EPA 600/R-09/140 | October 2009 | www.epa.gov/ord
United States
Environmental Protectio
Agency
Classification of Regional
Patterns of Environmental
Drivers and Benthic Habitats
in Pacific Northwest Estuaries
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Western Ecology Division
-------
DISCLAIMER
The information in this document has been funded wholly by the U.S. Environmental Protection
Agency (EPA). Deborah Reusser was partially funded through AMI/GEOSS TAG #DW-14-
92231501-0 from the U.S. EPA. The publication was subjected to review by the National Health
and Environmental Effects Research Laboratory's Western Ecology Division and the U.S.
Geological Survey (USGS), and is approved for publication. However, approval does not signify
that the contents reflect the views of the U.S EPA or the USGS. The use of trade, firm, or
corporation names in this publication is for the information and convenience of the reader; such
use does not constitute official endorsement or approval by the U.S. Department of Interior, the
USGS, or the EPA of any product or service to the exclusion of others that may be suitable.
Preferred Citation
Lee II, H. and Brown, C.A. (eds.) 2009. Classification of Regional Patterns of Environmental
Drivers And Benthic Habitats in Pacific Northwest Estuaries. U.S. EPA, Office of Research and
Development, National Health and Environmental Effects Research Laboratory, Western
Ecology Division. EPA/600/R-09/140.
ACKNOWLEDGMENTS
The authors wish to thank the following for their contributions to this report:
Dynamac, Inc. for field and laboratory work; Steve Rumrill and Alicia Helms, South Slough
National Estuarine Research Reserve, for Coos Estuary salinity data; Ralph Garono, Earth
Designs, Inc., Steve Rumrill and Derek Sowers, South Slough National Estuarine Research
Reserve, and Mark Trenholm and Don Reynolds, Tillamook Estuaries Partnership, for reviewing
the aerial photomap classification; George Priest, Oregon Department of Geology and Mineral
Industries, for providing bathymetric data; Peter Eldridge (deceased), U.S. EPA, for providing
isotope data from the Salmon River Estuary; Steven Ferraro, U.S. EPA, for information on food
webs; Robert Lackey, formerly of the U.S. EPA, for salmon information; Tony Olsen, U.S. EPA,
for assistance in sampling design; Tad Larsen of Raytheon Information Solutions for GIS
support; staff of the Integrated Stable Isotope Research Facility, U.S. EPA, for analyzing the
stable isotope samples; and Jimmie K. Cheney and Karen Ebert, Senior Environmental
Employment Program with U.S. EPA, for proofreading the document. Peter Eldridge, Melanie
Frazier, Jody Stecher, Katharine Marko, Allison Aldous of The Nature Conservancy, and Jeff
Weber of the Oregon Department of Land Conservation and Development reviewed an earlier
version of this document and provided insightful suggestions.
-------
TABLE OF CONTENTS
List of Figures vi
List of Tables xii
EXECUTIVE SUMMARY xv
CHAPTER 1: INTRODUCTION AND FRAMEWORK FOR THE PACIFIC NORTHWEST
REGIONAL CLASSIFICATION PROJECT 1
1.0 Scope 1
1.1 Problem Statement 1
1.2 Regulatory Framework for Estuarine Classifications 2
1.3 Overview of Eutrophic Conditions and Nutrient Dynamics in the Pacific Northwest 4
1.4 Biotic Response Measures in PNW Estuaries 5
1.5 Design and Objectives of the Pacific Northwest Regional Study 8
CHAPTER 2: REGIONAL CLASSIFICATION OF PACIFIC NORTHWEST ESTUARIES BY
WETLAND AND LAND COVER PATTERNS 10
2.0 Introduction 10
2.1 Estuarine, Wetland, and Landscape Data Sources and Methods 11
2.1.1 Estuary Definition and Inventory 11
2.1.2 Wetland Habitats Based on NWI 13
2.1.3 Watershed Delineation 22
2.1.4 Land Cover Patterns, Impervious Surfaces, and Population Density 23
2.1.5 Classification Strategy and Methods 23
2.1.6 Ecological Resources in Pacific Northwest Estuaries 30
2.2 Inventory of Estuaries on the Pacific Coast 31
2.3 Ecological Services of Small Pacific Northwest Estuaries 34
2.4 Classification by Geomorphology and Oceanic Exchange 41
2.4.1 Coastal Lagoons and Blind Estuaries 42
2.4.2 Tidal Coastal Creeks/Tidally Restricted Coastal Creeks 43
2.4.3 Marine Harbors/Coves 45
2.4.4 Drowned River Mouth and Bar-Built Estuaries 45
2.4.4.1 Tide- and River-Dominated Drowned River Mouth Estuaries 46
2.5 Classification of Estuaries by NWI Wetland Classes 53
2.6 Watershed Classification 61
2.6.1 Watershed Classes 61
2.7 Crosswalk of Wetland and Watershed Classifications 65
2.8 Watershed Alterations and Population Patterns 74
2.8.1 Flow-Path Analysis of Impervious Surfaces 79
2.9 Synthesis of Classification Approaches and Research Needs 80
CHAPTER 3: ENVIRONMENTAL CHARACTERISTICS OF THE SEVEN TARGET
ESTUARIES 82
3.0 Introduction 82
3.1 Pacific Northwest Climate 82
3.2 Coastal Upwelling in the Pacific Northwest 82
3.3 Criteria for Choosing Target Estuaries and Previous Estuarine Classifications 83
3.3.1 Management Classification of Oregon Estuaries 83
3.3.2 Geomorphological/Hydrological Classifications 83
11
-------
3.4 Estuary and Watershed Sizes 85
3.5 Watershed Characteristics and Land Cover 86
3.5.1 Watershed Slopes 86
3.5.2 Population Characteristics 86
3.6 Estuarine Hydrology 90
3.7 Nutrient Loading 91
CHAPTER 4: WATER QUALITY SURVEYS IN SEVEN TARGET ESTUARIES 96
4.0 Introduction 96
4.1 Methods 97
4.1.1 Water Quality Cruises 97
4.1.2 Datasonde Deployment 98
4.2 Riverine Nutrient Inputs During the Wet Season 99
4.3 Coastal Ocean and River Conditions During Dry Season Sampling 100
4.4 Estuarine Water Quality 102
4.4.1 Dry Season Chlorophyll a and TSS Patterns 102
4.4.2 Dry Season Nutrient Patterns 107
4.4.3 Dry Season Salinity Fluctuations 109
4.5 Comparison of Water Quality Data to EMAP Criteria 110
4.6 Synthesis 113
CHAPTER 5: ZONATION OF OCEAN AND RIVER DOMINANCE IN SEVEN TARGET
ESTUARIES 115
5.0 Introduction 115
5.1 Stable Isotope Sampling 117
5.2 Yaquina Estuary 118
5.2.1 Transport Model 118
5.2.2 Conservative Mixing Model 124
5.3 Alsea Estuary 128
5.4 Salmon River Estuary 134
5.5 Umpqua River Estuary 137
5.6 Coos Estuary 140
5.7 Tillamook Estuary 144
5.8 Nestucca Estuary 150
5.9 Synthesis and Uncertainties inZonation 154
CHAPTER 6: AERIAL MEASURES OF ESTUARINE INTERTIDAL AND SHALLOW
SUBTIDAL ZOSTERA MARINA COVERAGE 158
6.0 Introduction 158
6.1 Methods 158
6.1.1 Aerial Photography Sampling Frame and Design 158
6.1.2 Estuaries Surveyed 162
6.1.3 Ground Surveys 162
6.1.4 Classification Accuracy Assessments 163
6.1.5 Estuary Bathymetry 163
6.2 Results 164
6.2.1 Accuracy Assessments of Photomap Image Classifications 164
6.2.2 Among-Estuary Comparison of Zostera marina Coverage 164
6.2.3 Bathymetric Distribution of Zostera marina 165
in
-------
6.2.4 Within-Estuary Distribution oi Zostera marina 171
6.3 Synthesis 172
CHAPTER 7: AMONG AND WITHIN ESTUARINE DISTRIBUTIONS OF SEAGRASSES
AND ECOLOGICALLY IMPORTANT BENTHIC SPECIES IN PACIFIC NORTHWEST
ESTUARIES 173
7.0 Introduction 173
7.1 Methods 174
7.1.1 Probabilistic Sampling Frame and Design 174
7.1.2 Data Analysis 175
7.1.3 Field Methods 175
7.1.4 Comparison of 0.25 m versus 2.5 x 2.5 m Quadrats 189
7.2 Across-Estuary Patterns 189
7.2.1 Sediment Grain Size and TOC in the Target Estuaries 189
7.2.2 Among-Estuary Distribution of Zostera marina 193
7.2.3 Among-Estuary Distribution of Zostera japonica 194
7.2.4 Composition and Seasonality of Benthic Green Macroalgae 196
7.2.5 Among-Estuary Patterns of Benthic Macroalgae 199
7.2.6 Among-Estuary Patterns of Burrowing Shrimp 205
7.2.7 Bare Habitat 208
7.2.8 Multivariate Analysis of the Target Estuaries 208
7.3 Within-Estuary Distribution of Seagrasses and Other Benthic Resources 211
7.3.1 Within-Estuary Distributions of Seagrasses 212
7.3.2 Within-Estuary Distributions of Benthic Macroalgae 215
7.3.3 Within-Estuary Distributions of Burrowing Shrimp 215
7.4 Benthic Macroalgae as Diagnostic Indicator of Cultural Eutrophication 218
CHAPTER 8: LOWER DEPTH LIMIT OF ZOSTERA MARINA IN SEVEN TARGET
ESTUARIES 219
8.0 Introduction 219
8.1 Methods 220
8.1.1 Selection of Sampling Points 220
8.1.2 Sampling 221
8.1.3 Light Attenuation Coefficients 222
8.1.4 Epiphyte Biomass 223
8.2 Results 223
8.2.1 Sampling Points/Sampling Errors 223
8.2.2 Depth Distribution 224
8.2.3 Kdvs. Distance from Estuarine Mouth 229
8.2.4 Maximum Zostera marina Depth, ^and Light Relationship 233
8.2.5 Epiphyte Patterns and Impact on Light 234
8.3 Discussion 236
8.3.1 Sampling Method 236
8.3.2 Depth Distributions 236
8.3.3 Water Clarity Relationships 237
8.3.4 Zostera marina Light Criteria 239
8.3.5 Yaquina Estuary Zostera marina Light Criteria 240
CHAPTER 9: EMERGING PACIFIC NORTHWEST PARADIGM 242
IV
-------
9.0 Overview of the Pacific Northwest Environment and Regulatory Implications 242
9.1 Environmental Conditions 242
9.2 Classification of Pacific Northwest Estuaries 246
9.3 Regulatory Implications 247
9.4 Future Directions 247
REFERENCES 248
APPENDIX A: MAPS OF THE DISTRIBUTION OF ZOSTERA MARINA BASED ON
AERIAL SURVEYS 267
APPENDIX B: DETAILED DESCRIPTION OF QUALITY ASSURANCE
PROCEDURES 275
-------
List of Figures
Figure 2-1. Overview of the generation of the PNW estuary inventory and the wetland and
landscape data synthesis and analysis 12
Figure 2-2. Estuaries along the Washington coast from Cape Flattery to the Columbia River... 14
Figure 2-3. Estuaries in northern Oregon 15
Figure 2-4. Estuaries in southern Oregon 16
Figure 2-5. Estuaries in Northern California, north of Cape Mendocino 17
Figure 2-6. Size distribution of all 230 coastal estuaries in California, Oregon, and Washington
identified by the presence of aNWI estuarine polygon 32
Figure 2-7. Size distribution of the 103 PNW coastal estuaries identified by the presence of a
NWI estuarine polygon 32
Figure 2-8. Distribution of estuarine subtidal and intertidal habitats in the 103 PNW estuaries. 33
Figure 2-9. Distribution of the projected historical number of Coho smolts by estuary size
classes in 29 Oregon estuaries 41
Figure 2-10. Relationship between freshwater inflow normalized by estuarine area and
normalized by estuarine volume 50
Figure 2-11. Salinity versus distance from the mouth for the wet and dry seasons for estuaries
with area-normalized freshwater inflow ranging from about 11 to 695 m3 m"2 year"1 51
Figure 2-12. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the area
of the consolidated NWI habitats 55
Figure 2-13a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
area of the consolidated NWI habitats 56
Figure 2-13b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
area of the consolidated NWI habitats 56
Figure 2-13c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
area of the consolidated NWI habitats 57
Figure 2-13d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
area of the consolidated NWI habitats 57
Figure 2-14. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
relative proportion of the consolidated NWI habitats 58
Figure 2-15a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
relative proportion of the consolidated NWI habitats 59
Figure 2-15b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
relative proportion of the consolidated NWI habitats 59
Figure 2-15c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
relative proportion of the consolidated NWI habitats 60
Figure 2-15d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
relative proportion of the consolidated NWI habitats 60
Figure 2-16. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data 62
Figure 2-17a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data 63
Figure 2-17b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data 63
VI
-------
Figure 2-17c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
areas of the land cover classes in the NO AA 2001 watershed data 64
Figure 2-17d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data 64
Figure 2-18. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data 66
Figure 2-19a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data 67
Figure 2-19b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data 67
Figure 2-19c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data 68
Figure 2-19d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data 68
Figure 2-20. Cluster analysis of the 31 PNW estuaries based on six metrics of watershed
alteration 78
Figure 2-21. Percent impervious surfaces in the Yaquina watershed for six buffer zones of
different distances from the shoreline of the Yaquina Bay or Yaquina River 79
Figure 4-1. Flood-tide water temperature at Yaquina Estuary and outermost high tide station
from 2004 dry season classification cruises 101
Figure 4-2. Flood-tide water temperature at Yaquina and Coos estuaries and outermost high tide
stations from 2005 dry season classification cruises 102
Figure 4-3. Water temperature versus MV+MV relationship generated using data from a
station on the shelf and classification high tide outermost stations 103
Figure 4-4. Water temperature versus PO43" relationship generated using data from a station on
the shelf and classification high tide outermost stations 103
Figure 4-5. Dry season chlorophyll a versus salinity for all estuaries with filled symbols
representing high tide cruises and hollow symbols representing low tide cruises 106
Figure 4-6. Median dry season chlorophyll a versus freshwater inflow normalized by estuary
volume for eight PNW estuaries 106
Figure 4-7. Dry season phosphate versus salinity for all classification water quality stations.. 108
Figure 4-8. Dry season nitrate+nitrite versus salinity for all classification water quality
stations 108
Figure 4-9. Dry season salinities near the mouths of estuaries sampled in 2004 and data
collected inNestucca during 2006 to fill data gap from lost instrument Ill
Figure 4-10. Dry season salinities near the mouths of estuaries sampled in 2005 Ill
Figure 4-11. Difference between high and low tide dry season cruise salinity at station closest to
estuary mouth versus area-normalized freshwater inflow 112
Figure 5-1. Natural abundance nitrogen isotope ratio for various nitrogen sources 117
Figure 5-2. Location map for Yaquina Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTF 119
Figure 5-3. Macroalgae isotope data at five stations (N1-N5) for Yaquina Estuary with red boxes
indicating dry seasons 120
Figure 5-4. Comparison of modeled and observed isotope ratio at Station N2 in Yaquina Estuary
during 2004 122
vn
-------
Figure 5-5. Comparison of modeled and observed isotope ratio at Station N3 in Yaquina Estuary
during 2004 122
Figure 5-6. Modeled contribution of each source of dissolved inorganic nitrogen in Yaquina
Estuary during 2004 123
Figure 5-7. Location map for Alsea Estuary showing the locations of datasondes, water quality
stations, isotope samples, and WWTF 128
Figure 5-8. Monthly average observed isotope ratio at 5 locations in Alsea Estuary 129
Figure 5-9. Location map for Salmon River estuary showing the locations of datasondes, water
quality stations and isotope samples 136
Figure 5-10. Location map of Umpqua River Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTFs 138
Figure 5-11. Location map of stations in Coos Estuary showing the locations of datasondes,
water quality stations, isotope samples, and WWTFs 143
Figure 5-12. Location map for Tillamook Estuary showing the location of datasondes, water
quality stations, isotope samples, and WWTFs 146
Figure 5-13. Median dry season salinities in Tillamook Estuary 149
Figure 5-14. Location map of Nestucca Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTF 152
Figure 6-1. Intertidal and shallow subtidal distribution of Z marina from digital image
classification of aerial photos of Yaquina Estuary taken in April 2004 160
Figure 6-2. Track of lower margin of surface-visible distribution of Z. marina in the near-
subtidal zone and of upper margin in the Yaquina Estuary 161
Figure 6-3. Sectors of Yaquina Estuary covered by bathymetric surveys conducted by the U.S.
Army Corps of Engineers between 1998 and 2002 163
Figure 6-4. Comparison of substrate elevations from a bathymetric model and independent total
station survey measurements within Yaquina estuary 167
Figure 6-5. Interpolated bathymetry and Z. marina distribution for a section of the lower
Yaquina Estuary 167
Figure 6-6. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the
bathymetric model for Yaquina Estuary 168
Figure 6-7. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the
bathymetric model for Alsea Estuary 168
Figure 6-8. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the
bathymetric model for Tillamook Estuary 169
Figure 6-9. Frequency of occurrence values for Z. marina within one-foot depth intervals around
MLLW in Yaquina, Alsea, and Tillamook estuaries 170
Figure 7-1. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
in the Alsea Estuary 176
Figure 7-2. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
in the Coos Estuary 177
Figure 7-3. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
in the Nestucca Estuary 178
Vlll
-------
Figure 7-4. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
in the Salmon River Estuary 179
Figure 7-5. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
in the Tillamook Estuary 180
Figure 7-6. The locations of the probabilistic sampling strata and the Z. marina lower limit
surveys in the Umpqua River Estuary 181
Figure 7-7. The locations of the probabilistic sampling strata and the Z. marina lower limit
surveys in the Yaquina Estuary 182
Figure 7-8. Diagram of the intertidal sampling site layout with the 2.5 x 2.5 m quadrat and the
0.25 m2 plant and shrimp quadrats used in the probabilistic surveys 185
Figure 7-9a. Front of the field sheet used to record estimates of habitat characteristics and
burrowing shrimp hole counts within the 2.5 x 2.5 m quadrat 186
Figure 7-9b. Back of the field sheet used to record plant cover, point intercepts, shoot counts,
and blade widths in the 0.25 m2 plant quadrat and burrowing shrimp hole counts in the 0.25
m2 shrimp quadrat 187
Figure 7-10. Cumulative distribution function of the percent fines in the intertidal zone of the
seven target estuaries 191
Figure 7-11. Cumulative distribution function of the percent total organic carbon in the intertidal
zone of the seven target estuaries 192
Figure 7-12. Average percent cover values of benthic green macroalgae at six sites in the
Yaquina Estuary during 1999 197
Figure 7-13. Average biomass values of benthic green macroalgae at six sites in the Yaquina
Estuary during 1999 197
Figure 7-14. Average proportions of green benthic macroalgal biomass relative to the September
peak measured in Yaquina Estuary during 1999 198
Figure 7-15. Cumulative distribution function of the % cover of benthic macroalgae estimated
from the point-intercept method using the 0.25 m2 quadrats 201
Figure 7-16. Percent of the intertidal area covered by benthic macroalgae by density class. .. 202
Figure 7-17. Cumulative distribution function of the unadjusted benthic macroalgal biomass. 203
Figure 7-18. Cumulative distribution function of the seasonally adjusted benthic macroalgal
biomass 204
Figure 7-19. Cumulative distribution function of the number of TV. californiensis burrow holes
per 0.25 m in the seven target estuaries 206
Figure 7-20. Cumulative distribution function of the number of U. pugettensis burrow holes per
0.25 m2 in the seven target estuaries 207
Figure 7-21. Percent of intertidal area occupied by bare habitat as defined by the absence of Z.
marina, Z.japonica, benthic green macroalgae, N. californiensis, or U. pugettensis in the 2.5
x 2.5m quadrats 209
Figure 7-22. Non-metric multidimensional scaling of the seven target estuaries based on the
percent of intertidal area occupied by Z. marina, Z.japonica, benthic macroalgae, N.
californiensis, U. pugettensis, and bare habitat 210
Figure 7-23. Relative distributions of the total area occupied by Z. marina between the oceanic
and riverine segments 214
Figure 7-24. Relative distributions of the total area occupied by Z.japonica between the oceanic
and riverine segments 214
IX
-------
Figure 7-25. Relative distributions of the total area occupied by benthic macroalgae between the
oceanic and riverine segments 216
Figure 7-26. Relative distributions of the total area occupied by N. californiensis between the
oceanic and riverine segments 217
Figure 7-27. Relative distributions of the total area occupied by U. pugettensis between the
oceanic and riverine segments 217
Figure 8-1. Schematic of underwater camera apparatus 222
Figure 8-2. Lower limit of Z. marina versus distance from the mouth of the Alsea Estuary.... 225
Figure 8-3. Lower limit of Z. marina versus distance from the mouth of the Coos Estuary 226
Figure 8-4. Lower limit of Z. marina versus distance from the mouth of the Nestucca
Estuary 226
Figure 8-5. Lower limit of Z. marina versus distance from the mouth of the Salmon River
Estuary 227
Figure 8-6. Lower limit of Z. marina versus distance from the mouth of the Tillamook
Estuary 227
Figure 8-7. Lower limit of Z. marina versus distance from the mouth of the Umpqua River
Estuary 228
Figure 8-8. Lower limit of Z. marina versus distance from the mouth of the Yaquina
Estuary 228
Figure 8-9. Light attenuation coefficient versus distance from the mouth of the Yaquina Estuary
generated using long-term cruise dataset 229
Figure 8-10. Light attenuation coefficient versus distance from the mouth of the Yaquina
Estuary generated using continuous dataset 230
Figure 8-11. Light attenuation coefficient versus distance from the mouth of the Yaquina
Estuary for classification dataset 230
Figure 8-12. Light attenuation coefficient versus distance from the mouth of the Alsea
Estuary 231
Figure 8-13. Light attenuation coefficient versus distance from the mouth of the Nestucca
Estuary 231
Figure 8-14. Light attenuation coefficient versus distance from the mouth of the Salmon River
Estuary 232
Figure 8-15. Light attenuation coefficient versus distance from the mouth of the Tillamook
Estuary 232
Figure 8-16. Light attenuation coefficient versus distance from the mouth of the Umpqua River
Estuary 233
Figure 8-17. Relationship between Z. marina lower depth limit and light attenuation coefficient
in the Yaquina Estuary 234
Figure 8-18. Temporal relationship of epiphytic biomass per unit leaf area on Z. marina external
leaves in the Yaquina Estuary 235
Figure 8-19. Epiphyte biomass per unit leaf area on older and younger Z. marina leaves by
season in the oceanic and riverine segments of the Yaquina Estuary 235
Figure 8-20. Linear regression relationship between the percent of light reduction to log(x+l)
transformed epiphyte biomass per unit Z. marina leaf surface area 236
Figure 8-21. Relationship between Z. marina maximum depth limit and K^. 239
Figure A-l. Distribution of intertidal and shallow subtidal Z. marina in the Alsea Estuary 268
Figure A-2. Distribution of intertidal and shallow subtidal Z. marina in the Coos Estuary 269
-------
Figure A-3. Distribution of intertidal and shallow subtidal Z. marina in the Nestucca
Estuary 270
Figure A-4. Distribution of intertidal and shallow subtidal Z. marina in the Salmon River
Estuary 271
Figure A-5. Distribution of intertidal and shallow subtidal Z. marina in the Tillamook
Estuary 272
Figure A-6. Distribution of intertidal and shallow subtidal Z. marina in the Umpqua River
Estuary 273
Figure A-7. Distribution of intertidal and shallow subtidal Z. marina in the Yaquina Estuary. 274
XI
-------
List of Tables
Table 1-1. Ecological interactions of five taxa evaluated in the field surveys 7
Table 2-1. Consolidation of National Wetland Inventory habitat codes found in PNW
estuaries 19
Table 2-2. Area and geomorphology of PNW estuaries and watersheds 24
Table 2-3. Examples of ecological resources and services in PNW estuaries 35
Table 2-4. Freshwater sources into estuaries, normalized freshwater inflow, mouth width, and
structure of mouth for drowned river mouth and bar-built estuaries in the PNW 47
Table 2-5. Crosswalk of the cluster analyses by the area of NWI consolidated wetland classes
and the area of the NO AA land cover classes at the 5% significance level 70
Table 2-6. Crosswalk of the cluster analyses by the area of NWI consolidated wetland classes
and the area of the NO AA land cover classes at the 60% significance level 71
Table 2-7. Crosswalk of the cluster analyses by the relative proportion of NWI consolidated
wetland classes and the relative proportion of land cover classes at the 5% significance
level 72
Table 2-8. Crosswalk of the cluster analyses by the relative proportion of NWI consolidated
wetland classes and the relative proportion of land use classes at the 60% significance
level 73
Table 2-9. The mean, median, and maximum values for each of the metrics of watershed
alteration for the 89 coastal watersheds 75
Table 2-10. The mean, median, and maximum values for each of the metrics of watershed
alteration for the 31 coastal watersheds used in the multivariate analysis 76
Table 3-1. Current and historical classifications of the seven target estuaries 84
Table 3-2. Estuary and watershed size of the seven target estuaries 87
Table 3-3. Watershed attributes of the seven target estuaries 88
Table 3-4. Population size and density in the seven target estuaries 89
Table 3-5. Hydrographic characteristics of the seven target estuaries 93
Table 3-6. Wet and dry season freshwater inflow for the seven target estuaries 94
Table 3-7. Residence time calculated using the modified tidal prism and fraction of freshwater
approaches for the seven target estuaries 94
Table 3-8. Estimates of nutrient loading to six of the seven target estuaries 95
Table 3-9. Comparison of nitrogen sources during wet and dry seasons for the Yaquina
Estuary 95
Table 4-1. Dates of cruises and number of stations sampled for the seven target estuaries 97
Table 4-2. Number of short-term datasondes deployed during the dry season in each estuary and
year deployed 99
Table 4-3. Median wet season dissolved inorganic nitrogen and phosphate calculated using
stations with salinity < 2 psu 99
Table 4-4. Median dry season chlorophyll a and total suspended solids for each estuary and its
ranking 105
Table 4-5. Median dry season NOs'+NCV, NH4+, and PO43" for each estuary and its ranking. 105
Table 4-6. West Coast criteria for water quality parameters (U.S. EPA, 2004b) 113
Table 5-1. Summary of median observed salinities for Yaquina Estuary during May - September
of 2003 and 2004 121
xn
-------
Table 5-2. Data used in the mixing model and time interval of salinity data utilized for each
month 127
Table 5-3. Comparison of percent contribution of DIN sources estimated using transport model
and Isosource model 127
Table 5-4. Summary of dry season salinities in Alsea Estuary 132
Table 5-5. Ocean and river end members for Alsea calculations 132
Table 5-6. Observed and predicted isotope ratios in Alsea Estuary 133
Table 5-7. Summary of dry season salinities in Salmon River estuary 136
Table 5-8. Observed and predicted isotope ratios in Salmon River estuary 136
Table 5-9. Summary of dry season salinities in Umpqua River Estuary 139
Table 5-10. Observed and predicted isotope ratios in Umpqua River Estuary 139
Table 5-11. Observed and predicted isotope ratios in Coos Estuary 141
Table 5-12. Summary of dry season salinities in Coos Estuary 143
Table 5-13. Summary of dry season salinities in Tillamook Estuary 147
Table 5-14. Observed and predicted isotope ratios in Tillamook Estuary 148
Table 5-15. Summary of dry season salinities in Nestucca Estuary 153
Table 5-16. Observed and predicted isotope ratios in Nestucca Estuary 153
Table 5-17. Summary of the total estuarine area for each target estuary and the area and
percentage of total in the oceanic and riverine segments based on salinity and macroalgal
isotope ratios 157
Table 6-1. Mapping method utilized and dates of mapping for intertidal and shallow subtidal 2.
marina distribution in the target estuaries 162
Table 6-2. Summary of error matrix results for detection of Z. marina in five of the target
estuaries based on the ground truth surveys 164
Table 6-3. Estimates of Z. marina area in the target estuaries from the aerial/on-surface
surveys 166
Table 6-4. Cumulative percent area of Z. marina within four depth intervals relative to MLLW
in three PNW estuaries 169
Table 6-5. Relative distribution of the total Z. marina habitat area in the oceanic and riverine
segments of the estuaries based on the aerial surveys 171
Table 6-6. Percent coverage of Z. marina within the oceanic and riverine segments of the
estuaries based on the aerial surveys 172
Table 7-1. Sampling design for the probabilistic field surveys and areas of the probabilistic
frames 184
Table 7-2. Sediment characteristics of the target estuaries from the probabilistic surveys 190
Table 7-3. Percent of intertidal area with Z. marina estimated using different methods 195
Table 7-4. Percent of intertidal area with Z.japonica estimated using different methods 195
Table 7-5. Percent of intertidal area with green benthic macroalgae estimated using different
methods 200
Table 7-6. Biomass estimates for benthic macroalgae 200
Table 7-7. Relative cover of the five ecologically important benthic taxa and bare habitat in the
oceanic and riverine segments in the seven target estuaries 213
Table 8-1. Estuaries sampled for the lower depth limit of Z. marina and the datum adjustment
factor used to convert lower depth limit from the mean lower low water datum to mean sea
level datum 221
Xlll
-------
Table 8-2. Linear regression coefficients of maximum Z. marina depth versus distance from the
mouth of each of the seven target estuaries 224
Table 8-3. Overall mean depth, mean of deepest 1/3 of depth values, and maximum depth that Z.
marina was observed in each of the seven target estuaries 225
Table 8-4. Comparison of percent of surface irradiance needed to maintain Z. marina at its
maximum colonization depth from published data and from Yaquina Estuary data 239
Table B-l. Measurement quality objectives for data collected by Western Ecology Division. 276
Table B-2. Quality calibration and control checks for instruments and parameters 278
Table B-3. Accuracy and precision for nutrient analyses 282
Table B-4. Precision and accuracy for YSI Multiparameter Sondes 283
Table B-5. Precision and accuracy of 515N data 283
Table B-6. Precision and accuracy of TOC/N sediment analysis 284
Table B-7. Precision and accuracy of sediment particle size analysis 285
Table B-8. Data comparability between four field crew members for intertidal probabilistic
sampling 285
Table B-9. Image classification accuracy assessment 288
Table B-10. Bounding coordinates for the eight 2001 MRLC zones containing Pacific Coast
watersheds 293
Table B-l 1 Overall accuracy of NLCD 2001 land cover analysis 294
xiv
-------
EXECUTIVE SUMMARY
Increased anthropogenic nutrient loading and the subsequent eutrophication of coastal
ecosystems is a growing ecological and economic problem both in the United States and
globally. Eutrophication can result in a range of ecological impacts including hypoxic
conditions, fish kills, loss of submerged aquatic vegetation (SAV), degraded benthic conditions,
harmful algal blooms, and detrimental increases in benthic macroalgae. The nature and severity
of the impacts vary with the level of nutrient loading as well as with the estuary type and
regional drivers.
One tool to help address this problem is the development of classification schemes to allow
researchers and managers to extrapolate results from a limited number of well-studied estuaries
to the larger domain of estuaries within the same class. Several estuarine classification schemes
have been developed based on different approaches and endpoints. However, the ecological
reality of these classification schemes for the Pacific Northwest (PNW) is not clear, in part,
because of the limited baseline information available to evaluate the schemes. Additionally, the
available information gives "mixed messages" as to whether eutrophication is occurring in the
coastal PNW estuaries. Dissolved oxygen levels are generally high and chlorophyll a is
moderate to low, indicating a non-eutrophic condition. However, nutrient loading is high and
within the range of eutrophic estuaries on the East Coast. Additionally, benthic macroalgal
blooms, which have been used as a diagnostic indicator of eutrophication in other parts of the
world, are a seasonal event in the PNW.
To help determine the extent of eutrophication, the Pacific Coastal Ecology Branch (PCEB) of
the Western Ecology Division (WED) initiated a classification study of PNW estuaries. The
PNW was defined as including the coastal estuaries from Cape Mendocino in Northern
California (40.440°N) to Cape Flattery in Washington (48.383°N). Puget Sound was not
included in the present study. Two different approaches were used. The first component of the
research focused on a landscape analysis using existing data for the entire PNW. An inventory
of all of the PNW estuaries was generated based on the occurrence of estuarine habitat as defined
by the National Wetland Inventory (NWI), and all of the associated watersheds were delineated.
The PNW contains a surprisingly large number of estuaries, a total of 103 based on the present
NWI analysis. However, most of these estuaries are small, <1 km2, and only 13 of them are
>10 km2. These PNW estuaries appear to break out into seven general types based on
geomorphology, oceanic exchange, and riverine influence: coastal lagoons, blind estuaries,
tidally restricted coastal creeks, tidal coastal creeks, marine harbors/coves, drowned river mouth
estuaries, and bar-built estuaries. The first three of these have restricted connection with the
ocean on at least a periodic basis, and are potentially vulnerable to anthropogenic nutrient
additions and watershed alterations. Conversely, the marine harbors/coves have extensive
flushing and are presumably less sensitive to either nutrient enrichment or watershed alterations.
The drowned river mouth estuaries are sub-divided into tide-dominated, moderately river-
dominated, and highly river-dominated systems based on the extent of estuarine area or volume
weighted freshwater flow. Because of potentially higher exchange, the river-dominated estuaries
may be less sensitive to nutrient enrichment.
xv
-------
Geomorphology by itself did not cleanly separate groups with similar vulnerabilities among the
remaining 31 tide-dominated river mouth estuaries, river-dominated river mouth estuaries, and
bar-built estuaries. One approach taken to identify groups of estuaries was to evaluate spatial
and temporal salinity patterns based on historical data as well as a "normalized freshwater
inflow" index we developed. This index ranks estuaries by the relative amount of freshwater
entering the systems, and was used as a quantitative approach to separating tide- versus river-
dominated estuaries. Another approach was to conduct multivariate analyses to examine the
biotic and watershed similarities among the remaining 31 estuaries. The estuaries were clustered
based on the patterns of wetland habitats from the National Wetland Inventory (NWI) and
independently on land cover patterns in the associated watersheds. Several different
classifications were produced as well as a "crosswalk" between the two classifications to
determine which estuaries were grouped together in both analyses. Six pairs of estuaries were
similarly classified based on both wetland and land cover patterns, and these estuary pairs
potentially could form a framework for developing ecologically relevant nutrient criteria for the
PNW.
The second component of the research was to develop a classification scheme based on nitrogen
loading and sources, particularly as it related to the relative importance of oceanic- versus
riverine-derived nitrogen. Seasonal coastal upwelling from approximately April to October is
the major source of nitrogen and phosphorous to the near-coastal region. Previous research by
PCEB showed that high nutrient oceanic water advected into the Yaquina Estuary, Oregon was
the major nutrient source in the lower estuary during the dry season (May to October) and
riverine nutrients (i.e., terrestrially derived) were the dominant nutrient source only in the upper
Yaquina Estuary. In contrast, riverine inputs dominated through most of the estuary during the
wet season (November to April); however, there is little utilization of the wet season riverine
inputs due to low solar irradiance and short residence times associated with high river inflow.
Overlaying the distribution ofZostera marina L., the native seagrass species, over the spatial
pattern of nitrogen sources identified oceanic inputs as the major source for the bulk of the SAV
population in the Yaquina Estuary during the dry season.
To determine the generality of this pattern observed in the Yaquina Estuary, field studies were
initiated in seven target estuaries spanning a range of sizes (2.0 km2 to 54.9 km2) - Alsea, Coos,
Nestucca, Salmon River, Tillamook, Umpqua River, and Yaquina. We focused the field studies
on the dry season (May to October) because this is the primary period of biological nutrient
utilization. One objective was to use the nitrogen source model developed for the Yaquina
Estuary and site-specific water quality and stable isotope patterns in benthic macroalgae to
delineate the ocean- and river-dominated segments in a suite of estuary types. Results from this
component demonstrated that advection of high nutrient ocean water was the major nitrogen
source in the lower estuary for the seven target estuaries.
The next objective was to determine if the SAV populations and populations of other key
estuarine resources were primarily exposed to ocean- or river-derived nutrients. Five biotic
endpoints were evaluated, three primary producers, and two secondary consumers. The major
focus was on the perennial, rooted aquatic seagrass Z. marina, which was evaluated both by field
surveys and by aerial photography. In addition to Z. marina, probabilistic field surveys
evaluated the abundance and distributions of two additional benthic primary producers, the
xvi
-------
nonindigenous seagrass Zoster a japonica Aschers. & Graebn. and benthic green macroalgae.
The secondary consumers evaluated were two burrowing shrimp, Neotrypaea californiensis and
Upogebiapugettensis. One pattern that emerged from comparing across estuaries was that
macroalgal blooms occurred in all the estuaries, suggesting that it is a natural phenomenon, while
stable isotope data suggest the presence of these blooms is not an indication of eutrophication.
Another pattern was that the intertidal occurrence of the non-native Z. japonica exceeded that of
the native Z. marina in several of the target estuaries.
Overlaying the spatial distributions of these species on the spatial patterns of dry season nutrient
sources showed that the bulk of the Z. marina., benthic macroalgae and both burrowing shrimp
populations occurred in the oceanic segments of all seven estuaries. Thus, for two of the primary
producers, oceanic nitrogen is the dominant source during the principal growing season.
Likewise for the burrowing shrimp, the bulk of the populations would primarily be exposed to
oceanic nitrogen during the dry season. During the wet season, terrestrially derived nitrogen is
the major nutrient source for these estuaries. However, this is a period of low water column
chlorophyll a and Z. marina and macroalgal production suggesting that this terrestrially derived
nitrogen is not primarily driving production by these species.
The exception to the pattern of the primary producers occurring primarily in the lower estuary
was the non-native seagrass, Z. japonica, which was relatively abundant in both the oceanic and
riverine segments. Thus, this non-native seagrass would have a higher exposure to terrestrially
derived nitrogen than the native primary producers during the summer growing season. One
possibility that has yet to be explored is whether low levels of nutrient enrichment stimulate the
growth and establishment of this non-native seagrass.
The field surveys also evaluated the lower depth margin of Z. marina across the seven estuaries
and along estuarine gradients. The lower depth limit to which Z. marina grows decreases in the
upper estuary segments, which correlates with a general decrease in water clarity in the upper
segment of the estuaries. With further development, the lower depth limit of Z. marina could
potentially be used as an integrative indicator for assessing decreases in water clarity, as could
occur from nutrient stimulation of phytoplankton. Even though our research indicates that the
current levels of benthic macroalgae are natural, increases in macroalgae coverage or biomass in
the riverine segments could also be used as an indicator of nutrient enrichment especially if such
studies were coupled with measurements of stable isotopes (515N) to identify the nitrogen
sources associated with the blooms.
Results from our studies and a growing body of literature suggest several key points for the
management of nutrient enrichment in the PNW. First, environmental drivers such as coastal
upwelling strongly indicate that regional approaches to classification will be necessary to
generate ecologically relevant groupings. Second, PNW coastal estuaries are, in general, not
showing indications of cultural eutrophi cation. Additionally, the majority of the populations of
four of the five biotic resources we evaluated occurred in the oceanic segment of the estuaries,
indicating a lower vulnerability to terrestrially derived nitrogen. However, anecdotal
observations of phytoplankton blooms in the upper estuary segments of a few estuaries suggest
that the riverine segments of estuaries may be experiencing localized nutrient enrichment. Third,
the development of national estuarine nutrient criteria that do not take into account the naturally
xvn
-------
high nutrient concentrations resulting from upwelling are likely to result in numerous false non-
attainments of criteria in PNW coastal estuaries. Similarly, the development of Total Daily
Maximum Loads (TMDLs) for nutrients also needs to consider the effects of oceanic nutrient
loadings.
xvin
-------
CHAPTER 1: INTRODUCTION AND FRAMEWORK FOR THE
PACIFIC NORTHWEST REGIONAL CLASSIFICATION PROJECT
Henry Lee II and Walter Nelson
1.0 Scope
As part of the U.S. EPA's Aquatic Stressors Framework (U.S. EPA, 2002), a variety of
approaches to the classification of estuarine systems with respect to observed or predicted
responses to nutrient enrichment have been examined. This study explores approaches to
classification at a regional scale using distributions of submerged aquatic vegetation (SAV),
wetland classes, and other estuarine biological resources as classifying variables for Pacific
Northwest (PNW) estuaries. We also explore whether spatial and temporal patterns of salinity
and nutrient dynamics among PNW estuaries allow us to group them with respect to their
vulnerability to nutrient enrichment. This document is a revised version of a previous internal
EPA report (Lee et al., 2006), and includes new analyses and interpretations based on updated
wetland and salinity data. A companion study (Dettmann and Kurtz, 2006) explored the use of
empirical nutrient-SAV load-response models for estuaries of the New England region as well as
the response of phytoplankton biomass to nutrient concentrations in a suite of estuaries in the
eastern U.S. Together, these two studies provide assessments of a series of approaches to
estuarine classification that may be of value in setting national and regional nutrient criteria for
estuaries.
1.1 Problem Statement
Increased anthropogenic nutrient loading and the subsequent eutrophication of coastal
ecosystems is a growing ecological and economic problem both in the U.S. and globally (e.g.,
Nixon, 1995; NRC, 2000; Cloern, 2001; Bricker et al., 2003; Scavia and Bricker, 2006).
Eutrophication can result in a range of ecological impacts including hypoxic conditions, fish
kills, loss of SAV, degraded benthic conditions, harmful algal blooms, and detrimental increases
in benthic macroalgae. The nature and severity of the impacts vary with the level of nutrient
loading as well as the estuary type and regional drivers (e.g., Cloern, 2001; Bricker et al., 2003).
Ideally, each estuary would be evaluated independently as to the nature and extent of these
impacts and any mitigation/regulatory actions would be tailored to each estuary or watershed.
The reality is, however, that there are insufficient resources to conduct detailed scientific studies
on each estuary or to develop estuary-specific water quality criteria or management plans for
every coastal water body. Hence, approaches to reduce this complexity are needed.
One general approach to prediction across a range of water bodies is to use water quality models,
such as Basins (http://www.epa.gov/waterscience/basins/basinsv3.htm) or Sparrow
(http://water.usgs.gov/nawqa/sparrow/). While these models are a powerful approach to making
predictions in the better studied water bodies, they require extensive data input as well as
expertise, reducing their general utility. A different, and complementary, approach is to classify
estuaries that have similar responses to nutrient enrichment. If successful, such classifications
allow extrapolation from one estuary to others within the same class, thereby reducing the
amount of site specific data needed to make environmental decisions. Although lacking the
same level of site specific predictive ability as the complex water quality models, classification is
potentially a more practical approach to reducing the complexity inherent in extrapolating across
-------
a range of estuaries, especially for those regions with limited data. As pointed out by the
National Research Council, "a widely accepted estuarine classification scheme is a prerequisite
for a systematic approach to extending lessons learned and successful management options from
one estuary to another" (NRC, 2000, page 163).
A plethora of classification approaches have been applied to estuarine systems, with 26 different
classification frameworks identified in the review by Kurtz et al. (2006). Two classical
approaches are geomorphic classification, which classifies estuaries as drowned river valleys, bar
built, fjords, and tectonically formed estuaries (e.g., Pritchard, 1955, 1967), and the
hydrodynamic approach, which classifies estuaries by their circulation patterns and stratification
(e.g., Hansen and Rattray, 1966). Several recent efforts focused on grouping systems based on
physical attributes which may be relevant to expression of nutrient impacts, such as the EPA's
classification framework for coastal systems (U.S. EPA, 2004a) or the ASSESTS approach to
ranking the eutrophication status of coastal waters (Bricker et al., 2003). To build upon these
efforts, we initiated a classification research program focused on PNW estuaries. As described
below, this program was designed around both the potential management uses of an estuarine
classification scheme and the key environmental drivers in the PNW.
1.2 Regulatory Framework for Estuarine Classifications
Given the large number of approaches to estuarine classification, one of the early questions we
asked was how an estuarine classification scheme could be used in a management context,
specifically in relation to the development of water quality criteria or the management of excess
nutrients. That is, we viewed the initial question not as "How to classify?", but rather "Why
classify?" We identified five major approaches relating to the regulation of excess nutrient, as
summarized below. Using this framework and an assessment of the data needs and availability
in the PNW, we focused our research on management issues #2, #4, and #5 listed below.
1) Classification by Current Ecological Condition: A fundamental management need is
assessing the current condition of coastal water bodies, which is a type of classification when
multiple water bodies are compared. Such a comparison has been used to classify near-
coastal water bodies as having "good", "fair", or "poor" ecological condition at regional and
national scales in the National Coastal Condition Report (U.S. EPA, 2004b, 2006). Estuaries
can also be compared in terms of their attainment of designated-use criteria. Classifying
estuaries by their existing ecological condition requires: 1) a suite of ecologically relevant
indicator metrics that can be practically measured over a range of different water body types
and 2) field surveys that measure the metrics at the appropriate spatial and temporal scales.
2) Classification by Nutrient Loading and Sources: Estuaries can be classified by nutrient
loadings from non-point, point, and natural sources. Comparison of estuaries by the relative
magnitude of different sources is critical in developing effective management strategies to
reduce nutrient loadings (Driscoll et al., 2003) or, more fundamentally, evaluating the role of
anthropogenic versus natural loadings. Specifically, classification by loadings can help
prioritize remediation/enforcement efforts or develop Total Daily Maximum Loads (TMDLs).
The data needed to generate loading/source classifications include measurements of non-
point, point, and natural loadings across a suite of water body types with loading models used
to estimate missing values. These types of data are similar to those needed for a TMDL; the
-------
difference is that data to classify estuaries need to be taken over a suite of estuaries though
less extensive data collection per estuary can be used to classify estuaries compared to what
would be needed for regulatory actions. In addition to data on estuary nutrient concentrations,
classification by loading will usually require data on land cover in the associated watershed.
Accordingly, "Landscape conditions (e.g., % cover of land uses)" is listed as one of the
recommended core water quality indicators in EPA guidance for Sections 303(d), 305(b) and
314 of the Clean Water Act (http://www.epa.gov/owow/tmdl/2006IRG).
3) Classification to Derive/Validate Water Quality Criteria: Comparison of ecological
condition in a suite of estuaries along a nutrient gradient is one approach to deriving stressor-
response relationships (e.g., Latimer and Kelly, 2003). While more of a "natural experiment"
than a classification per se, the change in ecological condition in response to increased
exposure can identify nutrient thresholds. Such cross-estuary comparisons can also help to
identify the most sensitive endpoints and diagnostic indicators of nutrient stress. Perhaps the
greatest limitation of such natural experiments is the potential, and often unknown,
confounding factors when comparing across estuaries. Nonetheless, such cross-system
comparisons are a powerful tool in generating or validating water quality criteria under
realistic conditions. The necessary requirement for this approach is a suite of "reasonably"
similar estuaries that primarily differ in their nutrient concentrations or loadings. Finding
such a nutrient gradient can be challenging, as in some regions finding a true "reference"
estuary may be difficult while in other regions finding highly impacted estuaries may prove
challenging.
4) Classification by the Resources at Risk: A fundamental but often overlooked type of
classification is to categorize estuaries by the ecological resources at risk. The States
establish designated uses for water bodies as part of state water quality standards, and nutrient
or other protective criteria are determined to be able to meet these specific designated uses.
Given the need to protect designated uses, an evaluation of the resources at risk is an
important early step in the development of a water quality criterion, risk assessment, or
mitigation action. An obvious example is the application of nutrient criterion for SAV to
classes of estuaries devoid of SAV, which may be under- or overprotective of the actual
resources within such systems. More subtlety, the distribution of the resource within the
estuary may be a key factor in the exposure of the resource to anthropogenic nutrients or other
stressors. Classifying estuaries by their resources at risk requires: 1) identification of the
regionally high priority ecological resources; 2) identification of how these high priority
ecological resources are distributed within and across estuaries; and 3) evaluation of the
overlap of the resource(s) with anthropogenically derived nutrients or other stressors.
5) Classification by Estuarine Vulnerability: An approach related to classification by
resources at risk is to categorize estuaries by their inherent susceptibility to nutrient
enrichment (e.g., Bricker et al., 2003; U.S. EPA, 2004a). Identification of groups of estuaries
likely to display adverse impacts at low to moderate nutrient concentrations/loadings versus
those that have a higher assimilative capacity can assist managers in prioritizing monitoring,
remediation, or enforcement actions. To the extent that vulnerability is related to regional
drivers, such as climate or ocean conditions, a classification based on vulnerability can serve
as the framework for developing defensible regional water quality criteria. Predicting
-------
vulnerability is a complex process requiring consideration of natural and anthropogenic
nutrient concentrations/loadings, sensitivity of the specific endpoint(s), spatial/temporal
overlap of the endpoint(s) with excess nutrient concentrations, and mitigating or enhancing
factors. In many cases, it will be necessary to draw on data (e.g., dose-response relationships)
from other estuaries or from similar species.
1.3 Overview of Eutrophic Conditions and Nutrient Dynamics in the Pacific Northwest
Compared to the East and Gulf Coasts, relatively little baseline information on nutrient
enrichment is available for PNW estuaries. In the most comprehensive national review of
eutrophication, Bricker et al. (2007; also see Bricker et al., 2003) evaluated 138 estuaries across
the U.S., which included 17 coastal PNW estuaries exclusive of Puget Sound. Nine of these 17
estuaries were considered to have insufficient data for classification. Of the remaining eight
estuaries, two were classified as having low eutrophic conditions, six were considered to have
moderate low eutrophic conditions, and none were classified as having high eutrophic
conditions. Several Pacific estuaries were classified as having high eutrophic conditions in
central and southern California and in Puget Sound, but these are all in different biogeographic
ecoregions than the PNW estuaries (Spalding et al., 2007).
Water quality data from U.S. EPA's Coastal Environmental and Monitoring Assessment
Program (EMAP) 1999 and 2000 surveys can also be used to evaluate water quality on the
Pacific Coast. The 1999 survey sampled the "small" estuaries of California, Oregon, and
Washington while the 2000 survey sampled Puget Sound, San Francisco Estuary, and main stem
of the Columbia River Estuary. In 200 water quality samples from the 1999 survey, the lowest
dissolved oxygen (DO) value was 3.8 mg I"1 and less than 4% of the area of the small coastal
estuaries had DO concentrations less than 5 mg I"1 (Nelson et al., 2005b), the level indicative of
biological stress (Bricker et al., 2003). Even when the more urbanized estuaries from the 2000
survey were included, only two sites out of 371 stations had DO values <2 mg I"1 (U.S. EPA,
2004b), the level indicative of hypoxia (Bricker et al., 2003). Both of these low values occurred
in Hood Canal in southern Puget Sound, which is a deep fjord type estuary with limited
recirculation (http://www.hoodcanal.washington.edu). Likewise, chlorophyll a concentration on
the Pacific Coast was rated as good, and only two sites in Puget Sound and one in California
exceeded the threshold of 20 |ig I"1 for a "poor" rating (U.S. EPA, 2004b). The generally high
DO values and low to moderate levels of chlorophyll a suggested that eutrophi cation was not a
wide-spread problem on the Pacific Coast, in particular in the PNW coastal estuaries.
Even though the available evidence did not suggest that PNW estuaries were displaying
symptoms of eutrophication, there are indications of high nutrient conditions. Water quality on
the Pacific Coast was rated as "fair" because of high dissolved inorganic phosphorous (DIP) and
poor water clarity (U.S. EPA, 2004b). Nitrogen loading normalized to estuarine area for the
Yaquina Estuary, Oregon is 25 mmole N m"2 d"1 (Brown and Ozretich, 2009), which is as great
as or greater than loading in a number of eutrophic systems (Nixon et al., 2001). Though
phytoplankton blooms do not appear to be a wide-spread phenomenon, we have observed
phytoplankton blooms in the upper regions of a few estuaries as well as the import of high
chlorophyll a concentrations (> 40 ug I"1) from the ocean (Brown and Ozretich, 2009). Finally,
benthic macroalgae, an indicator of eutrophication (CENR, 2003; Bricker et al., 1999), is
abundant in several PNW estuaries (e.g., Thorn, 1984; Kentula and DeWitt, 2003). Thus, the
-------
emerging pattern for the PNW is high DIP concentrations and high nitrogen loadings, high levels
of benthic macroalgae yet low or moderate phytoplankton concentrations. Similar cases of high
nutrient loading with low chlorophyll a levels have also been observed in other Pacific Coast
estuaries (Cloern, 2001), and certain estuaries appear less sensitive to nutrient loading due to
"modulating filters", such as strong tidal forcing, high turbidity, and dominance of benthic
suspension feeders.
One key factor affecting PNW estuaries is the seasonal upwelling that occurs during the spring-
summer (Hickey and Banas, 2003). Coastal upwelling is such a strong regional driver that the
PNW coastal estuaries can be considered "extensions of the coastal ocean during the growing
season" (Hickey and Banas, 2003). Another key factor is the highly seasonal pattern in
precipitation, with dramatically reduced rainfall during the summer (Emmett et al., 2000). This
pattern of increased coastal nutrient concentrations and reduced runoff potentially increases the
relative contribution of nutrients from oceanic sources during the summer. Field and modeling
research (Brown et al., 2007; Brown and Ozretich, 2009) showed that nutrient-rich ocean water
advected into the estuary was the major source of nutrients and phytoplankton for the lower
portion of the Yaquina Estuary during the spring and summer. Overlaying the distribution of
seagrasses on the pattern of nitrogen sources within the estuary identified oceanic-derived
nitrogen as the major source for the major portion of the SAV population in the Yaquina during
the primary growing season. During the low river flow conditions of summer, riverine nutrients
(i.e., terrestrially derived) were the dominant nitrogen source only in the upper Yaquina, which
contains relatively little SAV. Consequently, the SAV in this estuary does not appear to be
highly vulnerable to anthropogenic nitrogen increases during the primary growing season
because the bulk of the population only has minor exposure to watershed-derived nutrients.
These observations highlight the major differences in the nutrient dynamics of PNW estuaries
compared to East and Gulf Coast estuaries largely resulting from summertime upwelling and
reduced summertime river flow in the PNW. These differences could have substantial effects on
regulatory strategies. For example, what is the ecological relevance and defensibility of water
quality criteria in systems dominated by natural nutrient sources that periodically exceed the
criteria? Another issue is how to develop criteria or implement TMDLs in estuaries with
seasonally, and even tidally, variable fluxes of nutrients. One difficulty in addressing these
issues is that research on the effects of upwelling on nutrient and phytoplankton dynamics in
coastal estuaries has been limited to a relatively few PNW estuaries, primarily the Columbia
River Estuary (e.g., Hamilton, 1984), Coos Estuary in Oregon (Fry et al., 2001; Roegner and
Shanks, 2001; Roegner et al., 2002), Willapa Estuary in Washington (e.g., Hickey et al., 2002;
Hickey and Banas, 2003; Newton and Horner, 2003), and Yaquina Estuary in Oregon (Brown et
al., 2007; Brown and Ozretich, 2009). In particular, the pattern of the SAV being primarily
exposed to ocean-derived nutrients has only been studied in the Yaquina Estuary, and without
studies in different types of systems it is not clear whether this is a general pattern across PNW
estuaries.
1.4 Biotic Response Measures in PNW Estuaries
In addition to the water quality patterns mentioned above, PNW estuaries are characterized by
several dominant benthic macrophytes and secondary consumers. The dominant intertidal
macrophytes in the PNW are the perennial, rooted aquatic seagrasses Zostera marina L. and
-------
Z.japonica Aschers. & Graebn. and the benthic green macroalgae guild, where a guild is a set of
species with similar ecological function. Two dominant benthic species in many PNW estuaries
are the burrowing shrimp Neotrypaea californiensis and Upogebia pugettensis. Each of these
primary producers and secondary consumers strongly affect other components of estuarine
ecosystems, and can be considered ecological engineering species (sensu Jones et al., 1997).
Specific ecological functions and interactions of these species/guilds are summarized below and
in Table 1-1. Because of their importance in nutrient dynamics and food webs, these five taxa
were chosen as practical biotic response measures, or endpoints, in our surveys to relate
distributions of estuarine resources to nutrient sources.
Z. marina is the dominant seagrass species in PNW estuaries, forming dense beds in the
intertidal and shallow subtidal zones (e.g., Kentula and Mclntire, 1986; Thorn, 1990; Thorn et
al., 2001). Seagrasses promote estuarine productivity and diversity by serving as nurseries and
foraging areas for recreationally and commercially important fish and shellfish, providing bird
habitat, increasing the density of benthic invertebrates, and contributing to estuarine primary
production (e.g., Bayer, 1979; Hemminga and Duarte, 2000; Jackson et al., 2001; Ferraro and
Cole, 2007). A second seagrass species, the nonindigenous Z. japonica, was targeted because it
appears to be spreading in the PNW (e.g., Young et al., 2008). The ecological role of this non-
native species is not well known, but appears to have both positive and negative ecological
impacts (Posey, 1988; Larned, 2003; Wonham, 2003; Ruesink et al., 2006; Kaldy, 2006), and
thus can be considered both a stressor and an ecological resource.
Another important primary producer assemblage in PNW estuaries is benthic macroalgae, which
primarily consists of green ulvoid species. Benthic macroalgae were targeted in our field
surveys because blooms of Ulva and Enteromorpha have been associated with eutrophication in
other parts of the world (e.g., Valiela et al., 1997; Hauxwell et al., 2003) and have been used as
an indicator of eutrophi cation (Bricker et al., 1999; CENR, 2003). However, benthic macroalgal
blooms are an annual event in many PNW estuaries and are likely to be an important component
of primary production (e.g., Phillips, 1984; Thorn, 1984; Kentula and Mclntire, 1986; Kentula
and DeWitt, 2003; Thorn et al., 2003; Bulthuis and Shull, 2006). Thus, as with Z.japonica,
benthic macroalgae can be considered both a resource and a stressor.
In addition to the primary producers, the ghost shrimp, N. californiensis, and the mud shrimp,
U. pugettensis, were selected as biotic indicators. Both species are harvested commercially and
recreationally though they also have a negative effect on oyster aquaculture (Feldman et al.,
2000). Both of these deep burrowing ecological engineers modify sediment properties by
turning over large quantities of sediment and by irrigating large amounts of water through the
sediment (DeWitt et al., 2004). In response to this intense bioturbation, sediment nitrogen fluxes
are increased and shrimp-dominated tide flats are second only to the ocean as a source of
dissolved inorganic nitrogen (DIN) during the summer in the Yaquina Estuary (DeWitt et al.,
2004). While potentially stimulating phytoplankton by adding nutrients to the water column,
burrowing shrimp simultaneously reduce phytoplankton concentrations by filtering large
quantities of overlying water, perhaps as much as the entire volume of the Yaquina Estuary daily
(Griffen et al., 2004). The intense bioturbation by N. californiensis also reduces the ability of Z.
japonica to spread into unvegetated areas (Dumbauld and Wyllie-Echeverria, 2003).
-------
Table 1-1. Ecological interactions of five taxa evaluated in the field surveys. The effects listed here are to illustrate the general types
of interspecific and trophic interactions expected in Pacific Northwest estuaries and are not a complete listing of all biotic interactions.
TAXON
Z. marina
Z. japonica
Benthic
macroalgae
Neotrypaea
Upogebia
EFFECTS ON
WATER
COLUMN
NUTRIENTS
Decrease by
uptake and
potential export
from estuary via
rafting
Decrease by
uptake and
potential export
from estuary via
rafting
Decrease by
uptake (release
during decay)
Increase benthic
flux
Increase benthic
flux
EFFECTS ON
PHYTOPLANKTON
Compete for water
column nutrient. Minor
effect in PNW estuaries?
Compete for water
column nutrient. Minor
effect in PNW estuaries?
Compete for water
column nutrient. Minor
effect in PNW estuaries?
Minimal
Decrease by feeding
EFFECTS
ON
Z. MARINA
NA
Competitor
in some
estuaries
Impact at
high biomass
Inhibits
colonization
Neutral or
may facilitate
colonization
EFFECTS ON
BENTHIC
INFAUNA
Structure enhances
species richness and
density over bare
sediment. Provides
total organic carbon
(TOC) to sediment.
species richness and
density over bare
sediment. Provides
TOC to sediment.
marina!
Provides TOC to
sediment.
Detrimental at high
levels due to
smothering, shading,
and H2S formation
Reduces certain
benthic guilds
Enhanced relative to
bare sediment
EFFECTS ON
JUVENILE
FISHES &
CRABS
Nursery
Nursery, though
to lesser extent
thanZ marina!
Potential "habitat
sink" for juvenile
crabs during
spring tides?
Minimal?
Minimal
Minimal
EFFECTS ON
HIGHER
TROPHIC
LEVELS
Habitat refuge
and foraging for
fish and crabs.
Consumed by
brant geese &
American
widgeon
Habitat refuge
and foraging for
fish and crabs?
Consumed by
brant geese &
American
widgeon
Consumed by
some amphipods
Prey
Prey
EFFECTS OF
NUTRIENT
ENRICHMENT
ON TAXON
Increase in turbidity
and increase in
epiphytes reduce
lower depth limit of
Z. marina.
Increase in turbidity
and increase in
epiphytes reduce
lower depth limit?
Stimulate growth at
lower enrichment
level?
Increase area
covered and
biomass, with
impacts on SAV and
benthos
Decrease if low
dissolved oxygen
Stimulate due to
higher
phytoplankton at
moderate levels?
Decrease if low
dissolved oxygen
-------
1.5 Design and Objectives of the Pacific Northwest Regional Study
Based on our understanding of the nutrient dynamics and distribution of estuarine resources
within the PNW, we focused our research on three of the previously identified management
needs: 1) classification by nutrient loading and sources; 2) classification by the resources at risk;
and 3) classification by estuarine vulnerability. Research to address these needs was conducted
at two spatial scales. The first was at the landscape scale of entire estuaries and associated
watersheds within the PNW, which is discussed in Chapter 2. The six primary objectives at the
landscape scale were:
1) Identify and delineate all estuaries and associated watersheds within the PNW.
2) Classify estuaries by geomorphology and extent of river versus ocean influence.
3) Classify estuaries by spatial and temporal salinity patterns
4) Classify estuaries by wetland patterns.
5) Classify coastal watersheds by land cover patterns.
6) Integrate wetland and watershed classifications to identify functionally similar estuaries.
A total of 103 estuaries were identified within the PNW based on the most recent NWI data, with
their sizes spanning more than five orders of magnitude. A variety of classification schema were
generated from these landscape-scale attributes based on the concept that different types of
classifications are useful for addressing different scientific or management issues. A
"crosswalk" of the wetland and land use classifications was then generated to identify
functionally similar estuaries.
The second spatial scale was at the level of the individual estuary where we conducted both field
and aerial surveys. We identified seven target estuaries in Oregon spanning a range of size (3.1
to 54.9 km ), geomorphology, and perceived ocean versus riverine influence. The seven target
estuaries were the Alsea, Coos, Nestucca, Salmon River, Tillamook, Umpqua River, and
Yaquina. One of the goals at the estuary scale was to determine patterns and drivers of water
quality across a range of estuary types. The specific objectives of the water quality surveys and
modeling were to:
1) Determine the within-estuary and among-estuary spatial patterns of water quality based on
nutrients, chlorophyll a, and dissolved oxygen (Chapter 4).
2) Assess the among-estuary patterns of spatial and temporal variation in salinity (Chapter 4).
3) Based on water quality patterns, models, and isotope ratios in macroalgae, divide each of
the estuaries into segments that are primarily dominated by ocean-derived nutrients versus
riverine-derived nutrients (Chapter 5).
-------
Another goal at the estuary scale was to quantify the within- and among-estuary distributions and
abundances of each of the five biotic endpoints. Specific objectives of this research were to:
1) Determine Z. marina'?, intertidal/shallow subtidal bathymetric distribution using aerial
survey data (Chapter 6).
2) Using the aerial surveys of Z. marina distributions, assess the potential exposure to
terrestrially derived nutrients by quantifying what proportions of the population occurred in
the ocean- and river-dominated segments of the target estuaries (Chapter 6).
3) Based on field probabilistic surveys, assess the potential exposure of each of the five biotic
endpoints to terrestrially derived nutrients by quantifying what proportions of the populations
occurred in the ocean- and river-dominated segments of the target estuaries (Chapter 7).
4) Classify the target estuaries based on similarities in the relative abundances of the five
biotic endpoints (Chapter 7).
5) Determine the lower depth limit of Z. marina to determine its relationship to ambient water
clarity and estuary type (Chapter 8).
The final objective was to summarize the key patterns and processes that differentiate PNW
estuaries from those on the East and Gulf coasts as well as those that result in different
vulnerabilities to nutrient enrichment among PNW estuaries (Chapter 9). We suggest that this
summary forms the nucleus of an emerging "Pacific Northwest Paradigm". This paradigm can,
and should, be refined as additional data are collected and as we gain better insights into the
oceanic, estuarine, hydrologic, climatic, watershed, and ecological processes affecting nutrient
dynamics and biotic distributions.
-------
CHAPTER 2:
REGIONAL CLASSIFICATION OF PACIFIC NORTHWEST ESTUARIES
BY WETLAND AND LAND COVER PATTERNS
Henry Lee II, Deborah A. Reusser (USGS), Patti Haggerty (Indus Corp.),
Cheryl A. Brown, and Patrick J. Clinton
Key Findings
A census of the coastal waterbodies on the Pacific Coast identified 230 estuaries
(exclusive of Puget Sound) of which 103 occurred in the Pacific Northwest
(PNW). The majority of the PNW estuaries are small, with 73 of the 103
estuaries less than 1 km2, and most have extensive intertidal habitat.
• PNW estuaries provide a suite of estuarine services and functions that vary by
estuary size, with small- and moderate-sized estuaries (< 10 km2)
disproportionately important to Coho smolts.
• PNW estuaries were classified based on geomorphology and extent of ocean
exchange. Eight estuaries were classified as coastal lagoons, 6 as blind estuaries,
53 as tidally restricted coastal creeks, 2 as tidal coastal creeks, 3 as marine
harbors/coves, 3 as bar-built estuaries, and 28 as drowned river mouth
estuaries.
• The drowned river mouth estuaries were further divided into river- or tidal-
dominated based on an "area-normalized freshwater inflow" metric, with 15
classified as "highly river-dominated", 7 as "moderately river-dominated", and
6 as tide-dominated estuaries. River-dominated estuaries have greater flushing
and, hence, are less vulnerable to nutrient enrichment.
• The 31 major river mouth and bar-built estuaries were then classified using a
clustering approach based on either wetland or watershed land cover patterns.
• A "crosswalk" was conducted to group estuaries that clustered together in both
the wetland and watershed analysis. This approach identified groups of
estuaries that presumably have similar nutrient dynamics and vulnerabilities.
• Most PNW estuarine watersheds display relatively low levels of alteration based
on development, agriculture, population density, and impervious surfaces.
2.0 Introduction
The overall goal of the regional classification was to evaluate the similarities and differences
among Pacific Coast estuaries based on estuarine wetland distributions, land cover patterns in the
10
-------
associated coastal watersheds, and spatial and temporal patterns of salinity distributions.
Wetland and watershed data were synthesized for all the coastal estuaries of the Pacific Coast
from the Tijuana Estuary in Southern California (32.5574°N) north to Cape Flattery in
Washington (48.383°N). The current effort focuses on the outer coastal estuaries of the Pacific
Northwest (PNW), which is defined as the coastal segment from Cape Mendocino in Northern
California (40.440°N) to Cape Flattery in Washington. The Strait of Juan de Fuca and Puget
Sound were not included in this analysis.
The first step in the analysis was a comprehensive delineation of all the coastal estuaries and
watersheds to assure a one-to-one relationship between each waterbody and its associated
watershed. Based on this inventory of PNW estuaries and watersheds, we addressed the
following objectives:
1) Identify the broad resource types and general ecological services provided by PNW
estuaries.
2) Group estuaries by their similarities in wetland patterns and estuarine landscape
attributes, including development of estuarine-scale metrics of key physical/climatic
drivers.
3) Group watersheds by their similarities in land cover patterns and other watershed
attributes.
4) Conduct a matrix match ("crosswalk") of the classifications by wetlands and watersheds
to identify estuaries overlapping in the two analyses. Use these groups to identify
functionally similar estuaries and to help develop a research framework to evaluate the
proposed classification schema.
5) Develop a baseline of estuarine and watershed landscape data for the PNW.
2.1 Estuarine, Wetland, and Landscape Data Sources and Methods
Achieving the five objectives of the regional classification required a synthesis of multiple types
of estuarine and watershed data and GIS analyses. Figure 2-1 provides an overview of this
synthesis to achieve objectives 1-4. In terms of the fifth objective, the summarized wetland and
watershed data are provided in Tables 2-2, 2-3, and 2-4.
2.1.1 Estuary Definition and Inventory
A national standard, the National Wetland Inventory (NWI) (http://www.fws.gov/nwi/; U.S. Fish
Wild. Sen, 2002), was used as the criterion for defining estuaries. NWI classifies aquatic habitat
types using a hierarchical set of attributes based on Cowardin et al. (1979), and includes marine,
estuarine, riverine, palustrine, and lacustrine areas with further subdivisions for tide height,
substrate type, and the presence of broad classes of wetland plants (e.g., emergent vs. aquatic
bed). These geospatial data were obtained from NWI digital databases and on-site digitization of
paper maps to fill in data gaps not available in digital format. The NWI codes found in PNW
estuaries are summarized in Table 2-1.
11
-------
Delineate
Watersheds
Section 2.1.3
Inventory of PNW Estuaries
Section 2.2
Ecological Resources
of PNW Estuaries
Sections 2.1.6 & 2.3
Classification Based on Geomorphology
and Oceanic Exchange
Section 2.4
Open
Marine Harbors
and Coves
Section 2.4.3
Restricted
Lagoons and Blind Estuaries
Section 2.4.1
Tidally Restricted Coastal Creeks
Section 2.4.2
Drowned River Mouth and Bar Built Estuaries
Section 2.4.4
Classification
by Watershed
Alterations
Section 2.8
Tidal vs. River
Dominated Estuaries
Section 2.4.4
Classification by
Land Use and %
Land Use Areas
Section 2.6
Classification by
Wetland Areas and
% Wetland Areas
Section 2.5
Crosswalk Between Landscape and
Wetland Classifications
Section 2.7
Figure 2-1. Overview of the generation of the PNW estuary inventory and the wetland and
landscape data synthesis and analysis.
12
-------
We defined estuaries as coastal water bodies that had a NWI estuarine polygon and that
discharged directly into the ocean. Rivers, creeks, tributaries, and other water bodies within an
estuary that do not discharge directly into the ocean were considered subestuaries. While only
the estuarine polygons were used to identify estuaries, the sum of the marine, estuarine, and tidal
riverine polygons was used to calculate the total estuarine area so as to capture the entire habitat
area likely to contain euryhaline flora and fauna as well as to minimize the effects of any
misclassifications of the salinity classes. The list of estuaries in the PNW was updated from
those in the earlier version (Lee et al., 2006) based on the 2007 NWI revisions in the Pacific
Northwest. In addition, any coastal water body that had an estuarine polygon in the 2001 NOAA
coverage (see Section 2.1.3) but did not have an estuarine polygon in NWI was identified and
delineated. These additional systems are listed for completeness but are not included in the
wetland or watershed analyses. Semi-enclosed harbors or bays with only marine polygons and
coastal streams with only tidal riverine polygons were delineated but were not classified as
estuaries and were not included in the analyses.
Based on these criteria, 103 PNW estuaries were identified (Figures 2-2 to 2-5). Sixteen of these
estuaries were in California, 24 in Washington, 62 in Oregon, and the Columbia River Estuary
which is divided between Washington and Oregon. The geomorphological class of each of these
estuaries is given in Table 2-2. The classification of estuaries as drowned river mouth, bar-built,
and blind comes from various sources (e.g., Bottom et al., 1979; Seliskar and Gallagher, 1983;
Cortright et al., 1987; Rumrill, 1998; Emmett et al., 2000) as well as our own analysis. The
classification of estuaries as coastal lagoons, tidally restricted coastal creek, and marine
harbor/coves is based on the arguments made in Section 2.4 as is our approach to quantitatively
separating tide- versus river-dominated estuaries.
2.1.2 Wetland Habitats Based on NWI
The NWI provides a regional-scale dataset to classify estuaries by wetland type. An advantage
of using wetlands to classify estuaries is that a biotic/habitat endpoint integrates a wide range of
environmental conditions, such as salinity patterns, flushing, and nutrient loading. Thus,
estuaries with similar NWI wetland patterns presumably will display similar responses to
nutrient enrichment.
While the NWI provides a dataset to evaluate estuarine habitats at both local and regional scales,
there are limitations. The NWI classifies habitats at a coarse resolution, such as "aquatic bed" or
"unconsolidated shore", and is thus unable to differentiate between benthic assemblages such as
burrowing shrimp beds versus sand flats without shrimp. A second limitation is that some of the
NWI data were generated in the late-1970s and early-1980s, and thus are historical snapshots of
estuarine conditions. However the NWI is continuing to update its habitat maps, including an
update of the Oregon estuaries in 2007. The list of estuaries and their sizes are based on these
updated data so that some estuarine areas will differ slightly from the values in the earlier version
of this document (Lee et al., 2006). A third limitation is that there was minimal field validation
and some of the classifications may be incorrect. For example, our observations in Beaver Creek
(Oregon) suggest that one potential error is the classification of tidal riverine wetlands as
estuarine emergent wetlands in smaller coastal systems.
13
-------
126°0'fr
124°0'W
123°0'W
122°OW
f-
ss
N
Cape Flattery
Waatch River ' Y"
SooesRiver :' '..-..
Ozette River. ,
Strait of Juan cle Fuca
Quillayute River
Goodman Creek
Hob River
Whale Creek
Mosquito Creek TtTedar Creek
, Kalaloch Creek
.Queets River
Raft River
Camp Creek :jj-flch Cr^k
Quinault River "~ -----
•»i i- «• "Wreck Creek
Moclins River " „ ,
1 Joe (. reek
Boone Creek -"
C°PalisRlver; Conner Creek
Grays Harbor Estuarv:..•- " .."'%v""-•---- -*~
Willapa Estuary1
Lootnis Lake
Columbia River Estuary
Clatsop Spit
Puget Sound
Washington
ifrlM
0 12.$, 25
SO
100
Figure 2-2. Estuaries along the Washington coast from Cape Flattery to the Columbia River.
-------
125°0'W
124°0'W
I
123°0'W
i
122°9'W
***"
?~
MS
>
1
G '
**j
"g*
>
"H
o ™
W;
**
*f'
^fc
ff-, «•
o
>fc
sf
s
i^ -
1-5*; *"
, * *^"^%
i 1 L-J \
^ ,r- .X
Columbia River Estuary " '«'^''*"
"H Clatsop Spit
\ Ci
Indian Point ,/-
Chapman Po*nty Eopjg
I
Cove Beach
Short Sanrt Creek
Rockaway Beach Creek Oualk*
Smith LakeJ Moctaw»fCIe»rLak*
>f
Tillamook Estuarm <»,
*'f ^&as
Netarts Bay |j *?-
Sand Lake
„ , , , Nestucca River i MHMfreelc
Daley Lake fc lljlies ^-»«
Neskowin Creek^
Salmon River K,
I
Siletz Estuarr •'"' _ „ , ,, _ ,
' - -f
Depoe Bay > ''
Yaquina Estuary --•, ^
;"
Alsea Estuarv t.-.
»chats River
Tennilfc
;:-L:
Sutton Creek ''«enyCi«ek
!
i
i
i
•
v
>
\
* ^--^ _
| s*~*~' " x
%, Jl^
-Ifci x-5^^
Oregon
^
0 12,1 2J 5* 73 Ui
Figure 2-3. Estuaries in northern Oregon. The Alsea, Nestucca, Salmon River, Tillamook, and Yaquina estuaries were five of the
seven target estuaries in the field surveys.
-------
126°0'W
12S°0'W
124°0'W
i
123°0'W
122°OW
N
Sutton Creek
Siuslaw River f
Siltcoos River,'
"
Umpqua River
Coos Estuarv •&
Coquille River
Elk River (OR)
Port Orford Head '•..
Hubbard C'reek
Mussel Creek
Rogue River
i" Twoiall*
- New
River
-
Hunter Creek
Mveri Creek
at
Pistol River
Burnt HiH Creek
Thomas Creek
Whaleshead Creek \
X
River (CA)|~~ " ~~™
Lake Earl /
» 12.5 25
•tot
75 li§
Figure 2-4. Estuaries in southern Oregon. The Umpqua River and Coos estuaries were two of the target estuaries in the field surveys.
-------
125
124"0'W
123°0'W
122°0'W
Thomas Creek I
Whaieshead Creek ^ CJieteo
"
Smith Mfver (CA)£~~ ' " "~
Lake Earl ;
Crescent City Harbor "\
Klamath River
.
Squashan Creek j
Redwood Creek
f""
Big Lagoon (N) i
(,a
i
Mad River i
J*"T
Humboldt Estuary jf--^
Eel River £
Bear Wver
Cape Mendocino \
* 12.5 25 Si 75
tkin
Figure 2-5. Estuaries in Northern California, north of Cape Mendocino.
-------
One procedural challenge was that some of the older maps either had obsolete or incorrect NWI
codes, though we were able to translate these into the current standards. Another challenge was
the inconsistent resolution in the use of special habitat modifiers for salinity and sediment type
among estuaries. For example, codes to indicate the specific salinity class (e.g., mesohaline)
were included in a few estuaries but were lacking in many others. When conducting statistical
clustering using the original NWI codes habitats with a higher level of resolution would be
considered different from the exact same habitat that did not contain the special modifiers.
We took a number of steps to reduce the effects of these limitations on the classifications. As
mentioned, we updated obsolete or incorrect codes; the updated codes are referred to as
"corrected NWI codes." To minimize the effect of misclassification of riverine tidal wetlands as
estuarine wetlands, we defined total estuarine area as the sum of marine, estuarine, and tidal
riverine polygons. Thus, the total estuarine area would be the same regardless of the specific
salinity designation within this range. Including the riverine tidal portion is ecologically
justifiable as inclusion of the low salinity habitat captures the full habitat range for euryhaline
and oligohaline species. To control for different levels of habitat resolution, we derived
"consolidated NWI codes" that merged the original and corrected NWI codes into broader
habitat classes (Table 2-1). For example, both "regularly flooded emergent wetland" and
"regularly flooded rooted aquatic bed" were consolidated into a "regularly flooded rooted" class.
The original 118 marine, estuarine, and tidal riverine NWI codes in the PNW estuaries were
merged into 48 consolidated codes (Table 2-1). These consolidated NWI codes were used in the
classification analysis to assure similar levels of resolution in the clustering. Additionally,
merging into broader habitat classes helped to minimize the effects of certain wetland
misclassifications (e.g., a site classified as emergent vegetation versus a rooted aquatic bed).
Merged NWI codes have been used in previous efforts, such as in NOAA's "Spatial Wetland
Assessment for Management and Planning" program (SWAMP; Sutler, 2001) and in wetland site
prioritizations (Brophy, 2003). Brophy (2003) suggested that merging NWI polygons better
identifies the "high biological value of a large, hydrologically interconnected wetland."
With these steps to standardize the data, the NWI should provide sufficient accuracy and
resolution to address regional scale wetland patterns. As stated in Oregon's "Wetlands Inventory
User's Guide" (Oregon Division of State Lands, no year), "The NWI provides excellent
information for a variety of planning purposes" which include the identification of "the general
location, extent, and type of wetlands on a regional basis, such as watershed or on tribal lands."
Also, as pointed out in the Oregon Watershed Assessment Manual (Brophy, 2007), the revised
NWI is a suitable base layer for estuarine assessments and is an alternative to hydrogeomorphic
(HGM) maps. Nonetheless, the NWI does not have the resolution of detailed site-specific
studies, such as conducted previously in Oregon (Cortright et al., 1987). It is beyond the scope
of this regional assessment to conduct a detailed comparison of the NWI wetland classes with the
previous habitat delineations in Oregon, but a cursory comparison suggests that there is
reasonable agreement on broad habitat classifications (e.g., aquatic beds, subtidal). The Oregon
study does, however, provide details as to specific habitat type, such as the importance of algae
growing on cobble/gravel in the Chetco River, which provide insights not possible from the
broader NWI classes.
18
-------
Table 2-1. Consolidation of National Wetland Inventory (NWI) habitat codes found in PNW estuaries. The NWI broad habitat
classes (estuarine, marine, tidal riverine) and tidal heights (subtidal, irregularly exposed, regularly flooded, irregularly flooded) are
given at the top in capitals and the consolidated habitat types falling within that salinity-tide height are given in the cells underneath in
bold. The original NWI codes making up the consolidated class are listed beneath the consolidated habitat. Original NWI codes that
are obsolete or incorrect are given in italics with our interpretation of the corrected code in parenthesis. The "rooted" classes in this
analysis are used to capture the NWI "EM" codes (emergent = "erect, rooted, herbaceous hydrophytes") and any other codes that are
identified as rooted or persistent. The "aquatic bed" classes are used to capture all "AB" codes (aquatic bed = "habitats dominated by
plants that grow principally on or below the surface of the water for most of the growing season in most years") other than those
identified as rooted or persistent. In PNW estuaries, areas classified as aquatic beds may be covered with SAV, macroalgae, or a
combination of both. Because of the uncertainty of the presence of vegetation with the "unconsolidated shore" codes, they are
classified as unvegetated though they may contain up to 30% vegetation. Discussion of NWI codes can be found in Cowardin et al.
(1979) and Smith (1991) while an online translator is available at http://www.fws.gov/wetlands/Data/webatx/atx.html.
ESTUARINE - SUBTIDAL
Unvegetated
E1UBL,
E1UB2L, E1UB3L,
EIOWL(EIUBL)
Unvegetated -
Excavated
ElUBLx
Unvegetated -
Diked/
Impounded
ElUBLh
Aquatic Bed
E1ABL
Rooted
E1AB3L
ESTUARINE - IRREGULARLY EXPOSED
Unvegetated
E2USM, E2US2M,
E2US3M, E2FLM
(E2USM)
Unvegetated -
Cobble-Gravel
E2UB1M
(E2US1M)
Unvegetated -
Streambed
E2SBM
Aquatic Bed
E2ABM,
E2AB/FLM
(E2ABM),
E2AB/USM
Rooted
E2AB3M,
E2EM1M
-------
ESTUARINE - REGULARLY FLOODED
Unnvegetated
E2USN, E2BBN
(E2USN), E2FLN
(E2USN), E2US2N,
E2US3N, E2FL3N
(E2US3N), E2FL6N
(E2USN)
Unvegetated -
Streambed
E2SBN
Unvegetated -
Spoils
E2USNS
Unvegetated -
Excavated
E2USNx
Aquatic Bed
E2AB1N,
E2EM/AB1N,
E2AB1/US3N,
E2US3/AB1N,
E2ABN,
E2AB/FLN
(E2ABN),
E2AB/USN,
E2AB2/US3N,
E2US/ABN
Rooted
E2EMN,
E2EM/ABN,
E2EM1N,
E2EM1/FLN
(E2EM1N),
E2EM1/US3N,
E2EM1/ABN,
E2EM/USN,
E2EM1/USN,
E2EM5N
Rooted -
Diked/
Impounded
E2EM5Nh
Rooted -
Spoil
E2EMNs
Rooted -
Excavated
E2EMNx
ESTUARINE - REGULARLY FLOODED (CONT.)
Shrub-Scrub
E2SSN
Rocky
E2RSN
ESTUARINE - IRREGULARLY FLOODED
Unnvegetated
E2US2P, E2BBP
(E2USP), E2FLP
(E2USP), E2USP
Unvegetated -
Cobble-Gravel
E2US1P
Rocky
E2RSPR
(E2RSPr)
Rooted
E2EM1/FLP
(E2EM1P),
E2EM1P,
E2EM5P,
E2EMP, E2EM1R
(E2EM1P)
Rooted -
Diked/
Impounded
E2EM5Ph
(E2EMPh)
Rooted - Spoils
E2EMPs
Shrub-Scrub
E2SSP,
E2US/SSP
(E2SSP),
E2SS/EM1P
(E2SSP),
E2SS1P
Forested
E2FOP,
E2FO1P,
E2FO5/EM1P,
E2FO4/1P,
E2FO4P,
E2FO4/EM1P
(E2FO4P)
MARINE - SUBTIDAL & REGULARLY FLOODED, & IRREGULARLY FLOODED
Irregularly
Flooded -
Unvegetated
M2USP,
M2US2P
Irregularly
Flooded - Aquatic
Bed
M2AB1/USN
Irregularly
Flooded - Rocky
M2RS2P,
M2RSNr, M2RSPR
(M2RSPr)
Regularly
Flooded -
Unvegetated
M2USN,
M2US2N
Regularly
Flooded -
Rocky -
Vegetated &
Unvegetated
M2RSN,
M2RS/ABN,
M2AB/RSN
Subtidal -
Rooted
M1AB3
Subtidal -
Unvegatated
M1UBL
-------
RIVERINE - TIDAL - PERMANENTLY & REGULARLY & SEMIPERMANENTLY FLOODED
Permanently
Flooded Tidal -
Unvegetated -
Subtidal
R1OWV
R1UBV
Permanently
Flooded Tidal -
Unvegetated -
Subtidal -
Excavated
RlUBVx
Permanently
Flooded Tidal -
Aquatic Bed
R1ABV
Regularly
Flooded -
Unvegetated -
Shore
R1USN, R1FLN
(R1USN)
Semipermanently
Flooded Tidal -
Unvegetated -
Subtidal
R1UBT
Semipermanently
Flooded Tidal -
Unvegetated -
Shore
R1UST,
R1FLT
(R1UST)
RIVERINE - TIDAL - SEASONALLY & TEMPORARILY & IRREGULARLY FLOODED
Seasonally
Flooded Tidal -
Unvegetated -
Shore
R1USR,
R1FLR(R1USR),
R1SS/FLR
(R1USR)
Seasonally
Flooded Tidal -
Unvegetated -
Shore - Spoil
R1USRS
(RlUSRs)
Seasonally
Flooded Tidal -
Rooted
R1EMR
Seasonally
Flooded Tidal -
Streambed
R1SS/USR
(R1SB/USR)
Temporarily
Flooded Tidal -
Unvegetated -
Shore
R1USS
Irregularly
Flooded -
Unvegetated -
Shore
R1FLY
(R1USP)
-------
2.1.3 Watershed Delineation
The watershed associated with each coastal estuary on the Pacific Coast was delineated, so that
there was a unique one-to-one relationship between the estuary and its watershed. The estuarine
watershed geospatial layer was derived and augmented from the watershed layer originally
generated forNOAA's Coastal Assessment Framework
(ftp://sposerver.nos.noaa.gov/datasets/CADS/GIS_Files/ShapeFiles/caf/). This layer was not
sufficiently detailed to represent the smaller estuaries on the Pacific Coast, requiring further
delineation of a number of coastal watersheds in California, Oregon, and Washington.
Watershed boundaries subtending these estuary sites were determined from several sources. In
Washington, Oregon, and northwest California, the sixth field hydrologic unit (HUC) sub-
watershed geospatial layer created by the U.S. Forest Service from 1:24,000 scale USGS maps,
digital elevation models, and other data sources
(http://www.reo.gov/gis/projects/watersheds/REOHUCvl_3.htm) were used as primary
references. A watershed layer refined to the seventh field HUC boundary lines
(http://www.fsl.orst.edu/clams/cfsl0233.html) was also used for most of coastal Oregon north of
the Rogue River. The digital basin layer CALWATER was used as the primary source for major
basin delineations in central and southern California. CALWATER is the "California
Interagency Watershed Map" produced in 1999 and updated in 2004, and represents the State of
California's working definition of watershed boundaries (http://gis.ca.gov/meta.epl?oid=22175).
In all states, final refinements to the drainage boundaries were based on review of the hydrologic
drainage patterns derived from digital elevation data (10-meter horizontal resolution in Oregon
and 30-meter horizontal resolution in Washington and California) and from USGS 1:24,000
scale quadrangle maps. In addition, photographs and digital imagery for coastal features were
examined from the California Coastal Records Project (http://www.californiacoastline.org), State
of Washington database of shoreline photos (http://apps.ecy.wa.gov/shorephotos/), Terraserver
(http://terraserver-usa.com/), and other on-line imagery sources. Digital boundaries for
extraordinary sites - the Columbia River Basin, the interior portion of the San Francisco
watershed, and the perimeter of the Tijuana River watershed - were located from additional
sources and incorporated as being the "best available data". Boundary lines and water bodies
were plotted and reviewed for accuracy of coding and fidelity to the original sources. In some
cases, adjustments were made to the attribute coding of water bodies to reflect judgments that
these were part of an estuarine system.
The entire watershed was delineated for each estuary (see Table 2-2). By delineating the entire
watershed, these drainage areas are equivalent to the sum of NOAA's Estuarine Drainage Area
(EDA, portion of watershed that empties directly into the estuary and is affected by tides) and
Fluvial Drainage Area (FDA, component of an estuary's watershed upstream of the EDA
boundary; http://www.csc.noaa.gov/crs/lca/gloss.html). Entire watershed areas were analyzed
rather than the area represented by ED As to capture the entire landscape contributing to the
nutrient loading of the estuary. For the Columbia, the tidal portion of the Columbia River
watershed was delineated up to the Bonneville Dam using the NOAA land cover data while the
entire Columbia Basin, including into Canada, was delineated by the Interior Columbia Basin
Ecosystem Project Management Project (ICBEMP, 1996; http://www.icbemp.gov). The NOAA
land cover data truncated the northeastern end of the Klamath watershed which was filled in
using the National Land Cover Data (NLCD; http://www.epa.gov/mrlc/nlcd-2001.html) data.
22
-------
r\
Because of the difficulty in delineating watersheds for the smallest estuaries (<0.1 km ), the
estimates for these systems may overestimate the actual area draining into these systems. We
also identified the coastal areas not containing an estuarine polygon that drain directly into the
ocean, which are referred to as Coastal Drainage Areas (CDAs) by NOAA. The watershed
associated with each CDA was delineated, but these results are not reported here. A number of
watershed areas calculated here differ from those previously reported by NOAA's Coastal
Assessment Framework because multiple estuaries contained within a single NOAA EDA were
split out to create a one-to-one relationship between each estuary and its watershed.
2.1.4 Land Cover Patterns, Impervious Surfaces, and Population Density
The land cover pattern of each coastal watershed was determined using the 1992 National Land
Cover Data (NLCD, http://www.mrlc.gov) and both the 1995 and 2001 land cover data from
NOAA's Coastal Change Analysis Program (C-CAP,
http://www.csc.noaa.gov/crs/lca/ccap.html). These data were derived from Landsat satellite
imagery and produced at a 30-meter spatial resolution with an overall target accuracy
requirement of 85%. The present analysis primarily utilized the 2001 NOAA data since they are
the most recent, though the earlier NLCD data were used to fill in gaps such as portions of the
Klamath watershed. The 2001 NOAA data are based on 22 land cover classes
(http://www.csc.noaa.gov/crs/lca/oldscheme.html), which are not exactly the same as used in the
NLCD. The percent impervious surface was calculated for each of the estuarine watersheds
from the 2001 NLCD land cover data (http://www.epa.gov/mrlc/nlcd-2001.html). This dataset
estimated the percent impervious surfaces on a scale of 0 to 100% by 30 meter cells over the
entire contiguous United States, the highest resolution of impervious surfaces available at a
regional scale. The earlier version of this report (Lee et al., 2006) calculated impervious surfaces
using the default impervious coefficients in Attila for different land cover classes (U.S. EPA,
2004c).
The population within each coastal watershed was calculated using both the 1990 and 2000
census data. This analysis was conducted at the census block scale, the smallest unit used by the
Census Bureau. Population was prorated by area for census blocks that were transected by a
watershed boundary. Population density (# per km2) was calculated for each watershed using the
total delineated watershed area and the percent population change from 1990 to 2000 was
calculated.
2.1.5 Classification Strategy and Methods
We used a "hybrid" classification approach that combined qualitative analysis with more formal
statistical methods. The qualitative analysis was used to initially separate out major types of
estuaries primarily based on nutrient forcing functions and geomorphology especially as it
related to oceanic exchange. An advantage of this type of analysis is that it allows the
identification of broad groups of estuaries based on factors that are recognized as critical drivers
but for which there is either a lack of quantitative data and/or a lack of suitable metrics to
incorporate into a statistical analysis. A qualitative analysis also allows us to incorporate
temporally variable or intermittent attributes, such as whether an estuary periodically closes off
at the mouth ("blind" or intermittent estuary).
23
-------
Table 2-2. Area and geomorphology of PNW estuaries and watersheds. This table lists all coastal water bodies with a NWI estuarine
polygon, with the areas (km2) for the marine, estuarine, and tidal riverine NWI habitats. "Total Estuary" is the sum of the area of the
three NWI habitat classes, and is the area used for estuaries in our analyses. The total watershed area (km2) from the 2001 NOAA
land cover analysis is given for the associated watersheds. The watershed area for the Columbia River is up to the Bonneville Dam.
The four italicized water bodies are those that have an estuarine emergent wetland or estuarine aquatic bed polygon in the 2001
NOAA land cover analysis (see Table 2-3) but do not have a NWI estuarine polygon. These additional "estuaries" are included for
completeness but are not included in the analyses. Alternative geomorphological classifications are given in parentheses. * = new
estuary not included in the previous analysis (Lee et al., 2006).
ESTUARY
Waatch River
Hobuck Creek
Sooes River
Ozette River
Quillayute River
Goodman Creek
Mosquito Creek
Hoh River
Cedar Creek
Kalaloch Creek
Queets River
Whale Creek
Raft River
Camp Creek
Duck Creek
Quinault River
LATITUDE
48.344
48.336
48.324
48.181
47.908
47.823
47.798
47.749
47.711
47.607
47.544
47.490
47.463
47.398
47.387
47.349
WATERSHED
AREA
(km2)
38.4
2.3
107.9
232.1
1,625.0
81.7
43.8
773.4
26.9
45.4
1,166.3
32.1
204.0
22.7
18.4
1,133.7
AREA (km2)
MARINE
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ESTUARINE
0.93
0.0
0.50
0.03
0.50
0.03
0.01
0.11
0.02
0.02
0.60
0.01
0.17
0.03
0.01
0.42
TIDAL
RIVERINE
0.24
0.0
0.07
0.0
0.46
0.0
0.0
0.73
0.0
0.0
0.84
0.0
0.07
0.0
0.0
0.24
TOTAL
ESTUARY
1.16
0
0.56
0.03
0.96
0.03
0.01
0.84
0.02
0.02
1.43
0.01
0.24
0.03
0.01
0.66
ESTUARY TYPE
Tide-dominated drowned river mouth
Tidally restricted coastal creek?
(Coastal lagoon?)
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidal coastal creek
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
(Blind - Drowned river mouth)
Tidally restricted coastal creek
Tidally restricted coastal creek?
Highly river-dominated drowned river
mouth
to
-------
ESTUARY
Wreck Creek
Moclips River
Joe Creek
Boone Creek
Copalis River
Conner Creek
Grays Harbor
Willapa
Loomis Lake Creek
Columbia River
Clatsop Spit*
Necanicum
Indian Creek
Chapman Point
Ecola Creek
Arch Cape Creek*
Cove Beach*
Short Sand Creek*
Nehalem
Lake Lytle
LATITUDE
47.284
47.248
47.206
47.159
47.126
47.091
46.950
46.373
46.490
46.263
46.277
46.011
45.931
45.915
45.899
45.803
45.794
45.760
45.658
45.636
WATERSHED
AREA
(km2)
17.9
84.2
60.8
20.3
105.3
36.9
6,981.3
2,774.5
5.6
14,520.8
(to Bonneville
Dam)
0.7
216.8
7.0
0.7
54.7
11.6
0.4
14.9
2,215.2
7.4
AREA (km2)
MARINE
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ESTUARINE
0.01
0.08
0.05
0.01
0.78
0.17
254.39
389.74
0.01
411.53
0.08
1.57
0.004
0.001
0.06
0.001
0.005
0.002
10.43
0.0004
TIDAL
RIVERINE
0.0
0.02
0.0
0.01
0.08
0.0
8.34
1.12
0.0
257.45
0.0
0.06
0.0
0.0
0.002
0.0
0.0
0.001
1.23
0.0034
TOTAL
ESTUARY
0.01
0.10
0.05
0.01
0.86
0.17
262.73
390.86
0.01
668.98
0.08
1.63
0.004
0.001
0.06
0.001
0.005
0.003
11.65
0.004
ESTUARY TYPE
Tidally restricted coastal creek
Tidally restricted coastal creek
(Tidal coastal creek)
Tidally restricted coastal creek
Tidally restricted coastal creek
Moderately river-dominated drowned
river mouth
Tidally restricted coastal creek?
Tide-dominated drowned river mouth
(Bar built)
Tide-dominated drowned river mouth
(Bar built)
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek?
(Tidal coastal lagoon?)
Moderately river-dominated drowned
river mouth
(Bar built)
Tidally restricted coastal creek
(Tidal coastal creek)
Tidally restricted coastal creek?
Tidally restricted coastal creek
Tidally restricted coastal creek?
(Tidal coastal lagoon?)
Tidally restricted coastal creek?
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Coastal lagoon
-------
ESTUARY
Rockaway Beach
Creek*
Rockaway Clear
Lake*
Smith Lake
Tillamook
Netarts Bay
Chamberlain Lake
Sand Lake
Sears Lake
Miles Creek
Nestucca
Daley Lake*
Neskowin Creek
Salmon
Siletz
Schoolhouse Creek
Depoe Bay
Yaquina
Beaver Creek
Alsea
Big Creek
Yachats
Tenmile Creek North
LATITUDE
45.613
45.605
45.596
45.513
45.402
45.308
45.276
45.247
45.231
45.182
45.124
45.100
45.046
44.903
44.873
44.808
44.620
44.524
44.422
44.370
44.309
44.225
WATERSHED
AREA
(km2)
2.2
1.6
9.2
1,455.3
46.4
0.5
51.5
1.5
2.7
826.3
5.1
53.3
192.6
954.8
2.9
13.4
650.5
87.2
1,221.6
21.8
112.8
60.6
AREA (km2)
MARINE
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ESTUARINE
0.0004
0.0005
0.009
36.98
10.43
0.04
4.28
0.002
0.06
4.65
0.001
0.009
3.08
7.48
0.01
0.04
18.97
0.53
12.49
0.09
0.11
0.04
TIDAL
RIVERINE
0.0
0.0
0.005
0.51
0.001
0.0
0.0
0.0
0.0
0.35
0.0
0.005
0.04
1.38
0.0
0.0
0.99
0.02
0.0
0.0
0.0
0.0
TOTAL
ESTUARY
0.0004
0.0005
0.015
37.48
10.43
0.04
4.28
0.002
0.06
5.00
0.001
0.014
3.11
8.86
0.01
0.04
19.96
0.55
12.49
0.09
0.11
0.04
ESTUARY TYPE
Tidally restricted coastal creek
Tidally restricted coastal creek
Coastal lagoon
(Tidally restricted coastal creek)
Tide-dominated drowned river mouth
Bar built
Coastal lagoon?
Bar built
Coastal lagoon
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
(Tidal coastal lagoon)
Tidally restricted coastal creek
Moderately river-dominated drowned
river mouth
(Bar built)
Moderately river-dominated drowned
river mouth
Coastal lagoon
(Tidally restricted coastal creek)
Marine harbor/cove
(Drowned river mouth)
Tide-dominated drowned river mouth
Tidally restricted coastal creek
Moderately river-dominated drowned
river mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
Tidally restricted coastal creek
-------
ESTUARY
Berry Creek
Sutton Creek
Siuslaw River
Siltcoos River
Tahkenitch Creek
Umpqua River
Tenmile Creek South
Coos
Sunset Bay
Twomile Creek North
Coquille River
Twomile Creek South
New River
Sixes River
Elk River
Port Orford Head
Hubbard Creek*
Brush Creek
Mussel Creek
Euchre Creek
Gregg Creek*
Rogue River
Hunter Creek
LATITUDE
44.094
44.060
44.017
43.873
43.815
43.669
43.561
43.429
43.335
43.236
43.123
43.044
43.001
42.853
42.793
42.746
42.734
42.685
42.616
42.564
42.546
42.422
42.386
WATERSHED
AREA
(km2)
9.4
38.5
2,008.7
185.3
93.1
12,146.2
222.9
1,575.5
14.7
8.9
2,729.8
40.7
329.0
347.5
236.4
0.3
17.5
28.5
26.6
96.6
6.7
13,500.9
115.2
AREA (km2)
MARINE
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ESTUARINE
0.02
0.15
15.59
0.33
0.07
27.73
0.04
54.20
0.12
0.0
5.08
0.11
1.63
0.31
0.51
0.01
0.01
0.02
0.02
0.11
0.02
1.32
0.07
TIDAL
RIVERINE
0.0
0.0
0.0
0.04
0.19
6.05
0.46
0.70
0.0
0.0008
1.81
0.0
0.04
0.08
0.16
0.0
0.0
0.0004
0.0
0.001
0.001
1.44
0.02
TOTAL
ESTUARY
0.02
0.15
15.59
0.36
0.26
33.78
0.50
54.90
0.12
0.0008
6.89
0.11
1.67
0.39
0.66
0.01
0.01
0.02
0.02
0.12
0.02
2.77
0.1
ESTUARY TYPE
Tidally restricted coastal creek
(Tidal coastal creek)
Tidally restricted coastal creek
Moderately river-dominated drowned
river mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Tide-dominated drowned river mouth
Marine harbor/cove
Tidal coastal creek
(no NWI estuarine polygon)
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Blind - Drowned river mouth
Blind - Drowned river mouth
Blind - Drowned river mouth
Tidally restricted coastal creek
(Tidal creek)
Tidally restricted coastal creek
Tidally restricted coastal creek
Tidally restricted coastal creek
Tidally restricted coastal creek
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
-------
ESTUARY
Myers Creek
Pistol River
Burnt Hill Creek*
Cove at Boardman
Park*
Thomas Creek*
Whaleshead Creek*
Chetco River
Winchuck River
Smith River
Lake Earl
Crescent City Harbor
Klamath River
Johnson Creek
Ossagon Creek
Squashan Creek
Redwood Creek
Freshwater Lagoon
Stone Lagoon
Dry Lagoon
Big Lagoon
Little River
Mad River
LATITUDE
42.307
42.28
42.232
42.216
42.166
42.144
42.045
42.005
41.945
41.826
41.744
41.547
41.463
41.442
41.389
41.292
41.269
41.244
41.224
41.174
41.027
40.942
WATERSHED
AREA
(km2)
14.7
271.9
7.0
0.4
7.0
0.0
911.5
184.6
1,942.6
73.1
29.0
40,580.9
1.4
2.5
0.9
731.7
5.5
19.8
3.2
136.4
117.3
1,286.8
AREA (km2)
MARINE
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.61
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.17
ESTUARINE
0.02
0.44
0.004
0.005
0.01
0.02
0.40
0.09
2.04
9.01
0.01
1.09
0.01
0.01
0.01
0.22
0.0
2.30
0.0
5.03
0.06
1.16
TIDAL
RIVERINE
0.0
0.11
0.0
0.0
0.0
0.0
0.32
0.03
0.34
0.0
0.0
1.18
0.0
0.0
0.0
0.09
0.0
0.0
0.0
0.05
0.0
0.0
TOTAL
ESTUARY
0.02
0.55
0.004
0.005
0.01
0.02
0.72
0.12
2.38
9.01
1.63
2.27
0.01
0.01
0.01
0.30
0.0
2.30
0.0
5.09
0.06
1.33
ESTUARY TYPE
Tidally restricted coastal creek
(Tidal coastal creek)
Blind - Drowned river mouth
Tidal coastal creek?
Tidally restricted coastal creek?
Tidally restricted coastal creek
Tidally restricted coastal creek?
Highly river-dominated drowned river
mouth
Blind - Drowned river mouth
Highly river-dominated drowned river
mouth
Coastal lagoon
Marine harbor/cove
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
(Coastal lagoon?)
Tidally restricted coastal creek
Blind - Drowned river mouth
Coastal lagoon
(permanently blocked)
Coastal lagoon
Blind - Tidally restricted coastal creek
Coastal lagoon
Tidally restricted coastal creek
Highly river-dominated drowned river
mouth
-------
ESTUARY
Humboldt
Eel River
Guthrie Creek
Bear River*
LATITUDE
40.75
40.641
40.542
40.476
WATERSHED
AREA
(km2)
472.1
9,535.9
22.6
214.5
AREA (km2)
MARINE
0.07
0.0
0.0
0.004
ESTUARINE
71.42
10.18
0.005
0.12
TIDAL
RIVERINE
0.0
5.43
0.0
0.05
TOTAL
ESTUARY
71.49
15.61
0.005
0.18
ESTUARY TYPE
Bar built
(Coastal lagoon)
Highly river-dominated drowned river
mouth
Tidally restricted coastal creek
Tidally restricted coastal creek
(Blind - Drowned river mouth)
to
VO
-------
After the initial identification of major estuarine types, hierarchical clustering based on group
averages (see Clarke and Warwick, 2001; McCune and Grace, 2002) was used to classify the
estuaries or watersheds based on similarities in wetland habitat or watershed land cover patterns,
respectively. The Bray-Curtis metric was used as the measure of similarity in these analyses.
Clustering was conducted using both the absolute area of the NWI or land cover classes as well
as the relative proportions of the classes using untransformed data. When clustering variables
measured in different units (e.g., population, % impervious surfaces), the data were normalized
by subtracting the mean from each value and dividing by the standard deviation to generate a
standard score, with Euclidean distance used as the similarity metric. All clustering was
conducted with Primer6 (Clarke and Gorley, 2006; http://www.primer-e.com/).
Two criteria were used to define estuary groups in the clustering analysis. The primary approach
was whether the cluster was significantly different from a random reordering of the data within
the branch in the dendrogram using the SEVIPROF test in Primer6 (Clarke and Gorley, 2006).
By testing for differences within each branch independently, significant differences can occur at
different similarity levels in separate branches of the dendrogram, so that it is not necessary to
choose a single similarity level. Clusters of estuaries that were not significantly different were
not further divided regardless of the level of similarity. A value of p=0.05 was used as the
default significance level though there is no inherent reason that this value is the ecologically
"correct" level to identify groups of estuaries with similar functional attributes. Accordingly, we
evaluated a range of significance values (p = 0.10, 0.20, 0.40, 0.60) in the SEVIPROF tests. As
the significance level is relaxed (i.e., p increases), a greater number of significant clusters will be
generated and within-cluster similarity will increase though there will be fewer estuaries or
watersheds per cluster. This sequential approach provides flexibility in developing a
management framework, allowing managers to balance the extent of variability within estuary
groups versus the practical issues of increasing the number of groups. A secondary criterion to
defining groups was to combine estuaries within the branch if similarity was high (>75% with
Bray-Curtis or <25% of the maximum dissimilarity with Euclidean distance) even if the branch
showed a significant difference with SIMPROF. While this introduces a degree of subjectivity,
it eliminates the problem of generating numerous classes that would likely require similar
management strategies.
While it would be desirable to have a single classification that captures all aspects of current
impacts and future vulnerability, the reality is that multiple classification systems and approaches
are needed to address different scientific and management issues. For example, a classification
based on resource availability (e.g., extent of wetlands) might group an estuary differently than
one based on loading (e.g., land cover pattern). Our approach was to evaluate several
classifications schema both in this chapter and in Chapter 7 based on different biotic and
landscape attributes related to nutrient dynamics or estuarine vulnerability. It should be
recognized, however, that any classification system should be considered a hypothesis until it is
evaluated with independent datasets demonstrating similar estuarine responses, nutrient
dynamics, or vulnerability to increased loadings.
2.1.6 Ecological Resources in Pacific Northwest Estuaries
The ultimate purpose of this classification exercise is to derive the insights required for the
efficient management of ecological resources in PNW estuaries. Different types of resources, for
30
-------
example salmonids and submerged aquatic vegetation, are likely to respond differently to
nutrient enrichment, thus requiring different management strategies. Additionally, the type and
extent of ecosystem services varies across estuaries, which may suggest different management
prioritizations among estuaries. Thus, as part of this effort, a preliminary evaluation of the
general types of ecological services provided by each of the estuaries was summarized in
Table 2-3 and further discussed in Section 2.3.
One class of ecological services relates to the occurrence of estuarine intertidal and subtidal
wetlands, which was evaluated using both the emergent wetland and aquatic bed classes in NWI
and the 2001 NOAA land cover analysis. The irregularly flooded estuarine forest and shrub
classes were not included since these semi-terrestrial wetland types are less vulnerable to
estuarine water quality. Three other important PNW resources are oyster aquaculture, native
salmon runs, and native and migratory birds. Oyster aquaculture was evaluated from known
facilities as well as from state permits. The presence of current or historic salmon runs was
based on reports in the literature, while Lawson et al. (2004) provided the predicted number of
Coho smolts historically present for Oregon estuaries based on a watershed model. Two
measures of an estuary's importance to birds were whether the estuary had been classified as an
Important Bird Area (IB A) by the Audubon Society
(http://www.audubon.org/bird/iba/index.html) and whether it is recognized as an important
shorebird site in the U.S. Shorebird Conservation Plan for the Northern Pacific Coast (Drut and
Buchanan, 2000; http://www.fws.gov/shorebirdplan/RegionalShorebird/RegionalPlans.htm).
The final function considered, but not quantified, was recreational support. Recreational use of
PNW estuaries is primarily limited to boating, fishing, crabbing, and clamming, with relatively
little direct water contact.
2.2 Inventory of Estuaries on the Pacific Coast
The starting point of our comprehensive classification of estuaries was to conduct an inventory
of the estuaries on the Pacific Coast. We found that the Pacific Coast contains a surprisingly
large number of estuaries - 230 based on the criterion that the waterbody contains an estuarine
NWI polygon and discharges directly into the ocean. This number does not include subestuaries
within larger water bodies nor does it include the Strait of Juan de Fuca or Puget Sound. In
addition, there are several harbors and coastal streams that do not contain NWI estuarine
polygons but which fall within the size range of the coastal estuaries. Although not counted as
estuaries, these additional water bodies may provide some of the same ecological functions as
similar sized estuaries. As a result of the 2007 NWI update, the number of PNW estuaries
increased from 216 in the earlier report (Lee et al., 2006) to 230 in the present analysis. The
number estuaries may continue to change as NWI adds or deletes estuarine polygons from the
smallest coastal waterbodies.
The naming convention used in this report is to refer to a waterbody as an "estuary" if it consists
of multiple named embayments, rivers, or creeks that constitute major geographic features. This
is to avoid confusion whether the name refers to the entire waterbody or a single component. For
example, "San Francisco Estuary" is used to refer to the combination of San Francisco Bay and
the Delta, including San Pablo Bay, Suisan Bay, and other components of the Delta. The use of
"San Francisco Bay" is restricted to identifying the bay proper. Similarly, we refer to the
"Yaquina Estuary" or "Yaquina Bay Estuary" rather than "Yaquina Bay" because the estuary is
-------
80 -,
60 -
0>
'_
CS
-*->
W 40 -
<*-
o
>_
| 20 -
Z
0
40
72
59
40
15
4
1 1
i i i i i
<0.01 0.01-0.1 0.1-1.0 1-10 10-100 >100
Estuary Size Class (km )
Figure 2-6. Size distribution of all 230 coastal estuaries in California, Oregon, and Washington
(exclusive of Puget Sound) identified by the presence of a NWI estuarine polygon.
40
35 -
30 -
GO
.«
1 25 -
3
-*->
W 20 -
<*-
o
S3 15 -I
jz
I 10-1
Z
5 -|
0
17
23
17
10
<0.01 0.01-0.1 0.1-1.0
1-10
10 - 100
>100
Estuary Size Class (km )
Figure 2-7. Size distribution of the 103 PNW coastal estuaries identified by the presence of a
NWI estuarine polygon.
32
-------
% of Estuarine Habitat
K^ 1\J U> -|^ Wl
D 0 0 0 0 0
1 1 1 1 1
48
7
42
3
| |
w III
Subtidal Irregularly Regularly Irregularly
Exposed Flooded Flooded
Figure 2-8. Distribution of estuarine subtidal and intertidal habitats in the 103 PNW estuaries.
Intertidal habitat is defined as the sum of the irregularly exposed, regularly flooded, and
irregularly flooded NWI estuarine and marine classes.
composed of both Yaquina Bay proper and lower portions of the Yaquina River. Names of
waterbodies composed of a single named component, such as Beaver Creek, are not modified.
Estuaries on the Pacific Coast vary greatly in size. The smallest is Liddell Creek in Southern
California (0.00017 km2) and the largest is the San Francisco Estuary (3346 km2). Most of the
Pacific Coast estuaries are small, with 171 (74%) having an estuarine area <1 km2 (Figure 2-6).
Although there are numerous small estuaries, the bulk of the estuarine habitat is concentrated
within a few systems. The San Francisco Estuary contains almost two thirds of the total
estuarine area on the Pacific Coast, exclusive of Puget Sound, with the four largest estuaries (San
Francisco, Columbia River, Willapa Estuary, and Grays Harbor) accounting for 88% of the
estuarine area.
A similar analysis for the PNW identified 103 coastal estuaries (see Figures 2-2 to 2-5) of which
Burnt Hill Creek and Rockaway Beach Creek are the smallest (0.0004 km2; Table 2-2). The
largest is the Columbia River Estuary with an area of 669 km2, of which 257 km2 is tidal riverine
habitat much of which is fresh. Most PNW estuaries are small, with 59 (66%) of the estuaries <1
km2 (Figure 2-7). Only the Columbia River, Willapa, and the Grays Harbor estuaries are larger
than 100 km2. On an area basis, these three largest estuaries account for 79% of the total PNW
estuarine area, exclusive of Puget Sound. In addition to these estuaries defined by NWI, the
NOAA 2001 land cover survey identified four small water bodies that they classified as
containing estuarine habitat but which did not contain a NWI estuarine polygon. These water
bodies are listed in Table 2-2 but were not included in the numerical tallies of estuaries or the
statistical analyses.
33
-------
One prominent characteristic of PNW estuaries is their extensive intertidal area. The intertidal
area for the 103 estuaries was calculated as the sum of the NWI estuarine polygons defined as
"irregularly exposed", "regularly flooded" and "irregularly flooded", which approximates tidal
heights from mean lower low water (MLLW) to mean higher high water (MHHW). Averaging
the NWI estuarine polygons across all of the PNW estuaries, intertidal habitat constituted 52% of
the estuarine area (Figure 2-8). There is no apparent pattern in the extent of intertidal habitat as a
function of estuary size, with the percent intertidal in the NWI estuarine habitats ranging from
about 39% to 57% across estuarine size classes.
2.3 Ecological Services of Small Pacific Northwest Estuaries
9 9
Estuaries smaller than about 1 km or even less than 10 km have generally been excluded from
previous estuarine classification efforts. For example, the National Estuarine Eutrophication
Assessment (Bricker et al., 1999) included all 13 estuaries in the PNW >10 km but only 4 of the
17 estuaries between 1 and 10 km2 and none <1 km2. However, the large number of small PNW
estuaries (Figure 2-7) suggests that it was important to include them in the regional analysis for
several reasons. These smaller estuaries are important for native salmon, a critical regional
resource and management issue (Lackey, 2004; Lackey et al., 2006a, b). Forty-six of the 73
estuaries <1 km2 have reported salmon runs (Table 2-3). An analysis of the predicted historical
number of Coho smolts produced per watershed along the Oregon coast (Lawson et al., 2004)
indicated that the small estuaries produced a proportionally greater number of smolts than the
larger estuaries when normalized to estuarine area (Figure 2-9). Estuaries <1 km2 were
contributing about 8% to total Coho runs while estuaries <10 km2 were contributing about 24%.
These smaller coastal watersheds and estuaries may also serve as a future refuge for wild salmon
with the increasing development and alteration of the larger estuaries and watersheds. Lackey et
al. (2006c) suggest that the coastal rivers in California, Oregon, Washington, and southern
British Columbia offer greater potential for preserving wild salmon runs than do the Sacramento-
San Joaquin, Columbia, and lower Fraser Rivers, places where the most expensive salmon
recovery efforts are now focused. In addition to salmon, small estuaries in Northern California
(e.g., Smith River and Stone Lagoon) provide critical habitat for the endangered tidewater goby
(Eucyclogobius newberryf) (Federal Register, 2002; Ahnelt et al., 2004), and many of the smaller
estuaries provide habitat for resident and migratory birds (Table 2-3), including bald eagles and
brown pelicans.
r\
In terms of other ecosystem services, several of the coastal lagoons <10 km contain extensive
emergent wetlands (e.g., Big Lagoon, Lake Earl) while many of the drowned river estuaries and
tidal creeks between about 0.5 km2 and 10 km2 also contain considerable wetlands (Table 2-3).
The wetlands in these small estuaries are likely important for coastal and migratory birds and to a
limited number of brackish water fishes such as the tidewater goby. Among the smallest
estuaries (<0.1 km2) only 7 of the 50 contain any emergent wetlands, none contain any aquatic
beds (Table 2-3), and these smallest estuaries are little utilized by recreational crab or clam
species. Thus, the ecosystem services provided by these smallest estuaries appear to be related
to supporting wild salmon runs, recreation, and perhaps as bird habitat. These smallest estuaries
and their associated rivers and creeks may also provide economic benefits in terms of increased
property value, though we made no attempt to estimate this.
34
-------
Table 2-3. Examples of ecological resources and services in PNW estuaries. Estuarine emergent wetlands and aquatic beds from the
NWI are the sums of classes with "EM" or "AB" codes, respectively. The estuarine emergent wetland and aquatic beds from NOAA
are areas for these two land cover classes. "Estuary Area" is the sum of the NWI estuarine, marine, and tidal riverine areas (see Table
2-2). "Oyster culture" indicates whether there is commercial oyster aquaculture or whether there are permits to allow oyster
aquaculture. "Salmon present" indicates whether salmon are currently or have historically been reported from the estuary or from the
streams and rivers flowing into the estuary. "N?" is used to indicate estuaries where we suspect salmon are absent based on estuary
size or other landscape attributes while "?" indicates that we are unaware of any reports of salmon from that estuary. The "# smolts"
is the historical potential number of Coho smolts predicted to have occurred within Oregon estuaries (Lawson et al., 2004). The five
italicized water bodies are those that have an estuarine emergent wetland or estuarine aquatic bed polygon in the 2001 NOAA land
cover analysis (see Table 2-2) but do not have a NWI estuarine polygon. SCP = listed as important site in the Shorebird Conservation
Plan with sites marked with an asterisk (*) supporting > 4000 birds. IBA = Important Bird Areas. NA = Estuary not identifiable in
NOAA dataset.
ESTUARY
Waatch River
Hobuck Creek
Sooes River
Ozette River
Quillayute River
Goodman Creek
Mosquito Creek
Hoh River
Cedar Creek
Kalaloch Creek
Queets River
Whale Creek
Raft River
Camp Creek
Duck Creek
ESTUARY
AREA
(km2)
1.16
0.0
0.56
0.03
0.96
0.03
0.01
0.84
0.02
0.02
1.43
0.01
0.24
0.03
0.01
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
0.93
0.0
0.17
0.0
0.05
0.0
0.0
0.0
0.0
0.0
0.15
0.0
0.08
0.0
0.0
NWI
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
0.0
0.003
0.01
0.0
0.03
0.0
0.0
0.004
0.0
0.002
0.09
0.0
0.002
0.002
0.0
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.0
0.002
0.002
0.0
0.0
0.001
0.0
0.0
0.01
0.0
0.002
0.0
0.0
OYSTER
CULTURE
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
SALMON
PRESENT
(# SMOLTS)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
BIRD
HABITAT
-------
ESTUARY
Quinault River
Wreck Creek
Moclips River
Joe Creek
Boone Creek
Copalis River
Conner Creek
Grays Harbor
Willapa
Loomis Lake Creek
Columbia River
Clatsop Spit
Necanicum
Indian Creek
Chapman Point
Ecola Creek
Arch Cape Creek
Cove Beach
Short Sand Creek
Nehalem
Lake Lytle
Rockaway Beach
Creek
Rockaway Clear
Lake
Smith Lake
ESTUARY
AREA
(km2)
0.66
0.01
0.10
0.05
0.01
0.86
0.17
262.73
390.86
0.01
668.98
0.08
1.63
0.004
0.001
0.06
0.001
0.005
0.003
11.65
0.004
0.0004
0.0005
0.015
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
0.01
0.0
0.003
0.0
0.0
0.40
0.0
15.24
39.44
0.0
28.63
0.06
0.42
0.0
0.0
0.02
0.0
0.002
0.0
2.34
0.0
0.0
0.0
0.0
NWI
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
147.59
181.47
0.0
4.07
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.14
0.0
0.0
0.0
0.0
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
0.003
0.001
0.0
0.002
0.004
0.001
0.0
11.98
29.04
0.06
18.99
NA
0.01
NA
NA
0.0
NA
NA
NA
2.76
NA
NA
NA
NA
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.01
0.0
0.0
0.0
0.0
0.0
0.0
1.11
2.67
0.003
0.79
NA
0.06
NA
NA
0.001
NA
NA
NA
0.02
NA
NA
NA
NA
OYSTER
CULTURE
N
N
N
N
N
N
N
Y
Y
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
SALMON
PRESENT
(# SMOLTS)
Y
Y
Y
Y
?
Y
Y
Y
Y
N?
Y
N
Y
(685,000)
Y
(100)
7
Y
(72,000)
N?
N?
N?
Y
(3,330,000)
7
N?
N?
7
BIRD
HABITAT
SPC*
SPC*
IBA
IBA
SPC
-------
ESTUARY
Tillamook
Netarts Bay
Chamberlain Lake
Sand Lake
Sears Lake
Miles Creek
Nestucca
Daley Lake
Neskowin Creek
Salmon River
Siletz River
Schoolhouse Creek
Depoe Bay
Yaquina
Beaver Creek
Alsea
Big Creek
Yachats River
ESTUARY
AREA
(km2)
37.48
10.43
0.04
4.28
0.002
0.06
5.00
0.001
0.014
3.11
8.86
0.01
0.04
19.96
0.55
12.49
0.09
0.11
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
4.29
1.07
0.0
2.37
0.0
0.0
0.88
0.0
0.0
2.13
2.16
0.0
0.0
3.03
0.42
2.49
0.08
0.0
NWI
ESTUARINE
AQUATIC BED
(km2)
0.02
2.60
0.0
0.07
0.0
0.0
0.09
0.0
0.0
0.004
0.0
0.0
0.0
3.65
0.0
3.25
0.0
0.02
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
4.51
1.16
NA
3.19
NA
NA
1.14
NA
0.002
2.54
1.54
0.0
0.002
1.11
0.03
2.43
NA
0.0
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.50
0.89
NA
0.02
NA
NA
0.36
NA
0.0
0.0
0.45
0.0018
0.0
0.72
0.01
0.99
NA
0.0
OYSTER
CULTURE
Y
Y
N
N
N
N
N
N
N
N
N
N
N
Y
N
N
(under
consider-
ation)
N
N
SALMON
PRESENT
(# SMOLTS)
Y
(3,288,000)
Y
(15,000)
?
Y
(123,000)
?
?
Y
(1,037,000)
N?
Y
(49,000)
Y
(168,000)
Y
(1,217,000)
Y
(2,000)
Y
(7,000)
Y
(1,217,000)
Y
(265,000)
Y
(1,628,000)
Y
Y
(110,000)
BIRD
HABITAT
SPC*/IBA
SPC*/IBA
IBA
IBA
SPC/IBA
SPC*/IBA
SPC/IBA
-------
ESTUARY
Tenmile Creek
North
Berry Creek
Sutton Creek
Siuslaw River
Siltcoos River
Tahkenitch Creek
Umpqua River
Tenmile Creek
South
Coos
Sunset Bay
Twomile Creek
North
Coquille River
Twomile Creek
South
New River
Sixes River
Elk River
Port Orford Head
Hubbard Creek
Brush Creek
ESTUARY
AREA
(km2)
0.04
0.02
0.15
15.59
0.36
0.26
33.78
0.50
54.90
0.12
0.0008
6.89
0.11
1.67
0.39
0.66
0.01
0.01
0.02
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
0.02
0.0
0.0
3.32
0.08
0.0
3.18
0.01
7.28
0.0
0.0
1.07
0.01
0.48
0.03
0.05
0.0
0.0
0.0
NWI
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.03
1.30
0.0
0.0
0.07
0.0
5.98
0.0
0.0
0.06
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
0.0
0.0
0.0
1.14
0.0
0.001
1.43
0.0
4.91
NA
0.01
1.38
0.02
0.27
0.05
0.01
NA
NA
0.0
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.01
0.09
0.06
0.07
0.94
0.01
4.20
NA
0.0
0.17
0.0
0.02
0.0
0.0
NA
NA
0.0
OYSTER
CULTURE
N
N
N
Y
N
N
Y
N
Y
N
N
N
N
N
N
N
N
N
N
SALMON
PRESENT
(# SMOLTS)
Y
(28,000)
Y
(54,000)
Y
(84,000)
Y
(2,674,000)
Y
(771,000)
Y
(228,000)
Y
(8,199,000)
Y
(525,000)
Y
(2,058,000)
7
N?
Y
(4,169,000)
Y
(134,000)
Y
(396,000)
Y
(372,000)
Y
7
7
Y
BIRD
HABITAT
SPC/IBA
IBA
IBA
IBA
SPC*/IBA
SPC*/IBA
SPC/IBA
-------
ESTUARY
Mussel Creek
Euchre Creek
Greggs Creek
Rogue River
Hunter Creek
Myers Creek
Pistol River
Burnt Hill Creek
Cove at Boardman
Park
Thomas Creek
Whaleshead Creek
Chetco River
Winchuck River
Smith River
Lake Earl
Crescent City
Harbor
Klamath River
Johnson Creek
Ossagon Creek
Squashan Creek
Redwood Creek
Freshwater Lagoon
Stone Lagoon
Dry Lagoon
Big Lagoon
Little River
ESTUARY
AREA
(km2)
0.02
0.12
0.02
2.77
0.10
0.02
0.55
0.004
0.005
0.01
0.02
0.72
0.12
2.38
9.01
1.63
2.27
0.01
0.01
0.01
0.30
0.0
2.30
0.0
5.09
0.06
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
0.0
0.02
0.002
0.06
0.0
0.0
0.01
0.0
0.0
0.0
0.0
0.01
0.004
0.07
0.0
0.0
0.03
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NWI
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
9.01
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.03
0.0
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
0.02
0.08
NA
0.25
0.01
0.0
0.14
NA
NA
NA
NA
0.03
0.05
0.56
4.31
0.0
0.01
0.0
0.0
NA
0.18
0. 10
0.08
0.12
0.57
0.14
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
NA
0.10
0.0
0.0
0.0
NA
NA
NA
NA
0.01
0.0
0.0
0.0
0.0
0.005
0.0
0.0
NA
0.0
0.0
0.0
0.0
0.0
0.0
OYSTER
CULTURE
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
Y
(Permits)
N
N
N
N
N
N
N
N
N
N
SALMON
PRESENT
(# SMOLTS)
Y
Y
?
Y
Y
Y
Y
N
N
?
7
Y
Y
Y
Y
No?
Y
7
7
7
Y
No
Y
No?
Y
Y
BIRD
HABITAT
-------
ESTUARY
Mad River
Humboldt
Eel River
Guthrie Creek
Bear Creek
TOTAL
ESTUARY
AREA
(km2)
1.33
71.49
15.61
0.005
0.18
1677.44
NWI
ESTUARINE
EMERGENT
WETLAND
(km2)
0.02
3.93
1.68
0.0
0.0
128.22
NWI
ESTUARINE
AQUATIC BED
(km2)
0.0
19.82
0.06
0.0
0.0
384.34
NOAA
ESTUARINE
EMERGENT
WETLAND
(km2)
0.33
9.35
4.95
0.0
0.003
111.15
NOAA
ESTUARINE
AQUATIC BED
(km2)
0.0
0.0
0.002
0.0
0.0
14.31
OYSTER
CULTURE
N
Y
N
N
N
10 Y
SALMON
PRESENT
(# SMOLTS)
Y
Y
Y
Y
?
67 Y
BIRD
HABITAT
IBA
-
-------
-2
"o
100
80 -I
60 -
40 -
20 -
0
% Estuarine Area
% Coho Smolt
86
24
13
68
<1
1-10
Estuary Size Class (km )
Figure 2-9. Distribution of the projected historical number of Coho smolts by estuary size
classes in 29 Oregon estuaries. Coho smolt data from Lawson et al. (2004; see Table 2-3).
The final reason for the inclusion of these small water bodies is that they are likely to require
different management strategies. Factors such as their smaller volumes, differences in seasonal
and tidal variability in exchange and salinity, and different biotic communities are all likely to
result in different exposures and vulnerabilities. Research focused on these largely ignored small
systems is required to better understand how they function, but it is possible to speculate that
their small size makes them more vulnerable to certain types of stressors while at the same time
offering greater opportunities for cost-effective protection and/or mitigation efforts
2.4 Classification by Geomorphology and Oceanic Exchange
The 103 PNW estuaries were initially separated into broad classes based on the potential extent
of oceanic exchange, a fundamental feature affecting vulnerability to terrestrial nutrient loading.
Since residence times are not available for most of these waterbodies, we qualitatively estimated
relative oceanic exchange from estuarine geomorphology based on the literature, aerial
photographs, and personal observations. Photographic sources included the aerial photographs
taken by the Pacific Coastal Ecology Branch (PCEB; see Chapter 6) as well as online sources
including the Oregon Coastal Atlas (http://www.coastalatlas.net), high resolution (1:12,000)
panchromatic imagery commercially available through GlobeExplorer ™, Terraserver
(http://www.terraserver.com), GoogleEarth (http://www.google.com), Washington Coastal Atlas
(http://www.ecy.wa.gov/programs/sea/SMA/atlas_home.html), Washington's Department of
Ecology Washington Shoreline Aerial Photos (http://apps.ecy.wa.gov/shorephotos/index.html),
and the California Coastal Records Project (http://www.californiacoastline.org/). Additionally,
the NWI coverages were used to evaluate the overall waterbody shape and structure.
41
-------
It is important to recognize that many estuaries display characteristics of several
geomorphological classes so assigning an estuary to a single class is somewhat artificial. In
recognition of this gradation, alternate classifications are given when appropriate. To the extent
practical, an estuary's structure and dynamics should be considered as well as its classification
when assessing vulnerability or developing a management strategy.
2.4.1 Coastal Lagoons and Blind Estuaries
We identified two types of moderate to large estuaries with restricted exchange. The first are
coastal lagoons, defined here as coastal water bodies located behind berms with NWI estuarine
polygons that are only intermittently open to the ocean. For example, some Southern California
lagoons remain open only about a third of the time (Elwany et al., 1998). Coastal lagoons can
breach during storms, after periods of heavy rain, or long-shore movement of dune sand (Elwany
et al., 1998). Coastal lagoons are also artificially breached, especially in California, to increase
circulation, improve water quality, and reduce flooding (e.g., Williams et al., 1999; Merritt Smith
Consulting, 1999). We classified eight PNW water bodies as coastal lagoons (Table 2-2). The
largest of these is Lake Earl while the smallest is the Sears Lake.
The second type of waterbody with restricted oceanic exchange is the blind or intermittent
estuary. Mouths of blind estuaries periodically close due to the formation of an ephemeral berm.
The mouth of a blind estuary may not close every year, but closure is frequent enough to be a
regular characteristic of the system. Blind estuaries are more common in the southern portion of
the Pacific coast, with many of the smaller southern California estuaries closing during the late
summer and early fall before the rains. We separate blind estuaries from coastal lagoons based
on these attributes: blind estuaries are open to the ocean more frequently than lagoons; there
appears to be more among-year variation in blind estuaries whether the mouth closes; and blind
estuaries tend to have a more riverine shape compared the more "pond" or "lake" shape of
lagoons that often run parallel to the dunes. Such a distinction is also made with the South
African "Temporarily Open/Closed Estuaries" (TOCE), which are divided into "Intermittently
Closed Estuaries" (ICEs), which have a connection to the ocean more than 50% of the time, and
"Intermittently Open Estuaries" (lOEs), which are closed to the ocean more than 50% of the time
(Whitfield and Bate, 2007).
Based on the aerial photographs, literature (e.g., Bottom et al., 1979; Cortright et al., 1987), and
online sources (http://www.coastalatlas.net), six PNW waterbodies were classified as blind
(Table 2-2). The smallest is the Winchuck River while the largest is the New River. Note that
this count of blind estuaries does not include the tidally restricted coastal creeks discussed below
that may also periodically close off at the mouth.
With both lagoons and blind estuaries, closure of the mouth eliminates any direct flushing with
the ocean though there may be subsurface flow between the estuary and the ocean. The absence
or greatly reduced exchange during closure can result in low dissolved oxygen, extended periods
of low salinity at least at the surface, increased temperatures, and stratification. Fish kills have
been reported during periods of closure in blind estuaries in Oregon due to the ponding of
freshwater (Clifton et al., 1973 in Bottom et al., 1979). Blooms of nuisance algae can occur
during periods of closure (e.g., John Gilchrist & Associates and Fall Creek Engineering, Inc.,
2005). Five dogs died after swimming in Big Lagoon in 2001, which was attributed to exposure
42
-------
to toxic blue green algae resulting from warm weather and the heavy nutrient load caused by the
lack of winter breaching (Humboldt County Dept. Health Human Sen, 2005). Bottom dissolved
oxygen in the Mattole River in northern California periodically decreased below 6 mg I"1 after
"permanent lagoon formation" and occasionally dropped to 4 mg I"1, though interestingly the
lowest dissolved oxygen levels occurred two weeks before the estuary closed off (Zedonis et al.,
2008). Salinity in these restricted waterbodies can vary drastically depending upon the extent of
breaching. When closed for extended periods, some of these waterbodies can transform into
freshwater systems. The Mattole River became a freshwater lagoon after the closure of the
mouth (Busby, 1991). Similarly, our measurements of salinity in the New River ranged from 0
to 1.6 psu in the summer of 2003 (H. Lee II, unpub. data).
As a class, both the lagoonal systems and blind estuaries are highly vulnerable to nutrient
enrichment during periods of restricted oceanic exchange. However, both lagoons and blind
estuaries experience dramatic among-year and seasonal variability in flushing and corresponding
variation in water quality and salinity, confounding how to manage these systems. One strategy
is to assume "worst case" scenarios of restricted or blocked ocean access. Since salmon are
present in many of these systems, an appropriate beneficial use defined in Oregon for these
estuaries is to support "Resident Fish and Aquatic Life, Salmonid Spawning & Rearing". The
dissolved oxygen threshold for this beneficial use in Oregon for estuaries is 6.5 mg I"1
(http://www.sos.state.or.us/archives/rules/OARs_300/OAR_340/340_041.html). In addition, the
Oregon narrative criteria states that "where a less stringent natural condition of water of the State
exceeds the numeric criteria ... the natural condition supersedes the numeric criteria and
becomes the standard for that water body." Dissolved oxygen levels of > 6.5 mg I"1 may be
difficult to obtain during periods of restricted flushing especially in the lagoonal systems. One
management option is to artificially breach the barriers, though that may result in other impacts,
such as the premature flushing of salmon smolts.
2.4.2 Tidal Coastal Creeks/Tidally Restricted Coastal Creeks
Tidal coastal creeks are creeks or streams discharging directly into the ocean and which
experience input of ocean water at least during high tide. These systems are the smallest type of
"estuary", some of which are essentially freshwater streams with a limited area influenced by
ocean waters. However, the hydrodynamics of these systems appears to be more complex than
their small size suggests. Based on field observations on a few coastal creeks and analysis of
aerial photographs, most have restricted connections with the ocean during a portion of the year.
We refer to these as "tidally restricted coastal creeks" which are defined by having one or more
of the following characteristics at least intermittently: 1) there is a narrowing of mouth from sand
movement sufficient to restrict exchange with the ocean; 2) formation of a berm that closes the
mouth, separating the creek from the ocean; or 3) there is a development of a sill near the mouth
sufficient to restrict exchange with the ocean.
We tentatively identified 55 waterbodies as tidal coastal creeks in the PNW, ranging from
Chapman Point Creek (0.001 km ) to Beaver Creek (0.55 km ) (Table 2-2). Analysis of aerial
photographs suggests that most of these systems have limited exchange during portions of the
year, and 53 are tentatively considered tidally restricted coastal creeks. There is no clear
demarcation between the larger tidally restricted coastal creeks and the tidal rivers. For example,
Bear River was classified as a tidally restricted coastal creek though it alternatively could be
43
-------
classified as a blind drowned river mouth estuary. Several of the smaller tidal rivers in
Washington, including the Waatch, Hoh, Quinault, Sooes, and Quillayute, have partially
restricted mouths and may represent the transition between the tidally restricted coastal creeks
and blind drowned river mouth estuaries. Part of the difficulty in cleanly separating these
systems is the paucity of studies on the estuarine segments of tidal creeks. Even a relatively
small field effort evaluating salinity patterns and mouth constrictions across a suite of tidal
creeks would greatly promote our understanding of their dynamics.
Based on our present knowledge, flushing is presumed to be very high in estuarine portion of
tidally restricted coastal creeks when they are open, and salinity near the mouth is likely to
undergo extreme tidal and seasonal fluctuations. An indication of the extent of these salinity
variations can be found in the larger drowned river mouth estuaries, where salinity at the mouth
can vary by more than 20 psu over a tidal cycle (Figures 4-8 and 4-9), and median salinity near
the mouth of highly river-dominated systems can vary by 20 psu seasonally (Figure 2-11). We
suspect that salinity near the mouth of the tidal creeks can cycle from essentially marine to fresh
over a tidal cycle during the periods when they are open. During such periods, this high tidal
exchange reduces the accumulation of terrestrially derived nutrients in the "estuarine" segments
of these waterbodies. However, these extreme salinity fluctuations also limit the ability of all but
the most euryhaline species to survive. A possible reflection of this salinity stress is the paucity
of estuarine emergent wetlands and aquatic beds in tidal coastal creeks <0.1 km2 (Table 2-3).
During other periods, restriction near the mouth limits exchange in these tidal creeks. If the
restriction is of sufficient duration, the "estuarine" portion of tidal creeks may develop low
dissolved oxygen conditions. Additionally, an ongoing study of the Yachats River indicates that
the formation of a sill near the mouth restricts exchange with the ocean to the spring high tides,
essentially acting as a micro-fjord (C. Brown, unpublished data). This oceanic water advected
during spring tides can form localized pockets of deeper, saline waters below a freshwater
surface layer. With no or minimal exchange or turnover, the dissolved oxygen in these saline
pockets declines until the next spring tide or storm. Qualitative observations on other tidal
creeks suggests that the Yachats River is not unique in its structure; thus a number of tidally
restricted coastal creeks may have localized areas of bottom water with low dissolved oxygen
during a portion of the month.
While there are uncertainties about the dynamics and classification of these coastal creeks, we
suggest that as with blind estuaries these systems could be managed on a "worst case" scenario.
The most sensitive resource is likely to be the juvenile salmon that utilize many of these systems
(Table 2-3). In these cases, the Oregon dissolved oxygen standard of 6.5 mg I"1
(http://www.sos.state.or.us/archives/rules/OARs_300/OAR_340/340_041.html) would be an
appropriate criterion, except for when natural conditions prevent attainment of this standard. An
ongoing study of Yachats indicates that dissolved oxygen conditions fall below 6.5 mg I"1
periodically in pockets of bottom water "trapped" by natural processes. The lack of typical
"estuarine" resources and salmon in many of the estuaries smaller than about 0.01 km2 suggests
that the application of estuarine criteria may not be the best approach to their management. In
fact, the smallest estuarine polygon in a tidal creek was only 35m long, making application of
estuarine criteria to this system dubious. A more practical strategy may be to manage the
44
-------
freshwater component of these systems and/or to apply human health end-points, such as water
contract criteria for fecal coliforms, to the water advected onto the beach.
2.4.3 Marine Harbors/Coves
At the opposite extreme of these restricted estuaries, there is a group of coastal waterbodies that
has an unobstructed connection with the ocean with relatively small freshwater inflow. These
systems are separated as marine harbors/coves, and as a class should have a relatively low
vulnerability to terrestrial nutrient loadings. Based on analysis of photographs and preliminary
field observations, two coastal waterbodies fall into this class - Sunset Bay and Crescent City
Harbor. A third system, Depoe Bay, the "world's smallest harbor", is tentatively classified as a
marine harbor/cove. With a mouth width of about 28 meters, Depoe Bay is less open to the
ocean than either Sunset Bay or Crescent City, however Depoe Bay appears to be relatively well
flushed. With an open connection to the ocean and limited freshwater input, salinities in these
systems will tend to remain high and undergo smaller tidal and seasonal variations than the
drowned river mouth estuaries.
2.4.4 Drowned River Mouth and Bar-Built Estuaries
The remaining 31 estuaries can be split into two general classes based on geomorphology. The
first are bar-built estuaries, which are formed when ocean currents and wind form coastal dunes
that trap estuarine water behind them. Bar-built estuaries resemble lagoonal systems in having
low freshwater inputs but differ in having greater exchange with the ocean. The two generally
recognized bar-built estuaries in the PNW are Netarts Bay and Sand Lake. We classify a third
estuary, the Humboldt Estuary, as bar built. Though the Humboldt has been considered a
drowned river mouth estuary (e.g., Rumrill, 1998) there is no major river discharging into the
estuary. More appropriately, Emmett et al. (2000) classified the Humboldt as a lagoon, and these
authors pointed out that many lagoons are closed to the ocean during a portion of the year. The
difficulty with classifying Humboldt as a lagoon is that it has permanent jetties maintaining an
open connection with the ocean. Therefore, we suggest that with its current configuration the
Humboldt more closely resembles a bar-built estuary.
The second class is the drowned river mouth estuaries, which include 28 PNW estuaries.
Drowned river estuaries were created by flooding of river valleys as sea level rose during the
Holocene marine transgression after the last ice ago about 10,000 years ago (Emmett et al.,
2000). Drowned river mouth estuaries constitute the largest estuaries in the PNW, including the
Columbia River, Willapa Estuary, and Grays Harbor. Although formed by flooding of river
valleys, the geomorphology and size of many of the drowned river mouth estuaries in the PNW
are influenced by the formation of ocean-built bars. The drowned river mouth estuaries with
prominent ocean bars share characteristics with bar-built estuaries, and some authors have
classified them as bar built. For example, Willapa and Grays (Seliskar and Gallagher, 1983) as
well as the Salmon River (http://www.coastalatlas.net) have been classified bar-built estuaries.
While recognizing the importance of these ocean bars to their dynamics, we classify these
systems as drowned river mouth estuaries based both on their historic formation and the presence
of one or more moderate to large sized rivers discharging into the estuary.
45
-------
2.4.4.1 Tide- and River-Dominated Drowned River Mouth Estuaries
Drowned river mouth estuaries vary in the extent that their geomorphology and flushing are
driven by tidal versus fluvial processes. For example, the Columbia River and Rogue River are
considered to be strongly influenced by river flow (Cortright et al., 1987; McCabe et al., 1988)
while Willapa has relatively little freshwater input for its size. While categorization of systems
like the Columbia and Willapa is relatively straightforward, the difficulty comes in classifying
the full suite of PNW estuaries without rigorous criteria. The Australians separate the two types
of systems depending upon whether river or tide energy "dominated the evolution" of the estuary
(http://dbforms.ga.gov.au/www/npm.Ozcoast.glossary?pType=Audit). In a similar vein, Elliott
and McLusky (2002) define river-dominated estuaries as microtidal systems where rivers
provide the majority of the sediment and tide-dominated estuaries as macrotidal systems where
"there is a steady infilling of sediment provided at differing times from the sea and from the
river." Such geologically based definitions help highlight the factors driving the formation of an
estuary but they provide little insight into flushing, nutrient retention, or salinity patterns.
A more relevant approach to assessing estuarine vulnerability to nutrient enrichment is to
separate tide- versus river-dominated systems based on the relative extent of freshwater flow
through the estuary. To capture this relative input, we propose "normalized freshwater inflow"
metrics. The first step in generating these metrics is to calculate the total average annual volume
of rainfall over the entire watershed using the precipitation data in PRISM
(http://prism.oregonstate.edu/; Daly et al., 2007), which has records from 1971 to 2000. This
volume represents the total amount of freshwater impinging on the watershed, which then
evaporates, percolates into the soil, or runs off into the estuary. As a first-order simplification,
we assume that evaporation and percolation are similar among PNW watersheds, so that a
similar fraction of the rainfall ultimately flows into each of the estuaries. Thus, this total volume
of precipitation is a relative measure of the average annual freshwater flow into the estuary.
Once in the estuary, this freshwater inflow mixes with ocean water transported into the estuary
with the extent of mixing dependent, in part, upon estuary volume. Therefore, the next step is to
normalize the total annual precipitation (m3 of freshwater per year) by the total volume of the
estuary (m3). This metric has units of year"1 and is referred to as the annual "volume-normalized
freshwater inflow". Higher values of this metric indicate a more freshwater (river) dominated
system. Among the 17 PNW estuaries where bathymetry is available, the volume-normalized
freshwater inflow varies from 4 year"1 in the Humboldt Estuary to 2501 year"1 in the Rogue River
(Table 2-4). Because bathymetry is not available for most PNW estuaries, we developed an
alternative approach of normalizing the total annual precipitation by estuarine area. This annual
"area-normalized freshwater inflow" metric has units of m3 of freshwater per m2 of estuary per
year and is directly related to the volume normalized values (Figure 2-10). Since it is available
for all the estuaries, we use the area-normalized metric as our approach to ranking estuaries by
the extent of riverine influence.
Advantages of these watershed-scale metrics of freshwater input compared to flow data from
gauged stations are that they integrate precipitation over the entire watershed rather than
measuring flow in a single or limited number of tributaries, they can be calculated for estuaries
that are not gauged, and they can be linked to climate change scenarios to predict altered
freshwater inflow under different precipitation regimes. Advantages over simply using the ratio
of the drainage area to the estuary size are that the normalized freshwater inflow metrics capture
46
-------
Table 2-4. Freshwater sources into estuaries, normalized freshwater inflow, mouth width, and structure of mouth for drowned river
mouth and bar-built estuaries in the PNW. Bar-built estuaries are italicized. Estuaries with area-normalized freshwater inflows less
than 175 m3 m"2 year"1 are classified as tide dominated and estuaries with greater values are classified as river dominated. The river-
dominated estuaries are further subdivided into "highly river-dominated systems" (dark gray) with area-normalized freshwater inflow
^91 ^ 9 1 ^ 9
values >400 m m" year" and "moderately river-dominated systems" (light gray) with values > 175 m m" year" and <400 m m"
year"1. Estuarine volume obtained from http://ian.umces.edu/neea/ with the exception of Salmon River which was from Johnson and
Gonor (1982). The estimates for the normalized freshwater inflows for the Columbia River are based on the annual river discharge
versus the total volume of precipitation in the watershed and are not directly comparable to the values in the other estuaries.
ESTUARY
Columbia
Klamath
Rogue
Quinault
Quillayute
Chetco
Hoh
Queets
Smith (CA)
MAJOR FRESFIWATER
SOURCES
Columbia River and multiple creeks
Klamath River
Rogue river
Quinault River
Quillayute River and
2 subsidiary creeks
Chetco River
Hoh River
Queets River
Smith River and
1 subsidiary creek
NORMALIZED FRESFIWATER
INFLOW
AREA
NORMALIZED
(m3 m"2 year"1)
370
18,082
6,537
6,286
5,220
3,771
3,468
2,796
2,159
VOLUME
NORMALIZED
(year"1)
86
2435
2501
MOUTH
WIDTH
(m)
5038
350
280
38
250
78
28
70
110
ESTUARY MOUTH STRUCTURE
Jetties on both side of the mouth. Note:
Freshwater inflow based on river discharge
and not precipitation.
Mouth constrained by bluffs to the north
and south. Dunes may constrict mouth
width.
Jetties on both side of the mouth. Mouth
may have migrated before jetties.
Bluff constrains mouth to the north. Dunes
may constrict mouth width.
Jetties on both side of the mouth. Dunes
direct estuary northward of mouth.
Jetties on both sides of the mouth.
Mouth constrained by bluffs to the north
and south. Dunes may constrict mouth
width.
Dunes may constrict mouth width. Forms
dendritic channels behind dunes and directs
estuary northward of mouth.
Mouth constrained by bluffs to the north.
Dunes direct estuary southward of mouth.
-------
ESTUARY
Mad
Eel
Coquille
Sooes
Nehalem
Umpqua
Nestucca
Necanicum
Copalis
Siletz
Siuslaw
Alsea
Salmon
Tillamook
MAJOR FRESHWATER
SOURCES
Mad River
Eel River and several subsidiary
creeks
Coquille River
Sooes River
Nehalem River
Umpqua River and Smith River
Nestucca River and
Little Nestucca River
Necanicum River, Neawanna Creek,
and several subsidiary creeks
Copalis River and Cedar Creek
Siletz River, Schooner Creek, and
3 subsidiary creeks
Siuslaw River and
North Fork Siuslaw River
Alsea River and several creeks
including Lint and Drift Creeks
Salmon River and several
subsidiary creeks
Kilchis, Wilson, Tillamook, Trask,
and Miami Rivers
NORMALIZED FRESHWATER
INFLOW
AREA
NORMALIZED
^9 1
(m m" year" )
1,779
997
695
539
499
484
420
387
318
284
227
211
177
116
VOLUME
NORMALIZED
(year"1)
524
143
293
219
137
125
141
392
62
MOUTH
WIDTH
(m)
137
150
175
50
180
425
110
60
20
100
225
140
40
360
ESTUARY MOUTH STRUCTURE
Dunes direct estuary southward of mouth
and form a "lagoon" to the north of the
current mouth. Mouth may occasionally
close .
Forms dendritic channels behind dunes.
Dunes direct estuary northward of mouth.
Jetties on both side of the mouth.
Dunes direct estuary southward of mouth.
Mouth constrained by land to north and
south.
Jetties on both side of the mouth. Dunes
direct estuary northward from mouth.
Jetties on both side of the mouth.
Dunes direct estuary northward from mouth.
Mouth constrained by land to the south.
Dunes direct estuary northward and then
southward from mouth.
Dunes direct estuary southward of mouth.
Dunes direct estuary southward of mouth.
Dunes may constrict mouth width. Mouth
constrained by land to the north.
Jetties on both side of the mouth. Dunes
direct estuary southward from mouth.
Mouth constrained by bluffs to the south.
Mouth constrained by land to the north.
Dunes direct estuary southward from mouth.
Jetties on both side of the mouth. Dunes
direct estuary southward from mouth.
Dunes result in accumulation of freshwater
flow from five rivers.
-------
VO
ESTUARY
Waatch
Yaquina
Grays
Coos
Sand Lake
Willapa
Humboldt
Netarts
MAJOR FRESHWATER
SOURCES
Waatch River
Yaquina River and several creeks
draining into sloughs
Chehalis River and Elk, Johns, and
Hoquiam Rivers and several
subsidiary creeks
Coos River and several creeks
discharging into sloughs including
South and Isthmus Slough, and
Catching and Palouse Creek
Sand Creek and Jewell Creek and 2
subsidiary creeks. Groundwater?
Bear, Bone, Cedar, Naselle, Neman,
Niawiakum, North, Palix, and
Willapa rivers
Freshwater Creek and Elk River and
to lesser extent Jacoby Creek and
Salmon Creek
13 minor creeks
NORMALIZED FRESHWATER
INFLOW
AREA
NORMALIZED
^9 1
(m m" year" )
101
63
58
53
28
17
12
11
VOLUME
NORMALIZED
(year"1)
42
32
14
6
4
8
MOUTH
WIDTH
(m)
32
290
2750
620
135
9165
842
114
ESTUARY MOUTH STRUCTURE
Mouth constrained by bluff to the north.
Dunes may constrict mouth width.
Jetties on both side of the mouth.
Jetties on both side of the mouth.
Jetties on both side of the mouth. Dunes
direct main stem of estuary northward from
mouth.
Dunes form mouth, with extensive sand
accumulation behind dunes. Mouth may
occasionally close off estuary.
Mouth constrained by land to the north.
Dunes direct estuary southward from mouth.
Dunes result in accumulation of freshwater
flow from rivers.
Jetties on both side of the mouth. Estuary
runs behind dunes both northward and
southward. Dunes result in accumulation of
freshwater flow from rivers and creeks.
Dunes direct estuary southward from mouth.
-------
log FW , = 0.90 * log(FW ) - 0.10
0 volume °^ Area'
r =0.91
i i i i i i 11 r
io3
io4
-I JJ
__ __
Freshwater Inflow Normalized by Estuarine Area (m m year
Figure 2-10. Relationship between freshwater inflow normalized by estuarine area and
normalized by estuarine volume.
regional differences in rainfall, monthly values can be generated to evaluate seasonal changes in
potential flushing, and, again, the metric can be linked to climate change scenarios.
The area-normalized freshwater inflow was calculated for all the drowned river mouth estuaries
and bar-built estuaries (Table 2-4) except for the Columbia Estuary. Total precipitation was not
available for the entire Columbia Basin, though it was possible to generate a rough estimate from
the Columbia River's total annual discharge of approximately 198,000,000 acre-feet of water
(http://vulcan.wr.usgs.gov/VolcanoesAVashington/ColumbiaRiver/description_columbia_river.ht
ml). This annual river discharge is equivalent to a normalized freshwater inflow of
approximately 370 m3 per m2 per year. This estimate is a substantial underestimate compared to
the values for other estuaries since it does not include any water lost to evaporation or
percolation, which would be substantial over the entire Columbia Basin. While this value is not
directly comparable to the other values, it indicates that the Columbia Estuary has a high river
input, which is consistent with the tidal riverine segment constituting over a third of the entire
estuarine area (see Table 2-2).
Area-normalized freshwater inflow values vary over a thousand-fold, from about 11 in Netarts
^91
Bay to 18,000m m" year" in the Klamath. The metric successfully separates estuaries
generally considered river-dominated systems (e.g., Klamath and Rogue) from those considered
tide-dominated systems (e.g., Yaquina) suggesting that it can be used to classify less well studied
systems. Additionally, spatial and seasonal patterns in salinity variations in these estuaries
appear to be related to normalized freshwater inflow. Figure 2-11 shows salinity versus distance
from the mouth for eight estuaries where both dry and wet season salinities are available. These
-2
eight estuaries range in area-normalized freshwater inflow from 11 (Netarts) to 695 m m" year
-i
50
-------
£. 20 -
•5 I? -
^ 10 -
I)
35 ,
30 -
| 25-
>, 20
i 15 -
1'°
5 -
0
a) Netarts
-I ^ 4
Moulh, km
cllTillaniook
5 10
Distance from Mouth, km
IS
g) I'mpqua
35 -I
30
|25-
;-". 20
-
10 H
3
0
5 10 15
Distance from Mouth, km
20
35 -i
| 25
*, 20 ]
i 15
'•5 10-
" 5
0
35
30 J
•S 15
I '°-
s
35 -i
30 -
25 -
20-
15 -
10
5 -
0
0
b) Coos
c) Yaquiua
f—t—*-
1)
-•- Dr>-
M-H t r i»
H-H.J.; ||
;. i
0 -)
ll
5 10 15 20
Distance from Mouth, km
e) Salmon
5 1(1 15
Distance from Mouth, km
0 Aiwa
35 -i
30 -
£20-
5
0
4
Distance from Mouth, km
h) Coqiiilh-
-168
Distance from Mouth, km
10
10
Distance from Mouth, km
Figure 2-11. Salinity versus distance from the mouth for the wet and dry seasons for estuaries with area-normalized freshwater inflow
ranging from about 11 to 695 m3 m'2 year"1. Salinity variations are provided for bar-built, tide-dominated, moderately river-
dominated, and highly river-dominated systems. The symbols indicate the median values and the error bars represent the 25* and 75*
percentiles. Salinity data are from the Oregon Department of Environmental Quality (http://deql2.deq.state.or.us/lasar2/) except for
panel e. Data for panel e are from short-term deployments of sondes described in Chapter 4 Note that the Salmon River has the
largest volume-normalized freshwater inflow, reflecting its shallowness.
-------
(Coquille). Systems with low freshwater inflow have less salinity variation along the main axis
of the estuary, salt penetrates further into these systems, and they have less seasonal variations in
salinity than river-dominated systems. For the two estuaries with the highest values (Umpqua
and Coquille), there is essentially a linear decline in dry season salinity with distance from the
mouth. Finally, there is a positive relationship between area-normalized freshwater inflow
values and variation in salinity over a tidal cycle at the mouth of the seven target estuaries
(Figure 4-11). Thus, it appears that the area-normalized freshwater inflow metric captures key
aspects of tide versus river domination as measured by salinity.
The next question becomes what value to use to separate tidal from riverine systems. A
reasonably distinct break in values occurs between the Salmon and Tillamook estuaries, with the
Tillamook value about two-thirds of the Salmon value (Table 2-4). This break is also associated
with a difference in wet and dry season salinity variations (Figure 2-1 le). Based on this pattern,
we propose a value of >175 m m"2 year"1 as a preliminary threshold for defining river-
dominated estuaries. There is also a moderate break in values between the Nestucca and
Necanicum (Table 2-4), which is also reflected in the salinity patterns in the Salmon and Alsea
versus the Umpqua and Coquille (Figure 2-11). This pattern in salinity variation suggests a
subdivision of river-dominated systems into "highly river-dominated" systems with inflow
values greater than approximately 400 m3 m"2and "moderately river-dominated" systems with
inflow values between 175 m3 m"2 and 400 m3 m"2 year"1.
This approach classifies 22 of the 28 drowned river mouth estuaries as river-dominated and 6 as
tide-dominated (Table 2-4). Of the river-dominated estuaries, 16 are classified as highly river-
dominated and 6 as moderately river-dominated. In terms of geomorphology, the river-
dominated systems, and the highly river-dominated systems in particular, tend to have fewer
freshwater inputs compared to the tide-dominated systems which tend to have smaller freshwater
inputs distributed across several tributaries. Another pattern is that all of the small (<5 km2)
river mouth estuaries are classified as river dominated with the exception of the Waatch, while
10 of the 14 estuaries <5 km are classified as highly river dominated. All three bar-built
estuaries have very low normalized freshwater inflows, indicating that functionally they more
resemble the tide-dominated systems.
In terms of vulnerability, our initial hypothesis is that, all other factors being equal, estuaries
with higher normalized freshwater inflow will be less susceptible to terrestrially derived nutrient
enrichment because of higher flushing. The higher river flow in riverine systems may actually
increase the total loading from the watershed, but the high flushing should reduce the ability of
phytoplankton to maintain high populations. This suggestion is consistent with the trend towards
a decrease in median dry season chlorophyll a with increasing volume-normalized freshwater
inflow (Figure 4-6). The data from this set of estuaries suggest that summer chlorophyll a levels
in the estuaries may be determined by flushing time. Evaluation of this proposed relationship
between normalized river inflow and chlorophyll a concentrations will require additional water
quality studies across a range of estuary types as well as developing more complete water quality
models to explain differences among estuaries with similar riverine inputs. Tide-dominated
systems have close coupling with the coastal ocean and as a result may experience high nutrient
and chlorophyll a levels and low dissolved oxygen conditions as a result of intrusion of ocean
water into the estuary.
52
-------
2.5 Classification of Estuaries by NWI Wetland Classes
The next step in classifying these drowned river mouth and bar-built estuaries was to determine
similarity among estuaries based on wetland distributions. Distribution of NWI wetland types
serves as an integrated response to multiple drivers, including estuarine geomorphology,
watershed size and type, freshwater inflow, oceanic exchange, salinity structure, and tidal
exposure. Thus, patterns of NWI wetlands classes are potentially a better indicator of
environmental similarity than a classification based on any of those environmental drivers
individually. Classification by NWI classes also provides a mechanism to identify estuaries with
similar wetland resources, which could be used in establishing management prioritizations or
assist in designing specific mitigation actions.
Utilizing the consolidated NWI codes (Table 2-1), the 31 estuaries were first clustered based on
the actual areas of the wetland/habitat classes. This and the following cluster analyses are based
on the NWI available in 2006 since there were only minor changes in these larger estuaries in the
recent update. Since this analysis groups estuaries by habitat areas, classification by area will
tend to group larger and smaller estuaries, and in terms of management is probably most useful
in identifying groups of estuaries with similar extents of wetland resources. The 31 estuaries
clustered into four significant classes using the default p value of 0.05 (Figure 2-12). As
expected, the clusters broke out by estuarine size. One group (Cluster A) consisted of the three
largest estuaries: Willapa, Grays Harbor, and the Columbia River. The next group (Cluster B)
consisted of the next four largest estuaries: Humboldt, Coos, Umpqua, and Tillamook. Cluster C
consisted of the 10 moderate-sized estuaries, ranging in size from 4.28 km2 (Sand Lake) to
19.96 km2 (Yaquina Estuary). The final group (Cluster D) consisted of the suite of smaller
estuaries (0.56 km2 to 3.1 km2). The effect of increasing the p value was to break out the
Columbia as a separate group from Grays Harbor and Willapa (at p=0.10), but further increasing
the significance values did not split out any additional groups (Figures 2-13a to 2-13d).
A similar analysis was conducted using the relative proportion of the area of each of the
consolidated NWI classes. Use of proportional areas removed the direct effect of estuary size
and should better identify groups of estuaries that are functionally similar compared to the
classifications by actual areas. Based on the relative proportions, the 31 estuaries grouped into
three clusters at p=0.05 (Figure 2-14). One group (Cluster A) consisted of eight estuaries
showing a wide range in size, from the Quinault (0.66 km2) to the Columbia River (669 km2).
Though varying greatly in size, members in this cluster consist of the eight estuaries with the
largest normalized freshwater inflow (Table 2-4), assuming that the Columbia ranks among the
top eight. This grouping indicates that there are characteristic wetland profiles in the most highly
river-dominated systems. Another group (Cluster C) consisted of the Salmon and Waatch, which
showed little similarity with the other groups. These two systems separated from the other
estuaries largely due to their high proportion of upper marsh habitat (estuarine, irregularly
flooded, rooted class of Table 2-1).
The third group (Cluster B) consisted of 21 estuaries including both river- and tide-dominated
river mouth estuaries as well as all three of the bar-built estuaries. Size varied widely in this
group, from 0.56 km2 (Sooes) to 390.9 km2 (Willapa). The estuaries making up this cluster are
so diverse that it does not seem useful in grouping systems in terms of nutrient dynamics.
However, increasing the significance levels helped to separate estuaries with similar attributes
53
-------
(Figures 2-15a to 2-15d). The 10% significance level separated Grays Harbor and Willapa Bay
as a group (Figure 2-15a), two estuaries that share many similarities. At a significance level of
40%, 12 clusters were identified (Figure 2-15c) while the 60% significance level identified 16
clusters (Figure 2-15d). Using the criterion of combining estuaries with > 75% similarity results
in 10 "functional" groups at the 40% significance level and 12 "functional" groups at the 60%
significance level (Figure 2-15d).
Choosing which of these classifications is the "best" is not a statistical question but rather
depends upon the goals of the classification. Our interpretation is that the classification based on
the 60% significance level along with combining estuaries with > 75% similarity best captures
ecological similarity in terms of drivers related to wetland patterns while avoiding separation of
very similar systems (Figure 2-15d). Compared to the 40% significance level, the classification
based on the 60% significance level separated all the tide-dominated estuaries from the highly
river-dominated systems, though there is still some mixing of the tide-dominated, moderately
river-dominated, and bar-built estuaries. However, the practical limitation of this classification
is that it results in 12 functional groups, which may constitute too fine a division for
management. An alternative might be to use the classification based on the 10% significance
level which consists of four groups (Figure 2-15a).
54
-------
0-r
20--
D
cc
40--
60--
B
Pi
80--
100-L
CD
CD CD -^ W
. . Q. 3 O O
CD CD D" O O
1 - - E O
-p ST £
fT\ :"=
CD
LU
COaJ -C
*-^=H.gi:i!8!53.g:=g)ffl
o c
- CD CD '5 O -^ :=
•< z1 I S"-i 5 w
as^-e
o
O
CD
O a5
CD
O
(3 <5
CD
E » E
^°^0
li 8
1 a
I
O
Estuaries
Figure 2-12. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the area
of the consolidated NWI habitats. Estuaries joined with red lines are not significantly different
(p>0.05) based on SIMPROF analysis. Clusters with >75% similarity (horizontal line) are
combined in the analysis.
55
-------
Figure 2-13a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
area of the consolidated NWI habitats. Estuaries joined with red lines are not significantly
different (p>0.10) based on SIMPROF analysis. Clusters with >75% similarity (horizontal line)
are combined in the analysis.
Figure 2-13b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
area of the consolidated NWI habitats. Estuaries joined with red lines are not significantly
different (p>0.20) based on SIMPROF analysis. Clusters with >75% similarity (horizontal line)
are combined in the analysis.
56
-------
Figure 2-13c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
area of the consolidated NWI habitats. Estuaries joined with red lines are not significantly
different (p>0.40) based on SIMPROF analysis. Clusters with >75% similarity (horizontal line)
are combined in the analysis.
Figure 2-13d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
area of the consolidated NWI habitats. Estuaries joined with red lines are not significantly
different (p>0.60) based on SIMPROF analysis. Clusters with >75% similarity (horizontal line)
are combined in the analysis.
57
-------
20-r
40-
«*>
J 6°--
cc
80-
B
1
100-L
•2 3 § « £ §
US*
<3° 5,
cc
E ^
^, ro ^
a> ^ O
w cc w cc cc w
CD >? 9- g c a? 5?
CD
O
2
O
= O
3 -^
O
IIII
i|Ss
Estuaries
a
o
O
LU -Q
-^ CC
O O
Q O
= CD
H
CC
O
CD
™ s
T3
CC
.
cc
W
Figure 2-14. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
relative proportion of the consolidated NWI habitats. Estuaries joined with red lines are not
significantly different (p>0.05) based on SEVIPROF analysis. Clusters with >75% similarity
(horizontal line) are combined in the analysis.
58
-------
Figure 2-15a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
relative proportion of the consolidated NWI habitats. Estuaries joined with red lines are not
significantly different (p>0.10) based on SEVIPROF analysis. Clusters with >75% similarity
(horizontal line) are combined in the analysis.
Figure 2-15b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
relative proportion of the consolidated NWI habitats. Estuaries joined with red lines are not
significantly different (p>0.20) based on SEVIPROF analysis. Clusters with >75% similarity
(horizontal line) are combined in the analysis.
59
-------
20-r
40--
2 60-
1
53
_E
80-4-L p
..ifn
K
?1
.g ffiZ
8 %':
I"
w m
> Q.
M = t
'
ISIS
o 8 | £ £ S §
e s .a | 5 11
111 Pis
3
Estuaries
Figure 2-15c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
relative proportion of the consolidated NWI habitats. Estuaries joined with red lines are not
significantly different (p>0.40) based on SEVIPROF analysis. Clusters with >75% similarity
(horizontal line) are combined in the analysis.
20-r
40--
£ 60-
1
100-L
F
ri
•S 3
||
Estuaries
Figure 2-15d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
relative proportion of the consolidated NWI habitats. Estuaries joined with red lines are not
significantly different (p>0.60) based on SEVIPROF analysis. Clusters with >75% similarity
(horizontal line) are combined in the analysis.
60
-------
This classification separated out the eight highly riverine systems, Willapa and Grays Harbor,
and the Salmon and Waatch. The limitation is that the fourth group consists of 19 estuaries
ranging from bar-built to highly riverine dominated which appear to have different nutrient
dynamics. One of the joint scientific and management challenges then becomes determining
"how similar is similar enough".
2.6 Watershed Classification
Grouping estuaries by NWI classes identified suites of estuaries with similar patterns of wetland
habitats and presumably similar vulnerabilities to nutrient enrichment. A different question is
how the watersheds of the 31 estuaries group together. Within the same climatic regime,
estuaries with similar land cover patterns in the watersheds presumably have generally similar
nutrient loadings. To evaluate similarity in land cover patterns, the 31 watersheds were clustered
using the land cover classes from the 2001 NOAA dataset. The NOAA coverage for the
Columbia extended to approximately the Bonneville Dam, which captures the watershed
adjacent to the Lower Columbia, representing about 2.5% of the entire Columbian Basin. Thus,
the results for the Columbia watershed need to be interpreted in light of the fact that only a small
portion of the entire land mass draining into the Columbia River was captured. The NOAA
coverage for the Klamath was truncated at the northeastern segment, but as mentioned this area
was filled in by using the 2001 NLCD data. A complexity in the PNW in that the presence of
nitrogen-fixing alder can be an important component of nitrogen dynamics in these watersheds
(see Sections 3.7 and 4.2). NOAA's "deciduous forest" and "mixed forest" classes capture alder
but do not separate them out from other deciduous trees. This level of detail should be sufficient
to capture general similarities among watersheds, though differences in alder coverage may
result in somewhat different nitrogen loadings among otherwise similar watersheds.
2.6.1 Watershed Classes
The analysis based on the areas of the land cover classes resulted in seven clusters at a
significance level of 5% (Figure 2-16). As expected, there was a tendency for larger watersheds
to cluster together. The five estuaries with the largest watersheds - the highly riverine
dominated Columbia, Klamath, Rogue, Umpqua, and Eel - all formed one group (Cluster A).
Similarly, the three estuaries with the smallest watersheds (<100 km2) grouped together in
Cluster G while the estuaries with the next smallest watersheds (105 km2 to 216 km2) formed a
related cluster (Cluster F). Most of the estuaries with moderate-sized watersheds grouped in
Clusters B, C, and D, which had a relatively high similarity (>55%) among the three groups.
The Humboldt formed a unique cluster (Cluster E) that had moderate similarity (<50%) to the
other moderate-sized watersheds. Using different significance levels in the cluster analysis did
not substantially change the classifications (Figure 2-17a to 2-17d), with the major changes being
the separation of the Eel Estuary from the functional group consisting of the other estuaries with
large watersheds (A-D in Figure 2-17d) and splitting the tide-dominated Waatch from the bar-
built Sand Lake and Netarts watersheds (Figure 2-17 c).
61
-------
O-i
20-
40-
co
E
w
60-
80-
100-
I
r
co co
f |
A
P
1
&. s°
^
B
E
D
C
r
1
1
i
5 fif if Siipl^3
•§
CD
O
O
1
z o 3=; w
D 0 ^ CD
" "K O O
Vi -P °
0 | w
'§.
o
0
w
F
canicum
5
G
..p
i
C CD W ^
§-^: ~C o
5 a TO
"m CD CO
wEz^
co
W
Estuaries
Figure 2-16. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data. Estuaries joined with red
lines are not significantly different (p>0.05) based on SIMPROF analysis. Clusters with >75%
similarity (horizontal line) are combined in the analysis.
62
-------
Figure 2-17a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data. Estuaries joined with red
lines are not significantly different (p>0.10) based on SIMPROF analysis. Clusters with >75%
similarity (horizontal line) are combined in the analysis.
Figure 2-17b. Cluster analysis of 31 PNW estuaries at the 20% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data. Estuaries joined with red
lines are not significantly different (p>0.20) based on SIMPROF analysis. Clusters with >75%
similarity (horizontal line) are combined in the analysis.
63
-------
100-L
Estuaries
Figure 2-17c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data. Estuaries joined with red
lines are not significantly different (p>0.40) based on SIMPROF analysis. Clusters with >75%
similarity (horizontal line) are combined in the analysis.
O-r
Estuaries
Figure 2-17d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
areas of the land cover classes in the NOAA 2001 watershed data. Estuaries joined with red
lines are not significantly different (p>0.60) based on SIMPROF analysis. Clusters with >75%
similarity (horizontal line) are combined in the analysis.
64
-------
The clustering based on relative proportion of the areas of the land cover classes will tend to
group estuaries associated with structurally similar watersheds. For example, clustering based on
relative proportions of land use classes could group highly urbanized watersheds. In the PNW,
there was a high degree of similarity (>70%) among all the watersheds (Figure 2-18). To a large
extent, this similarity reflects the high percentage of the evergreen land cover class, which
ranged from 36% to 80% in the watersheds. The mixed forest and scrub/shrub classes were also
common across the watersheds, contributing 7% to 44% of the area. The similarity also reflects
the relatively low extent of agriculture and urbanization in these PNW watersheds. Clustering at
the 5% significance level resulted in nine groups (Figure 2-18) though joining clusters with
>75% similarity reduced this to four functional groups. The first functional group (Clusters A-
D) largely consisted of watersheds associated with river-dominated estuaries, with the notable
exceptions of Grays Harbor and Willapa Estuary. Both watershed and estuary size within this
group varied almost 700-fold. The second functional group (Clusters E and F) consisted of the
Columbia, Eel, and Humboldt estuaries. The third functional group (Clusters G and H) was a
mix of estuarine types with moderate to large watersheds. The last group consisted of Netarts
Bay, a bar-built estuary, which formed a unique cluster (Cluster I). Netarts separated from the
other estuaries largely due to the high percentage of unconsolidated shore, 13.8% compared to
<1% in most other watersheds, but resembled other watersheds in the high proportions of
evergreen and mixed forest land classes.
Increasing the significance level further divides the watersheds, with 11 groups identified at the
40% significance level (Figure 2-19c) and 13 at the 60% significance level (Figure 2-19d).
These finer divisions appeared to better capture similar watersheds, such as splitting the Grays
Harbor and Willapa watersheds from the Copalis. However, all of these divisions occurred at
similarities >75% so they did not increase the number of functional groups from the four
identified at the 5% significance level. Until future research shows the need to more finely
divide watersheds based on the relative proportions of land cover classes, we suggest that the
four functional groups identified at the 5% significance level (Figure 2-18) are sufficient for an
initial analysis of estuarine vulnerability based on watershed structure.
2.7 Crosswalk of Wetland and Watershed Classifications
To further identify functionally similar estuaries, we conducted a matrix match ("crosswalk") of
the classifications by wetlands and watersheds. Estuaries were identified that overlapped both by
clustering on wetland and land cover areas (Tables 2-5 and 2-6) and by clustering on relative
proportions of wetlands and land cover (Tables 2-7 and 2-8), with groups with >75% similarity
joined into functional groups. The basic assumption of the crosswalks is that estuaries grouped
by both wetland and watershed characteristics share similar environmental conditions.
Specifically, estuaries grouped by area share similar extents of wetland resources and land cover
classes. These groups are most relevant in developing management approaches that scale to
area, such as calculating total watershed loadings or strategies to protect the greatest extent of a
wetland class. In comparison, grouping by relative proportions more closely captures functional
attributes, and is probably the better approach to developing management strategies related to
nutrient vulnerability at a system level. Crosswalks were conducted based on the clustering
using the 5% and the 60% significance levels. Crosswalks based on the 5% significance level
are at a coarser resolution and will tend to have fewer but more variable co-clustered estuaries
compared to crosswalks based on the 60% significance level.
65
-------
70-r
75-
80--
85--
90--
95--
E
B
H
D
n
100-L
W CD .W CD (DO
CD CD 5 D" D) "m
-^ CD
O O
5 o
CD
O
O
Estuaries
"m o~ c fi CD
w CD J w z
Figure 2-18. Cluster analysis of 31 PNW estuaries at the 5% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data. Estuaries
joined with red lines are not significantly different (p>0.05) based on SEVIPROF analysis.
Clusters with >75% similarity (horizontal line) are combined in the analysis.
66
-------
70-r
75
100-L
w ro w ro •§ _
agsffi|6
nil
Is
K>
8 gE
|S
| w£i
Estuaries
Figure 2-19a. Cluster analysis of 31 PNW estuaries at the 10% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data. Estuaries
joined with red lines are not significantly different (p>0.10) based on SEVIPROF analysis.
Clusters with >75% similarity (horizontal line) are combined in the analysis.
70 1
75-
80-
Js 85-
E
CO
90-
95-
100-
A
R
-1-
i
-•-i
i
V> [D WJ [O C1J -1— U U •'T- -1— '
j>. D= g. g^ B ro g ra '
S|8|<§|S |gc
ro
c
1
\
.
A
n
F
D
-
ffl ([i C~ ±
OJ -t^ O
"3
F
Q
T
H
n
T
ifllSoS^^row-S-^-^E
5§ -POg-~-l0.20) based on SEVIPROF analysis.
Clusters with >75% similarity (horizontal line) are combined in the analysis.
67
-------
70-r
Estuaries
Figure 2-19c. Cluster analysis of 31 PNW estuaries at the 40% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data. Estuaries
joined with red lines are not significantly different (p>0.40) based on SEVIPROF analysis.
Clusters with >75% similarity (horizontal line) are combined in the analysis.
70 T
Estuaries
Figure 2-19d. Cluster analysis of 31 PNW estuaries at the 60% significance level based on the
relative proportions of the land cover classes in the NOAA 2001 watershed data. Estuaries
joined with red lines are not significantly different (p>0.60) based on SEVIPROF analysis.
Clusters with >75% similarity (horizontal line) are combined in the analysis.
68
-------
The columns in Tables 2-5 to 2-8 are the clusters based on the watershed analysis, while the rows
are the clusters based on wetland analysis. Estuaries listed within the same cell are systems that
were grouped together in both the watershed and wetland analyses. For example, Coos and
Tillamook were clustered together in both the wetland and land cover analyses on areas (Table 2-
5). Clusters are ordered in the tables by the mean size of the watersheds or estuaries within the
cluster, though proximity in the table does not necessarily connote greater similarity and Figures
2-12 to 2-19 should be consulted for the actual similarities. To avoid confusion, the clusters
based on wetlands were designated as A', B', etc.
In the crosswalk based on areas at the 5% significance level (Table 2-5), the five highly river-
dominated estuaries with large watersheds (column A) were dispersed across the four wetland
clusters (rows A ' to D'), indicating that there are large watersheds with large estuaries (A.-A ') and
large watersheds with small estuaries (A-D'). The opposite pattern was not observed, and there
were no small watersheds with large estuaries (i.e., no estuaries falling into F-/4 ', G-A ', F-B', or
G-B'). Another pattern is that the remaining river-dominated estuaries fell into relatively few
cells, and were separated from the tide-dominated estuaries with the exception of the Yaquina
occurring in cell D-C". Additionally, nine of the highly river-dominated estuaries fell into three
cells representing moderate to small estuaries with moderate-sized watersheds (C-C", C-D\ and
D-D'). Four of the other tide-dominated estuaries formed two pairs (Grays Harbor & Willapa
and Coos & Tillamook), while the Waatch was not linked with any of the other estuaries. The
pattern based on areas at the 60% significance level (Table 2-6) was the same with the exception
that the Eel River and Waatch were broken into separate watershed clusters (columns).
One obvious pattern in the crosswalk by relative proportions of wetland and land cover classes
at the 5% significance level (Table 2-7) is the grouping of seven of the highly river-dominated
systems in a single cell (A-D-A'). Excluding the Columbia River, these seven estuaries had the
highest normalized freshwater inflow values (Table 2-4). Thus, estuaries with the highest
freshwater inflow appear to have wetland and watershed patterns that differ from other river-
dominated or tide-dominated systems. With the exception of the Salmon River, the remaining
moderate-sized river-dominated estuaries were grouped into two cells (A-D -B' and G-H -B'),
though they were mixed with tide-dominated estuaries. At the 60% significance level (Table 2-
8), the four estuaries with the highest normalized freshwater inflow values (exclusive of the
Columbia) still grouped together (A-E -A '-C') though the three estuaries with next highest
freshwater inputs were separated into two cells (A-E -E' and A-E -D') with lower average
estuary size. At this higher significance level, the tide-dominated Grays Harbor and Willapa
estuaries were grouped together, as were the Coos and Yaquina estuaries. Assuming that
grouping by relative proportions captures functional aspects of estuaries and watersheds, we
predict that the estuaries grouped at the 60% significance level will have similar nutrient
dynamics. The practical difficulty in using the crosswalk based on the 60% significance level is
the large number of groups, which is reduced by more than half with the crosswalk based on the
5% significance level. It is not clear whether these groups show sufficient similarity for them to
be managed in a similar fashion, though research to help resolve this question is outlined in
Section 2.9.
69
-------
Table 2-5. Crosswalk of the cluster analyses by the area of NWI consolidated wetland classes (rows) and the area of the NOAA land
cover classes (columns) at the 5% significance level. Letters refer to the cluster groups in Figures 2-12 for wetlands and 2-16 for
watersheds. Mean watershed area and mean estuary area within the clusters are given in parentheses, with the area of the Columbia
watershed for the Lower Columbia to the Bonneville Dam. The clusters are ordered by mean watershed or estuary size and proximity
in the table does not necessarily imply high similarity. The geomorphology is indicated by color with tide-dominated estuaries in blue,
moderately river-dominated estuaries in green, highly river-dominated estuaries in red, and bar-built estuaries in black.
Clusters on Area /
5% Significance
NWI WETLAND CLUSTER
(Mean Estuary Area for Group)
A'
(441 km2)
B'
(49 km2)
C'
(ll.lkm2)
D'
(1.5km2)
WATERSHED LAND USE CLUSTERS
(Mean Watershed Area For Group)
A
(15161km2)
Columbia
Umpqua
Eel
Klamath
Rogue
B
(3197km2)
Grays
Willapa
C
(2009 km2)
Coquille
Nehalem
Siuslaw
Quillayute
Smith
D
(1087km2)
Tillamook
Alsea
Nestucca
Siletz
Yaquina
Chetco
Hoh
Mad
Queets
Quinault
E
(571 km2)
Humboldt
F
(156km2)
Copalis
Necanicum
Salmon
Sooes
G
(45 km2)
Netarts
Sand Lake
Waatch
-------
Table 2-6. Crosswalk of the cluster analyses by the area of NWI consolidated wetland classes (rows) and the area of the NOAA land
cover classes (columns) at the 60% significance level. Letters refer to the cluster groups in Figures 2-13d for wetlands and 2-17d for
watersheds. Mean watershed area and mean estuary area within the clusters are given in parentheses, with the area of the Columbia
watershed for the Lower Columbia to the Bonneville Dam. The clusters are ordered by mean watershed or estuary size and proximity
in the table does not necessarily imply high similarity. The geomorphology is indicated by color with tide-dominated estuaries in blue,
moderately river-dominated estuaries in green, highly river-dominated estuaries in red, and bar-built estuaries in black.
Clusters on Area /
60% Significance
NWI WETLAND CLUSTERS
(Mean Estuary Area for Group)
A'
(669 km2)
B'
(327 km2)
C'
(49 km2)
D'
(ll.lkm2)
E'
(1.5km2)
WATERSHED LAND COVER CLUSTERS
(Mean Watershed Area For Group)
A-D
(16567
km2)
Columbia
Umpqua
Klaniath
Rogue
E
(9536
km2)
Eel
F
(3197
km2)
Grays
Willapa
G
(2009
km2)
Coquille
Nehalem
Siuslaw
Quillayute
Smith
H
(1087
km2)
Coos
Tillamook
Alsea
Nestucca
Siletz
Yaquina
Chetco
Hoh
Mad
Queets
Quinault
I
(571 km2)
Humboldt
J
(156km2)
Copalis
Necanicum
Salmon
Sooes
K
(49 km2)
Netarts
Sand Lake
L
(38 km2)
Waatch
-------
Table 2-7. Crosswalk of the cluster analyses by the relative proportion of NWI consolidated wetland classes (rows) and the relative
proportion of land cover classes (columns) at the 5% significance level. Letters refer to the cluster groups in Figures 2-14 and 2-18.
Clusters with high similarity (>75%) were combined for the analysis. Mean watershed area and mean estuary area within the clusters
are given in parentheses. The clusters are ordered by mean watershed or estuary size and proximity in the table does not necessarily
imply high similarity. The geomorphology is indicated by color with tide-dominated estuaries in blue, moderately river-dominated
estuaries in green, highly river-dominated estuaries in red, and bar-built estuaries in black.
Clusters on Relative Proportion /
5% Significance
8-5 3
%$
1 1 *-H
H^ O
hJ ^
C ^ c^
2
~i
H 3
L^ I*
[rl c/5
^ [T1
1 E
(85 km2)
Ac
(46 km2)
C'
(2.1 km2)
WATERSHED LAND COVER CLUSTERS
(Mean Watershed Area For Group)
E-F
(8209 km2)
Columbia
Eel
Humboldt
A-D
(4482 km2)
Chetco
Hoh
Klamath
Queets
Quillayute
Quinault
Rogue
Copalis
Grays
Mad
Smith
Sooes
Umpqua
Willapa
Waatch
G-H
(1175km2)
Alsea
Coos
x^ \J\JJ
Coquille
Necanicum
Nehalem
Nestucca
Sand Lake
Siletz
Siuslaw
Tillamook
Yaquina
Salmon
I
(46 km2)
Netarts
to
-------
Table 2-8. Crosswalk of the cluster analyses by the relative proportion of NWI consolidated wetland classes (rows) and the relative
proportion of land cover classes (columns) at the 60% significance level. Letters refer to the cluster groups in Figures 2-15d and 2-
19d. Clusters with high similarity (>75%) were combined for the analysis. Mean watershed and mean estuary area within the clusters
are given in parentheses. The clusters are ordered by mean watershed or estuary size and proximity in the table does not necessarily
imply high similarity. The geomorphology is indicated by color with tide-dominated estuaries in blue, moderately river-dominated
estuaries in green, highly river-dominated estuaries in red, and bar-built estuaries in black.
Cluster on Relative Proportion /
60% Significance
NWI WETLAND CLUSTERS
(Mean Estuary Area for Group)
F'
(327 km2)
A'-C'
(135 km2)
G'
(29 km2)
N'
(22 km2)
K'
(16 km2)
H'-J'
(13 km2)
M'
(8.2 km2)
O'
(4.3 km2)
P'
(2.2 km2)
L'
(1.8km2)
E'
(l.lkm2)
D'
(0.7 km2)
WATERSHED LAND COVER CLUSTERS
(Mean Watershed Area For Group)
F-G
(8209 km2)
Columbia
Humboldt
Eel
A-E
(4482 km2)
Grays
Willapa
Klamath
Quinault
Quillayute
Rogue
Sooes
Umpqua
Copalis
Waatch
Mad
Smith
Hoh
Queets
Chetco
H-L
(1175km2)
Alsea
Coos
Yaquina
Necanicum
Nestucca
Siletz
Tillamook
Siuslaw
Coquille
Nehalem
Sand Lake
Salmon
M
(46 km2)
Netarts
-------
2.8 Watershed Alterations and Population Patterns
The last type of landscape analysis was evaluation of the extent of watershed alteration using a
suite of metrics indicative of anthropogenic "pressure" on the PNW estuaries. The detailed
analysis is based on the 31 estuaries used in the multivariate analysis (Table 2-4), while the
regional analysis is based on the 89 estuaries used in the original report (Lee et al., 2006) rather
than the 103 estuaries from the NWI update. The 14 watersheds associated with the "new"
estuaries (see Table 2-5) only constitute 0.2% of the total coastal watershed area so that their
exclusion has only a minor effect on the mean values for the region.
One set of indicators of anthropogenic pressure is the relative percentages of high development,
low development, and cultivated classes from the NOAA land cover data. NOAA has a class for
grasslands but does not separate out natural grasses from anthropogenic grasslands such as
pasture. Not to bias the estimates, we combined the 1992 NLCD classes for pasture/hay and
urban/recreational grasses as the fourth land cover metric of alteration. No attempt was made to
evaluate the effects of logging, which can increase non-point runoff of sediments and nutrients
(e.g., Likens et al., 1970). A potential indirect effect of logging may be an increase in red alder
after disturbance which may increase nitrogen fluxes from watersheds (Sections 3.7 and 4.2).
While this analysis does not capture the direct and indirect effects of logging, it does identify
those watersheds experiencing the greatest anthropogenic impacts from development and
agriculture. Another measure of alteration is population density from the 2000 census. Densities
were calculated on a watershed basis, which often cut across city or county boundaries. One
threshold relating population density to nutrient fluxes is the 386 people per km used in the EPA
"Classification Framework for Coastal Systems" to identify watersheds with a high risk for
nitrate inputs (U.S. EPA, 2004a).
The last metric of alteration is the percent impervious surfaces. High percentages of impervious
surfaces within a watershed can affect water quality both by increasing the volume and rate of
surface runoff, which in turn can increase erosion as well as non-point loadings of sediment,
contaminants, and nutrients (U.S. EPA, 1997). Adverse impacts are first observed in surface
waters in watersheds with about 10% impervious surfaces and with major impacts occurring at
25-30% (Arnold and Gibbons, 1996). In tidal creeks in South Carolina, physical/chemical
alterations and changes in fecal coliforms were first detected at 10-20% impervious surfaces with
impacts on the benthos, shrimp, and food webs detected at values of 20-30% (Holland et al.,
2004). NOAA's "Spatial Wetland Assessment for Management and Planning" (SWAMP) model
uses a threshold of impervious surfaces of <7.5% as the cutoff for "exceptional" environmental
condition (Sutler, 2001).
The values of these metrics for all 89 estuaries (Table 2-9) and for the 31 estuaries used in the
multivariate analysis (Table 2-10) indicate relatively low levels of alterations at the watershed
scale. The percent of high development land use only exceeded 2% in Crescent City Harbor
while the percent of low development only exceeded 5% in three watersheds and 1% in twelve
watersheds. Pasture and urban/recreational grasslands were more abundant than the other altered
land use classes in several of the watersheds. Anthropogenic grasslands exceeded 10% in the
Lake Earl and Humboldt watersheds and exceeded 5% in another five watersheds including the
74
-------
Table 2-9. The mean, median, and maximum values for each of the metrics of watershed alteration for the 89 coastal watersheds. The
"% High Development", "% Low Development", and "% Cultivated" are the percentage of the watersheds in these land cover classes
from the NOAA 2001 data. The "% Pasture & Urban Grasslands" is the combination of the pasture/hay and urban/recreational
grasses classes from the 1992 NLCD data. Population Density is from the 2000 census, while the "% Impervious Surfaces" was
calculated using the 30-m grids from NLCD (http://www.epa.gov/mrlc/nlcd-2001.html). The % impervious surface values and the
estuary with the highest percent impervious surfaces generated from Attila (U.S. EPA, 2004c) are given in parentheses for reference.
Mean
Median
Maximum
Watershed With
Maximum
% HIGH
DEVELOPMENT
0.13
0.01
4.98
Crescent City
% LOW
DEVELOPMENT
0.91
0.18
15.76
Loomis Lake
%
CULTIVATED
0.12
0.00
1.82
Twomile Creek
South
% PASTURE &
URBAN
GRASSLANDS
1.19
0.04
11.1
Lake Earl
%
IMPERVIOUS
SURFACES
(values from
Attila)
1.16
(2.54)
0.59
(2.02)
12.72
(16.11)
Loomis Lake
(Loomis Lake)
POPULATION
DENSITY
(# km'2)
18.09
4.24
288.79
Crescent City
-------
Table 2-10. The mean, median, and maximum values for each of the metrics of watershed alteration for the 31 coastal watersheds
used in the multivariate analysis. The "% High Development", "% Low Development", and "% Cultivated" are the percentage of the
watersheds in these land cover classes from the NOAA 2001 data. The "% Pasture & Urban Grasslands" is the combination of the
pasture/hay and urban/recreational grasses classes from the 1992 NLCD data. Population Density is from the 2000 census, while the
"% Impervious Surfaces" was calculated using the 30-m grids from NLCD (http://www.epa.gov/mrlc/nlcd-2001.html). The %
impervious surfaces value for the Columbia is based on the watershed up to the Bonneville Dam. The % impervious surface values
and the estuary with the highest percent impervious surfaces generated from Attila (U.S. EPA, 2004c) are given in parentheses for
reference.
Mean
Median
Maximum
Watershed With
Maximum
% HIGH
DEVELOPMENT
0.13
0.03
1.19
Humboldt
% LOW
DEVELOPMENT
0.56
0.28
3.71
Humboldt
%
CULTIVATED
0.17
0.00
1.29
Columbia
% PASTURE &
URBAN
GRASSLANDS
1.96
1.34
10.43
Humboldt
%
IMPERVIOUS
SURFACES
(values from
Attila)
0.83
(2.25)
0.57
(2.05)
3.61
(4.67)
Columbia
(Humboldt)
POPULATION
DENSITY
(# km'2)
14.73
5.96
119.19
Humboldt
-------
Columbia. Population densities were low to moderate across all the watersheds, and no
watershed exceeded the threshold of 386 people per km2. Only six estuaries had watersheds with
a population density exceeding 100 people per km2, with a maximum density of 289 people per
km in the Crescent City Harbor watershed.
The percent impervious surface integrates several different types of watershed alterations and the
recently available 30-m grid data from the NLCD (http://www.epa.gov/mrlc/nlcd-2001.html)
allows detailed analysis of this driver. In general, the percent impervious surface is low in the
coastal PNW watersheds (Tables 2-9 and 2-10). With about 12% impervious surfaces, only the
watersheds associated with Crescent City Harbor and Loomis Lake Creek exceeded the lower
10% threshold value from Arnold and Gibbons (1996). As a marine harbor/cove, Crescent City
Harbor is well flushed and this level of impervious surfaces is not likely to have a major effect
on its water quality. We do not have information on extent of flushing in Loomis Lake Creek
but Loomis Lake is classified as eutrophic based on total phosphorus concentrations (O'Neal et
al., 2001). Thus, it is possible that Loomis Lake Creek may have relatively high nutrient
concentrations during periods of restricted exchange with the ocean, though it is not clear what
estuarine resources would be exposed to these elevated nutrient concentrations (see Table 2-3).
No other watershed exceeded the NOAA SWAMP threshold of 7.5% for exceptional condition
and only six other watersheds had impervious surfaces >3%. Among the 31 estuaries, the
Columbia had the highest percent impervious surfaces at 3.6% but this is based on the 2001
NLCD data that extends only up to the Bonneville Dam and includes the city of Portland.
Presumably this percentage would decline if more of the Columbian Basin were included.
To evaluate the similarity in alterations, the 31 watersheds were clustered using the six metrics of
alteration. Because these variables are measured in different units, the values were normalized
and Euclidean distance was used as the metric of similarity. Classification by watershed
alterations resulted in four significant clusters (Figure 2-20). One group (Cluster A) that split off
at a high degree of dissimilarity consisted of the lower Columbia River and Humboldt Estuary.
These two systems have the highest percentages of high and low development land classes,
anthropogenic grassland land cover classes, the highest percent impervious surfaces, and the
highest population densities (Table 2-10). The Necanicum (Cluster B) broke out at a moderate
degree of dissimilarity from the other estuaries. Though not as altered as the Columbia and
Humboldt, the Necanicum watershed has the third highest values for the percentages of high
development and low development land cover classes, population density, and percent
impervious surfaces. Another group (Cluster C) that split out at a moderate degree of
dissimilarity consisted of four estuaries, the Klamath, Grays, Eel, and Rogue. These four
estuaries appeared to break out based on their moderately high values of both cultivated and
pasture/urban grassland land cover classes. The last group (Cluster D) consists of the remaining
24 estuaries, which as a group had no discernable pattern of watershed alteration. Thus, the
watersheds of the major PNW estuaries separate into four patterns of alteration, with the
Columbia and the Humboldt the most altered due to higher proportions of development and
agriculture.
77
-------
03
w
b
c
CO
03
T3
LU
9--
8--
7--
6--
5--
4--
3--
0-L
B
D
n
mr^
il
0 X
i'ii
03
J 3
o~o
CO
CO
CO
Estuaries
Figure 2-20. Cluster analysis of the 31 PNW estuaries based on six metrics of watershed
alteration. Similarity among clusters is measured as Euclidean distance. Estuaries joined with
red lines are not significantly different (p>0.05) based on SIMPROF analysis. The horizontal
lines indicate 25% dissimilarity of the maximum observed dissimilarity for reference.
78
-------
2.8.1 Flow-Path Analysis of Impervious Surfaces
While the percent impervious surfaces for the entire watersheds are low, the spatial distribution
of watershed alterations is likely to have an important effect on the extent of nutrient, pollutant,
and sediment runoff. The 30-m grid cells from the NLCD allow a flow path analysis of the
percent impervious surfaces as a function of the distance from the estuary or nearest river or
stream flowing into the estuary. Because of its complexity, we have only conducted a flow path
analysis for the Yaquina watershed which is presented here as a potential future direction.
Figure 2-21 shows the percent impervious surfaces within six concentric zones around the
estuary, rivers, and streams, increasing in distance from a 0-108 meter zone to the nearest water
to a 0-31,130 meter zone that represents the entire watershed. The key observation is that the
percent impervious surface is higher in the portion of the watershed closer to the water than in
land more distant from the water's edge. The maximum percent impervious surface of 3.78%
occurs within a band 0-500 meters around the estuary, river, and streams, a value more than 4-
times higher than the watershed average of 0.89%. Similar analyses should be conducted in
additional watersheds, but it is likely that this general pattern of higher impervious surfaces near
the water occurs in many, if not most, PNW watersheds as a result of the concentration of tourist
facilities, commercial and recreational fishing activities, and housing along bay and river fronts.
4_,
"* s
- 1
1
CO
•S 7* -
« 2'5
o
u
1
3 ">
* L
%
ao
£ 1 ^
S i-5
p.
HH
S® t
0s 1
OS
0-108 0-500 0-1000 0-5000 0-10000 0-31130
Buffer from Nearest Water (m)
Figure 2-21. Percent impervious surfaces in the Yaquina watershed for six buffer zones of
different distances from the shoreline of the Yaquina Bay or Yaquina River. The 0-31,130 m
buffer represents the entire Yaquina watershed.
79
-------
2.9 Synthesis of Classification Approaches and Research Needs
Five classification schemas are presented in this chapter - classifications based on
geomorphology and ocean exchange (Section 2.4), wetland clustering (Section 2.5), watershed
clustering (Section 2.6), crosswalks between the wetland and watershed clustering (Section 2.7),
and groupings on watershed alterations (Section 2.8). Which is the "best" approach depends
upon the goals of the user. Even with the specific goal of identifying estuaries with similar
vulnerabilities to nutrient enrichment, it is not clear which schema is best, in part because no
specific criteria have been developed as to what constitutes "similar enough" for management
purposes. It is possible, however, to provide some general guidelines on how the various
schemas could be used for management purposes as well as future research directions.
Of the various schemas, geomorphology provides the most general approach to grouping
estuaries with similar nutrient dynamics. Assuming the ideal case of "all other factors being
equal", potential differences in the extent of flushing based on ocean exchange and freshwater
inflow suggest the following relative vulnerabilities (from most to least vulnerable) to equivalent
increases from terrestrial runoff:
lagoon > blind estuary > tide-dominated river mouth > river-dominated river mouth > coves
This suggested ranking is for increased nutrient concentrations within the estuarine portion of the
system and not for total loading, which may show a pattern more closely tied to the normalized
freshwater inflow and presence of red alder in the watershed. Nor does this ranking account for
differences in the types of resources potentially at risk across these different classes of estuaries.
The tidally restricted coastal creeks are more complex than suggested by their small size and do
not easily fit into this ranking because of the potential occurrence of localized pockets of low DO
water resulting from periodic advection and trapping of ocean water (C. Brown, unpublished
data). Nor do bar-built estuaries readily fit into this scheme. The three PNW bar-built estuaries
do not share many similarities (Tables 2-7 and 2-8) and may function more like similar size tide-
dominated estuaries than as a separate class of estuaries per se.
Individual clustering by wetland or land cover classes provides a more detailed breakout of
estuaries than that provided by the geomorphologic classes. While this approach is useful in
identifying similarities in the resources at risk (wetland clustering) or differences in potential
loading (watershed clustering), we suggest that the crosswalks are more likely to identify
functionally similar estuaries than clustering by either attribute alone. As discussed in
Section 2.7, the crosswalks identified fine resolution separations using the 60% significance level
(Tables 2-6 and 2-8) and coarser level separations using the 5% significance level (Tables 2-5
and 2-7). One of the factors in resolving which of these schemas are sufficient for management
is how similar estuaries are within the groups compared to among-group differences. If, for
example, variation in a key attribute within a group exceeds the differences among groups, then
the classification may not be sufficiently robust for certain management purposes. Therefore, we
suggest one avenue of research should be to compare within- and among-group variations in
nutrient concentrations and dynamics across sets of highly similar estuaries, using the water
quality survey framework described in Chapter 4. Because nutrient dynamics are influenced by
drivers related to both the areas of estuaries and watersheds and to the relative proportions of
80
-------
wetland or watershed classes, we propose that estuaries that co-clustered based on both the areas
(Tables 2-5 and 2-6) and on relative proportions (Tables 2-7 and 2-8) should show the highest
similarities. This is essentially a "crosswalk of the crosswalks" to identify the most similar
estuaries.
Based on the 60% significance levels, there are six pairs of estuaries that co-clustered on area
and relative proportions: 1) Klamath and Rogue; 2) Grays and Willapa; 3) Coquille and
Nehalem; 4) Alsea and Yaquina; 5) Hoh and Queets; and 6) Nestucca and Siletz. These pairs
span a range of estuary types, including large to small estuaries as well as tide-dominated,
moderately river-dominated, and highly river-dominated estuaries (Tables 2-2 and 2-4).
Variation in nutrient dynamics within these pairs represents the minimum variation that can be
expected from groupings based on this approach. Results from such comparisons should be
assessed by environmental managers to determine whether the within-group variation relative to
the among-group variation is acceptable in a regulatory context. If the levels of variation are not
acceptable, then: 1) a different classification approach is required; 2) management expectations
may need to be modified regarding the actual level of similarity among PNW estuaries or; 3) the
desire to manage groups of "similar" estuaries in a common fashion may be unrealistic.
If the relative variation within and among these six pairs of estuaries is acceptable, the next step
would be to conduct additional comparisons among sets of grouped estuaries. In particular, the
three large groups in the 5% significance crosswalk based on relative proportions (A-D -A',A-
D-B\ and G-H - B' in Table 2-7) should be evaluated. If the relative variation falls within
acceptable bounds, this classification could be used as a framework for developing management
strategies for the PNW, with the caveat that additional studies would be required for the estuaries
falling outside of these groups. However, failure to find ecologically relevant differences among
groups at this level of significance while finding them in the six paired estuaries suggests that a
finer resolution classification will be needed.
81
-------
CHAPTER 3:
ENVIRONMENTAL CHARACTERISTICS
OF THE SEVEN TARGET ESTUARIES
Cheryl A. Brown and Henry Lee II
3.0 Introduction
The purpose of this chapter is to provide an overview of the environmental conditions within the
Pacific Northwest (PNW) and of the seven target estuaries chosen for field surveys and aerial
photography. This chapter is not meant to be a comprehensive review and the reader is referred
to Emmett et al. (2000) for an overview of environmental conditions for Pacific Coast estuaries,
Hickey and Banas (2003) for a review of the effects of oceanographic conditions on PNW
estuaries, the Oregon Coastal Atlas (http://www.coastalatlas.net) for a description of Oregon
estuaries, and the Coastal Landscape Analysis and Modelling Study (CLAMS;
http://www.fsl.orst.edu/clams/) for detailed information on coastal watersheds in Oregon.
3.1 Pacific Northwest Climate
Though the PNW has a reputation for being "wet and rainy", the PNW coast essentially has a
wet and dry season (see Emmett et al., 2000). From approximately October through March, the
PNW experiences frequent rainfall with a peak in December through January. From
approximately April through September, rainfall substantially declines often with only marginal
rainfall in July through September. In addition to this seasonality, there is a geographic pattern
with greater rainfall in the more northern portions of the PNW. Average annual precipitation in
northern coastal Washington can reach as much as 350 cm compared to 140 cm in the Humboldt
watershed in northern California. In terms of temperature, the PNW has a Mediterranean
climate, with mild summers and winters. The difference between winter and summer air
temperatures is small, only about 5° C near the Humboldt Estuary (Emmett et al., 2000).
Freezes and snowfall at the lower elevations are relatively rare, and with the exception of the
Columbia and Klamath watersheds there is relatively little snow pack in most of the coastal
watersheds.
3.2 Coastal Upwelling in the Pacific Northwest
Estuaries in the PNW are adjacent to the California Current System, which exhibits strong
interannual, seasonal and event-scale variability (Hickey and Banas, 2003). In this region,
seasonal wind-driven upwelling advects relatively cool, nutrient rich water to the surface. The
upwelling season typically commences in April and continues through September (Kosro et al.,
2006). During this time period, upwelling-favorable winds from the north dominate. The
upwelling conditions are interrupted by brief periods of downwelling-favorable conditions,
which usually persist for several days. Previous studies have demonstrated that the oceanic
inputs of nutrients and phytoplankton are important for estuaries adjacent to coastal upwelling
regions, such as the west coast of the U.S. (e.g., de Angelis and Gordon, 1985; Roegner and
Shanks, 2001; Roegner et al., 2002; Newton and Horner, 2003; Colbert and McManus, 2003;
Brown and Ozretich, 2009). Hickey and Banas (2003) examined variations in temperature,
salinity, and alongshore wind stress for three estuaries along the Oregon and Washington coast,
spanning a distance of 400 km. They demonstrated that there was coherence in the fluctuations
in temperature and salinity among these estuaries during the summer resulting from the large-
82
-------
scale patterns in along-shelf wind forcing at large scales (100s of kms). As a result, we believe
that coastal ocean conditions are relatively uniform over the geographic extent of the seven target
estuaries, which span 250 km (Figures 2-3 and 2-4).
3.3 Criteria for Choosing Target Estuaries and Previous Estuarine Classifications
The seven target estuaries were chosen to cover a range of estuarine and watershed conditions in
the PNW. While the goal was to choose a suite of PNW estuaries representative of the range of
systems, we focused this initial effort on Oregon estuaries because of logistical constraints.
When this work was initiated, we had not completed our regional analysis of PNW (Chapter 2),
and so relied on previous classifications and our personal experiences with these estuaries for the
selection. The primary classification systems of the PNW estuaries that we utilized in estuary
selection are summarized below and in Table 3-1.
3.3.1 Management Classification of Oregon Estuaries
Oregon estuaries were classified under Oregon's Statewide Planning Goal 16 (Estuarine
Resources) as natural, conservation, shallow draft development, or deep draft development
(Cortright et al., 1987; http://www.inforain.org/oregonestuary/oregonestuary_page5.html) with
the following definitions:
Natural: Estuaries lacking maintained jetties or channels, and which are usually little
developed for residential, commercial or industrial uses. They may have altered
shorelines, provided that these altered shorelines are not adjacent to an urban area.
Shorelands around natural estuaries are generally used for agriculture, forestry, recreation
and other rural uses.
Conservation: Estuaries lacking maintained jetties or channels, but which are within or
adjacent to urban areas which have altered shorelines adjacent to the estuary.
Shallow Draft Development: Estuaries with maintained jetties and a main channel (not
entrance channel) maintained by dredging at 22 feet or less.
Deep Draft Development: Estuaries with maintained jetties and a main channel
maintained by dredging to deeper than 22 feet.
The target estuaries included one natural estuary (Salmon River Estuary), two conservation
estuaries (Alsea and Nestucca estuaries), two shallow draft development estuaries (Tillamook
and Umpqua River estuaries), and two deep draft estuaries (Coos and Yaquina estuaries). Thus,
the target estuaries represent a good cross-section of the Oregon management types.
3.3.2 Geomorphological/Hydrological Classifications
Based on the geomophological/ocean exchange classification we developed after the field
surveys (Section 2.4), the seven target estuaries included three tide-dominated drowned river
mouth estuaries (Coos, Tillamook, and Yaquina), two moderately river-dominated river mouth
estuaries (Alsea, Salmon) and two highly river-dominated river mouth estuaries (Nestucca and
Umpqua estuaries) (Table 3-1). Within this suite of estuaries, the normalized
83
-------
Table 3-1. Current and historical classifications of the seven target estuaries. The current classification we developed is discussed in
Section 2.4.
Estuary
Alsea
Coos
Nestucca
Salmon River
Tillamook
Umpqua River
Yaquina
Current
Classification
Moderately river-
dominated drowned
river mouth
Tide-dominated
drowned river mouth
Highly river-
dominated drowned
river mouth
Moderately river-
dominated drowned
river mouth
Tide-dominated
drowned river mouth
Highly river-
dominated drowned
river mouth
Tide-dominated
drowned river mouth
Oregon
Management
Classification
Conservation
Deep draft
development
Conservation
Natural
Shallow draft
development
Shallow draft
development
Deep draft
development
Bottom et al.
(1979)
Drowned river
valley, partially
mixed
Drowned river
valley, well mixed
Drowned river
valley, partially
mixed
Drowned river
valley, partially
mixed
Drowned river
valley, partially
mixed
Drowned river
valley, partially
mixed
Drowned river
valley, partially
mixed
Parrish et al. (2001)
Well-flushed drowned rivers,
predominantly freshwater
input?
or
Well-flushed drowned rivers,
predominantly oceanic input?
Well-flushed drowned rivers,
predominantly oceanic input
Well-flushed drowned rivers,
predominantly oceanic input
Well-flushed drowned rivers,
predominantly freshwater input
Well-flushed drowned rivers,
predominantly oceanic input
Well-flushed drowned rivers,
predominantly freshwater input
Well-flushed drowned rivers,
predominantly oceanic input
Rumrill (1998)
NA
Tide-dominated
drowned river
mouth
NA
NA
Tide-dominated
drowned river
mouth
River-dominated
drowned river
mouth
Tide-dominated
drowned river
mouth
Burt&
Mcalister
(1959)
Partially mixed
Well mixed
Partially mixed
NA
NA
Well mixed
Two-Layer
(high flow)
Two-Layer
(high flow);
Partially to well
mixed (low flow)
Well mixed to
partially mixed
oo
-------
freshwater inflow metric varies about 9-fold from the Coos Estuary to the Umpqua River
Estuary. However, none of the estuaries with highest normalized freshwater inflow values were
included (see Table 2-4). Thus, our selection of estuaries represents a reasonably wide range of
drowned river mouth estuaries, but does not capture the full range expected if, for example, the
Rogue River or Klamath River had been included. Additionally, no blind estuaries, coastal
lagoons, marine harbor/coves, or tidally restricted coastal creeks, as defined in Chapter 2, were
included.
The various geomorophological classifications of Oregon estuaries that we initially used to
choose our target estuaries are summarized in Table 3-1 along with our classification schema.
There is agreement among all the classifications that the seven target estuaries are drowned river
mouth estuaries. In terms of riverine versus oceanic influences, there is also agreement on the
three tide-dominated estuaries and two of the river-dominated systems, the Umpqua and Salmon
estuaries. However, Parrish et al. (2001) classified the Nestucca Estuary as having
"predominantly oceanic input" versus our classification as highly riverine dominated. It is
possible that the Nestucca's division into two separate arms may affect how the normalized
freshwater inflow values should be interpreted. Parrish et al. (2001) were uncertain whether to
classify the Alsea Estuary as predominantly oceanic or freshwater input while we classified it as
moderately river-dominated.
3.4 Estuary and Watershed Sizes
As discussed in Chapter 2, estuary size was defined as the sum of the NWI marine, estuarine, and
tidal riverine polygons (http://www.fws.gov/nwi/; U.S. Fish Wild. Sen, 2002). Using this
definition, estuary size among the target estuaries varied almost 20-fold, from about 3 km2 for
the Salmon River Estuary to 55 km2 for the Coos Estuary (Tables 2-2 and 3-2). The Coos,
Tillamook, Umpqua, and Yaquina are the four largest estuaries in Oregon other than the
Columbia River. The only other large PNW estuaries are Grays Harbor and the Willapa Estuary
in Washington and the Humboldt Estuary in northern California. The smaller target estuaries
(Alsea, Salmon River, and Nestucca) were considered to capture much of the environmental
range in the moderate-sized estuaries in the PNW.
As discussed in Chapter 2, PNW estuaries have extensive intertidal zones. The average percent
intertidal area across all 89 PNW estuaries was 52% (Figure 2-8). The seven target estuaries
bracketed that average, ranging from 32% in the Umpqua River Estuary to 87% in the Salmon
River Estuary.
The sizes of the associated watersheds varied 63-fold (Table 3-2), from about 190 km2 in the
Salmon River Estuary to over 12,000 km2 in the Umpqua River Estuary. The ratio of estuarine
size to watershed size also varied by more than an order of magnitude. The Coos, Tillamook,
and Yaquina all had relatively large estuaries compared to the watershed (>2.5%) while the
Alsea and Salmon River had moderate-sized estuaries compared to the watershed (ca. 1%). Even
though they differed 7-fold in estuary size, the Nestucca and Umpqua River had the smallest
estuaries in relation to their watersheds, with estuary-to-watershed percentages of 0.61% and
0.28%, respectively.
85
-------
3.5 Watershed Characteristics and Land Cover
The percent land cover values for selected classes of the watershed associated with each of the
target estuaries are given in Table 3-3 based on the 2001 NOAA land cover data (see Chapter 2).
Averaged across the entire watershed, none of these watersheds are highly developed. The
maximum "high intensity development" land cover class was only 0.23% of the Coos watershed.
The combined "high intensity development" and "low intensity development" classes exceeded
0.5% only in the Coos and Yaquina watersheds. The extent of cultivated land was also very low,
with a maximum of 0.22% in the Alsea watershed. While there is little cultivated land, pasture
for cattle grazing occurred within some of the coastal watersheds, in particular the Tillamook
watershed. The NOAA survey does not have a land cover class specifically for pasture, but does
have a class for grasslands which includes both natural grassland and pastures. To better capture
the extent of pasture land, we used the 1992 National Land Cover Data (NLCD) land cover data,
which has a pasture land cover class. Based on the NLCD data, the Tillamook, Umpqua River,
and Nestucca watersheds each had about 3% to 4.5% pasture.
An integrative measure of land cover alteration is the percent impervious surfaces (see
Chapter 2). The target estuaries showed a small range in this metric, with all the values about
2% (Table 3-3). These values are well below the 7.5% cutoff for "exceptional" environmental
conditions used in the NOAA's "Spatial Wetland Assessment for Management and Planning"
(SWAMP) model (Sutler, 2001). These low percentages of altered land use and percent
impervious surfaces among the target estuaries are representative of PNW watersheds in general
(Tables 2-6 and 2-7).
3.5.1 Watershed Slopes
The median degree slope and median percent slope of the target estuaries are given in Table 3-3.
For details on methods and datasets used to calculate watershed slope see Section B.12.4. We
prefer degrees slope, which is bounded between 0 and 90 degrees, in comparison to percent
slope, which can vary from 0 to infinity. Though they are the more commonly reported values,
percent slopes can obtain very high values in areas with cliffs, which can skew the distribution
and the mean compared to using degrees slope. The EPA Watershed Academy states that "high
relief watersheds may have increments of 5-20 percent."
(http://www.epa.gov/owow/watershed/wacademy/wam/erosion.html). Using this criterion, all
the target watersheds have a high relief, with median percent slopes ranging from 24.9% in the
Salmon River watershed to 40.8% in the Tillamook watershed. The range in degrees slope was
from about 14 degrees in the Salmon River watershed to over 22 degrees in the Tillamook
watershed. Regardless of how it is measured, high relief is characteristic of PNW coastal
watersheds
3.5.2 Population Characteristics
The Umpqua River and Coos watersheds had the largest human populations, about 100,000 and
39,000, respectively (Table 3-4). Although the Umpqua watershed has the largest population,
the majority of the population (about 74,000) is located in the Roseburg area, which is 125 km
from the coast. In contrast, the bulk of the population in the Coos watershed is located adjacent
to the estuary. Normalized to watershed area, the population density in the Coos watershed was
about 25 people per km , about 66% to 6-fold greater than the watersheds in the other target
estuaries. To put the Coos watershed density in perspective, the Coos population density is one
86
-------
Table 3-2. Estuary and watershed size of the seven target estuaries. Estuary area includes the marine, estuarine, and tidal riverine
NWI classes. Estuarine intertidal and subtidal areas and the percent intertidal only include the estuarine NWI classes, so that the sum
of the intertidal and subtidal may not equal the total estuary area. The intertidal area includes "irregularly exposed", "regularly
flooded" and "irregularly flooded" habitats (approximately MLLW to MHHW).
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
LATITUDE
(degN)
44.4227
43.4294
45.1827
45.0469
45.5130
43.6694
44.6205
ESTUARY
AREA
(km2)
12.49
54.90
5.00
3.11
37.48
33.78
19.96
ESTUARINE
INTERTIDAL
AREA
(km2)
7.83
29.89
2.88
1.72
25.61
8.85
9.05
ESTUARINE
SUBTIDAL
AREA
(km2)
4.66
24.32
1.79
0.25
11.23
18.88
9.77
% INTERTIDAL
OF ESTUARINE
AREA
(km2)
62.7
55.1
61.7
87.3
69.6
31.9
48.1
WATERSHED
SIZE
(km2)
1222
1575
826
193
1455
12146
650
ESTUARY SIZE
AS % OF
WATERSHED
1.02
3.48
0.61
1.61
2.57
0.28
3.07
oo
-------
Table 3-3. Watershed attributes of the seven target estuaries. Slope is given in degrees and percent slope. Land cover data for
selected classes from the NOAA 2001 dataset except for the percentage for "pasture" which is from the 1994 NLCD data. The "%
Impervious Surfaces" was calculated using the 30-m NLCD data (http://www.epa.gov/mrlc/nlcd-2001.html). The % impervious
surface values generated from the NOAA land cover classes using Attila (U.S. EPA, 2004c) are given in parentheses for reference.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
LAND COVER (% WATERSHED AREA)
High
Intensity
Development
0.02
0.23
0.02
0.01
0.12
0.08
0.10
Low
Intensity
Development
0.13
0.86
0.14
0.14
0.36
0.40
0.40
Cultivated
0.22
0.00
0.00
0.00
0.00
0.12
0.00
Grasslands /
Pasture
3.31 /
1.30
7.09 /
1.39
5.49/
3.01
4.65 /
0.84
6.69 /
4.48
11. 19/
4.53
6.24 /
0.36
Deciduous
3.07
1.45
4.63
3.43
5.85
0.49
5.19
Evergreen
57.07
48.64
44.60
45.60
46.02
62.79
36.18
Mixed
Forest
25.86
15.89
29.22
24.97
28.37
6.94
27.24
Shrub
6.82
15.43
12.15
13.27
6.19
14.70
16.73
%
Impervious
Surfaces
0.51
(2.01)
1.16
(2.65)
0.55
(1.98)
0.85
(2.15)
0.87
(2.30)
0.46
(2.20)
0.89
(2.38)
Median
slope
(degrees/
percent)
18. 11
32.7
18.5 /
33.5
14.9 /
26.6
14.0 /
24.9
22.2 /
40.8
18.0/
32.5
16.5 /
29.7
oo
oo
-------
Table 3-4. Population size and density in the seven target estuaries. Population estimates based on 2000 census for the entire
watershed. Population densities (# per km2) are normalized both to the area of the watershed and to the area of the estuary.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
POPULATION
IN
WATERSHED
4825
38,950
3404
2881
7600
99,401
7970
POPULATION DENSITY (# per km^)
NORMALIZED TO
WATERSHED SIZE
3.9
24.7
4.1
14.9
9.7
8.2
12.3
NORMALIZED TO
ESTUARY SIZE
386
709
681
926
203
2943
399
% POPULATION
CHANGE FROM
1990 TO 2000
5.78
2.67
12.28
20.10
11.84
6.09
-4.85
oo
VO
-------
to two orders of magnitude lower than in many Central and Southern California coastal
watersheds, including Elkhorn Slough (339 km"2), South San Francisco Bay (745 km"2), San
Diego (780 km"2), and Anaheim (2841 km"2). Population density was also normalized to the
estuarine area, with the densest normalized population occurring in the Umpqua River watershed
with over 2900 persons per km2 of estuary. This high value for the Umpqua River Estuary, in
part, reflects the small estuary size compared to the large size of the Umpqua River watershed
(Table 3-2). However, this value is still at least an order of magnitude smaller than those the
moderate and large sized estuaries in Central and Southern California.
The percent population change from 1990 to 2000 varied among the watersheds. The largest
percent growth was in the Salmon River watershed which increased by 20% within a decade. At
the opposite extreme, the Yaquina watershed actually showed a decrease in population of about
5%. For the Yaquina watershed it is important to recognize that other than the bay front, most of
the City of Newport lies outside of the watershed. The population decrease in the Yaquina
watershed over this period probably reflects a population decrease in City of Toledo, located in
the upper portion of the Yaquina watershed.
3.6 Estuarine Hydrology
The seven target estuaries are relatively shallow with mean depths of 1.5-3 m (Table 3-5). The
volumes of the target estuaries range from 1 x 106 m3 (Salmon) to 2 x 108 m3 (Coos), and are
considered small estuaries compared to many in the U.S. (Hickey and Banas, 2003). These
estuaries are classified as mesotidal with a mean tidal range of about 2 m (Table 3-5) and have
mixed semidiurnal tides. There is a close coupling between the estuaries and the coastal ocean
as a result of the large tidal prism relative to estuary volume (Table 3-5 and Hickey and Banas,
2003). Additionally, there is considerable variability in the mouth width of the estuaries, with
the Salmon River Estuary having the smallest mouth (~ 40 m) and Coos and Umpqua estuaries
having the largest (400 - 600 m, Table 3-5).
Reflecting the seasonal pattern in rainfall, freshwater flow is about 5-fold to almost 10-fold
higher in the wet season compared to the dry season (Table 3-6). The wet season is defined as
November through April while the dry season is defined as May through October. This
seasonality in riverine inflow is several times greater for estuaries in the PNW compared to those
along the Atlantic coast of the U.S. (Hickey and Banas, 2003). The Simmons Ratio, which is the
ratio of riverflow per tidal cycle to tidal prism, is often used to classify estuaries by stratification
(Simmons, 1955). When the Simmons Ratio is greater than 1, the estuary is highly stratified.
When the ratio is 0.2-0.5, the estuary is considered partially mixed and when it is less than 0.1, it
is considered well mixed. Using Simmons ratio (calculated using annual average inflow), Alsea
and Salmon River estuaries are considered partially mixed, and the remainder of the target
estuaries are considered well mixed.
Residence (or flushing) time is a measure of the retention of water within a defined boundary
(Monsen et al., 2002). Residence time is often considered a mediating factor in assessing
estuaries susceptibility to nutrient loading (e.g., Quinn et al., 1991; National Research Council,
2000). Estuaries with long residence times are considered more susceptible to nutrient
enrichment than estuaries with short residence times. Estuarine residence times are influenced
by freshwater inflow, tides, wind, mixing, stratification, and system topography. There are
90
-------
numerous terms and methods for calculating these time scales of transport, including flushing
time, residence time, local residence time, turnover time, freshwater replacement time, and
transit time (Monsen et al., 2002; Abdelrhman, 2005). Two commonly used methods for
calculating residence time are the fraction of freshwater method and the modified tidal prism
approach (Dyer and Taylor, 1973). The fraction of freshwater method calculates the amount of
time for the freshwater inflow to replace the freshwater in the system. It can be interpreted as the
average transit time for freshwater in the system and is an appropriate time scale for materials
that are input to the estuary from the river (Sheldon and Alber, 2002). The modified tidal prism
is more appropriate for the dry season estimates when tidal forcing dominates (riverflow small
compared to tidal flow) and for estuaries such as those in the PNW where the volume is small
compared to the tidal prism. The fraction of freshwater method is more appropriate during the
wet season when freshwater inflows dominate. Presented in Table 3-7 are wet and dry season
residence times for the seven estuaries calculated using the modified tidal prism and fraction of
freshwater inflow approaches. The residence times of the seven target estuaries are short
(typically less than 1 month). During the wet season, the residence time calculated using the
fraction of freshwater method ranges from 2 to 10 days; while during the dry season, the
residence time calculated using the modified tidal prism method varies from 1 to 48 days.
3.7 Nutrient Loading
Previously published estimates are available for nitrogen and phosphorous loading for six of the
target estuaries. The nitrogen loading varies 26-fold for the seven target estuaries with the
Salmon River Estuary having the lowest nitrogen loading and the Umpqua River Estuary having
the highest (Table 3-8). This variability in nitrogen loading is related to differences in stream
flow among the estuaries (Table 3-6). The loadings in Table 3-8 represent point sources as well
as nonpoint and upstream sources. The Umpqua River estuary had the highest phosphorous
loading, while the Yaquina had the lowest (Table 3-8). Point source inputs represented the
largest phosphorous source in Coos and Yaquina estuaries, while in Tillamook, and Umpqua
estuaries non-point inputs from forested land was the dominant source of phosphorous (Quinn et
al., 1991). Nonpoint source inputs associated with forested land was the dominant nitrogen
source in Coos, Tillamook, Umpqua, and Yaquina estuaries (Quinn et al., 1991). Upstream
sources were the major source of nitrogen and phosphorous in the Alsea Estuary (Quinn et al.,
1991). Nutrient loading estimates are often normalized by estuary area and volume to examine
sensitivity to nutrient loading. When the nitrogen loading estimates are normalized to area,
Yaquina and Coos have the lowest loading, while Umpqua and Alsea have the highest.
The loadings presented in Table 3-8 do not include the input of nutrients from the coastal ocean.
Brown and Ozretich (2009) compared the major sources of dissolved inorganic nitrogen (DIN) to
Yaquina Estuary during the wet and dry seasons (Table 3-9). There are strong seasonal
differences in the nitrogen sources to the estuary. During the wet season, riverine sources
dominate, while during the dry season oceanic nitrogen inputs associated with coastal upwelling
dominate. In the dry season, benthic flux of DIN from the sediments into the water column is the
second largest source of nutrients. Atmospheric deposition of nitrogen is a minor nitrogen
source with direct deposition on the estuary only representing about 0.05% of the inorganic
nitrogen input to the estuary. In addition, atmospheric deposition on the watershed is small (8%)
compared to the watershed input associated with nitrogen fixing red alder in the watershed.
Annual nitrogen input from wastewater treatment facility effluent is estimated to be 0.4% of the
91
-------
total input to the estuary. The ocean is also a large source of phosphorous to PNW estuaries;
however, at this time the oceanic phosphorous loading has not been quantified.
Oregon Coast Range streams have high nitrate concentrations relative to other forested PNW
watersheds (Compton et al., 2003). Wigington et al. (1998) hypothesized that forest vegetation,
in particular the presence of red alder (Alnus rubra\ is the primary control of spatial variability
in stream nitrate concentrations in the Oregon Coast Range (including the Salmon River, Siletz,
and Alsea watersheds). Red alder is a native species in the PNW that colonizes areas disturbed
by fires, logging and landslides. Red alder have symbiotic N2-fixing bacteria that can fix 50-200
kg N ha"1 y"1 in pure stands (Binkley et al, 1994). Compton et al. (2003) found a significant
relationship between alder cover in the watershed and nitrate concentration in streams in the
Salmon River watershed. Naymik et al. (2005) found a similar relationship between stream total
nitrogen and broadleaf cover (which is primarily red alder in the Coast Range) in the Tillamook
watershed. Brown and Ozretich (2009) estimated that > 80% of the riverine nitrogen loading to
Yaquina Estuary is related to red alder cover.
92
-------
Table 3-5. Hydrographic characteristics of the seven target estuaries. Sources were (1) Shirzad et al. (1988), (2) Johnson and Gonor
(1982), (3) Percy et al. (1974), (4) Coastal Inlets Research Program (http://cirp.wes.army.mil/databases/inletsdb/inletsdbinfo.html),,
(5) http://tidesandcurrents.noaa.gov/tides04/tab2wclb.html, and (6) http://ian.umces.edu/neea/siteinformation.php. Mouth widths are
from the Pacific Coast Ecosystem Information System (PCEIS; Lee and Reusser, 2006).
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
TIDAL RANGE
(m)
1.8(1,5)
1.7(5)
1.8(5)
1.6(2,3)
1.6-1.9(5)
1.6(5)
1.8-1.9(5)
TIDAL
PRISM
(m3)
1.15x 10' (1)
5.27 x 10' (3)
9.5 x 105 (2)
4.81 x 10' (1)
6.23 x 10 '(4)
2.38 x 10' (4)
ESTUARJNE
VOLUME
(m3)
1.9 x 10' (6)
2.1 x 10s (6)
1.4x 10b(2)
7.0 x 10' (6)
7.5 x 10' (6)
3. Ox 10' (6)
RATIO OF
ESTUARJNE
VOLUME
TO TIDAL
PRISM
1.6
3.9
1.5
1.5
1.2
1.3
MOUTH
WIDTH
(m)
140
620
110
40
360
425
290
MEAN
DEPTH
(m)
2.0(1)
1.5(3)
1.8(1)
3.0(1)
DREDGED
No
Yes
No
No
Yes
(primarily
prior to
1979)
Yes
Yes
-------
Table 3-6. Wet and dry season freshwater inflow for the seven target estuaries. Sources: (1) Shirzad et al. (1988); (2) Quinn et al.
(1991), and (3) Percy et al. (1974).
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
WET
SEASON
(m3 s-1)
112
51
18
189
342
48
DRY
SEASON
(m3 s-1)
15
8
4
31
71
5
ANNUAL
AVERAGE
(m3 s-1)
65
82
30
43
11
110
209
210
19
27
SIMMONS RATIO =
RIVERFLOW PER TIDAL
CYCLE / TIDAL PRISM
0.3
0.1
0.5
0.1
0.1
0.1
ESTUARY TYPE
BASED ON
SIMMONS RATIO
Partially Mixed
Well Mixed
Partially Mixed
Well Mixed
Well Mixed
Well Mixed
SOURCE/NOTES
(1)
(2)
Nestucca River near Beaver
Estimate of flow at mouth (3)
Discharge near Otis
(1)
Umpqua discharge near Elkton
Umpqua River from (3)
Smith River from (3)
(1)
VO
Table 3-7.
estuaries.
Residence time calculated using the modified tidal prism and fraction of freshwater approaches for the seven target
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
MODIFIED TIDAL PRISM
WET
SEASON
(days)
1
7- 13
13-16
NA
NA
2-3
4-5
6
DRY
SEASON
(days)
4-9
11 - 16
40-48
NA
1-2
3-4
5-10
6-9
FRACTION OF
FRESHWATER INFLOW
WET
SEASON
(days)
NA
2- 10
3-11
NA
NA
3
1-7
O
9
DRY SEASON
(days)
14- 17
34
19-31
NA
1-2
9
5-32
5-8
20 - 106
SOURCE
Choi (1975)
Choi (1975)
Arneson (1976)
Askrenetal. (1976)
Choi (1975)
Colbert and McManus (2003)
Choi (1975)
Choi (1975)
-------
Table 3-8. Estimates of nutrient (nitrogen and phosphorous) loading to six of the seven target estuaries. Load estimates are not
available for Nestucca Estuary, and phosphorous loads are not available for Salmon River estuary. Total load estimates for Alsea,
Coos, Tillamook, Umpqua and Yaquina are from Quinn et al. (1991), while nitrogen load estimate for Salmon River Estuary estimated
from flow and DIN from Compton et al. (2003).
ESTUARY
Alsea
Coos
Salmon
Tillamook
Umpqua
Yaquina
NITROGEN LOADING
TOTAL
(tons y"1)
3875
3054
337
4315
8870
984
NORMALIZED
BY ESTUARY
AREA
(tons km2 y"1)
310.3
55.6
108.4
115.1
262.6
49.3
NORMALIZED
BY ESTUARY
VOLUME
(tons m3 y"1)
2.0 x 10'4
1.5 x 10'5
2.4 xlO'4
6.2 x 10'5
1.2x 10'4
3.3x 10'5
PHOSPHOROUS
TOTAL
(tons y"1)
61
95
NA
73
163
48
NORMALIZED
BY ESTUARY
AREA
(tons km2 y"1)
4.9
1.7
NA
2.0
4.8
2.4
NORMALIZED
BY ESTUARY
VOLUME
(tons m3 y"1)
3.2 x 10'b
4.5 x lO'7
NA
l.Ox 10'b
2.2 x 10'b
1.6x 10'b
VO
Table 3-9. Comparison of nitrogen sources during wet and dry seasons for the Yaquina Estuary (Brown and Ozretich, 2009).
SOURCE
River
Ocean
Wastewater
Benthic Flux1
Atmospheric Deposition
On Estuary
On Watershed
NITROGEN INPUT (mol DIN d'1)
WET SEASON
2.6 x 105
8.8x 104
1.8 x 103
-
2.2 x 102
l.lx 104
DRY SEASON
2.3 x 104
3.8-5.1 x 105
1.5 x 103
4.3 x 104
1.2x 102
6.0 x 103
ANNUAL
AVERAGE
1.6 x 105
2.3- 3.0 x 105
1.6 x 103
-
1.7 x 102
8.5x 103
Source: 'DeWitt et al. (2004)
-------
CHAPTER 4:
WATER QUALITY SURVEYS IN SEVEN TARGET ESTUARIES
Cheryl A. Brown and Christina L. Folger
Key Findings
• Water quality conditions in the target estuaries during the dry season were
dependent upon ocean conditions at time of sampling and fresh water inflow.
• The coastal ocean was the primary source of phosphate during the dry season.
• Estuaries received nitrogen from coastal ocean and watershed.
• Dry season chlorophyll a levels were relatively low in target estuaries, with 85%
of the stations sampled having chlorophyll a levels < 5 jig I"1.
• No incidence of low dissolved oxygen (< 5 mg I"1) occurred at any of the stations
sampled.
• There were some instances of poor water quality conditions (based on high
dissolved inorganic nitrogen) in Tillamook Estuary.
• Wet season dissolved inorganic nitrogen levels appear to be related to the red
alder cover within the watersheds.
4.0 Introduction
We collected water quality data in seven target estuaries to evaluate current water quality
conditions and for use in dividing each estuary into ocean- and river-dominated segments with
respect to nitrogen sources. These estuaries (Alsea, Nestucca, Yaquina, Salmon River, Coos,
Umpqua River and Tillamook) vary in size from about 3 km2 to 55 km2, and from river to ocean
dominated (Table 3-1 and 3-2). Our sampling consisted of high tide and low tide water quality
cruises and of short-term deployments of water quality datasondes during the months of June
through September of 2004 and 2005.
Most of this chapter focuses on the dry season (May to October), because this is the time period
of biological nutrient utilization. During the wet season (November to April), there is little
nutrient utilization due to short residence times (associated with high freshwater inflow) and low
solar irradiance. Mixing diagrams from Yaquina (unpublished data) and Tillamook estuaries
(Colbert and McManus, 2003) reveal conservative transport of nutrients during the wet season.
In addition, water column chlorophyll a and macroalgae biomasses are minimal during the wet
season.
96
-------
Previous research (e.g., de Angelis and Gordon, 1985; Roegner and Shanks, 2001; Roegner et
al., 2002; Colbert and McManus, 2003) has demonstrated that PNW estuaries are strongly
influenced by conditions occurring on the shelf, in particular wind-driven coastal upwelling
during the spring and summer. The NOs, PC>43", and temperature of water entering estuaries
during flood tides respond rapidly to changes in along-shore wind stress. High levels of
dissolved inorganic nitrogen (primarily in the form of NOs ) and PC>43" entering the Yaquina
Estuary lags upwelling-favorable winds by about 2 days (Brown and Ozretich, 2009). The input
of phytoplankton lags upwelling favorable winds by approximately 6 days and typically occurs
during downwelling conditions (Brown and Ozretich, 2009). Variations in water properties
determined by ocean conditions propagate approximately 11-13 km into the Yaquina Estuary.
Recently, there have been occurrences of severe hypoxia on the inner continental shelf of Oregon
(Grantham et al., 2004; Chan et al., 2008). At times low oxygen water from the inner shelf is
advected into the Yaquina Estuary (Brown et al., 2007).
4.1 Methods
4.1.1 Water Quality Cruises
Water quality cruises were conducted in the seven target estuaries during the summers of 2004
and 2005. During each cruise between 10 and 17 stations were sampled in each estuary,
depending upon the size of the estuary, and high and low tide cruises were conducted. The time
required to complete each cruise depended upon the size of the estuary, with small estuaries
(Salmon River, Nestucca, and Alsea) taking 2-3 hours and larger estuaries (Coos and Tillamook)
taking about 4-5 hours. Due to logistical constraints some of the riverine stations in Coos and
Tillamook were sampled up to 7 hours after the beginning of the cruise. To minimize the time
required to complete each cruise, two boats were used to sample the three largest estuaries
(Coos, Tillamook and Umpqua River). Both the low and high tide cruises proceeded from the
estuary mouth upriver to follow the propagation of the tide. For location of water quality cruise
stations in each target estuary see Chapter 5. The cruises extended from the marine to the tidal
fresh regions for all estuaries except Coos Estuary. For Coos Estuary, the lowest salinity
sampled during the dry season was 14 psu. The dates of the cruises and the number of stations
occupied in each estuary are presented in Table 4-1. We didn't sample during May or October
because these are transitional months that may experience high freshwater inflow events.
Additional cruises were conducted to characterize winter conditions during the wet seasons of
2006 and 2007.
Table 4-1. Dates of cruises and number of stations sampled for the seven target estuaries.
ESTUARY
Alsea
Salmon
Yaquina
Coos
Nestucca
Tillamook
Umpqua
DATES OF DRY SEASON
CRUISES
HIGH TIDE
9/24/04
7/20/04
6/15/04
8/9/05
8/1 1/04
7/22/05
6/24/05
LOW TIDE
9/28/04
7/16/04
6/21/04
8/8/05
8/20/04
7/23/05
6/25/05
DATES OF WET SEASON
CRUISES
HIGH TIDE
3/11/07
3/21/07
2/9/06
2/16/07
1/25/07
3/2/07
2/1 1/07
LOW TIDE
3/12/07
3/27/07
2/15/07
1/27/07
3/1/07
2/12/07
NUMBER
OF
STATIONS
10
10
10
17
15
15
13
97
-------
At each station profiles of conductivity, temperature and depth (CTD; SEE 19 SEACAT Profiler,
Sea-Bird Electronics, Inc., Bellevue, Washington), turbidity (Seapoint Turbidity Sensor,
Seapoint Sensors, Inc., Kingston, New Hampshire), photosynthetically active radiation (PAR,
LI-193 spherical sensor, LI-COR Biosciences, Lincoln, Nebraska) and in situ fluorescence
(WETStar Chlorophyll Fluorometer, WET Labs, Philomath, Oregon) were measured. All
instruments on the Sea-Bird profiler were factory calibrated by the manufacturer. The profile
measurements were taken at 0.5-sec intervals from the water surface to 0.5 m above the bottom,
and during post-processing the data were binned into 0.25-m intervals. If the water depth at the
station was shallow (typically less than 4 m), a YSI6600 multiparameter sonde (YSI, Yellow
Springs, Ohio) was used for water quality measurements (temperature, conductivity,
fluorescence, turbidity, depth, and dissolved oxygen). Dissolved oxygen measurements for the
profiles were obtained using a YSI 6600 attached to the Seabird frame. Dissolved oxygen
measurements (surface, mid-depth, and bottom) were collected in Tillamook, Coos and Umpqua
River estuaries during both the low and high tide cruises, and during the low tide cruise in
Salmon River Estuary. No dissolved oxygen measurements were collected in the Alsea, Yaquina
and Nestucca estuaries. The YSI datasondes were calibrated using the methods presented in
Section B.4.
Water samples were collected from mid-depth at each station using a hand-operated pump,
filtered (0.45 |im filter), and frozen until analysis. The samples were analyzed for dissolved
inorganic nutrients (nitrate+nitrite, ammonium, phosphate and silicate) by MSI Analytical
Laboratory, University of California-Santa Barbara, CA. One-liter surface water samples were
collected from each station and analyzed for chlorophyll a. These water samples were filtered
within 2 to 4.5 hours of sample collection using 47-mm GF/F filters. The volume of water
filtered varied between 250 and 500 ml. Chlorophyll a was extracted by sonicating the filters
and soaking them overnight in 10 ml of 90% acetone. The next morning the samples were
centrifuged and analyzed for chlorophyll a content using a fluorometer (10 AU Fluorometer,
Turner Designs, Inc., Sunnyvale, CA). Four-liter water samples were collected from each cruise
station and analyzed for total suspended solids (TSS). Each sample was agitated prior to
filtration and filtered using ashed Whatman 47 mm GF/F filters. Filters from samples were dried
overnight in an oven at 70°C and allowed to reach room temperature in a desiccator prior to
weighing. More information on the water column sample analyses can be found in Section B.2.1
and B.2.2
4.1.2 Datasonde Deployment
YSI 6600 multiparameter sondes (YSI, Yellow Springs, Ohio) and Mini-CTDs (Star-Oddi,
Iceland) were deployed in the seven target estuaries along the salinity gradient (see Table 4-2 for
number of instruments deployed in each estuary and Chapter 5 for maps illustrating deployment
locations). The YSI datasondes measured conductivity, temperature, turbidity, chlorophyll a (in
situ fluorescence), dissolved oxygen, pH, and depth. Mini-CTD units measured depth, salinity
and temperature. Both instruments collected data at 15 minute intervals. During 2004, we lost
two instruments deployed in the lower portion of the Nestucca Estuary. In order to fill this data
gap, in 2006 an additional YSI 6600 CTD was deployed at the mouth and three additional Mini-
CTDs were deployed in the mesohaline region of the estuary. Calibration procedures for all
instrumentation are presented in Sections B.2-B.4. In addition to the datasondes that we
deployed, there were additional YSI datasondes deployed in Coos Estuary associated with the
98
-------
South Slough Estuarine Research Reserve (Charleston Bridge and Valino Island) and Oregon
Institute of Marine Biology (OIMB). The locations of these instruments (labeled Y5-Y7) are
presented in Figure 5-10.
Table 4-2. Number of short-term datasondes deployed during the dry season in each estuary and
year deployed.
ESTUARY
Alsea
Coos
Nestucca
Tillamook
Salmon
Umpqua
Yaquina
NUMBER OF YSI
DATASONDES
2 (2004)
7 (2005)
2 (2004)
1 (2006)
4 (2005)
2 (2004)
4 (2005)
5 (2004)
NUMBER OF MINI-
CTDS
3 (2004)
3 (2005)
3 (2004)
3 (2006)
2 (2005)
3 (2004)
3 (2005)
0
4.2 Riverine Nutrient Inputs During the Wet Season
To examine differences in riverine nutrient levels for the target estuaries, we calculated median
wet season dissolved inorganic nitrogen (DIN) and PO43" using stations with salinity < 2 psu
(Table 4-3). During the wet season, the Yaquina estuary had the highest riverine DIN levels,
while Umpqua had the lowest. In general, estuarine differences in wet season DIN were similar
to variations in deciduous cover among the estuaries (Table 4-3). This suggests that variations in
riverine DIN among the estuaries primarily result from variations in red alder cover, which is
similar to the findings of Compton et al. (2003) and Wigington et al. (1998). The Nestucca
estuary had the highest median wet season PO43" (Table 4-3). The highest wet season DIN levels
observed in the target estuaries were measured at Stations C4 and C12 in the Tillamook Estuary
(Figure 5-12). Highest median NH4+ levels were measured in the Tillamook Estuary and the
highest NH4+ values measured occurred at Stations C4 and C12 in the Tillamook Estuary.
Table 4-3. Median wet season dissolved inorganic nitrogen and phosphate calculated using
stations with salinity < 2 psu. The table entries are listed from highest to lowest DIN levels.
ESTUARY
Yaquina
Tillamook
Nestucca
Alsea
Salmon
Coos
Umpqua
MEDIAN WET SEASON
DIN (uM)
92.6
69.8
61.9
35.0
34.9
33.3
17.6
NH4+ (uM)
0.93
2.66
0.87
0.99
0.88
1.04
1.60
P04J- (uM)
0.35
0.40
0.59
0.38
0.25
0.24
0.28
DECIDUOUS
COVER (%)
5.19
5.85
4.63
3.07
3.43
1.45
0.49
99
-------
4.3 Coastal Ocean and River Conditions During Dry Season Sampling
The seven target estuaries span about 250 km along the Oregon Coast (Figures 2-3 and 2-4).
Flood-tide water temperature can be used as an indicator of ocean conditions during our surveys
and to assess the variability in upwelling along the Oregon Coast. Low flood-tide water
temperatures (8 - 10°C) indicate the input of high-nutrient (NOs and PO43") water to the
estuaries associated with upwelling while warm temperatures indicate downwelling conditions
and input of low-nutrient water from the ocean. Figure 4-1 shows the flood-tide water
temperature near the mouth of Yaquina Estuary (Station Yl, which has been previously been
demonstrated as an indicator of upwelling/downwelling; Nelson and Brown, 2008) and water
temperature for the high tide cruises for the outermost estuarine stations sampled for the 2004
classification effort. Figure 4-2 shows the flood-tide water temperature near the entrance of
Yaquina and Coos estuaries (which are about 140 km apart) and the water temperature from the
high tide cruises from the 2005 classification effort. The close agreement between Yaquina and
Coos estuaries flood-tide water temperatures and water temperatures from the classification
cruises demonstrates that the ocean conditions are relatively uniform over the geographic range
of estuaries that we sampled during 2004 and 2005. This agrees with the analysis by Hickey and
Banas (2003) that demonstrated that water temperatures at the entrance of three estuaries along
the Oregon and Washington coasts (Coos, Grays Harbor and Willapa), which spanned 400 km,
were highly correlated during the upwelling season. During 2004, the Salmon River and
Nestucca estuaries were sampled during low oceanic nutrient conditions, with the high tide
cruise outermost stations having NOs'+MV levels of < 1.3 uM and PO43" levels of < 0.5 uM.
There was a delayed onset of coastal upwelling along the Oregon coast during 2005 with weak
upwelling conditions occurring from late May to mid July and strong upwelling conditions not
commencing until mid July (Kosro et al., 2006). During 2005, all estuaries that we sampled
during high tide coincided with upwelling conditions, which is indicated by the relatively high
NOs'+MV and PO43" conditions at the outermost stations. The upwelling conditions were
stronger during the Tillamook and Coos estuaries cruises than the Umpqua cruises (Figure 4-2).
This is evident in the NOs'+MV and PO43" levels during the high tide cruise at the outermost
stations of the Tillamook and Coos estuaries compared with the Umpqua River Estuary.
Nitrate+nitrite concentrations at the outermost stations were 19.1 and 23.6 uM for the Coos and
Tillamook estuaries, respectively, compared to 12.9 uM in the Umpqua River Estuary.
We used water temperature versus NO3"+NO2" and PO43" relationships generated using data from
the inner shelf off of Newport, Oregon (Wetz et al., 2005), to determine whether the nutrient
concentrations in the water entering the estuaries during flood tides were consistent with values
associated with coastal upwelling (Figures 4-3 and 4-4). For all of the estuaries, except for
Alsea, the high tide outermost stations fell along the relationships developed using offshore data,
suggesting that the MV and PO43" levels near the estuary mouths were consistent with coastal
upwelling.
The elevated NOs'+MV concentrations at the mouth of the Alsea Estuary were associated with
high freshwater input. There was a relatively high freshwater inflow event on September 19,
2004, which was 5 days prior to the high tide Alsea cruise. On September 19, the Alsea River
gauge at Tidewater (http://waterdata.usgs.gov/or/nwis/sw) had a discharge of 1 120 cubic feet s"1
(cfs), which is almost an order of magnitude higher than the long-term mean daily discharge for
September (130 cfs). We had additional cruises of the Alsea Estuary on September 21, 2004,
100
-------
which had an NCV+NCV concentration at the most riverine station of 73.8 uM (with salinity =
0 psu). The Oregon Department of Environmental Quality (http://deql2.deq.state.or.us/lasar2/)
sampled the Alsea River at a station about 20 miles upriver from the mouth of the estuary on
September 22, 2004 which had aNOs'+NCh" concentration of 49.4 uM, which was the highest
value measured at this station during September in the past 11 years. The median NCb'+NCV
concentration at this station for the month of September was 6.3 uM (data from 1994-2005,
n = 9). The anomalous NOs'+NCh" concentrations at the Alsea Estuary were a result of this first
freshwater inflow event of the fall. Interestingly, the PC>43" concentrations during the 2004 Alsea
cruise were not anomalous.
20
Yaquina
Classification
Salmon
18 -
6/1/2004
1
7/1/2004
r
8/1/2004
Date
T
9/1/2004
Figure 4-1. Flood-tide water temperature at Yaquina Estuary (solid line) and outermost high tide
station from 2004 dry season classification cruises (filled circle). Each datapoint from the
classification cruises is labeled with its estuary name.
101
-------
18 -
8
Yaquina
Coos
Classification
T
T
T
T
T
6/1/2005 6/15/2005
6/29/2005 7/13/2005
Date
7/27/2005 8/10/2005
Figure 4-2. Flood-tide water temperature at Yaquina (black line) and Coos (gray line) estuaries
and outermost high tide stations from 2005 dry season classification cruises (filled symbols).
4.4 Estuarine Water Quality
4.4.1 Dry Season Chlorophyll a and TSS Patterns
In Tables 4-4 and 4-5 we present the median dry season chlorophyll a, TSS, NOs'+MV, NH4+,
and PCM3" for the target estuaries and rank the estuaries by these median values. Dry season
chlorophyll a within the seven target estuaries was low with a median value of 2 ug I"1 (all
estuaries and dry season cruises). Chlorophyll a was in the low to medium category for
eutrophication symptoms (Bricker et al., 2003). Yaquina Estuary had the highest median
chlorophyll a (5.6 ug I"1), while the Umpqua River Estuary had the lowest (0.9 ug I"1). The
highest chlorophyll a measured was 17.3 ug I"1, which occurred near the mouth of the Coos
Estuary (at Station C2). The Coos Estuary had the highest median total suspended solids
(13.1 mg I"1), while the Umpqua River Estuary had the lowest (1.6 mg I"1). Our data suggests
that ocean-dominated estuaries have higher chlorophyll a than river-dominated estuaries. It is
not possible to separate out the chlorophyll a associated with oceanic import from that growing
in situ in the estuary from this dataset.
102
-------
40 -n
30 -
o" 2°-
z
o"
Z 10 -
0
(
%NH-5
o Classification
•
0
°
o ^
5 8 10 12 14 16 18 2
0
Water Temperature (deg C)
Figure 4-3. Water temperature versus NOs'+MV relationship generated using data from a
station on the shelf (NH-5, Wetz et al., 2005; gray circles) and classification high tide outermost
stations (hollow circles). All data are from the dry season. Data point from Alsea is identified.
3 -
~ 2-
0*
PH
1 -
0
(
o Classification
^l*
00
OO
0
1 I 1 I 1 , 1 , 1 , 1
5 8 10 12 14 16 18 2
0
Water Temperature (deg C)
Figure 4-4. Water temperature versus PC>4 " relationship generated using data from a station on
the shelf (NH-5, Wetz et al., 2005; gray circles) and classification high tide outermost stations
(hollow circles). All data are from the dry season.
103
-------
Roegner and Shanks (2001) demonstrated that oceanic chlorophyll a was imported into South
Slough in the Coos Estuary and that chlorophyll a concentrations were variable ranging from 0
to 15 ug I"1. Previous research in the Yaquina Estuary (Brown and Ozretich, 2009) has
demonstrated that water entering the estuary during flood tides can have peak chlorophyll a
concentrations of 40-50 ug I"1, and that the chlorophyll a levels in the ocean-dominated portion
of the estuary are dependent upon coastal wind forcing and are quite variable. The median
chlorophyll a imported into the Yaquina Estuary during May-September of 2002 and 2003 was
4 ug P (n = 298).
There was no clear pattern in chlorophyll a (from water samples) versus salinity, except for
during the high tide cruise in Coos Estuary (Figure 4-5). During the high tide cruise in the Coos
Estuary there was evidence of elevated chlorophyll a levels at higher salinities. In situ
fluorescence data from two of YSI datasondes in the Coos Estuary (Stations Yl and Y3) showed
the import of chlorophyll a from the coastal ocean to the estuary (indicated by a significant
relationship between in situ fluorescence and salinity) and a strong tidal signal. The low
chlorophyll a values in Umpqua were consistent with the in situ fluorescence data. The median
values of in situ fluorescence at Stations Yl and Y2 in Umpqua were 1.5 ug I"1, and a tidal signal
was not apparent. For comparison, the median in situ fluorescence values at Stations Yl and Y3
in Coos Estuary were 8.2 and 3.7 ug I"1, respectively. The median value of chlorophyll a
observed in the Yaquina Estuary during this study was consistent with historical dry season data
from this estuary (median = 4.9 ug I"1, n = 1205; Brown et al., 2007). Additionally, the median
value in the Alsea Estuary was similar to that from monthly sampling conducted from May to
September 2004 (median of all sampling events at six stations = 1.6 ug I"1, n = 30).
For PNW estuaries, we compiled all available chlorophyll a data, including other data collected
by our laboratory, National Coastal Assessment, Oregon Department of Environmental Quality,
and Washington Department of Ecology. Figure 4-6 shows the median dry season chlorophyll a
for eight PNW estuaries (including six of the target estuaries) and the freshwater inflow
normalized by estuary volume (Table 2-4). The median chlorophyll a patterns obtained by
compiling data from numerous sources is consistent with values we obtained in our surveys.
Median dry season chlorophyll a in the target estuaries appears to be related to freshwater inflow
normalized by estuary volume. Estuaries with the low freshwater inflow normalized to volume
have relatively high chlorophyll a with median values ranging from 3.3 - 4.9 ug I"1 for these tide-
dominated estuaries (Yaquina, Coos, Netarts and Willapa). Estuaries with relatively high
freshwater inflow normalized to volume (e.g., Alsea, Salmon and Umpqua) have low median
chlorophyll a (1-2 ug I"1). The tide-dominated estuaries may have relatively high chlorophyll a
compared to the river-dominated system because they have less flushing due to low freshwater
inflow and as a result phytoplankton growing inside the estuary reach higher levels. An alternate
explanation may be that the higher chlorophyll a levels in tide-dominated estuaries are due to
import of chlorophyll a from the coastal ocean. Newton and Horner (2003) demonstrated that
high productivity phytoplankton blooms are imported into Willapa Estuary from the coastal
ocean and have species that are of oceanic origin; however, moderate blooms also occur within
the estuary which are combination of phytoplankton species of oceanic and estuarine origin.
This suggests that the high chlorophyll a levels in the ocean-dominated estuaries maybe due to a
combination of the two explanations.
104
-------
Table 4-4. Median dry season chlorophyll a and total suspended solids (TSS) for each estuary
and its ranking. In the ranking, 1 represents the estuary with the highest median value and 7
represents the estuary with the lowest.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
MEDIAN CHL a
(ug r1)
1.0
2.4
2.8
1.6
1.7
0.9
5.6
RANKING BY
CHL a
6
3
2
5
4
7
1
MEDIAN TSS
(mgr1)
7.4
13.8
4.9
5.1
3.7
1.6
9.7
RANKING BY
TSS
3
1
6
5
4
7
2
Table 4-5. Median dry season
, NH4+, and PO43" for each estuary and its ranking.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
MEDIAN
NO3"+NO2"
(uM)
30.9
6.1
10.4
3.3
34.0
1.9
6.1
RANKING
BY
NO3"+NO2"
2
5
3
6
1
7
4
MEDIAN
NH4+
(uM)
4.6
3.2
2.2
1.4
2.0
0.9
1.7
RANKING
BY
NH4+
1
2
3
6
4
7
5
MEDIAN
PO43"
(uM)
0.7
0.9
0.7
0.5
0.8
0.3
0.8
RANKING
BY
PO43'
5
1
4
6
3
7
2
105
-------
15 -
^
S/j
i
V 10 -
^
o.
o
•_
8 5-
0
^
• •
oo *
O XX •
^ o
£ (J A^f
m w ^ D(57 P5^ #n *, ^
J£p^^> v^7 & ^ ^^PW
i I i I i I i
• Tillamook - High
n Tillamook - Low
• Coos - High
o Coos - Low
A Umpqua - High
A Umpqua - Low
* Alsea - High
O Alsea - Low
• Yaquina - High
o Yaquina - Low
T Salmon - High
v Salmon - Low
* Nestucca - High
* Nestucca - Low
0 10 20 30
Salinity (psu)
Figure 4-5. Dry season chlorophyll a versus salinity for all estuaries with filled symbols
representing high tide cruises and hollow symbols representing low tide cruises.
5 -
1 4-
1 3-
o.
o
Median Chloi
O H- tO
1,1,1
• Yaquina (n = 1205)
f Netarts(n = 21)
• Willapa (n = 606)
• Coos (n = 150)
9 •
Tillamook (n = 134) TT , _„, •
Umpqua (n = 52)
A1 , ^ Salmon (n = 168)
Alsea (n = 75)
1 , | , | , |
100 200 300 400 5(
)0
Freshwater Inflow Normalized by Volume (year )
Figure 4-6. Median dry season chlorophyll a versus freshwater inflow normalized by estuary
volume for eight PNW estuaries. The chlorophyll a includes data from this study, NCA,
Washington Department of Ecology, and Oregon DEQ.
106
-------
4.4.2 Dry Season Nutrient Patterns
The Tillamook Estuary had the highest median dry season concentration of NOs +NO2 , while
the Umpqua River Estuary had the lowest (average of all stations, low and high tide cruises;
Table 4-5). The Alsea Estuary had similar median NCV+ NO2" (rank = 2) as Tillamook, but this
was due to the high freshwater inflow event prior to the cruises. Additional sampling in the
Alsea Estuary suggested that the NOs'+MV and NH4+ levels were anomalously high during the
dry season cruises. Monthly cruises conducted in the Alsea Estuary from May to September
2004 had a median NO3"+NO2" value of 7.5 uM and an NH4+ value of 1 .7 uM (n = 30). The
median values of the nutrients in the Yaquina Estuary were similar to medians calculated from
weekly cruises (with 6 stations) conducted from May 5 - September 27, 2004 (median NCb
+NO2" = 10.9 uM, NH4+ = 2.7 uM, and PO43" = 0.9 uM). In the Tillamook Estuary about 44% of
the stations were located in low salinity regions of the estuary (with salinity < 5 psu), and these
low salinity stations had high NCb'+NCV levels (median = 40.3 uM). Excluding low salinity
stations (< 5 psu), the median concentration of MV+MV in Tillamook was about 21 uM. We
recalculated the median NCV+MV only using stations with salinity > 5 psu and the ranking of
the estuaries by median NCb'+NCV was Alsea, Tillamook, Nestucca, Coos, Yaquina, Umpqua
River, and Salmon River (highest to lowest).
The Salmon River Estuary had the lowest MV+MV probably due to the low nutrient
conditions in the coastal ocean during the sampling of this estuary. The relatively high NC>3
+NO2" levels (calculated from stations with salinities > 5 psu) in the Tillamook Estuary were
probably associated with strong upwelling conditions prior to the cruises (see Section 4.3)
combined with relatively high watershed inputs. Colbert (2004) found median NOs +NO2 levels
of ~ 8 uM in Tillamook Estuary (using data from May - September 1998 and 1999 with
salinities > 5 psu, n = 46), suggesting there is substantial interannual variability. Historical data
in Umpqua Estuary from the Oregon Department of Environmental Quality confirmed the low
median dry season nutrients in this estuary (median NOs +NO2 =1.6 uM and PO43" = 0.3 uM , n
= 15).
To examine differences in riverine nitrogen inputs among the target estuaries during the dry
season, we calculated the median dissolved organic nitrogen (DIN) using only stations with
salinity < 5 psu. Unfortunately, for the Coos Estuary our sampling did not extend far enough
upriver to examine riverine nitrogen input. Alsea, Nestucca, Tillamook and Yaquina estuaries
had similar DIN levels at the low salinity stations (< 5 psu) with median values of about 40 uM,
while the Umpqua River Estuary had the lowest (2.4 uM), and the Salmon River was
intermediate with median values of about 19 uM. Dry season riverine DIN input appeared to
follow the trends in deciduous cover in the watersheds (Table 3-3), with Alsea, Nestucca,
Tillamook and Yaquina watersheds having the highest deciduous cover ranging from 3.4 to
5.9%, and the Umpqua watershed having the lowest 0.49%.
During the dry season, there was a significant relationship between PO43" and salinity (Figure 4-
7) which demonstrates that the ocean was the dominant PO43" source for the target estuaries. A
substantial amount of the variability in the Figure 4-7 is due to temporal variations in ocean
conditions. In the PO43" versus salinity plots for individual cruises, the variability was reduced
107
-------
2.0 -
r = 0.40, n = 166
p< 0.0001
O
O.
0.0
0
10 15 20
Salinity (psu)
25
30
35
Figure 4-7. Dry season phosphate versus salinity for all classification water quality stations (all
estuaries, both years, high and low tide cruises).
r = 0.2, n = 172
p< 0.0001
0
0
10
15
20
25
30
35
Salinity (psu)
Figure 4-8. Dry season nitrate+nitrite versus salinity for all classification water quality stations
(all estuaries, both years, high and low tide cruises).
108
-------
9 Q
and the r was as high as 0.97 for some cruises. There was evidence of a PO4 " source at low
salinities in all the target estuaries. Plotting NOs'+NCh" versus salinity (Figure 4-8) demonstrates
that the estuaries receive NCV+MV from both the river and the ocean and that the riverine
concentrations are higher than those associated with the ocean. EMAP data for Oregon estuaries
exhibited a significant correlation between PC>43" and salinity, but no relationship between
dissolved inorganic nitrogen and salinity (Nelson and Brown, 2008). There were two
anomalously high (-90 uM) NCb'+NCV datapoints that occurred at low salinities, which
correspond to two sloughs in Tillamook Estuary. These two datapoints will be discussed in
Section 4.5.
4.4.3 Dry Season Salinity Fluctuations
Salinity conditions near the mouth of the seven target estuaries were compared to assess the
importance of riverine versus marine dominance during the dry season. For this analysis, we
examined the data from the short-term datasondes for the station nearest the mouth in each
estuary. Since we lost the datasonde deployed near the mouth of the Nestucca Estuary during
mid August 2004, we used data collected in late August 2006. To confirm that this was an
adequate substitution, we compared the high and low tide salinity data from the cruises of the
outermost station (Station CIO) to the data collected in 2006. The high and low tide data from
Station CIO were comparable to the high and low tide values from the datasondes deployed in
2006 (Station Y2) suggesting that this is an appropriate substitution. The datasondes were
located at varying distances from the estuary mouths. The datasonde deployed at Station Y2 in
the Nestucca Estuary was located at the mouth. The ones deployed in the Umpqua River and
Salmon River estuaries were about 1 km from the mouth. The datasonde deployed at Station Yl
in Tillamook Estuary was about 2 km from the estuary mouth, while the ones deployed in
Yaquina and Alsea estuaries were about 3 km from the mouth. In the Coos Estuary, we had a
Mini-CTD deployed near the mouth; however, it malfunctioned, so our datasonde closest to the
mouth was 6 km from the inlet. Additional data were provided from OEVIB, which is about 1 km
from the estuary mouth.
In the Yaquina Estuary, the salinity varied about 5 psu over a tidal cycle, even though this station
is located about 3 km from the estuary mouth (Figure 4-9). In the Alsea Estuary, the salinity
varied about 24 psu over a tidal cycle and this datasonde was deployed at a similar distance from
the estuary mouth as the one in the Yaquina. This difference between the salinity variations
between Yaquina and Alsea was probably due to differences in freshwater inflow. The dry
season freshwater inflow to the Alsea Estuary is three times that into the Yaquina Estuary
(Table 3-6). The influence of the high freshwater inflow event on September 19, 2004 was
evident in the decrease in low tide salinities near the estuary mouth. Salinity data from Station
Cl near the mouth of the Alsea Estuary differed by 7 psu between the high and low tide cruises,
which was less than that at Station Yl due to proximity to the estuary mouth. Even though the
Salmon River datasonde was located close to the estuary mouth the salinity varied about 23 psu
over a tidal cycle. These salinity variations were comparable to those in Alsea; however, the
Alsea datasonde was deployed about 2 km further upstream than the one in Salmon River
estuary. The salinity variations near the mouth of the Nestucca Estuary varied about 13 psu over
a tidal cycle (Station Y2). Salinities near the mouth of the Umpqua River Estuary varied about
19 psu over a tidal cycle. Using the datasonde near the mouth of Tillamook (Station Yl), the
salinity varied about 14.6 psu over a tidal cycle. When plotted on the same axis, the salinity
109
-------
variations at Tillamook were comparable to those at Nestucca (but the Nestucca datasonde was
located closer to the mouth of the estuary). The salinity variations near the mouth of Coos
Estuary are similar to (or slightly less than) those in the Yaquina Estuary.
Due to variability in the distance of the short-term datasondes from estuary mouths, we
examined the salinity differences between the high and low tide of the station closest to the
estuary mouth to rank the estuaries in terms of riverine dominance. Salinity variations at the
mouth appear to be related to normalized freshwater inflows presented in Table 2-4 (Figure 4-
11). Based on salinity variations and normalized freshwater inflow, the ranking of estuaries in
terms of riverine dominance would be Umpqua (most riverine), Nestucca, Alsea, Salmon,
Tillamook, Yaquina and Coos (most marine).
4.5 Comparison of Water Quality Data to EMAP Criteria
Dry season water quality data were compared to EPA's Environmental Monitoring Assessment
Program (EMAP) West Coast criteria (Table 4-6). Data collected in the seven target estuaries
were consistent with the EMAP data discussed in Section 1.3; however, since our water quality
stations were not probabilistically sampled, we cannot express our results as percent of estuarine
area. There was no incidence of low dissolved oxygen (< 5 mg I"1) at any of the stations that we
sampled. All stations (both low and high tide cruises of all estuaries) had dissolved inorganic
phosphorous levels in the fair range. As discussed previously, this was primarily resulting from
oceanic input of PC>43" to the estuaries (Figure 4-7). The majority (85%) of the stations sampled
(all estuaries) had chlorophyll a levels in the good range, with the remainder in the fair range.
The chlorophyll a was in the good range (< 5 ug I"1) for all stations and both sampling dates in
the Alsea, Salmon River, and Umpqua River estuaries. In the Coos, Nestucca, and Tillamook
estuaries most of the stations had chlorophyll a in the good range, but there were a few instances
of chlorophyll a in the fair range (Coos had 4 samples, Nestucca had 6 samples, and Tillamook
had 1 sample in the fair range). In Yaquina Estuary, most of the chlorophyll a samples were in
the fair range (14 out of 20). Interestingly, the occurrences of chlorophyll a in the fair category
occurred near the mouth in the Coos and Nestucca estuaries suggesting import of high
chlorophyll a water from the coastal ocean. In contrast, the relatively high chlorophyll a
concentrations in the Yaquina Estuary extended from the mouth to the tidal fresh portion of the
estuary.
Most (85%) of the stations sampled had dissolved inorganic nitrogen (DIN) levels in the good
category (with 14% in the fair category and 1% in the poor category). All samples in the Coos,
Salmon River and Umpqua River estuaries had DIN concentrations in the good category. All of
the occurrences of DIN in the fair category occurred in the riverine portions of the estuaries with
salinities less than 6.5 psu. The only estuary that had DIN concentrations in the poor range was
Tillamook (Stations C4 & C12 during low tide cruise which are located in Hathaway and
Dougherty sloughs with salinities of 0.2 and 11.6 psu). Unlike the anomalously high NCV
conditions in the Alsea Estuary during the classification cruises, there was not high river inflow
to the Tillamook Estuary prior to this cruise. There was additional nutrient sampling in the
vicinity of C4 and C12 by the Oregon Department of Environmental Quality during 1997 and
1998, which had similar high dry season DIN (peak values of DIN of about 100 uM compared to
110
-------
Yaqmna(Yl)
Salmon (Yl)
Alsea(Yl)
Nestucca (Y2)
Nestucca (CIO)
Alsea(Cl)
7/1/2004
7/22/2004
8/12/2004
Date
9/2/2004
9/23/2004
Figure 4-9. Dry season salinities near the mouths of estuaries sampled in 2004 and data
collected in Nestucca (Y2) during 2006 (shifted by 2 years) to fill data gap from lost instrument.
The dashed line indicates the high freshwater inflow event on September 19, 2004.
10 -
0
o
Yaquina(Yl)
•Umpqua(Yl)
Coos(Yl)
• Coos (Y7)
•Tillamook(Yl)
Umpqua(Cl)
Tillamook(Cl)
Coos(Cl)
6/15/2005
I
6/29/2005
I
7/13/2005
Date
I
7/27/2005
r
8/10/2005
Figure 4-10. Dry season salinities near the mouths of estuaries sampled in 2005. Lines indicate
short-term datasonde deployments, while symbols represent data from high and low tide cruises
for station nearest the mouth of the estuary with station name present in parentheses of legend.
Ill
-------
£
2 psu difference between
surface and bottom readings. Due to the shallow water depth during the low tide cruises in the
Salmon River and Nestucca estuaries, only one data point at mid-depth was collected; therefore,
stratification could not be determined for these cruises.
112
-------
Table 4-6. West Coast criteria for water quality parameters (U.S. EPA, 2004b).
PARAMETER
Dissolved Oxygen
Chlorophyll a
Dissolved inorganic
nitrogen
Dissolved inorganic
phosphorous
RANKING
GOOD
> 5 mg I'1
< 5 ug I'1
<0.5mgr1
<0.01 mgl'1
FAIR
2-5 mg I'1
S-lOugr1
0.5-1.0 mgl'1
0.01- 0.1 mgl'1
POOR
< 2 mg I'1
>2o^igr1
> 1 mg I'1
>0.1 mgl'1
In aggregate (all cruises, all estuaries, all stations) about 29% of the stations sampled exhibited
strong stratification. The estuary with the highest number of stations exhibiting strong
stratification was the Alsea Estuary (45% of stations). The estuaries with the lowest number of
stations exhibiting strong stratification were the Tillamook and Coos estuaries (13 and 19%,
respectively). The high degree of stratification in Alsea was probably related to the high
freshwater inflow event discussed earlier in this chapter. For estuaries where we had
stratification data for low and high tide cruises (all estuaries except for Salmon River and
Nestucca), strong stratification occurred more during the high tide sampling compared to the low
tide (19 occurrences during high tide versus 8 during low tide cruises). Of the stations that
showed strong stratification, the majority were upriver in the low salinity region (<15 psu). This
pattern was particularly evident in the Salmon River, Nestucca, and Alsea estuaries. The
Yaquina and Coos estuaries exhibited strong stratification in the ocean-dominated regions.
4.6 Synthesis
During the wet season, DIN levels were highest in the Yaquina and lowest in the Umpqua.
Variations in the wet season DIN appear to be related to deciduous cover. This suggests that
variations in red alder are resulting in differences in riverine DIN input to the estuaries. Dry
season water quality conditions in PNW estuaries were dependent upon the ocean conditions
during the surveys. Flood-tide water temperature as well as nutrient versus salinity relationships
generated using data from the inner Oregon shelf were useful for confirming that ocean
conditions were relatively uniform over the geographic extent of the estuaries we sampled and
for assessing the ocean conditions during the dry season water quality cruises. The results from
the Alsea Estuary demonstrated that when assessing nutrient levels of the estuaries it is important
to examine freshwater inflow immediately prior to sampling events (even during the dry season)
to ensure that there was not a strong freshwater inflow event that might result in elevated DIN
levels. Dry season nutrient data from the cruises demonstrated that the estuaries received PO43"
mainly from the coastal ocean, while both the ocean and rivers were DIN sources. In the seven
target estuaries, there were no occurrences of low dissolved oxygen, and chlorophyll a levels
were relatively low. The tide-dominated estuaries appear to have higher chlorophyll a levels
than the river-dominated estuaries. The dry season PO43" conditions were in the EMAP fair
category but this is due to the input of high PO43" water from the coastal ocean. Generally, most
of the DIN concentrations observed were in the good category (85%). There were two
occurrences of poor DIN conditions in two of the sloughs of Tillamook Estuary. Umpqua River
Estuary had the lowest NOs'+MV and PO43" of the seven target estuaries. There are several
potential explanations for the low nutrients in this estuary. Of the seven target estuaries, the
113
-------
Umpqua has the least amount of red alder in the watershed as evident by the lowest % Deciduous
and % Mixed and the highest % Evergreen (Table 3-3). Red alder is the primary deciduous tree
in the Oregon Coast Range, and NCV in streams has been related to its presence (see
Section 3.7). In addition, Umpqua has the highest freshwater inflow of the seven target estuaries
(even during the dry season, Table 3-6) and as a result is river dominated, which may result in
less oceanic nitrogen and phosphorous input to the estuary.
114
-------
CHAPTER 5:
ZONATION OF OCEAN AND RIVER DOMINANCE
IN SEVEN TARGET ESTUARIES
Cheryl A. Brown and James E. Kaldy III
Key Findings
Estuaries were divided into oceanic and riverine segments in terms of nitrogen
sources during the dry season.
Zonation was based upon natural abundance of nitrogen stable isotopes of green
macroalgae and transport modeling.
Analyses suggest that oceanic and riverine inputs are the dominant nitrogen
sources for the target estuaries.
Coos Estuary had the highest isotope ratios with multiple stations having isotope
ratios elevated above the oceanic end member, suggesting that wastewater
inputs are important in this estuary.
Five estuaries (Alsea, Coos, Nestucca, Tillamook, and Yaquina) had >50% of the
total estuarine area classified as ocean dominated.
5.0 Introduction
To assess the susceptibility of estuarine resources to watershed-derived nutrient loading, we
estimated the contribution of nitrogen sources to water column dissolved inorganic nitrogen
(DIN). Previous research at Yaquina Estuary has demonstrated that during the dry season (May-
October), the ocean is the dominant nitrogen source, while during the wet season (November -
April) riverine inputs dominate (Brown and Ozretich, 2009; Table 3-9). In addition, most of the
riverine nitrogen inputs are believed to be related to the presence of red alder in the watershed
(see Section 3.7). Point source input associated with a wastewater treatment facility increases in
importance in the up-estuary region, particularly during periods of minimal riverine inputs. The
purpose of this chapter is to assess the importance of oceanic versus riverine nitrogen sources in
the seven target estuaries.
We divided each target estuary into two segments, ocean dominated and river dominated, based
upon the dominant dry season DIN sources. We used multiple types of data and analyses to
derive the zonation of the estuaries and then translated the zonation into median salinities for use
in estuaries with limited data. For the Yaquina Estuary, we had the most extensive dataset and
analysis techniques for development of zonation based upon dry season nitrogen sources. We
used a transport model combined with natural abundance stable isotopes (515N) of green
macroalgae to identify the dominant nitrogen sources within the estuary as a function of time and
location. The transport model was validated by comparing predicted isotope ratios (using the
115
-------
transport model to mix isotopic end members) to observed macroalgal isotope ratios at five
locations. There were several reasons why attached green macroalgae were used as the
bioindicator in this study. In estuaries with high flushing, macroalgae are often the dominant
primary producers (Valiela et al., 1997), and their presence is often used as an indicator of
eutrophication (Bricker et al., 2003). During the dry season, there are dense blooms of green
macroalgae in many estuaries in the region. In addition, the attached nature of the green
macroalgae is beneficial in that it integrates the nitrogen sources over time, which is
advantageous when there is substantial temporal variability in nitrogen sources and loadings.
Nitrogen stable isotope ratios of green macroalgae are a useful indicator because the macroalgae
take up DIN from the water column and incorporate this nitrogen into their tissue. Using the
natural abundance isotope ratio of nitrogen, macroalgae can be linked to nitrogen sources with
distinct isotopic signatures (McClelland and Valiela, 1998; Costanzo et al., 2001). Fixation of
atmospheric nitrogen (including that by N2-fixing root nodules associated with red alder, see
Section 3.7) has 515N values slightly less than 0%o. Industrially produced fertilizer has a similar
isotope ratio due to the atmospheric source of nitrogen in the Haber Process. Stable isotope
signatures are particularly useful when the various sources contributing to a system are
isotopically distinct (Figure 5-1). As an example, nitrogen stable isotopes may be used to
distinguish fertilizer inputs from animal or human waste, but they are not useful for
distinguishing human wastes (septic or wastewater treatment facility) inputs from animal wastes
due to the overlap of isotopic signatures. Distinguishing nitrogen sources with overlapping
isotopic signature (e.g., human waste from animal wastes) is often accomplished by comparison
of nitrogen sources and land use patterns. Previous research has shown that green macroalgae
integrate water column conditions that have occurred over a two week interval (Aguiar et al.,
2003; Cohen and Fong, 2005). An indicator that integrates water column conditions is desirable
due to the variability of coastal ocean nutrient conditions associated with variations in wind
stress (see Section 4.3) and variability in water column DIN within estuaries (see Section 4.4.2).
Since it was impractical to develop a transport model for each estuary due to data limitations, we
used a conservative transport two end member mixing model to estimate the contribution of
riverine and oceanic DIN sources, which is possible due to their different salinity and nitrogen
stable isotope signatures. Similarity between observed and predicted isotope ratios suggests that
the oceanic and riverine inputs are the primary nitrogen sources to the estuary. If the observed
isotope ratio is greater than that predicted based upon conservative mixing of oceanic and
riverine nitrogen sources, this suggests that there may be additional nitrogen sources or
denitrification may be important. If there is a wastewater treatment facility (WWTF) source in a
given estuary, then we inferred the contribution of WWTF sources using Isosource (Phillips and
Gregg, 2003), which calculates all feasible combinations of the three sources that can produce
the observed isotope ratio at a given location and sampling time. To constrain the solutions, we
selected a subset of feasible solutions, which were consistent with the salinity variations at the
site. This method was validated for Yaquina Estuary through comparison to the results from the
transport model. Once validated, this technique was used to estimate the contribution of DIN
sources for estuaries sampled in this classification effort. In addition, salinity data were analyzed
for Yaquina Estuary to determine if zonation predicted from the isotope and transport models
could be approximated using salinity data. For estuaries where there were limited data available
116
-------
we used salinity data (collected as part of the classification effort as well as historic salinity data)
to aid in the zonation of the estuaries.
Ocean
Animal waste
WWTF
Septic
Fertilizer
Alder
-5
0
10
615N
15
20
25
Figure 5-1. Natural abundance nitrogen isotope ratio (515N expressed as %o) for various
nitrogen sources. Sources: ocean represents the 515N of green macroalgae from the Oregon
coast (Fry et al., 2001, and Kaldy unpublished), animal waste and fertilizer is from Kendall and
McDonnell (1998), septic is from Cole et al. (2004), and red alder is from Hobbie et al. (2000),
Tjepkema et al. (2000), Cloern et al. (2002).
5.1 Stable Isotope Sampling
In the Yaquina and Alsea estuaries, 5 sites were sampled monthly for the isotope ratio of
macroalgal tissue. Sites were located along the salinity gradient from the mouth of the estuary to
the region where the system becomes mesohaline. As part of the classification effort, samples
were collected for the isotope ratio of green macroalgae at two sites in the Alsea, Nestucca,
Salmon River, and Yaquina estuaries during 2004. In 2005, green macroalgae were sampled at 5
sites in the Umpqua River and Coos estuaries and 4 sites in the Tillamook Estuary.
Each site was characterized by the presence of either rock or other hard substrate (e.g. pilings,
tree stumps, docks) to support green macroalgae. No attempt was made to identify algae to
species. The dominant macroalgae were species from the family Ulvaceae, which includes a
variety of genera. At each site, five replicate samples were collected by hand from the top or
sides of rocks and pilings. Only thallus material that did not contact mud sediments was
117
-------
collected. Individual samples were rinsed in the field with milli-Q water, transferred to labeled
plastic bags and stored on ice. In the laboratory, samples were rinsed with milli-Q water to
remove adhering sediments, frozen and lyophilized. Dried samples were then pulverized with a
mortar and pestle. Percent nitrogen and the nitrogen isotope ratio (515N) of dried tissue were
measured using a Finnigan Mat Delta Plus XP isotope ratio mass spectrometer. All isotope
ratios are presented in standard 5 notation.
5.2 Yaquina Estuary
5.2.1 Transport Model
A two-dimensional, laterally averaged hydrodynamic and water quality model (Cole and Wells,
2000) was used to simulate the transport of riverine, oceanic and WWTF effluent DIN sources.
The transport model incorporates the temporal variability in nitrogen loading, the location of the
nitrogen inputs, and the time that each nitrogen source spends in the estuary. In the model
simulations, Yaquina Estuary was represented by 325 longitudinal segments spaced
approximately 100 m apart with each longitudinal segment having 1-m vertical layers. The
model domain extended about 37 km from the tidal fresh portion of the estuary at Elk City,
Oregon to the mouth of the estuary (Figure 5-2). Model simulations were performed for 2003
and 2004 and included tidal and wind forcing, as well as freshwater inflow. Parameters
simulated included water surface elevation, salinity, water temperature, and dissolved inorganic
nitrogen. Each nitrogen source, riverine (NR), oceanic (No), and WWTF effluent (Nw), was
modeled as a separate component. The nitrogen sources were modeled as
dN
- = transport - juN (5.1)
dt
where TV is the DIN source and // is a loss/uptake rate. The same value of// was used for all three
nitrogen sources and the value of// was determined by fitting total modeled DIN (No+NR+Nw) to
observations of DIN within the estuary. The best fit to observations was found with // = 0. 1 d"1.
Simulations were also performed with no uptake (ju = 0) which is equivalent to conservative
transport of the sources.
The results from the transport model were used to mix the three nitrogen sources using the
following equation
fR+fw+fo=l (5-3)
where /R,/W, and/o are the fractions of riverine, WWTF, and oceanic DIN, respectively, and SR,
Sw, and 5o are the isotope ratio of end members for riverine, WWTF effluent, and oceanic
sources, respectively. The mixture of the three end members (SM) is assumed to be that which
macroalgae utilizes. Estimates of the oceanic and riverine end members were obtained by
examination of the observed isotope ratios at Stations Nl and N5. The riverine end member (SR
= 0 to +2 %o) was estimated from the wet season isotope ratios at Station N5, while the oceanic
end member (So = +8 to +9 %o) was estimated from dry season isotope ratios at Station Nl.
The initial estimate for the WWTF end member (Sw = +15 to +22 %o) was determined from the
literature (Jones et al., 2003). To arrive at the final end member isotope ratios, model
118
-------
Mini CTD
15
§ N macroalgae
Figure 5-2. Location map for Yaquina Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTF. The red dashed line shows boundary between
oceanic and riverine segments.
simulations were performed varying each end member over the range estimated from the data
and literature. The final isotope ratio of end members for the three sources (5R=+2 %o,
5w=+20 %o, and 5o=+8.4 %o) was determined from the best fit (minimum root mean square error,
RMSE) between predicted and observed isotope ratio at the five stations N1-N5 during 2003 and
2004. The final oceanic end member selected is consistent with marine end members for the
west coast of the U.S. (Fry et al., 2001), while the riverine end member is consistent with the
isotope ratio expected for nitrogen associated with red alder (leaf tissue ranges between -3 and -
0.5%0; Hobbie et al., 2000; Tjepkema et al., 2000; Cloern et al., 2002).
The 515N of green macroalgae collected from Yaquina Estuary (locations shown in Figure 5-2)
during 2003 and 2004 exhibited strong seasonal patterns (Figure 5-3) associated with shifting
nitrogen sources. During winter conditions with large freshwater inputs, there is a large degree
of isotopic separation (A = 4%o) between samples collected near the ocean (Nl) and near the
riverine end-member (Station N4 or N5). This isotopic difference suggests that algae from these
sites are utilizing different nitrogen sources. During the dry season, as river flow decreases and
ocean upwelling becomes more pronounced, the isotope ratio of green macroalgae collected
119
-------
from the end-member stations becomes isotopically indistinguishable, suggesting that
macroalgae are utilizing one nitrogen source. In some months during the dry season, the data
suggest that a third nitrogen source, the WWTF, can become locally important (Figure 5-3).
Data from 2004 also exhibit this same general pattern.
The model (Equations 5.1-5.3) reproduces much of the observed temporal and spatial variability
observed in the macroalgae isotope dataset. Comparisons between simulated and observed
isotope ratios at Stations N2 and N3 are presented in Figures 5-4 and 5-5. The RMSE between
observed and predicted isotope ratios calculated for Stations Nl - N5 for each year ranges from
0.8 %o to 2.0 %o (with an average value of 1.3 %o). This agreement between simulated and
observed isotope ratios suggests that the model was correctly simulating the transport and mixing
of the three nitrogen sources. Figure 5-6 shows the monthly average contribution of each of the
three nitrogen sources (fn,fw, and^b) at Stations N1-N5 and adjacent to the discharge point of
the WWTF (region of maximum contribution of WWTF source).
10-
f
CO
0
Month
Figure 5-3. Macroalgae isotope data at five stations (N1-N5) for Yaquina Estuary with red boxes
indicating dry seasons. The dashed line indicates 515N = 5.2 %o which is the demarcation
between riverine and oceanic dominance, assuming two end member mixing of oceanic and
riverine sources.
120
-------
There was interannual variability in the boundary between the oceanic and riverine segments.
The exact location of this boundary varies with ocean conditions (e.g., El Nino, La Nina
conditions) as well as freshwater inflow. Using the modeled contributions integrated over May -
September, we estimated the position of this boundary during 2003 and 2004. During 2003 the
freshwater inflow was less than in 2004 and as a result the oceanic DIN reached further upstream
with the boundary between the oceanic and riverine segments being located between Station N3
(which received 57% of DIN from ocean) and N4 (which received 44% of DIN from ocean).
During 2004, the boundary was located slightly upstream of Y3 (Y3 received 53% of DIN from
the ocean, while N3 received 32% from the ocean). In addition, during the dry season this
boundary progressed up estuary as the freshwater inflow declined. This can be seen in
simulation results from 2004 (Figure 5-6) which show that Station Nl was ocean dominated
(fraction > 0.5) during the entire dry season (May - September), while Station N2 was river
dominated during May and ocean dominated from June - September. At Stations N3, N4, and
N5 ocean inputs increased in importance from May - August, but didn't dominate. This was
also evident in that the salinity at up estuary sites increased from May - September. The results
of this analysis may appear to contradict the loading comparisons presented in Table 3-9, which
shows that ocean inputs dominate during the dry season. This difference is because the analysis
presented in Table 3-9 is comparing total loading to the estuary, while the analysis described in
this section examined contributions to nitrogen concentrations at specific locations, which
represent the nitrogen sources that would be available to primary producers. Based on
comparison of salinities (modeled and observed, Table 5-1) and modeled contribution of sources,
we concluded that Yaquina Estuary was ocean dominated in areas where the median salinity was
> 25 psu and river dominated in areas where the median salinity < 25 psu (calculated using
salinity data from May - September). We used these salinity criteria to aid in the zonation of
other estuaries with limited data.
Table 5-1. Summary of median observed salinities for Yaquina Estuary during May - September
of 2003 and 2004. NA denotes data not available and * denotes extensive gaps in record.
STATION
Yl
Y2
Y3
Y4
Y5
MEDIAN SALINITY (psu)
2003
31.9 (surface)
29.1 (bottom*)
30.7
27.3
25.5
18.0
2004
32.9 (bottom)
NA
27.3
NA
19.3
121
-------
10
9
to
6
5
^ RMS Error =1.50 %o
River = 2 %o
Ocean = 8.4 %o
WWTF = 20 %o
50
100 150
Julian Day
200
250
300
Figure 5-4. Comparison of modeled and observed isotope ratio at Station N2 in Yaquina Estuary
during 2004. The grey line is model predicted isotope rate and the symbols and error bars
represent mean and standard deviations of observed macroalgae isotope ratio.
12
10
to
4
2
River = 2 %o
Ocean = 8.4 %o
WWTF = 20 %o
50 100 150 200 250 300
Julian Day
Figure 5-5. Comparison of modeled and observed isotope ratio at Station N3 in Yaquina Estuary
during 2004. The grey line is model predicted isotope rate and the symbols and error bars
represent mean and standard deviations of observed macroalgae isotope ratio.
122
-------
N1
to
100
80
% 60
° 40
20
0
1
0.8
gO.6
i? 0.4
0.2
100
80
60
40 f
20
0
N2 N3
100
80
60
40
20
0
N5
fl
1
Jill
100
80
60
I
40
20
• n. .
i
I
•
1
.M.
-100:
80
60
40
20
1 n 1
•
•
•
i •
1.
100
80
60
40
20
• n .
1
h
1
1 H
In
[
24681012 24681012 24681012 24681012 24681012 24681012
Month
QJUU
0.8
o.e
0.4
0.2
Ouuuuu
1
0.8
0.6
0.4
0.2
b 0J
0.8
0.6
0.4
0.2
Ojuy
1
0.8
0.6
0.4
0.2
H
p 1
0.8
0.6
0.4
0.2
24681012 24681012 24681012 24681012 24681012 24681012
Month
Figure 5-6. Modeled contribution of each source of dissolved inorganic nitrogen in Yaquina Estuary during 2004. The upper panels
shows concentrations and the lower panels show fractions of each source with shading indicating riverine (green), oceanic (red), and
WWTF effluent (blue) nitrogen sources.
-------
5.2.2 Conservative Mixing Model
Since transport models are not available for the other target estuaries, we used a two-end member
conservative mixing model to estimate the dominant nitrogen sources for different portions of
each estuary. To test this analysis technique, we performed this analysis for Yaquina Estuary
and compared the results to estimates obtained from the transport model (described above).
Previous research has suggested that it is difficult to identify nitrogen sources due to the
temporal variability of salinity and water nutrients in estuaries (e.g., Cohen and Fong, 2005). In
this analysis, we incorporated temporal variability in nitrogen sources by utilizing time series of
salinity data at multiple stations within the estuary to mix nitrogen sources. If time-series data
were not available in the vicinity of the macroalgae sample collection, then salinity data from the
high and low tide cruises from water quality stations in the vicinity were utilized. Briefly,
nitrogen concentration (CM) was modeled as the mixture of riverine and marine end-members
using the following equations (Fry, 2002)
CM=fCR+(\-f)C0 (5.4)
where C denotes concentration of DIN, the subscripts R and O indicate river and ocean end-
members, respectively, and/represents the fraction of freshwater in each sample calculated from
salinity (time series or high and low tide values)
where So and SR are the salinity (psu) at the river and ocean end-members, respectively and SM is
the salinity measured at a specific station. The isotope ratio of mixed estuarine samples was
calculated as
(56)
The same end members were used as in Section 5.2.1. Observed isotope ratios were compared to
isotope ratios predicted from conservative mixing (Equations 5.4-5.6) using salinity data
collected in each region of the estuary. The mean of the predicted isotope ratio (SM) was
compared to the observations. Similarity between observed and predicted isotope ratios suggests
that the oceanic and riverine inputs are the primary nitrogen sources to the estuary. If the
observed isotope ratio is greater than that predicted based upon mixing of ocean and riverine
nitrogen sources, this suggests that there may be additional nitrogen sources or denitrification
may be important. If there is a WWTF source in a given estuary, then we inferred the
contribution of WWTF sources using Isosource (Phillips and Gregg, 2003) which calculates all
feasible combinations of the three sources that could produce the observed isotope ratio at a
given location and sampling time. To constrain the solutions, we selected the subset of feasible
solutions which were consistent with salinity variations at the site. For example, if the estimate
for the ocean contribution using the conservative mixing model was 50% averaged over the two
weeks prior to the macroalgae sampling, then we selected solutions with ocean contributions of
40-60%. Through this technique we can identify the segments that are dominated by oceanic
nitrogen sources and identify regions where other nutrient sources (riverine and wastewater
treatment effluent) dominate.
To test this technique of interpreting isotope data using temporal variations in salinity data
(Equations 5.4-5.6), we compared results from this analysis to the estimate of the contribution of
124
-------
sources from the transport model. For both estimates of the contribution of nitrogen sources we
integrated the results (transport model and conservative mixing model) over the 2 weeks prior to
the macroalgae sampling date (if salinity data were available). Due to gaps in the salinity time-
series or biofouling of the sensor, we were not able to compare all months at all sites. Presented
in Table 5-2 is the interval of salinity data used in the conservative mixing model. To maximize
the number of data points for comparison, if there were gaps in the salinity time series we used
salinity data within about 4 weeks of the sampling date. Salinity data from Stations Y2, Y4 and
Y5 were used to predict the isotope ratios at Stations N2, N3, and N4 (Figure 5-2). We assumed
that the salinity of the river end member was zero (Sn = 0), and used the maximum salinity at
Station Yl during the 2 weeks prior to the isotope sampling as the ocean salinity (So). Riverine
and oceanic end members were determined from samples collected at Yl and Elk City (tidal
fresh region of the estuary). Presented in Table 5-2 are the values of So, CR and Co used for each
month.
There was good agreement between the results calculated using the transport model and those
using the conservative mixing model and Isosource (Table 5-3). Using the nitrogen stable
isotope ratio of macroalgae combined with short time series of salinity, we feel we are able to
estimate the contribution of 3 DIN sources within ± 10%, which is acceptable for the needs of
this study.
We performed sensitivity analyses to determine the effect of integration time period of the
macroalgae and uptake rate using the results of the transport model (calculated for each monthly
sampling at Stations N2, N3, and N4). The maximum difference in percent contribution of
nitrogen sources calculated using a 2 week versus a 3 week integration period is 6% and the
average absolute difference is 1%. Cohen and Fong (2005) found that Enteromorpha intestinalis
carries its nutrient history for weeks, suggesting that a 2-3 week integration period is appropriate.
The largest differences in percent contribution calculated using 2 and 3 week integration
intervals occurred at Station N2, which was probably related to the greater temporal variability in
predicted isotope ratio at this site (Figure 5-4). Results calculated with no uptake (\i =0)
compared to those calculated with u^ = 0.1 d"1 had minimal differences during the wet season
months (January - April). This is because there is little utilization of water column DIN during
the wet season due to short residence time (associated with high freshwater inflow). The
maximum difference in the percent contribution of the sources calculated with the two uptake
rates occurred during May and June. Additionally, the riverine contribution is underestimated in
the lower portion of the estuary for calculations using u^ = 0.1 d"1 compared to conservative
transport. This is because the transit time for riverine DIN is maximal in the lower portion of the
estuary and there is more time for loss of water column DIN. During May to September the
maximal difference of riverine contribution was 14% and the average (of Stations N2, N3 and
N4) underestimate of the riverine contribution was 6% (with using n = 0.1 d"1 compared to
conservative transport). Using an uptake rate of 0.1 d"1 tends to overestimate the contribution of
WWTF in the upper estuary (maximal difference of 7%) and had little effect in the lower estuary
(0-1%). The average absolute difference during May to September in the WWTF contribution
was 2%.
As a confirmation of the zonation based upon median salinities, we used the conservative mixing
model (Equations 5.4-5.6) combined with mean dry season end members (€R and Co) to
125
-------
determine the salinity (Su) where the nitrogen sources would switch from ocean to river
dominance. During 2003, the mean dry season ocean end member DIN is 13.5 uM, while the
riverine end member is 26 uM. Using these values the boundary between the oceanic and
riverine segments occurred at a salinity of about 21 psu, which is similar to the value of 25 psu in
determined in Section 5.2.1. In addition, assuming two end member (ocean and river) mixing,
isotope ratios above 5.2%o indicate oceanic DIN dominance while isotope ratios below this value
indicate riverine DIN dominance. Therefore, the salinity demarking the zonation would range
from 21-25 psu.
There are several assumptions used in this analysis that need to be valid for us to apply this
technique to the seven target estuaries. This analysis assumes that macroalgae do not exhibit
fractionation (preferentially take up and incorporate 14N relative to 15N), hence the isotope ratio
of the macroalgae reflects the isotope ratio of the DIN in the water column. Cohen and Fong
(2005) showed that E. intestinalis did not fractionate during DIN uptake and assimilation. In
addition, it assumes that variability of the isotope ratio of the DIN results from a mixture of three
nitrogen sources (ocean, river and WWTF) and not from fractionation associated with other
biogeochemical transformations (e.g. denitrification). Another assumption is that the isotope end
members used in the mixing model are applicable at a regional scale (i.e., isotope end members
are the same for all seven estuaries). Stable isotope analysis of samples from across the region
suggests that this is a reasonable first approximation. Also, this analysis assumes that there is
minimal temporal variability in the isotope ratios of the end members. We also assume that the
freshwater inflow is primarily associated with riverine inputs and WWTF inputs don't represent
a significant source of freshwater. Based on riverine and WWTF flow rates, this was a good
assumption even during periods of minimal riverflow (Table 3-6).
126
-------
Table 5-2. Data used in the mixing model and time interval of salinity data utilized for each
month. NA indicates data are not available.
SAMPLING
DATE
April 29
May 29
June 26
July 31
Aug 14
Sept 26
TIME INTERVAL OF SALINITY DATA
UTILIZED IN MIXING MODEL
STATION
N2
April 15-29
May 1-15 &
30-31
June 12-26
July 17-31
NA
Sept 1 - 16
STATION N3
April 15-29
NA
June 20 - 28
NA
July 31 -Aug 14
Sept 4- 16
STATION
N4
April 15-29
May 15-29
June 12-26
July 17-24
NA
Sept 23 -30
END MEMBER
VALUES
So
(psu)
31.3
31.7
32.3
34.0
34.0
33.6
Co
(uM)
3.5
(n=33)
4.1
(n=30)
9.4
(n=30)
18.8
(n=31)
14.5
(n=29)
19.1
(n=31)
CR
(uM)
74.4
(n=3)
49.5
(n=5)
27.8
(n=6)
9.0
(n=4)
11.1
(n=6)
15.6
(n=2)
Table 5-3. Comparison of percent contribution of DIN sources estimated using transport model
and Isosource model. Value presented for the transport model represent averages, while
Isosource results are ranges of feasible solutions. * denotes observed isotope ratio below river
end member, therefore solution is not feasible.
MONTH
TRANSPORT MODEL
OCEAN
RIVER
WWTF
ISOSOURCE
OCEAN
RIVER
WWTF
Station N2
April
May
June
July
September
24.4
42.3
80.6
91.9
93.5
74.3
55.8
16.2
5.0
3.7
1.3
1.9
3.2
3.1
2.8
8-28
20-40
63-83
86-100
87-98
71-85
48-62
5-19
0-9
0-7
0-8
11-19
11-19
0-6
2-7
Station N3
April
June
August
September
8.5
47.2
86.8
75.1
89.8
43.7
6.2
14.0
1.7
9.1
7.0
10.8
0-1
32-52
74-94
76-83
99-100
35-49
4-18
0-4
0
12-20
1-9
17-21
Station N4
April
May
June
July
September
4.5
6.5
32.1
48.2
61.1
93.8
90.4
56.6
31.0
22.4
1.6
3.1
11.3
20.8
16.5
*
0-15
19-39
65
70-85
*
79-90
46-60
35
0-10
*
5-12
14-22
0
15-22
127
-------
5.3 Alsea Estuary
In addition to the data collected as part of this study, we also had other data sources available for
the Alsea Estuary. For Alsea Estuary, we had monthly 5 15N of green macroalgae at five
locations during 2004 extending from the ocean-dominated to the river-dominated portions of the
estuary (labeled N1-N5 in Figure 5-7). As part of this study, Stations Nl and N4 were sampled
for 5 15N of macroalgae on September 15, 2004. In addition, we had YSI datasondes deployed at
three locations in this estuary from June - October 2004. Water quality cruises with six stations
(S1-S6) were performed weekly from May through the end of September 2004 (Table 5-4).
Historic salinity data were also available at twelve stations from the Oregon Department of
Environmental Quality (with sample size at each station varying between 22 and 47).
CTD Cruise
YSI datasonde
Mini CTD
i^
6 N macroalgae
Minor Sewage Facilit>
Figure 5-7. Location map for Alsea Estuary showing the locations of datasondes, water quality
stations, isotope samples, and WWTF. The red dashed line shows the boundary between the
oceanic and riverine segments.
128
-------
From June to October, the green macroalgae 5 15N at Stations Nl and N2 showed similar
temporal patterns (Figure 5-8) and an isotope ratio which suggested that oceanic input dominated
at these two stations (average values of 7.9 and 7.5 %o at Stations Nl and N2, respectively). The
isotope ratio at Station N3 was slightly elevated above Nl and N2, suggesting that there may be
an additional nitrogen source. Station N4 had isotope ratios below 5 %o from January through
July suggesting that riverine nitrogen sources dominated. At Station N4 during August the 5 15N
of the macroalgae exceeded the ocean end member suggesting that there may be a third nitrogen
source or denitrification may be important during the low overflow period. At Station N5, the
isotope ratio suggested that riverine nitrogen sources dominated from March through July and
October, and there may have been an additional nitrogen source during August and September or
denitrification had increased in importance. Riverine inflow averaged over the interval of one
month prior to the isotope sampling was at its minimum during the months of August and
September. During periods of minimal riverine inflow, minor DIN sources (e.g., nonpoint inputs
associated with septic systems along the river) may increase in importance. In addition,
denitrification may increase in importance during this period of longer residence times (although
see Discussion, Section 5.9). There is a WWTF that discharges into the Alsea Estuary; however,
it is located downstream of Stations N3 - N5 (Figure 5-7). Nonpoint septic inputs may have been
the cause of increased macroalgal 5 15N since the Oregon Department of Environmental Quality
(DEQ) recently initiated a program to repair failing septic systems along the Alsea River.
10-
-2
Jan 04 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
Figure 5-8. Monthly average observed isotope ratio at 5 locations in Alsea Estuary (dry season
is indicated by red box). The dashed line indicates 515N = 5.2 %o which is the demarcation
between riverine and oceanic dominance, assuming two end member mixing of oceanic and
riverine sources.
129
-------
Alsea Estuary is located 20 km south of Yaquina Estuary. In the conservative mixing model, we
used data from Yaquina Estuary for the ocean end member (Co) DIN concentrations (sampling
from Station Yl). Coastal ocean boundary conditions respond to regional phenomena such as
upwelling and downwelling which encompass the entire Oregon coast. Consequently, ocean
boundary conditions from a well characterized system such as Yaquina are applicable over a
broad region. In Section 4.3, we demonstrated that ocean input (as indicated by flood-tide water
temperature) was relatively uniform over the geographic range of the estuaries that we sampled.
Riverine end member DIN (C«) data were obtained from water samples collected at Station S6
(Figure 5-7). End member DIN concentrations were averaged over the 2 weeks prior to
macroalgal sample data collection (Table 5-5). During September there were no DIN samples
from Station S6 during the 2 weeks prior to our sampling so we used data from an Oregon DEQ
Station (located about 20 miles from the mouth of the estuary) during the month of September
from 1994-2006. We used salinity from Station Yl (bottom) in the Yaquina Estuary averaged
over the 2 weeks prior to the sampling date as the ocean salinity (So). Salinity data from the
stations nearest to the macroalgae sampling locations were used in the analysis (Table 5-4) and
in the calculations we salinity measured during the 2 weeks prior to the macroalgae sampling.
The similarity between the observed nitrogen stable isotope ratios and those predicted from the
conservative mixing model at Stations Nl and N2 suggested that oceanic and riverine sources
were the dominant DIN sources (Table 5-6). There was more uncertainty in the results
calculated using the cruise data (Station Nl) than those using the salinity time series (Station
N2). At Station N2, the analysis suggested that this station receives 76-92% of DIN from the
ocean. Unfortunately, there were not time series salinity data in the vicinity of Station N3 and
this station is in a region of salinity changing from marine to mesohaline (compare the median
salinities at Stations Yl and Y2, Table 5-4). The limited salinity data prevented a thorough
analysis of the nitrogen stable isotope ratio at this location; however, the similarity in temporal
patterns in the 5 15N during the dry season at Station N3 compared to Stations Nl and N2 (Figure
5-8), suggests that the ocean was the dominant nutrient source at this location, but there may be
an additional nitrogen source as evident by 5 15N of Station N3 > 5 15N of Stations Nl and N2.
The macroalgal 5 15N at Stations N4 and N5 were consistent with the two end member mixing
model during the months of April through July. During August and September the observed 5
15N at Stations N4 and N5 were elevated above that predicted from the two end member mixing
model. During April and May, the mixing model suggested that Station N4 was river dominated
(88-97%). During June at Station N4, there was considerable variability in the observed isotope
ratio with two of the samples reflecting isotope ratios (2.8 %o) similar to that predicted using the
two end member mixing model and three having elevated isotope ratios (5.6-7.4 %o). Assuming
that the mixing model was more representative than the observed isotope ratios, Station N4 was
river dominated (80%) during June. During July, the two end member mixing model suggests
that Station N4 received 58% of the DIN from oceanic sources and 42% from riverine sources.
There was high variability in the isotope ratio predicted using the two end member mixing model
and the mean value was about l%o above the observed isotope ratio. Using Isosource the
maximum ocean contribution that could produce the observed isotope ratio during July at Station
N4 is 46%. Based on these two analyses, this station was probably located in the transition
region during July. Isosource suggested that during August Station N4 received 35-55% of DIN
from the ocean, 21-35% from river, and 23-31% from WWTF/septic inputs. During September,
130
-------
Station N4 became ocean dominated, with it receiving 50-70% of DIN from ocean, 28-42% from
river, and 1-10% from WWTF. This suggested that Station N4 is in a transition region and the
dominant nutrient source depends upon the month sampled (even within the dry season).
Similarity between the observed nitrogen stable isotope ratio and the two end member mixing
model during the months of April through July at Station N5 suggested that the river and the
ocean are the two major nitrogen sources. In addition, the observed isotope ratio suggests that
this station is river dominated (70-98% of DIN) during these months (particularly during April -
June). Station N5 is located between Stations Y2 and Y3, and there were considerable
differences in isotope ratio predicted using these two different stations, particularly during the
months of July - September. Results from Isosource suggested that during August Station N5
receives 57-71% from the river, 9-29% from ocean, and 13-21% from WWTF/septic. During
September, Station N5 received 51-65% of DIN from riverine sources, 16-36% from the ocean,
and 12-20% from WWTF/septic.
We used the conservative mixing model (Equations 5.4-5.6) combined with mean dry season end
members (€R and Co) to determine the salinity (Su) where the nitrogen sources would switch
from ocean to riverine dominance. During 2004, the mean dry season ocean end member (Co)
was 11.7 uM (using data from Station Yl in Yaquina with n = 214), while the riverine end
member (C«) was 15.4 uM (using data from Station S6 in Alsea with n =19). The riverine end
member was similar to dry season average of 15.7 uM for an Oregon DEQ Station which is
located at rivermile 20. Using these values the boundary between the oceanic and riverine
segments would occur at a salinity of about 20 psu, which is similar to the value of 21 psu
calculated for Yaquina Estuary. In addition, assuming two end members (ocean and river),
nitrogen isotope ratios above 5.2 %o indicate that the ocean is the dominant nitrogen source while
5 15N ratios below this indicate river is the dominant N source.
From the stable isotope data and the conservative mixing model, it is clear that Stations Nl and
N2 were ocean dominated and Station N5 was river dominated during the dry season and the
boundary between ocean and riverine dominance in nitrogen sources was located between
Stations N2 and N4. Using the salinity criterion developed in the above paragraph, median
salinities from Stations Y2 and S3, suggested that Station N4 was river dominated during the dry
season. We used historical salinity (median dry season) from Oregon DEQ at several stations
located between S2 and N3 combined with the salinity criterion developed above to determine
the boundary between the oceanic and riverine segments (Figure 5-7). We have a large degree of
confidence in these zonation patterns because they are based on multiple lines of evidence.
131
-------
Table 5-4. Summary of dry season salinities in Alsea Estuary.
STATION
MEDIAN
SALINITY (psu)
51HPERCENTILE
(psu)
951HPERCENTILE
(psu)
Datasondes (Deployed June 17 - October 12, 2004)
Yl
Y2
Y3
31.1
12.3
4.6
16.6
3.1
0.1
33.7
25.3
11.0
Mini-CTDs (Deployed September 14-October 6, 2004)
Ml
M2
NA
7.5
NA
0.0
NA
13.1
Water Quality Cruises (June - October 2004)
SI
S2
S3
S4
S5
S6
31.4
25.0
13.4
9.0
1.5
0
20.8
4.7
1.3
0.0
0.0
0.0
33.4
33.2
30.1
19.9
12.5
7.3
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C8
C9
CIO
LOW TIDE
SALINITY (psu)
25.4
20.0
12.4
8.8
6.6
4.3
2.9
2.5
1.5
0.2
HIGH TIDE
SALINITY (psu)
32.7
32.5
32.2
29.1
17.5
6.4
5.1
2.8
0.4
0
Table 5-5. Ocean and river end members for Alsea calculations.
MONTH
April
May
June
July
August
September
C0 (uM)
5.9(n = 4)
7.7(n=ll)
13.8(n = 25)
10.1 (n = 23)
6.2 (n = 6)
14.5 (n = 20)
CR (uM)
27.5 (n = 2)
29.2 (n = 3)
9.5 (n = 3)
4.1 (n = 3)
6.9 (n = 2)
9.5(n=10)
132
-------
Table 5-6. Observed and predicted isotope ratios (from two end member model) in Alsea
Estuary.
MONTH
OBSERVED
ISOTOPE RATIO
(%o)
SMIX
(%o)
STATION USED
Station Nl
April
May
June
July
August
5.4 ±0.9
6.2 ±0.5
7.5 ±0.2
8.3 ±0.4
7.9 ±0.2
5.0 ± 1.6
4.5 ± 1.3
7.9 ±0.2
7.9 ±0.4
7.4 ±0.8
SI
SI
SI
SI
SI
Station N2
June
July
August
September
6.8 ±0.5
7.9 ±0.2
7.9±0.1
8. 3 ±0.2
7.4 ± 1.5
7.9 ±0.6
7.5 ±0.9
7.7 ±0.7
Yl
Yl
Yl
Yl
Station N3
September
8.6±0.1
6.0 ±2.5
C4
Station N4
April
May
June
July
August
September
3.2 ±0.2
2.6 ±0.6
5.0 ±2.0
4.8 ±0.6
9.8 ±0.7
6.8± 1.3
2.2 ±0.3
2.7 ± 1.3
2.5 ±0.2
3.3 ±0.8
5.9± 1.1
5.7 ± 1.0
6.1 ± 1.5
4.6 ± 1.0
5. 8 ±0.9
S3
S3
S3
Y2
S3
Y2
S3
Y2
Y2
Station N5
April
May
June
July
August
September
2.3 ±0.2
2.1 ±0.4
2.8 ±2.5
4.4 ±0.8
6.3 ±0.2
6.6 ±0.4
2.2 ±0.2
2.1 ±0.2
2.3 ±0.4
3. 3 ±0.8
2.1±0.1
4.9± 1.2
5.9± 1.1
3.9 ±0.7
4.5 ± 1.0
4.6± 1.0
3.2 ±0.4
3.1 ±0.4
5. 8 ±0.9
3.7 ±0.7
S4
S4
S4
Y2
Y3
S4
Y2
Y3
S4
Y2
Y3
M2
Y2
Y3
133
-------
5.4 Salmon River Estuary
Data used to determine the zonation in Salmon River Estuary included salinity data from short-
term YSI and Mini-CTD deployments in 2004 (Table 5-7), macroalgae isotope data (Table 5-8)
collected as part of the classification effort at two sites (Nl & N7) and previous data collected at
five locations (N2 - N6) (DeWitt and Eldridge, PCEB, unpublished data). The locations of all
sampling stations are presented in Figure 5-9.
The macroalgae 5 15N did not show as strong an ocean signal in Salmon River Estuary as in the
Yaquina and Alsea estuaries (maximum in 5 15N near mouth of about 6%o at Salmon compared to
8 %o in Alsea and Yaquina). This is because the estuary switches from ocean- to river-dominated
over a tidal cycle with the salinities at the mouth changing from 33 to 10 psu (Figure 4-8). As a
result, there was not a strong oceanic signal in the nitrogen stable isotope data even at the mouth
(Table 5-8). Using the salinity data from the classification data collection effort and the
conservative mixing model (Equations 5.4-5.6), we examined the importance of oceanic and
riverine nitrogen sources within the estuary. The concentration of DIN in the riverine end
member was determined from Station CIO on the low tide cruise (Cn = 21.6 uM). This riverine
end member was consistent with other data collected by U.S. EPA (Ozretich, PCEB, unpublished
data), which had a mean DIN of 21.3 uM during July and August of 2001. The DIN
concentration for the ocean end member was determined from flood-tide nutrient samples
collected at Yaquina Estuary every flood tide during June-August 2004 (Co =13.4 uM).
Salmon River Estuary is located 45 km north of Yaquina Estuary, and in Section 4.3 we
demonstrated that ocean conditions are relatively uniform over the geographic extent of the
target estuaries. We used an average value for the ocean end member since we were using
isotope data from different years (2000 and 2004).
Based on the conservative mixing model, the isotope ratio predicted using the salinity data from
Yl, Ml and M2 averaged between 6.5 %o and 7.0 %o, which is similar to observed isotope ratios
at Stations N1-N6. Using the conservative mixing model and salinity data from Y2, the
predicted isotope ratio was 4.6 %o, which is similar to the observed isotope ratio at N7. The
agreement between predicted and observed isotope ratios suggested that oceanic and riverine
inputs are the two main sources of nutrients to this estuary. There was better agreement between
the predicted and observed isotope ratio when a short-term salinity time series was used rather
than data from the high and low tide cruises. The nitrogen stable isotope ratio predicted using
the average of the high and low tide cruises appeared to underpredict the macroalgal 5 15N. The
isotope mixing model suggested that Stations N1-N6 received between 70-88% of the nitrogen
from the ocean, while Station N7 received about 60% of the nitrogen from the river. Based on
the median salinity of 25.9 psu at Station M2 and the macroalgal 5 15N indicating ocean
dominance at Station N6, the boundary between the oceanic and riverine segments was located
slightly above Station N6. This is consistent with the boundary being located approximately at a
median dry season salinity of 25 psu.
134
-------
Figure 5-9. Location map for Salmon River estuary showing the locations of datasondes, water
quality stations and isotope samples. The red dashed line shows the boundary between oceanic
and riverine segments.
135
-------
Table 5-7. Summary of dry season salinities in Salmon River estuary.
STATION
MEDIAN 51H 95 1H
SALINITY PERCENTILE PERCENTILE
(psu) (psu) (psu)
Datasondes (Deployed July 13-20, 2004)
Yl
Y2
30.6 14 33.4
16.5 2.6 29.7
Mini-CTDs (Deployed July 9-29, 2004)
Ml
M2
M3
29.61 12.1 36.1
25.9 15.2 31.5
16.7 4.5 33.8
Evidence of biofouling after 7/18/2004. Median calculated using data
from July 9- 18.
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C8
C9
CIO
LOW TIDE
SALINITY (psu)
22.7
14.8
13.3
10.1
7.9
4.4
1.8
0.4
HIGH TIDE
SALINITY (psu)
30.0
30.1
29.9
29.5
29.1
26.7
26.8
4.0
1.8
0.5
Table 5-8. Observed and predicted isotope ratios (from two end member model) in Salmon
River estuary.
STATION
Nl
N2
N3
N4
N5
N6
N7
SAMPLING
DATE
7/14/2004
8/23/2000
8/23/2000
8/23/2000
8/23/2000
8/23/2000
7/14/2004
OBSERVED
ISOTOPE RATIO
mean ± standard
deviation (%o)
5.9 ±0.3
6.6 ±0.4
5.7 ±0.4
6.7 ±0.3
6.9 ±0.9
6.5 ±0.4
3.6 ±0.5
SMK
mean ±
standard
deviation (%o)
6.5 ± 1.3
5. 8 ±2.4
7.0 ± 1.5
7.0 ± 1.5
5. 8 ±2.4
6.8±2.1
5.6 ±2.5
5. 3 ±2.8
6.5 ±2.4
5.1 ±2.9
4.6 ± 1.8
STATION
USED FOR
MIXING
MODEL
Cl
C2
Yl
Yl
C2
Ml
C3
C4
M2
C5
Y2
136
-------
5.5 Umpqua River Estuary
Data used to determine the zonation in Umpqua River Estuary included salinity (Table 5-9) and
macroalgae isotope ratios (515N) collected at five sites as part this study in 2005 (Table 5-10).
Unfortunately, the GPS had an error during the sampling of N2 and as a result we do not know
the exact sampling location, only that the sample was collected between Stations Nl and N3.
For the salinity data, we had short-term datasondes and Mini-CTD data collected in 2005 and
historic salinity available from Oregon DEQ.
Using the salinity data from the classification data collection effort and a conservative mixing
model (Equations 5.4-5.6), we examined the importance of oceanic and riverine nitrogen sources
within the estuary. The DIN concentration of the riverine end member was determined from the
Station Cl 1 low tide cruise (€R = 15.8 uM). The DIN concentration for the ocean end member
was the mean of the high tide cruise at Stations Cl - C3 (Co = 13.8 uM), which was similar to
dry season mean values of DIN from Yaquina Estuary for 2002-2004 (average = 14 uM).
Comparison of predicted and observed isotope ratios suggested that Stations N1-N3 received a
mixture of ocean and riverine nitrogen (particularly using Station Yl) with the ocean
contributing between 80-90% of the nitrogen at these sites. It was not possible to distinguish the
effect of the WWTF located near the entrance to Umpqua Estuary (Figure 5-10) due to the
variability in predicted nitrogen stable isotope ratios associated with salinity variations at Station
Yl. The disagreement between observed and predicted macroalgal 5 15N at Station N5 suggested
that there may be a third nitrogen source in this estuary. The City of Reedsport has a WWTF in
the vicinity of Station N5 (distance from WWTF to Station N5 is about 2 km, Figure 5-10).
Using Isosource, we calculated all of the possible combinations of the three nitrogen sources
which could produce the observed isotope value of 5.4 %o at Station N5. Using the observed
salinity at this site, we rejected the results from Isosource that had ocean contributions greater
than 10% based on salinity data and the two end member mixing model. Based on this analysis,
we estimated that the nitrogen sources at Station N5 were about 74-82% riverine and 15-20%
associated with WWTF effluent. The observed macroalgal 5 15N at Station N4 was slightly
elevated above the predicted isotope ratio from two end member model; however, the variability
in the predicted ratio made it difficult to determine whether this elevation was associated with an
additional nitrogen source or variability in the isotope ratios associated with the salinity varying
from 2 to 32 psu over a tidal cycle. Based on the nitrogen isotope data, the transition from ocean
to riverine dominance occurred between Stations N3 and N4. This was consistent with the data
from the high tide cruise, which have high salinity (> 30 psu) water reaching Station C4 and a
dramatic decrease in salinity above this station (decrease of about 17 psu between Stations C4
and C5). In addition, the high tide cruise showed the input of oceanic nutrients (NOs'+NCV and
PO43") to Stations Cl - C4. This was evident in that NO3"+NO2" decreases from 10 uM to 1.5 uM
at between Stations C4 and C5.
Using the oceanic and riverine DIN end members (€R =15.8 uM and Co = 13.8 uM) and the two
end member conservative mixing model, the boundary between the oceanic and riverine
segments occurred at a salinity of about 18 psu. This agrees fairly well with the salinity
(including historical salinity from Oregon DEQ) and macroalgae isotopes. The Oregon DEQ
database has a salinity station by Station C3 that has a median dry season salinity of 19 psu,
while a salinity station by Station C4 has a median dry season salinity of 16 psu.
137
-------
M3
012
10
5 km
O CTD Cruise
d YSI datasonde
A Mini CTD
• 5 ?rs macroalgae
• Minor Sewage Facility
Figure 5-10. Location map of Umpqua River Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTFs. The red dashed line shows the boundary
between the oceanic and riverine segments.
138
-------
Table 5-9. Summary of dry season salinities in Umpqua River Estuary.
STATION
MEDIAN SALINITY
(psu)
5 1H PERCENTILE 95 1H PERCENTILE
(psu) (psu)
Datasondes (Deployed June 21 - 26, 2005)
Yl
Y2
Y3
Y4
30.7
6.2
1.1
0.1
19.8 33.7
0.7 21.3
0.2 4.9
0.0 0.5
Mini-CTDs (Deployed June 20 - 26, 2005)
Ml
M2
M3
14.8
0.0
0.0
3.7 28.2
0.0 4.7
0.0 0.0
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C8
C9
CIO
Cll
C12
C13
LOW TIDE
SALINITY (psu)
23.4
15.3
9.8
5.3
1.9
1.8
1.3
0.1
0.8
0.2
0.0
0.0
0.0
HIGH TIDE
SALINITY (psu)
32.7
32.7
32.3
30.5
13.7
4.21
1.46
5.6
6.7
2.6
0.6
0.9
0.1
Table 5-10. Observed and predicted isotope ratios (from two end member model) in Umpqua
River Estuary.
STATION
Nl
N2
N3
N4
N5
SAMPLING
DATE
6/24/2005
6/23/2005
6/22/2005
6/22/2005
6/26/2005
OBSERVED
ISOTOPE RATIO
mean ± standard
deviation
7.8 ±0.5
7.3 ±0.2
7.2 ±0.2
5.7 ±0.1
5.4 ±0.2
SMIX
mean ± standard
deviation
(%o)
7.6 ±0.9
7.3 ± 1.4
6.5 ±2.5
5.9 ±3.2
4.9 ± 1.6
2.2 ±0.3
2. 5 ±0.7
STATION
USED FOR
MIXING
MODEL
Yl
Cl
C2
C3
Ml
M2
C8
139
-------
5.6 Coos Estuary
Coos Estuary had the highest isotope ratios of the seven estuaries that we sampled with three
stations (N3-N5) having 515N values of about 10 %o (Table 5-11). This elevation of the isotope
ratio above the marine end member may be associated with the presence of multiple WWTFs in
the watershed. There are 3 major WWTFs adjacent to the estuary (Figure 5-11), each of which is
permitted to discharge between 2 to 5 million gallons per day of treated effluent into the estuary
(http://www.deq.state.or.us/wq/sisdata/facilitycriteria.asp). This is the only estuary that has
multiple major WWTFs adjacent to the estuary.
Using the salinity data from the classification data collection effort (Table 5-12) and the
conservative mixing model (Equations 5.4-5.6), we examined the importance of oceanic and
riverine nitrogen sources within the estuary. Since our most riverine station (C19) had a salinity
of 13 psu, we could not use our data for the river end member. We used mean dry season DIN
(CR =13.7 uM, n = 28) from the Oregon DEQ database for a station at the Anson Roger Bridge
on the Coos River (5 km upriver of Station C19). This value is similar to the concentration
during the low tide cruise at Station C19 (10.3 uM). For the ocean end member, we used the
mean high tide DIN (Co = 21.3 uM) from Stations Cl, C4 and C5. This is probably a good
estimate of the ocean end member since there were strong upwelling conditions for about 3
weeks prior to our data collection effort (based on water temperature, see Figure 4-2).
Unfortunately, the Mini-CTD near the mouth of Coos Estuary (Ml) malfunctioned. Using the
datasondes closest to the mouth of the estuary (Yl and Y7) and the two end member
conservative mixing model using ocean and river end members, we predict that the isotope ratio
near the estuary mouth was 8.2-8.3 %o. Unfortunately, we did not have macroalgae isotope ratio
data at the estuary mouth, but Stations Nl and N2 are fairly close to the mouth and had similar
isotope ratios. Assuming two end member mixing of riverine and ocean sources, the observed
isotope ratios at Stations Nl and N2 suggest that the ocean provides 78 to 87% of the nitrogen at
these sites. Fry et al. (2001) found that the nitrogen stable isotope ratio of green macroalgae
during the dry season outside the mouth of Coos Estuary was 8.3 - 9.0 %o (sampled during
October 1998 and July 1999) and inside the mouth of Coos Estuary the isotope ratio was 7.7 -
7.8 %o sampled during the same months, which is similar to our data.
The observed isotope ratios at Stations N3-N5 exceed those expected from oceanic nitrogen
sources suggesting there is an additional nitrogen source. Isosource was used to calculate all
possible combinations of ocean, riverine, and WWTF nitrogen sources that could produce the
observed isotope ratios at Stations N3-N5. Since the short-term salinity time series provided
better estimates of the isotope mixing than the average of low and high tide cruises, we used the
two end member mixing model with data from Y2, Y3, and M3 to estimate the contribution of
the ocean. The two end member mixing model of ocean and river sources using Station Y2 and
Y3 suggested that the ocean provided about 97% of the DIN at Station N3 and N4. Using
Isosource and selecting only those solutions with an ocean contribution > 87%, the contribution
of nitrogen sources were estimated to be 87-91% oceanic, 9-13% WWTF and < 2% riverine.
The two end member mixing model of ocean and river sources using Station M3 suggested that
the ocean provided about 65% of the DIN at Station N5. Using Isosource and selecting only
those solutions with ocean contribution within ± 10% of that predicted from the two end member
mixing model, the contribution of nitrogen sources at Station N5 were estimated to be 55-75%
oceanic, 17-26% WWTF and 7-21% riverine. This may be an underestimate the contribution of
140
-------
the ocean and an overestimate of the contribution of the WWTF since M3 is located slightly
upriver of N5. Predicted isotope ratios calculated using Stations CIO and C16 were higher than
those using M3, and yield oceanic contributions of about 90%.
Based on this analysis, all of the macroalgal 5 15N sampling locations (N1-N5) were dominated
by oceanic nitrogen sources, and the elevation of the isotope ratio above ocean end member at
Stations N3 and N5 is due to the nitrogen sources being a mixture of oceanic and WWTF
sources. We used a median salinity of 25 psu (as a conservative estimate) as the boundary
between the oceanic and riverine segments. We used historical salinity from Oregon DEQ and
South Slough Estuarine Reserve to aid in the zonation.
Table 5-11. Observed and predicted isotope ratios (from two end member model) in Coos
Estuary.
STATION
Nl
N2
N3
N4
N5
SAMPLING
DATE
8/7/2005
8/7/2005
8/5/2005
8/6/2005
8/4/2005
OBSERVED
mean ± standard
deviation
(%o)
7.6±1.8
7.0 ±1.9
9.6 ±0.3
10.0 ±0.4
10.0±0.1
SMK
mean ± standard
deviation
8.4 ±0.0
8.2±0.1
8.1±0.1
8. 3 ±0.1
8.2±0.1
8. 3 ±0.1
8.2±0.1
8.2±0.1
8.2±0.1
8.1 ±0.2
6.1 ±0.9
7.9±0.1
7.8±0.1
STATION
USED FOR
MIXING
MODEL
C2
Y5
Y6
Yl
Y7
C4
Y2
C7
Y3
C6
M3
C16
CIO
141
-------
^ 4 o Oceanic
CTD Cruise
YSI dntiisonde
Miui CTD
15
6 N macroalgae
Major Sewage Facility
Figure 5-11. Location map of stations in Coos Estuary showing the locations of datasondes,
water quality stations, isotope samples, and WWTFs. The red dashed line shows the boundary
between the oceanic and riverine segments.
142
-------
Table 5-12. Summary of dry season salinities in Coos Estuary.
STATION
MEDIAN
(psu)
5mPERCENTILE
(psu)
95mPERCENTILE
(psu)
Datasondes (Y1-Y4 Deployed August 3 - 10, 2005; Y5 And Y6 Deployed July 26 -
August 24, 2006; Y7 Deployed July 6 - 20, 2006).
Yl
Y2
Y3
Y4
Y5
Y6
Y7
33.5
32.5
32.4
28.9
32.5
32.1
32.4
31.9
31.7
30.8
27.0
31.4
29.4
30.4
34.0
33.5
33.9
30.0
33.1
33.0
33.4
Mini-CTDs (Deployed August 3 - 10, 2006)
Ml
M2
M3
NA
27.8
20.6
NA
24.9
8.6
NA
30.8
24.5
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C9
CIO
Cll
C12
C14
CIS
C16
C17
CIS
C19
LOW TIDE SALINITY (psu)
33.1
NA
NA
32.5
31.7
30.4
32.0
29.8
29.0
27.9
26.3
31.9
31.4
29.4
20.6
23.2
12.5
HIGH TIDE SALINITY (psu)
33.8
34.2
33.6
33.9
33.8
33.1
33.1
32.3
30.6
29.9
28.3
33.9
NA
30.9
20.7
28.3
18.5
143
-------
5.7 Tillamook Estuary
Data used to determine the zonation in Tillamook Estuary included salinity data from short-term
YSI and Mini-CTD deployments in 2005 (Table 5-13) and macroalgae isotope data (Table 5-14)
collected as part of this study at four sites (N1-N4, Figure 5-12). Historic salinity data were also
available at 23 stations from the Oregon DEQ (see Figure 5-13;
http://deql2.deq.state.or.us/lasar2/). Using the salinity data from the classification data
collection effort (Table 5-13) and the conservative mixing model (Equations 5.4-5.6), we
examined the importance of oceanic and riverine nitrogen sources within the estuary. For the
ocean end member in the conservative mixing model, we used the high tide DIN (Co = 24.1 uM)
from Station Cl. This is probably a good estimate of the ocean end member since there were
strong upwelling conditions (based on water temperature in Figure 4-2) for about 1 week prior to
the macroalgae isotope sampling. Since there are five rivers discharging into Tillamook Estuary,
we used different river end members depending upon the macroalgae sampling location.
Stations Nl and N2 are located in the vicinity of the Miami River. For Stations Nl and N2, we
used the mean DIN concentrations from Station C3 (€R = 52.4 uM), which is in the Miami
River. Station C3 had similar high and low tide DIN concentrations and a salinity of <1 psu.
Colbert (2004) sampled the riverine end members of Tillamook in 1998 and 1999; their mean
DIN for the Miami River during May-August was 51.4 uM, which was comparable to our river
end member from Station C3. The Tillamook, Trask and Wilson rivers discharge into the
southern portion of Tillamook Estuary. For Station N4, we used the mean of the low tide cruise
for Stations C13-C15 (€R = 39 uM), which had salinities ranging from 0.1-1.3 psu. Colbert
(2004) sampled the Tillamook, Trask and Wilson rivers in 1998 and 1999 and found similar dry
season end members (mean = 32.2 uM). Since Station N3 is located in the central portion of the
estuary, we calculated the isotope ratios using the two different values for the river end member
(CR = 39 and 52.4 uM) and present results of both in Table 5-14. Interestingly, Tillamook
Estuary had the highest values of riverine end members of the seven target estuaries that we
sampled.
The observed isotope ratio at Station Nl was about l%o less than the isotope ratio predicted using
the two end member mixing model of ocean and river sources (using salinity data). Based on
this analysis, the contribution of oceanic and riverine nitrogen sources was 85% and 15%,
respectively. Using Isosource to mix the three end members, we found that the maximum ocean
contribution that could produce the observed isotope ratio was 67% (with the remainder of the
nitrogen being riverine). Based on both of these analyses, Station Nl was ocean dominated.
In the vicinity of Station N2 there was high variability in the salinity which is evident in the large
difference between salinities at Stations Yl and Y2 and C2 and C3. Stations Yl and C2 are
marine, while Stations Y2 and C3 are fresh. Since we did not have salinity observations at
Station N2, and considering the high variability in the nearby stations, the two end member
mixing model did not produce useful results. Isosource was used to calculate all possible
combinations of ocean, riverine, and WWTF nitrogen sources that could produce the observed
isotope ratio. There were limited results that could produce the observed isotope ratio, so the
salinity constraint was not needed. Results from Isosource suggested that the primary nitrogen
source was riverine (80-94%) with oceanic and WWTF being minor components (10-20% and 3-
144
-------
7%, respectively). Based on this analysis, riverine inputs were the dominant nitrogen source at
Station N2.
Unfortunately, we did not have acceptable salinity time series data in the vicinity of Station N3.
For the two end member mixing model, we used low and high tide cruise data from Stations C5
and C6. The isotope ratios predicted using the two end member mixing model were similar to
those observed, particularly for Station C6. Based on the two end member mixing model for
Station N3, the ocean contribution ranged from 65 to 82% and the riverine contribution ranged
from 18 to 34%, depending upon which station and river end member was used in the analysis.
Using Isosource and only selecting those results that had oceanic contributions >55%, the
oceanic contribution ranged from 55 to 68%, riverine ranged from 32 to 42% and WWTF was
less than 4%. Both of these analyses suggested that Station N3 was ocean dominated.
At Station N4 the observed isotope ratio was 1.7-2.6 %o higher than the isotope ratio predicted
using the two-end member mixing model (depending upon which station is used for the salinity)
suggesting that there may have been an additional nitrogen source. Based on a two end member
mixing model, the oceanic contribution ranged from 9 to 23%, while the riverine contribution
ranged from 76 to 91% (depending upon which station was used). There are two WWTF
facilities (one major and one minor) within 4 km of Station N4 (Figure 5-12). Using Isosource
and selecting only those solutions with the ocean contribution within ±10% of that predicted
from the two end member mixing model, we estimated that the riverine contribution ranged from
60 to 83%, oceanic contributions were less than 30%, and WWTF contributed between 5 and
18%. Therefore, based on these analyses Station N4 was river dominated.
Due to the uncertainty in the analysis of the isotope data due to multiple riverine end members,
multiple WWTF inputs, and non-ideal availability of salinity data for the mixing model relative
to macroalgae sampling locations, we used recent and historic salinity data in the derivation of
the zonation. Using the values of the riverine and oceanic end members from Tillamook Estuary
(CR = 39 -52 uM and C0 = 24.1 uM) and the conservative mixing model, the boundary between
oceanic and riverine dominance occurred at a salinity of about 20-23 psu. The value of the ocean
end member was relatively high compared to averaged dry season values for Yaquina Estuary.
Using C0 = 13.4 uM (which is the June to August average for 2004 at Yaquina Estuary), the
boundary would be located at about 25 psu. To be conservative in our estimates, we used a
salinity criterion of 23-25 psu to divide the estuary into oceanic and riverine segments (Figure 5-
13). The boundary between the oceanic and riverine segments in the northern portion of
Tillamook Estuary was based upon the analysis of the isotope data at Station N3.
145
-------
Y2n
N2 t3C3
Riverine
C CTD Cruise
G YSI datasoude
A Mini C TD
O 8 N macroalgae
• Major Sewage Facility
• Minor Sewage Facility
CIS
o
4
6
Figure 5-12. Location map for Tillamook Estuary showing the location of datasondes, water
quality stations, isotope samples, WWTFs. The red dashed line shows the boundary between the
oceanic and riverine segments.
146
-------
Table 5-13. Summary of dry season salinities in Tillamook Estuary (NA denotes missing or bad
data).
STATION
MEDIAN
(psu)
51HPERCENTILE
(psu)
Datasondes (Deployed July 19-25, 2005)
Yl
Y2
Y3
Y4
33.9
2.4
29.6*
7.8
25.0
0.0
10.6*
0.5
951HPERCENTILE
(psu)
34.9
27.2
34.1*
23.7
* Evidence of biofouling, summary statistics computed using 2 day record.
Mini-CTDs (Deployed July 19-25, 2005)
Ml
M2
M3
2.4
11.9
17.2
2.2
2.6
2.5
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C8
C9
CIO
Cll
C12
C13
C14
CIS
LOW TIDE SALINITY (psu)
29.7
27.3
0.1
11.6
25.2
19.4
NA
NA
11.0
NA
0.0
0.2
0.1
1.3
0.1
12.7
26.3
25.1
HIGH TIDE SALINITY (psu)
33.4
33.2
0.1
9.1
33.1
32.1
16.5
7.4
27.7
21.8
0.0
0.7
1.6
4.0
0.1
147
-------
Table 5-14. Observed and predicted isotope ratios (from two end member model) in Tillamook
Estuary.
STATION
Nl
N2
N3
N4
SAMPLING
DATE
7/20/2005
7/21/2005
7/22/2005
7/23/2005
OBSERVED
mean ± standard
deviation
(%o)
6.1 ±0.9
3.2 ±0.5
6.2 ±0.6
5.2 ±0.4
SMK
mean ± standard
deviation
(%o)
7.5 ± 1.0
7.7 ± 1.0
7.3 ± 1.4
7.3 ± 1.4
2.0
7.5 ± 1.0
2.6 ± 1.4
7.0-7.2
6.2-6.5
3.5± 1.3
2.6 ±0.5
STATION
USED FOR
MIXING
MODEL
Yl
Cl
C2
C2
C3
Yl
Y2
C5
C6
Y4
Ml
148
-------
30.0%o (99)
28.5%» (30)
26.0%o (117) \
Q ° \
29.0%o(198) \
Riverine
O 28.8%o (187)
27.9%o (26)
© DEQ Salinity Stations
0 0.5 1
\km
Figure 5-13. Median dry season salinities in Tillamook Estuary (with number of samples
indicated in parentheses). The red dashed line shows the boundary between the oceanic and
riverine segments.
149
-------
5.8 Nestucca Estuary
Data used to determine the zonation in the Nestucca Estuary included salinity data from short-
term YSI and Mini-CTD deployments in 2004 (Table 5-15) and macroalgae isotope data
(Table 5-16) collected as part of the classification effort at two sites (Nl & N2). During our
2004 field work, we lost two instruments in the lower portion of the estuary. During September
2006, we collected additional data in this estuary for use in confirming zonation. A YSI
datasonde was deployed near the mouth (Y2) and Mini-CTDs were deployed at 3 locations (M3-
M5).
As discussed in Section 4.3, the Nestucca Estuary was sampled during low ocean nutrient
conditions, which is reflected by DIN concentration of about 2 uM at Stations C8-C10 during the
high tide cruise. Flood-tide water temperature from Yaquina during the 2 weeks prior to the
isotope sampling indicated that the coastal ocean conditions were variable switching between
upwelling and downwelling conditions (Figure 4-1). There were limited oceanic end member
DIN data from Yaquina Estuary during the two weeks prior to the collection date (8/19/04) of the
isotope sampling in Nestucca; the mean ocean end member DIN concentration from Yaquina
during August 5th - 7th was 5.7 uM (n=6), and averaged over one month prior to the isotope
sampling was 15.7 uM (n = 38). Due to this variability in ocean end member DIN
concentrations and limited end member data available for the 2 weeks prior to sampling, we
present results using C0 values of both 5.7 and 15.7 uM (Table 5-16). Since the river end
member at Station Cl was fairly constant between the low and high tide cruises, we used the
mean of the low and high tide cruises as the riverine end member (CR = 38.4 uM). We did not
have short-term time series salinity data in the vicinity of Station Nl during 2004 due to the loss
of instruments. We used salinity data from Station C9 for the isotope mixing model and as an
approximation, we also used time-series salinity data from Station Y2, which was collected in a
similar time period (month) as the macroalgae samples from Nl but two years later. Using
salinity data from Y2, the observed isotope ratio at Nl was similar to that predicted using the
two-end member conservative transport mixing model. Based upon isotope ratio, Station Nl
received about 70-90% of nitrogen from oceanic sources.
The observed isotope ratio at Station N2 was substantially higher than that predicted from the
two-end member conservative mixing model (regardless of which value of C0 was used in the
calculation), suggesting that there was an additional nitrogen source in the system or that
denitrification was important (although see discussion Section 5.9). There is a WWTF located in
the vicinity of Station N2, which discharges into the Nestucca River about 1.5 miles upstream of
its confluence with Nestucca Estuary. In addition, there are two other minor WWTFs that
discharge into the Nestucca River about 7-10 miles upstream of the mouth of the estuary. The
presence of these three facilities suggests that WWTF input may be responsible for the observed
deviation from the two-end member mixing model. Using Isosource constrained by the salinity
data, the ocean contribution ranged between 8-13%, the riverine contribution ranged between 70-
80%, and the WWTF contribution ranged between 10-20% (depending upon which end member
is used).
Due to limited data availability, the southern boundary between the oceanic and riverine
segments was based upon the salinity criterion (using data from Station M2). The northern
150
-------
boundary was based upon data from the high tide classification water quality cruises (salinity and
dissolved inorganic nutrient data). Data from the high tide cruises suggested that ocean water (as
indicated by high salinity and nutrient levels comparable to Station CIO) reached Station C5 and
there was a dramatic decrease in salinity between Stations C5 and C4 (a difference of about 19
psu). In addition, the high tide cruise NCV+NCV concentrations were low (< 2 uM) between
Stations CIO and C5, reflecting the low ocean nutrient conditions during this cruise. There was
an increase in MV+MV (of about 13 uM) between Stations C5 and C4 reflecting an increase in
importance of riverine (or WWTF) nitrogen sources.
Using the conservative mixing model and the ocean and river end members for DIN (€R =38 uM
and Co = 15.7 uM), the boundary between ocean and river dominance occurred at a salinity of
24 psu. Based on median salinity, the estuary should be river dominated in the vicinity of
Station Ml (median salinity of 12 psu). The salinity data collected during 2006 from Stations
M3 and M4 confirmed that they were river dominated and that Station M5 was ocean dominated.
There is some uncertainty in the zonation of this estuary (in particular the northern line) due to
limited data.
151
-------
CTD Cruise
YSI datasonde
Mini CTD
15
N macroalsae
Minor Sewage Facility
Riverine
0 0,25 0.5
Figure 5-14. Location map of Nestucca Estuary showing the locations of datasondes, water
quality stations, isotope samples, and WWTF. The red dashed line shows the boundary between
the oceanic and riverine segments.
152
-------
Table 5-15. Summary of dry season salinities in Nestucca Estuary.
STATION
MEDIAN
(psu)
5mPERCENTILE
(psu)
95mPERCENTILE
(psu)
Datasondes (Yl Deployed July 28 - August 19, 2004, And Y2 Deployed From August 29 -
September 1, 2006)
Yl
Y2
6.4
33.2
1.0
22.4
18.1
33.5
Mini-CTDs (Ml And M2 Deployed July 29 - August 19, 2004; M3 Deployed August 29-
September 1, 2006, M4 Deployed September 1-6, 2006, M5 Deployed August 29 -
September 6, 2006.)
Ml
M2
M3
M4
M5
11.6
23.0
8.1
19.1
28.5
0.0
14.6
3.1
1.0
17.6
32.6
31.8
28.3
31.9
32.3
Classification Low And High Tide Cruises
STATION
Cl
C2
C3
C4
C5
C6
C7
C8
C9
CIO
Cll
C12
C13
C14
C15
LOW TIDE SALINITY (psu)
1.6
3.3
5.2
6.8
7.2
11.0
16.5
18.1
21.3
23.9
19.2
17.7
17.7
19.9
19.8
HIGH TIDE SALINITY (psu)
3.8
5.1
9.0
12.2
31.0
30.5
32.2
32.7
32.7
32.7
31.6
30.7
27.5
20.5
15.2
Table 5-16. Observed and predicted isotope ratios (from two end member model) in Nestucca
Estuary.
STATION
Nl
N2
SAMPLING
DATE
8/19/2004
8/19/2004
OBSERVED
mean ± standard
deviation
6.6 ±0.1
5.3 ± 1.4
SMK
calculated using
C0=5.7uM
3.6 ±0.4
7.0 ± 1.7
2.3 ±0.3
2.1 ±0.1
SMK
calculated using
C0= 15.7 uM
5.1 ±0.5
7.6 ±1.2
2.8 ±0.7
2.2 ±0.1
STATION
USED FOR
MIXING
MODEL
C9
Y2
Yl
Cl
153
-------
5.9 Synthesis and Uncertainties in Zonation
Utilizing 515N of green macroalgae combined with short-time series of salinity, we divided the
target estuaries into oceanic and riverine segments. In addition, using Isosource constrained by
the salinity data we estimated the contribution of WWTF inputs. Based on our analysis, oceanic
and riverine inputs were the dominant nitrogen sources within the target estuaries. The
maximum contribution of WWTF estimated from stable isotope data was 30%. We feel that our
analysis is robust since it uses multiple lines of independently gathered evidence which lead to
similar conclusions. The analysis utilizing nitrogen stable isotopes as a tracer of oceanic and
riverine nitrogen inputs appears to be consistent with zonation based upon median salinities.
Depending upon the DIN concentration in the riverine end member, the zonation occurred at
median salinities ranging from 18 to 25 psu. Shirzad et al. (1988) divided Oregon estuaries into
three zones (seawater (> 25 psu), mixing (0.5-25 psu), and tidal fresh (0-0.5 psu) using average
annual and depth-averaged salinity data. They included zonations for Alsea, Yaquina and
Tillamook estuaries. Our oceanic zone extends further upriver (1.6, 3.2, and 8 km for Alsea,
Tillamook and Yaquina, respectively) than the seawater zones presented in Shirzad et al. (1988).
This is probably a result of our zonation being based upon dry season conditions, whereas the
ones in Shirzad et al. (1988) are based upon average annual conditions.
Treated effluent associated with WWTF is discharged into six of the target estuaries. In
Yaquina, Umpqua River and Nestucca estuaries, the WWTF effluent is discharged into the
riverine segment, while in the Alsea and Coos estuaries the discharge is in the oceanic segment.
Tillamook Estuary has multiple WWTFs discharging into it, one into the oceanic segment and
two into the riverine segment. Effluent discharged into oceanic segments would experience
strong tidal mixing, and the effluent would remain in the estuary for a shorter time period. One
method that may aid in determining the susceptibility of estuaries to nutrient enrichment would
be to overlay point source inputs on the ocean-river zonation.
There is some uncertainty in the estuarine zonation as a result of data gaps as well as interannual
variability. This analysis was based upon one-time sampling of natural abundance stable isotope
of green macroalgae. When possible, we compared our stable isotope values to other data
available for the target estuaries. Coastal systems are highly dynamic, responding to
environmental forcing at scales ranging from minutes (e.g. changes in tidal elevation) to days
(e.g. storm tracks) to decades (e.g. the Pacific Decadal Oscillation). The zonation described here
is based primarily on single sampling events or a few days of sampling. Consequently, it is
imperative to recognize that these zonation schemes are intended to provide guidance at a gross
scale. We are most confident in the results for simpler estuaries (e.g. one riverine source) and
systems where extensive datasets were available (e.g., Alsea, Salmon and Yaquina). When
possible, we compared our value of riverine and oceanic end members to values from other
sources. These comparisons revealed that the river end members we utilized were consistent
with historical data. For estuaries where there are multiple rivers flowing into the estuary (e.g.,
Coos and Tillamook) there is more uncertainty in the results given the logistical and financial
constraints of sampling multiple riverine inputs.
One of the limitations of this analysis is that we cannot distinguish human (WWTF or septic)
from animal waste because their nitrogen stable isotope ratios are similar. Other chemical tracer
analyses can make this distinction; however, these methods were not used in this study.
154
-------
Additionally, we assume that macroalgae do not fractionate (preferentially take up 14N relative to
15N) during nitrogen assimilation, which is supported by the recent experimental work of Cohen
and Fong (2005). We are currently conducting laboratory experiments to critically evaluate this
assumption for green macroalgae collected from Yaquina Estuary. In addition, our method
assumes that the nitrogen stable isotope ratio is a result of a mixture of oceanic, riverine and
WWTF sources, and that the end member values are the same for all of the target estuaries. We
believe this is a good assumption due to the similarity in land cover within the watersheds of the
seven estuaries (Table 3-3) and the similarity in ocean conditions over the geographic extent of
our study (Section 4.3). Of the target estuaries, Coos and Umpqua had the least amount of alder
in the watershed (See Section 4.2). Fry et al. (2001) sampled water column nitrate and green
macroalgae in the Coos Estuary. The 5 15N of nitrate at a riverine site averaged +1.7%o and of
macroalgae averaged 1.8 %o, which is consistent with our riverine end member. Umpqua had the
lowest amount of red alder in the watershed (< 0.5%), and the most riverine macroalgae
sampling sites there had a minimum isotope ratio of+5.4 %o. If the riverine end member for
Umpqua was not +2 %o, we may have over-estimated the contribution from WWTF inputs. The
watersheds of the seven estuaries are primarily forested and have low population densities. If
there is an additional source that we have neglected or if denitrification is important, we may
have incorrectly calculated the contribution of the sources. Kendall and McDonnell (1998)
recommend that the contribution of nitrogen sources from natural abundance stable isotopes be
confirmed through an independent non-isotopic method. For Yaquina Estuary, we were able to
confirm the isotopic contribution of sources using the transport model.
Denitrification can result in elevation of the 5 15N, which can be misinterpreted as WWTF input
in our analysis. Kendall and McDonnell (1998) estimated that denitrification of fertilizer nitrate
with a 815N of l%o can result in the residual nitrate having an isotope ratio of 15%o, which is
similar to animal or human waste. The importance of denitrification in nitrogen removal in
estuaries is a function of residence time (Dettmann, 2001). Denitrification represents a
significant loss of nitrogen in estuaries with relatively long residence times. The residence times
of the target estuaries are relatively short (less than 1 month) due to their small volume, strong
tidal forcing, and high freshwater inflow (see Tables 3-6 and 3-7). We would expect that the
importance of denitrification would increase as the overflow declines. However, based on the
analysis of Dettmann (2001) and residence times (Table 3-7), less than 10% of the nitrate in the
target estuaries would be denitrified. Furthermore, the ability of the Yaquina transport model to
reproduce the spatial and temporal patterns in the isotope data strongly suggests that the nitrogen
stable isotope ratio of the macroalgae is being determined primarily by a mixture of the three
nitrogen sources (oceanic, riverine, and WWTF) rather than by denitrification. Our analysis also
assumes that the 5 15N of the sources is the same across the region. Comparisons of macroalgal
nitrogen stable isotope ratios between Yaquina and Alsea are similar for both the ocean and river
end members (Figures 5-3 and 5-8). In addition, macroalgae 5 15N from Coos Estuary by Fry et
al. (2001) further support the validity of this assumption. This analysis also assumes that there is
minimal temporal variability in the 5 15N of the end-members. The ability of the Yaquina
transport model to predict the 5 15N reliably using fixed values for the end-members (Figures 5-4
and 5-5) indicates that this is a reasonable assumption. Finally, we assume that the WWTFs
represent a minor contribution of the total freshwater input to these systems. Riverine flow rates
are many times larger than WWTF inputs even during periods of low river flow, suggesting that
this is a good assumption.
155
-------
Since we sampled intertidal macroalgae, there is some potential that our analysis would be
biased (over estimating oceanic contribution of nitrogen) due to elevation of the samples. If
macroalgal samples are collected at a high elevation relative to mean lower low water, then they
may only be exposed to water during higher water elevations, which would tend to occur during
flood tides. Since we primarily are concerned with nitrogen sources that seagrass habitat is
exposed to and most of the seagrass habitat is located in the intertidal, this is probably not a
significant source of error. This error would be more significant for highly stratified estuaries,
which is typically not the case for Oregon estuaries during the dry season (Tables 3-1 and 3-6).
Model simulations also reveal that this error would be most important in the lower portions of
the estuaries compared to the upper.
In our study area, most of the nitrogen sources to the estuary are isotopically distinct (Figure 5-1)
which leads to a relatively clean 2 or 3 end-member mixing solution. As discussed in
Section 3.7, watershed nitrogen inputs in the area are generally related to red alder (Alnus rubrd)
a nitrogen fixing species characterized by 5 15N ranging between -3 and -0.5 %o (Hobbie et al.,
2000; Tjepkema et al., 2000; Cloern et al., 2002). Samples from the Yaquina River indicate that
the 5 15N of the nitrate in the river water ranges from +0.4 to +2.4 %o (Kaldy, unpublished),
which is consistent with macroalgal 5 15N values for our riverine sites. Literature values for the 5
15N of oceanic nitrate (the dominant nutrient associated with upwelling) along the PNW coast of
the United States range between +6.6 and +7.7 %o (Kienast et al., 2002; Wankel et al., 2006).
The 5 15N of the nitrate for water samples collected from 25 miles offshore of the Yaquina
Estuary range from +6.7 to +7.6 %o (Kaldy, unpublished). The 5 15N values of green macroalgae
utilizing recently upwelled water along the Oregon coast typically range between +7 and +9 %o
(Fry et al., 2001; and Kaldy, unpublished). WWTF effluent generally has high 5 15N with values
ranging between +7 and +25 %o (Heaton, 1986; Jones et al., 2003). Future work includes the
measuring the isotope ratios of WWTF effluent from the local sources. Agricultural fertilizer
inputs to coastal Oregon estuaries are minimal since there are no major agricultural crops
cultivated along the coast. The amount of cultivated land in the watersheds of the seven target
estuaries varied from 0 to 0.22% (Table 3-3). In some localized areas animal waste inputs may
be substantial, notably Tillamook Estuary; however our isotope analyses cannot distinguish
between animal and human waste.
The use of macroalgal 5 15N to identify nitrogen sources in estuaries has received attention in the
recent scientific literature (Cohen and Fong, 2005). Unfortunately, the use of 515N is not always
clear cut. Nitrogen dynamics are extremely complex, mediated by a variety of microbes.
Different microbial biochemical transformations have specific isotope fractionation factors such
that the products have different 5 15N values than the initial reactants (Fry et al., 2003).
Consequently, caution must be used in the interpretation of 5 15N data, in particular neglecting
the ocean end member. For example, the direct application of the 5 15N regression equations
from Cole et al. (2004) to the data presented here would erroneously suggest that WWTF have a
much greater impact on estuarine nitrogen dynamics. The background marine signal associated
with upwelling is a departure from the general nutrient loading paradigm (e.g. nutrients primarily
from anthropogenic watershed sources). The potential error can result from not accounting for
all nitrogen sources and is likely to occur in areas where ocean upwelling is a dominant feature
of seasonal nutrient inputs as it is along the Pacific Coast of the U.S. (Fry et al., 2003; Cole et al.,
156
-------
2004). We have used multiple lines of evidence to support our conclusions. Specifically, we
have used two-end member mixing models of 5 15N and salinity as well as Isosource (Phillips
and Gregg, 2003) and a transport model. This approach was validated for the Yaquina Estuary
and we are encouraged that all three of these approaches provide very similar zonation patterns.
Based on the zonation scheme described in this chapter, five of the target estuaries had >50% of
the total estuarine area classified as ocean dominated. The Salmon and Umpqua River estuaries
were classified as having 29 and 30%, respectively, of total estuarine area classified as ocean
dominated. Salmon River Estuary is the smallest of all the systems examined encompassing
only 3.1 km2 of total estuarine area. As discussed in Section 4.4.3, both of these estuaries switch
from ocean- to river-dominated over a tidal cycle and based on this we classified them as having
the strongest river influence (Figures 4-8 and 4-9). Most systems were ocean dominated,
indicating that the influence of the coastal ocean and the dynamics of oceanographic phenomena
cannot be ignored. The strong influence of the coastal ocean on PNW estuaries is a departure
from the usual paradigm associated with nutrient loading.
Table 5-17. Summary of the total estuarine area for each target estuary and the area and
percentage of total in the oceanic and riverine segments based on salinity and macroalgal isotope
ratios.
AREA
(KM2)
Total
Estuarine
Oceanic
Segment
Riverine
Segment
YAQUINA
19.9
13.4
6.58
ALSEA
12.5
7.75
4.72
SALMON
3.11
0.91
2.19
UMPQUA
33.8
10.1
23.7
COOS
54.9
42.7
12.1
TILLAMOOK
37.5
23.4
14.0
NESTUCCA
5.00
4.07
0.93
%in
Oceanic
Segment
%in
Riverine
Segment
67
33
62
38
29
71
30
70
78
22
63
37
82
18
157
-------
CHAPTER 6:
AERIAL MEASURES OF ESTUARINE INTERTIDAL AND SHALLOW SUBTIDAL
ZOSTERA MARINA COVERAGE
David R. Young, Patrick J. Clinton, Henry Lee II, David T. Specht, and T Chris Mochon Collura
Key Findings
• Intertidal and shallow subtidal distributions of Zostera marina were mapped in
the seven target estuaries via aerial and on-ground surveys.
• The extent of Z. marina varied among the estuaries, ranging from non-
detectable via aerial surveys to about 11% of the intertidal area.
• The majority of the intertidal and shallow subtidal Z. marina is found in the
oceanic segments of the estuaries.
• Most of the Z. marina habitat is found at depths ranging from -3 to 3 feet (-0.9 m
to 0.9 m) above MLLW.
6.0 Introduction
Zostera marina is a flowering marine plant that can form thick meadows or beds of perennial
plants in the intertidal and subtidal sections of estuaries, providing a critical habitat and food
source for numerous taxa including commercially important fish and shellfish (Heck et al., 1989;
Sogard and Able, 1991; Dennison et al., 1993; Bostrom and Bonsdorff, 1997). Z. marina is one
of the many species of seagrasses that have been severely impacted by anthropogenic activities
around the world (Hemminga and Duarte, 2000; Short et al., 2001). One cost-effective approach
of assessing current seagrass distributions as well as changes in distributions is through the use
of aerial surveys. However, the intertidal and subtidal distribution of Z. marina presents
challenges in mapping its distribution. Therefore, an aerial survey method of mapping the
intertidal and shallow subtidal portions of the Z. marina habitat in PNW estuaries has been
developed by Clinton et al. (2007). A method of mapping the deeper subtidal portion of the
distribution not visible from the surface is still under development and testing (Dr. Ted DeWitt,
pers. comm.).
The objectives of this aerial photomapping study were to 1) assess the extent and distribution of
intertidal and shallow subtidal Z. marina in the seven target estuaries, and 2) determine the
relative distribution of Z. marina in the oceanic and riverine segments of each of the estuaries.
6.1 Methods
6.1.1 Aerial Photography Sampling Frame and Design
Owing to the extensive eelgrass distribution in PNW estuaries, this component of the project
used a remote sensing technique. The method employed aerial photographs and false-color near-
infrared (color infrared, CIR) film, which provide better contrast than full-color film in
distinguishing submersed aquatic vegetation (SAV) distributions (Young et al., 1999). However,
158
-------
because CIR film cannot resolve images more than a few cm below the water surface, the
photographs must be taken during exposed intervals (i.e., daylight low tides). In addition, the
weather must be cloud-free (or uniform high overcast) to obtain uninterrupted photo survey
coverage. In addition, the presence of benthic green macroalgae can confound the interpretation
of the SAV signals; thus, it is important to acquire the photography of target estuaries during late
spring or early summer, when there is enough Z. marina growth and sunlight for imaging, but
before the summertime bloom of benthic green macroalgae (see Section 7.2.4). Although most
of the photography was conducted when the tide level was between 0 and -2 ft (-0.6 m) relative
to Mean Lower Low Water (MLLW), the upper portion of the immersed eel grass plants in the
shallow subtidal zone (down to about -6 ft or -1.8 m) was floating on the surface and could be
detected by the CIR film. Thus, the range of the intertidal/shallow subtidal zone sampled by this
technique was from the upper margin of the zone at about +6 ft (+1.8 m) to about -6 ft (-1.8 m),
the approximate lower depth limit of detection. (English units are used for most available tide
level and bathymetry data in the PNW, and thus are the primary units used here). The lateral
extent of a study area generally ranged from about the ocean entrance of the estuary to the
upriver termination of the reported distribution of intertidal Z. marina in that system. In the
Yaquina Estuary, the aerial photography extended only about 11 km upriver from the ocean
entrance; thus, this study area did not include the entire riverine segment. However, numerous
boat surveys by the authors have shown that only a small proportion of the intertidal/shallow
subtidal Z. marina occurs upriver of the study area.
The photosurveys were conducted under contract by a commercial vendor. Maps of the required
flightlines and photocenters, tables of daylight low tide windows, and specifications of the
photography (e.g., large-format camera focal length, front and side overlap, maximum aircraft
tilt, camera calibration, etc.) were provided by PCEB. The photoscale used in the 2004
photography was 1:10,000, and in 2005 was 1:20,000. The resultant film (or diapositive copy)
was digitally scanned, and the digital photographs then were converted to digital photomaps with
pixels corresponding to 0.25 mx 0.25 m on the ground using ERMapper® desktop
orthorectification. These orthophotos were mosaiced for each estuary and then each pixel was
classified into one of two classes, defined as: (1) eelgrass bed (>10% cover by Z. marina) or (2)
bare substrate (< 10% cover by Z. marina).
Image classifications were accomplished with a combination of digital image processing and
manual techniques. Terrestrial portions of the image were masked in ERMapper® using vector
polygons, and a vegetation index algorithm was applied to mask patently unvegetated areas in
the remaining imagery. A seven class unsupervised classification using ERMapper® ISOCLASS
algorithm was applied to the masked imagery and the results converted to seven Arclnfo® format
binary grids representing each band of the isoclassification. The grids were manually edited in
ArcMap® by the photointerpreter using ArcScan® raster editing tools to remove false-positive
pixels and recompiled to form the seagrass map.
An example of the distribution of Z. marina coverage in the Yaquina Estuary obtained from the
classification of the April 2004 digital photomaps is presented in Figure 6-1. The aerial
photography did not yield acceptable images of Z. marina beds in the Nestucca and Salmon
River estuaries. For these estuaries, a vessel equipped with a differential-corrected global
positioning system (DGPS) was used to position the visible edges of intertidal and shallow
159
-------
f\J
Bare Training Site
>10% Seagrass Training Site
Seagrass margin - Boat GPS
Seagrass margin - Walking GPS
>10% Seagrass Classification
0.5
0
0.5
1 Kilometers
Figure 6-1. Intertidal and shallow subtidal distribution of Z. marina from digital image
classification of aerial photos of Yaquina Estuary taken in April 2004. The eastern (upriver)
edge of the survey area was selected because little intertidal Z. marina occurs beyond the
boundary shown.
160
-------
GPS Tracklines for Seagrass Delineation
Lower Boundary Boat Track Upper Boundary Foot Track
A SO 0 50 100 150 200 250 Meters
Figure 6-2. Track of lower margin of surface-visible distribution of Z. marina in the near-
subtidal zone (using DGPS in a small boat), and of upper margin (walking with DGPS), in the
Yaquina Estuary.
161
-------
subtidal Z. marina beds within each estuary. In some cases, upper margins also were positioned
by carrying a DGPS while walking this margin during exposed conditions. The root mean
square error of the DGPS system used in this study was ± 0.6 m. An example of such
positionings of the visible edges of an intertidal and shallow subtidal Z. marina meadow in
Yaquina Estuary is presented in Figure 6-2.
6.1.2 Estuaries Surveyed
The sampling dates and mapping method utilized for the seven target estuaries are presented in
Table 6-1.
Table 6-1. Mapping method utilized and dates of mapping for intertidal and shallow subtidal
Z. marina distributions in the target estuaries.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
DATE
April 9, 2004
Summer 2005
May 26, 2005
Summer 2004
Summer 2004
April 9, 2004
July 24, 2005
April 9, 2004
July 23, 1997 (above river mile 7)
MAPPING METHOD
Aerial Orthophotography/
Surface Positioning
Aerial Orthophotography
Surface Positioning
Surface Positioning
Aerial Orthophotography
Aerial Orthophotography
Aerial Orthophotography
6.1.3 Ground Surveys
"Ground truth" surveys were conducted in the five estuaries surveyed by aerial photography
usually in the same season that the photosurvey was conducted but sometimes a year or two
later. The methodology was based upon the recommendations of Congalton and Green (1999).
First, the most reliable map available of the target estuarine intertidal zone was used to separate
expected Z. marina from bare substrate strata. One hundred stations then were positioned
randomly within each stratum. Utilizing hovercraft transport and the high-resolution DGPS, 30-
70 stations in each class were located during low tide, positioned, and surveyed for percent cover
of Z. marina. A 1.25 m x 1.25 m quadrat, equipped with two orthogonal sets of five taut strings,
was placed successively in the four compass quadrats around the target position. The percent
cover of native Z. marina., green macroalgae, non-native Z.japonica, or bare substrate then was
quantified using the point-intercept method. Following the criterion established by the NOAA
Coastal Change Analysis Program (Dobson et al., 1995), stations with Z. marina coverage
greater than 10% were classified as Z. marina sites. Interference by benthic green macroalgae or
Z.japonica in the separation of ground stations into Z. marina or bare substrate classes was
negligible in all five estuaries surveyed. Approximately 10% of the stations surveyed in each
class were randomly withdrawn and used for training the geographical information system (GIS)
photointerpreter conducting the image classification. The data for these stations were not used in
the accuracy assessments of the resultant classifications.
162
-------
6.1.4 Classification Accuracy Assessments
Following the classification of the digital photomaps, positions of the non-training stations
determined on-site by DGPS were provided to the GIS photointerpreter for the five estuaries in
which ground surveys were conducted. An area equivalent to 2.5 m x 2.5 m on the ground
around a station's DGPS position was subsampled from the digital classification into a
10 pixel x 10 pixel array. Each pixel had been classified either as "Z. marina" or "bare
substrate." If the number of pixels classified as Z. marina was greater than 10 (>10% of the
total), that station was classified as Z. marina. Otherwise it was classified as a bare substrate
station. The same criterion was applied to the ground survey data. The results from the digital
classification and ground survey then were incorporated into a classical error matrix (Congalton
and Green, 1999).
6.1.5 Estuary Bathymetry
In view of the importance of substrate elevation to estuarine ecology, a bathymetric model was
developed for Yaquina Estuary. Several surveys of the main channel have been conducted by
the U.S. Army Corps of Engineers (USAGE) in recent years, but little information existed for the
extensive tide flats of the estuary. Therefore, in 2002 WED/PCEB contracted with USAGE to
extend their surveys into the shallow sectors of the estuary. Soundings were conducted at
extreme high tides near the end of the year along transect lines every 200 feet (67 m) over the
general area surveyed by aerial photography. Measured water depths were related to those
recorded by tide gauges within the estuary, and adjusted relative to Mean Lower Low Water.
The total area covered by the several bathymetric surveys utilized in this study is illustrated in
Figure 6-3. The data resulting from these surveys were used by PCEB to construct a bathymetric
model of Yaquina Estuary. Survey easting, northing, and depth values were interpolated using
the TOPOGRTD method provided in Arclnfo Workstation. This model is discussed further in
Section 6.2.3.
USACE 2002
USACE 2000
USACE 1999
USACE 1998
NOS 1953
I'DS 1>EM 1998
USGS 10m DEM
A
Figure 6-3. Sectors of Yaquina Estuary covered by bathymetric surveys conducted by the U.S.
Army Corps of Engineers between 1998 and 2002.
163
-------
6.2 Results
6.2.1 Accuracy Assessments of Photomap Image Classifications
Error matrices were prepared from the digital classification and ground survey results for the
Alsea, Coos, Tillamook, Umpqua, and Yaquina estuaries. Errors of omission (failing to include
a ground-truth station in the classification), commission (erroneously including a station in the
classification), and overall error for the classifications of the five orthophotographs are listed in
Table 6-2. These results for five of the seven target estuaries indicate that the Z. marina habitat
and bare substrate classifications from the estuary orthophotographs have accuracies of 83% or
greater, and the median overall accuracy was 89% (median overall error of 11%). For more
details on the accuracy assessment of the mapping see Section B.9 in Appendix B.
Table 6-2. Summary of error matrix results for detection of Z. marina in five of the target
estuaries based on the ground truth surveys.
ESTUARY
CLASSIFIED
Alsea
Z. marina
Bare Substrate
Coos
Z. marina
Bare Substrate
Tillamook
Z. marina
Bare Substrate
Umpqua
Z. marina
Bare Substrate
Yaquina
Z. marina
Bare Substrate
ERRORS OF
OMISSION
0%
0%
26%
12%
12%
16%
10%
11 %
2%
4%
ERRORS OF
COMMISSION
0%
0%
12%
26%
16%
12%
11 %
10%
4%
2%
OVERALL
ERROR
0%
17%
14%
11 %
3%
NO. OF
STATIONS
44
60
38
77
100
81
41
91
51
28
6.2.2 Among-Estuary Comparison of Zostera marina Coverage
Interpretation of the distribution and abundance of Z. marina from the aerial surveys and from
the probabilistic field surveys discussed in Chapter 7 requires an understanding of how areas
were sampled in the two approaches. The aerial photography detects Z. marina in both the
intertidal and shallow subtidal zones. As discussed above (Section 6.1.1), it appears that the
aerial photography detects Z. marina to a depth of approximately 6 feet (1.8 meters) below Mean
Lower Low Water (MLLW). In comparison, the probabilistic field surveys used an intertidal
sampling frame, approximately from Mean Lower Water (MLW) to Mean Higher Water
(MHW). Thus, the probabilistic surveys would not capture the subtidal portion of the SAV
population that was included in the aerial surveys. To compare directly the areal extents of Z.
marina between the two surveys, the sampling frames used in the probabilistic surveys were
overlaid on the aerial photography, generating estimates of the areas of Z. marina from the aerial
164
-------
photography within the same areas as sampled by the field surveys. Additionally, to calculate
the percent of the area covered by Z. marina (relative cover) it is necessary to have an accurate
estimate of the area actually surveyed by the aerial photography. Such areas could be calculated
assuming detection to 6 feet (1.8m) if bathymetric data were available. However, bathymetric
data are currently available for only three of the seven target estuaries. Therefore, the percent
cover of Z. marina was calculated using the portion of the aerial surveys that fell within the
probabilistic sampling frames, which have known areas. These percent cover values only apply
to the intertidal zone; future efforts will be directed at obtaining bathymetry for all the target
estuaries, allowing percent cover estimates that also include the shallow subtidal zone.
To compare the extent of variation of Z. marina among estuaries (Objective 1), the absolute and
relative areas of Z. marina in the five digitally classified estuaries, and the two surface-mapped
estuaries were calculated (Table 6-3). Maps of the aerial extent of Z. marina for each of the
target estuaries are presented in Appendix A. One comparison of interest is the percent coverage
values for intertidal Z. marina within the probabilistic frame. Three of the estuaries - Coos,
Tillamook, and Yaquina - show agreement within a factor of three (4.7 to 11.5% cover). For the
Umpqua River Estuary, the percent coverage value (1.3%) is substantially lower. The Alsea,
Nestucca, and Salmon River estuaries have very little intertidal Z. marina coverage within the
probabilistic frame. At present we have no complete explanation for the near absence of
intertidal Z. marina in the Nestucca and Salmon River estuaries. However, one possible
contributing factor may be the high wave energy in these estuaries. Nestucca and Salmon River
estuaries have the lowest values for median percent fines (1.9% and 7.9%, respectively; see
Figure 7-10), suggesting that the intertidal zones of these estuaries are high energy environments,
which may result in Z. marina being eroded or buried. Additionally, high tidal currents and
shifting sands were commonly observed at the mouth of Nestucca during peak flood and ebb
tides (see Section 8.2.5). Salinity may also be a contributing factor limiting Z. marina in these
estuaries as discussed in Section 7.2.2.
6.2.3 Bathymetric Distribution of Zostera marina
A comparison of elevations predicted by the bathymetric model for Yaquina estuary with those
obtained from an independent total station survey of a large embayment there (Idaho Flat) is
presented in Figure 6-4. This comparison indicates that there is agreement between the
measured and modeled elevations to within about 0.3 ft. Specifically, for the 167 points within
the depth interval -3.0 ft to +3.0 ft, the median value for the difference between the survey and
model elevations is -0.32 ft, indicating that, on average, the actual substrate elevations in this
interval may be approximately 0.3 ft (0.1 m) lower than the model elevations.
Bathymetric data collected by Professor Chris Goldfinger at Oregon State University for Alsea
Estuary (2002) and by the USAGE for Tillamook Estuary (1995) similarly were used to obtain
bathymetric models for these estuaries. The results presented in Figures 6-6 to 6-8 indicate
similar bathymetric distributions for intertidal/shallow subtidal Z. marina in these three PNW
estuaries. The aerial percentages of Z. marina occurring between specific depth intervals around
MLLW are summarized in Table 6-4. Again, the population included in this study is Z. marina
that is visible from the surface at low tide (MLLW), falling within the approximate depth range:
-6 ft (-1.8 m) to +6 ft (+1.8 m).
165
-------
Table 6-3. Estimates of Z. marina area in the target estuaries from the aerial/on-surface surveys.
The "Total Area of Z. marina" is the total area of Z. marina detected in the aerial/on-surface
surveys in the intertidal and shallow subtidal areas. The "Area of Z. marina within Probabilistic
Frame" is the area of Z. marina found within the intertidal probabilistic frame, which does not
include the shallow subtidal area sampled by the aerial photographs. The "% Coverage of Z.
marina within Probabilistic Frame" is the percent coverage of Z. marina relative to the area of
the probabilistic frame. A similar metric cannot be calculated for the entire aerial frame since
the exact area sampled by the photographs is unknown. The areas of Z. marina in the Nestucca
and Salmon River estuaries were determined by on-surface mapping rather than aerial
photography.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
TOTAL AREA OF
Z MARINA IN
AERIAL/ ON-
SURFACE SURVEY
(km2)
0.026
2.141
0.004
0.004
3.27
0.338
0.809
AREA OF Z MARINA
WITHIN
PROBABILISTIC
FRAME (km2)
0.005
1.238
0.0
0.0003
2.38
0.095
0.635
% COVERAGE OF
Z MARINA
WITHIN
PROBABILISTIC
FRAME
0.09
4.66
0.00
0.07
11.46
1.33
9.91
4 -
2 -
OH
«
>•
3
-o
—
•o
o
-2 -
-4-
y = 0.75 x +0.42
r2 = 0.94
1 i ' i ' i ' i ' r
-6 -4 -2024
Measured Elevations (ft, MLLW)
Figure 6-4. Comparison of substrate elevations from a bathymetric model and independent total
station survey measurements within Yaquina estuary.
166
-------
•iTTOOD
417500 413000
473500 419000
•115500
(lest above MLLW} (1««l above MLLW)
-43 --39
-39 --36
-36 --33
-33 --27
-30 --27
-24
-24 - -21
-21 "
-19 --15
iou.c..- UiACOf F9*i,
Afttll Photo • US CM 1M
417000
419500
129000
Figure 6-5. Interpolated bathymetry and Z. marina distribution for a section of the lower
Yaquina Estuary.
167
-------
40 -
- 30 -\
cs
-*•>
O
H
*o 20 -I
10 -
0
45.94
10.65
0.88 1.68
2'62
4.58
6.93
1 1
19.93
\ \ \ \
2 59
0.92 0.76 0.61 0.45
I I
.6-5-4-3-2-10 1 2 3 4 5 6
Depth Interval, ft (MLLW)
Figure 6-6. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the bathymetric
model for Yaquina Estuary (-6 ft to +6 ft).
40 -
- 30 -
cs
-*•>
o
H
*o 20 -I
10 -
0
45.38
5.35
6.86 7.14
1.02 1.37 2.00
1 1
13.29
\ \ \ \
8.08
3.27
2-17 1.19 0.80
\\
-6-5-4-3-2-10 1 2 3 4 5 6
Depth Interval, ft (MLLW)
Figure 6-7. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the bathymetric
model for Alsea Estuary (-6 ft to +6 ft).
168
-------
50 -
40 -
5 30 -
o
H
t*.
I 20-
10 -
0 -
49.63
0.39 0.79 1.41 ,2-27,
3.71
6.52
17.52
11.61
4.36
°-83 0.12 0.10
-6 -5-4.3-2-10 1 2 3 4
Depth Interval, ft (MLLW)
Figure 6-8. The percentage distribution of intertidal and shallow subtidal Z. marina habitat area
classified from aerial photography that occurs in one foot intervals predicted by the bathymetric
model for Tillamook Estuary (-6 ft to +6 ft).
Table 6-4. Cumulative percent area of Z. marina within four depth intervals relative to MLLW
in three PNW estuaries.
ESTUARY
Yaquina
Alsea
Tillamook
DEPTH INTERVAL (ft)
-i'to+r
76.6
65.9
73.6
-2' to +2'
86.0
80.3
88.9
-3' to +3'
91.4
88.8
95.5
-6' to +6'
97.7
96.8
99.0
As shown in Table 6-4, approximately 90 percent of the intertidal and shallow subtidal Z. marina
classified from the orthophotography images occurred within the depth range -3.0 ft to +3.0 ft (-
0.9 m to +0.9 m) around the MLLW datum. In comparison, Borde et al. (2003) reported that the
upper limit of the Z. marina meadows observed in Coos Bay during summer 1999 was
approximately +0.8 m (MLLW), in good agreement with our results. Corresponding
observations for the maximum depths range from -0.2 m to -1.5 m (median: -0.85 m; Thorn et
al., 2003), compared to our approximate lower limit value of-0.9 m. Further, these authors
reported maximum depths ranging from -0.4 m to -1.2 m (median: -0.80 m) in Willapa Bay, WA
(for location see Figure 2-2), again consistent with our findings. In a subsequent study, Thorn et
al. (2008) reported that Z. marina shoot densities decreased to zero below -1.5 m in Coos Bay
and Willapa Bay.
169
-------
Orth and Moore (1988) concluded from their studies in lower Chesapeake Bay that both
optimum and maximum depths for this species can vary considerably within a particular region,
depending upon water clarity. The same observation was made for the maximum depth of Z.
marina at numerous locations throughout the seven target estuaries surveyed in this study
(Chapter 8). In general, deepest limits occurred near the mouths of the estuaries where the
clearest water usually is found. The median of the maximum depth readings, relative to MLLW,
for each estuary ranged from -1.82 m (for Alsea Estuary) to +0.21 m (for Coos Estuary), with an
overall median of -0.76 m. (For the Tillamook, Yaquina, and Alsea estuaries alone, the overall
median was -1.19 m). Again, these findings are in reasonable agreement with the lower limit
value of-0.9 m we obtained for approximately 90 percent of the Z. marina in our bathymetric
distributions for three of the seven target estuaries.
This comparison presented in Table 6-4 suggests a method of estimating the suitability of
intertidal and shallow subtidal habitat for Z. marina as a function of bottom depth in a given
estuary. The estimate, for a specified depth interval, is calculated as the area in which Z. marina
occurs divided by the total area of the target estuary within that depth interval. These "frequency
of occurrence" values obtained for one-foot (0.3 m) depth intervals around MLLW between -6 ft
and +6 ft are illustrated in Figure 6-9.
| 40 -
>_
3
u
0 30-
o
u
§ 20 -
3
cr
4)
'_
2 10 -
s
i
§
nnR
^N U | |
-6 -5
n
/
1
-4
/
/
-3
i
/
/
/
-2
i — |
Pi
.
f
/
'
'
m
/
/
'
'
/
/
/
/
''
i i i
-1 0
—
Depth Interval,
—
./
/
/
/
/
n
i1
1 1
/
1 1 Yaquina
^^1 Alsea
\/ / \ Tillamook
R
3456
ft (MLLW)
Figure 6-9. Frequency of occurrence values for Z. marina within one-foot depth intervals around
MLLW in Yaquina, Alsea, and Tillamook estuaries (-6 ft to +6 ft).
Because the interval -3 ft (-0.9 m) to +3 ft (+0.9 m) contains approximately 90% of the Z. marina
classified in the color-infrared aerial photomaps for each of these three estuaries, this zone is
termed the primary depth interval for Z. marina. The average (± 1 standard error of the mean)
frequency of occurrence values obtained for this interval in Yaquina, Alsea, and Tillamook
estuaries are 13.9 ± 3.6%, 0.44 ± 0.12%, and 14.4 ± 3.3%, respectively. The averages for
170
-------
Yaquina and Tillamook estuaries are identical (14%), but are higher than that for the Alsea
Estuary (0.44%) by about a factor of 30. These relative values are consistent with those obtained
from the data presented in Table 6-3. Again, further studies are required to elucidate the causes
of differences in percent coverage values for Z. marina among PNW estuaries.
6.2.4 Within-Estuary Distribution of Zostera marina
As discussed in Chapter 5, the seven target estuaries were divided into riverine and oceanic
segments based on their primary nutrient sources. One of the objectives of the aerial surveys as
well as the probabilistic surveys (Section 7.3.1) was to assess the distribution of Z. marina
between these two estuarine segments. To address this question, the percent of total area of
intertidal Z. marina habitat classified from the aerial surveys (or surface mapping in the case of
the Nestucca and Salmon River estuaries) was apportioned within each segment, using both the
Total Aerial Frame and the Probabilistic Frame (Table 6-5).
Table 6-5. Relative distribution of the total Z. marina habitat area in the oceanic and riverine
segments of the estuaries based on the aerial surveys. Estimates are based both on the Z. marina
detected within the entire area covered by the aerial surveys and the portion of the Z. marina
detected within the subset of the intertidal probabilistic frame. NA = not applicable because no
Z. marina was detected within the probabilistic frame.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
% TOTAL Z MARINA FROM AERIAL SURVEY
CALCULATED USING TOTAL
AERIAL FRAME
OCEANIC
79.6
96.2
100.0
87.3
99.8
52.8
98.5
RIVERINE
20.4
3.8
0.0
12.6
0.2
47.2
1.5
CALCULATED USING
PROBABILISTIC FRAME
OCEANIC
98.8
96.6
NA
99.6
99.7
88.5
99.7
RIVERINE
1.2
3.4
NA
0.4
0.3
11.5
0.3
A major implication of these results is that for all of the target estuaries, the majority of the
intertidal and shallow subtidal Z. marina classified from the aerial (or surface mapping) surveys
is found in the oceanic segment. Generally, the distributions between the oceanic and riverine
segments calculated using the aerial and probabilistic frames are similar, with the exception of
the Umpqua River Estuary. In this estuarine system, Z. marina habitat using the Total Aerial
Frame is evenly divided between the oceanic (52.8%) and riverine (47.2%) segments. However,
the percent-of-total value for the oceanic segment of the Umpqua using the Probabilistic Frame
(88.5%) is much higher. One possible explanation for the difference between the two
approaches is that there may be mis-classification of Z. marina in the riverine segment. In
particular, two other seagrasses, Z. japonica and Ruppia maritima (widgeon grass), that occur in
lower salinity areas are difficult to differentiate from Z. marina using remote sensing.
Nonetheless, the presence of at least 53% and up to 100% of the total Z. marina mapped from the
aerial surveys occurring within the oceanic segments demonstrates the importance of this
segment to this species in the PNW.
171
-------
The distribution of Z. marina between the oceanic and riverine segments was also evaluated
using the relative percent cover of Z. marina as determined from the aerial photomapping
(Table 6-6). As discussed in Section 7.3, relative cover is defined as the percent of the area of
the oceanic segment or the riverine segment occupied by Z. marina. Calculation of relative
cover is limited to the probabilistic frame portion of the aerial frame since the area sampled is
needed to calculate the percent cover value. Higher relative cover values are assumed to indicate
a better habitat for a species given that it occupies a greater percentage of the total habitat area.
Based on this assumption, the oceanic segment constitutes a substantially better habitat for
Z. marina, with relative cover 7-fold to almost 200-fold higher in the oceanic segments in the six
estuaries with detectable Z. marina habitat. These results are consistent with the relative
distributions of the total Z. marina population (Table 6-5) and support the view that the lower
estuary is the prime habitat for Z. marina in PNW estuaries.
Table 6-6. Percent coverage of Z. marina within the oceanic and riverine segments of the
estuaries based on the aerial surveys. Estimates are based on the relative cover in the
probabilistic portion of the aerial survey frame. NA = not applicable because no Z. marina was
detected within the probabilistic frame.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
RELATIVE % COVER OF Z MARINA
OCEANIC
0.10
5.63
NA
0.08
15.28
2.86
12.03
RIVERINE
0.01
0.79
NA
0.002
0.08
0.26
0.18
6.3 Synthesis
Color infrared aerial photosurveys of the target estuaries yielded orthophotographs that were
digitally classified into Z. marina and bare substrate classes. The classifications were checked
via ground surveys using a stratified random sampling design. Classical accuracy assessments
yielded individual accuracies of 83% or greater, with an overall median accuracy of 89%. This
provided intertidal and shallow subtidal Z. marina distributions within five target estuaries, while
two other estuaries were mapped via ground surveys because of the limited areal extent of Z.
marina habitat. Comparison with bathymetric models for three estuaries showed that the one-
foot depth interval bracketing MLLW contained 66% to 77% of the intertidal Z. marina., and that
89% to 96% of the measurable Z. marina occurred within the primary depth interval -3 ft to +3 ft
(-0.9 m to +0.9 m). The ratio of the area occupied by Z. marina relative to the total estuarine
area within this interval provided frequency of occurrence values representative of habitat
suitability that showed Alsea Estuary contained a much lower percentage cover of Z. marina in
the primary depth interval than in Yaquina and Tillamook estuaries. A similar result was
obtained from probability-based ground surveys. Causes for such differences presently are
unknown. However, the maps of Z. marina distribution clearly indicated that a large majority of
intertidal and shallow subtidal Z. marina occurred in the oceanic segments of the target estuaries.
172
-------
CHAPTER 7:
AMONG AND WITHIN ESTUARINE DISTRIBUTIONS OF SEAGRASSES
AND ECOLOGICALLY IMPORTANT BENTHIC SPECIES IN PACIFIC
NORTHWEST ESTUARIES
Henry Lee II, David R. Young, Christina L. Folger, Cheryl A. Brown,
Janet O. Lamberson, Katharine M. Marko, and Faith A. Cole
Key Findings
• The areal extents of Z. marina, Z. japonica, benthic green macroalgae, two
burrowing shrimp (N. californiensis and U. pugettensis), and "bare" habitat in
the intertidal zone were estimated in the seven target estuaries using
probabilistic field surveys.
• All three tide-dominated estuaries had moderate to extensive coverage of
Z. marina. In comparison, Z. marina was not found in three of the river-
dominated estuaries and there was only limited coverage in the fourth.
Differences in median salinity and/or salinity variability among estuary classes
appear to be a major factor determining the extent of Z. marina though
sediment movement/energy and high densities of N. californiensis may also be
contributing factors.
• The nonindigenous Z. japonica occurred in six of the estuaries and the area
occupied exceeded that of Z. marina in four of them.
• Blooms of green macroalgae in coastal PNW estuaries appear to be a natural
process and not an indication of cultural eutrophication.
• Burrowing shrimp were a major component of the intertidal in all the estuaries,
though the species had different distributions among estuaries.
• A multivariate analysis on the benthic habitats grouped the three tide-
dominated estuaries together along with the river-dominated Alsea Estuary.
The river-dominated Nestucca and Salmon formed a second group, while the
river-dominated Umpqua separated out from the other estuaries.
7.0 Introduction
Probabilistic field surveys of the intertidal habitats were conducted in the seven target estuaries
with the objectives of: 1) assessing the among-estuary patterns of distribution and abundance of
Zostera marina, Z. japonica, benthic macroalgae, and two burrowing shrimp, Neotrypaea
californiensis and Upogebiapugettensis and 2) determining the within-estuary distribution of
each of these taxa especially in relationship to their distribution between the oceanic and riverine
segments within each estuary. An additional objective was to develop practical methods to
173
-------
quantify seagrasses, other biotic resources, and key environmental factors in regional-scale
surveys.
The first objective addresses the classification of PNW estuaries based on biotic structure,
providing information at finer spatial scale and with a higher level of biotic resolution than the
wetland classifications conducted at a regional scale (Chapter 2). The second objective
addresses one component of vulnerability, the potential exposure of the five taxa to
anthropogenically derived nutrients. In particular, the objective is to test the hypothesis that the
majority of the Z. marina occurs in the oceanic segment across a range of different types of PNW
estuaries. Probabilistic survey designs are well suited to addressing this objective as they
provide statistically unbiased estimates of the intertidal area occupied by each of the taxa. The
distribution of Z. marina was also evaluated through aerial photography (Chapter 6), and each
technique has its own strengths and limitations. It is beyond the scope of this document to
contrast the two approaches, and in the context of assessing vulnerability, the key question is
whether both approaches showed the same general patterns in Z. marina distributions.
In addition to the two Zostera species, benthic macroalgae and the two burrowing shrimp are
ecologically important benthic species in PNW estuaries, affecting both the physical and
chemical structure of the benthic environment and the structure of estuarine food webs (see
Section 1.5 and Table 1-1). Comparison of their coverage and/or abundance among estuaries
quantifies how these food web components vary by estuary type. By determining their
distributions among the oceanic and riverine segments, the probabilistic surveys identify which
of these food web components are most likely to be exposed to terrestrially derived nutrients, and
hence be more vulnerable to nutrient enrichment.
7.1 Methods
7.1.1 Probabilistic Sampling Frame and Design
A probabilistic sample design, as utilized by the Environmental Monitoring and Assessment
Program (EMAP), was used to generate statistically robust estimates of the abundance and
distribution of seagrasses and other benthic resources. An overview of the probabilistic sampling
approach used in previous coastal EMAP surveys in California, Oregon, and Washington can be
found in Nelson et al. (2005a, b). Additional details on the probabilistic approach to describing
the condition of ecological resources can be found in Diaz-Ramos et al. (1996), Stevens (1997),
Stevens and Olsen (1999), and at http://www.epa.gov/nheerl/arm.
The probabilistic field surveys used the same sampling frame as the coastal EMAP survey of
intertidal wetlands conducted in 2002 (Nelson et al., 2007). The estuarine intertidal sampling
frame was defined as all estuarine NWI polygons that were "regularly flooded", which was
interpreted as being between Mean Lower Low Water (MLLW) and Mean Higher High Water
(MHHW). NWI polygons labeled "irregularly exposed" were interpreted as being below MLLW
while polygons labeled "irregularly flooded" were interpreted as being above MHHW, neither of
which was included in the sampling frame. Thus, the sampling frame did not include either the
shallow subtidal or emergent marshes. Areas of the sampling frame and the post-sampling
division between the oceanic and riverine segments are summarized in Table 7-1.
174
-------
The Alsea, Nestucca, Yaquina and Salmon River estuaries were sampled in 2004, with the Coos,
Umpqua River, and Tillamook estuaries sampled in 2005 (Table 7-1). The field surveys were
completed during the index period of June through mid-September, with the exception of 15
samples in the Tillamook that were taken in December 2005 because of tidal constraints. These
sites were resampled in August 2006, and the updated results are used in the analysis. Nine
additional samples were taken in the Alsea Estuary in September 2006 to sample a portion of the
original sample frame that was inadvertently omitted. Measures of water quality and 515N stable
isotopes in macroalgae were collected synoptically with the probabilistic surveys as discussed in
Chapters 4 and 5.
A total of 100 random sites and a corresponding number of random replacement sites were
allocated to each of the seven estuaries. In an attempt to have a sufficient number of samples in
both the oceanic and riverine segments, the estuaries were a priori divided into sampling strata
(Figures 7-1 to 7-7) with a designated number of samples allocated to each stratum (Table 7-1).
The four estuaries sampled in 2004 were divided into two sampling strata based on an initial
estimate of the demarcation between the oceanic and riverine segments of the estuaries. In 2005
we modified our approach, and divided the Coos, Tillamook, and Umpqua estuaries into a
greater number of sampling strata based on salinity distributions. Sites that were found to be
located in inappropriate habitat (e.g., marsh, rocky areas) or were inaccessible were relocated
within 30m of the original location site using a protocol developed for the EMAP wetland
survey (Lamberson and Nelson, 2002). If it was still not possible to obtain a sample, the site was
abandoned and the first replacement site within the same sampling stratum was chosen.
7.1.2 Data Analysis
Data from the probabilistic surveys were primarily analyzed using cumulative distribution
functions (CDFs). Coastal EMAP studies have used CDFs to evaluate local and regional
patterns of coastal resources (e.g., Summers et al., 1993; Strobel et al., 1995; Hyland et al., 1996;
Nelson et al., 2005a, b). Details on the methods for calculating CDFs and the 95% confidence
intervals can be found in Diaz-Ramos et al. (1996). Briefly, CDFs describe the percentage of the
area of the sampling frame across the full range of the indicator values (e.g., percent cover,
burrow hole counts) based on the cumulation of the weighted samples. The sample weight of
each sample was determined by the area of the stratum divided by the number of samples taken
within stratum. The CDFs were generated for each target estuary over the entire sampling frame
(i.e., intertidal area of the estuary). The CDFs were also generated independently for each
oceanic and riverine segment within an estuary based on the areas determined to fall into these
segments after the zonation of the estuaries (Chapter 5). Since the demarcation line for the
oceanic and riverine segments transected some of the original sample strata, the sample weights
were recalculated based on these new strata areas and sample numbers.
7.1.3 Field Methods
Seagrass, macroalgae, and burrowing shrimp presence and abundance were estimated using two
quadrat sizes (Figure 7-8). Visual estimates were made in large quadrats (2.5 m on a side or 6.25
m ). A length of braided cord ten meters long with markers every 2.5 m was used to lay out the
2.5 m x 2.5 m sampling plot. Polyvinyl chloride (PVC) stakes were used to secure the four
corners and to define the site boundaries. The presence/absence of Z. marina, Z.japonica,
175
-------
c .3 ,0 "5 "e
X —
DDDD »000
Figure 7-1. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
(Chapter 8) in the Alsea Estuary. There were three sampling strata in the Alsea, with the third
stratum sampled in 2006 indicated by the hatched area. Additional lower limit points sampled in
2006 are not indicated in the figure. The locations where Z. marina was detected by aerial
photography are indicated in green. Also see Figure A-l in Appendix A for a map of the
Z. marina distribution based on the aerial surveys.
176
-------
Probabilistic
Sample Strata
Lower Limit Survey
Zostera marina
Terrestrial
O Subtidal
Intertidal
012 4
Figure 7-2. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
(Chapter 8) in the Coos Estuary. There were five sampling strata in the Coos. The locations
where Z. marina was detected by aerial photography are indicated in green. Also see Figure A-2
in Appendix A for a map of the Z. marina distribution based on the aerial surveys.
177
-------
/
o
1
2 Probabilistic
Sample Strata
r~is
® Lower Limit Survey
Zosfera marina
CH> Terrestrial
G Subticlal
Intertidal
0 0.25 0.5
kin
Figure 7-3. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
(Chapter 8) in the Nestucca Estuary. There were two sampling strata in the Nestucca. The
locations where Z. marina was detected by the on-ground survey are covered by the lower limit
symbols and the reader is referred to Figure A-3 in Appendix A for a map of the Z. marina
distribution based on the aerial surveys.
178
-------
1
2
3
4
Probabilistic
Sample Strata
CH5
® Lower Limit Survey
CD Zostera manna
CD Terrestrial
CD Subtidal
CD Intel-tidal
0 0.25 0.5 u^
km
Figure 7-4. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
(Chapter 8) in the Salmon River Estuary. There were two sampling strata in the Salmon. The
locations where Z. marina was detected by the on-ground survey are obscured by the lower limit
symbols and the reader is referred to Figure A-4 in Appendix A for a map of the Z. marina
distribution based on the aerial surveys.
179
-------
Probabilistic
Sample Strata
Lower Limit Survey
Zostera marina
CD Terrestrial
O Subtirtal
CD Intel-tidal
0 0.5 1 2
km
Figure 7-5. Locations of the probabilistic sampling strata and the Z. marina lower limit surveys
(Chapter 8) in the Tillamook Estuary. There were five sampling strata in the Tillamook. The
locations where Z. marina was detected by aerial photography are indicated in green. Also see
Figure A-5 in Appendix A for a map of the Z. marina distribution based on the aerial surveys.
180
-------
Figure 7-6. The locations of the probabilistic sampling strata and the Z. marina lower limit
surveys (Chapter 8) in the Umpqua River Estuary. There were five sampling strata in the
Umpqua. The locations where Z. marina was detected by aerial photography are indicated in
green. Also see Figure A-6 in Appendix A for a map of the Z. marina distribution based on the
aerial surveys.
181
-------
Figure 7-7. The locations of the probabilistic sampling strata and the Z. marina lower limit
surveys (Chapter 8) in the Yaquina Estuary. There were two sampling strata in the Yaquina.
The locations where Z. marina was detected by the aerial photograph are indicated in green.
Also see Figure A-7 in Appendix A for a map of the Z. marina distribution based on the aerial
surveys.
182
-------
benthic macroalgae, burrowing shrimp, and other indicators of biological activity were visually
estimated in each of these large quadrats. In 2005 and 2006, the percent cover of both seagrass
species in the large quadrats was assigned to one of five cover classes (0%, 1-10%, 10-40%, 40-
70%, and >70%). Two smaller 0.25 m quadrats were then placed within the larger 2.5 x 2.5 m
plot site, with one quadrat designated for percent plant cover and the other for burrowing shrimp.
The plant quadrat was placed in the center of the sampling plot to avoid placement bias. The
following measurements were taken within the plant quadrat: 1) percent plant cover of both
seagrasses and macroalgae; 2) number of vegetative, reproductive and seedling shoots; 3) leaf
measurements on five vegetative shoots; 4) macroalgal volume. The shrimp quadrat was placed
in an area within the 2.5 m x 2.5 m plot with less macroalgae or seagrass so that thick cover
would not obscure burrow holes.
We found that the design of the field sheets was critical to having the field crews accurately
collect the pertinent habitat information from quadrats. To assist in future investigations, the
current version of the field sheets is reproduced in Figure 7-9. Additional details on the field
methods follow.
Point-intercept estimate of plant cover - Point intercepts within the 0.25 m2 plant quadrat were
used to estimate the cover of the two seagrasses and benthic macroalgae. The point-intercept
estimate was based on the number of macroalgae and seagrass occurrences under 25 intercepts of
two sets of equally spaced five strings on the plant quadrat. The percent cover of seagrass and
macroalgae were measured independently in three dimensions (i.e., captured occurrence of plants
overlaid by other plants), thus the sum could exceed 100%.
r\
Number of seagrass shoots - All Z. marina shoots were counted within the 0.25 m plant quadrat
in 2004-2006. Due to the potentially high number ofZ.japonica shoots present in a 0.25 m2
area, the number ofZ.japonica shoots was only approximated in the plant quadrats in 2004. In
2005 and 2006, the number ofZ.japonica shoots was counted in a 0.01 m2 cell within the plant
quadrat. For Z. marina, vegetative and reproductive shoots were counted independently and a
determination was made as to whether it was an annual or perennial form by the presence or
absence of rhizomes.
Leaf measurements - Leaf measurements were taken on five haphazardly chosen shoots of both
seagrass species within each plant quadrat when present. Leaf measurements consisted of the
length of the longest blade to the nearest 1 cm, width of the same leaf at the widest point, and
number of leaves per shoot. Where fewer than five vegetative shoots were present within a
quadrat, the measurements were collected from all vegetative shoots present in the quadrat.
Macroalgal volume - Macroalgal volume was measured by removing all the algae from the
0.25 m2 plant quadrat, and measuring the volume of algae using a four-liter volumetric cylinder
following the method of Robbins and Boese (2002). Macroalgal volume was measured to the
nearest 100 ml after macroalgae were compressed using a plunger to a point where interstitial
water had been removed. A volume of 50 ml (half the detection limit) was used when
macroalgae was present but not in sufficient quantity to be measured in the four-liter cylinder.
183
-------
Table 7-1. Sampling design for the probabilistic field surveys and areas of the probabilistic frames. Estuarine intertidal area is the
sum of the NWI "regularly flooded" estuarine polygons, which is approximately from MLLW to MHHW. The Probabilistic Frame is
the actual area that was sampled. The areas of the oceanic and riverine frames are the areas within the oceanic versus riverine
segments (Chapter 5) that resulted from the post-survey division of the estuaries. The samples taken in 2006 in the Alsea and
Tillamook are included in the analysis.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
NWI
ESTUARINE
REGULARLY
FLOODED
AREA
(km2)
6.97
26.96
2.59
1.56
24.6
7.45
7.87
PROBABILISTIC FRAME
TOTAL
AREA
(km2)
5.20
26.56
2.01
0.45
24.38
7.16
6.40
OCEANIC
FRAME
(km2)
4.77
21.25
1.95
0.37
14.50
2.95
5.26
RIVERINE
FRAME
(km2)
0.44
5.31
0.06
0.07
9.88
4.22
1.15
NUMBER
OF
SAMPLING
STRATA
3
5
2
2
5
5
2
NUMBER OF SAMPLES
TOTAL
109
101
101
100
97
99
100
OCEANIC
FRAME
50
86
99
52
56
55
48
RIVERINE
FRAME
59
15
2
48
41
44
52
YEARS
SAMPLED
2004/2006
2005
2004
2004
2005/2006
2005
2004
oo
-------
2.5m
PSA
o
TOC
o
Salinity
0
Figure 7-8. Diagram of the intertidal sampling site layout with the 2.5 x 2.5 m quadrat and the
0.25 m2 plant and shrimp quadrats used in the probabilistic surveys. PSA = Sample for particle
size analysis. TOC = sample for total organic carbon.
185
-------
DATE:
WEATHER:
TIME (PDT):
(Current Time)
CREW:
Northing:
SITE #:
Easting:
2.5 X 2.5 Meter Quadrat
Percent Cover
(from look up list)
Zostera
marina
Absent (0%) Very Sparse (1-10%) Sparse (10-40%)
Moderate (40-70%) Dense (70-100%)
Zostera (0%) (1-10%) (10-40%)
japonica (40-70%) (70-100%)
Other (0%) (1.10%) (10-40%)
Rooted
Plants (40-70%) (70-100%)
Type
Debris Type
Green
Macroalgae
Other
Algae
Type
(0%) (1-10%) (10-40%)
(40-70%) (70-100%)
(0%) (1-10%) (10-40%)
(40-70%) (70-100%)
Sediment Characteristics
Color:
Black
Tan/Brown
Red
Gray I ght/white Olive-green
mottled other
Firmness** VS S/L F/MD H/VD
**Sed. Firmness Classes: VS = very soft, S/L = soft/loose F/MD = firm/ med
Density, H/VD = Hard/very dense
Sediment NQ Qdor
Odor:
Marsh Rip-rap
Shoreline
Classes- Embankment F
Notes
Hydrogen sulfide Petroleum
Houses
orest Other
Indicators Of Unvegetated sediment B
_. . . . Worms Burrc
Biological
. .... Mounds Sea
Activity: A . . .
* Amphipods Othe
Wave Frequency
Sediment Temp. (C)
Crests/1 m
Road
Dunes
Iodine
Levee
NR
Agriculture
•own Diatoms Macroalgae
wing Shrimp Clams
grass Dead Shellfish
r
If Longer than 1 m:
Distance between
Crests
Sediment Cores Circle when
TOC/PSA
collected:
Salinity
NOTES:
Figure 7-9a. Front of the field sheet used to record estimates of habitat characteristics and
burrowing shrimp hole counts within the 2.5 x 2.5 m quadrat.
186
-------
SITE #:
Quadrat #2 Plants
Point Intercept
Zm
Zj
GM
Dominant Algae
Ulva
Enteromorpha
Other
Open
Other Algae:
Other Plant:
Volume (ml):
Z. marina Blade Measurements
Shoot
1
2
3
4
5
Length of
longest
Blade (cm)
Width of
longest
Blade (mm)
No. leaves
per shoot
Seagrass Shoot Counts Zm orZj
Type
Vegetative
Reproductive
Seedlings
Perennial
# of shoots
Count marked cell only
Quadrat #3 Burrowing Shrimp
Hole Counts
Neo :
Upo:
% w/o algae in shrimp quad
QA Neo :
QA Upo:
NOTES:
Figure 7-9b. Back of the field sheet used to record plant cover, point intercepts, shoot counts,
and blade widths in the 0.25 m2 plant quadrat and burrowing shrimp hole counts in the 0.25 m2
shrimp quadrat.
187
-------
Burrowing shrimp abundance - The number of burrow holes of TV. californiensis and U.
pugettensis were each counted within the shrimp quadrat. Macroalgae cover was recorded
within the shrimp quadrat as a possible correction factor for obscured burrow holes. At every
tenth site, the number of shrimp burrow holes was counted by a second crew member as a
quality assurance check.
Habitat measures - Sediment samples for grain size and total organic carbon and nitrogen
(TOC/N) analysis were collected within the large quadrat (Figure 7-8). An 8-cm x 20-cm PVC
core was inserted into the sediment within the 2.5 x 2.5 m quadrat to obtain a sample for particle
size analysis (PSA). The top 10-cm of the sediment in the core was collected into a clean
resealable plastic bag. The sediment sample for TOC/N analysis was collected in a 4-cm x 15-
cm core inserted into the sediment adjacent to the PSA core and the top 5 cm were collected. A
sediment sample was collected for water content by scraping the sediment surface into a 50-cc
centrifuge tube. All sediment samples were stored on ice immediately after collection and frozen
as soon as was practical.
Sediment particle size was determined using a Beckman Coulter LS-100Q laser diffraction
particle size analyzer. The data are presented as percent fines which is the sum of percent silt
and clay. The TOC/N samples were analyzed using a Carlo Erba EA 1108 elemental analyzer
according to manufacturer's instructions as modified for sediment. The surface sediment sample
collected in the centrifuge tube was centrifuged to extract interstitial water, and salinity of the
supernatant water was measured with a bench-top salinity meter.
Sand ripples and waves were measured by laying a meter stick on the sediment surface. If sand
ripples were present, the number of wave crests along the 1-meter distance was counted. If the
sand ripples were larger than a meter, the distance between crests was measured using a 100-
meter transect tape. Data recorded in number of crests per meter were converted to distance
between crests. Sand waves were only recorded in the 2005 and 2006 field surveys. At each
sampling location, photographs were taken of the quadrats, including the meter stick or transect
tape to document surface structure. In addition, the type of shoreline development was visually
assessed, recorded, and photographed.
Integrated Measure of Presence/Absence (P/A) ofSAV, macroalgae, and burr owing shrimp and
bare habitat - Several different techniques were combined to obtain an "integrated" measure of
the presence/absence of Z. marina, Z.japonica., benthic green macroalgae, burrowing shrimp,
and other plant species over a 2.5 x 2.5 m area. It was first determined if the taxon was recorded
as present in either the 0.25 m quadrat or the 2.5 x 2.5 m quadrat. If the taxon was not recorded
in either quadrat, then the field notes and the site photographs were evaluated for indications of
the presence of the taxon within the large quadrat. The taxon was recorded as present if it was
observed in the site photograph. Use of the integrated P/A reduced any errors from
misclassifications in the field, and is used as the best indicator of a species presence or absence.
"Bare habitat" was defined as any 2.5 x 2.5 m quadrat that did not contain any Z. marina,
Z.japonica, benthic green macroalgae, or burrowing shrimp.
188
-------
7.1.4 Comparison of 0.25 m2 versus 2.5 x 2.5 m Quadrats
r\
As part of this study, we evaluated the efficacy of using 0.25 m (small) versus 2.5 x 2.5 m
(large) quadrats in estimating the distribution and abundance of seagrasses. The 0.25 m2
quadrats are small enough to allow point-intercept estimates or counts of seagrass shoots. The
2.5 x 2.5 m quadrats have the advantages that they are same size as the minimum mapping unit
in the aerial photography and are of sufficient size to "average out" small scale patchiness. Both
methods are valuable, but the results need to be interpreted in light of the differences in scale.
For all seven estuaries, estimates of the area occupied by Z. marina and Z.japonica from the
point intercepts were lower than those from the large quadrats (Tables 7-3 and 7-4). The larger
quadrats appear to generate a more comprehensive estimate of seagrass distribution because they
have a higher "capture" probability for relatively sparse species. Given this higher capture
probability, the large quadrats were judged the preferred approach to estimating the area
occupied by the two seagrasses. While the quantitative estimates of the density class from the
large quadrat are preferred over presence/absence, the percent cover classes were only collected
in the large quadrats in 2005 and 2006. Accordingly, the integrated measure of presence/absence
in the 2.5 x 2.5 m quadrats was used as the primary method of quantifying the intertidal area
occupied by Z. marina and Z. japonica. The point-intercept and shoot-count methods were used
in a supplemental fashion to estimate seagrass density.
As discussed in Chapter 6, the aerial photography has a detection limit of approximately 10%
seagrass cover within a 2.5 x 2.5 m minimum mapping unit. Therefore, the most direct
comparison between the probabilistic and the aerial surveys are the estimates of seagrass area
from the >10% density class from the large quadrats used in Tillamook, Coos, and the Umpqua.
7.2 Across-Estuary Patterns
7.2.1 Sediment Grain Size and TOC in the Target Estuaries
Two key sediment characteristics quantified were the percent total organic carbon (TOC) and
grain size, which was summarized as percent fines. The "muddiest" estuaries were the Yaquina
and Coos with 50% of the area having percent fines >30% (Table 7-2, Figure 7-10). The
sandiest estuary was the Nestucca with median percent fines of less than 2%. The Umpqua,
Salmon, Alsea, and Tillamook estuaries were intermediate with median percent fines ranging
from about 8% to 20%.
The median values for the percent total organic carbon (TOC) generally mirrored the results for
percent fines with Yaquina and Coos having the highest median values and Nestucca the lowest
(Table 7-2, Figure 7-11). The National Coastal Condition Report II (U.S. EPA, 2004b) used
TOC values >5% as an indication of "poor" sediment condition and values between 2% and 5%
as an indication of "fair" condition. Only 2 out of almost 700 stations across all the estuaries
exceeded 5% TOC, with a maximum value of 5.3% in the Coos Estuary. All seven estuaries had
TOC values >2% and <5%, though most values were at lower end of the "fair" range (<3%).
The highest coverages within this range occurred in the Yaquina and Coos, where 21% and 17%
of the intertidal sites were classified as "fair", respectively. Based on the minimal development
in the associated watersheds (Chapter 2) and the lack of major point sources, these higher TOC
values do not appear to be the result of anthropogenic loading.
189
-------
Table 7-2. Sediment characteristics of the target estuaries from the probabilistic surveys. Median values based on the areal extent of
the parameter in the intertidal. The 75th percentile was used for distance between wave crests because all the median values were 0.
Total organic carbon (TOC) values between 2% and 5% are classified as "fair" in the National Coastal Condition Report II (U.S. EPA
2004b) while values >5% are classified as "poor".
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
MEDIAN
% FINES
12.1
30.7
1.9
7.9
12.4
19.7
37.9
MEDIAN
% TOC
0.69
0.90
0.16
0.34
0.41
0.51
1.30
% OF AREA
WITH
2%< TOC <5 %
2.5
17.0
7.1
7.3
2.6
9.6
21.4
% OF AREA
WITH
TOC >5%
0
<1
0
<1
0
0
0
MEDIAN
%N
0.06
0.07
0.02
0.07
0.04
0.05
0.10
75th
PERCENTILE OF
DISTANCE
BETWEEN
WAVE CRESTS
(cm)
NA
<1
NA
NA
12
14
NA
MEDIAN
SEDIMENT
SALINITY
(psu)
28.0
29.7
29.6
32.0
25.1
19.4
31.2
VO
o
-------
100 -
90 -
-------
VO
to
CS
OJ
OJ
OJ
u
O)
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
10
0
0
% TOC
Figure 7-11. Cumulative distribution function (CDF) of the percent total organic carbon (TOC) in the intertidal zone of the seven
target estuaries. The nine samples taken in the Alsea in September, 2006 are not included. The red dashed vertical lines indicates the
2% and 5% TOC thresholds used in the National Coastal Condition Report II (U.S. EPA, 2004b) with "fair" sediment conditions
defined as between 2% and 5% and "poor" sediment conditions as >5%.
-------
7.2.2 Among-Estuary Distribution of Zostera marina
Tillamook had the greatest areal extent of Z. marina, occurring over 34% of the intertidal zone
(Table 7-3), while the Coos and Yaquina had coverages of about 12-17%. In contrast, the
Umpqua had only about 6% coverage and no Z. marina was found in the probability surveys in
the Alsea, Salmon, or Nestucca estuaries. Though the probability surveys did not detect
Z. marina in these estuaries, the aerial photography detected small amounts of Z. marina in the
Alsea and on-ground mapping found small beds in the Salmon and Nestucca (see Table 6-3). As
with any sampling method, the probabilistic survey has a detection limit and even a survey of
100 sites per estuary can fail to detect seagrasses with high confidence at low densities (e.g.,
Bailey et al., 1998). While the probabilistic surveys failed to detect these small amounts of
seagrass, the surveys were qualitatively correct in classifying these estuaries as having low
coverage.
The relatively wide expanses of seagrass found in Tillamook, Coos, and Yaquina as well as the
large expanses reported in Willapa Bay and Grays Harbor (e.g., Wyllie-Echeverria and
Ackerman, 2003) suggest that moderate to large tide-dominated estuaries contain proportionally
large areas of suitable habitat for Z. marina. In comparison, river-dominated estuaries have
substantially less Z. marina habitat. Only small beds occur in the Alsea, Nestucca, and Salmon
estuaries, while the Umpqua only contains one-half to one-sixth the percent cover compared to
the tide-dominated estuaries (Table 7-3). Estuarine size alone does not explain this pattern, as
the Umpqua and Alsea are roughly comparable in size to the Coos and Yaquina, respectively.
Additionally, the Netarts Estuary, which is smaller than either the Yaquina or Alsea, has
extensive seagrass coverage (Kentula and Mclntire, 1986; Wyllie-Echeverria and Ackerman,
2003).
The occurrence of this pattern is supported by the low percent cover of NWI aquatic beds in the
river-dominated estuaries. The average percent cover of NWI "aquatic beds" classes in the 22
river-dominated estuaries was 1.8% compared to 21.5% in the 6 tide-dominated estuaries (see
Table 2-3). Limiting this comparison to estuaries >10 km2, the percentages were 6.1% versus
25.7% in river- and tide-dominated estuaries, respectively. Though not as large a difference, the
same pattern was observed in the mid-1970's survey of Oregon estuaries (Cortright et al., 1987).
In this study, the average percent cover for seagrass and seagrass/algae classes was 15.7% in the
three tidal estuaries compared to 7.8% in eleven river-dominated estuaries. It appears, then, that
our definition of tide- and river-dominated estuaries helps to separate estuaries that support more
extensive Z. marina populations.
We hypothesize that the most important drivers resulting in these among-estuary patterns are
differences in average salinity values and, perhaps more importantly, differences in the temporal
variation in salinity. In Puget Sound, Z. marina appears to grow best at 20-32 psu (Phillips,
1984), and higher growth rates occurred at 30 psu compared to 10 psu in experiments (Thorn et
al., 2001, 2003). Additionally, diversions of fresh water into estuaries have been associated with
declines in seagrass (Estevez, 2000), while Montague and Ley (1993) found that increasing
salinity variation was associated with declines in seagrass biomass in Florida. Coastal PNW
estuaries are subject to substantial seasonal salinity variations in response to the seasonal pattern
in precipitation, with winter salinities dropping below 10 psu or even 5 psu over much of the area
193
-------
of many estuaries (e.g., ORDept. State Lands, 1985; Brown, unpubl. data; Figure 2-11). In
particular, river-dominated estuaries appear to experience greater seasonal variations than tide-
dominated systems (Figure 2-11; ORDept. State Lands, 1985; Haertel and Osterberg, 1967).
These extreme seasonal fluctuations are well illustrated in the Salmon River where differences in
summer and winter median salinities approach 20 psu and where much of the estuary has a
salinity <5 psu during the winter (Figure 2-11). In the Umpqua, much of the estuary has a winter
salinity <10 psu (Figure 2-11). In addition to this seasonal variability, dry season salinity near
the mouths of PNW estuaries can vary 10-25 psu over a tidal cycle (Figures 4-9 and 4-10). The
extent of this tidal variation is related to the normalized river inflow (Figure 4-11), with the
river-dominated systems demonstrating about a 2- to 10-fold greater variation in high and low
tide salinities than the tide-dominated estuaries. The extensive tidal variation within the Alsea
(Figure 4-10) may contribute to the paucity of seagrass in that system even though seasonal
fluctuations are not as great as in the Salmon or Umpqua estuaries. At the opposite extreme, the
occurrence of extensive seagrass beds in Netarts Bay (Kentula and Mclntire, 1986), which is
characterized by minimal seasonality in salinity (Figure 2-11), is consistent with the hypothesis
that salinity variations limit seagrass abundance and distribution.
A different and potentially complementary mechanism reducing Z. marina in river-dominated
systems is the greater current energy and sediment transport in these high flow systems, as
suggested in Section 6.2.2. Yet another possible contributing factor may be the presence of the
burrowing shrimp N. californiensis. N. californiemis can inhibit Z. marina (Dumbauld and
Wyllie-Echeverria, 2003), and its extensive coverage in both the Salmon and Nestucca estuaries
(Figure 7-19) may make it more difficult for Z. marina to become established. While these
suggestions need to be evaluated further, several independent data sets indicate that river-
dominated systems within the PNW have reduced Z. marina cover compared to tide-dominated
and bar-built estuaries.
7.2.3 Among-Estuary Distribution of Zosterajaponica
The regional distribution of the nonindigenous Z.japonica differed from the native seagrass in
several respects. Z.japonica was found in six of the target estuaries in the probability surveys,
including the Salmon and Nestucca (Table 7-4). The only target estuary where Z. japonica was
not found was the Alsea, and we are unaware of any reports of this species in the Alsea. In
contrast to Z. marina, Z.japonica occupied moderately large areas in both tide-dominated
estuaries, such as the Coos, and the river-dominated Umpqua. Besides these six estuaries,
Z. japonica has been reported in a wide range of PNW estuaries including the river-dominated
Columbia, Siletz, Necanicum, and Coquille estuaries; the tide-dominated Willapa and Grays
Harbor estuaries; and the bar-built Humboldt and Netarts (Lee and Reusser, 2006). In the
probabilistic surveys, the intertidal extent of Z.japonica actually exceeded that of Z. marina in
the Salmon, Nestucca, Coos, and Umpqua estuaries. Coverage of Z.japonica in these four
estuaries ranged from 4% to 23% compared to 0% to 12% for Z. marina based on the integrated
measure of presence/absence (Tables 7-3 and 7-4). This comparison is only for the intertidal
populations, and inclusion of the shallow subtidal Z. marina could result in more similar
percentages among the two species since Z.japonica is rare or absent in the shallow subtidal in
these estuaries.
194
-------
Table 7-3. Percent of intertidal area with Z. marina estimated using different methods. Estimates for presence/absence (P/A) integrate
several types of data over the 2.5 x 2.5 m quadrat (see Section 7.1.3). Estimates for "Large Quadrat" are the percent of the intertidal
area with >1% cover or >10% cover within the 2.5 x 2.5 m quadrats. Estimates from the point intercepts are the areas where there was
at least one intercept with Z. marina in the 0.25 m plant quadrats. NA = not available.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
# SITES WITH
Z. MARINA
P/A
0
12
0
0
28
8
11
% INTERTIDAL
AREA
P/A
0
11.7
0
0
34.2
5.5
17.4
% INTERTIDAL AREA
LARGE QUADRAT
COVER >1%
NA
10.4
NA
NA
34.2
5.5
NA
% INTERTIDAL AREA
LARGE QUADRAT
COVER > 10%
NA
2.4
NA
NA
15.6
0.5
NA
% INTERTIDAL AREA
POINT INTERCEPT >0
0
1.4
0
0
18.6
0.5
12.2
VO
Table 7-4. Percent of intertidal area with Z.japonica estimated using different methods. Estimates for presence/absence (P/A)
integrate several types of data over the 2.5 x 2.5 m quadrat (see Section 7.1.3). Estimates for "Large Quadrat" are the percent of the
intertidal area with >1% cover or >10% cover measured within 2.5 x 2.5 m quadrats. Estimates from the point intercepts are the
percent intertidal areas where there was at least one intercept with Z.japonica in the 0.25 m2 plant quadrats. NA = not available.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
# SITES WITH
Z JAPONICA
FROM P/A
0
17
19
O
9
22
18
% INTERTIDAL
AREA
P/A
0
19.4
23.4
3.6
10.5
20.7
11.9
% INTERTIDAL AREA
LARGE QUADRAT
COVER>1%
NA
19.4
NA
NA
10.5
19.3
NA
% INTERTIDAL AREA
LARGE QUADRAT
COVER > 10%
NA
10.0
NA
NA
5.0
3.2
NA
% INTERTIDAL AREA
POINT INTERCEPT >0
0
12.3
8.6
2.0
4.3
9.0
7.5
-------
Zoster a japonica has obtained these extensive within- and among-estuary distributions within the
last 30 to 50 years since it was first recorded on the Pacific Coast in 1957 and in Yaquina in 1976
(Harrison and Bigley, 1982; Bayer, 1996). Further, Z. japonica is continuing to rapidly expand
its distribution within portions of the Yaquina (Young et al., 2008). While it is apparent that
Z. japonica is expanding, the ecological consequences of this expansion are not clear. Both the
native and nonindigenous seagrass species occurred within the same coastal PNW estuaries we
surveyed, but they are not presently competing within these estuaries. Of the 138 occurrences of
seagrasses in the probabilistic surveys, only 2 sites (1.5%) contained both species. The low
overlap of the two species is due to their different intertidal distributions. Z. marina was
primarily found in the shallow subtidal and lower intertidal zones, while Z. japonica tended to
colonize a band higher in the intertidal and, based on our observations, around freshwater
streamlets (also see Kaldy, 2006; Ruesink, 2006). However, as Z. japonica continues to expand,
it is possible that the two species will overlap as reported for areas in Puget Sound and British
Columbia (Nomme and Harrison, 1991) and in Netarts (Dudoit, 2006), and then potentially
compete for space and light (Wonham, 2003; Bando, 2006). Additionally, as Z. japonica
expands it could eventually become sufficiently abundant to have an effect on primary
production and other components of estuarine food webs (e.g., Posey, 1988; Wonham, 2003;
Ruesink et al., 2006; Kaldy, 2006).
7.2.4 Composition and Seasonally of Benthic Green Macroalgae
Determining benthic green macroalgal distribution and abundance presents two additional
challenges compared to the seagrasses. The first is that green macroalgae are composed of
several species. In the Yaquina Estuary in 2001, benthic green macroalgae was found to be
comprised on average of taxa most closely resembling Ulva lima: -60%; U. fenestrata: -30%;
U.flexuosa: -10%; U. intestinalis: < 5% (D. Young, unpubl. data). Given the difficulty in
separating the species, all the green macroalgae are treated as a single taxon.
Another issue is the seasonality of the macroalgal blooms, which could potentially confound
differences among estuaries if they are sampled at different times. To address macroalgae
seasonality, we utilized data from a 1999 survey in the Yaquina that quantified seasonal changes
to derive a seasonal adjustment for biomass. The 1999 survey measured percent cover and
biomass of Z. marina and benthic green macroalgae monthly from June through December as a
function of distance from estuary mouth and elevation above MLLW at six sites within the lower
Yaquina Estuary (D. Young, unpubl. data). For this analysis, we took the mean of the monthly
averages of both percent cover and biomass (grams dry weight [gdw] m" ) for all six sites. These
data illustrate the strong seasonality in macroalgae in both cover (Figure 7-12) and biomass
(Figure 7-13), which increase from June through September and decline rapidly after October.
At its peak in September and October, the average macroalgal dry weight exceeded 200 gdw m"
and average surficial cover exceeded 50% at these non-random stations.
While aware of this seasonality, the extensive logistic effort required in the probabilistic surveys
necessitated sampling from early June through mid September (Figure 7-14). The Yaquina
Estuary was sampled earliest in the growing season, while the Alsea Estuary was sampled during
September, the period of peak macroalgal biomass (Figure 7-14). To minimize underestimation
of macroalgal occurrence during the initial phase of its growth spurt, the integrated
196
-------
0
June
Aug
Sept Oct
Month
Nov
Dec
Figure 7-12. Average percent cover values (± 1 std. error) of benthic green macroalgae at six
sites in the Yaquina Estuary during 1999.
300
S 250 -
-o
OX)
200 -
o 150 -
M
Jf> 100 -I
o
o
CS
OX)
50 -
0
T
T
June July Aug
Sept Oct
Month
Nov
Dec
Figure 7-13. Average biomass values (± 1 std. error) of benthic green macroalgae at six sites in
the Yaquina Estuary during 1999.
197
-------
1.2
1 i-o-l
S
o
2 0.8 H
I 0.6 -|
e
o
0.4 -
! 0.2 -I
o.
2
0.0
Y = -0.000128 X + 0.0616 X - 6.41
r2 = 0.84
Nestucca^-'
Tillamook / ^ Coos
Umpqua Salmon
/ Yaquina
150
200
T
250
Julian Day
300
350
Figure 7-14. Average proportions of green benthic macroalgal biomass relative to the September
peak measured in the Yaquina Estuary during 1999. Dates are given in Julian days. The solid
line connects the monthly averages while the dashed line is the polynomial regression fit to the
monthly data. The polynomial regression was used to derive the seasonally adjusted macroalgal
biomass in all of the target estuaries based on sampling date. The horizontal gray bars indicate
the sampling dates for each target estuary.
presence/absence from the 2.5 x 2.5 m quadrats was used as the primary data source for
distribution since its larger sample area is more likely to detect sparse populations. To correct
for seasonality in biomass, we derived a polynomial regression from the 1999 Yaquina dataset
relating the relative percentage of biomass to the monthly maximum value in September
(Figure 7-14). This polynomial was used to derive seasonally adjusted biomass estimates based
on actual sampling dates with the assumption that similar seasonal patterns occur in all Oregon
estuaries. This approach generated estimates of standing stock within the range previously
observed in Yaquina in six of the estuaries. However, the model predicted biomasses greater
than 12,000 gdw m"2 at a few Yaquina sites, several-fold higher than have been observed in our
previous studies or reported in other studies in Yaquina (Davis, 1981; Kentula and DeWitt, 2003;
D. Young and B. Boese, unpubl. data). It is possible that the model over predicted biomass at
these stations since these predictions utilized the steepest portion of the adjustment curve
(Figure 7-14). In any case, the maximum seasonally adjusted values in the Yaquina need to be
evaluated cautiously
198
-------
7.2.5 Among-Estuary Patterns of Benthic Macroalgae
Six of the target estuaries had areal coverages of macroalgae in the range of 41% to 64% of the
intertidal zone while macroalgae occurred over 22% of the intertidal zone in the Umpqua
(Table 7-5). Benthic macroalgae has been reported from a number of other coastal estuaries in
the PNW as well as from Puget Sound (Phillips, 1984; Thorn, 1984; Kentula and Mclntire, 1986;
Nelson and Lee, 2001; Thorn et al., 2003), and we have observed macroalgal mats in a number
of additional moderate-sized estuaries such as the Siletz. The presence of benthic macroalgae in
systems ranging from Puget Sound to the Salmon Estuary indicate that macroalgae is a wide-
spread feature of PNW estuaries greater than about 3 km2 in size. Limited observations suggest
that green benthic macroalgae is sparse or absent in estuaries smaller than about 0.5 km2, while
additional observations are needed for estuaries between 0.5 and 3.0 km2 in size.
At a sufficiently high density, macroalgae can result in anoxic sediment conditions, smother
benthos and seagrasses, and reduce light availability for seagrasses (e.g., Hull, 1987). In lieu of
an established threshold relating percent cover to these effects, we use >70% cover as the
threshold for "high" macroalgal cover. At this percent cover, macroalgae "blanket" the sediment
surface though we have no direct evidence it results in detrimental effects to seagrasses. Based
on the point intercepts (Figure 7-15), the estuaries divided into three groups with the Umpqua
having no high macroalgal cover, the Alsea, Coos, Nestucca, Salmon, and Tillamook estuaries
having moderate extents (4-7%) of high cover, and the Yaquina Estuary having an extensive
area (18%) of high macroalgal cover. This high cover occurred within the Yaquina Estuary
even though it was sampled the earliest during the macroalgal growth season (Figure 7-14).
Results from the 2.5 x 2.5 m quadrats show a similar pattern in the three estuaries where the data
are available. High macroalgal cover occurred over 6-8% of the intertidal area in the Coos and
Tillamook estuaries and was not detected in the Umpqua (Figure 7-16).
Another indicator of potential macroalgal impact is standing stock, with a value of 100 gdw m"2
identified as a potential threshold for impacts on seagrasses in Chesapeake Bay (Bricker et al.,
2003). This threshold is in the general range of observed effects from an experimental removal
of the green macroalgae Ulvaria obscura on shoot production in subtidal Z. marina in Puget
Sound (Nelson and Lee, 2001). In this experiment, sequential removals of about 144 gdw m"2,
40 gdw m"2, and 115 gdw m"2 over a month resulted in a slower decline in shoot production than
in the controls. The percentages of the intertidal area exceeding this threshold using the
unadjusted biomass (Figure 7-17) and seasonally adjusted biomass (Figure 7-18) were similar in
six of the target estuaries (Table 7-6). In the Yaquina Estuary, there was about a two-fold
difference between the estimates, with about 20% of the intertidal exceeding the threshold using
the unadjusted biomass versus about 42% using the seasonally adjusted values. Based on either
the unadjusted or adjusted values, the target estuaries break into three relatively distinct groups.
The first is the Yaquina Estuary with a high percentage of the intertidal zone (20-42%)
exceeding the biomass threshold, which is consistent with the pattern based on percent cover
(Figure 7-15). The second group consists of the Alsea, Coos, Nestucca, and Tillamook estuaries
with a moderate percentage (4-6%) of the intertidal exceeding the biomass threshold. The third
group consists of the Salmon and Umpqua estuaries which had minor areas (<1%) exceeding the
threshold. The Salmon differed from the Umpqua in having an extensive area of macroalgal
cover though the biomass did not reach high standing stock.
199
-------
Table 7-5. Percent of intertidal area with green benthic macroalgae estimated using different methods. Estimates for
presence/absence (P/A) integrate several types of data over the 2.5 x 2.5 m quadrat (see Section 7.1.3). Estimates for "Large Quadrat"
are the percent of the intertidal area with >1% cover or >10% cover measured within 2.5 x 2.5 m quadrats. Estimates for the point
intercept are intertidal areas where there was at least one intercept with macroalgae in the 0.25 m plant quadrats. NA = not available.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
# SITES WITH
MACROALGAE
FROM P/A
44
60
44
71
65
26
35
% INTERTIDAL
AREA
P/A
54.4
54.9
41.4
56.2
63.6
22.2
58.4
% INTERTIDAL AREA
LARGE QUADRAT
COVER>1%
NA
53.7
NA
NA
54.1
22.1
NA
% INTERTIDAL AREA
LARGE QUADRAT
COVER>10%
NA
10.4
NA
NA
14.9
3.2
NA
% INTERTIDAL
AREA
POINT INTERCEPT >0
35.2
18.3
12.9
43.2
17.8
2.6
48.2
to
o
o
Table 7-6. Biomass (MB) estimates for benthic macroalgae. The percent of the intertidal area withMg >100 gdw m"2 was interpolated
from the benthic macroalgal cumulative distribution functions (CDFs, Figures 7-16 and 7-17). The number of sites with unadjusted
and seasonally adjusted MB >100 gdw m"2 indicates the number of sampling sites that increased above the threshold after the seasonal
adjustment. Note that the slightly different percent areas with the unadjusted and adjusted values, even though there are the same
number of samples, is due to interpolating values from different CDF curves.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
# SITES WITH MB
MEASUREMENTS
(MB >0 gdw m"2)
24
31
14
50
38
12
31
# UNADJUSTED
SITES WITH
MB > 100 gdw m"2
3
6
5
0
11
1
12
# SEASONALLY
ADJUSTED SITES WITH
Ms>100gdwm"2
O
6
5
0
12
2
26
% INTERTIDAL
AREA WITH
UNADJUSTED
Ms>100gdwm"2
4.0
4.9
3.9
0
5.8
<0.1
18.0
% INTERTIDAL AREA
WITH SEASONALLY
ADJUSTED
MB > 100 gdw m"2
5.7
5.5
4.0
0
6.6
0.3
42.3
-------
to
CS
-------
Coos
Tillamook
Umpqua
0
0
1 - 10% 10 - 40% 40 - 70%
Macroalgae Density Class
>70%
Figure 7-16. Percent of the intertidal area covered by benthic macroalgae by density class. The
data are from estimates in the 2.5 x 2.5 m quadrats in the 2005 and 2006 surveys. The threshold
for "high" macroalgal cover is >70%.
The overall pattern that emerges is that seasonal macroalgal blooms are a common feature across
a range of different types and sizes of PNW estuaries. However, estuaries vary several fold in
the extent of their intertidal areas with high percent cover (>70%) or standing stocks
>100 gdw m" , with the highest values occurring in the Yaquina. One possible reason for the
high macroalgal biomass in the Yaquina Estuary is its high nutrient concentrations as indicated
by having the highest wet season dissolved inorganic nitrogen (DIN) concentrations as well as
the second highest cover of red alder in its watershed (Table 4-3). It is possible that the
macroalgae respond to this high nitrogen concentration early in the growing season when light
limitation is removed, rapidly building up a high biomass. The importance of DIN is further
suggested by the system with the lowest cover of macroalgae (Umpqua) also having the lowest
wet season DIN (Table 4-3). Regardless of the cause(s), macroalgae may reach biomasses
sufficient to adversely impact seagrasses over a moderate to substantial portion of the intertidal
in the Yaquina Estuary and over lesser areas in the Alsea, Coos, Nestucca, and Tillamook
estuaries, assuming the Chesapeake Bay biomass threshold is appropriate for the PNW. As
discussed in Section 7.4, these blooms appear to be a natural phenomenon and not an indication
of cultural eutrophication.
202
-------
to
CS
-------
to
CS
-------
7.2.6 Among-Estuary Patterns of Burrowing Shrimp
The spatial extent and density of burrowing shrimp were measured as burrow hole counts in the
0.25 m2 quadrats, visual estimates using five density classes in the 2.5 x 2.5 m quadrats, and
from an integrated measure of presence/absence. Because of their high density, the burrow hole
counts in the 0.25 m2 quadrats were an effective method to determine both the presence and
density of both species.
Neotrypaea californiensis beds were a major feature in all the target estuaries, occurring over
21% to 74% of the intertidal area (Figure 7-19). The Salmon River had both the greatest area of
the intertidal occupied by N. californiensis and the densest population, with burrow hole counts
exceeding 100 per 0.25 m2 at several sites. Neotrypaea also occupied a large extent of the
intertidal in the Nestucca though populations were not as dense as in the Salmon. The areal
extent and abundance of this burrowing shrimp in the other five estuaries varied from the largest
in the Alsea to the lowest in the Coos Estuary. The other burrowing shrimp, U. pugettensis,
showed a different pattern among the estuaries (Figure 7-20). Maximum coverage of U.
pugettensis occurred in the Alsea with 53% of the intertidal area occupied and a maximum
density exceeding 50 burrow holes per 0.25 m2. The Coos, Nestucca, Tillamook, and Yaquina
estuaries had moderate coverages, ranging from 17% to 33% of the intertidal area. In
comparison, U. pugettensis occurred over less than 4% of the area in the Salmon and none were
found in the Umpqua.
Our results and other studies (e.g., Feldman et al., 2000; DeWitt et al., 2004) demonstrate that
burrowing shrimp are major components of the intertidal zones of estuaries larger than about 3
km2. Of the 698 quadrats taken in the target estuaries, one or both of the burrowing shrimp
occurred in 424 samples (61%). The absence of U. pugettensis in the Umpqua and its low
coverage in both the Salmon and Nestucca (Figure 7-20) suggests that this species is less
abundant in highly river-dominated estuaries, which is consistent with an earlier assessment that
concluded that N. californiensis and U. pugettensis were probably absent from the Rogue
Estuary (ODFW, 1979). The among-estuary pattern for TV. californiensis does not appear to be
as closely linked to extent of freshwater flushing. It has low coverage in the river-dominated
Umpqua but high coverage in two other river-dominated estuaries, the Salmon and Nestucca, and
its among-estuary distribution may be controlled by a number of factors. In terms of estuary
size, limited observations suggest that both burrowing shrimp are sparse or absent in estuaries
less than about 0.5 km2, while additional observations are needed for estuaries between 0.5 and
3.0km2.
Both species directly and indirectly affect estuarine food webs within the PNW. Various fishes,
birds, and even whales prey on these species (e.g., Stenzel et al., 1976; Armstrong et al., 1995;
Feldman et al., 2000; Dumbauld et al., 2008). Perhaps more important are the indirect effects
these bioengineering species have on nutrient fluxes and phytoplankton. The intense
bioturbation and irrigation activities of these species increase the benthic flux of nitrogen
(D'Andrea and DeWitt, 2003), which contributes to the total estuarine nutrient loading (Brown
and Ozretich, 2009). At the same time, these species filter large quantities of the overlying water
(Griffen et al., 2004), potentially decreasing phytoplankton concentrations and turbidity.
205
-------
to
CS
-------
to
S 70
OJ
e
OJ
u
O)
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
0
0
Density (# of holes per 0.25 m )
Figure 7-20. Cumulative distribution function (CDF) of the number of U. pugettensis burrows holes per 0.25 m in the seven target
estuaries.
-------
The overall effect of these species on nutrient and phytoplankton concentrations will depend
upon a number of factors, including the relative proportion of the two species and the among-
estuary differences in the advection of oceanic nutrients and phytoplankton. It is important to
note, however, that the role of U. pugettensis in the regional food webs may be diminished in
response to a recent invasion by a parasitic bopyrid isopod (Orthione griffenis) (Markham,
2004), which may substantially reduce the density of the host shrimp within PNW estuaries
(Smith et al., 2008).
7.2.7 Bare Habitat
The last benthic habitat quantified was "bare habitat", which was defined as habitat devoid of
any seagrasses, macroalgae, or burrowing shrimp in the 2.5 x 2.5 m quadrats. As used here, it is
meant to capture the area of the estuary that is unsuitable or at least poor habitat for these five
ecologically important benthic taxa. It is not meant to imply that it is poor habitat for all species,
and a number of benthic infauna characteristically inhabit "bare sediment" in the PNW (Ferraro
and Cole, 2007). The coverage of bare habitat varied several fold, with only 1% of the intertidal
zone of the Nestucca not having identifiable biotic structure compared to over 20% of the
intertidal zone in both the Coos and Umpqua estuaries (Figure 7-21). These data indicate that at
least one of the five target taxa occupied >70% to 99% of the intertidal area of these PNW
estuaries.
7.2.8 Multivariate Analysis of the Target Estuaries
The differences in the areal extents of the benthic habitat types presented above allow managers
to group the estuaries based on the extent of a single resource (e.g., Z. marina) or on a potential
stressor (e.g., macroalgae). Another approach is to evaluate the overall similarity in the patterns
of these benthic habitat types. Since there are a limited number of estuaries and variables
(habitat types), ordination by non-metric multidimensional scaling (nMDS; Clarke and Warwick,
2001; McCune and Grace, 2002) was used instead of clustering to evaluate relationships. The
nMDS analysis was conducted with Primer6 (Clarke and Gorley, 2006;
http://web.pml.ac.uk/primer/primer6.htm) on the percent area occupied by the five target taxa
and the bare habitat in each of the estuaries using the Bray-Curtis similarity index.
The nMDS ordination had a very low stress (<0.01) indicating that the relationships among the
estuaries are well represented in two-dimensional space (Figure 7-22). The tide-dominated
Coos, Tillamook, and Yaquina estuaries form a group with a very high degree of similarity
(>80%), with the Alsea associated with this group at >75% similarity. The river-dominated
Nestucca and Salmon estuaries form a second cluster with high similarity. These six estuaries
form a single group at a moderately high (65%) similarity, with the Umpqua separate from the
other estuaries. This pattern indicates that tide-dominated estuaries have very similar relative
distributions of the six benthic habitats. The river-dominated Alsea is more closely aligned to
this group than to the smaller river-dominated Nestucca and Salmon, suggesting that there are
physical drivers related to estuary size or geomorphology that effect the biotic composition of
river-dominated estuaries. The separation of the Umpqua suggests that larger highly river-
dominated estuaries have different proportions of these intertidal benthic habitat types, in part
because of the low areal extent of macroalgae and absence of U. pugettensis.
208
-------
This analysis by benthic habitat type is based on a finer breakout of the intertidal habitat types
than the classifications by NWI wetland classes (Chapter 2). In particular, differentiating the two
burrowing shrimp habitats from unvegetated habitat and separating Z. marina and Z. japonica as
discrete habitats offers a much finer resolution of the regularly flooded NWI wetland habitats.
However, this analysis does not include the subtidal or emergent wetland classes that were
included in the NWI classification. Even with these differences, there are a number of
similarities. The Coos, Yaquina, and Alsea estuaries showed a high degree of similarity and the
Umpqua was separated from the tide-dominated and moderately river-dominated estuaries in
both analyses (Figures 2-14d and 7-22). The major difference is that the Salmon River showed
little similarity with the other estuaries based on the NWI analysis in comparison to its relatively
high similarity in the present analysis. The Salmon River separated from the other PNW
estuaries in the NWI analysis largely based on its high percentage (61%) of irregularly exposed
marshes, habitats not included in the present analysis.
30 -
•3 20
."2
'-S
-*•>
c
hH
<+•
® 10
c
u
0
28.1
21.5
6.6
16.4
11.8
1.5
7.1
Alsea Coos Nestucca Salmon Tillamook Umpqua Yaquina
Figure 7-21. Percent of intertidal area occupied by bare habitat as defined by the absence of
Z. marina, Z. japonica, benthic green macroalgae, N. californiensis, or U. pugettensis in the 2.5 x
2.5 m quadrats.
209
-------
Resemblance: S17 Bray Curtis similarity
to
2D Stress: 0.01
Similarity
65
75
or\
OU
Figure 7-22. Non-metric multidimensional scaling (nMDS) of the seven target estuaries based on the percent of intertidal area
occupied by Z. marina, Z.japonica, benthic macroalgae, N. californiensis, U. pugettensis, and bare habitat. Contours join estuaries at
different levels of similarity as measured by the Bray-Curtis similarity index.
-------
The comparison of the classifications based on benthic habitats versus NWI wetland classes
highlights the fact that different schema can be generated depending on the suite of estuaries and
variables used in the analysis. While identifying which is the "best" classification depends upon
the goals of the user, a general characteristic of a successful classification is that it maximizes
differences among groups while minimizing differences within groups. As discussed in
Section 2.9, one approach to identifying similar estuaries for evaluation of within-group
variability is to focus on estuaries that clustered together by more than one method. Of the seven
target estuaries, the Coos and Yaquina estuaries show high similarity based on benthic habitats
(Figure 7-22) and in the NWI and land cover crosswalk (Table 2-8). To the extent that these
classifications are related to nutrient dynamics, these two estuaries should have similar nutrient
and salinity patterns, which appear to be the case. Among the target estuaries, the Coos and
Yaquina were ranked sequentially in the dry season concentrations of PC>43", NCV+NCV, and
total suspended solids (TSS) and were within two ranks for chlorophyll a (Tables 4-4 and 4-5).
Yaquina and Coos showed a very similar extent of tidal variation at the mouth of the estuaries
when plotted against the area-normalized freshwater inflow (Figure 4-11), suggesting similar
short-term salinity dynamics. Additionally, Coos and Yaquina showed similar wet and dry
season salinity patterns along the estuarine gradient (Figure 2-11), indicating similar seasonal
dynamics. In contrast, the Umpqua was the least similar of the target estuaries in the nMDS
analysis and was separated from the other target estuaries in the NWI and land use crosswalk
(Table 2-8). This highly river-dominated estuary had the lowest median values for all the water
quality parameters among the target estuaries during the dry season (Tables 4-4 and 4-5) and
lowest wet season DIN and red alder cover in the watershed (Table 4-3), supporting the
suggestion that river-dominated estuaries have different nutrient dynamics and, presumably,
vulnerabilities.
7.3 Within-Estuary Distribution of Seagrasses and Other Benthic Resources
One of the objectives of the probabilistic surveys was to test the hypothesis that the majority of
the Z. marina and other target taxa occurred within the oceanic segments (see Chapter 5) of
different types of estuaries. Two nonexclusive mechanisms could generate this pattern. The first
is that the total intertidal area is greater in the oceanic segments, resulting in the majority of the
population occurring within the oceanic segment even if it did not constitute high quality habitat.
To evaluate whether the population primarily occurred in the lower or upper estuary, we
calculated the relative distributions of each tax on, which was calculated as the percent of the
total area occupied by each taxon occurring in the oceanic versus riverine segments. This metric
represents how the total population is split between the two estuarine segments, and sums to
100%. The second mechanism is that either the lower or upper estuary represents better habitat
independent of its area. Habitat quality was estimated from each taxon's relative cover which
was independently calculated as the percent of the area of the oceanic segment occupied by the
target taxon and the percent the area of the riverine segment occupied by the target taxon. For
example, a taxon could occupy 10% of the area of the oceanic segment versus 25% of the
riverine segment, indicating that the riverine segment provided a better relative habitat. We
attempted to define sample frames a priori within each of the estuaries so that there was an
adequate number of samples in both the oceanic and riverine segments. This initial distribution
was successful in six of the seven estuaries, with >15 samples in each segment (Table 7-1).
211
-------
However, only two samples were located in the riverine segment of the Nestucca, so any
percentage coverages for this portion of the estuary are preliminary.
7.3.1 Within-Estuary Distributions of Seagrasses
Zostera marina was nearly exclusively found in the oceanic segments in the four estuaries where
it was detected (Figure 7-23). The lowest proportion of the total area occupied by Z. marina in
the oceanic segment was 79% in the Tillamook with 98-100% of the seagrass occurring in the
oceanic segments in the Coos, Umpqua, and Yaquina. These results are qualitatively similar to
those from the aerial surveys (Table 6-5). Thus, both approaches demonstrate that the majority
of the native seagrass population occurs in the oceanic segments of both tide-dominated and
river-dominated estuaries.
While difference in sizes of the oceanic and riverine segments contributed to this pattern
(Table 7-1), the distribution of Z. marina is not simply a consequence of the oceanic segments
being larger as indicated by its relative cover. Z. marina occupied very little of the intertidal
zone (0-1%) in the riverine segments of the Coos, Umpqua, and Yaquina estuaries compared to
its relative cover in the oceanic segments (13-17%; Table 7-7). There was a smaller difference
between the two segments in the Tillamook (45% vs. 18%) but even here the relative cover in the
riverine segment was less than half that in the oceanic segment. The higher relative cover
occupied in the oceanic segments of all four estuaries indicates that the oceanic segments provide
substantially better habitat for Z. marina per unit of intertidal area.
Compared to Z. marina, the non-native Z.japonica displayed both a wider within-estuary range
and more variability among the estuaries. The majority of the area occupied by Z. japonica
occurred in the riverine segments of the Tillamook and Umpqua (Figure 7-24). In contrast, the
majority of the Z.japonica population occurred in the oceanic segments in the Coos, Salmon,
and Yaquina estuaries. Averaged across all seven estuaries, 53% of the area occupied by
Z. japonica occurred in the oceanic segments versus 47% in the riverine segments. The relative
cover of Z.japonica was about 2- to 10-fold higher within the riverine segment than the oceanic
segment in the Coos, Tillamook, and Yaquina (Table 7-7). In the Umpqua and Salmon estuaries
comparable relative covers occurred in both the oceanic and riverine segments. Thus, in contrast
to Z. marina, the upper estuary constitutes a high quality environment for this non-native species,
which presumably reflects Z.japonica'?, ability to tolerate lower salinities. The lowest sediment
salinity at which Z.japonica was found was 0.4 psu while several sites had salinities between 5
and 10 psu. In comparison, the lowest sediment salinity for Z. marina was 14 psu. The greater
among-estuary variation in the relative distributions of Z.japonica compared to Z. marina may
be a result of the non-native species not yet reaching an "equilibrium" population within some or
all of the estuaries.
Because a substantial portion of the Z.japonica population is in the riverine segments, this
species has more exposure to terrestrially derived nutrients and hence is more vulnerable to
anthropogenic nutrient sources than Z. marina. One possible scenario is that a moderate level of
nutrient enrichment would stimulate Z.japonica'?, growth in the riverine segment while having
little effect on the lower estuary Z. marina population. Such a stimulation could potentially
promote the establishment and spread of this invader either through enhanced vegetative growth
or seed production.
212
-------
Table 7-7. Relative cover of the five ecologically important benthic taxa and bare habitat in the oceanic and riverine segments in the
seven target estuaries. Relative cover is the percent area of the oceanic segment or riverine segment occupied by each tax on, and
ranges from 0 to 100% for each segment.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
RELATIVE COVER FOR EACH HABITAT
Z. marina
Oceanic
0
15
0
0
45
13
21
Riverine
0
0
0
0
18
0
1
Z. japonica
Oceanic
0
17
24
4
2
19
10
Riverine
0
30
0
2
22
22
18
Macroalgae
Oceanic
52
65
40
49
61
29
71
Riverine
80
15
0
94
67
17
1
Neotrypaea
Oceanic
36
23
64
78
34
61
46
Riverine
20
11
50
58
34
37
61
Upogebia
Oceanic
54
31
18
4
42
0
40
Riverine
1.3
0
0
2
16
0
3
Bare Habitat
Oceanic
6
13
2
14
15
20
2
Riverine
8
50
0
2
19
34
30
to
-------
100
CS
| 80-
1
£ 60 -
S
•3 40-
>^
o
H
o on
0 ZU —
o
0 0
100
1 1 Oceanic
IXX1 Riverine
0 0000
79
21
S
S. \
100
0
99
1
" 1 1 1 1 1 1 1
Alsea Coos Nestucca Salmon Tillamook Umpqua Yaquina
Figure 7-23. Relative distributions of the total area occupied by Z. marina between the oceanic
and riverine segments. No Z. marina was found in the Alsea, Nestucca, or Salmon in the
probabilistic surveys.
cs
100 -
80 -
60 -
73 40 ~
o
H
*S 20 -
0
100
69
0 0
31
91
86
Oceanic
Riverine
14
72
61
39
28
Alsea Coos Nestucca Salmon Tillamook Umpqua Yaquina
Figure 7-24. Relative distributions of the total area occupied by Z.japonica between the oceanic
and riverine segments. No Z. japonica was found in the Alsea. Only two samples were taken in
the Nestucca riverine segment.
214
-------
At high levels of nutrient enrichment, excess nutrients could result in eutrophic conditions in the
riverine segment that could potentially reduce Z. japonica in the upper estuary. Such reductions
would be considered detrimental to the extent that Z. japonica provides ecosystem services.
Thus, development of management strategies to reduce nutrient impacts on seagrasses will
depend, in part, upon the assessment of the ecosystem services provided by Z. japonica versus
any potential detrimental effects such as the potential for competition with Z. marina (e.g.,
Bando, 2006).
7.3.2 Within-Estuary Distributions of Benthic Macroalgae
The majority of the intertidal area occupied by benthic macroalgae occurred in the oceanic
segment in all seven target estuaries (Figure 7-25). At least 88% of the intertidal area occupied
by macroalgae was within the oceanic segments of the Alsea, Coos, Nestucca and Yaquina. A
smaller proportion of the macroalgae occurred in the oceanic segments of the Tillamook,
Umpqua, and Salmon but still >55% of the total macroalgal cover occurred in the lower estuary
in these systems. Thus, the bulk of the macroalgal populations are exposed to ocean-derived
nutrients during the growing season.
In terms of its relative cover, the extent of occurrence of macroalgae in the oceanic segments
ranged from 29% in the Umpqua to 71% in the Yaquina (Table 7-7). In comparison to this
relatively small range, the percentage of the riverine segment occupied by macroalgae varied
widely across estuaries. Macroalgae occurred over approximately 94% of the riverine segment
of the Salmon River versus only about 1% of the riverine segment in the Yaquina even though
the Yaquina had the highest coverage within the oceanic segment. The reasons for these
differences among estuaries are not clear though it is possible that differences in sampling dates
may have contributed to some extent. However, eutrophication does not appear to be the cause
based on the landscape analysis of these estuaries (Chapter 3). It is possible that differences in
sampling dates may have contributed to some extent these differences in within-estuary spatial
patterns. For whatever reason, under certain conditions the river-dominated segments provide
suitable habitat for benthic macroalgae, and these upper estuary populations are more likely to be
exposed to any increases in terrestrially derived nutrients.
7.3.3 Within-Estuary Distributions of Burrowing Shrimp
The populations of both burrowing shrimp species primarily occurred in the oceanic segments
(Figures 7-26 and 7-27) though the two species differed in their relative concentration within the
lower estuary. Seventy-nine to 100% of the estuarine area occupied by U. pugettensis occurred
in the oceanic segments in all seven estuaries (Figure 7-27). The high percentage in the oceanic
segment reflects that the riverine segments constitute a poor habitat for this species with <3% of
the upper estuary being occupied in six of the seven estuaries (Table 7-7). Only the Tillamook
had more than 10% of the intertidal area occupied by U. pugettensis in the riverine segment.
215
-------
CS
S3
J5P
"«
o
100 -
80 -
60 -
cs
- 40 -
o
H
20 -
0
95
100
12
Oceanic
Riverine
100
73
57
55
27
43
45
Alsea Coos Nestucca Salmon Tillamook Umpqua Yaquina
Figure 7-25. Relative distributions of the total area occupied by benthic macroalgae between the
oceanic and riverine segments. Only two samples were taken in the Nestucca riverine segment.
The majority of the intertidal area occupied by N. californiensis also occurred in the oceanic
segments (Figure 7-26), though it exhibited a wider range in its distribution among estuaries.
About 78% to 98% of the total area occupied by N. californiensis was located in the oceanic
segment in the Alsea, Coos, Nestucca, Salmon, and Yaquina estuaries. In comparison, about 55-
60% of the occupied area occurred in the oceanic segments of the Tillamook and Umpqua.
Another difference from U. pugettensis was that the riverine segments constituted relatively
good habitat for TV. californiensis. N. californiensis occupied more than 10% of the riverine
segment in all seven estuaries and occupied at least 50% of the upper estuary in the Nestucca,
Salmon, and Yaquina (Table 7-7). As a consequence of this greater riverine population, N.
californiensis might be expected to have a greater vulnerability to terrestrially derived nutrient
enrichments than U. pugettensis.
216
-------
100 -
j! 80 -
| 60 ^
•Ki
I
_ 40-1
o
H
20 -
0
95
89
98
11
87
\ Oceanic
\ Riverine
78
60
13
55
40
45
22
Alsea Coos Nestucca Salmon Tillamook Umpqua Yaquina
Figure 7-26. Relative distributions of the total area occupied by N. californiensis between the
oceanic and riverine segments. Only two samples were taken in the Nestucca riverine segment.
100
S3
S 80 -
<
.«
1 6°-
&
a 40-
o
H
^ 20 -
S
n
100
0
100
0
100
\ \ Oceanic
0
91
9
\x
" \ \ \
Alsea Coos Nestucca Salmon
k\
79
3 Riverine
21
S
S. X
0 0
98
2
( ( (
Tillamook Umpqua Yaquina
Figure 7-27. Relative distributions of the total area occupied by U. pugettensis between the
oceanic and riverine segments. Only two samples were taken in the Nestucca riverine segment.
217
-------
7.4 Benthic Macroalgae as Diagnostic Indicator of Cultural Eutrophication
High levels of macroalgae have been used in several areas of the world as a diagnostic of
eutrophic conditions (e.g., Valiela et al., 1992; Bricker et al., 1999, 2003). However, applying
this diagnostic to PNW coastal estuaries as an early indicator of cultural eutrophication is
problematic as the seasonal macroalgal blooms appear to be a natural regional phenomenon.
Summer blooms of macroalgae were observed in all seven target estuaries with macroalgae
occurring over 40% of the estuarine intertidal area in six of the estuaries (Table 7-5; Figure 7-
15). Macroalgal blooms have also been reported from other PNW coastal estuaries and Puget
Sound (e.g., Phillips, 1984; Thorn, 1984; Kentula and Mclntire, 1986; Nelson and Lee, 2001;
Thorn et al., 2003; Bulthuis and Shull, 2006). We suggest that such a widespread occurrence of
blooms is more indicative of a regional response than of localized anthropogenic enrichment.
Furthermore, it seems unlikely that all seven target estuaries are experiencing cultural
eutrophi cation given the low level of urbanization, cultivation, and percent impervious surfaces
as well as low population densities in the coastal watersheds (Section 2.8; Tables 3-3 and 3-4).
The possibility that red alder may increase after logging somewhat complicates this conclusion
as this nitrogen fixer contributes to the nitrogen loadings in the riverine segments (Section 3.7).
However, as discussed in this chapter, and supported by the 515N ratios (Chapter 5), the
macroalgae in the target estuaries were primarily exposed to oceanic- versus riverine-derived
nitrogen. Thus, under present conditions, terrestrially derived nutrients could only stimulate a
minority of the macroalgae in these PNW estuaries.
Based on these complementary lines of evidence, we conclude that the present level of benthic
macroalgae is not an indicator of cultural eutrophi cation in PNW estuaries. This conclusion is
specifically for coastal PNW estuaries, and certain subestuaries within Puget Sound may be
experiencing anthropogenically driven increases in macroalgae (e.g., Shaffer, 2001) as are some
estuaries within Southern California (e.g., Cohen and Fong, 2006). While the present level of
macroalgae is not an indicator of eutrophi cation, increases over the existing baselines could serve
as a diagnostic indicator of increasing eutrophic conditions. Increases in macroalgae in the upper
estuary, where riverine-derived nutrients are the dominant source, would be the better diagnostic
than changes in the lower estuary, which are more likely to reflect among-year differences in
oceanic inputs. It would be important to couple measurements of abundance and extent of
macroalgae with measurements of 515N stable isotope ratios to differentiate oceanic versus
terrestrial nutrient sources. Measurement of macroalgal tissue nitrogen concentrations could
complement the abundance and 515N measurements, especially as an integrative measure of
nutrient pulses (Fong et al., 1998; Cohen and Fong, 2001). The practical limitations, however, of
utilizing changes in macroalgal abundance as an indicator include its highly seasonal pattern,
among-year variations, and paucity of quantitative baselines. These limitations could be
mitigated as we generate additional baselines and develop a better understanding of the relative
importance of the factors driving the seasonal pattern and the differences among estuaries. Until
that time, however, we caution against simply using macroalgae as an eutrophi cation indicator in
the PNW.
218
-------
CHAPTER 8: LOWER DEPTH LIMIT OF ZOSTERA MARINA
IN SEVEN TARGET ESTUARIES
Bruce L. Boese, Walter G. Nelson, Cheryl A. Brown, Robert J. Ozretich, Henry Lee II,
Patrick J. Clinton, Christina L. Folger, T Chris Mochon-Collura, and Theodore H. DeWitt
Key Findings
• Lower depth limit for colonization of Z. marina and light attenuation coefficients
(Kd) were determined at multiple locations within the seven target estuaries.
• Leaf epiphyte biomass was determined seasonally at multiple locations in
Yaquina Estuary.
• Z. marina generally tended to colonize deeper toward the mouth of each estuary.
• Lower depth limit for Z. marina generally followed trends in light attenuation
(Kd), which tended to be greater in upriver estuarine reaches.
• Minimum light levels needed to maintain Z. marina in these estuaries was
estimated to be -13% of ambient surface irradiance.
• Epiphyte loads were greatest in summer and fall in the oceanic segment of the
Yaquina Estuary resulting in light reductions at the leaf surface by as much as
60%.
8.0 Introduction
Seagrass meadows are essential to the health and function of estuaries, providing critical habitat
to economically and ecologically important fish, invertebrates, and birds (den Hartog, 1977;
Thayer et al., 1975; Thayer and Phillips, 1977; Dennison et al., 1993; Batiuk et al., 2000).
Seagrass depth distribution is dependent upon light penetration, with coastal seagrasses
extending to depths receiving about 11% of the irradiance at the water's surface (Duarte, 1991).
If the maximum depth that seagrasses grow in an estuary is a result of water clarity alone, then
the maximum colonization depth may be a useful integrative water quality assessment measure
(Dennison et al., 1993). Some have suggested its use as a monitoring tool (Sewell et al., 2001;
Virnstein et al., 2002). Additionally, understanding the minimum light requirements for
seagrasses is necessary for protection of existing seagrass meadows as well as restoration (Batiuk
et al., 2000; Dennison et al., 1993; Fonseca et al., 1998).
In the U.S., the vast majority of seagrass research has been conducted on the Atlantic and Gulf
coasts with a goal to establish water quality criteria to assure the survival and restoration of
seagrass meadows (Batiuk et al., 2000). In contrast, there have been few studies on the light
requirements of U.S. Pacific Coast seagrasses, predominately Zostera marina, with the exception
219
-------
of the work of Zimmerman and his colleagues in San Francisco Bay (Zimmerman et al., 1991,
1995) and Thorn et al. (1998) in Puget Sound.
Light criteria have been proposed as part of the guidelines for restoring and maintaining
2. marina habitat in Chesapeake Bay (Batiuk et al., 2000). However, applying these values to
the PNW Z. marina populations is problematic due to differences in tidal amplitude which tend
to narrow the depth range of seagrasses (Koch and Beer, 1996) and other factors including
temperature and high estuarine flushing rates. Criteria for Chesapeake Bay Z. marina were
derived for the growing season (spring through fall) (Batiuk et al., 2000), where carbohydrates
are accumulated and used to maintain plants during the winter when plants cannot sustain a
positive carbon balance (Zimmerman et al., 1989). In contrast, for Z. marina in PNW estuaries,
winter irradiance appears to be sufficient for the maintenance of a positive carbon balance and as
a result plants continue to grow through the winter, albeit at a slower rate (Boese et al., 2005).
The present study measured the lower depth limit of Z. marina at multiple locations within seven
target estuaries. These depth measurements were then compared to determine if there were
spatial differences within and across estuaries. Light profiles were measured within these
estuaries to determine the water clarity, expressed as a light attenuation coefficient. These
measurements were then used to determine if the lower depth limits were correlated with water
clarity differences within and across estuaries. Finally these depth and water clarity values were
compared to literature values obtained for Z. marina in other estuaries.
8.1 Methods
8.1.1 Selection of Sampling Points
During the summers of 2004 and 2005, the seven target estuaries (Table 8-1) were sampled to
determine the maximum depth of Z. marina. In each estuary, 25-45 sampling points were
determined a priori using digital habitat maps from the Oregon Estuarine Plan Book (Cortright et
al., 1987). Sampling locations were determined by randomly selecting points on a line running
on the channel side of each seagrass-containing polygon using the Arc View v3.3 extensions
Random Point Generator vl.2 and Mila Utilities. Randomly selected sampling points were never
closer than 10 m to each other. If no seagrass was present at one of these pre-selected sampling
points, while on site, a replacement sampling point was selected from the nearest seagrass bed or
patch that was closest to the original a priori sampling point. The Yaquina Estuary was sampled
in both 2004 and 2005 with the additional points randomly selected to increase spatial coverage.
The Alsea Estuary was sampled in 2004 and again in 2006 to increase spatial coverage. In
practice, the number of sampling points varied with the estuary and conditions found on the
sampling days. The actual number of points sampled is presented in Table 8-1 with their
approximate locations within each of the seven estuaries shown on Figures 7-1 through 7-7.
220
-------
Table 8-1. Estuaries sampled for the lower depth limit of Z. marina and the datum adjustment
factor used to convert lower depth limit from the mean lower low water (MLLW) datum to mean
sea level (MSL) datum. * Value is mean of Oregon estuaries as specific correction was not
available for this estuary.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
ESTUARINE
AREA
(km2)
12.49
54.90
5.00
3.11
37.48
33.78
19.96
# SAMPLE
POINTS
39
34
30
30
24
27
64
YEAR SAMPLED
2004, 2006
2005
2004
2004
2005
2005
2004, 2005
MLLW TO MSL
DATUM
ADJUSTMENT
(m)
+0.90
+1.33
+1.25*
+1.25*
+1.36
+1.25*
+1.39
8.1.2 Sampling
The lower margin of seagrass at a sampling point was determined using a color underwater video
camera (Sea-Drop 650 series, SeaViewer Cameras, Tampa, FL). The camera was secured to a
fin stabilizer that was on the end of a 5-m aluminum pole which was attached to a davit mount
on the research vessel (Figure 8-1). For each sampling, the boat with attached camera was
initially positioned at the predetermined point and the boat was then moved approximately 50 m
offshore from the visible edge of a seagrass bed. The boat would then slowly proceed shoreward
on the transect line until seagrass became visible on the monitor. In 2004, when seagrass was
encountered a pole was used to anchor the boat while data were collected. In 2005 and 2006
reverse thrust would be applied to the boat while the depth and GPS data were collected as
quickly as possible.
In areas where the lower margin of a seagrass bed was found in water <1 m in depth, this margin
was most easily found by visual inspection. As the boat approached the lower edge of a seagrass
bed, an individual would exit the boat and hold it at that position while data were collected.
Occasionally sites were sampled in shallow water by wading from shore to the lower bed margin.
Data recorded included the site location, water depth, and the date and time of observations.
Water depth was determined using a lead line and/or hand-held depth sounder (deeper water) or
using a meter stick if the depth was <0.8 m. Depth sounder accuracy was checked using a lead
line during ideal conditions (i.e., no current, wind or waves). Depth values were adjusted to
actual tidal heights determined from the nearest NOAA tide station. This was done by entering
the estuary specific time and location adjustment factors (see:
http://hmsc.oregonstate.edu/weather/tides/tideadj.html) for each measurement into the tide
prediction software WXTIDE32 v4.4 and subtracting the resulting predicted tide height relative
to Mean Lower Low Water (MLLW) from the depth measurement. To be consistent with
seagrass lower limit values in the literature, lower limits relative to the MLLW datum were
221
-------
converted to lower limits relative to mean sea level (MSL) using the datum adjustment factors
for each estuary if available (Table 8-1). If no correction values were available for a given
estuary, the average correction value for Oregon estuaries (+1.25 m) was used. For more details
on the tidal corrections see Section B. 10.1.
Figure 8-1. Schematic of underwater camera apparatus.
8.1.3 Light Attenuation Coefficients
Photosynthetically active radiation (PAR) is defined as irradiance in the 400- to 700-nm
waveband, which encompasses the portion of the light spectrum that plant pigments use in
photosynthesis. PAR profile measurements were made using LI-COR® spherical sensors. These
measurements were made during single day cruises both at high and low tides at multiple
locations in each estuary. At each of these locations PAR was determined at 0.25 m depth
intervals from the surface to the bottom on both the down and up casts. Estimates of the overall
water column light attenuation coefficient (Kj) values were determined from these profile data
based on the slope of the linear regression of In PAR vs. depth.
The Kd values were also determined in the Yaquina Estuary from long-term datasets. These
values were determined in two ways. A series of nearly monthly cruises measured PAR vertical
depth profiles at multiple sites from June 1998 though September 2004. These 0.25-m interval
profiles were taken usually during flood tides. The Kd values for these profiles were determined
by the slope of the regression of In PAR vs. depth for the 0.50 m to 1.25 m intervals and also for
the 1.00 m to 3.75 m intervals, representing near surface and near bottom light attenuations.
222
-------
The second set of long-term Yaquina Estuary PAR measurements was taken at five fixed sites
located at 3.9, 4.5, 11.9, 16.3, and 18.9 km from the estuary mouth1. Measurements at these sites
were taken nearly continuously from 1999 though 2001 using two PAR sensors placed 0.75 m
apart in depth. Kd values for this continuous measurement were determined from the difference
between PAR measurements for the two sensors.
Once Kd was determined, the fraction of surface irradiance reaching a given depth was calculated
as
7-= *-*'* (8.1)
*0
where / is the irradiance at depth (m below MSL), I0 is the irradiance at the surface, Kd is the
light attenuation coefficient (m"1), and z is depth (m below MSL).
8.1.4 Epiphyte Biomass
A study of epiphytes growing on Z. marina leaves was conducted within the Yaquina Estuary
from 2000 though 2004. Data were collected at six stations distributed between 3.5 and 17 km
upriver from the estuary mouth. Leaves from collected plants were subdivided into outer (older)
and inner (younger) leaves. Epiphytes were scraped from these leaf groups, and dry weights (24-
36 hours at 60-70 °C) of the removed material determined for each individual plant. The effect
of epiphyte cover on light (PAR) availability to Z. marina was estimated in the laboratory using
a LI-COR® LI-190S A quantum sensor. Freshly removed epiphytes from a single leaf were
washed into a Plexiglas cylinder with distilled water (60 mL). A light source was placed above
this cylinder with the PAR sensor below the chamber and the amount of irradiance was
determined. This value was then compared to a similarly measured irradiance value obtained
using the same cylinder containing 60 mL of distilled water without epiphytes.
For analysis, dry weight data were partitioned by collection date into wet season (November-
April) and dry season (May - October). Stations were also combined into two groups (oceanic
and riverine segments) each with three stations based on the zonation presented in Chapter 5.
8.2 Results
8.2.1 Sampling Points/Sampling Errors
Inclement weather, tidal constraints, and channel navigation difficulties allowed for the
completion of only 24 sampling stations in the Tillamook Estuary. Seven of these points were
missing corresponding GPS data and therefore the sampling times for correcting depth were
estimated from field notes. Two points in the Coos Estuary also had no corresponding GPS
values. These latter locations were estimated from field notes using their relative positions to
other sampling points where GPS values were obtained.
Electronic depth sounders were tested in the laboratory against a lead line to determine accuracy.
Depth sounder readings were consistently within 0.1 and 0.2 meters of the lead line reading. The
1 The distance from the mouth of an estuary is here defined as the distance from the apparent ocean shore line, not
from the end of the projecting jetties.
223
-------
sonar scattering effect of benthic macroalgae and seagrass was a possible source of error in depth
data. To help minimize this effect depth soundings were taken from within 50 cm of visible
Z. marina shoots and not from directly over them. This precaution could not be taken when
shoots were not clearly visible to the naked eye; roughly 40% of transect depth soundings were
too deep (i.e. > about 1.5 meters, depending on turbidity) to consider this precaution reliable. In
many cases seagrass occurred on or just above benthic slopes estimated to exceed 45 degrees.
When this occurred, given the difficulty in stopping the boat it was unlikely that depth was
measured exactly above the edge of the seagrass bed. In these worse case conditions, we
estimate that the accuracy of the depth measurement was ± 60 cm.
8.2.2 Depth Distribution
Figures 8-2 through 8-8 show the lower depth limits of Z. marina as a function of distance from
the mouth within the seven target estuaries. The lower depth limit of Z. marina varied
considerably within a given estuary, but tended to be greater toward the mouth of each estuary,
with the exception of the Nestucca Estuary (Table 8-2).
The lower depth limit was also significantly different among estuaries (Table 8-3). The overall
mean lower depth limits for Z. marina in the Coos and Nestucca estuaries were not statistically
different from each other, but were different from the other five estuaries. The lower limit of
Z. marina at any particular site might depend upon several factors other than depth such as
sediment type, current velocity, salinity and bathymetry profiles, and the increased among-
estuary variance resulting from these factors makes it more difficult to detect the effects of light
availability. We attempted to reduce this variance added by other factors by reanalyzing only the
deepest 1/3 of the depth limit data. Although this procedure changes mean depth, standard error,
and the rank order of the estuaries, the re-analysis did not alter the results of the pairwise
comparisons (Table 8-3).
Table 8-2. Linear regression coefficients of maximum Z. marina depth (meters below MSL)
versus distance (km) from the mouth (apparent shore line) of each of the seven target estuaries.
Significant constants and slopes are indicated by bold text and probability value. Note that
Nestucca's significant regression is opposite of the general trend in that the lowest depths
observed are farthest from the estuary mouth.
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
CONSTANT
4.33 (p<0.001)
1.36 (p<0.001)
1.11 (p<0.001)
3.46 (p<0.001)
3.21 (p<0.001)
2.62 (p=0.001)
3.70 (p<0.001)
SLOPE
-0.297 (p<0.001)
-0.0056 (p=0.77)
0.134 (p<0.001)
-0.889 (p<0.001)
-0.137 (p=0.049)
-0.0251 (p=0.505)
-0.118 (p<0.001)
r2
0.50
0.003
0.58
0.57
0.16
0.0018
0.32
224
-------
Table 8-3. Overall mean depth (meters below MSL), mean of deepest 1/3 of depth values, and
maximum depth that Z. marina was observed in each of the seven target estuaries. Mean values
are mean ± standard error (n). Bold indicates that two of the seven estuaries were statistically
different from the other five estuaries but not significantly different from each other (Home-
Sidak method of pairwise comparisons, p < 0.05).
ESTUARY
Alsea
Coos
Nestucca
Salmon
Tillamook
Umpqua
Yaquina
OVERALL MEAN
DEPTH (n)*
2.95 ±0.19 (39)
1.28 ± 0.12 (34)
1.49 ± 0.07 (30)
2.25 ±0.16 (30)
2.46 ± 0.23 (24)
2.45 ±0.16 (27)
2.44 ±0.13 (64)
DEEPEST 1/3 OF
DEPTH VALUES (n)*
4.26 ±0.26 (13)
2.09 ±0.15 (11)
1.85 ± 0.02 (10)
3. 37 ±0.06 (10)
3. 72 ±0.39 (8)
3. 36 ±0.28 (9)
3. 62 ±0.11 (24)
MAX. DEPTH
(m)
6.55
3.03
1.94
3.70
5.17
4.71
4.53
*statistical difference in mean values (ANOVA, p<0.001).
6 -
= 4
n,
o>
Q
2 -
246
Distance from Mouth (km)
10
Figure 8-2. Lower limit of Z. marina (m below MSL) versus distance (km) from the mouth of
the Alsea Estuary.
225
-------
J
•s
S
3
6
a.
4*
Q
3.5 -I
3.0 -
2.5 -
2.0 -
1.5 -
1.0 -
0.5 -
0.0 -J
•
•
• •
Q
•
^_
*J
^ • * *
. ' * *" '
V
*
1 1 1 1 1
0 5 10 15 20 25 30
Distance from Mouth (km)
Figure 8-3. Lower limit of Z. marina (m below MSL) versus distance (km) from the mouth of
the Coos Estuary.
2.5 -I
2.0 -
, ,
X L5 '
3
o, 1-0 -
Q
0.5 -
0.0 -
(
• • *
-------
4 -,
3 -
J
-^
JS
o.
Q
1 -
0 -
*
:••
-------
5.0 -1
4.0 -
^ 3.0 -
1
a 2.0 -
Q
1.0 -
0.0 -
(
•
. • .
••'
•
) 2 4 6 8 10 12 14 16 18
Distance from Mouth (km)
Figure 8-7. Lower limit of Z. marina (m below MSL) versus distance (km) from the mouth of
the Umpqua River Estuary.
5 -,
4 -
J
JS
"S, 2
Q
1 -
0 -
(
* 1
• • *
••
• * • .1
•
.•*:**•• *
• • • •
•
• • •
) 2 4 6 8 10 12 14 16 18 20
Distance from Mouth (km)
Figure 8-8. Lower limit of Z. marina (m below MSL) versus distance (km) from the mouth of
the Yaquina Estuary.
228
-------
8.2.3 Kd vs. Distance from Estuarine Mouth
The long-term irradiance datasets from the Yaquina Estuary showed strong linear relationships
between K^and distance from the mouth (Figures 8-9 and 8-10). These results indicate that
water clarity generally decreases from the ocean-dominated mouth into river-dominated reaches
of the estuary. These annualized results are similar to data obtained on high and low tide cruises
conducted in the Yaquina Estuary on June 15 and June 21, 2004 (Figure 8-11).
Unfortunately, the same trend in ^observed during the one day cruises within the Yaquina
Estuary were generally not seen or were not statistically significant in the other estuaries
examined in the same manner (Figures 8-12 through 8-16)). In the Alsea Estuary, there was a
large tidal difference in Kd values (Figure 8-12). While the Kd values tended to increase in the
upriver portions of the Nestucca and Salmon River estuaries (Figures 8-13 and 8-14), these
trends were not significant and low tide Kd values were not determined in the Nestucca due the
shallowness of this estuary during the low tide cruise. In the Tillamook Estuary, low Kd values
on both the high and low tide cruises were observed at the station farthest up the estuary
resulting in a non-significant regression (Figure 8-15). Tillamook Estuary Kd values were further
complicated by highly variable weather conditions which resulted in an extremely variable
ambient light field during many of the PAR sensor casts. Few Kd values were determined for the
Coos Estuary due to instrument malfunction. Only in the Umpqua River Estuary was a
significant relationship found between distance from the estuary mouth and Kd (Figure 8-16).
2.0
1.6 -
1.4 -
1.2 -
0.6
Kd = (0.0355*Distance) + 0.74, r2 = 0.74, p<0.001
10 15 20 25
Distance from Mouth (km)
30
Figure 8-9. Light attenuation coefficient (Kd, m"1) versus distance from the mouth (km) of the
Yaquina Estuary generated using long-term cruise dataset. Kd values are means determined from
PAR depth profile measurements taken approximately monthly from June 1998 though August
2004.
229
-------
1.6 -
1.4 -
1.2 -
1.0 -
0.8 -
0.6
Kd = (0.0603*Distance) + 0.48, r2 = 0.81, p =0.037
6 8 10 12 14 16
Distance from Mouth (km)
18 20
Figure 8-10. Light attenuation coefficient (Kd, m ) versus distance from the mouth (km) of the
Yaquina Estuary generated using continuous dataset Kd values are means determined from PAR
measurements that were taken at five sites which were monitored continuously from 1999
though 2001.
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
= (0.0512*Distance)+0.390, r2 = 0.86, p<0.001
5 10 15 20 25 30
Distance from Mouth (km)
Figure 8-11. Light attenuation coefficient (Kd, m"1) versus distance from the mouth (km) of the
Yaquina Estuary for classification dataset. Kd values determined from light vs. depth profiles
measured in the Yaquina Estuary during a single high (•) and low (o) tide in June 2004.
230
-------
1.8 -1
1.6 -
1.4 -
1.2 -
£ i.o -
0.8 -
0.6 -
0.4 -
0.2 -1
o
o
o
o
"•
Kd= (0.012*Distance) +0.89, r2 = 0.003, p = 0.892
*
. •
1 1 1 1 1 1
01234567
Distance from Mouth (km)
Figure 8-12. Light attenuation coefficient (Kj, m"1) versus distance (km) from the mouth of the
Alsea Estuary. Values were determined during high (•) and low (o) tides.
2.5 -i 1
2.0 -
1.5 -
1.0 -
0.5 -
0.0
K= (0.27*Distance)-0.0314, r2 = 0.64, p=0.104
2345
Distance from Mouth (km)
Figure 8-13. Light attenuation coefficient (Kj, m ) versus distance (km) from the mouth of the
Nestucca Estuary. Values were determined during high tide.
231
-------
Figure 8-1'
Salmon Ri
1.1 -i
1.0 -
0.9 -
0.8 -
h4° 0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0
0
Kd = (0.115*Distance)+0.485, r2 =0.324, p = 0.182
^^^
• ^^^^^
^^^^^ •
0
0 0.5 1.0 1.5 2.0 2.5 3
0
Distance from Mouth (km)
\. Light attenuation coefficient (Kj, m"1) versus distance (km) from the mouth of the
ver Estuary. Values were determined during both high (•) and low (o) tides.
2.5 -,
2.0 -
1.5 -
•o
1.0 -
0.5 -
0.0 -1
o
o
o
o •
Kd = (0.00 13*Distance)+ 1.102, r2 = 0.000, p = 0.986
0
*
111111
0 2 4 6 8 10 12 14
Distance from Mouth (km)
Figure 8-15. Light attenuation coefficient (K& m"1) versus distance (km) from the mouth of the
Tillamook Estuary. Values were determined during high (•) and low (o) tides.
232
-------
1.8
1.6 -
1.4 -
1.2 -
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
0.0
= (0.0252*Distance)+0.427, rz =0.293, p = 0.030
0
10 15 20
Distance from Mouth (km)
25
30
Figure 8-16. Light attenuation coefficient (Kd, m"1) versus distance (km) from the mouth of the
Umpqua River Estuary. Values were determined during high (•) and low (o) tides.
8.2.4 Maximum Zostera marina Depth, Kd and Light Relationship
As there was a significant relationship between Kd and distance from the mouth in the Yaquina
and Umpqua River estuaries, it was possible to attempt to derive a relationship between the
lower limit for Z. marina and the estimated Kd values (Figure 8-9 and 8-16). Although there was
a significant relationship between Kd and lower depth limit of Z. marina for the Yaquina Estuary
(Figure 8-17), the relationship was not significant within the Umpqua Estuary (r2 = 0.034,
p=0.34). Using Equation 8.1, the amount of surface irradiance reaching the observed lower
depth limits for Z. marina within the Yaquina Estuary was estimated. These values ranged from
2 to 85% with a mean ± standard error of 13.4 ± 0.2 % (n = 64). Similar values for the other
estuaries sampled in this study were not determined due to the lack of reliable Kd estimates.
233
-------
a.
QJ
Q
4 -
3 -
2 -
1 -
0 -
Depth = 5.7 - (2.98 * Kd), r2 = 0.32, p<0.001
0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Figure 8-17. Relationship between Z. marina lower depth limit (m below MSL) and light
attenuation coefficient (Kd, m"1) in the Yaquina Estuary.
8.2.5 Epiphyte Patterns and Impact on Light
In the Yaquina Estuary, there was a general annual pattern in 2000 though 2003 in which
epiphyte biomass increased in the spring to maximal amounts in the summer and fall. This
statistically significant parabolic relationship was most clearly seen on the older, external
seagrass blades (Figure 8-18). In the Yaquina Estuary, epiphyte biomass per unit surface area of
seagrass leaves was higher in the oceanic segment than in the riverine segment in both wet and
dry seasons (Figure 8-19), although only the dry season differences were statistically significant.
Epiphyte biomass per unit leaf surface area was higher in the dry season than the wet season
within both segments. There is a significant positive linear relationship between percent light
reduction and log+1 transformed epiphyte biomass (Figure 8-20).
234
-------
12
10 -
«
E
o
3
6
2 -
y = -0.0002 x + 0.0757 x- 1.1484
r2 = 0.60
50 100 150 200 250
Julian Day
300
350
400
Figure 8-18. Temporal relationship of epiphytic biomass per unit leaf area on Z. marina external
leaves in the Yaquina Estuary. Values are for 2000 though 2003.
Oceanic Segment
Riverine Segment
0
Wet Season Dry Season
External
Wet Season Dry Season
Internal
Figure 8-19. Epiphyte biomass per unit leaf area on older (external) and younger (internal)
Z. marina leaves by season (wet or dry) in the oceanic and riverine segments of the Yaquina
Estuary. Wet season is from November through April. Dry season is from May through
October.
235
-------
100
80-
c
1 60
*- 40
WD
20
y = 79.044x +4.8381
r2 = 0.80
• ^ »
0.2
0.4
0.6
0.8
1.2
1.4
-2\
log (biomass (mg cm") +1)
Figure 8-20. Linear regression relationship between the percent of light reduction to log(x+l)
transformed epiphyte biomass per unit Z. marina leaf surface area.
8.3 Discussion
8.3.1 Sampling Method
The video camera sampling technique was difficult to conduct in typical field conditions. High
winds and strong currents made maintaining position or tracking a straight line difficult.
Stopping the vessel in time to mark the location of a Z. marina patch was frequently challenging,
as the vessel would drift off location during the time required to set anchor or stabilize position
before recording GPS position. Water clarity was also a factor which introduced variability in
estimating the lower limit of Z. marina because visibility was often as low as 0.2 to 1 m.
Measuring the lower depth margin of Z. marina using manual sampling methods appeared to be
more precise than the video collection technique. While this method could only be employed
during low tides and in areas where the Z. marina lower margin was shallow (< 1.5 m), when the
researcher exited the vessel quickly, the position and depth of the edge of the seagrass bed could
be determined with a very high degree of accuracy and precision.
8.3.2 Depth Distributions
In general, the maximum colonization depth (Zc) limit of Z. marina is ultimately determined by
water clarity (Duarte, 1991). However, we observed a high variability (Figure 8-2) in these
236
-------
maximum depth values within the seven target estuaries which cannot be explained by water
clarity relationships alone. Factors, such as current velocity, sediment characteristics, and other
non-optical water properties, play a role in determining where and how deep seagrasses grow
(Koch, 2001; Virnstein et al., 2002), possibly restricting the plants to shallower waters.
Although some variability in the depth distribution data were reduced by only analyzing data for
the deepest one-third of all the patches found within in each estuary (Table 8-3), this method
introduces a degree of subjectivity. However, given this caveat, there was a general trend (i.e., in
six of the seven estuaries) for the maximum depth of Z. marina colonization to be deepest
nearest the mouth (Figures 8-2 through 8-8) and a statistically significant linear model could be
fit to four of these estuaries (Table 8-2). The exception to this trend was within the Nestucca
Estuary where the opposite trend appeared to occur. This latter result may be due to the overall
shallowness of this estuary when compared to the others, which may have contributed to its
general lack of extensive Z. marina beds. Relative to the other estuaries in the present study,
Z. marina was not found near the mouth of the Nestucca Estuary where presumably mean water
clarity may be greater. This lack of Z. marina was most likely due to high tidal current velocities
and shifting sands, which we commonly observed at the mouth of this estuary during peak flood
and ebb tides. In addition, the loss of multiple instruments near the mouth of the Nestucca
Estuary supports this assessment that this is a high-energy environment. Upriver Z. marina was
limited to fringing patches next to deeper water channels, where it is also possible that high
seasonal current velocities may preclude deeper colonization depths.
Comparing these maximum colonization depths to other PNW estuaries is difficult. Although
several studies have measured these values (Thorn et al., 2001; Thorn et al., 2003; Selleck et al.,
2005), the data are presented in graphical form making it difficult to compare mean and
maximum colonization depths directly. Data from Thorn et al. (2003) indicate that the deepest
Z. marina observed in the Coos and Willapa estuaries was approximately 1.8m and 2.8 m below
MSL, respectively. Although these depths are shallower than those observed in the present
study, there is no indication that Thorn et al. (2003) made a concerted effort to find the maximum
colonization depth. Their study was instead focused on determining the depth of maximum
shoot density and relating this to salinity and irradiance.
Nonetheless, based on the present study and those of Thorn and his colleagues (Thorn et al.,
2001; Thorn et al., 2003), maximum colonization depths for Z. marina within coastal PNW
estuaries are shallower than those observed in Puget Sound (Thorn et al., 1998; Dowtry et al.,
2005; Selleck et al., 2005), where Z. marina is commonly found 5 m below MSL, with some
areas having shoots present to three times that depth (Selleck et al., 2005).
8.3.3 Water Clarity Relationships
A priori we assumed that overall water clarity would decrease in the more riverine portions of
estuaries and this trend in water clarity would be apparent in light attenuation measurements
taken over an interval of only a few days. This assumption appeared to be true for the Yaquina
Estuary, as long-term water clarity measures averaged over a number of years were similar to
those measured during two days in 2004. However, this trend was not consistently observed in
the other estuaries examined. It is likely that our short-term measurements of irradiance at depth
are not adequate to define the spatial variability within PNW estuaries. For example, in the
Alsea Estuary, Kj values at a given location were almost 4-fold greater during low tide than high
237
-------
tide. This variability is probably a natural occurrence in many PNW estuaries. At the four
locations within the Yaquina Estuary where Kj values were determined nearly continuously from
1999 to 2001, the coefficient of variation ranged from 33 to 66%, indicating considerable tidal
and seasonal variability.
Only within the Yaquina Estuary were there sufficient long-term measurements of Kj to derive a
relationship with Z. marina colonization depth. This linear relationship (Figure 8-17) was
consistent with the idea that some portion of the reduction in colonization depth with increasing
distance from the mouth of the Yaquina Estuary is related to a reduction in water clarity. This
upriver trend is consistent with that reported by Thorn et al., (2003) who noted an upriver
reduction in Z. marina abundance in the Coos and Willapa estuaries, which they attributed to
increased turbidity and reduced salinity.
Differences in Z. marina colonization depth between PNW outer coast estuaries and Puget Sound
bays are also likely due to water clarity, as K
-------
Table 8-4. Comparison of percent of surface irradiance needed to maintain Z. marina at
its maximum colonization depth from published data and from Yaquina Estuary data.
SOURCE
Duarte, 1991
Current Study
Batiuk et al., 2000
MEAN
20.5
13.4
15a
STANDARD
ERROR OF MEAN
2.1
1.9
N
29
64
MAX
43.9
85.8
MIN
4.7
1.5
a Value is not a mean but is based on an analysis of literature and on an evaluation of
monitoring and modeling research.
10 -,
8 -
1 6 -
a „
3 4-
€
ft
4) 9
Q 2
0 -
•••/ •
• On
• • o o<*e
• •• o0o &>,
• • o°^ s5° oQ
• • @O O CQ O (Q O
A ^M A O O n O
• °° o °°<9 0<8 ° ^
°oo 9 o
o
1 1 1 1 1 1 1
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1
6
Figure 8-21. Relationship between Z. marina maximum depth limit (m) and
Duarte (1991) and the Yaquina Estuary (o) from the present study.
& (•) Data from
8.3.4 Zostera marina Light Criteria
Minimum light requirements for maintaining and restoring seagrass have been proposed for
Chesapeake Bay (Batiuk et al., 2000) and Puget Sound (Thorn et al., 1998). Chesapeake Bay
light criteria values were empirically estimated by measuring the maximum depth of seagrass
annually and associated K22% of surface irradiance) regions
than for fresh and oligohaline regions (>15% of surface irradiance). These proposed criteria
were modified (Batiuk et al., 2000) by the amount of light absorbed by epiphytes encrusting
seagrass leaf surfaces, which reduced these criteria values to 9 and 15 %, respectively. These
239
-------
proposed criteria are also applicable only to the seagrass growing season (typically spring though
fall).
In contrast, light requirements for Z. marina in Puget Sound, Sequim Bay, Willapa Bay, and
Coos Bay were reported as integrated light intensity levels (Thorn et al., 1998; Thorn et al.,
2008). These were estimated using maximum seagrass depth measures, Kj values and
production-irradiance (P vs. I) relationships. Based on this methodology, Thorn et al. (2008)
suggested that the minimum light requirements to maintain Z. marina is -300 umoles photons
91 91
m" s" for at least three hours daily or ~3 moles photons m" d" during May through September.
However, growth would be light limited during the spring and summer at light levels < 7 moles
photons m"2 d"1. Thorn et al. (1998) went on to suggest that for Z. marina to minimally persist
would require mid-day minimum irradiance values at the maximum depth limit to be
9 1
approximately 150 jimoles photons m" s" during the year. Assuming that mid-day surface
irradiance is in the range of 1000-2000 jimoles photons m"2 s"1, this corresponds to
approximately 15-30% of surface irradiance which is consistent with other published criteria
values and the present study (Table 8-4). At a lower margin Z. marina site in the Yaquina
9 1
Estuary, the mean daily irradiance value was approximately 3.8 moles photons m" d" (Kaldy
and Lee, 2007). Although this value exceeds the Thorn et al. (1998) criterion, irradiance values
were highly variable, ranging from 0.5 to 7 moles photons m"2 d"1, with extended periods of
apparently inadequate lighting at depth from October to December (Kaldy and Lee, 2007).
However, even during these periods of apparently inadequate irradiance, Yaquina Estuary
Z. marina continues to grow (Boese et al., 2005; Kaldy and Lee, 2007). This continued growth
and survival under conditions of negative carbon balance is most likely maintained by the
utilization of rhizome carbohydrate reserves (Cabello-Pasini et al., 2002).
8.3.5 Yaquina Estuary Zostera marina Light Criteria
While it is tempting to directly apply the existing light criteria values to PNW estuaries, there are
several additional factors that need to be considered. Chesapeake Bay, and the estuaries from
which Duarte (1991) and Thorn et al. (1998) derived their relationships are generally less turbid
(mean Kj ~ 0.5 m"1) than the Yaquina (mean Kj- 1.1 m"1) and the other estuaries surveyed in the
present study. Zostera marina has been shown to adapt to lower winter irradiance by increasing
chlorophyll content (Zimmerman et al., 1995). Although we are not aware of any study that
documents an analogous response to turbidity, a similar response to chronically more turbid
water might allow for deeper colonization.
Temperature is a possible confounding factor. The range of near-surface temperatures within
Chesapeake Bay, Puget Sound, and in the estuaries used in Duarte's (1991) review are likely
greater than those observed within the Yaquina (Boese et al., 2005) due to the latter's twice daily
flushing with cold ocean water. Increased respiration rates due to higher summer temperatures
in these other estuaries would potentially need to be offset by increased irradiance for plants not
only to maintain themselves but to store carbohydrates in rhizomes which could then be used to
maintain the plant during the winter when irradiances may be less than optimal (Zimmerman et
al., 1995; Burke et al., 1996; Zimmerman and Alberte, 1996). Therefore, it is possible that
Z. marina in PNW coastal estuaries may require less spring and summer irradiance to perform
the same function because of the generally cooler water temperatures in these systems.
240
-------
Our study of epiphytes growing on Z. marina leaves in the Yaquina Estuary revealed a reduction
in the amount of epiphyte biomass in the riverine segment (lower salinity region). With the
exception of 2004 there appeared to be a seasonal pattern in epiphyte biomass such that the
greatest biomass occurred in the summer and fall. This accumulation of epiphytes reduced the
amount of light reaching leaf surfaces by about 60% in the oceanic segment of the Yaquina
Estuary where the greatest densities of Z. marina occur. At present we are not sure how the
epiphyte load and its impact on light availability compares to that found in other estuaries, but
such variation will need to be considered in future efforts to derive water column light criteria for
Z. marina.
241
-------
CHAPTER 9:
EMERGING PACIFIC NORTHWEST PARADIGM
Henry Lee II and Cheryl Brown
9.0 Overview of the Pacific Northwest Environment and Regulatory Implications
In an effort to see the "forest" from the "trees" presented in the previous eight chapters, we
offer a synopsis of the environmental conditions and nutrients dynamics in PNW coastal
estuaries. This overview is based on our long-term research in the Yaquina Estuary, the
classification study presented here, and a growing body of literature on the dynamics of
PNW estuaries. Some of the environmental conditions presented here will not apply to
every PNW estuary, in particular the estuaries with restricted tidal flushing. Nor does this
synopsis apply to Puget Sound, which would require a separate analysis. Additionally, in
most cases we had to extrapolate from a limited number of estuaries. Nonetheless, we
believe these regional elements form the basis of an emerging "Pacific Northwest
Paradigm." We then evaluate how these regional characteristics affect the classification of
PNW estuaries as well as their potential regulatory implications.
9.1 Environmental Conditions
Climate and River Flow
The PNW has a mild Mediterranean climate with a wet winter and dry summer. River
flows reflect this seasonal precipitation pattern.
Average annual rainfall is on the order of 140 to 350 cm y"1, with higher rainfall in the
northern portion of the PNW.
Watersheds and Land Cover
Watersheds are primarily forested, with low population densities and percent impervious
surfaces.
Logging is common in many coastal watersheds but this alteration is not well captured in
the standard land cover data.
Estuarine Inventory and Characteristics
There are 103 estuaries in the PNW. Most of these are small, with 73 estuaries less than
9 9
1 km in size. Thirteen estuaries are larger than 10 km , of which Grays Harbor and the
Willapa Estuary in Washington and the Columbia River Estuary are larger than 100 km .
PNW estuaries are characterized by large intertidal zones, which on average equal
approximately 50% of the estuarine area.
PNW estuaries are classified as mesotidal, with tidal ranges of about 2 meters.
242
-------
In general, PNW estuaries have high flushing with short residence times. Flushing is higher
in the winter due to increased river flow. In the summer, flushing is driven by tidal
exchange.
The PNW estuaries can be classified into seven classes (see Section 9.2), with the drowned
river mouth estuaries sub-divided into tide-dominated, moderately river-dominated, and
highly river-dominated systems.
Nutrient Concentrations and Loadings
Total nitrogen loading into PNW estuaries is high, often equivalent to loadings in eutrophic
systems.
Dissolved inorganic phosphorous (DIP) concentrations are high, and on a national scale are
rated as "fair".
Differences in the extent of flushing and freshwater inflow suggest that estuary classes vary
in their relative vulnerability to terrestrial nutrient loading in the following order (from
most to least vulnerable):
lagoon>blind estuary>tide-dominated river mouth>river-dominated river mouth>coves/harbors
Nutrient Sources
Seasonal coastal upwelling from approximately April to October is the major nutrient
source to the near-coastal region. The intensity of upwelling is highly variable over scales
of days to decades.
Advection of oceanic water into the lower estuary is the major nutrient (nitrate and
phosphate) source for the lower estuarine segments of PNW estuaries during the summer
growing season. This portion of the estuary is referred to as the "ocean-dominated"
segment of the estuary.
Terrestrially derived nutrients are the dominant nitrogen source for the upper estuary
portions of PNW estuaries during the summer growing season and for most of the estuary
during the winter. This portion of the estuary is referred to as the "river-dominated"
segment of the estuary.
Nitrogen fixation associated with red alder, a native tree, may be an important terrestrial
nitrogen source. Red alder populations may increase after fire, logging, and other
disturbances but for the purpose of this analysis they are considered a "natural" nutrient
source.
Point sources, such as sewage discharges may be locally important nitrogen sources but do
not appear to be major sources for PNW estuaries in general.
243
-------
Phytoplankton Concentrations and Sources
Chlorophyll a concentrations in PNW estuaries are generally low to moderate, and were
rated as "good" on a national scale.
Advected oceanic phytoplankton is a major source of phytoplankton in the lower portion of
the estuaries.
Ocean-dominated estuaries appear to have higher chlorophyll a levels than river-dominated
estuaries.
Grazing by the benthic communities and by burrowing shrimp in particular, may
substantially reduce phytoplankton concentrations and help to ameliorate enrichment
effects.
Submerged Aquatic Vegetation
The native Zostera marina is the dominant species of submerged aquatic vegetation (SAV)
in PNW estuaries in the lower intertidal and shallow subtidal.
The areal extent of Z. marina varies substantially among estuaries. In general, the largest
SAV beds are in the tide-dominated and bar-built estuaries.
The lower depth limit to which Z. marina grows decreases in the upper estuary segments,
which correlates with a general decrease in up estuary water clarity. There are differences
among estuaries in the lower limit to which Z. marina grows, which are also likely related
to differences in water clarity but additional factors, especially current velocities may
contribute to these differences.
The non-native Zoster a japonica has become abundant in a number of Pacific Coast
estuaries over the last 20 to 50 years, and its populations appear to be expanding. In several
estuaries, it covers a greater percentage of the intertidal zone than the native Z. marina.
The research to date suggests that Z. japonica may have both beneficial and detrimental
impacts on Z. marina, estuarine food webs, nutrient dynamics, and maintenance of
ecological services.
Benthic Macroalgae
Seasonal blooms of benthic macroalgae are a natural phenomenon in many of the PNW
estuaries and are important primary producers in the estuaries larger than about 3.0 km .
244
-------
Benthic macroalgae cover substantial portions of the intertidal in many estuaries, and reach
standing stocks associated with adverse impacts on seagrasses in Chesapeake Bay. It is not
known to what extent, if any, benthic macroalgae impacts SAV in the PNW.
The primary source of nitrogen for the intertidal benthic macroalgae blooms is natural,
primarily consisting of oceanic water transported into the estuary.
Benthic Communities and Burrowing Shrimp
The burrowing shrimp, Neotrypaea californiensis and Upogebia pugettensis, are two
frequently occurring ecological engineering species in moderate- to large-sized PNW
estuaries. These species affect sediment structure, benthic community composition, and
benthic nutrient fluxes through intense bioturbation and irrigation of the sediment.
Neotrypaea californiensis extends further up estuary than U. pugettensis. Additionally,
U. pugettensis is either less abundant or absent in river-dominated estuaries.
Exposure of SAV and Other Resources to Oceanic vs. Riverine Nutrients
A fundamental question in evaluating vulnerability to anthropogenic nutrient loading is
whether a species is primarily exposed to oceanic or riverine nutrients during the summer
growing season. Populations that are primarily exposed to oceanic nutrients would have a
relatively low vulnerability to terrestrially derived nutrient enrichments. Conversely,
species that have a large proportion of its population in the riverine segment would be
exposed to any increases in terrestrially derived nutrients, and hence have a higher
vulnerability to nutrient loading from watersheds.
The majority of the Z. marina, benthic macroalgae, N. californiensis, and U. pugettensis
populations occurred in the oceanic segments. While there were some differences among
estuaries, this basic pattern held for both river- and tide-dominated estuaries and for large
and small estuaries. Thus, the bulk of the populations of these species are primarily
exposed to a natural nutrient source during the summer growing season.
The exception to this pattern was the nonindigenous Z.japonica which was abundant in
both the oceanic and riverine segments. As a result of this distribution, Z.japonica
populations are more likely to be exposed to anthropogenic nutrient enrichments. This
increases their vulnerability to high anthropogenic nutrient loading though moderate
nutrient enrichment could potentially stimulate their growth and establishment.
Expressions of Eutrophic Conditions
Dissolved oxygen concentrations in PNW estuaries are generally above levels indicative of
eutrophic conditions. However, at times hypoxic shelf water can be advected into the lower
portion of PNW estuaries (Brown et al., 2007).
245
-------
The current levels of benthic macroalgae appear to be natural and not an indication of
eutrophi cation.
The high nutrient loadings of PNW estuaries appear to be due to advection of high nutrient
ocean water into the lower estuary and the effects of the nitrogen-fixing red alder in the
watershed.
Based on these observations, there is little evidence of cultural eutrophi cation at the
estuarine scale in PNW estuaries.
There are some observations suggesting that eutrophi cation may be occurring in localized
areas of some estuaries. In particular, phytoplankton blooms and reduced dissolved oxygen
may be occurring in sloughs with reduced flushing in the riverine segments of some
estuaries.
While not presently showing eutrophic conditions over large areas, increases in nutrient
loading from anthropogenic sources could potentially result in harmful algal blooms and
depressed oxygen in the upper estuary segments of these systems. However, more detailed
studies will be required to accurately assess the vulnerability of the upper estuary segment
and to predict the nutrient concentrations or loadings at which such deleterious impacts
would be expected.
Diagnostics of Eutrophication
With further development, the lower depth limit of Z. marina could potentially be used as
an indicator for assessing estuarine condition in the PNW.
Increases in macroalgae especially in the riverine segment could be used as an indicator of
nutrient enrichment, though there are a number of sampling issues.
Analysis of stable isotopes (515N) in the benthic macroalgae would help determine whether
ocean-derived or terrestrially-derived nitrogen was the major nutrient source associated
with macroalgal blooms.
9.2 Classification of Pacific Northwest Estuaries
Based on the importance of upwelling, large tidal ranges and flushing, wet/dry seasons, and
other environmental characteristics of the PNW, we suggest that a regional classification
system that captures these drivers is required to group the estuaries in an ecologically
meaningful manner. The classification schemes we presented are specifically for the PNW,
and as with any classification scheme, should be considered hypotheses until it is
demonstrated that they adequately predict types of biotic assemblages, nutrient
concentrations/loadings, impacts, or vulnerability.
Pacific Northwest estuaries break out into seven types based on geomorphology, ocean
exchange, and river influence: coastal lagoons, blind estuaries, tidal coastal creeks, tidally
restricted coastal creeks, marine harbors/coves, bar-built estuaries, and drowned river
246
-------
mouth estuaries. Using metrics normalizing total precipitation in the watershed to either
estuarine volume or area, the drowned river mouth estuaries are further divided into tide-
dominated river mouth estuaries and river-dominated river mouth estuaries.
r\
Small PNW estuaries <1 km are an important resource for salmon. Many of these systems
were tentatively classified as "tidally restricted coastal creeks" but additional information is
needed to evaluate the accuracy of this grouping.
Classification of estuaries by the absolute or relative areas of wetland classes from the
National Wetland Inventory (NWI) offers the advantage of grouping estuaries with similar
biotic habitats, which presumably integrates a wide range of environmental conditions. In
the same way, classification of estuaries by similarity in land cover in the associated
watersheds generates insights into potential anthropogenic loadings.
Six pairs of estuaries showed a high degree of similarity based on the intersection of the
clustering by NWI wetland habitat type and the clustering by land cover patterns. These six
estuary pairs - Grays Harbor and Willapa, Yaquina and Alsea, Coquille River and Siuslaw
River, Rogue River and Klamath River, Quillayute River and Smith River, and Hoh River
and Queets River - could potentially form the basis for developing management
frameworks in the PNW.
9.3 Regulatory Implications
The primary regulatory implication of our results is that the development of national
estuarine nutrient criteria that do not take into account the naturally high nutrient
concentrations resulting from upwelling are likely to result in numerous false non
attainments of nutrient criteria in PNW coastal estuaries (Brown et al., 2007). These false
non attainments are most likely to occur in the lower estuary during the summer (Brown et
al., 2007).
Another implication is that the development of total daily maximum loads (TMDLs) for
nutrients need to take into account the intrusions of high nutrient ocean water into estuaries
during a portion of the year. The influence of nitrogen fixation by red alder also needs to be
evaluated in the development of TMDLs.
Our results strongly suggest that a regional approach is required both for the development
of nutrient criteria and in the formulation of TMDL strategies.
9.4 Future Directions
This report describes a pilot effort at classifying PNW estuaries with regards to landscape
attributes and their susceptibility to nutrient enrichment. In addition to the seven target
estuaries described in this report, we have continued the sampling effort and to-date have
sampled an additional 8 estuaries (Coquille, Humboldt, Klamath, Necanicum, Netarts,
Siletz, Sixes, Yachats) using a modification of the methods described in this report. As
these and any new data are analyzed, the classifications and models presented here will be
tested and refined as needed.
247
-------
REFERENCES
Abdelrhman, M.A. 2005. Simplified modeling of flushing and residence times in 42
embayments in New England, USA, with special attention to Greenwich Bay, Rhode
Island. Estuarine, Coastal and Shelf Science 62:33 9-3 51.
Aguiar, A.B., J.A. Morgan, M. Teichberg, S. Fox, and I. Valiela. 2003. Transplantation and
isotopic evidence of the relative effects of ambient and internal nutrient supply on the
growth of Ulva lactuca. Biological Bulletin 205:250-251.
Ahnelt, H., J. Goschl, M.N. Dawson, and D.K. Jacobs. 2004. Geographical variation in the
cephalic lateral line canals of Eucyclogobius newberryi (Teleostei, Gobiidae) and its
comparison with molecular phylogeography. Folia Zoologica 53:385-398.
Armstrong, J.L., D.A. Armstrong, and S.B. Mathews. 1995. Food habits of estuarine staghorn
sculpin, Leptocottus armatus, with focus on consumption of juvenile Dungeness crab,
Cancer magister. Fishery Bulletin 93:456-470.
Arneson, RJ. 1976. Seasonal variation in tidal dynamics, water quality and sediments in the
Coos Bay estuary. M.S. Thesis, Oregon State University, Corvallis, OR. 250 p.
Arnold, C.L. and CJ. Gibbons. 1996. Impervious surface coverage: the emergence of a key
environmental indicator. Journal of the American Planning Association 62:243-258.
Askren, D., R. Hansen, B. Higgins, S. Noble, and R. Platt. 1976. A physical description of the
Salmon river estuary. Ocean Engineering Program, Oregon State University, Corvallis, OR.
67 p.
Bailey, A., H. Berry, B. Bookheim, and D. Stevens. 1998. Probability-based estimation of
nearshore habitat characteristics. In Proceedings of Puget Sound Research '98 Conference,
March 12-13, p. 580-588. Puget Sound Water Quality Action Team, Olympia, WA.
Bando, KJ. 2006. The roles of competition and disturbance in a marine invasion. Biological
Invasions 8:755-763.
Batiuk, R.A., B. Bergstrom, E. Koch, L. Murray, J.C. Stevenson, R. Bartleson, V. Carter, N.B.
Rybicki, J.M. Landwehr, C. Galleogos, L. Karrh, M. Naylor, D. Wilcos, K.S. Moore,
S. Ailstock, and M. Teichberg. 2000. Chesapeake Bay submerged aquatic vegetation
water quality and habitat-based requirements and restoration targets: a second technical
synthesis. United States Environmental Protection Agency, Chesapeake Bay Program,
Annapolis, MD. 174 p. and appendices
Bayer, R.D. 1979. Inter-tidal zonation of Zostera marina in the Yaquina Estuary, Oregon.
Syesis 12:147-154.
Bayer, R.D. 1996. Macrophyton and tides at Yaquina Estuary, Lincoln County, Oregon.
Journal of Oregon Ornithology 6:781-795.
248
-------
Binkley, D., K. Cromack, D. Baker, Jr. 1994. Nitrogen fixation by red alder: biology, rates, and
controls pp. 57-72. In D.E. Hibbs, D.S. DeBell, R.F. Tarrant (eds.) The Biology and
Management of Red Alder. Oregon State University Press, Corvallis, OR.
Boese, B.L., B.D. Robbins, and G. Thursby. 2005. Desiccation is a limiting factor for eelgrass
(Zostera marina L.) distribution in the intertidal zone of northeastern Pacific (USA)
estuary. Botanica Marina 48:275-283.
Borde, A.B., R.M. Thorn, S. Rumrill and L.M. Miller. 2003. Geospatial habitat change analysis
in Pacific Northwest coastal estuaries. Estuaries 26(4B): 1104-1116.
Bostrom, C. and E. Bonsdorf 1997. Community structure and spatial variation of benthic
invertebrates associated with Zostera marina beds in the northern Baltic Sea. Journal of
Sea Research 37:153-166.
Bottom, D., B. Kreag, F. Ratti, C. Roye, and R. Starr. 1979. Habitat classification and inventory
methods for the management of Oregon estuaries. Estuary Inventory Report, Oregon
Department of Fish and Wildlife. 109 p.
Bricker, S.B., C.G. Clement, D.E. Pirhalla, S.P. Orlando, andD.R.G. Farrow. 1999. National
estuarine eutrophication assessment: effects of nutrient enrichment in the nation's estuaries.
United States Department of Commerce, National Oceanic and Atmospheric
Administration, National Ocean Service, Special Projects Office and the National Centers
for Coastal Ocean Science. Silver Spring, MD. 71 p.
http://ian.umces.edu/neea/pdfs/eutro_report.pdf
Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks, and J. Woerner. 2007.
Effects of Nutrient Enrichment In the Nation's Estuaries: A Decade of Change. NOAA
Coastal Ocean Program Decision Analysis Series No. 26. National Centers for Coastal
Ocean Science, Silver Spring, MD. 328 pp.
http://ian.unices.edu/pdfs/neea_2004_report.pdf
Bricker, S., G. Matlock, J. Snider, A. Mason, M. Alber, W. Boynton, D. Brock, G. Brush, D.
Chestnut, U. Claussen, W. Dennison, E. Dettmann, D. Dunn, J. Ferreira, D. Flemer, P.
Fong, J. Fourqurean, Hameedi, D. Hernandez, D. Hoover, D. Johnston, S. Jones, K. Kamer,
R. Kelty, D. Keeley, R. Langan, J. Latimer, D. Lipton, R. Magnien, T. Malone, G.
Morrison, J. Newton, J. Pennock, N. Rabalais, D. Scheurer, J. Sharp, D. Smith, S. Smith, P.
Tester, R. Thorn, D. Trueblood, R. Van Dolah. 2003. National Estuarine Eutrophi cation
Assessment Update: Workshop summary and recommendations for development of a long-
term monitoring and assessment program. Proceedings of a workshop September 4-5,
2002, Patuxent Wildlife Research Refuge, Laurel, Maryland. NOAA, National Ocean
Service, National Centers for Coastal Ocean Science. Silver Spring, MD. 19 pp.
Bricker, S.B., J.G. Ferreira, and T. Sims. 2003. An integrated methodology for assessment of
estuarine trophic status. Ecological Modelling 169:39-60.
Brophy, L. 2003. Wetland site prioritization Lower Elk and Sixes Rivers, Curry County, OR.
Produced for Oregon Trout by Laura Brophy, Green Point Consulting, Corvallis, OR. 52 p.
249
-------
http://www.ortrout.org/images/8success/Cape%20Blanco%20Assessment.pdf#search=%22
%22wetland%20site%20prioritization%22%22.
Brophy, L.S. 2007. Estuary Assessment: Component XII of the Oregon Watershed Assessment
Manual. Prepared for the Oregon Department of Land Conservation and Development,
Salem, OR and the Oregon Watershed Enhancement Board, Salem, OR.
http://www.oregon.gov/OWEB/docs/pubs/wa_estuary/Estuary_Assessment_2007.pdf
Brown, C.A., W.G. Nelson, B.L. Boese, T.H. DeWitt, P.M. Eldridge, I.E. Kaldy, H. Lee II, J.H.
Power, and D.R. Young. 2007. An Approach to Developing Nutrient Criteria for Pacific
Northwest Estuaries: A Case Study of Yaquina Estuary, Oregon. USEPA Office of
Research and Development, National Health and Environmental Effects Laboratory,
Western Ecology Division. EPA/600/R-07/046.
http://www.epa.gov/wed/pages/publications/authored/EPA600R-
07046AnApproachToDevelopingNutrientCriteria.pdf
Brown, C.A. and RJ. Ozretich. 2009. Coupling between the coastal ocean and Yaquina Bay,
Oregon: importance of oceanic inputs relative to other sources. Estuaries and Coasts
32:219-237.
Bulthuis, D.A. and S. Shull. 2006. Monitoring the distribution of submerged aquatic vegetation
in Padilla Bay, NERR-SWMP biomonitoring pilot site, 2004. A final report to Estuarine
Reserves Division, OCRM, NOAA as partial fulfillment of Research & Monitoring Task 4
"Monitoring the Distribution of Submerged Aquatic Vegetation in Padilla Bay, NEER-
SWMP Biomonitoring Pilot Site, 2004" as part of the Operations Award for 2004-2005 to
Washington State Department of Ecology. Padilla Bay National Estuarine Research
Reserve, Shorelands and Environmental Assessment, Department of Ecology, Mt. Vernon,
WA. 28 p.
Burke, M.K., W.C. Dennison, and K.A. Moore. 1996. Non-structural carbohydrate reserves of
eelgrass Zostera marina. Marine Ecology Progress Series 137:195-201.
Burt, W.V. and W.B. McAlister. 1959. Recent studies in the hydrography of Oregon estuaries.
Research Briefs, Fish Commission of Oregon 7:14-27.
Busby, M.S. 1991. The abundance of epibenthic and planktonic macrofauna and feeding habits
of juvenile fall Chinook salmon (Oncorhynchus tshawytscha) in the Mattole River
estuary/lagoon, Humboldt County, California. M.S. Thesis, Humboldt County, Arcata, CA.
130 p.
Cabello-Pasini, A., C. Lara-Turrent, and R.C. Zimmerman. 2002. Effect of storms on
photosynthesis, carbohydrate content and survival of eelgrass populations from a coastal
lagoon and adjacent open ocean. Aquatic Botany 74:149-164.
Chan, F., J.A. Barm, J. Lubchenco, A. Kirincich, A. Weeks, W.T. Peterson, and B.A. Menge.
2008. Emergence of anoxia in the California Current large marine ecosystem. Science
319,920.
250
-------
Choi, B. 1975. Pollution and tidal flushing predictions for Oregon's estuaries. M.S. Thesis,
Oregon State University, Corvallis, OR. 163 p.
Clarke, K.R. and R.N. Gorley. 2006. Primer Version 6 User Manual, Tutorial. PRIMER-E,
Plymouth, UK. 190 p.
Clarke, K.R. and R.M. Warwick. 2001. Change in Marine Communities: An Approach to
Statistical Analysis and Interpretation. 2nd edition: PRIMER-E, Plymouth, UK. 172 p.
Clinton, P.J., D.R. Young, D.T. Specht, and H. Lee, II. 2007. A guide to mapping intertidal
eelgrass and nonvegetated habitats in estuaries of the Pacific Northwest USA. USEPA
Office of Research and Development, National Health and Environmental Effects
Laboratory, Western Ecology Division. EPA/600/R-07/062.
Cloern, I.E. 2001. Our evolving conceptual model of the coastal eutrophication problem. Marine
Ecology Progress Series 210:223-253.
Cloern, I.E., E.A. Canuel, and D. Harris. 2002. Stable carbon and nitrogen isotope composition
of aquatic and terrestrial plants of the San Francisco Bay estuarine system. Limnology and
Oceanography 47:713-729.
Cohen, R.A. and P. Fong. 2005. Experimental evidence supports the use of 515N content of the
opportunistic green macroalga Enteromorpha intestinalis (Chlorophyta) to determine
nitrogen sources to estuaries. Journal ofPhycology 41:287-293.
Colbert, D.L. 2004. Geochemical cycling in a Pacific Northwest Estuary (Tillamook Bay,
Oregon, USA), PhD. Dissertation, Oceanography Department, Oregon State University.,
Corvallis, OR.
Colbert, D. and J. McManus. 2003. Nutrient biogeochemistry in an upwelling-influenced
estuary of the Pacific Northwest (Tillamook Bay, Oregon, USA). Estuaries 26:1205-1219.
Cole, M.L., I. Valiela, K.D. Kroeger, G.L. Tomasky, J. Cebrian, C. Wigand, R.A. McKinney,
S.P. Grady, M.H. Carvalho da Silva. 2004. Assessment of a 5 15N isotopic method to
indicate anthropogenic eutrophi cation in aquatic ecosystems. Journal of Environmental
Quality 33:124-132.
Cole, T.M. and S.A. Wells. 2000. CE-QUAL-W2: A two-dimensional, laterally-averaged,
hydrodynamic and water quality model, Version 3, Instruction Report EL-2000. United
States Army Engineering and Research Development Center, Vicksburg, MS.
Committee on Environment and Natural Resources (CENR). 2003. An Assessment of Coastal
Hypoxia and Eutrophication in U.S. Waters. National Science and Technology Council
Committee on Environment and Natural Resources, Washington, D.C. 74 p.
Compton, I.E., M.R. Church, S.T. Larned, and W.E. Hogsett. 2003. Nitrogen export from
forested watersheds in the Oregon Coast Range: The role of N2-fixing red alder.
Ecosystems 6:773-785.
251
-------
Congalton, R.G. and K. Green. 1999. Assessing the Accuracy of Remotely Sensed Data:
Principals and Practices. Lewis Publishers, CRC Press, Inc.
Cortright, R., J. Weber, and R. Bailey. 1987. The Oregon estuary plan book. Oregon
Department of Land Conservation and Development. Salem, OR. Digital Version Available
at: http://www.inforain.org/mapsatwork/oregonestuary/
Costanzo, S.D., MJ. O'Donohue, W.C. Dennison, N.R. Loneragan, and M. Thomas. 2001. A
new approach for detecting and mapping sewage impacts. Marine Pollution Bulletin
42:149-156.
Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of Wetlands and
Deepwater Habitats of the United States. United States Department of the Interior, Fish
and Wildlife Service, Washington, D.C. 131 p.
D'Andrea, A.F. and T.H. DeWitt. 2003. Density-dependent impacts of burrowing shrimp on
benthic fluxes in Pacific Northwest estuaries: Applicability for estuarine-scale models of
nitrogen cycling. Invited presentation, special session on nutrient cycling: benthic-pelagic
interactions, Estuarine Research Federation (ERF) 2003 Meeting, Seattle, WA, September
14-18,2003.
Daly, C., J.W. Smith, J.I. Smith, and R.B. McKane. 2007. High-resolution spatial modeling of
daily weather elements for a catchment in the Oregon Cascade mountains, United States. J.
AppliedMeterology and Climatology 46:1565-1586.
Davis, M.W. 1981. Production dynamics of sediment-associated algae in two Oregon estuaries.
Ph.D. Dissertation. Oregon State University, Corvallis, OR. 135 p.
De Angelis, M.A. and L.I. Gordon. 1985. Upwelling and river runoff as sources of dissolved
nitrous oxide to the Alsea estuary, Oregon. Estuarine, Coastal and Shelf Science 20:375-
386.
den Hartog, C. 1977. Structure, function and classification in seagrass communities. In C.P.
McRoay and C. Heifferich (eds.) Seagrass Ecosystems: A Scientific Perspective. Marcel
Dekker, Inc., NY. pp. 89-121.
Dennison, W.C. 1987. Effects of light on seagrass photosynthesis, growth and depth
distribution. Aquatic Botany 27:15-26.
Dennison, W.C., RJ. Orth, K. Moore, J.C. Stevenson, V. Carter, S. Kollar, P.W. Bergstrom, and
R.A. Batuik. 1993. Assessing water quality with submersed aquatic vegetation.
BioScience 43:86-94.
Dettmann, E.H. 2001. Effects of water residence time on annual export and denitrification of
nitrogen in estuaries: A model analysis. Estuaries 24:481-490.
Dettmann, E.H. and J.C. Kurtz. 2006. Proposed Classification Scheme for Coastal Receiving
Waters Based on SAV and Food Web Sensitivity to Nutrients. Volume 1: Responses of
252
-------
Seagrass and Phytoplankton in Estuaries of the Eastern United States to Nutrients:
Implications for Classification. Office of Research and Development, National Health and
Environmental Effects Research Laboratory. EPA Report.
DeWitt, T.H., A.F. D'Andrea, C.A. Brown, B.D. Griffen, and P.M. Eldridge. 2004. Impact of
burrowing shrimp populations on nitrogen cycling and water quality in western North
American temperate estuaries, pp. 107-118. In A. Tamaki (ed.) Proceedings of the
Symposium on Ecology of Large Bioturbators in Tidal Flats and Shallow Sublittoral
Sediments—From Individual Behavior to their Role as Ecosystem Engineers. Nagasaki
University, Nagasaki, Japan. November 1-2, 2003.
Diaz-Ramos, S., D.L. Stevens Jr., and A.R. Olsen. 1996. EMAP Statistics Methods Manual.
United States Environmental Protection Agency, Office of Research and Development,
National Health and Environmental Effects Research Laboratory, Western Ecology
Division, Corvallis, OR. EPA/620/R-96002
Dobson, J., E.A. Bright, R.L. Ferguson, L.L. Wood, K.D. Haddad, H. Iredale III, J.R. Jenson,
V.V. Klemas, R.J. Orth, and J.P Thomas. 1995. NOAA Coastal Change Analysis Program
(C-CAP). Guidance for Regional Implementation. United States Department of Commerce,
National Oceanic and Atmospheric Administration, Charleston, SC.
Dowtry, P., B. Reeves, H. Berry, S. Wyllie-Echeverria, T. Mumford, A. Sewell, P. Milos, and R.
Wright. 2005. Puget Sound submerged vegetation monitoring project 2003-2004
monitoring report. Washington Department of Natural Resources. Olympia WA. 65 p.
and appendices.
Driscoll, C.T., D. Whitall, J. Aber, E. Boyer, M. Castro, C. Cronan, C.L. Goodale, P. Groffman,
C. Hopkinson, K. Lambert, G. Lawrence, and S. Ollinger. 2003. Nitrogen pollution in the
northeastern United States: sources, effects, and management options. BioScience 53:357-
374.
Drut, M.S. and J.B. Buchanan. 2000. U.S. Shorebird Conservation Plan, Northern Pacific Coast
Regional Shorebird Management Plan. United States Department of the Interior, Fish and
Wildlife Service. 31 p.
http://www.fws.gov/shorebirdplan/RegionalShorebird/RegionalPlans.htm
Duarte, C.M. 1991. Seagrass depth limits. Aquatic Botany 40:363-377.
Dudoit, C.M. 2006. The distribution and abundance of a non-native eelgrass, Zosterajaponica,
in Oregon estuaries. Senior thesis, Oregon State University, Corvallis, OR.
Dumbauld, B.R., D.L. Holden, and O.P. Langness. 2008. Do sturgeon limit burrowing shrimp
populations in Pacific Northwest estuaries? Environmental Biology of Fishes 83:283-296.
Dumbauld, B.R. and S. Wyllie-Echeverria. 2003. The influence of burrowing thalassinid
shrimps on the distribution of intertidal seagrasses in Willapa Bay, Washington, USA.
Aquatic Botany 77:27-42.
253
-------
Dyer, K.R. and P. A. Taylor. 1973. A simple, segmented prism model of the tidal mixing in
well-mixed estuaries. Estuarine and Coastal Marine Science 1:411-418.
Elliott, M. and D.S. McLusky. 2002. The need for definitions in understanding estuaries.
Estuarine, Coastal and Shelf Science 55:815-827.
Elwany, M.H., R.E. Flick, and S. Aijaz. 1998. Opening and closing of a marginal Southern
California lagoon inlet. Estuaries 21:246-254.
Emmett, R., R. Llanso, J. Newton, R. Thorn, M. Hornberger, C. Morgan, C. Levings, A.
Copping, and P. Fishman. 2000. Geographic signatures of North American West Coast
estuaries. Estuaries 23:7'65-792.
Estevez, E.D. 2000. Matching salinity metrics to estuarine seagrasses for freshwater inflow
management, pp. 295-308 In S.A. Bortone (ed.) Seagrasses: Monitoring, Ecology,
Physiology, and Management. CRC Press. Boca Raton, FL
Federal Register. 2002. Endangered and Threatened Wildlife and Plants; Withdrawal of
Proposed Rule To Remove the Northern Populations of the Tidewater Goby From the List
of Endangered and Threatened Wildlife. Volume 67:67803-67818.
Feldman, K.L., D.A. Armstrong, B.R. Dumbauld, T.H. Dewitt, and D.C. Doty. 2000. Oysters,
crabs, and burrowing shrimp: review of an environmental conflict over aquatic resources
and pesticide use in Washington State's (USA) coastal estuaries. Estuaries 23:141-176.
Ferraro, S. P. and F.A. Cole. 2007. Benthic Macrofauna-habitat associations in Willapa Bay,
Washington, USA. Estuarine, Coastal and Shelf Science 71:491-507.
Fielding, A.H. and J.F. Bell. 1997. A review of methods for the assessment of prediction errors
in conservation presence/absence models. Environmental Conservation 24(1): 38-49
Fong, P., K.E. Boyer, and J.B. Zedler. 1998. Developing an indicator of nutrient enrichment in
coastal estuaries and lagoons using tissue nitrogen content of the opportunistic alga,
Enteromorpha intestinalis (L. Link). Journal of Experimental Marine Biology Ecology
231:63-79.
Fonseca, M.S., WJ. Kenworthy, and G.W. Thayer. 1998. Guidelines for the Conservation and
Restoration of Seagrasses in the United States and Adjacent Waters. United States
Department of Commerce, National Oceanic and Atmospheric Administration, Coastal
Ocean Program Decision Analysis Series No. 12. NOAA Coastal Ocean Office, Silver
Spring MD. 222 p.
Fry, B. 2002. Conservative mixing of stable isotopes across estuarine gradients: a conceptual
framework for monitoring watershed influences on downstream fisheries production.
Estuaries 25:264-271.
Fry, B., A. Gace, and J.W. McClelland. 2001. Chemical indicators of anthropogenic nitrogen
loading in West coast NERR estuaries. Final Report submitted to NOAA/UNH
254
-------
Cooperative Institute for Coastal Environmental Technology (CICEET). 79 p.
Fry, B., A. Gace and J.W. McClelland. 2003. Chemical indicators of anthropogenic nitrogen
loading in four Pacific Estuaries. Pacific Science 57:77-101.
Grantham, B.A., F. Chan, KJ. Nielsen, D.S. Fox, J.A. Earth, A. Huyer, J. Lubchenco, and B.A.
Menge. 2004. Upwelling-driven nearshore hypoxia signals ecosystem and oceanographic
changes in the northeast Pacific. Nature 429:749-754
Griffen, B.D., T.H. DeWitt, and C. Langdon. 2004. Particle removal rates by the mud shrimp
Upogebiapugettensis, its burrow, and a commensal clam: effects on estuarine
phytoplankton abundance. Marine Ecology Progress Series 269:223-236.
Haertel, L. and C. Osterberg. 1967. Ecology of zooplankton, benthos and fishes in the
Columbia River Estuary. Ecology 48(3):459-472.
Hamilton, P. 1984. Hydrodynamic Modeling of the Columbia River Estuary. Report to the
Columbia River Estuary Data Development Program (CREST). 354 p.
Hansen, D.V. and M. Rattray. 1966. New dimensions in estuary classification. Limnology and
Oceanography 11:319-326.
Harrison, P.G. and R.E. Bigley. 1982. The recent introduction of the seagrass Zostera japonica
to the Pacific Coast of North America. Canadian Journal of Fisheries and Aquatic
Sciences 39:1642-1648.
Hauxwell, J., J. Cebrian, and I. Valiela. 2003. Eelgrass Zostera marina loss in temperate
estuaries: relationship to land-derived nitrogen loads and effects of light limitation imposed
by algae. Marine Ecology Progress Series 247:59-73.
Heaton, T.H.E. 1986. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere:
a review. Chemical geology 59:87-102.
Heck, K.L. Jr., K.W. Able, M.P. Fahay, and C.T. Roman. 1989. Fishes and decapod crustaceans
of Cape Cod eelgrass meadows: species composition, seasonal abundance patterns and
comparison with unvegetated substrates. Estuaries 12:59-65.
Hemminga, M.A. and C.M. Duarte. 2000. Seagrass Ecology. Cambridge University Press,
New York, NY.
Hickey, B.M. and N.S. Banas. 2003. Oceanography of the U.S. Pacific northwest coastal ocean
and estuaries with application to coastal ecology. Estuaries 26:1010-1031.
Hickey, B.M., X. Zhang, and N. Banas. 2002. Coupling between the California Current System
and a coastal plain estuary in low riverflow conditions. Journal of Geophysical Research
107 (CIO), 3166, doi: 10.1029/1999JCOOO160.
Hobbie, E.A., S.A. Macko, and M. Williams. 2000. Correlations between foliar 815N and
255
-------
nitrogen concentrations may indicate plant-mycorhizal interactions. Oecologia 122:273-
283.
Holland, A.F., D.M. Sanger, C.P. Gawle, S.B. Lerberg, M.S. Santiago, G.H.M. Riekerk., L.E.
Zimmerman, and G.I. Scott. 2004. Linkages between tidal creek ecosystems and the
landscape and demographic attributes of their watersheds. Journal of Experimental Marine
Biology and Ecology 298:151-178.
Homer, C., C. Huang, L. Yang, B. Wylie, and M. Coan. 2004. Development of a 2001 national
land-cover database for the United States. Photogrammetic Engineering & Remote Sensing
70(7):829-840.
Hull, S.C. 1987. Macroalgal mats and species abundance: Afield experiment. Estuarine,
Coastal and Shelf Science 25:519-532.
Humboldt County Department of Health and Human Services. 2005. Blue green algae (EGA)
detailed fact sheet. Prepared by: Harriet Hill, Division of Environmental Health, revised
June 2005.
http://www.swrcb.ca.gov/bluegreenalgae/docs/workshopll0805/bgadetailedfactsheet.pdf.
Hyland, J.L., T.J. Herrlinger, T.R. Snoots, A.H. Ring-wood, R.F. Van Dolah, C.T. Hackney,
G.A. Nelson, J.S. Rosen, and S.A. Kokkinakis. 1996. Environmental Quality of Estuaries
of the Carolinian Province: 1994. Annual Statistical Summary for the 1994 EMAP-
Estuaries Demonstration Project in the Carolinian Province. United States Department of
Commerce, National Oceanic and Atmospheric Administration, Office of Ocean Resources
Conservation and Assessment, Silver Spring, MD. NOAA Technical Memorandum NOS
ORCA97. 102 p.
Interior Columbia Basin Ecosystem Management Project (ICBEMP). 1996. Interior Columbia
Basin Ecosystem Management Project. Columbia River Basin (CRB) Boundary.
http://www.icbemp.gov/spatial/hydro/.
Jackson, E.L., A. A. Rowden, M.J. Attrill, S.J. Bossey, and M.B. Jones. 2001. The importance
of seagrass beds as habitat for fishery species. Oceanography and Marine Biology 39:269-
303.
John Gilchrist & Associates and Fall Creek Engineering, Inc. 2005. Technical Report Moran
Lake Water Quality Study & Conceptual Restoration Plan. Prepared For: Santa Cruz
County Redevelopment Agency, Santa Cruz, CA. http://sccounty01.co.santa-
cruz.ca.us/red/Moran%20Lake%20Report.htm.
Johnson, G.E. and J.J. Gonor. 1982. The tidal exchange ofCallinassa californiensis (Crustacea,
Decapoda) larvae between the ocean and the Salmon River estuary, Oregon. Estuarine,
Coastal and Shelf Science 14:501-516.
Jones, A., W. Dennison, F. Pantus. 2003. Assessment of sewage and septic derived N in the
Choptank and Patuxent Rivers. Final Report. Maryland Coastal Zone Management
256
-------
Program, Department of Natural Resources pursuant to NOAA award No. NA17OZ1124.
http://ian.umces.edu/pdfs/chop_pat_final_report.pdf
Jones, C.G., Lawson, J.H. and Shachak, M. 1997. Positive and negative effects of organisms as
physical ecosystem engineers. Ecology 78:1946-1957.
Kaldy, I.E. 2006. Production ecology of the non-indigenous seagrass, dwarf eelgrass (Zostera
japonica Ascher & Graeb) in a Pacific Northwest Estuary, USA. Hydrobiologia 553:201-
217.
Kaldy, I.E. and K.S. Lee. 2007. Factors controlling Zostera marina L. growth in the eastern and
western Pacific Ocean: Comparisons between Korea and Oregon, USA. Aquatic Botany
87:116-126
Kendall, C. and JJ. McDonnell (eds.). 1998. Isotope Tracers in Catchment Hydrology. Elsevier
Science, Amsterdam, The Netherlands, pp. 519-576.
Kentula, M.E. and C.D. Mclntire. 1986. The autoecology and production dynamics of eelgrass
(Zostera marina L.) in Netarts Bay, Oregon. Estuaries 9:188-199.
Kentula, M.E. and T.H. DeWitt. 2003. Abundance of seagrass (Zostera marina L.) and
macroalgae in relation to the salinity-temperature gradient in Yaquina Bay, Oregon.
Estuaries 26:1130-1141.
Kienast, S.S., S.E. Calvert, and T.F. Pedersen. 2002. Nitrogen isotope and productivity
variations along the northeast Pacific margin over the last 120 kyr: Surface and subsurface
paleoceanography. Pale oceanography 17(4): 105 5.
Koch, E.W. 2001. Beyond light: Physical, geological, and geochemical parameters as possible
submersed aquatic vegetation habitat requirement. Estuaries 24:1-17.
Koch, E.W. and S. Beer. 1996. Tides, light and the distribution of Zostera marina in Long
Island Sound, USA. Aquatic Botany 53:97-107.
Kosro, P.M., W.T. Peterson, B.M. Hickey, R.K. Shearman, and S.D. Pierce. 2006. Physical
versus biological spring transition. 2005. Geophysical Research Letters ?>?>, L22S03,
doi:10.1029/2006GL027072.
Kurtz, J.C., N.D. Detenbeck, V.D. Engel, L.M. Smith, SJ. Jordan, and D. Campbell. 2006.
Classifying coastal waters: current necessity and historical perspective. Estuaries 29:107-
123.
Lackey, R.T. 2004. A salmon-centric view of the twenty-first century in the western United
States, pp. 131-137, In P Gallaugher and L. Wood (eds.) Proceedings of the World Summit
on Salmon, Simon Fraser University, Burnaby, British Columbia, Canada.
Lackey, R.T., D.H. Lach, and S.L. Duncan. 2006a. Policy options to reverse the decline of wild
Pacific salmon. Fisheries 31:344-351.
257
-------
Lackey, R.T. , D.H. Lach, and S.L. Duncan. 2006b. Salmon 2100: The Future of Wild Salmon.
American Fisheries Society, Bethesda, MD. 629 p.
Lackey, R. T., D.H. Lach, and S.L. Duncan. 2006c. Wild salmon in western North America: the
historical and policy context, pp. 13-55. In R.T. Lackey, D.H. Lach, and S.L. Duncan
(eds.) Salmon 2100: The Future of Wild Pacific Salmon. American Fisheries Society,
Bethesda, MD.
Lamberson, J.O. and W.G. Nelson. 2002. West Coast Field Sampling Methods - Intertidal
2002. Unpublished modification to the EMAP Project Plan . United States Environmental
Protection Agency.
http://yosemite.epa.gOv/r 10/OEA.NSF/Monitoring/Coastal+Indicators/$FILE/intertidal+pr
otocol+.pdf
Landis, J.R. and G.C. Koch. 1977. The measurement of observer agreement for categorical
data. Biometrics 3 3:15 9-174.
Larned, S.T. 2003. Effects of the invasive, nonindigenous seagrass Zoster a japonica on nutrient
fluxes between the water column and benthos in a NE Pacific estuary. Marine Ecology
Progress Series 254:69-80.
Latimer, J.S. and J.R. Kelly. 2003. Proposed classification scheme for predicting sensitivity of
coastal receiving waters to effects of nutrients. United States Environmental Protection
Agency, Office of Research and Development, National Health and Environmental Effects
Research Laboratory, Atlantic Ecology Division, Narragansett, RI. AED-03-04-001. 14 p.
Lawson, P.W., E. Bjorkstedt, M. Chilcote, C. Huntington, J. Mills, K. Moore, T.E. Nickelson,
G.H. Reeves, H.A. Stout, and T.C. Wainwright. 2004. Identification of historical
populations of Coho Salmon (Oncorhynchus kisutch) in the Oregon coast evolutionarily
significant unit. Review Draft. Oregon Northern California Coast Technical Recovery
Team. United States Department of Commerce, National Oceanic and Atmospheric
Administration NOAA/NMFS/NWFSC. 129 p.
Lee II, H. and D. A. Reusser, with contributions from K. Welch, R. Fairey, and L. Hillmann.
2006. Pacific Coast Ecosystem Information System (PCEIS) V. 1.0. United States
Environmental Protection Agency and United States Geological Survey (Microsoft Access
2003 database).
Lee II, H., C.A. Brown, B.L. Boese, and D.R. Young (eds.). 2006. Proposed Classification
Scheme for Coastal Receiving Waters Based on SAV and Food Web Sensitivity to
Nutrients. Volume 2: Nutrient Drivers, Seagrass Distributions, and Regional Classifications
of Pacific Northwest Estuaries. Office of Research and Development, National Health and
Environmental Effects Research Laboratory. U.S. EPA Report.
Likens, G.E., F.H. Bormann, N.M. Johnson, D.W. Fisher, and R.S. Pierce. 1970. Effects of
forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-
ecosystem . Ecological Monographs 40:23-47.
258
-------
Markham, J.C. 2004. New species and records of Bopyridae (Crustacea: Isopoda) infesting
species of the genus Upogebia (Crustacea: Decapoda: Upogebiidae): the genera Orthione
Markham, 1988, and Gyge (Cornalia & Panceri, 1861). Proceedings of the Biological
Society of Washington 117:186-198.
McCabe, G.T., R.L. Emmett, T.C. Coley, and RJ. Mcconnell. 1988. Distribution, density, and
size-class structure of Dungeness crabs in the river-dominated Columbia River estuary.
Northwest Science 62:254-263.
McClelland, J.W. and I. Valiela. 1998. Linking nitrogen in estuarine producers to land-derived
sources. Limnology and Oceanography 43:577-585.
McCune, B. and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software,
Gleneden Beach, Oregon, USA With a contribution by Dean L. Urban. 304 p.
Merritt Smith Consulting. 1999. Biological and water quality monitoring in the Russian River
Estuary. Third Annual Report, 15 March 1999. Prepared for Klamath Resource Information
System (KRIS).
http://www.kri sweb.com/krisrussian/kri sdb/html/krisweb/biblio/russian_scwa_merrittsmith
_1999_estmon3.pdf
Monsen, N.E., I.E. Cloern, L.V. Lucas, and S.G. Monismith. 2002. A comment on the use of
flushing time, residence time, and age as transport time scales. Limnology and
Oceanography 47:1545-1553.
Montegue, C.L. and J.A. Ley. 1993. A possible effect of salinity fluctuation on abundance of
benthic vegetation and associated fauna in northeastern Florida Bay. Estuaries 16:707-
717.
National Research Council. 2000. Clean coastal waters: understanding and reducing the effects
of nutrient pollution. Ocean Studies Board and Water Science and Technology Board,
Commission on Geosciences, Environment, and Resources, National Research Council.
National Academy Press, Washington, D.C. 405 p.
Naymik, J., Y. Pan, and J. Ford. 2005. Diatom assemblages as indicators of timber harvest
effects in coastal Oregon streams. Journal of the North American Benthological Society
24:569-584.
Nelson, T.A. and A. Lee. 2001. A manipulative experiment demonstrates that blooms of the
macroalgae Ulvaria obscura can reduce eel grass shoot density. Aquatic Botany 71:149-
154.
Nelson, W.G. and C.A. Brown. 2008. Use of probability-based sampling of water-quality
indicators in supporting development of quality criteria. ICES Journal of Marine Science
65:1421-1427.
Nelson, W.G., H. Lee II, J.O. Lamberson, V. Engle, L. Harwell, and L.M. Smith. 2005a.
Condition of estuaries of California for 1999: a statistical summary. United States
259
-------
Environmental Protection Agency, Office of Research and Development, National Health
and Environmental Effects Research Laboratory, Western Ecology Division. EPA/620/R-
05/004.
Nelson, W.G., H. Lee II, J.O. Lamberson, V. Engle, L. Harwell, and L.M. Smith. 2005b.
Condition of estuaries of the western United States for 1999: a statistical summary. United
States Environmental Protection Agency, Office of Research and Development, National
Health and Environmental Effects Research Laboratory, Western Ecology Division.
EPA/620/R-04/200.
Nelson, W.G., H. Lee II, J.O. Lamberson, F. Cole, C. Weilhoefer, and P. Clinton. 2007. The
condition of tidal wetlands of Washington, Oregon, and California - 2002. United States
Environmental Protection Agency, Office of Research and Development, National Health
and Environmental Effects Research Laboratory, Western Ecology Division. EPA/620/R-
07/002.
Newton, J.A. and R.A. Horner. 2003. Use of phytoplankton species indicators to track the
origin of phytoplankton blooms in Willapa Bay, Washington. Estuaries 26(48): 1071-1078.
Nielsen, S.L., J. Borum, O. Geertz-Hansen, and J. Sand-Jensen. 1989. Marine bundplanters
dybdegraense. VandMiljo 5:217-220.
Nixon, S., B. Buckley, S. Granger, and J. Bintz. 2001. Response of very shallow marine
ecosystems to nutrient enrichment. Human and Ecological Risk Assessment 7:1457-1481.
Nixon, S.W. 1995. Coastal marine eutrophication: A definition, social causes, and future
concerns. Ophelia 41:199-219.
Nomme, K.M. and P.G. Harrison. 1991. Evidence for interaction between the seagrasses
Zoster a marina and Zoster a japonica on the Pacific Coast of Canada. Canadian Journal of
Botany 69:2004-2010.
O'Neal, S.L., D. Hallock and K. Smith. 2001. Water quality assessments of selected lakes
within Washington State: 1999. Washington State Department of Ecology, Olympia, WA.
Publication no. 01-03-009. 196 p.
Oregon Department of Fish and Wildlife. 1979. Natural Resources of Rogue Estuary. Estuary
Inventory Report. Vol. 2, No. 8. 33 p.
Oregon Division of State Lands. No year. Wetlands Inventory User's Guide: National Wetland
Inventory and Local Wetlands Inventories.
http://www.oregon.gov/DSL/WETL AND/docs/wi_user_guide.pdf
Orth, R J. and K.A. Moore. 1983. Chesapeake Bay: An unprecedented decline in submerged
aquatic vegetation. Science 222:51-53.
260
-------
Orth, RJ. and K.A. Moore. 1988. Distribution of Zostera marina L. and Ruppia maritima L.
sensu lato along depth gradients in the lower Chesapeake Bay, U.S.A. Aquatic Botany
32(3):291-305.
Parrish, J.K., K. Bell, E. Logerwell, and C. Roegner. 2001. Indicators of West Coast Estuary
Health, pp. 1-20. In J.K. Parrish and K. Litle (eds.) Pacific Northwest Coastal Ecosystems
Regional Study, 2000 Annual Report. United States Department of Commerce, National
Oceanic and Atmospheric Administration, Coastal Ocean Programs.
Percy, K.L., D.A. Bella, C. Sutterlin, and P.C. Klingeman. 1974. Descriptions and information
sources for Oregon estuaries. Sea Grant College Program, Oregon State University,
Corvallis, OR. 294 p.
Phillips, D.L. and J.W. Gregg. 2003. Source partitioning using stable isotopes: Coping with too
many sources. Oecologia 136:261-269.
Phillips, R.C. 1984. The ecology of eelgrass meadows in the Pacific Northwest: a community
profile. United States Department of the Interior, Fish and Wildlife Service, FWS/OBS-
84/24. 85 p.
Posey, M.H. 1988. Community changes associated with the spread of an introduced seagrass,
Zostera japonica. Ecology 69:974-983.
Pritchard, D. 1955. Estuarine circulation patterns. Proceedings of the American Society of Civil
Engineers 81:1-11.
Pritchard, D. 1967. What is an estuary: physical viewpoint. In G. Lauff (ed.) Estuaries.
Washington, D.C. American Association for the Advancement of Science. Publication 83.
Quinn, H., D.T. Lucid, J.P. Tolson, CJ. Klein, S.P. Orlando, and C. Alexander. 1991.
Susceptibility and status of West coast estuaries to nutrient discharges: San Diego Bay to
Puget Sound. Summary Report. NOAA/EPA. Rockville, MD. 35 p.
Reusser, D.A., H. Lee II, L. Hillmann, and D.A. Kluza. 2004. The Pacific Coast Estuarine
Information System: creating a baseline for the future [abstract]. The 13th International
Conference on Aquatic Invasive Species, Ennis, Ireland, September 20-24, 2004.
Robbins, B. and B. Boese. 2002. Macroalgal volume: A surrogate for biomass in some green
algae. BotanicaMarina 45:586-588.
Roegner, G.C. and A. Shanks. 2001. Import of coastally-derived chlorophyll a to South Slough,
Oregon. Estuaries 24:224-256.
Roegner, G.C., B.M. Hickey, J.A. Newton, A.L. Shanks, and D.A. Annstrong. 2002. Wind-
induced plume and bloom intrusions into Willapa Bay, Washington. Limnology and
Oceanography 47:103 3-1042.
Ruesink, J.L., B.E. Feist, CJ. Harvey, J.S. Hong, A.C. Trimble, and L.M. Wisehart. 2006.
261
-------
Changes in productivity associated with four introduced species: ecosystem transformation
of a'pristine'estuary. Marine Ecology Progress Series 311:203-215.
Rumrill, S. 1998. Habitat variability and function in Pacific Northwest Estuaries. Meeting
Summary: Protecting and Restoring Pacific Northwest Estuaries: The Task Before Us.
Pacific Northwest Coastal Ecosystems Regional Study (PNCERS), Troutdale, OR,
December 8-9, 1998. pp. 12-23.
Scavia, D. and S.B. Bricker. 2006. Coastal eutrophication in the United States.
Biogeochemistry 79:187-208.
Seliskar, D.M., and J.L. Gallagher. 1983. The ecology of tidal marshes of the Pacific Northwest
coast: a community profile . U.S. Fish and Wildlife Service, Division of Biological
Services, Washington, D.C. FWS/OBS-82/32. 65 pp.
Selleck, J.R., H.D. Berry, and P. Dowty. 2005. Depth profiles of Zostera marina throughout
greater Puget Sound: Results from 2002-2004 monitoring data. Washington Department of
Natural Resources. Olympia, WA. 14 p.
Sewell A.T., J.G. Norris, and S. Wyllie-Echeverria. 2001. Eelgrass monitoring in Puget Sound:
Overview of the submerged vegetation monitoring project.
www.psat. wa.gov/Publicati ons/01_proceedings/sessions/oral/3a_sewel.pdf.
Shaffer, J.A. 2001. Macroalgae Blooms and Nearshore Habitat and Resources of the Strait of
Juan de Fuca. Proceedings of Puget Sound Research 2001. Fifth Puget Sound Research
Conference, Bellevue, Washington. February 12-14, 2001. lip.
Sheldon, I.E. and M. Alber. 2002. A comparison of residence time calculations using simple
compartment models of the Altamaha River Estuary, Georgia. Estuaries 25:1304-1317.
Shirzad, F.F., S.P. Orlando., CJ. Klein., S.E. Holliday., M.A. Warren, and M.E. Monaco. 1988.
National Estuarine Inventory: Supplement 1, Physical and Hydrologic characteristics, The
Oregon estuaries. United States Department of Commerce, National Oceanic and
Atmospheric Administration, Rockville, MD.
Short, F.T., R.G. Coles, and C. Pergent-Martini. 2001. Global Seagrass Distribution. In F.T.
Short, R.G. Coles, and C.A. Short (eds.), Global Seagrass Research Methods. Elsevier
Science, Amsterdam, The Netherlands. 473 p.
Simmons, H.B. 1955. Some effects of upland discharge on estuarine hydraulics. Proceedings
of the American Society of Civil Engineers 81:792/l-20.
Smith, A.E., J.W. Chapman, and B.R. Dumbauld. 2008. Population structure and energetics of
the bopyrid isopod parasite Orthione griffenis in mud shrimp Upogebiapugettensis.
Journal of Crustacean Biolology 28:228-233.
Smith, G. A. 1991. NWI maps made easy - A user's guide to National Wetlands Inventory maps
of the Northeast Region. U.S. Fish and Wildlife Service, 13 p.
262
-------
Sogard, S.M. and K.W. Able. 1991. A comparison of eelgrass, sea lettuce macroalgae, and
marsh creeks as habitats for epibenthic fishes and decapods. Estuarine, Coastal, and Shelf
Science 33:501-519.
Spalding, M.D., H.E. Fox, G.R. Allen, N. Davidson, Z.A. Ferdana, M. Finlayson, B.S. Halpern,
M.A. Jorge, A. Lombana, S.A. Lourie, K.D. Martin, E. McManus, J. Molnar, C.A. Recchia,
and J. Robertson. 2007. Marine ecoregions of the world: A bioregionalization of coastal
and shelf areas. Bioscience 57:573-583.
Stenzel, L.E., H.R. Huber, and G.W. Page. 1976. Feeding behavior and diet of the long-billed
Curlew and Willet. The Wilson Bulletin 88:314-332.
Stevens, Jr., D.L. 1997. Variable density grid-based sampling designs for continuous spatial
populations. Environmetrics 8:167-195.
Stevens, Jr., D.L. and A.R. Olsen. 1999. Spatially restricted surveys over time for aquatic
resources. Journal of Agricultural, Biological and Environmental Statistics 4:415-428.
Strobel, C.J., H.W. Buffum, S.J. Benyi, E.A. Petrocelli, D.R. Reifsteck, and D.J. Keith. 1995.
Statistical Summary: EMAP Estuaries-Virginian Province-1990 to 1993. United States
Environmental Protection Agency, National Health and Environmental Effects Laboratory,
Atlantic Ecology Division, Narragansett, RI. EPA/620/R-94/026. 72 p. and appendices
Summers, J.K., J.M. Macauley, P.T. Heitmuller, V.D. Engle, A.M. Adams, and G.T. Brooks.
1993. Annual Statistical Summary: EMAP-Estuaries Louisianian Province-1991. United
States Environmental Protection Agency, Office of Research and Development, Gulf
Ecology Division, Gulf Breeze, FL. EPA/620/R-93/007. 63 p. and appendices.
Sutler, L.A. 2001. Spatial Wetland Assessment for Management and Planning (SWAMP):
Technical Discussion. United States Department of Commerce, National Oceanic and
Atmospheric Administration, Coastal Services Center. Publication Number 20129-CD.
Charleston, SC. 56 p.
Thayer, G.W. and R.C. Phillips. 1977. Importance of eelgrass in beds in Puget Sound. Marine
Fisheries Review 39:18-22.
Thayer, G.W., D.A.Wolfel, and R.B. Williams. 1975. The impact of man on seagrass systems.
American Scientist 63:288-296
Thorn, R.M. 1984. Composition, habitats, seasonal changes and productivity of macroalgae in
Grays Harbor Estuary, Washington. Estuaries 7:51-60.
Thorn, R.M. 1990. Spatial and temporal patterns in plant standing stock and primary production
in a temperate seagrass system. BotanicaMarina 33:497-510.
Thorn, R.M., A.B. Borde, S. Rumrill, D.L. Woodruff, G.D. Williams, J.A. Southhard, and S.L.
Blanton. 2001. Factors influencing the spatial and annual variability of eelgrass (Zostera
marina L.) meadows in Pacific Northwest, USA, systems, pp.38-51. In J.K. Parrish and K.
263
-------
Litle (eds.) Pacific Northwest Coastal Ecosystems Regional Study, 2001 Annual Report.
United States Department of Commerce, National Oceanic and Atmospheric
Administration, Coastal Ocean Programs.
Thorn, R.M., A.B. Borde, S. Rumrill, D.L. Woodruff, G.D. Williams, J.A. Southhard, and S.L.
Sargeant. 2003. Factors influencing spatial and annual variability in eelgrass (Zostera
marina L.) meadows in Willapa Bay, Washington, and Coos Bay, Oregon, estuaries.
Estuaries 26:1117-1129.
Thorn, R.M., L.D. Antrim, A.B. Borde, W.W. Gardiner, O.K. Sheffler, P.O. Farley, J.G. Norris,
S. Wyllie Echeverria, and T.P. McKenzie. 1998. Puget Sound's eelgrass meadows: factors
contributing to depth distribution and spatial patchiness. Puget Sound Research '98
Proceedings. Seattle, WA March 12-13, 1998.
Thorn, R.M, S.L. Southard, A.B. Borde and P. Stoltz. 2008. Light requirements for growth and
survival of eelgrass (Zostera marina L.) in Pacific Northwest (USA) estuaries. Estuaries
and Coasts 31:969-980.
Tjepkema, J.D., C.R. Schwintzer, R.H. Burris, G.V. Johnson, and W.B. Silvester. 2000. Natural
abundance of 15N in actinorhizal plants and nodules. Plant and Soil 219:285-289.
U.S. Environmental Protection Agency. 1997. Urbanization and Streams: Studies of Hydrologic
Impacts. United States Environmental Protection Agency, Office of Water, Washington,
D.C. EPA 841-R-97-009.
U.S. Environmental Protection Agency. 2002. Aquatic Stressors: framework and
implementation plan for effects research. EPA 600/R-02/074. United States Environmental
Protection Agency, National Health and Environmental Effects Research Laboratory,
Research Triangle Park, NC.
U.S. Environmental Protection Agency. 2004a. Classification framework for coastal systems.
EPA Report 600/R-04/061. United States Environmental Protection Agency, National
Health and Environmental Effects Research Laboratory, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 2004b. National Coastal Condition Report II. United
States Environmental Protection Agency, Office of Research and Development, Office of
Water. EPA-620/R-03/002.
U.S. Environmental Protection Agency. 2004c. Analytical Tools Interface for Landscape
Assessments (Attila). User Manual. EPA/600/R-04/083. 2004 Version. 39 p.
http ://www. epa.gov/esd/land-sci/attila/pdf/user_guide.pdf.
U.S. Environmental Protection Agency. 2006. National Estuary Program Coastal Condition
Report. United States Environmental Protection Agency, Office of Research and
Development, Office of Water. EPA-842/B-06/001.
U.S. Fish and Wildlife Service. 2002. National Wetlands Inventory: A Strategy for the 21st
Century. 14 p.
264
-------
Valiela, I., J. McClelland, J. Hauxwell, P.J. Behr, D. Hersh, and K. Foreman. 1997.
Macroalgal blooms in shallow estuaries: controls and ecophysiological and ecosystem
consequences. Limnology and Oceanography 42:1105-1118.
Valiela, I. , K. Foreman, M. Lamontagne, D. Hersh, J. Costar, P. Peckol, B. Demeo-Andreson, C.
D'avanzo, M. Babione, and C.-H. Shams, J.B.L. 1992. Couplings of watersheds and
coastal waters: sources and consequences of nutrient enrichment in Waquoit Bay,
Massachusetts. Estuaries 15:443-457.
Virnstein, R.W., E.W. Carter IV, LJ. Morris, and J.D. Miller. 2002. Utility of seagrass
restoration indices based on area, depth, and light, pp. 69-80. In H.S. Greening (ed.).
Seagrass Management: It's Not Just Nutrients! Tampa Bay Estuary Program, proceedings
of a symposium. St. Petersburg, FL.
Wankel, S.D., C. Kendall, C.A. Francis, and A. Paytan. 2006. Nitrogen sources and cycling in
San Francisco Bay Estuary: A nitrate dual isotopic composition approach. Limnology and
Oceanography 51(4): 1654-1664.
Wetz, J.J., J. Hill, H. Corwith, and P. A. Wheeler. 2005. Nutrient and extracted chlorophyll data
from the GLOBEC Long-Term Observation Program, 1997-2004. Data Report 193, COAS
Reference 2004-1, College of Oceanic and Atmospheric Sciences (COAS), Oregon State
University, Corvallis, OR.
Whitfield, A. and G. Bate (eds.). 2007. A review of information on temporarily open/closed
estuaries in the warm and cool temperate biogeographic regions of South Africa, with
particular emphasis on the influence of river flow on these systems. Interim report to the
Water Research Commission. WRC Report No 1581/1/07. 214 p.
Wigington, P.J., Jr., M.R. Church, T.C. Strickland, K.N. Eshleman, and J. Van Sickle. 1998.
Autumn chemistry of Oregon coast range streams. Journal of the American Water
Resources Association 34:1035-1049.
Williams, G.D., J. West, M. Cordrey, and K. Ward. 1999. The physical, chemical, and
biological monitoring of Los Penasquitos Lagoon (1999). Annual Report for Los
Pefiasquitos Lagoon Foundation, Cardiff, CA. 44 p.
Wonham, M.J. 2003. Ecological gambling: expendable extinctions vs. acceptable invasions, pp
179-205. In P. Kareiva and S. Levin (eds.) The Importance of Species. Princeton, NJ,
University Press Princeton.
Wyllie-Echeverria, S. and J.D. Ackerman. 2003. The seagrasses of the Pacific Coast of North
America, pp. 199-206. In E.P. Green and F.T. Short (eds.) World Atlas of Seagrasses:
present status and future conservation. Prepared by the UNEP World Conservation
Monitoring Centre. University of California Press, Berkeley, CA. 298 p.
Yang, L, C. Huang, C.G. Homer, B.K. Wylie, and M.J. Coan. 2003. An approach for mapping
large- area impervious surfaces: synergistic use of Landsat-7 ETM+ and high spatial
resolution imagery. Canadian Journal of Remote Sensing 29:230-240.
265
-------
Young, D. R., D.T. Specht, B.D. Robbins, and P.J. Clinton. 1999. Delineation of Pacific
Northwest SAVs from aerial photography: natural color or color infrared film? Proceedings
of the Annual Conference of the American Society of Photogrammetry and Remote
Sensing, Bethesda, MD. pp. 1173-1178.
Young, D.R., P.J. Clinton, and D.T. Specht. In press. Mapping intertidal eelgrass (Zostera
marina L.) in three coastal estuaries of the Pacific Northwest USA using false-color near-
infrared aerial photography. InternationalJournal of Remote Sensing.
Young, D.R., P.J. Clinton, P.J., D.T. Specht, T.H. DeWitt, and H. Lee II. 2008. Monitoring the
expanding distribution of nonindigenous dwarf eelgrass Zostera japonica in a Pacific
Northwest USA estuary using high resolution digital aerial orthophotography. Journal of
Spatial Science 53:87-97.
Zedonis, P., R. Turner, and N. Hetrick. 2008. Water Quality Dynamics of the Mattole River
Lagoon and Suitability for Rearing Salmonids in 2006. U.S. Fish and Wildlife Service,
Arcata Fish and Wildlife Office, Arcata Fisheries Technical Series Report Ts 2008-01,
Arcata, California. 54 p.
Zimmerman, R.C., J.L. Reguzzoni, and R.S. Alberte. 1995. Eelgrass (Zostera marina L.)
transplants in San Francisco Bay: Role of light availability on metabolism, growth and
survival. Aquatic Botany 51:67-87.
Zimmerman, R.C., J.L. Reguzzoni, S. Wyllie-Echeverria, M. Josselyn, and R.S. Alberte. 1991.
Assessment of environmental suitability for growth of Zostera marina L. (eelgrass) in San
Francisco Bay. Aquatic Botany 39:353-366.
Zimmerman, R.C. and R.S.Alberte. 1996. Effect of light/dark transition on carbon translocation
in eelgrass Zostera marina seedlings. Marine Ecology Progress Series 136:305-309.
Zimmerman, R.C., R. Smith, and R.S. Alberte. 1989. Thermal acclimation and whole-plant
carbon balance in Zostera marina L. (eelgrass). Journal of Experimental Marine Biology
and Ecology 130:93-109.
266
-------
APPENDIX A: MAPS OF THE DISTRIBUTION OF ZOSTERA MARINA
BASED ON AERIAL SURVEYS
The maps in this appendix show the areal extent of the intertidal and shallow subtidal Zostera
marina in seven target estuaries. Distributions were mapped using aerial photography combined
with ground survey results (see Chapter 6). The superimposed boxes show the extent of
orthophotography used in mapping.
267
-------
Figure A-l. Distribution of intertidal and shallow subtidal Z. marina in the Alsea Estuary.
268
-------
Aerial Frame
J> Zosiem marina
dD Terrestrial
O Subtidal
-------
Riverine
Zostera marina
d^ Terrestrial
O Subtidal
Intertidal
Riverine
Figure A-3. Distribution of intertidal and shallow subtidal Z. marina in the Nestucca Estuary.
270
-------
Zoster a marina
Terrestrial
O Subtidal
-------
Riverine
I I Aerial Frame
Zostera marina
Terrestrial
-------
Figure A-6.
Estuary.
Distribution of intertidal and shallow subtidal Z. marina in the Umpqua River
273
-------
Figure A-7. Distribution of intertidal and shallow subtidal Z. marina in the Yaquina Estuary.
274
-------
APPENDIX B: DETAILED DESCRIPTION OF QUALITY ASSURANCE
PROCEDURES
B.I Introduction
The quality assurance/quality control (QA/QC) program for this study is defined by the "Western
Ecology Division Data Quality Management Plan (QMP) (US EPA, 2006). Measurement
Quality Objectives (MQO's) establish the data user's requirements for precision and accuracy.
The Measurement Quality Objectives for each parameter in this study are presented in Table B-
1. Quality control measures were incorporated to assure data reliability and comparability and
are described in the QMP plan. All contributing research was performed in compliance with an
approved Quality Assurance Project Plan (QAPP). In addition, Standard Operating Procedures
(SOP's) were followed to standardize routine data collection, processing and analysis for specific
parameters. All procedural documents and QA/QC plans are approved by the WED Quality
Assurance Manager.
Standard QMP protocols include routine instrument calibrations, measures of analytical accuracy
and precision (e.g., analysis of standard reference materials, spiked samples, and field and
laboratory replicates), overall data, range checks on the various types of data, cross-checks
between original data sheets (field or lab) and the various computer-entered datasets, and
participation in intercalibration exercises. Additionally, QA/QC included ensuring field and
laboratory personnel were properly trained and experienced. Specific QA procedures are
detailed in the following sections relative to each data parameter.
Accuracy and precision are indicators of MQO's and were established from considerations of
instrument manufacturer's specifications, scientific experience, and/or historical data. Measures
of systematic error (accuracy) and of random error (precision) are used to evaluate the quality of
the results. Accuracy is a measure of how close measured values are to true values. In this
appendix, accuracy is calculated using the following equation:
Accuracy (%) = (1.0 - ((£( |Vt - Vm )/ n )/ Vt) * 100
where Vt is the true or standard value, Vm is the measured values, and n is the number of
measured values. Precision is an indication of the similarity of repeated analyses or sampling.
Precision is calculated with the following equation:
Precision (%) = (1.0-(SD / J)) * 100
where SD is the standard deviation, and X is the mean.
275
-------
to
Table B-l. Measurement quality objectives for data collected by Western Ecology Division.
Parameter
units
Expected range
Accuracy
Precision
SOP(s) or other
Seabird SBE-19 instrumentation
Conductivity
Salinity
Temperature
Sea Point Turbidity
Depth
mS cm"1
PSU
°C
NTU
meters
0-100
0-35 PSU
0 - 25 °C
0-125 NTU
0-15 m
±0.5% of reading
± 0.5% of reading or 0.1 PSU
±0.15°C
± 2% of reading or 2 NTU
± 0.018m
0.01 mS cm"1
0.01 PSU
0.01°C
0.1 NTU
0.001 m
SOP FSP.03, Manual
SOP FSP.03, Manual
SOP FSP.03, Manual
SOP FSP.03, Manual
SOP FSP.03, Manual
YSI 6600 Multiparameter Datasonde
Conductivity
Salinity
Temperature
Turbidity
Depth
Dissolved Oxygen
mS cm"1
PSU
°C
NTU
meters
mg I"1 or %
Saturation
0-lOOmScm"1
0-35 PSU
0 to 25 °C
0 to 1,000 NTU
0-15 m
0-20 mg I"1 or 0-200%
saturation
± 0.5% of reading or ± 0.001
mS cm"1
±l%or0.1PSU
±0.15°C
±2% of reading or 0.3 NTU
± 0.018m
± 0.2 mg I"1 or ±2% of reading
0.01 mS cm"1
0.1 PSU
0.01°C
0.1 NTU
0.001 m
0.01 mg I"1
SOP IOP.09, YSI manual
SOPIOP.09, YSI manual
SOP IOP.09, YSI manual
SOP IOP.09, YSI manual
SOP IOP.09, YSI manual
SOP IOP.09, YSI manual
Star ODDI Mini-CTD
Salinity
Temperature
Depth
PSU
°C
meters
0-40 PSU
-l°Cto+40°C
100m
± 0.75 PSU
±-0.1 °C
±0.4% of range
0.02 PSU
0.032°C
0.03% of full scale
Manufacturer Specifications
Manufacturer Specifications
Manufacturer Specifications
Discrete Water Samples
Dissolved NO3"+NO2"
Dissolved NH4+
Dissolved PO43"
Dissolved Si(OH)4
uM
uM
uM
uM
0-100
0-5
0-3
1-100
±5%
±5%
±5%
±10%
5%
5%
5%
10%
MSIAL UCSB, 2005
MSIAL UCSB, 2005
MSIAL UCSB, 2005
MSIAL UCSB, 2005
-------
to
Table B-l. Measurement quality objectives for data collected by Western Ecology Division.
Parameter
Total Suspended Solids
Water Column chlorophyll a
units
mgr1
Ugl'1
Expected range
1-150
0-200
Accuracy
15%
±15%
Precision
15%
±15%
SOP(s) or other
WRS 14B.2
MES SOP06.revO
Zostera marina and Macroalgae Data
Seagrass shoot density
Z. marina shoot length
Z. marina shoot width
Z. marina shoot dry wt.
Z. marina lower limit depth
Epiphyte biomass dry wt.
% Plant cover (Visual)
Point intercept (Frequency of
Occurrence) SAV
Algal volume (wet)
Macroalgae nitrogen isotope
ratio (515N)
shoots m"2
mm
mm
mg
cm
mg
% area
# of intercepts
ml
%o
1-1500 m'2
5-1200 mm
1-7 mm
0-1 500 mg
0-lOmMLLW
0-1 500 mg
0-100%
0-25
0-20,000
-2 to +25 %o
± 2 shoots 0.25 m"2
±2 mm
± 1 mm
±1%
<60cm
±1%
85%
10%
100ml
0.5%o
1 shoot per 0.25 m2 quad
± 1 mm
± 1 mm
±0.1 mg
N/A
±0.1 mg
±15%
11%
50ml
0.5%o
QAPP02.01
QAPP 04.01
QAPP 04.01
QAPP 04.01
Dynamac 2005 Work Plan
QAPP 04.01
QAPP 1.04
Dynamac 2005 Work Plan
QAPP 01.06, Robbins and
Boese 2002
QAPP02.01;
QAPP 01.02
Burrowing Shrimp
Shrimp Hole Density
holes m"2
0-1 500m'2
± 10%
30%
QAPP 2000.01
Physical Data
Estuary bathymetry
Sediment waves crests
ft
m
-40 to +10 MLLW
0.1 -5m
±0.5
±0.1
N/A
±2%
Survey contract
Dynamac 2005 Work Plan
Sediment Properties
Sediment total organic carbon
and nitrogen content
Sediment grain size
distribution
Percent of dry wt
Percent of dry wt
0-12%
0-100% (0.4 |im - 948nm)
±10% of standard
±5% of dry wt
± 10%
10%
QAPP 2000.01
QAPP 2000.01
-------
Table B-2. Quality calibration and control checks for instruments and parameters.
Instrument
Seabird CTD
Conductivity
Seabird CTD
Temperature
Seabird CTD Depth
(pressure)
Seapoint turbidity
YSI Multiparameter
Sondes
(6600 models)
YSI Hand-held
meters
CMT-DGPS
LiCorLI-193SA
sensor
Balance
(5 -place)
Turner 10-AU
Fluoro meter
Calibration procedure
SeaBird factory calibration
SeaBird factory calibration
SeaBird factory calibration
Seapoint factory calibration
Factory calibrated for depth;
calibrations performed to
manufacturer's specifications for
remaining parameters
Force to temperature corrected
standard
Serf-calibrates with satellites with
every use
Li-Cor factory calibration
performed every 2 years
Factory representative adjusts on
annual basis
2-point calibration using solid
secondary standard
Quality Control Check
Solution of known conductivity
Place in water bath with NIST
traceable thermometer
In air reading compared to Fortin type
Hg barometer (Nat. Wea. Serv. type)
1. Laboratory 3 -levels of spherical
particle solutions
2. Field total suspended solids vs.
reading
All parameters checked against factory
standards upon retrieval from
deployment
All parameters checked against factory
standards
Visit type 1 USGS reference site with
multiple readings
Solar noon clear sky exposure
Check with standard weights-Class 'S'
Check solid secondary standard
reading (low and high settings) for
instrument drift.
Frequency
If drift is
suspected
If drift is
suspected
If drift is
suspected
1. annually
2. quarterly to
monthly
Immediately
prior to
deployment
and upon
retrieval
Monthly
Annually
Bi-annually
Before each
session
At start,
middle and end
of every run
Acceptance
criteria
MQO
MQO
MQO
Linear R2>0. 95;
slope ±25% of
initial
Performed to
factory
specification,
methods and
standards
Performed to
factory
specification
MQO
MQO
Within specific
class tolerance
If high setting of
secondary standard
reading differs by ±
1% from true value
Action if values are
unacceptable
Factory return
Factory return
Factory return
Clean and re-test
Clean and re-calibrate or
return to factory
Clean and re-calibrate or
return to factory
Check post processing
values re-test
Return to factory
Contact balance
maintenance personnel
Re-calibrate and re-run
samples
-------
to
^1
VO
Table B-2. Quality calibration and control checks for instruments and parameters.
Microbalance
Isotope Ratio
Mass Spectrometer
Mettler AT250
balance
Coulter LS 100 Q
Carlo Erba LS 100 Q
1108 elemental
analyzer
Check calibration with internal
weights
Background and reference gas
stability
Working standards calibrated to
NIST or equivalent certified
standards
Annual calibration by contractor
(Quality Control Services, Inc.)
Beckman-Coulter factory
calibrated
Annual calibration by factory
technician
Readings of 5 class 'S' masses that
cover expected range
Zero enrichment test
Nitrogen and carbon isotope ratios of
working standards
Check accuracy with 'S' class
weights
Analyze grains of known size.
Instrument compares background
with factory established reference
Run standards of known TOC
Before each
session
Before each
session
24 QC
working
standards per
run of 72
samples
With each use
1-2 times a
year
Every sample
With each run
±2mg
±0.07%o precision;
<0.05%« offset
from zero
±0.2 %o precision
within a run
Contractor
determination
t-test comparing
results; Visual
evaluation of curve
differences
MQO
Clean weighing pan and
catch plate, recalibrate;
have serviced
Re-focus mass spec
source, repeat zero
enrichment test; service
if necessary
Repack combustion and
reduction columns,
reanalyze samples;
service if necessary
Check cell holder
alignment, optic
cleanliness, repeat
Realign laser beam to
adjust to offsets and
clean diffraction cell;
Factory service call.
Dismantle and clean
diffraction cell or
request company
technician visit
-------
B.2 Water Quality CTD Profiles and Grab Samples
Water quality cruises were conducted in each of the target estuaries at both high and low tides.
Depending on the depth, either a Seabird SEE 19 CTD or YSI sonde 6600 was used to measure
water quality parameters. For deep stations (>1.5 m) water quality parameters were measured
continuously through the water column and processed into 0.25-m depth intervals. At shallower
stations, readings were recorded at the surface, mid, and bottom. Variables measured by the
CTD and associated instruments include depth, temperature, conductivity, turbidity,
photosynthetically active radiation (PAR), and in situ fluorescence, and calculated variables
include salinity and density. In addition to the PAR sensor attached to the Seabird unit, PAR
readings were also taken in the air with a flat sensor and at 15 cm below the surface for profile
interpretation. Dissolved oxygen was measured at discrete depths with a YSI sonde or a DO
meter attached to the frame of the Seabird CTD.
For each water quality profile, the CTD was lowered to the bottom and a water sample collected
from mid-depth during the up-cast. This water sample was analyzed for chlorophyll a, total
suspended solids (TSS), and dissolved inorganic nutrients. Near-bottom conditions were
measured at 0.5 m above the bottom. Data were collected every second and binned with
'Seasoff software into 0.25 m discrete intervals. The YSI sonde was calibrated prior to use
following the manufacturer's specifications. Light attenuation coefficients (K) were calculated
for the water column on the downcast. Prior to analysis, the data were reviewed to eliminate any
false reading caused by reflection from the aluminum boat. Care was taken so that the PAR
sensor on the CTD was not in the shadow of the boat.
The water column was sampled at each site for dissolved inorganic nutrients (Si(OH)4, NOs
+NO2 , NH4+, and PC>43"), chlorophyll a concentration, and total suspended solids (TSS). Water
column samples were collected and prepared per MES SOP09.rev 0 (2003). Water quality
samples were filtered and processed on board the boat or upon return to the laboratory or hotel
within several hours of collection. Water samples were stored in coolers on ice prior to filtration.
A performance-based approach was used for evaluating the quality of the chemical analysis.
Depending upon the compound, laboratory practices included 1) continuous laboratory
evaluation through the use of Certified Reference Materials (CRMs) and/or Laboratory Control
Materials (LCMs), 2) laboratory spiked sample matrices, 3) laboratory reagent blanks, 4)
calibration standards, and 5) laboratory and field replicates.
B.2.1 Chlorophyll a and Total Suspended Solids
Water samples for chlorophyll a and total suspended solids (TSS) analyses were collected in
duplicate and filtered on board (if possible) or upon return to the hotel or laboratory. Typically,
the samples were filtered within 1-2 hours of sample collection. For TSS analysis, 1-liter of
unfiltered seawater was collected at mid-depth as described above, filtered on board the boat and
further processed according to SOP WRS 14B.2. The complete procedure for sample processing
and analysis of chlorophyll a samples is detailed in WED SOP06.rev 0. The samples were stored
in the freezer until analyses. Samples were extracted in 90% acetone solution. The fluorometer
was calibrated with a 10-AU solid secondary standard and "blanked" with freshly prepared 90%
acetone solution prior to each sample set analyzed. The high setting of the solid secondary
standard was used in calibration and the low setting was used as a quality control check after
280
-------
calibration. During analyses, the solid secondary standard and 90% acetone blank were checked
midway through and at the end of a sample set to verify that the fluorometer performance had
not changed. If the solid secondary standard high setting differed from true values by ±1%, the
instrument was re-calibrated and the previous half-set of samples were reanalyzed.
The solid secondary standard was calibrated to chlorophyll a concentrations using fresh
chlorophyll a standards provided by the manufacturer (Turner Designs). The chlorophyll a
standards supplied by the manufacturer (Turner Designs) consisted of a high concentration
(181 jig I"1) and a low concentration (18.2 jig I"1). The solid secondary standard was calibrated
with newly purchased standards in 2002. During 2006, the solid secondary standard was
checked using additional set of chlorophyll a standards. This quality control check revealed that
the accuracy was 99.2%. To assess the accuracy of the chlorophyll a measurements, we
compared the known value of the solid chlorophyll a standard (low setting) to the actual
measured values of the solid standard. Accuracy analysis was performed for all chlorophyll a
data used in this report. The accuracy and precision for the chlorophyll a data reported in this
study are estimated to be 99.7 %.
B.2.2 Nutrient Data
Nutrient analysis for nitrate+nitrite, ammonium, phosphate and silicate was performed by the
Marine Science Institute Analytical Laboratory (MSIAL) of University of California at Santa
Barbara. Nutrient analysis was carried out with a Lachat Instruments Model QuickChem 8000
Flow Injection Analyzer. Data quality indicators include representative calibration data, reagent
blanks, replicate analyses and percent recovery analyses of spiked and control samples.
Acceptable levels for these parameters are detailed in the MSIAL QA guidelines and provide
means of monitoring data quality (MSIAL Quality Assurance Manual 2005). In addition to these
internal QA checks, samples obtained from the National Institute of Sampling and Technology
and other producers of certified reference material were analyzed periodically to audit
performance. Deionized (DI) water blanks and sea water blanks (low-nutrient natural sea water,
aged to allow nutrient values to drop to near-zero levels) were also run. An independently
prepared "control" solution containing an intermediate concentration of each of the nutrients was
also run. The chemistries used in determining the various nutrient species on this instrument
have been developed by the manufacturer to have little or no salt effect, so the analytical
response is the same for fresh, DI water samples, and standards as for salt water samples and
standards. Saltwater samples, however, exhibit a refractive index-related response in the flow-
through detector, so the sea water blank was used to adjust the measurement timing parameters
to compensate for the refractive index effect. Instrument calibration was checked at the
beginning of a sample-batch run, at the end of the run, and periodically during the run. Each
calibration sample was analyzed in duplicate, and the resulting data were used to establish
calibration curves for each nutrient species. If the mean of the two replicates of any standard
differed from the known concentration of that standard by more than ten percent or more than
one-half the concentration of the lowest standard, whichever, was greater, the calibration for that
species was considered invalid and the calibration run was repeated. (See MSIAL Quality
Assurance Manual 2005 for more details). The precision of the replicates and the accuracy of
the standards for the data used in this report are presented in Table B-3.
281
-------
Table B-3. Accuracy and precision for nutrient analyses.
Precision
Accuracy
PO43"(%)
98.0
99.1
NO3"+NO2" (%)
99.5
98.5
NH4+ (%)
96.8
98.2
B.3 Handheld YSI Meters
The handheld multiparameter Yellow Springs Instruments (YSI) meters were checked on a
monthly basis and prior to use as needed with the manufacturer's (YSI) conductivity standard.
To check the dissolved oxygen (DO) reading, the DO probe was placed in a 100% saturated
environment for several minutes. If the percent saturation value differed by more than 2% from
100% saturation, the electrodes were cleaned, a new KC1 solution was applied and the membrane
was replaced. A second DO reading was taken after the DO probe was serviced to ensure that it
was operating properly. All QA/QC checks and calibrations were recorded in a database
(Microsoft® Office Access 2003).
B.4 YSI Multiparameter Sonde
The data presented in this report were collected with YSI 6600 Multiparameter Sondes. The
sondes were calibrated prior to deployment following the manufacturer's recommendations.
Conductivity was calibrated with a one-point calibration using standards with conductivity
values closest to the expected salinity range (50 mS cm"1 for high salinity stations, 10 mS cm"1
for mesohaline and 1 mS cm"1 for low salinity stations). Turbidity was calibrated with a two-
point calibration; using reverse osmosis water (RO) followed by a 123 NTU YSI standard
solution. The DO sensor was calibrated for DO in air at sea level using the saturated air in water
method. The DO anode was cleaned and fresh KC1 solution added prior to applying a new
membrane film. The probes were set in a calibration cup with a small amount of water for
maximum water vapor saturation for 15 minutes before the calibration reading was taken. The
barometric pressure was determined from either a mercury barometer or from a YSI 650 or 556
hand-held meter. Temperature cannot be calibrated but its performance was checked. The
datasondes were set in a flow through seawater bath in the laboratory for multiple readings
immediately before deployment and upon return to the lab. The temperature and salinity of the
water bath were cross-checked using an independent YSI handheld unit (YSI 650 or 556). All
QA/QC calibration data and ancillary metadata are recorded in an Access database. The
calibration accuracy for conductivity and turbidity was defined as the accuracy of the probe in a
standard solution. Post deployment accuracy was defined as the accuracy of the probes after
they are retrieved from the field and were tested against the known standard solutions. A
database was used to calculate the pre- and post-deployment accuracy of the readings when
compared to the known standards (Table B-4).
282
-------
Table B-4. Precision and accuracy for YSI multiparameter sondes. *The post-calibration
data were not recorded for conductivity.
Calibration Accuracy
Pre-deployment
Post Deployment
Accuracy (includes the
effect of biofouling)
Temperature
(%)
98.1
92.8
Salinity
(%)
98.6
96.5
Conductivity
(%)
99.5
N/A*
Turbidity
(%)
100.0
91.9
Dissolved Oxygen
(%)
99.7
98.5
B.5 Stable Isotope Data
Five replicate macroalgae samples were collected from each sampling location in each of the
estuaries. Algal material was collected from hard substrates to eliminate contamination from any
additional nutrient sources other than the water column. Samplers wore sterile lab gloves while
collecting to prevent contamination. Each algae sample was washed thoroughly in Milli-Q
water, frozen and lyophilized. The dried material was ground into a fine powder for isotope
analysis. Grinding mortar and pestles were thoroughly rinsed with acetone and allowed to
completely dry between samples (QAPP 02.01, 2002).
The EPA Integrated Stable Isotope Research Facility (ISIRF) analyzed the macroalgae samples
for 515N according to SOP CL-6 (1999). Samples were combusted using a Carla Erba elemental
analyzer (model # 1108) equipped with a 4 meter poraplot Q gas chromatograph column directly
coupled to an isotope ratio mass spectrometer operating in a continuous flow mode (Delta S,
Finnigan MAT, San Jose, CA, USA). This continuous flow mode also provides a direct
measurement of nitrogen content. Protocol and methods for operation of the mass spectrometer
are all based on published approaches that have been verified through multiple approach
analyses and inter-lab comparisons. A concentration, calibration and reference standard were
run at the beginning, mid and end of each run. Additionally, a spike, concentration standard, and
blank were run every ten samples. Standard material included NBS tomato leaves (1573) for
concentration standard; NBS spinach (1570) for reference standard; NIST Corn Stalk for
calibration standard and spike recovery. Results for precision and accuracy for stable isotope
data used in this study are presented in Table B-5.
Table B-5. Precision and accuracy of 515N data.
Precision
Accuracy
615N (%)
99.1 (n=23)
96.6 (n=63)
283
-------
B.6 Total Organic Carbon and Nitrogen of Sediment Sample
At all intertidal stations, a small sediment core was collected for total organic carbon and
nitrogen (TOC/N) analysis. Sediment cores were stored on ice until they could be frozen upon
return to the lab. One ml of 10% Reagent Grade HC1 in Milli-Q water was added to the sediment
vial prior to analysis. The acid was mixed thoroughly into the sample using a glass rod. The
sediment samples were then left at room temperature overnight and checked in the morning to
confirm that pH was below 2. Sediment samples were then dried overnight in a 70° C oven with
the last hour at 105° C. After drying the samples were homogenized on the roller mill for 24
hours, sub-sampled and pelletized. Samples were analyzed using a Carlo Erba EA 1108
elemental analyzer. Work was performed by Dynamac Corporation on-site at the WED
laboratory in Corvallis. Blanks, unknowns and three different standards were analyzed at the
beginning of the analytical run and after every tenth sample. Every 10th experimental sample
was also analyzed in duplicate and a 3rd aliquot was spiked with the Euro Vector Soil 3 SRM.
Results are reported as weight %.
The empirically determined limit of quantitation (LOQs) for carbon and nitrogen are 0.38 and
0.08 wt. %, respectively. The empirically determined limit of detection (LODs) for carbon and
nitrogen are 0.08 and 0.02 wt. %, respectively. These LOQs and LODs values are based on
sediment samples weighing approximately 30 mg. If the C or N concentration of an
experimental sample is below the detection limit, the value is flagged but reported. If this
condition is true for duplicate samples, the DQOs for precision are ignored. The actual
calculated value for precision is reported. Acceptable precision and accuracy values for both C
and N are > 90% and < 110% for the BCR 101 spruce needle, Euro Vector Soil 3, and MESS-2
marine sediment SRMs. Acceptable precision and accuracy values for both C and N are > 90%
and < 110% for acetanilide. All samples met the DQOs and the r2 values for the acetanilide
standard curves ranged from 0.9991 to 0.9998 for N and from 0.9994 and 0.9998 for C. Results
for precision and accuracy for TOC/N data used in this study are presented in Table B-6.
Table B-6. Precision and accuracy of TOC/N sediment analysis.
Precision (duplicates, %)
Spike Recovery (Accuracy, %)
Nitrogen
97.5
98.4
Carbon
95.2
101.7
B.7 Particle Size Analysis (PSA)
Sediment aliquots were analyzed for particle size by laser diffraction on a Beckman Coulter LS
100 Q series. Aliquots of sediment samples introduced into the LS100Q sample chamber were
well-mixed in water by a high speed, reversing mixer to assure representativeness. For each
sample, two aliquots were analyzed, with a minimum of two runs of each. The final value was
an average of all replicates. The precision calculations were run on the sum of all size fractions
in each sediment size class; clay: 0.38 - 3.86 jim and silt: 3.86 - 63.41 jim. The data were
reported as percent fines which were calculated as the sum of the silt and clay size fractions.
Control standards of nominally 15 um garnet and 500 um glass particles were purchased from
the manufacturer. They are provided with a certification of their average size and the standard
deviations of the size distributions. A certified Coulter technician evaluated the accuracy of the
instrument using these standard materials on seven service calls between 2001 and 2005
284
-------
(Table B-7). During operation, the instrument performs a number of self-checks. One of these is
a measure of the background reading prior to introduction of every sample, and this background
measurement is compared to a reference background reading file established at the factory. The
Coulter LS100Q software is programmed to automatically check and align its laser beam and to
measure the sensor offsets at startup and every 60 minutes thereafter.
Table B-7. Precision and accuracy of sediment particle size analysis.
Precision (duplicates) %
Accuracy (control recovery) %
Silt
96.2
Clay
95.0
98.6
B.8 Intertidal Probabilistic Sampling
QA/QC information for the probabilistic intertidal sampling was collected by comparing repeat
measurements taken by four individuals from the same quadrat. Either seven or fourteen 2.5 m2
quadrats were laid out in intertidal habitats spanning the natural range of seagrass and
macroalgae densities (low to high densities). Two smaller 0.25 m2 quadrats were laid out within
the larger 2.5 m2 quadrats in the same arrangement as in the actual field sampling and additional
measurements were collected. Four individuals independently recorded data from the same plots
and quadrats without discussing their estimates. After all quadrats were examined the data were
analyzed for precision. The results of the independent analysis are presented in Table B-8.
Table B-8. Data comparability between four field crew members for intertidal probabilistic
sampling.
Measure
(units)
Number of Seagrass Shoots
(Number shoots per quadrat)
Point Intercept Seagrass
(Number of intercepts)
Point Intercept Macroalgae
(Number of intercepts)
# Leaves Z. marina per shoot
(Number per shoot)
Longest Z. marina Leaf Length
(cm)
Z. marina Leaf Width
(mm)
Sediment Wave Crests
(cm)
Precision
(%)
82.3
78.6
98.1
88.4
94.2
92.2
91.6
Avg. STDEV
between reported
values
5.8
7.6
0.1
0.7
1.5
0.4
0.9
Avg. Max
Difference
between
reported
readings
5.5
2.7
0.1
1.3
3.0
0.7
1.7
#of
Quads
(n)
14
14
12
7
7
7
7
285
-------
Percent cover data were collected for five class categories (i.e., 0 [absent], 1-10%, 11-40%, 41-
70%, 71-100%), therefore, precision could not be calculated as with other numeric data.
However, in comparing the individual data assessments from four field crew participating in an
exercise to evaluate consistency between individuals the following statements can be made about
the categorical data. Percent Cover Green Macroalgae: Out of the 14 plots evaluated there was
no more than one class difference between the four independent assessments in ten of the plots
with four plots having 100% agreement. Percent Cover Seagrass: Out of 14 plots evaluated there
was no more than one class discrepancy between the four evaluations at eight of the plots with
six plots having 100% agreement. Percent Cover Other Algae: Out of seven plots examined there
was 100% agreement in six plots with only one of the four data collectors reporting one size
class difference in the eighth plot.
In assessing sediment color, several individuals combined the categories to encompass color
variations in the surface sediment so calculating a precision value was not straightforward;
however, for five of the seven plots all four crew included one or more of the same colors in their
assessment. In the other two plots, three of the four individuals agreed 75% of the time.
Sediment firmness was more subjective. In three of the seven plots all four crew agreed 100% of
the time, in two plots they agreed 75% of the time and in two other plots the four crew was split
between two different firmness values.
Macroalgae biomass was determined by the volumetric cylinder method (Robbins and Boese
2002). At each site all macroalgae were removed from each quadrat and measured
volumetrically in-situ using a 4000-ml plastic cylinder. The algae were pressed with a plunger to
remove excess water before recording the algal level in ml. This quick field method was
determined to be an accurate surrogate for biomass dry weight determination with a linear
relationship yielding an r value of 0.78 to 0.88 for macroalgae species (Robbins and Boese,
2002).
B.9 Aerial Mapping of Z. marina
The remote sensing procedure used in this study to map the intertidal and shallow subtidal
distribution of Z. marina utilizes aerial photography with false-color near-infrared (color
infrared, CIR) film. This allows an aerial survey to be conducted during daylight low tides (and
clear or uniform overcast weather) when a large majority of the Z. marina distribution in PNW
estuaries is exposed or visible. The CIR film provides substantially better spectral resolution of
exposed intertidal vegetation than true color film (Young et al., 1999). To map perennial Z.
marina habitat, the surveys were conducted in late spring or early summer before the summer
bloom of benthic green macroalgae that can interfere with the classification of Z. marina.
Photoscales utilized ranged from about 1:6,000 to 1:20,000. The aerial photographs were
digitally scanned and georectified while correcting for terrain and camera distortions to produced
digital orthophotos. The spatial accuracy of the photomap for this estuary (photoscale: 1:10,000)
was assessed by comparing 14 Root Mean Square Error (RMSE) offset values for positions of
photovisible objects obtained from the photomap, referenced to published National Geodetic
Survey (NGS) positions. The mean offset was 0.72 m ± 0.27 m (95% CI; Young et al., in press).
The digital orthophotos were classified into Z. marina (defined as >10% cover) and bare
substrate habitats (defined as < 10% cover). On-the-ground resolution of 0.25 m was obtained in
this process. A hybrid technique using both unsupervised and supervised classification steps has
been developed for this habitat mapping project (Clinton et al., 2007). The technique requires
286
-------
training data from ground truth surveys, with station positioning accomplished by a differential-
corrected global positioning system (GPS). The RMSE of GPS positions obtained at an NGS
first-order monument in Yaquina Estuary was 0.62 m.
The ground survey technique provides percent cover measurements or estimates from quadrat
placement or 35 mm snapshots within each stratum being classified, preferably augmented by
GPS traces of sections of the intertidal meadow upper margins, as training data used in the image
classification process. Another part of the ground survey employed a detailed procedure based
upon the recommendations of Congalton and Green (1999) to provide accuracy assessment data
from randomly positioned stations within each stratum (Young et al., in press). The results were
analyzed with the aid of an error matrix (also know as "confusion matrix"). This is a square
array of numbers set out in columns and rows that expresses the number of sample units assigned
to a particular category in one classification relative to the number of sample units assigned to a
particular category in another classification (Congalton and Green, 1999). Here, the number of
stations (units) classified as 2. marina or bare substrate habitat based on the ground survey
measurements (listed in the table's "column total" cells) are considered to be correct (i.e.
reference data), while the values listed in the individual table cells are the corresponding number
of stations in each class generated from the remotely sensed data. The quotient (as a percentage),
obtained by dividing the total number of correctly classified sample units in a given category
(e.g. Z. marina stratum) by the total number as indicated by the reference data, is termed
"producer's accuracy". Similarly, the quotient obtained by dividing the total number of correct
units in the given category by the total number of units classified as belonging to that category
(listed in the table's "row total" cells) is termed "user's accuracy". A third accuracy value,
obtained by dividing the sum of the "table diagonal" units by the total number of sample units in
the matrix, is termed "overall accuracy" (Congalton and Green, 1999). To assess the
effectiveness of this classification method, values for the Kappa index KHAT (and their
estimated 95% confidence intervals) for these matrices also were calculated as a measure that
assesses improvement over chance (Fielding and Bell, 1997). Results were obtained in the
spring of 2004 from 51 randomly positioned stations within intertidal Z. marina meadows and 28
randomly positioned stations within bare substrate strata of Yaquina Estuary.
Based upon a comparison of results from the image classification with those from the ground
survey (taken as the reference), application of the classical error matrix analysis yielded an
overall accuracy for Yaquina Estuary of 97%, with a Kappa Index value of 0.94 ± 0.002
(Table B-9). The high index value attained indicates excellent agreement (Landis and Koch,
1977). The investigators attribute this very high accuracy level to the extensive training data
provided via GPS mapping of the intertidal Z. marina meadow margins (Young et al., in press).
For corresponding classifications of intertidal Z. marina and bare substrate strata in Tillamook,
Umpqua and Coos estuaries, obtained without meadow margin mapping for training of the image
analyst, overall accuracy values were 86%, 89% and 83%, respectively. In contrast, in Alsea
Bay where extensive Z. marina meadow margin mapping was conducted, an overall accuracy of
100% was obtained (Table B-9). Due to the subtidal habitat of seagrass in Nestucca and Salmon
River estuaries, seagrass patches were outlined using either a boat or by foot with a GPS unit. A
"best fit" line was created from three replicate traces.
287
-------
Table B-9. Image classification accuracy assessment.
Tillamook Estuary
Image Class (Predicted)
Producer's Accuracy
User's Accuracy
Overall Accuracy:
Kappa Index KHAT*:
Yaquina Estuary
Image Class (Predicted)
Producer's Accuracy
User's Accuracy
Overall Accuracy:
Kappa Index KHAT*:
Alsea Estuary
Image Class (Predicted)
Producer's Accuracy
User's Accuracy
Overall Accuracy:
Kappa Index KHAT*:
UniDqua Estuary.
Image Class (Predicted)
Producer's Accuracy
User's Accuracy
Overall Accuracy:
Kappa Index KHAT*:
Coos Estuary
Image Class (Predicted)
Producer's Accuracy
User's Accuracy
Overall Accuracy:
Kappa Index KHAT*:
Z. marina
Bare Substrate
column total
86%
0.72 ± 0.003
Z. marina
Bare Substrate
column total
97%
0.94 ± 0.002
Z. marina
Bare Substrate
column total
100%
1.0
Z. marina
Bare Substrate
column total
89%
0.76 ± 0.004
Z. marina
Bare Substrate
column total
83%
0.62 ±0.005
Z. marina
88
12
100
88%
87%
50
1
51
98%
98%
44
0
44
100%
100%
37
4
41
90%
79%
28
10
38
74%
76%
Bare
Substrate
13
68
81
84%
85%
1
27
28
96%
96%
0
60
60
100 %
100 %
10
81
91
89%
95%
9
68
77
88%
87%
row
total
101
80
181
51
28
79
44
60
104
47
85
132
37
78
115
* 95% confidence interval assuming KHAT is normally distributed
288
-------
B.10 Zoster a marina Lower Depth Limit
Lower depth margin of Zostera marina was determined by georeferencing the position where the
deepest seagrass was encountered. Transects were randomly selected in distinct Z. marina beds
(identified from aerial photography) and were approached either by boat or foot depending on
the water depth. Sampling was conducted on the lowest tides possible to increase the accuracy
of locating plants growing at the lowest depth limit. In the deeper systems, an underwater video
camera was mounted on a long PVC pole linked to a video monitor on deck. When the first
Z. marina patch was seen on the monitor the pole was quickly thrust into the sediment to stop the
momentum of the boat. A GPS reading was taken and a lead line was used to record the depth to
the closest centimeter. In shallower waters, seagrass blades were clearly apparent on the water
surface and could easily be approached by foot. Personnel walked along the transect from shore
to the deepest Z. marina patch, recorded a GPS location, measured the depth with a lead line.
B.10.1 Tidal Corrections for Lower Depth Limits
Tidal corrections were applied to account for variations in tide elevation at time of lower limit
depth observations, with corrected lower depth limits expressed as depth below mean lower low
water (MLLW). Tidal predictions were used to make these tidal corrections. Tidal predictions
are least accurate during storms and extreme low and high tides. Review of weather and tidal
conditions during time periods when the lower depth limit was measured, indicated that
conditions were relatively calm during the sampling and not collected during extreme tides.
Tidal heights relative to MLLW were calculated using WXTIDE 32 for each depth site using the
time of data collection and the nearest tide prediction location. The largest source of error in the
tidal corrections resulted from not having predicted tides available for all locations along the
longitudinal axis of the estuary and having to use the closest tide prediction station. To estimate
the error associated with the tidal correction, differences in tidal corrections between the two
stations that are located upstream and downstream of the observations were calculated
B.10.1.1 Yaquina
The difference between predicted and observed tidal heights at the South Beach tide gauge was
less than 0.15 m on all data collection days. Tidal predictions that take into account variations in
tidal amplitude and phase lags were available for four locations (South Beach, Yaquina, Winant,
and Toledo) in the Yaquina Estuary (http://co-ops.nos.noaa.gov/tides05/tab2wclb.html#132).
The maximum distance along the river at any given depth station from a tide station was 4.5 km.
The error associated with the tidal correction is estimated to be 0.1-0.2 m. The error associated
with the tidal correction increases with distance from the mouth of the estuary, being a minimum
of 0.1 m in the lower estuary and as high as 0.2 m near Toledo.
B.10.1.2 Tillamook
The maximum difference between predicted and observed tidal heights at the Garibaldi tide
gauge was 0.3 m on data collection days. Tidal predictions that take into account variations in
tidal amplitude and phase lags were available at five locations (Barview, Garibaldi, Miami Cove,
Bay City and Hoquarten Slough) in the Tillamook Estuary
(http://co-ops.nos.noaa.gov/tides05/tab2wclb.html#132). The maximum distance along the river
any given depth station from a tide station was -3.6 km. The error associated with the tidal
correction is estimated to be 0.1- 0.3 m. The error associated with the tidal correction increases
289
-------
with distance from the mouth of the estuary, being a minimum of 0.1 m in the lower estuary and
as high as 0.3 m between Bay City and Hoquarten Slough.
B.10.1.3 Nestucca
The difference between predicted and observed tidal heights at the Nestucca tide gauge on data
collection days is unknown. The tidal predictions for Nestucca used parameters for variations in
tidal amplitude and phase lags from a tidal reference station in Crescent City, CA for a single
station at the bay entrance available at: http://co-ops.nos.noaa.gov/tides04/tab2wclb.html#132.
The maximum distance along the river at any given depth station from the tide station was ~ 4.7
km. The error associated with the tidal correction increases with distance from the mouth of the
estuary. To estimate the error associated with the tidal correction, we compared the maximum
distance above to the maximum distances from tide prediction stations in Yaquina and Tillamook
Bays and estimate the error to be between O.lm and 0.3 m.
B.10.1.4 Salmon
The difference between predicted and observed tidal heights at the Salmon River on data
collection days is unknown. No tidal prediction stations are available in the Salmon estuary. The
nearest tidal prediction station is at Taft in the Siletz Estuary which is 11.1 km to 12.5 km from
the sample locations in the Salmon Estuary. The tidal predictions for Taft use parameters for
variations in amplitude and phase lags of tides from a tidal reference station (Crescent City, CA)
for a single station at the bay entrance available at: http'.//co-
ops, nos.noaa.gov/tides04/tab2wclb.html#132 . The error associated with the tidal correction is
unknown.
B.10.1.5 Alsea
The difference between predicted and observed tidal heights at Alsea Bay on data collection days
is unknown. The tidal predictions for Alsea used parameters for variations in tidal amplitude and
phase lags from a reference station at Crescent City, CA for two single stations available at:
http://co-ops.nos.noaa.gov/tides04/tab2wclb.html#132. The maximum distance along the river
at any given depth station from the tide station was -1.2 km. To estimate the error associated
with the tidal correction, we compared the predicted tide lag between the Waldport, Alsea Bay
station and the Drift Creek, Alsea River station (which are -7.6 km apart) during hours of
upstream data collection. The error associated with time lag in the predicted tides for the data
collected is less than 0.1 m.
B.10.1.6 Umpqua
The difference between predicted and observed tidal heights at Umpqua River on data collection
days is unknown. The tidal predictions for Umpqua used parameters for variations in tidal
amplitude and phase lags from the tidal reference station in Humboldt Bay, North Spit, CA for
three stations available at: http://co-ops.nos.noaa.gov/tides04/tab2wclb.html#132. The
maximum distance along the river at any given depth station from the tide station was -5.5 km.
To estimate the error associated with the tidal correction, we compared the predicted tide lag
between the Umpqua River, Entrance station, the Umpqua River, Gardiner station and the
Umpqua River, Reedsport station (which are -10.5 km and - 4.0 km apart) during hours of data
collection. The error associated with time lag in the predicted tides for the data collected is less
than 0.37 m.
290
-------
B.10.1.7 Coos
The maximum difference between predicted and observed tidal heights at the Charleston, Coos
Bay tide station was 0.1 m. Historic tide data for the Charleston, OR station is available at:
(http://tidesandcurrents.noaa.gov/geo.shtml?location=9432780). The tidal predictions for Coos
use parameters for variations in amplitude and phase lags of tides from a tidal reference station
within the estuary (Charleston, OR) and for two more stations, Empire and Coos Bay available
at: http://co-ops.nos.noaa.gov/tides04/tab2wclb.html#132. The maximum distance along the
river at any given depth station from the tide station was -10 km. The error associated with time
lag in the predicted tides for the data collected is unknown because -70% of the depth stations
are upstream of the tide stations.
B.ll Field Locations and Distance Upriver Calculations
The geoposition of each station was collected in the field with a Global Positioning System
(GPS) and differentially corrected in post processing with data from the nearest National
Geodetic Survey (NGS) Continuously Operating Reference Station (CORS). The GPS data
collection device (CMT March II) published post-processed differentially corrected two
dimension root mean square error (spatial accuracy) ranges from 1 m to 5 m.
The distance from the mouth of each estuary to each station was calculated using GIS mapping
software ArcMap. For each estuary, National Hydrologic Dataset (NHD) center line features
extending the length of the rivers from the ocean shore (disregarding jetties) were converted into
route files. The ArcToolBox linear referencing tool, "LocateFeaturesAlongRoutes", was used to
calculate the distances in meters for each station from the ocean shore to the nearest point along
the centerline route feature. The distance from each station to the nearest point along the
centerline route was calculated simultaneously by the geoprocessing tool.
B.12 Estuarine and Watershed Landscape Analysis
To capture the diversity of estuary types on the Pacific Coast, a national standard, the National
Wetland Inventory (NWI) developed by the U. S. Fish & Wildlife Service (U.S. Fish Wild. Sen,
2002; http://www.fws.gov/nwi) was used as the criterion for defining estuaries. The NWI
classifies aquatic habitat types using a hierarchical set of attributes based on Cowardin et al.
(1979), which includes marine, estuarine, riverine, palustrine, and lacustrine areas with further
subdivisions for tide height, substrate type and the presence of broad classes of wetland plants
(e.g., emergent vs. aquatic bed). In an effort to inventory all estuaries, an estuary was defined as
any coastal water body that contained an NWI estuarine polygon.
B.12.1 Estuary Inventory
The estuary watershed geospatial layer generated for NOAA's Coastal Assessment Framework
(ftp://sposerver.nos.noaa.gov/datasets/CADS/GIS_Files/ShapeFiles/caf/) was used as the initial
source to identify the larger Pacific Coast estuaries. However, this layer was not sufficiently
detailed to represent the moderate to small estuaries. To develop a complete inventory of coastal
estuaries, estuary water bodies were identified from previously mapped estuarine polygons from
the NWI. Additionally, visual inspection of coastal sites either in the field or from aerial
photographs was used to identify any additional coastal water body that appeared to have
"estuarine" characteristics but that did not have a NWI estuarine polygon. Geospatial data from
the NWI were obtained from digital databases and from on-site digitization of paper maps not
291
-------
otherwise available in digital formats. Sites classified by NWI as estuarine, and in some cases,
marine, palustrine or lacustrine, were inserted into the estuarine/watershed GIS layer, retaining
the NWI classification system attributes. In some cases, adjustments were made to the attribute
coding of water bodies to reflect judgments that these were part of an estuarine system.
B.12.2 Watershed Delineation
The Sixth Field hydrologic unit code (HUC) subwatershed geospatial layer created by the Forest
Service from 1:24,000 scale USGS maps, digital elevation models and other data sources
(http://www.reo.gov/gis/projects/watersheds/REOHUCvl_3.htm) was used as a primary
reference. In Oregon, the Forest Service and Oregon State University have produced a
watershed layer refined to the 7th field HUC boundary lines for most of coastal Oregon
(http://www.fsl.orst.edu/clams/cfsl0233.htmi) north of the Rogue River.
In all states, final refinements to the drainage boundaries between coastal and estuarine basins
were often based on review of the hydrologic drainage patterns derived from digital elevation
data (10-meter resolution in Oregon) and from USGS 1:24,000 scale quadrangle maps.
The watersheds delineated in this project capture the entire drainage area for each of the
estuaries. By delineating the entire watershed, these watershed areas are equivalent to the sum of
NOAA's Estuarine Drainage Area (EDA, portion of watershed that empties directly into the
estuary and is affected by tides) and Fluvial Drainage Area (FDA, portion of an estuary's
watershed upstream of the EDA boundary; see http://coastalgeospatial.noaa.gov).
B.12.3 Land Cover Sources
The estuary watersheds and CD As were used as clipping boundaries for several land use/land
cover datasets that are available for the Pacific coastal region at this time. The National Land
Cover Data (NLCD, http://www.mrlc.gov) represents land cover circa 1992 and its extent is
nationwide. This dataset was clipped to the watershed boundary for each estuary or CDA and is
complete for all watersheds except for the interior Columbia River basin in Canada, and the
Tijuana River estuary in Mexico. The 1992 NLCD data contains 21 classes of land cover
(http://erg.usgs.gov/isb/pubs/factsheets/fsl0800.pdf) with some modification in the 2001 NLCD
data (http://www.epa.gov/mrlc/defmitions.html#1992). The area of each land cover class in
square kilometers and as a percentage of the watershed was calculated and entered into an
Access database.
Two additional land cover datasets have been created by the NOAA's Coastal Services Center
(C-CAP, http://www.csc.noaa.gov/crs/lca/ccap.html) program. The more recent data were
derived from late 2000 and 2001 Landsat TM (thematic mapper) imagery. NOAA also produced
a layer from imagery collected circa 1995-1996 and the earlier dataset was used to generate a
land cover change layer. As the NOAA data were generated primarily to investigate land cover
effects on coastal resources, the Pacific coast datasets did not completely cover the entire
drainage areas of the largest three estuary watersheds. The interior portion of the San Francisco
Bay estuary and the eastern portion of the Klamath River drainage were not included in the
NOAA datasets. The lower Columbia River basin downstream of Bonneville Dam is included,
but much of the interior Columbia River basin is not within the extent of the NOAA datasets.
The 2001 NOAA data are based on 22 land cover classes
292
-------
(http://www.csc.noaa.gov/crs/lca/oldscheme.html), which are not exactly the same as those used
in the NLCD.
In January 2007, the Multi-Resolution Land Characteristics Consortium (MRLC,
http://www.mrlc.gov) released a new national land cover data, NLCD 2001. The areas classified
by the NOAA C-Cap program were incorporated into the 2001 release. Procedures used in the
development of the 2001 land cover data layer are presented in Homer et al. (2004). The land
cover in NLCD 2001 is based on 30-meter resolution data derived from Landsat imagery and
uses 21 classes that are a modified version of the land classes used in the 1992 NCLD analysis
(see http://www.epa.gov/mrlc/defmitions.html#1992). The 2001 data were analyzed by zone,
with the analysis of the Pacific Coast watersheds utilizing data from eight zones. The bounding
coordinates of these eight zones are given in Table B-10.
Table B-10. Bounding coordinates for the eight 2001 MRLC zones containing Pacific Coast
watersheds.
MRLC
Zone
1
2
O
4
5
6
7
12
Longitude, degrees
West
-125.274529
-126.118507
-125.110557
-122.777682
-123.051273
-122.184297
-124.458076
-121.009858
East
-118.634037
-121.051106
-121.036959
-115.735672
-118.480769
-117.601669
-119.43724
-113.825176
Latitude, degrees
North
49.66092
46.59432
42.36827
39.03779
41.04824
41.35061
45.86137
42.42003
South
45.14402
41.63202
37.2476
31.62669
34.4985
34.52835
40.46927
36.26734
B.12.4 Land Cover Patterns and Watershed Characteristics
Land cover data from the 1992 and 2001 MRLC NLCD data and from the NOAA 1995 and 2001
data were used to calculate the area and percentage of the watershed for each of the 21 (NLCD)
or 22 (NOAA) land cover classes. Accuracy of the 1992 NLCD data by EPA region is presented
at http://landcover.usgs.gov/accuracy/index.php. Based on this analysis, users were cautioned
about applying the data to highly localized studies, such as over a small watershed. Accuracy of
the 1992 NLCD can be found at http://www.epa.gov/mrlc/accuracy.html while discussion of the
2001 NLCD data accuracy can be found at http://www.epa.gov/mrlc/accuracy-2001.html. The
overall accuracy by MRLC zone ranged from about 86% to 98% (Table B-l 1)
These datasets were used to generate estimates of the area of impervious surfaces in each
watershed using default coefficients from the Analytical Tools Interface for Landscape
Assessments (ATtlLA) software (U.S. EPA, 2004c). The MRLC 2001 impervious surface layer,
which represents an estimate of developed impervious surface per pixel by percent
imperviousness, was clipped and summarized for each watershed. Overall accuracies for the
impervious surfaces from the 2001 MRLC data range from 83 to 91 percent (Homer et al., 2004;
Yang et al., 2003), and represent a higher resolution estimate of impervious surfaces than
available from ATtlLA. Estimates of nitrogen and phosphorus loadings from land cover were
calculated from the watershed land cover data using coefficients from the ATtlLA program.
293
-------
Estimates of slope were calculated for each watershed from slope surfaces generated from IO-
meter (Oregon) digital elevation models (DEMS). Mean slope by percent and by degrees for
land surfaces above the mean high water level were calculated and all slope values were
exported to an Access database. The 30 meter DEMS were obtained from the National Elevation
Dataset (NED, http://ned.usgs.gov), a seamless mosaic of the best elevation data. The 10-meter
elevation data for Oregon was obtained by the USDA Forest Service for the Coastal Landscape
Analysis and Modeling Study project (CLAMS, http://www.fsl.orst.edu/clams) from USGS
drainage enforced digital elevation models.
Table B-ll. Overall accuracy of NLCD 2001 land cover analysis. Zones refer to the
MRLC zones containing Pacific Coast estuarine watersheds.
MRLC Zone
1
2
O
4
5
6
7
12
Overall Accuracy (%)
86.1
86.1
88.0
88.0
85.2
88.3
98.9
84.2
B.12.5 Population Density
Human population estimates from the 1990 and 2000 censuses (http://www.census.gov/) were
generated for each drainage unit. Area weighted estimates of total population by census block
were summed for each drainage and population density (individuals km" ) was calculated from
the total drainage population estimate.
B.12.6 Climate
Estimates of air temperature and precipitation were made for each watershed from several
sources of long-term climate data. Annual mean temperature and mean precipitation were
obtained from the 1 kilometer resolution climate summaries from 1980 to 1997 from the Daily
Surface Weather and Climatological Summaries (DAYMET; http://www.daymet.org). Annual
mean, annual mean minimum, annual mean maximum temperature and mean annual
precipitation were also obtained from the 800 meter resolution PRISM climate model for the
period 1971-2000 (http://www.climatesource.com). Comparisons were made between the two
climate datasets and diverging values were examined.
B.12.7 National Wetland Inventory
As part of the development of the sampling frame for the Western Coastal EMAP 2002 survey,
the available National Wetlands Inventory digital layers for coastal estuaries in Oregon,
Washington and northern California were collected by EPA's onsite GIS staff in 2001. Some
significant areas of Oregon estuaries were not included in the existing NWI digital dataset and
paper maps were obtained for these areas and scanned. The scanned maps were registered and
were used for "heads up" digitizing of the missing features.
294
-------
In July 2005, all of the available NWI datasets for the Pacific coast were retrieved from the
National Wetlands Inventory website. Digital maps were downloaded from the NWI website,
mostly individual ESRI shapefiles of 1:24,000 scale quadrangle maps. The zipped versions of
the shapefiles were created by NWI in the spring of 2003, according to the file creation dates. In
some cases where shapefiles for specific quadrangles were missing, older Arclnfo coverages
were available and retrieved from the NWI download website. All of these datasets are what the
current NWI mapping team calls the "static" data. Where the circa "2005" data differed from the
final version of the EPA NWI aggregation, the most recent data from the NWI website
superseded the previous data for inclusion in the estuary geospatial layer. The NWI targeted
mapping unit for the scale of photography most often used in the Pacific estuary classification
varied from 1 to 3 acres.
Currently the NWI Mapping project is transit!oning to the ESRI geodatabase format with
wetlands data being available in larger seamless blocks (1:100,000 scale) or as selected views
from an online mapping tool. The latest data for the Pacific Northwest (as of mid August 2006)
shows some alterations to the previous wetlands in portions of the Oregon coast. These changes
were captured in a revision of the Oregon estuary watershed geospatial layer in December 2007.
Descriptions for parsing the NWI attribute field codes to a habitat type were obtained from the
NWI documentation and from queries to the online Map Codes Search form
(http://www.fws.gov/wetlands/Data/webatx/atx.html). Some of the polygons in the NWI data
selected for inclusion in the estuary classification spatial layer have codes that are considered
obsolete under the current system. These will be kept in the database until revised geospatial
data are available from the NWI. The coding used in the wetlands classification system used is
based upon the Cowardin et al. (1979).
B.12.8 Estuarine Characteristics
Strahler stream order, which represents a hierarchical view of stream complexity based on the
number of tributary junctions in a stream system from headwaters to mouth, was recorded for the
largest order stream entering an estuary. A summarization of all the stream orders entering each
estuary has been completed for the Pacific Northwest and will be completed for all the Pacific
Coast estuaries. Order was recorded for the streams included in the National Hydrological
Dataset (NHD; http://nhd.usgs.gov/), which corresponds to streams on the 1:100,OOOK map
series and often is complete with streams mapped at the 1:24,000 level. Mouth width for each
estuary was estimated using aerial photography from Terraserver (http://www.terraserver.com)
and other imagery sources.
B.13 Other Data Sources
Historical salinity and nutrient data were obtained from the Oregon Department of
Environmental Quality web-based database. Data from this site were previously reviewed
internally by DEQ and assigned a data quality grade. Only data with grades of A and A+ were
used as part of this report.
295
-------
B.14 Hydrodynamic and Nutrient Source Model
All field and laboratory data used in the development of hydrodynamic and nutrient source
model were collected in accordance with WED's SOPs and QAPPs. A two-dimensional,
laterally averaged hydrodynamic and water quality model (Cole and Wells, 2000) was used to
simulate the transport of riverine, oceanic and wastewater treatment facility (WWTF) effluent
dissolved inorganic nitrogen (DIN) sources. This model is well suited for long-narrow estuaries,
such as Yaquina Bay, where there are minimal lateral variations in water column properties.
U.S. EPA (2001) suggested that this model may be useful in the estuarine nutrient criteria
development and has been used in developing estuarine Total Maximum Daily Loads (TMDLS).
In the model simulations presented in this study, Yaquina Estuary was represented by 325
longitudinal segments spaced approximately 100-m apart with each longitudinal segment having
1-m vertical layers. The model domain extended about 37 km from the tidal fresh portion of the
estuary at Elk City, Oregon to the mouth of the estuary. Model simulations were performed for
the interval January 1 to October 1 of 2003 and 2004 and included tidal and wind forcing as well
as freshwater inflow. Parameters simulated included water surface elevation, salinity, water
temperature, and DIN.
Model calibration is the process of determining model parameters that are appropriate for the
specific study location and time interval being simulated. The model used in this study was
calibrated through adjustment of friction coefficient, eddy viscosity, and eddy diffusivity. To
assess the model performance at simulating the hydrodynamics, we compared simulated and
observed water level variations at two locations in the estuary and salinity and water temperature
at four locations utilizing data from the YSI datasondes. Since the datasondes used at these
stations were not leveled in we could only compare relative water level fluctuations, not absolute
water level (referenced to MLLW). In addition, temperature and salinity from the CTD cruises
were compared to simulated values. The model was assessed by calculating the root mean
square error between observed and predicted variables.
Each nitrogen source, riverine (Nriver\ oceanic (Nocean), and WWTF effluent (Nwwf\ was modeled
as a separate component. The nitrogen sources were modeled as
dN
- = transport - juN
where TV is the DIN source and u^ is a loss/uptake rate. The same value of u^ was used for all three
nitrogen sources and the value of [j, was determined by fitting total modeled DIN
(NOCean+Nriver+Nwwtf) to observations of DIN within the estuary. The best fit to observations was
found with [^ = 0.1 d"1. Simulations were also performed with no uptake (u^ = 0) which is
equivalent to conservative transport of the sources. The results from the transport model were
used to mix the three nitrogen sources using the following equation
where fn,fw, and^b are the fractions of riverine, wastewater treatment facility, and oceanic DIN,
respectively, and SR, 5w, and SR are the isotopic end members for riverine, wastewater treatment
facility effluent, and oceanic sources, respectively. Estimates of the oceanic and riverine end
members were obtained by examination of the observed isotope ratios at the stations located near
296
-------
the mouth of the estuary and in the riverine portion of the estuary and comparison to the
literature. The initial estimate for the WWTF end member (Sw = 15-22%o) was determined from
the literature. To arrive at the final end member isotope ratios, model simulations were
performed varying each end member over the range estimated from the data and literature. The
final isotope ratio of end members for the three sources (5R=2%o, 5w=20%o, and 5o=8.4%o) was
determined from the best fit (minimum root mean square error, RMSE) between predicted and
observed isotope ratio at the five isotope sampling stations during 2003 and 2004. The final
oceanic end member selected is consistent with marine end members for the west coast of the
United States (Fry et al., 2001). While the riverine end member is consistent with the isotope
ratio expected for nitrogen associated with red alder (leaf tissue ranges between -3 and -0.5%o;
Hobbie et al., 2000; Tjepkema et al., 2000; Cloern et al., 2002).
B.15 Quality Assurance Project and Quality Management Plans (QAPPs) Used in This
Study
QAPP 2000.01. Changes in the Abundance and Distribution of Estuarine Keystone Species in
Response to Multiple Abiotic Stressors. T.H. DeWitt, EPA, 2000.
QAPP 01.02. Modeling of Landscape Change Effects on Estuarine Trophodynamics: an
Optimization Approach Using Inverse and Forward Modeling. P. Eldridge, EPA, 2001.
QAPP 01.04. Assessment of the Spatial and Temporal Distribution of Submersed Aquatic
Vegetation and Benthic Amphipods within the Intertidal Zone of Yaquina Bay Estuary,
Oregon via Color Infrared Aerial Photography. D.R. Young, EPA, 2001.
QAPP 01.06. Upper Margin Expansion: Influences on Seagrass, Zostera marina L. B.L. Boese,
EPA, 2001.
QAPP 02.01. Autecological studies of marine macrophytes including the sea grasses Zostera
marina and Z.japonica in Yaquina Bay, Oregon. J. Kaldy, EPA, 2002.
QAPP 04.01. Seagrass Research - Epigrowth Light Attenuation Task: Estimation of spatial and
temporal variation in light attenuation due to epigrowth on Zostera marina in Yaquina Bay.
W. Nelson, EPA, 2004.
QAPP-06.01. Development of the Pacific Coast Ecosystem Information System (PCEIS). H.
Lee II and D. Reusser, 2006.
Marine Science Institute Analytical Laboratory University of California, Santa Barbara. Quality
Assurance Manual-Draft. 2005.
U.S. Environmental Protection Agency. 2006. Quality Management Plan. NHEERL. Western
Ecology Division.
B.16 Standard Operating Procedures (SOPs) Used in This Study
CL -6. V.2. Standard Operating Procedures for Stable Isotope Ratio Mass Spectrometer Analysis
of Organic Material. EPA. 1999.
GPEP SOP 3.01 - TERA SOP. Carbon/Nitrogen Elemental Analysis. Rick King, Dynamac
Corporation. 1989.
MES EPOl.rev 0. Draft. Standard Operating Procedure for Collecting and Processing Zostera
marina and Associated Epiphytes for Light Attenuation Measurements. Dynamac
Corporation. 2004.
MES SOP09.rev 0. Standard Operating Procedure for Preparing Water Samples for Nutrient
Analysis. Dynamac Corporation. 2003.
297
-------
MES SOP02.rev 0. Standard Operating Procedures for Weighing Food Web Samples and
Submitting them to ISIRF for Stable Isotope Analysis. K. Rodecap, Dynamac Corporation.
2002.
WED SOP06.rev 0. Standard Operating Procedure for Preparation and Analysis of Estuarine
Water Samples for Determination of Chlorophyll-a Content. Dynamac Corporation. 2004.
SOPCMT.02, Sampling and Preparation for TOC Analysis, Sercy, Kathleen, et al, US EPA,
Western Ecology Division. 1991.
SOPIOP.09. Operating Procedure For YSI Series 6 Multiparameter Water Quality Meters, Model
#s 6000UPG and 6600, 6600EDS. D.T. Specht. EPA. 2004.
SOPFSP.01. Use of The Seabird Seacat (SBE-19) CTD Package. R. J. Ozretich. EPA. 1999.
WRS 14B.2. Standard Operating Procedure for the Determination of Total Suspended Solids
(Non- Filterable Residue). Dynamac Corporation. 2005.
SOPPMP.04. Measurement of sediment grain size distribution using a laser diffraction particle
size analyzer. James H. Power. EPA. 2007.
SOP-NHEERL/WED/PCEB/PC/2006-Ol-rl. Standard Operating Procedures for producing
digital aerial photomaps of estuarine intertidal ecosystems using color infrared film,
classifying eel grass and non-vegetated habitats, and assessing the accuracy of the
classifications. Clinton, PJ. and D.R. Young. EPA. 2006.
298
------- |