The Effects of Habitat Alteration by Estuarine Stressors on Ecological Resources of
Pacific Northwest Estuaries
Research Plan
Coastal Ecology Branch
Western Ecology Division
National Health and Environmental Effects Research Laboratory
United States Environmental Protection Agency
February 1999
Investigators
B. Boese, F. Cole, T. DeWitt, S. Ferraro, J. Lamberson,
H. Lee, W. Nelson, B. Ozretich, J. Power, B. Robbins,
A. Sigleo, D. Specht, D. Young
-------
The Effects of Habitat Alteration by Estuarine Stressors on Ecological Resources of
Pacific Northwest Estuaries
Research Plan
Coastal Ecology Branch
Western Ecology Division
National Health and Environmental Effects Research Laboratory
United States Environmental Protection Agency
February 1999
Investigators
B. Boese, F. Cole, T. DeWitt, S. Ferraro, J. Lamberson,
H. Lee, W. Nelson, B. Ozretich, J. Power, B. Robbins,
A. Sigleo, D. Specht, D. Young
The information in this document has been funded wholly by the U.S. Environmental
Protection Agency. It has been subjected to the Agency's peer and administrative
review, and it has been approved for publication as an EPA document. Mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
Signature Page
ii
The Effects of Habitat Alteration by Estuarine Stressors on Ecological Resources of
Pacific Northwest Estuaries
Research Plan
Coastal Ecology Branch
Western Ecology Division
National Health and Environmental Effects Research Laboratory
United States Environmental Protection Agency
Approvals:
Walter G. Nelson, Branch Chief 'Dat
Coastal Ecology Branch
h/o fr/9
Thomas A. Murphy, Director Date
Western Ecology Division
-------
iii
Table of Contents
Title page
Signature page ii
Table of Contents iii
Acronyms v
Abstract ... . vi
1.0 Introduction 1
1.1 Context of the Research Program 1
1.2 Rationale for the Research Program 2
1.2. 1 Drivers and Stressors Affecting Pacific Northwest Estuaries 5
1.2. 2 Estuarine Scaling issues 8
1.3 Scientific Questions 11
2.0 Regional Characteristics and Study Area 13
2.1 Pacific Northwest Estuaries 13
2.2 Study Area ; 15
3.0 Research Approach 19
3.1 Research Theme Organization 19
3.2 Key Infrastructure Support Elements 22
3.3 Research Integration, Interactions and Continuous Improvement 23
4.0 Detailed Research Project Descriptions 28
4.1 Research Theme A. Indicators of Ecological Condition for PNW Estuaries ... 28
Project A1 - Estuary Biota-habitat Relationships 28
Project A2 - Dynamics of an Estuarine Landscape: Spatial and 41
Temporal Patterns with Regard to Coastal Shoreline Development
Project A3 - Changes in The Abundance And Distribution of Estuarine 49
Keystone Species in Response to Multiple Abiotic Stressors
Project A4 - The Impact of Disturbances on an Eelgrass Habitat 68
Project A5 - Comparison of Factors Affecting the Distribution of the Non- .... 80
Indigenous Seagrass Zostera japonica with Those Controlling the Native
Zostera marina in Yaquina Bay, Oregon
Project A6 - Spatial Variation of Growth and Condition in Juvenile English ... 90
Sole Relative to Substrate Characteristics
4.2 Research Theme B - Stressor-Response Modeling 103
Project B1 - An Evaluation of the Geometry of Stress Using Spatially 103
Explicit Population Models
4.3 Research Theme C - Estuarine Physical-Chemical Stressors 121
Project C1 - Relationships Between Suspended and Bottom Sediment .... 121
-------
iv
Conditions and Macrophyte Distributions in Yaquina Bay Estuary
Project C2 - Evaluation of the Susceptibility of Eelgrass Beds in Oregon ... 135
Estuaries to Changes in Watershed Uses
Project C3 - Nutrient Processes: Watershed versus Oceanic Inputs to 154
PNW Estuaries
5.0 Research Linkages, Users and Participants 164
5.1 Research Group Linkages 164
5.2 Graduate and Postdoctoral Research Collaborators 165
5.3 Users and Participants 165
6.0 Quality Assurance 168
7.0 Literature Cited 169
8.0 Appendices 210
8.1 Appendix A Technical Details of Experiments for Project A.4 210
8.2 Appendix B Technical Details of Model Development for Project B.1 231
-------
V
ACROMYMS
AVS - Acid Volatile Sulfides
CEB - Coastal Ecology Branch
CIR - Color Infrared
CTDS - Conductivity-temperature-depth Sensors
CWAP - Clean Water Action Plan
DEM - Digital Elevation Model
DGPS - Differential-corrected Geographical Positioning System
DIN - Dissolved Inorganic Nitrogen
DIP - Dissolved Inorganic Phosphorous
EMAP - Environmental Monitoring and Assessment Program
FC - Full Color
GIS - Geographic Information System
GPS - Geographic Positioning System
LMER - Land Margin Ecosystem Research
MLLW - Mean Lower Low Water
NCR - National Research Council
NHEERL - National Health and Environmental Effects Research Laboratory
NOAA - National Oceanic and Atmospheric Administration
ORD - Office of Research and Development
ODFW - Oregon Department of Fisheries and Wildlife
OSU - Oregon State University
PNCERS - Pacific Northwest Coastal Ecosystem Study
PNW - Pacific Northwest
PAR - Photosynthetically Active Radiation
QA/QC - Quality Assurance/Quality Control
REB - Regional Ecology Branch
RPD - Redox Potential Discontinuity
SAV - Submerged Aquatic Vegetation
SAVEWS - Submerged Aquatic Vegetation Early Warning System
SEPM - Spatially Explicit Population Models
SOP - Standard Operating Procedures
STP - Sewage Treatment Plant
TOC - Total Organic Carbon
TSS - Total Suspended Sediment
TSS - Total Suspended Solids
US EPA - United States Environmental Protection Agency
USFWS - United States Fish and Wildlife Service
USGS - United States Geological Survey
WQV - Proposed Water Quality Values
WED - Western Ecology Division
YBE - Yaquina Bay Estuary
-------
vi
Abstract
The Coastal Ecology Branch (CEB), Western Ecology Division of the US Environmental
Protection Agency will initiate a research program to evaluate the effects of alterations of
estuarine habitats resulting from multiple stressor sources. The research will concentrate on
stressor effects on the ecological resources of estuaries of the Pacific Northwest (PNW). The
research program is designed to support the general mission of the US EPA which includes
the safeguarding of the natural environment upon which the health and well being of the
nation's population ultimately depends.
The goal of CEB research is to improve the ability to make key policy decisions on coastal
environmental issues by defining key ecological processes and by developing models to
predict stress-response relationships for ecological resources within Pacific Northwest
estuaries at a range of spatial and temporal scales. CEB research objectives are to 1)
evaluate how specific estuarine habitats respond to a range of potential stressors which may
lead to habitat alteration, 2) understand the influences of these stress factors at spatial scales
from local to regional, and 3) develop indicators of ecological condition which may be used to
evaluate estuarine status across multiple spatial scales.
The research effort will concentrate on two habitats 1) submerged aquatic vegetation (SAV),
and 2) burrowing shrimp, with lesser effort on other types of estuarine habitats. SAV and
shrimp are selected as focal research habitats and important assessment endpoints because
in each case the characteristic species which define the habitat do so because of their
"physical ecosystem engineering" activities.
To accomplish the objectives of the CEB research plan, research will be organized in three
thematic elements: A. Indicators of Ecological Condition for PNW Estuaries; B. Stressor-
Response Modeling; C. Estuarine Physical-chemical Stressors.
The research projects under Research Theme A address the questions: 1) what are the biotic
constituents of major estuarine benthic habitats of PNW estuaries, 2) what effects do various
abiotic and biotic stressors have on the biotic composition of principal habitat types, 3) what
role do biotic and abiotic stressors have in controlling the spatial extent and distribution
patterns of major estuarine habitat types, and 4) what are appropriate indicators of ecological
condition at the population, species, community and landscape levels for PNW estuarine
systems. Stressors that will be examined include anthropogenic physical disturbances such
as clam and burrowing shrimp harvesting, and sedimentation, salinity, and water column light
field alterations potentially generated by elevated runoff resulting from changes in landscape-
use patterns. Biotic stressors that will be examined include disturbances such as the
smothering of seagrass habitat by mat-forming algae potentially promoted by elevated
nutrients, and biotic stress induced by competition between native and exotic seagrass
species and between burrowing shrimp and seagrasses.
-------
vii
Projects under Research Theme B will work at the population and community levels to
develop modeling techniques to integrate the detailed studies of biological effects of estuarine
stressors of Theme A with the spatial-temporal stressor distribution studies of Theme C. The
principal current project is the development of spatially explicit modeling tools for estuarine
benthic populations to allow predictions of population responses to the imposition of multiple
stressors.
The research projects under Research Theme C will address the questions 1) what are the
spatial and temporal distribution patterns of the primary physical and chemical factors
determining estuarine habitat composition, 2) how are spatial variations in the physical-
chemical stressors associated with variations in target estuarine habitats, 3) how do
anthropogenic alterations of watershed characteristics influence the transport of dissolved and
particulate materials into the estuary, and 4) what are the relative roles of oceanic versus
riverine inputs of dissolved and particulate materials on water column physical-chemical
processes.
-------
1
1.0 Introduction
1.1 Context of the Research Program
The Coastal Ecology Branch (CEB), Western Ecology Division of the US Environmental
Protection Agency proposes to initiate a research program to evaluate the effects of
alterations of estuarine habitats resulting from multiple stressor sources. The research will
concentrate on stressor effects on the ecological resources of estuaries of the Pacific
Northwest. The research program is designed to support the general mission of the US EPA
which includes the safeguarding of the natural environment upon which the health and well
being of the nation's population ultimately depends.
CEB is a part of the National Health and Environmental Effects Laboratory (NHEERL), Office
of Research and Development (ORD) of EPA. The CEB research program is a response to
one of the six, high priority research areas identified in ORD's recent strategic evaluation of
research needs (US EPA, 1997). The CEB research program will focus on the priority need
for research to improve ecological risk assessment. Ecological risk assessment is defined as
a determination of "the nature and likelihood of effects of our actions on animals, plants, and
the environment" (SETAC, 1997). An increased focus on ecological risk assessment has
emerged from the reorganization of the environmental research laboratories of ORD within the
paradigm of risk assessment and risk management. The risk assessment approach provides
a means to focus agency research to address increasingly complex environmental research
issues.
The CEB research effort is also a response to the ORD strategic need to conduct research
that allows EPA to improve its ability to identify and respond to emerging environmental
issues. Agency research supporting development of environmental regulations has had
considerable success in reducing the most extreme forms of environmental damage from toxic
discharges into the nation's air, soil, fresh and marine waters, particularly from point source
discharges. The critical need has now become to "better understand the vulnerability and
sustainability of our ecological resources within the context of multiple stresses affecting
multiple endpoints at multiple scales" (US EPA, 1997).
At the same time, the recent National Research Council review of EPA research (NRC, 1997)
has pointed out that there is a need for an increased fundamental understanding of
ecosystems and has recommended that EPA establish a core research agenda to advance
such knowledge. To achieve this goal, the NRC review committee advocated development of
more effective environmental research tools, including new instruments and measurement
platforms, and development of more sophisticated environmental models (NRC, 1997).
The CEB research program will help support key action items of the Clean Water Action Plan
which forms the core of the Presidential Clean Water Initiative (see section 3.0 below).
Branch research will particularly help support actions for projects to Restore and Protect
-------
2
America's Wetlands, Protect Coastal Waters, and Reduce Nutrient Over-enrichment.
Thus CEB research will operate within the broad strategic context of improved ecological risk
assessment and in the support of key action requirements of the Clean Water Action Plan.
The goal of CEB research is to improve the ability to make key policy decisions on coastal
environmental issues by defining key ecological processes and by developing models to
predict stress-response relationships for ecological resources within Pacific Northwest
estuaries at a range of spatial and temporal scales. CEB research objectives are to 1)
evaluate how specific estuarine habitats respond to a range of potential stressors which may
lead to habitat alteration, 2) understand the influences of these stress factors at spatial scales
from local to regional, and 3) develop indicators of ecological condition which may be used to
evaluate estuarine status across multiple spatial scales.
1.2 Rationale for the Research Program
Nationwide, growth of the human population is disproportionally concentrated in the coastal
zone (Culliton et al., 1990). Human population growth is a principal driver for many ecological
stressors such as habitat loss, pollution, and nutrient enhancement which alter coastal
ecosystems and affect the sustainability of coastal ecological resources. Increased
globalization of the economy is a major driver influencing the introduction of exotic species
into port and harbors. Major environmental policy decisions at local, state and federal levels
related to land use planning, growth management, habitat restoration and resource utilization
will determine the future trajectory for estuarine conditions of the western U.S. The results of
policy decisions will have direct economic impacts on jobs, income, and population through
effects on fish and shellfish harvest, property values, shipping and transportation, and quality
of the coastal recreation experience. To optimize policy decisions, research is needed on the
complex nature of the interactions among multiple stressors in estuarine systems to allow a
greater ability to predict the outcome of policy choices.
Within the Pacific Northwest region, greatest population expansion to date has been in the
major urban areas of Seattle and Portland. Although both cities have inland locations, both
also border the largest coastal water bodies in their regions, Puget Sound and the Columbia
River, respectively. While development around the outer coastal estuaries has been less
intense, the coastal counties of Oregon have been growing from 1-2% per year over the last
decade (R. L. Johnson, OSU, unpublished data), which if maintained, translates into nearly a
40% increase over the next 20 years. The coastal Oregon growth rate demonstrates the same
inexorable development pressure found associated with estuarine systems in other areas of
the country. With increased development comes associated environmental problems such as
the loss of estuarine habitat, increased coastal resource utilization of all types, and increases
in both point source and non-point source contamination (Copping and Bryant, 1993).
Because of the current range of PNW estuaries in terms of state of development, CEB is
ideally situated to conduct research on the impacts of development on estuarine systems of
\
-------
3
the region.
Within estuaries, benthic environments are areas where stressor impacts will tend to
accumulate. Deposition of toxic materials, accumulation of sediment organics, and oxygen
deficiency of bottom waters typically have a greater impact on benthic organisms than on
planktonic and nectonic organisms because of their more sedentary nature. Long-term
studies of the macrobenthos (Reish, 1986, Holland and Shaughnessey, 1986) demonstrate
that macrobenthos is a sensitive indicator of pollutant effects. Benthic assemblages are also
closely linked to both lower and higher trophic levels, as well as to processes influencing
water and sediment quality, and therefore appear to integrate responses of the entire
estuarine system ( Leppakoski, 1979; Holland and Shaughnessey, 1986). In the Pacific
Northwest, the large tidal amplitude of the region means that a large proportion of total
estuarine area is intertidal, and thus benthic resources are of primary importance. For
example, up to 60% of the bottom of Yaquina Bay, OR and 90% of the bottom of Netarts Bay,
OR may be exposed at low tide (Shirzad et al 1988). Benthic resources of PNW estuaries
also provide important nursery habitat for juveniles of some salmonid species, provide direct
economic and recreational benefits through oyster culture and clam harvest, and provide
critical food resources for commercially and recreationally important fishes, waterfowl and
shorebirds.
A secondary rationale for selecting estuarine benthic systems for study is the professional
expertise of the research scientists of the Coastal Ecology Branch. The majority of branch
scientists have spent much of their professional careers studying environmental toxicology
related to contaminated marine sediments. The current redirection of research at CEB to
focus on broad scale ecological issues means that the most efficient cross over for these
individuals is to a focus on estuarine benthic resources, which is a study area where a
considerable body of past professional knowledge of sediment properties and processes may
be drawn upon. Recent hires have been specifically targeted to specialized expertise in areas
such as fish population dynamics, marine benthic ecology, landscape ecology, estuarine
ecological modeling, in order to provide a greater breadth and depth to the CEB research
team.
While benthic habitats within Pacific Northwest estuaries may be classified in a variety of
ways, general categories in the lower intertidal and subtidal regions include unvegetated
intertidal sand and mud, two habitats dominated by species of burrowing shrimps (Neotrypaea
californiensis and Upogebia pugettensis), intertidal and subtidal submerged aquatic vegetation
(SAV; including the seagrasses Zostera marina, Zostera japonica, and the widgeon grass
Ruppia maritima), and subtidal sand and mud. Additionally, there are at least eight categories
of tidal marsh found from lower high water upward to the terrestrial fringe (Akins and
Jefferson, 1973). In some estuaries, additional habitats include oyster beds and the
introduced, emergent, smooth cordgrass, Spartina alterniflora. In order to achieve a goal of
improving estuarine ecological risk assessment, it is necessary to first achieve some degree
of mechanistic understanding of the processes which influence the development and
-------
4
maintenance of these estuarine ecological assemblages. Resource limitations preclude a
simultaneous focus on all habitat types, and research expertise is greatest for lower intertidal
systems. Therefore the research effort will concentrate on two habitats 1) submerged aquatic
vegetation, and 2) burrowing shrimp, with lesser effort on other types of estuarine habitats.
SAV and shrimp are selected as focal research habitats and important assessment endpoints
because in each case the characteristic species which define the habitat do so because of
their "physical ecosystem engineering" activities (Jones et al., 1997). These activities alter the
habitat in ways that have strong cascading ecological effects, either positive or negative, on
other species in the habitat. In the broadest sense of the term, both SAV and burrowing
shrimp can be considered keystone species (Menge et al., 1994) in that their removal from the
habitat will result in a replacement by an alternate ecological system (Thayer et al., 1975).
However, the multiple uses of the term "keystone species" has been criticized (Mills et al.,
1993), and it is more useful to view the focal research organisms as ecosytem engineers.
Both taxa exert important controls on estuarine biodiversity, with eelgrass and other SAV
species having an enhancement effect and burrowing shrimp a negative effect. SAV is well
known to affect nutrient cycling, sediment stability, and water turbidity (Dennison et al., 1993),
and the extensive bioturbation activities of burrowing shrimp may also have strong effects on
ecosystem properties (DeWitt et al., 1997). Mud shrimp create deep burrows and process
massive amounts of water and sediment. As the major bioturbators in PNW estuaries, mud
shrimp can affect nutrient regeneration, sediment deposition and transport, and water quality.
By reducing sediment stability, they degrade habitat quality for oyster production and change
the benthic community. In an attempt to control mud shrimp for the enhancement of the oyster
industry, hundreds of acres in Willapa Bay are sprayed with carbaryl annually, which may
have additional ramifications for estuarine communities.
The value of SAV habitats as nursery areas and sources of food to economically and
recreationally important estuarine species is well established in many areas of the US
(Dennison et al., 1993), and the more limited data from the Pacific Northwest also supports a
similar importance in this region (Thorn, 1987). Data from carbon isotope analysis has
demonstrated that seagrass beds and their associated algae and animals are the food source
for outmigrating juvenile chum salmon (Oncorhynchus keta) in an arm of Puget Sound
(Simenstad and Wissmar, 1985, cited in Thorn, 1987). Juvenile Dungeness crabs appear to
utilize eelgrass meadows as refugia from predators (Armstrong et. al. 1982; cited in Thorn,
1987), and are found in large numbers is these habitats. Eelgrass is an important substrate
for the attachment of Pacific herring eggs (Thorn, 1987). Eelgrass habitat can make up a
large percentage of total estuarine habitat, for example representing 83% of the bottom
habitat in Padilla Bay, WA (Thorn, 1990). Coastline surveys in Washington state (Thorn and
Hallum, 1991) provide a probable underestimate of 25% of the shoreline as containing
seagrass.
Although the assessment endpoints of our studies are primarily benthic, many of the stressors
are transmitted to the benthic community through the water column (e.g. nutrients, suspended
-------
5
sediments). Thus we will also examine the spatial and temporal patterns of stressors within
the estuarine water column.
At present, our ability to predict how estuarine ecosystems of the PNW will respond to single
and multiple stressors is at a rudimentary level. Historically much of the marine research at
major academic institutions of the region has focused either on blue water oceanography or
open coast rocky intertidal systems, thus there are deficiencies in our basic knowledge about
PNW estuaries. Our proposed research will contribute to addressing the basic informational
deficiencies while evaluating the effects of multiple stressors on PNW estuaries.
Our predictive capabilities are also limited by a lack of appropriate ecosystem-level
approaches. During the last two decades, there have been major advances in methods to
evaluate the effects of point-source pollutant discharges to coastal systems. Many of these
methods are based on toxic effects or bioaccumulation by a few laboratory bioassay species
(e.g., Swartz 1987, Lee et al. 1993). Although the use of toxicity tests and bioaccumulation
measurements of surrogate species has proven useful in regulating individual chemicals, toxic
pollutants appear to play a fairly minor role in most PNW outer-coast estuaries. In addition,
the classical approaches developed for single pollutants are generally inadequate to
determine the effects of nonchemical stressors, the impact of multiple interacting stressors,
the cumulative effects of habitat alteration, overall ecosystem response, or the linkages
between the management of coastal watersheds and estuarine ecosystems. We have little
quantitative information on the effects of watershed alteration and resultant loadings of
estuaries by a variety of potential stressors, such as sediments, nutrients, and toxic
substances. Knowledge of circulation, sedimentation, and runoff in Pacific Northwest coastal
estuaries has not been well integrated, making it difficult to predict direct physical effects on
biota or the physical transport of particulate-associated or dissolved stressors, such as
nutrients. Similarly, quantitative relationships between the loss or alteration of specific
habitats and changes in estuarine structure and functions generally are lacking.
If federal, state, tribal and local governments are to effectively manage this region, existing
knowledge must be synthesized and analyzed from an ecosystem perspective, and critical
data gaps must be filled. Management needs such as the requirements of the CWAP (section
3.0 below), together with the combination of societal and ecological relevance of estuarine
benthic resources, make them an appropriate and important focus of CEB research.
1.2.1 Drivers and Stressors Affecting Pacific Northwest Estuaries
The great natural beauty and relatively low current population density of outer coast estuaries
of the PNW gives an impression of systems that are less altered than those in other areas of
the US. However, PNW estuaries are far from pristine. The list of potential disturbance
agents that have affected PNW estuaries includes habitat loss or alteration, exotic species
introduction, chemical contaminants, watershed alteration with changes in fluxes of
freshwater, nutrients, and sediments, pathogens, over harvest of species, and various other
-------
6
forms of anthropogenic alteration of estuarine resources.
Over the last 200 years, at least six great waves of extractive industry have washed across
the Pacific Northwest, constituting societal drivers producing a variety of stressors which have
altered ecological systems. The nature of the ecological alterations are difficult in some
cases to evaluate because of a lack of data from periods prior to exploitation (Durning, 1996).
In the first extractive wave, sea otters, a known ecological keystone species (Simenstad et al,
1978), were largely removed from coastal ecosystems in the PNW by 1810, and populations
have never recovered. The consequence may have been a persistent transformation of near
shore habitat from kelp forest to urchin barrens, although data are currently lacking to support
this hypothesis.
In contrast to inland locations, the wave of western mining had little direct effect on the outer
coast in terms of altering estuaries or causing chemical pollution. Open coast estuaries are
believed to have generally low concentrations of toxic pollutants such as heavy metals,
chlorinated hydrocarbons and polycyclic aromatic hydrocarbons because of relatively low
population densities and low levels of heavy industry (Copping and Bryant, 1993). However,
data for most estuaries are sparse, and there may be locations, e.g. Coos Bay, where
chemical contamination such as the presence of tributyl tin associated with shipping does
occur.
In the third wave, exploitative fishing devastated coastal fish and shellfish resources. Salmon
catch peaked in Willapa Bay on the Washington coast by 1902, and on the Oregon Coast by
1911 (Durning, 1996). Native oyster populations were largely wiped out by the late 1800's,
leading to attempts at replacement first by the unsuccessful introduction of the eastern oyster
(Crassostrea virginica) followed by the successful introduction of the Pacific oyster
(Crassostrea gigas){S'menstad and French, 1995).
The following exploitative waves of agriculture, logging and damming each resulted in
massive changes to land use practices throughout the region. In the Chesapeake Bay region,
deforestation associated with human settlement and agricultural clearing led to a 100%
increase in sediment accumulation rates (Cooper and Brush, 1991) during the 1800's. It is
likely that similar changes in sediment accumulation rates may have occurred in PNW
estuaries. Sedimentation problems associated with land use changes may be especially
acute in the Pacific Northwest because of the combination of steep coastal watersheds, high
rainfall, and timber harvesting. Dairy farming along streams entering PNW estuaries removed
riparian vegetation and grazing damaged stream banks, generating additional anthropogenic
modifications to estuarine sediment flux. The potential extent of changes due to
sedimentation is illustrated in Tillamook Estuary, where the volume and average depth have
been estimated to have declined by ~60% since the 1930s, as a result of high sediment loads
caused by major forest fires in the watershed (referred to as the Tillamook Burn) and
subsequent salvage logging and road building (James 1970).
-------
7
To improve navagablity of coastal rivers, high numbers of large drift logs were removed and
stream banks were clear cut to remove the source of logs. In the Coos River alone, the ASCE
removed nearly 8,600 logs and blasted over 1,700 boulders prior to 1920 (Gonor et al., 1988).
Removal of such structural elements must have inevitably altered the flux rates of sediments
and organic particulates into the estuary, with concomitant alterations of estuarine water and
habitat quality. While removal of large woody debris from freshwater streams is now known to
have contributed to the decline in salmonid populations in the PNW, no information in
available on what impacts debris removals may have had on estuarine ecological resources.
SAV may have been particularly vulnerable, since it is especially susceptible to decreases is
water column light penetration as a result of increased sediment loadings (Dennison et al.,
1993). For a seagrass such as Zostera marina growing in a marginal habitat, even pulses of
high turbidity of a month's duration may make the difference between local survival or
extinction of a patch (Moore et al. 1996, 1997). Thus the development of a conceptual
relationship between landscape and stream alterations, turbidity flux, and estuarine
populations needs development.
To the six great waves of ecological resource exploitation in the PNW may be added two less
immediately obvious processes, non-point source pollution and introduction of nonindigenous
species, both of which may alter or degrade the estuarine ecological systems in the region.
Non-point source pollution is primarily driven by the steady growth of the resident human
population, which tends to concentrate near water bodies such as estuaries. Increases in the
proportion of impervious surfaces (pavement, buildings, etc.) in the landscape, improperly
designed septic tank drain fields, and storm water drains all may contribute to the problem.
Runoff into the estuary will therefore tend to increase, particularly in high rainfall areas such
as the PNW, bringing with it increased non-point source additions of nutrients (Osborne and
Wiley, 1988), other chemical pollutants, and sediments (Howarth et al., 1991).
Nutrient and sediment loadings from population centers will augment the increased flux of
these materials already resulting from the larger scale watershed alterations associated with
logging of the coastal mountains (Howarth et al., 1991). The typical effects of nutrient
elevation in coastal waters include an increase in hazardous algal blooms (Lancelot et al.,
1987), and a shift in estuarine primary production from rooted SAV to benthic macroalgae and
phytoplankton in the water column. Increased organic and inorganic particulate matter from
runoff causes decreased light penetration (Stross and Sokal, 1989), and reduction in the
depth range of the SAV. Increased macroalgae form large mats on the bottom where they
may smother seagrass, cause localized oxygen depletion of sediments, and consequently
result in major alterations of estuarine benthic habitats (Raffaelli et al., 1998). This alteration
may have direct consequences to other trophic levels such as birds, marine invertebrates,
and fishes dependent on the benthic resources that are degraded.
Increased flux of sediment can also drastically alter estuarine habitats. Most benthic species
have a defined sediment grain size distribution which constitutes the preferred habitat
(Rhoads, 1974; Gray, 1974). Additions of sediments may alter the sediment grain size,
-------
8
generating associated changes in benthic communities. Inorganic suspended material will
also alter the light field, decreasing the ability of SAV to grow. A feedback then occurs, as
vegetation, which binds sediments with roots and causes hydrodynamic damping to cause
sediment deposition, is lost and physical resuspension of sediments becomes greater, further
decreasing SAV. Synergistic effects of suspended sediments and organic particulates may
greatly exacerbate the loss of SAV.
Increased sedimentation may also alter the bathymetric profile of an estuary, shifting the
zones of tolerance for benthic species in the intertidal and potentially eliminating or
significantly modifying the spatial distribution of some species. High sediment loadings may
also clog the feeding structures of suspension feeders, causing their elimination (Young and
Rhoads 1971). There is evidence that this has occurred in Yaquina Bay for oysters planted
on the bottom (Fasten, 1931; cited in Baker et al. 1995).
The second of the more insidious stressors of PNW estuaries is the introduction of
nonindigenous species. While the devastating effects of introduced species in island and
other terrestrial ecosystems has been well described, the effects of nonindigenous species on
estuarine habitats has only recently come under scrutiny (Carlton and Geller, 1993). The
potential for ecological transformation is massive. Some 367 marine invertebrate taxa were
recorded in the ballast water of ships arriving in Coos Bay, Oregon from Japan (Carlton and
Geller, 1993). Introduction of a single species of comb jelly in the Black Sea resulted in the
virtual destruction of pelagic fisheries valued at $250 million annually within only 10 years
(Harbison and Volovik, 1994). Extensive natural and anthropogenic disturbance within
estuarine systems may provide favorable conditions for invasion of exotic species (Cohen and
Carlton, 1998), and thus there may be synergistic interactions among important estuarine
stressors. In the PNW, one by-product of the attempted commercial introduction of the
eastern oyster appears to have been the introduction of smooth cordgrass, Spartina
alterniflora, into Willapa Bay, WA where the spread of this exotic species has resulted in the
conversion of hundreds of hectares of mud flat to salt marsh habitat with consequences to the
ecosystem that have not yet been fully defined (Simenstad and Thorn, 1995).
In an independent assessment conducted by the Tillamook Bay National Estuary Program of
EPA, priority management issues were habitat loss and alteration, sedimentation, introduced
pest species, species loss, and pathogens. The CEB research program will contribute directly
to addressing each of these management issues except pathogens, for which we have no
current expertise.
1.2.2 Estuarine Scaling Issues
Many anthropogenic and natural stressors interact with estuarine habitats in ways that alter
the normal spatial distribution pattern or mosaic of habitat patches. For example, channel
dredging may dissect a shallow sand bottom habitat into sub patches separated by a deeper
depositional environment characterized by finer sediments and altered biological
-------
9
communities. Various impacts to vegetated habitats will alter patch distribution and degree of
fragmentation. Dredging, propeller scars, removal of vegetation for oyster planting,
competitive displacement through bioturbation by burrowing shrimp, water column
eutrophication and increased turbidity will alter patch size, contiguousness, and spatial
distribution. Such alterations have important implications for estuarine ecosystem function.
Recent work from terrestrial systems (Wardle et a!., 1997) has shown that island area is
related to relative carbon partitioning, humus accumulation, nitrogen acquisition and
mineralization, standing biomass, and litter decomposition. There is reason to believe that
there may be important relationships of habitat patch size to estuarine ecosystem functions as
well (Robbins and Bell, 1994).
For example, the patch integrity of SAV may influence sedimentation rates through relative
degree of hydrodynamic damping, which may in turn influence passive larval recruitment and
sedimentary accumulation. Sediment accumulation rates may affect local accumulation rates
of particulate associated toxic substances. SAV patch integrity may influence habitat
resistance to some forms of disturbance. For example, large clonal patches may have greater
nutrient reserves in rhizomes as compared to small patches, which can be shunted to
damaged areas receiving either natural (burrowing shrimp, feeding pits from fishes or crabs)
or anthropogenic (clam digging, prop damage ) disturbance. Similarly, large patches may be
more tolerant of lowered light levels due to seasonal pulse events causing elevated turbidity
and lowered light levels. Other ecosystem properties that can be influenced by patch size for
SAV are:
• relative proportion of exogenous organic carbon accumulated in sediments,
• percentage of in situ Net Primary Production retained within the patch as detritus
versus exported from the patch,
• sediment chemistry properties such as depth of the Redox Potential Discontinuity
(RPD) or Acid Volatile Sulfides (AVS),
• characteristics of associated animal populations (abundance, size structure),
relative contribution of different function groups (detritivores, carnivores, suspension
feeders),
• detrital breakdown rates,
• relative degrees of epiphytic algal cover,
• relative degree of drift algal trapping.
The influence of the relative scale of habitat patches on functional properties within the
patches may also have a cascading effect to other trophic levels. For example, larval
trapping and detrital accumulation may be greater in larger habitat patches, resulting in
superior success for juvenile fish and shellfish.
These examples suggest that overall estuarine ecosystem response to stressors will be a
function both of the tolerance of individual species and habitat elements to a given stressor,
and the spatial distributional aspects of the habitat. Habitat fragmentation has become a
-------
10
crucial issue in terrestrial conservation ecology. As examples, forest fragmentation has been
shown to negatively impact reproductive output of migratory birds (Robinson et al. 1995),
above ground biomass of Amazonian trees (Lawrence et al., 1997), and old field
fragmentation affects small mammal population dynamics and the population persistence of
clonal plants (Robinson et al., 1992). For clonal SAV species, habitat fragmentation is
similarly of potentially great importance in estuarine systems with linked effects to other faunal
components associated with the habitat. The understanding of estuarine habitat scaling
relative to stressor impacts is crucial to the development of a predictive evaluation capability
for stressor response within estuarine systems. One approach to dealing with habitat scaling
relative to stressor impacts is the development of spatially explicit models of estuarine
populations which allow both the spatial extent of the target population and the stressor to be
varied. One research element of the proposed research plan will be to use spatially explicit
models to examine the two dimensional geometry of stressor effects on model populations of
estuarine organisms.
In addition to the use of spatially explicit modeling to deal with issues of scaling and estuarine
stressor responses, CEB has recently hired two principal investigator with research skills that
may be applied to landscape scale issues. A landscape ecologist with research interests in
estuarine habitat distribution patterns and organism responses has been hired. A
biogeographer that is a shared position with the Regional Ecology Branch of WED located in
Corvallis has considerable expertise in dealing with regional scale research questions. An
element of the CEB research plan will develop linkages between estuarine habitat
distributions and landscape alterations and land use patterns surrounding the estuary. As the
research develops, we plan to link estuarine system responses into land use patterns in the
estuarine watershed with the help of the biogeographer, and by means of developing research
collaboration with the Regional Ecology Branch of WED located in Corvallis.
In attempting to extrapolate research results it is important to define the set of estuaries to
which the results of the CEB research plan will apply. There are many physical types of
estuaries in the PNW region, and differences in physical aspects of estuaries clearly must be
considered. At the same time, the geographic region from panhandle Alaska to Northern
California is classed as a single biogeographic province based on similarities of biotic
composition. Estuaries within this province will thus share broad similarities of species
composition. For example, the same seagrass species are distributed across much of the
biogeographic province, and thus CEB research on stressor effects on SAV will generally be
relevant across the region. None the less, placement of CEB research within the context of
an estuarine classification scheme based on physical properties will be helpful.
There is an existing literature which may help to provide a better definition of estuarine
typology in the PNW (Hansen and Rattray, 1966; jay and Smith, 1988; Dethier, 1990;
Montgomery and Buffington, 1993, Jay et al, 1997). CEB will compile in data base format as
much physical information on PNW estuaries as possible and will investigate potential
approaches to categorization of estuaries in order to allow the CEB research effort to be
-------
11
placed into a proper framework for extrapolation. For example, work has shown that the ratio
of drainage basin area to area of the estuary proper is highly correlated with estuarine fish
species richness (Bottom and Jones, 1990). This ratio provides a metric for categorizing the
relative amount of riverine versus marine influence. Other such metrics will be investigated in
relation to the types of parameters which CEB is investigating wherever data sets are
available. Some compilation of comparative data for PNW estuaries has already been done
by other organizations (e g NOAA, 1988, 1991).
1.3 Scientific Questions
The overall research goal is to improve the assessment of risk associated with alterations of
ecological resources of PNW estuaries by multiple physical, chemical, and biological
stressors at a range of temporal and spatial scales.
The principal scientific questions to be addressed are:
• How do stressors affect "physical ecosystem engineering" species, and thereby lead to
alterations in estuarine habitats?
• What is the range of habitat alteration outcomes which result from individual and
multiple stressor effects?
• How may results of small scale, process-oriented research on stressor effects leading
to habitat alteration, be scaled to provide predictions for entire estuaries?
• What indicators of ecological status can be used to assess the effects of multiple
stressors on estuarine habitats?
• What are the consequences of estuarine habitat change on economically and
ecologically important estuarine species?
Each broad scientific question will be examined by means of a series of individual research
elements which are integrated among principal investigators. Research elements are often
mutually supporting, with data collected by one project providing important information to other
projects. The initial research approach will include field observations of stressor magnitude,
temporal variability, and spatial distribution concentrated at single estuary scale. Stressor
measures will be combined with a determination of the distributional patterns and organismal
composition of major habitats within the estuary. Use of remote sensing technologies to
acquire estuarine- scale habitat distributions will be explored. Multivariate, correlational and
geostatistical analytical tools will seek to associate large-scale, within-estuary distributions of
target habitats to stressor distributions. In conjunction with field observations, both field and
laboratory experimentation will be used to test specific hypotheses (defined in individual
research elements, Section 3) related to the interactions between specific abiotic and biotic
-------
12
stressors and habitat types. As the research program matures, additional emphasis will be
placed on comparative studies among estuaries to determine how specific stressor-response
relationships vary among PNW estuaries. In order to provide the capability to develop
predictive stressor response models at estuarine scales, CEB has hired an estuarine modeler
to develop models that couple the spatial and temporal distribution of stressors to predicted
habitat alteration outcomes. The expanded modeling effort will be supported by current
research elements which are developing spatially explicit population models and which are
testing existing estuarine models of water column light field responses to nutrient additions.
-------
13
2.0 Regional Characteristics and Study Area
2.1 Pacific Northwest Estuaries
The Pacific Northwest coastal region and the associated Columbian biogeographic province
extends from Puget Sound in the north to approximately Cape Mendocino in California (Fig.
1). The region contains at least 23 estuaries and coastal embayments with a total estuarine
surface area of 4,504 km2 (NOAA, 1998). The region contains a wide variety of types and
sizes of estuaries, ranging in scale from the fjord like Puget Sound where numerous mid-sized
streams form subestuaries as they drain into the main body of the sound, to small coastal
streams emptying out into the ocean with estuarine basins less than a fraction of a square
kilometer. The majority of estuaries along the outer coast of the Pacific Northwest are either
of the shallow river mouth type or the drowned river valley type, including the large Columbia
River estuary (Copping and Bryant, 1993). Because of their large size, geomorphology,
hydrology, and proximity to large population centers, Puget Sound and the Columbia River
estuary differ considerably from the smaller coastal estuaries of the Pacific Northwest
(Copping and Bryant, 1993).
The estuaries of the Pacific Northwest are geologically young (Simenstad and Fresh, 1995).
Estuarine geomorphology and ecology is periodically affected by tectonic events which may
result in catastrophic changes. For example, subsidence events of 0.5 - 2.0 m along the coast
appear to have taken place at least six times in the last 7,000 years (Atwater, 1987).
Geological studies indicate that as little as three hundred years ago, a massive tsunami
impacted coastal estuaries of the Pacific Northwest, causing transformation of coastal forests
into salt marsh (Yamaguchi et al., 1997).
The climate of the region is strongly marine dominated, with seasonal weather patterns being
heavily influenced by offshore current patterns. The dominance of northerly winds during the
summer results in the southerly flow of the California current with associated Ekman transport
of surface waters away from the coast generates coastal upwelling. The nutrient-rich
upwelling water reaches the photic zone and stimulates the surface primary production near
shore, helping support the rich commercial fisheries of the region. Summer fog as a result of
the upwelling is also a common summer occurrence. In winter, wind direction is primarily from
the south or southwest, generating a seasonal countercurrent, the Davidson current, which
flows northward near the coast. The northward flow of warmer water ameliorates winter
temperatures in the region. Strong winter storms from the southwest off the Pacific produce
high rainfall amounts.
The proximity of the Coast Range of mountains to the outer coast along the Pacific Northwest
means that, with the exception of the massive Columbia River, watershed drainage areas of
coastal estuaries are relatively small. However, there are steep stream gradients and large
volumes of freshwater runoff typically enter the estuary shortly after rainfall events (Copping
-------
14
Figure 1. Map of estuaries of the west coast of the United States. Outer coast estuaries in
the Columbian biogeographic province include those from Eel River CA to Grays Harbor, WA
(numbers 16-32). Map from National Oceanic and Atmospheric Administration, 1998.
Pacific
Northwest
Coast
Central
California
94) Hood Canal
32) Grays Harbor
31)WillapaBay
30) Columbia River
•29) Nehalem River ,
28) TIBunook Bay - ¦
27) Netarts Bay
2£}S3etz Bay
25) Yaquina Bay
24) Alsea Rive
23) Siuslaw River
22) Umpqua River
21) Coos Bay
20) Coquille River
19) Rogue River
18) IQamath River
17) Humboldt Bay
1$) Eel River
15) Tomales Bay
14) Drakes Eslero
13) Central San Francisco Bay /
San Pablo/Suisun Bay
12) San Francisco Bay
10) Monterey Bay
Southern
California
Coast
} Waihineton Cotital Bavi
35) Whidbey Basin/
Skagit Bay
33) Puget Sound
37) Port Orchard Syitem
36) South rogtt Sound
Wuhington
California
l)Hkhom Slough
9) Motto Bay
6) Alamitof Bay
*©©»- I 7) Anaheim Tiay
8) Santa Monica Bay ' Newport Bay
5) San Pedro gay'
3) Mission ^y
2) San Diego Bay
1) Tijuana Estuary
-------
15
and Bryant, 1993). Freshwater inflow to the coastal zone in the Pacific Northwest is totally
dominated by the Columbia River which accounts for 90% of flow in the summer, and 60% of
total flow in the winter (Simenstad et al., 1990).
Much of the land comprising the watersheds of PNW estuaries is still forested, with silviculture
being the dominant land use practice. For some estuaries, a high percentage of the forested
watershed is in public ownership, which means that it may be possible to deal with some
estuarine stressor problems through public policy changes.
With the exceptions of the Columbia River estuary and Puget Sound, the remaining coastal
estuaries appear to differ primarily in estuarine area, rather than in the fundamental nature of
the physical processes and stressors which influence the environmental conditions within the
estuary (Shirzad et al., 1988; NOAA, 1997). Detailed comparative studies of benthic infauna
among Pacific Northwest estuaries is presently lacking, but occurrence of benthic
invertebrates and fish of commercial importance is broadly similar among estuaries of the
region, although relative abundances vary considerably (Monaco et al., 1990). Variations
among estuaries of the Pacific Northwest in the extent of tidal flushing, retention times of
water masses, relative contributions of river runoff and marine waters to estuarine nutrient
loadings, and land use patterns within the estuarine watershed offer the potential to achieve
significant insights into estuarine risk assessments by comparative studies among estuaries.
2.2 Yaquina Bay Study Area
In order to take advantage of the body of historical research carried out by a variety of state
agencies, federal agencies, and university scientists housed at the Hatfield Marine Science
Center, the Yaquina Bay (Fig. 2) estuarine system has been selected as the primary study
area. This choice also allows optimum utilization of limited logistical support funds available
to CEB. After the research methods and approaches have been tested in Yaquina Bay,
supplementary observations to broaden the inference base will be obtained from Siletz Bay
and Alsea Bay, estuaries located immediately to the north and south of Yaquina Bay (Fig. 1).
Some chemical and biological observations in Willapa Bay, WA (Fig. 1) will also be continued
in order to take advantage of a longer time series of data than is available from Yaquina Bay,
and to provide the basis for an initial comparisons of ecological conditions among estuaries.
A further advantage of focusing CEB research on Yaquina Bay is that various research
agencies are currently conducting or initiating research in other Pacific Northwest estuaries,
and it will be possible to develop a broader spatial extrapolation of results from inter-estuary
comparisons by coordination with these research efforts. South Slough at Coos Bay, a part of
the National Estuarine Sanctuary Program, is the site of an estuarine research program
conducted under sponsorship of NOAA. Research in Tillamook Bay is being conducted as
part of the EPA National Estuary Program, and ancillary research in this estuary is being
conducted by McManus et al. of Oregon State University under sponsorship of the EPA STAR
grants program. The effects of removal of salt marsh dikes on the ecosystem function of the
-------
16
Salmon River estuary in being conducted by Simenstad et al. of the University of Washington.
The Pacific Northwest Coastal Ecosystem Study (PNCERS), funded by the NOAA Coastal
Ocean Program, will focus on Willapa Bay and Coos Bay. CEB is a collaborator on the
PNCERS project. Willapa Bay is also the site of previous research under the EPA Pacific
Northwest Research Program.
Yaquina Bay estuary lies at 44° 37' north latitude in the cool temperate Columbian
biogeographic province. The estuary, formed by the submergence of a portion of the Yaquina
River drainage basin, is approximately 17 km2 with an intertidal area representing
approximately 35% of estuarine area on average, but up to 60% on extreme tides. The river
drainage basin is approximately 655 km2, with a mean tidal range at the entrance of the
estuary of 1.8 m and a range of 1.92 m at Toledo, 20 km up estuary. Mean flow of the
Yaquina River is 30.5 m3 s'1, although flows may range from 1.3 to 87 m3s"1 (Callaway and
Specht, 1982). Rainfall in the Newport area averages 152 cm yr'1, with most rainfall occurring
in the period of November through March (Good, 1975). The estuary is classified as well
mixed under low flow summer conditions, but is classed as partially mixed with some vertical
stratification of waters of differing salinity under winter high-flow conditions (Shirzad et al.,
1988). The estimated retention time of freshwater within the estuary is approximately five
days (Shirzad et al., 1988).
Land use within the watershed of Yaquina Bay is currently approximately 95% forest, with the
remainder as urban, agricultural, and rangeland uses. Estimates suggest that inputs of
nitrogen are primarily nonpoint source runoff from forests, while phosphorus comes
predominantly from wastewater treatment plants (NOAA/EPA, 1991). Yaquina Bay is
estimated to be in the high range among west coast estuaries in terms of potential for
concentrating dissolved substances, but is in the medium range for existing levels of nitrogen
and phosphorus (NOAA/EPA, 1991). Based on limited data, the 1998 NOAA Estuarine
Eutrophication Survey stated that turbidity and chlorophyll a concentrations are at medium
levels in Yaquina Bay relative to other west coast estuaries, and that anoxia and hypoxia have
not been observed within the bay. These data are almost anecdotal in nature, and it is
probably premature to conclude that the ecological systems of Yaquina Bay are not being
stressed at least seasonally by these factors.
While the primary focus of the Research Plan in the near term will be in and around the
Yaquina Bay ecosystem, future comparisons to other Pacific Northwest estuaries wilt be
important in the long term. Research results from Yaquina Bay should be relevant in a fairly
broad geographic context, including many of the outer coast estuarine systems from the
panhandle of Alaska to northern California. The suggested inference space should not be
construed as an attempt by CEB to investigate all estuaries in this region. The logistical
details of the research plan (Section 4) clearly focus research primarily on Yaquina Bay
estuary in the initial phases of research. Some research on Willapa Bay is being maintained
by two projects (Sections 4.1 :A1 Ferraro/Cole, 4.3:C3, Sigleo). Minimal resources are being
committed to the Sigleo sampling effort, primarily to allow continuation of a time series of data
-------
17
she has collected in Willapa. A greater level of effort for the Ferraro/Cole project in Willapa
Bay has been provided because a comparison of habitat composition among estuaries is an
important component of the research project. The overall commitment by CEB to estuarine
locations other than Yaquina Bay for research is presently low.
There is the potential for a research effort in Tillamook Bay in FY 99 in support of a Region 10
interest in an EMAP intensive study of this estuary. Expansion of research efforts to other
PNW estuaries is being approached cautiously by CEB, and will only be initiated in a major
way after research approaches are thoroughly confirmed by work in Yaquina Bay.
-------
Figure 2. Map of the Yaquina Bay estuarine system. Map from National Oceanic and
Atmospheric Administration, 1998.
Yaquina Bav OR
Pacific
Ocean
Oregon
Kng
Sbvgh
Pooh Slough
N
t
Legend
SaBnHy Zones Salinity Zone Boundarie*
Hi Tidal Fresh (<0.5ppt) EZIH LowVarfabWy I ~ I Tide gauge
MMng (05 • 25.0 ppt)
[3 Seawater (» 25 ppt)
Medhim Variability
HghVariafctty
Head of Tide
-------
19
3.0 Research Approach
3.1 Research Theme Organization
To accomplish the objectives of the CEB research plan, research will be organized in three
thematic elements:
A. Indicators of Ecological Condition for PNW Estuaries
B. Stressor-Response Modeling
C. Estuarine Physical-chemical Stressors
Within each of the above research elements, a number of individual research projects are
proposed. Individual research projects integrate both within and across research themes. A
schematic indicating the organization of the research themes is provided in Figure 3.
Research projects are described in detail in sections 4.1 - 4.3 below.
The research projects under Research Theme A seek to address the questions: 1) what are
the biotic constituents of major estuarine benthic habitats of PNW estuaries, 2) what effects
do various abiotic and biotic stressors have on the biotic composition of principal habitat
types, 3) what role do biotic and abiotic stressors have in controlling the spatial extent and
distribution patterns of major estuarine habitat types, and 4) what are appropriate indicators
of ecological condition at the population, species, community and landscape levels for PNW
estuarine systems. A list of research projects for this theme is given in Table 1. Stressors
that will be examined include anthropogenic physical disturbances such as clam and
burrowing shrimp harvesting, and sedimentation, salinity, and water column light field
alterations potentially generated by elevated runoff resulting from changes in landscape-use
patterns. Biotic stressors that will be examined include disturbances such as the smothering
of seagrass habitat by mat-forming algae potentially promoted by elevated nutrients, and
biotic stress induced by competition between native and exotic seagrass species and between
burrowing shrimp and seagrasses.
Projects under Research Theme B will work at the population and community levels to
develop modeling techniques to integrate the detailed studies of biological effects of estuarine
stressors of Theme A with the spatial-temporal stressor distribution studies of Theme C. The
principal current project is the development of spatially explicit modeling tools for estuarine
benthic populations to allow predictions of population responses to the imposition of multiple
stressors. While initial model development will focus on the specific example of a bivalve
population, the modeling approach is a general one which can be applied to fish, crustacean,
or even seagrass populations. The ecological modeler currently being hired by CEB will
conduct modeling efforts to integrate research results at the community and landscape levels
of organization. Until this individual has joined the research team, a precise research program
can not be defined.
-------
20
Figure 3. Schematic diagram of interactions among research plan elements. Names placed
on the lines of the Integration section indicate projects with a secondary modeling component.
Names in italics indicate CEB staff currently developing additional research elements.
INFORMED POLICY DECISIONS
THROUGH IMPROVED RISK
ASSESSMENT
A - Estuarine Indicator B - Stressor-ResponsdS - Physidal-Cherrical
Development Modeling Stressors
Ferraro-
Cole
Ycxng
Specht
Boese
Lee
DeWitt
Elctidge
Robbins
Sigleo
etich
Nelson
Specht
Lambecson
Research Element Interactions - Coastal Ecology Branch
-------
21
Table 1. Summary of individual research projects in Research Theme A, Indicators of
Ecological Condition for PNW Estuaries, indicating the ecological level of the estuarine
indicator being studied, the primary stressors affecting the indicator, and selected assessment
endpoints.
Project PI Level of ecological Stressors Assessment
No. indicator Endpoints
A1
Ferraro
Cole
community
landscape
altered habitat structure
altered water chemistry
(nutrients),
exotic species,
toxics,
watershed alterations
community
structure, mosaic
of ecosystem types
A2
Robbins
community,
landscape
altered sediment
dynamics, altered
hydrology,
altered geomorphology
altered habitat structure
relative coverage
of ecosystem
types, spatial-
temporal patterns
of habitat
A3
DeWitt
population,
species,
community,
landscape
altered sediment
dynamics, altered
hydrology,
altered geomorphology
altered habitat structure
seagrass habitat
quality and
quantity,
ecological
condition of
burrowing shrimp
(ecological role)
A4
Boese
species,
community
physical disturbance,
altered water chemistry
(nutrients)
seagrass habitat
quality and
quantity,
community
structure
A5
Specht
population,
species,
landscape
altered sediment
dynamics, altered water
chemistry (nutrients),
exotic species
seagrass habitat
quantity, relative
coverage of
ecosystem types
A6
Power
population,
species
altered sediment
dynamics, altered habitat
structure,
exotic species
ecological
condition of
population and
species, trophic
relationships
-------
22
The research projects under Research Theme C seek to address the questions 1) what are
the spatial and temporal distribution patterns of the primary physical and chemical factors
determining estuarine habitat composition, 2) how are spatial variations in the physical-
chemical stressors associated with variations in target estuarine habitats, 3) how do
anthropogenic alterations of watershed characteristics influence the transport of dissolved and
particulate materials into the estuary, and 4) what are the relative roles of oceanic versus
riverine inputs of dissolved and particulate materials on water column physical-chemical
processes. Collection of data is integrated among the various project elements to the greatest
extent possible. Collection of data for physical and chemical parameters and their spatial and
temporal variation within the estuary will consist of deployment of multiple instruments capable
of logging data along the upstream axis of the estuary. Fixed station data will be augmented
by periodic cruises to deploy equivalent instrument packages to obtain more extensive
information on spatial variation of these parameters. A state-of-the-art in situ nitrate analyzer
with data logger will be deployed at the mouth of the estuary to determine the temporal
variation in nutrient inputs to the estuary from offshore sources. Data collected by CEB
scientists will be supplemented from sources such as the State of Oregon Dept. of Water
Resources which maintains a river flow gauge on a main stem of the Yaquina River.
Arrangements to obtain these data from the state on a near real time basis have been made.
As one example of how interactions among projects are structured, the projects of Boese,
DeWitt and Kentula, and Specht all will provide data on various aspects of the population
dynamics of the seagrass Zostera marina. Boese will evaluate effects of physical disturbance
and algal smothering, DeWitt and Kentula will evaluate effects of salinity, sedimentation, and
burrowing shrimp, and Specht will examine sediment nutrients, temperature and
geomorphology as controlling factors on Zostera populations. Young and Ozretich will
provide data on water column light fields, sedimentation rates, and large scale distribution
patterns of Zostera relative to physical stressors. Lee, with the spatially explicit population
model will be able to input population characteristics, growth and mortality rates in response
to stressors, and the spatial distribution of stressors in a spatially explicit population model
which will allow specific predictions of population survival given different combinations and
spatial distributions of stressors. The community structure assessments of Ferraro and Cole
can be used to extrapolate the effects of stressors on the seagrass species to other co-
occurring species in the habitat.
3.2 Key Infrastructure Support Elements
CEB has as a part of its research plan the establishment of a series of 5 permanent
environmental monitoring stations for continuous acquisition of environmental data. The
instruments for these stations have been acquired, and the details of deployment are currently
being worked out (Section 4.3:C2, Ozretich). For a number of logistic reasons, the
instruments will be deployed from pilings rather than buoys. Instruments will measure
conductivity (salinity), temperature, depth, turbidity, and photosynthetically active radiation.
One CTD has been deployed at the HMSC dock, and additional CTD's have already been
-------
23
deployed in temporary moorings up river. CEB is also investigating the possibility of obtaining
instrumental data in real time using radio modems.
CEB has an in situ instrument for continuous measurement of nitrate deployed at the HMSC
dock, and an additional nitrate monitor scheduled for deployment in the freshwater area near
the head of tide. Deployment of additional continuous nitrate monitors within saline areas of
the estuary will require the location of a source of more cost effective instrumentation. The
instrument currently deployed cost nearly $25,000, and acquisition of 4 more instruments for
nitrate analysis at all stations would be prohibitive given current CEB Expense dollar budgets.
More cost effective instrumentation is currently in the developmental stage at several
institutions, and we are hopeful that in the near future they will be commercially available so
that CEB will be able to add a full array of nutrient monitors to the CTD instrumentation
currently available.
CEB will also commit internal resources to the establishment of a mesocosm facility for
controlled experiments which is viewed as an important part of the research program Section
4.1 :A3, DeWitt). However, CEB financial resources are insufficient to allow construction of a
mesocosm facility with optimal research capabilities in a single phase of construction.
Resource limitation will mean that expansion of the mesocosm will require 2-3 years to
complete, and the scientific capabilities of the facility will be concomitantly limited over this
period.
3.3 Research Integration, Interactions and Continuous Improvement
The importance of information exchange and research interactions among all CEB research
scientists in order to achieve the CEB research plan is recognized. CEB will conduct weekly
team meetings, and biweekly research update discussions by all team members will be
introduced. Additionally, ad hoc meetings to coordinate field work, to optimize sample
collection, and to plan for allocation of technical support resources are frequent. CEB also
plans to establish an annual research retreat which will be used to provide a forum for a more
extensive internal presentation of research results by all investigators. Annual progress
towards meeting individual project goals and objectives will be reviewed, and progress
towards overall CEB objectives will be summarized at this retreat.
Establishment of a variety of partnerships is also essential to the success of the CEB research
mission. Principal Investigators (Pi's) are encouraged to interact with academic institutions by
becoming adjunct faculty members in appropriate university departments. Individual Pi's have
established cooperative efforts with researchers at the USGS and US Army COE. CEB has
made a concerted effort to identify state and local agencies which have research interests in
PNW estuaries, and has some cooperative activities already in place. For example, CEB is a
cooperator with the NOAA funded Pacific Northwest Coastal Ecosystems Regional Study
(PNCERS), and has provided GIS support to habitat restoration efforts in Yaquina Bay
sponsored by the Mid-Coast Watershed Council. CEB is interacting with Washington Sea
-------
24
Grant, Oregon Sea Grant, and the Pacific Estuarine Research Society to sponsor a special
symposium session on research in outer coastal estuaries of the Pacific Northwest. CEB has
collaborated with Oregon Sea Grant to sponsor the establishment of an on line bibliography of
research publications on Yaquina Bay. CEB will continue to expand collaborative research
contacts.
An increase in contacts with outstanding scientists in a variety of relevant research areas
would also be valuable to branch scientists. We recognize the need to increase our
interactions with physical scientists in particular. CEB has already initiated discussions with
estuarine researchers with international reputations for potential short term visits. The most
likely mechanism to provide funding would be through the existing National Research Council
fellowship program by providing support to selected Senior Fellows. CEB will continue to
actively explore mechanisms for attracting visiting investigators to the laboratory.
CEB staff are strongly encouraged to present research results at scientific conferences and to
publish results in a timely manner. CEB strives to provide travel support to each scientist to
present at least one paper at a major scientific meeting annually. CEB will also cosponsor
the Pacific Estuarine Research Society meeting in April, 1997, and CEB scientists are being
encouraged to present their preliminary results at this conference. CEB scientists provide an
annual Individual Publication Plan, and overall performance evaluation of scientists is strongly
weighted to such items as meeting presentations and journal publications.
An important area for branch improvement is that CEB lacks in house scientific expertise in
physical oceanography and numerical circulation modeling, areas which are critical to an
integrated estuarine research program. One mechanism to achieve such in house expertise is
via an NHEERL postdoctoral term position. In the event such a position can not be obtained,
we believe that by establishing partnerships with other academic and agency groups, the
need can be met. CEB has already let a contract for production of a two dimensional
numerical model for Yaquina Bay circulation. Dr. Peter Eldridge, the new ecological modeler
at CEB, has worked with numerical models of estuarine circulation. While he is not a physical
oceanographer, he is able to apply modeling techniques to examine predictions of distribution
of chemical parameters in estuaries. Other potential cooperators include oceanographers at
the College of Oceanic and Atmospheric Sciences at Oregon State University, and numerical
modelers with other laboratories of EPA (NERL), the US Army Corps of Engineers, and
NOAA.
3.4 Programmatic Support of Research Plan
The CEB research plan will support important federal initiatives such as the Clean Water
Action Plan. Examples of the relationship between CEB research elements and a number of
specific action areas of the CWAP are given in Table 2. As one example, CEB research to
determine the factors controlling distribution of eelgrass, which is a major wetland type in
-------
25
Table 2. Coastal Ecology Branch research activities in support of Clean Water Action Plan
priorities.
Priority 1) Restore and Protect American's Wetlands
Key Action: EPA and the Corps will emphasize restoration and mitigation of wetlands as remedies for section
404 violations.
Key Action: EPA {and other agencies} will provide technical.... assistance to states and tribes to integrate
habitat considerations into geographic-based planning programs ...
CEB Research
Determination of parameters controlling distribution of submerged aquatic vegetation (SAV), a major wetland
type, in Pacific Northwest estuaries.
Development of spatially explicit models for assessing population-level effects of stressors on estuarine species
including (SAV).
Determine historical trends in landscape use, including riparian vegetation, around a west coast estuary and
relate alterations to changes in SAV distribution patterns.
Priority 2) Protect Coastal Waters
Key Action: EPA {and other agencies} will develop a plan by the end of 1999 for coordinated monitoring of
coastal waters and will, by the end of 2000, develop a comprehensive report to the public on the
condition of the nation's coastal waters
Key Action: NOAA and EPA will further develop and support partnerships with state, tribal, and local
governments and organizations to provide technical assistance and information to local decision makers
in coastal areas...
CEB Research
Development and evaluation of new indicators of estuarine condition at the population, community and
landscape levels of ecological organization.
Priority 3) Reduce Nutrient Over-enrichment
Key Action: {EPA and other agencies} will... model and produce estimates of inputs, nutrient utilization (by
major source category), transport, and net contributions of nitrogen and phosphorus in watersheds across
the nation.
CEB Research
Modification of water quality models for SAV growth for west coast estuaries.
Determine magnitude, variation and sources of inputs of nutrients within selected west coast estuaries.
Priority 4) Unified Watershed Assessments
Key Action: Federal agencies will provide technical assistance or funding support for state efforts to develop
unified assessments of watershed health.
CEB Research
Determine historical trends in landscape use, including riparian vegetation, around a west coast estuary and
relate alterations to changes in submerged aquatic vegetation distribution patterns.
Development and evaluation of new indicators of estuarine condition at the population, community and
landscape levels of ecological organization.
-------
26
Pacific Northwest estuaries, will support the CWAP action of encouraging the restoration of
wetlands. Without a detailed knowledge of limiting factors for eelgrass growth, it is difficult for
resource managers to determine appropriate restoration targets. Such knowledge is also
required to maximize the success of actual restoration projects which may be carried out.
The research program proposed is a five year program which will lead to the goal of
developing a framework for prediction of ecological changes to PNW estuaries resulting from
multiple stressors by the year 2003. The precise nature of the prediction framework can not
be stated at this time. Two pathways are being considered. One is the use of integrative
models to generate prediciton of resource alteration based on stressor levels. The seonnd
possible approach is to develop CEB research results into an "alternative futures framework.
In this approach, future scenarios are developed based on projections of current ecological
conditions and economic and management policies and contrasted with alternative scenarios
based on variations in important factors influencing ecological change. Development of an
alternative futures approach would require building significant interactions with academic
research programs outside of CEB.
Milestones for the production of scientific products from the research are outlined in Table 3.
Research will be initiated in the summer of 1998.
Individual project descriptions follow in Section 4. Each project description is structured as a
research proposal containing a statement of the research goal, the rationale for the individual
research component, the specific research objectives, the scientific approach describing the
methods to achieve the research objectives, and the expected benefits of the research.
-------
27
Table 3. CEB Research Goal and Product Milestones.
Goal:
Development of a Framework for Prediction of Ecological Changes to Pacific Northwest
Estuaries Resulting from Multiple Stressors - YR 03.
CEB Goal Milestones:
1) Selection of approach for development of prediction framework - YR 00
2) Development of prediction framework - YR 03
CEB Research Project Product Milestones:
Theme A. Indicators of Ecological Condition for PNW Estuaries
(1) Peer-reviewed scientific papers on estuarine habitats as ecological indicators - YR 00, 02,
03
(2) GIS maps of estuarine resources. Technical assistance to Mid-Coast Watershed Council -
YR 98, 99, 00.
(3) Peer-reviewed scientific papers on landscape scale ecological indicators - YR 98, 99, 00,
01.
(4) Peer-reviewed scientific papers on exotic species as indicators of estuarine stress - YR 00,
01, 02.
Theme B.Stressor-Response Modeling
(1) Peer-reviewed scientific papers on spatially explicit models for assessing population -level
effects of stressors on estuarine species including (SAV) - YR 01, 02, 03
(2) Peer-reviewed scientific papers on models to integrate multiple stressor effects on
estuarine habitats - YR 01, 02, 03
Theme C. Estuarine Physical-Chemical Stressors
(1) Peer-reviewed scientific papers on water quality models for submerged aquatic vegetation
for west coast estuaries - YR 00, 01, 02, 03
(2) Peer-reviewed scientific papers on the magnitude, variation and source of input for
nutrients and other estuarine physical-chemical stressors - YR 00, 01, 02
(3) Peer-reviewed scientific papers on parameters controlling distribution of submerged
aquatic vegetation - YR 99, 00, 01
-------
28
4.0 Detailed Research Project Descriptions
4.1 Research Theme A. Indicators of Ecological Condition for PNW Estuaries
Project A1 - Estuarine Biota-habitat Relationships
Principal Investigator: Steven P. Ferraro
Co-Principal Investigator: Faith A. Cole
Goals: This research project contributes to the programmatic goal of developing cost-
effective methods, measures and models for predicting the cumulative effects of natural
and anthropogenic stressors on ecologically and economically important biotic
resources and ecosystem services values (e.g., food production) (Costanza et al. 1997)
of Pacific Northwest (PNW) estuaries. Specifically, the primary goal of this research
project is to determine the functional value of major PNW estuarine habitats in terms of
selected measurement and assessment endpoints. The primary anticipated products
are empirical models of biota-habitat relationships determined at a sufficient level of
resolution for use in alternative futures analysis (U.S. EPA 1995) and large-scale
ecological risk ("ecorisk") assessments (e.g., Landis and Wiegers 1997; Wiegers et al.
1997). Within-estuary biota-habitat relationships will be determined by testing the null
hypothesis (H,): There are no significant differences on endpoints of interest between
habitats in a given estuary. Temporal variability of within-estuary biota-habitat
relationships will be determined by testing the null hypothesis (H2): For a given habitat
type and estuary, there are no significant differences on endpoints of interest between
sampling events. Among-estuary biota-habitat relationships will be determined by
testing the null hypothesis (H3): For a given habitat type, there are no significant
differences on endpoints of interest among estuaries. The measurement and
assessment endpoints will include ecologically and economically important population
and community metrics (Hunsaker and Carpenter 1990), the densities of societal-
valued species and their prey (Simenstad et al. 1991), ecosystem-level distress
syndrome indicators (Rapport et al. 1985), and indicators of biotic integrity (Nelson
1990; Weisberg et al. 1997; Deegan et al. 1997; Karr and Chu 1997).
The basic approach and sampling and statistical design we will use is described in
detail in Ferraro and Cole (1996a,b). In summary, an EMAP-type (U.S. EPA 1992a)
stratified (by habitat) random sampling design will be used to estimate summary
statistics estuary-wide for each of the habitats investigated. Single classification
ANOVA (Sokal and Rohlf 1995) will be used to test the null hypotheses of no
differences in endpoints among habitats (H,), time intervals (H2), and estuaries (H3).
Within-estuary biota-habitat relationships will be determined by the magnitude of the
mean difference, if any, in endpoints (A, B,...) among habitats (X, Y,...) in a given
estuary. If H, is accepted, endpoint A is statistically indistinguishable in habitats X and
Y in estuary Q, and the common mean value of endpoint A is the best estimate for
-------
29
habitats X and Y in estuary Q. If H, is rejected, endpoint A is statistically different in
habitats X and Y in estuary Q, and there are different best estimates of endpoint A for
habitats X and Y in estuary Q. The ability to extrapolate biota-habitat relationships
over time is tested by H2. If H2 is accepted, the biota-habitat relationship holds over the
time period tested; if H2 is rejected, they do not. The ability to extrapolate biota-habitat
relationships across estuaries is tested by H3. Fidelity of an among-estuary biota-
habitat relationship for a given habitat is indicated if H3is accepted, i.e., there is no
significant difference in the mean values of a given endpoint in habitat X among
estuaries (Q, R), and the common value of endpoint A is the best estimate for habitat X
in estuaries Q and R. If H3is rejected, endpoint A is statistically distinguishable in
habitat X in estuary Q and R, and best estimates of endpoint A in habitat X differ in
estuaries Q and R.
Biota-habitat relationships will be determined based on the statistical outcomes of tests
of H1f H2 and H3 The strengths of the biota-habitat relationships will be assessed by
significance testing at three alpha levels: a = 0.05, 0.20, 0.50, with the three a-levels
taken to be indicative of strong, intermediate, and weak relationships, respectively.
Since each of the four parameters of statistical inference (a, p, n, 6) are a function of
the other three, comparing statistical inferences at different a-levels, as we propose, is
analogous to comparing the ability to detect "small," "medium," and "large" effect sizes
(6) (Cohen 1977), respectively.
Rationale: Alternative futures analyses and ecorisk assessments depend on estimating
exposures on receptors or habitats and estimating the effect of the exposure on the
receptors or habitats to predict future conditions (futures analysis) or to assess risk
(risk assessment) (U.S. EPA 1992b, 1995; Freemark et al. 1996; Landis and Wiegers
1997; White et al. 1997). Estimates of exposure and effects may be obtained by
measurements or models. The typical futures analysis and ecorisk assessment
components are:
exposure effect
Stressor > Receptor or Habitat > Response.
In futures analyses, and in ecorisk assessments where the receptors are habitats,
habitat area may be an assessment endpoint, if the habitat itself is of direct
environmental value, or habitat area may be a measurement endpoint for one or more
assessment endpoints associated with the habitat. In the latter case, knowledge of
biota-habitat associations provide one with the ability to translate observed or predicted
changes in habitat areas to changes in habitat-associated assessment and
measurement endpoints.
Three premises underlie the rationale for using a habitat-based approach for futures
analyses and ecorisk assessments in PNW estuaries. Premise 1: Most estuarine
-------
30
species exhibit habitat preferences and are distributed non-randomiy among habitats.
There is much evidence which supports this premise (see, e.g., Briggs and O'Connor
(1971), den Hartog (1977), Bayer (1979, 1981) Albright and Bouthiiiette (1982), Kneib
(1984), Zimmerman and Minello (1984), Phillips (1984), Wenner and Beatty (1988),
Dethier (1990), Heck et al. (1995), Zipperer (1996), and Bostrom and Bonsdorff
(1997)). Premise 2: The major ecological stressors in PNW estuaries are
sedimentation, nutrients, and the spread of nuisance exotic species, e.g., the Atlantic
smooth cordgrass, Spartina alterniflora. This was the consensus of a group of experts
(Williams and Zedler 1992); also see Baker et al. (1995). Premise 3: First-order effects
of the major stressors on PNW estuaries will be to change the type (e.g., from
unvegetated to Spartina, subtidal to intertidal, etc.) and area of habitats in PNW
estuaries. If premises 1-3 are correct, biota-habitat relationships can be used in
ecorisk and futures analyses to translate observed or predicted changes in habitat
areas into estuary-wide biotic effects by simple mathematics. For example, if estuary
area is lost by sedimentation or diking, the biota in the estuary is decreased in
proportion to the area of the habitat lost. If habitat areas change in the estuary, there
will be biotic winners and losers. An increase in Spartina habitat, for example, results
in an increase in the biota associated with Spartina and a decrease in the biota
associated with the habitat(s) displaced by Spartina. In habitat-based ecorisk
assessments, risk is inferred where stressors intersect habitats, and if they intersect,
the magnitude of the risk is inferred from the importance of the habitat (Landis and
Wiegers 1997), which may be quantified by the magnitude of the endpoints of interest
associated with the habitat. Other projects (e.g., NOAA 1997; Simenstad et al.
1997a,b; Young 1998) are being implemented to map PNW estuarine habitats. The
Coastal Ecology Branch (CEB) and others (Grue 1995) also plan to map habitat
distributions in PNW estuaries using data obtained from historical records.
Estuarine habitats are determined by bio-geophysical parameters, in particular, salinity,
temperature, substrate type (sediment grain size, vegetated versus unvegetated),
bathymetry, and the presence of keystone species (Kendall 1983; Kneib 1984; Phillips
1984; Dethier 1990; Zipperer 1996). We have tentatively identified eight major PNW
estuarine habitats- (1) Spartina alterniflora, (2) Zostera marina, (3) muddy sand, (4)
ghost shrimp (A/eof/ypaea)-dominant, (5) mud shrimp {Upogebia)-dom\nan\, (6) oyster,
(7) subtidal, undisturbed, and (8) subtidal, dredged- which together account for most of
the area of most PNW estuaries. We chose this habitat classification because we
believe these habitats support substantially different biota, their sizes and areal
distributions are likely to change as a function of the major stressors, and because they
can be relatively easily identified and mapped. The subtidal, dredged habitat will not
be included in our studies as a considerable amount of information already exists on
the effects of dredging (U.S. Army Corps of Engineers 1996). Other estuarine habitats,
such as macroalgae-covered, mud, and algal mats, may also be included in our study if
they are a significant component of the estuary. We will critically evaluate the results
of our analyses on our tentatively defined habitats, and may redefine "habitats," by
-------
31
combining our habitats which do not support appreciably different biota, or by splitting
our habitats into finer categories based on post stratification analyses with other
potential habitat-defining variables (e.g., grain size, vegetation density, salinity,
temperature).
Our measurement and assessment endpoints will include population and community
metrics of ecologic and economic importance (e.g., valued fish and crustacean species
and their primary prey (Simenstad et al. 1991)), keystone species (Hunsaker and
Carpenter 1990), ecosystem-level distress syndrome indicators (Rapport et al. 1985),
anthropogenic-effects indicators (e.g., abundance and % non-indigenous species), and
components of Nelson's (1990) index of biotic integrity, the Estuarine Benthic Index of
Biotic Integrity (Weisberg et al. 1997) and the Estuarine Biotic Integrity Index (Deegan
et al. 1997). Specific measurement endpoints will include the total number of species,
numerical abundance, total biomass, numerical abundance and biomass by major taxa
(crustaceans, polychaetes, molluscs, fish), diversity (e.g., Brillioun's, Mcintosh's,
Simpson's, and Swartz's indices), and number and biomass offish prey, bird prey, and
habitat-engineering species (e.g., Neotrypaea and Upogebia). Metrics based on life
history traits (e.g., density of short-lived opportunistic species and long-lived persistent
species), trophic level, or feeding mode (e.g., density of deposit and suspension
feeders) may also be used to assess habitat differences with respect to faunal types.
These endpoints were chosen for their ecological and/or societal relevance and their
susceptibility to the major stressors (U.S. EPA 1992b; Suter 1993). We focus on
benthic fauna because they are generally more sedentary and have stronger links to
habitats, they are easier to quantitatively sample, they are sensitive to stressors, and
they often mediate effects at higher trophic levels (U.S. EPA 1991). Strongly habitat-
related endpoints will be identified as those which can be reliably (a = 0.05, 1-p > 0.80)
detected as statistically different among habitats using the optimum sampling protocol
(see "Optimum Sampling Protocols" section, below). The relative sensitivity of the
endpoints to habitat type will be determined by power analysis (Ferraro et al. 1989,
1994).
Objectives: The primary objective of this research project is to determine biota
(macroinfauna, megainfauna, megaepifauna)-habitat relationships for major PNW
estuarine habitats. Other complimentary objectives are to determine optimum sampling
protocols and designs for determining biota-habitat relationships, and to identify cost-
effective, habitat-associated indicators of estuarine productive capacity, ecosystem
services values, and integrity.
Scientific Approach:
Optimum Sampling Protocols: The sampling objective of this study is to obtain
sufficiently accurate estuary-wide ("universe") estimates of the macroinfauna,
megainfauna, and megaepifauna in major PNW estuary habitats for ecorisk
assessments and futures analysis. Many different types of sampling gear and
-------
Figure 1. Chart of Willapa Bay showing intertidal habitats (based on best
available information) and random station locations within habitats. Due to
inaccuracies in habitat maps, some stations appear to be located outside habitats.
Willapa Bay Estuary -1996 Benthic Survey
9
Field Sample Points
Zostera Habitat
Upogebia Habitat
Neotrypaea Habitat
Spartina Habitat
Mudflats
1 0 1 2 3 4 Kilometers
-------
-------
Figure 2. Chart of Yaquina Bay showing Neotrypaea and Upogebia habitats (based on
best available information) and random station locations within habitats. Due to
inaccuracies in habitat maps, some stations appear to be located outside habitats.
1996 Yaquina Estuary Berrthic Survey
0.5 0 0.5 1 Kilometers
-0- Neotrypaea Sample Station
Upogebia Sample Station
Neotrypea and Upogebia Habitat
Neotrypaea
f~1 Upogebia
^ 17"
•owftunan
-------
-------
34
protocols have been used to sample different target species in different estuarine
habitats (e.g., Puget Sound Estuary Program 1987, 1990; Simenstad et al. 1989, 1991;
Rozas and Minello 1997). Since large-scale field studies of the type necessary to
determine estuary-wide biota-habitat associations are likely to be costly, and the
optimum (i.e., most cost-effective) sampling protocol is unknown, we will, as part of this
study, test and compare a variety of alternative sampling protocols using Ferraro et
al.'s (1989) power-cost efficiency approach. The optimal sampling protocol will be
determined as the least costly sampling protocol capable of reliably detecting
significant differences (a = 0.05, 1-3 * 0.80) among habitats on most of the
measurement and assessment endpoints of interest (Ferraro et al. 1989, 1994).
Standard statistical methods (e.g., Krebs 1989; p. 213-223) will be used to determine
the optimal allocation of samples within habitats. By identifying and then implementing
the optimum sampling protocol and design early in the study, we will maximize the
probability of meeting our research objectives and minimize cost (Green 1979; Ferraro
et al. 1989, 1994).
Details of our sampling and statistical design for determining macroinfauna-habitat
relationships and the optimum macroinfauna sampling protocol for detecting
differences among PNW estuary habitats are in Ferraro and Cole (1996a,b) and Boese
et al. (1996). In brief, benthic habitat maps will be drawn using best available
information (published charts, written reports, scientific papers, personal interviews with
experts familiar with the estuary, aerial photographs, etc.). As the benthic habitat maps
may be inaccurate, prior to sampling a sufficiently large number (~40) of random station
locations will be identified within each of the habitat map areas using the EMAP
protocol. Macroinfaunal samples will be collected at the first -10 (in 1996) or -15 (in
1998 and thereafter) random stations correctly identified with respect to habitat. The
difference in number of samples per habitat in 1996 and 1998 and thereafter is a result
of a preliminary determination of the optimum sampling protocol capable of reliably
detecting differences in the number of species and numerical abundance among
habitats in Willapa Bay in 1996 (Ferraro and Cole 1998; QAPP98.05 Estuary Biota-
Habitat Relationships, revised 8/11/98).
In August-September 1996 nine or ten stratified random samples were taken in each of
four habitats in Willapa Bay (Spartina, Zostera, Neotrypaea, Upogebia) (Figure 1) and
two habitats in Yaquina Bay (Neotrypaea, Upogebia) (Figure 2). The samples collected
at each station were divided into 28 subsamples [six 8-cm internal diameter (0.005 m2)
cores x 2 depth strata (0-5 cm and 5-10 cm) x 2 sieve mesh sizes (1.0 mm and 0.5 mm)
= 24 + 4 (2 cores x 2 depth strata x 1 mesh size (0.25 mm) = 28], The original plan was
to sequentially process (processing includes sorting, taxonomic identification,
enumeration, and quality assurance) the subsamples (core 1 from each station, then
core 2 from each station, etc.) and to continue processing the subsamples until
compositing subsamples into progressively larger sample units no longer appreciably
increased the statistical power to distinguish differences among habitats on the majority
-------
35
of the endpoints of interest. In light of preliminary data showing deeper (5-10 cm)
strata had far fewer individuals and few new species (Table 1), and in response to a
management decision to place new discrete time and resource limits on ongoing
research, sample processing was limited to four (of the six) 0-5 cm deep cores, 1.0 and
0.5 mm mesh samples and one (of the two) 0-5 cm deep core, 0.25 mm mesh samples
per station. Our future macroinfaunal sampling protocol and target endpoints will be
determined based on the results of this currently ongoing investigation (see preliminary
results in the "Expected Results and Benefits" section, below). To the extent possible,
the timing and location of the field sampling of macroinfauna, megainfauna and
megaepifauna components of this study will be same.
DeWitt (1998) is conducting a pilot study to determine the effectiveness and relative
cost efficiency of three megainfauna (>3 mm) sampling protocols, including two core
sizes and 2 extraction and processing methods for 40-cm diameter x 80-cm deep cores,
in burrowing shrimp habitat in Yaquina Bay. CEB staff will also test a suction "flushing-
coring" sampler (van Arkel and Mulder 1975; Grussendorf 1981). The sampling gear
and protocol determined as best in CEB pilot studies will be used in our project to
determine megainfauna-habitat relationships.
For our investigations of megaepifauna-habitat relationships, we will use a drop
sampler similar to that used by Zimmerman et al. (1984) if, after testing, it proves to be
an effective quantitative sampler of the primary target species in the intertidal habitats
of our investigation. A drop sampler is being considered first because it was the only
sampling gear recommended by Rozas and Minello (1997) for quantitative sampling in
all major shallow-water estuarine habitats and because Simenstad (pers. comm.)
recommended it for our study. Other possible samplers are a beach seine and a 3-m
staffbeam trawl (Gunderson and Ellis 1986). The primary target megaepifaunal species
will be juvenile Dungeness crabs (Cancer magister) and English sole (Parophrys
vetulus), as they are valued species which rely heavily on estuaries as nursery areas
and they appear to have specific estuarine habitat preferences (Bayer 1979, 1981;
Hedgepeth and Obreski 1981; Gunderson et al. 1990).
Macroinfauna-Biota Habitat Relationships: The presence and strength of macroinfauna-
habitat relationships will be determined by tests of differences on endpoints of interest
among estuarine habitats (see "Goals" section). In August-September 1996,
macrofaunal samples were collected at ten random stations from four habitats
(Spartina, Zostera, Neotrypaea, Upogebia) in Willapa Bay and two habitats
(Neotrypaea and Upogebia) in Yaquina Bay (Figures 1 and 2) (Ferraro and Cole
1996a,b). In a similar manner, we will periodically (Table 2) collect macroinfaunal
samples using the optimum sampling protocol at approximately fifteen random stations
from targeted habitats in Willapa Bay and Yaquina Bay and at least one other PNW
estuary in order to test the ability to extrapolate the relationships over time and across
estuaries. All macroinfauna sampling will occur in the index period July-September to
-------
36
avoid seasonal variability, for consistency with other estuarine research programs (e.g.,
the EMAP-Estuaries program's sampling index period is June-September (U.S. EPA
1992a)), and because most fish and invertebrates, including the primary target
megaepifauna species (Dungeness crab and English sole), are usually most abundant
during that time of year in PNW estuaries (Swartz et al. 1974; Bayer 1979, 1981;
DeBen et al. 1990; Gunderson et al. 1990; Hinton et al. 1992; McCabe et al. 1993).
Sampling more than one index period is impractical due to resource constraints, and
field sampling during the winter/early spring is logistically much more difficult due to
frequent inclement weather and infrequent daylight low tides. Inferences from our
research, therefore, will necessarily be limited to our sampling index period. The
amount and frequency of sampling will ultimately be dictated by available resources.
We anticipate that macroinfaunal sampling will be conducted biannually, at a minimum,
according to the priorities and the tentative schedule in Table 2. Resources permitting,
we will increase the frequency of macroinfaunal sampling from biannually to annually.
Macroinfaunal sampling will be performed in cooperation with another CEB research
project (Lamberson 1998).
Megainfauna-Biota Habitat Relationships: The presence and strength of megainfauna-
habitat relationships will be determined by tests of differences on endpoints of interest
among estuarine habitats (see "Goals" section). The sampling gear and protocol found
to be the most cost-effective in CEB pilot studies will be used. Megainfaunal sampling
will occur during the same times (July-September) and locations as the macroinfauna
sampling according to the priorities and the tentative schedule in Table 3. Resources
permitting, we will increase the frequency of megainfauna sampling from biannually to
annually. Megainfaunal sampling will be performed in cooperation with two other CEB
research projects (Boese 1998; DeWitt 1998).
Megaepifauna-Biota Habitat Relationships: The presence and strength of
megaepifauna-habitat relationships will be determined by tests of differences on
endpoints of interest among estuarine habitats (see "Goals" section). Since the
temporal within-season variability of the target megaepifauna species can be high (De
Ben et al. 1990; Gunderson et al. 1990), megaepifaunal sampling will occur monthly
(July, August, and September) at approximately the same locations as the infauna
sampling and according to the priorities and tentative schedule in Table 4. Resources
permitting, we will increase the frequency of megaepifauna sampling from biannually to
annually. Megaepifaunal sampling will be performed in cooperation with another CEB
research project (Power 1998).
Scientific Merit: Estuaries have high ecosystem services value, providing services
(food production, habitat refugia (e.g., nursery grounds), recreation, etc.) with a
estimated value of at least $22,832/ha/yr (Costanza et al. 1997). As estuaries are
stressed, habitat areas may change resulting in increases or decreases in their
ecosystem services value. Since ecosystem services values are not distributed equally
-------
37
Table 1. Willapa Bay macroinfauna samples, core 1, stations 1-5.
0.5 mm sieve
Individuals
Top
Bottom
5 cm
5 cm
Neotrypaea
91
58
Spartina
2127
115
Upogebia
352
134
Zostera
1253
64
TOTAL
3823
371
Species
Species
not in
Top
Bottom
core top
Neotrypaea
20
13
6
Spartina
36
16
0
Upogebia
44
25
8
Zostera
48
23
4
1.0 mm sieve
I
Individuals
Top
Bottom
5 cm
5 cm
Neotrypaea
29
25
Spartina
278
88
Upogebia
397
60
Zostera
524
338
TOTAL
1228
511
Species
Species
not in
Top
Bottom
core top
Neotrypaea
15
6
4
Spartina
31
13
2
Upogebia
41
23
7
Zostera
50
22
3
-------
38
Table 2. Macroinfauna field sampling completed (+) and planned in priority order by
habitat type. indicates absent or minor habitat.
Macroinfauna
Habitats
Willapa Bay
Yaquina Bay
Other
PNW Bay
Year:
1996
1998
1996
1998
2000
2000
Spartina
+
5
-
-
-
-
Zostera
+
2
1
1
1
Neotrypaea
+
7
+
4
4
3
Upogebia
+
6
+
3
3
2
oyster
1
-
-
-
-
mud
3
2
2
4
subtidal
4
5
5
5
Table 3. Megainfauna field sampling locations in priority order by habitat type,
indicates absent or minor habitat.
Megainfauna
Intertidal
Habitats
Yaquina Bay
Other
PNW Bay
Year:
1998
2000
2000
Spartina
-
-
-
Zostera
1
1
1
Neotrypaea
2
2
2
Upogebia
3
3
3
oyster
-
-
-
mud
4
4
4
-------
39
Table 4. Megaepifauna field sampling locations in priority order by habitat type.
indicates absent or minor habitat.
Megaepifauna
Intertidal
Habitats
Yaquina Bay
Other
PNW Bay
Year:
1998
2000
2000
Spartina
-
-
-
Zostera
1
1
1
Neotrypaea
3
3
3
Upogebia
2
2
2
oyster
-
-
-
mud
4
4
4
Table 5. Mean (SE) numerical abundance and mean (SE) number of macroinfauna per
0.005 m2 in four habitats in Willapa Bay. Analyses below are based on collections
using a 0.005 m2 x 5 cm deep core sample unit, ^0.5 mm animal size fraction, and n = 5
samples per habitat.
Mean (SE) Abundance
Mean (SE) Number Species
Spartina (S)
480.4 (82.8)
17.8 (2.32)
Zostera (Z)
342.4 (207.6)
26.8 (3.09)
Upogebia (U)
149.4 (116.4)
19.8 (2.60)
Neotrypaea (N)
21.4 (22.8)
7.6 (2.31)
Mean Square Error
22868
33.7
Tukey's test:
S> N
Z> N
(p < 0.05)
S>U
U> N
Z> N
-------
40
among habitats, valuations on habitats within estuaries are needed to determine
stressor effects on estuaries.
The primary purpose of this research is to quantify the functional value of habitats
within estuaries so that important effects of stressors on estuaries can be predicted
using the tools of ecorisk assessment and futures analysis (U.S. EPA 1992b; Landis
and Wiegers 1997; Wiegers et al. 1997). This information can then be conveyed to the
public and used by resource managers to make better informed decisions regarding
actions to minimize or eliminate the risks. This research is consistent with EPA/ORD's
high priority and long-term mission goals to conduct "research to improve ecosystem
risk assessment," "to develop scientifically sound approaches to assessing and
characterizing risks to human health and the environment," and "to provide common
sense and cost effective approaches for preventing and managing risks" (U.S. EPA
1996, 1997). This research will also identify optimal sampling protocols and designs
for estuary-wide habitat sampling, determine the sensitivity of measurement and
assessment endpoints in relation to habitat type, and increase basic knowledge in
estuarine ecology.
Expected Results and Benefits: The principal expected results of this research are
empirical models of macroinfauna-, megainfauna-, and megaepifauna-habitat
relationships for major PNW estuary habitats. Since neither sufficient data nor models
currently exist to conduct habitat-based ecorisk analyses or futures analyses, these
models will significantly improve our ability to perform ecorisk assessments and futures
analyses in PNW estuaries. The expected primary beneficiary is the public, on whose
behalf better informed decisions can be made to protect the environment.
Optimum sampling protocols and designs will be identified for determining biota-habitat
relationships in PNW estuaries, thus maximizing the probability of the success of
determining biota-habitat relationships and minimizing the cost. Preliminary results
suggest that useful macroinfauna-habitat relationships may be established among the
four habitats investigated in Willapa Bay with a smaller sampling effort than is typically
used in macroinfaunal studies (Table 5). Additional analyses will be performed with
sample units of up to four composited cores (total surface area = 0.02 m2), ^1.0 mm
versus *0.5 mm animal size fraction, and n = 10 to determine the overall optimal
sampling protocol and design for determining macroinfauna-habitat relationships for a
suite of measurement and assessment endpoints of interest.
This research will also increase our basic scientific knowledge and understanding of
estuarine ecology, especially the spatial and temporal distribution of estuarine
organisms and their habitat preferences.
-------
41
Project A2 - Dynamics of an Estuarine Landscape: Spatial and Temporal Patterns
with Regard to Coastal Shoreland Development
Principal Investigator: Bradley D. Robbins
Goals: The goal of the proposed research is to quantify the spatial organization (i.e.
heterogeneity) and the temporal dynamics of the habitats (elements) that define the
Yaquina Bay landscape and to evaluate historic and current impacts on this landscape
by anthropogenic stressors. Because the landscape mosaic is formed and maintained
by processes acting across a suite of spatial and temporal scales, information on
landscape heterogeneity and the spatial and/or temporal distribution of biological
patterns provide a new perspective on the dynamics of habitat change and the impact
this change may have on associated landscape elements (Forman and Godron 1984).
Investigations aimed at the detection and quantification of patterns at the landscape
level of analysis (i.e. coastal change detection analysis; C-CAP; Dobson et al. 1995)
may provide insight into processes that influence these patterns which themselves
appear only at larger levels of spatial aggregations. Principles drawn from terrestrial
landscape ecology (see Robbins and Bell 1994) will be used to analyze and interpret
my data because of the expected interdependence of landscape elements.
Specifically, this proposed research will explore how the heterogeneous combination of
landscape elements is structured (space), how it functions (interactions), and how it
changes (time).
Rationale: The research proposed here is intended to address ORD's primary goal to
develop scientifically sound approaches to assess and characterize risks to the
environment (EPA 1996). Specifically, this research will use a large-scale or a
landscape ecology-based perspective to assess the risks of anthropogenic (non-
chemical and/or chemical) stressors to ecosystems within the Yaquina Bay landscape
introduced from adjacent shoreland development. Because natural systems are
hierarchically structured and thus are inherently complex (Allen and Starr 1982), the
recognition of the interdependence of ecosystems and the evaluation of environmental
risk across multiple scales is important. Specifically, biological patterns seen at one
scale may control or be controlled by factors operating at another larger or smaller
scale. However, identification of the appropriate scale of inquiry for environmental
studies is often difficult and is necessarily related to the target organism and/or habitat
(Weins 1989; Robbins and Bell 1994). Alternatively, the ability to extrapolate results
across spatial scales may not be possible. For example, McNeill and Fairweather
(1993) found that the smaller patches supported significantly more species offish and
macroinvertebrates than larger patches because the increased perimeterarea ratio of
smaller patches allowed greater recruitment to occur. Others examining the influence
of seagrass patch size on the accumulation of drift algae (Bell et al. 1995), the growth
and survivorship of infaunal bivalves (Irlandi 1996, 1997) and patch use by hermit
crabs and gastropods (Robbins 1998) also found that directly scaling from a small area
-------
42
(1m2) to a large area (4m2) was not possible. These data suggest that extrapolation
from the small to the large scale may lead to an erroneous interpretation. These
difficulties may be alleviated by taking a multi-scale approach to ecosystem risk
assessment and management. For example, seagrass-dominated landscapes can be
viewed as an hierarchical arrangement of floral patches (i.e. seagrass and attached
macroalgae) embedded in a matrix of unvegetated sediment (i.e. sand and/or mud)
(Robbins and Bell 1994; Robbins 1998). The floral patches, in turn, can be considered
hierarchically from individual short shoots to kilometer-wide areas of contiguous cover
(Robbins and Bell 1994). Knowledge of the shoot to km patterns of seagrass
distribution allows questions to be asked such as what is the impact of a process which
occurs at one scale on a pattern seen at another? Specifically, does the propagation of
disturbance effects (stressors) depend upon the characteristics {i.e. patch distribution,
size, shape, or type) of the landscape?
In contrast to efforts directed toward quantifying spatial patterns in terrestrial habitats,
analogous information on intertidal/subtidal soft bottom marine habitats is limited
(Robbins and Bell 1994). However, the available results from estuarine studies have
demonstrated the feasibility of large scale (km) studies of seagrass patches and their
associated fauna (e.g. Ferguson and Korfmacher 1997; Lehmann et al. 1997; Malthus
and George 1997; Norris et al. 1997; Robbins 1997,1998; Schmieder 1997; Ward et al.
1997; Williams and Lyon 1997). These studies show how trophic organization, as
evidenced by fish feeding guilds or the linkage between macroinvertebrates and their
resources, varied with landscape features. Using the underlying principles of
landscape ecology, questions concerning processes which influence ecological spatial
heterogeneity can be addressed in spatially explicit terms (Weins et al. 1993) which will
contribute to our overall understanding of the functional aspects of the landscape.
Presently, it is not well known how factors responsible for creating estuarine landscape
patterns may have changed over time, especially in areas where anthropogenic effects
have not been documented. Therefore, utilizing an approach to measure spatial and
temporal patterns of the landscape, as well as the processes that produce these
patterns, may help identify sites that are similar in ecosystem characteristics beyond
the mere presence or absence of the target habitat and/or organism or interest.
The purpose of this study is to develop an understanding of the structure and function
of a PNW estuarine landscape as well as its temporal dynamics (change).
Anthropogenic activities along the estuary's adjacent shorelands have altered the
structure of the estuary and subsequently, its function. Thus, an understanding of the
coastal shoreland development over time will result in a better conceptual picture of the
estuary. My initial efforts must necessarily be directed to the development of a spatial
representation of a PNW estuary (i.e. Yaquina Bay). It is my intent to document the
number, size, and location of the Yaquina Bay landscape's structural elements across
the subtidal, intertidal, and coastal shorelands (defined as boundary lands within
1000m of MLLW). Because the tempo of habitat change, and the speed of associated
-------
43
processes whether subtle or complex (Franklin 1989) are critical to the understanding
of landscape dynamics and function I am proposing the use of historical aerial
photographs to extend the temporal component of this study. Temporal studies not
only allow for comparisons to be made across time but also provide insight into the
placement of appropriate field experiments and/or sampling sites for measuring and
evaluating ecosystem function.
Yaquina Bay, the fifth largest estuary in Oregon, encompasses an area of
approximately 1582ha with 35% intertidal and 65% subtidal (Oregon Estuaries 1973).
Yaquina Bay has been classified as a deep developed estuary (Estuary Plan Book)
with the city of Newport (population -9500) situated at its mouth and the city of Toledo
(population -3500) located nine miles upriver. Shorelands immediately adjacent to the
estuary can be classified as either urban-developed, rural-developed (agriculture), or
rural-undeveloped (forested) following protocols developed by NOAA's C-CAP (Dobson
et al. 1995). Within the Yaquina Bay estuary, seagrass patches are arranged either as
fringing patches along the lower (Zostera marina) or upper (Z japonica) intertidal
margins or as large contiguous or small highly convoluted patches (Z. marina) found
both intertidally and subtidally across the mud/sandflats. A second dominant
landscape element within Yaquina Bay are extensive aggregations of burrowing shrimp
(Neotypaea californiensis and Upogebia pugettensis). These two element types offer
an interesting dichotomy in that they may be spatial competitors while both have
economic and ecological importance. Seagrasses act as sediment traps, stabilizing
sediments and filtering suspended particulates from the water column (Dawes 1981)
and are nursery areas for many commercially important invertebrates and vertebrates.
Alternatively, the shrimp destabilize the substrate by excavating large quantities of
sediment while digging their extensive burrows and thus change the physical and
chemical properties of the sediment (Posey et al. 1991). The interaction between
eelgrass (Zostera marina) and burrowing shrimp (Neotypaea californiensis and
Upogebia pugettensis) within Yaquina Bay may provide an unique system in which to
explore questions of landscape dynamics both spatially and temporally across a suite
of spatial scales. For example, a stress such as increased sediment deposition as the
result of clear cutting along the boundary of the bay may result in an areal decline in
seagrass and subsequent increase in the areal extent of burrowing shrimp.
Alternatively, burrowing shrimp may increase both oxygen and nutrient availability
within sediment porewaters thus enhancing the growth of seagrass. Macroalgae
compose a third element found within the bay. However, because of the ephemeral
nature of macroalgae in Yaquina Bay, it is unclear what role this vegetation may play in
influencing the distribution and spatial organization of seagrass and/or burrowing
shrimp.
The time series maps developed in this study will be used to address how the spatial
arrangement of landscape elements within Yaquina Bay have been impacted by coastal
shoreland development. For example, the dredge and fill activities which have
historically taken place within Yaquina Bay have resulted in the loss of several
-------
44
hectares (~7%) of the intertidal region. The loss of intertidal flats to fill and the
subsequent blockage of natural drainage channels can modify both current patterns
and velocities which may alter sediment erosion and deposition patterns resulting in the
loss of valuable habitat. Alternatively, the construction of mooring areas, docks, and
wing dams within the bay may also have a negative impact on seagrass habitats either
because of shading after a dock has been built or the direct removal of seagrass during
construction (see Simenstad etal. 1997). Thus the obvious question is how has the
areal extent of seagrass changed over time and can this change be correlated with
anthropogenic behaviors (development) along the coastal shorelands?
Objectives:
1: To develop of a detailed depiction of the spatial organization of the Yaquina Bay
marine landscape;
2: To evaluate the impact of both historic and modern coastal shoreland development
on the landscape's spatial heterogeneity;
3: To examine the power of mapped variables for predicting the distribution of a
dominant landscape element within the Yaquina Bay landscape.
Methods: The project will have two phases with the acquisition of historic aerial
photographs and collateral materials (e.g. reports) of the Yaquina Bay estuary
(Objective 1) occurring first. This phase of my proposal will be coordinated with the
research of Dr. David Young (CEB/EPA, Newport, OR). Aerial photographs have been
successfully used to produce historic land-use maps to monitor terrestrial development
(Adeniyi 1980) and should also be suitable for reconstructing the historical distribution
of aquatic habitats and for subsequently monitoring coastal shoreland development
(see Benner 1991; Coulton etal. 1996; Benner and Sedell 1997). Historical
photographs are especially useful in identifying potentially hazardous sites that have
become inactive and/or subsequently developed for other purposes (Evans and Mata
1984) yet still represent an environmental threat (stressor). A description of the spatial
heterogeneity of the Yaquina Bay intertidal/subtidal landscape will be accomplished
using information derived from the interpretation of historic and modern aerial
photography. The length of the time period studied will be dependent upon the
availability and quality of historic aerial photographs. The source media for the historic
maps will most likely be black and white monoscopic aerial photographs taken at a
scale of 1:24000 or greater which may limit the recognition of some landscape element
types (see Robbins 1997). Other potential problems associated with historic aerial
photographs include the ambient environmental conditions (e.g. sun angle, tidal height,
wind speed, etc.) at the time the photographs were taken and the lack of
groundtruthing; in situ sampling is used to check the accuracy of interpretations made
on the basis of sensor data. Archived photographs are available from the City of
Newport, Oregon State University, the Oregon Department of Fish and Wildlife. Other
potential sources of photographs are the USGS, NOAA, Lincoln County DMV, Lincoln
County Planning Office, the University of Oregon, the Oregon Department of
-------
45
Transportation, and the US Army Corp of Engineers.
The source media for the modern aerial photographs will be first generation positive
transparencies (diapositives) from color infra-red (CIR) negatives at a scale of 1:7200.
Modern aerial photographs will be taken based on the following specifications: 1) a
standard calibrated aerial photographic camera with an antivignetting on a 6in lens with
the final product a series of 1:7200 stereoscopic color-infrared diapositives with a 60%
forward lap on 9x9in film; 2) low tide; 3) minimal haze and cloud cover with a water
surface glint of <5%/frame; 4) wind speeds of <5mph, no visible white caps, offshore
winds preferable to onshore winds; 5) water clarity should be high such that seagrass
patch boundaries can be clearly delineated. Although the NOAA Coastal Change
Analysis Program (Dobson et al. 1995) recommends the use of natural color film for the
detection of SAVs this is based on the need to detect subtidal seagrasses and macro
algae. In Yaquina Bay as in many PNW estuaries, the majority of seagrass and
macroalgae is intertidal and thus exposed during summer daylight hours. Thus CIR,
extensively utilized in the detection of terrestrial vegetation, is an appropriate choice for
the detection of exposed vegetation.
All modern aerial photographs will be scanned at 20 to 25micron resolution using a
high precision/ high accuracy laser scanner resulting in digital images with a nominal
image resolution ranging from 0.29m2 to 0.36m2. All scanned images will be
orthorectified using the appropriate digital elevation model (DEM) and entered into the
Branch GIS for digital data interpretation and classification. Landscape attributes
including patch type (i.e. SAV, burrowing shrimp), size, shape, and position within the
landscape will be measured. In situ (groundtruth) samples will be used to assess the
accuracy of each element classification. The number of necessary groundtruth
samples is dependent upon several factors including the number and size of each
category. A general rule of thumb is that enough groundtruth samples need be taken to
cover at least 5% of the overall area and 10% of each category. The estimated number
of groundtruth points will be further modified to remain within logistic sampling
constraints.
The second phase of this project will be further subdivided into two parts. First, I will
evaluate the impact of coastal shoreland development on the Yaquina Bay landscape
(Objective 2) examining two hypotheses:
H,: The spatial organization of landscape elements is not correlated with historic
large- scale anthropogenic disturbances (e.g. dredge/fill operation); and
H2: The development of coastal shorelands along the boundaries of the estuary
is not correlated to a change in the areal extent of SAVs and/or
the addition of upper marshland.
-------
46
Thematic data layers will be created based on a land-use classification scheme
modified from those proposed by C-CAP (Dobson et al. 1995). These data will be
analyzed using a composite mapping analysis technique which entails combining the
separate thematic data layers using their spatial coincidence (Lowery et al. 1995).
Changes in land-use classification with time within a thematic data layer will be
correlated to the spatial adjacency of classifications within other data layers. Several
additional approaches can be utilized to evaluate spatially representative data. I will
employ three analytical techniques: 1) simple pattern comparison between maps
generated at different times; 2) spatial autocorrelation and temporal cross-correlation
analyses; and 3) trend and residual analysis. The first step in my analytical approach
will be a comparison of maps by calculating the difference (isopachs) between maps
representing different time periods. For example, the seagrass map generated from
1997 aerial photographs will be compared to the seagrass map generated from the
1998 aerial photographs by "subtracting" 1997 seagrass areas from the 1998 seagrass
areas which will reveal the annual addition of seagrass. The second step is to measure
neighborly influence, defined as the spatial autocorrelation or the self-similarity of
adjacent data. Analogously, the relationship between spatial points through time can
be examined using cross-correlation. Both of these procedures assume that data
points lying in close proximity (whether in time or space) to one another are more
similar than widely separated data points. Specifically, one can ask whether an area
has changed through time in terms of the elements that define the area and those
areas which are adjacent. Third, if trends are found to be present in the spatial and/or
temporal configuration of the data, these can be removed and relationships can be
examined using residual trend analysis. Differences in the areal extent and
configuration of the seagrass will be compared using parametric (regression and/or
correlation analyses), non-parametric (correspondence analysis), and/or geostatistical
(semivariography) techniques.
The second part of Phase 2 will be to examine the power of mapped variables for
predicting the distribution of a dominant landscape element within Yaquina Bay
(Objective 3).
H3: The distribution of landscape elements can not be predicted by measurable
environmental variables.
To achieve this objective, I will develop a map of predicted seagrass coverage based
on differences in seagrass areal extent between 1997 and 1998 as derived from the
interpretation of aerial photographs projected one year into the future. This predicted
seagrass map will then be compared to observed seagrass aerial extent derived from
1999 aerial photographs. A comparison of predicted and observed maps will result in a
residual map which will be interpreted using thematic layers of environmental variables
which should have some influence on seagrass distribution. These thematic maps will
be derived from data collected by other EPA scientists at the Newport facility:
-------
47
burrowing shrimp (Dr. Ted DeWitt), water quality (Dr. Bob Ozretich), suspended and
bottom sediments, bathymetry (Dr. David Young), and nutrient availability (Dr. Ann
Sigleo). In essence I will answer the following questions:
1) What is the total area and patch size distribution of observed seagrass habitat
in 1997 and 1998?
2) What is the total area and patch size distribution of seagrass habitat predicted
on observed differences in seagrass habitat from 1997 to 1998?
3) How are areas where predicted and observed maps disagree distributed with
respect to other landscape elements (e.g. burrowing shrimp and/or
macroalgae), water quality, differences in sediments attributes, and nutrient
availability?
The correspondence between maps will be measured by testing for the non-random
distribution of residuals using spatial measures of contiguity or spatial autocorrelation
(Cliff and Ord 1981) or by developing a contingency or correlation matrix without regard
for spatial position (Phipps 1981).
An alternative to the use of an aerial platform to remotely sense intertidal and subtidal
landscape elements will be the use of videography. While this technology is
dependent upon water clarity and thus may not be suitable for use in Yaquina Bay, the
technique has been used in other estuaries and may prove feasible here. The method
is as follows. Video data will be collected using an 8mm camcorder housed in an
underwater video housing mounted on a modified roller trawl or a flying sled from six
100x100m areas located haphazardly in Sally's Bend, Idaho Flats, and Raccoon Flats.
The system will be towed behind a small skiff at a fixed rate of speed along
predetermined transects. Also attached to the system will be a depth transducer and a
DGPS. Depth data will be collected using a commercial data logger and a laptop
computer connected via a RS-232 cable, at a rate of once every 0.3sec. DGPS
readings will be recorded on the video every 0.5m. This research will be conducted in
cooperation with CEB scientist Dr. DeWitt and compared to similar data collected using
a 420-KHz hydroacoustic echolocation system operated by Dr. Bruce Sabol (USACE
Waterways Experiment Station) a collaborator on DeWitt's proposed research.
Comparisons between the methods will be made by directly comparing the
hydroacoustic signature to the video recording by transect and scoring the percentage
of agreements. Ground truthing will be conducted at all six sites following the protocol
outlined in Appendix A.
Expected Results/benefits: The proposed research will quantify the spatial
heterogeneity of an estuarine landscape and document its temporal dynamics. Change
detection analysis is a useful tool in detecting and evaluating areas particularly
vulnerable to environmental risk. Once these areas have been identified, they can then
be targeted for experimental manipulations or as areas requiring restoration and/or
-------
48
mitigation. This research is unique in that few if any studies exist which document the
spatial and temporal dynamics of a marine estuary at this scale of resolution.
An assessment of ecological risks to the landscape via anthropogenic stressors will
also be made and addressed. Specifically, the impact of historical and modern coastal
shoreland development to the abundance and distribution the major landscape
elements within Yaquina Bay will be evaluated. These data will be useful for the
development of future risk management plans. The model developed by this research
will be a useful tool in the recognition and management of future environmental risks
and will allow predictions concerning the distribution of the landscape's dominant
elements based on several environmental variables such as water quality, nutrient
availability, and sediment chemistry, organization, and dynamics to be made.
APPENDIX A: IN SITU MAPPING PROTOCOL
Six sites (100x100m) will be mapped by a 4 member team of researchers working in
pairs. Each pair will consist of a "caller" and a "recorder." The caller's job is to identify
the dominant element(s) along each transect. The recorder maintains a written record
on provided data sheets by crossing out the appropriate code signifying the presence
of an element. Recorders are also responsible for recording Quadrat (site) and
Transect number, Date and Time, and the names of both the caller and recorder. The
team works in concert with each pair mapping half of the site (50x100m area).
Specifically, 3150 data points will be collected by each team from each site. Collected
data will consist of identifying the presence of five elements: seagrass, Zostera sp.(G),
green macroalgae (Ag), shrimp (S), bare substrate (B), and brown/red algae (Ab) for all
sites. In addition, at Site 6 the shrimp classification will be further subdivided into two
categories by shrimp genera: Upogebia (Su) and Neotrypaea (Sn).
Data will be collected across each site by walking 21 transects positioned at five meter
intervals. Each transect will be 3m wide and delineated by a 100m polypropylene rope
marked at 1m intervals. The rope will be initially positioned at the first transect (A-D).
One caller will begin at their end of the transect (corner A) and will walk 50m to the
middle of the transect (toward corner D) while the second caller will begin at the 50m
mark and walk to their 0m mark (corner D). A recorder will be at each end of the
transect line. Data will be collected every 2m in the direction of the transect and every
1 m across the width of the transect. When each caller completes their 50m walk, the
transect rope will be moved 5m. The presence of a caller in the middle of the transect
will aid in the movement of the transect line. After positioning the rope along the next
transect the callers will again walk 50m stopping every 2m for data collection. At the
end of this transect the callers will be in their original positions on the transect line.
This procedure will be repeated until each site is mapped.
-------
49
Project A3 - Changes in The Abundance And Distribution of Estuarine Keystone
Species in Response to Multiple Abiotic Stressors
Principal Investigator: Ted DeWitt
Coprincipal Investigators: Mary Kentula, Bob Ozretich, Brad Robbins, David Young
Goals:The goal of the proposed research project is to assess the ecological risks
posed by multiple abiotic stressors to eelgrass and burrowing shrimp, which are
ecologically- and economically-important, habitat-creating, keystone species living in
Pacific coast estuaries. The proposed research will use field and mesocosm
experiments to measure the effects of abiotic stressors on the population dynamics of
eelgrass and burrowing shrimp and to measure the effects of stressors on the outcome
of competition among these keystone species.
Rationale: Eelgrass (Zostera marina and Z. japonica) and burrowing shrimp
(particularly Neotrypaea californiensis and Upogebia pugettensis) create or extensively
modify large areas of intertidal and subtidal benthic habitat within Pacific estuaries.
These species occur in estuaries from Alaska to Baja California where they are often
the ecologically dominant benthic species. They are keystone species because they
determine the structure of associated benthic communities (Thorn 1987, DeWitt et al.
1997). Eelgrass beds are habitats for resident species offish and invertebrates,
provide temporary refuge and foraging habitat for migratory fish (e.g., salmonids), and
provide feeding habitats for waterfowl and migratory shorebirds (Griffin 1997). The
economic value of eelgrass habitats in Pacific Northwest estuaries includes providing
habitat for salmonids, herring, and Dungeness crabs (Thorn 1987, (Doty et al. 1990,
Berkeley 1998, Simenstad and Fresh 1995).
Estuarine sandflats and mudflats often harbor high densities of burrowing shrimp.
These crustaceans dig extensive galleries below the sediment surface, excavating
large quantities of sediment, thereby changing the physical and chemical properties of
the sediment and the stability of the substrate. Sessile species are often excluded from
burrowing shrimp beds, whereas abundances of mobile species can be enhanced
(Posey 1986a, Posey et al. 1991, Brooks 1995). Densities of epibenthic megafauna,
such as crabs, are reduced over burrowing shrimp beds because bioturbation by the
shrimp buries or kills objects that provide physical structure for cover or foraging habitat
(i.e., eelgrass, oysters) (Armstrong et al 1989, Doty et al. 1990). As compared to
eelgrass or oyster-bed habitats, burrowing shrimp habitat harbors much lower diversity
of benthic invertebrates (Posey et al. 1991, Brooks 1995). Modification of benthic
habitats by burrowing shrimp has direct negative economic impact to commercial oyster
mariculture in the PNW (DeWitt et al. 1997, Griffin 1997). The economic damage to
oyster culture is so great that since the 1950's growers apply the pesticide, carbaryl, to
over 320 ha of intertidal, estuarine oyster beds in Willapa Bay and Grays Harbor (WA)
(DeWitt et al. 1997).
-------
50
Eelgrass beds and burrowing shrimp beds affect the flux and storage of nutrients and
carbon within estuaries (Phillips 1984). Zostera and associated epiphytes take up
nutrients from the water column, and release it seasonally in the form of dead leaves
and blades. As zones of hydrodynamic damping (i.e., depositional areas), eelgrass
beds may also trap particulate organic matter from the water column which then
accumulates within the bed (Fonseca et al. 1982). This can have important
consequences for the deposition and spatial distribution of persistent hydrophobic or
particulate-bound contaminants (e.g., metals, PAHs, PCBs, etc.) bound to sediments
within the estuary. Conversely, bioturbation and burrow irrigation by burrowing shrimp
may result in the flux of nutrients and other chemicals (including persistent
contaminants) from the sediment into the water column. Resuspension of sediments by
the actively burrowing Neotrypea leads to the winnowing of fine-grained sediments from
tide flats, thereby changing the textural properties of the sediments as well as the
concentration of organic carbon (DeWitt et al. 1997). Mud shrimp (Upogebia) pump
large volumes of water through their burrows and feed on suspended particulate
matter; large populations of mud shrimp may be significant consumers of planktonic
primary production and thus possibly affect the productivity of other suspension feeders
within the estuary.
Relatively large annual fluctuations in the distribution and abundance of eelgrass and
burrowing shrimp have been observed in outer-coast estuaries of Oregon and
Washington over the last 20-50 years (Thorn and Hallum 1990; Ellis 1997; Griffin
1997). These variations are unusual as compared to other areas such as Puget Sound
(Thorn and Hallum 1990). In some cases, long-term reduction (e.g. 10+ yr) in the
distribution and abundance of eelgrass beds has been recorded (Ellis 1997; Griffin
1997), whereas the distribution and abundance of burrowing shrimp (particularly
Neotrypea) is reported to have increased within several estuaries (DeWitt et al. 1997).
In Yaquina Bay eelgrass beds, mapped in the mesohaline portions of Yaquina estuary
in 1968 and 1979 (FWS 1968, Bayer 1979), have disappeared. Proposed causes
include regional environmental changes (i.e., El Nifto, drought cycles, Columbia River
water management) and local anthropogenic disturbance such as dredging,
channelization, shoreline modification, construction of docks, commercial and
recreational shellfish farming and harvesting, sedimentation associated with upland
erosion, flood control and water diversion (Thorn and Hallum 1990, Simenstad and
Fresh 1995, Rumrill and Christy 1996). An important scientific need is a greater
understanding of how abiotic stressors affect the population dynamics of these
keystone species. Research is also needed to identify and rank the abiotic stressors
that affect eelgrass and burrowing shrimp populations, in order to manage water quality
for enhanced resource value or to classify sites by their suitability for habitat restoration
(Thorn 1990a).
In addition to abiotic stressors, interspecific interactions undoubtably affects the
distribution and abundance of eelgrass and burrowing shrimp, which are limited in
-------
spatial distribution to the intertidal and shallow subtidal by light availability (eelgrass;
Phillips 1984) and predation (burrowing shrimp; Posey 1986b). Competition for space
potentially occurs between eelgrass and burrowing shrimp, and large-scale changes in
the distribution and abundances of these species may have been caused by
competition (Dumbauld 1998, Brooks 1997, Thorn 1997, Bennett 1997, Wilson 1996).
Burrowing by these shrimps may be responsible for substantial reduction in the
abundance of eelgrass in PNW estuaries (Swinbanks and Luternauer 1987, Dumbauld
et al. 1997), and such effects occur in other marine ecosystems (Suchanek 1983,
Woods and Shiel 1997). Eelgrass populations in Willapa Bay, WA, often expand into
intertidal oyster beds following the use of pesticide to control burrowing shrimp
(Dumbauld 1998, Dumbauld etal. 1997, Bennet 1997, Wilson 1996, DeWitt et al.
1997). Potential interaction mechanisms are sediment resuspension which reduces
light levels and thereby eelgrass growth, and depletion of sediment nutrients by burrow
irrigation. Conversely, Harrison (1987) showed that eelgrass may out-compete
burrowing shrimp following an increase of water clarity, possibly because the increased
water clarity allowed eelgrass to overgrow Neotrypea burrows before the shrimp
became active. It is also known that Z. marina rhizome mats can impede burrowing by
Neotrypea and Upogebia (Brenchley (1982).
Competition is most evident between Zostera and Neotrypea, and although their spatial
distributions are largely disjunct, populations can abut one another (Dumbauld et al.
1997). Zostera and Upogebia populations often overlap, but the nature of interactions
is unknown. Neotrypea and Upogebia populations are largely disjunct, but do overlap
along their common borders. Aggressive interactions occur between individuals that
inadvertently come to share the same burrow (personal observation). There may be
ecologically important three-way interactions may change under different environmental
conditions. If Upogebia has no effect on Zostera, but competes for space with
Neotrypea, then Upogebia may be beneficial to Zostera by suppressing Neotrypea.
One of the scientific contributions of the proposed research will be to provide a better
understanding of the mechanisms and dynamics of competition between Zostera and
burrowing shrimp in Pacific estuaries.
Laboratory and field experiments will be conducted to determine whether these species
do in fact compete for space, and to identify the mechanisms by which competitive
displacement occurs. Subsequent experiments will focus on whether natural and
anthropogenic stressors, particularly those associated with water quality, affect the
outcome of competition. Given that changes in water quality or climate may have
contributed to changes in the distribution and abundance of eelgrass and burrowing
shrimp (discussed above), it is important to know whether natural or human-caused
changes in the abiotic environment can influence the outcome of competition among
these species. Changes in water quality, reduction in the abundance of predators or
herbivores, increase in physical disturbance, or combinations of these stressors can
provide one species with a competitive advantage over the other. These interactions
-------
52
are not weli understood, and can not yet be incorporated into estuarine management
practices.
Light and salinity appear to be the dominant abiotic factors affecting eelgrass and
burrowing shrimp populations in PNW estuaries (Phillips 1984, Thorn 1990, Bird 1982,
Posey 1987). Light probably controls the spatial distribution of Z. marina populations
(Olson et al. 1997), and the light penetration regime in PNW estuaries may have
changed historically due to increased input of fine suspended particulates (associated
with erosion) or phytoplankton abundance (associated with increased input of nutrients,
especially nitrogen) (Thorn and Hallum 1990). Light stressor studies will be
coordinated with Dr. Robert Ozretich (CEB) who is testing the efficacy of water quality
(including light) based models to predict the performance of Z. marina populations.
Low salinity is strongly suspected to limit the up-estuary distribution of eelgrass and
burrowing shrimp (Phillips 1984, Posey 1987) and episodic freshets may affect
population distribution in the lower estuary too. The salinity dynamics of Pacific
Northwest estuaries are likely being altered because of long-term changes in rainfall,
increased runoff due to land use alterations, and water diversion. Other abiotic
stressors will be studied in future experiments as their relative importance is
determined.
Objectives: The goal of the proposed research is to assess the ecological risks posed
by multiple abiotic stressors to eelgrass and burrowing shrimp. This will be
accomplished through:
1. Field experiments to evaluate coring and remote sensing methods to detect,
identify, quantify, and map eelgrass and burrowing shrimp;
2. Laboratory and field experiments to identify, measure, and map the population-
level effects of single and multiple abiotic stressors on eelgrass and burrowing
shrimp;
3. Laboratory and field experiments to measure competition among eelgrass and
burrowing shrimp species, and to measure how abiotic stressors affect the
outcome of that competition;
Methods: The research questions that will be addressed are:
1) What are efficient and accurate methods to sample and map the abundances and
distributions of eelgrass and burrowing shrimp?;
2) Which population-level characteristics of eelgrass and burrowing shrimp show the
strongest and least variable responses to abiotic stressors?;
3) Does competition between eelgrass and burrowing shrimp influence the abundance
or spatial distribution of these species?; and
4) What are the most important stressors affecting the abundance and distribution of
eelgrass and burrowing shrimp in Pacific Northwest estuaries?
-------
1. Population Sampling Methods Evaluation: Experiment 1.A. Comparison of Coring
Methods to Sample Burrowing Shrimp: A limited-effort pilot study will be conducted to
compare at least four methods for collecting quantitative core samples for burrowing
shrimp. These methods are 1) clam gun cores (15-cm diameter x 60-cm length, which
are extracted by manually lifting the corer and sample from the sediment), 2) manually
excavated megacorer (40-cm diameter x 80-cm length, in which sediments are
manually dug out of the core barrel), 3) hydraulically flushed megacorer (40-cm
diameter x 80-cm length, in which sediments flushed from the corer using high pressure
water jet, with the water, sediment, and animals directed through a spout and through a
screen to collect the sample), and 4) a suction sampler (15-cm to 20-cm diameter x
100-cm length, excavated by venturi-action caused by high pressure water jet;
Grussendorf 1981). Each corer will be used to collect replicate samples of Neotrypea
and Upogebia within intertidal habitats. Corers will be compared on the basis of the
among-sample variability of shrimp abundance and their ease of use. The mega-
infaunal sampling method selected in this study will be used in many of the field
experiments described below, and by other researchers at CEB (i.e., Boese, Cole,
Ferraro, and Robbins).
Additionally, we will evaluate the accuracy and precision of burrow-hole density as a
means of measuring the abundance of Neotrypaea and Upogebia. Burrow-hole density
is used widely as a simple measure of shrimp abundance, but is subject to inaccuracy .
due to incorrect species identification, seasonal variation, and habitat variation (i.e,
grain size, current strength) (DeWitt et al. 1997). Burrow-hole density is determined by
counts of burrow openings per unit area (typically a 0.25m2, 0.5m2, or 1,0m2 quadrat)
(Dumbauld 1994). The relationship between burrow-hole density and shrimp density
has only been determined for Willapa Bay estuary (Dumbauld 1994). We will
characterize this relationship for Yaquina estuary by measuring burrow-hole density
(using a 0.25m2 quadrat) at all sites where mega-infaunal cores samples are collected;
burrow-hole counts will be made prior to coring. Additionally, temporal and
environmental data will be collected (i.e., date, sediment grain size, overlying water
salinity and temperature, sediment temperature). These data will be stored in a
database which subsequently will be used to statistically evaluate (i.e., using
regression analysis) how species, season, and environmental factors affect the
relationship between burrow-hole density and burrowing shrimp density.
Experiment 1.B. Comparison of Acoustic, Photographic, and Videographic Remote-
Sensing Methods to Identify and Map Estuarine Submerged Aquatic Vegetation: An
experiment will be conducted to compare the accuracy of three remote sensing
techniques (acoustic, photographic, and videographic) in terms of their ability to
represent the spatial distribution of SAV and burrowing shrimp within Yaquina estuary.
The abundance patterns of each target organism, Zostera marina, macroalgae
(including Ulva, Enteromorpha, Chaetomorpha, and Fucus), and burrowing shrimp
burrow openings will be mapped using each of the remote sensing methods and then
-------
54
comparing these maps to ground-truth data. Six study sites 10,000m2 in area {i.e.
100m x 100m) will be selected within the central bay of Yaquina estuary (Figure 1)
based on 1997 aerial photographs. Site selection will be stratified to include areas with
continuous and patchy flora and fauna.
The accuracy of each map produced by the three remote sensing methods will be
compared to the corresponding map generated from ground-truth data by three
methods, across four spatial scales: 1-m2, 10-m2, 100-m2, and 1000-m2. First, the 1-m2
maps will be smoothed by resampling. At each spatial scale, pixels will be reclassified
by the floral or faunal type of greatest proportion. A correlation coefficient will then be
calculated between all common points among maps to estimate the overall
correspondence between map pairs. At each spatial scale, we will test whether there is
agreement between the remotely sensed data and the ground-truth data in terms of
element type within each pixel. Second, difference maps (isopachs) will be constructed
by subtracting one map from another, and will be used to look for small areas of large
deviation between map pairs which may have unduly influenced the correlation
coefficient (Davis 1984). Third, map pairs will be examined using a cross-tabulation
matrix which will be analyzed for similarities in pixel type using the kappa coefficient, k
(Carstensen 1987). This direct comparison of mapping accuracy will give us the ability
to recommend a method of remote sensing best suited for the study of estuarine SAV.
Additional technical details can be found in Appendix A.
Experiment 1.C. Evaluation of Remote Sensing Methods to Detect, Identify, and
Quantify Burrowing Shrimp: This project will compare several acoustic remote sensing
methods and videographic remote sensing for their abilities to detect, identify, and
quantify Neotrypea and Upogebia in intertidal sediments. Preliminary trials (summer-
fall 1998) will be conducted with each method to determine whether any signal
corresponding to the presence or absence of burrowing shrimp can be detected. If any
of the methods show capability to detect the shrimp, then methods-comparison
experiment will be conducted (summer-fall 1999).
Test sites (patches approximately 2-m x 2-m) will be designated within Yaquina estuary
(Figure 1), pre-classified by ground survey into six patch types: 1) uninhabited sand
{Neotrypaea habitat), 2) low density Neotrypaea (<100 holes m"2), 3) high density
Neotrypaea (>300 holes m"2), 4) uninhabited muddy-sand (Upogebia habitat), 5) low
density Upogebia (<100 holes m*2), and 6) high density Upogebia (>300 holes m2).
Minimally, eight sites of each type will be identified; three sites will be used for
calibration of the classification method, and the remaining sites will be used for testing
the accuracy of the classification. The sites will also be selected to represent a wide
variety of sediment characteristics (i.e., sediment grain size, organic matter content,
water content, and compaction). Field staff will pre-classify sites at low tide by
identifying the shrimp (i.e., by the configuration of the burrow openings; qualitative core
samples can also be collected to verify the species identity of shrimp) and by
-------
measuring the density of burrow openings (i.e., number of burrow holes per 0.25-m2).
Crew operating the remote sensing methods will be blind to identity of each site. The
boat pilot will maneuver and anchor the boat such that the acoustic transponder or
video camera is positioned directly over a site, and the investigator will be allowed to
sample the site for a fixed of period time (nominally, two minutes).
After all sites have been sampled, the classification of the "calibration" sites will be
revealed to each investigator. Each investigator will then analyze data they collected,
classify the remaining "unidentified" sites, and report the results (along with raw data
and the rationale for making the classification). After the investigators report their
findings, the correct classification of the unidentified sites will be revealed. The
accuracy of classifying burrowing shrimp patches by each remote sensing method will
be assessed by chi-square analysis. Each investigator also will be asked to evaluate
how their system could be modified to improve its accuracy, to discuss any logistical or
analytical difficulties that were encountered, and to assess the feasibility and cost of
adapting their remote sensing method to a geographically-accurate system for mapping
burrowing shrimp. Additional technical details can be found in Appendix A.
Experiment 1.D. Evaluation of Remote-Sensing Methods to Map Burrowing Shrimp
Populations: If experiment 1.C. is successful, a study will be conducted (summer-fall
2000) to determine the accuracy of maps of burrowing shrimp populations generated by
one or more remote sensing methods. Selection of the remote sensing method(s) used
in this study will be based on the results of experiment 1 .C. The specific experimental
design and methods will be similar to those used in experiment 1.B. First, we will
compare the accuracy of the remote sensing methods. Test sites will be designated
throughout estuary, encompassing a variety of shrimp population densities, and maps
of each site will be prepared using each remote sensing method and by ground-truth
surveys. Mapping accuracy of each method will be evaluated using the methods of
experiment 1.B. Second, we will evaluate whether seasonal changes in shrimp
burrowing activity affects the accuracy of the maps. Shrimp beds will be sampled and
mapped by remote sensing and ground survey in summer (when burrowing activity is
maximum) and in winter (when burrowing is minimum, and Neotrypea burrow openings
can be filled with sediment), and map accuracy will be determined as above. And third,
shrimp populations inside and adjacent to eelgrass beds will be included to evaluate
whether SAV affects the accuracy of the maps produced by remote sensing. The
results of this study will be used to select a remote sensing method to use in mapping
the large-scale spatial distribution of burrowing shrimp.
II. Population Dynamics of Eelgrass and Burrowing Shrimp Along Stressor Gradients:
Experiment 2. A. Population Characteristics of Eelgrass and Burrowing Shrimp Along the
Dominant Salinity-Temperature Gradient in Yaquina Estuary, Oregon: This study will
provide a baseline of eelgrass and burrowing shrimp population dynamics, and will
evaluate population-level response variables that may be used for subsequent work.
-------
p^pipii
¦y?w' * *r iAh&Jmmi*l§-
w^^wfiwr ^gpg
Central Yaquina Estuary
1997 CIR Aerial Photos
100 0 100 200 300 400 500 Meters
fiANf SH) ( l&SJOOSV-flgi/IHWCtNTBAT AHR
Figure 1. Aerial color infrared photograph-mosaic of central Yaquina estuary showing three
areas where remote-sensing methods-comparisons studies may be conducted.
-------
-------
57
The ability to assess the ecological risks posed by multiple stressors to eelgrass and
burrowing shrimp depends on defining the current condition of the resource to serve as
a benchmark from which to judge their responses to natural or anthropogenic stressors.
Defining these benchmarks is important for field experiments proposed later in this
proposal, as well as for research proposed by other CEB scientists (i.e., Boese,
Ozretich, Young, and Robbins). Because salinity and temperature are important
natural "stressors" affecting the distribution of organisms within estuaries, we will
characterize the population dynamics of Zostera, Nerotrypaea, and Upogebia along the
of temperature-salinity gradient of Yaquina Bay (as defined by DeBen et al. 1990;
Figure 2). Specifically, the population studies on these three species will: 1) document
the demographic characteristics for each species along a natural stressor-gradient; 2)
investigate the utility of various measures of population characteristics as indicators of
population-level response to natural stressors; 3) evaluate the frequency at which
various demographic characteristics need to be sampled; and 4) evaluate methods that
could be used for rapid assessments of condition for Zostera.
These objectives will be met by using "intensive" and "extensive" sampling approaches.
The "intensive" sampling effort will be done at large patches of Zostera, Neotrypea, and
Upogebia co-occurring in Zones I, II, and III as classified by DeBen et al. (1990) (Figure
2). No intensive sampling site will be located in section IV as these species are
uncommon in this part of the estuary. The rationale for sampling large patches is to
characterize the population dynamics of each species where it grows best as indicated
by extent of the patch and to sample an area large enough that sampling and the
associated trampling are unlikely to destroy the patch. The "extensive" sampling effort
will employ collecting random samples of Zostera, Neotrypea, and Upogebia
throughout the entire estuary taken in conjunction with the sampling being done by
Ferraro and Cole (CEB). The intensive samples will be used to develop the detailed
characterization of the population dynamics of each species, while the extensive
sample will provide insight into whether population information from the intensive
sample sites can be extrapolated throughout the estuary. Both types of samples will
provide information on the relative utility of various demographic variables and methods
used to collect population data. Technical details for this study can be found in
Appendix A.
The DeBen et al. (1990; Fig. 2) salinity-temperature stratification represents annual
average environmental conditions, which undoubtably vary seasonally. Ideally, the
stressor gradient will be sufficiently steep that eelgrass and shrimp at each study site
will experience distinctly different salinity and temperature regimes. However,
preliminary data is not available to make this determination, and physical/chemical
environmental measurements made during each sampling will be used to characterize
the seasonal variation of environmental conditions at each site. If the environmental
conditions are not distinctly different among sites, the study may be changed after the
first year to include other eelgrass or shrimp beds.
-------
58
Fig 2. Stratification of the Yaquina estuary by salinity and temperature regime, using the
studies will be conducted in Sections (zones) I, II, and III.
¦ 24*00
1 •SECTION
0BIOCC
I YAQUINA
BAY
TOLEDO
TlAHD
rtouinA atct-
COOS 5*r
OREGON
I o 1 ZKm
-------
Experiment 2.B. Development and Testing of Biophysical Population Models: This
study will develop and test "biophysical population models" that predict whether
Zostera marina, Neotrypea, or Upogebia should be present or absent at a site based on
physical and chemical properties of the site. These models only classify sites as to
their potential as suitable habitat for the target organism, and cannot unequivocally
identify the factors that cause an organism to be absent. These models are also useful
for ranking the importance abiotic stressors that affect the distributions of eelgrass and
burrowing shrimp populations within estuaries, and are useful for guiding habitat
restoration efforts (Thorn 1990). The biophysical population models will consist of the
range of conditions in which populations of each species have been found in Pacific
Northwest estuaries, as reported in previous studies. A preliminary biophysical model
for Z marina is presented in Table 1, which was derived by Thorn (1990) as a list of the
habitat characteristics for eelgrass in Pacific Northwest estuaries. The model for Z.
marina will be updated by review of the literature since Thorn (1990), and similar
models will be prepared similarly for Neotrypea and Upogebia based on review of the
literature, much of which has been compiled already (see DeWitt et al. 1997; Feldman
et al. in prep.).
Table 1. Preliminary list of components for a biophysical model for Zostera marina
populations in the Pacific Northwest. Table from Thorn (1990).
Depth
0.0 to -6.6 m MLLW
optimal
Light
r>f ci irfarA irrartianre
maY hinmactc rprnrrifarl
70-175 FE rrr2 s"1
saturates photosynthesis; temp.
0.95 X mean annual Secchi depth
max. depth limit
Nutrients
ample inorganic nitrogen and
phosphate
growth can be nutrient limited;
sources are sediment and water
column; excess nutrients can
reduce growth due to high
er>ir>h\/te hinmacc
Salinity
10-30%o
optimal
Sediment
mixed sand and mud
optimal
Slope
flat to very slight incline
optimal
-------
60
Temperature
10-20°C
optimal
To test the biophysical population models, we will sample for the presence of each
species (i.e., Z marina, Neotrypea, and Upogebia) and benthic and water-column
environmental parameters (see Table A.4, Appendix A) at approximately 100 sites
within Yaquina estuary. Geographic positioning system (GPS) coordinates for
sampling sites will be designated using a tesselated stratified-random design (Stevens
1997), with the estuary stratified by depth and salinity-temperature zones (Figure 2).
Water-column parameters will be measured three or four times per year (spring,
summer, fall, and possibly winter), whereas the ground-based habitat abiotic and biotic
variables will be measured once during the active growing and burrowing seasons for
eelgrass and shrimp (i.e., between April and October). Water column parameters may
be collected over the span of two years to obtain a better measure of their temporal
variation. Site-selection, water-column sampling, and benthic sampling will be
coordinated with Bob Ozretich and David Young (CEB) whose research call for similar
site dispersion and data.
One unresolved issue is how best to represent the temporal variability in water column
parameters at each site. Clearly, water column parameters will be more variable
temporally than sediment-associated variables. Single samples at high tide collected
once per season will not be sufficient to characterize this variability. One approach is
to report the maximum and minimum extremes for each parameter. We will also
investigate linking measurements made at the sites to 1) water quality data collected
continuously by moored CTD/PAR units (see proposal by David Specht, CEB) using
regression techniques, and 2) to a circulation model for the Yaquina estuary, such as
the 2-dimensiona! model used by the NOAA/PMEL Tsumami Project (Kamphaus 1998).
In any case, we will consult with other estuarine scientists to identify other ways to
integrate these temporally-variable data. These approaches could may eventually
allow CEB to have the capability to accurately model water column parameters at sites
within the estuary.
The goodness of fit of the predicted and measured presence and absence data for
each species will be tested using the G-test (Sokal and Rohlf 1981). The null
hypothesis to be tested is that the measured data are not different from the predicted
data (i.e., that the model correctly predicts the presence and absence of each species).
Additional technical details of this experiment may be found in Appendix A.
Experiment 2.C. Effects of Multiple Abiotic Stressors on the Population Biology of
Keystone Species: Laboratory (mesocosm) and field experiments will be conducted to
measure the effects of single and multiple abiotic stressors on population
characteristics (i.e., survival, growth, and fecundity) of eelgrass and burrowing shrimp
I
-------
(
61
under controlled conditions. The purpose of these experiments will be to 1) determine
whether the interacting effects of the stressors is greater than the sum of the effects of
each independent stressor, and 2) to determine which stressor(s) has the greatest
impact on eelgrass or shrimp survival or growth. Mesocosms are simplistic mimics of
the natural environment that offer an opportunity to conduct stressor-response
experiments under controlled conditions. Field experiments face the risk that
unanticipated, uncontrolled, local events may confound the experimental treatments,
such as vandalism, accidental human disturbance, erosion, sediment deposition, or
rapid changes in water quality. Mesocosm experiments will be conducted first, followed
by field experiments to verify the results of the laboratory studies.
Mesocosm Experiments: CEB plans to construct mesocosms that are appropriate for
conducting these experiments during FY1998 (Appendix A). Experiments to measure
the presence and outcome of competition (expt. 3.A. and 3:B.) will be conducted in the
mesocosms before the multiple stressor-response experiments are conducted. The
schedule for starting the multiple-stressor experiments is dependent on the availability
of the mesocosms and the results of the competition experiments, and could thus start
either in spring/summer 1999 (if competition is not measurable) or sometime in 2000.
Three experiments that are anticipated are:
1) Zostera (1 density) x salinity (3 conc.) x light (2 intensities) x temperature (3 levels) x
4 replicates = 72 compartments;
2) Neotrypea (1 density) x salinity (3 conc.) x temperature (3 levels) x 4 replicates =
72 compartments; and
3) Upogebia (1 density) x salinity (3 conc.) x temperature (3 levels) x 4 replicates = 72
compartments.
Aside from specifying the species and stressors, the methods used for all
stressor-response mesocosm experiment will be identical. Population response
variables will be similar to those used in experiment 2. A. Technical details of the
experiments are described in Appendix A.
Field Experiments: Field studies to measure the responses of eelgrass and burrowing
shrimp to multiple stressors will verify results of the mesocosm experiments. To the
greatest extent possible, fully-factorial designs will be used in the field experiments,
although that might be difficult to achieve for stressors that co-vary spatially or
temporally (i.e., depth and mean temperature, salinity and temperature, light and
temperature). Natural gradients in abiotic stressors will be used when possible,
particularly for water-column variables (i.e., salinity, nutrients, turbidity, temperature,
dissolved oxygen). Some variables are amenable to experimental manipulation, such
as light (by shading), substrate-type, sedimentation rate, sediment nutrient
concentration, and depth. Selection of stressor treatments will be guided by the
mesocosm multiple-stressor experiments, the biophysical models (expt. 2.B.), and the
-------
62
logistics of finding or creating spatially and temporally reliable gradients of the stressor.
The population-level response variables will be similar or identical to those used in
previous mesocosm and field experiments. Technical details of the experiments are
described in Appendix A.
Container or Cage Artifacts: The cylinders in the mesocosm experiments (and
below-ground cages in field experiments) may affect the growth of eelgrass or
burrowing behavior of the shrimp. It is suspected that the latter will be more of an issue
than the former because eelgrass has been successfully grown in very small containers
(i.e., 1-gallon pots) in laboratory and field experiments (pers. obs. of work by Ron
Thorn). Controlling for container effects with burrowing shrimp may be very difficult,
especially in the mesocosm studies. However, as the same chambers (cylinders or
cages) will be used for all treatments within the mesocosm and field experiments, any
artifacts will be equally distributed among replicates and treatments. Additionally, I will
compare the behavior and physiological/ecological properties of shrimp in experimental
chambers (mesocosm or field experiments) with those of shrimp in field populations to
determine the magnitude of laboratory artefacts. Variables measured could include the
surface area: volume ratio and the number of angular turns of burrows (determined
from resin casts at the end of experiments), sediment deposition rate, oxygen
consumption rate, burrow irrigation rate, or elemental flux rates. If experimental
artefacts are minimal, I would expect the characteristics of shrimp measured in the
experiments to fall within the range of characteristics measured in the field populations,
as determined by non-parametric statistical comparisons of median values. (This also
assumes that important environmental variables, such as temperature, salinity, and
sediment grain size are comparable in the experimental and field populations). It would
be impractical to change the size or physical characteristics of the cylinders used in the
mesocosm and field experiments, as suggested by the reviewers, without sacrificing
other important experimental design considerations, particularly the number of
replicates. Thus, I will first measure the magnitude of artefacts, and then adjust the
design of the chambers or experiments only if artefact effects are substantial.
Experiment 2.D. Mapping Changes in Eelgrass and Burrowing Shrimp Populations
Along Stressor Gradients: This study will measure whether the spatial distributions of
eelgrass or burrowing shrimp change along abiotic stressor gradients in the manner
predicted from laboratory and field experiments, and to determine whether changes in
the spatial distribution of a species are characteristic of particular classes of stressor.
One possible product from this work could be a landscape-scale indicator of stress to
estuarine benthic habitats (i.e., eelgrass or burrowing shrimp beds). The design of this
project is incomplete and will be developed with other CEB collaborators (Drs. Brad
Robbins (landscape ecology), Denis White (biogeographer), and David Young (aerial
photography)). Very broadly, our goals are 1) to develop population change maps
SAVs and burrowing shrimp for portions of Yaquina estuary, 2) to develop similar maps
for other estuaries in order to measure differences in rates of population change across
-------
63
the region, and 3) to develop population change maps along stressor gradients to test
hypotheses concerning detection and measurement of changes in species' spatial
distribution in relation to abiotic stress. Population change maps (= difference maps,
isopachs) will be constructed using remote sensing methods and the techniques for
comparing maps described in experiments 1.B. and 1.D.
Several issues will need to be addressed prior to testing hypotheses about population
spatial changes in relation to stress. First, we need to know what spatial scale to map
for each species. The mapping accuracy study (expt. 1.B.) will examine how accuracy
changes with spatial scale for each of three remote-sensing methods. Results from
that study will be balanced against the resolution of available data (to be acquired from
other researchers or public agencies) and the cost of obtaining high-resolution spatial
data. Second, we need to know the temporal scale at which to map population change.
We anticipate that annual or biennial maps may be adequate for these purposes, but
also recognize that eelgrass and burrowing shrimp distributions can change
dramatically seasonally. One approach to resolving this question would be to generate
population change maps for select sites on a seasonal basis for 2+ yr, measure the
rate of population change over different temporal scales, and select the scale at which
the rate of change reaches an asymptote. Third, we need to know whether natural
rates of population change vary geographically across the PNW. It seems likely that
latitudinal gradients in day length, temperature, rainfall, or storm intensity could affect
natural rates of population expansion and contraction. Underlying natural
environmental gradients will have to be identified and their influence measured before
we can confidently use population change maps to detect and measure the effects of
anthropogenic stress.
A plan for this project will be developed over the next year. In the meantime, we can
start to measure rates of annual change in spatial distribution of eelgrass using aerial
photographic data for Yaquina estuary obtained by David Young (CEB) in 1997 and
1998. We will also seek data from previous aerial photographs of Yaquina estuary (an
inventory of which is being prepared by David Specht, CEB), and time-sequenced
remote-sensing data from other estuaries. We anticipate forming collaborations with
researchers within the region who have similar interests in mapping the large-scale
distributions of these species in other estuaries. These include, potentially, Roxanna
Hinzman (Tillamook Bay National Estuary Project), Steve Rumrill (South Slough
National Estuarine Research Reserve), Si Simenstad (Univ. Washington), Doug
Bulthuis (Padilla Bay National Estuarine Research Reserve), Tom Mumford
(Washington Dept. of Natural Resources), and Ron Thorn (Battelle Marine Sciences
Laboratory). Additional technical issues are discussed in Appendix A.
III. Effects of Multiple Abiotic Stressors on Competition Among Eelgrass and Burrowing
Shrimp - Laboratory and field experiments will be conducted to measure the presence
of interspecific competition for space among Zostera, Neotrpea, and Upogebia (expt.
-------
64
3.A.). The goals and design of subsequent laboratory and field experiments will
depend on the presence of competition among these species. If competition is
important, the next laboratory and field experiments will be directed to measuring
whether water quality affects the outcome of the competitive interaction (expt. 3.B.). If
competition is not important, laboratory and field experiments will focus on measuring
the effects of multiple stressors on each species individually (i.e., experiment 2.C.).
Experiment 3.A. Measuring Interspecific Competition Among Eelgrass and Burrowing
Shrimp: Laboratory and field experiments will be conducted to test the null hypothesis
that interspecific competition does not occur between Zostera and Neotrypea, between
Neotrypea and Upogebia, and between Zostera and Upogebia. Mesocosm experiment
to test this hypothesis will be conducted prior to conducting field experiments.
Definitive laboratory experiments will be conducted during fall 1998 - summer 1999;
while the mesocosms are being constructed, preliminary laboratory and field
experiments may be conducted during summer 1998.
Mesocosm Experiment: The methods used for these experiments will be identical to
those described for the mesocosm stress-response experiments (expt. 2.C.), with the
experimental treatments consisting of pairs of species at different population densities.
The first experiment will measure competition between Zostera and Neotrypaea
because this interaction is expected to have the greatest impact on the distribution and
abundance of benthic habitats (and associated species) in PNW estuaries. The
experimental design will be: Zostera (4 densities: 0, 100, 200, 400 shoots m'2) x
Neotrypaea (4 densities: 0, 100, 200, 400 shrimp m'2) x 4 replicates. The second and
third experiments will measure competition between Zostera and Upogebia, and
between Neotrypaea and Upogebia. The densities will be comparable to those used in
the Zostera-Neotrypaea competition experiment.
The competition experiments will be very similar to those described for the multiple
stressor experiment, employing the same methods as described for experiment 2.C. to
harvest, acclimate, and plant test organisms, and to sample population variables (
,Appendix A, Table A.4). The duration of the competition experiments will be 8 to 16
weeks, depending on whether effects are measurable by non-destructive sampling.
Results will be analyzed using analysis of variance to test the null hypothesis that
changes in the density of one species did not affect the demographic variables of the
second species. If competition between Zostera and either species of shrimp is
significant, we will probably move ahead to conducting experiments on the effects of
water quality on the outcome of competition, rather than immediately conduct the
experiments to measure competition between the shrimp species. We will return to
conduct that experiment after completing experiments on the effects of abiotic stressors
on the competition between eelgrass and shrimp. Additional experiments may be
conducted to examine three-way interactions among these species after the two-way
interspecific interactions have been explored.
-------
65
Field Experiment: Reciprocal-transplant field experiments will be conducted to verify
that interspecific competition can be measured in the field. Zostera will be transplanted
at different densities into Neotrypaea or Upogebia beds, Neotrypaea within Zostera or
Upogebia patches, and Upogebia within Zostera or Neotrypaea patches. All possible
pairings of these three species will be tested simultaneously. For each transplanting,
pairs of competitors will be identified as "invader" (the species that is placed into the
patch) and "resident" (the species occupying the patch prior to the experiment).
Eelgrass or shrimp "invaders" will be harvested from nearby populations, sorted to
uniform size and condition, and "planted" inside caged plots (40-cm diameter) within
large patches of the "resident" species. The plots will be created by excavating a 40-
cm diameter x 60-cm depth core of sediment, lining the hole with 3-mm mesh plastic
screen "underground fence" to a depth of 60 cm, and refilling the hole with the
excavated sediment which will be sieved to <3-mm to remove burrowing shrimp,
eelgrass, and other megafauna. Initially, the plots will covered with 3-mm mesh cages
to prevent immigration or emigration of conspecifics, predators, or competitors. After
an acclimation period of a few days, all cages will be removed, and underground
fences will be removed from some plots but retained in others.
Three treatment-types will be deployed for each species pair: control, fenced invader,
and unfenced invader (Figure 4). For controls, shrimp or eelgrass will be back-
transplanted into beds of their own species, which will measure the effects of
transplant-shock on the invader. Fenced out-plant plots will measure both transplant-
shock effects and the suitability of the "resident" habitat to support the survival and
growth of the invader. The un-fenced invader treatments will measure the combined
effects of the presence of the "resident" species (i.e., the competition), habitat
suitability, and transplant shock. The replication rate will be based on results of a
power analysis of key response variables (i.e., %cover and net primary productivity for
Zostera, abundance for the shrimp) using variance estimates from field populations
(i.e., Kentula and Mclntire 1986, Dumbauld et al. 1996). The duration of the
experiments could be extended for as long as 1-yr if there is no apparent effect of
competition; in this case, a bi-monthly sampling frequency will be used. The long
duration may be required for competition to manifest itself, although we expect the
effects to be revealed within weeks, particularly between Neotrypaea and Zostera
because of the high rate of sediment turnover caused by ghost shrimp.
Population-level responses of resident species measured within experimental plots will
be similar or identical to those used in previous mesocosm and field experiments
(Appendix A, Table A.4). Non-destructive samples will be measured approximately
every 4 weeks (depending on site accessibility due to tides and weather). The field
experiment will be conducted for 6 to 20 weeks, depending on the rate at which
responses were observed in mesocosm experiments and on whether responses are
observed and are measurable in the field. Analysis of variance will be used to test the
null hypothesis that the effect of competition on population characteristics is the same
-------
66
as the effects of transplant shock and habitat suitability (i.e., that population
characteristics of the invader in the un-fenced treatment are not significantly different
from those of the invader in the fenced or control treatments).
Experiment 3.B. Effects of Abiotic Stressors on the Outcome of Competition Between
Keystone Species: Laboratory experiments, and subsequent field experiments, will
measure the effects of water quality or benthic habitat variables on the outcome of
competition between eelgrass and burrowing shrimp. The laboratory experiments will
be very similar to those described above (expt. 3.A.), except that fewer population-
density treatments will be used, and one or more abiotic stressor levels will be used.
The first abiotic stressor to be studied is light, based on the suggestion by Harrison
(1987) that improvement of water clarity led to a increase in eelgrass, to the detriment
of burrowing shrimp.
Mesocosm Experiments: The experimental design will be something along the lines of:
Zostera (3 densities) x shrimp (3 densities) x light (2 intensities) x 4 replicates. The
species to be used will be determined from experiments in 3.A. The methods to
conduct this experiment and to measure responses of the eelgrass and shrimp will be
the same those described for experiment 3.A. Subsequent mesocosm experiments
may be conducted to examine the effect of salinity, temperature, nutrients, or substrate-
type on competition between eelgrass and the shrimp (or between the two species of
burrowing shrimp). Results of the biophysical population-model study (expt. 2.B.) will
be used to select the stressors. Later experiments may examine the effects of two
interacting stressors on competition.
Field Experiments: If the above mesocosm experiment demonstrates that abiotic
stressors can affect the outcome of competition, field experiments will be conducted to
verify this finding in the presence of "strong" and "weak" abiotic stressors. It is
impractical to plan these experiments in detail at this time, but they are expected to
employ either the reciprocal-transplant methods described above (expt. 3.A.) using
"resident" populations distributed along natural stressor gradients, or to outplant pairs
of species at different population densities into plots distributed along natural or
artificial stressor gradients. Varying abiotic stressor levels in the field can be achieved
by either conducting the experiment along existing stressor gradients (i.e., salinity-
temperature, depth, nutrient, etc.) or by manipulating habitat characteristics (i.e., using
shade cloth to reduce light levels; Bulthuis 1997). The duration of these experiments,
the measurement variables, and statistical analytical methods will likely be the same as
those used in field experiments described previously (expt. 2.C. and 3.A.).
Expected Results and Benefits: The research proposed here will contribute to the
understanding of the population ecology of keystone species in Pacific temperate
estuaries, and their responses to abiotic stressors, which is required to assess risks of
multiple stressors to estuarine ecosystems and to manage the living resources therein.
-------
67
One strength of the proposed work is that it integrates ecological processes occurring
at several spatial and temporal scales, from responses measured after several weeks
in laboratory and field experiments, to annual (or longer) changes in landscape-scale
patterns of populations among coastal estuaries of Pacific Northwest. The proposed
research project uses several different approaches to study these problems (including
laboratory and field experiments, estuarine-scale population surveys, and remote
sensing) and integrates the resulting data into GIS layers that will be incorporated into
the Yaquina estuary GIS database maintained by the US EPA WED/Coast Ecology
Branch. The proposed experiments are highly integrated, sharing methods and
information, conserving resources by coordinating efforts, and organized to be
conducted in a hierarchical fashion (i.e., baseline studies conducted first, which build
into more complex studies). Results from these studies will be complied into
manuscripts prepared for submission to peer-reviewed scientific and environmental
management journals.
From a basic-science perspective, the proposed research will provide important
and unique information about the population and community ecology of estuarine
keystone species. In particular, this research will evaluate the importance of
competition among keystone species which is believed to be a fundamental, but
overlooked, mechanism that determines the distribution and abundance of valuable
eelgrass habitat. From an environmental-management perspective, the proposed
research will provide critical information on the effects of abiotic stressors on the
dynamics of habitat dominated by eelgrass or burrowing shrimp. These are arguably
the most ecologically and economically important benthic species within Pacific
estuaries because of their habitat-generating and habitat-modifying activities. Data
from this research may be useful in the establishment of water quality guidelines or
other management tools, for siting ecological restoration projects, or for habitat
enhancement programs. New information on the ecology of burrowing shrimp may lead
to improved methods to control their population growth within commercial oyster beds,
and thereby lead to reduced need for applying Sevin pesticide in Willapa Bay and
Grays Harbor, WA. Additionally, the contemporary and historical data collected for
Yaquina and other Pacific estuaries, stored in GIS-database format, will be useful to
community-based environmental management of Pacific coast watersheds and
estuaries.
-------
68
Project A4 - The Impact of Disturbances on an Eelgrass Habitat
Principal Investigator: Bruce L. Boese
Goals: Project goals will be to determine the benthic effects of disturbance (i.e.,
recreational clam harvesting and algal smothering) on eelgrass (Zostera marina)
habitat. The hypothesis is that disturbance of an eelgrass patch reduces shoot density
resulting in a reduction in eelgrass primary productivity and the diversity and
abundance of associated mega and macrobenthos. Results of the clam harvesting
experiments will be placed in a larger spatial context by comparison to the results of a
field survey conducted in an eelgrass meadow which is subjected to heavy recreational
clamming activity.
Rationale: Eelgrass is an important habitat in Pacific Northwest (PNW) estuaries.
Zostera marina meadows serve as a nursery ground for juveniles of commercially
important species such as herring and as a refuge for juvenile salmonids (Griffin, 1997;
Simenstad and Wissmar, 1985; Levings, 1990; den Hertog , 1977). Eelgrass meadows
are significant sites of primary production and eelgrass shoots can be utilized directly
for food by some water fowl such as the western black brant (Griffin, 1997; Kentula,
1982) and indirectly by many species via consumption of detritus (Thayer etai, 1975).
Eelgrass roots stabilize the sediment (Thayer etal. 1975) and the presence of
eelgrass dampens wave energy which may serve to reduce erosion and to enhance
larval settlement (Orth, 1992). Because of these characteristics species abundances
in eelgrass patches are usually greater than in other estuarine habitats (Everett et a!.,
1995).
Mechanical disturbances associated with commercial shell fish and bait harvesting
operations have been shown to reduce mudflat biodiversity (Brown and Wilson, 1997)
and to adversely affect SAV growth (Fonseca etal., 1984; Peterson etal., 1987; Everett
et al., 1995). Although the obvious factor effecting the upper limit of SAV growth is
tidal exposure (Thayer, et al., 1975), bioturbation by a burrowing sand shrimp
[Neotrypaea) may have an effect on the upper growth limit of Z. marina and may
prevent a high intertidal eelgrass species (Z. japonica) from inhabiting the full extent of
its intertidal range (Dumbauld et al., 1997). The establishment of shrimp beds has
been shown to be capable of excluding seagrasses (Suchanek, 1983), conversely,
dense eelgrass roots have also been shown to inhibit the burrowing ability of shrimp
(Benchley, 1982). In addition, natural eelgrass colonization of sandflats and eelgrass
transplant experiments have been shown to reduce Neotrypaea abundance (Harrison,
1987).
A relatively unexplored eelgrass disturbance factor is the potential for macroalga (Ulva
and Enteromorpha) to overgrow and smother relatively large eelgrass patches.
Macroalgal blooms which are often associated with increased nutrients may be a
-------
69
contributing factor to the apparent world wide decline in seagrasses (see review by
Raffaelli et al., in press). For example, den Hartog (1994) documented a sudden and
complete disappearance of Zostera species from a British estuary and concluded that
increased nutrient loads from a new sewage treatment plant and a warm/dry fall
resulted in a massive increase in Enteromorpha which appeared to smother and kill
eelgrass patches. While observations such as this suggest a direct linkage between
macroalgal increases and eelgrass declines, increased nutrient loads may have a
direct adverse effect on eelgrass by stimulating epiphyte growth on eelgrass leaves
which serves to reduce available light for photosynthesis and by reducing water clarity
via increased phytoplankton densities (Dennison eta!., 1993). Thus it may be difficult
to separate the direct effects macroalgal growth on eelgrass from nutrient effects
(Raffaelli et al., in press). In Oregon estuaries, floating Ulva mats have been observed
to similarly cover and kill Z. marina patches (Kentula, 1982), especially during late
summer low tides (Kentula, pers. comm.), however, the mechanism by which this
occurs, its spatial impact and recovery time are unknown.
As eelgrass has been shown to affect current velocity and turbulence (Gambi et al.,
1990), reductions in eelgrass densities by mechanical disturbances or by smothering
with Ulva mats or storm deposited sediments (Onuf, 1987) may serve to alter the wave
or current energy at the affected site (Peterson, etal., 1987). Colonization of these
sites by Neotrypaea and the potential for increased wave energy at these denuded
sites may perpetuate the disturbance, making recolonization by eelgrass difficult. This
sort of interaction has led to speculation that sandflats and seagrass habitats may be
alternative stable states (Peterson et al., 1987).
The habitat relationships between sandflats dominated by Neotrypaea and eelgrass is
of particular interest in the PNW. The oyster culturing methods used in Yaquina Bay
(stake and rack culture) have been shown to be destructive to eelgrass (Everett et al.,
1995), however, a definitive study of oyster ground culture (method used in Willapa
bay) on eelgrass patches, has not been done (John Johnson, ODFW, pers. comm.).
One possibility is that persistent disturbance from activities associated with oyster
culture creates a disturbed environment in which burrowing shrimp can thrive at the
expense of eelgrass (Simenstad and Fresh, 1995) indirectly affecting the utility of the
habitat for continued oyster culture. Regardless of the cause, Neotrypaea and to a
lesser extent burrowing mud shrimp (Upogebia) are considered pest species by oyster
growers and are periodically sprayed in Washington State with carbaryl as a control
measure. In Oregon, carbaryl spraying for the same purposes is a recurring proposal
but is currently illegal.
Recreational clamming is a significant activity in PNW estuaries which may have an
impact on eelgrass. Locally, gaper (Tresus capax) and butter clams (Saxidomus
giganteus) are harvested by digging in lower intertidal Z. marina meadows while
cockles (Clinocardium nuttalli) are often raked from the sediment surface of higher
-------
70
intertidal areas. Clamming activity is not limited to intertidal areas of eelgrass
meadows as some clam diggers wade at low tide into two or three feet of water to dig
subtidal clams. Thus, this mechanical disturbance could affect eelgrass to -6 ft MLLW.
An additional potentially disruptive activity involves the taking of bait shrimp. This is
done both recreationally using a hand operated slurp gun and commercially in which a
high pressure stream of water is directed into the sediment. Commercial shrimp
operations are highly destructive as the sediment is liquefied, collapsing burrows with
the surviving shrimp harvested as they are force to the surface by the injected water.
Although most of the local commercial activity has been directed at Neotrypaea
populations in Alsea Bay, some Upogebia are harvested from eelgrass meadows in
Yaquina and Netarts Bays.
It is apparent from the above discussion that physical disturbance of eelgrass meadows
has the potential to cause deleterious effects on an estuarine scale. As human coastal
populations and associated tourism increases, the potential for physical disturbance
from recreational and commercial activities are also likely to increase. The proposed
research describes field experiments which will address the potential for some of these
physical disturbance to affect eelgrass patches and their associated biota.
Objectives: 1. To examine the short and long-term effects of mechanical disturbance
(recreational clamming) and algal smothering on eelgrass patches and associated
biota.
2. To compare recreational clamming experimental results to the results of a
field survey conducted at a site which has been subjected to prolonged and intense
recreational clamming.
Experimental Methods:
Recreational Clamming Simulation Experiments: The effects of two recreational
clamming methods (raking and digging) on Z marina and associated biota will be
examined beginning in May 1998. Clamming treatments will be applied to experimental
plots multiple times to simulate a high level of recreational clamming activity. The
digging treatment will be applied to Z marina experimental plots where T. capax and S.
giganteus are likely to be present and the raking treatment will be applied to
experimental plots at which C. nuttalli are likely to be present. As T. capax and S.
giganteus are found in low intertidal eelgrass meadows, the two treatment types will be
applied in vertically separate areas which are also likely to have differing eelgrass
densities.
-------
71
The locations of the two study sites will be selected using the existing 1997 aerial
survey of Yaquina Bay and its associated ground truth survey. Preliminary surveys will
also be conducted in March and April 1998 to locate clam beds within these eelgrass
habitats. The most ideal location for both of these experiments is the Sally's Bend mud
flat (Figure 1) which contains the most extensive eelgrass meadows in Yaquina Bay.
Sally's Bend is relatively flat, thus it should be possible to locate all plots within a
narrow tidal range. Public access to this eelgrass meadow is limited as the central
areas of the meadow are accessible at low tide only via hovercraft.
Once the locations of the two study areas are known, transects or plot grids will be
established in each. Along each transect or within each of the plot grids a series of 1.5
m X 1.5 m contiguous potential sampling plots will be numbered and randomly
designated as control or treatment plots. Each of these plots will visited in order and
visually accessed using the following acceptance criteria: 1) eelgrass is present, 2) the
plot is undisturbed (e.g., no obvious signs of clam harvesting), and 3) the plot is at
least 1.5 m away from previously selected control or treatment plots. If the plot meets
these criteria it will be marked with a numbered wooden stake located at a specified
distance and direction from the center of the plot and the GPS coordinates noted.
Locating marking stakes away from plots will help ensure that drift alga entrapped by
the stake will not effect the plots.
Clam Raking Experiment: The clam raking experiment will begin in May 1998 and
consist Of 50 treatment and 50 undisturbed j Figure 1: Major Yaquina Bay
control plots which are randomly assigned \ Eelgrass Meadows
along a transect or within a plot grid. All 50
treatment plots will be raked using a four-
tined hoe (each tine ~ 20 cm in length) to
remove cockles. This treatment will be
repeated on these same 50 plots two
additional times (June and July 1998).
Concurrent with the July raking treatment, ten
control and ten treatment plots will be
selected for destructive sampling in August
1998. Fifteen eelgrass shoots will be marked
in each of these plots for net primary production measurements. Shoot marking
techniques are described in Kentula (1982), Kentula and Mclntire (1986), and Zieman
(1974). The first five of these shoots which appear undamaged will be collected when
these plots are destructively sampled at the next sampling interval (~1 month) and
analyzed (leaf width, length, growth) using the procedures described in Zieman(1974)
and Kentula and Mclntire (1986).
ally's Bend
ty ClammM
Destructive sampling of randomly selected plots will begin in August 1998 (Table 1).
-------
72
This sampling will accomplished by placing a 1.0 m2 quadrat in the center of each 2.25
m2 plot (Figure 2). Single cores for macrofauna (8 cm dia., 5 cm deep) and grain
size/TOC (3.4 cm dia., 5 cm deep) will be taken at random sites within the quadrat.
Also at random locations within each of these quadrats Z. marina will be harvested from
four 15 cm diameter areas (Figure 2) and transported to the laboratory for biomass
measurements. Megafauna will then be sampled from these harvested areas using
four cores (15 cm dia., 60 cm deep)1. These cores will be combined and sieved (3.0
mm) in the field. All samples will be numbered as to plot and transported on ice to the
laboratory. Macrofaunal cores will be sieved (0.5 mm) in the laboratory. Material
retained on sieves from mega and marcofaunal cores will be preserved (10% formalin)
then transferred into 70% ethanol for later sorting, taxonomic identification (species
level), and biomass determinations. This sampling pattern will continue until recovery
(no significant difference between treatments/controls) or for up to two years post-
treatment (Table 1).
Figure 2: Plot Design
Destructive Sampling
1.0 m
Macrofauna Core
1.0 m
1.5 m
EeJgrass Shoots Marked For
Primary Production!
Eelgrass Biomass and
Megafauna Cores
1.5 m
1An alternative method is currently under development at our laboratory. This method uses a flushing-
coring device similar to one used to collect bivalves to a depth of 70 cm (Gmssendorf, 1981). This should allow us
to sample a larger area to depth with less effort than that required when using large diameter manual corers.
-------
Table 1: Sampling Strategy for the Clam Raking Experiment
Date
Plots
destruc-tively
sampled
Macrofauna
Cores taken
Megafauna
Cores taken
# Plots
harveste
d
Grain
Size TOC
cores
1° Production
Measurements
# Plots
Raked
May
1998
50
June
1998
50
July
1998
10 plots
marked
10 plots
marked
50
Aug.
1998
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
May
1999
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
July
1999
10 plots
marked
10 plots
marked
Aug.
1999
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
May
2000
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
July
200
0
10 plots
marked
10 plots
marked
Aug.
2000
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
Total
50 Control
SO
Treatment
50
50
50
composites
50
composites
50
50
50
50
30
30
50
plots
3 X
each
-------
74
Clam Digging Experiment: During spring and summer 1998 low tides, lower intertida!
eelgrass patches will be visited to locate relatively undisturbed areas were T. capax
and S. giganteus populations are present. As with the clam raking experiment, 50
treatment and 50 control plots will be randomly assigned along a transect or within a
plot grid. All 50 treatment plots will be dug using a shovel to remove ail large,
recreationally important clam species. This treatment will be repeated on these same
50 plots two additional times (June and July 1998). Although the tentative sampling
design for this experiment (Table 2) is essentially identical to that used in the clam
raking experiments, it will be more difficult to accomplish as these low intertidal areas
are exposed for a shorter duration and treating plots by digging will involve more time
than raking. Hopefully the number of plots to be sampled can be reduced based upon
the preliminary results of the clam raking experiment and those of an ongoing study
being conducted at our laboratory to determine the optimal sampling strategy for
defining habitat-biota relationships (Ferraro and Cole, 1996). As with the clam raking
experiments, sampling will continue until recovery (no significant difference between
treatments and controls) or for up to two years post-treatment.
Table 2: Tentative Sampling Strategy for the Clam Digging Experiment
Date
Plots
destruc-tively
sampled
Macrofaun
a Cores
taken
Megafauna
Cores taken
# Plots
harveste
d
Grain
Size TOC
cores
1° Production
Measurements
# Plots
Dug
May
1999
50
June
1999
50
July
1999
10 plots
marked
10 plots
marked
50
Aug.
1999
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
May
2000
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
July
2000
10 plots
marked
10 plots
marked
-------
75
Aug.
2000
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
May
2001
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
July
2001
10 plots
marked
10 plots
marked
Aug.
2001
10 Control
10 Treatment
10
10
10
composites
10
composites
10
10
10
10
10 measured
10 measured
Total
50 Control
50
Treatment
50
50
50
composites
50
composites
50
50
50
50
30
30
50 plots
3 X each
Field Survey: Concurrent with the clam digging experiment, a similar sampling strategy
(Figure 2) will be employed in an eelgrass meadow which is harvested by the public for
T. capax and S. giganteus during summer low tides. Comparison between
experimental and field survey plot data will at least qualitatively indicate how well the
experimental manipulations mimic actual long-term clamming activities. The sampling
design for this study (Table 3) differs from the previous designs as an adjacent
undisturbed site is not available for a control and primary productivity measurements
will not be attempted due to the likelihood of disturbance by clam harvesters. If
recovery occurs in the clam digging experiment, the field survey will be discontinued.
Table 3: Sampling Strategy for the Field Survey
Date
Plots
Destruc-
tively
sampled
Macrofaun
a Cores
taken
Megafauna
Cores
# Plots
harvested
Grain Size
/TOC cores
Aug. 1999
10
10
10
composites
10
10
May 2000
10
10
10
composites
10
10
Aug. 2000
10
10
10
composites
10
10
-------
76
May 2001
10
10
10
composites
10
10
Aug. 2001
10
10
10
composites
10
10
Total
50
50
50
composites
50
50
Locating a suitable site for the field survey in Yaquina Bay may prove difficult.
The most extensively harvested area in the bay is located on the south shore from the
public fishing pier west to the nearest wing dam on the south jetty (Figure 1). However,
this site is approximately 2 km from the probable experimental sites in Sally's Bend and
may not be readily comparable. Another possible site is located on the far east side of
Sally's Bend. Although considerably closer to the primary study area, the site is only
occasionally clammed and is considerably smaller in size than the south jetty area.
Raking for cockles is not a major activity in Yaquina Bay, although this technique is
used at other sites on the Oregon Coast (e.g., Alsea Bay). A comparative site for the
clam raking experiment may be added to this project if a suitable site is located.
Algal Smothering Experiment: Preliminary work will be done in 1998 and 1999 to
determine: 1) when algal mats are present, 2) where algal smothering occurred, 3) the
duration of smothering, and 4) a practical method to experimentally smother eelgrass
with algae. Algal smothering experiments will begin in mid to late summer, 1999 or
2000 depending the continuing need to sample clam raking and digging plots.
In this experiment the effects of single, acute smothering events will be assessed, and
site recovery monitored over 1 to 3 years. As before, 2.25 m2 plots will be selected
along a transect or within a plot grid, and control (no disturbance) or treatments (algal
smothering) randomly assigned to the plots. The smothering treatment will be
accomplished by covering plots with algal mats which will be collected from adjacent
areas and restrained over the treatment plots in mesh bags or under nets which will be
staked down at the periphery of each plot. This method should allow the alga to settle
on the eelgrass at low tide yet provide for some circulation of oxygenated water at high
tide. These smothering treatments will be maintained for a duration that is similar to
that observed in pre-experimental surveys.
After the smothering treatment is removed, control and experimental plots will be
sampled using the same sampling methods and plot design used in the previous
experiments (Figure 2). Measurement procedure will be repeated within each plot until
recovery (no significant difference between treatment and control) or for up to two
years post-treatment. A proposed sampling strategy is presented in Table 4.
-------
77
Table 4: Tentative Sampling Strategy for Algal Smothering Experiment
Date
Plots distruc-
tively
sampled
Macrofaun
a Cores
taken
Megafauna
Cores taken
# Plots
harvested
Grain
Size TOC
cores
1° Production
Measurements
# Plots
smothered
July
2000
50
Aug.
2000
10 Control*
10 Treatment
10
10
10
Composites
10
Composites
10
10
10
10
May
2001
10 Control
10 Treatment
10
10
10
Composites
10
Composites
10
10
10
10
July
2001
10 plots
marked
10 plots
marked
Aug.
2001
10 Control
10 Treatment
10
10
10
Composites
10
Composites
10
10
10
10
10 measured
10 measured
May
2002
10 Control
10 Treatment
10
10
10
Composites
10
Composites
10
10
10
10
July
2002
10 plots
marked
10 plots
marked
Aug.
2002
10 Control
10 Treatment
10
10
10
Composites
10
Composites
10
10
10
10
10 measured
10 measured
Total
50 Control
50
Treatment
50
50
50
Composites
50
Composites
50
50
50
50
20
20
50 plots
1 time
each
•Depending on the method used to smother plots, treatment control plot (i.e., algal entrapment method but without
algae) maybe needed in addition to undisturbed control.
Data Analysis: Comparisons between controls and treatments within each of these
experiments is basically the same. Each experiment is a completely randomized
design which will be analyzed by two-way ANOVA. Sources of variation to be analyzed
are as follows:
-------
78
Source of Variation Number Degrees of Freedom
Time (Fixed) a=5 a-1
Treatment (Fixed) b=2 b-1
Raking and Control
Algal Smothering and Control
Interaction (a-1 )(b-1)
Error n=10 ab(n-1)
The exception to this will be the clam digging experiment in which samples taken from
the field survey will be included as an additional treatment. Although comparing
controls and digging plots located in Sally's Bend to a heavily clammed area located
two kilometers away is far from ideal, it does allow for some interpretation of the field
survey results. Sources of variation for this two-way ANOVA will be as follows:
Source of Variation Number Degrees of Freedom
Time (Fixed) a=5 a-1
Treatment (Fixed) b=3 b-1
Digging, Field Survey, Control
Interaction (a-1)(b-1)
Error n=10 ab(n-1)
The plot design (Figure 2) gives the appearance that multiple sampling strategies are
being employed within a plot. This is not the case as samples will either be
composited into one sample (e.g., megafaunal cores and harvested Zostera) or mean
values determined (e.g., primary production values) and compared among plots.
Expected Results and Benefits: Experimental treatments will initially reduce eelgrass
biomass and possibly primary production. This is likely to have deleterious effects on
macrofaunal abundance and diversity. Effects are likely to be enhanced and of greater
duration at the higher intertidal sites which generally have lower initial eelgrass
densities. Reductions in the eelgrass shoot density may result in the colonization of
the disturbed plots by opportunistic species such as exotic amphipods (e.g.,
Grandidierella japonica) and Neotrypaea. The extent and time course of colonization
by Neotrypaea would increase our understanding of the possible amensalistic
relationship between eelgrass and burrowing shrimp. In addition measurements made
on eelgrass productivity under stressed and unstressed conditions may be useful in
developing a method for assessing the relative health of eelgrass patches.
The present experiments are being conducted to determine if a significant treatment
effect could be observed in an area in which physical/chemical variation is minimal.
This design maximizes the ability to detected a statistically significant effect. If
ecologically important treatment effects are detected within the Sally's Bend study
-------
79
site, experiments would need to be conducted to expand the inference space, either
across Yaquina Bay and/or among several estuaries. If treatment effects are not
observed in the present study (as suggest by preliminary results), the likelihood of
finding a significant effect across a more heterogeneous environment would be minimal
and the next experiment (clam digging, algal smothering) would be started ahead of
schedule.
Results of the clamming experiments and the associated field survey would be useful in
predicting the effects of increased utilization of eelgrass meadows for recreational
clamming and may help in estimating the effects of other mechanical disturbances (e.g.
those associated with oyster culture and commercial shrimping). These data could be
combined with Oregon Department of Fisheries and Wildlife (ODFW) data on the
extent of recreational clamming activity in Yaquina Bay and other Oregon estuaries in
order to make projections on how this activity could impact eelgrass meadows as
human population and tourism increases. However, the proposed experiments are only
designed to determine if a significant treatment effect can be observed in an area in
which physical/chemical variation is minimal. The design maximizes the ability to
detected a statistically significant effect. If ecologically important treatment effects
are detected within the Sally's Bend study site caution should be used in extrapolating
these results across the entire Yaquina Bay estuary or among several estuaries.
Additional experiments may need to be conducted to expand the inference space in
order to make statistically valid large scale impact predictions. However, if treatment
effects are not observed in the present study, the likelihood of finding a significant
effect across a more heterogeneous environment would be minimal and an expanded
experimental protocol would not be necessary.
Results of algal smothering experiments could be scaled up to estimate the effect on
the eelgrass in the entire Yaquina Bay estuary using existing (1997) aerial
photographs which provide a snap-shot of the extent of algal rafts in Yaquina Bay. This
sort of information would be useful in estuarine risk assessments. The link between
increased nutrients and enhanced drift algae suggests that as the nutrient loads of
PNW estuaries increase with increasing human population, the smothering of eelgrass
by drift algae is likely to increase. The present study would provide a first look at the
associated benthic effects.
The results of this research will be communicated to the scientific community through
three or more peer reviewed journal publications and several regional and national
meeting presentations. This research should significantly advance our understanding
of how physical disturbance effects eelgrass and associated biota.
-------
80
Project A5 - Comparison of Factors Affecting the Distribution of the Non-
indigenous Seagrass Zostera japonica with Those Controlling the Native Zostera
marina in Yaquina Bay, Oregon.
Principal Investigator: David T. Specht
Goals: The project goal will be to determine the predominant factors affecting the
distribution of the non-indigenous seagrass Zostera japonica as compared to the native
seagrass Zostera marina in Yaquina Bay, Oregon.
Rationale: The native seagrass Zostera marina L. and the non-indigenous Z. japonica
Aschers. & Graebn. co-occur in many Pacific Northwest (PNW) estuaries. Phillips and
Mefiez (1988) establish the introduction of the non-indigenous Z japonica on the
Pacific Coast of North America concurrent with the importation of Japanese oysters and
spat to Willapa Bay, Washington, in 1925, and it has spread south to at least Coos
Bay, Oregon (Harrison and Bigley, 1982). Bayer (1996) concludes that Z. japonica was
introduced to Yaquina Bay in the early to mid-1970s; he first collected specimens in
1976, but had not encountered any in his prior eelgrass survey of 1974-75 (Bayer,
1979). There appear to be no other local records of recognition of or collection of Z
japonica (cf. Z. nolti) prior to that time.
In Yaquina Bay Zostera species are present in the intertidal zone along a salinity
gradient upriver from the area of the Yaquina Bay Bridge, approximately 1 river mile
(RM) inland from the mouth of the estuary, to a point upstream near Boone Slough,
approximately RM 10, ~3 statute miles south and west of Toledo (Figure 1). Lateral
distribution differences among the two species are evident. The vertical distribution of
Z. japonica in Yaquina Bay appears to be concentrated in the upper intertidal,
immediately adjacent to emergent aquatic rooted species of grasses, reeds and sedges
(seaward edge of the emergent salt marsh), occurring in bands paralleling the shore
from ~1 m to as much as 30-100 m wide, at elevations -+0.4 to +1.8 m above mean
lower low water (MLLW), the bulk occurring at ~+1.3 m (Bayer, 1996). Harrison (1979,
1982) describes Z. japonica as an opportunist with an ability to grow in the subtidal,
and to "escape" direct competition with a taller, light-shading Z marina, by enduring
dessication at low tide at the upper intertidal level. Nomme and Harrison (1991),
studying Zostera spp. distributions just south of Vancouver, B.C., using transplant
experiments, assert that abiotic conditions could not be responsible for the absence of
Z. japonica from deeper zones typically occupied by Z marina, because transplanted
monospecific clones grew well at all cross-transplanted inter- and subtidal elevations
during the growth season. Unlike Yaquina Bay habitats, their study site typically
endures considerably harsher winter exposures, and the Z japonica population is
largely annual, growing from seed each season. Thorn (1990), studying Z. japonica
populations in Padilla Bay, WA (northern Puget Sound), states that it is virtually absent
from the bay in winter. Thorn (1990) also observed considerable overlap of vertical
-------
81
distribution of Z. marina and Z japonica, the maximum biomass of the combined
populations occurring at ~+0.3m above MLLW; peak standing stocks of both species
were centered at about +0.5 m (Z japonica) and MLLW (Z marina).
Yaquina Bay distribution patterns differ from those previously described, in that the
populations do not overlap. Z marina habitats grade from immediately adjacent to Z
japonica populations (typically up-bay), to widely separated intervals (meters to
hundreds of meters laterally and ~0.0 to -1.5 meters vertically - typically lower-bay).
This distribution gap may be due to several factors, among them are wind-generated
wave action in the intertidal area (Phillips and Mertez, 1988), species-specific
dessication and insolation effects (McRoy and McMillan, 1977), predation pressure by
herbivorous waterfowl (Keller, 1963; McRoy, 1966, 1970) and non-herbivorous
waterfowl, which uproot and ingest eelgrass covered with herring eggs (Bayer, 1979),
and differential temperature tolerance of intertidal vs. subtidal populations (Biebl and
McRoy, 1971). The influence of shading by the more physically-imposing Z marina is
not likely to be the operant mechanism in Yaquina Bay populations where, in most
cases, the two are physically separated by unvegetated expanses.
Bayer (1979, 1996) concluded that in Yaquina Bay, Z marina reproduced by seed in
the upper intertidal, while lower intertidal and subtidal beds reproduced vegetatively.
The subtidal limit of distribution of Z. marina in Yaquina Bay may be dictated by
Figure 1. Yaquina Bay Study Site
Newport
RM 2
Yaquina
Bay Bridge
EPA LAB
RM 0 1! Idaho Point
Sally's Bend
Z. japonica
Toledo
Coquille
Point
Sawyer's \\
Landing Criteser's \ |rm li\
Pacific
Ocean
King Slotfdn
Oregon Boone
Oyster Slough
Rive rB end
Z japonica
& X^Oyw^'le \
Craigie Pt.
Z. marina
2. japonica
N
Yaquina Bay
Oregon
-------
insufficient light penetration and/or physical disturbance (episodic hydraulic erosion of
substrate along dredged or natural tidal channels due to storm or wind events). Bayer
(1996) further inferred that because of its higher intertidal elevation, the ability of Z
japonica to persist throughout the winter season could be due to lack of the uprooting
suffered by Z marina from wind-generated wave erosion in the intertidal and less
intensive waterfowl grazing. Whereas Z japonica is reported to exist as an annual,
reproducing by seed in other Pacific Northwest estuaries (Thorn, 1990; Nomme &
Harrison, 1991), it persists overwinter as a perennial in most habitats in Yaquina Bay
(Bayer, 1996, Specht, pers. obs.) in apparent good health, although declining in
biomass during the winter. Phillips and Menez (1988) conclude that Z marina "is
known to remain active all winter along a broad latitudinal gradient in North America",
and that, "Above 22 °C the plants either produce flowers and seed, becoming annual, or
become moribund. According to Phillips et al. (1983), flowering is a response to
warming water temperatures which interact with local genotypes". Bayer's (1996)
observations might be interpreted to conclude that the intertidal exposure during
summer months in Yaquina Bay provides enough warming to provide such stimulus for
flowering, whereas subtidal populations, which are subjected to tidal prism flooding
influenced by cooler waters provided by summer upwelling, are inhibited from
flowering.
Kentula (pers. obs., May 1, 1998) observed Z marina flowering at several subtidal
locations in the upper bay, at Craigie Point (~RM 9), Oysterville (~RM 7) and the west
shore south of King Slough opposite Sawyer's Landing (~RM 4) (where a specimen
was collected), but no flowering was observed seaward of RM 4. The surface water
temperature at Sawyer's Landing was 14°C. The maximum surface water temperature
(1 m depth) on May 1 was <14.5°C, averaging -12.5 °C. The average temperature in
the prior month was ~12.5°C, with a range of 10-14°C (recorded at the HMSC small
boat dock, RM 2). Water column temperature varies seasonally, with colder water
during the winter rainy season draining off the watershed (typically -8 °C upstream,
-9.5 °C downbay); summer wind-induced upwelling reduces temperatures in the lower
bay subject to the tidal prism (typically -22 °C upstream, -13 °C down-bay). Periodic
anomalies, such as "el nino" weather, cancel this pattern during the summer (no
upwelling), with lower bay temperatures on the order of -20°C; severe events are
relatively infrequent. Historical records (Specht, unpublished data, 1976-78) for that
zone show surface water temperatures not exceeding ~13°C during March and April
periods; surface temperatures in late June did not exceed 20°C at Criteser's Dock (RM
11) or seaward. Surface temperatures approached 20-22°C from RM 11 or landward
only during the period of July-August; lower bay temperatures average 12-14°C during
the same period. The observations suggest that the guidelines of Phillips and Menez
(1988) on flowering of Z. marina may require further examination.
Thum (1972), investigating the ecology of an estuarine flatworm in Yaquina Bay
adjacent to the Hatfield Marine Science Center (intertidal flats - RM 2), characterized
-------
83
the production of macrophytes (to include Z. manna, Enteromorpha tubulosa, E.
intestinalis and Ulva "angusta") as approaching "100 gm/30 cm2" (dry weight basis) (-
1.1 kg dw/m2, assuming Thum meant a 30 cm x 30 cm plot). He observed that while
stations in the upper intertidal were "devoid of macrophytic material by January,...
young Zostera marina, 1 to 2 cm in length were present a month later. The Zostera at
the lower stations (-2.0 and 0.0 feet) was cropped severely by the black Brant (Branta
nigricans) during the winter, allowing sediment erosion to proceed." Notwithstanding
this observation, his data indicate that the subtidal populations maintained a minimum
biomass level of ~"20 gm/30 cm2" (- 0.22 kg dw/m2, using the previous assumption)
over winter. Thum also stated that permanent, mature Zostera beds occurred below
0.0 feet, while seasonally transitory beds occurred between 3.0 and 0.0 feet. Kentula
(1982) measured Zostera marina maximum standing crop in Netarts Bay, a similar
environment, as ranging from a low of 143 g dw m"2 (high intertidal) to a high of 463 g
dw m"2 (low intertidal). She reported net annual production, including above- and below
ground parts, as 3.1 kg dw m*2 yr"1.
A variety of factors may interact to explain spatial distribution patterns of seagrasses.
That Zostera japonica flourishes in the high intertidal while Z. marina does not may
indicate that Z. japonica is more tolerant of high levels of insolation and desiccation
during daylight low-tide exposures. Alternately it may be more tolerant of exposure to
freshwater flows off the land than Z marina, although Z. marina is known to be
remarkably tolerant of low (~6 PSU - practical salinity units) salinity (Phillips and
Menez, 1988).
Controlling factors may differ at each end of the estuarine depth gradient. Since much
of the lateral growth of beds of Zostera marina is by vegetative extension of roots and
rhizomes, growth of shoots at any given depth descending a steep gradient may not be
light-dependent, as nutrition would initially be supplied by the parent plant through the
rhizomes until the leaf blade growth is sufficient to reach into the euphotic zone. The
physical environment at the lower edge of distribution, being on the edge of a channel,
is likely quite severe compared to life on the intertidal flat. Although the average
channel current speed is approximately 1.5 knots (Goodwin et at. 1970), episodic runoff
events approach 15 knots (Callaway and Specht, 1982), which could induce substantial
bedded sediment erosion, especially in low-stability sediment; high-flow induced
turbidity is known to suppress eelgrass photosynthesis (Stephan and Bigford, 1997).
Nomme and Harrison (1991) report that the ability of Z. marina to resume growth after
transplanting to subtidal depths was reduced with increasing depth, and assumed this
was due to reduced light availability. While light reduction may be the ultimate cause
of growth reduction, isolation of the transplant plug may be significant. Isolation from
an extensive network of rhizomes of a parent plant will prevent transfer of nutrients,
resulting in growth inhibition of plugs at depths where connected plants might grow
successfully.
-------
84
Knowledge of the potential role of nutrient distribution patterns in directly influencing
seagrass distribution patterns in west coast estuaries is limited. Estuarine circulation
patterns may be important. Kentula (1982) describes circulation patterns in Netarts
Bay that result in two distinct water masses: "bay water" and "ocean water". Little
mixing occurs between the two water masses within single tidal cycles. Factors
affecting the degree of mixing include substantial differences in temperature and
salinity. The clear implication is that these circulation pattern phenomena (typical of
PNW estuaries subject to upwelling influence) can reflect the sources of and affect the
distribution of nutrients for the estuarine primary producers.
It is likely that nutrient distribution patterns along the estuary have changed
considerably in the last 100 years. A significant proportion of the nutrient supply in
Yaquina Bay in earlier times (Hate 1800's through -1980) may have come from
disposal of untreated and digested process waste from the Georgia Pacific Kraft Paper
Mill and predecessors, sawmill wood processing waste (e.g., sawdust, bark), leachate
from failing shore side septic tank drain fields (e.g., from Coquille Point upriver to the
city of Toledo), and normal and storm overflow from the chronically under capacity
municipal sewage treatment plant (STP) at Toledo. Due to repairs and upgrading of
septic tank drain fields bordering the estuary, improvements of the domestic sewage
treatment plant at Toledo and industrial effluent treatment systems at the Georgia-
Pacific Mill since the 1980's, this seasonally variable fraction of the supply of nutrients
to the estuary may have significantly declined (although significant episodes of rainfall-
related STP combined overflows still occur at Toledo with some regularity).
Winter and spring growth of both Zostera sp. in Yaquina Bay is likely sustained by
sediment-bound and microbially-generated N, in addition to runoff-derived nutrients,
but growth of most primary producers is limited by light availability (Davis, 1982). Late
summer growth of Zostera marina, epiphytes, phytoplankton and macroalgae may be
largely fueled by additional N and P contained in the coastal upwelling water in the
flooding tidal prism, and ammonia generated from infauna feeding on plankton and
detritus (Reusch and Williams, 1998). Zostera japonica, with its location higher in the
intertidal, may have less opportunity for extraction of nutrients from the flooding tidal
prism than Z marina. It may thus be more efficient in intercepting ground-water borne
nutrients flowing across the exposed upper intertidal or arising from land-based
seepage, or in obtaining nutrients directly from recycled excreta from waterfowl feeding
in the area. In the maritime climate of Yaquina Bay, Z. japonica is exposed to rainfall
inundation for more extended periods of time during winter-spring months than Z
marina because of its elevated position in the upper intertidal. The relatively high
levels of excrement-derived nutrients from migratory wildfowl present in the estuary
during the fall months (Bayer, 1979, and Thum, 1972) will tend to cycle as slowly
converted refractory material in the sediment, in addition to nutrients supplied by
populations of resident wading birds, fish, invertebrates and marine mammals (e.g.,
harbor seals). Much of this material, because it occurs in relatively shallow water, is
-------
85
probably retained in the intertidal during tidal ebb, and may be tidally pumped upstream
as either dissolved or particulate matter as it is released along the tide interface, and
thus may also influence upstream distributions.
Zostera marina, however, may be able to compete in relatively N-poor sediment areas
because of a difference in microbial associations that provide N-fixation, and/or
scavenge critical nutrients from the salt wedge during upwelling. Additionally, episodic
river floods supply nitrogen off the watershed. The results of such floods are seen most
clearly in the "spike" in primary production following the first fall rains and in the growth
response in populations of phytoplankton-consuming bivalves (e.g., Macoma nasuta)
(Specht and Lee, in prep.). Laboratory phytoplankton biostimulation bioassays
(Specht, 1974, 1976, and unpublished data, 1975-76 samples) also suggest such rapid
responses to nutrient pulses. These nutrient sources should be relatively independent.
In order to assess the degree in which physical factors may be controlling plant
distribution, it is necessary to have indicators of plant stress. In addition to
measurements of overall plant growth reflected in shoot counts or biomass
measurements, Phillips & Mefiez (1988) cite a number of studies in which leaf width
appears to be an indicator of stress (Phillips and Lewis, 1983; McMillan, 1979;
McMillan and Phillips, 1979). Setchell (1920, 1929) noted that leaves of Z. marina
were narrower in the intertidal than subtidal, and that leaves are narrower in winter than
in summer. Setchell suggested that these narrower leaves were due to stress from
greater annual or even tidal exposure or ranges of temperature. Ostenfeld (1908)
recorded narrower subtidal leaves when the plants grew in sand than in mud. In this
substrate example, the stress could also be a nutrient one. Field transplants across
gradients have confirmed some of these observations. In certain cases populations
show phenotypic plasticity and adapt readily to new sites with an accompanying
increase or decrease in leaf width. In other cases populations show little change in leaf
width and are said to be genotypically differentiated. It is thought that the latter
populations are native to stressed locations (Phillips & Menez, 1988).
If we are able to show what mechanisms are involved in restricting the distribution of
the exotic Zostera japonica to the upper intertidal in this habitat, we may be able to
isolate those mechanisms eligible for control strategies that would be effective in
restricting the establishment of other non-indigenous invasive plant species, such as
the smooth cordgrass Spartina alterniflora, a significant threat to the ecological health
of PNW estuaries.
Objectives: The overall goal of the research effort is to determine factors responsible
for the differential distribution patterns of the exotic seagrass Zostera japonica versus
the native seagrass Zostera marina. Because other research at Coastal Ecology
Branch (CEB) will focus on the role of light availability in the water column on seagrass
-------
86
distributions patterns, the focus of my research will be on nutrients as a potential factor
affecting seagrass distribution. Specific objectives and hypotheses to be examined are
described below.
Objective 1. To compare the sediment nutrient (N species, P) concentrations for
Zostera japonica and Zostera marina habitats within the root/rhizome horizon in
selected sample sites within Yaquina Bay.
Hypotheses 1.1 - Sediment nutrient (nitrate-N, ammonia-N, soluble reactive
phosphate) concentrations and their bioavailability to resident SAV are not significantly
different between populations of Zostera marina and Z. japonica at equivalent locations
along the salinity gradient from Yaquina Bay Bridge upstream to the vicinity of Boone
Slough. Factors which may affect this bioavailability include a) sediment temperature
regimes, b) infauna distribution and composition and c) nutrients from land-margin/high
marsh groundwater drainage. These factors are addressed below.
Objective 2. To compare the a) water column nutrient (N species, P) concentrations,
b) PAR penetration through the overlying water column, and c) sediment temperature,
sediment grain size distribution, and salinity of interstitial water within the root/rhizome
mat among selected Yaquina Bay Z. japonica and Z. marina beds. Sampling will be on
a monthly basis over at least one calendar year, with higher frequency sampling
(weekly) at selected sites during the period of maximum growth (May through June).
Site-specific values of water column parameters will be compared to longer term
records of water quality parameters across the estuarine gradient using moored
conductivity-temperature-depth sensors (CTDs), at surface or bottom locations axially
distributed upstream from the OSU dock (~RM 2). Parameters to be measured will
include salinity, temperature, turbidity, tide level, and, in selected installations,
dissolved oxygen, PAR light intensity, and in-situ fluorometric determination of
chlorophyll "a" concentration. To benchmark current conditions, comparisons will be
made to water column nutrient concentrations from a 1976 survey of the Yaquina
estuary (Callaway and Specht, 1982; Callaway, Specht and Ditsworth, 1988). Specific
hypotheses are:
Hypotheses 2.1 - Surficial sediment temperature ranges and averages are not
significantly different between equivalent locations for populations of Z. marina and Z.
japonica along the estuarine salinity gradient.
Hypotheses 2.2 - Salinity, temperature, and light availability to the seagrass
canopy are not significantly different in the water-column at tidal extremes between
populations of Z. marina and Z japonica at similar locations along the estuarine salinity
gradient.
-------
87
Objective 3. To compare the growth, survival and reproductive success of reciprocally
transplanted specimens of Z japonica and Z marina over a minimum two-year period,
to demonstrate potential physiological differences in local habitat-adapted
monospecific clones of the two species. Stress measures may include reduction in
blade width compared to control, time of onset of flowering and seed set, and shoot
density and mean shoot length.
Hypotheses 3.1 - Reciprocal transplants of monospecific clones of Z. japonica
and clones from MLLW elevation of Z marina will show no difference in growth or
"stress" measures at equivalent locations along the estuarine salinity gradient.
Methods:
Objective 1 - A - Using Yaquina Bay aerial photography from 1997 (Young et al. in
prep.) And field survey, we will locate appropriate sites of monospecific populations of
Zostera marina and Z japonica to sample for sediment nutrient concentrations,
sediment characterization, and sources of transplantation clones. I propose to
establish lateral transects at three locations at which both Zostera species co-occur (at
MHHW and MLLW): the north shore at Sally's Bend, the north shore at RiverBend, and
the south shore south of Craigie Point (Fig. 1).
A survey of the sites will be conducted by the PI or an outside contractor to establish
the vertical elevations of the sampling sites to an appropriate tolerance level. Vertical
zonation is perceived as critical to establishment and maintenance of these populations
(see D. Young, CEB, proposal). This survey will be conducted in cooperation with on-
site contractor and EPA staff concerned with geographic information system (GIS),
geodetic control, and photogrammetric aspects of the FY97 Yaquina Bay mapping
exercise (Young et al., in prep.).
Objective 1 - B - Sediments will be sampled within the selected habitats and analyzed
for N and P species (nitrate-nitrogen, ammonia-nitrogen, ortho- or soluble reactive
phosphate) in both interstitial water and that fraction sorbed to sediment particulates.
Objective 1 - C - Infaunal distribution and abundance will be addressed in the DeWitt et
al. proposal surveys, and data compared to Boese et al. (1997,1998) flood survey site
results, where coincident with Zostera japonica occurrence.
Objective 1 - D - Interstitial water will be sampled periodically at the upper intertidal
margin in sediment a depth of ~5 cm. at the upper terminus of transplant transects for
N-P analyses.
Objective 2 - A -We will deploy CTDs at strategic locations in the estuary to
characterize the salinity and temperature structure over time. These instruments will
undergo quality assurance/quality control (QA/QC) checks and will be deployed,
-------
88
operated and maintained by the on-site contractor, at the technical direction of the PI
(see also Ozretich and Young research proposals). Stations are proposed at the
following locations:
1) OSU pump dock (-RM 2), at which a surface unit (i.e., attached to float dock to read
1.0 m below the surface) and a bottom unit (located at approximate bottom of the
euphotic zone, i.e., the mean lower level of Zostera marina beds), this unit will
also have a PAR light sensor attached, a fluorometer and turbidity sensor; the
surface unit will also measure turbidity.
2) Between Idaho Point and Sawyer's Landing (- RM 3.5), at which a surface unit with
a turbidity sensor will be deployed to float 1 m below the surface.
3) RiverBend (-RM 5.5), at which a surface unit with a turbidity sensor will be
deployed to float 1 m below the surface.
4) Oregon Oyster, or Oysterville (-RM 7), at which a surface unit with a turbidity
sensor will be deployed to float 1 m below the surface.
5) Criteser's Moorage (~RM 11), at which a surface unit with a turbidity sensor will be
deployed to float 1 m below the surface.
6) One unit equipped with a PAR light sensor and a fluorometer will be reserved for
vertical casts, to profile the water column at optimal times preceding and
immediately following episodic climatic events (heavy rain and flooding, for
instance).
Objective 2 - B - Sediment temperature over time will be recorded by in-situ placement
of temperature loggers at ~5 cm depth along each transect, with an additional transect
at the west bank south of King Slough ("Racoon Flats"), where Zostera marina grows
all the way up through the intertidal to the MHHW line, and a transect on the north
shore of Idaho Point. Initially, five loggers will be placed on the Sally's Bend transect,
three at RiverBend, three at Craigie Point, two at Racoon Flats, and two on the north
side of Idaho Point. The distribution is designed to measure sediment temperature
variance during tidal submersion and intertidal emersion in a number of different
exposures. We will also read instantaneous sediment temperature at the root/rhizome
level (~5 cm depth) during sampling operations
Objective 3 - At three locations along a salinity gradient, we will transplant
monospecific clones of Zostera japonica and Z. marina (from MLLW habitat)
reciprocally on a lateral transect to determine survival, growth and reproductive
success in non-native habitats, monitoring the transplanted clones over at least a two-
year period. Transplanting should be accomplished prior to the maximum growth
season (May). Transect lines will be located by Geographic Positioning System (GPS)
and marked. Elevations will be determined subsequent to transplanting activity by GPS
and laser range-finder in conjunction with standard surveying techniques. Reaction to
stress will be measured by determining significant differences in leaf blade width in new
growth immediately above the intercalary meristem, shoot density and mean shoot
-------
89
length on a periodic basis (Nomme and Harrison, 1991).
Expected Results and Benefits: Determination of the physical parameters
characterizing the present habitats and sources of nutrient supply for the different
populations of Zostera species could allow the potential for control strategies of the
non-indigenous Z. japonica, if that is determined to be a desirable end. Such
knowledge may be of significant import in controlling or preventing the establishment of
other potential invading species, such as Spartina alterniflora (smooth cordgrass, a
significant problem in a growing number of west coast estuaries), which could utilize
the same nutrient source or tolerate the physical environment. Success in determining
levels of stress in local populations of Zostera spp. could prove a useful tool in
evaluation of declining SAV populations.
-------
90
Project A6 - Spatial Variation of Growth and Condition in Juvenile English Sole
Relative to Substrate Characteristics
Principal Investigator: James H. Power
Goal: The goal of the proposed research is to determine how the growth rates of an
important estuarine flatfish, the juvenile English sole (Pleuronectes vetulus, formerly
Parophrys vetulus), varies in time and space, and particularly to relate these patterns to
spatial variations in the sediment substrate of Yaquina Bay. I will further interpret this
information with respect to the consequences of natural or anthropogenic alterations in
the Yaquina Bay substrate.
Estuaries, as the region of transition between rivers and the coastal ocean, are
influenced by the combined action of freshwater inflow, tidal processes, watershed
features, and anthropogenic effects. The result is an environment where important
physical, chemical, and biological parameters can vary greatly in both space and time
(Virnstein 1990; Moore et al. 1996; Jassby et al. 1997). Any natural or anthropogenic
process that alters those spatial or temporal patterns, or imposes greater variability, will
in turn affect the ecology of the residents and their role in the estuarine ecosystem.
Recent studies have begun to highlight the importance of spatial patterns in the benthic
flora and fauna of estuaries, and the causative agents that might give rise to those
patterns (Kneib 1994; Brandt and Mason 1994; Deegan and Garritt 1997; see also the
Journal of Experimental Marine Biology and Ecology Volume 216, Nos. 1-2 which is
devoted to a series of papers organized under the theme "The ecology of soft-bottomed
habitats: Matching spatial patterns with dynamic processes"). With regard to estuarine
fishes, Sogard (1994) discusses how a number of factors, such as competition,
predator avoidance, or habitat degradation, might act to constrain estuarine fishes to
suboptimal habitats. This proposal concerns the association between estuarine-
watershed processes, as reflected by the sediment substrate composition, and the
proximal growth and condition of an important estuarine flatfish that lives in association
with those sediments during part of its life cycle.
This research is being proposed both for its intrinsic scientific merit and because the
EPA Office of Research and Development (EPA-ORD) has identified "Research to
Improve Ecosystem Risk Assessment" as a high priority research topic (EPA 1997).
This EPA-ORD Strategic Planning document lists the four components to ecosystem
risk assessment as: 1) Hazard identification; 2) Dose-response assessment;
3) Exposure assessment; and 4) Risk characterization. The research proposed here is
intended to identify a heretofore little recognized hazard to an important estuarine
species: the relationship between variations in bottom type and the important
physiological process of growth. It will do this by undertaking a field-based
"dose-response" assessment of the effects of the hazard. Both the identification and
assessment can be then used to characterize the risks associated with alterations to
-------
91
the estuary and its surrounding watershed. It is then expected that the results of this
work will contribute to a regional scale "exposure assessment", whereby the
consequences of natural or anthropogenic alterations to the distribution and
composition of estuarine sediments might be anticipated.
Rationale: The growth of juvenile fish has been extensively studied because of its
importance in fish population biology. A fish's rate of growth is perhaps one of the most
important indicators of its condition (physiological health), and the growth rate in turn
reflects the relative suitability of the fish's local environment. Fish that are growing at
comparatively higher rates can do so because they are obtaining a sufficient amount
and quality of food, and they do not have to divert much energy towards coping with
external stressors. Conversely, fish that are not growing well are either not locating
sufficient food resources, are being subjected to some energetically taxing stress, or
both. The rate of growth is believed to be especially important for the larval and
juvenile (pre-reproductive) stages in fish. Rapid growth to larger size in larvae and
juveniles is beneficial, because with growth the larger fish then become unavailable to
predators that cannot successfully handle them. Growth may also confer greater
foraging ability or the ability to avoid stressors.
The results provided by DeAngelis et al. (1993) can be used to illustrate how spatial
variations in growth can have important implications for a fish population. They provide
an analytical solution to a partial differential equation that expresses how the size-
frequency distribution of a fish cohort develops over time under specified growth and
mortality conditions. In this model (their "Case 1") all fish begin growth at the same
time, and at the same size. An important feature of the model is that fish mortality rate
is inversely proportional to fish size. Each of the individual fish grows at a constant
growth rate, but it's assumed that growth rates vary among the fish, being normally
distributed with mean g0m and standard deviation b. Below are the population
frequency distributions that are predicted by this model after 60 days of growth. These
curves were created keeping all parameter values provided by DeAngelis et al. (1993)
constant, except for the standard deviation of the growth rate (b).
-------
92
50d
40.
b = 0.16
= 30.
o
| 20.
3
Z
b = 0.32
10.
25
35
45
55
65
75
85
95
105
115
5
15
Length of fish (mm)
Note that when variability in growth rate is increased (to b= 0.32) there is a greater
proportion of larger fish in that population, and a larger mean fish size. Further, there
are more fish surviving: the curve for £*=0.32 represents 911 fish, while there are 506
and 596 survivors for 6=0.08 and b=0.16 respectively. This result occurs in spite of the
fact that the mean growth rate was identical and constant for all three of these
"populations". What occurs is that the few fish fortunate enough to have higher growth
rates (those from the right hand tail of the growth rate distributions) can capitalize on
this advantage by the combination of rapid growth and by escaping the higher mortality
at smaller sizes. The vertical line at a size of 60 mm in the graph is intended to
illustrate the further consequences of variability in growth rate. Hypothetical^, this
could represent the size at which an important size-dependent life history event might
occur, such as spawning or emigration from an estuarine nursery area. Variability in
growth rate again has important implications for the numbers of fish that achieve this
critical size in a timely manner.
These theoretical calculations underlie the premise of the research proposed here: that
the magnitude of growth rate variability within a real-world estuary, possibly occurring
because of spatial patterns in the substrate and food resources available, has a
significant impact on the real fish population. These ideas are in accordance with
those expressed by Underwood (1991), who points out that a natural or anthropogenic
stress on an ecosystem may not necessarily alter the mean value of an ecological
parameter. Instead, such stress may affect the spatial or temporal variability of that
parameter (in this case fish growth rate), with important consequences for the
population and ecosystem.
One such ecological parameter of importance to benthic fish is the substrate they live
upon and from which they obtain their food. It is well known that a wide variety of
natural and anthropogenic processes act to alter the bottom sediment characteristics in
estuaries, and this is true of Pacific Northwest estuaries. Changes in the watershed
that affect sediment distribution include the clear cutting of timber and other agricultural
-------
93
practices, alterations of the stream flows, or changes in land use such as residential
development and urbanization that affects runoff and erosion of the adjoining land.
The proposed research will be done in Yaquina Bay, Oregon, and the Bay's channel is
periodically dredged, which also redistributes the sediments and affects the circulation
patterns. However, detailed, historical data on the spatial distribution of infaunal
species and the associated substrate characteristics of Yaquina Bay appears to be
lacking. Data on grain size distribution at 88 sites in Yaquina Bay were provided by
Bruce Boese (U.S. EPA) in the form of the percentages of sand, silt, and clay. These
data were summarized in a principal component analysis, and the first principal
component axis accounted for 96.6% of the variability in the data. As a result, the first
principal component score of each location in Yaquina Bay is a convenient scalar to
summarize sediment types in Yaquina Bay, and these are plotted below with the size of
the circle being proportional to the component score.
Sally's Bend
Idaho Flats
Smaller circles in this plot indicate locations where the sediments have a high sand
content, while larger circles indicate sediments that are mixtures of silt and clay. Note
that the intertidat sediments in the Idaho Flats region are dominated by sand, while the
substrate directly across the channel in Sally's Bend has substantial percentages of silt
and clay. These variations in substrate composition can in turn be expected to
influence the abundance and persistence of the infaunal organisms that are available
to English sole as food. J. Chapman (Oregon State Univ.) and B. Boese (pers.
commun.) have indeed noted considerable small scale variations in the abundances of
infaunal amphipods (Corophium sp.) sampled within Yaquina Bay. Similar small-scale
patchiness is also apparent in the distribution of submerged aquatic vegetation, such
-------
94
as Zostera. English sole, however, are widely distributed in the Bay and can be
captured at essentially all locations in lower Yaquina Bay (De Ben et al. 1990).
In spite of the spatial structuring evident in estuaries, there have been comparatively
few studies that have actually examined differences in fish feeding and growth rates
wholly within an estuary, especially at spatial scales of several kilometers or less.
Shaw and Jenkins (1992) found that feeding conditions for the flounder Rhombosolea
tapirina were much better in a small embayment than they were in more open waters
that were several kilometers away. The flounders in the better environment were
feeding at a rate 3-4 times greater than that of the poorer location. Using caging
experiments, Sogard (1992) found differences in growth, measured as change in fish
length over 15 days, of three species at four New Jersey locations separated by
roughly 15 km. Her study sites varied in bottom type and in the amount and kind of
vegetation, but vegetation did not appear to be related to the growth of two of the
species. Working at a larger scale, Sogard and Able (1992) found that winter flounder
at one of four locations on the New Jersey coast were growing at a lower rate
compared to the other three sites. They calculated flounder growth rates from otolith
increment counts and widths. Berghahn et al. (1995) also used otolith microstructural
analysis to find that plaice (Pleuronectes platessa) juveniles grew faster in a food-rich
area when compared with a food-poor area about 5 km distant. In a controlled
laboratory experiment Gibson and Batty (1990) did not find a direct effect of substrate
type on plaice growth rates. However, they do emphasize that under field conditions
the substrate will affect the food resources available to this flatfish, and so will have an
indirect effect on its growth. The premise of the proposed research, suggested by
these studies, is that variations in the substrate and immediate environment occupied
by benthic fish can be regarded as a stressor, or indirect (proxy) indicator of associated
stressors, particularly reduced food resources, that affect the fish's recent growth rate.
The adult English sole occupies coastal waters, but the larvae and juveniles utilize
Pacific coast estuaries as nursery grounds (Boehlert and Mundy 1987; Gunderson et
al. 1990; Chamberlain and Barnhart 1993). English sole enter Yaquina Bay as larvae
beginning in late February, and benthic juveniles later emigrate from the estuary
beginning in late May after overwintering one season (Boehlert and Mundy 1987).
Previous studies have shown that English sole are widely distributed throughout
Yaquina Bay, and can be captured at locations ranging from the dredged central
channel to mudflats that are exposed at low tides (Westrheim 1955; Toole 1980;
Krygierand Pearcy 1986; De Ben et al. 1990). Although direct study of juvenile English
sole substrate preferences have not been reported, Becker (1988) did find that adults
in Puget Sound appear to segregate by sex depending on bottom grain size
characteristics. As a flatfish, the juvenile English sole lives in close association with
the bottom sediments, and food habit studies have shown that it relies on infaunal
organisms for a majority of its food (Hogue and Carey 1982; Becker and Chew 1987;
Barry et al. 1996). Growth rates of juvenile English sole have been reported in a
-------
95
number of studies (Misitano 1976; Williams and Caldwell 1978; Rosenberg 1982;
Rosenberg and Laroche 1982; Kreuz et al. 1982; Laroche et al. 1982; Shi et al. 1997),
but none of these have reported growth rates relative to bottom type, or have made
explicit comparisons among specific localities. The English sole's comparatively
sedentary habit, numerical abundance, and widespread occurrence at a variety of
within-estuary locations, all combine to make it an ideal subject for this study.
There are several approaches to evaluating fish growth. Mark-recapture methods are
often used with adults, but are impractical with juveniles that are not the target of a
fishery. Observation of the progression of modal values in length-frequency
distributions has also been used to evaluate English sole growth rates, but that
approach has coarse resolution and does not permit spatial comparisons. The
principal approach used in this study will be a biochemical analysis offish muscle
tissue, specifically an evaluation of the animal's relative concentrations of RNA and
DNA. The DNA content of an animal's tissues remains constant. However, the RNA
content of that same tissue will vary according to the amount of protein synthesis that
was being done at the time the animal was sacrificed. Hence, a measure of RNA
content, scaled by the DNA content of the same tissue, provides a snapshot of recent
animal protein synthesis rate and nutritional status (see Ferron and Leggett 1994 for a
recent review). The RNA:DNA ratio was used by Malloy and Targett (1994a, b) to
compare growth of juvenile summer flounder from Delaware and North Carolina.
Although data are limited, the time course of changes in fish RNA:DNA ratios in
response to changes in food is on the order of several days to a week, depending on
the species and its developmental stage. For example, Clemmesen and Doan (1996)
observed a significant change in larval cod RNA:DNA ratios after 5 days of starvation
immediately after yolk absorption. Grant (1996) noted a significant decline in adult
brown trout RNA:DNA ratios after one week in fish given rations of 1% body weight day
1 when compared to fish receiving 5% of their body weight day"1.
The RNA:DNA ratio provides a measure of the fish's recent growth performance. It is
of considerable interest to determine whether the fish otoliths might provide a more
permanent record of fish growth rates. Daily increment formation has been reported in
English sole otoliths (Rosenberg 1982), and these provide a measure of the time that
has elapsed during their deposition. Measures that relate otolith to somatic growth
have used either overall otolith size, or the comparative widths of the otolith
increments, as indicators of fish somatic growth. However, there is continuing debate
as to whether otolith and somatic growth in fish are coupled or decoupled, and whether
the relationship is linear or nonlinear (Campana 1990), and in fact slow growth in some
fish results in larger otoliths. Investigators such as Folkvord et al. (1997) assert that
there is an otolith-somatic growth relationship in herring (Clupea harengus). However,
others such as Mosegaard (1988), Milicich and Choat (1992), and Fitzhugh et al.
(1997) have noted that temperature variations will affect, and possibly obscure, any
otolith-somatic growth relationship. Wright et al. (1990) also did not find any relation
-------
96
between somatic and otolith growth in salmon parr, and further argued (Wright 1991)
that salmon otolith growth is more closely related to fish metabolic rates, as measured
by oxygen consumption, rather than somatic growth. Molony and Choat (1990) noted
there may be a considerable (ca. 15 days) lag between reductions in somatic growth
and otolith growth, because a fish placed under reduced rations may first mobilize
stored food reserves for an extended period, allowing it to sustain otolith growth for
some time before the effect of reduced food resources is reflected in the otolith growth
pattern. A key assumption in using otoliths to indicate fish growth is that increments
are formed daily, but another confounding difficulty in this usage is that slow growing
fish may not deposit daily increments, or the increments may be too narrowly spaced to
be resolved (Bailey and Stehr 1988). It spite of this controversy, it appears worthwhile,
at least on a limited basis, to explore whether English sole otoliths provide an indication
of somatic growth that can be used as an alternative to the RNA:DNA ratio.
Objectives: The objectives of the proposed research are:
1. To examine the spatial and temporal patterns of two indicators of recent growth
rates (the RNA:DNA ratio and otolith increment deposition) in an important
estuarine fish, the English sole.
2. To evaluate those patterns with respect to the fish's immediate environment,
especially the characteristics of the sediment where the animals are collected.
3. To interpret the combined dataset of fish growth and sediment parameters at
larger spatial scales, particularly with respect to anthropogenic effects.
Scientific Approach: Two kinds of field observations will be conducted in this study,
and the first is the periodic collection of free-living fish. Collection of fish will be
coordinated with the research of Ferraro and Cole of the EPA Newport laboratory, who
will undertake sampling to evaluate estuary biota-habitat relationships in Yaquina Bay.
A sampling method that Ferraro and Cole plan to use is the "drop sampler". The drop
sampler is a fiberglass cylinder, five feet in diameter and five feet in height. It is
suspended from the bow of a boat, and the boat is allowed to drift to the sample site so
as to not startle the resident organisms. The sampler is released and allowed to
quickly fall to the bottom, enclosing a sample of the water column and the substrate. It
is pushed further into the sediment after deployment, and a pump is used to evacuate
the water from the interior. The organisms are then collected from the interior, so that it
captures a quantitative sample of the resident epifauna and infauna. The drop sampler
was compared with other sampling methods by Rozas and Minnelo (1997), and they
advocate its use as the most effective gear for sampling shallow estuarine habitats.
Ferraro and Cole plan on deploying the sampler in a variety of contrasting habitats in
Yaquina Bay, and these contrasts will contribute to the objective of this study: to make
spatial comparisons using English sole captured at a variety of locations in Yaquina
Bay. Up to ten individual English sole will be collected at each deployment of the drop
-------
97
sampler, and transported alive back to the laboratory for analysis. Temperature and
salinity will be recorded, and sediment samples will also be collected for later analysis
of grain size distribution and total organic content using methodology reviewed by
Larson et al. (1997). Further, Ferraro and Cole plan to characterize the benthic infauna
as part of their work, and this will contribute information to this project concerning the
food resources available to the co-occurring English sole.
We fully anticipate that the drop sampler will be as effective a sampler in Yaquina Bay
as it has been in other regions, but we are presently unaware of others that have used
it in Pacific Northwest estuaries. It is therefore appropriate to discuss alternative
sampling strategies to be used in case the drop sampler is ineffective for some
unforeseen reason. De Ben et al. (1990) have reported on year-round trawling for
bottom fish in Yaquina Bay at a sequence of stations from near the Bay mouth to nearly
freshwater locations. Their data indicates that English sole were readily captured at
their first six stations, and as an alternative to the drop sampler we may choose to
reoccupy three of those (De Ben et al. stations 1, 3, and 5) on a monthly basis between
May and October. These stations are near, or in, the main channel. In this event we
would also sample three additional stations lateral to the channel, but near to shore,
using a beach seine or trawl depending on tide height and bathymetry. Again, water
temperature and salinity will be recorded for each trawl, and a core of bottom
sediments will be collected for later grain-size and total organic carbon analysis. Up to
ten individual sole will be selected from each trawl, and this selection will be done so
as to collect a range of the fish sizes captured in each sample.
Finally, National Marine Fisheries Service scientists have indicated they will be
conducting beach seine sampling for juvenile salmon in twelve west coast estuaries
(including Yaquina Bay) ranging from San Francisco Bay in California to the Nisqualiy
River in the state of Washington. They have generously offered to provide English sole
they might incidentally capture to this project, and so we will opportunistically examine
fish collected from these other locations as they might become available. If sufficient
numbers of fish are obtained this way then larger scale, regional inferences about
variation in sole growth rates may be possible.
Regardless of how they are captured, the sole will be held in separate buckets
associated with the individual sampling site and returned to the laboratory. They will
be anaesthetized in a 100 ppm solution of tricaine methanesulfonate (MS222), after
which they will be weighed, measured, and sacrificed. Samples of muscle tissue will be
collected and frozen at -80° for later biochemical analysis, and sagittal otoliths will be
removed from the fish for later examination. Some additional fish will also be collected
for experiments described below.
It is not known whether juvenile English sole exhibit any sort of site fidelity, although
anecdotal information indicates that estuarine flatfish do not move much between
-------
98
adjoining habitats (S. Sogard pers. commun.). Continual movements of English sole
between differing substrate types would inhibit detecting any relationship between
bottom type and fish growth. Unfortunately, undertaking a conclusive field assessment
of juvenile English sole site fidelity is problematic. This would ordinarily be done by
conducting a mark-recapture experiment, and the recapture of a significant number of
previously marked fish at or near the same site where they were marked would suggest
the sole had maintained residence in that area over the intervening time. However, the
large areal extent of Yaquina Bay habitat that is available to the sole is in contrast to
the limited areal extent that can be effectively resampled using a trawl or beach seine.
This makes it doubtful that sufficient numbers of sole could be marked, and then
recaptured at a later date, to allow a conclusion about the fish's site fidelity to be made,
even if fish dispersal from the site of marking is minimal.
With these considerations in mind it is also proposed to undertake caging experiments
as the second set of field observations, so that juvenile English sole can be restricted
to selected substrates for known periods of time. This portion of the research will follow
the methodology of Sogard (1992). Cages will be constructed with open bottoms to
allow the fish access to the substrate, and then anchored at two locations of contrasting
substrate composition: one in the Sally's Bend area and the other in the Idaho Flats
region (see previous figure on sediment types in Yaquina Bay). A "skirt" will be sunk
around each cage's perimeter to prevent escape of the fish that might locate an
opening under the bottom edge. Some of the sole collected during the trawling
described above will be held in the laboratory and fed ad libitum for periods no less
than one week, to allow them to equilibrate from their previous growth and feeding
history while in the wild. These fish will be marked by subcutaneous injection of
colored acrylic paint, so that individual fish identities can be known during the
experiments. The otoliths of these fish will also be marked by immersing them in a
solution of oxytetracycline (Secor et al. 1991a). Uptake of oxytetracycline into the
otoliths leaves a fluorescent mark in the otolith that can later be viewed under a
microscope and used as a reference point.
After being held in the laboratory, the fish will be weighed and measured, and one to
five individuals will be placed in the cages and held there for varying periods
depending on anticipated tide heights. Tidal height patterns for Yaquina Bay have
been examined, and a series of windows during the first field season have been
identified where a very low tide is followed by extended periods of higher low tide levels
(corresponding to the times between spring tides). The plot below shows the heights of
lower low tide versus date in the 1998 field season. Fish will be placed in a cage
during a period of especially low spring tides (e.g. late May), and recovered during the
next occurrence of equivalent low water levels, roughly 14 days later.
-------
99
4
3
2
1
0
CB
1
2
-3
0 1 M
D at*
Choosing these dates for caging experiments will help ensure that the cages, and fish,
are not exposed during intervening low tides (of course anomalous weather or river
discharge conditions may alter predicted tide height, but these cannot be anticipated
and countered).
Sample sizes (number offish/cage) will initially be low (one or two per cage), and then
revised upward as we gain an understanding offish growth rate patterns and can be
assured that fish growth is not depressed due to overcrowding of the cages. At the
conclusion of a caging experiment the fish will be recovered, returned to the laboratory,
and processed the same way as the fish captured during the trawl sampling: weighed,
measured, and otoliths and tissue collected. An expectation of this research is that
spatial differences in growth rate among cage locations will be detected. If this is found
to be the case, then the magnitude of the observed growth rate differences will permit
at least a qualitative assessment of how much between-location movement must occur
to equilibrate the fish growth rates between sites, and this information can then further
be compared with the growth rate variations observed among the free-living fish
captured with the drop sampler or trawl.
The quantitation of fish tissue RNA and DNA content from both trawl samples and fish
from caging experiments will follow the methodology developed by Burger (1998; his
protocol is presently being revised to form a SOP). Briefly, the fish tissue is sonicated,
then frozen, both of which disrupt the cells and free the nucleic acids. The sample is
then treated with proteinase to further remove any nucleoproteins bound to the nucleic
acids that might interfere with the assay. At this step the sample is divided into two
equal aliquots. The fluorochrome dye thiazole orange is added to one of these
aliquots. This dye binds to both RNA and DNA, and the amount bound to the nucleic
acids is determined using a spectrofluorometer. The enzyme RNase is added to the
second aliquot to digest the RNA in the sample, leaving the DNA behind. The DNA
remaining in this sample is also determined using thiazole orange. The amount of RNA
-------
100
that was in the tissue is then calculated as the difference between the two
determinations, and the ratio of RNA to DNA content is interpreted as an index of
recent protein synthesis and growth.
Otoliths will be prepared and examined using the standard methodology described by
Secor et al. (1991 b). They will be cleaned, embedded in Spun- media, mounted on
microscope slides, and polished until thin enough so that the daily increments that were
deposited in them can be viewed using a video microscopy setup. The number of
recently deposited rings in each otolith will be counted and interpreted as the number
of days, while the widths of a span of one or more increments will be interpreted as
proportional to the fish's growth during that time period. English sole, like other flatfish,
form secondary otolith primordia when they metamorphose to the benthic form, and the
increments formed around the secondary primordia are especially easy to count
(Sogard pers commun.). Xiao (1996), Schirripa and Goodyear (1997), and Gallego and
Heath (1997) provide mathematical analyses of the process under which otolith and
somatic growth might be linked and recognizable, and their results will be used as
guidance in this study.
The data from these experiments will be examined as follows:
1. The RNA:DNA ratio determined for a fish will be compared with the recent otolith
increment growth of the same fish, to establish the degree of correlation
between them. The two approaches provide different representations of recent
fish growth, and each has its own advantages and limitations. There presently
appear to be only a few studies that have made explicit comparisons between
biochemical and otolith growth measures (Clemmesen and Doan 1996; Suthers
1996).
2. The RNA:DNA ratio and otolith growth information will be evaluated in both
wild-caught and caged fish with respect to both bottom type (location) and date
to characterize the spatial and seasonal patterns of juvenile English sole growth
in Yaquina Bay. It is anticipated the same principal component analysis
approach of sediment samples described previously can be usefully applied to
sediment samples collected as part of this study, i.e. that grain size distribution
can be summarized as a principal component axis score. The sediment PCA
score can then be used as a metric describing the substrate that can be used in
a correlative analysis of that score and sole growth data.
3. The expectation of this research is that, after controlling for season, the growth
rates will differ significantly among locations in wild fish, and it will also differ
significantly among locations in caged fish. If this is found to be the case, then
comparisons between growth rates of caged fish and wild fish caught at the
same location and time will be made with regard to the sign and magnitude of
-------
101
their correlation. If wild fish are remaining near a site, and not moving among
substrate types, then the expectation is that growth between these will be
positively correlated: wild fish caught near rapidly growing caged fish will also
have elevated growth rates, while fish caught near slow growing caged fish will
have a similarly depressed growth rate.
4. The combination of growth rate information and substrate characteristics will be
used to make inferences about how changes in sediment input, dredging, or
other changes to the surrounding watershed that will potentially affect the
substrate, might in turn enhance or diminish the growth of juvenile English sole
while resident in the estuary.
5. Finally it is appropriate to also include an evaluation of the statistical properties
of the observed growth data as an early outcome of the work proposed here. A
detailed statistical design for the above measurements hasn't been provided
because this work represents an effort to measure variability in growth rates, and
since that variability is presently unknown it is not possible to define the sample
sizes necessary for statistical comparisons. The continuing controversy over the
utility of otoliths as a direct measure of growth has been mentioned previously.
Although others have measured the RNA:DNA ratio in fish, the majority of the
published work has been done on larvae and/or as tightly controlled lab
experiments. Because of these considerations the proposed research must
proceed adaptively. The anticipated approach is to continually reevaluate the
data as it is acquired, so that the validity of statistical comparisons (or requisite
sample sizes) can be ascertained as early as possible.
Expected Results and Benefits: The proposed research will examine a vital rate
process in an important fish of Yaquina Bay, and will relate variations in that process to
an attribute of the Bay (substrate composition). Sediment composition is subject to
change from both natural and anthropogenic processes; the changes in substrate
characteristics will affect changes in the English sole food supply that substrate
supports. There is ample evidence in the existing literature that spatial differences in
fish growth rate occurs, but these differences have not been studied at smaller spatial
scales and have not been linked to environmental variations at those same scales.
The methodology proposed for this study is well established, and the combined
application of otolith and biochemical approaches at smaller spatial scales represents a
significant strength of this research.
It is anticipated that the proposed research will document the relative sensitivity of
juvenile English sole to habitat alterations. The substrate composition is viewed as an
indicator of the juvenile English sole's habitat quality. This proposal does not include
studies of infaunal community composition, or English sole feeding habits, in order to
keep the scope of work within manageable limits. However, it is easily recognized that
-------
a suite of environmental factors associated with substrate composition, especially
associated flora and fauna, will have bearing upon English sole growth. As a result,
this study will integrate with companion EPA studies concerning infaunal and
submerged aquatic vegetation community dynamics. Finally, the results of this
research can be utilized by management personnel responsible for decisions
concerning sediment input and alterations in Pacific Northwest estuaries.
-------
103
4.2 Research Theme B. - Stressor-Response Modeling
Project B1. An Evaluation of the Geometry of Stress Using Spatially Explicit
Population Models
Principal Investigator: Henry Lee II
Co-Investigators: Chuck Bodeen, Ted DeWitt
Goals: The long range research goal is to generate insights into how marine/estuarine
populations respond to single and multiple anthropogenic stressors, including the
interactions among the spatial and temporal patterns of the populations and the
stressors. Multiple stressors denotes both multiple sources of the same stressor and
the combined effects of multiple chemical and/or non-chemical agents. The primary
objective is to evaluate the effects of the spatial distribution of single and multiple
sources of a stress on the qualitative patterns of population dynamics of a
representative benthic resource species.
Rationale: A frequently stated maxim is "dilution is the solution to pollution." Based on
this strategy, tens if not hundreds of millions of dollars have been spent designing and
constructing high performance diffusers for municipal and industrial discharges.
Maximizing dilution is inherent in several sections of the Clean Water Act (e.g., Section
301H) that compare ambient water concentrations with Water Quality Criteria at the
edge of a zone of dilution. However, dilution is not considered the solution to all
pollution. Dispersal of plutonium is rarely, if ever, suggested as the solution to
radioactive waste disposal. The management strategy for dredge materials is to
minimize loss during dredging and disposal (see U.S. EPA/ACE, 1991). While a few
dispersive sites exist (e.g., the Alcatraz dredge disposal site), the potential for
containment is a prime consideration in choosing most dredge disposal sites,
especially for materials containing sediment contaminants. Similarly, a containment
strategy is inherent in most of Superfund's remedial actions.
The point is that two diametrically opposed disposal strategies have been used in the
management of anthropogenic stressors in marine and estuarine environments. These
two strategies potentially differ dramatically in cost and in level of protection afforded to
ecological resources. For example, the costs and near-field versus far-field ecological
effects of a sewage outfall designed to minimize diffusion would be considerably
different from one with a high performance diffuser. Nonetheless, there does not
appear to have been a systematic evaluation of "dispersal" versus "containment"
strategies. One key component of such an evaluation would be to predict the effects of
different strategies on the population dynamics and sustainability of high value species,
the assessment endpoints (U.S. EPA, 1991). This research addresses a subset of this
problem by posing the question, "What is the relative risk to populations of benthic
resource species exposed to a dispersed versus a contained stressor?"
-------
104
The proposed use of population assessment endpoints differs from most
marine/estuarine risk assessments that are usually conducted using indicators, or
measurement endpoints (U.S. EPA, 1991), at the individual level (e.g., toxicity
benchmarks) or the community/ecosystem level (e.g., benthic community structure). In
part, this is due to the difficulty in quantitatively sampling populations. Population
models are one potential method of addressing this problem both by integrating
available knowledge and as an exploratory tool. Besides the obvious example of
fisheries models (e.g., Ricker, 1975; Hilborn and Walters, 1992; Higgins et al., 1997),
population endpoints have been used to address a variety of anthropogenic stressors.
Population models have been used to predict the effects of contaminants or thermal
pollution on fish populations (e.g., Van Winkle, 1977; Power and Power, 1995; Black et
al., unpublished), population demographics have been used in laboratory toxicology
studies (Gentile et al., 1982; Levin et al., 1996; DeWitt, unpublished), and simulation
studies have been used to predict the effects of individual and multiple stressors on fish
populations (Barnthouse et al., 1987, 1989, 1990; Marschall and Crowder, 1996).
A standard assumption of most population models, such as the ones mentioned above,
is of a single, well-mixed population. This approach is inappropriate for the question
posed in this research, which inherently implies a non-homogeneous habitat. Point
discharges, such as sewage outfalls, create strong stress gradients while dredge
disposal forms "localized" patches of disturbance. Besides the limitations due to
stressor heterogeneity, there has been a growing recognition of the role habitat
heterogeneity and interconnectiveness among subpopulations have on population
dynamics (e.g., Roughgarden and Iwasa, 1986; Hastings and Harrison, 1994; Karelva
and Wennegren, 1995). Spatially explicit population models (SEPMs) are a recent
advancement in this area, where SEPMs "combine a population simulator with a
landscape map that describes the spatial distribution of the landscape features"
(Dunning et al., 1995). SEPMs are emerging as important management tools for
terrestrial populations in fragmented habitats (e.g., spotted owls). However, this
approach has been applied to only a few marine/estuarine resource species (Quinn, et
al., 1993; Botsford etal., 1994; Botsford, 1995).
To address the relative risk associated with these dispersive versus containment
disposal strategies, we propose to develop a SEPM. This proposal describes a two-
year research project using this SEPM to generate insights into the relative ecological
risks associated with exposure scenarios with different spatial configurations of single
and multiple sources and to identify the relative importance of various environmental
and demographic parameters in determining these risks. By focusing on multiple
sources, this research is responding directly to the Agency's need for methods for
assessing cumulative risk (U.S. EPA, 1997a).
Objectives : The goal of the research is to elucidate general patterns and relationships
and not to predict population trends with sufficient accuracy for management of the
-------
105
target species. The research is framed in the context of a generalized stressor, though
it can be conceptualized as a fixed mass of a contaminant discharged into a body of
water. Given this mass of contaminant, what geometry of the pollutant source(s) (e.g.,
discharges, sediment hot spots) minimizes effects on the target population? The
specific questions addressed are:
Under what exposure scenarios should stressors be concentrated versus
dispersed to minimize effects on benthic populations?
What demographic characteristics determine whether benthic populations are
more sensitive to a dispersed versus a contained stressor?
What characteristics of the stressor determine whether benthic populations are
more sensitive to a dispersed versus a contained stressor?
How does the optimum mitigation strategy for maintenance of benthic
populations vary with the total intensity of the stressor (e.g., total mass of
contaminant)?
METHODS:
I. Model Overview: A theoretical model is used to address the above questions. A
theoretical model allows us to focus on discovering the general qualitative patterns
rather than on the details of a particular pollutant, species, or site. A theoretical
approach also facilitates the integration of information and relationships from a diversity
of pollutants and species. Addressing the effects of the geometry of stress requires
sub-models for the habitat, stressor dispersal, stressor-response function, and
population dynamics. To the extent practical, these sub-models are dimensionless, so
many parameters are unitless and functions related to density are normalized to the
equilibrium density. Detailed mathematical formulations are given in the Appendix
while the model parameters are summarized in Appendix Table 1 and key assumptions
are summarized in Appendix Table 2.
The habitat is modeled as independent, identical cells. In Phase I the habitat is models
as a one-dimensional space while in the other Phases it is modeled as a two-
dimensional space with circular symmetry. In Phase I the cells are of unit area spaced
one unit apart along a line (i.e., a linear habitat). In other Phases cells are of unit area
and placed in annular rings. This means that at radius r there are 2nr cells. Fractional
cells are accounted for by the numerical integration process (see Appendix).
The circular model was chosen for simplicity of analytical solutions and ease of
computer programming. Every cell at radius p from a source is exposed to the same
level of stress from that source. In spaces based on squares or hexagons, the number
-------
106
of cells at a particular distance from the source is a complicated function. If there are
compelling reasons, the results calculated with our annular model can be translated
into hexagonal cells within a nearly circular domain after the fact, but with additional
programming.
The habitat is initially assumed to be homogeneous except for spatial differences in
anthropogenic stress. Using life history characteristics for an equilibrium population,
an equal number of the target species, the clam Mya arenaria, are introduced into each
cell. Recruitment is simulated by summing the larval output from each cell into a single
larval pool, which is evenly distributed among all cells. An age-structured population
model (Leslie matrix) is used to predict natural population dynamics and the effects of
anthropogenic stress.
A generalized stressor is then imposed in a "source" cell (point source). Depending on
the exposure scenario, a single or multiple source cells will be distributed throughout
the habitat. Stress level decays exponentially with distance from each source cell, with
the steepness of the decline determined by a lapse rate. The stress level at the origin
is adjusted so that the total stress from each stressed cell in the habitat (summed
across all cells) is constant among the different exposure scenarios, which can be
conceptualized as a fixed mass of a contaminant discharged into an estuary. This
constraint makes the stress level at the origin a function of the lapse rate. For
example, under high lapse rate (high dispersion) the stress (e.g., the fixed mass of
contaminant) will be widely dispersed over the habitat resulting in a lower stress level
in the source cell. Once the stressor level in a cell has been determined, its effect on
post-metamorphic survival and/or fecundity is predicted using a logistic
stressor-response function.
Both spatial and temporal variability and density-dependent interactions will be phased
in as the research progresses, thereby relaxing many of the initial simplifying
assumptions. Simulations will be used to predict the relative effects of different
geometries of the stress (source cells) on population dynamics and sustainability. In
the simplest case, all the stress is introduced into a single source cell. Other exposure
scenarios simulate the stress being "discharged" into multiple source cells with different
decay rates. The importance of exposure, stressor, and life-history parameters on
population dynamics will be assessed by sensitivity analyses. The ultimate goal of the
research is to approximate the stress pattern in the Southern California Bight, San
Francisco Bay, and/or Yaquina Bay.
II. Research Phases: Our strategy is to divide the two-year research project into
discrete conceptual and programming phases that will be addressed sequentially
(Appendix Table 3). Phase I includes problem formulation and development of a one-
dimensional model. These tasks have been or are nearly completed. Phase II
develops the techniques for the two-dimensional model and derives preliminary
-------
107
functions for density dependence. These are ongoing tasks that will be completed by
the end of April - early May 1998. Phase III begins the simulations with adult-adult
density-dependent interactions but no stochastic variation. Stochastic variation on
benthic habitat quality and life history parameters will be incorporated in Phase IV. We
anticipate completing most of these tasks by late August 1998. Phase V incorporates
adult-juvenile and larval-larval density-dependent interactions with a monthly time step.
Because of the potentially large number of simulations, we do not anticipate completing
this phase before the end of November or December 1998. In particular, comparing
predicted and observed temporal variability may take longer. The incorporation of
actual field data in Phase VI is the final suite of tasks, and is the most problematical
both because it builds upon all the other phases and because of the need to gather and
analyze field data from a variety of sources. We anticipate finishing the simulations in
this phase by the end of June, 1999. As mentioned in the Budget/Resources section,
full completion of Phase VI may require additional resources. These phases are a
"road map" to help guide the research and will be modified as needed.
III. Habitat and Stressor Distribution: The model domain is broken into cells that
represent uniform, independent sections of habitat. The size of the cell is sufficiently
large that migration of post-metamorphic Mya among cells is assumed to be negligible.
In the two-dimensional configuration, the model domain is a circle with a radius of 100
units (see Figure A-2 in Appendix). A circular domain was chosen to simplify the
analytical calculations though the techniques can be adapted to rectangular,
hexagonal, or even spatial approximations of actual estuaries. The assumption that the
natural (unstressed) habitat quality is homogeneous across all cells will be relaxed as
spatial and temporal variability are incorporated into the model (Appendix Table 3).
A generalized anthropogenic stressor is applied to one or more source cells. The
source is assumed to be constant so the level of stress does not vary over time. Each
source cell can be conceptualized as a point source, such as sewage discharge or
contaminated sediment hot spot, emanating pollutants in all directions. The stress
level in surrounding cells decays exponentionally from each source cell. Rapidity of
the decay over space is determined by the lapse rate, bs. In the one-dimensional case,
the normalized stress level in cells surrounding a source cell (cell 0) is calculated as:
S© = S0 e ¦"
where:
S© = stress level in cell c
S0 = stress at the source cell
c = cell number (0-100 indicating distance from source cell)
bs = lapse rate
-------
108
The Appendix gives the methods for calculating stress levels in the two-dimensional
case and for multiple source cells with overlapping stressor distributions.
So as not to confound the effects of total stress and the geometry of stress, the total
amount of stress in the model domain is constant regardless of the number of source
cells. This can be conceptualized as discharging a fixed mass of effluent into an
estuary. As more outfalls are added, the discharge per outfall declines proportionally
but the total mass of contaminants remains constant. Total stress is a more
appropriate measure to predict wide-spread effects than a localized measure of stress
level, such as the pollutant concentration at the edge of the zone of dilution. For
example, Mearns and O'Connor (1984) showed that the spatial scale of impacts on
marine/estuarine environments was related to the mass discharged for a variety of
pollutants.
One complexity with multiple source cells is how to calculate the dispersion of stress
when some source cell is not in the center of the habitat (circle). The stress from such
a source is dispersed exponentially over the circular domain, but is truncated at the
domain edge. The strength of this source is determined by the fact that it contributes
the same fixed mass of effluent as if it were at the center of the circle. This can be
conceptualized as having two sewage discharges, one in the center of the estuary and
other at the edge. Since the effluent can not leave the estuary (model domain), the
effluent concentrations in cells surrounding the discharge at the edge are higher than
those surrounding the discharge in the center. Mathematical details on addressing this
problem are given in the Appendix.
Our initial approach to varying the spatial pattern of the stress is to expand the number
of source cells at the center of the habitat in a geometric series, resulting in an
expanding "circle" of source cells. At the limit, all of the cells in the domain are stress
sources, with each receiving stress from all of the others. A major objective of these
single-source simulations will be to determine how the effects of dispersing a stressor
spatial geometry of the stressor interacts with stressor characteristics (e.g., slope of
stressor-response function - see below) and population characteristics.
As multiple sources are added, the stress from each will be normalized so that the total
stress to the domain will be constant in each scenario (e.g., fixed mass of effluent
discharged through multiple outfalls). We will evaluate various exposure scenarios
using different spatial configurations of source cells coupled with different decay rates.
For example, we could evaluate the effects of locating discharges along the boundary
of the estuary versus in the center with high and low decay rates. Another example is
evaluating the effects of clustering multiple dredge disposal sites in a sector versus
distributing them equi-distant throughout the estuary. Our ultimate goal is to
approximate the distribution of stress patterns in San Francisco Bay, Yaquina Bay,
and/or Southern California Bight, as discussed below.
-------
109
IV. Population Biology of Target Species: While it is not our purpose to forecast
population changes in any particular species, the model requires life history character-
istics of a "target" species. We focus on commercial/ recreational species because
they are important to the public and are heavily weighted in risk management
decisions. The difficulty with mobile species, such as fishes and crabs, is their duration
of exposure becomes a confounding factor that we felt was best not addressed in this
initial effort. Of the potential clam species, Mya arenaria was chosen because of the
availability of age-specific vital rates (Brousseau, 1978; Brousseau et al., 1982),
including an estimate of equilibrium population structure and "larval" mortality
(Appendix Table A-1). Mya is an important commercial species on the Atlantic coast
and a recreational species on the Pacific coast, though historically there has been
commercial fishing in Grays Harbor (David Armstrong, pers. comm.). On the Pacific
coast, Mya occurs from Vancouver Island to San Diego and is locally abundant in
several estuaries, including Yaquina Bay and San Francisco Bay.
Since we are not attempting to simulate Mya per se, we may modify the life history
characteristics reported by Brousseau to simplify the modeling as needed. For
example, when we go from the yearly time step to a month time step (Appendix), we
may assume that the larval period lasts only one month. We recognize that life-history
parameters can vary with habitat (e.g., Appeldoorn, 1995). By conducting a sensitivity
analysis on the life-history parameters, it should be possible to determine to what
extent spatial/temporal variations in Mya's vital rates qualitatively affect the
conclusions. These results will also offer insights into the confidence in extrapolating
the conclusions to other clam species with generally similar life history characteristics.
Because of fundamental differences in life histories, extrapolation to taxa that brood
their young or are highly mobile is problematical.
V. Population Model: Leslie Matrix Model: An age-structured Leslie matrix, is used to
model the population dynamics of Mya. The Leslie matrix predicts the population size
in the next generation from age-specific vital rates for survival and fecundity (see
Emlen, 1973; Caswell, 1989; Akgakaya, 1997). (An alternate model is a stage matrix
which groups individuals by size or biological stage (egg, pupa, adult) instead of age).
Anthropogenic effects are expressed through alterations in these vital rates. Leslie or
stage matrix models have been applied to a variety of terrestrial and aquatic
populations (e.g., Van Winkle, 1977; Caswell, 1989; Emlen 1989), including estuarine
bivalves (Brousseau et. al., 1982; Weinberg, 1985,1989). Recently, the approach was
used to assess stressor effects on a marine polychaete (Levin et al., 1996).
Emlen (1989) concluded that matrix models "work well over broad circumstances."
Though Power and Power (1995) concluded that an individual-based model (IBM) for
brook trout was better than a Leslie matrix model, the Leslie matrix actually predicted
total population size slightly more accurately. In any case, most of the differences were
-------
110
only on the order of a few percent. The U.S. EPA Risk Assessment Forum (U.S. EPA,
1991) concluded that Leslie or stage matrix models coupled with sensitivity analyses
can be used to: 1) understand the effects of changes in the vital rates on population
dynamics; 2) estimate the direct effects of stressors on population dynamics; 3) serve
as a tool for evaluating alternative management strategies; 4) serve as a framework for
interpreting the consequences of acute and sublethal stressors; and 5) provide a
method of incorporating demographic properties into ecological risk characterization.
Application of Leslie Matrix Model to Mya: Details of applying the Leslie matrix to Mya
and of incorporating a generalized stressor are given in the Appendix, while key
assumptions are summarized in Appendix Table 2. Two specific modifications are
summarized here. First, the production of female eggs from each cell is summed into a
common "larval pool." It is initially assumed this larval pool is completely mixed and
that the surviving larvae are distributed equally among all the cells. In other words,
there is equal recruitment into each cell, though not necessarily equal survival of
recruits as discussed below.
The second modification is a result that Brousseau et al. (1982) used a mesh too large
to capture the 0-2 month post-settlement class (i.e., early recruits) when sampling the
benthic population. Therefore, their calculated estimate of "larval" survival at
equilibrium (their r^ our p^; Appendix Table 1) includes both larval and early juvenile
(post-settlement) mortality. Modeling adult-juvenile or larval-larval density-dependent
interactions requires partitioning this total mortality, though as pointed out by Rumrill
(1990), no one has accurately partitioned larval and early juvenile mortality. To
address this uncertainty, we define a ratio (Rjp) of survival in the 0-2 month class to
survival in the larval phase. Rjp changes the partitioning of the mortality among life
history stages, not the total amount. Values for Rjp will be based on literature values
for larval survival (Rumrill, 1990), though these reported values are likely to be lower
estimates either because the studies did not cover the full larval stage or did not
include losses due to unfertilized eggs.
Inhouse Model and RAMAS-GIS: We initiated this research with one of us (C.B.)
writing a Fortran program to calculate Leslie matrices. Since then, we have obtained
RAMAS-GIS (Ak^akaya, 1997), which links age-structured population models with
spatial data on habitat quality. Strengths of RAMAS-GIS include the number of built-in
functions and links to GIS data. While a powerful program, the present version does
not meet all our requirements. One limitation is that RAMAS-GIS is limited to 160
populations (= subpopulations in cells), which is insufficient for our two-dimensional
simulations. Another limitation is that RAMAS has an upper limit of 20 x 20 age- or
stage-matrices, which is insufficient if we use a month time step for Mya, which requires
a 144 x 144 matrix (12 year life span of Mya X 12 months/year). An alternative to the
month time step would be to convert the data to a stage matrix, using larvae and early
juveniles as stages. However, converting the existing Brousseau Leslie matrix to a
-------
Ill
stage matrix would be difficult, or impossible, even for an unstressed population. For
these reasons, we will continue the development of the inhouse program as the primary
tool for this research. As appropriate, we will use RAMAS-GIS to address specific
questions. To date, we have used RAMAS-GIS as a QA check on our inhouse
program, and the comparisons have been excellent.
VI. Stressor-Response Function: A unitless generalized anthropogenic stressor is
used to simulate the effects of chemical pollutants (e.g., neutral ogranics, heavy
metals) as well as many non-pollutant stressors (e.g., temperature, hypoxia, organic
enrichment). Natural mortality and the additional anthropogenic mortality are assumed
to be independent. Therefore, their combined effect is calculated as the product of the
survival rates and the order in which the stresses are applied does not affect the total
mortality. The assumption of independence could be violated if one stressor increased
susceptibility to new stress (e.g., pollutant reducing predator avoidance) or if one of the
stressors operates in a highly density-dependent fashion (e.g., invasions by exotic).
These cases will not be explored at this time.
The logistic function (Barnthouse et al. 1989, 1990) will be used as the
stressor-response function for the generalized stressor. The specific equation used
(equation #8 in Barnthouse et al. 1990) is:
P=e(a+^/[1 +e< a +0") ]
where
P= fractional response in population
a = fitted parameter
/?= fitted parameter
X = Iog10 concentration of contaminant
As explained in the Appendix, x is transformed from concentration to a unitless level of
the generalized stress normalized to the stress level the source cell. The parameter a
determines the extent of the effect (e.g., mortality) for a given stress level (e.g., toxicity
of a discharge for a given concentration). f3 determines the slope of the
stressor-response curve, with larger 0 s resulting in steeper stressor-response curves.
Stressors with a high P have a narrow range between no or minimal effect and high
effect. By altering a and/or /?, it is possible to generate a family of sigmoid curves
describing the dose-response relationship for individual pollutants, complex mixtures,
and selected non-chemical stressors (e.g., temperature).
We will initially assume that all adult stages (> 1 year) have the same sensitivity to the
stressor. However, the effect of increased sensitivity of juveniles (<1 year) will be
-------
112
examined by increasing the value of a relative to the value used for adults while
maintaining /? constant. The importance of acute versus sublethal effects will be
examined by varying the values of a for mortality (aq) and fecundity (aj. We will
review the literature to obtain acute/chronic or acute/sublethal ratios as guides for the
relative difference in sensitivity in these two vital rates.
The steepness of the stressor-response function, as determined by p, is likely to be a
major driver in determining the cumulative effect of low to moderate levels of stress.
Barnthouse et al. (1989) determined the range in /?for mortality based on 77 chronic
exposures offish to organic pollutants or metals. They found no significant difference
in the Rvalues for eggs, larvae, and post-larval life stages, so the same range will be
applied to all life history stages of Mya. Barnthouse et al. (1989) did not evaluate
sublethal effects but as an initial assumption we will apply the same range in /? to
stressor effects on fecundity.
At this stage, it is assumed that the generalized stressor does not cause larval
mortality. In general, water column effects should be minimized because of the high
dilution with overlying water in marine/estuarine systems. One of us (H.L.) reviewed
numerous 301H applications (variance for secondary sewage treatment in
marine/estuarine waters), and effects on phytoplankton or larvae were rarely, if ever,
documented. Even if there are effects on the water column from a discharge or
contaminated sediment, they are likely to be localized around the source and thus have
a negligible effect on overall larval survivorship.
It is important to recognize limitations of the stressor-response function as used in this
model. First, the model does not consider indirect effects. Indirect effects may be
physiological in nature (e.g., increased sensitivity to low salinity due to pollutant stress)
or ecological (e.g., competitive release). To the extent that indirect effects are
important, the model predicts the potential effects on the population rather than an
actual population trajectory. Because the objective of the research is to rank order
exposure scenarios, the occurrence of indirect effects alters the conclusions only if they
change the relative risk associated with various exposure scenarios. For example, the
ability of a pollutant-sensitive keystone predator to regulate the population of the target
species may be a function of the scale of the contaminated area, and hence the lapse
rate of the stressor. No attempt is made to simulate these complex ecological-
toxicological interactions at this time. Second, the logistic function is likely to be
inappropriate for many biotic stressors (e.g., exotic species) because of non-linear
effects and feedbacks. The third case is habitat alteration, which is used here to
denote multi-generational changes in the "natural" physical/chemical properties of the
environment. Short-term changes (e.g., low salinity after a flood) would be modeled
using the logistic stressor-response model. Long-term changes probably would be
better modeled by redefining habitat quality by altering the natural rates of survival and
fecundity.
-------
113
VII. Density-Dependent Interactions: Density-dependent interactions can occur during
three life-history stages in benthic organisms: larval, early juvenile, and adult (£1 year).
Since arbitrary densities are used, density dependence will be scaled to the equilibrium
densities of larvae or adults, as appropriate. Our strategy is to address adult-adult
interactions first as these can be modeled using a yearly time step. Later in the
research, we will incorporate adult-juvenile and larval-larval interactions, both of which
require a monthly time step to be able to separate larval and early juvenile densities. In
general, we will attempt to bracket the potential importance of density-dependent
interactions by modeling three levels: none (base case), moderate, and high. If
inclusion of density dependence qualitatively changes the conclusions, more detailed
simulations will be conducted. The specifics for each life history stage are given below.
Adult-Adult Density-Dependent Interactions: Field experiments have documented
cases where growth rates, fecundity, and survival of benthic organisms have declined
at densities approximating natural levels (e.g., Peterson, 1982; Peterson and Black,
1988; Peterson and Black, 1993; Olafsson et al., 1994). It appears that fecundity is
more sensitive to "crowding" than survival, so we will primarily focus density-dependent
effects on fecundity and secondarily on adult survival. Because of the impracticality of
converting the existing age-based Leslie matrix for Mya to a size-based stage matrix,
we do not consider effects on growth. The direct changes in fecundity should capture
indirect effects growth rate, and size, would have on fecundity. Based on Peterson's
(1982, his Fig. 5) results with the clam Protothaca staminea, a ramp function is used to
approximate the density-dependent effects on fecundity (Appendix). We will use
Peterson's (1982) work and other lab/field experiments reported in the literature to
bracket the input values for the ramp function.
Adult-Juvenile Density-Dependent Interactions: Adults clams may reduce recruitment
success by directly preying on settling larvae, altering habitat suitability for early
recruits, or through competition. While the evidence for intraspecific adult-juvenile
interactions at natural densities is mixed (though very early recruits often are not
sampled quantitatively), Olafsson et al. (1994, page 95) concluded that "there is some
evidence that juvenile invertebrates also experience density-dependent mortality as a
consequence of food limitation." Therefore, we will model survival of the youngest
benthic age class (= recruitment success) as a function of the localized adult (£1 year)
density. Unless a better function is found in the literature, we will assume a ramp
function to describe this interaction. At this stage we will not consider facilitation,
where adults enhance recruitment (e.g., Peterson and Black, 1993).
Larval-Larval Density-Dependent Interactions: Density-dependent interactions affecting
larval survival are a function only of the size of the larval pool. Fisheries scientists
have spent considerable effort evaluating larval density-dependent interactions and
models (e.g., Ricker, 1973, 1975; Hall, 1988; Hilborn and Walters, 1992; Higgins et al.,
1997). Given the current state-of-the-science, our approach is to use functions
-------
114
generally accepted among fisheries scientists, in particular the Ricker and the
Beverton-Holt equations (Ricker, 1973, 1975). Ak^akaya (1997) suggests that the
Ricker equation better describes contest competition while the Beverton-Holt equation
better describes scramble competition. Both functions will be used in selected cases to
determine whether the nature of the relationship qualitatively affects the conclusions.
VIII. Spatial and Temporal Variability: Spatial and temporal variation is inherent in
estuarine and marine populations and environments (e.g., Ulanowicz et a)., 1982;
Lipcus and Van Engel, 1990; Nelson and Virnstein, 1995). This variation can have
"deterministic" components, such as the spatial patterns of habitat types, and
stochastic components. While recognizing these sources of uncertainty, we will initially
focus on deterministic, homogeneous models, which allows us to identify key variables
and qualitative trends before adding additional layers of programming and conceptual
complexity. After completion of the first suite of deterministic simulations, temporal and
spatial variability in the "natural" vital rates will be phased into the simulations
(Appendix Table 3). This natural variation is in addition to any variation due to the
•effects of the anthropogenic stressors.
Considerable effort is required to incorporate stochastic variation with precision.
Ak^akaya (1997) calculated that approximately 1000 simulations are needed to
generate results with a 95% confidence interval of about ±3% in models incorporating
environmental stochasticity. To keep the number of simulations to a reasonable
number, the importance of parameters related to stressors (e.g., lapse rate, /J) will be
evaluated primarily through deterministic sensitivity analyses (Appendix Table 1).
Parameters defining natural habitat quality and the associated vital rates will be varied
both deterministically and stochastically. Because many of these vital rates are likely
to have small mean values and high variation, a lognormal distribution for the rates is
likely to be used (see Akgakaya, 1997). As appropriate, we will follow the EPA
guidance on Monte Carlo analysis (U.S. EPA, 1997b).
The indicators used for habitat quality and the types of variation in are discussed
below. The sources are broken into classes for clarity but several sources of variation
will be combined in simulations (Appendix Table 3).
Indicators of Habitat Quality: Survival of early recruits and fecundity should be
sensitive indicators of benthic habitat quality in terms of food availability, sediment
suitability, physical-chemical environment, and interspecific interactions. Therefore, we
will model the effects of spatial and temporal variation in benthic habitat quality
primarily through changes in these rates, and secondarily on adult survival. Recruits
are defined either as the 0-year class or the 0-2 month class, depending on whether
yearly or monthly time steps are used, respectively. Survival rate is the only measure
of habitat (water column) quality used for larvae. A more realistic approach that may be
explored is to use instantaneous rates of mortality (see Rumrill, 1990) and make
-------
115
duration of the larval phase a function of habitat quality (e.g., temperature).
Spatial Variation in Benthic Habitat Quality: The first step will be to impose stochastic
variation on the homogeneous benthic habitat, simulating within-habitat variation.
Survival of recruits and fecundity within each cell will be varied randomly. The
equilibrium values will be used for the means and variance estimates will be derived
from studies of within-habitat variation, though we recognize that such estimates can be
scale dependent. We assume that fecundity and juvenile survival are correlated (i.e.,
what is good for adults is good for juveniles), though different scenarios will be
explored as time permits. Both spatial stochasticity and adult-adult density
dependence will be imposed concurrently in these simulations. A refinement is to
impose a spatial pattern of differing habitat types (e.g., high intertidal, deep water) on
the model domain. Each habitat type would have different qualities as expressed
through the vital rates. The habitat types would be discrete, though it is possible, and
more complicated, to model gradations in habitat quality. Stochastic variation could
then be imposed upon these habitat types.
We will explore the possibility of modeling habitat quality in San Francisco Bay,
Yaquina Bay, or the Southern California Bight. Habitat types would be ranked for
quality for bivalves based on expert judgement using factors such as water depth,
salinity, and substrate type. Although this approach is admittedly crude, it is an initial
step in incorporating field data. The specific geographic area would be chosen based
on data availability, the extent to which the site conforms to the model assumptions,
and Agency needs. Habitat data for Yaquina Bay are available from previous studies
and from ongoing research projects at the Coastal Ecology Branch (CEB). Data for
San Francisco Bay are available from the various Region 9 monitoring programs,
previous research programs (e.g., Lee et al., 1994; NOAA, 1997), the San Francisco
Estuarine Institute, and the CalFed program. Data from the Southern California Bight
are available from the 1994 and 1998 Southern California Bight Monitoring Study
(Janet Hashimoto, pers. comm.) as well as the extensive studies conducted around the
Southern California outfalls (e.g., Swartz et al., 1986; Ferraro and Cole, 1990; Stull,
1995; ongoing Region 9 Superfund assessment of Palos Verdes shelf). Field data
would be incorporated into the model through Arclnfo, which may require using
RAMAS-GIS. Though it is a goal to simulate one or more of these sites, we will be
cautious not to "get lost in the detail" for any particular site.
Spatial Pattern of Larval Recruitment: In the methods described above, benthic
conditions drive the spatial variation in recruitment. Another approach to relaxing the
assumption of homogeneous recruitment is to randomly distribute the larval pool
among cells. This approach varies the supply of larvae to each cell, simulating
stochasticity due to localized circulation patterns. The two approaches can be
combined to generate a total spatial variation in recruitment.
-------
116
Temporal Variation in Benthic Habitat Quality: Benthic habitats can vary from year to
year due to climatic conditions affecting factors such as microalgal production or
sediment type. This temporal variation in habitat quality will be modeled by randomly
varying fecundity and juvenile survival among years. As with spatial variation, year-to-
year variations in fecundity and recruit survival will be co-varied. Temporal variation
will be applied uniformly across all cells and habitat types.
Temporal Variation in Larval Survival: Year-to-year variability in the recruitment of a
variety of fishes and epibenthic invertebrates (e.g., Ulanowicz et al., 1982; Lipcus and
Van Engel, 1990) suggests that temporal variation in larval survival has a greater
impact on population dynamics than temporal changes in benthic habitat quality.
Therefore, we will place a greater effort on capturing this source of uncertainty. This
effect will be simulated by stochastically varying survival of the larval pool among
years. Landing records for Mya or similar bivalves (e.g., 1890-1992 Maine soft-shell
clam landings available on http://www. maine.com/mer/bpproj.htm; Ulanowicz et al.,
1982) will be used as a guide to the variability in these vital rates. Landing records
include variation due to both larval survival and survival to marketable size classes, so
it will be necessary to estimate the relative contributions of these two components.
Stochastic variation can operate either in a density-independent or density-dependent
fashion. In the former, the natural larval survival rate is varied randomly from year to
year. By varying the probability of survival from the larval pool to the 3-12 month class
(Pp«q). 11 is possible to impose this type of variation while using the one-year time step.
With density-dependent variation, the larval carrying capacity (input into the Ricker or
Beverton-Holt equations) is varied randomly from year to year. This approach requires
a monthly time step. It is not obvious which of these is more realistic or whether they
would generate qualitatively different conclusions.
IX. Model Outputs and Analysis: The primary objective of this research is to generate
rankings for the risk associated with various spatial distributions of stressors. To
achieve this objective, we will examine a suite of population metrics that emphasize
different life history characteristics or aspects of population dynamics. Qualitative
agreement among outputs would increase our confidence in the conclusions.
Disagreement would suggest that the results are sensitive to model assumptions or that
different components of population dynamics respond differently to the spatial
distribution of stress. In either case, further analysis would be needed.
Population Size and Age Structure: Total population size and age structure will be
examined at specific times (e.g., 5, 10, 15 years) under various exposure scenarios.
Additionally, the temporal variance in population size and vital rates will be examined
since strongly fluctuating populations may have a greater probability of extinction. The
effects of anthropogenic stress on the spatial pattern of the population will be examined
by plotting population density on a cell-by-cell basis at various time steps. We will
-------
117
explore the possibility of importing the model outputs into geostatistic software for
spatial analysis, including calculating varigrams and plotting out changes using kriging
techniques (see Rossi et al.t 1992). Changes in age structure of the total population
and subpopulations (individual cells or spatial aggregates of cells) will also be
examined, especially in relationship to the equilibrium age structure and densities of
"marketable" clams.
Population Finite Rate of Increase: A population's finite rate of increase is a predictor
of its trajectory. Compared to population size, the finite rate of increase integrates the
anthropogenic effects on age structure and age-specific fecundity. Although changes
in the finite rate of increase must be interpreted with caution especially in terms of
projecting long-term population trends (Akgakaya, 1997), it is useful in analyzing near-
term trends. It can be calculated from the Leslie matrix for the total population or
subpopulations. These values are amenable to the same types of temporal and spatial
analyses as population size. In addition, the relative importance of changes in specific
vital rates on the finite rate of increase will be determined by sensitivity analysis
(Weinberg, 1985; Caswell, 1989).
Quasi-Extinction: The susceptibility of a population to extinction is of direct
management relevance. Because the model uses arbitrary values for population size,
it cannot predict absolute extinction. Calculating quasi-extinction is possible, however,
which is the probability that the population will fall below some threshold size. The
thresholds will be set at decimal fractions of the equilibrium population size (e.g.,
0.001). The deterministic simulations generate predictions, not probabilities, of
whether the population will fall below the threshold, while the stochastic simulations
generate probabilities of quasi-extinction within a specified time.
Size of Larval Pool: The size of the larval pool represents the total reproductive output
or "fitness" of the population. As such, it is one indicator of the ability of the population
to recover from stress and to colonize denuded or new habitats. How well this indicator
predicts population trends is likely to depend on the nature and strength of
density-dependent interactions.
Contiguous Areas with Population Above Management Thresholds: As time permits,
we will examine two management thresholds. The economics of exploiting a resource
such as Mya depend, in part, on the spatial distribution of "dense" beds. If dense beds
are aggregated, costs associated with locating and fishing will be minimized. In
contrast, if the dense beds have a "checker board" distribution, these costs will
increase. This economic endpoint will be qualitatively evaluated by determining the
fragmentation of dense beds as a function of stressor geometry. Various values will be
used for the definition of "dense" and of the minimum fishing area (management unit).
These analyses will be conducted by importing the model outputs into ArcView.
-------
118
Due to public concern, it may not be possible to allow a localized population to fall
below some minimum value. The public may not accept a "dead zone" regardless of
population viability. To determine if the conclusions are sensitive to this political
reality, we will evaluate the effect of maintaining minimum population densities on the
conclusions. These analyses resemble the economic-based analyses except that they
would use different thresholds and different scale for defining management units.
Again, these analyses will be conducted by importing the model outputs into ArcView.
X. Model Validation: Approaches to Model Validation: The proposed model can be
"validated" in several ways. The most straightforward is algorithm validation in which
the specific mathematical functions are shown to produce the correct outputs. This
ongoing QA/QC is being conducted by comparing the inhouse model predictions with
those of RAMAS-GIS and by reproducing outputs for functions to the published results
(e.g., duplicating the life history of Mya to that published in Brousseau et. al., 1982)
One approach to validating ecological models is given by Caswell (1976) who
distinguishes between theoretical models, whose goal is to generate insights, and
predictive models, whose goal is to forecast. Caswell concluded that statistical
comparison of outputs to time series of data is not applicable with theoretical models.
Rather, he suggests testing hypotheses generated by the model. Because of the
complexity, it is often more practical to evaluate predictions from sub-models rather
than the overall model. Caswell also suggests explicitly comparing alternate models.
Though Caswell does not believe that sensitivity analysis can validate the "truth" of a
theoretical model, it can generate estimates of uncertainty (see Miller, 1974; U.S. EPA,
1997b). We will combine these approaches to derive a qualitative estimate of
confidence in the model predictions (Appendix Table 3).
Sensitivity Analysis and Stochastic Variation: As previously stated, deterministic
sensitivity analyses will be conducted on a suite of parameters and stochastic variation
will be added to another suite. These analyses will determine how robust the
conclusions are to various assumptions or model inputs.
Alternate Models: Several sub-models are amenable to comparisons of alternate
configurations. Generating the same, or very similar, predictions with different
sub-models increases confidence in the robustness of the conclusions. The alternate
sub-models will be evaluated on selected scenarios that capture the range in the
biological and exposure parameters. The following alternate sub-models will be
considered for evaluation:
Stressor-Response Function: probit, quadratic (simulates hormetic or protective
dose range), empirical dose-response function for sediment contaminants
derived from Southern California sewage discharges (see below).
-------
119
Stressor Decay from Source: Inverse relationship (stress proportional to
1/(distance to source))
Larval Density-Dependent Interactions: Ricker equation, Beverton-Holt
equation.
Adult-Adult and Adult-Juvenile Density-Dependent Interactions: Alternate
model(s) to the ramp function will be determined from a literature search.
Hypothesis Testing: The most rigorous, and most difficult, type of validation is testing
model hypotheses. While conducting field or mesocosm experiments to test
predictions from the overall model is theoretically possible, such tests would be
prohibitively expensive. Therefore, any hypothesis testing must be conducted using
available data sets and, most likely, on sub-models. We will test the following
hypotheses, which are ordered in approximate order of their ability to reject the model
or sub-model.
Population Variation: There is no significant difference in the coefficient of
variation (or other measure of temporal variation) between that predicted for the
stochastically varying unstressed population and that observed in natural
populations of Mya. Predictions from the Leslie matrix would be compared with
long-term catch records for Mya (e.g., Ulanowicz et al., 1982;
http://www.maine.com/mer/bpproj.htm). The limitation of this test is that it is
likely that the same data used to bracket input values would be used to evaluate
the hypothesis, so this approach blends into curve fitting. An independent test is
possible only if two independent data sets from geographically similar areas can
be found.
Stressor Dispersion from Source Cell: There is no significant difference
between the shape of the one-dimensional stressor distribution predicted by the
exponential lapse rate and the observed pattern for sediment pollutants along a
gradient away from Southern California sewage discharges. The Southern
California outfalls will be used both because of the extensive data and because
of the relatively uniform habitat along the 60 meter contour (e.g., Swartz et al.,
1986; Ferraro and Cole, 1990; Stull, 1995).
Relationship of Population Distribution to Stressor Distribution: There is no
significant difference in the spatial pattern of relative population density
predicted by the one-dimensional model and the pattern in benthic populations
observed in a gradient away from Southern California sewage discharges. The
specifics of how to conduct this test need to be developed further, but our
preliminary idea is to use the empirical relationship between population density
of target benthic species and sediment contamination along an outfall gradient to
-------
120
develop an empirical "dose-response" function. The dose may be expressed as
the summation of toxic units to account for mixtures of pollutants. Substituting
this empirical dose-response for the logistic stressor-response function, the
model would then be used to predict the shape of the population distribution
along the gradient from a second, independent outfall. Comparison of the
predicted and observed population patterns would evaluate the robustness of
the model to predict responses at different sites with similar stressor types.
Population Extinction: The most rigorous test would be to evaluate model
predictions concerning population extinction (or quasi-extinction) and stressor
geometry. We will review the literature to determine if such data sets exist.
Expected Results and Benefits: One immediate benefit of the proposed research
would be the development of a SEPM that captures the key characteristics of benthic
populations and the direct effects of many of the dominant stressors in marine and
estuarine ecosystems. This tool would be available for use by the CEB and Region 9.
The model is a research tool and additional development would be required to make it
"user friendly." Another near-term output is a draft of peer-reviewed publication on the
effects of the geometry of stress which will be completed in FY98 or early FY99.
Additional publications will follow in FY99 and FY00.
The proposed research directly addresses the Agency's need for the development of
cumulative risk assessment methods. Regions 9 identified multiple stressors and
cumulative effects as one of their priority scientific needs. The Agency's guidance on
cumulative risk assessment (U.S. EPA, 1997a) stated that"... EPA's risk assessment
emphasis has shifted increasingly to a more broad based approach characterized by
greater consideration of multiple endpoints, sources, pathways and routes of exposure
" In future research, the model could be modified to address other components of
cumulative risk, such as evaluating the adequacy of standard ecological indicators
(e.g., sediment bioassay) to detecting cumulative effects at the population level.
Eventually, it should be possible to use SEPM's to make first-order predictions on the
combined effects of a mosaic of stressor sources with multiple types of stressors.
Besides generating general insights into cumulative risk, the proposed research
directly relates to the management of sewage discharges and dredge material. The
research could suggest, for example, that one or the other of the present management
strategies does not minimize risk to benthic populations. Any such conclusion would
be considered preliminary, and additional lab-field and theoretical research would be
required before changing management practices. Nonetheless, the proposed research
could indicate whether this line of investigation should be pursued.
-------
121
4.3 Research Theme C - Estuarine Physcial-chemical Stressors
Project C1 - Relationships Between Suspended and Bottom Sediment Conditions
and Macrophyte Distributions in Yaquina Bay Estuary
Principal Investigator: David R. Young
Goals: The goals of this research project are to determine the relationship of
suspended and bottom sediment conditions to the distribution patterns of principal
habitat-defining macrophytes such as eelgrass and green algae, in a small non-
urbanized estuary typical of those found in the Pacific Northwest (PNW). Emphasis will
be placed on identifying sediment-related physical/geochemical stressors that may
affect such distributions. As part of a Coastal Ecology Branch (CEB) synthesis effort,
results of related past and present studies on these and other candidate stressors
(e.g., low salinity, excess nutrients, benthic competitors, grazing by waterfowl, natural
and anthropogenic physical disturbances) will be compared and evaluted. Sources of
those stressors determined to significantly impact macrophyte habitats in this test PNW
estuary then can be sought to support improved ecosystem risk assessments required
by Environmental Protection Agency (EPA) and other environmental managers.
Rationale: Macrophytes constitute major habitats in estuaries. The reduction of
Submerged Aquatic Vegetation (SAV) distributions in Atlantic and Gulf Coast estuaries
over the last few decades has been a major source of concern (Orth et al., 1995).
Extensive research into causative or associated factors has been reported (Ogata and
Matsui, 1965; den Hartog and Polderman, 1975; Jupp and Spence, 1977; Orth, 1977;
Rasmussen, 1977; Phillips et al., 1978; Thayer et al., 1984; Dawes and Lawrence,
1979; Whitmann and Ott, 1982; Bulthuis, 1983; Phillips, 1984; Dennison and Alberte,
1985; Libes, 1986; Wetzel and Neckles, 1986; Muehlstein et al., 1988; Orth and Moore,
1988; Giesen etal., 1990; Zimmerman et al., 1991; Batiuketal., 1992; Burkholderet
al., 1992; Dennison et al., 1993; Neckles et al., 1993; Stevenson et al., 1993; Thorn et
al. 1995; Bayer, 1996a, 1996b; Moore et al., 1996; Wetzel, 1996). A major finding is
that reduced subsurface light (Photosynthetically Active Radiation or PAR) caused by
increased turbidity has a major impact on SAV distributions. Moore et al. (1996)
utilized growth and survival of transplanted eelgrass (Zostera marina) to investigate
potential causative factors for the absence of this species in previously-vegetated
sectors of the lower Chesapeake Bay. They suggested that water quality conditions
causing seasonally-high turbidity and resultant water column light attenuation (K^)
was a major stressor. Accumulation of epiphytes also may have contributed to this
stress. Suspended sediment (principally inorganic particulates) and Chlorophyll a best
correlated with K,,. However, deposited sediments also can be a major stressor on
eelgrass. Anthropogenic alterations to a small coastal lagoon in southern California
apparently resulted in abnormal accumulation of storm-related sediments, permanently
removing this eelgrass habitat (Onuf, 1987). Such findings suggest that natural events
-------
122
(e.g., storms; floods; coastal upwelling) and anthropogenic activities (e.g., enhanced
watershed release of soil and nutrients; dredging; breakwaters; wastewater discharge)
may interact in a complex fashion to alter water column and substrate conditions,
potentially resulting in significant alterations of macrophyte-habitat distributions in PNW
estuaries. One of the EPA Office of Research and Development (ORD) High Priority
Research Topics is "Research to Improve Ecosystem Risk Assessment" (USEPA,
1997). The target areas of concern include causes of the destruction of critical habitat.
Cost-effective actions to prevent or mitigate such habitat destruction require an
understanding of the vulnerability and sustainability of these ecological resources
regarding the effects of multiple stressors on multiple endpoints at multiple scales.
This research project will focus on the relative importance of sediment-related stressors
on one such critical habitat (SAV), as an assessment endpoint, at the scale of the
coastal non-urbanized PNW estuary.
Objectives and Hypotheses; One of the persistent questions regarding alterations of
PNW watersheds involves possible impacts on living estuarine resources. SAV is
known to be an important habitat and/or food source for economically-important
fisheries and waterfowl species in PNW estuaries. The following objectives (in
conjunction with those of other Branch researchers) address possible relationships
between living estuarine resources, SAV distributions, turbidity (regarding effects on
PAR and SAV distributions), suspended solids (regarding effects on turbidity),
deposited solids (regarding effects on substrate composition, elevation, and SAV
distributions), and watershed alterations (regarding effects on sediment input to the
estuary):
Objective 1. Document annually (1997-2000) the summer season distributions of the
major SAV taxa throughout (or in selected sectors of)Yaquina Bay estuary (YBE). The
primary focus will be on eelgrass (principally Zostera marina), and the secondary focus
on mixed green algae (principally Enteromorpha spp. and Ulva spp.). Test the
hypotheses:
H0(1): There are no detectable annual changes in the summer distributions of eelgrass
or mixed green algae in any of the target YBE sectors over the four year study
period.
H0(2): There is no association between total suspended sediment (TSS) concentration
or turbidity and the summer distribution of eelgrass in YBE.
H0(3): There is no association between bottom sediment composition (grain size
distribution; total organic carbon [TOC]; total nitrogen [N]) and the summer
distribution of eelgrass in YBE.
Objective 2. During summer/fall 1998, document the bathymetry of YBE. Determine
-------
123
the substrate elevation ranges within which (1) eelgrass and (2) mixed green algae
occurred during summer 1998.
H0(4): There is no association between substrate elevation and the summer distribution
of eelgrass and mixed green algae in YBE.
Objective 3. Document bottom sediment accumulation rates and composition during
the last century within key intertidal habitats in YBE. Test the hypothesis:
H0(5): There is no association between historic alterations to the YBE watershed and
any changes measured in bottom sediment accumulation rates or composition
inYBE.
Scientific Approach:
Objective 1: The Remote Sensing (Aerial Photography - Ground Data) Pilot Project
survey of Yaquina Bay estuary conducted in the summer of 1997 will be completed.
Approximately 170 Color Infrared (CIR) photographs (scale of 1:7200) were taken of
the entire estuary (Figure 1) at an extreme low tide (-2 to +1 feet around Mean Lower
Low Water) on July 23, 1997. Another -50 Full Color (FC) photographs were taken of
the center part of the Bay within about one hour of the CIR photography. As part of the
planned digital photogrammetry project, diapositives from about 90 of the CIR
photographs providing stereo coverage of the intertidal and subtidal portions of the
estuary between the Ocean Breakwater and the City of Toledo will be scanned at a
resolution of about 25 microns. Based on ground control points (located by Differential-
corrected Geographical Positioning System [DGPS] to within +/- 0.5 m), the images will
be geocoded and then orthorectified by a professional digital photogrammetry
contractor.
As part of the ground data ("truth") component of the project, we selected 10 categories
of substrate cover which we termed "visually dominant taxon-density classes". By
"visually dominant" we mean that classification of substrate cover that is most visable
from above. The term "taxon-density" is included because we pre-selected 4 classes of
estimated "percent cover" values (0-25, 26-50, 51-75, 76-100 percent) for quantifying
the abundance of the endemic and dominant eelgrass Zostera marina. For each of 10
visually-dominant taxon-density classes, 6 stations (3 each on opposite sides of a
sector) were preselected in each of 5 sectors (A-E) of YBE between Hatfield Marine
Science Center (HMSC) and Boone Slough (Figure 2). As described above, 4 target
density classes were assigned to Zostera marina. For the other 6 taxa (the exotic
eelgrass Zostera japonica, mixed green algae, benthic diatoms, bare substrate control,
mud shrimp, ghost shrimp), stations characterized as "heavy" density were selected.
The ground data survey was initiated the day preceeding that of the aerial photography
(July 23, 1997). Virtually all of the 240 SAV-related stations, constituting 80 percent of
the 300 total ground data stations, were surveyed within 7 days of the aerial
-------
124
photograph survey. (The surveys of the two shrimp habitats, believed to have the least
temporal variability, were completed later in the summer). A 0.25-m2 quadrat strung
with a 5 x 5 wiregrid was placed at one corner of a numbered stake (with a colored
side north-facing, adjacent to a card indicating which compass sector [SE, SW, NW,
NE] was being sampled). After taking a near-vertical color photo, an estimate of the
Percent Cover of each of the 7 target taxa was made . Then a Percent Occurance
measure of each taxon was made by assessing which taxon was "visually-dominant" at
each of the 25 point-intercepts of the wire grid. Voucher specimens, and samples of
surficial sediment (0-2 cm) once per station, also were collected. The frame then was
moved to the other three compass positions and the sequence repeated, so that a total
of 4 Percent Cover estimates and 100 Point-Intercept counts were made for the 7 target
taxa (including bare substrate control) at each station (see Field Data Sheet, Figure 3A,
and Ground Data Survey Instructions, Figure 3B).
Quite satisfactory agreement was obtained between the Percent Occurance values
from the point-intercept measures and the Percent Cover estimates for eelgrass (Z.
marina, Figure A) and mixed green algae (principally Enteromorpha spp. and Ulva spp.,
Figure 5). Given that remote sensing strategies usually include a significant ground
data component, and the relatively narrow exposure time windows usually available for
lower-intertidal habitats, a rapid and reliable method of quantifying the visually-
dominant taxa recorded in an aerial photography is critical.
For each sector, half of the stations (15 total for YBE) sampled in the ground data
survey for a given taxon-density class will be selected randomly for the digital image
interpretation\analysis (supervised classification) step. The remaining 15 stations will
be used to provide some general guidance regarding the reliability of the classification.
To provide initial calibration data for the summer 1997 aerial photography test project,
station positions were selected on the basis of a pre-flight visual assessment of taxon
presence and density, as opposed to being probability-based. To improve the basis for
comparison, the two lower and the two higher density classes of Z marina may be
combined to yield 30 classification and 30 general assessment stations each for the 0-
50 percent and 51-100 percent cover classes. Based on a preliminary evaluation of
the CIR photographs, it is anticipated that reliable identifications of at least "dense"
beds of the endemic eelgrass (Z marina) and mixed green algae in the 5 target
sectors of YBE will be produced (Figure 6). (Beds of the exotic eelgrass Z. japonica
also appear to be distinguished in the CIR photos). Thus, the results of the intensive
ground data survey of the 120 Z marina eelgrass stations and the 30 mixed green
algae stations (all surveyed within 7 days of the aerial photography) are expected to
provide a reasonably reliable classification, and resultant digital maps (projected image
resolution: 0.5 m) of these two dominant SAV habitats in Yaquina Bay estuary during
July 1997, to the limit of the edge of tide at the time of a given CIR photograph.
During summer 1998, aerial photography of YBE again will be conducted as part of a
-------
125
comparative study of three remote sensing methods for delineating SAV (and possible
burrowing shrimp) populations in the intertidal zone. Based on a preliminary evaluation
of the summer 1997 CIR aerial photographs, it appears that a large majority of the Z
marina growing in the estuary occurs within a 4.5 km x 4.5 km zone (Zone I) located
near the mouth of the estuary (Figure 7). This zone encompasses the study areas for
the underwater video survey proposed by Dr. Brad Robbins, and the study areas for the
accoustical survey proposed by Dr. Ted DeWitt. Both studies will include ground data
surveys. (Related studies of SAV and burrowing shrimp habitats proposed by Dr.
Bruce Boese, and by Dr. Steven Ferraro and Faith Cole, also are expected to provide
relevant ground data results). The total area of intertidal zone targeted for the five
related studies in Zone I does not exceed about 10 km2. To aid the planning process, a
general matrix relating key variables of aerial photography (Figure 8) has been
prepared from published algorithms (Avery and Berlin, 1992; ASPRS, 1996). The
assumptions made in preparing this matrix are (1) use of a calibrated high-precision
aerial camera (9 in. x 9 in. film) with a 6 inch focal length (available on short notice of a
break in cloud/fog cover); (2) an average ground elevation of 3 m; (3) a standard
square study area 5 km x 5 km; and (4) a digital scanner resolution (for diapositives) of
20 microns. The independent variable is photo scale (ranging from 1:1000 to 1:7500).
The dependent variables related to product usefulness are minimum ground line
separation, pixel size, and nominal image resolution in ground units (m). The limiting
variables related to cost are the total number of photos (regarding acquisition and
diapositive production/scanning costs) and digital file size (regarding minimum
computer specifications).
For reference, the summer 1997 aerial photography survey of the entire YBE (1:7200)
yielded a total of about 220 photos (-170 CIR + -50 Full Color), at an acquistion cost
of about $11,000. The minimum flight altidude (above ground level) over populated
areas in Oregon is 1000 feet. Thus, the highest resolution available to us is
tabululated under the 1:2000 photo scale column (for a flight altitude of 1010 feet).
According to the results obtained from the theoretical relationships of aerial
photography, this scale should yield (in ground units) a pixel size of about 4 cm x 4 cm,
and a nominal image resolution (for blocks of 4 pixels) of about 8 cm x 8 cm (3 in. x 3
in). However, the projected file size per color (3 band) photo of almost 400 megabytes
(MB) - near our present operational limit - may require that we have the diapositives
scanned at a digital scanner resolution of 25 rather than 20 microns; this would yield a
photo file size of about 250 MB, a pixel size of about 5 cm x 5 cm, and a nominal image
resolution of about 10 cm x 10 cm. This range of 8 -10 cm nominal Image resolution
appears to encompass our present operational limit for CIR or FC aerial photography.
Although a "realistic" image resolution limit probably is several-fold higher, with enough
ground data collection and digital photo interpretation effort, it may be possible to
detect sediment ripple/mound patterns above dense beds of burrowing shrimp. (This
possibility might be enhanced even futher by using an intermediate-grade metric
camera to take low-altitude, rear sun position - low sun-angle oblique photographs,
-------
126
discussed below). Given that habitats related to populations of SAV and burrowing
shrimp appear to account for a substantial percentage of the total area of intertidal
mudflats in at least some PNW estuaries, acquiring an operational capability to detect
and map these habitats via aerial photography, and other, remote sensing methods is a
high-priority objective.
To obtain an estimated cost of acquisition of digital images for the ~10 km2 intertidal
mudflat study area of Zone I, we have scaled the projected 546 photos for a 25 km2
area (Figure 7) by a factor of 0.4, yielding about 220 photos, similar to the total number
obtained in the summer 1997 aerial survey. Thus, we accept the cost of about $11,000
(~$50 each) for that photo acquisition as our estimate for Zone I for 1998. Estimated
costs of producing and scanning (to 20 or 25 micron resolution) a diapositive also is ~
$50; the corresponding estimated combined cost for 220 CIR photos is about $11,000..
Thus the estimated total cost of obtaining ~220 digital images, ready to enter into the
Branch Geographical Information System (GIS), is about $22,000, or about $100 each
(prorating the fixed aircraft mobilization cost).
Based on costs estimates made to date, converting the images from these 220 digital
files into a geocoded, orthorectified state (to a ground positional accuracy of ~ 50 cm)
via a commercial photogrammetry contractor could cost approximately $100,000. This
would appear to be beyond the Branch Fiscal Year (FY98 or FY99) budgets. Costs
could be reduced by decreasing the photo scale (moving to the right on the matrix of
Figure 8), thereby increasing the pixel size and minimum ground separation/image
resolution values while reducing the number of photos required to provide standard
stereophotography coverage (60 percent frontlap; 30 percent sidelap). For example,
decreasing the photo scale by a factor of 2 (from 1:2000 to 1:4000) reduces the
minimum ground separation also by a factor of 2 (from 1.3 cm to 2.7 cm), but the
number of photos (and proportional costs) by almost a factor of 4 (to 172 photos/25 km2
or -70 photos/10 km2). This is similar to the number (~90) of summer 1997 CIR photos
(1:7200) of the YBE (up to the City of Toledo) now being scanned, geocoded, and
orthorectified for entry into the Branch GIS for digital image interpretation and
classification. The commercial cost of producing digital orthophotos from such a
summer 1998 set could be substantially lowered (perhaps by as much as 50 percent)
by utilizing the Digital Elevation Mode! (DEM) to be generated as part of the present
(1997) digital orthophotography project. Similar savings with substantially higher
ground postitional accuracy may be obtainable by utilizing the X,Y,Z data (DEM) to be
generated via the YBE 1998 bathymetry project, described below. Based on this logic,
it is proposed that, during summer 1998, approximately 70 CIR photos of the ~ 10 km2
I intertidal mudflat area of Zone I will be collected at a photo scale of 1:4000. Utilizing
the DEM obtained either from the 1997 CIR digital orthophoto production, or from the
1998 bathymetry project, the estimated cost of obtaining GIS-ready geocoded,
? orthorectified digital photographs of the intertidal mudflats of Zone I is ~ $15,000.
Adding this to an estimated cost of $7000 to obtain and scan diapositives of the
j
i
-------
127
estimated 70 CIR aerial photos (~ $100 each) yields a combined photo acquistion -
digital orthophotography cost for the -10 km2 study area of about $22,000.
As mentioned above, the square (4.5 km x 4.5 km = 20.25 km2) Zone I contains most of
the present distribution of Z. marina. (It includes the summer 1997 Sectors A-C, and
the northern half of Sector D, Figure 2). However, a second rectangular (7.5 km x 2.5
km = 18.75 km2) Zone II (Figure 7), beginning approximately halfway between
Sawyer's Landing and River Bend Marina and extending to Babcock Creek (Figure 10),
contains the remaining beds of this native eelgrass (southern half of Sector D, and
Sector E). In addition, from the summer 1997 ground data survey and the
corresponding CIR aerial photos, David Young concluded that a previously-vegetated
area in Sector E between Grassy Point and Craigie Point (USFWS, 1968; Bayer, 1979)
did not contain any substantial beds of Z. marina in summer 1997 (Figure 10). This
fact makes this Zone II an especially important section of the YBE, as a natural
laboratory for investigating the question: What changes in this part of the estuary have
occurred that may have contributed to the loss (or substantial decrease) of the Z.
marina beds reported there in the 1960's and 1970's? In addition, David Specht and
David Young observed extensive distributions of the exotic eelgrass Z. japonica in the
upper intertida! of Sector E (in Zone II) during their summer 1997 ground data survey.
Further, David Specht has concluded from a preliminary evaluation of the
corresponding CIR aerial photos that he is able to detect beds of this exotic eelgrass in
these photos.
These facts support the argument that, at the least, the summer 1998 CIR photography
planned for Zone I also should be conducted in Zone II. The estimated area of the river
channel and intertida! mudflat area of this zone does not exceed about 10 km2. Thus,
based on the derivation above that about 70 aerial photos (1:4000) would be needed to
provide stereo coverage of the -10 km2 intertidal mudflat area of Zone I, approximately
another 70 CIR aerial photo (at an added acquisition cost of about $3000) would be
needed for the river portion (and portions of the two major southern sloughs) in Zone II.
The developed film from this set could be archived until a decision whether or not to
order corresponding digital orthophotos, based on results of past or proposed surveys,
is made. The estimated annual combined cost (through outside contractors) of
conducting this two-tier aerial photography acquisition and partial digital
orthophotography is about $25,000. A decision regarding whether on not to repeat (or
modify) the summer 1998 surveys during summer 1999 and/or summer 2000, in order
to provide sufficient data for a SAV Change Detection analysis (Lyon et al., 1998),
would be made on the basis of results obtained from the 1997 and 1998 mapping (and
associated ground data) projects, as well as those from the other SAV/benthic shrimp
remote sensing projects proposed by Dr. Ted DeWitt and by Dr. Brad Robbins. The
results of these three collaborative research projects will be used to test H0(1).
As mentioned above, for the initial test project on aerial photography - digital
-------
128
photogrammetry of YBE conducted during summer 1997, the ground data sampling
component was based on stations that were selected before the photography on the
basis of desired taxa and estimated densities. This initial strategy was used to insure
that enough relevant information on ground distributions of the 7 target taxa (including
bare substrate control) would be obtained to adequately train the image analyst in
recognizing the digital images of differing taxa (or habitats) obtained by this remote
sensing project, which was new to our Branch scientists. However, because these
ground data sites were not obtained using a probability-based sampling strategy, the
validity of using them to conduct a traditional accuracy assessment of the image
classifications is questionable.
Therefore, beginning in summer 1998, as part of the collaborative effort with Dr. Robert
Ozretich and Dr. Ted DeWitt, a probability-based sampling program targeting the water
column, bottom sediment, and distribution patterns of SAV and salt marsh will be
conducted. To minimize the overall cost of sampling, the following common approach
between the Principal Investigators has been developed. To provide satisfactory
geographical coverage, the division of the estuary into the four "salinity" zones
proposed by De Ben et al. (1990) will be adopted. (Their fourth, lowest salinity zone
ending near Toledo City has been extended upriver somewhat to the Toledo Bridge
near the mouth of Olalla Slough, Figure 10). Within each of these four zones, polygons
within the classes listed below have been constructed using the Branch GIS:
1. Salt Marsh (generally higher than + 10 ft relative to MLLW)
2. +5 ft to +10 ft
3. +2 ft to + 5 ft
4. - 6 ft to + 2 ft
5. Deeper than - 6 ft
• All of these polygon classes except that for "salt marsh" are based on the bathymetry
summarized in 1953 (Figure 11). The salt marsh polygons are based on maps
provided in The Oregon Estuary Plan Book (ODLCD, 1987). The elevation zone +5 ft
to +10 ft is believed to contain the major distributions of the exotic eelgrass Z japonica.
| The major distributions of mixed green algae are believed to occur betwee +2 ft and +5
ft, while most of the endemic eelgrass Z marina is believed to occur between - 6 ft and
+ 2 ft. The lowest elevation zone (deeper than - 6 ft) will be sampled only for water and
sediment.
The GIS-based polygons have been sent to Dr. Anthony Olson, the EPA/ORD Western
Ecology Division (WED) statistician. Using established prochedures (Stevens,
1994; 1997) developed for EPA's Environmental Monitoring and Assessment Program
(EMAP), Dr. Olson will supervise production of the number of probability-based
» stations specified by each researcher within each polygon class for each of the four
salinity zone strata. Dr. Ozretich has developed a station allocation strategy that is
-------
129
expected to yield more than 30 water column - surficial sediment stations each within
and outside of beds of Z marina, throughout the estuary. Drs. DeWitt and Young have
specified a common suite of 8 stations each for elevation zone numbers 2, 3, and 4
above, for each of the four salinity zones, and an additional 8 stations per salinity zone
have been specified for the salt marsh polygons. (The largest practical overlap
between the probability-based stations of Ozretich and DeWitt & Young will be sought).
Thus, at least 32 probability-based stations will be obtained for each elevation level in
the estuary. (To insure that this minimum number of stations is obtained for each target
polygon class, the production of three times as many probability-based station
positions have been requested for each class). If, for a given sampling activity, the
randomly selected station is judged to be inadequate for that sampling purpose, a
standard procedure for obtaining an adequate station as close as possible to the
specified postition will be developed in advance of the field sampling.
For the ground data ("truth") survey, a modification of the benthic sampling strategy
successfully used during the summer 1997 ground data survey (Figures 4 and 5) will
be employed to obtain Percent Cover and Percent Occurance data for SAV (two
eelgrass species and two mixed green algae taxa) in a sufficiently rapid manner to
permit the sampling of the 96 benthic stations (32 each for elevation zone numbers 2,
3, and 4 above) within two weeks of the planned date of aerial photography (weather
permitting).
Based on sufficiently low tides (~ -1.7 ft. to -1.0 ft. below Mean Lower Low Water -
MLLW) occuring during periods of sufficent daylight (8 a.m. to 4 p.m.), the present
target period for the photography is the second week of either July or August. There
are approximately 16 days of sufficiently low tides between late June and late July, and
between late July and late August. Thus, using the CEB hovercraft for rapid transport
over the mudflats, two benthic surveyors could be expected to complete 6 stations per
low tide interval (~ 1.5 hr./day) if the time to sample and move on to the next station did
not exceed about 15 minutes.
To meet this criterion, a 1.0 m x 1.0 m quadrat strung with a 5 x 5 wire grid (25 point
intercepts) will be placed at the pre-marked probability-based positions, and a Percent
Cover estimate and Percent Occurance measure of the target SAV taxa (at least Z
marina, Z.japonica, Enteromorpha spp. and Ulva spp.) will be obtained. Voucher
specimens, and a sample of surficial sediment, also will be collected at each of the
probability-based benthic stations. This strategy should enable the surveying of the 96
probability-based intertidal ground data stations as part of the summer 1998 aerial
photography project. This is expected to support a valid accuracy assessment of the
image classification resulting from this project (Congalton, 1991; Stehman, 1995;
1996). Based on an evaluation of this assessment, modifications will be made, as
necessary, to the ground data surveys supporting any summer 1999 and summer 2000
aerial photography surveys.
-------
130
C1R photography has the distinct advantage of distinquishing between different types of
vegetation, such as eel grass and green algae. However, this remote sensing method
has the disadvantage that it cannot reliably detect underwater images (owing to the
very limited transmission of infrared radiation through water). Thus, once digital
orthophotos have been produced from the summer 1997 CIR aerial photographs,
efforts will be made to compare selected digitized Full Color aerial photos (taken the
same morning) with their CIR "mates" (photocenters of the CIR and FC pictures often
are separated by only about 30 m). The objective of such a comparison is to determine
if the offshore boundaries of the eelgrass bed images from the CIR and FC photos
systematically differ (i.e., does the CIR deep boundary generally occur inshore of the
FC deep boundary). If so, by what percentage does the aerial distribution of eelgrass
determined from the FC photography exceed that determined from the CIR
photography. This exercise should provide some insight into the degree of bias
associated with low-tide CIR aerial mapping of eelgrass beds, relative to the aerial
distributions obtained from FC aerial photography.
A corresponding effort to evaluate the degree to which low tide CIR aerial maps of
eelgrass beds underestimate the actual distribution will be made via a ground survey.
During calm, full daylight periods of summer 1998 at the lowest navigatable tide, the
lower edge of 2. marina beds in different sectors of Zones I will be mapped by driving
a vessel-mounted DGPS unit slowly over the clearly visible deep boundary of the beds.
The resultant deep boundaries will be compared to the apparant deep boundaries
obtained from analysis of the CIR digital othrophotos from the summer 1998 aerial
photography survey. (If substantial bias is measured, attempts to extend this
comparison to Zone II also will be made).
Another important part of remote sensing via aerial photography is the collection and
evaluation of historical photographs. This is more difficult for intertidal vegetative
habitats, where aerial photographs taken for other purposes may be of limited
usefulness owing to (1) black and white, monoscopic photographs; (2) photos taken at
other than a low tide; (3) lack of any ground truth data. However, there still is the
possibility of locating some historical aerial photos that could provide, at the least, an
opportunity for estimating SAV distributions which are approximately consistent with
present distributions. To date, David Specht and David Young have located potentially
useful photographs archived by the City of Newort, the Oregon State University (OSU)
Library, and the OR Dept. of Fish and Wildlife. Contacts also have been located for
probable sources from the OR Dept. of Transportation in Salem, and the U.S. Corp of
Engineers in Portland. During the three year study period, these and other candidate
sources of historical photographs will be systematically contacted and surveyed. Any
photos of the intertidal mudflat zones of Yaquina Bay Estuary, or other PNW estuaries
that may be studied by CEB in the future, will be obtained and evaluated. If sufficiently
promising photo images are located, a contract to delineate such images and then
geocode and orthorectify the resulting polygons (assuming post-positioning of a
-------
131
sufficient number of photo-visable ground control points) may be sought. The
methodology for this approach, using either mono or stereo aerial photographs in
conjuction with computer-assisted analytical plotters, has been developed over the last
decade by OSU's aerial photogrammetry expert, Professor Ward Carson (Carson,
1985; Warner and Carson, 1991, 1992; Warner et al. 1993). This technique also may
be employed for geocoding and orthorectifying polygons (delineated SAV or
burrowing shrimp distributionsons) on CIR or FC 35-mm oblique photos taken by CEB
scientists from a small plane at low altitude over YBE. This technique already is being
used by WED scientists to monitor flooding of riparian zones in the Willamette Valley
(S. Klein, personal communication). Alternatively, a Windows-NT computer program to
produce orthorectified digital photos (from geocoded photo-visable points and the
Digital Elevation Model of the estuary) will be obtained for use by the Branch GIS
specialists.
In collaboration with the research proposed by Dr. Robert Ozretich and by David
Specht, beginning in summer 1998 near-surface and near-bottom water column
measurements (Callaway and Specht, 1982; Callaway et al., 1988) of Turbidity will be
initiated at 5 selected continuous sampling sites where the in situ sensors can be
securely deployed (Figure 10). Collections of water for gravimetric determination of
Total Suspend Solids (TSS) also will be made randomly at these stations in an effort to
establish a significant relationship between Turbidity and TSS. Replicate samples of
surficial bottom sediment will be collected quarterly at these stations for measurement
of Total Organic Carbon and Nitrogen (Percent TOC/N) and Grain Size Distribution
(Percent Sand, Silt, and Clay). Near-surface and near-bottom Turbidity measurements
(again with random samples for TSS) also will be made twice during each quarter at
more than 60 probability-based stations throughout the estuary. As discussed above,
these stations are expected to include approximately equal numbers of sites within and
outside areas containing beds of Z marina. A surficial bottom sediment sample for
determination of TOC/N concentrations and Grain Size Distribution also will be
collected at the time of the water column in situ measurements. An empirical model
relating the data from the continuous sampling sites and the probability-based sites will
be developed under the guidance of WED statisticians. Based upon the
recommendations of these statisticians, the results of these water column and surficial
bottom sediment measurements made during the three year study period (1998 - 2000),
in conjunction with those obtained from the corresponding three annual (summer) SAV
distribution surveys described above, will be used to test hypotheses H0(2) and H0(3).
Objective 2: The most recent comprehensive bathymetry of YBE was compiled by the
National Oceanic and Atmospheric Administration's (NOAA) National Ocean Survey
(NOS) in 1953. This information has been used by the Branch GIS group to produce a
bathymetric chart of the estuary (Figure 11). Because the data for this chart are
approximately 50 years old, during a week of extreme high (daylight) tides during Fall
1998, a bathymetric survey of YBE will be conducted. Using DGPS for continuous
-------
132
positioning (accurate to within +/- 0.5 m), depth soundings via a fathometer mounted on
a shallow-draft boat will be conducted along predetermined transects. A filter to remove
most of the variation in depth measurements due to waves will be used, resulting in an
estimated accuracy of 0.04- 0.08 m in the fathometer readings. Absolute depths will be
determined from tidal height measurements obtained from the continuous tide gauge
situated at the HMSC dock. Using laser-survey techniques, the tide height at various
sectors of YBE during the survey will be related to a local Vertical Datum, thus
permitting corrections for spatial (up-estuary) vertical displacement owing to substrate
grade and atmospheric pressure differential. The proposed extent of this bathymetric
survey is from the mouth of YBE at the Ocean Breakwater to the mouth of Olalla Slough
southeast of the City of Toledo (Figure 10). Thus, the area off Boone Slough also will
be obtained. This area includes the sites of the "eelgrass" bed reported in 1968 by the
U.S. Fish & Wildlife Service (USFWS, 1968), and the Z marina bed studied in 1974 by
Bayer (1979); as described above, both beds were determined (from the ground data
survey and CIR photographs) to be virtually absent in 1997. Thus, bathymetry for areas
of both extant and previously-existing beds of Z. marina will be determined. The
absolute depths will be compiled in a digital map and entered into the CEB
Geographical Information System for comparison with corresponding summer 1998
SAV distributions, to determine their substrate ranges and test hypothesis H0(4).
Objective 3: Beginning in summer 1998, cores of bottom sediment will be collected
from different sectors of YBE for geochronology (sediment core layer dating). A
traditional approach in such a project is to seek out likely depositional sites in the
vicinity of the various habitats of interest. An example of such a conceptual sampling
plan is illustrated in Figure 10. In this plan an effort initially was made to select areas
"to the side" of the main flow of water within various sectors of the estuary, in the hope
of finding areas sufficiently exposed to sediment-laden waters to capture a long-term
record of net deposition rates, but sufficiently removed from areas of periodic or
episodic high-velocity currents that such sedimentary records could persist. For
example, one likely area to find an undisturbed depositional record is the uppermost
part of the intertidal zone, in the salt marshes at the estuary's edge. The approach was
used in a preliminary study conducted in Willapa Bay, WA. In collaboration with Dr.
Brian Atwater (USGS - University of WA), Professor James Phipps (Grays Harbor
College, WA), and Professor Brent McKee (Tulane University, LA), large-volume cores
were collected from shallow (~1.5 m) pits dug into natural marshes at river sites known
to have experienced a subsidence of about .1 meter approximately 300 years ago
(Yamaguchi et al., 1997). Preliminary results from Lead-210 (Pb-210) geochronology
(Figure 12) show that, below a mixed surface layer a few cm deep, there is a regular
decrease in the concentration of Pb-210 indicating a sediment accumulation rate of
approximately 0.25 + 0.05 (wet) cm/yr. This is consistent with the clearly visible "buried
marsh soil" horizon at a depth of about 120 cm at this site (and the probability that the
sediment accumulation rate was higher during the first century following the January
1700 subsidence - when the drowned estuarine marsh was inundated frequently, than
-------
133
was the rate during the last century - when the accumulated sediment had raised the
substrate elevation and tidal inundation was much less frequent). Unfortunately, the
practical limit of Pb-210 geochronology for estuarine cores (with relatively high
deposition rates) is about 100 yr BP (Before the Present). This precludes a direct
measurement of hypothesized alterations in sediment accumulation rate between the
time of initial European colonization (-1850 A.D. - 150BP) and about 1900, during
which interval the Old-Growth forests of the Pacific Northwest first were logged.
However, the introduction of much more efficient logging techniques since the turn of
the century, and the increased demand for timber since then, could have led to a
significant increase in soil erosion, and mass "sediment wasting" via slides. This in
turn could have caused an increase in estuarine sediment accumulation in YBE since
the early 1900's that may be measurable.
Thus, the sediment core sampling plan proposed here will include salt marshes of YBE,
seeking evidence of a change in sediment accumulation rate, and a correlation
between the time such a change is initiated and the time of a major alteration in the
YBE watershed. However, to permit conclusions to be made regarding changes in the
very recent sediment history of the estuary as a whole, the probability-based sampling
strategy described above for the other elevation zones also will be utilized. Owing to
the limited number of cores that can be analyzed during the study period, the
probability-based stations will be adjusted according to pre-established rules to obtain
nearby sites that include, where feasible, the following classes:
1. Salt marshes (supra intertidal mudflats)
2. Beds of exotic eelgrass Z. japonica (upper intertidal mudflats)
3. Beds of mixed green algae (intermediate intertidal mudflats)
4. Beds of endemic eelgrass Z. marina (lower intertidal mudflats)
5. Sites of known or suspected previous distributions of Z marina
6. Subtidal non-dredged "slopes"
The Channel proper is not proposed for sampling, in view of the low probability of long-
term (decadal) sediment accumulation owing to flood runoff scouring or dredging.
If consistent with the recommendations of the WED statisiticians assisting us in the
development of the sampling design, cores will be collected throughout the estuary at
(or near) probability-based stations which fall within the above-described classes.
At subtidal stations, in collaboration with Dr. Dixon Landers (WED), a custom coring
device deployed from the WED sediment coring vessel will be used to obtain cores up
to 1 m in length. At mudflat stations, a 6-inch diameter core barrel will be used to
obtain the 1-m sediment cores by hand. At salt marsh stations, pits will be dug and
stainless steel three-sided box corers will be inserted horizontally to obtain the 1-m
cores. The subtidal and mud flat samples will be maintained in a vertical position until
-------
134
returned to the CEB laboratory clean room and completely extruded; all cores will be
sampled at 1-cm intervals and frozen for analysis of Pb-210 by Dr. McKee via chemical
separation and alpha-ray spectrometry (Clifton and Hamilton, 1979; Nittrouer et al.,
1979, 1983/84; Carpinter et al., 1982; Huh et al., 1990), or at CEB via gamma-ray
spectrometry of whole samples in sealed containers at secular equilibrium (Cutshall et
al, 1983; Appleby et al., 1986). Cesium-137 (Cs-137) profiles also will be obtained at
the CEB laboratory on selected core samples, using the nuclear fallout peak of 1963 as
an independent horizen to check the dating via the Pb-210 method.
These data will be used to document approximate dates of deposition of individual
sediment core layer samples during about the last 100 years. The principal objective of
this phase of the research plan is to determine if there have been any significant
changes in sediment accumulation rate or composition over the last century, and the
degree to which any increases in accumulate rate may have elevated the substrate in
parts of YBE. In particular, is there evidence that areas previously vegetated by the
endemic eelgrass Z marina, or present beds of mixed green algae, have shoaled
enough to exceed the present optimal range of Z. marina here. A comparison of the
present bathymetry in such areas with that of the 1953 NOS survey, and with early
bathmetric data, also will be made in an effort to answer, or shed light on, such
questions.
As a part of this reseach project, the collaboration of a specialist in the "historical
reconstruction" of PNW watersheds will be sought to document the history of such
alterations in the YBE watershed. The goal will be to detect sufficiently precise data,
from both the geochronology and the historical reconstruction, to obtain statistically-
significant changes in rates, so that correlations between the times of such changes
may be sought. For example, is there a correlation between the time of a major
watershed alteration (e.g., a distinct change in logging rate or method), and the time
when a significant increase in sediment accumulation rate occured. The guidance of
WED statisticians will be sought in an effort to obtain data adequate to permit the
testing of hypothesis H0(5).
Expected results and benefits: This research project is expected to answer/address
the above-stated Question and Hypotheses which are of direct relevance to the
Agency's goal of reducing uncertainty in the risk assessment process. In particular, the
results should benefit EPA and other environmental scientists and managers
concerned with minimizing the impact of anthropogenic activities on the relatively-small,
non- or lightly-urbanized estuaries of the Pacific Northwest.
-------
135
Project C2 - Evaluation of the Susceptibility of Eelgrass Beds in Oregon
Estuaries to Changes in Watershed Uses
Principal Investigator: Robert Ozretich
GOALS: The overall goal of the proposed research is to assess the susceptibility of
eelgrass beds to negative changes in water quality parameters caused by watershed
alterations. This requires:
1. Determination of the vulnerability of Zostera marina (Zm) habitat in coastal Oregon
estuaries to changes in water column conditions.
2. Determination of a sufficient set of parameters for the calculation of Zm production
in Yaquina Bay.
3. Determination of the sources of suspended matter and nutrients entering the water
column of these estuaries.
4. Determine the water column response to changes in the anthropogenic sources and
masses of suspended solids and nutrients entering Pacific coastal estuaries.
RATIONALE: Submerged aquatic vegetation (SAV) provides food for waterfowl and
critical habitat for fish and shellfish. It also affects nutrient cycling, sediment stability
and water clarity. The extensive experience of east coast researchers with the impacts
of urbanization on SAV has resulted in a variety of models explaining the relationships
between eutrophication, suspended solids and the health of SAV. Dennison et al.
(1993) proposed water quality values (WQVs) that were estimated using
correspondence analysis on five water quality parameters that were associated with
sites of varying SAV cover in Chesapeake Bay. The proposed WQVs for Chesapeake
Bay are salinity-dependent, and include light attenuation, total suspended solids (TSS),
chlorophyll a, dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphorous
(DIP) (Table 1). As an initial assessment of the current status of Oregon estuaries as
suitable habitat for SAV, these water quality parameters could be determined and
individually compared to the proposed WQVs. Interpretation of the results would need
to consider that the dominant estuarine SAV in this region is Zm.
A recent areal and ground survey of the intertidal area of Yaquina Bay in 1997 has
discovered that a complete loss of a significant Zm bed in Yaquina Bay has occurred
since this bed was last reported in 1974 (Bayer, 1979), and apparent reductions in the
extent of other beds have occurred since they were last described in 1971 (as reported
in Good, 1975). Near the missing Zm bed (river distance -15 km) nutrient and TSS
WQVs were exceeded during the early spring of 1977 (Callaway et al., 1988 and
unpublished nutrient data). The spring growth period has been shown to
-------
136
be critical for continued survival of Zm in Chesapeake Bay (Moore et al., 1996 and
1997). Average river flow in April in Yaquina Bay is 50% of the winter maxima (NOAA,
1988) which is the month that the growth season of Zm begins (Kentula and Mclntire,
1986). A critical issue is whether watershed use practices have changed in the
Yaquina watershed between the 1970s and the present in ways that contribute more
solids to the still-significant river flows at this critical time of year. A second critical
issue
Table 1
from Dennison, et al. (1993)
Water quality characteristics that are necessary to sustain SAV to a depth of -1 m
MLW in Chesapeake System.
Salinity regime
Light
attenuation
coefficient
(K.inr1)
Total
suspended
solids
(mg/L)
Chlorophyll a
(M9/L)
Dissolved
inorganic
nitrogen
(pM)
Dissolved
inorganic
phosphoroL
(MM)
Tidal
freshwater (0-
0.5%o)
<2.0
<15
<15
—
<0.7
Oligohaline
(0.5-5%o)
<2.0
<15
<15
--
<0.7
Mesohaline
(5-18%o)
<1.5
<15
<15
<10
<0.3
Polyhaline
(>18%o)
<1.5
<15
<15
<10
<0.7
is whether changes have occurred in the point and non-point discharges of algae-
stimulating nutrients to the upper reaches of the estuary.
In a recent compilation of existing published and anecdotal information regarding the
eutrophication status of Pacific coast estuaries, (NOAA, 1997 draft) no SAV information
was available for 14 of the 38 estuaries (11% of the region's estuarine surface area)
and no information was available on an additional 8 estuaries (26% of area) related to
changes in SAV coverage between 1970 and 1995. Most of the estuaries for which a
minimal amount of information was available were the smaller, outer coastal estuaries
of the region, including Yaquina, Siletz, Alsea, Netarts and Tillamook Bays. Among
these estuaries there is a five-fold difference in freshwater residence times and a range
in extent of watershed urbanization.
-------
137
With the baby-boom population of the U.S. approaching retirement age and the
attractiveness of coastal areas for retirement, the pressures on coastal Oregon to
accommodate a growing population will only increase over the next 25-40 years. As
these people arrive their needs for housing, sewage disposal, drinking water, shopping
centers, roads, etc. will impact the water quality of the estuaries around which they will
live. The extent of water quality changes in response to added nutrients and/or solids
would need to be estimated by accounting for source-dependent nutrient processes
and the unique hydrodynamics of each estuary. These considerations would be part of
a comprehensive watershed plan and would include estimates of the total maximum
daily loads (TMDLs). A component of such an analysis would be an estimation of the
response of the estuary to these inputs that could come from a comprehensive
estuarine ecosystem model such as that of Hopkinson and Vallino (1995). This model
was designed to account for residence times, sources and rates of nitrogen inputs and
losses, and includes both settling algorithms for particulate matter and a 1-dimensional
advection-dispersion model representing the hydrodynamics of the estuary. Future
responses in the Dennison et al. water quality parameters (chlorophyll a and nitrogen
concentrations) could be predicted with this model by varying those anthropogenically-
controlied parameters to which these parameters respond.
OBJECTIVES: There are four major objectives of the proposed research. These are:
1. Determine in Yaquina estuary the temporal and spatial distribution of five water
quality parameters (Table 1) that have been associated with growth of SAV.
Differences between these parameters and the WQVs proposed by Dennison et al. will
be compared with the distribution of Zm. Regional WQVs for Zm will be proposed if
there are significant differences between Zm distributions and those predicted from
measured water column parameters using Chesapeake Bay-derived WQVs. In
addition, the role of light attenuation and desiccation in the distribution of Zm will be
evaluated.
2. Determine the consequences of using different sources and kinds of data in
calculating seasonal production of Zm. The hypotheses to be tested include equality of
discrete and continuously determined light attenuation coefficients, equality of
calculated and measured in situ irradiances. In addition, the comparability of full
spectrum light sensor readings (under defined conditions) and PAR measurements.
Determination of the consequences of using averaged attenuation coefficients in
calculating seasonal production of Zm.
3. Determine the proportions of different sources of suspended matter and nutrients
entering the water column using a variety of means including isotope ratios,
carbon/chloropyll a composition, multiple component mixing models.
4. Provide a model that accurately reproduces the distribution of water column
components that are linked to the habitat requirements of Zostera marina. Such a
-------
model should account for the most important physical and biological processes that
control the water column parameters affecting Zm growth. The accuracy of the
Hopkinson and Vallino (1995) model in reproducing measured distributions of
phytoplankton, dissolved inorganic nitrogen, and salinity in a single estuary (Yaquina)
will be determined. If sufficient agreement is obtained, algorithms for tides and the
coupling of suspended solids and phytoplankton to water column light attenuation will
be added.
SCIENTIFIC MERIT OF OBJECTIVE 1: Given a suitable substrate, the distribution of
sea grasses appears to be controlled primarily by the availability of light that in turn is
affected by water quality parameters. Light attenuation, nutrient, chlorophyll a and
suspended solids concentrations have been found to correlate with the presence or
absence of sea grasses in Chesapeake Bay (Dennison et al. 1993; Batiuk et al., 1992)
and the success or failure of Zm transplants in that system (Moore et al., 1996; Batiuk
et al., 1992). Water quality values (Table 1) that are associated with permanent SAV
beds to a depth of -1 m (MLW) in the Chesapeake estuary have been proposed
(Dennison et al. 1993) and strategies for restoration of beds to greater depths have
been articulated (Gallegos, 1994; Gallegos and Kenworthy, 1996; Batiuk et al., 1992).
Although nutrient limited systems would, generally, not respond with measurable
increases in concentrations of the limiting nutrient (Ryther and Dunstan, 1971), the
interactions of temporally competing limiting constituents apparently results in the
associations of elevated nutrient concentration and diminished SAV abundance in the
Chesapeake system.
The utility of using water quality to define sea grass habitat requirements is that the
total maximum daily loads (TMDLs) of the identified water quality parameters could be
calculated. TMDLs for various tributaries and reaches of an estuary could be
established through waste load allocation modeling in comprehensive watershed plans
either to maintain current SAV distributions or to attain more extensive distributions that
were characteristic of an earlier, less impacted, time. Watershed models that are able
to incorporate physics and geology, as well as, land-use are only now beginning to
become available (Jay, personal communication). In calculating TMDLs the
contributions of natural estuarine processes to the water quality parameters need to be
assessed before waste loads are allocated to anthropogenic processes (Jaworski,
1981). For example, nitrogen and phosphorous flux continuously from sediment due to
benthic respiration, with water column concentrations varying inversely with the tidal
range, i.e. spring vs neap (D'Elia et al., 1981). Light attenuating particles can be
present in the water column due to particle trapping by estuarine circulation (turbidity
maximum), episodic bottom current shear and wind-induced resuspension (Briggs and
Cronin, 1981). Unique to the west coast is the possible transport of nutrient-rich deep
water into estuaries that was upwelled onto the shelf during periods of continuous
northerly winds (usually during the summer). Each of these processes can contribute
to nutrient and/or particle loading to estuaries and need to be assessed before land
-------
139
use changes in the watersheds are considered.
Objective 1a: The objective of this goal is to determine the vulnerability of Zm habitat in
coastal Oregon estuaries to changes in water column conditions. The WQVs that are
associated with healthy SAV beds in Chesapeake Bay may not be quantitatively
applicable to other regions and may not be entirely applicable to the distribution of a
single species e.g. Zostera marina, even in the Chesapeake system. However,
according to Dennison et al. (1993), these WQVs represent the "absolute minimum
water-quality characteristics necessary to sustain plants in shallow water exceeding
any of the five water-quality characteristics seriously compromises the chances of SAV
survival."
Each estuary will be stratified and sampling stations will be selected using probabalistic
sampling protocols (Stevens, 1994). The length of each estuary will be determined
from the mouth to a predominantly freshwater upstream location. Each will be
segmented into 4, roughly equal (river mile) intertidal reaches that may correspond to
areas of different salinities (if known), or merely distances from the mouth.
Each of the 4 salinity-based (DeBen et al., 1990) reaches of Yaquina Bay (Figure 1)
was segregated into 3 depth intervals, (deeper than -6', channels), low (-6' to +2'), and
medium (+3' to +5'). Sampling locations were selected from the depth intervals in
eachreach using the Stevens (1994) protocols. The deepest (channel) locations would
permit virtually unlimited sampling access but the tidal range of Oregon's estuaries
(10-14') would limit access of the shallowest intertidal locations to a few spring tidal
cycles a year.
-------
140
Figure 1
n
§
~
~
tHVBC
bic^I 4 III* ~
mm
iRvErBErdlVBtB
>y.<4.
883®%
msBmZnbGCB
reach#
ocrt'rm£Fy^aTltijrt3(%ste
=sfc^kTBooaped,7S77
-------
141
The specific null hypotheses to be tested are based on the following:
As a first approximation of this vulnerability, current water column conditions will be
compared to the WQVs of Dennison et al. (Table 1) that were proposed to be sufficient
to permit SAV to grow to a depth of-1m , MLW, equivalent to -1.7' MLLW in Yaquina
Bay. This comparison is reflected in the general null hypothesis, H,: the individual
water quality parameter means are not significantly different from Dennison's WQVs.
This would be tested by comparisons of estuary-wide and reach parameter means
against the WQV. Student's t-test would be used. An interpretation of the null
hypothesis generally being upheld, would be that water column conditions are near to
those that could cause problems for Zm growth. Alternative hypothesis H^a,, where
estuary-wide or reach mean values were generally less than the WQVs, would suggest
that conditions for Zm growth to this depth were in good shape. Alternative hypothesis
. where the estuary-wide or reach mean values generally exceeded the WQVs
would suggest that conditions for Zm growth were presently in poor shape. The
inability to reject the null hypothesis, H, and the possibility of alternative hypothesis
would suggest that causes for elevated water column values should be sought
(Goal 3).
During each sampling interval 3 samples from each depth interval per reach will be
taken. Prior to testing H1( the values from the three depth intervals within each reach
will be compared to one another using ANOVA. This will allow testing of the horizontal
homogeneity of these parameters (H2). Three samples are sufficient to detect
differences of 50% with an assumed depth interval SD of 15%. Although 9 samples will
be taken within each reach, only 6 per reach per sampling time would be needed to
detect differences (H3) of 25% with an assumed reach SD of 15%. All these estimates
are based on a power of 0.8 to detect significant differences at an alpha level of <0.05.
If no differences are found among depth intervals in any reach the parameter values
from all depths will be used to represent the reach in comparisons among reaches, also
using ANOVA. If among depth differences within a reach are found, comparisons of all
reaches will be by common depth. There will be 3 sampling stations for each of the 3
depths and 4 reaches for a total of 36 stations. These stations will be sampled bi-
monthly throughout the year.
Objective 1b: Establishing a relationship between Zm coverage and water column
processes. The WQVs of Dennison et al. resulted from estimates of breakpoints in
multidimensional renderings of measured water column concentrations and the
presence, absence, or transience of SAV populations in Chesapeake Bay. The much
larger tidal range on Oregon's coast compared to the Chesapeake system suggests
that exposure could play a significant role in the vertical distribution of Zm in Oregon
estuaries and needs to be considered. Probit analysis will be used to investigate the
roles of light attenuation and desiccation in Zm distributions because the response
-------
142
variable, percent coverage, is reported categorically.
The percent coverage of sediment by Zm at the water column stations will be
determined by other investigators. Their results will be reported as 0%, 50%, or 100%
coverage and will come from the 3 depth strata. The predictor variables will be light
attenuation (rrv1) and exposure (hours*day1). The daily exposure to air will be
calculated using the tidal model developed by Frick (1996) coupled with the bay's
updated bathymetry (contracted to be done fall-98, D. Young, PO).
SCIENTIFIC MERIT OF OBJECTIVE 2. If the PNW-WQVs are found to be
significantly different from the Chesapeake-derived WQVs it may not be due to
biological differences between the Atlantic and Pacific Zm populations. Although
carbon fixation rates by Zm have been shown to vary among habitats, possibly in
response to variability of the light regime (Zimmerman, et al., 1994, and references
within), little difference in minimal light requirements have been reported (21%±3% of
surface incident, mean±SE) among specimens from the Atlantic and Pacific Oceans
and the Baltic (Dennison et al., 1993). Little is known about the light requirements of
eelgrass in the Pacific Northwest (Olson and Doyle, 1995); it is known that Zm has not
been found deeper than about -6.6 m MLLW in Puget Sound (Phillips, 1984) and in
Yaquina Bay it was found in the 1997 survey between -2 m and +0.5 m of MLLW.
What is very different is the much greater summer tidal range of PNW estuaries, 3 m vs
<1 m in the Chesapeake system. The excursion of highest and lowest tides through
the daylight hours of each lunar month is an imposition of additional variability (Dring
and Luning, 1994; Zimmerman et al., 1994) on the flux of photosynthetically active
radiation (PAR, 400 to 700 nm). Continuous light measurements are the best way to
document the consequences of this varying light field.
Continuous PAR measurements have shown that the in situ light regime in estuaries
can be much more dynamic than expected from sidereal considerations (Zimmerman et
al., 1994) and that production estimates based on average monthly or seasonal
attenuation coefficients can miss periods of elevated attenuation during the spring that
are critical for Zm survival (Moore et al., 1996; Bulthuis, 1997). Although the expected
vegetative growth of Zm in this locale is the Apr-Oct interval (Kentula and Mclntire,
1986) continuous instrument deployments covering the estuary will be throughout the
year. Seasonal Zm production will be computed using continuous PAR measurements
with Chesapeake-derived and Pacific coastal estuary-derived mean attenuation
coefficients. Photosynthesis vs irradiance (P vs I) parameters will be taken from the
literature (Zimmerman et al. 1994) for these calculations. Shoot growth (from the
proposed studies of Boese and DeWitt) will be correlated with nearby continuous PAR
measurements (this study) in Yaquina Bay as an independent estimate of the above-
ground fraction of the seasonal production. The acquisition by the Branch of a
mesocosm facilitly for manipulating seagrass would facilitate the determination of P vs I
and other physiological parameters that may be specific to the race of Zm, or other
-------
143
SAVs present in Yaquina Bay or PNW estuaries, in general.
Objective 2a: The objective of this goal is to evaluate the applicability of periodic
monitoring in delineating the submarine irradiance of PAR in coastal Oregon estuaries.
Failure of Zm to survive from one season to another has been attributed to short
periods (2-4 weeks) of elevated light attenuation during its early growth period. The
occasional late spring freshet on the Oregon coast coupled with the early onset of Zm's
growing season (April) are a combination of circumstances that could be inimical to Zm
survival. Continuous monitoring of PAR irradiance would be ideal in identifying such
events (if they occur) and quantifying their consequences. However, the unit cost of
continuous monitoring of PAR is high and deployment sites are limited because of
security considerations. Our interest in outer coastal Oregon estuaries, some of which
have no docks for deployment, compels us to evaluate how representative are
temporally discrete measurements made at easily accessed locations (channels) and,
can inexpensive light meters that could be continuously deployed give us PAR-
equivalent data. To address these questions continuous deployments of PAR sensors
and the evaluation of a light intensity meter are components (Goal 2b) of this research
project.
Continuous deployments of PAR sensor arrays will be placed on accessible docks or
Coast Guard daymarks within each reach of Yaquina Bay (Figure 1). Reach 2 will have
2 arrays deployed for a total of 5 arrays, bay wide. Each array will consist of 2 PAR
sensors that will obtain 4 to 12 measurements per daylight hour. Four of the arrays
will have the sensors a fixed distance apart (1.0 m) at a fixed vertical position. The ~
Hatfield Marine Science Center (HMSC) array will consist of a near-surface floating
sensor and a deeper, fixed-position sensor. The geometry and structure of these
arrays has not been designed at this time and each may be site specific. Attenuation
coefficients will be calculated from the fixed- distance paired PAR units directly and this
distance will be calculated for the HMSC site from the Frick tide model that is most
accurate (time and amplitude) at the HSMC location. The HMSC floating sensor will be
used in concert with a roof mounted PAR sensor (EPA building ~200 m south). The
submarine and air sensors will be used to document the loss of light through the water-
air interface. This information would be used to evaluate the utility of PAR
measurements obtained from continuously monitoring air-mounted units in simulating
the submarine light field obtained from simultaneously determined submarine
irradiance measures at the 5 sites around the bay.
The specific null hypotheses to be tested are based on the following:
The following null hypotheses will test the representativeness of temporally discrete
light attenuation coefficients in Yaquina Bay. H4: the discrete attenuation coefficients
for a reach are not significantly different from those from the continuously monitored
sites. Daily averaged coefficients from the moored sensors for the time interval of
-------
144
interest (e.g. monthly, n=30) will be compared to the coefficients obtained discretely on
the same time interval. Student's t-tests and will be performed on data from each reach.
The discrete sites will be the channel sites from the random selection process sampled
bi-monthly during the midday highs of the neap and spring tides. H5: the discretely
obtained light attenuation coefficients are positively correlated with the results from the
moored sensors. The general expectation is that light attenuation would be positively
related to reach number, with reach 4 receiving the river load of suspended solids.
This expectation should result in a range of attenuation coefficients. Mean discrete
and continuous values from each reach for each and all sampling periods would be
tested for significant correlations. The expectation is that the short-term variability in
attenuation coefficients from the moored arrays would be greatest in reach 4 as the
consequences of storms as elevated suspended material in the river pass by the
arrays. Longer term variability could result from the growth of phytoplankton in
response to river-borne nutrients (sewage and watershed) or upwelling-injected
nutrients from the ocean. Temporal plots of attenuation coefficients from the reaches
would give qualitative information about the duration of light attenuating events or
processes. Tests of equality of variances at different times within a reach and among
reaches would permit statistically based confirmation of general statements about
variability. Differences between and among variances will be looked for using tests of
homogeneity of variance (H5) such as F-tests for 2 comparisons, or F-max or Bartlett's
tests for several.
Independent of the comparisons of discretely determined attenuation coefficients
among reaches (H3) would be those obtained by comparisons of attenuation
coefficients among reaches from the moored arrays (H6) using ANOVA.
The following null hypothesis will test the utility of the roof-obtained PAR
measurements. H7: the total submarine PAR irradiance at a fixed depth obtained from
continuous submarine monitoring is not significantly different from the irradiance
computed from the in-air PAR unit. These irradiance values will be computed on a
daily basis and compared over weekly, monthly, and growth season (March-Sept)
intervals. The comparisons will be by reach; the daily computed irradiance will utilize
the daily air-to-water loss value coupled with the reach-averaged attenuation coefficient
from the discrete water column sampling. Student's t-tests will be used.
Objective 2b: Instrumentation for continuous, in situ, PAR monitoring is expensive, less
expensive alternatives are needed. Full spectrum (350 to 1100 nm) light intensity
meters (HOBO StowAway LI, Onset Computer Corp., Bourne, MA) will be deployed at
the same depths as the LI-COR spherical meters (see IMPLEMENTATION section) to
evaluate the feasibility of using Onset meters in future, more extensive data collections
at more remote sites. Intercalibration between the LI-COR and Onset sensors will use
in situ measurements of the spectral irradiance at the sites using a LI-COR
spectroradiometer because the Onset sensor is strongly dependent on the spectral
-------
145
characteristics of the incident light and in situ light attenuation is strongly wavelength
dependent above 700 nm (Jerlov, 1968). This attenuation of radiation at longer
wavelengths than the PAR waveband suggests that the Onset sensors may be
integrating radiation essentially in the PAR range at depths exceeding 1 m (Jerlov,
1968). Integration of spectral radiance will be used to calibrate the Onset meters.
The range of the Onset meters is into the infrared and they are twice as sensitive to
900 nm radiation as 650 nm radiation making comparisons to PAR (400-700 nm)
sensors in direct or diffuse sunlight meaningless. However, pure water strongly
attenuates light greater than 800 nm to the extent that, for example, below 3 m in a
Nebraska lake no radiation >750 nm or <375 nm was detected. Using a
spectroradiometer that yields irradiance per nm of radiation we will determine the depth
below which only radiation in the PAR range penetrates. This will be done in each
reach and the maximum depth of penetration of radiation exceeding 700 nm will be
determined. This depth would represent the minimum depth of water that could be
above a deployed Onset sensor for it to be comparable to PAR sensors.
The specific null hypotheses to be tested are based on the following:
Response factors to PAR for individual meters will be determined (significant inter unit
variability, Dr. Annette Olson, personal communication). An array accommodating
several Onset sensors, a spherical PAR unit and a spectroradiometer will be deployed
at the HMSC dock at a fixed depth during a tidal cycle. The fixed depth will be
determined by the expected tidal range and the PAR-only penetration criterion. The
predictably changing water depth provides the varying irradiance to which each sensor
would respond. Under these calibration conditions the following null hypothesis will be
tested H„: there is a positive correlation between the Onset sensor output and the PAR
radiation determined by the spherical sensor and spectroradiometer. If this positive
correlation is found, individual Onset sensors will be attached to the moored PAR
arrays and longer-term correlations between sensors will be ascertained before they
would be considered for solo deployment.
SCIENTIFIC MERIT OF OBJECTIVE 3. If any one of the WQVs are close to being
exceeded, or are exceeded in locations within the estuary, identification of the
cause(s) of the excedence(s) would be needed for rational management of the
watershed for optimum SAV growth. If nutrient WQVs were exceeded, direct sewage
sources would need to be sampled and septic drain field contributions would need to
be measured or estimated from adjacent land uses (Valiela, et al., 1992; Short and
Burdick, 1996), and the contribution of benthic flux would need to be measured or
estimated. Yaquina Bay has municipal sewage (secondary treatment) from Toledo
discharged into it. If light attenuation was found to be excessive, partitioning of the
TSS into inorganic and organic fractions would be needed. The organic components
could be assigned to autochthonous sources (algae or fecal pellets containing algae),
-------
146
or allochthonous sources (originating from terrestrial plants and/or sewage). This
partitioning could be done by determining the carbon and nitrogen isotope ratios in the
TSS in combination with an analysis of the lignin and chlorophyll a content. Particles
from different sources and, the water itself, contribute to light attenuation in ways that
have been modeled (Gallegos, 1994; Gallegos and Kenworthy, 1996). Knowing which
sources are contributing directly or indirectly to the shading of SAV by TSS would aid
managers in making decisions for current or future land use practices.
Terrestrial plants fix carbon through the C3- carbon fixation pathway resulting in tissue
carbon containing less 13C (513C approximately -25%o) than the C4- carbon fixation
pathway that yields 5 13C values around approximately -12%o (Lajtha and Michener,
1994). The del (6) notation in isotope research refers to the following relationship:
^sa " Rstd
5(%o) = x 1000 or (R^R^ -1) x 1000
^std
where is the ratio of the heavier to lighter isotope of the sample and RsW the ratio in
the standard material.
Marine plants tend to have 613C signatures intermediate between the terrestrial values
and can be species-specific (Lajtha and Michener, 1994, Chapter 7). The use of S13C
values alone to distinguish most carbon sources to sub-estuaries of Hood Canal in
Puget Sound, WA was found to be insufficient (Simenstad and Wissmar, 1985). The
concomitant use of 513C and 615N measures have been shown to provide better
resolution in tracing sources of estuarine organic matter (Peterson et al. 1985), with
terrestrial sources containing 51SN values of approximately -4%o and estuarine detritus
approximately +5%o (Craft et al., 1988). Particulate nitrogen originating from sewage
treatment plants would tend to be more enriched in 15N than the original nitrogen
source material (ie. food) because degradative processes tend to leave behind the
heavier isotopes while producing 15N-depleted dissolved products, such as urea and
ammonia. For example, dissolved nitrogen species from sewage sources (ammonia
depleted in 1SN) may be the dominant nitrogen source for up-estuary plankton in the
early summer. It is likely that the salt wedge containing dissolved nitrogen species
from upwelled water (nitrate enriched in 15N, Lajtha and Michener, 1994, Chapter 7)
reaches the up-estuary area later in the summer following spring tides. The 615N
signature of planktonic algae in the upper reach of the estuary may be expected to
become more positive in response to the changing sources of nitrogen as the growing
season progresses. The lignin content of TSS would further elucidate the source of
these solids because the plankton and macro algae, including Zm (but not Spartina
alterniflora), contain virtually no lignin (Hedges et al., 1979; Haddad and Martens,
1987; Ciefuentes, 1991; Goni and Hedges, 1995). In addition, from the lignin
parameters, the plant source of terrestrial material, gymnosperm (conifers) or
angiosperm (broad leafed) could be elucidated (Hedges et al., 1979).
-------
147
Objective 3: The object of this goal is the determination of the sources of suspended
matter and nutrients entering the water column of these estuaries. The sources of
these water quality impacting materials are varied and the goal of such a determination
would be to allocate the load among the identified sources. The approach to be taken
in this study is to determine unique characteristics or combinations of characteristics of
the possible sources and looking at the position of samples along mixing curves of the
sources. For suspended solids, stable isotopes of carbon 13C and nitrogen 1SN,
chlorophyll a, total lignin content, and the distribution of lignin decomposition products
will be used to characterize sources of suspended organic matter. Source solids
(endmembers) that would be considered include river borne material, settled fines from
river backwaters, leaves, needles and wood of vegetation inhabiting the riparian zones
of the river, soil litter from riparian zones of the river, undiluted sewage, axenic cultures
of local phytoplankton, fecal pellet algae concentrates from the reaches of the estuary
(herbivores concentrate phytoplankton but are relatively inefficient consumers), fine
grained surficial sediment from different reaches of the estuary, and various
macrophytes of the estuary, including Zm. Sources of nutrients that would be sampled
include river water, sewage, septic field seeps to high intertidal channels, upwelled
oceanic water, and sediment interstitial water. To help elucidate the sources of dissolve
inorganic nitrogen species the 15N/14/N ratios would be determined in the nitrate and
ammonium components using the methods of Sigman, et al., (1997) and Holmes, et al.,
(1998), respectively. Estimates of source-specific loads will be made using mass
budget calculations based on linear mixing models. Collaboration with watershed
modelers will be sought to help asses TMDLs for the estuary.
SCIENTIFIC MERIT OF OBJECTIVE 4: Segments of coastal Oregon estuaries
generally fall into one of two categories (1a or 1b) in the estuarine classification
scheme of Hansen and Rattray (1966). Category 1a is well-mixed with slight salinity
stratification, generally near the estuary mouth, while category 1b has appreciable
stratification at the head of the estuary, closest to the river input. Given the
hydrodynamic similarities of these estuaries and our desire to link changes in
watershed land use to estuarine processes that control particle formation and/or
distribution, the generalized estuarine model of Hopkinson and Vallino (1995) is a good
choice to evaluate. This one-dimensional, advection-dispersion reaction model
incorporates the generalized hydrodynamics of an estuarine system at steady-state,
using river input, estuary geometry, and dispersion coefficients from the literature.
These physical process parameters are joined with coupled nutrient uptake and
release, and organic matter production/oxidation reactions.
The model is based on the premise that water column community metabolism is
primarily controlled by three forcing functions affecting nutrient availability, and the
flushing of food web components from the estuary. The first is residence time, which
influences the time for uptake of dissolved inorganic nitrogen (DIN), the
-------
148
remineralization of allochthonous organic matter and the flushing of phytoplankton and
zooplankton from the estuary. The second is the DIN to organic nitrogen loading ratio.
DIN is readily available to the algae autotrophs whereas the organic form requires a
remineralization step. If the rate of the remineralization reaction is on the order of the
flushing rate, it would result in much of this nitrogen not being "fixed" in the estuary, but
being lost to the ocean. The third is the lability and nitrogen content of the organic
matter entering the estuary. Sewage and forest litter would be at opposite extremes of
lability for nitrogen, and both are entering many Pacific coastal estuaries. These
physical and biological processes of the Hopkinson and Vallino model lead to
predictions of steady-state distributions of salt, algae, and nutrients which can be
measured throughout the estuary to assess the validity of the model in general, as well
as the specific choices made for the various rates, yields, mortalities, and parametric
ratios.
Although this model is based on assumptions such as steady-state for physical and
biological processes that may be approximately valid only for limited times, the
functions included and the predicted endpoints are appropriate for simulating some
responses to different loading scenarios that would result from changes in watershed
use. As written, the model does not explicitly address light attenuation which is directly
relevant to SAV. Incorporation of the solids-specific (algae, clays) and water-specific
(DOC and color) attenuation coefficients of Gallegos (1994) and a time-varying water
column depth by adding a tidal subroutine (e.g. Vorosmarty and Loder, 1994) would
increase the ecological realism of the model, would increase its utility, and the
predicted light attenuation could be validated by our in situ PAR measurements
(Objective 1a). The authors of the model are in the process of modifying it, and to the
extent possible, we will work with them in making any changes.
Objective 4a: The object of this goal is the determination of estuarine water column
responses to changes in the anthropogenic sources of suspended solids and nutrients.
The model that we will evaluate for these purposes is that of Hopkinson and Vallino
(1995). This model is a 1-dimensional advection diffusion model with biological
components responding to carbon and nitrogen inputs, only. If we find that its
simulated distributions closely mimic measured distributions of phytoplankton,
dissolved inorganic nitrogen and salt we will use it as a backbone to add explicit
algorithms to simulate inorganic particle distributions and tides.
The current model will be run using estuary-specific parameters such as river flow,
dispersion coefficients, estuary length, and those from the literature, such as various
uptake and growth rates. Although dissolved nutrient input from the river is solely
discharge rate dependent the ability of moving water to support TSS is a power
function of this rate and can be watershed dependent (Jay, personal communication).
Boundary conditions (endmember concentrations) will be determined or estimated in
the ocean and river waters entering the estuaries. These will include: autotrophs (as
-------
149
chlorophyll a), heterotrophs, dissolved inorganic nitrogen (DIN), labile organic material
(OM), refractory OM (each in mg parameter nr3) and salt (%o). Model accuracy will be
evaluated through comparisons to concentrations determined at 10 evenly distributed
channel locations along the length of the estuary Figure 1 (Jassby et al., 1997).
The accuracy of the simulations of the conservative parameter salt would be a test of
the basic advection-dispersion elements of the model and is critical to all others.
Salinity is the most easily determined parameter in this model as it would come directly
from our CTD instruments. Physical parameter adjustment/selection to obtain the best
match for this parameter will anchor these critical parameters. Repeated runs with
random-normal distributions (Abramowitz and Stegun, 1965) of biological process
parameters will be used in the evaluation phase using the literature to establish mean
values and their uncertainties. ANOVA will be used to detect the parameters to which
this model is most sensitive by comparing at several single locations the model outputs
for, say, 25 runs for each parameter. The effects of the ranges of parameters available
in the literature will be evaluated with those parameter combinations yielding the
smallest deviation from the measured parameters selected as the likely, estuary-
specific "true°values. Measurements of autotroph concentration as chlorophyll a and
dissolved inorganic nitrogen species concentrations will be made for comparisons to
model run outputs and those parameters affecting these concentrations will be
manipulated for "best fit" to narrow the range of possible values. At this time, outputs:
refractory and labile organic matter, and heterotrophic concentrations are not measured
directly. Dissolved organic carbon, and particulate organic carbon and nitrogen will be
determined and will put bounds upon these components and ratios but they are not
explicit model outputs.
Ojective 4b: Once parameter selection has resulted in a minimum overall difference
between model and measurements, estimates of precision at any location can be made
by compiling the range of results obtainable for reasonable ranges in parameters
including the variation in the measured endmember concentrations. The use of runs
with combinations of random-normal parameter distributions would be equivalent to a
propagation of errors analysis and would yield uncertainties of simulated output
parameter means.
IMPLEMENTATION
Objective 1a, Probabilistic Station Selection: We will work with staff of the Regional
Ecology Branch, who are familiar with the EMAP probabalistic sampling protocols
(Stevens 1994) in selecting stations in the Yaquina estuary. In general, we will break
an estuary into three depth zones that are continuous throughout. These stations will
be sampled during optimum sun angles (local solar noon ± 2-3 h, Miller and
McPherson, 1993) with a minimum of 5 feet of water over the bottom to obtain a profile
at least 1 meter in length. This would result in sampling these sites during a 4-6 hour
period with a sampling water depth of at least 6.5 feet. The number of stations in each
-------
150
zone will likely be in the range of 30 to 50. Since all the required samples will take
more than one 4-hour interval, approximately an equal number of samples from each
zone will be sampled during each interval over up to four 4-6-hour intervals. These
stations will be sampled every other month starting in May 1998. Supplemental
stations will be selected that will provide supportive data to other researchers involved
in Zm out-plantings and productivity measurements (DeWitt and Boese projects)
Objective 1b, Water Quality Parameter Methodology: The stations selected above
will be sampled using a SeaCat Profiler with an attached inlet to a shipboard pump. In
general, instrumental data (CTD, fluorescence, PAR, turbidity, DO) will be collected
through the water column and discrete water samples from the pump will be collected
from mid-depth, but not to exceed 0 ft MMLW. The water subsamples will be
quantitatively filtered for TSS (gravimetric collection) and chlorophyll a. Filtered water _
will be analyzed for water nutrients (ammonia, nitrate, nitrite, orthophosphate, and
silicate), water color, and DOC. Unfiltered water will be analyzed for water nutrients
and chlorophyll a. Analytical methods are listed in Table 2. Turbidity and fluorescence
responses from the SeaCat will be calibrated using the measured TSS and chlorophyll
a concentrations.
Objective 2a, Continuous PAR Methodology: We will obtain a continuous in situ
record of scalar PAR irradiance and attenuation coefficients at sites along the estuary
axis using 4n sensors (LI-193SA with LI-1400 datalogger, LI-COR, Lincoln, NB) at two
fixed depths (-1 m to -2m MLLW). Turbidity, temperature and salinity will also be
monitored continuously at a near-surface depth. Spherical sensors are used because
Zm blades are able to intercept light from all angles (Morris and Tomasko, 1993). Five
relatively secure sites have been chosen for continuous deployments of the PAR
arrays. The locations are the Hatfield Marine Science Center dock complex, Sawyer's
Landing marina, River Bend marina, Oregon Oyster dock, and Critesers Marina (Figure
1). Negotiations for the use of Coast Guard day marks for these deployments are
ongoing. A floating and fixed PAR array at the Hatfield Marine Science Center dock
will permit the calculation of the attenuation coefficient of the entire water column minus
the top 1m. A 2n PAR sensor will be continuously mounted on the roof of the EPA
building to provide in-air irradiance.
Computation of productivity from integration of the continuous PAR records will allow
us to compute carbon fixation during the growing season and compare productivity from
literature P vs I relationships to those obtained at different, or the same, sites in the
estuary using other methods such as leaf growth (see Boese Research Project). The
need for continued continuous deployments of paired PAR sensors and the utility of
PAR profiles will be assessed from the. variability of daily integrated PAR and
attenuation coefficients.
Objective 2b, Alternative to PAR Methodology: The LiCor arrays are expensive
-------
151
(-$2000) and with their surface dataloggers they are especially vulnerable to
interference by the public. Relatively inexpensive, full spectrum (350 to 1100 nm) light
intensity meters (HOBO StowAway LI, Onset Computer Corp., Bourne, MA) will be
deployed at the same depths as the LI-COR spherical meters to evaluate the feasibility
of using these meters in future, more extensive data collections at more remote sites.
Intercalibration between the LI-COR and Onset sensors will use in situ measurements
of the spectral irradiance at the sites using a LI-COR spectroradiometer because the
Onset sensor is strongly dependent on the spectral characteristics of the incident light
and in situ light attenuation is strongly wavelength dependent above 700 nm (Jerlov,
1968).
Objective 3, Anthropogenic Sources of Water Quality Parameters: From a second
subsample of the water pumped from selected stations a sufficient mass of particulate
material will be collected for the analysis of TOC/N, lignin, carbon and nitrogen
isotopes. These results will be combined with the analyses of materials that would be
considered pure end members of a two or more source mixing problem. End member
candidate materials for excess nutrients might be sewage effluent, estuarine fecal
material, decayed Zm and leaves of local deciduous trees. Analysis of water from
estuary side channels, septic seeps, sewage out falls, and benthic flux chambers would
contribute to the segregation of anthropogenic from "natural" contributions of nutrients.
Objective 4a, Model Implementation and Verification. The model of Hopkinson and
Vallino (1995) was developed at the Ecosystem Center of the Marine Biological
Laboratory in Woods Hole, MA and represents the synthesis of their thoughts on
linking runoff and water column metabolism. This model was in part funded by the
Land Margin Ecosystem Research (LMER) program of the National Science
Foundation and was validated using parameters and values from Parker River/Plum
Island Sound, MA, a LMER site. We have the model up and running and have
reproduced some of the published distributions.
a) Discrete water sample concentrations and SeaCat instrument values will be used
for boundary (end member) concentrations of salt, nutrients, etc. Initially, default model
parameters, Yaquina Bay geometry and diffusion coefficients (Callaway and Specht,
1982) will be used for model runs. The need to add a second or third dimension to our
modelling considerations will be made in consultation with EPA collaborator Pete
Eldridge and consultant Dick Callaway. In addition, coupling of this bay model to a
shelf-regional model may be necessary to capture the dynamics of upwelling and net
tidal flux of model components. This will be accomplished through collaboration with
new offshore oceanographic-meteorologic program that is to be administered through
Oregon State University. Boundary concentrations will be determined during maximum
and minimum river flow periods and periods of upwelling. In addition, hourly sampling
of physical water column characteristics will be sampled for 24 hrs during neap and
spring tides during the winter and summer seasons.
-------
152
b) To assess the accuracy of the model in computing concentrations of algae, nutrients,
etc. (Table 3) down the length of the estuary we will deploy the SeaCat Profiler at
evenly-spaced stations (Jassby et al., 1997) at midchannel with discrete water samples
from selected depths at each station.
Objective 4b, Model Modification, Light and Tides, a) We will incorporate light
attenuation linked to inorganic particles and phytoplankton through PAR absorption
and scattering relationships (Gallegos, 1994). This would result in a length-of-estuary
distribution of attenuation coefficients that, through a P vs I relationship, would
delineate the light environments sufficient to sustain SAV growth. We could use the P
vs I algorithms presented by Zimmerman, et al.(1994), and others. For the most
accurate, locally-applied model, growth dynamics of the local race of Zm will ultimately
need to be determined (Objective 1111, above).
b) We will incorporate algorithms to represent tides. With the modeled light attenuation
this would generate a more realistic in situ light environment as the tide-induced water
column depth changes exceed the minimum depth and leaf extension of many Zm beds
in Yaquina Bay.
c) Repeated model runs with random-normal distributions (Abramowitz and Stegun,
1965) of the newly created parameters will be used to evaluate the relative sensitivity
of the modified model.
EXPECTED RESULTS AND BENEFITS: The major products from this project will
be an evaluation of the current susceptibility of eelgrass beds in Yaquina Bay and a
model that could be used to provide results for different future development scenarios.
Estimates of the composition of the suspended material that is controlling light
attenuation will be given. With these results, the effectiveness of different mitigation
efforts to enhance water quality related to eelgrass habitat could be evaluated.
-------
153
Table 2. Parameters needed for calculating Water Quality Values of Dennison et al.
(1993).
Description
light attenuation coefficient
total suspended solids
chlorophyll a
dissolved inorganic nitrogen
dissolved inorganic phosphorus
salinity distribution
Analysis
photosyn. active radiation (PAR) 2 depths/station
mass collection on preweighed membrane filters
acetone/water extract with fluorometer quantitation
AutoAnalyzer
AutoAnalyzer
CTD
Table 3. Parameters needed for Hopkinson and Vallino (1995) estuarine ecosystem model.
Description
Analysis
dissolved inorganic nitrogen
C:N ratios of allochthonous solids
particulate organic C and N
dissolved organic carbon
total suspended solids
flocculation of DOC to POC
water residence time
river and ocean concentrations
estuary geometry
dispersion coefficients
river flow
NH4, N03, N02 AutoAnalyzer
TOC/N
TOC/N (filters and sediment)
DOC
mass collection on preweighed membrane filters
DOC, TOC/N on glass fiber filters
river gauge data with estuary geometry
endmembers for the hydrodynamic model
tidal analysis, GIS
literature values
river gauges (USGS)
-------
154
Project C3. Nutrient Processes: Watershed versus Oceanic Inputs to PNW Estuaries
Principal Investigator: Anne C. Sigleo
Goal: The goal of the proposed work is to define annual dissolved nitrogen and suspended
nitrogen budgets for two estuaries in the northwestern part of the U.S. under various
scenarios for watershed rainfali/run off conditions, oceanic inputs and anthropogenic activities
in the area. A further goal is to establish predictive relationships between readily measured
nutrient characteristics and to use these data as indicators of ecosystem change, or condition,
in PNW estuaries by linking predictive estuarine circulation and watershed models.
RATIONALE: Willapa Bay, Washington is a highly productive estuary that supports over
50% of the commercial oyster production on the west coast. In addition, the bay serves as a
nursery for Dungeness crab and English sole, and as a spawning area for Chinook, chum and
coho salmon. It provides extensive clamming opportunities and is home to shrimp, salmon,
and steelhead and cuthroat trout.
Water quality, including nutrient load, is of concern for many of the stakeholders in Pacific
Northwest estuaries, particularly those involved with aquaculture, who are dependent on water
quality for a marketable product. To capitalize on the nutrient data base developed from
research over the past three years, work will continue in Willapa Bay, Washington. Data from
Yaquina Bay, Oregon will be used to provide a comparative framework, thus strengthening
the potential applicability of the results to multiple PNW estuaries. These two estuaries
encompass a size range of 17 km2 in Yaquina Bay, to 320 km2 in Willapa Bay, one of the
larger PNW estuaries. The respective watersheds for these estuaries vary from 655 km2 in the
Yaquina to over 2000 km2 (1,865-2,849 km2 reported range) in Willapa Bay (Baker et al.,
1995). Both estuaries support aquaculture (oysters, hardshell clams), fishing, and migratory
waterfowl. The Willapa watershed is dominated by timber plantations, along with some
agriculture, beef ranching and dairy farming. Urbanization is minimal, with the exception of
Long Beach Peninsula and the lower Willapa river. The Yaquina estuary lies at the mouth of
the Yaquina River and appears more urbanized with a population of 9,785 (July '96) located in
the city of Newport and Toledo (3,200). The population of the Willapa area (6,884 (1996)) is
distributed around the bay at Long Beach (1,400), llwaco (864), South Bend (1,660) and
Raymond (2,960).
In spite of their importance, sufficient data on nutrients and energy flow through the primary
producers for resource assessments and future planning are lacking for both estuaries.
Hydrographic input to Pacific Northwest estuaries is related to seasonal variations in rainfall
(Peterson et al., 1982 and 1984). During winter months of peak river discharge, terrestrial
material is transported downriver into the estuaries. During the summer months, however,
beach sand from the continental shelf is transported into the estuaries by tidal currents
(Scheidegger et al., 1971; Peterson et al., 1984; Schwatrz et al., 1985). Nutrient inputs to
-------
155
PNW estuaries are thought to be similarly seasonal, although only sporadic data are available
to support this theory.
Sufficient levels of nutrients are required for organism growth, health and reproduction. Above
the optimum levels, however, high nutrient concentrations can alter the ecosystem by favoring
species that flourish in high nutrient conditions. The recent toxic bloom of Pfiesteria sp. in
subestuaries of Chesapeake Bay was the result of excessive nutrient loading (Milot, 1997).
The reoccurrence of toxic algal blooms also is a major concern in Pacific coastal estuaries
(Sayce and Horner, 1996; Horner et al., 1997). The initiator or stimulus for the organisms to
produce the toxin is organism stress, particularly that due to eutrophic levels of nutrients (Pan
et al, 1996a and 1996b; Sivonen, 1996; Kotaki et al., 1996; Bates et al., 1996; Paerl, 1996;
Paerl and Millie, 1996; Christoffersen, 1996 ). In Willapa Bay, Yaquina Bay and the adjacent
ocean coastal zones, toxins above a legal threshold result in closure of shellfish beds and the
sale of oysters is prohibited. This results in a substantial loss of income for the oyster growers
as well as a loss of food source for members of the Shoalwater tribe in Willapa bay.
In addition to harmful algal blooms, the consequences of high nutrient loading include oxygen
depletion as a secondary effect due to oxygen uptake by microorganisms during the
breakdown of large amounts of algal biomass. Thus, a nutrient budget, including nutrient
sources, would benefit planners considering environmental impacts such as sewage disposal
from increasing populations. A method of quantifying and stating nutrient loadings are TMDLs
(total maximum daily loadings). The data for mass balance goals also provide fundamental
data for supporting TMDLs.
The results of analyses to date provide data on nutrients affecting the primary producers at
the base of the aquatic food web for resource assessment and balanced resource
management for Willapa Bay. Nutrient concentrations, phytoplankton biomass concentrations
and phytoplankton species abundances were determined in samples from at least five stations
for over 34 months in collaboration with K. Sayce, Shoalwater Botanical Laboratory, Nahcotta,
WA. The results indicate that high loadings of nitrate (up to 1.2 mg/L) were entering the
estuary from coastal waters, rather than the rivers of the Willapa watershed. The 1995, 1996
and 1997 water years were unusually cool and wet with significantly higher rainfall, nitrate
nitrogen concentrations (up to 1 mg/L) and settled plankton biomass (up to 150 ml/L
phytoplankton + zooplankton). Dissolved nitrate values decreased steadily throughout the
spring and summer, but did not disappear entirely. Ammonium and phosphorous were not
detected in the main channels of the estuary, although traces of ammonium were measured in
the Willapa and Palix Rivers. Phytoplankton, dominated by diatoms, also continued to bloom
steadily throughout spring and summer. The 1995 nitrate values are comparable to those
reported by the USGS for water years 1969 and 1970, also wet years with similarly high
runoff. The results provided nutrient data of sufficient resolution to detect at least two primary
nitrogen nutrient sources for Willapa Bay. Future work will continue to document the
quantitative input from each of the sources. The water column baseline data on nutrient
concentrations and variability can be utilized in circulation modeling and oyster condition
-------
156
index studies.
OBJECTIVES: The purpose of this research project is to assess the magnitude of each
component of the nitrogen biogeochemical cycle, along with supporting information on
phosphorous, carbon and silica, to evaluate potential nutrient limitations or sufficiencies for
phytoplankton productivity and species composition of the above elements via Redfield ratios.
Availability of nutrients is a major factor controlling phytoplankton primary productivity and
species composition (Golterman et al, 1983), including species that can produce toxic blooms
(Pan, et al. 1996a and b).
The specific objectives of this work are:
1) to determine the net fluxes of nitrate, nitrite, ammonium, dissolved organic nitrogen
(DON), reactive phosphate, and dissolved silica between estuarine, river and
ocean waters.
2) to determine whether PNW estuaries act as net sources or sinks of inorganic
nutrients at the current levels of nutrient loading in Willapa and Yaquina
estuaries.
3) to examine source-sink relationships within intertidal communities and estimate net
fluxes of nutrients across the sediment-water interface.
4) to examine wet depositional inputs of nutrients. This is not a likely PNW source
(Weathers et al., 1988), however, some measurements are required to prove or
disprove the idea.
Scientific Approach: 1) To determine nutrient fluxes, samples will be collected weekly.
Phytoplankton in estuaries respond rapidly to light and changes in water column nutrients so
that weekly samplings are the longest time interval appropriate to capture the effects of
nutrients on plankton biomass and productivity. We have previously supplemented the weekly
sampling scheme with hourly to two-hourly diel cycle samplings over three day periods to
follow individual water masses (Sigleo et al, 1997) and were able to document an upwelling
event. A similar approach is proposed again, along with the additional data collection via a
continuous (every 15 minutes) nitrate monitor. The five stations used for weekly sampling are
representative of the major parts of Willapa bay. The stations are located at 1) Naselle
Channel, representative of the Naselle River outflow; 2) Nahcotta Channel, representative of
the central-southern bay; 3) Palix River BCM dock for the Palix River that is available at all
times to Bay Center Mariculture personnel; 4) Riddle Spit Channel between Riddle Spit and
Nemah Spit, representative the northern center of the bay; and 5) Stackpole Slough west of
Nahcotta Channel is located over oyster flats and represents oyster finishing beds (Fig. 1).
Samples were collected at high slack tide for phytoplankton species, salinity, temperature and
turbidity during 1993 and 1994. For data continuity, samples were collected for nutrient
analyses (1.5 m depth) simultaneously with samples for phytoplankton species identifications
and counts from the same five stations in 1995, 1996, and 1997. In 1997, additional stations
were added at 1) the Pacific Ocean to quantify the nutrient content of water masses entering
-------
157
and exiting Willapa Bay, 2) Toke Point at the northern end of Willapa Bay in the North River
and Willapa River mixing zone, 3) Naselle hatchery, 4) Wiilapa hatchery, 5) Loomis Lake as a
measure of long term atmospheric input, and 6) Island Lake, a recreational lake surrounded
by homes, to study the septic tank input. The hatchery samples (collected above the hatchery
water inflow) are representative of the river nutrient inflow and are adjacent to the USGS river
gaging stations above the head of tide. The Yaquina Bay sampling station is at the terminus of
the OSU seawater intake dock and has the advantage that nutrients can be analyzed within
minutes of collection. All samples are collected at high slack tide, except for specific
experiments or where otherwise noted.
To determine the input of nitrate nitrogen in PNW estuaries, and particularly the net input flux
of oceanic nitrogen requires continuous measurement over seasonal cycles. Because the
oceanic water enters and exits twice daily on tidal cycles the method of estimating the amount
entering, utilized within the estuary and leaving the system requires continuous
measurements to model the appropriate functions from the acquired data. To accomplish this
goal, we propose using an Ocean Systems Ltd nutrient monitor that can measure nitrate
hourly up to 30 days. After discussing the problem with numerous oceanographers and
oceanographic modelers, it was concluded that continuous monitoring of the major nitrogen
species, namely nitrate, over all tidal cycles, seasons, and rainfall/runoff conditions was a
reasonable means for obtaining the required data. To determine the amount of seawater in a
water mixture, the element bromide is being used as the indicator of seawater (ie, percent
seawater). Bromide can be measured very precisely and accurately by ion chromatography in
our laboratory, along with the nitrate from manually collected samples. By measuring nitrate
at a specific point just inside the mouth of Yaquina Bay continuously, the amount of nitrate can
be approximated using Newport's available tide model by Walter Frick. This tide model is able
to predict velocity under the Yaquina bridge to within 90% accuracy (Frick, pers. com.). The
Yaquina calculations are simplified due to the very confined channelization of the water by
jetties. From the simplest perspective, tide flat circulation modeling can wait until additional
resources become available, and refinements to data collection and calculations can be
made.
The instruments for more detailed modeling include acoustic Doppler profilers (ADPs)
for 3-D current profiling on the tidal flats and in the channels. Drogues used in conjunction
with the current profilers will provide actual water paths at desired depths to refine models for
both surface currents and deeper ones. For example, unknown circular loops or eddies
concentrating nutrients, suspended particulate material and biota within a bay have been
identified in other estuaries this way (Science, v281:196, 1998). The tide flats themselves
are a known source and sink of nitrogen species, and to address that issue, an NRC postdoc
will begin working in April of FY99.
In order to calculate a flux, concentrations are typically measured at a single point (ie, a
discrete point). One requires a fixed point to make any sense of the data. To relieve
concerns that the station chosen for deployment of the NAS nitrate analyzer may not be
-------
158
representative of the Bay entrance, we plan several cruises at different seasons and tidal
cycles to measure nitrate around the analyzer location at different depths, and under the
Yaquina bridge to the jetty entrance. A second nitrate analyzer on a navigation buoy 15 or 20
miles outside the mouth of the Yaquina would improve the data set substantially.
Nitrogen fluxes are equal to the nutrient concentration times the mass water flow and
the mass balance of an analyte is equal to the flux times time. In Willapa Bay, we propose
mounting the nitrate monitor on a navigation buoy between Riddle Spit and Nemah Spit.
Although the bay entrance would be ideal, we have yet to devise a workable mooring because
the entrance is continuously shifting. In Yaquina Bay, the nitrate monitor is mounted on the
terminus of the OSU dock at the mouth of the bay.
The nutrients studied are listed in Table A.2.3-1 and include nitrate, phosphate, and
ammonium, all of which may be quantified reliably by ion chromatography using EPA Test
Method 300.0 or spectrophotometric methods (Pfaff et al., 1991; Fong et al, 1995; Strickland
and Parsons, 1977). Bromide, measured with the nutrients by ion chromatography, is used as
an indicator of the oceanic component in a water mass. The assessment of nutrients also will
include silica (Pan et al., 1996a).
During winter months of peak river discharge, terrestrial material is transported downriver into
the estuaries. To quantify the river sediment loads and the nutrients contained therein,
suspended sediments will be sampled intensively over the first and subsequent rainfall/high
river flow events of the fall and winter ( between October 1, 1998 and March 1, 1999 ) at the
Willapa River station near Willapa, the Naselle River station near Naselle and the Chitwood
station on the Yaquina River. Samples will be collected according to procedures outlined in
USGS Techniques of Water Resource Investigations (TWRI) publication "Field Methods for
Measurement of Fluvial Sediment". Samples shall be analyzed for suspended sediment
concentration by weight, and relative grain-size fractions. Filtered water samples (60 ml) will
be collected for dissolved nutrient analysis and frozen as soon as possible. These data will be
combined with the hydrograph data obtained at the gaging stations by the USGS (Tabel 2) to
produce estimates of suspended sediment loadings to the estuaries.
-------
159
L4€»llc-n
r,«;
x> ¦»< ¦» ¦< < :
4;
X* < < •* x> > >
>~ >vs •< **¦>>>:
* \f*. •» »•
& <*"*";<* s< <:
IIS*
5,5; ^ : '"V,;
Vsl^r
ci : -
r ?r "?- «v«
cm?;":
'.' «\,v c
wy:<-;o .•x-»::v;o;>
|^'i^:':-:':":^::;v:;?;iv!;:;:
U.KXV
Shm?
* •¦>< -<9..-*•>»•> >
gssss
' «. -.V
hhhihOiihi'
Willapa Bay
Watershed
Figure 1. Willapa Bay sample collection locations at Naselle Channel (NS), Nahcotta Channel
(NH), Palix River (PR), Riddle Spit Channel (RS), Stackpole Slough(SP), Toke Point (TK),
and Pacific Ocean (PO). The USGS water flow measurement stations are labeled with a G.
-------
Table 1. Water Sample Parameters
Sample type
Parameter
measured
How measured or
obtained
Percent of
samples
Nutrient
N, P,
Si, NH3
HN03
Ion chromatograph
Spectrophotometer
Nitrogen monitor
100
20
80
Temperature
Degrees C
Thermometer
100
Turbidity
Depth
Secchi disc
50
Salinity
Density
Hydrometer
75
Phytoplankton
group
Pigments
HPLC
10
Ocean vs River
source
indicators
Stable isotopes
Bromide
Isotope mass spec
(P.O., Coop, etc)
Ion chromatograph
15 (contract)
100
Table 2. Willapa Watershed and Yaquina River Hydrographic Data
Name:
Gage Number
Drainage Area
Data Period
Notes:
North River
12017000
12016600
219 mi2
188 mi2
1927-1977,
1995-1997
1965
Sed data 1965
Willapa River
12013500
130 mi2
1947-present
Chemical data
1965-1986
Naselle River
12010000
54.8 mi2
1929-present
Chemical data
1965-1980
Yaquina River
14306030
71.0 mi2
1973-present
Sed data 73-
74
-------
161
2) To determine whether PNW estuaries act as net sources or sinks, the source(s) of water
column nutrients will be further investigated during summer sampling using diagnostic
indicators such as stable isotopes of carbon and nitrogen whose abundance can be
indicative of unique sources or processes. In particular the oceanic vs river source of nitrate
nitrogen will be examined using nitrogen isotopes (Sigleo and Macko, 1985; Bilby et al.,
1996). We found in other studies that the nitrogen source differed seasonally, with winter
inputs coming from terrestrial runoff of nitrate, whereas the summer source of nitrogen was
remineralized benthic ammonium (Sigleo and Macko, 1985). Suspended sediments for
nutrient studies will be collected via sediment traps (Sigleo and Shultz, 1993) at 2 m depth,
timed to collect over 6-hour tidal cycles. Previously in the Potomac estuary we employed
duplicate vertical arrays of three (surface, pycnocline, and near bottom). In Pacific coastal
estuaries, the water column is generally well mixed vertically. Since vertical features such as
stratification are uncommon, single traps at one intermediate depth are proposed initially.
3)Benthic fluxes, particularly benthic ammonium release from the Pacific intertidal and surf
zones (Lewin, 1977; Sigleo et al., 1997), will be monitored by benthic chambers for four hour
deployments (Boynton et al, 1980; Sigleo, 1985; Cowan and Boynton, 1996) using the
benthic chamber design of Davis and Mclntire (1983) or others. The oceanic input and exit
for an individual parcel of water will be further examined using drogues (Power, 1996). This
technique could be combined with larval recruitment studies for the estuarine fish research of
Power, since larvae and oceanic phytoplankton are transported into estuaries in incoming
water masses, as are oceanic nutrients from upwelling.
A major sink for fixed aquatic nitrogen in many coastal estuaries is denitrification (Seitzinger
et al. 1984), or the conversion of organic nitrogen to N2 by denitrifying bacteria. Seitzinger et
al. (1984) found that the annual N2 production for Narraganestt Bay, R.I., was equal to
approximately 50% of the fixed inorganic N loading to the bay. Denitrification accounted for
about 35% of the organic nitrogen mineralized and removed from the sediments as N2
Denitrification is commonly measured using a 15N tracer technique to quantify in situ
denitrification rates (Xue et al., in press). Gas samples collected over discrete time intervals
are analyzed on an isotope mass spectrometer to quantify the rate of N2 production (Xue et
al., in press).
4) Nutrient input from wet deposition will be measured with water from the rain gages at the
Nahatta, Cranberry center and ridge weather stations. Cloudwater for the Northwest has the
lowest ion content measured (Weathers et al, 1988), suggesting that rain is not a source of
nutrients.
Expected Results and Benefits: The data produced already are being used to predict
oyster growth and condition for the existing scenario for marketing purposes by the Willapa
Bay Oyster Growers Commission, Share Bank Enterprises, Inc., and Taylor United. The
Washington State Department of Fish and Wildlife both funds and uses our phytoplankton
data to monitor for toxic blooms of Pseudo-nitzschia. The results ideally will include a model
-------
162
(HSPF) that can predict environmental impacts based on changes in nutrient inputs (Lumb,
1994), possibly via a postdoc or graduate student.
The results for 1995 through 1997 indicated that winter storms bring nutrients, particularly
phosphorous, downstream. During the spring and summer, however, the primary source of
nutrients is oceanic upwelling. Because oceanic upwelling is a natural process with a wide
range of variability, both the process and the natural range of nutrient supply must be known
to assess anthropogenic effects. It is clear that there is a threshold effect, above which
anthropogenic perturbations would have a significant effect. In other words, anthropogenic
inputs must exceed a nominal threshold to adversely affect the estuary. Determining the
threshold above upwelled nutrients by measuring the nutrients from upwelling and
bioconcentration is fundamental to understanding nutrient processes in Pacific Northwestern
estuaries.
Cross Linkages: Tide models for tidal exchange were developed for the Pacific coast from
Baja to Alaska by Frick, Chapman and Clinton in EPA and are available. A circulation model
developed for Puget Sound by Walters at the USGS in Tacoma can be developed for
Willapa Bay via a 34K IAG and be verified using Walt Frick's Acoustic Doppler Current
Profiler data and Barbara Hickey's current data (supported by Sea Grant and EPA, Newport).
The circulation model is critical to predicting the tidal range of estuarine parameters such as
salinity and turbidity. With respect to circulation modeling, the only way of organizing and
making deductions, including forecasting, with large quantities of data is with a
comprehensive circulation model. Salinity zones within the bays, for example, change
drastically with 1) the tidal input or output, and 2) the rainfall/runoff parameters. Storm
forcing also affects the salinity from the seaward side. To achieve prediction capability,
physical parameters such as the channels that seawater enters versus exits (no, they're not
the same usually, except in confined cases), along with speeds and direction are required.
Then there is the chemistry and sediment transport. Fortunately, Athens considers sediment
transport algorithms a priority item and may provide help with a "test" estuary. Chemical and
watershed models such as HSPF exist and can be combined into the circulation model.
Determining terrestrial inputs into the estuaries from upland watersheds can be modeled to
organize and analyze present and historical data, and to simulate or project effects of future
changes. One widely utilized program in both EPA and the USGS is HSPF (Hydrological
simulation program -Fortran) and auxiliary prediction programs such as HSPEXP (Dinicola,
1990; Lumb et al, 1994). The HSPF watershed model could be accomplished as a student
thesis.
In Yaquina Bay and Coos Bay, nutrients, larval recruitment of crabs, barnacles and several
other estuarine organisms are presently being studied at OIMB (Linda Shapiro, director).
Offshore currents, nutrients and oceanographic processes are being studied by several
faculty (Barth and Smith, Wheeler) in the Department of Oceanic and Atmospheric Sciences
at OSU from Yaquina Head to Cape Blanco. Bob Emmett of NOAA, NMFS is studying the
-------
163
anchovy population from Grays Harbor to Tillamook Bay, and we are coordinating our
sampling at Willapa Bay during the summer of 1998 to coincide with his oceanic sampling off
of Willapa Bay.
Actual data exist that have not been utilized. These data include detailed current surveys by
the USGS of Willapa Bay in 1976 and 1977, U. Washington current surveys in 1995 and
1996, and extensive ADP collections on 3 navigation buoys for 1998 by a contractor for
Washington Department of Transportation.
-------
164
5.0 Research Linkages, Users and Participants
5.1 Research Group Linkages
The individual projects and themes of the proposed CEB research plan are interlinked
in many ways. DeWitt's work on interactions between burrowing shrimp and eelgrass
habitats will complement that of Boese who is examining the effects of additional types
of physical and biological disturbance on eelgrass, and the data sets should be
strongly complimentary. Specht will make measurements in beds of the exotic
seagrass species Zostera japortica which will parallel the types of measurements being
made by DeWitt in beds of the native Zostera marina. Specht's studies will require
bathymetric contouring and he will collaborate with Young to attain digital terrain
elevation models of the estuarine intertidal and subtidal zones.
The Boese and DeWitt benthic studies also share sample locations with the habitat-
faunal relationship studies of Ferraro and Cole. The Boese experiments will provide
insight into how habitat-faunal relationships change as the eelgrass habitat is
disturbed, while Dewitt's work will examine the effects of additional stressors on
eelgrass and burrowing shrimp habitats. Both studies thus provides complementary
information to Ferraro and Cole. Part of the work proposed by Ferraro and Cole seeks
to develop optimal sampling designs for estuarine habitats, which may potentially
benefit all other benthic research components.
The landscape scale research of Robbins is particularly aligned with research
proposed by DeWitt, Ozretich, Sigleo, and Young. Research proposed by other CEB
scientists concerned with habitat use and faunal distribution (Ferraro, Cole, and Power)
can also use the landscape ecology framework as a vehicle for making predictions
concerning the distribution of target organisms. More specific research questions, such
as those asked by Boese and Specht, can also take advantage of the information
developed by landscape pattern analysis. Results of landscape scale research within
the estuary can also be interpreted in the context of the watershed research findings
conducted by WED scientists in the Regional Ecology Branch. The integration of
watershed processes will facilitate the understanding of the linkages between terrestrial
and aquatic landscapes.
Basic physical and geochemical data for the water column and surficial sediment
components of the ecosystem will be generated by the research of Ozretich, Sigleo,
Young, and Specht. An accurate digital map of the Submerged Aquatic Vegetative
habitats, and detailed digital maps of estuarine bathymetry, also will be available to
support CEB research efforts.
The spatially explicit modeling effort involving Lee, Bodeen and DeWitt can directly
interact with the population level development of ecosystems indicators proposed by
-------
165
Power. Power's work deals with spatial variability of stressor effects on juvenile fish
growth rates, and is thus similar in concept to the modeling effort. Field data collected
by Power on fish, and DeWitt on burrowing shrimp populations, will serve as potential
inputs to the spatially explicit modeling effort. Conversely, the modeling effort will serve
as an analytical tool to help understand the results of the population level studies of
stressor effects.
5.2 Graduate and Postdoctoral Research Collaborators:
CEB is greatly expanding its committment to the support of graduate and postdoctoral
research positions to allow additional research questions related to the general
research direction to be addressed. At present, there is one National Network for
Environmental Management Studies (NNEMS) graduate research fellow beginning
work at CEB. Ms. Kelly Chapin will interact with Ferraro's work on habitat-biota
relationships with a study of comparative habitat utilization of native and exotic
brachyuran crabs in Yaquina Bay. A National Research Council (NRC) Postdoctoral
Fellow, Dr. Gail Dethloff will also collaborate with Ferraro in a project examining
intertidal habitat use by fish and decapod species is Pacific Northwest estuaries.
In addition to Chapin's work, NRC fellow Dr. Kama Almasi will conduct a comparative
study involving an examination of range expansion and community dynamics of the
native (Z. marina) and non-native (Z. japonica) eelgrass species in Yaquina Bay.
NHEERL Postdoctoral Fellow Dr. Scott Larned will examine the interactions between a
native burrowing shrimp and the introduced Z japonica to determine their effects on
sediment and nutrient dynamics. Alamsi and Larned's work will interact with that of
Specht, DeWitt, Sigleo and Young.
5.3 Users and Participants:
The principal expected users of the results of the CEB research effort are the many
organizations involved in conducting habitat-based ecorisk assessments and futures
analyses on PNW estuaries. These include ecosystem risk assessors, estuarine
ecologists, and local, state, and federal agencies with estuarine resource management
responsibilities. Local planners, particularly the Willapa Alliance have already begun
using nutrient and phytoplankton data collected by CEB in Willapa Bay, and other
important local contacts with groups such as the Central Oregon Coast Watershed
Council are being developed. At the regional level, EPA Regions IX and X, the
Washington State Department of Fish and Wildlife (collaborators 1995 -1997),
Washington Department of Ecology, and the Pacific Northwest District, US Geological
Survey, Water Resources Division are all potential users of the research data. At the
national level, an important client is the Office of Water of EPA.
There are numerous potential collaborators in the research program both regionally
-------
166
and nationally. Potential agency collaborators on-site at HSMC include the Oregon
Department of Fish and Wildlife and National Marine Fisheries Service. CEB is funding
one National Network for Environmental Management Studies (NNEMS) Fellow and
two National Research Council Research Associates to conduct independent research
projects in areas related to CEB research interests during FY 98. The Tillamook Bay
National Estuary Program is conducting research on eelgrass-burrowing shrimp-oyster
interactions in Tillamook Bay which are of interest to the CEB program. CEB is a
collaborator in the Pacific Northwest Coastal Ecosystem Research Study (PNCERS)
led by David Armstrong and Julia Parrish (Univ. Washington) and will exchange data
with this large group of researchers. CEB has also provided facilities in support of
PNCERS research through the guest worker program.
Additional scientific interactions will be sought with federal and academic experts for
various components of the research plan. For seagrass these may include Mark
Fonseca (NOAA/NMFS), Paul Harrison (Univ. British Columbia), Steve Rumrill (South
Slough National Estuarine Research Reserve), Fred Short (Univ. New Hampshire), Ron
Thorn (Battelle Marine Sciences Laboratory), and Sandy Wyllie-Echeverria (Univ.
Washington), and for burrowing shrimp may include David Armstrong (Univ.
Washington), Brett Dumbauld (Washington Department of Fisheries), and Kris
Feldman (Univ. Washington). Important interactions to expand CEB expertise on
acoustic remote sensing methods may include Whitlow Au (Univ. Hawaii), Nick
Chotiros (Univ. Texas), Darrell Jackson (Univ. Hawaii), and Bruce Sobal (US Army
Corps of Engineers, Waterways Experiment Station). Potential participants in defining
estuarine biota- habitat relationships include the Oregon Department of Fish and
Wildlife (ODF&W: Bob Buckman, Dan Bottom) and National Marine Fisheries Service
(NMFS: Bob Emmett).
Interactions for the physical-chemical process investigations may include the principal
investigators (McManus et al.) on an EPA/NSF grant to study Tillamook Bay, the
Georgia Pacific Corporation, the Sanitation Districts of Toledo, Siletz, Walport and
Tillamook, as well as, the State Department of Environmental Quality and Lincoln and
Tillamook County Health Division in providing current and historical waste loadings,
including septic tanks, to the four estuaries and their major rivers. All local and state
land use planners and shell fish producers would be potential users for the project
products.
The spatially explicit modeling effort will interact with EPA Region IX personal to help
define the key questions, exposure scenarios, and to find data sources. Particular
contacts include Janet Hashimoto of the Water Division, who has the responsibility for
regulating or monitoring sewage discharges and dredge disposal in California, and GIS
personnel. As the models that are developed move to the field verification stage,
interactions will occur with the major regional research organizations in San Francisco
Bay (San Francisco Estuarine Institute) and Southern California (Southern California
-------
167
Coastal Water Research Program).
One potential group of users of the spatially explicit modeling research results will be
the Headquarters and Region 9 risk managers who have regulatory responsibility for
sewage discharges (e.g., 301H), industrial discharges (NPDES permits) and dredge
material disposal.
-------
168
6.0 Project Management and Quality Assurance
The research program is an integrated effort representing contributions from all
scientific research staff of the Coastal Ecology Branch. The overall research program
is subdivided into thematic elements representing groups of related individual Principal
Investigator research projects. Each thematic element will be administered by a Project
Coordinator who will hold weekly meetings with Pi's to insure integration of research
within the thematic element. Project coordinators will be responsible for integration of
research among the different research themes, and every other week all branch
research groups will meet together to achieve among-group integration. The Branch
Chief will insure that overall programmatic research goals are being met.
All research data will be collected according to the Quality Assurance standards of the
Western Ecology Division of NHEERL. Quality Assurance Project Plans will be
prepared for each research project and approved by the WED Quality Assurance
officer. Existing CEB Standard Operating Procedures (SOP's) for chemical and
biological analyses will be followed, and new SOP's will be prepared by principal
investigators as required to document new research activities.
All research work will be performed under the general safety guidelines provided by the
CEB Employee Health and Safety Manual, and the CEB Chemical Hygiene Plan, with
additional guidance provided by the Health and Safety Handbook of the Western
Ecology Division. All field work in support of the research effort will be conducted
under the "Health and Safety Plan for Field Research Concerning the Ecology of
Pacific Northwest Estuaries" (approved 4/27/98).
All research projects will undergo review for environmental compliance issues and will
operate under an approved environmental compliance document.
-------
169
7.0 Literature Cited
Abramowitz, M. and I.A. Stegun, eds. 1965. Handbook of Mathematical Functions,
with Formulas, Graphs, and Mathematical Tables. Dover, NY.
Akcakaya, H.R. 1997. RAMAS GIS: Linking Landscape Data with Population Viability
Analysis (version 2.0). Applied Biomathematics, Setauket, NY.
Akins, G. J., and C.A. 1973. Jefferson. Coastal Wetlands of Oregon. A Natural
Resource Inventory Report to the Oregon Coastal conservation and
Development Commission, pp.159.
Adeniyi, P.O. 1980. Land-use change analysis using sequential aerial photography
and computer techniques. Photogrammetric Engineering and Remote Sensing.
46:1447-1464.
Albright, R., and P.K. Bouthillette. 1982. Benthic invertebrate studies in Gray's Harbor,
Washington. U.S. Army Corps of Engineers, Seattle, WA, Contract Number
DACW67-80-C-0091.
Allen, T.F.H. and T.B. Starr. 1982. Hierarchy: perspectives for ecological complexity.
University of Chicago Press, Chicago IL
Appeldoorn, R.S. 1995. Covariation of life-history parameters of soft-shell clams (Mya
arenaria) along a latitudinal gradient. ICES Marine Science Symposium, 199:19-
25.
Armstrong, D.A., B.G. Stevens, and J.C. Hoeman. 1982. Distribution and abundance
of Dungeness crab and crangon shrimp and dredging-related mortality of
invertebrates and fish in Grays Harbor, WA. Grays Harbor and Chehalis River
Improvements to Navigation Environmental Studies. Seattle, WA, U.S. Army
Corps of Engineers.
Armstrong, D.A., B. Dumbauld, and D. Doty. 1989. Oyster culture and crab habitat:
Conflicts over use of the insecticide Sevin in coastal estuaries. The Northwest
Environmental Journal Vol. 5, No. 1.
Atkinson, R.J.A., and C.J. Chapman. 1984. Resin casting: a technique for
investigating burrows in sublittoral sediments. Progress in Underwater Science
10: 109-115.
-------
170
Baker, J.P., D.H. Landers, H. Lee, II, PL. Ringold, R.R. Sumner, P.J. Wigington, R.S.
Bennett, E.M. Preston, W.E. Frick, A.C. Sigleo, D.T. Specht, and D.R. Young.
1995. Ecosystem Management Research in the Pacific Northwest. U.S.
Environmental Protection Agency Document EPA/600/R-95/069, 251 p.
Barnthouse, L.W., G.W. Suter, A.E. Rosen, and J.J. Beauchamp. 1987. Estimating
responses of fish populations to toxic contaminants. Environmental Toxicology
and Chemistry 6:811-824.
Barnthouse, L.W., G.W. Suter II, and A.E. Rosen. 1989. Inferring population-level
significance from individual-level effects: An extrapolation from fisheries science
to ecotoxicology. in Aquatic Toxicology and Environmental Fate: Eleventh
Volume, ASTM STP 1007, G.W. Suter (ed.), American Society of Testing
Materials, Philadelphia PA. pp 289-300.
Barnthouse, L.W., G.W. Suter, and A.E. Rosen. 1990. Risks of toxic contaminants to
exploited fish populations: Influence of life history, data uncertainty and
exploitation intensity. Environmental Toxicology and Chemistry. 9:297-311.
Barry, J.P., M.M. Yoklavich, G.M. Cailliet, D.A. Ambrose, and B.S. Antrim. 1996.
Trophic ecology of the dominant fishes in Elkhorn Slough, CA, 1974-1980.
Estuaries. 19:115-138.
Bates, S.S., C. Leger and K.M. Smith. 1996. Domoic acid production by the diatom
Pseudo-nitzschia multiseries, as a function of division rate in silicate-limited
chemostat culture. In Harmful and Toxic Algal Blooms, ed. T. Yasumoto, Y.
Oshima, and Y. Fukuyo. UNESCO. P. 163-166.
Batiuk, R.A., R.J. Orth, K.A. Moore, W.C. Dennision, J.C. Stevenson, L.W. Staver, V.
Carteer, N.B. Rybiki, R.E. Hickman, S. Kollar, S. Bieber, and P. Heasly. 1992.
Chesapeake Bay submerged aquatic vegetation habitat requirements and
restoration goals; a technical synthesis. U.S. Environmental Protection Agency
CBP/TRS83/92, Chesapeake Bay Program, Annapolis, MD
Bayer, R.D. 1979. Intertidal zonation of Zostera marina in the Yaquina Estuary, OR.
Syesis 12:147-154.
Bayer, R.D. 1979. Intertidal shallow-water fishes and selected macroinvertebrates in
the Yaquina Estuary, Oregon. Unpublished manuscript. Copy on file at the
Oregon State University Marine Science Library, Newport, OR. 134 pp.
Bayer, R.D. 1981. Shallow-water intertidal ichthyofauna of the Yaquina Estuary, OR.
Northwest Science 55: 182-193.
-------
171
Bayer, R.D. 1996. Censusues of black brant at Yaquina Estuary, Lincoln County, OR.
Journal of Oregon Ornithology 6:723-780
Bayer, R.D. 1996. Macrophyton and tides at Yaquina Estuary, Lincoln County, OR.
Journal of Oregon Ornithology 6:781-795.
DeBen, W.A., W.L. Clothier, G.R. Dittsworth, and D.J. Baumgartner. 1990. Spatio-
temporal fluctuations in the distribution and abundance of demersal fish and
epibenthic crustaceans in Yaquina Bay, OR. Estuaries 13:469-478
Becker, D.S. 1988. Relationship between sediment character and sex segregation in
English sole, Parophrys vetulus. Fisheries Bulletin 86:517-524.
Becker, D.S., and K.K. Chew. 1987. Predation on Capitella spp. by small-mouthed
pleuronectids in Puget Sound, WA. Fisheries Bulletin 85:471-479.
Bell, S.S., M.O. Hall, and B.D. Robbins. 1995. Toward a landscape approach in
seagrass beds: using macroalgal accumulation to address questions of scale.
Oecologia 104:163-168.
Benner, P.A. 1991. Historical reconstruction of the Coquille River and surrounding
landscape. Near Coastal Waters National Pilot Project: The Coquille River, OR.
Action Plan for Oregon Coastal Watersheds, Estuary and Ocean Waters, 1988-
1991.
Benner, P.A. and J.R. Sedell. 1997. Upper Willamette River landscape: an historical
perspective, pp 23-47. InA. Laenen and D.A. Dunnette (eds.) River Quality:
Dynamics and Restoration. Lewis Publishers, Boca Raton
Bennett, L. 1997. Personal communication. R&B Oysters, Inc., Bay Center, WA.
Berghahn, R., K. Ludemann, and M. Ruth. 1995. Differences in individual growth of
newly settled 0-group plaice (Pleuronectes platessa L.) in the intertidal of
neighboring Wadden Sea areas. Netherlands Journal of Sea Research
34:131-138.
Berkeley, S. 1998. Personal communication. Oregon State University, Hatfield Marine
Science Center, Newport, OR.
Biebl, R. and C. P. McRoy. 1971. Plasmatic resistance and rate of respiration and
photosysnthesis of Zostera marina at different salinities and temperatures.
Marine Biology. 8:48-56.
-------
172
Bilby, R.E., B.R. Fransen, and P.A. Bisson. 1996. Incorporation of nitrogen and
carbon from spawning coho salmon into the trophic system of small streams:
evidence from stable isotopes. Canadian Journal of Fisheries and Aquatic
Sciences 53:164-173.
Bird, E.M. 1982. Population dynamics of thalassinidean shrimps and community
effects through sediment modification. PhD dissertation, University of Maryland.
Black, D. et al. unpublished. U.S. Environmental Protection Agency, Atlantic Ecology
Division, Narragansett, R.I.
Boehlert, G.W., and B.C. Mundy. 1987. Recruitment dynamics of metamorphosing
English sole, Parophorys vetulus, to Yaquina Bay, OR. Estuarine Coastal Shelf
Sciences 25:261-281.
Boese, B.L., F.A. Cole, S.P. Ferraro, J O. Lamberson, H. Lee II, and S. Pozarycki.
1996. QAPP for Pacific Northwest Estuaries Project Areas III and IV. U.S.
Environmental Protection Agency, Coastal Ecology Branch, Newport, OR.
Boese, B.L. 1998. The impact of disturbances on an eelgrass habitat. ]n WED/CEB
FY98 Research Plan. U.S. Environmental Protection Agency, Coastal Ecology
Branch, Newport, OR.
Bostrom, C., and E. Bonsdorff. 1997. Community structure and spatial variation of
benthic invertebrates associated with Zostera marina (L.) Beds in the northern
Baltic Sea. Journal of Sea Research 37: 153-166.
Botsford, L.W., C.L. Moloney, A. Hastings, J.L. Largier, T.M. Powell, K. Higgins, and
J.F. Quinn. 1994. The influence of spatially and temporally varying
oceanographic conditions on meroplanktonic metapopulations. Deep-Sea
Research II 41:107-145.
Botsford, L.W. 1995. Population dynamics of spatially distributed, meroplanktonic,
exploited marine invertebrates. ICES Marine Science Symposium 199:118-128.
Bottom, D.L. and K.K. Jones. 1990. Species composition, distribution, and
invertebrate prey of fish assemblages in the Columbia River Estuary. Progress
in Oceanography Vol. 25:243-270.
Boynton, W.R., J.H. Garber, R. Summers and W.M. Kemp. 1995. Inputs,
transformations, and transports of nitrogen, phosphorus in Chesapeake Bay and
selected tributaries. Estuaries 18:285-314.
-------
173
Boynton, W. R., W.M. Kemp and C.G. Osbourn. 1980. Nutrient fluxes across the
sediment-water interface in the turbid zone of a coastal plain estuary, pp. 93-
109. In V.S. Kennedy (ed.) Estuarine Perspectives. Academic Press, NY.
Brandt, S.B., and D.M. Mason. 1994. Landscape approaches for assessing spatial
patterns in fish foraging and growth, pp 211-238. In Stouder, D.J., K.L. Fresh,
and R.J. Feller, {eds.}. Theory and application offish feeding ecology.
University of South Carolina Press, Columbia, SC.
Brenchley, G.A. 1982. Mechanisms of spatial competition in marine soft-bottom
communities. Journal of Experimental Marine Biology and Ecology 60: 17-33.
Briggs, P.T., and J.S. O'Connor. 1971. Comparison of shore-zone fishes over
naturally vegetated and sand-filled bottoms in Great South Bay. New York Fish
and Game Journal 18:15-41.
Briggs, R.B. and L.E. Cronin. 1981. Special characteristics of estuaries. B.J. Neilson,
and L.E. Cronin (eds.), International Symposium on Nutrient Enrichment in
Estuaries. Humana Press, NJ
Brooks, K.M. 1995. Long-term response of benthic invertebrate communities
associated with the application of carbaryl (Sevin) to control burrowing shrimp,
and an assessment of the habitat value of cultivated Pacific oyster (Crassostrea
gigas) beds in Willapa Bay, Washington, to fulfill requirements of the EPA
carbaryl data call in. Final Report. Aquatic Environmental Services, Port
Townsend, WA.
Brooks, K M. 1997. Personal communication. Aquatic Environmental Sciences, Port
Townsend, WA.
Brousseau, D.J. 1978. Spawning cycle, fecundity, and recruitment in a population of
soft-shell clam, Mya arenaria, from Cape Ann, MA. Fisheries Bulletin 76:155-
165
Brousseau, D.J., J.A. Baglivo, and G.E. Lang. 1982. Estimation of equilibrium
settlement rates for benthic marine invertebrates: Its application to Mya arenaria
(Mollusca: Pelecypoda). Fisheries Bulletin 80:642-644.
Brown, B. and W. H. Wilson Jr. 1997. The role of commercial digging of mudflats as
an agent for change on infaunal intertidal populations. Journal of Experimental
Marine Biology and Ecology 218:46-61.
-------
174
Bulthuis, D.A. 1983. Effects of in situ light reduction on density and growth of the
seagrass Heterozostera tasmanica (Martens ex Aschers.) Den Hartog in
Western Port, Victoria, Australia. Journal of Experimental Marine Biology and
Ecology 67:91-103
Bulthuis, D.A. 1997. Effects of short-term light reduction during different seasons on
survival of intertidal and subtidal eelgrass, Zostera marina, in Padilla Bay,
Washington. 14th Biennial Estuarine Research Federation International
Conference, October 12-16, Providence, Rl.
Burger, M.J. 1998. Temporal variation in larval fish condition indices. Ph.D. Thesis,
Louisiana State University, Baton Rouge, LA.
Burkholder JM, K.M. Mason and H.B. Glasgow, Jr. 1992. Water column nitrate
enrichment promotes decline of eelgrass Zostera marina L.: evidence from
seasonal mesocosm experiments. Marine Ecology Progress Series 61.163-178
Callaway, R.J., and D.T. Specht. 1982. Dissolved silicon in the Yaquina Estuary,
Oregon. Estuarine and Coastal and Shelf Sciences 15:561-567.
Callaway, R.J., D.T. Specht, and G.R. Ditsworth. 1988. Manganese and suspended
matter in the Yaquina Estuary, OR. Estuaries 11:217-225.
Campana, S.E. 1990. How reliable are growth back-calculations based on otoliths?
Canadian Journal of Fisheries and Aquatic Sciences 47:2219-2227.
Carlton, J.T. and J.B. Geller. 1993. Ecological roulette: the global transport of
nonindigenous marine organisms. Science 261:78-82.
Carstensen, L.W. Jr. 1987. A measure of similarity for cellular maps. American
Cartographer 14:345-358
Caswell, H. 1976. The validation problem. |n Systems Analysis and Simulations in
Ecology. B.C. Patten (ed.), Academic Press, N.Y. pp 313-325.
Caswell, H. 1989. Analysis of life table response experiments - decomposition of
effects on population growth rate. Ecological Modeling 46:221-237.
Chamberlain, R.H., and R.A. Barnhart. 1993. Early use by fish of a mitigation salt
marsh, Humboldt Bay, CA. Estuaries 16:769-783.
i
i
-------
175
Chotiros, N.P. 1997. Modeling of penetration of dolphin sonar into ocean sediments.
Abstract. 13th Annual Meeting of the Acoustical Society of America. Journal of
the Acoustical Society of America 102: 3124.
Christoffersen, K. 1996. Ecological implications of cyanobacterial toxins in aquatic
food webs. Phycologia 35:42-50.
Cifuentes, L.A. 1991. Spatial and temporal variations in terrestrially-derived organic
matter from sediment of the Delaware estuary. Estuaries 14:414-429.
Clemems, K.E. and P.D. Komar. 1988. Oregon beach-sand compositions produced
by the mixing of sediments under a transgressing sea. Journal of Sedimentary
Petrology 58:519-529
Clemmesen, C. and T. Doan. 1996. Does otolith structure reflect the nutritional
condition of a fish larva? Comparison of otolith structure and biochemical
(RNA/DNA ratio) determined on cod larvae. Marine Ecology Progress Series
138:33-39.
Cliff, A.D. and J.K. Ord. 1981. Spatial processes. Models and Applications, Pion,
London. 266 pp
Cohen, A.N. and J.T. Carlton. 1998. Accelerating invation rate in a highly invaded
estuary. Science 279:555-558.
Cohen, J. 1977. Statistical power analysis for the behavioral sciences. Revised
edition. Academic Press, Orlando, FL.
Congalton, R.G. 1991. A review of assessing the accuracy of classifications of
remotely sensed data. Remote Sensing of the Environment 37:35-46
Cooper, S.R. and G.S. Brush. 1991. Long-term history of Chesapeake Bay anoxia.
Science 254:993-996.
Copping, A. and B.C. Bryant. 1993. Pacific Northwest Regional Marine Research
Program, Vol. 1, Research Plan, 1992-1996. Office of Marine Environmental
and Resource Programs, University of Washington, Seattle.
Costanza, R., R. D'Arge, R. De Groot, S. Farber, M. Grasso, B, Hannon, K. Limburg, S.
Naeem, R.V. O'Neill, J. Paruelo, R.G. Raskin, P. Sutton, and M. Van den Belt.
1997. The value of the world's ecosystem services and natural capital. Nature
387:253-260.
-------
176
Coulton, K.G., P.B. Willliams and P.A. Benner. 1996. An environmental history of the
Tillamook Bay estuary and watershed. Tillamook Bay National Estuary Project
Technical Report.
Cowan, J.L. and W.R. Boyton. 1996. Sediment-water oxygen and nutrient exchanges
along the longitudinal axis of Chesapeake Bay: seasonal patterns, controlling
factors and ecological significance. Estuaries 19:562-580.
Craft, C.B., S.W. Broome, E.D. Seneca and W.J. Showers. 1988. Estimating sources
of soil organic matter in natural and transplanted estuarine marshes using stable
isotopes of carbon and nitrogen. Estuarine and Coastal and Shelf Sciences
26:633-641.
Culliton, T.J., M.A. Warren, T.R. Goodspeed, D.G. Remeer, C.M. Blackwell, and
J.J. McDonough, III. 1990. 50 Years of Population Change along the Nation's
Coasts, 1960-2010. NOAA, Office of oceanography and Marine Assessment,
National Ocean Service, Coastal Trends Series, Rockville, MD. 41 pp.
Daniel, W.W. 1983. Biostatistics: a foundation for analysis in the health sciences. 3rd
Edition, J. Wiley and Sons, New York. 534 pp
Davis, J.C. 1986. Statistics and Data Analysis in Geology. 2nd Edition. J.Wiley and
Sons, NY. 646 pp.
Davis, M.W. and C.D. Mclntire. 1983. Effects of physical gradients on the production
dynamics of sediment-associated algae. Marine Ecology Progress Series
13:103-114.
Dawes, C.J. and J.M. Lawrence. 1979. Effects of blade removal on the proximate
composition of the rhizome of the seagrass Thalassia testundinum Banks ex
Konig. Aquatic Botany 7:255-266
Dawes, C.J. 1981. Marine Botany. John Wiley and Sons, New York. 628 pp
Day, J.W., C.A.S. Hall, W.M. Kemp, and A. Yanez-Arancibia. 1989. Estuarine ecology.
John Wiley & Sons, NY.
DeAngelis, D.L., K.A. Rose, L.B. Crowder, E.A. Marschall, and D. Lika. 1993. Fish
cohort dynamics: Application of complementary modeling approaches. American
Naturalist 142:604-622.
-------
177
De Ben, W.A., W.D. Clothier, G.R. Ditsworth, and D.J. Baumgartner. 1990. Spatio-
temporal fluctuations in the distribution and abundance of demersal fish and
epibenthic crustaceans in Yaquina Bay, OR. Estuaries 13:469-478.
Deegan, L.A., and R.H. Garritt. 1997. Evidence for spatial variability in estuarine food
webs. Marine Ecology Progress Series 147:31-47.
Deegan, L.A., J.T. Finn, S.G. Ayvazian, C.A. Ryder-Kieffer, and J. Buonaccorsi. 1997.
Development and validation of an estuarine biotic integrity index. Estuaries 20:
601-617.
D'Elia, C.F., K.L. Webb, and R.L. Wetzel. 1981. Time varying hydrodynamics and
water quality in an estuary. In B.J. Neilson, and L.E. Cronin (eds.), International
Symposium on Nutrient Enrichment in Estuaries. Humana Press, NJ.
den Hartog, C. and P.J.G. Polderman. 1975. Changes in the seagrass populations of
the Dutch Waddenzee. Aquatic Botany 1:141-147
den Hartog, C. 1977. Structure, function and classification in seagrass communities,
in McRoy, C.P. and C Helfferich, eds. Seagrass ecosystems: A scientific
perspective. Marcel Dekker, Inc., NY. pp 89-121.
Dennison W.C. and R.S. Alberte. 1985. Role of daily light period in the depth
distribution of Zostera marina eelgrass. Marine Ecology Progress Series
25:51-61
Dennison, W.C. 1987. Effects of light on seagrass photosynthesis, growth and depth
distribution. Aquatic Botany 27:15-26.
Dennison, W.C., R.J. Orth, K.A. Moore, J.C. Stevenson, V. Carter, S. Kollar, P.W.
Bergstrom, and R.A. Batiuk. 1993. Assessing water quality with submerged
aquatic vegetation (Habitat requirements as barometers of Chesapeake Bay
health). Bioscience 43:86-94.
Dennison, W.C., R.J. Orth, K.A. Moore, J.C. Stevenson, V. Carter, S. Kollar,
R.W. Bergstrom, and R.A. Batiuk. 1993. Assessing water quality with
submersed aquatic vegetation: habitat requirements as barometers of
Chesapeake Bay health. Bioscience 43:86-94.
Dethier, M.N. 1990. A marine and estuarine habitat classification system for
Washington State. Washington Natural Heritage Program, Dept. Natural
Resources. Olympia, WA. 56 pp.
-------
178
DeWitt, T.H., K.F. Wellman, T. Wildman, D.A. Armstrong, and L. Bennett. 1997. An
evaluation of the feasibility of using integrated pest management to control
burrowing shrimp in commercial oyster beds. Final Report to Washington
Department of Ecology, Olympia, WA.
Dewitt, T.H, B. L. Boese, J.O. Lamberson and B.D. Robbins. 1998. Changes in the
abundance and distribution of estuarine keystone species in response to
multiple stressors. Research Preproposal, Coastal Ecology Branch,
WED/NHEERL, Newport, OR.
Dinicola, R. S. 1990. Characterization and simulation of rainfall-runnoff relations for
headwater basins in Western King and Snohomish Counties, WA. USGS WRI
Report 89-4052. 52 pp.
Doty, D.C., D.A. Armstrong, and B.R. Dumbauld. 1990. Comparison of carbaryl
pestioide impacts on Dungeness crab (Cancer magister) versus benefits of
habitat derived from oyster culture in Willapa Bay, WA. Report to the Willapa
Bay/ Grays Harbor Oyster growers Association. 69 pp.
Dring, M.J. and K. Luning. 1994. Influence of spring-neap tidal cycles on the light
available for photosynthesis by benthic marine plants. Marine Ecology Progress
Series 104:131-137.
Duarte, C.M., N. Marba, N. Agawin, J. Cebrian, S. Enriquez, M.D. Fortes, M.E.
Gallegos, M. Merino, B. Olesen, K. Sand-Jensen, J. Uri, and J. Vermaat. 1994.
Reconstruction of seagrass dynamics: age determinations and associated tools
for the seagrass ecologist. Marine Ecology Progress Series 107:195-209.
Dumbauld, B. R. 1994. Thalassinid shrimp ecology and the use of carbaryl to control
populations on oyster ground in Washington coastal estuaries. Ph.D.
Dissertation, University of Washington, Seattle.
Dumbauld, B.R., D.A. Armstrong, and K.L. Feldman. 1996. Life-history characteristics
of two sympatric thalassinidean shrimps, Neotrypea californiensis and Upogebia
pugettensis with implications for oyster culture. Journal of Crustacean Biology
16:689-708.
Dumbauld, B., S. Wyllie-Echeverria, and M. Herrle. 1997. (Abstract) Ghost shrimp
control and eelgrass (Zostera japonica) distribution. Presented at PERS Annual
Meeting May 15-17 Tillamook , OR.
Dumbauld, B.R. 1998. Personal communication. Washington Department of Fisheries,
Nacotta, WA.
-------
179
Dunning, J.B., B.J. Danielson, B.R. Noon, T.L. Root, R.H. Lamberson, and
E.E. Stevens. 1995. Spatially explicit population models: Current forms and
future uses. Ecological Applications 5:3-11
Durning, A T. 1996. The six floods. WorldWatch November/December 1996. pp. 28-
36.
Ellis, R. 1997. Draft scientific and technical report on the biological resources of
Tillamook Bay. Unpublished Tillamook Bay National Estuary Project Technical
Report.
Emlen, J.M. 1973. Ecology: An Evolutionary Approach. Addison-Wesley Publishing,
Reading, MA.
Emlen, J.M. 1989. Terrestrial population models for ecological risk assessment: a
state-of-the-art review. Environmental Toxicology and Chemistry 8:831-842.
Engst, N. Unpublished thesis data - Netarts (see Kentula thesis [1982]).
Evans, B.M. and L. Mata. 1984. Aerial photographic analysis of hazardous waste
disposal sites, Hazardous Waste and Environmental Emergencies, Hazardous
Materials Control Research Institute, Maryland.
Everett, R.A., G.M. Rutz, and J.T. Carlton. 1995. Effect of oyster mariculture on
submerged aquatic vegetation: An experimental test in a Pacific Northwest
estuary. Marine Ecology Progress Series 123:205-217.
Fasten, N. 1931. The Yaquina oyster beds of Oregon. The American Naturalist
LXV(700):434-468.
Feldman, K.L., B.R. Dumbauld, D.A. Armstrong, and T.H. DeWitt. Ecology of the
burrowing shrimp, Neotrypea califomiensis and Upogebia pugettensis, from
estuaries of the US Pacific Northwest, (in prepj
Ferguson, R.L. and K. Korfmacher. 1997. Remote sensing and GIS analysis of
seagrass meadows in North Carolina, USA. Aquatic Botany 58:241-258.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA.
Canadian Journal of Fisheries and Aquatic Sciences 46:2157-2165.
-------
180
Ferraro, S.P. and F.A. Cole. 1990. Taxonomic level and sample size sufficient for
assessing pollution impacts on the Southern California Bight macrobenthos.
Marine Ecology Progress Series 67:251-262.
Ferraro, S.P., R.C. Swartz, F.C. Cole, and W.A. De Ben. 1994. Optimum
macrobenthic sampling protocol for detecting pollution impacts in the Southern
California Bight. Environmental Monitoring and Assessment 29:127-153.
Ferraro, S .P and F.A. Cole. 1996. Fidelity of habitat-community relationships in
Yaquina and Willapa Bays. A research plan prepared for the Pacific Northwest
Estuaries Project. U.S. Environmental Protection Agency. Coastal Ecology
Branch. Newport, OR.
Ferraro, S.P., and F.A. Cole. 1996a. Research plan for developing within Willapa Bay
habitat-community relationships. Task#: III,A,1,a. ]n Program Plan Pacific
Northwest - Estuaries. FY1996 and FY1997 Research. U.S. Environmental
Protection Agency, Coastal Ecology Branch, Newport, OR.
Ferraro, S.P. and F.A. Cole. 1996b. Research plan for testing fidelity of habitat-
community relationships in Yaquina and Willapa Bay. Task#: lll,A,1,b. ]q
Program Plan Pacific Northwest - Estuaries. FY1996 and FY1997 Research.
U.S. Environmental Protection Agency, Coastal Ecology Branch, Newport, OR.
Ferron, A. and W.C. Leggett. 1994. An appraisal of condition measures for marine fish
larvae. Advances in Marine Biology. 30:217-303.
Fish and Wildlife Service (FWS). 1968. Preliminary survey offish and wildlife in
relation to ecological and biological aspects of Yaquina Bay, Oregon. Special
Report. U.S. Fish and Wildlife Service, U.S. Department of the Interior,
Portland, OR.
Fitzhugh, G.R., S.W. Nixon, D.W. Ahrenholz, and J.A. Rice. 1997. Temperature
effects on otolith microstructure and birth month estimation from otolith
increment patterns in Atlantic menhaden. Transactions of the American
Fisheries Society 126:579-593.
Fong, A., M. Johnson, D. Tingey, and R. Rygiewicz. 1995. Macro-nutrients analysis by
suppressed ion chromatography with conductivity detection. Project SOP
Number 3.02, Corvallis, OR.
Fonseca, M.S., J.S. Fisher, J.C. Zieman, and G.W. Thayer. 1982. Influence of the
seagrass, Zostera marina L. on current flow. Estuarine and Coastal and Shelf
Sciences. 15:351-364.
»
i
-------
181
Fonseca, M.S., G.W. Thayer, A.J. Chester, and C. Foltz. 1984, Impact of scallop
harvesting on eelgrass (Zostera marina) meadows: Implications for
management. North American Journal of Fisheries Management 4:286-293.
Fonseca, M.S., W.J. Kenworthy, F.X. Courtney, and M.O Hall. 1994. Seagrass
transplanting in the southeastern United States: methods for accelerating habitat
development. Restoration Ecology 2:198-212.
Fonseca, M. 1994. A Guide to Planting Seagrasss in the Gulf of Mexico. Texas A&M
University Sea Grant College Program. TAMU-SG-94-601.
Fonseca, M.S., W.J. Kenworthy, and F.X. Courtney. 1996. Development of planted
seagrass beds in Tampa Bay, Florida, USA. I. Plant components. Marine
Ecology Progress Series 132:127-139.
Fonseca, M.S. 1996a. The role of seagrasses in nearshore sedimentary processes: a
review, in, K.F. Nordstrom and C.T. Roman (eds.) Estuarine Shores: Evolution,
Environments and Human Alterations. John Wiley and Sons, Ltd.
Chapter 11, pp 261 -286
Fonseca, M.S. 1996b. Scale dependence in the study of seagrass systems, pp 95-104
in J. Kuo, R.C. Phillips, D.I. Walker, and H. Kirkman (eds.) Seagrass Biology:
Proceedings of an International Workshop, Rottnest Island, Western Australia.
Freemark, K., C. Hummon, D. White, and D. Hulse. 1996. Modeling risks to
biodiversity in past, present and future landscapes. Technical Report No. 268,
Canadian Wildlife Service, Headquarters, Environment Canada, Ottawa K1A
0H3.
Forman, R.T.T. and M. Godron. 1984. Landscape Ecology. John Wiley and Sons, NY
619 pp
Franklin, J.F. 1989. Importance and justification of long term studies in ecology.in,
G.E. Likens (ed.) Long Term Studies in Ecology: Approaches and alternatives.
Springer-Verlag, NY. pp 3-19
Frick, W. 1966. A model of emergence patterns affecting Pacific Northwest intertidal
zones. WED-CEB QAPP, signed 4/11/96.
Gallego, A., and M. Heath. 1997. The effect of growth-dependent mortality, external
environment and internal dynamics on larval fish otolith growth: an individual-
based modelling approach. Journal of Fish Biology. 51:121-134.
-------
182
Gallegos, C.L. 1994. Refining habitat requirements of submerged aquatic vegetation:
role of optical models. Estuaries 17:187-199.
Gallegos, C.L. and W.J. Kenworthy. 1996. Seagrass depth limits in the Indian River
Lagoon (Florida, U.S.A.): Application of an optical water quality model.
Estuarine and Coastal and Shelf Sciences 42:267-288.
Gambi, M.C., A.R.M. Nowell, and P.A. Jumars. 1990. Flume observations on flow
dynamics in Zostera manna (eelgrass beds). Marine Ecology Progress Series.
61:159-169.
Gentile, J.H., S.M. Gentile, N.G. Hairston, and B.K. Sullivan. 1982. The use of
life-tables for evaluating the chronic toxicity of pollutants to Mysidopsis bahia.
Hydrobiologia 93:179-187.
Gibson, R.N. and R.S. Batty. 1990. Lack of substratum effect on the growth and
metamorphosis of larval plaice Pleuronectes platessa . Marine Ecology
Progress Series 66:219-223.
Giesen, W.B.J.T., M.M. van Katwijk, and C. den Hartog. 1990. Eelgrass condition and
turbidity in the Dutch Waddden Sea. Aquatic Botany 37:71-85
Golterman, H.L., P.G. Sly, and R.L. Thomas. 1983. Study of the relationship between
water quality and sediment transport. International Hydrological Programme,
UNESCO, Paris. 231 pp.
Goni, M. A. and J.I. Hedges. 1995. Sources and reactivities of marine-derived organic
matter in coastal sediments as determined by alkaline CuO oxidation.
Geochimica et Cosmochimica. Acta. 59:2965-2981.
Gonor, J.J., J.R. Sedell, and P.A. Benner. 1988. What we know about large trees that
fall into streams and rivers, pp. 83-112. in Maser, C. , R.F. Tarrant, J.M.
Trappe, and J.F. Franklin, eds. From the Forest to the Sea. A story of fallen
trees. U.S. Forest Service, General Technical Report PNW-GTR-229.
Good, J.W. 1975. Estuary development with implications for management: A case
study of Northwest Natural Gas Company's liquified natural gas (LNG) project at
Yaquina Bay, Oregon. Unpublished report, School of Oceanography, Oregon
State University. 107 pp.
Grant, G.C. 1996. RNA-DNA ratios in white muscle tissue biopsies reflect recent
growth rates of adult brown trout. Journal of Fish Biology 48:1223-1230.
-------
183
Gray, J.S. 1974. Animal-sediment relationships. Oceanography and Marine Biology
Annual Review 12:223-261.
Green, R.H. 1979. Sampling design and statistical methods for environmental
biologists. Wiley, New York, NY.
Griffin, K. 1997. Eelgrass ecology and commercial oyster cultivation in Tillamook Bay,
OR. A literature review and synthesis. Tillamook Bay National Estuary Project,
Garibaldi, OR. July 1997.
Grue, C.E. 1995. Evaluation of the protection of aquatic and terrestrial biodiversity in
Willapa Bay watersheds, jn A companion proposal to the U. S. Environmental
Protection Agency Pacific Northwest Ecosystem Research Program. Ecosystem
research priorities and opportunities in the Willapa Ecosystem, Southwest
Washington State. The Willapa Bay Alliance.
Grussendorf, M.J. 1981. A flushing-coring device for collecting deep-burrowing
infaunal bivalves in intertida! sand. Fisheries Bulletin, U.S. 79:383-385.
Gunderson, D.R. and I.E. Ellis. 1986. Development of a plumb staff beam trawl for
sampling demersal fauna. Fisheries Research 4:35-41.
Gunderson, D.R., D.A. Armstrong, Y-B. Shi, and R.A. McConnaughey. 1990. Patterns
of estuarine use by juvenile English sole (Parophrys vetulus) and dungeness
crab (Cancer magister). Estuaries 13 59-71.
Haddad, R.I., and C.S. Martens. 1987. Biogeochemical cycling in an organic rich
coastal marine basin. IX. Sources of vascular plant-derived sedimentary organic
material. Geochimica et Cosmochimica Acta. 51:2991-3001.
Hall, C.A.S. 1988. An assessment of several of the historically most influential
theoretical models used in ecology and of the data provided in their support.
Ecological Modeling 43:5-31.
Hansen, D.V. and M. Rattray, Jr. 1966. New dimensions in estuary classification.
Limnology and Oceanography 11:319-326.
Harbison, R. and Volovik. 1994. The ctenophore, Mneiopsis leydyi, in the Black Sea: A
holoplanktonic organism transported in the ballast water of ships, pp. 25-37.
Proceedings of the Conference and Workshop: Nonindigenous estuarine &
marine organisms (NEMO). National Oceanic and Atmospheric Administration
(NOAA), U.S. Dept. Of Commerce 125 pp.
-------
184
Harlin, M.M. and B. Thorne-Miller. 1982. Nutrient enrichment of seagrass beds in a
Rhode Island coastal lagoon. Marine Biology 65:221-229.
Harper, J.R., B. Bornhold, D. Howes, D. McCullough, and B. Emmett. 1998. Seabed
Imaging and Mapping System - seabed classification of substrate, epiflora, and
epifauna. Unpublished report, Coastal & Ocean Resources, Inc., Sydney, British
Columbia, Canada.
Harrison, P. G. 1979. Reproductive Strategies in Intertidal Populations of Two Co-
occurring Seagrasses (Zostera spp.). Canadian Journal of Botany 57:2635-
2638.
Harrison, P. G. 1982. Seasonal and Year-to-Year Variations in Mixed Intertidal
Populations of Zostera japonica Aschers. and Graebn. and Ruppia maritma L.
Aquatic Botany 14:357-371.
Harrison, P. G. and R. E. Bigley. 1982. The recent introduction of the seagrass
Zostera japonica Aschers. and Graebn. to the Pacific Coast of North America.
Canadian Journal of Fisheries and Aqatic Sciences 39:1642-1648.
Harrison, P.G. 1987. Natural expansion and experimental manipulation of seagrass
(Zostera spp.) Abundance and the response of infaunal invertebrates. Estuarine
and Coastal Shelf Sciences 24:799-812
Hastings, A. and S. Harrison. 1994. Metapopulation dynamics and genetics. Annua!
Review of Ecolgy Systematics. 25:167-188.
Heck, K.L., K.W. Able, C.T. Roman, and M.P. Fahay. 1995. Composition, abundance,
biomass, and production of macrofauna in a New England estuary: comparisons
among eelgrass meadows and other nursery habitats. Estuaries 18:379-389.
Hedges, J.I. and D.C. Mann. 1979. The characterization of plant tissues by their lignin
oxidation products. Geochimica et Cosmochimica Acta. 43:1803-1807.
Hedgpeth, J.W. and S. Obrebski. 1981. Willapa Bay: a historical perspective and a
rationale for research. Office of Biological Services, U.S. Fish and Wildlife
Service, Washington, D.C. FWS/OBS-81/03. 52 pp.
Higgins, K., A. Hastings, J.N. Sarvela, and L.W. Botsford. 1997. Stochastic dynamics
and deterministic skeletons: population behavior of Dungeness crab. Science
276:1431-1435.
-------
185
Hilborn, R. and C.J. Walters. 1992. Quantitative Fisheries Stock Assessment: Choice,
Dynamics, and Uncertainty. Chapman and Hall, New York.
Hinton, S.A., R.L. Emmett, and G.T. McCabe, Jr. 1992. Fishes, shrimp, benthic
invertebrates, and sediment characteristics in intertidal and subtida! habitats at
Rice Island and Miller Sands, Columbia River Estuary, 1991. Final Report for
U.S. Army Corps of Engineers, Portland, OR (Contract E96910025) and Coastal
Zone and Estuarine Studies Division, NMFS, Seattle, WA.
Hogue, E.W. and A.G. Carey Jr. 1982. Feeding ecology of 0-age flatfishes at a
nursery ground on the Oregon coast. Fisheries Bulletin 80:555-565.
Holland, A.F. and A.T. Shaughnessey. 1986. Separation of long term variation in
benthic organisms into major components, h Oceans 86 Conference Record.
Vol. 3. Monitoring strategies symposium. Institute of Electrical and Electronic
Engineers, Piscataway, NJ. pp. 1056-1061.
Hopkinson, C.S., and J.J. Vallino. 1995. The relationships among man's activities in
watersheds and the estuaries: a model of runoff effects on patterns of estuarine
community metabolism. Estuaries 18:598-621.
Horner, R.A., D.L. Garrison, and F.G. Plumley. 1997. Harmful algal blooms and red
tide problems on the U.S. west coast. Limnology and Oceanograohy 42:1076-
1088.
Howarth, R.W. J.R. Fruch and D. Sherman. 1991. Inputs of sediment and carbon to an
estuarine ecosystem: influnece of land use. Ecological Applications 1:27-39.
http://www.maine.com/mer/bpproj.htm. 1997. Personal communication by Dana
Wallace reported on the web site on 06/13/1997.
Hunsaker, C.T., and D.E. Carpenter, eds. 1990. Ecological indicators for the
Environmental Monitoring and Assessment Program. EPA 600/3-90/060. U.S.
Environmental Protection Agency, Office of Research and Development,
Research Triangle Park, NC.
Ibarra-Obando, S.E. and C.F. Boudouresque. 1994. An improvement of the Zieman
leaf marking technique for Zostera marina growth and production assessment.
Aquatic Botany 47:293-302.
Irlandi, E.A. 1996. The effects of seagrass patch size and energy regime on growth of
a suspension-feeding bivalve. Journal of Marine Research 54:161-185.
-------
186
Irlandi, E.A. 1997. Seagrass patch size and survivorship of an infaunal bivalve. Oikos
78:511-518.
James, W. 1970. A photographic analysis of Oregon's estuaries. Dept. of Civil
Engineering, Oregon State University.
Jassby, A.D., B.E. Cole, and J.E. Cloern. 1997. The design of sampling transects for
characterizing water quality in estuaries. Estuarine and Coastal and Shelf
Sciences 45:285-302.
Jaworski, N.A. 1981. Sources of nutrients and the scale of eutrophication problems in
estuaries. ]n B.J. Neilson, and L.E. Cronin (eds.), International Symposium on
Nutrient Enrichment in Estuaries. Humana Press, NJ
Jay D.A., W.R. Geyer and D.R. Montgomery. 1997. An ecological perspective on
estuarine classification, submitted to Estuarine Synthesis: the Next decade, J. E.
Hobbie (ed.), SCOPE publication.
Jay, D.A. and J.D. Smith. 1988. Circulation in and classification of shallow, stratified
estuaries. In: Physical Processes in Estuaries, J. Dronkers and W. van
Leussen, Springer-Verlag, Heidelberg, pp. 21-41.
Jensen, S.L., B.D. Robbins, and S.S. Bell. 1996. Predicting population decline:
seagrass demographics and the reconstructive technique. Marine Ecology
Progress Series 136:267-276.
Jerlov, N.G. 1968. Optical Oceanography. Elsevier, NY. 194 pp.
Jones, C.G., J.H. Lawton, and M. Shachak. 1997. Positive and negative effects of
organisms as physical ecosystem engineers. Ecology 78:1946-1957.
Jupp, B.P., D.H.N. Spence. 1997. Limitations on macrophytes in an eutrophic lake,
Loch Leven. Journal of Ecology 65:175-186
Kamphaus, R. 1998. NOAA/PMEL Tsunami Project, Hatfield Marine Science Center,
Newport, OR.
Karelva, P. and U. Wennegren. 1995. Connecting landscape pattern to ecosystem
and population processes. Nature 373:299-373.
Karr, J.R. and E.W. Chu. 1997. Biological monitoring and assessment: using
multimetric indexes effectively. EPA 235-R97-001. University of Washington,
Seattle
-------
187
Keller, M. 1963. Growth and distribution of eelgrass (Zostera marina L.) in Humboldt
Bay, CA. M.S. thesis, Humboldt State College. 53 pp.
Kendall, D.R. 1983. The role of physical-chemical factors in structuring subtidal
marine and estuarine benthos. Final Report. Environmental Laboratory, U.S.
Army Engineer Waterways Experiment Station, Vicksburg, Ml.
Kentula, M.E. 1982. Population dynamics of a Zostera marina L. Bed in Netarts Bay,
Oregon. Ph.D. Thesis Oregon State University, Corvallis, OR. 158 pp.
Kentula, M.E., and C.D. Mclntire. 1986. The autecology and production dynamics of
eelgrass (Zostera marina L.) in Netarts Bay, OR. Estuaries 9:188-199.
Kneib, R.T. 1984. Patterns of invertebrate distribution and abundance in the intertidal
salt marsh, causes and questions. Estuaries 7:392-412.
Kneib, R.T. 1994. Spatial pattern, spatial scale, and feeding in fishes, pp 171-185 In
Stouder, D.J., K.L. Fresh, and R.J. Feller, {eds.}. Theory and application offish
feeding ecology. Univ. South Carolina Press, Columbia, SC.
Kotaki, Y., K. Koike, T. Ogata, S. Sato, Y. Fukuyo, and M. Kodama. 1996. Domoic
acid production by an isolate of Pseudo-nitzschia multiseries, a possible cause
for the toxin detected in bivalves in Ofunato Bay, Japan. In; Harmful and Toxic
Algal Blooms, ed. Yasumoto, T., Oshima, Y., and Fukuyo, Y. UNESCO, pp 151-
154.
Krebs, C.J. 1989. Ecological methodology. Harper and Row, New York, NY.
Kreuz, K.F., A.V. Tyler, G.H. Kruse, and R.L. Demory. 1982. Variation in growth of
Dover soles and English soles as related to upwelting. Transactions of the
American Fisheries Society 111:180-192.
Krygier, E.E., and W.G. Pearcy. 1986. The role of estuarine and offshore nursery
areas for young English sole Parophrys vetulus Girard of Oregon. Fisheries
Bulletin, U.S. 84:119-132.
Lajtha, K. and R.H. Michener (eds.). 1994. Methods in Ecology: Stable Isotopes in
Ecology and Environmental Science. Blackwell Scientific Publications. London.
316 pp.
Lamberson, J.O. 1998. Amphipods as indicators of ecosystem stress and integrity in
Pacific Northwest estuaries. In WED/CEB FY98 Research Plan. U.S.
Environmental Protection Agency, Coastal Ecology Branch, Newport, OR.
-------
188
Lancelot, C. G. Billen, A. Sournin, T. Weisse, F. Colljin, M.J.W. Veldhuis, A. Davies,
and P. Wassman. 1987. Phaeocystis blooms and nutrient enrichment in the
continental coastal waters. Ambio 16:38-46.
Landis, W.G., and J.A. Wiegers. 1997. Design considerations and a suggested
approach for regional and comparative risk assessment. Human and Ecological
Risk Assessment 3:287-297.
Laroche, J.L., S.L. Richardson, and A.A. Rosenberg. 1982. Age and growth of a
Pleuronectid, Parophrys vetulus, during the pelagic larval period in Oregon
coastal waters. Fisheries Bulletin, U.S. 80:93-104.
Larson, R., A. Morang, and L. Gorman. 1997 . Monitoring the coastal environment;
Part II: Sediment sampling and geotechnical methods. Journal of Coastal
Research 13:308-330.
Laurance, W.F., S.G. Laurance, L.V. Ferreira, J.M. Rankin-de Merona, C. Gascon, and
T.E. Lovejoy. 1997. Biomass collapse in Amazonian forest fragments. Science
278:1117-1118.
Lee II, H., B. Boese, J. Pelletier, M. Winsor, D. Specht, and R. Randall. 1993.
Guidance Manual, Bedded Sediment Accumulation Tests. Epa/600/R-93/183.
U.S. Environmental Protection Agency.
Lee, II, H., A. Lincoff, B.L. Boese, F.A. Cole, S.P. Ferraro, J.O. Lamberson, R.J.
Ozretich, R.C. Randall, K.R. Rukavina, D.W. Schults, K.A. Sercu, D.T. Specht,
R. Swartz, and D.R. Young. 1994. Ecological Risk Assessment of the Marine
Sediments at the United Heckathom Superfund Site. ERL-N N269. EPA Region
IX Report. U.S. Environmental Protection Agency, Newport, OR.
Lehmann, A., J.M. Jaquet, and J.B. Lachavanne. 1997. A GIS approach of aquatic
plant spatial heterogeneity in relation to sediment and depth gradients, Lake
Geneva, Switzerland. Aquatic Botany 58:347-362.
Leppakoski, E. 1979. The use of zoobenthos in evaluating effects of pollution in
brackish-water environments, jn The use of ecological variables in
environmental monitoring. The National Swedish Environment Protection Board,
Report PM 1151. pp 151-157.
Levin, L., H. Caswell, T. Bridges, C. DiBacco, and G. Plaia. 1996. Demographic
responses of estuarine polychaetes to pollutants: Life table experiments.
Ecological Applications 6:1295-1313.
-------
189
Levings, C.D. 1990. Strategies for restoring and developing fish habitats in the Strait
of Georgia-Puget Sound inland sea, Northeast Pacific Ocean. Marine Pollution
Bulletin 23:417-422.
Lewin, J. 1977. Persistent blooms of surf diatoms along the Northwest coast. In The
Marine Plant Biomass of the Pacific Northwest Coast, ed. Krauss, R. W., OSU
publication, pp 81-92.
Libes, M. 1986. Productivity-irradiance relationship of Posidonia oceanicia and its
epiphytes. Aquatic Botany 26:285-306
Lipcius, R.N. and W.A. Van Engel. 1990. Blue crab population dynamics in
Chesapeake Bay: Variation in abundance (York River, 1972-1988) and stock-
recruit functions. Bulletin of Marine Science 46:180-194.
Lowery, J.H. JR, Miller, and G.F. Hepner. 1995. A GIS-based sensitivity analysis
of community vulnerability to hazardous contaminants on the Mexico/US border.
Photogrammetric Engineering and Remote Sensing 61:1347-1359.
Lumb, A., R.B. McCammon and J.L. Kittle, Jr. 1994. Users Manual for an expert
system (HSPEXP) for calibration of the hydroiogical simulation program.
USGS WRI 94-4168. 97 pp.
Malloy, K.D., and T.E. Targett. 1994a. The use of RNA:DNA ratios to predict growth
limitation of juvenile summer flounder (Paralichthys dentatus) from Delaware and
North Carolina estuaries. Marine Biology 118:367-375.
Malloy, K.D. and T.E. Targett. 1994b. Effects of ration limitation and low temperature
on growth, biochemical condition, and survival of juvenile summer flounder from
two Atlantic coast nurseries. Transactions of the American Fisheries Society
123:182-193.
Malthus, T.J. and D.G. George. 1997. Airborne remote sensing of macrophytes in
Cefni Reservoir, Anglesey, U.K. Aquatic Botany 58:317-332.
Marschall, E.A. and L.B. Crowder. 1996. Assessing population responses to multiple
anthropogenic effects: a case study with brook trout. Ecological Applications
6:152-167.
McCabe, G.T. Jr., S.A. Hinton, and R.L. Emmett. 1993. In-water restoration between
Miller Sands and Pillar Rock Island, Columbia River: Biological Surveys, 1992.
Coastal Zone and Estuarine Studies Division, NMFS, Seattle, WA.
-------
190
McMillan, C. 1979. Differentiation in Response to Chilling Temperatures among
Populations of Three Marine Spermatophytes, Thalassia testudinum,
Syringodium filiforme, and Halodule wrightii. American Journal of Botany
66:810-819.
McMillan, C. and R. C. Phillips. 1979. Differentiation in Habitat Response among
Populations of New World Seagrasses. Aquatic Botany 7:185-196.
McRoy, C. P. 1966. The standing stock and ecology of eelgrass, Zostera marina,
Izembeck Lagoon, Alaska. M. S. thesis, University of Washington, Seattle.
138 pp.
McRoy, C. P. 1970. On the biology of eelgrass in Alaska. Ph.D. thesis, University of
Alaska, Fairbanks. 156 pp.
McRoy, C. P. and C. McMillan. 1977. Production ecology and physiology of
seagrasses. Jn C. P. McRoy and C. Hefferich (Eds.). Seagrass Ecosystems: A
Scientific Perspective. Marcel Dekker, Inc., NY. pp. 53-87.
Meams, A.J. and T.P. O'Connor. 1984. Biological effects versus pollutant inputs: The
scale of things, in Concepts in Marine Pollution Measurements. H.H. White
(ed.) Maryland Sea Grant College, College Park, pp 694-722.
Menge.B.A., E.L. Berlow, C.A. Blanchette, S.A. Navarette, and S B. Yamada. 1994.
The keystone species concept: variations in interaction strength in a rocky
intertidal foodweb. Ecological Monographs 65:249-286.
Miles, E.L. and R.B. Whitlatch. 1997. "Priscilla": a portable in situ suction sampling
device. In: Maney, E.J., Jr. and C.H. Ellis, Jr. (eds.), Diving for Science -1997.
Proceedings of the American Academy of Underwater Sciences, Northeastern
University, Boston, pp 117-121.
Milicich, M.J., and J.H. Choat. 1992. Do otoliths record changes in somatic growth
rate - conflicting evidence from a laboratory and field study of a temperate reef
fish, Parika scaber. Australian Journal of Marine Freshwater Reseach
43:1203-1214.
Miller, D.R. 1974. Sensitivity analysis and validation of simulation models. Journal of
Theoretical Biology 48:345-360.
Miller, R.L. and B.F. McPherson. 1993. Causes of light attenuation in estuarine waters
of southwestern Florida; In: Morris, L.J. and D.A. Tomasko (eds). ]n
Proceedings and Conclusions of Workshops on: Submerged Aquatic Vegetation
-------
191
and Photosynthetically Active Radiation. Special Publication SJ93-SP13. St.
Johns River Water Management District, Palatka, FL.
Miiot, C. 1997. The rise in toxic tides. Science News 152:202-204.
Misitano, D.A. 1976. Size and stage of development of larval English sole, Parophrys
vetulus, at time of entry into Humboldt Bay. California Fish and Game.
62:93-98.
Molony, B.W., and J.H. Choat. 1990. Otolith increment widths and somatic growth
rate: The presence of a time-lag. Journal of Fish Biology 37:541-551.
Monaco, M.E., D.M. Nelson, R.L. Emmettand S.A. Hinton. 1990. Distribution and
abundance of fishes and invertebrates in west coast estuaries. Vol. 1: Data
summaries. ELMR Rep. No. 4. NOAA/NOS Strategic Environmental
Assessments Division, Silver Spring, MD, 232 pp.
Montgomery, D.R. and J.M. Buffington. 1993. Channel classification, prediction of
channel response, and assessment of channel condition. Washington State
Dept. of Natural Resources Report TFW-SH10-93-002 for the SHAMW
Committee of the Washington State Timber/Fish/Wildlife Agreement. 84 pp.
Moore, K.A., H.A. Neckles, and R.J. Orth. 1996. Zostera marina (eelgrass) growth and
survival along a gradient of nutrients and turbidity in the lower Chesapeake Bay.
Marine Ecology Progress Series 142:247-259.
Moore, K.A., R. L. Wetzel and R.J. Orth. 1997. Seasonal pulses of turbidity and their
relations to eelgrass (Zostera marina L.) survival in an estuary. Journal of
Experimental Marine Biology and Ecology 215:115-134.
Morris, L.J. and D.A. Tomasko (eds.). 1993. Proceedings and Conclusions of
Workshops on: Submerged Aquatic Vegetatioin and Photosynthetically Active
Radiation. Special Publication SJ93-SP13. St. Johns River Water Managment
District. Palatka, FL.
Mosegaard, H., H. Svedang, and K. Taberman. 1988. Uncoupling of somatic and
otolith growth rates in arctic char (Salvelinus alpinus) as an effect of differences
in temperature response. Canadian Journal of Fisheries and Aquatic Sciences
45:1514-1524
Muehlstein, L.K., D. Poeter, F.T. Short. 1988. Labyrinthula sp., a marine slime mold
producing the symptoms of wasting disease in eelgrass, Zostera marina. Marine
Biology 99:465-472
-------
192
Mukai, H., Aioi, and Y. Ishida. 1980. Distribution and biomass of eelgrass (Zostera
marina L.) and other seagrasses in Odawa Bay, Central Japan. Aquatic Botany
8:337-342.
National Oceanic and Atmospheric Administration. 1988. National Estuarine
Inventory: Supplement 1. Physical and Hydrologic Characteristics, The Oregon
Estuaries. Rockville, MD. Office of Oceanography and Marine Assessment.
National Oceanic and Atmospheric Administration/U.S. Environmental Protection
Agency. Team on Near Coastal Waters. 1991. Susceptibility and status of west
coast estuaries to nutrient discharges: San Diego Bay to Puget Sound.
Summary Report. 35 pp.
National Oceanic and Atmospheric Administration. 1997. Review of sediment quality
investigations in San Francisco Bay. NOAA Tech. Monogr. NOS ORCA 116.
Seattle, NOAA Hazardous Materials Response and Assessment Division.
National Oceanic and Atmospheric Administration. 1997. C-CAP-Changes in land
cover in the Columbia River Estuary: 1989-1992. Charleston, S.C.: U.S.
Department of Commerce, National Oceanic and Atmospheric Administration
Coastal Services Center. 1997. NOAA CSC/1-97/001.
National Oceanic and Atmospheric Administration. 1998. NOAA's Estuarine
Eutrophication Survey. Vol. 5: Pacific Coast Region. Silver Springs, MD.
National Oceanic and Atmospheric Administration, Office of Ocean Resources
Conservation Assessment. 68 pp.
Neckles HA, R.L. Wetzel and R.J. Orth. 1993. Relative effects of nutrient enrichment
and grazing on epiphyte-macrophyte (Z. marina L.) dymanics. Oecologia
93:285-295
Nelson, W.G. 1990. Prospects for development of an index of biotic integrity for
evaluating habitat degradation in coastal systems. Chemistry and Ecology
4:197-210.
Nelson, W.G. and R.W. Virnstein. 1995. Long-term dynamics of seagrass
macrobenthos: asynchronous population variability in space and time, in
Biology and Ecology of Shallow Coastal Waters. A. Eleftheriou et al. (eds.)
Olsen and Olsen, Denmark, pp 185-190.
NRC. 1997. Building a foundation for sound environmental decisions. Committee on
Research Opportunities and Priorities for U. S. Environmental Protection
Agency. National Academy Press, Washington.D.C. 87 pp.
-------
193
Nomme, K. M. and P. G. Harrison. 1991. Evidence for interation between the
seagrasses Zostera marina and Zostera japonica on the Pacific coast of
Canada. Canadian Journal of Botany 69(9):2004-2010.
Norris, J.G., S. Wyllie-Echeverria, T. Mumfored, A. Bailey, and T. Turner. 1997.
Estimating basal area coverage of subtida! seagrass beds using underwater
videography. Aquatic Botany 58:269-288.
Ogata, E. and T.Matsui. 1965. Photosynthesis in several marine plants of Japan as
affected by salinity, drying and pH, with attention to their growth habitats.
Botany Maine 8:199-217
Olafsson, E.B., C.H. Peterson and W.G. Ambrose. 1994. Does recruitment limitation
structure populations and communities of macro-invertebrates in marine soft
sediments: The relative significance of pre- and post-settlement processes.
Oceanography and Marine Biology: Annual Review 32:65-109.
Olson, A.M., and E.G. Doyle. 1995. Light requirements of eelgrass, a literature survey.
Draft Interim Report. School of Marine Affairs, University of Washington, Seattle,
WA.
Olson, A.M., E.G. Doyle, and S.D. Visconty. 1997. Light requirements of eelgrass: a
literature survey, in Simenstad, C.A., R.M. Thorn, and A.M. Olson. Mitigation
between regional transportation needs and preservation of eelgrass beds.
Volume 2, Appendicies. Report no. QA-RD 421.2, Washington State
Department of Transportation, Olympia, WA.
Onuf, C.P. 1987. The ecology of Mugu Lagoon, California: An estuarine profile.
Biological Report.. Marine Science Institute, University California, Santa
Barbara, CA. 85:7-15
Oregon Department of Land Conservation and Development. 1987. Oregon estuary
plan book of the Oregon coastal management program, Oregon Department of
Land Conservation and Development, Salem, OR.
Orth, R.J. 1977. The importance of sediment stability in seagrass communities. ]q
Coull BC (ed) Ecology of marine benthos. University of South Carolina Press,
Columbia, pp 281-300
Orth, R.J. and KA. Moore. 1988. Distribution of Zostera marina L. and Ruppia
maritima L. sensu lato along depth gradients in the lower Chesapeake Bay,
USA. Aquatic Botany 32:291-305
-------
194
Orth, R.J. 1992. A perspective on plant-animal interactions in seagrasses: Physical
and biological determinants influencing plant and animal abundance, pp 147-
164. ]n Plant-Animal Interactions in the Marine Benthos ( D. M. John, S.J.
Hawkins, and J.H. Price eds), Clarendon Press, Oxford.
Orth, R.J., J.F. Nowak, G.F. Anderson, D.J. Wilcox, J.R. Whiting, and L.S. Nagey.
1995. Distribution of submerged aquatic vegtation in the Chesapeake Bay and
tributaries and Chincoteague Bay. Final report summitted to: U.S. Environmental
Protection Agency (grant CB993267-01-0).
Osborne, L.L. and M.J. Riley. 1988. Empirical relationships between land use/cover
and stream water quality in an agricultural watershed. Journal of Environmental
Management 26:9-27.
Ostenfeld, C. H. 1908. On the Ecology and distribution of the grass-wrack (Zostera
marina) in Danish Waters. Report of the Danish Biology Station 16:1-62.
Ozretich, R. 1998. Susceptibility of Eelgrass Beds in Oregon Estuaries to Changes in
Watershed Uses. Research Preproposal, Coastal Ecology Branch,
WED/NHEERL, Newport, OR.
Paerl, H.W. and D.F. Mille. 1996. Physiological ecology of toxic aquatic
cyanobacteria. Phycologia 35:160-167.
Paerl, H.W. 1996. A comparison of cyanobacterial bloom dynamics in freshwater,
estuarine and marine environments. Phycologia 35:25-35.
Pan, Y., D.V.S. Rao, K.H. Mann, W.K.W. Li, and W.G. Harrison. 1996a. Effects of
silicate limitation on production of domoic acid, a neurotoxin, by the diatom
Pseudo-nitzschia multiseries. II Continuous culture studies. Marine Ecology
Progress Series 131:235-243.
Pan, Y., D.V.S. Rao, K.H. Mann. 1996b. Changes in domoic acid production and
cellular chemical composition of the toxigenic diatom Pseudo-nitzschia
multiseries under phosphate limitation. Journal of Phycology 32:371-381.
Patriquin, D. 1973. Estimation of growth rate, production and age of the marine
angiosperm Thalassia testudinum Konig. Caribbean Journal of Science 13:111-
123.
Peterson, B.J., R.W. Howarth, and R.H. Garritt. 1985. Multiple stable isotopes used to
trace the flow of organic matter in estuarine food webs. Science 227:1361 -1363.
-------
195
Peterson, C.H. 1982. The importance of predation and intra- and interspecific
competition in the population biology of two infaunal suspension-feeding
bivalves, Protothaca staminea and Chione undatella. Ecological Monographs
52:437-475.
Peterson, C.H., K. Scheidegger, P. Komar. 1982. Sand dispersal patterns in an active
margin estuary of the northwestern United States as indicated by sand
composition, texture, and bedforms. Marine Geology 59:77-96.
Peterson, C.H., K. Scheidegger, P. Komar, and W. Niem. 1984. Sediment composition
and hydrography in six high-gradient estuaries of the Northwestern United
States. Journal of Sedimentary Petrology 54:86-97.
Peterson, C.H. and R. Black. 1988. Density-dependent mortality caused by physical
stress interacting with biotic history. American Naturalist 131:257-270.
Peterson, C.H. and R. Black. 1993. Experimental tests of the advantages and
disadvantages of high density for two coexisting cockles in a Southern Ocean
lagoon. Journal of Animal Ecology 62:614-633.
Pfaff, J. D., C.A. Brockhoff, and J. W. O'Dell. 1991. The determination of inorganic
anions in water by ion chromatography - Method 300.0 U.S. Environmantal
Protection Agency, Cincinnati OH.
Phillips, G.L., D. Eminson, and B. Moss. 1978. A mechanism to account for
macrophyte decline in progressively euthrophicated freshwaters. Aquatic
Botany 4:103-126
Phillips, R. C. and R. R. Lewis. 1983. Influences of Environmental Gradients on
Variations in Leaf Widths and Transplant Success in North American
Seagrasses. Marine Technological Society Journal 12:59-68.
Phillips, R. C., C. McMillan and K. W. Bridges. 1983. Phenology of Eelgrass, Zostera
marina L., along Latitudinal Gradients in North America. Aquatic Botany 15:145-
156.
Phillips, R.C. 1984. The ecology of eelgrass meadows in the Pacific Northwest: a
community profile. U.S. Fish and Wildlife Service, Washington, DC, FWS/OBS-
84/24. 85 pp.
Phillips, R. C. and E. G. Mefiez. 1988. Seagrasses. Smithsonian Contributions to
Marine Science. No. 34. Smithsonian Institution Press, Washington, DC.
104 pp.
-------
196
Posey, M. H. 1986a. Changes in a benthic community associated with dense beds of
a burrowing deposit feeder, Callianassa califbmiensis. Marine Ecology Progress
Series 31:15-22.
Posey, M. H. 1986b. Predation on a burrowing shrimp: distribution and community
consequences. Journal of Experimental Marine Biology and Ecology 103:143-
161.
Posey, M. H. 1987. Effects of lowered salinity on activity of the ghost shrimp
Callianassa califomiensis. Northwest Science 61:93-96.
Posey, M.H. 1990. Functional approaches to soft-substrate communities: how useful
are they? Review of Aquatic Science 2:343-356.
Posey, M.H., B.R. Dumbauld, and D.A. Armstrong. 1991. Effects of a burrowing mud
shrimp, Upogebia pugettensis (Dana), on abundances of macro-infauna.
Journal of Experimental Marine Biology and Ecology 148:283-294.
Power, J.H. 1996. Simulation of the effect of advective-diffusive processes on
observations of plankton abundance and population rates. Journal of Plankton
Research. 18:1881-1896.
Power, J.H. 1998. Spatial variation of growth and condition in juvenile English sole
relative to substrate characteristics. Id WED/CEB FY98 Research Plan. U.S.
Environmental Protection Agency, Coastal Ecology Branch, Newport, OR.
Power, M. and G. Power. 1995. A modeling framework of analyzing anthropogenic
stresses on brook trout (Salvelinus fontinalis) populations. Ecological Modeling.
80:171-185.
Puget Sound Estuary Program. 1987. Recommended protocols for sampling and
analyzing subtidal benthic macroinvertebrate assemblages in Puget Sound. ]n
Recommended protocols for measuring selected environmental variables in
Puget Sound. Puget Sound Water Quality Action Team, Olympia, WA.
Puget Sound Estuary Program. 1990. Recommended guidelines for sampling soft-
bottom demersal fishes by beach seine and trawl in Puget Sound. ]n
Recommended protocols for measuring selected environmental variables in
Puget Sound. Puget Sound Water Quality Action Team, Olympia, WA.
Quinn, J.F., S. Wing, and L.W. Botsford. 1993. Harvest refugia in marine
invertebrates: models and applications to the red sea urchin Strongylocentrotus
franciscanus. American Zoologist 33:537-550.
-------
197
Raffaelli, D., J. Raven, and L. Poole. 1998. Ecological impact of green macroalgal
blooms. Annual Review of Oceanography and Marine Biology. In press.
Rapport, D.J., H.A. Regier, and T.C. Hutchinson. 1985. Ecosystem behavior under
stress. American Naturalist 125: 617-640.
Rasmussen, E. 1977. The wasting disease of eelgrass (Zostera marina) and its
effects on environmental factors and fauna, jn McRoy CP, Helfferich C (ed)
Seagrass ecosystems: a scientific perspective. Marcel Dekker, NY
Reish, D.J. 1986. Benthic invertebrates as indicators of marine pollution: 35 years of
study. In: Oceans 86 Conference Record. Vol. 3. Monitoring strategies
symposium. Institute of Electrical and Electronic Engineers, Piscataway, NJ.
pp. 885-888.
Rhoads, D.C. 1974. Organism-sediment relations on the muddy sea floor.
Oceanography and Marine Biology Annual Review 12:263-300.
Ricker, W.E. 1973. Critical statistics from two reproduction curves. In "Fish Stocks
and Recruitment", B.B. Parrish (ed). Rapports et Proces-Verbaux des Reunions.
Vol. 164, pages 333-340.
Ricker, W.E. 1975. Computation and Interpretation of Biological Statistics of Fish
Populations. Bulletin of the Fisheries Research Board of Canada Bulletin 191.
382 pp.
Robbins, B.D., and S.S. Bell. 1994. Seagrass landscapes: a terrestrial approach to
the marine subtidal environment. Trends in Ecology and Evolution 9:301-304.
Robbins, B.D. 1997. Quantifying temporal change in seagrass areal coverage: the use
of GIS and low resolution aerial photography. Aquatic Botany 58:259-267.
Robbins, B.D. 1998. The hermit crab, Pagurus maclaughlinae and live gastropods: an
investigation on resource provision in a subtidal seagrass landscape. Ph.D.
Dissertation, University of South Florida, Tampa, FL. 167 pp.
Robinson, G.R., R.D. Holt, M.S. Gaines, S.P. Hamburg, M.L. Johnson, H.S. Fitch, E.A.
Martinko. 1992. Diverse and contrasting effects of habitat fragmentation.
Science 257:524-526.
Robinson, S.K., F.R. Thompson III, T.M. Donovon, D.R. Whitehead, and J. Faaborg.
1995. Regional forest fragmentation and the nesting success of migratory birds.
Science 267:1987-1990.
-------
198
Roitblat, H.L., W.W.L. Au, P.E. Nachtigall, R. Shizumura, and G. Moons. 1995. Sonar
recognition of targets embedded in sediment. Neural Networks 8:1263-1273.
Rossbach, K.A. 1997. Distinguishing inshore and offshore communities of bottlenose
dolphins (Tursiops truncatus) near Grand Bahama Island, Bahamas. Masters
Thesis, Dept. Fisheries and Wildlife, Oregon State University, Corvallis, OR.
Rosenberg, A.A. and J.L. Laroche. 1982. Growth during metamorphosis of English
so\e, Parophrys vetulus. Fisheries Bulletin, U.S. 80:152-155.
Rosenberg, A. A. 1982. Growth of juvenile English sole, Parophrys vetulus, in
estuarine and open coastal nursery grounds. Fisheries Bulletin, U.S.
80:245-252.
Rossi, R.E., D.J. Mulla, A.G. Journel, and E.H. Franz. 1992. Geostatistical tools for
modeling and interpreting ecological spatial dependence. Ecological
Monographs 62:277-314.
Roughgarden, J. and Y. Iwasa. 1986. Dynamics of a metapopulation with
space-limited subpopulations. Theoretical Population Biology 29:235-261.
Rozas, L.P. and T.J. Minello. 1997. Estimating densities of small fishes and decapod
crustaceans in shallow estuarine habitats: a review of sampling design with
focus on gear selection. Estuaries 20:199-213.
Rumrill, S.S. 1990. Natural mortality of marine invertebrate larvae. Ophelia 32:163-
198.
Rumrill, S. and J. Christy. 1996. Ecological impacts of oyster ground culture within
estuarine tidelands: South Slough National Estuarine Research Preserve.
Report prepared for the Oregon Dept. of Land Conservation and Development.
Ryther, J.H., and W.M. Dunstan. 1971. Nitrogen, phosphorous, and euthrophication in
the coastal marine environment. Science 171:1008-1013.
Sabol, B. and R.E. Melton. 1995. Development of an automated system for detection
and mapping of submersed aquatic vegetation with hydroacoustic and global
positioning system technologies. Report 1: The submersed aquatic vegetation
early warning systems (SAVEWS) - system description and user's guide (ver.
1.0). Joint Agency Guntersville Project Aquatic Plant Management.
Sabol, B., E. McCarthy, and K. Rocha. 1997. Hydroacoustic basis for detection and
characterization of eelgrass (Zostera marina). Presented at 4th Conference on
-------
199
Remote Sensing for Marine and Coastal Environments, Orlando, FL, 17-19
March.
Sayce, K.A. and R. Horner. 1996. Pseudo-nitzschia spp. in Willapa Bay,
Washington, 1992 and 1993. In Harmful and Toxic Algal Blooms, ed. T.
Yasumoto, Y. Oshima, and Y. Fukuyo. UNESCO pp. 131-134.
Scheidegger, K.M., L.D. Kulm, and E.J. Runge. 1971. Sediment sources and dispersal
patterns of Oregon continental shelf sands, Journal of Sedimentary Petrology,
41:1112-1120.
Schirripa, M.J., and C.P. Goodyear. 1997. Simulation of alternative assumptions of
fish otolith-somatic growth with a bioenergetics model. Ecological Modeling
102:209-223.
Schmieder, K. 1997. Littoral zone—GIS of Lake Constance: a useful tool in lake
monitoring and autecological studies with submersed macrophytes. Aquatic
Botany 58:333-346.
Schwartz, M. L., J. Mahala, and H.S. Bronson III. 1985. Net shore-drift along the
Pacific Coast of Washington state. Shore and Beach, Vol. 53:21-25.
Secor, D.H., M.G. White, and J.M. Dean. 1991a. Immersion marking of larval and
juvenile hatchery-produced striped bass with oxytetracycline. Transactions of
the American Fisheries Society 120:261-266.
Secor, D.H, J.M. Dean, and E.H. Laban. 1991b. Manual for otolith removal and
preparation for microstructural examination. Technical Publication 1991-01,
Belle W. Baruch Institute for Marine Biology and Coastal Research, University of
South Carolina.
Seizinger, S.P., S.W. Nixon and M.E.Q. Pilson. 1984. Denitrification and nitrous oxide
production in a coastal marine ecosystem. Limnology and Oceanography
29:73-83.
Setchell, W. A. 1920. Geographical Distribution of the Marine Spermatophytes.
Bulletin of Torrey Botany CI. 47:563-579.
Setchell, W. A. 1929. Morphological and Phenological Notes on Zostera marina L.
University of California Publcation on Botany 14:389-452.
Sharp, W.C., W.A. Lellis, M. Pardee-Woodring, M.J. Butler IV, W.F. Herrnkind, J. H.
Hunt, T.R. Matthews. 1998. The Use of Coded Microwire Tags for
-------
200
Mark-Recapture Studies of Juvenile Caribbean Spiny Lobster, Panulirus argus.
Abstract. 27th Annual Benthic Meetings, Melbourne, FL, March 12-15.
Shaw, M. and G.P. Jenkins. 1992. Spatial variation in feeding, prey distribution and
food limitation of juvenile flounder Rhombosolea tapirina Gunther. Journal of
Experimental Marine Biology and Ecology 165:1-21.
Shi, Y., D.R. Gunderson, and P.J. Sullivan. 1997. Growth and survival of 0+ English
sole, Pleuronectes vetulus, in estuaries and adjacent nearshore waters off
Washington. Fisheries Bulletin 95:161-173.
Shirzad, F.F., S.P Orlando, C.J. Klein, S.E. Holliday, M.A. Warren, and M.E. Monaco.
1988. National Estuarine Inventory: Supplement 1. Physical and Hydrologic
Characteristics the Oregon Estuaries. National Oceanic and Atmospheric
Administration. Ocean Assessments Division, Rockville MD. 25 pp.
Short, F. T. 1987. Effects of sediment nutrients on Seagrasses: Literature review and
mesocosm experiment. Aquatic Botany 27:41-57.
Short, F.T., and D.M. Burdick. 1996. Quantifying eelgrass habitat loss in relation to
housing development and nitrogen loading in Waquoit Bay, MA. Estuaries
19:730-739.
Sigleo, A.C. and S.A. Macko. 1985. Stable isotope and amino acid composition of
estuarine dissolved colloidal material. In Marine and Estuarine Geochemistry,
ed Sigleo and Hattori, Lewis Publishers, Chelsea, Ml pp. 29-46
Sigleo, A. C. 1985. Organic nitrogen fluxes at the sediment water interface as
determined by benthic chamber studies in the Patuxent estuary. American
Chemical Society Meeting. Program with Abstracts.
Sigleo, A. C. and D.J. Shultz. 1993. Amino acid composition of suspended particles,
sediment trap material and benthic sediment in the Potomac Estuary. Estuaries
16:405-415.
Sigleo, A. C., K.A. Sayce, and P. Stotts. 1997. Multiple nutrient sources in an
unurbanized Pacific Northwest estuary, Abstract, Estuarine Research Federation
Meeting, October! 2 -16, 1997
Simmenstad, C.A., J.A. Estes, and K.W. Kenyon. 1978. Aleuts, sea otters, and
alternate stable state communities. Science 200:403-41
-------
201
Simenstad, C.A. and R. C. Wissmar. 1985. 5 13C evidence of the origins and fates of
organic carbon in estuarine and nearshore food webs. Marine Ecology Progress
Series 22:141-152.
Simenstad, C.A., C.D. Tanner, and R.M. Thom. 1989. Estuarine wetland restoration
monitoring protocol: Appendices. U.S. Environmental Protection Agency,
Region 10, Office of Puget Sound, Seattle, WA.
Simenstad, C.A., L.F. Small, C.D. Mclntire, D.A. Jay and C. Sherwood. 1990.
Columbia River estuarine studies: an introduction to the estuary, a brief history,
and prior studies. Progress in Oceanography 25:1-14.
Simenstad, C.A., C.T. Tanner, R.M. Thom, and L.L. Conquest. 1991. Estuarine habitat
assessment protocol. EPA 910/9-91 -037.
Simenstad, C.A. and K.L. Fresh. 1995. Influence of intertidal aquaculture on benthic
communities in Pacific Northwest estuaries: Scales of disturbance. Estuaries
18:43-70.
Simenstad, C.A. and R. Thom. 1995. Spartina alterniflora (smooth cordgrass) as an
invasive halophyte in Pacific Northwest estuaries. Hortus Northwest 6:9-12, 38-
39.
Simenstad, C.A., R.M. Thom, and A.M. Olson. 1997. Mitigation between regional
transportation needs and preservation of eelgrass beds. Report no. QA-RD
421.1, Washington State Department of Transportation, Olympia, WA.
143 pp and Appendices
Simenstad, C.A., D. Armstrong, and L. Conquest. 1997a. Estuarine Landscape
Structure Interaction with Ecosystem Processes. Project Summary #11. ]n
Pacific Northwest Research Program. May 1997. Peer Review. U.S.
Environmental Protection Agency, Western Ecology Division, Corvallis, OR.
Simenstad, C.A., R. Edwards, D. Montgomery, and L. Conquest. 1997b. Development
and Mapping of Pacific Northwest Watershed-Estuary Landscape Typology.
Project summary #7a. ]n Pacific Northwest Research Program. May 1997. Peer
Review. U.S. Environmental Protection Agency, Western Ecology Division,
Corvallis, OR.
Sivonen, K. 1996. Cyanobacterial toxins and toxin production. Phycologia 35:12-24
Smith, R. E., D.H. Peterson, S.W. Hager, D.D. Harmon, L.E. Scheme!, and R.E.
Herndon. 1985. Seasonal and interannual nutrient variability in northern San
-------
202
Francisco Bay. In Marine and Estuarine Geochemistry, ed Sigleo and Hattori,
Lewis Publishers, Chelsea, Ml. pp.137-159.
Society of Environmental Toxicology and Chemistry (SETAC). 1997. Ecological Risk
Assessment Technical Issue Paper. Pensacola, FL.
Sogard, S.M. 1992. Variability in growth rates of juvenile fishes in different estuarine
habitats. Marine Ecology Progress Series 85:35-53.
Sogard, S.M., and K.W. Able. 1992. Growth variation of newly settled winter flounder
(Pseudopleuronectes americanus) in New Jersey estuaries as determined by
otolith microstructure. Netherlands Journal of Sea Research 29:163-172.
Sogard, S.M. 1994. Use of suboptimal foraging habitats by fishes: Consequences to
growth and survival. Pages 103-131 ]n Stouder, D.J., K.L. Fresh, and R.J.
Feller, {eds.}. Theory and application of fish feeding ecology. University of
South Carolina Press, Columbia, SC
Sokal, R.R., and F.J. Rohlf. 1995. Biometry. The principles and practice of statistics
in biological research. 3rd ed. W.H. Freeman and Company, San Francisco,
CA.
Specht, D. T. 1974. The use of Standardized algal bioassays for nutrient assessment
in coastal Oregon Estuaries. Proc. 4th Annual Technology Conference on
Estuaries of the Pacific Northwest, Oregon State University Engineering
Experimental Station Circ. 50:15-31.
Specht, D. T. Unpublished data. 1975-76 field survey, Yaquina Bay, Oregon.
Specht, D. T. 1976. Seasonal variation of algal biomass production potential and
nutrient limitation in Yaquina Bay, Oregon, pp. 149-174. ]n Biostimulation and
Nutrient Assessment. (Middlebrooks, E. J., D. H. Falkenborg, and T. E.
Maloney, Eds.). Proc. Biostimulation and Nutrient Assessment Symposium,
Utah State Univ., Logan, September, 1975, PRWG168-1. (Subsequently
republished by Ann Arbor Press: Ann Arbor, 1976.)
s
Specht, D. T. and H. Lee II. (In prep.) Field (intertidal) vs. laboratory growth rates of
the tellinid clam, Macoma nasuta (Conrad)in Yaquina Bay, OR USA.
Stevens, D.L. 1994. Implementation of a national monitoring program. Journal of
Environmental Management. 42:1-29
-------
203
Stevens, D.L. 1997. Variable density grid-based sampling designs for continuous
spatial populations. Environmetrics 8:167-195
Stevenson, J.C., L.W. Staver, K.W. Staver. 1993. Water quality associated with
survival of submersed aquatic vegetation along an estuarine gradient. Estuaries
16:436-361.
Stout, H. (ed.). 1976. The Natural Resources and Human Utilization of Netarts Bay,
Oregon. NSF Student Originated Studies Program. Oregon State University,
Corvallis. 247 pp.
Strickland, J.D.H. and T.R. Parsons. 1977. A Practical Handbook of Seawater
Analysis, Bulletin 167 (2nd Edition), Fisheries Research Board of Canada,
Ottawa. 310 pp.
Stross, R. G. and R.C. Sokal. 1989. Runoff and flocculation modify underwater light
environment of the Hudson River estuary. Estuarine and Coastal Shelf Science
29:305-316.
Stull, J.K. 1995. Two decades of marine biological monitoring, Palos Verdes,
California, 1972-1992. Bulletin of the Southern California Academy of Science
94:21-45.
Suchanek, T.H. 1983. Control of seagrass communities and sediment distribution by
Callinassa (Crustacea, Thalassinidea) bioturbation. Journal of Marine Research
41:281-298
Suter, G.W. II, ed. 1993. Ecological risk assessment. Lewis Publishers. Chelsea, Ml.
Suthers, I.M. 1996. Spatial variability of recent otolith growth and RNA indices in
juvenile Diaphus kapalae (Myctophidae): An effect of flow disturbance near an
island? Marine Freshwater Research 47:273-282.
Swartz, R.C., W.A. De Ben, and A.J. McErlean. 1974. Comparison of species diversity
and faunal homogeneity indices as criteria of change in biological communities.
E P A-600/4-74-004.
Swartz, R.C., F.A. Cole, D.W. Schults, and W.A. DeBen. 1986. Ecological changes in
the Southern California Bight near a large sewage outfall: benthic conditions in
1980 and 1983. Marine Ecology Progress Series 31:1-13.
Swartz, R.C. 1987. Toxicological methods for determining the effects of contaminated
sediment on marine organisms. Pp. 183-198. in K.L. Dickson, A.W. Maki, and
-------
204
W.A. Brungs, (eds). Fate and Effects of Sediment Bound Chemicals in Aquatic
Systems.
Swinbanks, D. D., and J.L. Luternauer. 1987. Burrow distribution of thalassinidean
shrimp on a Fraser Delta tidal flat, British Columbia. Journal of Paleontology
61:315-332.
Thayer, G.W., D.A. Wolfe, and R.B. Williams. 1975. The impact of man on seagrass
systems. American Scientist 63:288-296.
Thayer, G.W., W.J. Kenworthy, and M.S. Fonseca. 1984. The ecology of eelgrass
meadows of the Atlantic coast: A community profile FWS/OBS 84/02
Thorn, R.M. 1987. The biological importance of Pacific Northwest estuaries. Northwest
Environmental Journal 3:21-42.
Thorn, R. M. 1990. Spatial and temporal patterns in plant standing stock and primary
production in a temperate seagrass system. Marine Botany 33(6):497-510.
Thorn, R. M. 1990. A review of eelgrass (Zostera maraina L.) transplanting projects in
the Pacific Northwest. Northwest Environmental Journal 6:121-137.
Thorn, R.M. and L. Hallum. 1990. Long-term changes in the areal extent of tidal
marshes, eelgrass meadows and kelp forests of Puget Sound. Final report to
U.S. Environmental Protection Agency, EPA 910/9-91-005. Fisheries Research
Institute, University of Washington, Seattle. FRI-UW
Thorn, R., B. Miller, and M. Kennedy. 1995. Temporal patterns of grazers and
vegetation in a temperate seagrass system. Aquatic Botany 50:201-205
Thorn, R.M. and A.B. Borde. 1996. Human intervention in the coastal ecosystems.
Report prepared for the Pacific Northwest Coastal Ecosystems Regional
Research Study Program, FY 1996 Workshop.
Thorn, R. 1997. Personal communication. Pacific Northwest National Laboratory,
Battelle Marine Sciences Laboratory, Sequim, WA.
Thum, A.B. 1972. An ecological study of Diatomovora amoena, an interstitial acoel
flatworm, in a estuarine mudflat on the Central Coast of Oregon. A thesis
submitted to Oregon State University. 185 pp.
-------
205
Thrush, S.F., R.B. Whitlatch, R.D. Pridmore, J.E. Hewitt, V.J. Cummings, and M.R.
Wilkinson. 1996. Scale-dependent recolonization: the role of sediment stability
in a dynamic sandflat habitat. Ecology 77:2472-2487.
Toole, C.L. 1980. Intertidal recruitment and feeding in relation to optimal utilization of
nursery areas by juvenile English sole (Parophrys vetulus: Pleuronectidae).
Environmental Biology of Fisheries 5:383-390.
Ulanowicz, R.E., M.L. Ali, A.V. Vivian, D.R. Heinle, W.A. Richkus and J.K. Summers.
1982. Identifying climatic factors influencing commercial fish and shellfish
landings in Maryland. Fishery Bulletin 80:611-619.
Underwood, A. J. 1991. Beyond BACI - Experimental designs for detecting human
environmental impacts on temporal variations in natural populations. Australian
Journal of Marine and Freshwater Research. 42:569-587.
U.S. Army Corps of Engineers. Eelgrass transplanting techniques.
U.S. Army Corps of Engineers. 1996. Environmental effects of dredging. Technical
notes. Volume II. U.S. Army Engineer Waterways Experiment Station,
Vicksburg, MS.
U.S. Environmental Protection Agency/ACE. 1991. Evaluation of Dredged Material for
Ocean Disposal (Testing Manual). U.S. EPA Report No. 503/8/91/001, Office
of Marine and Estuarine Protection, Washington, D.C.
U.S. Environmental Protection Agency. 1991. National Estuary Program: Monitoring
Guidance Document. Draft. EPA 503/8-91-002.
U.S. Environmental Protection Agency (USEPA). 1992. EMAP-Estuaries. Virginia
Province 1990 Demonstration Project Report. EPA/600/R-92/100.
U.S. Environmental Protection Agency. 1992a. EMAP-Estuaries. Virginia Province
1990 Demonstration Project Report. EPA/600/R-92/100.
U.S. Environmental Protection Agency. 1992b. Framework for Ecological Risk
Assessment. EPA/630/R-92/001.
U.S. Environmental Protection Agency. 1995. An SAB Report: Ecosystem
management- imperative for a dynamic world. EPA-SAB-EPEC-95-003.
U.S.
Environmental Protection Agency. 1996. Strategic plan for the Office of Research
and Development. EPA/600/R-96/059.
-------
206
U.S. Environmental Protection Agency. 1997. Update to ORD's strategic plan.
EPA/600/R-97/015. US U. S. Environmental Protection Agency, Center for
Environmental Research Information, Cincinnati, OH. 74 pp.
U.S. Environmental Protection Agency. 1997. Update to ORD's strategic plan. Office
of Research & Development (ApriH 997) EPA/600-R-97/015. Washington DC
74 pp.
U. S. Environmental Protection Agency. 1997a. Guidance on Cumulative Risk
Assessment. Part 1. Planning and Scoping (Available at:
www.epa.gov/ORD/spc/cumrisk2.htm).
U.S. Environmental Protection Agency. 1997b. Guiding Principles for Monte Carlo
Analysis. EPA/630/R-97/001.
Valiela, I., K. Foreman, M. LaMontagne, D. Hersh, J. Costa, P. Peckol, B. DeMeo-
Anderson, C. D'Avanzo, M. Babione, C. Sham, J. Brewley, and K. Lajtha. 1992.
Couplings of watersheds and coastal waters: Sources and consequences of
nutrient enrichment in Waquoit Bay, MA. Estuaries 18:443-457.
van Arkel, M.A., and M. Mulder. 1975. A device for quantitative sampling of benthic
organisms in shallow water by means of a flushing technique. Netherlands
Journal of Sea Research 9:365-370.
Van Winkle, W. 1977. Proceeding on Conference on Assessing the Effects of Power-
Plant Induced Mortality on Fish Populations. May 3 -6, 1977. Sponsored by
Oak Ridge National Laboratory, Energy Research Development Administration
and Electric Power Research Institute. W. Van Winkle (ed). Pergamon Press.
Virnstein, R.W. 1990. The large spatial and temporal biological variability of Indian
River lagoon. Florida Scientist. 53:249-256.
Vorosmarty, C.J. and T.C. Loder III. 1994. Spring-neap tidal contrasts and nutrient
dynamics in a marsh-dominated estuary. Estuaries 17:537-551.
Ward, D.H., C.J. Markon, and D.C. Douglas. 1997. Distribution and stability of
eelgrass beds at Izembek Lagoon, Alaska. Aquatic Botany 58:229-240.
Wardle, D.A., 0. Zackrison, G. Hornberg, and C. Gallet. 1997. The influence of island
area on ecosystem properties. Science 277:1296-1299.
Weathers, K.C., G.E. Likens, F.H. Bormann, S.H. Bicknell, B.T. Bormann, B.C. Daube,
Jr., J.S. Eaton, J.N. Galloway, W.C. Keene, K.D. Kimball, W.H. McDowell, T.G.
-------
207
Siccama, D. Smiley, and R.A. Tarrant. 1988. Cloudwater chemistry from ten
sites in North America. Environmental Science and Technology 22:1018-1026.
Weinberg, J.R. 1985. Factors regulating population dynamics of the marine bivalve
Gemma gemma: intraspecific competition and salinity. Marine Biology 86:173-
182.
Weinberg, J.R. 1989. Predicting population abundance and age structure: testing
theory with field data. Marine Ecology Progress Series 53:59-64.
Weins, J.A. 1989. Spatial scaling in ecology. Functional Ecology 3:385-397.
Weins, J.A., N.C. Stenseth, B. van Home, and R.A. Ims. 1993. Ecological
mechanisms and landscape ecology. Oikos 66:369-380.
Weisberg, S.B., J. Ananda Ranasinghe, D.M. Dauer, L.C. Schaffner, R.J. Diaz, and
J.B. Frithsen. 1997. An estuarine benthic index of biotic integrity (B-IBI) for
Chesapeake Bay. Estuaries 20:149-158.
Wenner, E.L., and H.R. Beatty. 1988. Macrobenthic communities from wetland
impoundments and adjacent open marsh habitats in South Carolina. Estuaries
11:29-44.
Westrheim, S.J. 1955. Size composition, growth, and seasonal abundance of juvenile
English sole (Parophrys vetulus) in Yaquina Bay. Oregon Fish Commission
Research Briefs. 6:4-9.
Wetzel, D.J. 1996. Brant use of Yaquina Estuary, Lincoln County, Oregon in the
Spring of 1976. Journal of Oregon Ornithology 6:715-722
Wetzel. R.L. and H.A. Neckles. 1986. A model of Zostera marina L. photosynthesis
and growth:simulated effects of selected physical-chemical variables and
biological interactions Aquatic Botany 26:307-323.
White, D., P.G. Minotti, M.J. Barczak, J.C. Sifneos, K.E. Freemark, M.V. Santelmann,
C.F. Steinitz, A.R. Kiester, and E.M. Preston. 1997. Assessing risks to
biodiversity from future landscape change. Conservation Biology 11:349-360.
Whitman, K.J. and J.A. Ott. 1982. Effect of cropping on growth in the Mediterranean
seagrass Posidonia oceanica L Delile. PSZN I: Marine Ecology 3:151-159.
Wiegers, J., W.G. Landis, L. Mortensen, and V.J. Wilson. 1997. The use of a habitat
based multiple stressor and ranking ecological risk assessment technique for
-------
208
prediction of environmental risk in Port Valdez, Alaska, in Abstract Book, #342,
Society of Environmental Toxicology and Chemistry, SETAC 18th Annual
Meeting.
Wiggins, W. D., G. P. Ruppert, R. R. Smith, L.L. Reed, L.E. Hubbard, and
M.L. Courts. 1995. Water Resources Data, Washington Water Year 1994, U.S.
Geological Survey Water-Data Report WA-94-1. 466 pp.
Wilkinson. 1996. Scale-dependent recolonization: the role of sediment stability in a
dynamic sandflat habitat. Ecology 77:2472-2487.
Williams, D.C. and J.G. Lyon. 1997. Historical aerial photographs and a geographic
information system (GIS) to determine effects of long-term water level
fluctuations on wetlands along the St. Marys River, Michigan, USA. Aquatic
Botany 58:363-378
Williams, S.F., and R.S. Caldwell. 1978. Growth, food conversion and survival of
O-group English sole (Parophrys vetulus - Girard) at five temperaures and five
rations. Aquaculture 15:129-139.
Williams, S.L., and J.B. Zedler. 1992. Restoring sustainable coastal ecosystems on
the Pacific Coast. Establishing a research agenda. Report. No. T-CSGCP-026.
California Sea Grant College, University of California, La Jolla, CA.
Wilson, R. 1996. Personal communication. Bay Center Maricutture, Bay Center, WA
Woods, C.M.C. and D.R. Shiel. 1997. Use of seagrass Zostera novazelandica
(Setchell, 1933) as a habitat and food by the crab Macrophthalmus hirtipes
(Heller, 1862) (Brachyura: Ocypodidae) on rocky intertidal platforms in southern
New Zealand. Journal of Experimental Marine Biology and Ecology 214:49-65.
Wright, P.J. 1991. The influence of metabolic rate on otolith increment width in
Atlantic salmon parr, Salmo salar L. Journal of Fisheries Biology 38:929-933.
Wright, P.J., N.B. Metcalfe, and J.E. Thorpe. 1990. Otolith and somatic growth rates in
Atlantic salmon parr, Salmo salar L.: evidence against coupling. Journal of
Fisheries Biology 36:241-245
Xiao, Y.S. 1996. How does somatic growth rate affect otolith size in fishes? Canadian
Journal and Fisheries Aquatic Scientist 53:1675-1682.
-------
209
Xue, Y., D.A. Kovacic, M.B. David, L.E. Gentry, R.L. Mulvaney and C.W. Lindau. (In
press) In situ measurements of denitrification in constructed wetlands. Journal
of Environmental Quality.
Yamaguchi, D.K., B.F. Atwater, D.E. Bunker, B E. Benson, M.S. Reid. 1997. Tree-ring
dating the 1700 Cascadia earthquake. Nature 389:922-923.
Young, D.K. and D.C. Rhoads. 1971. Animal-sediment relations in Cape Cod Bay,
MA. I. A transect study. Marine Biology 11:242-254.
Young, D.R. 1997. Quality Assurance Project Plan (QAPP) for aerial photographic
remote sensing of Yaquina estuary study. Unpublished report, US EPA National
Health and Environmental Effects Laboratory, Western Ecology Division,
Coastal Ecology Branch.
Young, D.R. 1998. Relationships between sediment substrate conditions and
macrophyte distributions in Pacific Northwest estuaries. ]n WED/CEB FY98
Research Plan. U.S. Environmental Protection Agency, Coastal Ecology
Branch, Newport, OR.
Zieman, J.C. 1974. Methods for the study of the growth and production of turtle grass,
Thallassia testudinum Konig. Aquaculture 4:139-143.
Zimmerman, R.J., and T.J. Minello. 1984. Densities of Penaeus aztecus, Penaeus
setiferus, and other natant macrofauna in a Texas salt marsh. Estuaries 7 421-
433.
Zimmerman, R.J., T.J. Minello, and G. Zamora, Jr. 1984. Selection of vegetated
habitat by brown shrimp, Penaeus aztecus, in a Galveston Bay salt marsh.
Fisheries Bulletin, U.S. 82:325-336.
Zimmermann, R.C., J.L. Reguzzone, S. Wyllie-Echeverria, E. Josselyn, R.S. Alberte.
1991. Assessment of environmental suitability for growth of Zosteria marina L
(eelgrass) in San Francisco Bay. Aquatic Botany 39:353-366.
Zimmerman, R.C., A. Cabello-Pasini, and R.S. Alberte. 1994. Modeling daily
production of aquatic macrophytes from irradiance measurements: a
comparative analysis. Marine Ecology Progress Series 114:185-196.
Zipperer, V.T. 1996. Ecological effects of the introduced cordgrass, Spartina
alterniflora, on the benthic community structure of Willapa Bay, WA. MS thesis,
University of Washington.
-------
210
8.0 Appendices
8.1 Appendix A. Technical Details of Experiments for Project A4
Experiment 1.B. Comparison of Acoustic, Photographic, and Videographic Remote-
Sensing Methods to Identify and Map Estuarine Submerged Aquatic Vegetation: The
acoustic remote sensing method will be "SAVEWS" (Submersed Aquatic Vegetation
Early Warning System), which is a 420-KHz hydroacoustic echolocation system used to
map Zostera marina and other rooted vegetation in shallow waters (Sabol and Melton
1995; Sabol et al. 1997). SAVEWS will be operated by Dr. Bruce Sabol (USACE
Waterways Experiment Station) who is collaborating on this project. If equipment is
available, we may also evaluate other acoustic remote sensing systems, such as
sidescan sonar and the RoxAnn system (Stenmar Group; Aberdeen, Scotland). The
videographic remote sensing method will consist of an underwater video camera, linked
to a GPS unit and depth sensor, which will be towed along a systematic series of
transects across the study sites. The videography will be provided by Dr. Brad Robbins
(CEB), a collaborator on this project. Hydroacoustic and videographic sampling will
occur along transects located 10-m apart that will traverse each patch at high tide. An
orthogonal set of hydroacoustic and videographic transects also will be made across
each site. Thus, hydroacoustic and videographic data will be collected in a grid of
10-m width on each site. Ideally, the hydroacoustic and videographic data will be
collected simultaneously, providing nearly 100% overlap of data. Both data sets will be
interpolated across each study site using an inverse weighted interpolation algorithm.
The photographic remote sensing method will be high-resolution, aerial infrared or
natural color photography (either 1:2000 or 1:3600 scale) which will be provided by Dr.
David Young (CEB), a collaborator on this project. These data will be digitized and
photointerpreted across the six study sites. To minimize any potential temporal bias,
aerial photographs will be taken at low tide on the earliest available date contemporary
with the acoustic, videographic and ground-truth sampling. The success of the aerial
photography is constrained by the availability of equipment and a pilot, and a
"photographic window" unobscured by clouds. For these reasons, the study will be
conducted in latesummer when the risk of cloudy weather is reduced. Ground-truth
data will be obtained by sampling the study sites using a grid on 1-m2 centers. Field
staff will traverse thirty-three 3m wide transects recording the presence/absence of
SAV by species as well as the presence/absence of burrowing shrimp (i.e. burrows) at
each grid point.
The videographic remote sensing method will consist of an underwater video
camera, linked to a GPS unit and depth sensor, which will be towed along a systematic
series of transects across the study sites. The videography will be provided by Dr.
Brad Robbins (CEB), a third collaborator on this project. Hydroacoustic and
videographic sampling will occur along 1-m-wide transects located 10-m apart that will
-------
211
traverse each patch at high tide. An orthogonal set of hydroacoustic and videographic
transects also will be made across each site. Thus, hydroacoustic and videographic
data will be collected in a grid of 10-m width on each site. Ideally, the hydroacoustic
and videographic data will be collected simultaneously, providing nearly 100% overlap
of data. Both data sets will be interpolated across each study site using an inverse
weighted interpolation algorithm.
The photographic remote sensing method will be high-resolution, aerial infrared
or natural color photography (either 1:2000 or 1:3600 scale) which will be provided by
Dr. David Young (CEB) who is also collaborating on this project. These data will be
digitized and photointerpreted across the six study sites. To minimize any potential
temporal bias, aerial photographs will be taken at low tide on the earliest available date
contemporary with the acoustic, videographic and ground-truth sampling. The success
of the aerial photography is constrained by the availability of equipment and a pilot, and
a "photographic window" unobscured by clouds. For these reasons, the study will be
conducted in July, August, or September when the risk of poor weather is reduced.
Ground-truth data will be obtained by sampling the study sites using a grid on 1-
m2 centers. Field staff will traverse (33) 3m wide transects recording the
presence/absence of SAV by species as well as the presence/absence of burrowing
shrimp (i.e. burrows) at each grid point.
Experiment 1.C. Evaluation of Remote Sensing Methods to Detect, Identify, and
Quantify Burrowing Shrimp: Acoustic remote sensing methods will include a variety of
echo-sounder methods (using different transducer frequencies, electronics, or software
to sample or interpret signals) and possibly side-scan sonar. Interest in participating in
this project has been expressed by Drs. Whitlow Au (Univ. of Hawaii), Nicholas
Chotiros (Univ. Texas), Darreli Jackson (Univ. Washington), Bruce Sobal (USACE
Waterways Experiment Station) and sonar development companies, such as BioSonics
(Seattle, WA) and Marine Sonics (White Marsh, VA). Underwater videography will
consist of a GPS-coupled video system towed behind a boat, and will be provided by
Dr. Brad Robbins (CEB) who is collaborating on this project. Each video frame will
image approximately 1-m2 of the seafloor. Post-processing involves a trained individual
classifying and quantifying objects on each video frame.
production (Table A.1). While processing the Zostera shoots, the number of leaf scars
will be counted and recorded to reconstruct plant demography and leaf and rhizome
production (Duarte et al. 1994) (Table A.1). The technique developed by Duarte et al.
assumes that the PI is a continuous, linear process. Kentula and Mclntire (1986)
reported that PI changes throughout the growing season for
Prior to sampling for burrowing shrimp at the test sites (described below) each
investigator will be allowed to spend at least one day testing their system over
burrowing shrimp beds in Yaquina estuary. This preliminary trial period will provide an
-------
212
opportunity to "fine tune" the systems so that they have the best opportunity to be
successful in the experiment.
Each remote sensing method will sample the same patches of intertidal
sediments (study sites), some of which will be used to calibrate the signal, and the
others to test the accuracy of the signal interpretation (i.e., habitat classification). All
sites will be located in the intertidal zone of the central portion of Yaquina estuary in
unconsolidated sediments inhabited by Neotrypea or Upogebia. Field staff will pre-
classify sites at low tide by identifying the shrimp (i.e., by the configuration of the
burrow openings; qualitative core samples can also be collected to verify the species
identity of shrimp) and by measuring the density of burrow openings (i.e., number of
burrow holes per 0.5-m2). If uninhabited sediment patches cannot be located within or
adjacent to burrowing shrimp habitats, 2-m x 2-m patches of shrimp beds will be
defaunated by smothering as per Thrush et al. (1996). Each site will be marked with a
numbered float to guide the boat pilot there at high tide.
Minimally, eight sites of each type will be identified; three sites will be used for
calibration of the classification method, and the remaining sites will be used for testing
the accuracy of the classification. Ideally a higher replication rate (such as 10 per
patch-type) will be used, but that will be determined by the logistics of ground-truth
sampling, the time required to sample sites acoustically, and the time to travel between
sites. All crew on-board the boat from which sampling is conducted will be "blind" to
the identity of each site. Since the acoustic and videographic methods can only be
used at high tide, concealing the identity of the sites will not be difficult. The boat pilot
will maneuver and anchor the boat such that the acoustic transponder or video camera
is positioned directly over a site, and the investigator will be allowed to sample the site
for a fixed of period time (nominally, two minutes).
Experiment 2.A. Population Characteristics ofEelgrass and Burrowing Shrimp Along
the Dominant Salinity-Temperature Gradient in Yaquina Estuary, Oregon: This study is
designed to provide a baseline of eelgrass and burrowing shrimp population dynamics,
and to evaluate population-level response variables that may be used for subsequent
work. Despite the fact that Zostera marina L. (eelgrass) is the most common and
abundant seagrass in the northern temperate Pacific (Phillips 1984), information on the
distribution and growth of eelgrass is limited, especially for Oregon. The distribution of
Zostera in Oregon was reported by Proctor et al. (1980) and aspects of its growth were
reported by Bayer (1979) for Yaquina Bay and by Stout (1976) and Kentula and
Mclntire (1986) for Netarts Bay. Although there is a relatively detailed database on the
autecology of Zostera for Netarts Bay, the ability to extrapolate that information to other
estuaries in Oregon may be limited. Netarts does not have a major river flowing into it,
and, therefore, does not have freshwater inputs and the related temperature and
salinity gradients similar to other Oregon estuaries. Although low salinity limits the up-
river distribution of eelgrass world-wide, very little is known about changes in the
-------
213
autecology and demography of Z marina along salinity-temperature gradients of
Pacific Northwest estuaries (Phillips 1984, Thorn 1990). Bird (1982), Posey (1986a,
1987) and Dumbauld et al. (1996) conducted limited population studies on Neotrypea
in the mesohaline portions of various Pacific Northwest estuaries (including sites in
Yaquina estuary studied by Bird). Dumbauld (1994) also studied the population biology
of mud shrimp, Upogebia, in Willapa Bay, Washington, but it is uncertain whether the
information from that study can be extrapolated to Oregon estuaries. Posey (1987)
demonstrated that low salinity can be lethal to both Neotrypea and Upogebia, but no
study has examined whether the population structure of either species changes along
the natural, salinity-temperature stressor gradient common to Pacific Northwest
estuaries.
Intensive Sampling Approach: Intensive-sampling study sites will be sampled monthly
during day-time low tides year-round for burrowing shrimp and from May through
September for Zostera. Sampling will be limited to times when the sites are exposed
long enough to be sampled. From October through April, the sites will be visited as
weather and the tides permit to determine what sampling can be done for Zostera. At
minimum, descriptive notes and pictures of the sites will be taken any time a site is
visited. Physical and chemical data will be collected in conjunction with the biophysical
model-testing study (Expt. 2.D.) described below. In addition, we will attempt to
measure the tidal elevation of the places where samples are taken within each site in
conjunction with the bathymetric mapping work proposed by Dr. David Young (CEB).
Ten permanent transects will be equally spaced across each site and will run
parallel to the tidal gradient, i.e., perpendicular to the shore (Figure A.1). The first and
tenth transects will be located 5 m from the edge of the Zostera bed. A grid of
permanent transects will be used as the sample design to facilitate finding marked
shoots of Zostera on the sample date following marking. (See below for details on
shoot marking.) Burrowing shrimp populations will be sampled along these transects
also.
Zostera will be sampled along each transect approximately 1m from the upper
and lower edge of the bed, i.e., two sample points per transect (Figure A.1). Kentula
and Mclntire (1986) found that Zostera had distinct characteristics at the upper and
lower limits of its distribution intertidally in Netarts. At each sample point, a 1-m2
quadrat will be placed perpendicular to the transect and percent cover will be
measured using both visual estimate and point count methods as was done by CEB
scientists in the 1997 aerial photography study (Table A.1) (Young 1997). At the
sample points on transects 2, 4, 6, 8 and 10, a circular quadrat of comparable size to
that being used for the shrimp sampling will be placed in the lower left-hand quarter of
the 1-m2 quadrat (Figure A.2) and the number of vegetative shoots, number of flowering
shoots, and number of seedlings counted (Table A.1). These measures will provide
information on standing crop and sexual reproduction (timing, proportion of the shoots
-------
214
flowering, proportion of shoot growing from seed). To measure eelgrass growth
dynamics and net primary production, three vegetative shoots typical of the plot will be
chosen from the lower right-hand quarter of the 1-m2 quadrat (Figure A.2) and marked
according to methods developed by Kentula and Mclntire (1986). Approximately four
weeks later during the next tide suitable for sampling, one of the three marked shoots
from each sample point will be harvested. Three shoots are marked to assure that at
least one shoot will be found and in good condition, e.g., leaves intact. The rhizome for
each shoot will be clipped from the shoot(s) growing in front of it and at the point where
it originated (i.e., where either where the rhizome branches or another shoot is
growing). This procedure will harvest enough of the rhizome to test the application of
the methods of Duarte et al. (1994) which utilizes the patterns of leaf scars on seagrass
rhizomes to estimate historical seagrass productivity from a one-time destructive
sample. Shoots from each plot will be placed in a labeled plastic bag, transported on
ice to the laboratory where they will be stored at 4°C until processing (within 96h).
Eelgrass leaf-area per shoot will be determined by measuring the width and
length of each leaf. The methods of Kentula and Mclntire (1986) will be used to
determine plastochrone interval (PI) (the time interval between initiation of successive
leaves on a shoot) (Patriquin 1973), export interval (El) (the time between sloughing of
two successive leaves on a shoot) (Kentula and Mclntire 1986), and net primary
production (Table A.1). While processing the Zosterai shoots, the number of leaf scars
will be counted and recorded to reconstruct plant demography and leaf and rhizome
production (Duarte et al. 1994) (Table A.1). The technique developed by Duarte et al.
assumes that the PI is a continuous, linear process. Kentula and Mclntire (1986)
reported that PO changes throughout the growing season for Zostera in Oregon.
Therefore, as recommended by Jensen et al. (1996), we will use the leaf marking data
from the intensive sites in conjunction with the extensive sampling event in July
(described below) to empirically establish the PI that will be used to generate Zostera
demographic information (Duarte et al. 1994).
-------
Figure A.l. Diagram of transect and quadrat placement to be used at each of the intensive study sites.
Diagram is not to scale.
Neotrypeabed
Upogebia bed
k ®—"k ®—k (g)
10 m
100m transect
channel
ft -Comtshrinpholes * -Mark shoots, estimate oover
-Take shrimp cores
- Mark shoots estimate ccwer,
count shoots
-------
216
Table A.1. List of Zostera population variables, sample units, and sample sizes per
intensive-study site per month. Zone refers to the upper and lower intertidal limits of
the Zostera bed on the site, Three intensive-study sites will be sampled.
Variable
Sampled
Sampl
e Unit
Number of
Samples
per Zone
Number of
Samples per
Transect
Number of
Samples
per Site
% Cover
1-m2
quadrat
10
2
20
Density vegetative shoots
circular
quadrat
5
2
10
Density reproductive
shoots
circular
quadrat
5
2
10
Density of seedlings
circular
quadrat
5
2
10
Density of eelgrass
circular
quadrat
5
2
10
Leaf area
short
shoot
10
2
20
PIastochrone interval
short
shoot
10
2
20
Export interval
short
shoot
10
2
20
Net primary production
(above ground)
short
shoot
10
2
20
Shoot demography (Duarte
et al. 1994)
short
shoot
10
2
20
Leaf production (Duarte et
al. 1994)
short
shoot
10
2
20
Rhizome growth (Duarte et
al. 1994)
short
shoot
10
2
20
-------
Figure A.2. Diagram of placement of the 1-m2 quadrat and the samples to be taken
within it.
Marked shoots
x
\
x
X
Circular quadrat
-1-m2
quadrat
t
i
-------
218
Neotrypea and Upogebia will be sampled along each of the 10 transects of the
study grid. The center of the spatial distribution of each shrimp will be located by visual
inspection of the distribution of burrow holes along each transect. Neotrypea burrows
are readily identified by the cone-shaped deposit of sediment around the opening,
whereas Upogebia burrows are identifiable by their larger diameter opening, lack of a
sediment cone, and mucus lining (identified by touch). Burrow-hole density data (i.e.,
the number of burrow openings per m2) will be obtained for both shrimp species on
every transect using a 0.5-m2 quadrat (Table A.2). These data will provide information
on the spatial and temporal variability of shrimp density across each site. Mega-
infaunal core samples and young-of-the-year (YOY) core samples will be collected
along five transects and will be processed to provide data on the absolute density, size
distribution, sex ratio, reproductive condition, and recruitment of each shrimp. Cores
will be collected along every-other transect each month (i.e., even-numbered transects
one month, odd-numbered transects the next month), and the location of core sites will
alternate between the left and right side of the transect to avoid re-sampling the same
patch of sediment. Along one transect, four mega-infauna cores will be collected using
a 15-cm to 20-cm diameter x 1-m depth corer (method to be determined by Expt. I.A),
sieved through a 3-mm mesh screen, and combined to form one sample; the material
collected on the screen will be preserved in buffered formalin until picking, and each
shrimp will be measured for carapace length, its gender determined, and reproductive
condition of females assessed by the presence and developmental stage of eggs (as
per Dumbauld 1994). Adjacent to the mega-infauna cores, four YOY core samples will
be collected using a 15-cm diameter x 10-cm deep corer, sieved through a 0.5-mm
mesh, and combined into one sample; the material collected on the screen will be
preserved in buffered formalin containing rose bengal stain, then picked for small
burrowing shrimp, which will be identified to species and counted. Thus, each month
we will collect ten burrow-hole count samples, five mega-infauna core samples, and
five YOY core samples at each site.
Extensive-Sampling Approach: Sample locations for the extensive sampling effort will
be the same as those selected by Steve Ferraro and Faith Cole (CEB) within eelgrass,
ghost shrimp, and mud shrimp habitats throughout Yaquina estuary. Ten sites within
each habitat will be selected by a random-site generator which weights the distribution
of sites by the spatial distribution of the habitats: more sites are located where the
greatest proportion of habitat is present. This is the same random-sampling strategy
employed by EMAP (USEPA 1992). We will use the mega-infauna core samples
collected for Ferraro and Cole (using the same protocol described above) to quantify
the demographic characteristics of Neotrypea and Upogebia at each site. Burrowing
shrimp hole-count data also will be collected at each site. At sites within Zostera beds,
four 15-cm diameter x 10-cm deep cores will be taken to collect samples of eelgrass.
In addition, percent-cover of eelgrass will be measured by the point-intercept method
and by visually estimating percent cover by species using 1-m2 quadrats. To the extent
possible, the Zostera core samples will be collected in July when
-------
219
Table A.2. List of burrowing shrimp population variables, sample units, and sample
sizes per intensive-study site per month. Three intensive-study sites will be sampled.
Variable
. Sampled
Sample Unit
Number of
Samples per
Transect
Number of
Samples
per Site
Neotrypea burrow-hole
density
0.5-m2 quadrat
1
10
Neotrypea abundance
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Neotrypea size-frequency
distribution
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Neotrypea sex ratio
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Neotrypea female
reproductive condition
mega-infaunal core
sample*
0.5(1 per 2
transects)
5
Neotrypea young-of-the -
year abundance
(recruitment)
composite of 25 5-cm
diameter cores
0.5 (1 per 2
transects)
5
Upogebia burrow-hole
density
0.5-m2 quadrat
1
10
Upogebia abundance
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Upogebia size-frequency
distribution
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Upogebia sex ratio
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Upogebia female
reproductive condition
mega-infaunal core
sample*
0.5 (1 per 2
transects)
5
Upogebia young-of-the -
year abundance
(recruitment)
composite of 25 5-cm
diameter cores
0.5 (1 per 2
transects)
5
* coring method for burrowing shrimp will be determined in preliminary studies (Expt.
1.A)
-------
220
Table A.3. Population variables and sample units for eelgrass and burrowing shrimp to
be collected in July at each extensive-sampling study site. One sample will be
collected per site. The samples will be collected from the 10 randomly selected sites
used by Ferraro and Cole and the 3 sites used in the intensive-sampling study (i.e., a
total of 13 sites).
Variable Sampled
Sample Unit
Zostera % cover
1-m2 quadrat
Zostera: density vegetative shoots
circular quadrat
Zostera: density reproductive shoots
circular quadrat
Zostera: density of seedlings
circular quadrat
Zostera: total density
circular quadrat
Zostera: above ground biomass
circular quadrat
Zostera: below ground biomass
circular quadrat
Zostera: total biomass
circular quadrat
Zostera shoot demography {Duarte et al.
1994)
circular quadrat
Zostera leaf production (Duarte et al. 1994)
circular quadrat
Zostera rhizome growth (Duarte et al. 1994)
circular quadrat
Neotrypea or Upogebia burrow-hole density
0.5-m2 quadrat
Neotrypea or Upogebia abundance
mega-infaunal core sample*
Neotrypea or Upogebia size-frequency
distribution
mega-infaunal core sample*
Neotrypea or Upogebia sex ratio
mega-infaunal core sample*
Neotrypea or Upogebiafemale reproductive
condition
mega-infaunal core sample*
Neotrypea or Upogebia young-of-the -year
abundance (recruitment)
composite of 25 5-cm diameter
cores
* coring method for burrowing shrimp will be determined in preliminary studies (Expt.
1.A)
-------
eelgrass primary productivity is expected to be maximum (Kentula and Mclntire 1986).
The sampling methods and intensities for each population variable are summarized in
Tables A.2 and A 3.
In addition to the samples taken from the cores harvested by Ferraro and Cole,
four 15-cm diameter x 10-cm deep cores will also be collected in July (i.e., when
Ferraro and Cole sample their sites) in the upper and lower intertidal limits of the
Zostera at each of the three intensive sites. The Zostera and associated sediment
within the circular quadrat will be collected into a 3-mm mesh sieve. The sediment will
be washed from the sieve and the plants placed into labeled plastic bags, transported
on ice to the laboratory and refrigerated until processed (two weeks maximum).
Eelgrass core samples will be processed to determine number of vegetative and
flowering shoots and seedlings; and above and below ground biomass (Kentula and
Mclntire 1986). While sorting the shoots to determine biomass, information needed to
determine Zostera demographics will be recorded, including number of leaf scars,
attached and unattached shoots, and number of leaves per shoot (Duarte et al. 1994,
Jensen et al. 1996). As stated above, these results will be compared with those
obtained from the intensive sample to determine the best way to measure Zostera,
Neotrypea, and Upogenia demographics in Oregon.
Experiment 2.B. Development and Testing of Biophysical Population Models: This
effort will be conducted in collaboration with Dr. Bob Ozretich (CEB) and in consultation
(and possibly collaboration) with other researchers studying eelgrass and burrowing
shrimp ecology (such as, Drs. Mary Kentula [WED], Ron Thorn [Battelle Marine
Sciences Laboratory], Brett Dumbauld [Washington Department of Fisheries], and
David Armstrong [Univ. Washington]). Dr. Ozretich will be testing a water quality-
based model to predict the distribution of eelgrass, and we plan on combining our
resources by selecting the same sampling sites and using the same water column data
and eelgrass presence-absence data. Bob Ozretich and I plan to combine the best
aspects of our two models in the future and to test the accuracy of the new model
across Pacific Northwest estuaries. Environmental variables to be measured are listed
in Table A.4.
One unresolved issue is how best to represent the temporal variability in water
column parameters at each site. Clearly, water column parameters are likely be much
more variable temporally (on tidal and seasonal time scales) than sediment-associated
variables. Single samples at high tide collected once per season will not be sufficient
to adequately characterize this variability. One approach is to report the extremes of
the 95% confidence interval for each parameter at each site as the maximum and
minimum values; however, we will consult with other estuarine scientists to identify
other ways to integrate these temporally-variable data. We will also investigate linking
-------
222
measurements made at the sites to 1) water quality data collected continuously by
moored CTD/PAR units (see proposal by David Specht, CEB) using regression
techniques, and 2) to a circulation model for the Yaquina estuary, such as the 2-
dimensional model used by the NOAA/PMEL Tsumami Project (Kamphaus 1998).
These approaches could lead to allowing us the capability to accurately model
temporal variability of water column parameters at each of the study sites.
At low tide, we will visit each site to sample eelgrass and burrowing shrimp (i.e.,
presence or absence within a 4-m2 area) and to measure (or sample for) benthic habitat
variables. The latter include porewater salinity (measured by a hand-held
refractometer), sediment temperature (measured by a digital thermometer), and slope
(measured using a 2-m long level and protractor or inclinometer). Bathymetric
elevation will be obtained from the digital elevation map being developed by Dr. David
Young (CEB). Water depth also will be recorded each time water quality variables are
measured (i.e., three times per year), and bathymetric elevation for each site can be
estimated by averaging the tide-corrected depth measurements. Sediment grain size
(percent gravel, sand, silt-clay) and sediment total organic carbon (TOC) will be
measured in the laboratory following current CEB SOPs.
Future Model and Experiment: The biophysical population model will be combined with
the habitat suitability model being developed and tested by Bob Ozretich (CEB), and
the new model will be tested in Yaquina estuary and one or more other Pacific
Northwest estuaries. It is premature to speculate on the design of that study, but
certain features are foreseen. First, it will be advantageous to conduct the study in
estuaries in which other ecologists are studying the distribution and population biology
of eelgrass and burrowing shrimp, and in which capability exists to measure water
column parameters. Second, it will be logistically advantageous to form collaborations
with the ecologists studying eelgrass and burrowing shrimp in those estuaries so that
their staff may participate in the sampling effort. Candidate estuaries and researchers
include: Tillamook Bay, OR (Roxanna Hinzman, Tillamook Bay National Estuary
Project), Coos Bay, OR (Steve Rumrill, South Slough National Estuarine Research
Reserve), Willapa Bay, WA (Brett Dumbauld, Washington Dept. of Fisheries; Si
Simenstad or David Armstrong, Univ. Washington), and Padilla Bay, WA (Doug
Bulthuis, Padilla Bay National Estuarine Research Reserve). Third, it will be important
to include Yaquina estuary in the study in order to test whether the new model provides
the same accuracy as its predecessor.
Experiment 2.C. Effects of Multiple Abiotic Stressors on the Population Biology of
Keystone Species - Methods for Mesocosm Experiments: Using data from the
mesocosm competition experiments (expt. 3.A. and 3,B.), power analysis will be
employed to determine the replication rate that will allow us to measure changes in the
responses in critical population variables, such as %cover, abundance, and growth or
productivity. The levels and replication rates listed here are presented only to illustrate
-------
possible experimental designs.
-------
Table A.4. Environmental parameters to be measured for the biophysical population
models.
¥sMranie{erV;V;v: ^
Analytical Method
Sampled Medium
light attenuate coefficient
photosynthetic active
radiation (PAR) 2
depths/station
water column
overlying water salinity
^(conductivity) ,
CTD
water column
water temperature
CTD
water column
depth
CTD
water column
total suspended solids
mass collection on
preweighed membrane filters
water column
dissolved inorganic nitrogen
AutoAnalyzer
water column
dissolved inorganic,
phosphorus
AutoAnalyzer
water column
Zostera or burrowing shrimp
presence or absence
sediment
porewaier salinity
refractometer
sediment
sediment temperature
hand-held digital
thermometer
sediment
sediment-surface slope
inclinometer
sediment
sediment grain size (% sand/
silt, day) i1 ,
sieve method
sediment
sediment organic carbon and
nitrogen •• >.
TOC/N analyzer
sediment
Sathymetnc elevation
Digital elevation map or
depth-sounder
measurements
water column and
sediment
Fiberglass or PVC cylinders (30.5 cm diameter x 91 cm height [= 12" x 36"]) will
be filled with 76 cm (= 30") of fine-grained sediment, sieved to <1-mm, obtained from
-------
225
intertidal habitats adjacent to eelgrass and burrowing shrimp beds. Each cylinder will
be placed within one of the 61 cm x 61 cm (= 2' x 2') compartments of a holding tank
(Figure B1). Zostera will be planted uniformly across the sediment surface of each
cylinder at densities of-200 shoots m"2 (i.e., -15 shoots per cylinder; this is a
moderately high shoot density for Pacific Northwest populations [Phillips 1984]) and
allowed to establish rooting during an acclimation period (duration of the acclimation
period will be set based on advice from experts such as Drs. Fred Short [Univ. New
Hampshire] and Ron Thorn [Battelle Marine Science Laboratory]). Eelgrass will be
harvested from Yaquina estuary or nearby estuaries, transported damp and shaded in
coolers to the EPA laboratory, held in running seawater at ambient salinity and
temperature at the laboratory, and planted within 48 hr of harvest. Salinity, light, or
temperature will be changed gradually to allow acclimation to the experimental
conditions. Similarly, burrowing shrimp will be harvested at least 1 week prior to the
start of the experiment and acclimated to experimental conditions in seawater.
Neotrypea and Upogebia will be added to the cylinders at densities of -100 m'2, which
are moderately high densities for these species (Dumbauld 1994; unpubl. data).
Shrimp used in the experiments will be sorted to uniform size, and sub-samples of
shrimp will be measured for initial carapace length, wet weight, and dry weight.
The mesocosm experiments will be conducted for 6 to 20 weeks. The duration
will depend on whether effects can be observed during the experiment; if effects are not
apparent during non-destructive sampling, the experiment will be terminated after the
longer period. Eelgrass or burrowing shrimp in the cylinders will be sampled non-
destructive^ every two weeks and destructively at the end of the experiment.
Population variables measured non-destructively for eelgrass will include percent
cover, density of vegetative and reproductive shoots, and growth as indicated by the
change in leaf surface area (for the longest leaf in each of 5 shoots) (Table 2). Four
weeks before the end of the experiment, 5 shoots per cylinder will be marked for
above-ground net primary productivity measurements using methods developed by
Kentula and Mclntire (1986). At the end of the experiment, eelgrass will be sampled
destructively for above-ground net primary productivity, above and below ground
biomass, plastichrone interval, and export interval (see expt. 2.A.). Non-destructively
sampled population variables measured for Neotrypea and Upogebia include the
density of burrow openings within a 20-cm diameter quadrat and the maximum height of
the sediment-water interface; the latter will be measured as the distance from the rim of
the cylinder to the top of the highest mound of sediment produced by the burrowing
shrimp. At the end of the experiment, the sediment will be pumped out of each cylinder
(using a venturi aspirator or a suction sampler such described by Miles and Whitlatch,
1997) and washed through a 3-mm mesh screen to capture the shrimp. Destructively-
collected variables for the shrimp will include total abundance of shrimp, carapace
length, sex and reproductive condition, and wet and dry weights (Table A.5).
Additional variables that could be studied in these factorial-design, multiple-
-------
226
stressor experiments include nutrient concentration (eelgrass), substrate-type,
sedimentation, dissolved oxygen concentration, or persistent chemical contaminants.
Selection of the variables to be used will be guided by the results of testing the
biophysical models for each species (expt. 2.B.). Four-way multiple stressor-response
experiments could be conducted also, but using central-composite experimental
designs rather than fully-factorial experimental designs. Multi-factorial (i.e. 3-way or 4-
way) analysis of variance will be used to analyze the data to test the hypotheses that 1)
no significant interaction exists in the effects caused by multiple stressors (for all
possible combinations of the stressors tested), 2) that individual stressors have no
significant effect on the survival or growth of the test organism, and 3) that each
stressor has the same magnitude of effect on the test organism.
Field Experiments: Eelgrass or shrimp will be harvested from nearby populations,
sorted to uniform size and condition, and "planted" inside caged patches (40-cm
diameter) at experimental sites. The patches will be created by excavating a 40-cm
diameter x 60-cm depth core of sediment, lining the hole with 3-mm mesh plastic
screen (i.e., underground fence) to a depth of 60 cm, and refilling the hole with the
excavated sediment which will be sieved to <3-mm to remove burrowing shrimp and
eelgrass. Population densities of each species will be similar to those used in the
mesocosm experiments. A cage will be placed over each patch to prevent burrowing
shrimp emigration or immigration, predator immigration, and to provide protection from
scouring. After an acclimation period of a few days, the above-ground cage will be
removed from all plots, but the below-ground fence will be retained to prevent
burrowing shrimp from leaving or entering the plot by lateral burrowing.
The population-level response variables will be similar or identical to those used
in previous mesocosm and field experiments (Table A.5). Non-destructive samples will
be measured approximately every 4 weeks (depending on site accessibility due to tides
and weather). The field experiment will be conducted for 6 to 20 weeks, depending on
the rate at which responses are observed in mesocosm experiments and on when such
responses are measurable in the field. Data from the experiments will be analyzed by
analysis of variance to test the same hypotheses as in the analogous mesocosm
experiment: 1) no significant interactions in effects of co-occurring stressors, 2) no
effect of each stressor alone, and 3) all stressors have the same magnitude of effects.
Beyond this, it is premature to design these experiments in further detail as they are not
likely to be started until the above mesocosm and competition experiments are
concluded (i.e., not until sometime in 2000).
Experiment 2.D. Mapping Changes in Eelgrass and Burrowing Shrimp Populations
Along Stressor Gradients: Several issues will need to be addressed prior to testing
hypotheses about population spatial changes in relation to stress. First, we need to
know what spatial scale to map for each species. The mapping accuracy study (expt.
1.B.) will examine how accuracy changes with spatial scale for each of three remote-
-------
227
sensing methods. Results from that study will be balanced against the resolution of
available data (to be acquired from other researchers or public agencies) and the cost
of obtaining high-resolution spatial data. Second, we need to know the temporal scale
at which to map population change. We anticipate that annual
-------
228
Table A.5. List of population variables, sample units, and sampling frequencies for
eelgrass and burrowing shrimp that may be collected in mesocosm and field stress-
response experiments. One sample will be collected per replicate experimental
chamber (cylinder in mesocosm, plot in field experiments).
Sampling
Frequency
VclllalJIc gallipivU
UIIJl
Zostera % cover
20 cm diam. quadrat
every 2-4 weeks*
Zostera: density of vegetative
shoots
20-cm diam. quadrat
every 2-4 weeks*
Zostera: density reproductive shoots
20-cm diam. quadrat
every 2-4 weeks*
Zostera: total shoot density
20-cm diam. quadrat
every 2-4 weeks*
Zostera leaf area
5 shoots
every 2-4 weeks*
Zostera above-ground biomass
20-cm diam. quadrat
start and end of
expt.
Zostera below-ground biomass
20-cm diam. quadrat
start and end of
expt.
Zostera total biomass
20-cm diam. quadrat
start and end of
expt.
Zostera plastochrone interval
20-cm diam. quadrat
start and end of
expt.
Zostera export interval
20-cm diam. quadrat
start and end of
expt.
Zostera net primary productivity
(above ground)
5 shoots
start and end of
expt.
Neotrypea or Upogebia burrow-hole
density
20-cm diam. quadrat
every 2-4 weeks*
Neotrypea or Upogebia maximum
height of sediment deposit
(mesocosm only)
1 mound per
replicate
every 2-4 weeks*
Neotrypea or Upogebia abundance
all shrimp per
replicate
start and end of
expt.
-------
229
Variable Sampled
Sample Unit
Sampling
Frequency
Neotrypea or Upogebia size-
frequency distribution
all shrimp per
replicate
start and end of
expt.
Neotrypea or Upogebia wet and dry
wt.
all shrimp per
replicate
start and end of
expt.
Neotrypea or Upogebia sex ratio
all shrimp per
replicate
start and end of
expt.
Neotrypea or Upogebia female
reproductive condition
all shrimp per
replicate
start and end of
expt.
* variables sampled every 2 wk in mesocosms, and approximately every 4 in the field
(depending on accessibility to study sites due to tides and weather).
or biennial maps may be adequate for these purposes, but also recognize that eelgrass
and burrowing shrimp distributions can change dramatically seasonally. One approach
to resolving this question would be to generate population change maps for select sites
on a seasonal basis for 2+ yr, measure the rate of population change over different
temporal scales, and select the scale at which the rate of change reaches an
asymptote. Third, we need to know whether natural rates of population change vary
geographically across the Pacific Northwest. It seems likely that latitudinal gradients in
day length, temperature, rainfall, or storm intensity could affect natural rates of
population expansion and contraction. Underlying natural environmental gradients will
have to be identified and their influence measured before we can confidently use
population change maps to detect and measure the effects of anthropogenic stress.
Experiment 3. A. Measuring Interspecific Competition Among Eelgrass and Burrowing
Shrimp: Methods for Mesocosm Experiments: The experiments will be very similar to
those used for the multiple stressor experiment (Expt. 2.C). Zostera will be planted
uniformly across the sediment surface of each cylinder at densities of 0 to 400 shoots
m"2 (i.e., natural densities for the Pacific Northwest; Phillips 1984) and allowed to
establish rooting during an acclimation period (this acclimation period will be set based
on advice from experts such as Drs. Fred Short [Univ. New Hampshire] and Ron Thorn
[Battelle Marine Science Laboratory]). Burrowing shrimp will be harvested at least 1
week prior to the start of the experiment and acclimated to experimental conditions in
seawater. Neotrypea, harvested and sorted to size ~1 wk prior to the experiment, will
be added to the cylinders at densities ranging from 0 to 450 m'2, whereas Upogebia will
be added at densities ranging from 0 to 300 m*. The higher densities are equivalent to
those measured recently in Yaquina estuary (unpubl. data). Sub-samples of shrimp will
-------
230
be measured for initial carapace length, wet weight, and dry weight.
The mesocosm competition experiments will be conducted for 8 to 16 weeks.
The duration will depend on whether effects can be observed during the experiment; if
effects are not apparent during non-destructive sampling, the experiment will be
terminated after the longer period. Population variables for both species in each
cylinder will be sampled non-destructively every two weeks and destructively at the end
of the experiment, and will be the same as those described for the stressor-response
experiments (Table A.5).
Field Experiment: The patches will be created by excavating a 40-cm diameter x 60-cm
depth core of sediment, lining the hole with 3-mm mesh plastic screen "underground
fence" to a depth of 60 cm, and refilling the hole with the excavated sediment which will
be sieved to <3-mm to remove burrowing shrimp and eelgrass. The plots will covered
with a 3-mm mesh cage to prevent immigration or emigration of conspecifics or
competitors. After an acclimation period of a few days, all cages will be removed, and
underground fences will be removed from some plots but retained in others. The
replication rate will be based on results of a power analysis of key response variables
(i.e., %cover and net primary productivity for Zostera, abundance for the shrimp) using
variance estimates from field populations (i.e., Kentula and Mclntire 1986, Dumbauld et
al. 1996). The duration of the experiments could be extended for as long as 1-yr if
there is no apparent effect of competition; in this case, a bi-monthly sampling frequency
will be used. The long duration may be required for competition to manifest itself,
although we expect the effects to be revealed within weeks, particularly between
Neotrypea and Zostera because of the high rate of sediment turnover caused by ghost
shrimp.
-------
8.2 Appendix B Technical Details of Model Development for Project B1
-------
232
Appendix Table 1: Model Parameters. The "Range in Values" is an approximation of the range in values that may be
used in simulations without the addition of any stochastic variation. Note that the stressor-response parameters and the
vital rates calculated from these values are listed as a function of age ("age-specific") even though at this stage we
assume that all adult classes 1 year) have the same sensitivity (see text and Table 2).
PARAMETER
DEFINITION
UNITS
POTENTIAL
RANGE
HOW IT WILL BE
VARIED
a,M
age specific parameter determining level of
anthropogenio-related increased mortality in logistic
function
Dimensionless
-7 +7
Deterministic
age specific parameter determining level of
anthropogenic-related reduction in fecundity in
logistic function
Dimensionless
-7 +7
Deterministic
4(x)
age specific parameter determining slope of
increased anthropogenic-related mortality in logistic
function
Dimensionless
1 16
Deterministic
A(x)
age specific parameter determining slope of reduced
anthropogenic-related fecundity in logistic function
Dimensionless
1 "+ 16
Deterministic
bs
lapse rate for stress function
Dimensionless
.0187 .251
Deterministic
c
cell number
Dimensionless
0 "*31,416
Definition
d
distance from center of domain to a stress source
Dimensionless
o 100
Deterministic /
Stochastic
dq(x,c)
anthropogenic-related increase in mortality for age x,
cell c, time t
Dimensionless
0 1
Calculated
dm(x,c)
anthropogenic-related relative decrease in fecundity
for age x, cell c, time t
Dimensionless
0 1
Calculated
-------
233
eggs(c.t)
total eggs produced by all females in cell c at time t
eggs
0 2.84 10®
Calculated
m„(x)
natural fecundity for age x
eggs / time
0 76485
Stochastic
m(x,c)
total fecundity for age x.cell c,
(= m„(x) r 1 - dm(x,c) 1)
eggs / time
0 76485
Calculated
n(x,c,t)
number of individuals of age x in cell c at time t
individuals
o -+ 100
Calculated
no-2(c,t) •
number of individuals in 0-2 month class in cell c at
time t
individuals
0 "* 10,000
Calculated
number of individuals in 3-12 month class in cell c at
time t
individuals
0 41.5
Calculated
nc(r)
number of cells at distance r from a stress source
cells
1 -+ 628
Calculated
n cells
total number of cells in the domain
cells
31,416
Definition
nst
number of stress sources
cells
1 -*31,416
Deterministic
ntat(c,t)
total number of individuals cell c at time t
(= nn.,(c,t) + Y. n(i,c,t),i=i,i2)
individuals
0 -+ 10,100
Calculated
P
Bamthouse logistic function
Dimensionless
o 1
Calculated
P(x,c)
total probability of survival from age x to age x+1 in
cell c
( = Pn(x) 11 - dq(x,c) ] = [ 1 - qn(x) ] [ 1 - dq(x,c) ])
Dimensionless
0 -> 0.984
Calculated
P„(x)
natural probability of survival from age x to age x+1
(= f 1 - a_(x) 1)
Dimensionless
0 -~ 0.984
Stochastic
-------
234
Pp '
survival rate from the larval pool to the 0-2 month
class
Dimensionless
0 .004
Stochastic
pool(t)
total number of eggs produced by all cells at time t
(= £ eggs(c.t))
eggs
0 1.78 1013
Calculated
Ppeq
probability of survival from the pool to the 3-12
month class (= "rj in Brousseau's paper)
Dimensionless
.00001462
Stochastic
CNJ
©
Q.
probability of survival from 0-2 month class to 3-12
month class
Dimensionless
.004 1.0
Stochastic
P3-12
probability of survival from 3-12 month class to 2nd
year
Dimensionless
0.177
Stochastic
% •
mortality from the pool to the 0-2 month class (= [ 1 -
Pnl)
Dimensionless
.996 1.0
Stochastic
qn(x)
natural age specific mortality for age x
Dimensionless
.016 -» 1.0
Stochastic
q(x.c)
total mortality for age x, cell c
( = H - p(x,c) ] =1-|1-q„(x) 111 dq(x,c)l)
Dimensionless
.016 1.0
Calculated
p
distance from center of domain to a cell
Dimensionless
0 -1,100
Calculated
r
distance from stress source to a cell
Dimensionless
0 200
Deterministic
Rjp
ratio of probability of survival in the 0-2 month stage
to the probability of survival in the pool (= p/ p„)
Dimensionless
1 68382
Deterministic
S©
normalized stress level in cell c
Dimensionless
1.6 10-1J -1, 10-2
Calculated
C/>
0
stress level at stress source
Dimensionless
1.6 10"1J -~ 10 2
Deterministic
^tot
total stress from one source cell
{ = ST/n,.= 1 /nrf)
Dimensionless
3 105 1.000
Deterministic
-------
235
ST
total stress in the domain
( = ZSM= 1.000)
Dimensionless
1.000
Definition
0
angle between r and p
Dimensionless
0 -*¦ 2tt
Calculated
t
time
year or month
1 50
Definition
X
age
year or month
1 -*¦ 12 or
1 144*
Definition
X
Bamthouse stress parameter = log10 [S/So]
Dimensionless
-4 "*0
Calculated
DETERMINISTIC = Only varied deterministically during sensitivity analyses (no stochastic variability added)
STOCHASTIC = Stochastic variability added, may also be varied deterministically.
CALCULATED = Calculated in model (may have deterministic and/or stochastic components)
DEFINITION = Defined in model.
* = used only if time unit is one month rather than one year
-------
236
Appendix Table 2: Key Assumptions in the Spatially Explicit Population Model. See Table 3 for a summary of the
research phases and how specific assumptions will be relaxed.
ASSUMPTION
NOTES
WHEN ASSUMPTION
RELAXED
HABITAT CHARACTERISTICS
Benthic habitat spatially homogeneous.
Violated both due to within-habitat and among-habitat variability.
Phases IV, V. and VI.
Benthic habitat temporally homogeneous.
Likely to be violated though not apparent whether it is as important
as spatial variation except in extreme cases.
Phase IV.
Water column habitat, and its affect on larval survival,
temporally homogeneous.
Poor assumption - temporal variation in water column habitat
quality is likely to be a major driver for variability in larval survival.
Phases IV and V.
Habitat types are discrete.
Discrete habitat types imposed as spatial structure in Phase IV.
Effect of discrete habitats vs. gradients scale dependent.
Not addressed in this research.
STRESSOR CHARACTERISTICS
Stressor decays exponentially with distance from
source
Alternate models to be tested.
Phase III and VI.
Secondary ecological effects of stress are
unimportant.
Use of field derived vital rates incorporates natural levels of
interspecific interactions for that location and time. Potentially can
have a large effect if food or key stone predator more sensitive than
target species.
Not addressed during this
research.
Stress level constant over time.
Should be approximated in short term for sewage discharges and
sediment contaminants. Less valid over long term.
Not addressed during this
research.
Logistic stressor function describes stressor effects on
survival and fecundity with Mya.
General dose-response equation. Evaluated for mortality with fish
with 77 compounds, not evaluated for fecundity. Should
approximate any sigmoid response; questionable for biotic
stressors.
Alternate models evaluated in
Phase III and VI.
Stressor does not impact the larval (water column)
stage.
Assumption good for sediment stressors. May be violated locally
for effluents but likely to have negligible effect on larval pool.
Not addressed in this research.
Effects of natural stress and anthropogenic stress are
independent
Synergetic effects on mortality not well documented in aquatic
toxicology. Additivity of toxicity is a base assumption in many risk
assessments.
Not addressed in this research.
-------
237
POPULATION CHARACTERISTICS
Population is at equilibrium at time=0.
Often violated in short-term, approximated true in long-term.
Violation has greatest effect on "larval" survival rate since rate was
calculated assuming equilibrium population structure. Addressed
by adding stochastic variation
Phases IV and V.
Age-specific vital rates resemble those determined for
Mya.
Used field derived rates for adults though likely to vary naturally with
habitat and temporally. Addressed through sensitivity analysis and
stochastic variation.
Phases III, IV, V.
Leslie matrix is suitable population model.
Leslie matrix is broadly accepted age-structured model. IBM
approach might be preferred if modeling mobile species with
strongly density-dependent migration.
Not addressed in this research.
Larvae are well mixed in water column to form a larval
pool and are distributed equally to all cells.
Likely to be violated both because of habitat heterogeneity and
stochastic processes.
Phase IV and V.
No migration of aduft clams.
Good for Mya which is deep burrowing. Needs to be evaluated for
other benthic species.
Not addressed in this research.
All adult (>1 year) age classes have same sensitivity.
Should be approximately true. Effects of increased sensitivity of
juveniles (<1 year) will be evaluated.
Adults: Not addressed in this
research.
Juveniles: Phase III and IV.
DENSITY-DEPENDENT INTERACTIONS
Adult-adult density-dependent
interactions only affect fecundity and not
adult survival
Based on literature, fecundity appears more
sensitive than mortality. Effects on growth
captured through changes in fecundity.
Not addressed in this
research.
Adult-adult and adult-juvenile density-
dependent interactions are predicted by
a ramp function.
Experimental results for adult-adult approximate
ramp function. Uncertain about adult-juvenile.
Phase III and IV.
Larval-larval density-dependent
interactions predicted by Ricker or
Beverton-Holt functions.
Common approach in fisheries models.
Phase V.
-------
238
Appendix Table 3: Phases of the Research. Only changes in an approach are listed under a phase.
HABITAT 1 STRESSOR
POPULATION
DENSITY
DEPENDENCE
TEMPORAL/
SPATIAL
VARIABILITY
VALIDATION
PHASE 1: Develop Preliminary One-Dimensional Spatially Explicitly Population Model for Benthic Clam
1-
Dimensional,
homogeneou
s
Generalized stressor-
response function
(logistic) for post-
metamorphic survival
& fecundity
Develop Leslie
matrix for Mya for
equilibrium
population
None.
None
Algorithm (inhouse
program compared
to RAMAS-GIS)
Single source cell
Single larval pool
w/equal recruitment
in all cells
Exponential decay
rate of stressor
PHASE II: Develop Preliminary 2-D SE
PM Incorporating Multi
pie Point Sources wi
th Constant Total Stress
2-
dimensional
circle, homo-
geneous
Develop methods for
multiple source cells
w/overlapping stress
Use yearly time
step (mortality in
larval and 0-2
month classes not
separated)
Derive functions
for adult-adult;
adult-juv.; larval-
larval
None
Algorithm (functions
compared to
published results)
PHASE III: Cc
No Stochastici
>nduct Simulations Using 2-D SEPM Incorporating Multiple Sources; Adult-Adult Density-Dependent Interactions;
ty
Simulate multiple source cells
with different spatial
confiaurations
Adult-adult effects on
fecundity
Preliminary simulation of
discrete habitat types
Sensitivity analyses on
lapse rate, p, a
-------
239
Alternate stressor-response
and decav sub-models
-------
240
PHASE IV: Conduct Simulations Using 2-D SEPM Incorporating Multiple Sources; Density Dependence on Adult-Adult; Spatial &
Temporal Stochasticity
Simulate a range
of discrete
habitat types
Random spatial variation in
habitat quality on fecundity
and 0-year class survival
Sensitivity analysis vital
rates
Random temporal variation
in habitat quality
Random density-indepen-
dent temporal variation on
larval survival (by varying
P™)
PHASE V: Conduct Simulations Using 2-D SEPM Incorporating Multiple Sources; Density Dependence on Adult-Adult, Adult-
Juvenile, Larval-Larval; Spatial & Temporal Stochasticity
Leslie matrix w/month
time step (mortality in
larval and 0-2 month
classes separated)
Adult-Juvenile using
ramp function
Random spatial variation in
larval recruitment
Ho: Predicted temporal
variation = observed in
catch records
Larval-Larval using
Ricker and/or Beverton
- Holt equations
Random density-indepen-
dent & density-dependent
temporal variation on larval
survival (by varying Pp^, and
carrying capacity)
Compare alternate models
for larval density
dependence
PHASE VI: Simulate Geographical Site
Larval Stages; Spatial & Temporal Stoc
Using Multiple Sources; Density Dependence on Adult-Adult, Adult-Juvenile, and Larval-
hasticity
Simulate habitat
quality based on
field data
Empirical stressor distribution
around S. Calif, outfalls.
Ho: Predicted 1-D stressor
pattern = pattern observed
in S. California
Empirical dose-response
based on pop. distributions
around S. Calif, outfalls.
Ho: Predicted 1-D pop.
distribution = distribution
observed in S. Calif.
-------
241
MATHEMATICAL DETAILS
A-l. LESLIE MATRIX APPLIED TO MY A
Brousseau et al. (1982) provided a life table for the clam Mya arenaria assuming an
equilibrium population structure (Table A-1). As discussed in the text and below, this work
did not separate larval and early juvenile mortality. Also as discussed in the text, we use the
concept of a larval pool to sum the reproductive output from all the cells (subpopulations).
The Leslie matrix for a species with a life span of twelve years, including the concept of a
fully mixed larval pool, with a year time step, is:
eggs
n Pm
0 2
p m
2 3
P m
11 12
pool/ncells
nl
p
fpeq
n1
"2
n3
—
P!
P2
X
"2
"3
-12_
pn_
"l 2 _
(1)
c,t+1
C,t
or, for the eggs produced by cell c at time t:
eggsCiW = p„ m2c n ,et +p2c m 3c n 2et + ...
+P 11, m 12, n 11,,t (2)
The larval pool population is defined as
pool M = leggs c M (3)
C
and, for survival to the next generation:
n i,,fi = Ppeq P°ol, / ncells
^ 2,.t+1 = P 1, n 1,c,t (4)
n 12,c,t+1 = P 11 ,c n 11,c,t
where n x = number of female individuals, x = age = 1,2,..12
-------
242
P x = probability of survival from age x to age x+1
Ppeq = probability of survival from the larval pool to age 1
nee I Is = total number of cells in the domain
m x = annual number of female eggs produced by a female of age x
C = cell number
t = time step (year)
Note that
a) both age-specific survival, p, and age-specific fecundity, m, may be modified by
the anthropogenic stress
b) the presence of p in equation (2) implies reproduction after survival
P
n
m
eggs
pool/ncells
2.84E+06
0
0
3-12 mo
P peq
41.51
0
0
2 yr
0.177
7.35
3744
27508
3
0.912
6.70
17170
115048
4
0.904
6.06
31159
188739
5
0.952
5.77
39957
230414
6
0.949
5.47
50341
275489
7
0.969
5.30
62460
331213
8
0.984
5.22
76465
398991
9
0.911
4.75
76465
363481
10
0.911
4.33
76465
331131
11
0.911
3.95
76465
301661
12
0.911
3.59
76465
274813
Total Pop.
Total eggs
(exel pool)
100.00
=pool
2.84E+06
Table A-1: Life History Data for Mya arenaria. Data from Brousseau et al. (1982).
Calculation of "larval" survival is based on an equilibrium population.
-------
243
A-ll. CONVERSION OF LESLIE MATRIX TO MONTH TIME STEP
The field technique used by Brousseau et al. (1982) did not sample juveniles less than 2mm
in length, the 0-2 month old class. Thus, their estimate of "larval" survival includes both
losses in the larval stage and in the early juvenile stage. Initially, we use a one year time
step in the model, and use the life history data for the 3-12 month age class as
representative for the first year class. Note that this approach does not change the total
mortality but does allocate more of it to the larval stage and less to the post-metamorphic
(benthic) stage.
A future step in the refinement of the model will be to treat the pool of eggs and each of the
first twelve months of life in the benthos separately. The reasons are:
1. Brousseau's data starts with month 3 and the age step in the Leslie matrix should
be the same for all ages;
2. Modeling adult-juvenile and larval-larval density dependence will require treating
the pool and the first few months of the benthic stage, the recruits, individually.
3. With a month time step and the corresponding life history and stressor data, it
would be possible to model seasonal variations in effects.
Therefore, for the entire life cycle, the model eventually will use a month as a time unit rather
than a year. On a modern computer there is no noticeable difference between applying the
Leslie equations 12 times or 144 times at each time step. The model will be designed to
preserve certain properties of the Brousseau yearly life table. After the first year, probability
of survival from one month to the next will be the Brousseau measured value raised to the
1/12 power so that after being applied 12 times the same probability of survival will obtained
in each year and the equilibrium population distribution also will be maintained.
Similarly, in the ten months from three through twelve, the Brousseau survival ratio for the 3-
12 month class to the second year will be taken to the 1/10 power and applied 10 times.
Finally, the survival rate, pw from the pool to month three will be artificially split into a
survival from pool to month one and the remaining survival to month three. The ratio of
survival in the 0-2 month class to survival in the larval phase (Rjp) is used to partition the
larval and juvenile survival. The survival from month one to month three will be raised to the
Vi power and applied twice.
The result is that at the third month and during the first month of each year the month-based
population calculation has the same value as the year-based calculation. Fecundity will be
allowed for only one month each year, with the fecundity values being the same as those
presented in Brousseau's annual data (Table A-1).
Using the fecundities and probabilities of survival from the life history characteristics (Table
A-1) the probability, ppeq , of survival from the pool to the 3-12 month class for an
unstressed population in equilibrium can be determined by simple algebraic steps which are
-------
equivalent to setting the principal Eigenvalue of the Leslie matrix =1.0
Consider a single cell and Brousseau's life table. Eliminating 6ggS and pool between
equations (2), (3), and (4) yields the required value of the probability of survival from the
pool to the 3-12 month class (assuming no juvenile fecundity) as
12 1-1
Ppeq=10/ [Pi(m2+£m, n P J) ]
(5)
I = 3 j = 2
In the case of the Brousseau data ppeq = .00001462. This value will vary with different
environmental conditions for the same species and among species.
A-lll. DENSITY-DEPENDENT RAMP FUNCTION
A ramp function will be used to model adult-adult and adult-juvenile density-dependent
interactions. Two parameters are needed to describe the relationship, a threshold density at
which density-dependent interactions begin and a slope of the line (see Figure A-1). Both
density and fecundity are expressed as "normalized" values which are the ratios of the
predicted values to the equilibrium values. The normalized fecundity can then be calculated
as:
If normalized adult density < threshold,
normalized fecundity = 1.0.
If normalized adult density ^ threshold,
normalized fecundity =1- slope *( normalized adult density - threshold ).
If normalized density > threshold + 1 / slope,
normalized fecundity = 0.
Similar calculations are used to predict effects on juvenile survival. Application of density-
dependent effects on both fecundity and juvenile survival generate the net effect on
recruitment success. Multiplication of these two linear functions can result in a quadratic
(non-linear) effect on recruitment success over certain ranges. The implications of this
observation will be explored.
-------
245
Example of normalized fecundity as a functon
of adult density (after Peterson, 1982)
threshold = .75
slope = 2
*
1
a
0.6 -
*
a 0.4
IB
E
6
c
0
1
1.5
0.5
normalized adult density
Figure A-1: Reduction in normalized fecundity as a function of adult density
normalized to "natural" density.
IV. APPLICATION OF SINGLE STRESSOR SOURCE IN CIRCULAR DOMAIN
Consider a single source of stress located at the center of the circular domain (See Figure A-
2).
The stress function used in this proposal is
S = S0 e"bs' (6)
For a single stress source a the center of the circular domain,
^ max
-------
246
Sto, = S0 Jebsr2 n rdr (7)
0
i.e. integrating rings of constant stress.
-------
247
Figure A-
Single
cell in
domain.
where
r = the
e from
source,
ring inconstant
stress
TOP VIEW
CROSS SECTION
2:
source
circular
distanc
the
radius
= 100,
stress
1 max !
of domain L-
bs =
lapse rate,
S0 = stress level at the center (source cell),
S = stress level at r,
Stot = total stress in the domain.
For comparing cases of different lapse rates, bs, we define the total stress in the system
as
s,„t = 1 .o
Then
S0 =bs2 / [2 TT {e -100bs (-100 bs -1) + 1 }]
-100 bs
(8)
which shows that the lapse rate and S0 are intimately connected. To compare different
stress profiles (e.g., dispersed vs. contained) it is convenient to select a set of S0 and
examine the stress and accumulated stress as functions of radial distance from the source.
-------
248
Equation (8) can be used to produce Table A-2. These results show that as dispersion
increases (smaller bs), the stressor level increases in the source cell.
So
bs
r*for S / S0 = .15
1.0 E-2
0.25066
8
1.0 E-3
0.07914
24
1.0 E-4
0.01874
100
Table A-2: Relationship between stress and lapse rate.
Here P* is defined, as an example, as the point where the stress is approximately 15% of
the stress at the origin.
r* - -ln(S/S0) / bs = -ln(.15)/bs = 1.897 / bs (9)
Figure A-3 shows the raw stress as a function of r. This is the stress level associated with
each cell at distance r from the source cell. By definition, the total stress introduced into the
model
domain is a constant (1.0). This is illustrated in Figure A-4 which shows the integrals of each
curve.
-------
249
Stress as a function of distance from concentic source cell
i
1.E+00
80
11)0
.E-01
distance
15%
E-02
15%
.E-03 14.
15%
.E-05
E-06
V)
V)
at
U)
So=1.E-2
A So=1.E-3
~ So=1.E-4
.E-07
.E-08
.E-09
.E-10
.E-11
.E-12
E-13
Figure A-3: Stress level as a function of distance from source cell for three different lapse rates in
a circular domain. Lapse rates from Table A-2. "15%" indicates cell where stress level equals
15% of the source cell. Total stress in domain is 1.0 in all cases.
-------
250
TolaJ stress as a function of distance from oonbentric stressed cdl
1.0
y
as-
07
0.6
0.4
0.3
0.0
100
Figure A-4: Accumulated stress as a function of distance from the origin. Total stress =1.0
for all cases.
V. NON-SYMMETRICAL AND MULTIPLE STRESSOR SOURCES
If the cell domain and the stressed cells are symmetrically located about the center of a
circle, then all cells at the same distance from the center are affected equally by the
stressor(s). This concept restricts the stressed region to circles or circular rings all
concentric with the center of the circular domain of cells. Stresses can be calculated to the
cells in each annular ring from the center to the edge and the Leslie matrix can be applied to
what has essentially been reduced to a one-dimensional problem.
-------
251
Since the cells are defined to have unit area, the number of cells at any radius, r, is
nc(r) = 2 tt r (10)
and the total number of cells in the domain is
ncells = n rmax2 = nioo2 = 3i,4i6 (11)
In actual fact, the problem is solved by integrating the annular rings rather that each cell, so
the potential problem of fractional cells in equations (10) and (11) vanishes.
If a stress source is not at the center of the domain (see Figure A-5), the stress still depends
on the distance, r, from the source. Such a stress source has the potential of spreading
stress to the limits of the grey circle in Figure A-5, but its effect will be limited to the area of
the domain. The total stress from such an off-center source, Stot, will still be defined as
1.00 (for a single source cell in the domain), but S0 and bs cannot be calculated from
Equations (7) and (8), since they are based upon the stress source being located at the
center of the domain. Stress can be numerically integrated over the grey circle, but S will be
set to zero if p > 100, i.e. if the location of a cell is not in the domain.
-------
domain «s
center "
stress
source
S stressed d$\
cell F Tdp
Figure A-5: Stressor source not at center of model domain.
Using the trigonometric "Law of Cosines"
r = V d2+p2 + 2dp cos 0 (12)
and 100 n
Stol = 2Sj Je bsrpd0dp (13)
0 0
100 n
or S„ = .5/ J Je ^'pdGdp (14)
0 0
-------
If multiple stress sources are studied, the total stress from each will be accumulated and
then each will be divided by the total number of stress cells, nst, (since each source produces
the same total stress) so that in comparative cases, the total stress to the domain will still be
ST = ISJ nst = 1.00.
nst
VI. APPLICATION OF LOGISTIC STRESSOR-RESPONSE FUNCTION
Barnthouse et al. (1990, his equation #8) gives the fractional response, P, to exposure
concentration , X, as
p= e(cr+(VQ I [1 + g(cr+ @X) j
In Barnthouse's work the concentration term, X, is understood to be the log10 of a physical
concentration (|jg/L). We will use X- log10 [S/S0] Instead and then adjust the response to
the stress by specifying increases in mortality and decreases in fecundity.
"/?" determines the slope and curvature of the logistic function and "a ", a function only of
dm0
or , determines its level. The response function can be obtained by specifying either a
and P or /?and the response at a particular cell. We will specify Bas /?q for increased
mortality and /?m for reduction in fecundity and Barnthouse states that compared to other
uncertainties there is no reason to try to use different fi's for different age classes. In the
earlier Phases we will use Barnthouse's nominal value of 6.0 for both /?'s and for all ages.
We replace Barnthouse's Pwith a notation of increased mortality due to anthropogenic
stress, dq(x,c).
The aq parameter can be calculated following Barnthouse. In general, for age x
aq(x) - In [ dq(x,c*) / (1-dq(x,c*)) - /?q(x) Xq*]
X* = log10 (stress) for a fractional increase in mortality, dq(x,c*).
c* is the cell where we wish to perform the calculation of aq.
We will define the response at the source cell where X* - log10 (1.0 ) = 0
since S/So=1.0 by definition of our normalised stress curve and
(15)
-------
254
Then for dq(x, c*) = .9 (say)
CTq(x) = In (.9/1) = 2.197
Now we have
dg =g{aq(x) + ft(x) log10(S/So]}y +g{aq(x) + ft(x) Iog10[ S/So]}j
(17)
Similarly, for fecundity, (see Figure A-6 and Table A-3 )
am(x) = In [ dm(x, c.) / (1-dm(x,,..))]
(18)
and
dm =g{am(x) + An(x) Iog10[ S/So)]} j ^ +g{am(x) + /*n(x) Iog10[ S/So]}j
Bamthouse reduction in fecundity
beta = 6
-dmO = 0.1
-dmO = 0.3
-dmO = 0.5
-dmO = 0.7
-dmO = 0.9
The natural probability of survival, p x = [i-q J. must be multiplied by the stress-related
reduction in
the
probability
of survival,
[1 - dq J, to
produce the
total
probability
of survival.
Natural
fecundity,
mx, must be
multiplied by
the stress-
related
reduction in
fecundity,
[1 - dm x,d
, to produce
the total fecundity.
-------
255
dmO
alpha
0.1
-2.1972
0.3
-0.8473
0.5
0.0000
0.7
0.8473
0.9
2.1972
Figure A-6: Change in reduction in fecundity, dm, as a function of the relative stress level.
Table A-3: Relationship between reduction in fecundity and Ctm(x)
The Leslie equations become
eggsct+1 = p 1>c [1 -dq(1,c)] m2 c [1 - dm(2,c)] n 1 ct
+ p 2c [1 - dq(2,c)] m 3c [1 - dm(3,c)] n 2ct + ...
+P [1 -dq(11,c)] m 12 c [1 -dm(12.c)] n 11ct (20)
The larval pool population is still defined as
pool t+i = I eggs c ,„
(21)
-------
and, for survival to the next generation:
n u,M = Ppeq pool t / ncells
n 2,c,t+1 = P 1,c
[1 -dq(1,c)] n
1,C,t
n3tc,t+i = P 2,c n ¦ dq(2,c)] n 2 ct
n 12,c,t+1 ~ P 11,c ^ 11,c,t
-------
NHEERL-COR-900R
lTTTT,T,„T „„„„ TECHNICAL REPORT DATA
NHEERL-COR - 900R /0. . . . .
(Please read instructions on the reverse before completing)
1. REPORT NO.
600/R-99/043
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE The effects of hatitat alteration by estuarine stressors on
ecological resources of Pacific Northwest estuaries.
5. REPORT DATE
6. PERFORMING ORGANIZATION
CODE
7. AUTHOR(S) B. Boese, F. Cole, T. Dewit, S. Ferraro, J. Lamberson, H. Lee, W.
Nelson, b. Ozretich, J. Power, B. Robbins, A. Sigleo, D. Specht, D. Young
8. PERFORMING ORGANIZATION REPORT
NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
US EPA ENVIRONMENTAL RESEARCH LABORATORY
200 SW 35th Street
Corvallis, OR 97333
13. TYPE OF REPORT AND PERIOD
COVERED
14. SPONSORING AGENCY CODE
EPA/600/02
15 SUPPLEMENTARY NOTES:
-------
1 6. Abstract: The coastal Ecology branch (CEB), Western Ecology Division of the US Environmental Protection Agency will
initiate a research program to evaluate the effects of alterations of estuarine habitats resulting from multiple stressor sources.
The research will concentrate on stressor effects on the ecological resources of estuaries of the Pacific Northwest (PNW). The
research program is designed to support the general mission of the US EPA which includes the safeguarding of the natural
environment upon which the health and well being of the nation's population ultimately depends.
The goal of CEB research is to improve the ability to make key policy decisions on coastal environmental issues by defining key
ecological processes and by developing models to predict stress-response relationships for ecological resources within Pacific
Northwest estuaries at range of spatial and temporal scales. CEB research objectives are to 1) evaluate how specific estuarine
habitats respond to a range of potential stressors which may lead to habitat alteration, 2) understand the influences of these
stress factors at spatial scales from local to regional, and 3) develop indicators of ecological condition which may be used to
evaluate estuarine status across multiple spatial scales.
The research effort will concentrate on two habitats 1) submerged aquatic vegetation (SAV) and 2) burrowing shrimp, with lesser
effort on other types of estuarine habitats. SAV and shrimp are selected as focal research habitats and important assessment
endpoints because in each case the characteristic species which define the habitat do so because of their "physical ecosystem
engineering" activities.
To accomplish the objectives of the CEB research plan, research will be organized in three thematic elements: A. Indicators of
Ecological condition for PNW estuaries; B. Stressor Response Modeling; C. Estuarine Physical-chemical stressors.
The research projects under Research theme A address the questions: 1) what are the biotic constituents of major estuarine
benthic habitats of PNW estuaries, 2) what effects do various abiotic and biotic stressors have on the biotic composition of
principal habitat types, 3) what role do biotic and abiotic stressors have in controlling the spatial extent and distribution patterns
of major estuarine habitat types, and 4) what are appropriate indicators of ecological condition at the population, species,
community, and landscape levels for PNW estuarine systems. Stressors that will be examined include anthropogenic physical
disturbances such as clam and borrowing shrimp harvesting, and sedimentation, salinity, and water column light field alterations
potentially generated by elevated runoff resulting from changes in landscape-use patterns. Biotic stressors that will be examined
include disturbances such as the smothering of seagrass habitat by mat-forming algae potentially promoted by elevated nutrients,
and biotic stress induced by competition between native and exotic seagrass species and between burrowing shrimp and
seagrasses.
Projects under Research theme B will work at the population and community levels to develop modeling techniques to integrate
the detailed studies of biological effects of estuarine stressors of theme A with the spatial-temporal stressor distribution studies
of theme C. The principal current project is the development of spatially explicit modeling tools for estuarine benthic populations
to allow predictions of population responses to the imposition of multiple stressors.
The research projects under Research Theme C will address the questions 1) what are the spatial and temporal distribution
patterns of the primary physical and chemical factors determining estuarine habitat composition, 2) how are spatial variations in
the physical anthropogenic alterations of watershed characteristics influence the
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED
TERMS
c. COSATI Field/Group
Estuarine Stressors, Pacific Northwest,
Yaquina Bay
18. DISTRIBUTION STATEMENT
1 9. SECURITY CLASS (This Report)
21. NO. OF PAGES: 256
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
------- |