of polysulfides (formed by the reaction of sulfide with
elemental sulfur) might enhance the rate of bacterial
methylation of Hg via the formation of mercury poly-
sulfide complexes, which would keep the Hg in solution
and diffuse rapidly through bacterial membranes. This
project found that mercury poly sulfide complexes
indeed form and enhance the solubility of Hg in sulfidic
waters and that the addition of elemental Hg (in the
form of polysulfides or particulate sulfur) to cultures of
sulfate reducers enhances markedly their rate of Hg
methylation. This result is complicated, however, by
the changing concentrations of sulfur species in the
cultures, and control experiments are under way.
Besides precipitation of mercuric sulfide in
sediments, the major pathway of elimination of Hg from
aquatic systems is by reduction of ionic Hg to elemental
Hg that, as a gas, is then volatilized into the atmo-
sphere. Much attention has been paid to the chemical,
photochemical, and microbial processes that reduce
ionic Hg, and it has generally been assumed that, once
formed, all the elemental Hg eventually escapes to the
atmosphere. This study has demonstrated that this is
not the case: elemental Hg can be oxidized back to
ionic Hg in oxygenated water. This process, which
competes with volatilization and decreases the rate of
elimination of Hg from aquatic systems, depends both
on the chloride concentration in the medium and the
presence of appropriate particles. This is illustrated in
Figure 1, which depicts oxidation of elemental Hg in
unfiltered freshwater after the addition of chloride ions,
whereas little oxidation was observed in unsalted filtered
freshwater.
These results modify radically our understanding
of the processes that control the concentration and the
methylation of Hg in aquatic systems. Once confirmed
and extended, the results should provide a fundamental
understanding of the chemical and biological parameters
that determine the extent of bioaccumulation of Hg in
various ecosystems.
68
-------
Formation and Propagation of Large-Scale
Sediment Waves in Periodically Disturbed Mountain Watersheds
Gary Parker
St. Anthony Falls Laboratory, Department of Civil Engineering, University of Minnesota, Minneapolis, MN
water surface showing
a weak hydraulic jump
original channel bed
imposed sediment wave
a new delta composed of
finer sediment develops in
the ponded area upstream
of the sediment wave
sediment wave
some time later
Figure 1. Evolution of a sediment pulse from a sample numerical run.
Human-induced disturbances in watersheds such
as mining or timber harvesting, combined with certain
hydrologic conditions, often result in the formation of
large-scale waves of sediment that are moved (pro-
pagated) downstream. These waves have deleterious
effects on aquatic and riparian habitat and can result in
bridge and pipeline failures at crossings as well as the
loss of infrastructure such as roads adjacent to the river.
The goal of this research is to develop a single,
global numerical model encompassing backwater ef-
fects, abrasion, gravel sorting, transport and deposition
of fines, point and distributed sources of water and
sediment, and input hydrographs to predict the evolution
of sediment waves in rivers. A combination of field
work, laboratory experiments, and numerical experi-
ments is employed. The purpose of the field work and
laboratory experiments is to provide verification of the
numerical model.
A preliminary numerical model has been de-
veloped and verified against two small-scale laboratory
experiments. The results of a sample run are shown in
Figure 1. Also, Figure 1 shows successful modeling of
the ponding of water upstream, the formation of a
hydraulic jump downstream, the overall upstream
propagation of the deposit, and the formation of a new
downstream-propagating delta composed of fine
sediment at the upstream of the ponded region.
It is easy to suppose that sediment waves in rivers
should propagate downstream. However, results from
the numerical model reveal that dispersion always
dominates the process. That is, the apex of a sediment
wave usually does not move too far from its original
position. For gravel bed rivers, which usually have
steep slopes, the position of the apex is more likely to
move upstream than downstream. This finding is im-
portant in determining the fate of landslides in rivers.
The next steps of this research include: (1) per-
forming more experiments to verify the findings of the
numerical model and (2) fine-tuning the model for an
actual river (i.e., Redwood Creek in California).
Corresponding field data for this case have been
collected. The model will be modified to a user-friendly
form so that interested parties can use it as a predictive
tool.
69
-------
Multiscale Statistical Approach to Critical-Area
Analysis and Modeling of Watersheds and Landscapes
G.P. Patil
Center for Statistical Ecology and Environmental Statistics, Department of Statistics, Pennsylvania State
University, University Park, PA
W.L. Myers
School of Forest Resources and Environmental Resources Research Institute, Pennsylvania State University,
University Park, PA
Public agencies and corporate landholders are
developing and maintaining major environmental data-
bases for computerized mapping and analysis. The cost
of such databases increases rapidly with increasing
spatial detail. This project undertakes the development
of objective and efficient methods for determining
patterns of spatial variation from such databases—
particularly relating to water resources and landscape
ecology—and for making pattern comparisons between
databases having different levels of detail. A compari-
son of patterns emerging from data at different levels of
detail will enable managers to select the appropriate
spatial resolution necessary for particular purposes.
Cost efficiency also will be increased by the ability to
indicate likely zones of spatial uncertainty and to detect
areas of disagreement between alternate hydrologic
models, so that detailed data acquisition can be better
targeted and modeling will be more precise. Applica-
tion of these approaches will provide a better basis for
districting watersheds, thereby reducing the cost of
monitoring and detecting the need for remediation.
The general nature of the approach to multiscale
characterization of landscapes in this project is through
analytically induced pattern deterioration. The basic
idea is akin to survival of the fittest in natural selection,
whereby strong elements persist and weaker ones are
progressively vanquished. Although spatial pattern is a
complex of properties, one can think in terms of large
coherent patches that stand in contrast to their neighbors
as being the stronger elements, whereas small internally
variable patches having relatively similar neighbors are
the weaker ones. As with a large field of contestants in
a marathon, the weaker will drop out early while the
stronger will withstand considerable stress before being
subordinated. Multiscale character is thus reflected in
the rate and extent of deterioration for pattern elements
under the influence of a progressive degenerative pro-
cess.
The project considers three kinds of spatial data
layers: (1) surface data such as elevations or concen-
trations; (2) signal data such as remotely sensed
reflected energy; and (3) categorical data such as land
cover, soil, or geology. Analytical methods appro-
priate to each kind of data have been conceptualized,
and software capability for their application is being
developed. Echelon hierarchies of hillforms organized
as dendrogram trees apply to surface data, with
pruning of the tree being the degenerative process.
Patchwork representation by our "PHASE" version of
statistical clustering is appropriate for signal data, with
information capacity and patch size constraints serving
to induce pattern deterioration. Random and modal
filters bring about progressive deterioration of spatial
pattern in categorical data.
Progressive filtering of categorical data has been
studied intensively, and entropy-based measures of
patch structure are used to capture the trajectory of
pattern deterioration across scales. Transition matrix
models have been formulated that exhibit pattern
deterioration similar to land cover maps of selected
watersheds in Pennsylvania. Model parameters are
currently being estimated for 104 major watersheds in
Pennsylvania. The fitted transition matrices will be
studied for their ability to characterize and differentiate
among watersheds.
PHASE formulation not only provides for
multiscale exploration of signal data, but also offers an
alternative to conventional methods of handling and
analyzing digital image data in remote sensing. It
compresses multiband image data into a single layer of
patch-type identifiers with accompanying table(s) of
characteristics for patch types. PHASE compressed
image data is a value-added product that is beyond the
scope of usual copyrights on image data, thus permit-
ting unlimited distribution (see Figure 1). The com-
pression is sufficient to accommodate 10 scenes of
Lands at Thematic Mapper formulations covering
Pennsylvania on a single CD-ROM, along with a
variety of supplemental vector geographic information
systems (GIS) data.
70
-------
Figure 1. Central Pennsylvania sample PHASE from Landsat Thematic Mapper.
71
-------
Contemporary Water and Constituent Balances for
the Pan-Arctic Drainage System: Continent to Coastal Ocean Fluxes
Bruce Peterson
Marine Biological Laboratory, Woods Hole, MA
Charles Vorosmarty and Richard Lammers
University of New Hampshire, Durham, NH
Current predictions of global climate change
indicate a greater sensitivity in the Arctic than in
temperate or tropical regions. During the next century,
average temperatures could be warmer by several
degrees Celsius, resulting in changes to snow ac-
cumulation and melting, depth of soil thaw, and
precipitation. Changing plant communities could alter
evapotranspiration rates and therefore the water balance
across the entire pan-Arctic. These changes, in turn,
influence the Arctic Ocean through feedbacks between
freshwater discharge, sea ice, and deepwater formation.
The primary goal of this research is to provide a
quantitative, contemporary estimate of spatially dis-
tributed water balance, discharge through rivers, and
associated fluxes of constituents from the entire ter-
restrial drainage system to the Arctic Ocean.
Our approach combines a data-rich geographic
information system (GIS) with redundant models
applied to establish the pan-Arctic freshwater and
material balances. Aerological estimates of water flux
conergence and precipitation minus evaporation (P-E)
are calculated using rawinsonde (a balloon-borne atmos-
pheric sounding instrument) archives. This has been
performed over the Mackenzie, Ob, Yenesei, and Lena
drainage basins. In parallel, ongoing development of
a permafrost water balance model (P/WBM) has
yielded a first set of estimates of water budgets at 0.5
degrees (latitude x longitude) resolution. Freshwater
runoff from each of these cells is then cascaded down
the river network and discharged into the Arctic Ocean.
A database of river discharge gauges (n = 3,500) has
been assembled covering land mass contributions from
the United States, Canada, and Russia. Additionally,
an improved digital stream network has been
established to provide the best possible routing of water
and constituents downstream to the Arctic Ocean.
About 555 gauges from large catchments, greater than
15,000 km2, were used to validate the network. Figure
1 shows the improvement in the comparison of ob-
served and simulated drainage areas using these gauges.
The new discharge data set from the large basins
represents an increase of data resolution by an order of
magnitude over previous data sets (from 89 to 646
gauges).
Successful simulations using P/WBM were
carried out for Imnavait Creek, Alaska. Upscaling of
the model is currently taking place with applications to
Boreal zone catchments and the entire pan-Arctic
region. Preliminary simulations have taken place that
span the full 30-year timeframe for this project
(1960-1990). High year-to-year variability in many
locations potentially masks any progressive climate
change.
Preliminary results from rawinsonde data
(1974-1992) show high interregional variability in P-E
for the Eurasian watersheds. This variability is driven
by an occasional large summer negative flux con-
vergence with southerly outflow of water vapor from the
basins. In some years, this negative flux is large
enough to cause annual P-E to turn negative. Winter
balances appear to be realistic. The summer problem
appears to stem from the inability of the rawinsonde
network to "capture" the high topography at the water-
shed divides, an effect that is currently being remedied.
Simulations using P/WBM over the entire
pan-Arctic region show large errors when compared
with observed river discharge records. A significant
source of this error is in precipitation, particularly over
mountainous regions due to: (1) bias in the locations of
the precipitation gauges (typically at low elevations) and
(2) the sparse network of meteorological stations in the
northern regions giving interpolated fields a high level
of uncertainty. Modifications to these interpolated fields
are under way and combine existing elevation and
rawinsonde data to improve temperature and precipi-
tation estimates.
There continues to be a need to better identify
the sources of error in the components of the pan-
Arctic water balance. This issue will be pursued in the
context of looking at the consistency between rain-
sonde, water balance model results, and discharge
station records.
72
-------
Drainage Area (sq. km) - Before Edit
CO
1e6
1e5
1e4
1e3
• •••-" - '.x.r
1e4
. .
"•' "
1e5
1e6
Observed
Drainage Area (sq. km) - After Edit
1e6
O
CO ie5
1e4
1e3 -
1e4
1e5
1e6
Observed
Figure 1. Contemporary water and constituent balances for the pan-Arctic drainage system.
73
-------
Modeling Temporal Rainfall
via a Fractal Geometric Approach
Carlos E. Puente
Hydrologic Science Program, University of California, Davis, CA
ox,
Figure 1. Alternative derived measures (DY's) arising from the same fractal interpolating function / using alternative parent multifractal measures
(DX's).
Accurate rainfall modeling is of vital importance
for the proper management of our environment.
Rainfall descriptions are required, among others, to
model pollution migration, address issues related to cli-
mate change (i.e., global circulation), estimate extreme
weather events, and manage our watersheds. Although
several sophisticated (stochastic) rainfall models exist,
they do not capture the variability observed at a fixed
location when a storm passes by. They approximate the
irregular and intermittent rain patterns by superimpos-
ing, for example, randomly arriving rectangular pulses
and consequently preserve only some statistical features
of the rainfall series (e.g., mean, variance, correlations,
etc.). Because these representations are typically
limited by their analytical tractability and because there
has been a recognition of chaotic and multifractal effects
in rainfall, a new class of rainfall models has been
developed by our group. The basis for these new mo-
dels is the belief that predictability could only be im-
proved when the overall trends and observed details
present in rainfall events are considered explicitly.
The goal of this project is to develop a better
understanding of temporal and spatial rainfall patterns
within the State of California under alternative climatic
conditions. Parsimonious representations of rainfall
records under alternative climatic conditions will be
derived so that a classification of observed rainfall pat-
terns may be elucidated. This task will be accomplished
by employing the fractal-multifractal representation to
encode rainfall records throughout the State of Cali-
fornia as suitable fractal transformations of appropriate
(turbulence-related) multifractal measures. Once the
time series is represented by this approach, a classifica-
tion of rainfall events will be attempted via surrogate
parameters that define the fractal transformation and the
multifractal measure. This research should lay a firm
foundation for a new approach towards hydrologic dy-
namics in terms of surrogate geometric information.
This methodology should be viable because it concen-
trates on capturing what is observed (i.e., the geometry
of rainfall series).
During this year, studying the nature of the
derived measures that are generated via the fractal-
multifractal representation has continued under a variety
of alternative scenarios. Simple modifications to the
original procedure could be made so that data may be
generated, similarly to data obtained in rainfall records,
transforming a variety of Cantorian measures (measures
with holes) via the same fractal interpolating function.
The existence of these extensions is most welcome
because they are quite parsimonious, requiring only five
surrogate parameters. Such extensions are being em-
ployed to encode actual data sets in California, solving
an inverse optimization problem via genetic algorithms
that minimizes squared differences of real and predicted
records. A study of the encodings obtained and their
relationship to climatic conditions will be performed
during the next year.
74
-------
Effects of Food Web Structure and Nutrient Loading
on Lake Productivity and Gas Exchange With the Atmosphere
Daniel E. Schindler, Stephen R. Carpenter, and James F. Kitchell
Center for Limnology, University of Wisconsin, Madison, WI
Jonathan J. Cole and Michael L. Pace
Institute for Ecosystem Studies, Gary Arboretum, Millbrook, NY
g T
o ™
20
Q- O)
E rl
2.8
2.6
2.4
2.2
2.0
A
^ ^
minnow.
__ _ A
bass
0 1.0 2.0 3.0
P input (mg C rrV2d'1)
160
80
x
13 ^
*= E 0
O
O
-80
-160
B
A O
0 400 800 1200
Primary production (mg C m"2d"1)
Figure la. Effect of P input rate on primary production in four
lakes with contrasting food web structures. Each symbol represents
a summer mean for 1 lake year combination from 1991 to 1995.
Lakes characterized by high planktivory and low-grazing rates are
shown by circles. Triangles represent lakes with piscivores and
high-grazing rates. Lakes that were experimentally fertilized are
denoted by filled symbols: gray symbols are lake years before
enrichment; black symbols are during enrichment.
Figure Ib. Relation between calculated carbon dioxide flux
between lakes and the atmosphere and the estimated primary
production rate in 1992-1995 (correlation -0.84). Positive values
of carbon dioxide flux represent net flow out of lakes, and negative
values represent flow into lakes. The dashed line represents con-
ditions in equilibrium with the atmosphere. Symbols are as de-
scribed for (la).
When lakes are viewed in a landscape per-
spective, lake ecosystem dynamics are viewed as the
outcome of interactions between drivers at the landscape
scale (e.g., water, nutrients, and organic carbon) and
internal processes (e.g., nutrient cycling and food web
interactions). In this project, whole-lake experiments
were performed to evaluate the interactions between
food web structure and nutrient loading in controlling
primary productivity of lakes. Food webs were manipu-
lated by changing the dominant fish species, which
resulted in a restructuring of the zooplankton grazer
communities. Lakes with piscivorous fishes (bass) have
large-bodied zooplankton that are very effective algae
grazers. Lakes with planktivorous fishes (minnows)
have small-bodied zooplankton that are much less
effective at suppressing algae growth. Experimental
additions of inorganic nitrogen and phosphorus resulted
in substantial increases in algae growth (primary pro-
duction) in all lakes. However, primary production in
lakes with piscivorous fishes was much less responsive
to nutrient addition than in lakes with planktivorous
fishes. This effect of food web structure on primary
production was sustained over phosphorus loading rates
that ranged from less than 0.2 mg P/m2/d to greater than
3 mg P/m2/d (see Figure la).
Food web effects on primary productivity also had
implications for carbon dioxide exchange between lakes
and the atmosphere. Prior to nutrient enrichment, all
study lakes were net sources of carbon dioxide to the
atmosphere. Nutrient enrichment increased algae de-
mand for carbon dioxide, causing lakes to become net
sinks for atmospheric carbon. At identical nutrient load-
ing, atmospheric carbon invasion was greater in a lake
with planktivorous fishes and low grazing than in a lake
with piscivorous fishes and high grazing (see Figure Ib).
Carbon stable-isotope distributions corroborated the
drawdown of lake carbon dioxide and traced atmospheric
carbon transfer from algae to top predators. Thus, top
predators altered ecosystem carbon fixation and linkages
to the atmosphere. Our study reinforces the idea that
exploitation of top predators and the introduction of
exotic species that cause changes in trophic structure are
an aspect of environmental change that have important
implications for structure and function of ecosystems.
75
-------
A Comparative Institutional Analysis of Conjunctive
Management Practices Among Three Southwestern States
Edella Schlager
University of Arizona, Tucson, AZ
Among the more popular contemporary
recommendations for improved watershed use and
protection is conjunctive use of surface and under-
ground water resources. Conjunctive use involves
the coordination of surface water supplies and
storage with groundwater supplies and storage for
purposes of sustainable watershed use and enhanced
watershed protection. Conjunctive use in a water-
shed requires a great deal of joint effort among
human beings, the prospects for which will be af-
fected strongly, though not exclusively, by insti-
tutional arrangements that define organizational
forms and jurisdictions and provide incentives and
disincentives to individuals. Using a comparison of
Arizona, Colorado, and California, this project
seeks to advance the theoretical and empirical
understanding of the relationships between insti-
tutional arrangements governing the allocation, use,
and protection of water resources and the develop-
ment, implementation, and performance of con-
junctive use programs.
The approach taken in this project is one of
institutional analysis. Institutional analysis is based
on in-depth comparative analyses of the per-
formance of diverse institutional arrangements. The
technique of institutional analysis was developed
during the last quarter-century at Indiana Uni-
versity, primarily by Elinor and Vincent Ostrom
and colleagues at the Workshop in Political Theory
and Policy Analysis (a complete description of the
approach is available in E. Ostrom, 1990). In
essence, institutional analysis involves: (1) the use
of a systematic framework for classifying insti-
tutional arrangements (conceived as sets of rules)
and levels of action; (2) viewing the development of
institutional arrangements as processes of inten-
tional (individual or collective) human choice and
action constrained by physical circumstances and
community context, recognizing that human beings
are capable of multiple levels of action (e.g., as
rule makers and rule followers, as organization
designers as well as organization members); and (3)
assessing the effects of institutional arrangements
on human choice and action as well as evaluating
the effects and effectiveness of institutional ar-
rangements relative to their intentions.
Given a well-developed and systematic
framework, institutional analysis allows for com-
parisons across situations on a variety of dimen-
sions. For instance, community characteristics and
physical settings may be held constant, while rules
are varied, permitting an understanding of how
different configurations of rules promote various
outcomes.
Although Arizona, California, and Colorado
face similar physical problems of water distribution
in the same geographic region, the three states have
devised markedly different approaches to con-
junctive water management. Arizona has taken a
more state-centered approach. The primary entities
engaged in conjunctive water management are the
Central Arizona Water Conservation District
(CAWCD), the operator of the Central Arizona
Project, and the Arizona Water Banking Authority
(AWBA). By far, the dominant conjunctive water
management approach is the exchange of surface
water for groundwater credits, or in lieu recharge.
CAWCD makes available to CAP water for
agriculture in exchange for agriculture ground-
water. CAWCD uses the groundwater for drought
protection and to enhance the reliability of the
CAP. If a pumping station or a portion of the canal
becomes inoperable, CAWCD can deliver ground-
water to its customers. Although California uses
conjunctive water management primarily as in-
surance against water shortages, it is primarily local
level jurisdictions, and not state entities, that
engage in it. The dominant conjunctive water
management approach is constructed recharge
projects whereby surface water is stored under-
ground through injection wells and spreading
basins. Due to recent changes in state law, in lieu
recharge is beginning to be used in California. Like
California, Colorado's conjunctive management
activities occur primarily among local level
jurisdictions. Unlike both Arizona and California,
however, conjunctive water management is used
primarily to protect the rights of senior surface
water users under the state's prior appropriation
doctrine, and not for drought protection. Colorado
jurisdictions use a combination of in lieu recharge
and constructed recharge projects.
Institutional arrangements govern the type,
development, and uses of conjunctive water
management projects. Data collection for this
project is in its preliminary stages. As further data
are collected, issues of institutional performance,
the allocation of costs and benefits, and the en-
vironmental impacts of conjunctive water manage-
ment projects will be explored.
76
-------
Water and Sustainable Development
in the Binational Lower Rio Grande/Rio Bravo Basin
Jurgen Schmandt
The Houston Advanced Research Center (HARC), The Woodlands, TX, and The Instituto Technologico y de
Estudios Superiores de Monterrey (ITESM), Monterrey, Nuevo Leon, Mexico
In its 1994 Regional Assessment of Water
Quality in the Rio Grande Basin, the Texas Natural
Resource Conservation Commission made two recom-
mendations for further research and policy develop-
ment: (1) an improved understanding of water issues
as they relate to the binational border dynamic and (2)
development of lasting links across the U.S.-Mexico
border. The goals of this project are to: develop
reliable watershed-based data sets from both Mexican
and U.S. sources to analyze the whole binational region;
analyze water resource issues as critical factors in the
region's long-run planning and sustainable development;
and engage researchers, policymakers, and the civic
community to help answer the question, "What actions
can be taken to achieve sustainable development in this
rapidly growing, drought-prone, environmentally fragile
watershed?"
Mexican and U.S. researchers are paired in teams
to perform data compilation and analysis in the follow-
ing areas: (1) water supply and demand, (2) water
quality, (3) population and socioeconomic conditions,
(4) ecology, (5) water management and institutions, and
(6) geographic information systems (GIS). The teams
produce an integrated "baseline report" of current
conditions in the region, which then provides the basis
for subsequent analysis of alternative future scenarios.
Scenarios are developed by combining regional demo-
graphic projections to the year 2030 with alternatives for
future water availability, irrigation technologies, and
management practices.
Involvement of the community is a key element
of the project, and mechanisms for stakeholder input
and consultation are incorporated into the research pro-
cess. Researchers and graduate students from Mexico
and the United States (as part of a joint University of
Texas-ITESM Policy Research Project) have conducted
a survey among stakeholders and community leaders.
In addition, workshops are scheduled involving re-
searchers and stakeholders from both sides of the
border. These outreach components gather insights into
the community's concerns, perceived threats, and
opportunities related to water and development in the
watershed.
The preliminary findings of the study include the
following: (1) The population will double by the year
2030, while surface water supply will stay at current
levels or decrease. (2) The single source of surface
water for 2 million people and intensive irrigated
agriculture in the Lower Rio Grande—the combined
Amistad-Falcon international reservoir system—is at its
lowest level since coming online 25 years ago. (3) The
current drought in the upstream parts of the watershed
began in 1994 and continues unabated. Mexican far-
mers lost two harvests, and several irrigation districts in
Texas went dry in 1996. (4) The adequacy of the exist-
ing water supply system under drought conditions is
being tested for the first time. During the drought of
record in the 1950's, only 500,000 people lived in the
same area where 2 million live today. Amistad reser-
voir did not yet exist, and Falcon reservoir came on line
in the middle of the drought. (5) Significant changes in
fauna appear to be correlated with decreasing stream-
flows, the proliferation of exotic species, and chemical
pollution. (6) GIS analysis indicates that irrigated land
area in the basin is significantly larger than official
records indicate. (7) Rapid growth and development in
the San Juan and Conchos River basins in Mexico could
significantly reduce the amount of water these tributaries
contribute to the main stem of the Lower Rio Grande/
Bravo. (8) Initial runs of the agricultural model suggest
that significant long-term reductions in water supply
would at most have a very small impact on the regional
economy because new irrigation technology and crop-
ping patterns would be adopted. (9) Transfers of agri-
cultural water to municipal use occur in Texas through
the use of water markets and joint infrastructure im-
provements on the part of cities and irrigation districts.
The significance of these findings is that the
system is flexible—small increases in irrigation ef-
ficiency can release significant amounts of water for
other uses, hi addition, a tremendous opportunity exists
for "win-win" arrangements in which municipal water
suppliers agree to finance improvement in irrigation
efficiency in return for the rights to the water saved.
Furthermore, the existing regional water market holds
great potential for more efficient use of water.
The research teams plan to meet in January 1998
to discuss the scenario analysis portion of this project.
A community workshop involving researchers and
stakeholders is scheduled for May 1998, and the
integration of scenario analysis is planned for spring
1998.
77
-------
'"51
Basin Boundary -
United States
of America
tad Reservoi
Falco'n. Reservoir
"1
McAllen
'*>
Estados Unidos
Mexicanos
Shaded area represents primary study region
TAMAUL1PAS
Figure 1. Lower Rio Grande/Rio Bravo Basin.
78
-------
Environmental Change and Adaptive Resource
Markets: Computer-Assisted Markets for Resource Allocation
Vernon Smith, S. Rassenti, E. Hoffman, R. Howitt, A. Dinar
University of Arizona, Tucson, AZ
Water delivery systems have traditionally been
operated by large centralized authorities due to the
natural monopoly nature of the industry and the quasi-
public good nature of a reliable and clean water supply.
In most countries, the era of meeting new demands by
increased supply development has been terminated by
economic and environmental costs, environmental
change, and the shortage of suitable sites. Like the
energy and natural gas industries, the water supply
industry will have to look to market mechanisms to
reallocate existing water supplies and increase efficiency
to meet future demands.
This project's goal is to develop a formal mech-
anism for evaluating alternative market structures for
integrated water systems. This goal will be achieved by
the following objectives: (1) developing a structural
market model for the California water industry, (2)
deriving value and cost functions for the different
components in the structural model, (3) developing an
experimental design that uses the water market model,
(4) running experiments on alternative market options
using economically motivated subjects in a replicable
design, and (5) developing a remote (Internet) ex-
perimental ability that allows extension of the market
experiments to market participants in the field.
A prototype of a smart water market for the
central California water industry has been designed and
tested at the Economic Science Laboratory in Tucson,
Arizona. The input needed for the player at each node
is in the form of a step function representing their
willingness to pay for different quantities and qualities
of water. Experiments on market price convergence and
market efficiency were conducted using students whose
earnings were based on their market success. Players
represented agricultural, urban and environmental
interests, and were given initial endowments of money
and water rights reflecting the current allocation in the
state. After initial training, the experiments were run for
15 trading periods in four experimental sessions.
Student participants were paid between $2 and $70 as a
result of their trading actions.
The first set of experiments with the simplified
model obtained realistic results. A "smart" computer-
assisted market with a double-blind auction structure
was used. Under this structure, buyers and sellers do
not have to reveal their true valuation for the goods to
achieve an outcome that is close to the theoretical
optimum under perfect competition. However, the
relative share of their defined supply or demand
schedule that the players were able to realize differed
significantly among players and groups. Alternative
allocations of water transport facilities made a signi-
ficant difference in the outcome between players. The
outcomes of the initial experiments shows that it is
possible to reproduce price efficiency and convergence
results in natural resource markets. The results were
obtained using players that included both users and an
environmental agent, who was only interested in the
level of flows through a certain sensitive node. The
initial results show that the novel concept of market
interactions between resource users and environmental
interests can be modeled in a replicable manner using
this approach.
The next steps are: (1) extending the model to a
larger number of participants in each sector of the water
economy; (2) incorporating a water quality component
into the supplies, demands, and the smart market soft-
ware; (3) developing Windows NT-based Internet soft-
ware to enable use by outside participants in the exper-
iments ; and (4) extending the smart market analysis to
water allocation problems in a developing country.
Through step (3), the performance of participants who
are actually working in the industry can be compared
with the student-based results. This latter work will have
benefits as a validation of the formal results from
student-based experiments and also will make the smart
market a practical tool for developing market skills by
members of the industry.
79
-------
Detecting Fecal Contamination
and Its Sources in Water and Watersheds
Mark D. Sobsey
University of North Carolina at Chapel Hill, Chapel Hill, NC
MS2
SW1
SW6
SW11
SW16
SW21
SW26
GA
SW2
SW7
SW12
SW17
SW22
SW27
QP
SW3
SW8
SW13
SW18
SW23
SW28
FI
SW4
SW9
SW14
SW19
SW24
SW29
SP
SW5
SW10
SW15
SW20
SW25
SW30
Probe nA Probe HB
Figure 1. Lysis zone hybridization for identification of 30 F RNA coliphage isolates from surface water by oligoprobes IIA and nB. Positive
controls consisting of 5 prototype F RNA coliphages (MS2, GA, Q(3, FT and SP) are shown on the top. Thirty field isolates (SW1-SW30)
are arranged on the key.
Current methods to detect, quantify, and predict
water and watershed quality with respect to fecal
contamination are inadequate and unreliable. Because
so little is actually known about the levels and sources
of fecal contamination in water, there are great
uncertainties about the human and environmental health
risks from the pathogens associated with contaminated
water. The objectives of this research are to further
develop, evaluate, and apply new and improved mea-
sures of human and animal sources of fecal contami-
nation in water and watersheds. Two main criteria will
be used to judge the value and reliability of the
candidate indicators: (1) their ability to predict the pre-
sence and concentrations of the pathogens relative to
the sources of the organisms and (2) their ability to
predict the presence, persistence, transport, and fate of
the pathogens in response to natural environmental
processes.
The investigators have developed reliable
methods for the simultaneous concentration of viruses,
bacteria, and protozoans from water using disposable,
hollow fiber ultrafilters. Recoveries of seeded organ-
isms from surface water are 50-75 percent and the
coefficients of variation (CVs) were relatively small.
The new methods previously developed to
concentrate human gastroenteritis viruses from water
were field tested on a groundwater supply of drinking
water that was implicated in an outbreak of viral
gastroenteritis. The virus isolated from the water was
genetically identical to the virus independently isolated
from stool specimens of ill persons by another lab-
oratory. This is the first time such a virus was suc-
cessfully recovered and detected in an incriminated
water supply, and the work was recently published
(Bellar et al., JAMA 278(7):563-568).
New and improved methods to detect indicator
viruses known as coliphages using "gene probes" were
improved and field tested (see Figure 1). Also, the
researchers found that it was possible to distinguish
fecal from nonfecal indicator viruses in water on the
basis of bacterial host range and growth temperature.
Coliphages of fecal origin grow well at temperatures of
42-45° C, and they have a narrow host range. Coli-
phages of possible nonfecal, environmental origin will
grow well at temperatures of 37° C or lower and not at
42-45° C. Some of these environmental coliphages have
a wider host range and will grow on some other
bacteria. In field samples of surface waters impacted by
known sources of human and animal fecal contamination,
levels of various fecal indicator organisms generally
increased when fecal waste sources were nearby. Even in
areas where animal wastes were managed by land
application, adjacent surface waters showed evidence of
fecal contamination. Levels of indicators were 10-1,000
times higher than background levels.
Coliphages in the water of stations near animal
waste sources were primarily animal groups, and
coliphages in the water of stations impacted by muni-
cipal wastewater discharges or nonpoint source human
wastes were primarily human groups. At stations where
both sources were impacting water quality, both groups
were detected. These results indicate that sources of
fecal contamination can be reliably identified and traced
by identifying and grouping the coliphages that are
detected using gene probes.
Studies on detecting and quantifying microbial
pathogens and indicators of fecal contamination in
waters and watersheds impacted by human and animal
fecal contamination will be continued in the watersheds
now being monitored. An additional watershed will be
added during the next 12-month period. Also, the new-
ly developed methods to detect human pathogens of
fecal origin will be applied to these waters to determine
if the known human and animal waste sources are
important contributors of these protozoan and bacterial
pathogens. The reliability of the various microbial
indicators in predicting the presence and concentrations
of these pathogens also will be examined.
80
-------
Ecoregion-Specific Comparison of Stream Community Responses to
Nutrient Gradients Using Both Survey and Experimental Approaches
R. Jan Stevenson
University of Louisville, Louisville, AT
Mike WUey
University of Michigan, Ann Arbor, MI
Joe Holomuzki
Transylvania University, Lexington, KY
Landscape Development
I
Nutrients
Algal mat &
filaments
Flow Stability
^. Scraping
, - ' Herbivores
Collector/
Gathers
Predators
Most evidence that exists for regulatory purposes
comes from toxicological studies, which are often
conducted in laboratory experiments. This project is
aimed at providing a credible scientific basis for setting
regulatory criteria relating to changes in stream
ecosystems along an environmental gradient.
Nutrient loading commonly causes an increase in
primary productivity and a replacement of clean-water
ecosystem algal-invertebrate communities with a new
suite of organisms that results in very different stream
structure and function. Quantitatively assessing the
relationships between nutrient concentrations and algal-
invertebrate communities in streams has been challeng-
ing because stream ecosystems are so physically
variable and biologically dynamic. This project is
designed to develop a better predictive understanding of
nutrient effects on stream communities in ecoregions
with different hydrological stability. In this project, the
investigators are using stream surveys with large within-
ecoregion sample sizes, experiments, and analyses of
large preexisting databases.
Initial results indicate that chances for high
accrual of algae with increasing nutrients is greater in
the hydrologically variable streams of Kentucky than in
the stable streams of Michigan. Invertebrate abun-
dances, much lower in Kentucky than in Michigan, do
not respond to nutrients in Kentucky but do show a
positive response to nutrients in Michigan.
Three groups of experiments were conducted
during the summer of 1997 to complement previous
work. In one experiment, the hypothesis that thresholds
in nutrients occur where algae can grow faster than
grazers can consume them was tested in experimental
Figure 1. Positive and negative effects on invertebrates in streams.
Nutrient effects on algae and invertebrates in streams are highly
dependent upon climate, geology, and flow stability of streams.
Nutrients stimulate growth rates of algae. If unregulated by grazers,
nutrients stimulate the change of algae from thin mats to thick mats
and filaments. High flow stability allows invertebrate densities to
increase and regulate accumulation of algal mats and some filamentous
algae. Low flow stability, either from scouring spates or drought,
constrain invertebrate densities so that algae can easily overgrow
substrata when nutrients are abundant. When these algae overgrow
substrata, changes in invertebrate communities due to the positive
effect of the abundant algae on some invertebrates are observed.
Little evidence has been found that overgrowths of algae have negative
effects on invertebrates in streams (solid lines indicate positive effects;
dashed lines indicate negative effects).
streams with 96 experimental reaches. Eight different
nutrient concentrations, 2 different grazer treatments,
and 6 replicates of each treatment were manipulated in
the 96 reaches. In a second experiment, nitrate and
phosphate concentrations were varied in 72 different
experimental streams to determine the thresholds in
nutrient concentrations that saturate algal growth rates
when biomass is low and the concentrations that cause
peak algal accrual on substrata (nuisance algal growths).
Nine different nutrient levels of nitrate and phosphate
were used with 4 replicates of each treatment. The third
set of experiments assessed the relationship between
invertebrates and chariging habitat conditions when
algae accumulate. Preliminary results indicated over-
growths of streams by long, filamentous green algae
(Cladophora) can have a positive effect on some stream
invertebrates, particularly hydropsychid caddisflies.
These experiments showed that these invertebrates
prefer rocks with Cladophora as compared to bare
rocks, and stoneflies preyed more successfully on these
caddisflies on bare rocks than on rocks covered with
Cladophora.
Future work will develop stronger quantitative
predictions of the effects of specific concentrations of
nutrients on algae and invertebrates in streams in diff-
erent ecoregions. This will be accomplished by evalu-
ating data from 2 years of algae and invertebrate data
from 130 streams in Kentucky and Michigan and from
two sets of experiments in which nutrient concentrations
were manipulated. In addition, relations between in-
dices of biotic integrity, indices of environmental
stressors, and nutrient concentrations will be assessed
with data from the 130 streams that were in the study.
81
-------
Holocene Floodplain Development as a Function of Climate Change
and Human Activities: The Arroux and Loire Rivers, Burgundy, France
Eric C. Straffin and Michael D. Blum
Department of Geosciences, University of Nebraska-Lincoln, Lincoln, NE
Holocene age floodplains of the Arroux and Loire
Rivers contain a regionally extensive record of fluvial
activity, which can be interpreted within established
archaeological, paleoenvironmental, and historically
documented climatological frameworks. River mor-
phology and associated deposits record the magnitude
and frequency of floods, associated with climate regimes
driven by global atmospheric circulation. The
reconstruction of past fluvial activity and correlated
discharge regimes facilitates the prediction of fluvial
adjustments to climatic and anthropogenic changes
expected in the near future.
The southern Burgundy region of France is
situated at the boundary between maritime, continental,
and Mediterranean climates. During the period of his-
torical monitoring, the position and dominance of these
climatic regimes has shifted as a function of changes in
atmospheric circulation. Discharge regimes have fluctu-
ated as well, but details differ regionally due to geologic
and antecedent conditions, and the latitudinal position
and style of storm tracks. Historic analog models sug-
gest that the dominance and persistence of a particular
climate regime should result in regionally circum-
scribed, but isochronous alluvial fills containing
differing styles of floodplain construction. For example,
the Little Ice Age in Europe (ca. 1300-1850 A.D.) was
marked by episodic cooling, resulting in large floods
and flood variability between basins, while Medieval
times (ca.700-1300 A.D.) were marked by warmer con-
ditions, and more steady flood regimes, although the
magnitude of floods varied between western and eastern
drainage basins.
Throughout Europe, enhanced fluvial activity
occurred during the period 1250 to 1550 A.D., es-
pecially during transitions from dominantly cold sub-
periods to warmer conditions. The Loire and Arroux
valleys follow the same regional trend, with pre-
dominantly overbank deposition of sands and silts
during the Neoglacial period, which resulted in
straightened, chute-cut-off channel patterns and the
burial of older alluvial surfaces. This style of fluvial
activity began ca. 1050 B.C., contrasting sharply with
earlier gravelly, laterally accreting, meandering rivers
characteristic of the mid-Holocene. This morphological
and sedimentological change predates significant human
impacts on the landscape, is regional in extent, and
when combined with proxy evidence supports the notion
that climate change has been the dominant mechanism
in driving fluvial adjustments.
The separation of climatic from human influences
on the response of rivers is often difficult to determine,
but there is increasing recognition that at regional
scales, climate has been the dominant mechanism for
changes in fluvial dynamics. Human activities have
conditioned these responses locally, and at regional
scales have served to amplify existing climatic in-
fluences. In southern Burgundy, Roman land use prac-
tices were introduced ca. 50 B.C., and noticeable hu-
man impacts in upper portions of the Loire basin
occurred even later. Roman and Medieval land use
practices may have increased the overall thickness of
overbank deposits but were most likely not responsible
for the overbank discharges permitting such deposition
at regional scales.
Land use changes in this rural region have
changed little over the last several hundred years, and
if we can use the past response of the Loire and
Arroux Rivers as analogs to what may happen to this
region with future climatic change, we may expect
that with decreased temperatures and increasingly
meridional circulation, variability in flood frequency
will increase, resulting in flood-prone valley bottoms.
Moderate increases in temperature and increased
zonal circulation may produce less variable discharge
regimes, but more regular inundation of lower
floodplain positions.
82
-------
Response and Compensation to a
Bivalve Invasion by an Aquatic Ecosystem
David L. Strayer, Nina Caraco, Jonathan J. Cole, Stuart Findlay, and Michael L. Pace
Institute of Ecosystem Studies, Millbrook, NY
% change due to zebra mussel
o
o
microzoopl.
phytoplankton -
macrozoopl. -
unionids
macrobenthos
susp. solids
DIN
secchi
bacteria
SRP
Figure 1. Variation in the response of North American freshwater ecosystems to the zebra mussel invasion. All ecosystems have estimated zebra
mussel filtration rates of 70-125 percent of the water column per day. The solid line shows the response of the Hudson River, while
the points show responses from Lake Erie, Lake St. Clair, Saginaw Bay, and Oneida Lake (from data of Holland [1993], Fahnenstiel
et al. [1995], Fanslow et al. [1995], Holland et al. [1995], Johengen et al. [1995], Maclsaac et al. [1995], Mellina et al. [1995], and
Horgan [1996]). DIN=dissolved inorganic nitrogen; SRP=soluble reactive phosphorus.
Bivalves are so abundant in many shallow aquatic
ecosystems that their filter-feeding regulates pelagic and
benthic variables. Humans cause sudden changes in bi-
valve populations through the introduction of alien
species, water pollution, and overharvesting, which may
have large effects throughout the ecosystem. The in-
vestigators postulate that the response of an ecosystem
to changes in bivalve populations will depend on the
characteristics of the ecosystem as well as on the
numbers and biological characteristics of the bivalve.
Thus, two ecosystems might respond very differently to
the same bivalve population. Specifically, the project
hypothesis is that the ecosystem response will depend on
some characteristics that are fixed (e.g., morphometry,
hydrology) as well as flexible compensatory pathways
(e.g., functional substitution of species, environmental
feedbacks, and induced defenses).
The investigators have been looking at ecosystem
responses to bivalve populations by studying the response
of the Hudson River to the ongoing zebra mussel invasion
and comparing its response to that of other ecosystems,
where large bivalve populations appeared or disappeared.
This project combines long-term field studies, models,
small-scale experiments, and cross-system comparisons.
Zebra mussels first appeared in the Hudson in 1991 and
became the dominant consumer in the ecosystem by the
end of 1992. Estimated clearance times for the entire
volume of water in the freshwater tidal Hudson River
have been in the range of 1-4 days. Consequently, large
changes have emerged through the ecosystem. Densities
of phytoplankton and small zooplankton, which are
consumed by zebra mussels, fell by 80-90 percent, and
the composition of the remaining phytoplankton shifted
markedly from preinvasion communities. Populations
of native bivalves fell by 50-70 percent in response to
the loss of their phytoplankton food. Bacterial densities
doubled, perhaps because of losses of small zooplankton
that feed on bacteria. Water clarity and dissolved re-
active phosphorus both rose as phytoplankton densities
fell. The density and extent of rooted plant beds may
have increased as the Hudson's water cleared. Densities
of other sediment-dwelling animals fell in deep waters
but rose in shallow waters, probably in response to
changing patterns of primary production.
Generally, other systems exposed to bivalve
grazers show qualitatively similar responses to the
Hudson, but with very large quantitative differences
across systems (see Figure 1). These differences pro-
bably reflect the extent to which ecosystem character-
istics constrain the strength of the interaction pathways
that connect the bivalve with the rest of the ecosystem.
Specifically, cross-system comparisons support the idea
that ecosystem responses depend on vertical mixing of
the water column; the size, structure, and specific com-
position of the plankton; the nature of factors that
otherwise limit growth rates of plankton (i.e., nutrients,
light, advection, etc.); the diet breadth of consumers;
the magnitude of allochthonous inputs that can substitute
as food for phytoplankton; inputs of physical energy to
the system; and system morphometry.
83
-------
Scaling Up Spatially Distributed
Hydrologic Models of Semi-Arid Watersheds
David G. Tarboton \ Christopher M. U. Neale 1, Keith R. Cooley 2, Gerald N. Flerchinger2,
Clayton L. Hanson 2, Charles W. Slaughter2, Mark S. Seyfried2, Rajiv Prasad 1, Charlie Luce ',
Greg Crosby 1, and Changyi Sun 1
1 Utah State University, Logan, UT
2 USDA Northwest Watershed Research Center, Boise, ID
3 a
OcU.92
Dec 1.92
Feb1.93
Apr 1,93
Jim 1.93
Aug1,93
Figure 1. Modeled basin average snow water equivalent in Upper Sheep Creek, 1992/93, using a fully distributed model on a 30-meter grid and
lumped model with snow-covered area fraction parameterized using a depletion curve. The good comparison indicates that the depletion
curve approach is promising for scaling up the modeling of snow accumulation and melt, allowing small basin size areas (26 Ha) to be
modeled as a single node in a distributed model with larger elements.
Semi-arid rangeland and forested watersheds
comprise a large portion of the Western United States.
The quality, quantity, and timing of runoff from these
watersheds is crucial for water supply and affects
agriculture, fisheries, recreation, and hydropower. The
purpose of this project is to understand interacting
watershed processes over a range of scales in the
Reynolds Creek Experimental Watershed (RCEW) in
southwest Idaho.
The investigators are developing a spatially
distributed modeling framework that accounts for
spatial variability in topography, vegetation, and soils
to quantify the complete water balance at a range of
spatial scales. This will provide a framework within
which to test hypotheses regarding the hydrology and
water balance at Reynolds Creek. As such, it is a
working model with alternative modules being switched
in and out for comparison against data. The model
development is proceeding in parallel with the acquisi-
tion and processing of the remotely sensed data. This
includes data from six aircraft overflights as well as
Landsat data. Field data being used consist of stream-
flow, meteorological data, soil moisture and
groundwater data, and evapotranspiration flux data.
Parameterization of snow subgrid variability has
focused on causes for this variability and how to model
it. The results to date indicate that the subgrid varia-
bility due to drifting was equally or more important than
subgrid variability in solar radiation for estimating the
quantity and timing of surface water inputs from
snowmelt. A parameterization linking snow covered
area to basin average snow water equivalent shows
promise as a tool in scaling the energy balance up model
to larger model elements. Figure 1 shows a comparison
of the time evolution of basin snow water equivalent
calculated using a 255-point distributed model and the
spatially integrated snowmelt model. The close com-
parison is indicative of the potential of this approach.
Leaf Area Index (LAI) measurements to date
indicate that: (1) there is approximately a twofold range
in LAI within the major plant communities in the
watershed; (2) the maximum LAI occurs early in the
growing season, shortly after snowmelt; and (3) the LAI
of all the plant communities (i.e., at all sites) decreases
steadily during the summer to an apparent minimum in
the fall. Different approaches for using remotely sensed
imagery to map the spatial distribution of the critical
vegetative communities within the watershed are
currently being tested. This will allow the incorporation
of vegetative dynamics into the surface energy flux
modeling.
This project will lead to a better understanding of
the spatial variability and scale dependence of
hydrologic processes in RCEW. Because this work is
aimed at gaining a better understanding of the physical
processes and their interactions, results will be
generalizable to other watersheds in the semiarid mount-
ainous Western United States.
84
-------
Traveling Wave Behavior During Subsurface Transport of
Biologically Reactive Contaminants: Implications for In Situ Bioremediation
Albert J. Valocchi
University of Illinois at Urbana-Champaign, Urbana, IL
Contamination of soil and groundwater by
hazardous chemicals is widely recognized as one of the
major environmental problems faced by modern society.
Although there is ongoing debate regarding the extent
and severity of contamination, there is no disagreement
over the staggering economic costs incurred in
monitoring and rehabilitating polluted groundwater
supplies. In-the-ground (in situ) remediation of con-
tamination costs a fraction of other approaches. The
objectives of this project are to investigate how transport
and mixing processes affect the overall performance of
engineered in situ bioremediation. Although these pro-
cesses play a key role in the ultimate success of actual
remediation projects, their significance cannot be ascer-
tained through typical laboratory-scale studies.
Mathematical models are being analyzed for a
typical bioremediation scenario in which a uniformly
distributed organic contaminant is degraded by in-
digenous soil microbes that are stimulated by an injected
material (e.g., an electron acceptor such as oxygen).
Our analysis starts with simple, one-dimensional homo-
geneous systems and progresses to more realistic
mutlidimensional heterogeneous aquifers.
The following cases have been studied: (1)
one-dimensional uniform flow in a homogeneous
aquifer; (2) ideal radial flow from an injection well in a
homogeneous aquifer; and (3) two-dimensional flow in
a stratified aquifer. For a wide variety of cases, the
system evolves to form traveling waves; that is, the
spatial profiles of the organic pollutant, electron ac-
ceptor, and biomass attain constant shapes that travel in
unison. These traveling waves form because of a
balanced interaction between solute mixing processes
and localized biodegradation processes. This results in
a very localized reaction zone where the pollutant and
electron acceptor mix together, as can be seen in Figure
1, which is for a two-dimensional stratified aquifer.
For conditions when traveling waves exist, simple
formulas have been derived to calculate the long-term
rate of pollutant removal due to biodegradation. The
removal rate expressions are similar for all the different
flow systems that we have examined. The pollutant re-
moval rate depends upon transport properties of the
aquifer, but it does not depend upon the initial biomass
concentration or upon the rate at which the biomass can
degrade the pollutant. Results from the simple formulas
were verified by comparison with numerical simula-
tions.
The traveling wave framework is a useful sim-
plifying tool for approximating the complexity of bio-
remediation modeling. Use of the analytical formulas
reduces significantly the computational burden of es-
timating the efficiency of a remedial design. The results
are significant because they indicate conditions for
which the long-term pollutant removal rate does not
depend upon the laboratory-determined rate of bio-
degradation. This helps explain the common observa-
tion that laboratory-determined degradation rates over-
estimate the degree of biodegradation attained in many
field projects.
In the future, the analysis will be extended to
more realistic patterns of aquifer heterogeneity where
the soil permeability changes randomly in all directions.
Detailed numerical simulations of the cases studied to
date indicate that the traveling wave behavior is not
established until a certain time elapses. The inves-
tigators plan to study this initial time period because the
biodegradation rate, which is often much larger than the
long-term rate, which is attained after the traveling
waves form.
85
-------
1.0-1
N 0.5-
N 0.5
N 0.5-
20.0
20.0
20.0
T
0.0 0.2 0.4- 0.6 0.8 1.0
Normalized concentration scale
Figure 1. Normalized concentration profiles of the organic substrate (pollutant), electron acceptor, and biomass characterizing
enhanced in situ bioremediation in a stratified aquifier. The aquifier consists of two layers, with the more permeable
layer on the bottom. The electron acceptor is injected from the left-hand boundary. The profiles maintain their
shapes as they migrate downstream (to the right) as traveling waves.
86
-------
Carbon Exchange Dynamics in a Temperate Forested Watershed
(Northern Michigan): A Laboratory and Field Multidisciplinary Study
Lynn M. Walter ', L.M. Abriola 2, J.M. Budai \ G.W. Kling 3, P.A. Meyers ', J.A. Teeri 3, and D.R. Zak 4
'Department of Geological Sciences;2 Department of Chemistry and Environmental Engineering;3 Department
of Biology; •* School of Natural Resources; The University of Michigan, Ann Arbor, MI
Beginning in the mid-1800's, humans started an
uncontrolled experiment with carbon on earth. It is now
apparent that the effects of increasing levels of atmos-
pheric CO2 resulting from this experiment will alter
carbon dynamics and the functioning of terrestrial and
aquatic ecosystems in ways that are not entirely under-
stood. The goal of this project is to determine the fate
of organic carbon produced in temperate forests. Such
forests constitute major potential sinks for anthropogenic
CO2. Importantly, studies of carbon allocation in
forests under enhanced and ambient CO2 growth
conditions have shown that above and below ground
carbon storage as well as root and microbial respiration
processes all increase with elevated CO2. The question
driving this research effort is, "How does the increased
rate of carbon fixation and processing in forest stands
and soils affect the overall carbon budget at the
watershed scale?" Is the additional organic carbon
merely recycled via respiration and returned to the
atmosphere, or is it transformed and transported from
the soil zone to the regional groundwater system for
longer term storage?
The project is taking place in a hydrologically and
physiographically constrained portion of the Cheboygan
watershed. Within the watershed confines are two well-
studied forest stands and an established elevated CO2
experiment in which aspens and sugar maples are grown
in mesocosms. Soil compositions and water chemistry
are characterized in each experimental mesocosm and in
instrumented natural forest stands (aspens vs. sugar
maples) to determine how root respiration and microbial
processing of organic matter may be linked to carbon
transport out of the rooted zone. As a direct measure of
mineral weathering effects on dissolved inorganic
carbon transport, prepared experimental arrays of
feldspar and carbonate minerals have been implanted in
natural forests and in experimental mesocosms.
The first year of the project was devoted to
hydrologic characterization of the watershed system,
establishment of soil water sampling arrays, and
general chemical characterization of soil waters,
surface waters, and groundwaters in the study area.
Soil waters exhibit large vertical chemical variations,
generally grading from dilute, dissolved organic
carbon-rich solutions in the upper 20 cm into
mineralized solutions chemically similar to regional
groundwaters by 4 meters in depth. Mass balance
among dissolved carbon species suggests that dissolved
organic carbon (DOC) originating from reactions in
the upper rooted zone is transformed to dissolved
inorganic carbon (DIG) via respiration and coupled
mineral solubilization reactions. Dissolved silica and
aluminum increase rapidly in soil waters suggesting
that aluminosilicate minerals, as well as carbonate
minerals, are dissolving as DOC (especially organic
acid anions) is transformed to DIG (see Figures la and
Ib). The significant solute acquisition and carbon
transformation evident in soil water profiles suggest
that organic processes active in the upper soil horizons
are closely linked to mineral dissolution mechanisms
and the overall rate of solute transport out of the soil
zone.
The next 2 years of the project will involve
generating an overall carbon budget for the watershed
and modeling carbon exchange rates among the main
carbon reservoirs. The experimental mesocosms will
yield more information on links among growth
conditions, soil water chemistry, and DOC/DIG
fluxes as the second summer of tree growth begins in
1998. Regional groundwater flow modeling together
with more detailed groundwater sampling transects
will better define the horizontal transport effects
along flow paths relative to the vertical effects in soil
water chemical profiles.
87
-------
Figure la.
Figure Ib.
Figures la and Ib.
As shown in scanning electron phomicrographs, "Minerals present in glacial drift soil horizons, such as
carbonates (a) or aluminosilicates (b), undergo dissolution and enhanced solubilization in organic acid and
CO2-rich soil waters."
-------
A Comparison of Agricultural vs. Forested Basins: Carbon
and_Nutrient Cycling Within the Hyporheic Ecotone of Streams
David S. White, Susan P. Hendricks, Timothy C. Johnston, George Kipphut, and William E. Spencer
Center for Reservoir Research and Hancock Biological Station, Murray State University, Murray, KY
Figure 1. Installing gravel bar wells at the Ledbetter Creek, Kentucky, hyporheic study site.
Processes and patterns within the hyporheic zone,
the interface between groundwater and surface water
beneath streams, may help explain differences in stream
functions between forested and agricultural basins.
Land use effects on subsurface processes mediated by
increased nutrient, carbon, and sediment loads are
largely unknown. This project's goal is to examine
differences in the function of hyporheic ecotones of
third order streams in pristine and agricultural water-
sheds. The hyporheic ecotone functions as a biogeo-
chemically active interface where biogeochemical
storage and processing are directly linked to the
longterm health and productivity of surface waters.
Anthropogenic alterations within watersheds (such as
increased sedimentation or runoff) potentially alter
hyporheic function and therefore the biotic integrity of
stream ecosystems. Ledbetter Creek (LC) is an
agricultural watershed located in Galloway County,
Kentucky; Panther Creek (PC) is a forested basin
located in Stewart County, Tennessee. The two study
streams both empty into Kentucky Lake. Hyporheic
zones within specific third order sites on both streams
have been mapped and monitored with wells, mini-
piezometers and seepage meters (see Figure 1).
Monitoring provides baseline data for specific investiga-
tions and manipulations under base flow and high
discharge conditions.
At base flow, LC surface discharge contains more
nitrogen but less phosphorus than PC. Also, LC carries
twice the suspended solids and four times the suspended
organic matter load. Fine sediment deposition on the
bed surface in LC (at approximately 10 times the rate
for PC) appears to decrease surface flow infiltration into
the bed and through gravel bars. Decreased surface-
subsurface interaction is demonstrated by several
measurements. Vertical hydraulic gradients (VHG) are
minimal or slightly positive throughout the LC site.
Conversely, VHG at PC are more distinct, indicating
extensive upwelling and downwelling areas through bed
sediments. Conservative tracer studies using bromide
and a one-dimensional transport model indicate a much
higher degree of surface-subsurface connectivity at PC.
Several gravel bar wells at PC have shown rapid
hydrologic exchange with surface water (100 percent),
while similar wells at LC have shown little hydrologic
exchange within a 9-hour period. There is less oxygen
in the LC hyporheic ecotone, commensurate with higher
concentrations of phosphate, ammonium, iron, and
manganese as well as methane, particularly within the
gravel bars. In general at base flow, the hyporheic
ecotone at LC is much more isolated from surface water
than at PC.
A 1-hour, half-inch rainfall in the LC basin
produces a steep hydrograph (0.03 to 2.0 m3 sec"1) within
2 hours but little measurable change in PC discharge
because of basin retention differences. With increasing
discharge, LC surface water contains exponentially more
suspended solids, nitrogen, and phosphorus, presumably
derived from runoff but potentially from hyporheic storage
as well. At PC, increased discharge causes a slight increase
in suspended solids, but surface water nitrogen and phos-
phorus concentrations decrease, demonstrating surface
water dilution. Storm flows in LC tend to remove settled
fine sediments, increase subsurface flow, and decrease
hyporheic concentrations of ammonium and phosphate.
Whether the changes in subsurface chemistry represent flux
into surface water or redox changes in the subsurface
environment is now being investigated. Concurrent studies
in PC and LC hyporheic ecotones are now being conducted
on microbial diversity and production, benthic algal
production, and sediment fauna (hyporheos) secondary
production. Initial results indicate that PC is more
biologically diverse with generally higher rates of second-
ary production.
-------
In Situ Assessment of the Transport and
Microbial Consumption of Oxygen in Groundwater
Tadashi Yoshinari
Wadsworth Center, New York State Department of Health and SUNY, Albany, NY
R.L. Smith
U.S. Geological Survey, Boulder, CO
J.K. Bohlke and K. Rtvesz
U. S. Geological Survey, Reston, VA
50
= 40
8.
o
oo
30
20
\AfelllD
473; E = 5.4
26-11; E = 10.0
343 top; E = 5.9
343bot;E = 1.r
13-16; E = 5.2
62-12; E = 3.6
593; E = 2.8
air
saturated
water (10 oq
-10123
In O2 concentration in mg/L
Levels of dissolved oxygen in groundwater
influence the rates and extents of biodegradation of
organic pollutants within an aquifer. Because microbial
respiration is the most important mechanism of oxygen
consumption in this environment and thus is a key
determinant of aquifer geochemistry, the investigators
have been conducting a study to quantify aerobic
respiration within a plume of dilute treated sewage in a
sand and gravel aquifer in Cape Cod, Massachusetts.
The approaches being used to assess aerobic
respiration include: (1) quantifying rates of oxygen
consumption using in situ tracer tests and laboratory
incubations with sediment core and groundwater
samples and (2) analyzing the concentrations and
isotopic composition of O2 in various geochemical zones
within the aquifer and determining the relative
importance of dispersion and respiration in controlling
oxygen levels within these zones.
Rates of oxygen consumption were determined by
three methods: (1) A natural-gradient tracer test was
conducted to directly measure in situ rate of oxygen
demand. (2) Oxygen consumption was measured in
laboratory incubations at in situ temperatures and
oxygen concentrations with sediment core slurries using
a gas chromatographic (GC) technique. (3) An assay of
bacterial electron transport system (ETS) activity was
conducted in parallel with the GC analysis by amending
Figure 1. Variations in the concentration and isotopic composition of
dissolved oxygen near the upper boundary of a contaminated
groundwater plume at Cape Cod. Epsilon (e), the apparent isotope
fractionation factor, is defined by a version of the Rayleigh distillation
equation; l,000*ln(R/Ro) = -e*ln(C/C0), where R = 18O/16O, C =
concentration, and RQ and C0 denote the initial condition (before
reaction). The 618O value on the y-axis, a measure of the relative
enrichment of 18O/'6O in a sample, is defined by: 618O = 1,000*
(R/RSTANDARD -1) ^d is approximately equal to l,000*ln(R/R°). Thus,
the e value is a measure of the rate of change of 618O during decline
in the dissolved O2 concentration. Here it is assumed that groundwater
had initial values around O2 = 340 ^M and 618O = +24.3 per mil (in
equilibrium with air at 10 • 12°C). The concentrations of O2
decreased while values of 6I8O-O2 increased from +24 to at least
+45 per mil, indicating that the O2 consumption by microbial
processes is taking place in the plume water to varying degrees.
sediment core slurries with the tetrazolium salt INT,
which acts as an artificial electron acceptor.
Rates of oxygen consumption in the sediment
core incubations, as measured by GC, were dramatically
higher than rates calculated from the in situ tracer test.
Conversely, rates of oxygen consumption from the INT
assay were significantly lower than the in situ rates.
The data suggest that incubation of aquifer samples
within bottles in the laboratory appreciably stimulates
microbial respiration, even when in situ temperature and
oxygen levels are maintained and that INT is toxic.
The main factors that determine the relative
enrichment of 18O/16O in O2 in an aquifer are: (1) microbial
reduction of O2 that causes an increase in the ratio of
18Q/16Q ^ ^ remaining fraction of O2 due to kinetic
isotope fractionation and (2) dispersion of O2, which could
result in concentration gradients without major isotopic
variation. Figure 1 shows an increase of 18O (see the
definition in the figure caption) with decreasing O2
concentration through a vertical O2 gradient at seven
locations in the aquifer, clearly indicating that O2
consumption has occurred at these sites. The low apparent
e values suggest that the vertical O2 concentration gradient
within the aquifer is caused by a combination of O2
consumption and dispersion in varying degree. Numerical
models are currently being applied to better define the
relationship between O2 consumption and dispersion.
90
-------
Index of Authors
Abriola, L.M., 87
Allen, D., 51
Anderson, S., 31
Andrews, R.N., 32
Arnosti, C., 45
Ashton, M., 3
Bain, M.B., 14
Ballard, S., 11
Barten, P., 3
Beck, M.B., 2
Benfield, E.F., 6
Bennett, L., 3
Benoit, G., 3
Bernert, J., 36
Beschta, R.L., 64
Blough, N.V.,45
Blum, M.D., 82
Bohlke, J.K., 90
Booth, D., 19
Booth, D.B., 28
Bosch, D.J., 6
Bostrom, G., 16
Botsford, L., 35
Bott, T., 38
Boumans, R., 49
Braden, J.B., 25
Brezonik, P.L., 26
Brooks, R.P., 12,47
Brown, W., 14
Budai, J.M., 87
Surges, S.J., 28
Bushek, D., 51
Caneday, L., 17
Caraco, N., 83
Carpenter, S.R., 75
Carper, W.R., 55
Clark, D., 19
Cleveland, C.J., 10
Colbert, D.L., 16
Cole, J.J., 75, 83
Cole, C.A.,47
Cooley, K.R., 84
Corbett, C., 51
Cosby, J., 11
Costanza, R., 49
Couch, C.A., 66
Cox, W.E., 6
Crosby, G., 84
Cruse, R., 36
Dame, R.F., 51
Danielson, B., 36
Deegan, L.A., 4, 53
Demissie, M., 25
DeVuyst, E., 25
Dinar, A., 79
Diplas, P., 6
Dixon, A., 55
Driscoll, C.T., 61
Easter, K.W., 26
Edwards, D., 51
Eilers, J., 36
Estes, M.K., 60
Evans, B.M., 12
Fernandez, I., 11
Ferreri, C., 38
Findlay, S., 83
Finley, J., 38
Flerchinger, G.N., 84
Focht, W., 17
Ford, M.S.J., 16
Fox, J., 29
Freemark, K., 36
Freyberg, D.L., 56
Galatowitsch, S., 36
Giambelluca, T., 29
Glotfelty, C., 38
Gordon, S.I., 57
Govindaraju, R.S., 59
Gowda, P.H., 8
Grant, S.B., 60
Greene, R.G., 6
Gregory, L., 51
Griffin, R., 19
Hairston, Jr., N.G., 61
Hamlett, J.M., 12
Hanson, C.L., 84
Haroa, R.J., 8
Hendricks, S.P., 89
Herman, J.S., 62
Herricks, E.E., 33
Hershey, R., 31
Hession, W.C., 38
Hoffman, E., 79
Holomuzki, J., 81
Hopkinson, C., 10
Hornberger, G.M., 62
Horwitz, R., 38
Howitt, R., 79
Hristov, T.N., 12
Jackson, J., 38
Johnson, J., 38
Johnson, M., 35
Johnston, T.C., 89
Kahl, J., 11
Kaplan, L., 38
Karr, J.R.,28
Kauffman, J.B., 64
Kearns, C.M., 61
Kellert, S., 3
Kibler, D.F., 6
Kipphut, G., 89
Kitchell, J.F., 75
Kjerfve, B., 51
Kling, G.W., 87
Knight, C.G., 12
Koepfler, E., 51
Komar, P., 16
Kramer, E.A., 66
Kremer, J., 4
Lammers, R., 72
Larive, C.K., 55
Lewis, V.P., 65
Lewitus, A., 51
Li, H.W., 64
Li, J.L.,64
Lincoln, D.E., 65
Lohani, V.K., 6
Loucks, D.P., 14
Lovell, C.R., 65
Luce, C., 84
Ludwig, P., 11
Lund, J., 35
Lynch, R., 17
Lynn, W.R., 14
Mageean, D., 11
91
-------
Index of Authors (continued)
Matzke, G., 36
Maxwell, T., 49
McDonnell, A.J., 12
McDowell, P., 64
Mclntosh, B.A., 64
McManus, J., 16
Meo, M., 17
Meyer, J.L., 66
Meyers, P.A., 87
Miller, W W., 31
Miller, G.C., 31
Mitra, P., 25
Morel, F.M.M., 68
Mostaghimi, S., 6
Moyle, P., 35
Mulla, D., 26
Myers, W.L., 70
Nagarkatti, P.S., 6
Napier, T.L., 8
Nassauer, J., 36
Neale, C.M.U., 84
Newbold, J.D., 38
Norton, B.C., 2
Norton, S., 11
Novotny, V., 19
Olson, T.M., 60
Orlob, G., 35
Orth, D.J., 6
Pace, M.L., 75, 83
Papellis, L., 31
Parker, G., 69
Patil, G.P., 70
Patten, B.C., 2
Paul, M.J., 66
Perry, J.A., 26
Petersen, G.W., 12
Peterson, B., 72
Pickett, S.T.A., 20
Polasky, S., 36
Porter, K.G., 2
Prasad, R., 84
Prins, T., 51
Puente, C.E., 74
Quinn, J., 35
Rambo, A.T., 29
Rasmussen, T.C., 2
Rassenti, S., 79
Rastetter, E., 10
Reice, S.R., 32
Revesz, K., 90
Rhoads, B.L., 33
Richerson, P., 35
Richey, I.E., 34
Roberts, P.V., 56
Rustad, L., 11
Sabatier, P., 35
Saiers, J.E., 62
Sankowski, E.T., 17
Santelmann, M., 36
Schauman, S., 28
Schindler, D.E., 75
Schlager, E., 76
Schmandt, J., 77
Schneider, D., 25
Schubauer-Berigan, J., 51
Seyfried, M.S., 84
Shabman, L.A., 6
Shaw, D., 31
Shepherd, A., 2
Sipes, J., 17
Skelly, D., 3
Slaughter, C.W., 84
Smith, C., 16
Smith, V., 79
Smith, R.L., 90
Sobsey, M.D., 80
Sparks, R.E., 25
Spencer, W.E., 89
Standley, L., 38
Steenhuis, T.S., 14
Stephenson, K., 6
Stevenson, R.J., 81
Straffin, E.G., 82
Strayer, D.L., 83
Suchanek, T., 35
Sun, C., 84
Sweeney, B., 38
Tarboton, D.G., 84
Taulbee, W.K., 66
Teeri, J.A., 87
Turco, R.P., 40
Tyler, S., 31
Vallino, J., 10
Valocchi, A.J., 85
Vieux, B., 17
Villa, F.,49
Voinov, H., 49
Voinov, A., 49
Vorosmarty, C., 72
Wainger, L., 49
Walter, L.M., 87
Ward, A.D., 8, 57
Wardrop, D.H., 47
Webler, T., 4
Weinberg, M., 35
White, D., 36
White, D.S., 89
White, D.C., 25
White, D., 57
Wiley, M., 81
Willett, K., 17
Wilson, D., 33
Woodin, S.A., 65
Xia, R.,25
Yoshinari, T., 90
Zak, D.R., 87
92
-------
The Foundation provides awards for research and education in the sciences and engineering. The
awardee is wholly responsible for the conduct of such research and preparation of the results for
publication. The Foundation, therefore, does not assume responsibility for the research findings
or their interpretation.
The Foundation welcomes proposals from all qualified scientists and engineers and strongly
encourages women, minorities, and persons with disabilities to compete fully in any of the
research and education related programs described here. In accordance with federal statutes,
regulations, and NSF policies, no person on grounds of race, color, age, sex, national origin, or
disability shall be excluded from participation in, be denied the benefits of, or be subject to
discrimination under any program or activity receiving financial assistance from the National
Science Foundation.
Facilitation Awards for Scientists and Engineers with Disabilities (FASED) provide funding for
special assistance or equipment to enable persons with disabilities (investigators and other staff,
including student research assistants) to work on NSF projects. See the program announcement or
contact the program coordinator at (703) 306-1636.
The National Science Foundation has TDD (Telephonic Device for the Deaf) capability, which
enables individuals with hearing impairment to communicate with the Foundation about NSF
programs, employment, or general information. To access NSF TDD dial (703) 306-0090; for
FIRS, 1-800-877-8339.
------- |