Predicting patterns of
vulnerability to climate
change in near coastal
species using an algorithm-
based risk assessment
framework
&EPA
United States
Environmental Protection
Agency
EPA/600/R-17/052 November 2017 www.epa.gov/ord
Office of
Research and Development
National Health and
Environmental Effects
Research Laboratory
Western Ecology Division
-------
THIS PAGE INTENTIONALLY BLANK.
-------
EPA/600/R-17/052
November, 2017
Predicting Patterns of Vulnerability to
Climate Change in Near Coastal Species
Using an Algorithm-Based Risk
Assessment Framework
By
Henry Lee II, U.S. EPA, Western Ecology Division
Christina Folger, U.S. EPA, Western Ecology Division
Deborah A. Reusser, USGS (Emeritus)
Patrick Clinton, U.S. EPA, Western Ecology Division
Rene Graham, CSS
Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Western Ecology Division
-------
Disclaimer
This document has been reviewed by the U.S. Environmental Protection Agency, Office of
Research and Development, and approved for publication. Any mention of trade names,
products, or services does not imply an endorsement by the U.S. Government or the U.S.
Environmental Protection Agency. The EPA does not endorse any commercial products,
services, or enterprises.
Recommended Citation
Lee II, H., Folger, C.L. Reusser, D.A., Clinton, P. and Graham, R. 2017. Predicting Patterns of
Vulnerability to Climate Change in Near Coastal Species Using an Algorithm-Based Risk
Framework. EPA/600/R-17/052. 299 pages.
li
-------
Contents
Disclaimer ii
Contents iii
Change Log xi
Acronyms and Abbreviations xiii
Acknowledgements xvi
Executive Summary xvii
Section 1. Introduction and Overview 1
1.1 Problem Statement 1
1.2 Goals and Objectives 6
1.3 Scope of Document 7
1.4 Geographic and Taxonomic Scope 8
1.4.1 Geographic Scope 8
1.4.2 Taxonomic Scope and Standardization 9
1.5 Algorithm-Based Approach to Climate Risk Assessment 11
Section 2. Risk Categories and Overall Vulnerability 16
2.1 Individual Risk Values 16
2.2 Resilience Traits 19
2.3 Overall Risk - "One Way to Live, A Thousand Ways to Die" 20
Section 3. Relative Abundance Estimates 22
3.1 Background 22
3.2 Abundant or Rare? 22
3.3 Hierarchical Relative Abundance Classification Schema 23
3.4 Importance of Habitat Area 25
3.5 Data Sources - Quantitative Data 28
3.5.1 Dominance Normalized Relative Abundance (DNRA) 28
3.5.2 Quantitative Cut Points for Abundance Classes 29
3.6 Data Sources - Online Biodiversity Databases 34
3.7 Data Sources - Text-Based Information 34
3.7.1 Parsing Natural History Texts 34
3.7.2 Negative Evidence: The Dog That Didn't Bark 35
3.8 Hybrid Approach to Estimating Ecoregion Abundances 36
3.8.1 Synthesizing Multiple Data Types 36
3.8.2 Checking Abundance Classifications at an Ecoregion Scale 38
Section 4. Baseline/Status Risks 39
4.1 Introduction 39
4.1.1 Observed versus Preferred Habitats and Environmental Ranges 47
4.2 Baseline/Status Traits - Biogeographic Distributions 50
in
-------
4.2.1 Introduction 50
4.2.2 Relationship of Range Size to Vulnerability 50
4.2.3 Endemics - Vulnerability Trait 51
4.2.4 Restricted Distribution - Vulnerability Trait 52
4.2.5 Wide Distributions - Resilience Trait 53
4.2.6 Arctic Endemics - Vulnerability Trait 58
4.2.7 Small Island Distributions - Resilience Trait 59
4.2.8 Nonindigenous Species - Resilience Trait 61
4.3 Baseline/Status Traits - Relative Abundance Patterns 62
4.3.1 Background on Relative Abundance Metrics 62
4.3.2 Hyper-Rare Species - Vulnerability Trait 63
4.3.3 Abundant Someplace/Rare Everywhere - Vulnerability and Resilience Traits 63
4.3.4 Population Trends - Vulnerability and Resilience Traits 64
4.3.5 Southern Ecoregion Rare and Ecoregion to North Abundant - Vulnerability Trait.... 67
4.3.6 Northern Transients - Resilience Trait 68
4.4 Baseline/Status Traits - Life History 69
4.4.1 Introduction 69
4.4.2 Symbiotic Relationships - Vulnerability Trait 69
4.4.3 Habitat Specialization 71
4.4.4 Trophic Specialization 73
4.4.5 Anadromous/Catadromous 77
4.4.6 Growth and Productivity 78
4.5 Climate-Adjusted Baseline/Status Risks - Linking Baseline/Status & Climate Risks 82
Section 5. Temperature Predictions 84
5.1 Introduction 84
5.2 Future Temperature Predictions 85
5.3 Ecoregional Thermal Windows Approach 87
5.3.1 ETW Approach 87
5.3.2 Conceptual Framework 90
5.3.3 Within-Ecoregion Temperature Risks ("Worst-Case Scenario") 91
5.3.4 Abundance-Normalized Temperature Risks ("Ecosystem Services Risks") 92
5.3.5 Evaluation of ETW Thermal Thresholds 93
5.3.6 Data Source and Analysis 94
5.4 Biogeographical Thermal Limit Approach 97
5.4.1 Introduction 97
5.4.2 BTL Approach 97
5.4.3 Comparison of ETW and BTL 103
5.5 Northern Colonization 104
Section 6. Ocean/Coastal Acidification 106
6.1 Background 106
6.2 Background and Projected pH and Aragonite Saturation State (Qa) Values 107
6.2.1 pH Values 107
6.2.2 Aragonite Saturation State (Qa) Values 108
6.3 Toxicology Approach to Establishing pH Effects Thresholds 111
6.3.1 Introduction 111
6.3.2 Use of Maximum Acceptable Toxicant Concentrations (MATCs) to Generate
Comprehensive Effects Thresholds 111
iv
-------
6.3.3 Ocean Acidification Population Viability Effects Thresholds 146
6.4 Biotic Traits Modifying Sensitivity and Temperature-Adjusted Ocean Acidification Risks.... 149
6.5 Risk Type and Risk Algorithm 153
6.5.1 Risk Type 153
6.5.2 Risk Algorithm and Assignment of Sensitivity Classes 154
Section 7. Sea Level Rise 155
7.1 Introduction 155
7.2 Overview of SLR Approach 156
7.3 Eustatic Rates 157
7.4 Regional Isostatic Rates 158
7.5 Duration 161
7.6 Constrained Versus Unconstrained Habitats 161
7.7 High and Low Exposure Habitats 162
7.8 Habitat Thresholds 162
7.8.1 Introduction 162
7.8.2 Rocky Intertidal and Mussel Beds 165
7.8.3 Open Coastal Beaches, Backshore Beach Zones and Algal Beach Wrack 172
7.8.4 Emergent Marsh 172
7.8.5 Submerged Aquatic Vegetation 185
7.8.6 Tide Flats - Unvegetated Sand/Mud & Oyster Beds & Macroalgal Mats 191
7.8.7 Mangroves 200
7.9 Sea Level Rise Risks for Invertebrate and Fish Species 205
7.9.1 Depth Preferences 206
7.9.2 Habitat Preferences 207
7.9.3 Final SLR Risk 207
Section 8. Uncertainty Analysis and Quality Assurance/Quality Control 209
8.1 Uncertainty Analysis - Overview 209
8.2 Sources and Levels of Uncertainty 209
8.3 Reporting of Uncertainty 214
8.4 EPA/ORD's Quality Assurance/Quality Control 214
Appendix A.Under The Hood - Hardware, Software, Access Levels, & Backups 217
A-1 Servers 217
A-2 Software 218
A-3 Access Leve Is 218
A-4 Backup Strategy 219
Appendix B.Outputting Risk Assessment Results 222
B-1 Vulnerability Summary Output 222
B-2 Outputting Individual Climate Risks 224
B-3 Outputting Results for Northern Colonization 229
Appendix C. Near-Coastal Habitat Areas and GIS Metadata 231
C-1 Introduction 231
C-2 Near-Coastal Habitat Areas - Patterns of Offshore and Estuarine Habitats 231
C-3 Near-Coastal Habitat Areas - Geospatial Analysis 237
C-4 Calculation and Metadata for Computing Habitat Thresholds for West Coat Intertidal Rocky
Habitats due to Sea Level Rise using LiDAR Topobathy 239
v
-------
Appendix D. Evaluation of Temperature as Determinant for Warm-Edge Range Limits of Marine
Species243
D-1 Physiological 243
D-2 Range Shifts 244
D-3 Impaired Fecundity/Recruitment 245
D-4 Trophic Dynamic Shifts 245
Appendix E.Metadata of GIS Analysis of Temperature and Ocean Acidification Values 250
E-1 Aragonite Saturation State Projections by MEOW Ecoregion GIS Process 250
E-2 NOAA Climate Projections by MEOW Ecoregion GIS Process 250
Glossary of Terms 253
Bibliography 263
vi
-------
List of Tables
Table 1-1. Approaches used to predict effects of climate change on aquatic species and habitats 4
Table 1-2. Objectives of the current risk analysis framework and risk analysis of near-coastal species 7
Table 2-1. Potential climate change effects on individuals and populations within an ecoregion 19
Table 3-1. Hypothetical example of species' abundances at an ecoregion scale 27
Table 3-2. Definitions and quantitative cut points for the three-level relative abundance classifications... 31
Table 4-1. Baseline/status climate rules derived from biogeographic distributions, relative abundance, life
history traits, and population trends 41
Table 4-2. Baseline/status risks derived from productivity index parameters for fish 45
Table 4-3. Guidelines to distinguish between observed versus preferred habitats and environmental
conditions 48
Table 4-4. Number of species with endemic, restricted, or wide distributions 51
Table 4-5. MEOW provinces in the four major temperature regimes 56
Table 4-6. Arctic Ecoregions 58
Table 4-7. Number of Arctic endemics, small island colonizers, and nonindigenous species 59
Table 4-8. Small island ecoregions 61
Table 4-9. The number of species identified by each of the relative abundance rules 63
Table 4-10. Population trend classes based on percent change in population size within an ecoregion. .65
Table 4-11. Number of species with symbiotic relationship, habitat specialization, trophic specialization
and anadromous/catadromous reproduction 71
Table 4-12. Unique habitats of limited distribution 72
Table 4-13. Guidelines for assigning levels of trophic specialization for single and multiple feeding
modes 75
Table 4-14. Productivity index parameter thresholds for fishes 78
Table 4-15. Sebastes productivity parameters 79
Table 4-16. Climate-adjusted baseline/status risk values 83
Table 5-1. CMIP5 annual mean surface air temperature anomalies (°C) from the 1986-2005 reference
period to 2081-2100 for the four RCPs 84
Table 5-2. Temperature ranges (°C) associated with different risk levels for ecoregion mean annual
SSTs 88
Table 5-3. Temperature ranges (°C) associated with different risk levels for ecoregion mean summer
SSTs 88
Table 5-4. Temperature ranges associated with different risk levels for ecoregion mean winter SSTs
(Jan.-Feb.-March) 89
Table 5-5. Predicted increases in annual, summer, and winter SSTs for 2050-2099 based on the
RCP 8.5 scenario (°C) 90
vii
-------
Table 5-6. Number of exceedances of the moderate and high risk thresholds for annual SST based on the
ETW approach 96
Table 5-7. Historical and projected mean annual SSTs (°C) 99
Table 5-8. Historical and projected mean annual air temperatures (°C) 100
Table 5-9. Historical and projected mean summer air temperatures (°C) 100
Table 5-10. Historical and projected mean winter air temperatures (°C) 101
Table 5-11. Historical and projected mean 30-m temperatures (°C) 102
Table 5-12. Historical and projected mean 100-m temperatures (°C) 102
Table 5-13. Comparison of risk predictions using the ETW versus the BTL approaches 103
Table 6-1. Historical and projected annual pH 109
Table 6-2. Historical and projected summer pH.' 109
Table 6-3. Historical and projected winter pH 110
Table 6-4. Historical and projected aragonite saturation state values 110
Table 6-5. Summary of pH exposure experiments with decapods 114
Table 6-6. MATCs for pH for each decapod species based on all endpoints (comprehensive analysis). 144
Table 6-7. Comprehensive pH thresholds for high, moderate, and low sensitivity decapods using most
sensitive MATCs 146
Table 6-8. MATCs for pH based on the population viability endpoints for each decapod species 148
Table 6-9. Population viability pH thresholds values for high, moderate, and low sensitivity decapods.. 149
Table 7-1. Eustatic sea-level rise scenarios used as default values for ecoregion SLR risk analysis 158
Table 7-2. Derivation of ecoregion-specific isostatic rates 160
Table 7-3. Habitat thresholds associated with different levels of percent habitat loss 163
Table 7-4. Habitat threshold classes based on the percentage of habitat lost to sea level rise 165
Table 7-5. Studies predicting percentage loss of rocky intertidal habitat due to sea level rise 167
Table 7-6. Summary of low marsh percent habitat change under different SLR rates 174
Table 7-7. Submerged Aquatic Vegetation (Zostera marina), summary of seagrass percent habitat
change under different SLR values 188
Table 7-8. Summary of tide flat percent habitat change under different SLR rates 192
Table 7-9. Summary of mangroves percent habitat change under different SLR values and rates 202
Table 7-10. Risk values assigned to each combination of depth, habitat and exposure classes for each
habitat threshold 206
Table 8-1. Preliminary analysis of the major sources of uncertainty in the climate risk framework 210
Table 8-2. Level of confidence adapted from the IPCC 213
-------
Table A-1. Summary of the privileges associated with each level of access in CBRAT 220
Table D-1. Summary of studies supporting the assumption that temperature sets the warm-edge range
limits of marine species 246
Table E-1. Historical climate (1956-2005) and Anomaly (2050-2099) data 251
List of Figures
Figure 1-1. Study area and MEOW ecoregions comprising the Northeast Pacific and U.S. Arctic 11
Figure 2-1. Conceptual view of increased likelihood of adverse impacts with increasing risk level 18
Figure 3-1. Three-level relative abundance schema for use at regional scales 25
Figure 4-1. Example of an endemic species, defined as occupying only one MEOW ecoregion 52
Figure 4-2. Example of a species with a restricted distribution, defined as species occupying two MEOW
ecoregions 53
Figure 4-3. Example of a species occupying multiple ecoregions but not classified as having a wide
distribution 54
Figure 4-4. Example of a species with a wide distribution, defined as occupying three MEOW
provinces 55
Figure 4-5. Example of an Arctic endemic, defined as a species that occurs only in Arctic ecoregions.... 59
Figure 4-6. Default levels of trophic specialization based on single and two feeding modes 76
Figure 5-1. Distribution of Chionoecetes bairdi illustrating WOE, NWUE, COE, and NCUE ecoregions... 86
Figure 5-2. Abundance pattern of Hemigrapsus nudus as example for calculation of abundance-
normalized temperature risks 93
Figure 5-3. Schematic of the derivation of thermal risk values with the BTL approach 98
Figure 6-1. Cumulative distribution of the MATCs for each decapod species based on all endpoints 146
Figure 6-2. Cumulative distribution of the MATCs based on population viability endpoints for each
decapod species 149
Figure 6-3. Temperature-Adjusted Ocean Acidification Risks 151
Figure 7-1. Generalized sea level rise approach to calculating relative risk 157
Figure 7-2. Low marsh habitat - Constrained 185
Figure 7-3. Tide flats - Unconstrained 199
Figure 7-4. Tide flats - Constrained 200
Figure B-1. Output Vulnerability Summary screen 223
Figure B-2. Portion of Vulnerability Summary CSV - Output for climate risks 225
Figure B-3. Portion of Vulnerability Summary CSV - Output for baseline/status risks 226
Figure B-4. Outputting vulnerability summary limited to species with a specific risk or resilience factor. 227
Figure B-5. Screen for outputting risks associated with a specific climate stressor 228
Figure B-6. Portion of the output from an individual climate risk output 229
IX
-------
Figure B-7. Northern Colonization Test screen 230
Figure B-8. Portion of the output from Northern Colonization Test 230
Figure C-1. Total estuarine area in the Southern California Bight, Northern California, Oregon,
Washington, Vancouver Coast and Shelf, and Puget Trough/Georgia Basin ecoregions 233
Figure C-2. Areas of major estuarine habitats from Puget Trough/Georgia Basin to Southern California
Bight ecoregions 234
Figure C-3. Areas of offshore versus total estuarine unconsolidated habitats by ecoregion 235
Figure C-4. Area of major habitat types from 0-200 m offshore in the Southern California Bight
Ecoregion 235
Figure C-5. Area of major habitat types from 0-200 m offshore in the Oregon, Washington, Vancouver
Coast and Shelf Ecoregion 236
Figure C-6. Area of major habitat types from 0-200 m offshore in the Northern California Ecoregion 236
Figure C-7. Calculation of habitat thresholds by ecoregion for rocky intertidal habitats 242
x
-------
Change Log
11/22/2017: 508 compliant version.
11/22/2017: Because of its resolution, the CMIP5 model was mixing terrestrial and ocean air
temperatures in the Puget Trough/Georgia Basin Ecoregion. Therefore, the Puget
Trough/Georgia Basin Ecoregion air temperatures and projections were approximated by taking
the average of the values in ecoregions to the north and south (North American Pacific Fjordland
and Oregon, Washington, Vancouver Coast and Shelf ecoregions). The likelihood of northern
colonization was recalculated using these average air values, and Tables 5-8, 5-9, and 5-10 were
modified to include these new air temperatures.
XI
-------
THIS PAGE INTENTIONALLY BLANK
xii
-------
Acronyms and Abbreviations
Qa Aragonite saturation state
ACE Air, Climate and Energy
(EPA/ORD research program)
AOO Area of occupancy
AVHRR Advanced Very High Resolution
Radiometer (remote sensing)
BTL Biogeographic Thermal Limit
approach
°C Degrees Celsius
CBRAT Coastal Biodiversity Risk
Analysis Tool
cm Centimeter(s)
CMIP5 Coupled Model Intercomparison
Project Phase 5
COE Coolest occupied ecoregion
DEM Digital elevation model
DNRA Dominance normalized relative
abundance
EOO Extent of occurrence
EIS Environmental impact statement
EMAP Environmental Monitoring and
Assessment Program
EPA Environmental Protection
Agency
ESLR Eustatic sea level rise
ESRI Environmental Systems
Research Institute
ETP Eastern Tropical Pacific
ETW Ecoregional Thermal Window
approach
FAO Food and Agriculture
Organization of the United
Nations
GAM Generalized additive models.
GBIF Global Biodiversity Information
Facility
GIS Geographic information system
GLM General linear model
hr Hour
IPCC Intergovernmental Panel on
Climate Change
IUCN International Union for
Conservation of Nature
k von Bertalanffy growth
coefficient
km Kilometer(s)
Lat. & Long Latitude and Longitude
LIDAR Light Detection and Ranging
LME Large Marine Ecosystems
LOAEL Lowest observed adverse effect
level
m Meter(s)
MATC Maximum acceptable toxicant
concentration
MEOW Marine Ecoregions of the World
MHHW Mean higher high water
Xlll
-------
MLLW Mean lower low water
mm Millimeter(s)
MTP Mexican Tropical Pacific
Ecoregion
NCUE Next coolest unoccupied
ecoregion
ND No data
NEP Northeast Pacific
netCDF Network Common Data Form
NIS Nonindigenous species
NOAA National Oceanic and
Atmospheric Administration
NOAEL No observed adverse
effect level
NWI National Wetlands Inventory
NWP Northwest Pacific
NWUE Next warmest unoccupied
ecoregion
OA Ocean acidification
OBIS Ocean Biogeographic
Information System
Pa Pascal
PaCOOS Pacific Coast Ocean Observing
System
PDF Portable Document Format
PFMC Pacific Fishery Management
Council
PHf Free pH scale
PHnbs National Bureau of Standards
pH scale
pHsws Seawater pH scale
pHi Total pH scale
PICES North Pacific Marine Science
Organization
PSU Practical salinity units
r Intrinsic rate of increase
RACE Resource Assessment and
Conservation Engineering
(NOAA)
RCP Representative concentration
pathways
RSLR Relative sea level rise
SAV Submerged aquatic vegetation
SCAMIT Southern California Association
of Marine Invertebrate
Taxonomists
SDMs Species distribution models
SLAMM Sea Level Affecting Marsh
Model
SLR Sea level rise
sp Single species
spp Multiple species
SST Sea surface temperature
TLS Terrestrial laser scanning
TOC Total organic carbon
TSCA Toxic Substances Control Act
xiv
-------
USFWS United States Fish and Wildlife
Service
USGS United States Geological Survey
WOE Warmest occupied ecoregion
WoRMS World Register of Marine
Species
xv
-------
Acknowledgements
Special thanks to the following people who have provided programming support and design of
CBRAT: Dylan McCarthy, Rachel Nehmer, Marshall Hanshumaker and Robert Reusser. Emily
Saarinen, Melanie Frazier and Katie Marko provided insights into the early development of the
framework. Thanks to all the students and contractors who helped populate CBRAT including
Rebecca Loiselle, Summer Maga, Tracy Hoblit, Anthony Pham, Rochelle Regutti, Micaela
Edelson, Alma Meyer, Erin Horkan and Maya Kaup. Carol DeLong, our technical editor,
assisted with data entry and editorial reviews. The authors would also like to acknowledge Tim
Counihan and Jill Hardiman of the USGS Western Fisheries Research Center for their assistance
in collating the abundance and distributions of rockfish and for coordinating interagency
cooperation through an EPA-USGS Interagency Agreement. Dayv Lowry of Washington Dept.
Fish and Wildlife provided helpful insights on Puget Sound rockfish. Workshops with the
Southern California Association of Marine Invertebrate Taxonomists (SCAMIT) provided expert
information on several taxa and usability of CBRAT; Don Cadien, Paul Valentich-Scott, Gene
Coan, Doug Eernisse, Nora Foster, Greg Jensen, Ron Velarde, Mary Wicksten, Rick Brusca, and
Roger Clark all shared their time and expertise. Mary Mahaffy of the U.S. Fish and Wildlife
Service helped co-sponsor a workshop on trait-based risk assessment, which provided key
insights into the design of CBRAT. Maggie Dutch and Valerie Partridge (Washington Dept.
Ecology) reviewed an earlier version of this document and hosted a Biotic Matrix Workshop
with expert taxonomists and ecologists, resulting in improvements for CBRAT. We would
especially like to thank our taxonomic contractor Dancing Coyote Environmental under the
leadership of Larry Lovell and Dean Pasko, for their efforts to synthesize and manage data
contributions from colleagues John Chapman, Maria del Socorro Garcia-Madrigal, Kenneth
Coyle, Doug Diener, Francisco Solis-Marin, Rich Mooi, Phil Lambert, Megan Lilly, Sandy
Lipovsky, Leslie Harris, Jerry Kudenov, and Tony Phillips, who all made valuable contributions
to the project. The following peer reviewers provided insightful suggestions that improved both
the document and the risk assessment approach: Rebecca Flitcroft, Thomas Hurst, Walter
Nelson, Tony Olsen, Steve Rumrill, and James Markwiese. Finally, Dr. Lee would like to
acknowledge the continued support of EPA's Air, Climate, and Energy (ACE) research program.
xvi
-------
Executive Summary
Goals and Objectives
With increasing temperatures, ocean acidification, and sea level rise (SLR), climate change is arguably
the greatest threat facing near-coastal ecosystems (0-200 m depth). For management to respond in a
scientifically-sound fashion, it is critical to have a basic knowledge of the extent and pattern of risk to
near-coastal species. To address these needs, we developed a rule-based framework to predict the relative
risk of near-coastal species to climate change at regional scales. The framework synthesizes risks from
biotic traits (baseline risks) and population status (trends) with risks predicted from increasing ocean
temperatures, ocean acidification, and sea level rise.
Within this overall goal, key objectives were to develop a framework capable of predicting: a) climate
risks for rare species as well as for better studied species; b) identifying major climate stressor(s) affecting
each species within each region; c) geographic patterns of the importance of different climate stressors;
and d) how risk changes under different climate scenarios. We developed an ecoinformatics website, the
Coastal Biodiversity Risk Analysis Tool (CBRAT), to conduct the climate risk analyses and to serve as a
practical tool for managers and researchers to address climate and species inquiries. As detailed in this
document, over thirty rules were used to predict a species risks due to temperature, ocean acidification, or
sea level rise. As discussed under "Uncertainty," we contend that the present framework is able to identify
high risk vs. low risk species and regional risk patterns but does not have the resolution required for
fisheries management.
Geographic and Taxonomic Scope of Current Framework
The Marine Ecoregions of the World (MEOW) is used as the biogeographic schema for evaluating
regional distributions and climate risks. The present effort focuses on species in the twelve MEOW
ecoregions that make up the Northeast Pacific (NEP) and U.S. Arctic, ranging from the Gulf of California
through the Beaufort Sea, however the main focus was from Southern California north. To evaluate the
efficacy of the framework, the current effort focuses on calculating preliminary risks for brachyuran and
lithodid crabs (417 species), rockfish (71 species), and bivalves (892 species) that occur within 200 m
depth.
Risk Categories and Overall Vulnerability
In the current framework, each risk rule generates one of four risk levels for each species: minor, low,
moderate, or high risk. As the risk level increases, the likelihood, severity, and types of adverse impacts
increase, as does the ability to detect such changes especially with the more abundant species. Climate
impacts may range from physiological changes to population impacts and while it is not possible to
predict the specific effects under each scenario, population declines are expected with high risk scenarios.
We attempted to standardize the risks across different traits and climate stressors so that a high risk for
one climate stressor is approximately equivalent to a high risk for another stressor. However, this proved
difficult for ocean acidification because of the predominance of laboratory exposures using physiological
xvii
-------
and behavioral endpoints not readily related to population viability. Even with this limitation, the overall
risk value for a species within an ecoregion is calculated as the single greatest risk among the climate
adjusted baseline/status risks and the risks calculated for temperature, ocean acidification, and sea level
rise.
Expert Opinion Versus Algorithm-Based Approaches
Most risk assessments of marine and freshwater organisms incorporating multiple climate stressors have
used expert opinion to determine risk. While useful when there is limited knowledge, relying on experts is
prone to a number of disadvantages including: a) experts are subject to at least nine types of cognitive
biases; b) lack of transparency; c) need to reconvene experts to evaluate new climate scenarios, taxa, or
locations; d) lack of consistency among experts; and e) limitations of human experts being able to
evaluate hundreds to thousands of species across wide geographical areas. To address these limitations,
we developed an algorithm approach where the risk is automatically generated from a centralized
knowledgebase stored in CBRAT and a set of explicit rules.
CBRAT - Web-Based Risk Analysis Tool
A distinguishing feature of the current effort is that the risk framework is implemented in an online
ecoinformatic tool, the "Coastal Biodiversity Risk Analysis Tool" (http://www.cbrat.org/'). CBRAT
serves as the platform to calculate the climate risks using the associated knowledge base of biotic traits
and rule sets along with the user input climate values. A key feature is that managers and researchers are
able to easily evaluate different climate scenarios and assumptions by changing the baseline or future
climate values and/or the effects thresholds for temperature, ocean acidification, and sea level rise.
CBRAT outputs all the biotic trait information for each species (e.g., depth preferences) as well as the risk
associated with each rule for each species by ecoregion. This output allows users to evaluate the details of
risk patterns as well as use the synthesized biotic trait for other types of analyses.
Ecoregion-Scale Relative Abundances
Biogeographic distributions identify where a species can survive while abundances help elucidate
preferred versus marginal environmental conditions. Because of the insights abundances provide, we
developed an approach to classifying the relative abundances of each species at an ecoregion scale using a
hierarchical abundance schema. A "hybrid" approach integrating regional and local quantitative survey
data, natural history texts, expert opinion, and online biodiversity databases was used to estimate relative
abundances. Using this synthesis of datatypes, it was possible to estimate relative abundances for
essentially all the crabs, rockfish, and bivalves in each ecoregion from Southern California through the
Beaufort Sea.
Baseline/Status Risks
The first method to identify species at risk was a set of rules using "baseline" biotic traits, such as a
species' range, and status metrics, such as population trends, which are associated with increased climate
vulnerability or resilience. Such baseline/status indicators are widely used in conservation and have the
advantage that the data are available for most species The main disadvantage is the difficulty of predicting
how the risk associated with a specific baseline/status trait changes under different specific climate
-------
scenarios, though it is possible to link the overall baseline/status risk to the overall level of climatic risk.
We generated 17 rules predicting vulnerability or resilience to climate change that can be applied to both
well-studied and lesser known species. Because abundance is an indicator of a population's viability,
relative abundance is used to modify the risk level for many of these rules. The baseline/status risks most
frequently indicating high vulnerability are endemicity, habitat specialization, symbiotic relationships,
current population declines, and population growth metrics, though the importance of these traits varies
geographically and among taxa.
Temperature Risks and Northern Colonization
The core method developed to predict risks associated with increased temperatures was the Ecoregional
Thermal Windows approach (ETW) that compares the projected sea surface temperature (SST) in each
ecoregion to the historic range of SST values in the "warmest occupied ecoregion" or WOE.
Temperatures in the WOE represent the warm range limit of a species and are assumed to represent the
upper ecological thermal limit for a species to maintain a viable population. For this analysis, the
ecoregion-scale historic SSTs were derived from an analysis of 28 years of "advanced very high
resolution radiometer" (AVHRR) remote sensing data while the future projections were extracted from
the CMIP5 model used by the IPCC served through NOAA's Climate Change Web Portal. The level of
risk is determined by comparing the projected SST in each ecoregion to the historic mean plus a number
of standard deviations (SD) in the WOE. Moderate risk is defined as a projected SST in a northern
ecoregion greater than the WOE mean + 2 SDs while high risk is defined as a projected temperature
greater than the WOE mean + 3 SDs. The "representative concentration pathway" (RCP) 8.5 is used as
the default in CBRAT but users are able to input ecoregion-specific temperatures associated with any
climate scenario.
To evaluate risks for species occurring at different depths, we developed the Biogeographical Thermal
Limit (BTL) approach that predicts risks for intertidal species using projected air temperature, shallow
subtidal species using projected temperatures at 30 m depth, and deep subtidal species using projected
temperatures at 100 m depth. For this analysis, both the baseline temperatures and future projections were
based on the CMIP5 model. The BTL approach compares the projected temperatures in the target
ecoregion to temperature thresholds for each depth based on four bins between the historic temperatures
in the WOE and the "next warmest unoccupied ecoregion" (NWUE). The NWUE is usually immediately
to the south of the WOE, and is assumed to be too warm to maintain a viable population of the target
species, with high risk defined as a projected temperature greater than the 3rd bin between the WOE and
NWOE. The BTL approach generated the same risks as the ETW over 87% of the time with the
brachyuran crabs when compared from the Beaufort Sea to Southern California. When there was a
deviation, the BTL was less sensitive. Because of this difference, the moderate and high risks generated
by the BTL were combined as "at risk species".
A geographic pattern emerging from a preliminary analysis of brachyuran crabs with both the ETW and
BTL approaches is that high thermal risks are primarily limited to the southernmost occupied ecoregion
of a species. The lack of substantial thermal impacts in the more northern range of a species assumes
either that warm-tolerant genotypes occur in the ecoregions north of the WOE or that warm-genotypes
from southern ecoregions migrate northward. To assess the "worst-case" scenario assuming no warm-
xix
-------
tolerant genotypes, CBRAT also calculates risks by comparing the projected temperature in the target
ecoregion to the historic temperate range within ecoregion. A third view of risk is to compare the target
ecoregion to the southern ecoregion where the abundance declines, presumably because it is too warm,
and is most applicable to evaluating impacts on commercial species. This approach is discussed in the
document but not implemented in the current version of CBRAT.
Climate change may also result in sufficient warming of cooler ecoregions to allow northern range
expansion. To evaluate this potential, we reversed the logic of the BTL approach and derived temperature
thresholds based on the bins derived between the "coolest occupied ecoregion" (COE) and the "next
coolest unoccupied ecoregion" (NCUE), which is usually immediately to the north of the COE. CBRAT
outputs the "suitability" for colonization in the unoccupied northern ecoregions based on the future
projected temperatures, while recognizing that other factors could limit a species' expansion.
Ocean Acidification
Though the least well understood of the climate stressors, it is possible to conduct a first-order regional-
scale risk assessment of ocean acidification by treating it like other contaminants. Specifically, we
propose deriving "maximum allowable toxicant concentrations" (MATCs) for pH and aragonite
saturation state from a synthesis of exposure experiments. MATC is the geometric mean of the "no
observed adverse effects level" (NOAEL) and the "lowest observed adverse effects level" (LOAEL), and
in the present context is the lowest "allowable" pH for a particular species. Because of the limited number
of exposure experiments, the proof-of-concept with the decapods takes a comprehensive approach and
uses the single most sensitive MATC for each species regardless of the specific endpoint or life history
stage. To generate ocean acidification risks more similar to the population associated risks for
temperature and sea level rise, we conducted a similar analysis just using endpoint directly related to
population viability; however, the number of studies is too limited to currently to generate reliable effects
thresholds.
Because species within ataxon vary greatly in their sensitivity, a cumulative frequency distribution curve
is generated from the most sensitive MATC for each species within a taxon. This frequency distribution is
then used to generate high, moderate, and low sensitivity thresholds to pH and aragonite saturation state.
After assigning a sensitivity class to a species, its risk is calculated by overlaying the specific pH effects
thresholds on ecoregion-scale projected values. Because of the reported interaction between elevated
temperatures and reduced pH, moderate ocean acidification risks are elevated to high acidification risk
under moderate to high temperature risk. Baseline and projected pH values for surface waters were
presented in the document and CBRAT from the CMIP5 model served through NOAA's Climate Change
Web Portal, using RCP 8.5 as the default. Baseline and projected aragonite saturation state values were
presented based on projections developed by Cao and Caldeira based on the University of Victoria Earth
System Climate Model version 2.8.
Based on a preliminary risk assessment with decapods, assignment of the high, moderate, or low
sensitivity threshold to a species has a major effect on its ocean acidification risk assignment. Which
raises the question on how to assign sensitivity classes for species lacking experimental studies. Ideally, it
will be possible to assign sensitivity classes based on readily available physiological or life history traits;
xx
-------
to date the single example is that species with brood protection and/or lecithotrophic larvae have low
sensitivities. Another approach is to classify species by their ecological and taxonomic similarity to
experimentally tested species. Alternatively, moderate sensitivities could be used a "restrained" analysis
and the high sensitivity thresholds as a "high risk" analysis.
Sea Level Rise
The objective of the sea level rise component is to estimate the population decline in the invertebrate and
fish species inhabiting intertidal habitats based on the assumption that population declines are
proportional to the extent of habitat loss. Predicting SLR risk species integrates four steps. The first is to
estimate a net ecoregion sea level rise value (mm) for each ecoregion from the global eustatic rate and
regional rates of isostatic adjustment. The second step is to generate "habitat thresholds" for each of the
major intertidal habitat types from the literature and SLR models; models included SLAMM for wetlands
and mangroves and a LIDAR/topobathy model we developed for the rocky intertidal. These thresholds
classify the percent loss of each habitat type as minor, low, moderate, or high based on the extent of net
SLR within each ecoregion. To account for inland migration of habitats, habitat thresholds were
developed for both "unconstrained" and "constrained" (coastal squeeze) scenarios with the constrained
thresholds used for Puget Sound through Southern California and the unconstrained thresholds used for
the less developed ecoregions. The proportion of a species' population at risk due to loss of intertidal
habitats depends upon its depth distribution. Thus, the third step is to generate risk values for the target
species based on the habitat thresholds and species' depth preferences. Because many near-coastal species
occupy multiple habitats, the final step is to assign the greatest SLR risk across all observed and preferred
habitats occupied by the species.
In a preliminary analysis with brachyuran crabs and an "intermediate-high" eustatic SLR rate of 12
mm/yr, moderate and high SLR risks were limited to primarily intertidal crabs from Puget Sound south
through the Cortezian ecoregion. The lack of risk in the northern ecoregions is due to high isostatic uplift
countering SLR in much of Alaska and the paucity of intertidal crabs in the Arctic.
Uncertainties and Limitations
Climate change predictions are subject to a number of uncertainties, and the current document lays out a
strategy for qualitative uncertainty analysis. The components of such an analysis are: 1) identification and
characterization of uncertainty sources; 2) estimates of the direction and relative magnitude the
uncertainty is likely to have on results; and 3) reporting of qualitative uncertainties in a non-technical
summary. In addition to the qualitative analysis, it is possible to conduct a quantitative uncertainty
analysis on the key climate projections and effects thresholds by changing input values in CBRAT.
A preliminary uncertainty analysis indicates that: 1) predictions from the current framework are sufficient
to identify the scope and patterns of risk and for regional-scale adaptation planning; 2) predictions are
sufficient to flag high risk commercial species but not for fisheries management; 3) lack of sufficient
spatial resolution in the current regional-scale climate models limits the ability to predict temperature and
pH changes within estuaries, increasing the uncertainty for estuarine organisms; and 4) the greatest
uncertainty appears to be associated with ocean acidification. It is not unsurprising that there are a number
of uncertainties in predicting the effects of multiple climate stressors on hundreds of species over the
xxi
-------
entire U.S. Pacific Coast. But with higher resolution climate models and additional effects research it
should be possible to reduce these uncertainties overtime.
XXll
-------
Section 1. Introduction and Overview
1.1 Problem Statement
Climate change is arguably the greatest threat facing near-coastal ecosystems (0-200 m depth) in
recent history. Over the last 100 years, climate change has resulted in documented increases in
ocean temperatures, reduced pH, and increased sea levels (Doney et al., 2009; Portner et al.,
2014; Blunden and Arndt, 2016). These climatic alterations have, or will, result in a host of
ecological impacts, ranging from species' range shifts to a loss of ecosystem functions to
regional/global extinctions (Harley et al., 2006; Hannah, 2012). The nature and extent of these
impacts can vary substantially among species according to their exposure to specific climate
stressors as well as their life history, physiological, and population traits. As a simple example, a
2-meter sea level rise (SLR) would have major effects on many intertidal species but a trivial
effect on continental shelf species. Besides these species differences, the ecological impacts of
climate change will vary regionally, both in response to geographical differences in the extent of
climate alterations as well as latitudinal differences in biodiversity and species composition.
For management to respond in a scientifically-sound fashion to climate change, it is critical to
have a basic knowledge of what species are at the greatest risk, what climate stressors represent
the largest threats, where risk is the greatest, and how risk varies with different climate scenarios.
Understanding the relationship between the nature and extent of risks with different climate
scenarios informs policy makers of the potential benefits to reducing emissions while a
knowledge of the geographical patterns of risk helps set regional adaptation priorities.
Knowledge of the vulnerability of rare species is important both because of their contributions to
ecological functions and genetic diversity (Balint et al., 2011; Prather et al., 2013) as well as
their central role in conservation and adaptation efforts (Raphael and Molina, 2007; Angulo et
al., 2009). However, a major challenge in addressing the full breadth of species along the U.S.
coast is the diversity of near-coastal species. Almost 1500 species of fish occur along the U.S.
West Coast (Love et al., 2005) and over 1000 bivalve species have been reported from Alaska
through northern Mexico (Coan et al., 2000; Coan and Valentich-Scott, 2012).
Not surprisingly for a threat of this scope and complexity, a number of different approaches are
being applied to identify vulnerable species and habitats (Table 1-1). Each has benefits and
limitations - some are better at revealing underlying mechanisms while others are better at
generating predictions for a large number of species. Some require massive amounts of
quantitative survey data, while others can be applied via data mining. While all the approaches in
Table 1-1 are complementary, only three potentially address our objectives of assessing multiple
species at regional scales: 1) species distribution models; 2) evaluation of species' "climate
velocities"; and 3) trait-based approaches.
1
-------
Species distribution models (SDMs) include a suite of statistical approaches based on associating
records of where a species occurs, or its abundance, with the environmental parameters at each
occurrence (Elith and Leathwick, 2009). They have been used to evaluate nonindigenous species
(e.g., Reusser and Lee, 2008; Herborg et al., 2009), native species for conservation purposes
(Pearson, 2010), and to evaluate distributional changes in response to climate change (e.g.,
Mellin et al., 2012; Reusser et al., 2016). Advantages are that SDMs can predict the potential
response of target species at a relatively fine-scale resolution and can evaluate species' responses
to different temperature scenarios, assuming the data set used to construct the model
encompasses all or most of the species' temperature range.
A disadvantage of SDMs is the number of samples required to generate robust models; based on
the "one in ten" rule of thumb, a minimum of 10 samples ("observations") containing the
targeted species ("event") is required for each predictor variable in the model (see Harrell et al.,
1996; Babyak, 2004), while other authors suggest at least 20 observations per variable is required
to avoid overfitting (Steyerberg et al., 2000). The sampling requirement generally limits SDMs
to more abundant species, though in one case the SDM was linked into online databases
(FishBase and SeaLifeBase) allowing an evaluation of over 1000 exploited fish and invertebrates
(Cheung et al., 2008, 2009). Besides the sample size limitations, we are unaware of any cases
where SDMs have been used with ocean acidification or sea level rise.
A novel approach is the evaluation of "climate velocity", or shifts in population centroids across
the landscape which are presumably in response to recent temperature changes (Pinsky et al.,
2013). These authors evaluated 360 marine taxa. However, even more than with the SDMs, the
large number of quantitative samples needed is a major limitation; the Pinsky et al. analysis used
a database of 128 million individuals primarily from the NOAA's RACE groundfish trawl
surveys. Molinos et al. (2016) expanded upon this technique by linking into modeled species
distributions from AquaMaps (http://www.aquamaps.org/main/home.php') combined with
projections of future sea surface temperatures (SSTs), resulting in predictions for over 12,000
near-coastal and oceanic species. While this approach provides important insights into regional
and global responses to ocean warming, it is not suitable for rare species, many of which have no
or very few records in the AquaMaps. Additionally, assessments of climate velocity do not
appear to be suitable for evaluating either ocean acidification or sea level rise.
The last of the three potential approaches, trait-based analyses, have been used to address a
number of conservation issues. Among marine fishes, traits have been used to evaluate the
effects of overfishing (e.g., Musick, 1999; Dulvy et al., 2004; Reynolds et al., 2005). Reynolds et
al. (2005) concluded, "Simple life history traits can be incorporated directly into quantitative
assessment criteria, or used to modify the conclusions of quantitative assessments, or used as
preliminary screening criteria for assessment of the 95% of marine fish species whose status has
yet to be evaluated either by conservationists or fisheries scientists."
2
-------
Trait-based approaches have also been applied to assess climate vulnerability. In a simulation
study with amphibians and reptiles, Pearson et al. (2014) inferred, "extinction risk due to climate
change can be predicted using a mixture of spatial and demographic variables that can be
measured in the present day without the need for complex forecasting models." With freshwater
and marine species, the IUCN evaluated 797 coral species globally, Moyle et al. (2013)
evaluated all the native and nonindigenous freshwater fishes in California while Hare et al.
(2016) evaluated 82 marine fishes and invertebrates on the Northeast U.S. continental shelf.
These studies indicate that trait-based risk assessments can be successfully conducted for a large
number of species over wide geographical locations. Further, combining risks derived from
projections of future temperatures, SLR, and pH with the trait analysis, as did Hare et al. (2016),
strengthens the predictions.
In reviewing these methodologies, we concluded the most rigorous approach to predicting effects
of multiple climate stressors, evaluating species with limited data, evaluating geographical
patterns of risk and conducting assessment on different climate scenarios was to integrate a trait-
based approach with climate effects thresholds. As discussed in Sections 5-7, effects thresholds
are numerical values indicating different levels of risk for temperature, pH and sea level rise that
are overlain on projected climate values. The analysis of biotic traits augments the climate
thresholds by identifying at risk species potentially missed by the comparison of regional climate
values with general impact levels. There are, however, a number of differences between our
framework and the previous trait-based efforts, including use of an algorithm-based risk
assessment versus expert solicitations (Section 1.5) and integrating relative abundance into the
analysis (Section 3). Another key difference is that the climate risk analysis and associated data
are available via an online tool, the Climate Biodiversity Risk Analysis Tool (CBRAT,
faftp://www.cbrat.ore), allowing managers and researchers to review the information and conduct
their own risk assessments.
3
-------
Table 1-1. Approaches used to predict effects of climate change on aquatic species and habitats.
# Species Analyzed = Number of species analyzed within a study. SDM = species distribution model. GLM = generalized linear model. GAM =
generalized additive models. NIS = nonindigenous species.
Approach
Primary Climate
Stressors Evaluated
# Species Analyzed
Comments
Examples
Laboratory exposures
Temperature, Ocean
acidification or both
Few
Variations in exposure conditions and response
variables make it difficult to compare studies; best
used as inputs into predictive models.
Johansen and Jones,
2011; Waldbusser et al.,
2013; Long et al., 2016
Biochemical
responses
Temperature,
Ocean acidification
Few
Utility as predictive vs. monitoring approach is
unclear. Extrapolation to other species unclear.
Helmuth and Hofmann,
2001; Tomanek, 2010
Field experiments
Primarily temperature
Few
Bias towards intertidal species. Usually at local scale.
Extrapolation to other species unclear.
Yamane and Gilman,
2009; Jones et al., 2012
SLAMM
Sea level rise
Wetland habitats
Site specific SLR model for wetland habitats, not the
associated species. Moderately high data
requirements. Primarily at local scale.
Glick et al., 2007; Craft
et al., 2009; Lee et al.,
2014
Mechanistic
population models
Temperature or
Ocean acidification
Few
High data requirements, largely limited to well-
studied commercial species. Potentially can model
temperature, ocean acidification, and/or SLR.
Buckley et al., 2010
SDM - Presence only
or with abundance
data (GLMs & GAMs)
Primarily temperature and
Habitat
One to dozens
Reasonably high data requirements, not suitable for
rare species. Can elucidate geographic patterns of
risk if sampled at appropriate scale and covers
adequate temperature range.
Brown et al., 2011;
Jones et al., 2013;
Reusser et al., 2016
SDM - Abundance
drawing on global
databases
Temperature
1066 fishes and
invertebrates
Special case of linking into global databases. Not
suitable for rare species. Temperature only.
Cheung et al., 2008,
2009
Climate velocity -
NOAA RACE data
Temperature
360 fishes and
invertebrates
High data requirement for quantitative samples -
linked into NOAA RACE data. Not suitable for rare
species. Temperature only.
Pinsky et al., 2013
Climate velocity -
AquaMap modeled
distributions
Temperature
12,796 fishes and
invertebrates
Links to modelled probability distributions, modeled
geographic patterns of risk. Not suitable for rare
species. Temperature only.
Molinos et al. 2016
Trait-based: IUCN
Temperature and Ocean
acidification
797 corals
Based on expert solicitation with climate projections.
Included rarer species.
Foden et al., 2008, 2013
Trait-based:
Freshwater fishes
Climate in general
121 native & 43 NIS
freshwater fish
Not geographically specific (all of CA). Based on
expert solicitation. Included rarer species
Moyleetal. 2013
4
-------
Approach
Primary Climate
Stressors Evaluated
# Species Analyzed
Comments
Examples
Trait-based & Climate
projections: NE
Atlantic shelf
Temperature, Ocean
acidification, Sea level rise,
Precipitation, Salinity,
Currents
82 coastal fishes &
invertebrates
Included greater range of climate stressors, based on
expert solicitation. Focused on common species. Did
not elucidate geographic patterns of risk.
Morrison et al., 2015;
Hare et al., 2016
Trait-based & climate
effects thresholds:
Pacific Coast
Temperature, Ocean
acidification, Sea level rise
387 crabs, 71 bottom-
associated rockfish, &
884 bivalves
Algorithm-based risk calculations. Included rare
species. Modeled geographic patterns of risk.
Current study
5
-------
1.2 Goals and Objectives
The overall goal of the current research is to predict the relative risk of near-coastal (0-200 m
depth) species to climate change at regional scales. As detailed in this document, we approached
this challenge by developing a climate risk assessment framework that synthesizes predictions
based on biotic traits or attributes ("baseline/status risks", Section 4) and predictions based on
overlaying effects thresholds on projected values for temperature, ocean acidification, and sea
level rise (Section 5-Section 7). Both the baseline/status risks and the climate risks are generated
via a set of rules that are detailed in the appropriate sections. The analysis is conducted and risks
reported at the spatial scale of the "Marine Ecoregions of the World" (MEOW, Spalding et al.,
2007), ranging across twelve ecoregions from the Gulf of California through the Beaufort Sea.
The "Coastal Biodiversity Risk Analysis Tool" (CBRAT; http://www.cbrat.org/) is the platform
used to calculate the climate risks using the associated knowledge base of biotic traits, climate
projections, and rule sets. Programming details and metadata on CBRAT are given in Appendix
A, while Appendix B provides an overview on how to conduct risk assessments in CBRAT.
Within the overall goal of predicting climate risk, there is a suite of more specific project
objectives that are listed in the Table 1-2.
6
-------
Table 1-2. Objectives of the current risk analysis framework and risk analysis of near-coastal species.
OBJECTIVES
Climate Risk Framework
• Develop a framework that predicts the relative risks associated with temperature, ocean acidification and sea
level rise for species within each MEOW ecoregion as well as an overall climate risk for each ecoregion.
• Develop a framework that identifies the major climate stressor(s) affecting each species within each
ecoregion.
• Develop a framework that predicts relative climate risks for rare species with limited data as well as for the
better studied species.
• Develop approaches to integrating multiple climate stressors, especially temperature and ocean acidification
CBRAT - Web-Based Risk Analysis and Research Tool
• Develop an online system, CBRAT, as the tool to integrate biotic traits, historical and projected environmental
values, and the rules to predict relative risk
• Design CBRAT such that approved managers and researchers, as well as the CBRAT administrators, can
evaluate different climate scenarios.
• Synthesize biotic trait information for use in addressing non-climate management and research
questions/issues
• To the extent practical, promote CBRAT as a public outreach tool for the informed public.
Geographical and Taxonomic Patterns of Risk
• Predict how climate risk varies regionally for each species.
• Predict the geographic patterns of the relative importance of different climate stressors from the Gulf of
California through the Beaufort Sea.
• Evaluate how risk varies among major taxa as well as the relative importance of different climate stressors for
different taxa.
Evaluate Different Climate Scenarios
• Evaluate the relative risks associated with different climate scenarios for temperature, ocean acidification, and
sea level rise.
Transparency/Uncertainty analysis
• Provide transparency in the data used and in the rules to predict risk.
• Document each rule such that a user could calculate the risk manually.
• Document the major assumptions associated with each rule.
• In CBRAT, document the sources of information used to assign biotic traits along with any associated
assumptions.
• Generate a qualitative uncertainty analysis.
1.3 Scope of Document
The focus of the current report is to document the climate risk framework we developed for near-
coastal species. The specific rules and key assumptions for the baseline/status risk, temperature
increases, ocean acidification, and sea level rise are given in Section 4, Section 5, Section 6, and
7
-------
Section 7, respectively. The rules are described in sufficient detail that a user could evaluate the
risk for a single species by hand. Examples of the types of data that are used in the risk
assessment are presented to illustrate the types of required data and how they are analyzed. For
example, a detailed review of the effects of pH on decapods is provided (Section 6) but a review
is not provided for aragonite saturation since the methods are the same for both.
In addition to the conceptual framework, this report documents how the web-based Coastal
Biodiversity Risk Analysis Tool (CBRAT; http://www.cbrat.oref) is used to conduct the risk
analyses. This document is not a user's manual, though Appendix B provides a guide on how to
conduct the risk assessments in CBRAT. For further information, the user is referred to the
"User's Guide & Metadata to Coastal Biodiversity Risk Analysis Tool (CBRAT): Framework for
the Systemization of Life History and Biogeographic Information" (Lee et al., 2015), which is
available on CBRAT.
The results of the climate risk assessments will be detailed in a separate document (Lee et al., in
progress). This initial risk assessment will analyze all the near-coastal brachyuran crabs (365
species), lithodid crabs (22 species), bottom-associated rockfish (71 species), and bivalves (884
species) reported from the Gulf of California through the Beaufort Sea.
1.4 Geographic and Taxonomic Scope
1.4.1 Geographic Scope
We use the "Marine Ecoregions of the World" (MEOW) (Spalding et al., 2007) as the
biogeographic framework for evaluating the distributions and abundances of near-coastal species
as well as for assessing climate risk. MEOW is a hierarchical schema for marine coastal waters
down to a depth of 200 m. The original three levels of MEOW include ocean basin realms
divided into smaller provinces and then smaller ecoregions. As defined by Spalding et al. (2007),
ecoregions are "Areas of relatively homogeneous species composition, clearly distinct from
adjacent systems. The species composition is likely to be determined by the predominance of a
small number of ecosystems and/or a distinct suite of oceanographic or topographic features." To
capture differences in the eastern and western sides of the Atlantic and Pacific, we previously
modified the MEOW schema by adding a fourth level, the region which is between a realm and
province (Reusser and Lee, 2011; Lee and Reusser, 2012). Detailed maps of the world's 232
ecoregions are available in CBRAT under the Documents tab; GIS shapefiles are available at
http://maps.tnc.org/gis data.html.
The present effort evaluates species' vulnerabilities in the 12 ecoregions that make up the
Northeast Pacific (NEP) and U.S. Arctic (Figure 1-1). The three U.S. Arctic ecoregions (Eastern
Bering Sea, Chukchi Sea, Beaufort Sea- continental coast and shelf) are in the Arctic Realm
(Arctic is not broken into provinces). The Cold Temperate Northeast Pacific Province is
composed of the six ecoregions ranging from Northern California up through the Aleutian
8
-------
Islands. The MEOW Warm Temperate Northeast Pacific Province is composed of the Cortezian,
Magdalena Transition, and Southern California Bight ecoregions. We extend the study into Baja
California and the Gulf of California so as to fully capture the Warm Temperate Northeast
Pacific Province. Additionally, it is likely that many of the near-coastal species in these
ecoregions will migrate northward with warming so their inclusion helps predict future colonists.
We chose the MEOW schema over the other existing biogeographic schema because it appears
to best capture biological reality across the globe. It has been used in a variety of biodiversity
and conservation studies, including a global assessment of human impacts on marine ecosystems
(Halpern et al., 2008), wetland conservation Ramsar Convention (Ramsar, 2008), IUCN's
assessment of global ocean protection (Toropova et al., 2010), assessing biogeographic patterns
of marine invaders (Molnar et al., 2008; Lee and Reusser, 2012) and in analyzing global
biodiversity patterns of various taxa (Piepenburg et al., 2011; Van Soest et al., 2012; Barboza
and Defeo, 2015; Molinos et al., 2016). Additionally, species' distributions can be viewed by
MEOW ecoregions in the online Ocean Biogeographic Information System (OBIS;
http://iobis.org/mapper/).
In terms of the appropriateness of this spatial scale for risk analysis, the ecoregion level is large
enough to capture population level responses to regional climate changes. Since population
declines resulting in rarity is a potential impact of climate change, Gaston's conclusion (1994)
that "the concept of rarity can be applied to almost any spatial scale, it is of primary interest and
has been most extensively studied at regional or biogeographic scales" supports our focus on
regional risk assessments. MEOW ecoregions are also large enough to incorporate both major
and minor near-coastal habitat types along with their associated taxa. Conversely, the MEOW
ecoregions are small enough to detect geographical patterns in risk. In contrast, with the
commonly used Large Marine Ecoregions (LMEs) schema, the California Current LME extends
from the entrance to Puget Sound to the entrance of the Gulf of California, an area that is divided
into four MEOW ecoregions. Lastly, MEOW ecoregions are an appropriate scale to inform many
management decisions, in particular those related to conservation and climate impacts on
populations (e.g., Ramsar, 2008; Toropova et al., 2010).
1.4.2 Taxonomic Scope and Standardization
The risk framework described in this document should, in theory, apply to all near-coastal fishes
and invertebrate taxa that occur within 200 m depth assuming the basic distributional and biotic
trait data are available. Having said that, each major tax on may require modification of some of
the rules or addition of new rules. As described in Section 4, the available information allows a
set of baseline/status risk rules for fish based on productivity, but such rules are not currently
available for invertebrates. Since many rules predicting risk are based on biogeographical
patterns, the general approach should be modifiable for submerged aquatic vegetation (e.g.,
Zostera spp.), marsh plants, and macroalgae (e.g., kelp), though new rules would likely be
9
-------
necessary to capture the physiology of primary producers (e.g., positive effect of increased CO2).
Lastly, relative abundances and the corresponding risk levels are generated for the portion of
species' populations that occur at depths from 0-200 m. Thus, there is greater uncertainty in the
risk predictions for oceanic species for which 0-200 m only constitutes a small portion of their
total population.
Our strategy is to evaluate the species tax on by taxon rather than by habitat. The major
conceptual reason for evaluating vulnerabilities by taxon is that we use relative abundance within
an entire taxon as one of the attributes to assess risk (Section 3), thus promoting evaluations of
an entire taxon for comparative purposes. A practical advantage is that much of the literature is
taxon based, thus it is more efficient to synthesize biotic traits by major taxon.
While it is a common refrain that there are not enough taxonomists (e.g., Kim and Byrne, 2006),
it has been our experience that there are more than enough to cause mischief for ecologists and
biogeographers. Because of differing views on taxonomy, we standardize using the World
Register of Marine Species (WoRMS, http://www.marinespecies.ore/index.php. Costello et al.,
2013). A downloaded version of the WoRMS database is incorporated into CBRAT so that every
new species is checked against the WoRMS higher level taxonomy as it is added. Species in
CBRAT and their synonyms are periodically compared against those in WoRMS using the
WoRMS "Match taxa" to find updates in taxonomy and errors such as valid species being
entered as synonyms. For a few taxa, we use regional authorities in lieu of WoRMS. In
particular, we use the regional treatises of Coan et al. (2000) and Coan and Valentich-Scott
(2012) for bivalves. Such species are identified in CBRAT on the species' taxonomy page as
deviating from WoRMS.
10
-------
§
A r ct i
c O c e a n * - V-;-.'- *
tr>
OwWMiM SeMifartSM
f ff*—U.S.A.
Z
b
s
¦
Urttrn / T .
ftcrinf 1*1
Canada
GuH of AIjUj
Trough/
American V Grorci« Bat>n
Air u tun kljntfi
Paofk FjonlUnd * ^ y
Of^WA.
Vancouvrr SCT"lt
§
in
-
Coast and
-<*
Pacific Ocean
SheM
« ^ USA
_ ism Franotco
K
Northern
California ^-AmAmg&c*
Z
\ \ —
8
Kawai'fr
California k " > \
Bight
J^whilu
M.»ud^rr4
Corterian
z
b
m
«r>
i j
1 1 1
180Of 160£TW 1400W 120 CW 100 0W
Figure 1-1. Study area and MEOW ecoregions comprising the Northeast Pacific and U.S. Arctic.
The U.S. Arctic consists of the Beaufort Sea - Continental Coast and Shelf, Chukchi Sea, and
Eastern Bering Sea ecoregions. The remaining ecoregions constitute the Northeast Pacific Region,
with the Aleutian Islands through the Northern California ecoregions making up the Cold Temperate
Northeast Pacific Province and the Southern California Bight, Magdalena Transition, and Cortezian
ecoregions making up the Warm Temperate Northeast Pacific Province. Hawaii is not part of the
Northeast Pacific and is not assessed as part of this effort.
1.5 Algorithm-Based Approach to Climate Risk Assessment
Previous trait-based assessments of climate change on marine and freshwater organisms have
used an expert opinion, or expert elicitation, approach (e.g., Foden et al., 2013; Moyle et al.,
2013; Hare et al., 2016). Such expert opinion approaches are useful when there is very limited
information and where there are no suitable models. An example is the evaluation of the
combined effects of melting of the Greenland and Antarctic ice sheets, reorganization of the
Atlantic Meridional Overturning Circulation, shift to a more permanent El Nino regime, and
dieback of the Amazon rainforest on climate change (Kriegler et al., 2009). Although there was
large uncertainty among the experts, the process was able to provide approximate bounds for
n
-------
triggering events. As pointed out by Drescher et al. (2013; also see Sutherland, 2006), expert
elicitations addressing ecological issues have gained momentum during the last two decades.
There are, however, a number of disadvantages with expert elicitations. Experts are subject to at
least nine types of cognitive biases, ranging from overconfidence to "motivational biases when
opinions are influenced for personal or research reasons" (O'Leary et al., 2008; also see Kynn,
2008). Other sources of uncertainty relate to issues such as linguistic differences in the
understanding the specific meaning of terms and "translation confusion" in translating a response
from one scale to another (e.g., categorical to numerical probabilities) (Kuhnert et al., 2009;
Drescher et al., 2013). These unintentional biases can be mitigated through the use of carefully
crafted expert elicitations such as the Delphi procedure (Rowe and Wright, 2011; Drescher et al.,
2013) or the procedure detailed in the NOAA climate change risk analysis (Morrison et al.,
2015). However, while such detailed procedures can mitigate the effects of these biases, they
cannot eliminate them since such biases are often not obvious (Kuhnert et al., 2009) and thus are
difficult to control.
There is also the question of the accuracy of expert elicitations compared to algorithm-based
predictions. In a meta-analysis covering 136 research studies comparing expert opinion versus
automated predictions, the experts were clearly better in only eight cases (Grove et al., 2000). A
number of other studies have found that automated methods outperformed experts in predicting a
variety of endpoints including medical diagnoses, psychiatric diagnoses of criminal behavior, job
and school performance, and eradication of aquatic nonindigenous species (Dawes et al., 1989;
Grove et al., 2000; iEgisdottir et al., 2006; Kuhnert et al., 2009; Kuncel et al., 2013; Drolet et al.
2015). In some cases, the improvements over the experts were slight but in other cases the
automated predictions were substantially better. In predicting job performance, the mechanical
and holistic data combination methods displayed a 50% improvement compared to experts
(Kuncel et al., 2013).
These results led McAfee (2013) in an article on the Harvard Business Review site titled "Big
Data's Biggest Challenge? Convincing People NOT to Trust Their Judgement" to state, "The
practical conclusion is that we should turn many of our decisions, predictions, diagnoses, and
judgments—both the trivial and the consequential—over to the algorithms. There's just no
controversy any more about whether doing so will give us better results."
Because of these limitations with expert solicitation, our objective was to design an algorithm-
based (rule-based) approach to assessing risk. Specifically, our objective was to create a "turn-
key" web-based tool by removing expert opinion from the final risk calculations. In designing
our framework, we recognized that expert opinion could be used in three different phases of the
risk analysis.
12
-------
• Phase 1: Synthesizing biotic and environmental traits and estimating values for
parameters with incomplete data (e.g., relative abundances, depth preferences).
• Phase 2: Generation of climate effects thresholds and rule sets used to predict risk.
• Phase 3: Calculation of climate risk from biotic and environmental traits, effects
thresholds and rules.
In Phase 1 and Phase 2, we utilized extensive literature review and synthesis as well as expert
opinion from several workshops with regional/national experts. However, by design, the experts
were not questioned about the potential climate risks to any particular species or the geographical
patterns of risk. Also, in most cases different experts were questioned regarding biotic traits and
rule generation, thus separating these two phases of the analysis. In Phase 3, CBRAT was
designed so that no expert intervention was required to calculate the risk from the synthesized
information and rule sets. That is, the calculation of the individual climate risks is independent of
user inputs and decisions other than to decide on a particular climate scenario. Additionally, we
note that with hundreds to thousands of species analyzed across 12 ecoregions and the
complexity of the rules, in most cases the calculated risk values were not apparent when
synthesizing the information or generating the rules.
Besides circumventing the limitations of expert elicitations, algorithm-based approaches coupled
with a web-based knowledge base offer a number of advantages:
• Transparency in the biotic trait values, climate exposure values, and effect thresholds
used in the analysis for each species in each ecoregion.
o Use of expert opinion in assessing biotic traits is documented in the comments
associated with each species in CBRAT.
• Transparency in the logic and rules used to calculate risk.
o The data and rules used to generate every risk estimate are explicitly defined,
o Clarity of the rules allows future improvements.
• Application of the rules to generate the risks is unbiased.
• Consistency in predictions compared to predictions made by multiple experts over time
or made by different groups of experts.
o Straightforward to evaluate multiple geographic areas once biotic data on species are
collected without need for different sets of regional experts.
13
-------
o Straightforward to conduct scenario modelling to evaluate range of risks associated
with different emission scenarios and uncertainty analysis to evaluate uncertainty
associated with the effects thresholds.
o Straightforward for managers/researchers to evaluate different scenarios via
unsupervised risk assessments.
• Linked database and rule sets capture institutional knowledge so it is not necessary to
start anew each time a new taxon or location is evaluated or to assess different climate
scenarios.
• Simple to incorporate new biological data or climate values as they become available.
• Practical to modify or generate new rules as new knowledge becomes available without
the need to reassemble groups of experts, though it does require new programming.
• Synthesized biotic and climate data are potentially useful in addressing other research and
management issues.
• Web-based systems can be used as an outreach tool for the informed public.
In addition to these advantages, it is tempting to argue that algorithm-based predictions for
climate risks are more accurate based on the studies mentioned above. However, expert
elicitation for climate and the algorithm-based approach in CBRAT are too new to compare
accuracy, so this is an open question. Regardless, we suggest that algorithm-based systems will
prove to be the more accurate approach as they are further developed and tested.
As with any approach, there are also some limitations to algorithm-based approaches. The two
main disadvantages we found while implementing CBRAT:
• Initially it is more time consuming to create the knowledge base, corresponding rule sets,
and web interface for an algorithm-based risk analysis than soliciting a panel of experts.
• Potentially, there is a greater time lag in incorporating the most recent information
compared to gathering and soliciting a panel of experts.
Another possible advantage of expert solicitation is the ability to establish levels of uncertainty
using self-estimated levels of confidence by the respondents or from estimates from other
experts. However, McBride et al. (2012) showed that there was no consistent relationship
between expert performance in predicting the outcome of scientific experiments and the expert's
publication record, years of experience, or self-assessment of expertise. This result combined
with the documented overconfidence of most experts suggests there may be substantial
14
-------
uncertainty around expert-generated confidence estimates. In any case, a strategy to assessing
uncertainty with CBRAT is discussed in Section 8.
15
-------
Section 2.
Risk Categories and Overall Vulnerability
2.1 Individual Risk Values
The present assessment uses a categorical approach to assigning climate risks. Other regional
studies using a categorical approach to assessing environmental quality include the EPA's
Environmental Monitoring and Assessment Program (EMAP) (e.g., Nelson et al., 2007, 2008)
and National Aquatic Resource Surveys (NARS) (U.S. EPA, 2015;
https://www.epa.gov/national-aquatic-resoiirce-siirveYs/ncca). A categorical approach has also
been used in several climate risk assessments (e.g., Morrison et al., 2015). In the current
framework, the risk associated with each individual trait or climate stressor is assigned to one of
four classes: minor, low, moderate, or high risk, with corresponding numerical values of 0 to -3.
Risks are scored on a negative basis because of the inclusion of resilience traits (Section 2.2) that
are scored with positive values. The negative scores can be considered a measure of the species'
population viability. "Minor risk" is used instead of "no risk" to acknowledge the uncertainty in
the predictions. Traits and stressors that do not apply to a particular species are assigned a null
value. Species missing critical trait information for a particular rule set are also assigned a null
value. For example, it is not possible to assign a sea level rise risk if the depth range of a species
is unknown. Null values do not affect the overall risk score for a species (Section 2.3).
The climate risk factors inherently incorporate both exposure and sensitivity attributes. Projected
changes in temperature (Section 5), pH/aragonite saturation state (Section 6), and sea level rise
(Section 7) constitute the exposure component. The sensitivity component is formalized via the
effects thresholds that associate minor to high risks to specific values of temperature, sea level
rise and pH/aragonite saturation state. Overlaying the thresholds on the projected climate values
for an ecoregion generates the risks associated with each climate stressor in the ecoregion, which
may be modified by specific species traits. As discussed in Section 4, baseline/status risks
capture species' inherent sensitivities or resiliencies to climate change. While individual baseline
risks and status metrics are not directly coupled with climate change, the overall baseline/status
risk is weighted by the overall degree of climate risk. We also note that baseline/status risks
based on or modified by a species' biogeographic distribution or abundance pattern likely
incorporate indirect effects, such as trophic interactions, that affect a species' range or
abundance.
The risk levels based on the climate effects thresholds are most simply viewed as resulting from
the direct effects of a particular climate stressor on one or more life history stages. However,
determination of the temperature risks incorporates biogeographic distributions and abundances,
16
-------
and thus are likely to capture indirect effects operating at an ecoregion scale (Section 5). Another
issue is possibility of interactions among climate stressors. As a first step in addressing such
interactions, we developed a simplified approach to approximate the interaction between risks
associated with ocean acidification and elevated temperatures (Section 6.4).
The risk classifications are a relative ranking to help identify the species most and least
vulnerable to climate change. As the risk level increases from minor to high, the likelihood of an
adverse impact increases, severity of impacts increases, number of different types of impacts
increases, as does the ability to detect such changes (Figure 2-1). A wide range of different
effects have been reported as a result of climate change (Table 2-1) and the specific effects will
depend upon the severity of the change, specific type of climate change, and attributes of the
species. For example, ocean acidification is likely to result in a suite of physiological effects
whereas sea level rise may result in increased population fragmentation through habitat loss but
is not likely to result in major physiological changes. Though these risk levels are not meant to
predict specific impacts, in many cases population declines are likely, especially at high risk
levels. While in the worst cases, extirpation from an ecoregion is possible, a high risk should not
be interpreted as implying regional extinction. Further, the risks are based on a long-term
response to climate change and it is likely that population size and other indicators of ecological
condition will show increased fluctuations with a changing environment, potentially including
periods of positive growth. Finally, we note that while not an adverse impact, the likelihood of
detecting physiological or ecological changes increases as the risk level increases, though the
ability to detect changes in abundance will largely be limited to more abundant species because
of number of samples needed to detect changes in rare species.
17
-------
Minor Risk
Low Risk Moderate Risk High Risk
A
Likelihood of Population Reduction
Likelihood of Other Ecologically Significant Impacts
Severity of Ecologically Significant Impacts
Number of Ecologically Significant Impacts
r
Figure 2-1. Conceptual view of increased likelihood of adverse impacts with increasing risk level.
Schematic of relationship between risk classes and the likelihood of adverse impacts on near-coastal
species. Not every type of impact is necessarily expected in all cases (e.g., an increase in severity may
not be accompanied by an increase in the number of impacts). The most likely types of effects are listed
in Table 2-1.
-------
Table 2-1. Potential climate change effects on individuals and populations within an ecoregion.
"Major climate stressors" are the climate stressors most often associated with a particular type of effect.
Effects are limited to those that occur within a single ecoregion, and the table does not include changes at
a biogeographical scale such as range contractions or expansions.
Response
Major Climate Stressor(s)
Examples
Ecoregion Population
Population decline
All
Laffoley and Baxter, 2016
Regional extirpation
All
Glynn, 2012; Maclean and
Wilson, 2012
Increased population fragmentation
Sea level rise, Temperature
Chu-Agor et al., 2012;
Hubbard et al., 2014
Increased population fluctuations
All
McLaughlin et al., 2002
Increased susceptibility to "events"
All
Wethey et al., 2011
Decreased genetic variability
All
Balint et al., 2011
Deepening of species to cooler waters
Temperature
Dulvy et al., 2008
Ecosystem Functions and Services
Population falls below level to provide ecosystem
functions
All
Bulling et al., 2010
Population falls below economically viable level
All
Sumaila et al., 2011; Branch
et al., 2013
Biochemical and Physiological
Increased susceptibility to disease
Temperature
(elevated and/or fluctuation)
Eisenlord et al., 2016; Kohl et
al., 2016
Change in calcification rate
Ocean acidification
Chan and Connolly, 2013;
Waldbusser et al., 2016
Altered individual growth rate
Temperature &
Ocean acidification
Thresher et al., 2007;
Sheridan and Bickford, 2011
Reproductive output
(fecundity, # reproductive events, hatching rate)
Temperature &
Ocean acidification
Lawrence and Soame, 2004;
Swiney et al., 2016
Reduction in maximum body size
Temperature &
Ocean acidification
Sheridan and Bickford, 2011;
Cheung et al., 2012
Physiological/biochemical alterations
Temperature & Ocean
acidification
Helmuth et al., 2010;
Hofmann and Todgham,
2010; Hanset al., 2014
2.2 Resilience Traits
Species may possess traits that provide an ability to cope with climate change and/or to recover
after being impacted by a climate event. Some authors separate these into resistance and
resiliency (e.g., McKinney, 1997), respectively, but we follow Bernhardt and Leslie (2013) and
use resilience as a general term encapsulating both. We identified seven baseline traits that
potentially indicate increased resilience to climate change in near-coastal species (see Table 4-1).
These factors are assigned values of low (+1), moderate (+2) or high resilience (+3).
19
-------
Resilience factors are not incorporated into the calculation of overall risk primarily because in
most cases it is unclear whether the mechanisms conveying resilience directly offset the risks.
For example, being a nonindigenous species someplace (Section 4.2.8) and the ability to colonize
small island ecoregions (Section 4.2.7) both indicate good colonizing ability, which should help
a population recover from a disruption. However, it is not clear how this colonizing ability would
offset the physiological impacts of ocean acidification or the habitat loss resulting from sea level
rise within a specific ecoregion. In some cases, there is a spatial mismatch between the resilience
traits and the risk factors. Species having a wide distribution may insulate the species from
global extinction (Section 4.2.5) but it is not apparent how this trait confers protection within an
ecoregion. In these cases, inclusion of the resilience factors in calculating the overall risk within
an ecoregion could incorrectly reduce the threat to the species. Though not currently used,
identification of these resilience traits allows them to be incorporated if future research identifies
scientifically-sound principles on how resilience traits offset specific risk factors.
2.3 Overall Risk - "One Way to Live, A Thousand Ways to Die"
As detailed in the remainder of this document, over 30 different risk and resilience values are
independently evaluated for each species in each occupied ecoregion. We attempted to
standardize the risks across different traits and climate stressors. The goal is that a high risk for
sea level rise would be approximately equivalent to a high risk for temperature in the sense that
in both cases that there would be ecologically significant impacts on population viability (Figure
2-1). While such equivalence is extremely difficult to calibrate or to demonstrate, the
temperature and SLR risks both evaluate changes to population viability. For ocean acidification,
however, much of the data currently available to generate risks are not directly related to
population viability (see Section 6.3.3). Even with this difference among the climate stressors,
we posit that there is at least a general correspondence among the risk classes calculated via
different rules. Because of this general correspondence, our approach to assigning an overall risk
for a species within an ecoregion is to take the single highest risk; the calculation of the overall
risk is not increased if there are multiple risk factors with the same value. Assuming that a
moderate risk is the greatest risk, multiple moderate risk values do not result in a high risk,
though multiple values can increase the confidence that the species is at some degree of risk.
Similarly, a single high risk factor is sufficient to assign an overall high vulnerability to the
species.
Basing the overall vulnerability on the greatest risk factor has a long history, dating back to the
nineteenth century with Liebig's law of the minimum, where a species is impacted by the single
most limiting constraint (see Jones et al., 2006). More recently, the same general approach has
been used in EPA's regional-scale monitoring programs that use multi-metric indicators of
ecological condition, with the component condition set to poor if any of the individual indicators
are poor (e.g., U.S. EPA, 2015). It is also conceptually similar to the multiplicative versus
additive approach of assessing habitat suitability in nonparametric multiplicative regressions
20
-------
(HyperNiche™; McCune, 2011). As pointed out by McCune, "if any one factor is lethal then no
level of any other independent factor can compensate for it."
A number of other climate risk assessments have used some type of summation or averaging of
risks (e.g., Moyle et al., 2013; Hare et al., 2016). Of these approaches, we most strongly disagree
with averaging individual factors especially if they include both risk and resilience factors. As
pointed out by Morrison et al. (2015), averaging tends to minimize the influence of high risk
factors, thus potentially underestimating the overall risk. Averaging risk and resilience values is
especially problematic because, as mentioned above, the mechanism or scale of the resilience
factor may be different than the risk factors impinging on the individuals or populations within
an ecoregion.
A stronger case can be made for basing the overall vulnerability on some type of summation of
the number of risk factors (e.g., Morrison et al., 2015; Hare et al., 2016). The summation of risk
factors is analogous to the ranking of the ecological condition of sites for water quality, sediment
quality, and fish tissue contamination by the number of component indicators rated as good, fair,
or poor (U.S. EPA, 2015). Nonetheless, we did not pursue this approach for the following
reasons:
Rules were developed with the objective of approximate equivalence in terms of risk.
Thus, a single high risk is sufficient to identify the species is at peril and reducing the
overall vulnerability based on lower risk factors would underestimate the threat.
We are unaware of any a priori ecological justification for choosing any particular
methodology to sum different levels of risks.
The results are process dependent, with the overall risk dependent upon the number of
risk factors incorporated into the analysis and how the different risk levels are weighted.
Certain baseline/status risk factors or resilience factors may be expressions of the same,
or similar, attributes (e.g., colonization ability), and summing them could double account
for these attributes.
The general lack of understanding of the interactions among different risk and resilience
factors argues against combining them into a single risk score.
While we contend that using the single greatest risk is the most scientifically defensible
approach, the vulnerability output from CBRAT (Appendix B) is designed to allow users to
evaluate different approaches such as summing the number of risks.
21
-------
Section 3.
Relative Abundance Estimates
3.1 Background
Abundance data provide more insights into potential vulnerabilities than are provided by
distributions alone. Distributions identify where a species can survive while abundances help
elucidate preferred and marginal environmental conditions; and several studies have indicated
that rare species are more susceptible to disturbances (e.g., Davies et al., 2000; Duncan and
Young, 2000; Davies et al., 2004). However, because the majority of species in a community or
a taxon are rare (e.g., Gray et al., 2005) we do not use rarity alone as an indicator of climate risk,
with the exception of Hyper-rare species (Section 3.3). Rather, relative abundance is used as a
modifier to the baseline rules predicting climate risk, as discussed in the next chapter.
As far as we are aware, estimating regional relative abundances for thousands of marine species
for an entire coast has not been previously attempted. Thus, we had to develop approaches to
generate such information. This section details: 1) an overview of some of the factors affecting
abundance estimates; 2) a hierarchical abundance schema for classifying relative abundances;
and 3) a hybrid approach to generating relative abundance estimates from a mix of quantitative
and qualitative information.
3.2 Abundant or Rare?
Everyone knows what abundant and rare species are, but quantifying the concept proves elusive.
A number of researchers, in particular Gaston and his colleagues (e.g., Gaston, 1994, 1997,
2011; Blackburn and Gaston, 1997) and conservation biologists (e.g., Hartley and Kunin, 2003;
Flather and Sieg, 2007; Marcot and Molina, 2007; Hercos et al., 2012), have tackled defining
population abundances. Nonetheless, no generally agreed upon set of definitions have emerged,
largely because of the complexities associated with defining different types of abundance across
multiple spatial scales and taxa. The major complexities include:
Scale Dependency: Assessment of species' abundances is strongly influenced by the
spatial scale evaluated, from that of a single sample (point scale) to global population
estimates. Species may be abundant at a small scale in a particular locality but rare at a
regional scale.
Temporal Dependency: Most near-coastal species show seasonal patterns in abundance,
and many are subject to strong short-term stochastic variations. Thus, when and over how
long a population is sampled will affect its estimated abundance, with longer timeframes
smoothing out the seasonal and short-term fluctuations.
22
-------
Absolute Abundances: Absolute estimates of abundance can be expressed as a density,
total number of individuals within some habitat or region, areal extent of the population
(e.g., marshes and corals), or as the frequency of occurrence in a set of samples. Absolute
abundance estimates are very sensitive to differences in sampling gear and sampling
design.
Relative Abundances: Relative abundances are abundances normalized to some measure
of the abundances of the other species in the taxon or guild. Relative abundances are
often used when comparing studies with different sampling techniques or when
comparing species across taxa or habitats. Values of relative abundances depend upon
what taxon or guild is used to relativize the abundances. In general, relative abundances
are larger the narrower the taxon or guild used to relativize the abundances. For example,
a rockfish's relative abundance will likely be greater if its abundance is relative only to
other rockfish rather than to all bottom fish.
Given that our spatial domain is the MEOW ecoregions, which includes multiple habitat types,
assessing absolute abundances for hundreds of common and rare species is essentially
impossible. It is possible, though challenging, to assess relative abundances within an ecoregion
by integrating different types of information, including habitat areas. Our approach is to assess
relative abundances within major taxonomic units for species either associated with the bottom
or the water column. For example, the relative abundances of brachyuran crabs are determined
by comparing them to other bottom-associated decapods. The rockfish are evaluated compared to
other bottom-associated fishes, and not to water-column species. Operationally, these bottom-
associated species are those normally captured in bottom trawls and grabs, while water-column
associated species are captured in mid-water or surface trawls.
It is important to emphasize that we generate relative abundance for the entire ecoregion and not
by habitat. Thus, the relative abundance of crabs in the rocky intertidal are compared to all
offshore crabs. The only ecosystem split is that offshore and estuarine species are evaluated
independently. To the extent practical, the relative abundances of estuarine crabs are relative to
other estuarine crabs while offshore crabs are compared to offshore species. Final estimates are
based on the adult stage.
3.3 Hierarchical Relative Abundance Classification Schema
To compare abundance estimates derived from quantitative surveys with those from natural
history texts, we needed a classification schema to systematize the abundance estimates. An
example of an early sample-scale schema is the ACFOR system (abundant, common, frequent,
occasional, or rare; Crisp and Southward, 1958) to describe invertebrate abundances in the rocky
intertidal. This system was then replaced with "ESACFOR" (Hawkins and Jones, 1992) that had
seven classes (extremely abundant, superabundant, abundant, common, frequent, occasional, or
rare). However, in evaluating such sample-scale schemas it became evident that they did not
23
-------
accommodate classifications at different levels of information, a common occurrence when
assessing multiple species at a regional scale. Additionally, these schemas, as well as others,
mixed abundance with frequency of occurrence. While frequency is related to a species'
abundance, we posit that abundance classes should not be defined by frequency of occurrence
since the ecological factors resulting in high/low frequency may be different than those resulting
in high/low population abundance.
To address the need for a regional-scale schema, we developed a three-tiered relative abundance
schema (Figure 3-1) that is flexible enough to accommodate both quantitative and qualitative
information, with the level of resolution determined by data availability. Level I classifications
include Present, Not Reported, Error/Extinct, and Transient. Not Reported indicates that there are
no records for the species in the ecoregion, and is the default, while Present indicates that there
are valid records. Species designated as Present presumably have reproducing populations and
are considered established. Present and Not Reported form the basis for describing species'
biogeographic distributions (i.e., "presence/absence").
Two additional Level I classifications are necessary to address ecological and taxonomic
complexities. The first are Transients, species that temporarily occur in an ecoregion due to
climatic or oceanographic events, and are further discussed in Section 4.3.6. The second
classification, Error/Extinct captures species that have been reported to occur in an ecoregion but
do not actually occur either because they were incorrectly reported or went extinct in the
ecoregion. Note that Error/Extinct was previously referred to as Absent (Lee et al., 2015).
Incorrect attribution can be due to incorrect taxonomy, taxonomic revisions, mislabeling of
samples, or incorrectly extending a species range. In our earlier assessment of coastal invaders in
the North Pacific (Lee and Reusser, 2012), we found that such "problem children" occur with
annoying regularity. One metric to evaluate the number of occurrences is the 'species X
ecoregion' occurrences where, for example, Metacarcinus magister in the Northern California
Ecoregion is one species X ecoregion occurrence. Of the 840 species x ecoregion occurrences of
brachyuran crabs in the NEP and U.S. Arctic (see www.cbrat.ore). 92 were classified as
Error/Extinct. Most of these were due to incorrect attributions. Species are also classified as
Error/Extinct when they previously occurred in the ecoregion but went extinct, such as
nonindigenous species that were introduced but never became established or native species
impacted by habitat loss or over exploitation (Dulvy et al., 2003). The Population Trends map in
CBRAT identifies whether a species is a mistake or extinct (Lee et al., 2015).
Level II is an assessment of a species' general relative abundance (Abundant, Moderate, or
Rare). Level III allows more detailed information to be captured, providing a higher resolution
classification of a species' abundance. Basically, each of the Level II classes is divided into two
subclasses at Level III plus one additional Hyper-rare class. We define Hyper-rare species as
those that have not been reported for >50 years within an ecoregion taking into account whether
24
-------
there has been at least a minimal sampling effort. The 50-year criterion is in accordance with
typical 53-year lag between the last sighting of a species and the reporting of a species'
extinction (Dulvy et al„ 2003).
As detailed below, a "hybrid" approach combining multiple lines of evidence is used to assign
relative abundances. Because many species only have limited information, Level II rather than
Level III relative abundances are used to modify the baseline rules (Section 4). As additional
information on coastal species becomes available, it should be possible to generate Level III
relative abundances for many if not most species, which in turn may provide higher resolution
risk predictions.
Level I
Present
Level II
Level III
Hyper-Rare
Error/Extinct
Not reported
(default)
T ransient
(Not Established)
r
Low Moderate
Moderate
^ J
Abundant
M odersrtel y Abun dant
Very Abundant
Figure 3-1. Three-level relative abundance schema for use at regional scales.
Level I is the basis for describing biogeographic distributions of species. Level II
describes the geographical pattern of relative abundance of a species. Level III
describes the relative abundance pattern with a greater resolution. Hyper-rare
species have not been observed in >50 years assuming at least a minimal
sampling effort.
3.4 Importance of Habitat Area
At an ecoregion scale, the total population size of a species is determined by the sum of its
abundance across all the habitats within the ecoregion. This can be expressed as:
Eq. 1: Total population in ecoregion = ^i"_l(Habitatl X Densityi)
As illustrated by this heuristic formula, the size of a species' population at a regional scale is as
dependent upon the total area of the occupied habitats as on the densities. Species that are
abundant in only a single habitat of limited extent would have a low relative abundance when
25
-------
averaged over the entire ecoregion. Conversely, a species that is moderately abundant in a very
wide-spread habitat might be ranked relatively abundant compared to the other species within the
ecoregion.
To illustrate the importance of habitat area, the total population size for a species is calculated as
the product of habitat area times density for an approximate 1000-fold range in both area and
density (Table 3-1). Using the simplifying assumption that a species only occupies a single cell,
the relative abundance of each species (cell) is classified using the hierarchical abundance
schema (Section 3.3) based on dominance normalized relative abundances (Section 3.5.1). The
point of this exercise is to illustrate that most area and density combinations result in a
classification of Rare. Of the 121 'area X density' combinations, 66 are classified as Very Rare
and 10 as Moderately Rare. In comparison, only 15 combinations are classified as Moderately
Abundant or Very Abundant. Even at the highest density, species are Rare if they occupy a
habitat of limited extent and are classified as Abundant only if the species occupies a habitat of
at least moderate spatial extent. The bottom line is that areas of the occupied habitats are as
important in determining a species' total abundance at an ecoregion scale as the more commonly
reported densities.
26
-------
Table 3-1. Hypothetical example of species' abundances at an ecoregion scale.
Each species is represented by a single cell, with its abundance calculated as the product of the area of the
habitat times the density in that habitat. Abundances are classified according to the hierarchical abundance
schema that ranges from Very Rare to Very Abundant (Section 3.3) based on their dominance normalized
relative abundance (Section 3.5.1) in each cell. There is a total of 121 species (cells) and a total of 4,190,209
individuals summed across all species. Both the median density and the median habitat area is 32; there are
no Abundant species at the median density regardless of habitat area or Abundant species at median habitat
area regardless of density.
Species Density
1
2
4
8
16
32
64
128
256
512
1024
re
•*->
n
re
1
1
2
4
8
16
32
64
128
256
512
1024
2
2
4
8
16
32
64
128
256
512
1024
2048
4
4
8
16
32
64
128
256
512
1024
4096
8
8
16
32
64
128
256
512
1024
8192
16
16
32
64
128
256
512
1024
16384
X
M-
o
re
£
<
32
32
64
128
256
512
1024
32768
64
64
1024
2048
512
1024
65536
128
128
256
512
1024
131072
256
256
512
1024
131072
262144
512
512
1024
131072
262144
524288
1024
1024
131072
262144
524288
1048576
Key
Very
Rare
Moderately
Rare
Low
Moderate
High
Moderate
Moderately
Abundant
Very
Abundant
To provide guidance to the areas of different habitats, the total areas of the major estuarine and
offshore habitats were determined using georeferenced marine/estuarine landscape data for
Oregon, Washington, and California. As detailed in Appendix C, data sources included the
National Wetland Inventory (NWI) and various offshore surveys. The major patterns described
in Appendix C are:
• Estuarine habitat area is substantially less than offshore area to a depth of 200 m.
• Intertidal and subtidal unconsolidated sediment combined are the major estuarine
habitats.
27
-------
• Submerged aquatic vegetation (SAV) constitutes a relatively small percentage of the
estuarine area from Puget Sound through Northern California, and a somewhat greater
percentage in Southern California.
• Emergent marshes are moderately abundant in Oregon and Northern California, and
relatively less abundant in Puget Sound and Southern California.
• Unconsolidated sediments are the major offshore habitat type from Oregon through
Southern California. Rocks and boulders constitute the second largest offshore habitat but
a minor habitat in estuaries.
• Kelp constitutes a relatively small proportion of the total area in Puget Sound through
Southern California.
For other ecoregions, we assumed generally similar patterns of major habitats (e.g.,
predominance of unconsolidated sediments and relatively small estuarine area. Two specialized
habitats, mangroves and corals, do not occur in the analyzed ecoregions. Mangroves appear to be
moderately abundant at least in certain areas in the Magdalena Transition and Cortezian
ecoregions (Glenn et al., 2006; Spalding et al., 2010). In comparison, isolated coral patches and
coral reefs only occur in the Cortezian Ecoregion, where they constitute a minor area.
Though not used in a formal algorithm, these patterns of habitat area were considered when
combining multiple sources of information in the "hybrid approach" to estimate relative
abundance (Section 3.8).
3.5 Data Sources - Quantitative Data
Quantitative biotic studies are a key information source to estimate relative abundances, but the
challenge is that these studies vary greatly in scale, from a single restricted location to regional
surveys. They also vary in sampling design, from fixed sites to randomized surveys, and in
sampling gear. This section describes how we mitigate the effects of these differences by
normalizing abundances and setting thresholds to convert these normalized abundances to the
relative abundance classes described in Section 3.3.
3.5.1 Dominance Normalized Relative Abundance (DNRA)
When quantitative data are available, the question becomes how to compare abundances across
different studies. Our first attempt to compare relative abundances across studies was to
normalize each species' abundance to the mean abundance of all the species of the target taxon
in the sample set. While normalization of abundances to the mean is intuitive, it has the
limitation that with increasing sample size new rare species are added to the species set relatively
faster than the total number of individuals increases. This results in the average abundance of all
28
-------
species decreasing since the denominator (number of species) increases faster than the numerator
(sum of individuals), which in turn affects the relative abundances of the species. Use of the
median abundance does not resolve this issue as it is even more sensitive to increasing sample
size and inclusion of new rare species.
To minimize the problem of sample size dependency, we normalized individual species
abundances to the mean abundances of the "numerical dominants," defined as those species
constituting >75% of the individuals. Use of 75% criterion is derived from Swartz's dominance
index (Swartz et al., 1986), which is the minimum number of species required to account for
75%) of the total individuals. Advantages of normalizing abundances to the numerical dominants
are that, in most cases, their mean abundance stabilizes with a moderate number of samples and
their mean abundance does not change monotonically with increasing sample size.
Consequently, the relative abundances of non-dominant species do not change systematically as
the number of rare species increases with increased sampling.
Dominance normalized relative abundance (DNRA) is calculated from a quantitative sample set
according to the following procedure:
1. Determine the relative abundance (%>) of all species within the target taxon in the
total collection.
2. Determine the species that make up 75% of the total individuals of the taxon.
3. Calculate the average abundance of the species constituting 75% of the
individuals. This is inclusive, so if the cumulative percentage of the first 4 species
is 74.9%) take the average abundance of the first 5 species.
4. Divide the abundance of each species by the average abundance of the
numerically dominant species. This value is the "dominance normalized relative
abundance" for that species.
Because of limitations in capturing the Very Rare species at the lower end of the abundance
range (see Table 3-2), dominance normalized relative abundances should be calculated with
sample sets with >1000 total individuals and preferably >5,000 individuals; otherwise the
approach may fail to identify the Very Rare species.
3.5.2 Quantitative Cut Points for Abundance Classes
After calculating dominance normalized relative abundances for a taxon, the next step is to
partition the species into the classes used in the hierarchical relative abundance schema, which
requires generating quantitative cutpoints. As pointed out by Gaston (1994, 1997), there is no
general theory to establish cutpoints for rare versus abundant species. While theory does not
offer any easy answers, it does provide guidelines. In particular, there is strong theoretical and
29
-------
empirical evidence that most species in assemblages are rare or moderate and only a few species
within a taxon are abundant (e.g., Gray et al., 2005).
Using previously described species abundance distribution patterns as a guide, we evaluated
various cutpoints within single datasets using quantitative abundances from two large databases.
The first was the bottom trawl data from NOAA's RACE program (www.afsc.noaa. gov/R ACE/)
from which we summarized the bottom trawl data from 1977 to 2006 by MEOW ecoregion in an
Access database (USGS_EPA RACE 1977-2006, 2013). The other large dataset was an Access
database of EPA's EMAP benthic surveys on the West Coast combined with other benthic
surveys (U.S. EPA, 2008). With the RACE data, we analyzed the relative abundance of all
bottom fish in the Eastern Bering Sea, while with the benthic data we analyzed the bivalves in
Puget Sound through Southern California. Additionally, we evaluated the breakout of Rare to
Abundant species at an ecoregion scale with the brachyuran crabs.
The maximum value for a dominance normalized relative abundance is around 5 (i.e., a
dominant species is 5-times as abundant as the average abundance of all the numerical
dominants). Such high values tend to occur in extreme environments where a few species
dominate the fauna. At the opposite extreme, the lowest value, 0.00000015, was derived for
bottom fish in the Bering Sea based on the extensive RACE dataset. In most cases, however, the
lower value is on the order of 0.001 to 0.00001 in a large dataset. After exploring the behavior of
different cutpoints with these datasets, we finalized the values for Level II and Level III relative
abundances given in Table 3-2. These cut points give ecologically realistic percentages of Rare
to Abundant species from the quantitative survey data in the sense that few species are classified
as Abundant and many as Rare as well as capturing the difference between high and low
diversity regions.
30
-------
Table 3-2. Definitions and quantitative cut points for the three-level relative abundance classifications.
Dominance normalized relative abundances (DNRA) are the values used with quantitative studies to assign relative abundance classes. Phrases
commonly used in relation to the abundance class both in terms of abundance and frequency of occurrence are provided as a guide. Most of the
phrases are not unique to a single level of abundance class; interpretation of these terms needs to be taken in context of scope and spatial scale
of the study. The approximate range of the number of species in each abundance class is based on our analyses of the relative ecoregional
abundances of brachyuran crabs and bivalves in the Southern California Bight, Northern California, Oregon, Washington, Vancouver Coast and
Shelf, and Puget Trough/Georgia Basin ecoregions. (Table modified from Lee et al., 2015). NA = Not applicable.
Abundance
Class
Qualitative Description
Common Key
Phrases -
Abundance
Common Key
Phrases -
Frequency of
Occurrence
Dominance
Normalized
Relative
Abundance
Outpoints
Approximate
Ranges of
Percentage of
Species in an
Ecoregion
Level 1
Present
Valid quantitative or qualitative records
exist for a species within an ecoregion.
Present, Observed,
Reported, Found,
Occurs
Frequency >0
>0
95 - 100%
Not Reported
There are no records known for the
species in an ecoregion. This is the
default.
No mention of the
species within the
ecoregion.
Frequency = 0
NA
NA
Error/Extinct
Species that have been incorrectly
reported as present in a region due to
incorrect taxonomy or taxonomic
revisions, or that have gone extinct within
the ecoregion.
Misidentified,
Taxonomic revision,
Extinct, Extirpated
NA
NA
0-5%
Transient
Species that temporarily occur in an
ecoregion due to unusual climatic or
oceanographic events but do not
establish a permanent population.
Transient, Extralimital,
Temporary, Migrant,
Not established,
Outside normal range
Varies
(Often low frequency)
>0
0-5%
31
-------
Abundance
Class
Qualitative Description
Common Key
Phrases -
Abundance
Common Key
Phrases -
Frequency of
Occurrence
Dominance
Normalized
Relative
Abundance
Cutpoints
Approximate
Ranges of
Percentage of
Species in an
Ecoregion
Level II
Abundant
Numerous and usually observed in
collections in suitable habitat(s). Often
inhabit a habitat of wide spatial extent
and/or multiple habitats.
Abundant, Common,
Plentiful
Widespread,
Frequently observed,
High rate of capture
>0.1
4-17%
Moderate
Includes both species that are abundant
in habitats of small to moderate spatial
extent as well as species that are
regularly found at multiple sites but which
do not normally constitute a major portion
of the individuals.
Moderate, Relatively
common, Not
uncommon, Collected
in reasonable numbers
Moderate rate of
capture, Often
observed
>0.01 <0.1
18-45%
Rare
Species with low total population sizes.
Often inhabit habitats of limited spatial
extent. May be relatively abundant in a
spatially limited habitat.
Rare, Uncommon,
Specialized
Infrequently observed,
Low frequency, Rarely
observed, Low rate of
capture, Not found very
often
<0.01
39-66%
Hyper-Rare
Species that have not been observed
within an ecoregion for 50+ years, with
the caveat that there has been at least a
moderate sampling effort.
Extremely rare,
Possibly extinct
Not observed, Not
seen for over 50 years
0
0-4%
32
-------
Abundance
Class
Qualitative Description
Common Key
Phrases -
Abundance
Common Key
Phrases -
Frequency of
Occurrence
Dominance
Normalized
Relative
Abundance
Cutpoints
Approximate
Ranges of
Percentage of
Species in an
Ecoregion
Level III
Very Abundant
The most numerous species within an
ecoregion, usually inhabit a habitat of
large spatial extent and/or multiple
habitats.
Numerical dominant,
Very abundant
Ubiquitous, Very
widespread, Nearly
always collected
>0.5
2-8%
Moderately
Abundant
Abundant species within an ecoregion,
but not numerically dominant.
Abundant, Very
common
Widespread, Regularly
captured
>0.1<0.5
2-12%
High Moderate
Species frequently observed in one or
several habitats though usually not
among the most numerous species.
Common, Not
uncommon, May be
abundant in suitable
habitats
Frequent, Often
observed, Routinely
collected
> 0.03<0.1
5-25%
Low Moderate
Species that occur in high abundances in
relatively spatially limited habitats.
Common, Not
uncommon, Common
in one locality
Regularly observed,
Not infrequent
>0.01<0.03
5-25%
Moderately Rare
Uncommon species, but often observed
in low numbers in large collections. May
inhabit specialized habitats and/or
habitats of limited area. May also include
generalist species at the end of their
biogeographic range.
Rare, Sparse
Infrequent
> 0.005<0.01
8-25%
Very Rare
The least abundant species in an
ecoregion, often inhabit specialized
habitats or habitats of limited area.
Usually sparse even in suitable habitats.
Can include species at the end of their
biogeographic range.
Rare, Very rare,
Unusual, Only one
specimen found
Rarely observed,
Seldom found
<0.005
10-50%
33
-------
3.6 Data Sources - Online Biodiversity Databases
Online biodiversity informatics databases are becoming an increasingly important source of
species' information (Edgar et al., 2016). Of the current databases, the Ocean Biogeographic
Information System (OBIS; http://iobis.org/mapper/) and Global Biodiversity Information
Facility (GB1F; https://www.gbif.org/) are particularly useful. Both sites plot individual
georeferenced sample points from museum records and quantitative surveys, and can be used to
fill in a species' distribution. OBIS maps occurrence data by MEOW ecoregion, making it easier
to extract the records. These sites can also provide insights into a species' geographical pattern of
abundance. A large number of reports within an ecoregion suggests a moderate or high relative
abundance for that species. However, the converse has to be interpreted cautiously; few or no
records do not necessarily indicate rarity as the site may not have captured the pertinent surveys.
Caution also has to be exercised in comparing among species because differences in the number
of occurrences may be a function of the surveys summarized rather than real differences in
abundance.
Of the two sites, GBIF tends to have more species than OBIS, though it also tends to have more
incorrect or suspicious records in our experience (see Robertson, 2008). Another caution is that
GBIF may include fossil records that are not obvious unless the specific record is viewed. For
suspicious records (e.g., a single report of a Pacific species in the Atlantic), it is important to
backtrack the suspect records to their original sources.
3.7 Data Sources - Text-Based Information
3.7.1 Parsing Natural History Texts
"When / use a word," Humpty Dumpty said in rather a scornful tone, "it means
just what I choose it to mean - neither more nor less."
—Lewis Carroll, Through the Looking-Glass.
The reality is that quantitative data are not available for many species, especially at regional
scales. There is, however, a wealth of information from taxonomists and natural historians that
date back over a 100 years on the Pacific Coast. For rare species not reported from quantitative
surveys, taxonomic and natural history texts may be the only source of information. We initially
developed a set of about 100 key words and phrases related to abundance (e.g., "dense," "not
common," etc.) with the objective to standardize, and perhaps automate, the parsing of natural
history text. However, after a gallant effort, we abandoned this approach. Natural history texts
are too context specific to use simple parsing of key words and phrases to generate relative
abundances at an ecoregion scale.
Another issue is that natural historians tend to report species' abundances from the species'
preferred habitats where they are most abundant. Thus, many species are referred to as
34
-------
"common", "frequent" or the dreaded "not uncommon". In Hart's (1982) "Crabs and Their
Relatives of British Columbia", "common" was used in association with 17 of the 35 brachyuran
crabs and "widespread" was used in association with an additional 6 species. Based on these
descriptors, one could conclude that 23 of the 35 crabs (65.7%) in British Columbia are
abundant, a much higher percentage than can be reconciled with ecological theory. We refer to
this tendency of natural historians to assign abundances based on the most favorable habitats as
the "Panda Bear Syndrome" - there is a bamboo forest somewhere in south central China where
Panda Bears are the most abundant mammal. If this is your point of reference, you view Panda
Bears as numerical dominants.
Even with these challenges, taxonomic and natural history texts are treasure troves of
information. We offer the following guidelines in using these texts to evaluate regional relative
abundances:
Table 3-2 provides a list of the commonly used words and phrases in relation to
abundance and frequency of occurrence in relation to Level I through Level III relative
abundance classes. Most of the phrases are not unique to a single abundance class, and
need to be interpreted in the context of the scope and spatial scale of the study (e.g., local
habitat or regional scale).
Texts that describe species at a regional scale are given greater weight than texts
describing only a local area.
Give greater weight to cases where several authorities describe similar abundances, with
the caveat that the more recent authors are not repeating results from earlier authors.
Unless the text is exclusively focused on a local area, it is generally more straightforward
to extrapolate a description of rarity to a regional scale than to extrapolate a report of high
abundance to a region.
If possible, compare the text-based abundance to known quantitative estimates to help
calibrate how the author uses various words and phrases.
3.7.2 Negative Evidence: The Dog That Didn't Bark
Sherlock Holmes in "Silver Blaze" (Sir Arthur Conan Doyle, 1892) observes there is much to be
learned when something that is expected doesn't happen:
Gregory: "Is there any point to which you would wish to draw my attention?"
Holmes: "To the curious incident of the dog in the night-time."
Gregory: "The dog did nothing in the night-time."
Holmes: "That was the curious incident."
Negative evidence has a long history in ecology and evolution, at least as far back as Darwin in
1854 (Darwin, 1854). While it needs to be used judiciously, negative evidence can help identify
35
-------
rare species. If a species has been reported from an ecoregion in a taxonomic treatise or a species
checklist but is not reported from a large dataset, its absence is suggestive that the species is rare.
One caveat is that the dataset is based on surveys using appropriate sampling methods and in
appropriate locations. Another caution is that if there are only one or two reports of a species in
an ecoregion, the accuracy of the identification should be evaluated.
3.8 Hybrid Approach to Estimating Ecoregion Abundances
3.8.1 Synthesizing Multiple Data Types
Recognizing that not all data are equal or that all data types are available for all species, we
developed a hybrid approach to assigning relative abundance classes at an ecoregion scale. The
following guidelines were used in weighting different types of information:
The most useful data are those from regional scale, randomized surveys such as the
previous EMAP surveys (e.g., Nelson et al., 2004, 2007, 2008) and the current National
Aquatic Resource Surveys (NARS, U.S. EPA, 2015; https://www.epa.gov/national-
aquatic-resource-survevsY Consideration needs to be given to any potential effects of
habitat biases or any sampling gear limitations. For example, neither EMAP nor NARS
sampled the rocky intertidal. These quantitative data can be converted into dominance
normalized relative abundance values in a csv file which allows the relative abundance
classifications for all the species in the survey to be automatically mapped into CBRAT
with an accompanying PDF for documentation. This functionality saves considerable
time compared to entering relative abundances species by species when a survey contains
hundreds to thousands of species.
Regional-scale, non-randomized quantitative surveys, such as NOAA's RACE surveys
(https://www.afsc.noaa.eov/RACE/defaiilt.php). are given second priority. Besides the
limitations mentioned above, non-random surveys are subject to nonrandom sampling
and spatial biases. As with the randomized surveys, the relative abundances can be
automatically mapped into CBRAT if the data are converted to dominance normalized
relative abundances in a CSV file.
Expert opinion specifically addressing the ecoregional abundance of a taxon, such as
occurred during EPA's "extreme natural history" workshops with SCAMIT (Cadien and
Lovell, 2012; Lovell and Cadien, 2013), is usually given third priority. The expert
opinion may be given higher priority if the experts can identify limitations with the
quantitative surveys or provide more up-to-date information. A key component of
working with experts is to provide adequate background information and training to help
standardize relative abundances across experts. To the extent practical, the experts should
explain their conclusions, especially any that deviate from quantitative surveys.
Local randomized and non-randomized surveys are, in general, given fourth priority.
Their weight depends upon the scale of the study and whether there are additional local
36
-------
studies with similar results. Consideration needs to be given to the effects of strong
spatial and habitat biases on the results. As with the randomized surveys, the relative
abundances can be automatically mapped into CBRAT if the data are converted to
dominance normalized relative abundances, though in these cases the abundances only
apply to a portion of the ecoregion.
Natural history and taxonomic texts are given fourth or fifth priority. They are given
more weight when several texts give similar independent answers or when there is no or
very limited quantitative data. In addition to the eccentricities of natural history texts, an
additional challenge is standardizing across the experts interpreting the text-based
information. Working directly with experts reduces differences among individual
interpretations as does initial training.
OBIS and GBIF, and frequency of occurrence data in general, are given the lowest
priority and are generally used as a supplement to other data types. A large number of
records in an ecoregion from OBIS or GBIF suggest a moderate or high abundance;
however, absence of records does not necessarily indicate absence of the species.
Combining these various information types is not formulaic, but several guidelines are possible:
Level III abundance classes should be used if the source(s) give sufficient resolution and
there is reasonable certainty in the results. If the sources do not provide sufficient
resolution or there is uncertainty (e.g., two authoritative sources disagree), the
abundances should be classified at Level II. The CBRAT comment function should be
used to document the reasoning and sources, especially for the more problematic cases.
As mentioned, the area of the habitat potentially occupied by the species needs to be
incorporated into assigning abundance estimates at an ecoregion scale. It's been our
experience that it takes some training to have experts "scale up" their view from local
habitats to ecoregions.
Species may only occupy the very northern or southern portion of an ecoregion (e.g.,
Alijos Rocks which is located at the edge of the southern border of the Southern
California Bight ecoregion); in such cases they would be considered rare or very rare
when averaged over the entire ecoregion.
One approach to assigning relative abundances with species that have limited or no
quantitative data is to initially define a set of "anchor species" that have sufficient
information to allow assigning relative abundances with reasonable confidence. Then, by
comparing the data-poor species to various upper and lower anchor species, it is often
possible to assign a Level II or even Level III relative abundance.
One initial step is to partition the species into abundance bins based on whatever
information is available. In most cases, the abundant species will be limited to the 1st
bins while the fourth bin will consist of rare species. The 2nd bin will primarily consist of
37
-------
species with moderate relative abundances. The 3rd bin likely will be a mix of moderate
and rare species. This approach is a guide and each species should then be evaluated
independently. Nonetheless it can make evaluating a large set of species less daunting.
Our experience is that trained interpreters are generally within a single Level III class (e.g.,
moderately rare versus very rare).
3.8.2 Checking Abundance Classifications at an Ecoregion Scale
A useful check after classifying a majority of the species is to evaluate the percentages of
abundant, moderate, and rare species within each ecoregion. In general, the percentages should
follow the order Rare > Moderate » Abundant. Additionally, the ecoregion-specific percentages
can be compared to the ranges in Table 3-2. This table summarizes the percentages of
brachyuran crabs and bivalves based on our initial analysis of the four ecoregions from Puget
Sound to Southern California. Approximately 98-100% of the species in these taxa are classified
to a Level II relative abundance in each of the ecoregions. Percentages for the Level III classes
had to be extrapolated as only 31-61% of the crabs and bivalves are currently classified at this
level of resolution. The percentages in Table 3-2 are presented as general examples, but
distributions that deviate greatly should be examined. However, as noted earlier, percentages
may deviate from those in Table 3-2, in particular in stressed or extreme environments and when
there is very low diversity of a taxon. An example is the Beaufort Sea-continental coast and shelf
Ecoregion which has only three brachyuran crabs, of which two (66%) are classified as abundant
and the third as high moderate (33%) based on their numbers.
Also, the distributions should be examined if adjoining ecoregions have substantially different
breakouts. If such deviations are noted, examine whether the differences were driven by results
from different experts or by a reliance on a particular study in one ecoregion. Another possibility
is that an ecoregion has a large number of species classified only as Present, which may
artificially reduce the percentage of rare species.
38
-------
Section 4.
Baseline/Status Risks
4.1 Introduction
In their analysis of climate change effects on freshwater fish, Moyle et al. (2013) defined
baseline vulnerabilities as species traits, or indicators of such traits, that identify which species
are most vulnerable to current environmental stressors other than climate change. Biotic traits,
such as life span and spatial distributions as well as measures of population trends, have been
used frequently in conservation research (e.g., Sodhi and Ehrlich, 2010), in predicting climate
risks with freshwater, and terrestrial species (e.g., Olden et al., 2008; Galbraith and Price, 2009;
Diamond et al., 2011; Chown, 2012; U.S. EPA, 2012; Moyle et al., 2013; Pearson et al., 2014)
and in predicting climate risks with marine species (Foden et al., 2013; Hare et al., 2016).
In the current analysis, we slightly modify the concept of baseline risks. While other authors
have included population trends with baseline risks (e.g., Moyle et al., 2013), we limit baseline
risks to inherent biotic traits of species. "Status" is used to capture changes in a species' viability
due to exogenous factors such as overfishing or non-climate related habitat loss. The
combination of these two is referred to as "baseline/status risks". The second difference is that
baseline/status risks are used herein to capture increased risk under climate stress. While
endemicity is an indicator of population vulnerability under current conditions, we evaluate it as
an indicator of increased vulnerability to climate change (e.g., Malcolm et al., 2006; Loarie et al.,
2008; Morueta-Holme et al., 2010). Specifically, the risk levels associated with the
baseline/status risks identified below are the risks under increased temperatures, reduced pH,
and/or sea level rise. Section 4.5 discusses how baseline/status risks are linked to different levels
of climatic stress.
A total of 17 baseline/status traits were identified that could be applied to both well-studied and
lesser known near-coastal species, with the rules are summarized in Table 4-1 and Table 4-2.
These rules are divided into three categories: biogeographic distributions, relative abundances,
and life history traits. The rules are further classified as either global or ecoregion specific.
Global rules are those that apply the same risk for a species across all ecoregions while
ecoregion-specific rules incorporate some ecoregion-specific trait, such as relative abundance,
that modify the risk geographically. The baseline/status rules in CBRAT assign a risk value
ranging from -3 (high risk) to 3 (high resiliency) depending on whether they increase
vulnerability or increase resiliency, respectively. It is not currently possible to change the risk
levels associated with particular traits in CBRAT, but users can change the values in the
vulnerability summary spreadsheet (see Appendix B).
39
-------
In the following sections, we list the number of crabs, rockfish, and bivalve species at risk due to
the various traits. These provide an assessment of the general applicability of the rules. Rules
that identify risk in a very small number of species have limited general utility while rules that
predict high risk in most species do not have sufficient resolution to differentiate taxonomic or
geographical patterns. These risks are based on a preliminary analysis, and may change with the
formal risk analysis (Lee et al., in progress).
40
-------
Table 4-1. Baseline/status climate rules derived from biogeographic distributions, relative abundance, life history traits, and population trends.
Rules are classified as Vulnerability or Resilience depending upon whether the trait results in increased or reduced risk from climate change,
respectively. Rules are also classified by the type of trait: biogeographical distributions, regional abundance patterns, or life history traits. A rule
that applies to all ecoregions is referred as "global", while geographically specific rules are referred to as "ecoregion specific". Risk is scored from
-3 (high risk) to 3 (high resilience).
Trait
Risk /
Resilient
Type
Global or
Ecoregion
Specific
Baseline/Status Rule
Comments &
Exceptions
Endemic
Vulnerability
Distribution
Ecoregion
If species present in only one ecoregion AND NOT Abundant => -3
If species present in only one ecoregion AND Abundant => -2
If species present in more than one ecoregion => 0
If species present in only two ecoregions AND Hyper-rare in one or both
=> -3
Restricted
Distribution
Vulnerability
Distribution
Ecoregion
If species present in only two ecoregions AND Rare in both => -2
If species present in only two ecoregions AND Present or Moderate or
Abundant in one or both => -1
If species present in more than two ecoregions => 0
Do not include
ecoregions where the
species is Transient.
Wide
Distribution
Resilience
Distribution
Global
If species occurs in Arctic/Southern Ocean realm & Cold Temperate &
Warm Temperate Provinces => 2
If species occurs in Cold Temperate & Warm Temperate & Tropical
Provinces => 2
If species does not occur in three Provinces with different temperature
regimes => 0
Do not include
ecoregions where the
species is Transient or
Hyper-rare. See Table
4-5 for MEOW
provinces by
temperature regime.
Arctic Endemic
Vulnerability
Distribution
Global
If species present only in Arctic ecoregions => -2
If species present in any ecoregion outside the Arctic => 0
Do not include
ecoregions where the
species is Transient or
Hyper-rare. See Table
4-6 for Arctic
ecoregions.
41
-------
Trait
Risk /
Resilient
Type
Global or
Ecoregion
Specific
Baseline/Status Rule
Comments &
Exceptions
Small Island
Distribution
Resilience
Distribution
Global
If species occupies a "Small Island Ecoregion" => 2
If species does not occupy a "Small Island Ecoregion" => 0
Do not include
ecoregions where the
species is Transient or
Hyper-rare. See Table
4-8 for Small Island
Ecoregions.
If species has a Master NIS classification anywhere globally with a
Master Established value => 2
Nonindigenous
Species (NIS)
Resilience
Distribution
Global
If species has a Master NIS classification but establishment is Not
Established OR Unknown OR only Stocked => 0
If species does not have a Master NIS classification anywhere => 0
Hyper-Rare
Vulnerability
Abundance
Ecoregion
If species is Hyper-rare => -3
If species is not Hyper-rare => 0
Rare
Everywhere
Vulnerability
Abundance
Global
If species is Rare or Hyper-Rare in all ecoregions => -1
If species is Present, Moderate, OR Abundant in one or more
ecoregions => 0
Do not include
ecoregions where the
species is Transient.
Abundant
Somewhere
Resilience
Abundance
Global
If species is Abundant in any ecoregion => 1
If species is not Abundant in any ecoregion => 0
42
-------
Trait
Risk /
Resilient
Type
Global or
Ecoregion
Specific
Baseline/Status Rule
Comments &
Exceptions
Population
Trend
Vulnerability
Abundance
Ecoregion
If Population Trend is "No Apparent Trend" => 0
If Population Trend is "Unknown" => Null
If Population Trend is "Moderate Decrease" (-30% to -49% decline) => -
1
If Population Trend is "Substantial Decrease" (-50 to -79% decline)
AND abundance is Present, Moderate, or Abundant =>-2
If Population Trend is "Substantial Decrease" (-50 to -79% decline)
AND abundance is Rare => -3
If Population Trend is "Extreme Decline" (> -80% decline) =>-3
If Population Trend is "Moderate Increase" (30% to 49% increase) => 1
If Population Trend is "Substantial Increase" (50 to 100% increase) => 2
If Population Trend is "Major Increase" (>100% increase) OR "Order of
Magnitude" (>10-fold increase) => 3
Do not include
ecoregions where the
species is Transient.
Southern
ecoregion Rare
- Northern
ecoregion
Abundant
Vulnerability
Abundance
Ecoregion
If a Rare ecoregion abuts an Abundant ecoregion to the north AND
there are no Present, Moderate, OR Abundant ecoregions to the south
of the Rare ecoregion =>-2
All ecoregions to the south of the Rare ecoregion abutting the Abundant
ecoregion are also Rare => -2
If not one of the above cases => 0
Limited to 12
ecoregions in U.S.
Arctic and NEP, and
the ecoregions of the
Tropical East Pacific
Province.
Northern
Transient
Resilience
Abundance
Ecoregion
If a Transient ecoregion occurs to the north of an occupied ecoregion
with an abundance of Present, Rare, Moderate, or Abundant AND there
are no other occupied ecoregions to the north of the Transient
ecoregion => 3
If not a Transient to the north of an occupied ecoregion => 0
Does not apply to
Transient ecoregions
to the south of an
occupied ecoregion.
43
-------
Trait
Risk /
Resilient
Type
Global or
Ecoregion
Specific
Baseline/Status Rule
Comments &
Exceptions
Symbiotic
Specialization
Vulnerability
Life History
Global
If "Strength of Relationship" (Symbiotic) is Incidental =>0
If "Strength of Relationship" (Symbiotic) is Facultative =>-2
If "Strength of Relationship" (Symbiotic) is Obligate AND abundance is
Present, Moderate, or Abundant =>-2
If "Strength of Relationship" (Symbiotic) is Obligate AND abundance is
Rare or Hyper-Rare => -3
If no symbiotic relationship => 0
Do not include
ecoregions where the
species is Transient.
Habitat
Specialization
Vulnerability
Life History
Global
If no Specialized Habitats => 0
Vulnerable Specialized Habitats
Obligate & Preferred Habitat => -3
Facultative & Preferred Habitat => -2
Incidental & Preferred Habitat => Data error
Obligate & Observed Habitat => Data error
Facultative & Observed Habitat => -1
Incidental & Observed Habitat => 0
Resistant Specialized Habitats
Obligate & Preferred Habitat => -2
Facultative & Preferred Habitat => -1
Incidental & Preferred Habitat => Data error
Obligate & Observed Habitat => Data error
Facultative & Observed Habitat => -1
Incidental & Observed Habitat => 0
If multiple specialized
habitats, take the
greatest risk. See
Table 4-12 for
vulnerable & resistant
specialized habitats.
44
-------
Trait
Risk /
Resilient
Type
Global or
Ecoregion
Specific
Baseline/Status Rule
Comments &
Exceptions
Trophic
Specialization
Vulnerability
Life History
Global
If Specialist Trophic Specialization => -3
If Moderate Trophic Specialization => -1
If Generalist Trophic Specialization => 0
If Unknown Trophic Specialization => Null
Anadromous /
Catadromous
Vulnerability
Life History
Global
If species is anadromous or catadromous AND Rare => -3
If species is anadromous or catadromous AND Present, Moderate, OR
Abundant => -2
If species is not anadromous or catadromous => 0
Productivity
parameters
Vulnerability
& Resilience
Life History
Global
See Table 4-2.
Currently only applies
to fish
Table 4-2. Baseline/status risks derived from productivity index parameters for fish.
Thresholds for "high" (green), "moderate" (yellow), "low" (orange), and "very low productivity" (red) productivity parameters are from Musick et al.,
2000. We combine two of the productivity parameters listed by Musick et al., maturation age of females (age of first reproduction) and maximum
life span, to generate a climate risk that is modified by the species' relative abundance in the ecoregion. These rules are for fish only. Risk is
scored from -3 (high risk) to 3 (high resiliency).
Maturation - Min. Age (Female only)
Max. Life Span
Relative Abundance
Risk
<12 months
0 to 36 months
Abundant
3
<12 months
0 to 36 months
Present, Rare, Moderate
2
<12 months
37 to 102 months
All
2
12 to 48 months
0 to 36 months
All
2
12 to 48 months
37 to 102 months
All
1
12 to 48 months
103 to 360 months
All
0
49 to 120 months
37 to 102 months
All
0
45
-------
Maturation - Min. Age (Female only)
Max. Life Span
Relative Abundance
Risk
12 to 48 months
>360 months
All
-1
>120 months
37 to 102 months
All
-1
49 to 120 months
103 to 360 months
All
-1
49-120 months
>360 months
All
-2
>120 months
103 to 360 months
All
-2
>120 months
>360 months
Abundant
-2
>120 months
>360 months
Present, Rare, Moderate
-3
If missing => Null
If missing => Null
All
Null
46
-------
4.1.1 Observed versus Preferred Habitats and Environmental Ranges
Several of the baseline/status rules (Table 4-1) as well as the SLR rules (Section 7) are modified
by whether a species occupies a "preferred" or "observed" habitat or environmental condition.
This section provides guidelines for distinguishing between the two (also see Lee et al., 2015).
Many marine and estuarine species are found across a wide range of habitats and environmental
conditions, yet the majority of the population occurs within a much more restricted range. For
example, several estuarine species are found in low abundances on the continental shelf but the
preferred habitat is intertidal estuarine soft bottoms. Classifying these species simply as estuarine
and oceanic is misleading about where the species primarily occurs, yet ignoring the oceanic
portion of the population truncates the species' environmental range. To address such cases, we
developed a natural history topology where many of the species' traits are classified as either
"observed" or "preferred" values. Observed and preferred classifications are used for regime,
habitat, salinity, depth, substrate, wave & current energy, adult & reproductive temperatures,
feeding type, and hosts for symbionts.
Preferred habitats or environmental ranges are those that the species "normally" occurs in.
Observed indicates that the species has been collected in a particular habitat or within an
environmental range but these condition may represent marginal conditions. All species have a
preferred habitat and environmental range though not all occur in marginal conditions or at least
have not been reported from marginal conditions. The preferred range can be conceptualized as
encompassing approximately 80% of the population while 20% of the population occurs under
the observed environmental conditions. In reality, such quantitative data are rarely available and
in lieu of such data we developed a set of guidelines to distinguish between the two (Table 4-3).
Note that observed is the default classification when there is insufficient information to decide on
the relative suitability of the habitat or environmental range for a species.
47
-------
Table 4-3. Guidelines to distinguish between observed versus preferred habitats and environmental conditions.
Observed indicates that the species has been collected within a particular environmental range and may represent marginal conditions. Preferred
indicates environmental conditions under which the species normally occurs. Observed is the default classification in absence of sufficient data.
The guidelines are listed in their approximate order of utility. The examples are for depth, which is classified as neritic (>0 to 200 m) with shallow
subtidal (>0 - 30 m) and deep subtidal (>30 - 200 m) subclasses.
Order
utility
Indicators of Preferred Environment
Indicators of Observed Environment
Comments
1
Experts classify environment as preferred (e.g.,
usually found, typically found, normally occurs at,
mostly found, common)
Experts classify environment as marginal
(e.g., rarely found, atypical, uncommon)
Such information often given in natural history
texts.
2
Moderate to high frequency of occurrence
Low frequency of occurrence
Relative to the specific species, requires
quantitative data.
3
Moderate to high abundance
Low abundance
Relative to the specific species, requires
quantitative data.
4
Environmental range reported multiple times across
multiple papers and databases
A particular environmental range is not or
only rarely reported in papers and
databases
Multiple reports of depths between 30 and 200 m
would indicate that the "deep subtidal" was a
preferred depth class. A caution is that the same
data are often repeated in different sources.
5
Species is observed in only in subclasses of an
environmental class, then the more general class is
preferred.
If species is observed in more than the two
higher level environmental classifications
If a species is observed in shallow subtidal and
deep subtidal, but nowhere else, the neritic is
classified as preferred.
6
If a species' observed quantitative range spans two
environmental subclasses and the occupied space
in one is > 80% of the subclass and < 20% in the
other, make the former preferred and the latter
observed
If a species' observed range does not follow
these criteria (i.e. more than 20% and/or
less than 80% in a subclass)
Shallow Subtidal: 20% of 30 m range = 6 m;
80% of 30 m range = 24 m.
Deep Subtidal: 20% of 170 m range = 34 m;
80% of 170 m range = 136 m
If a species range is 27-180 m, it has a 3 m
overlap in shallow subtidal which is < 6 m,
complies with the 20% rule. The 150 m in the
deep subtidal is > 136 m, complies with the 80%
rule. Therefore, shallow subtidal is classified as
observed and the deep subtidal as preferred.
7
If only the mean environmental value is given, use
as indicator of preferred level
No mean environmental value is given
Can indicate preferred environmental class but
not range.
48
-------
Order
utility
Indicators of Preferred Environment
Indicators of Observed Environment
Comments
8
Suitable for breeding, presence of gravid females,
breeding pairs, healthy nests
Marginal or unsuitable for breeding
Breeding can occur in marginal habitats, though
not as frequently.
9
Juveniles often found, juveniles present
Juveniles rarely found
10
Organisms found are normal to large in size
Evidence for stunted growth
Relative to the specific species.
11
Low levels of biochemical / physiological stress
markers
High levels of biochemical / physiological
stress markers
For example, HSP70 for temperature ranges.
49
-------
4.2 Baseline/Status Traits - Biogeographic Distributions
4.2.1 Introduction
While it is recognized that a species' vulnerability is related to its biogeographic distribution,
there is no uniform approach to defining ranges, and Gaston (1997) summarized 14 different
metrics used to describe geographic ranges. Our approach is to characterize distributions by
using the number of MEOW ecoregions occupied. This is a measure of the extent of occurrence
(EOO), or the "distance or area between the outermost limits to the occurrence of a species"
(Gaston, 1994, 1997; IUCN, 2001, 2016). We propose six distributional metrics, of which three
are based on the size of a species' range and three are based on where a species occurs (Table
4-1).
4.2.2 Relationship of Range Size to Vulnerability
There is considerable literature indicating that both marine (e.g., Roberts and Hawkins, 1999;
Musick et al., 2000; Polidoro et al., 2012) and terrestrial species (e.g., Cooper et al., 2008) with
smaller ranges are at a greater extinction risk. Similarly, paleontological evidence indicates that
taxa with small ranges were at greater extinction risk than those with wider ranges (see
McKinney, 1997 for review). In their simulation study, Pearson et al. (2014) found that the
"occupied area", defined as the total area of all occupied patches (= area of occupancy, AOO),
was the single most important variable predicting extinction due to climate change. Finally, the
IUCN uses both the AOO and EEO as key components of their Red List criteria (IUCN, 2016).
Lower risks associated with larger ranges result from a suite of non-exclusive factors. A wide
biogeographic distribution generally indicates that the species has wide physiological tolerances
to temperature and perhaps other environmental factors as well. Occupation of multiple regions
can mitigate the impact of local or regional perturbations by spreading of risk across a species'
geographical distribution (IUCN, 2016). Species with wide ranges often have a greater genetic
diversity than species with narrow ranges, suggesting a greater adaptability to environmental
changes. Conversely, a narrow range may indicate that the species has poor dispersal ability, is a
poor competitor, and/or is highly susceptible to predation, factors potentially increasing a
species' vulnerability to new stressors. Finally, species with small ranges often have low
abundances, though there are exceptions (see Gaston, 1994; Hobbs et al., 2011).
There are two cautions in interpreting the risks associated with small ranges. First, inadequate
sampling in certain regions may give the appearance of a limited distribution while, in fact, the
species extends over multiple ecoregions. Such overestimation of species with limited
distributions will be less pronounced for well-studied taxa like crabs, bivalves and fish than for
lesser studied taxa. Second, is the presence of cryptic species, "two or more distinct species
classified as a single species" (Bickford et al., 2006). Thus, what appears to be a widely
distributed species may consist of a number of localized, distinct species, potentially
50
-------
underestimating the number of species with limited distributions. Cryptic species will be most
prevalent among polychaetes (e.g., Nygren, 2014) and other taxonomically challenging taxa.
While it is not possible to generate rules to catch such errors, it is straightforward to modify
species' distributions as taxonomic revisions become available.
4.2.3 Endemics - Vulnerability Trait
As the smallest unit of occupancy in our analysis, we define endemic species as those occupying
a single MEOW ecoregion (e.g., Bonita mexicana, Figure 4-1). Based on the evidence discussed
above, endemic species are considered to be particularly vulnerable to both natural and
anthropogenic threats. However, as pointed out by Hobbs et al. (2011), "high abundance of
marine endemic species may buffer them from intrinsic characteristics that increase the
probability of extinction". Incorporating the potential for buffering, we generated the following
rules (see Table 4-1):
If an endemic species is abundant in the ecoregion => -2
If an endemic species has a moderate or rare abundance or is classified as present => -3
Approximately 13% of the brachyuran crabs are endemic (Table 4-4). Many of these endemics
occur in the Gulf of California (Cortezian Ecoregion), and the percentage of endemics is reduced
to 4.3% if this ecoregion is excluded. The other taxa range from 0 to 4.2% endemics, again with
many of the endemics in the Gulf of California. While a commonly used indicator of risk,
endemicity only identifies a relatively small number of at risk species.
Table 4-4. Number of species with endemic, restricted, or wide distributions.
The number of occupied ecoregions is evaluated globally. The values for the brachyuran crabs are also
calculated excluding the Cortezian Ecoregion because of the concentrations of endemics in the Gulf of
California. These values are preliminary and the bivalve results are based on incomplete trait analysis.
Taxon
#
Species
Average #
Ecoregions
Occupied
Median #
of
Ecoregions
Occupied
# Endemic
(%)
#
Restricted
(%)
Wide
Distribution
(%)
Brachyuran crabs
365
7.02
6
45
(12.57%)
29
(7.95%)
29
(7.95%)
Brachyuran crabs
- wo/ Cortezian
ecoregion
210
7.66
7
9
(4.29%)
22
(10.47%)
26
(12.38%)
Lithodid crabs
21
7.33
7
0
(0%)
1
(4.76%)
6
(28.57%)
Rockfish
71
4.94
5
3
(4.22%)
11
(15.49%)
12
(16.90%)
Bivalves
892
8.45
6
35
(3.92%)
48
(5.38%)
172
(19.28%)
51
-------
4.2.4 Restricted Distribution - Vulnerability Trait
Most coastal species have wide distributions. Crab, bivalve and rockfish species occupy an
average of about 5 to 8.5 ecoregions (Table 4-4). Based on these wide distributions, we created a
second metric, species with restricted distributions, defined as species occurring in only two
MEOW ecoregions (e.g., Lophopanopeus leucomanus (Figure 4-2). Species with restricted
distributions are subject to the same vulnerabilities as endemics, though not to the same severity
because of their wider span in temperature and other environmental conditions as well as a
greater spreading of risk. We apply analogous rules for restricted species as with the endemics
but the risk is reduced by one risk class based on the assumption of a lesser vulnerability (Table
4-1):
Species present in only two ecoregions and Hyper-rare in one or both > -3
Species present in only two ecoregions and Rare in both => -2
Species present in only two ecoregions and Present or Moderate or Abundant in one
or both => -1
Species present in more than two ecoregions => 0
Abundance
Very Abundant
M
High Moderate
| Abundant
u
Moderate
3 Moderately Abundant
Low Moderate
$ Moderately Rare
y
Present
| Rare
¦
Reported Absent
| Very Rare
~
Transient
| Hyper Rare
i
Conflict
Very Abundant
The most numerous species within an
ecoregion, usually inhabit a habitat of large
spatial extent and/or multiple habitats.
Figure 4-1. Example of an endemic species, defined as occupying only one MEOW ecoregion.
The pinnotherid crab Bonita mexicana has only been reported from Tortugas Bay, Mexico, which puts it in
the Southern California Bight Ecoregion. The color key is used in CBRAT to symbolize the ecoregion
relative abundance classes.
52
-------
Figure 4-2. Example of a species with a restricted distribution, defined as species
occupying two MEOW ecoregions.
The mud crab Lophopanopeus leucomanus has been reported only from the
Northern California and Southern California Bight ecoregions. The color key to
the ecoregion relative abundance classes is given in Figure 4-1.
The risk is increased if one of the ecoregions is classified as Hyper-rare since these populations
are so small that such species are functionally endemics. As with the endemics, risks are
modified by abundance, with a greater risk when the species is rare in both of the ecoregions.
Based on these rules, 29 brachyurans have restricted di stributions or 22 if the Cortezian
Ecoregion is excluded. In comparison, 1 lithodid crab, 11 rockfish, and 48 bivalves have
restricted distributions (Table 4-4).
4.2.5 Wide Distributions - Resilience Trait
Wide biogeographic distributions indicate that a species has wide environmental tolerances, and
thus should be less vulnerable to climate change. Wide distributions may also reduce
vulnerability by spreading of risks and indicate a larger total population size. One possible metric
for wide distributions is the number of ecoregions occupied compared to the average (or median)
for the taxon (Table 4-4). The limitation is that if all the ecoregions occur in regions with similar
temperatures, the number of ecoregions occupied may not accurately identify species with wide
environmental tolerances. For example, a number of the warm-water brachyuran crabs that reach
their northern limit in the Magdalena or Cortezian ecoregions extend southward through the
Tropical Eastern Pacific Province and into Indo-Pacific ecoregions. The coral gall crab,
Hapalocctrcinus mcirsnpialis, occurs in the Cortezian Ecoregion and twenty Jndo-Pacific
ecoregions. Similarly, many Arctic species extend over multiple Arctic and cold temperate
ecoregions, such as the Arctic lyre crab, Hyas coarctatus that occupies 27 ecoregions (Figure
53
-------
4-3). Although these two distribution patterns exceed the average number occupied for
brachyuran crabs, all the ecoregions tend to have broadly similar temperature regimes.
To better identify species with broad environmental tolerances, we define wide distributions as
those that encompasses at least three MEOW provinces with different temperature regimes. For a
species limited to the NEP and U.S. Arctic, it would be considered to have a wide distribution if
it occurred in the Arctic, Cold Temperate Northeast Pacific Province and the Warm Temperate
Northeast Pacific Province. Alternatively, a species would be considered widespread if it
occurred in the Cold Temperate Northeast Pacific, Warm Temperate Northeast Pacific, and the
Tropical Eastern Pacific provinces. An example is the sandflat elbow crab, Latirfambrus
occidentalis, that occurs from the Northern California Ecoregion to the Guayaquil Ecoregion,
encompassing three MEOW provinces (Figure 4-4).
Analysis for wide distributions is based on global di stributions of the species, with a list of
provinces in polar regions (Arctic/Southern Ocean), cold temperate, warm temperate, and
tropical temperature regimes in Table 4-5. We assign a moderate resilience level to species with
wide distributions, as defined by the following rules (Table 4-1):
Species occurs in Arctic & Cold Temperate & Warm Temperate Provinces => 2
Species occurs in Cold Temperate & Warm Temperate & Tropical Provinces => 2
Species does not occur in three Provinces with different temperature regimes 0
A total of 29 brachyuran crabs, 6 lithodid crabs, 12 rockfish, and 172 bivalves have wide
distributions (Table 4-4).
Figure 4-3. Example of a species occupying multiple ecoregions but not classified as having a wide
distribution.
The crab Hyas coarctatus occupies 27 MEOW ecoregions in the Arctic and Cold Temperature
provinces, which have broadly similar temperature regimes. The color key to the ecoregion relative
abundance classes is given in Figure 4-1.
54
-------
Figure 4-4. Example of a species with a wide distribution, defined as occupying
three MEOW provinces.
The elbow crab Latulambrus occidentalis has been reported from cold temperate,
warm temperate, and tropical provinces. The color key to the ecoregion relative
abundance classes is given in Figure 4-1.
55
-------
Table 4-5. MEOW provinces in the four major temperature regimes.
The Arctic does not have provinces and is listed at the MEOW realm level.
Arctic and Southern Ocean
Cold Temperate
Warm Temperate
Tropical
Arctic
Amsterdam-St Paul
Agulhas
Andaman
Continental High Antarctic
Black Sea (Ponto-Caspian)
Benguela
Arabian
Scotia Sea
Cold Temperate Northeast Pacific
East Central Australian Shelf
Bay of Bengal
Subantarctic Islands
Cold Temperate Northwest Atlantic
Lord Howe and Norfolk Islands
Central Indian Ocean Islands
Subantarctic New Zealand
Cold Temperate Northwest Pacific
Mediterranean Sea
Central Polynesia
Juan Fernandez and Desventuradas
Northern New Zealand
Easter Island
Lusitanian
Southwest Australian Shelf
Eastern Coral Triangle
Magellanic
Warm Temperate Northeast Pacific
Galapagos
Northern European Seas
Warm Temperate Northwest Atlantic
Guinea Current
Southeast Australian Shelf
Warm Temperate Northwest Pacific
Gulf of Guinea
Southeast Australian Shelf
Warm Temperate Southeastern
Pacific
Hawaii
Southern New Zealand
Warm Temperate Southwestern
Atlantic
Java Transitional
Tristan Gough
West Central Australian Shelf
Java Transitional
Marquesas
Marshall, Gilbert and Ellis Islands
North Brazil Shelf
Northeast Australian Shelf
Northwest Australian Shelf
Red Sea and Gulf of Aden
Sahul Shelf
Somali/Arabian
South China Sea
South Kuroshio
Southeast Polynesia
56
-------
Arctic and Southern Ocean
Cold Temperate
Warm Temperate
Tropical
St. Helena and Ascension Islands
Sunda Shelf
Tropical East Pacific
Tropical Northwestern Atlantic
Tropical Southwestern Atlantic
Tropical Southwestern Pacific
West African Transition
West and South Indian Shelf
Western Coral Triangle
Western Indian Ocean
57
-------
4.2.6 Arctic Endemics - Vulnerability Trait
Species limited to one or more of the 19 ecoregions comprising the MEOW Arctic Realm (Table
4-6) are defined as Arctic endemics. Limitation to the Arctic indicates that the species has a
narrow temperature range and would be vulnerable to temperature increases as well as having
limited opportunity to migrate northward. Limitation to the Arctic is considered a moderately
strong indicator of risk, with the following rules (Table 4-1):
Species present only in Arctic ecoregions => -2
Species present in any ecoregion outside the Arctic => 0
This is not a common distribution (see Josefson and Mokievsky, 2013). None of the brachyuran
crabs, lithodid crabs or rockfish are limited to the Arctic though six bivalves that occur in U.S.
Arctic ecoregions are so limited (Table 4-7). An example is Boreacola maltzani (Figure 4-5) that
occurs in the Chukchi Sea, Beaufort Sea, and in Europe and Russia Arctic ecoregions.
Table 4-6. Arctic Ecoregions.
The High Arctic ecoregion was not included in the MEOW schema,
but was added to capture species that occur in the highest portion
of the Arctic.
Arctic Ecoregions
Eastern Bering Sea
Chukchi Sea
Beaufort Sea - continental coast and shelf
Beaufort-Amundsen-Viscount Melville-Queen Maud
High Arctic Archipelago
Lancaster Sound
Baffin Bay - Davis Strait
Hudson Complex
Northern Labrador
West Greenland Shelf
East Greenland Shelf
North Greenland
North and East Iceland
North and East Barents Sea
White Sea
Kara Sea
Laptev Sea
East Siberian Sea
High Arctic
58
-------
Table 4-7. Number of Arctic endemics, small island colonizers, and nonindigenous species.
The values are preliminary for the bivalves.
Taxon
# Species
Arctic
Endemic
(%)
Small Island
Colonizer
(%)
Nonindigenous
Species
(%>
Brachyuran crabs
365
0
(0)
106
(29.04)
5
(1.36)
Lithodid crabs
21
0
(0)
o o
1
(4.76)
Rockfish
71
0
(0)
o o
o o
Bivalves
892
6
(0.67)
118
(13.23)
10
(1.12)
Figure 4-5. Example of an Arctic endemic, defined as a species that occurs only in Arctic ecoregions.
The clam Boreacola mattzani is limited to the Chukchi and Beaufort Sea - Continental Coast and Shelf
ecoregions and to Eurasian Arctic ecoregions. The color key to the ecoregion relative abundance
classes is given in Figure 4-1.
4.2.7 Small Island Distributions - Resilience Trait
There is a set of MEOW ecoregions that are surrounded by water with no direct contact with the
mainland, which we refer to as island ecoregions. Mainland species that occur on these island
ecoregions possess three key traits indicating a lower vulnerability to environmental change (see
Whittaker and Fernandez-Pelacios, 2007 for a discussion of traits associated with island
colonizers). First, they have good dispersal ability to initially colonize island ecoregions. Second,
they are able to establish a population with a relatively small number of initial colonizers. Third,
island colonizers are able to maintain populations in a relatively limited area. These traits should
reduce vulnerability by enhancing the species' ability to recover from environmental
perturbations. Note that this increased resilience refers only to species occurring both on the
mainland and island ecoregions; island endemics are considered to be especially vulnerable to
anthropogenic perturbations (e.g., Whittaker and Fernandez-Palacios, 2007).
59
-------
While there is a debate about how population dynamics on small islands differs from those on
larger islands ("small island effect"; see Lomolino and Weiser, 2001; Triantis et al., 2012), we
assume that the biotic traits associated with colonization and small population viability are more
pronounced with species able to occupy small islands. For this analysis, we use the
Revillagigedos, the largest of the three island ecoregions in the Tropical Eastern Pacific, as the
upper limit. The Revillagigedos have a land area of approximately 158 km2, which we round to
200 km2 as our upper threshold. In comparison, Hawaii has an area of 28,311 km2. Based on this
threshold, there are 29 small island ecoregions globally, three of which occur in the Tropical
Eastern Pacific (Table 4-8).
The upper size threshold of only 200 km2 is a stringent criterion and, accordingly, species that
occur in small ecoregions are assigned a moderate resilience. The specific rules (Table 4-1) are:
If species occupies a "Small Island" Ecoregion => 2
If species does not occupy a "Small Island" Ecoregion" => 0
A total of 106 brachyuran crabs and 118 bivalve species occupy small island ecoregions (Table
4-7). The majority of these are reported from the Gulf of California presumably reflecting the
subtropical/tropical nature of Revillagigedos, Clipperton, and Cocos Islands as well as the lack
of island ecoregions off more northern ecoregions. In comparison to the bivalves and
brachyurans, no lithodid crabs or rockfish occupy small island ecoregions.
60
-------
Table 4-8. Small island ecoregions.
Island ecoregions are defined as ecoregions surrounded by water with no direct contact
with the mainland, while small island ecoregions are defined as those with a land area of
200 km2 or less.
ECOREGION
PROVINCE
Amsterdam-St Paul
Amsterdam-St Paul
Bermuda
Tropical Northwestern Atlantic
Bounty and Antipodes Islands
Subantarctic New Zealand
Bouvet Island
Subantarctic Islands
Campbell Island
Subantarctic New Zealand
Cargados Carajos/Tromelin Island
Western Indian Ocean
Chagos
Central Indian Ocean Islands
Clipperton
Tropical East Pacific
Cocos-Keeling/Christmas Island
Java Transitional
Cocos Islands
Tropical East Pacific
Easter Island
Easter Island
Fernando de Naronha and Atoll das Rocas
Tropical Southwestern Atlantic
Juan Fernandez and Desventuradas
Juan Fernandez and Desventuradas
Kermadec Island
Northern New Zealand
Lord Howe and Norfolk Islands
Lord Howe and Norfolk Islands
Macquarie Island
Subantarctic Islands
Marshall Islands
Marshall, Gilbert and Ellis Islands
Ogasawara Islands
Tropical Northwestern Pacific
Peter the First Island
Subantarctic Islands
Phoenix/Tokelau/Northern Cook Islands
Southeast Polynesia
Rapa-Pitcairn
Southeast Polynesia
Revillagigedos
Tropical East Pacific
Sao Pedro and Sao Paulo Islands
Tropical Southwestern Atlantic
Snares Island
Southern New Zealand
South China Sea Oceanic Islands
South China Sea
Southern Cook/Austral Islands
Southeast Polynesia
Three Kings-North Cape
Northern New Zealand
Trindade and Martin Vaz Islands
Tropical Southwestern Atlantic
Tristan Gough
Tristan Gough
4.2.8 Nonindigenous Species - Resilience Trait
Nonindigenous species (NIS) is another group that has demonstrated both good dispersal ability
and the ability to establish populations with a small inoculant. Similar to the small island
occupants, these traits should reduce their vulnerability to environmental changes, and NIS may
61
-------
actually increase with climate change (e.g., Walther et al., 2009a). Because invaders may
undergo substantial declines after an initial "boom" (e.g., Delefosse et al., 2012) as well as being
potentially susceptible to climatic events themselves (McDowell et al., 2017), we consider
invasion of a non-native ecoregion a moderate rather than high resilience trait. To reduce
uncertainty about the population status of the invaders, we only include invaders that are
considered established in a non-native ecoregion. This excludes a number of stocked non-native
species in Asia where it is not clear if they have established a breeding population in the wild,
and which are assigned an Unknown establishment class (see Lee and Reusser, 2012). The
specific rules forNIS (Table 4-1) are:
Species has a Master NIS classification with a Master Established classification
anywhere globally => 2
Species has a Master NIS classification but is Not Established, Unknown Establishment,
or only Stocked classification => 0
Species does not have a Master NIS classification => 0.
We use our previous synthesis of the distribution of NIS in the North Pacific (Lee and Reusser,
2012) to identify species from the Northwest Pacific (NWP) that have invaded the NEP and,
conversely, native species from the NEP that have invaded the NWP. For invaders on the U.S.
East Coast, Europe, and other areas, we use previous summaries of NIS (e.g., Ruiz et al., 2000;
Streftaris et al., 2005). Using these sources, we identified five established NIS brachyuran crabs
and one nonindigenous lithodid crab that was purposely introduced by the Russians into the
Barents Sea (J0rstad et al., 2002) (Table 4-7). There are no established non-native rockfish and
ten established nonindigenous bivalves.
4.3 Baseline/Status Traits - Relative Abundance Patterns
4.3.1 Background on Relative Abundance Metrics
As mentioned, ecoregional abundance patterns provide additional insights into vulnerabilities
than those provided by biogeographical distribution patterns alone and we propose six sets of
rules based on relative abundance (Table 4-9). Because of the substantially greater percentage of
species classified at Leve II relative abundance compared to Level III, we utilize Level II in
generating the rules.
62
-------
Table 4-9. The number of species identified by each of the relative abundance rules.
Values in parentheses are percentage of the species in the taxon. All values are preliminary. NYA = not yet
analyzed.
Taxon
#
Species
Hyper
-rare
Abundant
Someplace
Rare
Everywhere
Population
Decline
S. Ecoregion
Rare -
N. Ecoregion
Abundant
Northern
Transient
Brachyuran
crabs
365
19
(5.20)
36
(9.86)
57
(15.61)
77
(21.10)
0
(0)
13
(3.56)
Lithodid
crabs
21
0
(0)
1
(4.76)
3
(14.28)
3
(14.29)
1
(4.76)
0
(0)
Rockfish
71
0
(0)
21
(29.58)
24
(33.80)
20
(28.17)
3
(14.28)
1
(1.41)
Bivalves
892
5
(0.56)
107
(12.00)
172
(19.28)
NYA
NYA
27
(3.02)
4.3.2 Hyper-Rare Species - Vulnerability Trait
Because of the high percentage of rare species (e.g., Gaston, 1994; Flather and Sieg, 2007), rarity
in itself is an insufficient trait to identify vulnerable species. The one exception are Hyper-rare
species, which are species that have not been observed in 50 years, assuming at least a minimal
sampling effort (see Section 3.3). We interpret this extreme rarity as an indicator of
environmental/biotic conditions unfavorable for the species, which in turn indicates a high
vulnerability to other stressors. The specific rules are (Table 4-1):
Species is Hyper-rare => -3
Species is not Hyper-rare => 0
No lithodid crabs or rockfish are classified as Hyper-rare and only five bivalves (Table 4-9). In
comparison, there are 19 brachyuran crabs classified as Hyper-rare in one or more ecoregions.
Part of the reason for the higher number of brachyurans may result from initial poor descriptions
of some of the species, especially pinnotherid crabs, reducing the likelihood that recent
researchers would report these species. Because of this possibility, researchers should check the
taxonomy of Hyper-rare species to help distinguish between true rarity and taxonomic
uncertainty.
4.3.3 Abundant Someplace/Rare Everywhere - Vulnerability and Resilience Traits
Species that are abundant someplace possess a suite of traits that allow them to effectively
exploit the available resources under the correct conditions, a suite of traits not shared by many
species as indicated by the relatively small percentage of abundant species in nearly all
assemblages (see Section 3; Gaston, 1994; Flather and Sieg, 2007). The life history attributes
promoting abundance are presumably related to those allowing a species to adapt to
63
-------
environmental changes. Additionally, the occurrence of a large population in one or more
regions provides at least a short-term buffer against detrimental environmental changes and a
greater opportunity for re-colonization of impacted regions. Accordingly, we predict species that
are abundant in at least one ecoregion are more likely to adapt to environmental changes. Species
that are rare everywhere are essentially the converse of this, and are predicted to have a higher
vulnerability and lesser ability to adapt to climatic changes. Because of the myriad of ecological,
historical, and sampling factors that can affect abundance estimates, we assign a low resilience or
vulnerability score to these two attributes. The specific rules are (Table 4-1):
Species is Abundant in any ecoregion => 1
Species is not Abundant in any ecoregion => 0
And for rarity:
Species is Rare or Hyper-Rare in all ecoregions => -1
Species is Present, Moderate, or Abundant in one or more ecoregions => 0
Ecoregions outside of the Northeast Pacific and U.S. Arctic are included in both analyses. The
criterion for rarity is applied strictly and species are not classified as "rare everywhere" if they
are classified as Present in any ecoregion. Thirty-seven crabs are abundant somewhere, including
36 brachyurans and 1 lithodid crab, while 21 rockfish and 107 bivalves are abundant somewhere
(Table 4-9). In terms of rarity, 57 brachyuran crabs are rare everywhere compared to 3 lithodid
crabs. A total of 24 rockfish and 172 bivalves are rare everywhere.
4.3.4 Population Trends - Vulnerability and Resilience Traits
Population trends are an important criterion in evaluating whether a species is at risk in
conservation ecology (Flather and Sieg, 2007) and is a key factor in determining extinction risk
in the IUCN's Red List (fattp://www.iucnredlist.ore/: Keller and Bollmann, 2004; Akgakaya et
al., 2006). Population trends are also used in evaluating risk to climate change; the evaluation of
freshwater fish vulnerability to climate change used four measures of population decline over
different time periods (Moyle et al., 2013).
Incorporating population trends into the present risk schema requires three steps, the first of
which is generating thresholds or cutpoints for different classes of population increases or
declines. To the extent possible, we harmonized our thresholds with the A2-A4 criteria of the
IUCN in their Red Book listing (IUCN, 2016) (Table 4-10). This resulted in four classes of
population decline with thresholds analogous to the A2-A4 criteria. Additionally, our Unknown
is generally equivalent to the IUCN's Data Deficient. However, an important difference is that
we assign "No Apparent Trend" in ecoregions with at least minimal background information on
the species if there is no indication of a decline instead of "Unknown". Our logic is that even
with poorly sampled species, a >30% decline in a species or a large loss in the species' habitat
64
-------
will often be noted by natural historians or observed in sampling programs. While this approach
has the advantage of moving beyond the refrain of insufficient information, it is not as
comprehensive as the procedure used by IUCN. Another difference is that the IUCN rule for
population reductions is that population declines should be calculated for the most recent three
generations or 10 years, whichever is longer. We attempted to follow this guideline, but in some
cases recent information was not available for non-commercial species and we had to rely on
older observations. Because of these differences, the population trend assignments in CBRAT
can be used to identify species of concern but they are not directly transferable to an IUCN Red
List assessment without additional analysis.
The IUCN does not have thresholds for population increases, which are important for capturing
climate change "winners". Accordingly, we generated increase trend cutpoints that mirror the
declines. The 100% increase is a doubling, which is proportionally equivalent to a 50% decline.
In some cases, there may be very large increases and the "Order of Magnitude Increase" class
was added to capture population "booms" in recent invaders as well as for the potential of large
initial increases in native species migrating into a northern ecoregion.
Table 4-10. Population trend classes based on percent change in population size within an ecoregion.
The closest equivalent IUCN A2-A4 criteria (IUCN, 2016) are given. The CBRAT assignments are not
directly transferable to an IUCN Red List assessment without additional analysis.
Population Trend Class
Population Trend
(% change in population
size)
Closest Equivalent IUCN
A2-A4 Criterion
Order-of-Magnitude Increase
>10X
None
Major Increase
>100% to <10X
None
Substantial Increase
50% to 100%
None
Moderate Increase
30% to 49%
None
No Apparent Trend
-29% to 29%
<30% decline
Moderate Decrease
-30% to -49%
Vulnerable (>30% decline)
Substantial Decrease
-50% to -79%
Endangered (>50% decline)
Extreme Decrease
-80% to -99%
Critically Endangered (>80% decline)
Extinct/Extirpated
-100%
Possibly Extinct & Extinct
Unknown
NA
Data Deficient
The second step in incorporating population trends is to generate a set of rules relating the
population trends classes to risk classes. The logic is that stress due to climate change will
exacerbate any current population declines due to habitat loss, overfishing, or other non-climate
drivers (e.g., Hewitt et al., 2016). While the exact nature of such interactions are generally
unknown, we assume that the greater the current population decline the greater the impact of
additional climate-related stress. Thus, we assign high, moderate, and low climate risks to the
IUCN's critically endangered, endangered, and vulnerable classes, respectively. This assignment
65
-------
is similar to that used in the climate risk analysis for freshwater fishes (Moyle et al., 2013),
which used >80% and >50% reduction as the two most severe classes of long-term population
trends. However, we modify the risk based on population abundance. The default risk for a
Substantial Decrease is Moderate, but increased to High when the species is Rare within an
ecoregion. This increased risk is based on the premises that a rare population has less of a buffer
to respond to an additional stressor and that the existing environmental conditions in the
ecoregion are unfavorable to rare species.
Using this logic, the following rule set is applied on an ecoregion-by-ecoregion basis (Table
4-1):
If Population Trend is
If Population Trend is
If Population Trend is
If Population Trend is
If Population Trend is
If Population Trend is
Abundant => -2
If Population Trend is
If Population Trend is
If Population Trend is
The third step is to estimate population trends across all the species and ecoregions being
evaluated. Ideally, quantitative trend data would be used for each species in each ecoregion,
however the reality is that such data are not available for the vast majority of near-coastal
species, with the exception of some commercial species. Rather than limit our analysis to
commercial species, we take a more liberal approach and use whatever population information is
available. The following are used as indicators of Substantial to Extreme declines: 1) closed
fishery and 2) species included on regional threatened or endangered lists. Other indicators used
to support quantitative population trend data, or used when such data are not available, include:
1) vulnerability to overfishing; 2) vulnerability to by-catch; 3) vulnerability to trawling damage;
4) documented or projected near-term habitat loss; 5) vulnerability to NIS; and 6) vulnerability to
pollution.
An example of using expert opinion is that several crabs in the Gulf of California occurring at
the depths of shrimp trawlers are considered to be declining due to their susceptibility to trawling
damage and as by-catch (R. Brusca, personal communication to Henry Lee, 2015). An example
of using habitat loss to identify likely declines is the pinnotherid crab Scleroplax granulata,
"Major Increase" OR "Order of Magnitude Increase" => 3
"Substantial Increase" => 2
"Moderate Increase" => 1
"No Apparent Trend" => 0
"Moderate Decrease" => -1
"Substantial Decrease" AND abundance is Present, Moderate, or
"Substantial Decrease" AND abundance is Rare => -3
"Extreme Decrease" => -3
"Unknown" => Null
66
-------
which is likely declining due to the decline of a major host, Upogebiapugettensis, resulting from
an invasive parasitic isopod (Griffen, 2009; Dumbauld et al., 2011).
Based on these multiple lines of evidence, we assign a population trend class. If there is
sufficient information to assign an abundance class but no evidence of a decline, the "No
Apparent Trend" class is assigned. If there is insufficient information to assign an abundance
class, "Unknown" is assigned. There is reasonable population trend data for rockfish in many of
the ecoregions, and 20 of the 71 bottom-associated rockfish in the Northeast Pacific are
classified as experiencing at least moderate declines in one or more ecoregions (Table 4-9). In
nearly all cases, these declines are the result of overfishing of these slow growing, long-lived
species (e.g., Musick et al., 2000, Drake et al., 2010). While there are less data for the crabs, 77
brachyuran crabs and 3 lithodid crabs are classified as undergoing population declines in one or
more ecoregions (Table 4-9).
4.3.5 Southern Ecoregion Rare and Ecoregion to North Abundant - Vulnerability Trait
A few species show a strong gradient in their relative abundance, being Rare in a southern
ecoregion and then Abundant in the ecoregion immediately to the north. Our interpretation is that
these species are not well adapted to the direct or indirect effects of the higher temperature
regime in the south compared to the cooler northern ecoregion. Thus, the population in the
southernmost ecoregion is considered to be moderately vulnerable to increased air and/or water
temperatures, whether through direct thermal effects or indirectly through altering ecological
processes (e.g., trophic dynamics, competitive interactions).
In the case of disjunct distributions, all disjunct ecoregions to the south of the Rare ecoregion
abutting the Abundant northern ecoregion are assigned a moderate climate risk if they are all
Rare. However, if any of these disjunct southern ecoregions are not classified as Rare, the risk
for all the ecoregions is set to 0. The Aleutians are considered north of the Gulf of Alaska
because the Aleutian Ecoregion has a lower mean sea surface temperature (Payne et al., 2012a).
Similarly, the Puget Trough/Georgia Basin Ecoregion is considered north of the Oregon,
Washington, Vancouver Coast and Shelf Ecoregion. With these definitions, the specific rules are
(Table 4-1):
If a Rare ecoregion abuts an Abundant ecoregion to the north and there are no Present,
Moderate, or Abundant ecoregions to the south of the Rare ecoregion => -2
All ecoregions to the south of the Rare ecoregion abutting the Abundant ecoregion are
also Rare => -2
If not one of the above cases => 0
67
-------
This is an uncommon pattern, and no brachyuran crabs and only one lithodid crab, Lithodes
aequispinus, displays this regional abundance pattern (Table 4-9). The pattern is slightly more
common among the rockfish, with three species showing this regional pattern.
4.3.6 Northern Transients - Resilience Trait
We define Transients as species that temporarily inhabit an ecoregion beyond their normal range
due to unusual climatic or oceanographic events. By definition, Transients are unable to maintain
a long-term viable population in the new ecoregion under present conditions, and thus die out in
one or a few generations. Species introduced outside of their natural range via anthropogenic
vectors, such as ballast water discharges, are considered NIS and not transients. Our concept of
Transient is similar to "vagrant", "visitor", "extralimital", or "ephemeral" species as used by
various authors (e.g., Rodrigues and Gaston, 2002). We further distinguish between "northern
transients" that occur to the north of the species' northern range limit versus the less frequent
"southern transients" that occur to the south of the species' southern range limit.
Transients are often rare in their "invaded" ecoregion, though this is not inherent in its definition.
Since abundance alone cannot be used to differentiate between the random sampling of rare,
native species versus transients, we utilize the species' association with an event, the presence of
only juveniles, higher abundances to the south, and expert opinion as the primary approaches to
identifying transients. On the Pacific Coast, the major oceanographic event resulting in northern
transient species is the occurrence of El Nino, which results in warmer waters off of Baja,
Mexico to Alaska (e.g., Chavez et al., 2002). For example, the 1997-1998 El Nino was
exemplified by "an unusually high occurrence of subtropical organisms along the California
coast" (Pondella and Allen, 2001; also see Lea and Rosenblatt, 2000; Engle and Richards, 2001).
Occurrence of these southern species in northern ecoregions during these warm-water events
demonstrates that they have the characteristics needed to migrate rapidly under favorable
conditions. Such migration may be active, which may be the case with swimming portunid crabs.
Alternatively, species may possess traits that promote their passive transport northward during
larval or adult stages. Occurrence of transients outside their normal range also demonstrates that
they can survive at least for short periods in the northern ecoregion(s) under the conditions
associated with an El Nino. Our projection is that with the advent of increased water
temperatures there is a high likelihood that these species will become established in these
northern ecoregions, assuming no other environmental limitation. The specific rules for northern
transients in their invaded ecoregion are (Table 4-1):
If a Transient ecoregion occurs to the north of an occupied ecoregion => 3
If not a Transient to the north of occupied ecoregion => 0
68
-------
Thirteen brachyuran crabs were associated with El Nino or other oceanographic events and are
classified as northern transients in one or more ecoregions (Table 4-9). There are no northern
transients among the lithodid crabs, though there is one southern transient, Hapalogaster
cavicauda, in the Gulf of California. Only one rockfish (Sebastes alutus) has a transient
population recorded in the Chukchi Sea, most probably carried from the Bering Sea by ocean
currents (Mecklenburg et al., 2002). Among the bivalves there are 27 northern transients.
4.4 Baseline/Status Traits - Life History
4.4.1 Introduction
Key life history attributes related to a species' vulnerability include degree of specialization,
reproductive strategies, and population growth rates. Addressing specialization first, a host of
studies have linked niche specialization with increased species' vulnerability. McKinney (1997)
summarized fossil evidence for terrestrial and marine taxa indicating that rare stenotopic
(specialist) species were more prone to extinction than rare eurytopic (generalist) species.
Studies on extant populations of birds (Jiguet et al., 2007) and butterflies (Warren et al, 2001)
found that specialists were more vulnerable than generalists. Specialization has also been used
specifically to assess vulnerability to climate change. "Specialized habitat and/or microhabitat
requirements" is one of five attributes used by the IUCN in assessing vulnerability to climate
change in bird, amphibian, and reef-building corals (Foden et al., 2008) while the "degree of
habitat specialization" is used as one of the key components of terrestrial species' sensitivity to
climate change (U.S. EPA, 2009).
We distinguish three types of specialization: habitat specialization, trophic specialization and
symbiotic relationships. Recognizing that there is a gradient in the biotic relationships between
habitat specialization and symbiotic relationships, we differentiate symbiotic relationships as
species that live directly on or in its hosts, while habitat specialists live in the general vicinity of
a particular biotic habitat, particularly macrophytes. A crab living within a polychaete tube
would be classified as a symbiont while a crab living in association with mangroves would be
classified as a habitat specialist. We also classify symbionts as habitat specialists, and trophic
specialists when appropriate, to highlight the nature of the biotic interactions.
Two other life history traits that we use to predict vulnerabilities are first the presence of a
diadromous reproductive strategy, where the species spends part of its life in freshwater and part
in saltwater, and second the potential population growth rate of fishes.
4.4.2 Symbiotic Relationships - Vulnerability Trait
Symbiotic interactions are often referred to as commensalism, defined as a biotic relationship in
which the commensal (symbiont) benefits and the host is not affected (+/0 relationship), or
mutualism where both symbiont and host benefit (+/+ relationship). In many cases, closer
69
-------
examination of these relationships shows that the symbiont benefits at the expense of its host (+/-
relationship). In particular, kleptoparasitism, where the symbiont steals food from the host, may
be relatively common among marine symbionts (e.g., Telford, 1982; Morissette and
Himmelman, 2000; Iyengar, 2005). Classic parasitism, such as a gill parasite, is also a type of +/-
symbiotic relationship. Regardless of the nature of the interactions, three key aspects of a
symbiotic relationship in terms of vulnerability are: 1) strength of the biotic relationship; 2)
vulnerability of the host to climate change; and 3) abundance of the symbiont.
In terms of the strength, we define an obligatory relationship as one where at least one life
history stage of the symbiont requires a host(s). Obligatory symbionts are nearly always found
(>90%) in association with a single host or suite of host taxa. We do not assume a 100%
association because symbionts may occasionally be found outside the host(s) because of
reproduction (e.g., male Pinnixa crabs searching for females) or when migrating to a new host.
With these highly dependent species, loss of the host(s) would result in a major population
decline, potentially resulting in local or regional extinction. In cases where the symbiont infests
multiple hosts, each host would be considered a facultative relationship but the symbiont would
be considered obligate if it is nearly always associated with a host. We define a facultative
relationship as when the target species occurs with its host(s) 10% to <90% of the time. While
less vulnerable, populations of facultative symbionts likely would experience declines with the
reduction or loss of their host(s). An "incidental" relationship is defined as one that occurs <10%
of the time; loss of such a host(s) would presumably have a minor impact on the symbiont
population. Note that while we pose these as quantitative thresholds, in most cases the strength
of the association has to be evaluated from qualitative data.
Another factor affecting the risk to a symbiont is the vulnerability of the host(s) to climate
change. Ideally, the risk analysis for the symbiont would explicitly incorporate the risk to each
host in each ecoregion. This is not currently possible because the climate risks of many of the
host taxa (polychaetes, echinoderms, and corals) have not yet been completed. In the interim, we
take a conservative approach of assuming that the hosts are vulnerable, which may overestimate
symbiont risk in some ecoregions. To at least identify the potential for such interactions, CBRAT
has a simple classification whether climate change is likely to impact the primary hosts. While
not currently utilized in the rules, users can evaluate our current general assessment of climate
impacts on the primary hosts.
The third factor impacting a symbiont's risk is its relative abundance, with abundant symbionts
having a greater buffer to environmental changes compared to rare symbionts. A high frequency
of occurrence of a symbiont with its host(s) does not in itself indicate that the symbiont is
abundant as the host(s) may be uncommon. Rather, the relative abundance of the host(s) needs to
70
-------
be factored into assessing a symbiont's abundance. After incorporating abundance, the rules for
symbionts become (Table 4-1):
If "Strength of Relationship" (Symbiotic) is Incidental => 0
If "Strength of Relationship" (Symbiotic) is Facultative => -2
If "Strength of Relationship" (Symbiotic) is Obligate
AND abundance is Present, Moderate, or Abundant => -2
If "Strength of Relationship" (Symbiotic) is Obligate
AND abundance is Rare or Hyper-Rare => -3
If no symbiotic relationship => 0
Symbiotic relationships are relatively common among the brachyuran crabs, with 51 species
having an obligate relationship (Table 4-11), the majority of which are pinnotherid crabs. In
comparison, no lithodid crabs or rockfish are symbiotic. The analysis for symbiotic relationships
for the bivalves has not been completed.
Table 4-11. Number of species with symbiotic relationship, habitat specialization, trophic
specialization and anadromous/catadromous reproduction.
Numbers are for high levels of specialization. Habitat specialization for the brachyurans does
not include the pinnotherid crabs, which are captured under symbiotic relationships.
Specialization classifications have not yet been completed for the bivalves. All values are
based on a preliminary analysis. NYD = not yet determined. Numbers in parentheses are
nprrpnt nf cnpripc
Taxon
#
Species
Obligate
Symbiotic
Relationship
High Habitat
Specialization
High Trophic
Specialization
Anadromous/
Catadromous
Brachyuran
crabs
365
51
(13.97)
8
(2.19)
5
(1.36)
1
(0.27)
Lithodid crabs
21
0
(0)
0
(0)
0
(0)
0
(0)
Rockfish
71
0
(0)
0
(0)
0
(0)
0
(0)
Bivalves
892
NYD
NYD
NYD
0
(0)
4.4.3 Habitat Specialization
Certain species occupy "unique habitats of limited distribution", which we define as spatially-
limited habitats with a physical/chemical structure distinct from other habitats, providing unique
environmental conditions. Spatially limited is used in comparison to the area of all the other
habitats within the ecoregion. Some unique habitats, such as tide pools, are extremely limited in
71
-------
area. Others, like marshes and mangroves, occupy much larger areas but are still relatively small
compared to other habitats averaged over the ecoregion, in particular unvegetated sand/mud (see
Appendix C).
A total of 24 habitats or ecosystems are considered unique habitats of limited distribution (Table
4-12). Because not all these habitats are equally vulnerable to climate change, we separate them
into "climate vulnerable" and "climate resilient" habitats. This differentiation is used in a
comparative sense to separate the unique habitats most susceptible to climate change; it does not
mean that the resilient habitats will not be impacted by climate change. Marshes, mangroves, and
SAV habitats are included among the climate resilient habitats because independent risks are
generated for SLR for each of these habitats (Section 7). For these habitats, the resilient
classification is meant to capture non-SLR effects thus avoiding double accounting for SLR
risks.
Table 4-12. Unique habitats of limited distribution.
Unique habitats are separated into those most susceptible to climate change, the "climate vulnerable"
habitats, and those less susceptible, the "climate resilient" habitats. Each habitat is classified as
whether it is unconsolidated, consolidated, pelagic, or a specialized system. Specialized systems are
unique and spatially limited ecosystems composed of more than one habitat type. Non-coral reefs
include sponge and polychaete reefs. As used here, oyster beds are limited to those on
unconsolidated sediments. Indicates classified as resilient because a separate SLR risk is calculated
for each habitat.
Climate Vulnerable
Climate Resilient
Burrowing shrimp
Unconsolidated
Algal mats
Consolidated
Coral reef
Consolidated
Cold seeps
Specialized systems
Kelp
Consolidated
Dune
Unconsolidated
Mussel beds
Consolidated
Emergent Marshes*
Unconsolidated
Non-coral reefs
Consolidated
Hydrothermal vents
Specialized systems
Oyster beds
Unconsolidated
Mangrove*
Unconsolidated
Phyllospadix
Consolidated
Pelagic systems
Pelagic
Rhodoliths / Maerl
Consolidated
Saline lagoons
Specialized systems
Sea ice
Specialized systems
Sea mounts
Specialized systems
Solitary corals
Consolidated
Submerged Aquatic
Vegetation*
Unconsolidated
Tide Pools
Consolidated
Whale falls
Specialized systems
Wrack
Unconsolidated
Wood
Consolidated
As with the symbiotic relationships, habitat specializations may be obligate, facultative, or
incidental; independently a species utilization of a habitat is classified as observed or preferred.
Obligate is used in the sense that loss of the habitat would result in a substantial population
decline. All obligate habitat specializations are also classified as preferred habitats. In
comparison, all incidental specialized habitat utilizations are classified as observed habitats.
72
-------
Facultative utilization covers a broader range of reliance on the habitat with its importance
classified as an observed or preferred habitat. The specialization classification is keyed to the
most sensitive life history stage, and for some species the juvenile stage utilizes the habitat of
limited distribution. For example, several rockfish preferentially settle from the planktonic stage
into kelp beds (Love et al., 2002). While only dependent upon the habitat for a portion of its life
cycle, loss or reduction of nursery habitats could result in a population bottleneck.
Based on these definitions, the following rules were generated (Table 4-1):
If no Specialized Habitats => 0
Vulnerable Specialized Habitats
Obligate AND Preferred Habitat => -3
Facultative AND Preferred Habitat => -2
Facultative AND Observed Habitat => -1
Incidental AND Observed Habitat => 0
Resilient Specialized Habitats
Obligate AND Preferred Habitat => -2
Facultative AND Preferred Habitat => -1
Facultative AND Observed Habitat => -1
Incidental AND Observed Habitat => 0
Some degree of utilization of these unique habitats is relatively common among brachyuran
crabs, with approximately 118 species utilizing one or more of the unique habitats listed in Table
4-12. However, only eight non-pinnotherid brachyuran crabs are classified as obligate habitat
specialists (Table 4-11). None of the lithodid crabs or rockfish are obligate habitat specialists,
though as mentioned above, some juvenile rockfish are facultative habitat specialists.
4.4.4 Trophic Specialization
Several lines of evidence indicate that trophic specialists are more vulnerable to environmental
changes. The fossil record shows that detritus-feeders (= deposit feeders) have lower background
extinction rates than other feeding types, which McKinney (1997) attributed to "a more
generalized diet and lack of feeding specialization." Among extant species, trophic specialists
have shown greater vulnerability to environmental changes or disturbances in both terrestrial
(e.g., Charrette et al., 2006) and marine (e.g., Graham, 2007) species. Thus, there is strong
support to assign a high baseline/status vulnerability to species with specialized feeding habits.
73
-------
Details on the feeding regimes of most marine/estuarine species are not well known, but a
limited number of trophic specialists have been documented. Tropical corallivores display
varying degrees of specialization (Graham, 2007) while in temperate systems a small number of
species that are closely associated with biotic structures have restricted diets. Examples include
crabs associated with kelp (e.g., Pugettiaproducta\ Hines, 1982; Jensen, 1995, 2014), limpets
living in association with kelp (e.g., Patella argenvillei\ Bustamante et al., 1995), and crabs
living in coral (e.g., Trapezia bidentata; Abele, 1976). Most marine herbivore gastropods are
generalists with the exception of ascoglossan sea slugs, in contrast to terrestrial herbivores where
many species are specialists (Trowbridge, 1994; also see Clark, 1994). However, such trophic
specialists appear to be the exception among marine/estuarine species.
In the absence of species-specific information, a set of guidelines are used to assign trophic
classifications (Figure 4-6). Based on these guidelines, species are automatically assigned a
trophic classification in CBRAT for a single feeding mode using the rules below or according to
those in Figure 4-6 for multiple feeding modes. These rules are an initial effort at classifying
trophic interactions, and the automatic assignment based on these rules can be changed by users
via a "Manual Override" in CBRAT.
Species with Symbiotic Algae => Specialist
Species with chemoautotrophic bacteria => Moderate
Parasites/disease => Unknown
Primary producer => Generalist
Herbivore => Unknown
Herbivore - grazer => Generalist
Herbivore - folivore => Depends on number of plants species consumed
Specialist if < 5 food items.
Moderate if >5 food items and <10 food items
Generalist if >10 food items
Predator => Depends upon the number of prey items consumed
Specialist if < 5 prey items.
Moderate if >5 prey items and <10 prey items
Generalist if >10 prey items
Scavenger => Generalist
Detrivore => Generalist
Decomposer => Generalist
74
-------
Suspension Feeder => Moderate
Deposit Feeder => Generalist
Osmotrophy => Generalist
Table 4-13. Guidelines for assigning levels of trophic specialization for single and multiple feeding modes.
Assigning Levels of Trophic Specialization
•
Assign an Unknown as the default for unusual feeding modes, and then classify each of these species by the
specifics of its feeding modes.
•
Parasites/diseases vary greatly in specificity so set to Unknown. Parasites may occasionally have other
feeding modes, but not generalizable. Use same guidelines for number of hosts as with predators.
•
Loss of symbiotic algae harms corals, which are suspension feeders. This is the rationale for Specialist with
that combination. However, Unknown with other combinations since it depends upon how strongly the species
relies on symbiotic algae (e.g., predatory nudibranchs with symbiotic algae).
•
Assume that chemosynthetic bacteria are more robust than symbiotic algae, so assign Moderate if only
feeding type. Suspension feeders are Moderate, so combination with chemosynthetic bacteria results in
Generalist classification. Osmotrophy is not sufficient to change from Moderate unless shown to be a
substantial nutritional source.
•
Primary producers include both photosynthesis and chemosynthesis but classification is limited to
macrophytes, not symbiotic algae. Set to Generalist. Macrophytes not commonly combined with other feeding
modes (e.g., Venus flytrap). Set combinations to Unknown.
•
Herbivore alone is Unknown.
o Grazers appear to be Generalists on microalgae, so addition of other feeding type is still a
Generalist.
o Folivores are Unknown and are classified by the number of macrophytes species consumed.
•
. Herbivore with osmotrophy falls under the unusual guideline (Unknown).
•
Specialization of predators depend upon the number of prey consumed. Combination of predation with other
active feeding types is Generalist, except parasites which depends upon number of prey/hosts and
osmotrophy, which depends upon how important it is, so Unknown.
•
Scavengers, detritivores, and decomposers tend to feed on what they find (= Generalist) and addition of other
feeding types would not decrease level of specialization unless they were the dominant feeding type.
•
Suspension feeders appear to be less general in their feeding than deposit feeders, so Moderate. Combined
with other feeding type assign a Generalist except for primary producer and parasite/disease which fall under
the unusual trophic type guideline of Unknown.
•
While deposit feeders may select particular particle sizes, they are classified as Generalists in sense that their
food source is not highly susceptible to climate change. Combined with other feeding type does not reduce
Generalist classification except for primary producer and parasite/disease which fall under the unusual trophic
type guideline of Unknown.
•
Uptake of DOC as a primary food source is considered a Generalist. However, uptake of DOC appears to be
a supplemental trophic mode in many cases. Classification of osmotrophy combined with other feeding modes
depends upon the relative importance of osmotrophy compared to other feeding mode.
75
-------
Parasite /
Disease
Symbiotic
Algae
Chemo-
autotrophic
Primary
Producer
Herbivore
Herbivore -
Grazer
Herbivore -
Folivore
Predator
Scavenger
Detritivore
Decomposer
Suspension
Feeder
Deposit
Feeder
Osmo-
trophy
Parasite / Disease
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Symbiotic Algae
Specialist
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Specialist
Unknown
Unknown
Chemo-autotrophic
Moderate
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Generalist
Generalist
Moderate
Primary Producer
Generalist
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Herbivore
Unknown
NA
NA
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Unknown
Herbivore - Grazer
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Herbivore - Folivore
Unknown
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Unknown
Predator
Unknown
Generalist
Generalist
Generalist
Generalist
Generalist
Unknown
Scavenger
Generalist
Generalist
Generalist
Generalist
Generalist
Generalist
Detritivore
Generalist
Generalist
Generalist
Generalist
Generalist
Decomposer
Generalist
Generalist
Generalist
Generalist'
Suspension Feeder
Moderate
Generalist
Generalist
Deposit Feeder
Generalist
Generalist
Osmotrophy
Generalist
Figure 4-6. Default levels of trophic specialization based on single and two feeding modes.
These are the default levels of specialization generated automatically from the feeding type, but they can be modified by users in CBRAT.
Unknown = level of specialization varies depending upon the specific feeding habitats of the species, including the number of different types of
prey consumed. Several of the feeding combinations are possible but rare (e.g., predator and herbivore). Guidelines used to assign the levels of
tropic specialization are given in Table 4-13.
76
-------
The climate risks are then assigned based on the degree of specialization. With moderate
specialization, a reduction in key food items may have some impact on population viability.
However, many marine/estuarine species can switch diets (e.g., Graham, 2007; Jumars et al.,
2015), including native predators learning to prey on introduced species (e.g., Inger et al., 2010;
Dijkstra et al., 2013). Accordingly, moderate trophic specialists are assigned a low risk. In
contrast, trophic specialists are assigned a high risk since they have limited options to switch
diet. The specific rules (Table 4-1) become:
If Generalist Trophic Specialization => 0
If Moderate Trophic Specialization => -1
If Specialist Trophic Specialization => -3
Not including the pinnotherid crabs, which may be kleptoparasites, five brachyuran crabs display
a high degree of trophic specialization (Table 4-11). No lithodid crabs or rockfish display a high
degree of trophic specialization.
4.4.5 Anadromous/Catadromous
Diadromous species have specialized reproductive strategies in which they migrate to or from
marine waters to reproduce. Anadromous species spend most of their adult life in marine waters
and then migrate to freshwater to breed. Archetypical anadromous species are Pacific Northwest
salmon, such as ocean-type Chinook (Oncorhynchus tshawytscha) and chum salmon ((). keta).
Catadromous species spend most of their adult life in freshwater and then migrate to the ocean to
breed. This life history strategy is less common (Allen et al., 2006), but one example is the
American eel, Anguilla rostrata, on the East Coast of the United States. These life history
strategies are vulnerable to anthropogenic impacts both on their ability to migrate between
freshwater and marine environments and to climate impacts on their freshwater, estuarine, and
marine habitats (Greene et al., 2009).
We assign a high vulnerability to both life history strategies if the population is rare. Because
larger populations provide a buffer to the effects of climate impacts, we assign a moderate risk
when the species is moderate or abundant. The specific rules (Table 4-1) are:
If species is anadromous or catadromous AND Rare => -3
If species is anadromous or catadromous AND Present, Moderate, or Abundant => -2
If species is not anadromous or catadromous => 0
There are no anadromous crabs, rockfish, or bivalves in the NEP or U.S. Arctic (Table 4-11) but
there is one catadromous brachyuran crab. The mitten crab (Eriocheir sinensis) was introduced
into the San Francisco Estuary, where it spends most of its adult life in the freshwater delta,
77
-------
migrating down to the estuary to spawn (Rudnick et al., 2005). It was considered a serious pest
species in the Bay-Delta region, but its population has been declining such that adults were rarely
observed in 2012 (http://nas.er.usgs.gov/queries/FactSheet.aspx7speciesICN182V
4.4.6 Growth and Productivity
Life history parameters related to population productivity, including fecundity, intrinsic rate of
increase, age at maturity, and maximum age, have been used to evaluate vulnerability among a
diverse group of vertebrates including whales, sharks, and bony fishes (Musick 1999; Musick et
al. 2000; Dulvy et al., 2004; Gallagher et al., 2012). Using these traits, Musick et al. (2000)
provided thresholds for different risk levels with a primary focus on fishing pressure. In this
section, we adapt these thresholds as climate vulnerability metrics for fish. Currently, no
equivalent productivity thresholds are available for invertebrate taxa.
Threshold values for these productivity metrics to evaluate extinction risk with fish are shown in
Table 4-14 When the intrinsic rate of increase (r) is not available, age at maturity is the next most
important factor along with maximum age (Musick 1999). These later two metrics are often
correlated with the von Bertalanffy coefficient (k) (Musick 1999). High fecundity rates are useful
in some cases but may be misleading for Pacific rockfish whose reproductive patterns indicate
very low larval survival as well as infrequent recruitment (Musick 1999; Parker, et al., 2000).
Additionally, Denney et al. (2002) and Reynolds et al. (2005) found no evidence that high
fecundity increases recruit production or reduces likelihood of extinction.
Table 4-14. Productivity index parameter thresholds for fishes.
Measures include intrinsic rate of increase (r), von Bertalanffy k, fecundity, age at maturity (Tmat), and
maximum age (Tmax). The thresholds are guidelines for the risk of extinction from Musick (1999) and
Musick et al. (2000). Lower risk is associated with high productivity and higher risk with very low
productivity. These classifications are primarily developed for the effects of fishing pressure.
Productivity Parameter
High
Productivity
Medium
Productivity
Low
Productivity
Very Low
Productivity
r(yr.-1)
> 0.50
0.16-0.50
0.05-0.15
< 0.05
von Bertalanffy k (yr.-1)
> 0.30
0.16-0.30
0.05-0.15
< 0.05
Fecundity per year
>104
102-103
101 -102
< 101
Age at maturity
(T mat)
< 1 yr.
2-4 yr.
5-10 yr.
> 10 yr.
Maximum age
(T max)
1-3 yr.
4 -10 yr.
11-30 yr.
> 30 yr.
We generated the productivity climate rules based on maximum age and age at maturity for
females (Table 4-2) because they are among the most readily available productivity measures for
Pacific rockfish (e.g., Love 2011). These rules are based on the assumption that short-lived
species that mature earlier are less vulnerable to climate change impacts than species that take
78
-------
longer to mature. Such short-lived, rapidly maturing species are also more likely to rebound from
short-term climatic events better than slower maturing species. Table 4-15 summarizes the
available productivity values for Sebastes.
There are a number of differences in how fishing pressure and climate change impact fish
populations. For example, while overfishing tends to extirpate the largest/oldest rockfish (Dulvy
et al. 2003), climate change and habitat alterations often have a greater impact on larval/juvenile
recruitment and survivorship, as observed from the last 20 years of warming oceans (Love and
Yoklavich 2006). Another difference is that while long-lived species are at risk due to slower
reproduction, their ability to survive over decades has demonstrated an ability to adapt to a range
of environmental conditions. The spatial patterns of the stressors are also different, with fishing
concentrated where the fish are most abundant, while climate change will impact rare
populations as well as abundant ones.
Because of these differences, we suggest that the risk levels developed primarily for fishing
pressure (Table 4-14) need to be modified, including incorporating the potential buffering of
climate effects when there are abundant populations. Incorporating abundance with the various
combinations of maximum life span and age at maturity results in 15 rules (Table 4-2). As
detailed in Table 4-15, both required productivity values are available for 46 of the 71 bottom-
associated Sebastes species, a well-studied tax on. Of these 46 species, 37 are considered
somewhat to highly vulnerable to climate change based on the productivity metrics in at least
one ecoregion.
Table 4-15. Sebastes productivity parameters.
Productivity parameters are rated high (green), medium (yellow), low (irangi), and very low (red) based
on Musick (1999) and Musick et al. (2000) (see Table 4-14). Von Bertalanffy k values are for females or
sexes combined. Climate vulnerability is based on Tmat and Tmax, with the von Bertalanffy coefficient
and fecundity given for comparison. Productivity values are from Love et al. (2002) and Love (2011). Gray
indicates species for which none of the productivity parameters are available. The table includes three
primarily pelagic species (S. peduncularis, S. sinensis, S. varispinis) that are not included in the risk
analysis. ND = no data.
Age at Maturity
Max age
von
(Female; Tmat)
(Tmax)
Bertalanffy
Fecundity
Species
(years)
(years)
(k)
(# eggs)
Sebastes aleutianus
20
205
0.108
ND
Sebastes alutus
4-10
104
0.175
2,000-505,000
Sebastes atrovirens
3-6
25
0.29
10,000-340,000
Sebastes auriculatus
3-10
34
0.16
55,000-339,000
Sebastes aurora
11-32
118
0.06
ND
Sebastes babcocki
3-19
106
ND
ND
79
-------
Age at Maturity
Max age
von
(Female; Tmat)
(Tmax)
Bertalanffy
Fecundity
Species
(years)
(years)
(k)
(# eggs)
Sebastes baramenuke
ND
ND
ND
ND
Sebastes borealis
21-23
160
0.03
ND
Sebastes brevispinis
9-18
82
0.093
181,000-
1,917,000
Sebastes carnatus
ND
24
0.253
ND
Sebastes caurinus
3-8
50
0.1
16,000-650,000
Sebastes chlorostictus
6-19
51
0.062
14,000-760,000
Sebastes chrysomelas
3-6
30
0.22
25,000-450,000
Sebastes ciliatus
11
67
ND
ND
Sebastes constellatus
6-14
32
0.09
33,000-228,000
Sebastes cortezi
ND
ND
ND
ND
Sebastes crameri
8-9
105
0.16
20,000-610,000
Sebastes dallii
ND
12
0.12
3,900-18,000
Sebastes diploproa
6-10
103
0.1
14,000-255,000
Sebastes elongatus
3333-12
54
0.079
11,000-295,000
Sebastes emphaeus
1-2
22
0.53
3,300-58,000
Sebastes ensifer
3
43
0.14
12,200-38,000
Sebastes entomelas
3-8
60
0.2
95,000-
1,113,000
Sebastes eos
ND
52
ND
ND
Sebastes exsul
ND
24
ND
ND
Sebastes flavidus
<15
64
0.17
56,000-
1,992,700
Sebastes gilli
ND
60
ND
ND
Sebastes glaucus
7-10
19
ND
ND
Sebastes goodei
3-8
39
0.17
18,000-538,000
Sebastes helvomaculatus
ND
87
0.1
ND
Sebastes hopkinsi
3-7
19
0.18
9,000-39,000
Sebastes jordani
2-4
32
0.198
50,000
Sebastes lentiginosus
ND
22
ND
ND
Sebastes levis
ND
55
0.06
181,000-
1,925,000
Sebastes macdonaldi
ND
20
ND
ND
Sebastes maliger
5-22
95
0.07
ND
Sebastes melanops
5-15
56
0.33
283,618-
1,135,457
Sebastes melanosema
ND
ND
ND
ND
80
-------
Age at Maturity
Max age
von
(Female; Tmat)
(Tmax)
Bertalanffy
Fecundity
Species
(years)
(years)
(k)
(# eggs)
Sebastes melanostictus
ND
205
0.108
ND
Sebastes melanostomus
13-26
90
0.04
152,000-769,000
Sebastes miniatus
4-9
60
ND
63,000-
2,600,000
Sebastes moseri
ND
29
ND
ND
Sebastes mystinus
4-11
44
0.149
525,000
Sebastes nebulosus
3-6
79
ND
ND
Sebastes nigrocinctus
ND
116
ND
ND
Sebastes notius
ND
ND
ND
ND
Sebastes ovalis
4-12
37
0.05
61,000-160,000
Sebastes paucispinis
3-6
58
0.163
20,000-
2,298,000
Sebastes peduncularis
ND
ND
ND
ND
Sebastes phillipsi
ND
53
ND
ND
Sebastes pinniger
7-20
84
0.163
260,000-
1,900,000
Sebastes polyspinis
6-13
88
0.178
ND
Sebastes proriger
7 or more
70
0.166
ND
Sebastes rastrelliger
2-5
23
0.11
80,000-760,000
Sebastes reedi
ND
100
0.25
ND
Sebastes rosaceus
ND
14
0.12
12,600-95,000
Sebastes rosenblatti
4-15
58
0.05
30,000-655,000
Sebastes ruberrimus
20
147
0.04
1,200,000-
2,700,000
Sebastes rubrivinctus
ND
38
ND
ND
Sebastes rufinanus
ND
ND
ND
ND
Sebastes rufus
10-20
53
0.04
65,000-608,000
Sebastes saxicola
2-9
38
0.06
15,000-230,000
Sebastes semicinctus
1-6
15
0.37
3,000-31,000
Sebastes serranoides
3-8
30
0.18
30,000-490,000
Sebastes serriceps
3-7
25
0.233
70,000
Sebastes simulator
ND
36
ND
20,880-63,700
Sebastes sinensis
ND
ND
ND
ND
Sebastes spinorbis
ND
45
ND
ND
Sebastes umbrosus
3-8
31
ND
ND
Sebastes variabilis
9
76
0.235
ND
81
-------
Age at Maturity
Max age
von
(Female; Tmat)
(Tmax)
Bertalanffy
Fecundity
Species
(years)
(years)
(k)
(# eggs)
Sebastes variegatus
ND
47
0.11
ND
Sebastes varispinis
ND
ND
ND
ND
Sebastes wilsoni
ND
26
ND
ND
Sebastes zacentrus
6-10
73
0.122
ND
4.5 Climate-Adjusted Baseline/Status Risks - Linking Baseline/Status & Climate
Risks
As discussed in Section 4.1, the baseline/status risks are defined as the increased risks under
climate stress. If there is no climate risk, these baseline/status risks would not contribute to the
overall climate risk score regardless of their value. This is not to state that traits like habitat
specialization or endemicity do not represent a vulnerability for these species under current
climatic conditions, but that the overall climate risk should not be increased if there is no
substantial stress from climate change. Conversely, in cases where a species is impacted by one
or more climate drivers, the baseline/status risk potentially contribute to the species' overall
climate vulnerability.
The procedure for linking baseline/status risks to the extent of climate change is to first
determine the greatest individual climate risk among temperature, ocean acidification, and sea
level rise and the greatest individual baseline/status risk. From these two values, the climate-
adjusted baseline/status risk is calculated as:
1. If greatest climate risk is Minor (0), the climate-adjusted baseline/status risk is set to
Minor (0) regardless of the individual baseline/status risk values.
2. If the greatest climate risk is Low (-1), the climate-adjusted baseline/status risk is the
greatest individual baseline/status risk minus one (e.g., from -3 to -2).
3. If the greatest climate risk is Moderate (-2) or High (-3), the climate-adjusted
baseline/status risk is equal to the greatest individual baseline/status risk.
4. Resilience baseline/status values (1 to 3) are ignored.
These rules are illustrated in Table 4-16 The climate-adjusted baseline/status risk value is used in
calculating the overall climate risk rather than the individual baseline/stature risk values. As
shown in Table 4-16, the climate-adjusted baseline/status risk factors increase the overall climate
risk only when the baseline/status risk is High (-3) and the greatest climate risk is Low (-1) or
82
-------
Moderate (-2). The values for the greatest climate risk, the greatest baseline/status risk, and the
climate-adjusted baseline/status risk are all output in the Vulnerability Summary (Appendix B).
Table 4-16. Climate-adjusted baseline/status risk values.
Values of the climate-adjusted baseline/status risk are based on the combination of the
greatest individual climate risk and the greatest individual baseline/status risk. Multiple
risks of the same value do not alter the calculation. The climate-adjusted
baseline/status risk is used in determining the overall risk for a species within an
ecoregion. Red = climate-adjusted baseline/status risk is greater than the greatest
climate risk, increasing overall risk. Blui = Climate-adjusted baseline/status risk same
as the greatest climate risk. Black = climate-adjusted baseline/status risk less than the
greatest climate risk.
Greatest Climate Risk
Greatest Baseline/Status risk
0
-1
-2
-3
0
0
0
0
0
-1
0
0
-1
-2
-2
0
-1
-2
-3
-3
0
-1
-2
-3
83
-------
Section 5.
Temperature Predictions
5.1 Introduction
Temperature is the public manifestation of climate change. To date, the bulk of the climate
change debate has focused on temperature increases as have policy discussions. As part of the
Copenhagen Accord, a general consensus was reached that the increase in global mean surface
air temperature should be limited to 2° C. Though there are concerns whether a cap of 2° C is
sufficiently protective (see examples in Portner et al., 2014), it has become a concrete rallying
point for the management of global climate change. A practical issue is the extent of emission
reductions required to stay under this cap. A recent summary of Coupled Model Inter-
comparison Project Phase 5 (CMIP5) results (Table 5-1) indicates that an emission scenario of
RCP 6.0 or 4.5 is required to reach this global air temperature goal. RCP 4.5 "is a stabilization
scenario in which total radiative forcing is stabilized shortly after 2100, without overshooting the
long-run radiative forcing target level" while RCP 6.0 scenario is a "stabilization scenario in
which total radiative forcing is stabilized shortly after 2100, without overshoot, by the
application of a range of technologies and strategies for reducing greenhouse gas emissions"
(Wayne, 2013). In comparison, RCP 8.5 is "A high scenario that assumes continued increases in
greenhouse gas emissions until the end of the 21st century" (Snover et al., 2013).
Table 5-1. CMIP5 annual mean surface air temperature anomalies (°C) from the 1986-2005 reference
period to 2081-2100 for the four RCPs.
Modified from Table 12.2 of Collins et al., 2013. "The multi-model mean ±1 standard deviation ranges
across the individual models are listed and the 5 to 95% ranges from the models' distribution (based on a
Gaussian assumption and obtained by multiplying the CMIP5 ensemble standard deviation by 1.64) are
given in brackets. Only one ensemble member is used from each model and the number of models differs
for each RCP"
Region
RCP 2.6
RCP 4.5
RCP 6.0
RCP 8.5
Global
1.0 ± 0.4 (0.3, 1.7)
1.8 ± 0.5 (1.1, 2.6)
2.2 ± 0.5 (1.4, 3.1)
3.7 ± 0.7 (2.6, 4.8)
Land
"Ocean
Tropics
Arctic
Antarctic
1.2 ± 0.6 (0.3, 2.2)
0.8 ± 0.4 (0.2, 1.4)
0.9 ±0.3 (0.3, 1.4)
2.2 ± 1.7 (-0.5, 5.0)
0.8 ± 0.6 (-0.2, 1.8)
2.4±0.6(1.3, 3.4)
1.5 ± 0.4 (0.9, 2.2)
1.6 ± 0.4 (0.9, 2.3)
4.2 ± 1.6 (1.6, 6.9)
1.5 ±0.7 (0.3, 2.7)
3.0 ±0.7 (1.8, 4.1)
1.9 ± 0.4 (1.1, 2.6)
2.0 ± 0.4 (1.3, 2.7)
5.2 ± 1.9 (2.1, 8.3)
1.7 ± 0.9 (0.2, 3.2)
4.8 ± 0.9 (3.4, 6.2)
3.1 ± 0.6 (2.1, 4.0)
3.3 ±0.6 (2.2, 4.4)
8.3 ± 1.9 (5.2, 11.4)
3.1 ± 1.2 (1.1, 5.1)
This section describes two approaches to predicting the effects of temperature increases on near-
coastal organisms, which we refer to as the Ecoregional Thermal Window (ETW) approach and
the Biogeographic Thermal Limit (BTL) approach. Both approaches generate ecoregion-specific
84
-------
risks for the NEP and U.S. Arctic, as well as allowing the assessment of different climate
scenarios. The two approaches make the basic assumption that biogeographic distributions
reflect the ecological thermal limits for a species. While a number of factors potentially affect a
species' vulnerability to temperature at a microscale (e.g., Helmuth et al., 2010), the bulk of the
evidence supports the contention that temperature is the overriding variable setting the
biogeographical range limits of most species as reviewed in Appendix D.
It is important to emphasize that the temperatures limiting species' ranges at an ecoregional scale
are not necessarily physiological thermal limits (e.g., CTmax). Besides the possibility of direct
thermal effects on adults, limitation of species in the warmer ecoregions could result from effects
on larval/juvenile stages, sublethal effects, such as reductions in fecundity, or indirect effects
such as changes in predator-prey relationships or loss of key ecosystem engineers (e.g.,
Weinberg et al., 2016; Lord et al., 2017). While analysis of temperature changes at an
ecoregional-scale does not identify the specific mechanism(s), it presumably captures the effects
on the population regardless of life-history stage or whether the effects are direct or indirect.
The ETW and BTL approaches use different methodologies to assign risk and different sources
for baseline temperature data. The ETW evaluates risk based on the range of sea surface
temperatures (SSTs) observed via remote sensing in the "warmest occupied ecoregion" (WOE)
(Figure 5-1), where "occupied" includes a classification of Present or any Level II or III
abundance classes. The BTL compares temperatures in the WOE to those observed in the "next
warmest unoccupied ecoregion" (NWUE), which is usually, but not always, the ecoregion
directly to the south (Figure 5-1). Use of two different ecological climate models was not
undertaken in the blind hope that there would be a complete one-to-one correspondence. Rather,
similarities in predictions provide greater confidence in the results while differences help identify
values and/or assumptions requiring additional research.
5.2 Future Temperature Predictions
Both approaches use NOAA's Climate Web Portal
(http://www.esrl.noaa.eov/psd/ipcc/ocn/ccwp.html; Scott et al., 2016) for the default projected
changes in temperature. Data served on the Climate Web Portal are based on the Coupled Model
Inter-comparison Project Phase 5 (CMIP5; Taylor et al., 2012; Bopp et al., 2013) that informed
the temperature predictions in the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change (Collins et al., 2013). Outputs from different models used in the CMIP5 are
interpolated to a 1-degree latitude/longitude grid to allow for intermodel comparisons.
85
-------
1 Next Coolest Unoccupied Ecoregion = CHUKCHI
Coolest Occupied Ecoregion = E. BERING
Warmest Occupied Ecoregion = NO. CALIFORNIA
Next Warmest Unoccupied Ecoregion = SO. CALIFORNIA
Figure 5-1. Distribution of Chionoecetes bairdi illustrating WOE, NWUE, COE,
and NCUE ecoregions.
WOE = warmest occupied ecoregion; NWUE = next warmest unoccupied
ecoregion; COE = coolest occupied ecoregion; NCUE = next coolest unoccupied
ecoregion. Purple shading indicates an occupied ecoregion.
The following options were chosen for the default temperature projections for SST, air
temperature, 30-m temperature, and 100-m temperature:
a. Historical period: 1956-2005 (1980/1981 average)
b. Future period: 2050-2099 (2074/2075 average)
c. RCP8.5
d. Average of all models
e. Statistic of change: Anomaly
f. Download entire year, summer (July-Aug.-Sept.), and winter (Jan.-Feb.-March) for
SST and air temperature
g. Download annual values for 30-m and 100-m depths
The number of models incorporated into the average predictions depends upon the parameter,
currently ranging from 37 for air temperature to 10 for subsurface water temperatures at 30 and
100 meters. The anomaly is the predicted difference in temperature between the future time
86
-------
period compared to the historical period. The anomaly is added to the historic baseline value to
generate the projected future temperature.
To generate ecoregion-scale temperature values, GIS was used to extract all the grids within each
ecoregion, and then the mean value of the climate parameter calculated within each of the
ecoregions. Details on the GIS techniques are given in Appendix E.
5.3 Ecoregional Thermal Windows Approach
5.3.1 ETW Approach
The ETW approach is based on comparing the projected SST in the target ecoregion to the
historic range of SST values observed in the warmest occupied ecoregion (WOE). Specifically,
the risk is determined by comparing the projected temperature in the target ecoregion to the
number of standard deviation (SD) units around the historic mean SST in the WOE, which is
based on 28 years of remote sensing SST data (Section 5.3.6). The rules to generate risk are:
a. Projected SST in target ecoregion < Historic mean SST + 1 SD in WOE = Minor risk
(temperatures normally experienced).
b. Projected SST in target ecoregion > Historic mean SST + 1 SD in WOE AND
< Historic mean SST + 2 SDs in WOE = Low risk (temperatures frequently
experienced).
c. Projected SST in target ecoregion > Historic mean + 2 SDs in WOE AND < Historic
mean + 3 SDs in WOE = Moderate risk (temperatures rarely experienced).
d. Projected SST in target ecoregion > Historic mean SST + 3 SDs in WOE = High risk
(temperatures very rarely if ever experienced in recent past).
Risks are analyzed for the annual average SSTs as well as for summer (July-August-September)
and winter (January-February-March) independently to gain insight into what season is likely to
be limiting.
We currently do not have ecoregion temperatures or projections in the tropics other than for the
Mexican Tropical Pacific (MTP) ecoregion. Thus, if a species occurs in any tropical ecoregion
globally, the temperatures in the MTP are used as the tropical WOE surrogate. A list of tropical
ecoregions is given in CBRAT. Examination of global baseline SST maps produced from
NOAA's Climate Web Portal indicates that the MTP is as warm as the other ecoregions
comprising the Eastern Tropical Pacific and Tropical Atlantic. However, the MTP appears to be
about 2°C cooler than much of the Indo-West Pacific and Indian Ocean. Use of temperatures in
the MTP for species that occur in these locations underestimates the actual upper thermal
window of these species. Consequently, it can overestimate the risk. Not many NEP species
occur in the Indo-West Pacific or Indian Ocean, but for those that do, users should evaluate the
87
-------
risk based on this possibility. A future improvement to CBRAT would be to identify a surrogate
ecoregion in the Indo-Pacific and/or Indian Ocean to be used as the default for species occurring
in these regions.
Table 5-2. Temperature ranges (°C) associated with different risk levels for ecoregion mean annual SSTs.
Analysis based on 28 years of AVHRR remote sensing data.
Ecoregion
Mean
Historical
Value
Minor Risk
Range
Low Risk
Range
Moderate Risk
Range
High Risk
range
Beaufort Sea -
Continental Coast/Shelf
0.03
<0.46
0.47-0.9
0.91 - 1.34
>1.35
Chukchi Sea
0.55
<1.23
1.24 - 1.92
1.93-2.61
>2.62
Eastern Bering Sea
3.75
<4.32
4.33-4.9
4.91 -5.47
>5.48
Aleutian Islands
5.67
<6.06
6.07-6.47
6.48-6.87
>6.88
Gulf of Alaska
7.42
<7.89
7.9-8.38
8.39-8.87
>8.88
North American Pacific
Fjordland
9.47
<9.92
9.93-10.38
10.39-10.84
>10.85
Puget Trough/Georgia
Basin
10.44
<10.93
10.94-11.43
11.44-11.94
>11.95
Oregon, WA, Vancouver
Coast/Shelf
11.51
<12.06
12.07-12.61
12.62-13.17
>13.18
Northern California
13.55
<14.16
14.17-14.78
14.79-15.4
>15.41
Southern California Bight
17.81
<18.39
18.4-18.99
19.0-19.58
>19.59
Magdalena Transition
22.61
<23.27
23.28-23.94
23.95-24.61
>24.62
Cortezian
24.79
<25.23
25.24-25.68
25.69-26.13
>26.14
Mexican Tropical Pacific
28.87
<29.22
29.23-29.58
29.59-29.94
>29.95
Table 5-3. Temperature ranges (°C) associated with different risk levels for ecoregion mean summer
SSTs.
Data source same as in Table 5-2.
Ecoregion
Mean
Historical
Value
Minor Risk
Range
Low Risk
Range
Moderate Risk
Range
High Risk
Range
Beaufort Sea -
Continental Coast/Shelf
1.77
<3.07
3.08-4.38
4.39-5.68
>5.69
Chukchi Sea
3.35
<4.81
4.82-6.27
6.28-7.74
>7.75
88
-------
Ecoregion
Mean
Historical
Value
Minor Risk
Range
Low Risk
Range
Moderate Risk
Range
High Risk
Range
Eastern Bering Sea
8.67
<9.64
9.65-10.62
10.63-11.6
>11.61
Aleutian Islands
8.44
<9.02
9.03-9.62
9.63-10.21
>10.22
Gulf of Alaska
11.82
<12.49
12.5-13.17
13.18-13.85
>13.86
North American Pacific
Fjordland
13.22
<13.94
13.95-14.66
14.67-15.38
>15.39
Puget Trough/Georgia
Basin
13.59
<14.92
14.93-16.26
16.27-17.6
>17.61
Oregon, WA, Vancouver
Coast/Shelf
14.12
<14.98
14.99-15.85
15.86-16.72
>16.73
Northern California
15.19
<15.96
15.97-16.74
16.75-17.52
>17.53
Southern California Bight
20.53
<21.4
21.41 -22.29
22.3-23.17
>23.18
Magdalena Transition
26.15
<27.22
27.23-28.3
28.31 -29.38
>29.39
Cortezian
30.22
<30.6
30.61 -30.99
31.0-31.38
>31.39
Mexican Tropical Pacific
30.49
30.91
30.92-31.34
31.35-31.77
>31.78
Table 5-4. Temperature ranges associated with different risk levels for ecoregion mean winter SSTs
(Jan.-Feb.-March).
Data source as in Table 5-2.
Ecoregion
Mean
Historical
Value
Minor Risk
Range
Low Risk
Range
Moderate Risk
Range
High Risk
Range
Beaufort Sea -
Continental Coast/Shelf
-1.26
<-0.76
-0.75--0.25
-0.24 - 0.26
>0.27
Chukchi Sea
-1.53
<-1.31
-1.3--1.07
-1.06--0.83
>-0.82
Eastern Bering Sea
1.0
<1.54
1.55-2.08
2.09-2.63
>2.64
Aleutian Islands
3.68
<4.15
4.16-4.64
4.65-5.13
>5.14
Gulf of Alaska
4.12
<4.73
4.74 - 5.34
5.35-5.95
>5.96
North American Pacific
Fjordland
6.68
<7.28
7.29-7.9
7.91 -8.52
>8.53
Puget Trough/Georgia
Basin
7.51
<7.94
7.95-8.38
8.39-8.82
>8.83
89
-------
Ecoregion
Mean
Historical
Value
Minor Risk
Range
Low Risk
Range
Moderate Risk
Range
High Risk
Range
Oregon, WA, Vancouver
Coast/Shelf
9.34
< 10.04
10.05-10.76
10.77-11.47
>11.48
Northern California
12.64
< 13.52
13.53-14.41
14.42-15.3
>15.31
Southern California
Bight
15.72
< 16.56
16.57-17.42
17.43-18.27
>18.28
Magdalena Transition
19.96
< 20.98
20.99-22.02
22.03-23.05
>23.06
Cortezian
19.57
< 20.57
20.58-21.58
21.59-22.59
>22.6
Mexican Tropical Pacific
27.53
< 27.97
27.98-28.42
28.43-28.87
>28.88
Table 5-5. Predicted increases in annual, summer, and winter SSTs for 2050-2099 based on the RCP 8.5
scenario (°C).
Predictions are based on an analysis of the CMIP5 climate models downloaded from the NOAA Climate
Web Portal (http://www.esrl.noaa.gov/psd/ipcc/ocn/ccwp.html').
Ecoregion
Annual Increase
Summer Increase
Winter Increase
Beaufort Sea - Continental
Coast/Shelf
2.29
5.55
0.16
Chukchi Sea
2.6
5.13
0.61
Eastern Bering Sea
3.56
4.03
2.92
Aleutian Islands
3.03
3.63
2.53
Gulf of Alaska
3.1
3.53
2.79
North American Pacific Fjordland
2.8
3.18
2.53
Puget Trough/Georgia Basin
2.15
3.12
1.8
Oregon, WA, Vancouver
Coast/Shelf
2.62
2.9
2.41
Northern California
2.54
2.83
2.34
Southern California Bight
2.4
2.38
2.34
Magdalena Transition
2.27
2.33
2.21
Cortezian
2.42
2.52
2.31
5.3.2 Conceptual Framework
The concept behind the ETW approach is that the historic temperatures in the WOE represent the
upper temperature range for the species to maintain a viable population. The species has
frequently experienced temperatures close to the WOE mean. However, the species has rarely, if
ever, experienced temperatures over ecological timeframes two or three standard deviations
warmer than the WOE mean. These higher temperatures become increasingly stressful until they
reach the mean temperature in the NWUE, a temperature at which the species no longer
90
-------
maintains a viable population. Thus, the temperature ranges in the WOE provide an ecological
thermal "window" on the upper temperature limits. Using the temperatures in the historically
warmest ecoregion to predict thermal ranges is analogous to the use of "environmental
matching" between a species' native range (donor region) and nonnative range (recipient region)
in invasive species risk assessments (e.g., Gollasch, 2006; Committee on Assessing Numeric
Limits for Living Organisms in Ballast Water, National Research Council, 2011). Though at a
different sampling scale, predicting suitable versus unsuitable temperatures is also analogous to
the incorporation of temperature in species' distribution models (SDMs) to predict range
changes.
It is possible that populations in more northern, cooler ecoregions consist of genotypes less
tolerant of warmer temperatures compared to those in the WOE (see Visser, 2008). To the extent
this occurs, using temperature ranges in the WOE may underestimate short-term risk in these
cooler ecoregions. However, over decades to a century, we assume that the more warm-adapted
genotypes within the northern ecoregion will replace the less tolerant genotypes. Additionally,
warm-tolerant genotypes in southern ecoregions may migrate northward. One line of support for
the potential for warm-adapted genotypes to migrate in relatively short time periods is the rapid
expansion of a whole host of nonindigenous species within decades (e.g., Sorte et al., 2010;
Pilgrim et al., 2013). A similar line of support is the rapid northward migration of marine species
in response to recent temperature increases (see Appendix D). Finally, many marine species have
a rapid rate of dispersal compared to terrestrial species (Kinlan and Gaines, 2003).
5.3.3 Within-Ecoregion Temperature Risks ("Worst-Case Scenario")
By comparing projected temperatures in the northern ecoregions to those in the WOE, the ETW
approach inherently assumes that genotypes adapted to the increased temperature either exist in
the northern ecoregion or will colonize from southern ecoregions. As mentioned above, there is
support that many if not most near-coastal species are able to migrate reasonably rapidly.
However, some species may have much slower migration rates either due to inherent properties
of the species or due to barriers. For example, species with short pelagic larval durations (PLD)
tend to have more genetic isolation-by distance, indicating less connectivity among the
populations (Selkoe and Toonen, 2011). Thus, species with short PLDs may colonize northern
ecoregions more slowly than those with longer PLDs. In cases with a lag in colonization, the
species' population may show an initial decline or other symptoms of stress but then recover as
the southern genotypes colonize.
To assess the risk associated with warm-genotypes either not existing or not colonizing in an
ecological relevant timeframe, we calculate thermal risks with the ETW approach using the
temperature range within the target ecoregion. For a species that ranges from the Gulf of Alaska
to Northern California, the projected temperature in the Gulf of Alaska is compared to the
historical temperature range in the Gulf of Alaska instead of the WOE (Northern California).
91
-------
Using the temperature within the ecoregion as the reference represents a "worst-case" scenario
for thermal risks. This within-ecoregion temperature risk is calculated using the annual, summer,
and winter SSTs, and is referred to as the "Within-Ecoregion Temperature" risk. Depending
upon a species' distribution, these risks can be substantially more sensitive, with a moderate risks
occurring in many cases with about a 1 °C increase. These annual and seasonal risks are output
in the Vulnerability Summary (Appendix B), but are not used in calculating the overall risk,
though the user is encouraged to evaluate whether they better capture the risks for a particular
species.
5.3.4 Abundance-Normalized Temperature Risks ("Ecosystem Services Risks")
Another method using the ETW approach is to define the thermal risks from the temperature in
the northernmost ecoregion at which there is a decrease in abundance. Figure 5-2 shows the
biogeographic abundance pattern for Hemigrapsus nudus, with the North Pacific Fjordland,
Puget Trough/Georgia Basin, and Oregon, Washington, Vancouver Coast and Shelf ecoregions
classified as abundant. Using the "abundance-normalized temperature" risks, the projected
temperatures in these ecoregions are compared to the historical range in the Northern California
Ecoregion, the northernmost ecoregion with a reduced Level II abundance. Similarly, the risk in
the Northern California Ecoregion is compared to the Southern California Bight Ecoregion,
which has a lower abundance (and is the WOE in this case). The abundance-normalized risks are
most informative when the primary interest is in reductions in ecosystem services, such as
commercial/recreational species, or with ecological dominants such as keystone species and
ecological engineers.
Hemigrapsus nudus is an ideal case for calculating abundance-normalized risks, and in many
cases the abundance patterns are "mixed" with a lower abundance near the center of the species'
distribution. Because such patterns make automation of the risks complicated and slow the risk
algorithm, we do not calculate these risks in CBRAT at this time. However, using the ecoregion
historical temperature ranges (Tables 5-2 to 5-5) and the abundance patterns from the
Vulnerability Summary or the Basic Export (see Lee et al., 2015), it is feasible for users to
calculate this risk by hand for a limited number of key species.
92
-------
Figure 5-2, Abundance pattern of Hemigrapsus nudus as example for
calculation of abundance-normalized temperature risks.
Projected temperatures in the three ecoregions where H. nudus is abundant
(Pacific Fjords, Puget, and Oregonian) are compared to the historical
temperature range in Northern California, the northernmost ecoregion with a
lower Level II abundance. Similarly, the risk for the Northern California is derived
by comparing the projected temperature to the historical range in the Southern
California Bight Ecoregion, which has a lower abundance. The color key to the
ecoregion relative abundance classes is given in Figure 4-1.
5.3.5 Evaluation of ETW Thermal Thresholds
A question in generating risk ranges was how much of a deviation from the mean ecoregion
temperature should constitute minor, low, moderate, and high risks? First, it was decided to use
ranges based on standard deviations rather than quartiles because standard deviations provide at
least an approximation of how frequently various temperature ranges might occur within an
ecoregion. Second, a two-tailed evaluation was used in generating the deviations around the
means because the same data can be used to evaluate deviations around the lower temperature
range to predict the potential for northward migration (Section 5.5).
Because organisms frequently experience temperatures within one SD of the mean, setting the
minor and low risks was relatively straight-forward. For the moderate and high risks, the issue
was whether the mean plus two SDs and the mean plus three SDs, respectively, were too
stringent. In evaluating the actual SSTs across all years and ecoregion (Table 5-6), the mean plus
two standard deviations criterion was exceeded 3.9% of the time. Thus, species in these
ecoregions only rarely experience temperatures this warm, which fits with our conception of a
moderate risk. In comparison, the mean plus three standard deviations was not exceeded in any
93
-------
of the 304 cases. These upper thresholds represent temperatures never, or very rarely,
experienced by the species in an ecological timeframe.
Since the historic temperature in the NWUE is sufficiently warm that the species is unable to
maintain a viable population, reaching or exceeding it could result in regional extirpation of the
species. Thus, as a check on using the mean plus three SDs as the high threshold, we compared
this value to the mean in the NWUE. In 9 of the 12 ecoregions, the mean plus three SDs was less
than the temperature in the next warmest ecoregion (Table 5-6), indicating that this threshold
identifies a high risk scenario though not necessarily one that would result in regional
extirpation. One of the three exceptions was the Beaufort, which may reflect both that the
Chukchi-Beaufort are aligned more longitudinally than latitudinally and the paucity of data in
these Arctic ecoregions. Another exception is Puget Trough/Georgia Basin, which may reflect
properties of inland seas and/or the influence of a terrestrial signal mixed with the ocean signal
(see Section 5.3.4). The last exception was the Pacific Fjords when compared to the
Puget/Georgia ecoregion. In this case, it might be more appropriate to compare it to the
Oregonian rather than to an inland sea, in which case the mean plus 3 SDs was below the next
warmest coastal ecoregion. In any case, exceeding the high risk threshold is predicted to have
measurable effects on populations in the affected ecoregion, though not necessarily regional
extirpation.
5.3.6 Data Source and Analysis
The Ecoregional Thermal Window approach relies on our previous analysis of nearshore SSTs in
the North Pacific and U.S. Arctic based on the Advanced Very High Resolution Radiometer
(AVHRR) data (Payne et al., 2011, 2012a, 2012b, and unpublished) for the historic mean and
standard deviations around the means. Raw data for the North Pacific are available in a USGS
Open File Report (Payne et al., 2011; https://pubs.usgs.eov/of/2.010/1251/) while the data for the
Arctic ecoregions are available in another report (Payne et al., 2012b;
http://pubs.uses. gov/of/2011/1246/).
Detailed GIS and data analysis procedures are given in Payne et al. (2011). Briefly, AVHRR data
from January, 1982 through December, 2009 were downloaded, generating 28 years of data. The
AVHRR-derived Pathfinder monthly-mean SSTs on 4x4 km grids were analyzed for each of the
North Pacific and Arctic ecoregions. Only high quality remote sensing data, according to
AVHRR data quality ranks, were used in the analysis. This quality criterion resulted in relatively
minor loss of data in the NEP ecoregions, but because of ice, cloud cover and fog, a majority of
the points did not meet the quality rank for inclusion in the Arctic ecoregions. This problem was
especially acute in the winter months and the analysis was limited to months with at least 10
points. For the NEP ecoregions, the analysis was limited to grids within 20 km of the coastline.
Because of the extensive loss of data points, the analysis for the Arctic included the entire
ecoregion and not just locations within 20 km of the coastline.
94
-------
The initial temperature analysis (Payne et al., 2012a) did not include the Puget Trough/Georgia
Basin or the Mexican Tropical Pacific ecoregions, both of which were subsequently analyzed
using the same procedures. Because of the 4x4 km grid size, it was not possible to remove all
grids with any terrestrial influence from the Puget Trough/Georgia Basin analysis.
Predicted annual, summer and winter SST increases for the 8.5 RCP scenario by ecoregion were
derived from CMIP5 downloaded from NOAA's Climate Web Portal
(http://www.esrl.noaa.eov/psd/ipcc/ocn/) (Appendix E). The predicted increases in each
ecoregion were added to the annual or seasonal historic mean SST for the ecoregion to generate
the projected ecoregion temperatures.
95
-------
Table 5-6. Number of exceedances of the moderate and high risk thresholds for annual SST based on the ETW approach.
The first two rows summarize the number of times the moderate and high risk thresholds for SST derived from standard deviation (SD) units were
exceeded. The third row summarizes the number of years used in the analysis. The number of years was reduced in the Beaufort and Chukchi
because of data loss due to ice, fog and clouds. In the next three rows, the mean + 3 SDs is compared to the temperature in the next warmest
ecoregion. In 9 of the 12 comparisons, the mean + 3 SDs is less than the next warmest ecoregion. NA = not applicable.
Years/Mean
Beaufort
Chukchi
Bering
Aleutians
GOA
Pac. Fjords
Puget
Oregonian
N CA
S CA
Magdalena
Cortezian
Sum of Yrs.
% of Yrs.
# years exceeding
mean + 3 SDs
0
0
0
0
0
0
0
0
0
0
0
0
0
0%
# years exceeding
mean + 2 SDs
1
1
2
1
0
1
0
1
1
2
2
0
12
3.9%
# Years in Analysis
9
17
28
27
28
28
27
28
28
28
28
28
304
NA
Mean SST (°C)
0.03
0.55
3.75
5.67
7.42
9.47
9.94
11.51
13.55
17.81
22.61
24.79
NA
NA
Mean + 3 SD (°C)
1.35
2.62
5.48
6.89
8.88
10.85
11.95
13.18
15.41
19.59
24.62
26.14
NA
NA
Mean SST in
Next Warmest
0.55
3.75
5.67
7.42
9.47
9.94
11.51
13.55
17.81
22.61
24.79
28.87
NA
NA
Ecoregion (°C)
96
-------
5.4 Biogeographical Thermal Limit Approach
5.4.1 Introduction
SSTs have frequently been used as indicators of temperature impacts of climate change (e.g.,
Hiddink et al., 2015; Molinos et al., 2016). However, deeper species experience different
temperature means and extremes, and may be exposed to a different relative rate of climate
change than shallow species. To evaluate how thermal risks potentially vary with the depth range
of the species, we developed the Biogeographical Thermal Limit (BTL) approach. The BTL
approach predicts temperature effects by comparing temperatures across ecoregions rather than
using observed temperature ranges within the warmest occupied ecoregion as with the ETW
approach. Specifically, the BTL thresholds are based on temperature bins between the historic
temperatures in the WOE and the NWUE (Figure 5-1) using values from the NOAA Climate
Web Portal. Use of bins is necessitated by the absence of multiple year data from the NOAA
Climate Web Portal, making the calculation of standard deviations around historic means
impossible. The main advantage is that climate risks can be evaluated for a range of species'
depth distribution, using temperatures from six different depths:
SST: General thermal stress
Annual Air Temperature: Intertidal thermal stress
Summer Air Temperature: Intertidal thermal stress
Winter Air Temperature: Intertidal thermal stress
30-m Temperature: Thermal stress in shallow subtidal (>0-30 m depth)
100-m Temperature: Thermal stress in deep subtidal (>30-200 m depth)
The primary assumption of the BTL is that the absence of a species in the NWUE is because one
or more of the temperature parameters is too warm for the species to maintain a viable
population. This assumption that the southern range limits are determined directly or indirectly
by temperature appears to be generally applicable, as discussed in Appendix D.
5.4.2 BTL Approach
As with the ETW approach, the ecoregion projected temperature is determined for each of the
temperature parameters by adding the predicted ecoregion-specific increase to the historic mean
(see Table 5-7 through Table 5-12). The BTL thermal effects thresholds are derived by dividing
the temperature difference between the historical WOE and NWUE into four equal bins (Figure
5-3). (Note that these are not quartiles since they are not derived from a distribution.) The BTL
effects thresholds are calculated for SSTs, and for air, 30-m, and 100-m temperatures. Then, the
97
-------
following base rules are used to generate the risk class from these bins:
Projected temperature minor risk
Projected temperature >lst bin and <2nd bin => low risk
Projected temperature >2nd bin and <3rd bin => moderate risk
Projected temperature >3rd bin => high risk
WOE (20 °C)
1st Bin (21°C)
Minor Risk
Low Risk
2nd Bin (22 °C)
3rd Bin (23 °C)
NWUE (24 °C)
Moderate Risk
High Risk
Figure 5-3. Schematic of the derivation of thermal risk values
with the BTL approach.
WOE = warmest occupied ecoregion; NWUE = next warmest
unoccupied ecoregion. Example given with an historical
temperature of 20 °C in the WOE and 24 °C in the NWUE.
Modifiers to these base rules and data are:
1) Downgrade the risk for the intertidal, shallow subtidal, or deep subtidal depths by
one risk class if the depth within these classes is classified as Observed rather than
Preferred.
a. As an example, if the risk was calculated as Moderate for the 30-m depth
using the base rules, but the organism had an Observed shallow subtidal depth
classification, the risk would be downgraded to Low.
98
-------
2) SST is not downgraded by depth preferences since it is an indicator of general
thermal stress.
3) If a species occurs in any tropical ecoregion, use the MTP as both the WOE and
the NWUE, assuming no other ecoregion is warmer than the MTP.
4) Ecoregions with an Error/Extinct or Transient classification for a species are not
included in the analysis.
5) No baseline data are available for the Puget Trough/Georgia Basin ecoregion for
the 30-m and 100-m temperatures, though values can be entered into CBRAT as
they become available.
a. For species whose southern range is the North American Pacific Fjordland
Ecoregion, the Oregon, Washington, Vancouver Coast and Shelf
Ecoregion is used as the NWUE.
Thermal risks for the population are calculated independently for each of the six depth-season
combinations, with the risks modified by the depth preferences of the species. A preferred depth
is taken as an indicator that a sizable portion of the total population occurs within that depth
range. Conversely, an observed depth class indicates that only a small fraction of the population
occurs within that depth range, and thus only a small portion of the total population would be at
risk. We account for these differences in population size by downgrading the risk by one class
when the depth class is observed versus preferred. The risk for each of these depth classes is
applied to the entire population within the ecoregion and it is possible for a species to have high
risk at one depth and a low risk at another. In this case, the species would be assigned a high risk.
The one exception to modifying the risks by depth are those resulting from the SSTs, which are
viewed as an overall predictor of temperature stress (e.g., Molinos et al., 2016).
Table 5-7. Historical and projected mean annual SSTs (°C).
Historical means derived from 1995-2005 baseline. Predicted increases are based on RCP 8.5 and 2050-
2099 future timeframe. Ecoregions are ordered by historic mean values. The historic means and
predicted increases are derived from the CMIP5 models downloaded from NOAA's Climate Change Web
Portal (http://www.esrl.noaa.gov/psd/ipcc/ocn/).
Ecoregion - Annual SSTs
SST Historical
Mean
SST Predicted
Increase
SST Projected
Temperature
Beaufort Sea - Continental Coast/Shelf
-1.07
2.29
1.22
Chukchi Sea
-0.73
2.6
1.87
Eastern Bering Sea
3.02
3.56
6.58
Aleutian Islands
5.95
3.03
8.98
Gulf of Alaska
6.95
3.10
10.05
North American Pacific Fjordland
9.77
2.80
12.57
Puget Trough/Georgia Basin
11.34
2.15
13.49
99
-------
Ecoregion - Annual SSTs
SST Historical
Mean
SST Predicted
Increase
SST Projected
Temperature
Oregon, WA, Vancouver Coast/Shelf
12.39
2.62
15.01
Northern California
16.37
2.54
18.91
Southern California Bight
20.67
2.40
23.07
Magdalena Transition
24.81
2.27
27.08
Cortezian
25.71
2.42
28.13
Mexican Tropical Pacific
27.64
2.52
30.16
Table 5-8. Historical and projected mean annual air temperatures (°C).
Data sources same as in Table 5-7. *Because of the resolution of the CMIP5 climate model, the air
temperatures and projections for the Puget-Georgia Ecoregion are averages of the values in the N Pac
Fijord Ecoregion and OR-WA-Vanc Ecoregion.
Ecoregion
Air Annual
Historical
Mean
Air Annual
Predicted
Increase
Air Annual
Projected
Temperature
Beaufort Sea - Continental Coast/Shelf
-11.66
8.34
-3.32
Chukchi Sea
-11.44
8.97
-2.47
Eastern Bering Sea
-0.67
5.56
4.89
Aleutian Islands
4.67
3.52
8.19
Gulf of Alaska
4.84
3.73
8.57
North American Pacific Fjordland
7.70
3.32
11.02
Puget Trough/Georgia Basin*
9.41
3.19
12.60
Oregon, WA, Vancouver Coast/Shelf
11.12
3.05
14.17
Northern California
15.32
2.88
18.2
Southern California Bight
19.38
2.78
22.16
Cortezian
23.22
3.28
26.5
Magdalena Transition
23.32
2.65
25.97
Mexican Tropical Pacific
26.15
2.78
28.93
Table 5-9. Historical and projected mean summer air temperatures (°C).
Data sources same as in Table 5-7. *Because of the resolution of the CMIP5 climate model, the air
temperatures and projections for the Puget-Georgia Ecoregion are averages of the values in the N Pac
Fijord Ecoregion and OR-WA-Vanc Ecoregion. The historical winter air temperatures based on the
average of these two ecoregions were 5.45 °C (RCP 8.5 model) and 5.95 °C (RCP 4.5 model) compared
to 6.2 °C based on the average of winter values from six NOAA buoys in the Puget Sound/Georgia Basin
Ecoregion
Air Summer
Historical
Mean
Air Summer
Predicted
Increase
Air Summer
Projected
Temperature
Chukchi Sea
1.36
5.04
6.40
Beaufort Sea - Continental Coast/Shelf
1.69
5.10
6.79
100
-------
Ecoregion
Air Summer
Historical
Mean
Air Summer
Predicted
Increase
Air Summer
Projected
Temperature
Eastern Bering Sea
6.87
4.28
11.15
Aleutian Islands
9.24
3.88
13.12
Gulf of Alaska
10.17
3.79
13.96
North American Pacific Fjordland
12.77
3.52
16.29
Puget Trough/Georgia Basin*
14.24
3.49
17.73
Oregon, WA, Vancouver Coast/Shelf
15.7
3.46
19.16
Northern California
19.19
3.15
22.34
Southern California Bight
23.02
2.81
25.83
Magdalena Transition
26.42
2.69
29.11
Cortezian
27.56
3.36
30.92
Mexican Tropical Pacific
27.60
3.02
30.62
Table 5-10. Historical and projected mean winter air temperatures (°C).
Data sources same as in Table 5-7. *Because of the resolution of the CMIP5 climate model, the air
temperatures and projections for the Puget-Georgia Ecoregion are averages of the values in the N Pac
Fijord Ecoregion and OR-WA-Vanc Ecoregion. The historical winter air temperatures based on the
average of these two ecoregions were 5.45 °C (RCP 8.5 model) and 5.95 °C (RCP 4.5 model) compared
to 6.2 °C based on the average of winter values from six NOAA buoys in the Puget Sound/Georgia Basin
Ecoregion
Air Winter
Historical Mean
Air Winter
Predicted
Increase
Air Winter
Projected
Temperature
Beaufort Sea - Continental Coast/Shelf
-26.14
10.96
-15.18
Chukchi Sea
-25.52
12.56
-12.96
Eastern Bering Sea
-8.56
7.73
-0.83
Gulf of Alaska
0.52
3.89
4.41
Aleutian Islands
1.09
3.33
4.42
North American Pacific Fjordland
3.45
3.33
6.78
Puget Trough/Georgia Basin*
5.45
3.10
8.54
Oregon, WA, Vancouver Coast/Shelf
7.44
2.86
10.30
Northern California
12.12
2.66
14.78
Southern California Bight
16.31
2.71
19.02
Cortezian
19.16
3.10
22.26
Magdalena Transition
20.62
2.58
23.20
Mexican Tropical Pacific
24.47
2.53
27.00
101
-------
Table 5-11. Historical and projected mean 30-m temperatures (°C).
Data sources same as in Table 5-7. ND = No data.
Ecoregion
30-m Historical
Mean
30-m Predicted
Increase
30-m Projected
Temperature
Beaufort Sea - Continental Coast/Shelf
-1.40
0.33
-1.07
Chukchi Sea
-1.09
0.33
-0.76
Eastern Bering Sea
1.75
2.82
4.57
Aleutian Islands
4.59
2.54
7.13
Gulf of Alaska
5.27
3.07
8.34
North American Pacific Fjordland
8.11
2.89
11.00
Oregon, WA, Vancouver Coast/Shelf
10.35
2.74
13.09
Northern California
14.94
2.51
17.45
Southern California Bight
19.55
2.51
22.06
Magdalena Transition
23.79
2.31
26.10
Cortezian
24.92
2.40
27.32
Mexican Tropical Pacific
26.88
2.43
29.31
Puget Trough/Georgia Basin
ND
ND
ND
Table 5-12. Historical and projected mean 100-m temperatures (°C).
Data sources same as in Table 5-7. ND = no data.
Ecoregion
100-m Historical
Mean
100-m Predicted
Increase
100-m Projected
Temperature
Beaufort Sea - Continental Coast/Shelf
-1.28
0.82
-0.46
Chukchi Sea
-1.43
0.75
-0.68
Eastern Bering Sea
1.89
3.10
4.99
Aleutian Islands
3.74
2.33
6.07
Gulf of Alaska
4.95
2.77
7.72
North American Pacific Fjordland
7.32
2.36
9.68
Oregon, WA, Vancouver Coast/Shelf
9.23
1.95
11.18
Northern California
11.96
1.54
13.5
Southern California Bight
15.15
1.14
16.29
Mexican Tropical Pacific
17.28
1.44
18.72
Magdalena Transition
17.29
0.89
18.18
Cortezian
17.38
1.89
19.27
Puget Trough/Georgia Basin
ND
ND
ND
102
-------
5.4.3 Comparison of ETW and BTL
Use of the temperature bins does not have a theoretical statistical underpinning as does the use of
standard deviations around the mean. However, bins measure the delta between a suitable
temperature (i.e., WOE) and a presumably unsuitable temperature (NWTJE) and thus provide a
first-order estimate of the risk. To evaluate the efficacy of the thresholds based on quantiles, we
compared the BTL predictions against those from the ETW approach for brachyuran crabs
(Table 5-13), using the results from the ETW as the standard.
From the Beaufort through Southern California, the BTL and ETW approaches agreed 87% of
the time. In 49 of the 50 cases when there was a disagreement, the ETW predicted a high risk
versus a moderate risk with the BTL. In only one case was the risk classification two levels
apart. The majority of the differences occurred in Southern California, but there was still 70%
agreement in this ecoregion. This level of concordance gives us confidence in using the BTL
approach to evaluate temperature risks associated with air and in subsurface waters. One
possibility to address the potential for the BTL to underestimate risk is to combine the moderate
and high risks into a "species at risk" classification.
The BTL did not perform well in the two southern ecoregions. Even though there was a high
correspondence between BTL and the ETW in the Magdalena, the BTL overestimated risk by
two levels in all the discrepancies. There was very poor correspondence between the two
methods in the Cortezian, and all the differences were three risk levels (minor vs. high risk).
Because of the high degree of overestimating risk, the BTL approach should not be applied to
these two ecoregions.
Table 5-13. Comparison of risk predictions using the ETW versus the BTL approaches.
Preliminary analysis of the temperature risks of brachyuran crabs by ecoregion using the ETW and BTL
approaches. The ETW results are taken as the standard. BTL risk classes less than the ETW values
underestimate risk; classes greater than the ETW values overestimate risk.
Ecoregions
#
Predictions
# ETW =
BTL
# ETW
(high)
vs. BTL
(moderate)
# ETW
(moderate)
vs. BTL
(minor)
# ETW
(minor)
vs. BTL
(moderate)
# ETW
(minor)
vs. BTL
(high)
Beaufort to
Southern
California
394
344
(87.3%)
49
1
0
0
Southern
California
131
92
(70.2%)
39
0
0
0
Magdalena
138
114
(82.6%
0
0
24
0
Cortezian
295
64
(21.7%)
0
0
0
231
103
-------
5.5 Northern Colonization
One of the likely effects of warming is the northern colonization of southern species (e.g., Jones
et al., 2012; Somero, 2012; Cahill et al., 2013). To evaluate the potential for such colonization,
we reversed the logic of the BTL approach to evaluate whether temperatures in unoccupied
northern (cooler) ecoregions will be become sufficiently warm to allow colonization. We choose
the BTL approach over the ETW to allow assessments for air, 30-m depth and 100-m depth
temperatures as well as SSTs, with the assumption that colonization would be unlikely if any
occupied temperature range was not suitable.
This analysis requires two new definitions (Figure 5-1):
COE = Coolest occupied ecoregion; usually the northernmost occupied ecoregion.
NCUE = Next coolest unoccupied ecoregion; usually just north of the COE.
The logic to predict the thermal suitability of northern unoccupied ecoregions is analogous to the
approach for predicting thermal risks. The projected future temperature in each of the
unoccupied northern ecoregions is compared to the bins derived from the historical temperatures
in the COE and NCUE, but the rules predict thermal suitability rather than risk:
Projected temperature in target ecoregion > 3rd bin of COE and NCUE => 3 (high suitability)
Projected temperature in target ecoregion < 3rd bin of COE and NCUE & >2nd bin => 2
(moderate suitability)
Projected temperature in target ecoregion <2nd bin of COE and NCUE & >lst bin => 1 (low
suitability)
Projected temperature in target ecoregion 0 (minor suitability)
The suitabilities are modified by depth similar to the risk calculations, with suitabilities
downgraded by one level if the species has an observed versus preferred depth class. The Puget
Sound/Georgia Basin Ecoregion did not have values for the 30-m and 100-m depth strata.
Additionally, because of the resolution of CMIP5 model, we used the average of the air
temperatures in the North Pacific Fijord and OR-WA-Vanc ecoregions. The BTL approach
should not be used in predicting colonization potential into the Magdalena or Cortezian
ecoregions given the issues with this approach in these areas.
High and moderate suitabilities in unoccupied ecoregions indicate that temperature is not likely
to be a limiting factor to northern colonization. However, the analysis does not evaluate whether
other factors, such as lack of suitable habitat or dispersal limitation, could limit a species' spread.
An example is an obligate coral specialist that would be unable to colonize northern ecoregions
104
-------
until corals colonized them. These factors should be evaluated for species with temperatures
suitable for colonization of northern ecoregions.
105
-------
Section 6.
Ocean/Coastal Acidification
6.1 Background
It has become increasingly clear over the last decade that ocean acidification is a major threat to
the biodiversity of marine/estuarine ecosystems and the associated ecosystem services. (Note we
use the term "ocean acidification" for consistency with the literature but the risk framework is
focused on "coastal acidification", which can have different dynamics than true oceanic systems;
see Waldbusser and Salisbury, 2014.) These threats have been expounded upon in a number of
reviews (e.g., Portner, 2008; Hendriks et al., 2010; Byrne, 2011; Ross et al., 2011; Whiteley,
2011; Bellard et al., 2012; Wicks and Roberts, 2012; Kroeker et al., 2013; Wittmann and Portner
2013; Waldbusser and Salisbury, 2014; Mathis et al., 2015; Ross et al., 2016; Foo and Byrne,
2016). However, even with these recent efforts, ocean acidification is the least understood of the
three climate drivers addressed in CBRAT. While it is beyond the scope of this document to
provide a detailed review of the uncertainties associated with ocean acidification, the following
highlights some of the challenges, from measurement to interpreting exposure studies:
• pH in marine waters has been measured using four different scales, which can differ by
more than 0.1 pH unit (see
http://pmel.noaa. gov/co2/story/Qualitv+of+pH+Measurements+in+the+>
chives). This affects both the intercomparability of laboratory studies as well as use of
historic pH measurements.
• Most high resolution pH measurements have been taken in the ocean or coastal waters,
with relatively few within estuaries. Further, current regional projections for pH and
aragonite saturation state (Qa) from the Gulf of California to the Beaufort are at too
coarse a scale to model estuaries.
• pH and aragonite saturation state both tend to decrease with depth, though the pattern
with depth varies geographically. However, regional projections for pH and aragonite
saturation state from the Gulf of California to the Beaufort are only available for surface
waters.
• A wide range of physiological, behavioral, and survival end-points, and exposure
durations have been used to evaluate impacts on eggs, larvae, juveniles, and adults. This
lack of uniformity makes it challenging to compare among studies and taxa.
106
-------
• Ninety-five percent of the ocean acidification laboratory studies reviewed up to 2014
"had interdependent or non-randomly interspersed treatment replicates, or did not report
sufficient methodological details" (Cornwall and Hurd, 2016).
• There is an inadequate understanding of the factors conveying resiliency to certain taxa,
such as their evolutionary history of exposure to low pH (e.g., estuarine species).
• In addition to direct effects, ocean acidification may impact food webs and trophic
interactions (Haigh et al., 2015).
• Ocean acidification may interact with other stressors, in particular hypoxia and
temperature increases (Harvey et al., 2013; Breitburg et al., 2015).
We identify these sources of uncertainty not to be disheartening but to provide a perspective on
the state-of-the-science. While recognizing these uncertainties, we believe there is sufficient
information to generate a first-order regional-scale risk assessment of ocean acidification.
Besides providing an overview of the patterns of risk, the risk assessment provides an ancillary
benefit of providing a framework to organize the data and to identify major information gaps.
However, the regional-scale risk assessments conducted in CBRAT are not sufficient to manage
fisheries within a locality, which are better addressed by higher resolution models (e.g., Mathis et
al., 2015; Punt et al., 2016).
The steps in the risk analysis are: 1) generate ecoregion-scale historical baseline values for pH
and aragonite saturation state; 2) generate ecoregion-scale projections for future pH and
aragonite saturation states; 3) identify high, moderate, and low sensitivity classes of species
within a taxon; 4) generate pH and aragonite saturation state thresholds for minor, low,
moderate, and high risks for each of the sensitivity classes; 5) identify whether pH, aragonite
saturation state, or both are the major stressor for each species; and 6) overlay the appropriate
tax on-specific threshold on the projected pH or aragonite saturation value to generate an
ecoregion-specific risk for each species. To evaluate the extent of the uncertainties in these
values, CBRAT was designed to allow users to change baseline and projected pH and aragonite
saturation state values, the sensitivity class of a species, threshold values for pH and aragonite
saturation state, and whether reductions in pH or aragonite saturation is the major stressor.
6.2 Background and Projected pH and Aragonite Saturation State (Qa) Values
6.2.1 pH Values
Both the historic mean sea surface pH values and the predicted sea surface pH values are from
the CMIP5 model downloaded from NOAA's Climate Web Portal
(fattp://www.esrl.noaa.eov/psd/ipcc/ocn/). As mentioned, the CM IPS was used in the Fifth IPCC
Report (Collins et al., 2013). Outputs from different models used in the CMIP5 are interpolated
107
-------
to a 1-degree latitude/longitude grid to allow for intermodel comparisons. The same default
settings were used for pH as for temperature:
a. Historical period: 1956-2005 (1980/1981 average)
b. Future period: 2050-2099 (2074/2075 average)
c. RCP8.5
d. Average of all models
e. Statistic of change: Anomaly
f. Download entire year, summer (July-Aug.-Sept.), and winter (Jan.-Feb.-March)
GIS was used to clip the gridded data to the MEOW ecoregion borders and then the mean pH
was calculated within each of the ecoregions. Details on the GIS techniques are given in
Appendix E. The CMIP5 models output pH in the total pH scale (log of total hydrogen ion
concentration; personal communication from James Scott (NOAA) to Henry Lee, 9/1/2016). The
default values for historic and projected pH values are given in Table 6-1 to Table 6-3.
6.2.2 Aragonite Saturation State (Qa) Values
NOAA's Climate Change Web Portal does not provide aragonite saturation state projections.
Therefore, we followed Foden et al. (2013) in their analysis of hermatypic corals and use the
aragonite saturation state projections developed by Cao and Caldeira (2008) based on the
University of Victoria Earth System Climate Model version 2.8. The aragonite saturation values
were provided by Dr. Cao (Long Cao on 7/19/2014 to Henry Lee). Simulation results were
provided for the 2010 (baseline), 2050 and 2100 for RCPs 2.6, 4.4, 6.0, and 8.5. We focus on
RCP 8.5 and 2100 scenario. The model has a resolution of 1.8 degrees latitude by 3.6 degrees
longitude. As with the NOAA temperature data, GIS was used to derive a weighted average for
each ecoregion (Appendix E). As pointed out by Cao and Caldeira (2008), the resolution of the
model is too coarse to resolve aragonite values in coastal regions though "changes in coastal
ocean chemistry should largely track corresponding changes in nearby open ocean waters." The
baseline and projected aragonite saturation state values are given in Table 6-4.
108
-------
Table 6-1. Historical and projected annual pH.
Historical means derived from 1995-2005 baseline. Predicted increases are based on RCP 8.5 and 2050-
2099 future timeframe. The historical means and predicted increases are derived from the CMIP5 models
downloaded from NOAA's Climate Change Web Portal (Wtpi//www.8srl.noaa.gov/psd/ipec/oen/). pH
reported in total pH scale. ND = no data.
Ecoregion
Historical
Annual pH
Predicted
Annual pH
Decline
Projected
Annual pH
Beaufort Sea - continental coast and shelf
8.12
-0.35
7.77
Chukchi Sea
8.11
-0.33
7.78
Eastern Bering Sea
8.11
-0.3
7.81
Aleutian Islands
8.09
-0.28
7.81
Gulf of Alaska
8.11
-0.3
7.81
North American Pacific Fjordland
8.1
-0.3
7.8
Puget Trough/Georgia Basin
ND
ND
ND
Oregon, Washington, Vancouver Coast and
Shelf
8.11
-0.3
7.81
Northern California
8.1
-0.27
7.83
Southern California Bight
8.09
-0.27
7.82
Magdalena Transition
8.09
-0.26
7.83
Cortezian
8.11
-0.25
7.86
Table 6-2. Historical and projected summer pH.'
Data sources same as in Table 6-1. ND = no data.
Ecoregion - Summer pH
Historical
Summer pH
Predicted Summer
pH Decline
Projected
Summer pH
Beaufort Sea - continental coast and shelf
8.15
-0.38
7.77
Chukchi Sea
8.15
-0 36
7.79
Eastern Bering Sea
8.1
-0.3
7.8
Aleutian Islands
8.1
-0.28
7.82
Gulf of Alaska
8.09
-0.29
7.8
North American Pacific Fjordland
8.07
-0.28
7.79
Puget Trough/Georgia Basin
ND
ND
ND
Oregon, Washington,
Vancouver Coast and Shelf
8.08
-0.28
7.8
Northern California
8.07
-0 27
7.8
Southern California Bight
8.04
-0 26
7.78
Magdalena Transition
8.05
-0.25
7.8
Cortezian
8.07
-0.25
7.82
109
-------
Table 6-3. Historical and projected winter pH.
Data sources same as in Table 6-1. ND = no data.
Ecoregion - Winter pH
Historical
Winter pH
Predicted Winter
pH Decline
Projected
Winter pH
Beaufort Sea - continental coast and shelf
8.1
-0.33
7.77
Chukchi Sea
8.09
-0.31
7.78
Eastern Bering Sea
8.1
-0.3
7.8
Aleutian Islands
8.07
-0.28
7.79
Gulf of Alaska
8.11
-0.3
7.81
North American Pacific Fjordland
8.12
-0.31
7.81
Puget Trough/Georgia Basin
ND
ND
ND
Oregon, Washington, Vancouver Coast and
Shelf
8.13
-0.3
7.83
Northern California
8.13
-0.28
7.85
Southern California Bight
8.12
-0.27
7.85
Magdalena Transition
8.13
-0.26
7.87
Cortezian
8.15
-0.25
7.9
Table 6-4. Historical and projected aragonite saturation state values.
Historical means derived from 2010. Predicted ecoregion decreases in aragonite saturation based on
RCP 8.5 and 2100 future timeframe. Values based on Cao and Caldeira (2008) analysis using the
University of Victoria Earth System Climate Model version 2.8.
Ecoregion - Annual aragonite
saturation state
Historical
Aragonite
Saturation State
Predicted Aragonite
Saturation State
Decline
Projected
Aragonite
Saturation State
Beaufort Sea - continental coast and
shelf
1.49
-0.84
0.65
Chukchi Sea
1.44
-0.77
0.67
Eastern Bering Sea
2.33
-0.88
1.45
Aleutian Islands
1.86
-0.85
1.01
Gulf of Alaska
1.97
-0.87
1.1
North American Pacific Fjordland
2.22
-1.0
1.22
Puget Trough/Georgia Basin
2.44
-1.02
1.42
Oregon, Washington, Vancouver Coast
and Shelf
2.4
-1.03
1.37
Northern California
2.48
-1.11
1.37
Southern California Bight
2.63
-1.18
1.45
Magdalena Transition
2.56
-1.25
1.31
Cortezian
2.61
-1.22
1.39
110
-------
6.3 Toxicology Approach to Establishing pH Effects Thresholds
6.3.1 Introduction
We propose that pH and aragonite saturation state effects thresholds can be generated using
approaches derived from toxicology. We provide the specifics for decapods with pH but the
approach would be analogous for other taxa and for aragonite saturation state. The initial step is
to synthesize the field and experimental exposures for a taxon. Such a summary for the decapod
is presented in Table 6-5; the full summary included additional parameters such as the pH scale
and life history traits of the test species (available from the authors). At this time, 34 studies
covering 25 decapod species have been synthesized, which will be updated as new information
becomes available. An examination of Table 6-5 reveals that a wide range of behavioral,
calcification, development, genetics, mortality, and physiological endpoints have been evaluated.
An initial evaluation of other taxa also indicate a wide range of endpoints were used in these
exposure experiments.
Several ocean acidification reviews have highlighted the differences in sensitivity among major
taxa, with the crustaceans less sensitive than corals, mollusks, or echinoderms (e.g., Kroeker et
al., 2013; Wittmann and Portner, 2013). However, examination of Table 6-5 indicates that there
also is a wide range of sensitivities among species within a taxon. Zoea of the European lobster,
Homarus gammarus, had nearly 50% less calcium in the carapace at a pH of 8.1 compared to the
control of 8.39 (Arnold et al., 2009), while the commercially important southern Tanner crab,
Chionoecetes bairdi, had a reduced survival at a pH of 7.8 compared to the control at 8.1
(Swiney et al., 2016). In contrast to these species, the burrowing shrimp, Upogebia deltaura, did
not experience significant mortality at a pH of 7.35 (Donohue et al., 2012).
Since one of our objectives is to predict differences in risk among species, we focused on
developing a framework capable of capturing such within-taxon differences. Because species-
specific thresholds are available for only a handful of species, our approach is to develop
thresholds for three classes: high, moderate, and low sensitivity species with each sensitivity
class having a different set of effects thresholds. While not as precise as species-specific
sensitivity levels, the three classes should provide sufficient resolution to identify the major
patterns. The challenge with three sensitivity classes is the number of threshold values required.
Three threshold values are needed to separate the four risk levels (minor, low, moderate, and
high) within each sensitivity class, so a total of nine threshold values are required.
6.3.2 Use of Maximum Acceptable Toxicant Concentrations (MATCs) to Generate
Comprehensive Effects Thresholds
Results from the pH exposure experiments are amenable to deriving multiple thresholds using
the "maximum acceptable toxicant concentration" (MATC). The MATC (or GMATC) is the
geometric mean of the "no observed adverse effects level" (NOAEL), the highest test level for
111
-------
which effects are not statistically different from the lowest effect concentration, and the "lowest
observed adverse effects level" (LOAEL), the lowest level at which the effects were significantly
different than the controls or non-significant exposure concentration (U.S. EPA Risk Assessment
Forum, 1998). Note that the NOAEL and LOAEL are defined as the "lowest" contaminant
concentrations, but for pH they are for the highest no effect pH and highest significant effect pH,
respectively. Because of their long-standing use, we continue to use the terms for pH. Also,
because pHs are in logio units, MATCs are calculated using the antilogs of the pHs and then back
transformed into a pH by taking the logio.
MATCs are used to estimate the "safe" concentration of a contaminant and have been used in
ecological risk assessments (U.S. EPA, 1998), in evaluating soil contamination for Superfund
(EPA, 2003), evaluating contaminants under the Toxic Substances Control Act (TSCA)
(Nabholz, 1991), and pesticides under the Federal Insecticide, Fungicide, and Rodenticide Act
(FIFRA) (EPA, 2004; Fairchild et al., 2009). MATCs have been derived in two ways; firstly, in a
more restricted sense limited to results from chronic tests and secondly, in a more general sense
based on less than chronic tests. Because of the limitations of the pH exposure data, we use
MATC in the general sense, and calculate the "comprehensive" threshold using the lowest pH
resulting in a significant effect, the LOAEL, for each species regardless of the endpoint or
exposure duration. The MATC was then calculated using the control pH or the lowest non-
significant pH value, whichever was lower. A single MATC was calculated for each species
using the most sensitive response, which generated 21 MATC values, ranging from a pH of 7.35
to 7.96 (Table 6-6).
While thresholds based on a range of mortality, physiological, and behavioral endpoints is not
ideal, aggregating results from multiple types of endpoints has a long history in toxicology. The
"effects range - median" (ERM) is the median sediment concentration resulting in any
significant effect based on all available information, ranging from bacterial responses
(Microtox™) to laboratory and field studies (Long et al., 1995). ERMs have been used
extensively to compare the toxicity among contaminants and to evaluate spatial patterns of
sediment toxicity (e.g., Belan, 2003; Hyland et al., 2004; Nelson et al., 2005; Hale and Heltshe,
2008; Dasher et al., 2015). We posit that a similar aggregation of results across different
endpoints is the best practical approach to synthesizing pH effects until ocean acidification
exposures and endpoints are at least quasi-standardized. However, in Section 6.3.3, we present
an approach limited to endpoints directly related to population viability.
After synthesizing the MATC values, the next step is to identify the low, moderate, and high
sensitive classes by identifying breaks in the cumulative frequency distribution of the MATCs
(Figure 6-1). There is a discrete break between 7.75 and 7.60 pH. Though not as discrete, there
was another break between 7.87 and 7.80. These breaks identify three sets of species based on
sensitivity. In taxa lacking discrete breaks in the cumulative distribution curve, sensitivity classes
112
-------
can be identified by dividing the data into thirds or by setting the break points at the upper and
lower quantiles (Figure 6-1).
After identifying the three sensitivity groups, three values within each group need to be
identified to generate the cutpoints for the four risk classes (minor, low, moderate and high)
within each sensitivity group. The threshold for minor risks for each sensitivity group is set equal
to or greater than the highest pH within each group. The threshold for high risks is set equal or
less than the lowest pH within each of the three groups. The cutpoint between the low and
moderate risks is calculated as the median of the values within each of the sensitivity groups.
Specifically, the median is set as the upper end of the range for moderate risks. For example,
within the high sensitivity group, the highest MATC was 7.96, which was set to the minor risk
threshold (Table 6-7). The lowest MATC, 7.87, was used as the high risk threshold. The median
of all the values within the high sensitivity groups was 7.90, which was set as the higher value
within the moderate risk range. From these cutpoints, it was then possible to define the low risk
range as 7.91 to 7.95.
The risks generated in this fashion should be reasonably protective since the MATCs are based
on the most sensitive significant response for each species. For example, the juveniles of both the
red king crab (Paralithodes camtschaticus) and the southern Tanner crab (Chionoecetes bairdi)
displayed significant mortality at pH 7.8 compared to the control at 8.0 (Long et al., 2013). Both
of these commercial crabs are classified as sensitive, and any pH equal to or less than 7.87 is
considered a high risk and a pH less than 7.9 is classified as a moderate risk. Exposure of the
larvae of the moderately sensitive Dungeness crab (Metacarcinus magister) to a pH of 7.5
resulted in reduced survival compared to a pH of 8.0 (Miller et al., 2016). This results in a
MATC of 7.75, which is at the threshold for a high risk for a moderate sensitivity species. While
the values appear protective, the proposed thresholds are considered preliminary because of the
limited number of exposure experiments and the lack of consistency in experimental procedures
and endpoints.
113
-------
Table 6-5. Summary of pH exposure experiments with decapods.
Taxa: Anom. = Anomuran; Asta. = Astacidea; Brae. = Brachyura; Cari. = Caridea; Gebi. = Gebiidea; Pena. = Penaeoidea.
Sources: 1 = Swiney et al., 2016; 2 = Long et al., 2016; 3 = Meseck et al., 2016; 4 = Rastrick et al., 2014; 5 = Taylor et al., 2015; 6 = Dodd et al., 2015; 7 =
Landes and Zimmer, 2012; 8 = Appelhans et al., 2012; 9 = de la Haye et al., 2011; 10 = Miller et al., 2016; 11 = Donohue et al., 2012; 12 = Long et al.,
2013; 13 = Kurihara et al., 2008; 14 = Paganini et al., 2014; 15 = Ceballos-Osuna et al., 2013; 16 = Walther et al., 2009b; 17 = Dissanayake and
Ishimatsu, 2011; 18 = Kim et al., 2015; 19 = Agnalt et al., 2013; 20 = Keppel et al., 2012; 21 = Arnold et al., 2009; 22 = Small et al., 2010; 23 = Bechmann
et al., 2011; 24 = Christmas et al., 2013; 25 = Fehsenfeld et al., 2011; 26 = Hans et al., 2014; 27 = Spicer et al., 2007; 28 = Metzger et al., 2007; 29 = Ries
et al., 2009; 30 = Carter et al., 2013; 31 = Arnberg et al., 2013; 32 = Miller et al., 2014; 33 = Small et al. 2016; 34 = Styf et al., 2013; 35 = Walther et al.,
2010; 36 = Schiffer et al., 2014. d = day; w = week; m = month. ND = No data.
Species
Taxon
X
Q.
Arag.
Saturation
"i
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Callinectes
sapidus
Brae.
8.03
2.13
409
Control
60 d
24.9
Juvenile
Survival rate
75%
Mortality
Control
29
Callinectes
sapidus
Brae.
7.85
1.53
606
Exposed
60 d
25
Juvenile
Survival rate
83%
Mortality
Not sig.
29
Callinectes
sapidus
Brae.
7.72
1.13
903
Exposed
60 d
25
Juvenile
Survival rate
75%
Mortality
Not sig.
29
Callinectes
sapidus
Brae.
7.31
0.47
2856
Exposed
60 d
25.1
Juvenile
Survival rate
67%
Mortality
Not sig.
29
Callinectes
sapidus
Brae.
8.03
2.13
409
Control
60 d
24.9
Juvenile
Calcification
rate
434%
Calcification
Control
29
Callinectes
sapidus
Brae.
7.85
1.53
606
Exposed
60 d
25
Juvenile
Calcification
rate
598%
Calcification
Sig.?
29
Callinectes
sapidus
Brae.
7.72
1.13
903
Exposed
60 d
25
Juvenile
Calcification
rate
601%
Calcification
Sig.?
29
Callinectes
sapidus
Brae.
7.31
0.47
2856
Exposed
60 d
25.1
Juvenile
Calcification
rate
724%
Calcification
Sig.?
29
114
-------
Carcinus
maenas
Carcinus
maenas
Cancer
pagurus
Cancer
pagurus
Cancer
pagurus
Cancer
pagurus
Species
CD
Q)
o
CD
Q>
o
CD
0)
o
CD
Q)
O
CD
Q>
O
CD
Q>
O
Taxon
o
CO
o
o
o
CD
CD
O
o
o>
CD
O
PH
D
D
D
D
D
D
Arag.
Saturation
D
D
sp
0s
o
o
K)
Ambient
sp
0s
o
o
K)
Ambient
pco2
(ppm / (jatm /
kPa)
Exposed
(acidification
Control
Exposed
Control
Exposed
Control
Control
or Exposure
cn
3
cn
3
ca. 3 d
ca. 3 d
ca. 3 d
ca. 3 d
Duration
of Exposure
8-18
(Seasonal
change)
8-18
(Seasonal
change)
10-22
10-22
10-22
10-22
Temp. °C
Adult
Adult
Adult
Adult
Adult
Adult
Life Stage
Cuticle
thickness,
break
resistance of
claw
Cuticle
thickness,
break
resistance of
claw
Heat tolerance
Heat tolerance
Temp,
dependent
max PaC>2
values in
haemolymph
(Pa02 (kPa))
Temp,
dependent
max PaC>2
values in
haemolymph
(Pa02 (kPa))
End point
No effect
Control
Temperature and
PH
Downward shift (5
°C) of upper
thermal limits of
aerobic scope
Control
Reduced
Control
Response
Calcification
Calcification
Physiological
Physiological
Physiological
Physiological
Response
Type
Not sig.
Not sig.
cp'
Control
cn
cp'
Control
Significant
(Sig./Not
sig.)
K>
00
K>
00
K>
00
K>
00
Source
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Carcinus
maenas
Brae.
8.00
ND
ND
Exposed
(warming)
5 m
13-23
(Seasonal
change)
Adult
Cuticle
thickness,
break
resistance of
claw
Control pH
Calcification
Not sig.
7
Carcinus
maenas
Brae.
7.70
ND
ND
Exposed (acid
+ warm)
5 m
13-23
(Seasonal
change)
Adult
Cuticle
thickness,
break
resistance of
claw
No effect
Calcification
Not sig.
7
Carcinus
maenas
Brae.
8.06
0.96
650
Control
10
12.9
Adult
Feeding rate
and behavior
Control
Behavior
Control
8
Carcinus
maenas
Brae.
7.84
0.53
1250
Exposed
10 w
12.9
Adult
Feeding rate
and behavior
No effect
Behavior
Not sig.
8
Carcinus
maenas
Brae.
7.36
0.2
3500
Exposed
10 w
12.9
Adult
Feeding rate
and behavior
41% reduction in
feeding
Behavior
Sig.
8
Carcinus
maenas
Brae.
8.0 -
8.12
0.96
650
Control
3 d to 11 w
13.0
Adult
Gill gene
expression
(multiple
genes)
Control
Genetic
Control
25
Carcinus
maenas
Brae.
7.24-
7.36
0.2
3500
Exposed
3 d to 11 w
13.0
Adult
Gill gene
expression
(multiple
genes)
Acidification does
not act as a
strong stressor on
the cellular level
in gill epithelia
Genetic
Sig.
25
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
1 y (adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
99%
Mortality
Control
1
116
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
1 year
(adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
99%
Mortality
Not sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
1 year
(adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
99%
Mortality
Not sig.
1
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
2 y (adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
87%
Mortality
Control
1
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
2 y (adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
68%
Mortality
Not sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
2 y (adult)
5.0 mean
(1-9)
Larval
Percent of
viable larvae
46%
Mortality
Sig.
1
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
development
Control
Development
Control
1
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
development
No effect either
year
Development
Not sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
development
No effect either
year
Development
Not sig.
1
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
morphometries
Control
Development
Control
1
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
morphometries
Affected both
years
Development
Sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
up to 2 y
5.0 mean
(1-9)
Embryonic
Embryonic
morphometries
Affected both
years
Development
Sig.
1
117
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
2 y
5.0 mean
(1-9)
Adult
Survival rate
63%
Mortality
Control
1
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
2 y
5.0 mean
(1-9)
Adult
Survival rate
38%
Mortality
Sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
2 y
5.0 mean
(1-9)
Adult
Survival rate
44%
Mortality
Sig.
1
Chionoecetes
bairdi
Brae.
8.10
1.42
396
Control
2 y
5.0 mean
(1-9)
Adult
% Calcium dry
wt.
ca. 15.5%
Calcification
Control
1
Chionoecetes
bairdi
Brae.
7.80
0.77
779
Exposed
2 y
5.0 mean
(1-9)
Adult
% Calcium dry
wt.
ca. 15%
Calcification
Not sig.
1
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposed
2 y
5.0 mean
(1-9)
Adult
% Calcium dry
wt.
ca. 11%
Calcification
Sig.
1
Chionoecetes
bairdi
Brae.
8.10
1.8
269 ± 20
Control
16 d
5.0
Larval
Survival wild-
brooded larvae
Control
Mortality
Control
2
Chionoecetes
bairdi
Brae.
7.80
0.8
810 ±23
Exposed
16 d
5.0
Larval
Survival wild-
brooded larvae
No effect,
mortality curve
similar to control
Mortality
Not sig.
2
Chionoecetes
bairdi
Brae.
7.50
0.4
1665±162
Exposed
16 d
5.0
Larval
Survival wild-
brooded larvae
No effect,
mortality curve
similar to control
Mortality
Not sig.
2
Chionoecetes
bairdi
Brae.
8.10
1.8
269 ± 20
Control
10 d
5.0
Larval
Morphology of
wild-brooded
larvae
Control
Development
Control
2
118
-------
Species
Taxon
X
Q.
Arag.
Saturation
"i
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Chionoecetes
bairdi
Brae.
7.80
0.8
810 ± 23
Exposed
10 d
5.0
Larval
Morphology of
wild-brooded
larvae
No effect
Development
Not sig.
2
Chionoecetes
bairdi
Brae.
7.50
0.4
1665±162
Exposed
10 d
5.0
Larval
Morphology of
wild-brooded
larvae
No effect
Development
Not sig.
2
Chionoecetes
bairdi
Brae.
8.1
Embryo
& 8.1
Larvae
1.76 ±
0.16
326 ± 34
Control
2 y
7.01 ±
0.55
Larval
LT50
12.42
Mortality
Control
2
Chionoecetes
bairdi
Brae.
8.1
Embryo
& 7.8
Larvae
0.81 ±
0.04
811 ±38
Exposed
(larvae)
2 y
7.12 ±
0.59
Larval
LT50
11.36 (1.06 d
shorted than
control)
Mortality
Sig.?
2
Chionoecetes
bairdi
Brae.
8.1
Embryo
& 7.5
Larvae
0.43 ±
0.02
1620 ±60
Exposed
(larvae)
2 y
7.06 ±
0.54
Larval
LT50
10.44 (1.98 d
shorter than
control)
Mortality
Sig.?
2
Chionoecetes
bairdi
Brae.
7.8
Embryo
& 8.1
Larvae
1.76 ±
0.16
326 ± 34
Exposed
(embryo)
2 y
7.01 ±
0.55
Larval
LT50
14.54 (2.12 d
longerthan
control)
Mortality
Sig.?
2
Chionoecetes
bairdi
Brae.
7.8
Embryo
& 7.8
Larvae
0.81 ±
0.04
811 ±38
Exposed
(embryo &
larvae)
2 y
7.12 ±
0.59
Larval
LT50
14.05 (1.63 d
longerthan
control)
Mortality
Sig.?
2
119
-------
o
O
O
O
O
O
O
O
O
C/>
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
onoecet
bairdi
"O
(D
O
CD*
>
CD
(/)
CD
(/)
CD
(/)
CD
(/)
CD
(/)
CD
0)
CD
O)
CD
O)
CD
O)
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Taxon
CO
o
CO
o
OS'Z
CO
o
CO
o
7.5
Embryo
& 7.5
Larvae
7.5
Embryo
& 7.8
Larvae
7.5
Embryo
& 8.1
Larvae
7.8
Embryo
& 7.5
Larvae
PH
o
00
4^
4^
O
4^
o
00
4^
4^
0.43 ±
0.02
0.81 ±
0.04
1.76 ±
0.16
0.43 ±
0.02
Arag.
Saturation
00
CO
CD
K>
1597
00
CO
CD
K>
1620 ±60
811 ±38
326 ± 34
1620 ±60
pco2
(ppm / (jatm /
kPa)
Exposure
Control
Exposure
Exposure
Control
Exposed
(embryo &
larvae)
Exposed
(embryo &
larvae)
Exposed
(embryo)
Exposed
(embryo &
larvae)
Control
or Exposure
hO
hO
hO
hO
hO
hO
hO
hO
hO
Duration
of Exposure
Mean =5;
1-9
Mean =5;
1-9
Mean =5;
1-9
Mean =5;
1-9
Mean =5;
1-9
o ^
l+
7.12 ±
0.59
7.01 ±
0.55
o ^
l+
Temp. °C
Adult
Adult
Adult
Adult
Adult
Larval
Larval
Larval
Larval
Life Stage
Total counts of
hemocytes
Total counts of
hemocytes
Survival ral
Survival ral
Survival ral
LT50
LT50
LT50
LT50
End point
CD
CD
CD
No effect
Control
44%
38%
63%
15.31 (2.89 d
longerthan
control)
13.28 (0.86 d
longerthan
control)
15.58 (3.16 d
longerthan
control)
12.06 (0.36 d
shorter than
control)
Response
~0
TJ
(/>
o"
o
(Q
(/>
o"
o
CQ
Mortality
Mortality
Mortality
Mortality
Mortality
Mortality
Mortality
Response
Type
Q)_
Q)_
Not sig.
Control
Not sig.
Not sig.
Control
cn
cq"
:o
Not sig.?
cn
cq"
:o
Not sig.?
Significant
(Sig./Not
sig.)
CO
CO
CO
CO
CO
K)
hO
hO
Source
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposure
2 y
Mean =5;
1-9
Adult
Total counts of
hemocytes
No effect
Physiological
Not sig.
3
Chionoecetes
bairdi
Brae.
8.10
1.44
392
Control
2 y
Mean =5;
1-9
Adult
# dead cells in
the
hemolymph
Control
Physiological
Control
3
Chionoecetes
bairdi
Brae.
7.80
0.78
781
Exposure
2 y
Mean =5;
1-9
Adult
# dead cells in
the
hemolymph
No Effect
Physiological
Not sig.
3
Chionoecetes
bairdi
Brae.
7.50
0.4
1597
Exposure
2 y
Mean =5;
1-9
Adult
# dead cells in
the
hemolymph
Increase
Physiological
Sig.
3
Chionoecetes
bairdi
Brae.
8.00
1.43
438
Control
200 d
4.4-11.9
Juvenile
Mortality rate
0.0010 day-1
Mortality
Control
12
Chionoecetes
bairdi
Brae.
7.80
0.87
792
Exposed
200 d
4.4-11.9
Juvenile
Mortality rate
0.0023 day-1
Mortality
Sig.
12
Chionoecetes
bairdi
Brae.
7.50
0.44
1638
Exposed
200 d
4.4-11.9
Juvenile
Mortality rate
0.0050 day-1
Mortality
Sig.
12
Chionoecetes
bairdi
Brae.
8.00
1.43
438
Control
200 d
4.4-11.9
Juvenile
Growth rate
Control
Development
Control
12
Chionoecetes
bairdi
Brae.
7.80
0.87
792
Exposed
200 d
4.4-11.9
Juvenile
Growth rate
Slower than
control, faster
than 7.5 treatment
Development
Not sig.
12
Chionoecetes
bairdi
Brae.
7.50
0.44
1638
Exposed
200 d
4.4-11.9
Juvenile
Growth rate
Slower than
control and 7.8
treatment
Development
Sig.
12
121
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Hemigrapsus
oregonensis
Brae.
7.80
ND
506
Control
5 d
15.9
Larval
Mean
swimming
speed
Decreased from
day 1 to day 5
Behavior
Control
24
Hemigrapsus
oregonensis
Brae.
7.53
ND
1031
Exposed
5 d
15.9
Larval
Mean
swimming
speed
No different than
control
Behavior
Not sig.
24
Hemigrapsus
oregonensis
Brae.
7.80
ND
506
Control
5 d
15.9
Larval
Avg. # prey
consumed in
24 h
Increased, control
higherthan
treatment on day
1, lower on day 5
Behavior
Control
24
Hemigrapsus
oregonensis
Brae.
7.53
ND
1031
Exposed
5 d
15.9
Larval
Avg. # prey
consumed in
24 h
No significant
impact on feeding
rate
Behavior
Not sig.
24
Homarus
americanus
Asta.
8.10
ND
400
Control
12 d
20
Larval
Carapace
length
Control
Development
Control
20
Homarus
americanus
Asta.
7.70
ND
1200
Exposed
12 d
20
Larval
Carapace
length
Shorter carapace
in acidified
treatment for
larval stages 2-4
Development
Sig.
20
Homarus
americanus
Asta.
8.10
ND
400
Control
12 d
20
Larval
Days to reach
larval stage III
10 d
Development
Control
20
Homarus
americanus
Asta.
7.70
ND
1200
Exposed
12 d
20
Larval
Days to reach
larval stage III
12 d
Development
Sig.
20
Homarus
americanus
Asta.
8.03
2.13
409
Control
60 d
24.9
Juvenile
Survival rate
25%
Mortality
Control
29
Homarus
americanus
Asta.
7.85
1.53
606
Exposed
60 d
25
Juvenile
Survival rate
25%
Mortality
Not sig.
29
122
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Homarus
americanus
Asta.
7.72
1.13
903
Exposed
60 d
25
Juvenile
Survival rate
42%
Mortality
Not sig.
29
Homarus
americanus
Asta.
7.31
0.47
2856
Exposed
60 d
25.1
Juvenile
Survival rate
58%
Mortality
Sig.?
29
Homarus
americanus
Asta.
8.03
2.13
409
Control
60 d
24.9
Juvenile
Net
Calcification
rate
353.0 wt% 60 d"1
Calcification
Control
29
Homarus
americanus
Asta.
7.85
1.53
606
Exposed
60 d
25
Juvenile
Net
Calcification
rate
349.5 wt% 60 d"1
Calcification
Not sig.
29
Homarus
americanus
Asta.
7.72
1.13
903
Exposed
60 d
25
Juvenile
Net
Calcification
rate
376.3 wt% 60 d"1
Calcification
Not sig.?
29
Homarus
americanus
Asta.
7.31
0.47
2856
Exposed
60 d
25.1
Juvenile
Net
Calcification
rate
606.1 wt% 60 d"1
Calcification
Sig.?
29
Homarus
gammarus
Asta.
8.07
1.7
497
Control
5 w
9.5
Juvenile
Survival
ca. 100%
Mortality
Control
33
Homarus
gammarus
Asta.
7.74
0.82
1086
Exposed
5 w
9.6
Juvenile
Survival
ca. 85%
Mortality
Sig.
33
Homarus
gammarus
Asta.
6.9
0.13
8773
Exposed
5 w
9.6
Juvenile
Survival
ca. 72%
Mortality
Sig.
33
Homarus
gammarus
Asta.
8.05
1.86
559
Control
5 w
13.1
(ocean
warming)
Juvenile
Survival
ca. 100%
Mortality
Control
33
123
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Homarus
gammarus
Asta.
7.73
0.92
1258
Exposed
5 w
13.1
(ocean
warming)
Juvenile
Survival
ca. 92%
Mortality
Not sig.
33
Homarus
gammarus
Asta.
6.89
0.14
8827
Exposed
5 w
13.2
(ocean
warming)
Juvenile
Survival
ca. 79%
Mortality
Sig.
33
Homarus
gammarus
Asta.
8.07
1.7
497
Control
5 w
9.5
Juvenile
Mean growth
ca. 6%
Development
Control
33
Homarus
gammarus
Asta.
7.74
0.82
1086
Exposed
5 w
9.6
Juvenile
Mean growth
ca. 7.5%
Development
Not sig.
33
Homarus
gammarus
Asta.
6.9
0.13
8773
Exposed
5 w
9.6
Juvenile
Mean growth
-0.35%
Development
Sig.
33
Homarus
gammarus
Asta.
8.05
1.86
559
Control
5 w
13.1
(ocean
warming)
Juvenile
Mean growth
17.10%
Development
Control
33
Homarus
gammarus
Asta.
7.73
0.92
1258
Exposed
5 w
13.1
(ocean
warming)
Juvenile
Mean growth
ca. 14%
Development
Not sig.
33
Homarus
gammarus
Asta.
6.89
0.14
8827
Exposed
5 w
13.2
(ocean
warming)
Juvenile
Mean growth
ca. 2%
Development
Sig.
33
Homarus
gammarus
Asta.
8.07
1.7
497
Control
5 w
9.5
Juvenile
Oxygen
consumption
ca. 0.18 pmol min-
1 g "1
Physiological
Control
33
Homarus
gammarus
Asta.
7.74
0.82
1086
Exposed
5 w
9.6
Juvenile
Oxygen
consumption
0.105 pmol min-1
g-1
Physiological
Sig.
33
124
-------
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Homarus
gammarus
Species
Asta.
Asta.
Asta.
Asta.
Asta.
Asta.
Asta.
Asta.
Asta.
Asta.
Taxon
00
00
CT>
00
PH
00
K>
CO
CO
o
cn
o
00
CD
CO
o
cn
CD
o
O
O
o
O
Arag.
Saturation
cn
00
CD
K>
00
CD
00
K>
4^
CD
K>
00
CD
CO
727
692
1258
559
1086
497
8827
1258
559
8773
pco2
(ppm / (jatm /
kPa)
Exposed
Control
Exposed
Control
Exposed
Control
Exposed
Exposed
Control
Exposed
Control
or Exposure
cn
cn
cn
cn
cn
cn
cn
cn
cn
cn
Duration
3
3
5
5
5
5
5
5
5
of Exposure
10 or 18
10 or 18
13.1
(ocean
warming)
13.1
(ocean
warming)
9.6
9.5
13.2
(ocean
warming)
13.1
(ocean
warming)
13.1
(ocean
warming)
9.6
Temp. °C
Larva
Larva
£=
<
CD
=3
£=
<
CD
=3
£=
<
CD
=3
£=
<
CD
=3
£=
<
CD
=3
<—
£=
<
CD
=3
£=
<
CD
=3
£=
<
CD
=3
Life Stage
CD
CD
CD
CD
CD
CD
CD
CD
% deformities
% deformities
Ca
concentration
in carapace
Ca
concentration
in carapace
Ca
concentration
in carapace
Ca
concentration
in carapace
Oxygen
consumption
Oxygen
consumption
Oxygen
consumption
Oxygen
consumption
End point
CO
4^
O
0)
O
0)
O
0)
O
0)
O
k>
CD
CO
o_
3.
=3
O
0)
23%
5% or 12%
CO
TZ
3
o
3
cq
K>
TZ
3
o
3
CQ
4^
TZ
3
o
3
CQ
4^
CD
TZ
3
o
3
CQ
o
CQ TZ
- 3
o
3
Z3
p
^ cn
CQ -c
-3
o
3.
ZS
o
k>
-*• 4^
CQ -c
-3
o
3.
=3
Response
D
CD
<
CD
O
~o
3
D
CD
<
CD
O
~o
3
O
0)
o
-h
o"
0)
t t
o
0)
o
-h
o"
0)
t 1
o
0)
o
-h
o"
0)
t t
O
0)
o
-h
o"
0)
t t
~0
=r
(/>
o"
o
CQ
~0
=r
C/)
o"
o
CQ
~0
=r
C/)
o"
o
CQ
~0
zr
C/)
o"
o
CQ
Response
Type
CD
=3
t t
CD
=3
t i
o
=3
o
=3
o
=3
o
=3
o
Q)_
o
Q)_
o
Q)_
o
Q)_
cn
cq'
o
o
=3
t i
Sig
O
o
=3
t t
o
t t
(/)
o
o
=3
t t
Sig
Sig
O
o
=3
* t
O
* t
(/)
Significant
(Sig./Not
¦o
o_
o_
CQ
o_
o_
CQ
sig.)
CD
CD
33
33
33
33
33
33
33
33
Source
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Homarus
gammarus
Asta.
7.62
1.02
1198
Exposed
5 m
10 or 18
Larval
% deformities
43%
Development
Sig.?
19
Homarus
gammarus
Asta.
7.84
1.81
692
Control
5 m
18
Larval
Survival from
stage 4 to 5
months
46%
Mortality
Control
19
Homarus
gammarus
Asta.
7.82
1.75
121
Exposed
5 m
18
Larval
Survival from
stage 4 to 5
months
17%
Mortality
Not sig.
19
Homarus
gammarus
Asta.
7.62
1.02
1198
Exposed
5 m
18
Larval
Survival from
stage 4 to 5
months
61%
Mortality
Not sig.
19
Homarus
gammarus
Asta.
7.84
1.81
692
Control
1 y
18
Juvenile
% deformities
33%
Development
Control
19
Homarus
gammarus
Asta.
7.82
1.02
111
Exposed
1 y
18
Juvenile
% deformities
44%
Development
Not sig.
19
Homarus
gammarus
Asta.
7.62
1.02
1198
Exposed
1 y
18
Juvenile
% deformities
21%
Development
Not sig.
19
Homarus
gammarus
Asta.
8.39
4.33
315
Control
28 d
17
Larval
Carapace
mass
Control
Calcification
Control
21
Homarus
gammarus
Asta.
8.10
4.38
1202
Exposed
28 d
17
Larval
Carapace
mass
Reduction in
mass at Zoea
stage 4
Calcification
Sig.
21
Homarus
gammarus
Asta.
8.39
4.33
315
Control
28 d
17
Larval
[Ca2+]and
[Mg2+] in the
carapace
Control
Calcification
Control
21
126
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Homarus
gammarus
Asta.
8.10
4.38
1202
Exposed
28 d
17
Larval
[Ca2+]and
[Mg2+] in the
carapace
Nearly 50% less
Ca in carapace at
Zoea stage 4
Calcification
Sig.
21
Homarus
gammarus
Asta.
8.39
4.33
315
Control
28 d
17
Larval
Survival
Control
Mortality
Control
21
Homarus
gammarus
Asta.
8.10
4.38
1202
Exposed
28 d
17
Larval
Survival
No effect
Mortality
Not sig.
21
Hyas araneus
Brae.
8.00
ND
380
Control
several
hours
Cooling
10-6
Adult
A in blood
haemolymph
pp02-kPa
3.75
Physiological
Control
16
Hyas araneus
Brae.
7.80
ND
710
Exposed
several
hours
Cooling
10-6
Adult
A in blood
haemolymph
ppC>2 - °C
4.84
Physiological
Not sig.
16
Hyas araneus
Brae.
7.30
ND
3000
Exposed
several
hours
Cooling
10-6
Adult
A in blood
haemolymph
ppC>2 - °C
4.6
Physiological
Sig.
16
Hyas araneus
Brae.
8.00
ND
380
Control
several
hours
Warming
10-25
Adult
A in blood
haemolymph
ppC>2 - °C
-6
Physiological
Control
16
Hyas araneus
Brae.
7.80
ND
710
Exposed
several
hours
Warming
10-25
Adult
A in blood
haemolymph
ppC>2 - °C
-5.5
Physiological
Sig
16
Hyas araneus
Brae.
7.30
ND
3000
Exposed
several
hours
Warming
10-25
Adult
A in blood
haemolymph
ppC>2 - °C
-5.53
Physiological
Sig.
16
127
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Hyas araneus
Brae.
8.11
ND
354
Control
Length of
dev. stage
(-10 - 80 d)
3
Zoea II
Duration of
larval stage
72 d - Helgoland
59 d - Svalbard
Development
Control
35
Hyas araneus
Brae.
7.81
ND
754
Exposure
Length of
dev. stage
(-10 - 80 d)
3
Zoea II
Duration of
larval stage
74 d - Helgoland
63 d - Svalbard
Development
Sig.
35
Hyas araneus
Brae.
7.33
ND
2378
Exposure
Length of
dev. stage
(-10 - 80 d)
3
Zoea II
Duration of
larval stage
76 d - Helgobard
68 d - Svaldbard
Development
Sig.
35
Hyas araneus
Brae.
8.12
ND
346
Control
Length of
dev. stage
(-10 - 80 d)
9
Zoea II
Duration of
larval stage
18 d - Helgoland
23 d - Svalbard
Development
Control
35
Hyas araneus
Brae.
7.81
ND
786
Exposure
Length of
dev. stage
(-10 - 80 d)
9
Zoea II
Duration of
larval stage
19 d - Helgoland
21 d - Svalbard
Development
Sig.
35
Hyas araneus
Brae.
7.35
ND
2443
Exposure
Length of
dev. stage
(-10 - 80 d)
9
Zoea II
Duration of
larval stage
20 d - Helgoland
22 d - Svalbard
Development
Sig.
35
Hyas araneus
Brae.
8.05
ND
401
Control
Length of
dev. stage
(-10 - 80 d)
15
Zoea II
Duration of
larval stage
11 d - Helgoland
12 d - Svalbard
Development
Control
35
Hyas araneus
Brae.
7.79
ND
846
Exposure
Length of
dev. stage
(-10 - 80 d)
15
Zoea II
Duration of
larval stage
11 d - Helgoland
12 d - Svalbard
Development
Sig.
35
Hyas araneus
Brae.
7.34
ND
2637
Exposure
Length of
dev. stage
(-10 - 80 d)
15
Zoea II
Duration of
larval stage
12 d - Helgoland
13 d- Svalbard
Development
Sig.
35
128
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Hyas araneus
Brae.
8.13
ND
420
Control
ND
ND
Zoea I
Mortality
15.5%
Mortality
Control
36
Hyas araneus
Brae.
ND
ND
3300
Exposure
ND
ND
Zoea I
Mortality
21.6%
Mortality
Not sig.
36
Hyas araneus
Brae.
ND
ND
420
Control
ND
ND
Zoea II
Mortality
14.7%
Mortality
Control
36
Hyas araneus
Brae.
ND
ND
3300
Exposure
ND
ND
Zoea II
Mortality
32.2%
Mortality
Not sig.
36
Lysmata
californica
Cari.
7.99
ND
462
Control
21 d
18.7
Adult
Mean cuticle
thickness
27.9 |jm
Calcification
Control
5
Lysmata
californica
Cari.
7.53
ND
1297
Exposed
21 d
18.7
Adult
Mean cuticle
thickness
23.66 |jm
Calcification
Not Sig.
5
Lysmata
californica
Cari.
7.99
ND
462
Control
21 d
18.7
Adult
Body
Transparency
Peak Range
630-910nm
Physiological
Control
5
Lysmata
californica
Cari.
7.53
ND
1297
Exposed
21 d
18.7
Adult
Body
Transparency
Peak Range
680-885nm
Physiological
Sig.
5
Metacarcinus
magister
Brae.
8.00
1.74
466
Control
45 d
12
Larval
Larval survival
57.90%
Mortality
Control
10
Metacarcinus
magister
Brae.
7.50
0.54
1781
Exposed
45 d
12
Larval
Larval survival
13.50%
Mortality
Sig.
10
Metacarcinus
magister
Brae.
7.10
0.25
3920
Exposed
45 d
12
Larval
Larval survival
21.10%
Mortality
Sig.
10
Metacarcinus
magister
Brae.
8.00
1.74
466
Control
45 d
12
Larval
Percent
reaching larval
stage 4
68%
Development
Control
10
129
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Metacarcinus
magister
Brae.
7.50
0.54
1781
Exposed
45 d
12
Larval
Percent
reaching larval
stage 4
25%
Development
Sig.
10
Metacarcinus
magister
Brae.
7.10
0.25
3920
Exposed
45 d
12
Larval
Percent
reaching larval
stage 4
25%
Development
Sig.
10
Metacarcinus
magister
Brae.
8.00
1.74
466
Control
34 d
12
Embryonic
Proportion
hatched
0.77
Mortality
Control
10
Metacarcinus
magister
Brae.
7.50
0.54
1781
Exposed
34 d
12
Embryonic
Proportion
hatched
0.59
Mortality
Sig.
10
Metacarcinus
magister
Brae.
7.10
0.25
3920
Exposed
34 d
12
Embryonic
Proportion
hatched
0.72
Mortality
Not Sig.
10
Metacarcinus
magister
Brae.
7.80
ND
506
Control
5 d
15.9
Larval
Mean
swimming
speed
Decreased from
day 1 to day 5
Behavior
Control
24
Metacarcinus
magister
Brae.
7.53
ND
1031
Exposed
5 d
15.9
Larval
Mean
swimming
speed
Increased
swimming speed
over control
Behavior
Sig.
24
Metacarcinus
magister
Brae.
7.80
ND
506
Control
5 d
15.9
Larval
Avg. # prey
consumed in
24 h
Increased, control
higherthan
treatment on day
1 and day 5
Behavior
Control
24
Metacarcinus
magister
Brae.
7.53
ND
1031
Exposed
5 d
15.9
Larval
Avg. # prey
consumed in
24 h
No significant
impact on feeding
rate
Behavior
Not sig.
24
Metacarcinus
magister
Brae.
8.10
ND
49.2 Pa
Control
7-10 d
14
Adult
Hemolymph
PH
pH 8.01
Physiological
Control
26
130
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Metacarcinus
magister
Brae.
7.40
ND
327.9 Pa
Exposed
7-10 d
14
Adult
Hemolymph
PH
pH 7.93
Physiological
Not sig.
26
Metacarcinus
magister
Brae.
8.10
ND
49.2 Pa
Control
7-10 d
14
Adult
Hemolymph
pC02
132.9 Pa/4.91
mM[5]
Physiological
Control
26
Metacarcinus
magister
Brae.
8.10
ND
49.2 Pa
Control
7-10 d
14
Adult
Hemolymph
[HC03-]
4.91 mM
Physiological
Control
26
Metacarcinus
magister
Brae.
7.40
ND
327.9 Pa
Exposed
7-10 d
14
Adult
Hemolymph
pC02
402.2 Pa
Physiological
Sig.
26
Metacarcinus
magister
Brae.
7.40
ND
327.9 Pa
Exposed
7-10 d
14
Adult
Hemolymph
[HC03-]
14.89 mM
Physiological
Sig.
26
Metapenaeus
joyneri
Pena.
8.14
2.07
0.04 kPa
Control
1-10 d
15
Adult
Metabolic
Scope
15.5 AAMR-RMR
Physiological
Control
17
Metapenaeus
joyneri
Pena.
6.91
0.16
0.92 kPa
Exposed
1-10 d
15
Adult
Metabolic
Scope
9.5 AAMR-RMR
Physiological
Sig.
17
Metapenaeus
joyneri
Pena.
8.16
2.58
0.04 kPa
Control
1-10 d
20
Adult
Metabolic
Scope
11.3 A AMR-RMR
Physiological
Control
17
Metapenaeus
joyneri
Pena.
6.90
0.17
0.92 kPa
Exposed
1-10 d
20
Adult
Metabolic
Scope
8.8 AAMR-RMR
Physiological
Sig.
17
Metapenaeus
joyneri
Pena.
8.14
2.07
0.04 kPa
Control
10 d
15
Adult
Muscle mass
(% of body
mass)
44.9%
Physiological
Control
17
Metapenaeus
joyneri
Pena.
6.91
0.16
0.92 kPa
Exposed
10 d
15
Adult
Muscle mass
(% of body
mass)
45.5%
Physiological
Not Sig.
17
131
-------
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Metapenaeus
joyneri
Metapenaeus
joyneri
Species
CD
Q)
O
CD
Q)
O
CD
Q>
O
CD
Q>
O
CD
Q>
O
CD
0)
O
CD
Q>
O
CD
Q>
O
CD
Q>
O
CD
Q>
O
Pena.
Pena.
Taxon
CO
CD
CD
CD
O
cn
CD
CO
CD
CD
CD
CD
cn
CD
CO
CD
CD
CD
CD
O
CO
CD
PH
D
D
D
D
D
D
D
D
D
D
O
hO
cn
00
Arag.
Saturation
0.25 kPa
o
o
00
7T
~0
Q)
6.04 kPa
1.1 kPa
0.25 kPa
0
CD
00
7T
TJ
Q)
6.04 kPa
1.1 kPa
0.25 kPa
0
CD
00
7T
TJ
Q)
0.92 kPa
0
CD
4^
~0
Q)
pco2
(ppm / (jatm /
kPa)
Exposed
Control
Exposed
Exposed
Exposed
Control
Exposed
Exposed
Exposed
Control
Exposed
Control
Control
or Exposure
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
O
Q.
O
Q.
Duration
of Exposure
cn
cn
cn
cn
cn
cn
cn
cn
cn
cn
K>
O
K>
O
Temp. °C
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Life Stage
Mortality
Mortality
Haemolymph
[HCOs-]
Haemolymph
[HCO3-]
Haemolymph
[HCO3-]
Haemolymph
[HCOs"]
Haemolymph
PCO2
Haemolymph
PCO2
Haemolymph
PCO2
Haemolymph
PCO2
Muscle mass
(% of body
mass)
Muscle mass
(% of body
mass)
End point
No apparent
effect
No apparent
effect
Increased
Increased
Increased
No change
Increased
Increased
No consistent
pattern
No change
40.56%
45.95%
Response
Mortality
Mortality
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Response
Type
Not sig.
Control
cn
cq'
cn
cq'
cn
cq'
Control
cq'
cq'
Not sig.
Control
cq'
Control
Significant
(Sig./Not
sig.)
K>
K>
K>
K>
K>
K)
K)
K)
K>
K>
Source
-------
u>
u>
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Species
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Brae.
Taxon
CO
CO
CO
o
CD
CO
hO
CO
CD
00
CO
CO
CO
CD
CD
6.05
6.74
CO
96'Z
6.05
6.74
PH
k>
en
CD
4^
CO
hO
k>
cn
CD
z
D
z
D
D
2
D
z
D
z
D
Arag.
Saturation
CD
00
00
cn
o
o
1136
cn
CT>
CD
00
00
cn
o
o
6.04 kPa
1.1 kPa
0.25 kPa
o
CD
00
7T
TJ
Q)
6.04 kPa
1.1 kPa
pC02
(ppm / (jatm /
kPa)
Exposed
Control
Exposed
Control
Exposed
Control
Exposed
Exposed
Exposed
Control
Exposed
Exposed
Control
or Exposure
4^
Q.
Q.
Q.
Q.
Q.
4^
Q.
CT>
Q.
CT>
Q.
CT>
Q.
O)
CL
CT>
Q.
CT>
Q.
Duration
of Exposure
O
O
cn
cn
O
O
cn
cn
cn
cn
cn
cn
Temp. °C
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Life Stage
Haemolymph
[HC03 "]e
Haemolymph
[HC03 "]e
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Haemolymph
PH
Mortality
Mortality
End point
9.19 mmol M
6.56 mmol I"1
pH 7.95
pH 7.89
pH 7.87
pH 7.84
Declined
(ca. 7.4)
No consistent
pattern
No consistent
pattern
No change (ca.
8.0 pH)
All died after 5
days
No apparent
effect
Response
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Physiological
Mortality
Mortality
Response
Type
cq'
Control
Not Sig.
Control
Not Sig.
Control
cq'
Not sig.
Not sig.
Control
cq'
Not sig.
Significant
(Sig./Not
sig.)
4^
K>
K>
K>
K>
K>
K>
Source
-------
Nephrops
norvegicus
Nephrops
norvegicus
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Necora puber
Species
>
(/>
11
Q)
>
(/>
11
Q)
CD
Q)
O
CD
Q)
O
CD
0)
O
CD
Q>
O
CD
Q)
O
CD
0)
O
CD
0)
O
CD
0)
O
Taxon
CD
00
CT>
CD
CD
Kj
CD
CO
en
CD
CD
CD
Kj
CD
CO
cn
CO
NJ
CO
CD
00
PH
D
D
O
O
CD
O
CO
NJ
NJ
O
CD
CD
o
CO
NJ
NJ
4^
CO
NJ
Arag.
Saturation
886-1787
(higher at
higher
temperatures)
CO
CO
o
1234
3205
CO
4^
1234
3205
CO
4^
1136
cn
CD
pco2
(ppm / (jatm /
kPa)
Exposed
(lower pH &
various
temperatures)
Control
Exposed
Exposed
Control
Exposed
Exposed
Control
Exposed
Control
Control
or Exposure
4^
3
4^
3
CO
o
Q.
CO
o
Q.
CO
o
Q.
CO
o
Q.
CO
o
Q.
CO
o
Q.
4^
Q.
Q.
Duration
of Exposure
Ul
00
en
CO
CO
CO
CO
CO
CO
cn
cn
Temp. °C
Embryonic
Embryonic
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Life Stage
Development
rate of
embryos
Development
rate of
embryos
[Ca2+]and
[Mg2+] in the
carapace
[Ca2+]and
[Mg2+] in the
carapace
[Ca2+]and
[Mg2+] in the
carapace
Oxygen uptake
Oxygen uptake
Oxygen uptake
Haemolymph
[HC03 "]e
Haemolymph
[HC03 "]e
End point
No pH effect
(temperature has
a sig. effect)
Control
No effect
No effect
Control
O
CD
cn
zr
O
K)
3
CO
O
CD
"C_
O
K)
3
cq
=r
O
cn
zr
O
K)
3
CQ
11.03 mmol M
7.14 mmol I"1
Response
Development
Development
Calcification
Calcification
Calcification
Physiological
Physiological
Physiological
Physiological
Physiological
Response
Type
Not sig.
Control
Not sig.
Not sig.
Control
CD
cq'
Not Sig.
Control
cq'
Control
Significant
(Sig./Not
sig.)
CO
4^
CO
4^
K>
NJ
NJ
NJ
NJ
NJ
Source
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Nephrops
norvegicus
Asta.
8
ND
330
Control
4 m
5
Embryonic
Oxygen
consumption
of eggs
Control
Development
Control
34
Nephrops
norvegicus
Asta.
7.6
ND
886-1787
(higher at
higher
temperatures)
Exposed
(lower pH &
various
temperatures)
4 m
5-18
Embryonic
Oxygen
consumption
of eggs
No pH effect
(p=0.051;
temperature has
sig. effect)
Development
Not sig.
34
Nephrops
norvegicus
Asta.
8
ND
330
Control
4 m
5
Embryonic
Oxidative
stress in eggs
Control
Development
Control
34
Nephrops
norvegicus
Asta.
7.6
ND
886-1787
(higher at
higher
temperatures)
Exposed
(lower pH &
various
temperatures)
4 m
5-18
Embryonic
Oxidative
stress in eggs
Reduced
compared to
controls
(temperature was
not sig.)
Development
Sig.
34
Pagurus
bernhardus
Asta.
8.20
2.89
375
Control
5 d
15
Adult
Latency to find
shell
10.7% failed to
find new shell
Behavior
Control
9
Pagurus
bernhardus
Anom.
6.80
0.18
12191
Exposed
5 d
15
Adult
Latency to find
shell
45.7% failed to
find new shell
Behavior
Sig.
9
Pagurus
tannreri
(bathyal)
Anom.
7.60
0.83
1379
Control
4 w.
6
Adult
Time for prey
detection after
4 wks.
exposure
250 seconds
Behavior
Control
18
Pagurus
tannreri
(bathyal)
Anom.
7.10
0.23
2366
Exposed
4 w
6
Adult
Time for prey
detection after
4 wks.
exposure
720 seconds
Behavior
Sig.
18
135
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Pagurus
tannreri
(bathyal)
Anom.
7.60
0.83
1379
Control
20 w
6
Adult
Antennular
flicking
Fairly constant
Behavior
Control
18
Pagurus
tannreri
(bathyal)
Anom.
7.10
0.23
2366
Exposed
20 w
6
Adult
Antennular
flicking
Decreased
throughout
experiment
Behavior
Sig. > 7 d
18
Pagurus
tannreri
(bathyal)
Anom.
7.60
0.83
1379
Control
9 w
6
Adult
Respiration
rate
Gradual decrease
Physiological
Control
18
Pagurus
tannreri
(bathyal)
Anom.
7.10
0.23
2366
Exposed
9 w
6
Adult
Respiration
rate
Increased 3
weeks, returned
to pretreatment by
8 weeks
Physiological
Not sig.
18
Palaemon
pacificus
Cari.
8.20
ND
380
Control
30 d
25
Adult
Survival
90%
Mortality
Control
13
Palaemon
pacificus
Cari.
7.89
ND
1000
Exposed
30 d
25
Adult
Survival
55%
Mortality
Sig.
13
Palaemon
pacificus
Cari.
8.20
ND
380
Control
15 d
25
Adult
Survival
95%
Mortality
Control
13
Palaemon
pacificus
Cari.
7.64
ND
1900
Exposed
15 d
25
Adult
Survival
65%
Mortality
Sig.
13
Palaemon
pacificus
Cari.
8.20
ND
380
Control
15 & 30 d
25
Adult
% increase in
length
Control
Development
Control
13
Palaemon
pacificus
Cari.
7.89
ND
1000
Exposed
30 d
25
Adult
% increase in
length
No effect
Development
Not sig.
13
136
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Palaemon
pacificus
Cari.
7.64
ND
1900
Exposed
15 d
25
Adult
% increase in
length
Slower growth
Development
Sig.
13
Pandalus
borealis
Cari.
8.10
ND
ND
Control
35 d
5
Larval
Mean
accumulated
mortality
37% on day 35
Mortality
Control
23
Pandalus
borealis
Cari.
7.60
ND
ND
Exposed
35 d
5
Larval
Mean
accumulated
mortality
25% on day 35
(lower than
control on last day
only)
Mortality
Not sig.
23
Pandalus
borealis
Cari.
8.10
ND
ND
Control
35 d
5
Larval
Development
time (to IV
zoea)
Control
Development
Control
23
Pandalus
borealis
Cari.
7.60
ND
ND
Exposed
35 d
5
Larval
Development
time (to IV
zoea)
Lower % than
control
Development
Sig.
23
Pandalus
borealis
Cari.
8.11
1.3-1.8
337-474
Control
up to 13 d
6.7
Larval
Larval
development
(Stage II, III,
and IV)
Control
Development
Control
31
Pandalus
borealis
Cari.
7.65
0.5-0.7
1038 - 1437
Exposed
up to 13 d
6.7
Larval
Larval
development
(Stage II, III,
and IV)
Decreased,
greatest decrease
in Stage IV
Development
Sig.
31
Pandalus
borealis
Cari.
7.60
0.5-0.7
1147 - 1751
Exposed (pH
and
temperature)
up to 13 d
9.5
Larval
Larval
development
(Stage II, III,
and IV)
Greater values
than at lower
temp.
Development
Not. sig
31
137
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Pandalus
borealis
Cari.
8.11
1.3-1.8
337-474
Control
up to 13 d
6.7
Embryonic
Hatching
success
98.7
Mortality
Control
31
Pandalus
borealis
Cari.
7.65
0.5-0.7
1038 - 1437
Exposed
up to 13 d
6.7
Embryonic
Hatching
success
ca. 98%
Mortality
Not sig.
31
Pandalus
borealis
Cari.
7.60
0.5-0.7
1147 - 1751
Exposed (pH
and
temperature)
up to 13 d
9.5
Embryonic
Hatching
success
96.1 (not dif. than
elevated temp,
alone)
Mortality
Sig.
31
Pandalus
borealis
Cari.
8.11
1.3-1.8
337-474
Control
up to 13 d
6.7
Larval
Feeding rate
(Stages II, III
and IV)
ca. 2.5 (Stage II)
Behavior
Control
31
Pandalus
borealis
Cari.
7.65
0.5-0.7
1038 - 1437
Exposed
up to 13 d
6.7
Larval
Feeding rate
(Stages II, III
and IV)
ca. 2.5 (Stage II)
Behavior
Not sig.
31
Pandalus
borealis
Cari.
7.60
0.5-0.7
1147 - 1751
Exposed (pH
and
temperature)
up to 13 d
9.5
Larval
Feeding rate
(Stages II, III
and IV)
ca. 3.3 (Stage II;
not different than
elevated temp,
alone)
Behavior
Sig.
31
Pandalus
borealis
Cari.
8.11
1.3-1.8
337-474
Control
up to 13 d
6.7
Larval
Oxygen
consumption
rate
ca. 50 nmol O2 h"1
mg-1 dry mass
Physiological
Control
31
Pandalus
borealis
Cari.
7.65
0.5-0.7
1038 - 1437
Exposed
up to 13 d
6.7
Larval
Oxygen
consumption
rate
ca. 49 nmol O2 h"1
mg-1 dry mass
Physiological
Not sig.
31
138
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Pandalus
borealis
Cari.
7.60
0.5-0.7
1147 - 1751
Exposed(pH
and temp.)
up to 13 d
9.5
Larval
Oxygen
consumption
rate
ca. 53 nmol O2 h"1
mg-1 dry mass
(not different than
elevated temp,
alone)
Physiological
Sig.
31
Panopeus
herbstii
Brae.
8.20
6.7
499
Control
71 d/48hrs
25.97
Adult
Percentage of
oysters
consumed
67.50%
Behavior
Control
6
Panopeus
herbstii
Brae.
8.04
5.1
785
Exposed
71 d/48hrs
25.97
Adult
Percentage of
oysters
consumed
41%
Behavior
Sig.
6
Panopeus
herbstii
Brae.
7.05
0.8
9274
Exposed
71 d/48hrs
25.97
Adult
Percentage of
oysters
consumed
1%
Behavior
Sig.
6
Panopeus
herbstii
Brae.
8.20
6.7
499
Control
72 d/48hrs
25.97
Adult
Time handling
prey
40%
Behavior
Control
6
Panopeus
herbstii
Brae.
8.04
5.1
785
Exposed
71 d/48hrs
25.97
Adult
Time handling
prey
20%
Behavior
Not sig.
6
Panopeus
herbstii
Brae.
7.05
0.8
9274
Exposed
71 d/48hrs
25.97
Adult
Time handling
prey
5%
Behavior
Sig.
6
Panopeus
herbstii
Brae.
8.04
5.1
785
Exposed
71 d/48hrs
25.97
Adult
Calcification
rate
No effect
Calcification
Not sig.
6
Panopeus
herbstii
Brae.
7.05
0.8
9274
Exposed
71 d/48hrs
25.97
Adult
Calcification
rate
No effect
Calcification
Not sig.
6
Paraiithodes
camtschaticus
Anom.
8.00
1.43
438
Control
200 d
4.4-11.9
Juvenile
Mortality rate
0.0023 day1
Mortality
Control
12
139
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Paralithodes
camtschaticus
Anom.
7.80
0.87
792
Exposed
200 d
4.4-11.9
Juvenile
Mortality rate
0.0047 day1
Mortality
Sig.
12
Paralithodes
camtschaticus
Anom.
7.50
0.44
1638
Exposed
200 d
4.4-11.9
Juvenile
Mortality rate
0.025 day1
Mortality
Sig.
12
Paralithodes
camtschaticus
Anom.
8.00
1.43
438
Control
200 d
4.4-11.9
Juvenile
Growth rate
61% higher mass
than 7.8 treatment
Development
Control
12
Paralithodes
camtschaticus
Anom.
7.80
0.87
792
Exposed
200 d
4.4-11.9
Juvenile
Growth rate
Slower than
control, faster
than 7.5 treatment
Development
Sig.
12
Paralithodes
camtschaticus
Anom.
7.50
0.44
1638
Exposed
200 d
4.4-11.9
Juvenile
Growth rate
Slower than
control and 7.8
treatment
Development
Sig.
12
Penaeus
plebejus
Pena.
8.03
2.13
409
Control
60 d
24.9
Adult
Survival rate
100%
Mortality
Control
29
Penaeus
plebejus
Pena.
7.85
1.53
606
Exposed
60 d
25
Adult
Survival rate
100%
Mortality
Not sig.
29
Penaeus
plebejus
Pena.
7.72
1.13
903
Exposed
60 d
25
Adult
Survival rate
100%
Mortality
Not sig.
29
Penaeus
plebejus
Pena.
7.31
0.47
2856
Exposed
60 d
25.1
Adult
Survival rate
100%
Mortality
Not sig.
29
Penaeus
plebejus
Pena.
8.03
2.13
409
Control
60 d
24.9
Adult
Calcification
rate
15.3 wt% 60 d"1
Calcification
Control
29
Penaeus
plebejus
Pena.
7.85
1.53
606
Exposed
60 d
25
Adult
Calcification
rate
17.3 wt% 60 d"1
Calcification
Not sig.?
29
Penaeus
plebejus
Pena.
7.72
1.13
903
Exposed
60 d
25
Adult
Calcification
rate
27.5 wt% 60 d"1
Calcification
Sig.?
29
140
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Penaeus
plebejus
Pena.
7.31
0.47
2856
Exposed
60 d
25.1
Adult
Calcification
rate
37.8 wt% 60 d-1
Calcification
Sig.?
29
Petrolisthes
cinctipes
Anom.
7.93
ND
574
Control
7-10 d
13
Embryonic
Metabolic rate
ca. 1.9 |jmol
O2 h"1
Physiological
Control
30
Petrolisthes
cinctipes
Anom.
7.58
ND
1361
Exposed
7-10 d
13
Embryonic
Metabolic rate
ca. 1.74 |jmol
O2 h"1
Physiological
Sig.
30
Petrolisthes
cinctipes
Anom.
7.93
ND
574
Control
4-10 d
13
Larval
Metabolic rate
ca. 11.75 |jmol
O2 h"1
Physiological
Control
30
Petrolisthes
cinctipes
Anom.
7.58
ND
1361
Exposed
4-10 d
13
Larval
Metabolic rate
ca. 12.25 |jmol
O2 h"1
Physiological
Not sig.
30
Petrolisthes
cinctipes
Anom.
7.93
ND
574
Control
3-5 d
13
Juveniles
Metabolic rate
ca. 22.75 |jmol
O2 h"1
Physiological
Control
30
Petrolisthes
cinctipes
Anom.
7.58
ND
1361
Exposed
3-5 d
13
Juveniles
Metabolic rate
ca. 22.8 |jmol
O2 h"1
Physiological
Not sig.
30
Petrolisthes
cinctipes
Anom.
7.93
ND
574
Control
6 d
13
Larval
C/N ratio
ca. 3.38 |jmol
O2 h"1
Physiological
Control
30
Petrolisthes
cinctipes
Anom.
7.58
ND
1361
Exposed
6 d
13
Larval
C/N ratio
ca. 3.58 |jmol
O2 h"1
Physiological
Sig.
30
Petrolisthes
cinctipes
Anom.
8.00
2.033
450
Control
10 d
14
Larval
Oxygen
consumption
ca. 0.75 O2 larva-1
hr1
Physiological
Control
32
Petrolisthes
cinctipes
Anom.
7.71
1.127
949.8
Exposed (CO2
only)
10 d
14
Larval
Oxygen
consumption
ca. 0.79 O2 larva-1
hr-1
Physiological
Not sig.
32
Petrolisthes
cinctipes
Anom.
7.71
1.127
949.8
Exposed (CO2
followed by
salinity stress)
10 d
14
Larval
Oxygen
consumption
ca. 0.89 O2 larva-1
hr-1
Physiological
Sig.
32
141
-------
~D
Q CD
Q CD
~t)
Q CD
~D
Q CD
~D
Q CD
Q CD
Q CD
~D
Q CD
~D
Q CD
~D
Q CD
C/>
o o
"5"
CD rr
W CD
C/)
o o
CD rr
0) CD
C/)
o o
CD rr
0) CD
C/)
o o
CD
0) CD
(/)
o o
CD
0) CD
C/)
o o
CD rr
0) CD
C/)
o o
CD rr
0) CD
(/)
o o
CD rr
0) CD
C/)
o o
CD rr
0) CD
C/)
o o
CD rr
0) CD
C/)
(D
O
CD*
>
Anom.
Anom.
Anom.
Anom.
Anom.
Anom.
Anom.
Anom.
Anom.
Anom.
Taxon
06Z
09Z
06Z
09Z
06Z
cn
09Z
cn
09Z
8.12
PH
z
D
z
D
z
D
z
D
z
D
o
CD
cn
o
CD
cn
4^
CD
Arag.
Saturation
574
1361
574
1361
574
4801
1476
4801
1476
461
pco2
(ppm / (jatm /
kPa)
Control
(Ambient)
Exposed
Control
(Ambient)
Exposed
Control
(Ambient)
Exposed
Exposed
Exposed
Exposed
Control
Control
or Exposure
CD
Q.
CD
Q.
CD
Q.
CD
Q.
CD
Q.
17 d
17 d
17 d
17 d
17 d
Duration
of Exposure
ambient
ambient
ambient
ambient
ambient
CO
o
K>
cn
-
-
-
Temp. °C
Larval
Embryonic
Embryonic
Embryonic
Embryonic
Adult
Adult
Adult
Adult
Adult
Life Stage
Larval survival
Embryo
development
Embryo
development
Hatching
success
Hatching
success
Respiration
rate
(pmolCh/min/g)
1 ^
o CD
o
Q) 32.
—r r+ —f
3 Q)
1 1
^ o"
(Q =3
Respiration
rate
(pmolCh/min/g)
Respiration
rate
(pmolCh/min/g)
1 ^
o CD
o
CD 32.
—r r+ —f
3 CD Q)
1 1
^ O
(Q =3
End point
Control
No volume
increase before
hatching
volume increased
15% before
hatching
No effect
Variable
CO
cn
CO
4^
CD
4^
00
4^
Response
Mortality
Development
Development
Mortality
Mortality
Physiological
Physiological
Physiological
Physiological
Physiological
Response
Type
Control
cq"
Control
Not sig.
Control
Sig.
(temp, x
PH)
Not sig.
(temp, x
PH)
Not sig.
Not sig.
Control
Significant
(Sig./Not
sig.)
cn
cn
cn
cn
cn
4^
4^
4^
4^
4^
Source
-------
Species
Taxon
X
Q.
Arag.
Saturation
E
re
n
CM ^
8 I?
q. -Ir
Control
or Exposure
Duration
of Exposure
Temp. °C
Life Stage
End point
Response
Response
Type
Significant
(Sig./Not
sig.)
Source
Petrolisthes
cinctipes
Anom.
7.60
ND
1361
Exposed
9 d
ambient
Larval
Larval survival
Routinely lower
survival than 7.9
treatment
Mortality
Not Sig.
15
Petrolisthes
cinctipes
Anom.
7.90
ND
574
Control
(Ambient)
40 d
ambient
Juvenile
Juvenile
survival
Control
Mortality
Control
15
Petrolisthes
cinctipes
Anom.
7.60
ND
1361
Exposed
40 d
ambient
Juvenile
Juvenile
survival
Reduced survival
after longer-term
exposure (but
95% CI overlap)
Mortality
Sig.
15
Upogebia
deltaura
Gebi.
7.99
1.52
607
Control
35 d
14
Adult
Haemolymph
PH
Control
Physiological
Control
11
Upogebia
deltaura
Gebi.
7.64
0.77
1396
Exposed
35 d
14
Adult
Haemolymph
PH
No effect
Physiological
Not sig.
11
Upogebia
deltaura
Gebi.
7.35
0.4
2707
Exposed
35 d
14
Adult
Haemolymph
PH
Reduced
Physiological
Sig.
11
Upogebia
deltaura
Gebi.
7.99
1.52
607
Control
35 d
14
Adult
Mortality
Control
Mortality
Control
11
Upogebia
deltaura
Gebi.
7.64
0.77
1396
Exposed
35 d
14
Adult
Mortality
No effect
Mortality
Not sig.
11
Upogebia
deltaura
Gebi.
7.35
0.4
2707
Exposed
35 d
14
Adult
Mortality
No effect
Mortality
Not sig.
11
Upogebia
deltaura
Gebi.
6.71
0.11
14110
Exposed
35 d
14
Adult
Mortality
100% mortality on
day 35
Mortality
Sig.
11
143
-------
Table 6-6. MATCs for pH for each decapod species based on all endpoints (comprehensive analysis).
NOAEL = no observed adverse effect level; LOAEL = lowest observed adverse effect level; MATC = maximum acceptable toxicant concentration. The
NOAEL is the control pH or the lowest non-significant exposure, identified by *. MATC is calculated as the geometric mean of the NOAEL and LOAEL. The
MATCs were calculated by first taking the antilog of the pHs and then taking the Iog10 of the geometric mean. A single MATC is calculated for each
species based on the most sensitive sublethal or mortality exposure. Only species with significant effects are included. Data from Table 6-5.
Species
NOAEL
LOAEL
Most
Sensitive
MATC
Duration of
Exposure
Life Stage
End point
Response
Type
Citation
Necora puber
8.09
7.83
7.96
14 days
Adult
Haemolymph [HC03 ~]
extracellular
Physiological
Rastrick et al.,
2014
Chionoecetes bairdi
8.1*
7.8
7.95
up to 2 years
(adults)
Embryonic
Embryonic
morphometries (egg size
and yolk dimensions)
Development
Swiney et al,
2016
Callinectes sapidus
8.03
7.85
7.94
60 days
Juvenile
Calcification rate
Calcification
Ries et al., 2009
Homarus gammarus
8.07
7.74
7.91
5 weeks
Juvenile
Survival
Mortality
Small etal. 2016
Hyas araneus
8.0
7.8
7.90
several hours
Adult
Change in blood
haemolymph oxygen
partial pressure °C
Physiological
Walther et al.,
2009b
Paralithodes
camtschaticus
8.0
7.8
7.90
199days
Juvenile
Mortality rate
Mortality
Long et al., 2013
Homarus americanus
8.1
7.7
7.90
12 days
Larval
Days to reach larval
stage III
Development
Keppel et al.,
2012
Pandalus borealis
8.11
7.65
7.88
up to 13 days
Larval
Larval development
(Stage II, III, and IV)
Development
Arnberg et al.,
2013
Penaeus plebejus
8.03
7.72
7.88
60 days
Adult
Calcification rate
Calcification
Ries et al., 2009
Nephrops norvegicus
8.0
7.6
7.80
4 months
(adults)
Embryonic
Oxidative stress in eggs
Development
Styfetal., 2013
144
-------
Species
NOAEL
LOAEL
Most
Sensitive
MATC
Duration of
Exposure
Life Stage
End point
Response
Type
Citation
Palaemon pacificus
7.89*
7.64
7.77
15 days
Adult
% increase in length
Development
Kurihara et al.,
2008
Lysmata catifornica
7.99
7.53
7.76
21 days
Adult
Body transparency peak
range
Physiological
Taylor et al.,
2015
Petrolisthes cinctipes
7.9
7.6
7.75
40 days
Juvenile
Juvenile survival
Mortality
Ceballos-Osuna
et al., 2013
Metacarcinus magister
8.0
7.5
7.75
45 days
Larval
Larval survival
Mortality
Miller et al., 2016
Carcinus maenas
7.84*
7.36
7.60
10 weeks
Adult
Feeding rate and
behavior
Behavior
Appelhans et al.,
2012
Panopeus herbstii
8.04*
7.05
7.55
71 d ays/48 hrs
Adult
Time handling prey
Behavior
Dodd et al., 2015
Metapenaeus joyneri
8.14
6.91
7.53
1 to 10 days
Adult
Metabolic scope
Physiological
Dissanayake &
Ishimatsu, 2011
Upogebia deltaura
7.64*
7.35
7.50
35 days
Adult
Haemolymph pH
Physiological
Donohue et al.,
2012
Pagurus bernhardus
8.2
6.8
7.50
5 days
Adult
Latency to find shell
Behavior
de la Haye et al.,
2011
Cancer pagurus
7.9
7.06
7.48
ca. 3 days
Adult
Heat tolerance
Physiological
Metzger et al.,
2007
Pagurus tannreri
7.6
7.1
7.35
4 weeks
Adult
Time for prey detection
after 4 weeks' exposure
Behavior
Kim et al., 2015
145
-------
Cumulative Percent of Most Sensitive MATC
for Each Species
8.00 7.90 7.80 7.70 7.60 7.50 7.40
PH
7.30
Figure 6-1. Cumulative distribution of the MATCs for each decapod species based on all
endpoints.
Three groups are identified: high sensitivity species (red), moderate sensitivity species
(brown) and low sensitivity species (green). The 1st and 3rd quartiles are shown as an
example of using percentiles to define the sensitivity classes. Data from Table 6-6.
Table 6-7. Comprehensive pH thresholds for high, moderate, and low sensitivity decapods using most
sensitive MATCs.
Thresholds derived from an analysis of the lowest MATC values for each species (Table 6-5 and Table
6-6Table 6-7).
Sensitivity Level
Minor Risk
Low Risk
Moderate Risk
High Risk
High Sensitivity
>7.96
7.91 -7.95
7.88-7.90
<7.87
Moderate Sensitivity
>7.80
7.77-7.79
7.76-7.76
<7.75
Low Sensitivity
>7.60
7.50-7.59
7.36-7.49
<7.35
6.3.3 Ocean Acidification Population Viability Effects Thresholds
A limitation of the comprehensive effects thresholds based on all endpoints is that there is a
disconnect between the risks for ocean acidification and those for temperature and sea level rise,
which are associated with population viability. Inclusion of physiological and behavioral
endpoints in the comprehensive analysis may result in more sensitive effects thresholds (higher
MATCs) than the population viability associated risks. This general issue has been recognized
previously, and in their review of ocean acidification effects on population survival, Busch and
McElhany (2016) weighted responses with "a known relationship to population persistence"
twice as heavily as "all else". We present here an example of how ocean acidification thresholds
146
-------
more closely aligned to those for temperature and sea level rise can be generated from MATCs
based on endpoints directly linked to population viability: morality, survival, and larval
development rate. Larval development rate was included because an extension of the larval phase
can reduce recruitment because of the very high larval mortality rates (Rumrill, 1990; Pedersen
et al., 2008).
Limiting the analysis to only population viability endpoints reduced the number of studies to 10
(Table 6-8), six of which were used in calculating the comprehensive MATCs (Table 6-6). As
with the comprehensive MATCs, the values are plotted as a cumulative frequency distribution
and high, moderate, and low sensitivity groups identified (Figure 6-2). The resulting population
viability thresholds for the high sensitivity class (Table 6-9) are essentially the same as those
based on the comprehensive MATCs (Table 6-7). However, there are only two species identified
in the moderate sensitivity class, both of which have the same value (7.75). We use the 7.75 pH
as the threshold for high risk, which is the same value as with the comprehensive moderate
threshold (Table 6-7). In lieu of any other data, we use the comprehensive thresholds for minor,
low, and moderate risks as a first-order approximation. Given the small range in the moderate
thresholds with the comprehensive MATCs (7.75-7.80), thresholds based on mortality endpoints
should not be substantially different. While there are only two species in the low sensitivity
class, there is a spread in the pH values. Taking the mid-point as the threshold for upper end of
the moderate risks allows the generation the low sensitivity thresholds (Table 6-9).
While there is considerable uncertainty due to the limited number of studies, users can enter the
thresholds in Table 6-9, or other population-based thresholds, to assess how the extent and
pattern of ocean acidification risk changes by using thresholds more comparable to the
temperature and sea level rise risks.
147
-------
Table 6-8. MATCs for pH based on the population viability endpoints for each decapod species.
NOAEL = no observed adverse effect level; LOAEL = lowest observed adverse effect level; MATC = maximum acceptable toxicant concentration.
The NOAEL is the control pH or the lowest non-significant exposure. MATC is calculated as the geometric mean of the NOAEL and LOAEL,
calculated by first taking the antilog of the pHs and then taking the log 10 of the geometric mean. A single MATC is calculated for each species
based on endpoints directly related to population viability, including mortality, survival, and larval duration. Only species with significant effects are
included. Data from Table 6-5. Indicates species used in calculation of the comprehensive MATCs (Table 6-6).
Species
NOAEL
LOAEL
Most
Sensitive
Population
MATC
Duration of
Exposure
Life Stage
End Point
Response
Type
Citation
Necora puber
7.74
6.05
6.90
16 days
Adult
Mortality
Mortality
Spicer et al., 2007
Upogebia
deltaura
7.99
6.71
7.35
35 days
Adult
Mortality
Mortality
Donohue et al.,
2012
Metacarcinus
magister *
8
7.5
7.75
45 days
Larvae
Larval survival
Mortality
Miller et al., 2016
Petrolisthes
cinctipes *
7.9
7.6
7.75
40 days
Juvenile
Juvenile survival
Mortality
Ceballos-Osuna et
al., 2013
Pandalus
borealis *
8.1
7.65
7.88
up to 13 days
Larvae
Larval development (Stage
II, III, and IV)
Development
Arnberg et al.,
2013
Chionoecetes
bairdi
8
7.8
7.90
199days
Juvenile
Mortality rate
Mortality
Long et al., 2013
Homarus
americanus *
8.1
7.7
7.90
12 days
Larvae
Days to reach larval stage III
Development
Keppel et al.,
2013
Paralithodes
camtschaticus *
8
7.8
7.90
199days
Juvenile
Mortality rate
Mortality
Long et al., 2013
Homarus
gammarus *
8.07
7.74
7.91
5 weeks
Juvenile
Survival
Mortality
Small etal. 2016
Hyas araneus
8.11
7.81
7.96
Length of dev. stage
(-10 - 80 d)
Zoea I & II
Duration of larval stage
Development
Walther et al.,
2010
148
-------
Cumulative % of Most Sensitive MATC for each Species
100
90
80
70 g
u
60 «
Q.
50 >
S
m
40 3
E
30 3
u
20
10
0
8.0 7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 7.1 7.0 6.9 6.8
pH
Figure 6-2. Cumulative distribution of the MATCs based on population viability endpoints for each
decapod species.
Three groups are identified: high sensitivity species (red), moderate sensitivity species (brown) and low
sensitivity species (green). Data from Table 6-9.
Table 6-9. Population viability pH thresholds values for high, moderate, and low sensitivity decapods.
Thresholds derived from an analysis of MATC values based on mortality, survival, and larval duration for
each species (Table 6-8, Figure 6-2). See text for limitations of these values.
Sensitivity
Minor Risk
Low Risk
Moderate Risk
High Risk
High Sensitivity
>7.96
7.91 to 7.95
7.89 to 7.90
<7.88
ft/loderate Sensitivity
>7.80
7.76 to 7.79
7.76 to 7.76
<7.75
Low Sensitivity
>7.35
7.13 to 7.34
6.91 to 7.12
<6.90
6.4 Biotic Traits Modifying Sensitivity and Temperature-Adjusted Ocean Acidification
Risks
Even with the increased interest in ocean acidification, only a handful of species will be
experimentally evaluated to determine if they have high, moderate, or low sensitivity to pH
•
~—J—
/T\
0
a
Vi/
/ •
©Hyas araneus
Homarus gammarus
Paralithodes camtscha
tic us
Homarus amercanus
• Chionoecetes bair
-------
changes. A potential solution to this data deficit is to use biotic traits to identify the species most
and least sensitive to ocean acidification. A number of factors have been suggested as affecting
sensitivity including: 1) osmoregulatory ability (Whiteley, 2011); 2) species being "pre-adapted"
to low pH by their evolutionary history in upwelling regions, areas of hypoxia, and/or estuaries
(e.g., Tseng et al., 2013; Pansch et al., 2014; Heinrich et al., 2016); 3) shell structure, especially
the difference between the more soluble aragonite versus calcite (Kleypas et al., 2006; Ries,
2011); and 4) pelagic duration by affecting the exposure time of a sensitive life history stage.
However, in our evaluation of the literature, none of these patterns were sufficiently clear to
generate general rules applying to multiple taxa across subtropical to Arctic environments at this
time.
One general pattern that occurs across taxa is a tendency towards increased impacts of reduced
pHs at elevated temperatures. In their review, Kroeker et al. (2013) highlighted "a trend towards
enhanced sensitivity to acidification when taxa are concurrently exposed to elevated seawater
temperature" while Harvey et al. (2013) noted in their meta-analysis that "four of five of the
biological responses measured (calcification, photosynthesis, reproduction, and survival, but not
growth) interacted synergistically when warming and acidification were combined." Enhanced
effects of pH at higher temperatures have been reported from a variety of taxa and habitats,
including stony corals (Anlauf et al., 2011, Edmunds et al., 2012), ophiuroids (Wood et al.,
2010), decapods (e.g., Paganini et al., 2014), bivalves (Matozzo et al., 2012; Ko et al., 2014), and
fish (Munday et al., 2009). While several studies have not demonstrated such an interaction or
only a very weak interaction (e.g., Horn et al., 2016; Zhang et al., 2016), enhanced temperature
appears to exacerbate the negative effects of reduced pH more frequently than not, especially
with greater thermal stress.
To capture this interaction, we calculate the "temperature-adjusted ocean acidification risk" by
pairing the greatest individual risk factor for temperature with the greatest risk for pH/aragonite
saturation state. The temperature-adjusted ocean acidification risk is considered the overall risk
for pH/aragonite saturation state and is used in calculating the overall vulnerability for a species
(see Appendix B). The combination of risks is illustrated in Figure 6-2 based on the following
rules:
1. Minor ocean acidification risk and any temperature risk => Minor temperature-adjusted
ocean acidification risk.
2. Low ocean acidification risk and any temperature risk => Low temperature-adjusted
ocean acidification risk.
3. Moderate ocean acidification risk and Minor temperature risk => Moderate
temperature-adjusted ocean acidification risk.
150
-------
4. Moderate ocean acidification risk and Low temperature risk => Moderate temperature-
adjusted ocean acidification.
5. Moderate ocean acidification risk and Moderate temperature risk => High temperature-
adjusted ocean acidification.
6. Moderate ocean acidification risk and High temperature risk => High temperature-
adjusted ocean acidification.
7. High ocean acidification risk and any temperature risk => High temperature-adjusted
ocean acidification.
These rules are based on the concept that minor and low temperature risks are not sufficiently
detrimental to elevate the ocean acidification risk. For example, the low temperature risk with the
ETW approach is defined as a projected temperature less than mean historical temperature and
two standard deviations, which occurs frequently (see Section 5.3.1). At the other extreme,
temperatures associated with moderate and high temperature risk occur infrequently, and are
presumably sufficiently stressful to aggravate effects of ocean acidification.
Ocean Acidification Risk
Minor
Low
Moderate
High
o
L.
Minor
0
-1
-2
-3
mperatu
Risk
Low
0
-1
-2
-3
Moderate
0
-1
-3
-3
V
1-
High
0
-1
-3
-3
Figure 6-3. Temperature-Adjusted Ocean Acidification Risks.
This figure illustrates the value of the "temperature-adjusted ocean acidification risk" based on the
interaction between the ocean acidification risk and the greatest temperature risk. Ocean acidification risk
is the larger of the risks associated with pH oraragonite saturation state. Moderate ocean acidification
risk is elevated to high risk under conditions of moderate or high temperature risk. Color key for the
temperature-adjusted ocean acidification risk: minor risk = green, low risk = yellow, moderate risk =
orange, high risk = red.
Another apparent pattern is that species with non-feeding larvae or offering some sort of
protection to larval/juvenile stages are more resistant to ocean acidification:
1) A field study assessed the life history traits of species along a pH gradient created by
a shallow vent system in the Mediterranean (Lucey et al., 2015). All 13 polychaetes
species occurring at the lowest pH site (6.4-7.8 pH) had some type of brooding and/or
direct development. The authors concluded, "long-term survival of marine species in
151
-------
acidic conditions is related to life history strategies where eggs are kept in protected
maternal environments (brooders) or where larvae have no free swimming phases
(direct developers)."
2) The larvae and juveniles of the lecithotrophic sea star Crossaster papposus grew
faster at the low pH exposure (7.7), and there was no effect on survival or
skeletogenesis (Dupont et al., 2010).
3) The eggs of the cuttlefish Sepia officinalis, which has lecithotrophic development,
developed successfully and developed an aragonite shell under low pH conditions
within the egg (Gutowska and Melzner, 2009).
4) Waldbusser et al. (2016) found that the brooding oyster Ostrea lurida was less
sensitive to low pH than Crassostrea gigas, which has planktotrophic larvae. The
actual cause for the enhanced resistance was the slower development of the O. lurida
embryos rather than brooding per se. Nonetheless, we assume that many if not most
brooding bivalves have slower embryo growth than their counterparts with
planktotrophic larvae, and should have similar resistance to lower pH.
5) Haliotis rufescens, the red abalone, has a lecithotrophic larvae. The expression pattern
of the two shell formation genes in the early life history stages were not affected at a
pH of 7.87 (Zippay and Hofmann, 2010). However, there was an interaction with
temperature, and two of the early life history stages had reduced thermal tolerance
with higher pH exposure.
Besides the pH exposures, several authors have suggested that species with lecithotrophic
development have a lower vulnerability to environmental perturbations than those with strict
planktotrophic development. Byrne (2011) suggested that invertebrates "may have evolved a
buffered non-feeding larval life history, free of the vagaries of planktonic food supply in
response to stressful conditions in the plankton". Supporting evidence included the extinction of
several planktotrophic lineages during paleo-climatic events (Valentine and Jablonski 1986,
Pechenik 1999, Uthicke et al. 2009). Based on their studies with Crossaster papposus, Dupont et
al. (2010) postulated that "lecithotrophy may be an advantage in an unpredictable and extreme
environment".
Based on these studies, we incorporate breeding and larval type as conditional factors modifying
the risk associated with coastal acidification. If the target species has any of the following
breeding or developmental strategies, the species is automatically assigned to the low sensitivity
class:
Life History +• Development +• Breeding Strategy +• Ovoviviparous
Life History +• Development +• Breeding Strategy +• Oviparous +• Eggs brooded in tube
152
-------
Life History +• Development +• Juvenile Development +• Direct Development
Life History +• Development +• Larval Phase +• Planktonic Larvae +• Planktonic-lecithotrophy
Life History +• Development +• Larval Phase +• Benthic Larvae +• Brooded
Life History +• Development +• Larval Phase +• Benthic Larvae +• Benthic-lecithotrophy
Several invertebrate species (Allen and Pernet, 2007) as well as rockfish (Berkeley et al., 2004)
display a mixed reproductive strategy, or facultative planktotrophy, where larvae have oil
reserves but actively feed as the larvae mature. Such a mixed strategy is coded in CBRAT by
selecting both planktotrophy and lecithotrophy; in terms of the risk calculation this mixed
reproductive strategy is treated as lecithotrophic.
As with nearly every generalization about marine species, there may be exceptions to this rule.
The gastropod Crepidula fornicata broods its embryos within capsules maintained by the
females. Based on a set of exposure experiments, it was concluded that encapsulation did not
protect them against lower pH (Noisette et al., 2014), though the authors noted "C. fornicata
larvae seemed less affected than other mollusk species." Another possible exception are small,
thin-shelled brooders, such as the bivalve Carditella marieta. To accommodate such exceptions,
users can change the sensitivity class on a species-by-species basis.
6.5 Risk Type and Risk Algorithm
6.5.1 Risk Type
Taxa vary in whether reductions in pH or aragonite saturation state is the primary stressor.
Hermatypic corals are very sensitive to reductions in aragonite saturation state to the point of
having their exoskeleton literally dissolve (e.g., Cohen and Holcomb, 2009). Larval and juvenile
bivalves are also sensitive to reductions in aragonite saturation state (Waldbusser et al., 2015).
As pointed out by (Whiteley, 2011), the calcification process in crustaceans is likely to be less
susceptible to ocean acidication than with bivalves because the exoskeletal is mostly composed
of calcite rather than aragonite. Based on this, we assign pH rather than aragonite saturation state
as the major stressor for crustaceans. However, this conclusion should be experimentally tested
since crustacean larvae initially deposit soluble amorphous calcium carbonate and high-
magnesium calcite (Ross et al., 2011). pH is also assigned as the main stressor for fish and other
taxa lacking carbonate shells (e.g., most polychaetes).
To assign the appropriate stressor in the risk assessments, pH or aragonite saturation state is
identified in CBRAT as the primary stressor for major taxa. For example, aragonite saturation
state is assigned as the primary stressor for bivalves while pH is assigned as the primary stressor
for decapods and fish. The major stressor assigned for polychaetes is pH, though serpulid
153
-------
polychaetes, as well as some sabellids and cirratulids, create calcium carbonate tubes. To address
this type of variation, individual families or species can be assigned a different ocean
acidification stressor. Additionally, both pH and aragonite can be identified as major stressors in
CBRAT, in which case the ocean acidification vulnerability is based on the greater of the two
risks.
6.5.2 Risk Algorithm and Assignment of Sensitivity Classes
The ocean acidification risk algorithm compares the ecoregion-specific projected pH and
aragonite saturation state values to the respective thresholds for the target species. For pH,
comparisons are conducted for projected annual, summer, and winter values, while only annual
values are available for the aragonite saturation state. Only the risk associated with the primary
stressor(s) for the assigned sensitivity class is output in the vulnerability summary (see Appendix
B), with the temperature-adjusted ocean acidication risk used in calculating the overall risk for a
species.
Based on a preliminary risk assessment with decapods, assignment of the high, moderate, or low
sensitivity threshold to a species has a major effect on its ocean acidification risk assignment. In
even the most optimistic scenario, only a limited number of species within a taxon will be tested
experimentally, necessitating assigning sensitivity classes in the absence of direct evidence. One
approach is to classify species by their similarity to experimentally tested species, taking into
account both taxonomic and ecological similarities. Another approach is to use moderate
sensitivity effects thresholds as a "restrained" analysis and the high sensitivity thresholds as a
"high risk" analysis. Currently, moderate sensitivity is used as the default in CBRAT, though we
suggest users evaluate risk using different sensitivity classes.
154
-------
Section 7.
Sea Level Rise
7.1 Introduction
Sea level rise (SLR) is a threat to near-coastal biotic communities as well as human well-being
and infrastructure (e.g., NRC, 2012; Wong et al., 2014;
http://www.corpsclimate.us/ccaceslciirves.cfm). Inundation of coastal lands could put upwards of
1.8 to 7.4 million people at risk in the United States and the GDP "could potentially decline by
USD 70-289 billion" (Haer et al., 2013). There is also concern regarding the effects of sea level
rise on intertidal habitats, such as wetlands and mangroves. Loss of these habitats puts species
that depend upon them at risk. Perhaps the most notable example of such an impact is that SLR
has resulted in the extirpation of the Australian Bramble Cay melomys, Melomys rubicola, from
its only known habitat (Gynther et al., 2016), the first documented case of a mammal extinction
due to climate change.
In assessing the potential impacts of climate change, it is important to distinguish between global
or eustatic SLR and local or relative SLR. Eustatic SLR is the global rise in the ocean level due
to changes in the volume of ocean water. There is only one eustatic SLR value for all the oceans.
Based on observed contributions to SLR from 1993 to 2010 (Church et al., 2013), the
contributors to observed eustatic sea level rise were:
• -34 % thermosteric expansion of sea water from the increased heat content of the ocean
• -24% Glaciers except in Greenland and Antarctica
• - 15.4% Glaciers and ice sheet in Greenland
• ~8%> Antarctic ice sheet
• -12%) Changes in land water storage
• -13%) Other and unexplained
The biggest uncertainties in predicting future eustatic rates are the extent of melting of the large
ice sheets of Greenland and Antarctica (Nicholls et al., 2011).
The actual extent of SLR at a location is modified by several local or regional factors, and the net
change in sea level at any particular location due to both eustatic SLR and local factors is often
155
-------
referred to as the relative sea level rise (RSLR). In most cases, the most important of these local
factors is isostatic or tectonic effects, which is the subsidence or uplift of the land relative to
mean sea level. In some locations, uplift is sufficient to offset the predicted levels of SLR, while
in other locations subsidence exacerbates SLR. Other local factors, such as wind, storms, and
barometric pressure, can also modify sea level (e.g., NRC, 2012). The effects of these factors are
usually temporary, and are not considered here.
In this section, we will first describe our general approach to predicting the relative risk of
species to sea level rise on an ecoregional scale. As with other climate drivers, we focus on the
regional loss of intertidal habitats and its regional effect on the associated target species. This
regional approach does not have the detail of localized models, such as SLAMM (e.g., Glick et
al., 2007) but we contend that it has sufficient resolution to identify both the species at greatest
risk and how SLR risk varies along the coast. In the remainder of the section, we provide a
synthesis of available information used to generate default input values for CBRAT.
7.2 Overview of SLR Approach
The SLR procedure consists of four steps (Figure 7-1). The first is to estimate a relative or net
ecoregion sea level rise value (mm) by adding the isostatic rate of a particular ecoregion to the
global eustatic SLR rate; that value is then multiplied by the number of years being modeled to
generate an estimate of net sea level rise (Sections 7.3-7.5). The second step is to estimate the
percentages of each occupied habitat that will be lost to SLR in each ecoregion which are the
basis of generating habitat threshold values (classes of the percent of habitat lost with SLR
values; Section 7.8). To account for the potential of intertidal habitats to migrate inland, habitat
thresholds are developed for both "constrained" and "unconstrained" scenarios (Section 7.6).
The third step is to generate risk values for the target species for each occupied habitat from the
habitat thresholds and depth preferences of the species (Section 7.9.1). For species occupying
multiple habitats, the fourth step is to modify the risk factors generated in step 3 based on its
habitat preferences (Section 7.9.2). As with other risks, the SLR risks are classified from
"Minor" (0) to "High" (-3). While the approach should be generally applicable in other areas, the
habitat thresholds were calibrated for the Northeastern Pacific and would likely need to be
adjusted for other regions.
156
-------
Steps for determining a species risk to Sea level Rise (SLR)
Calculate
Relative
SLR =
(Eustatic
SLR +
Isostatic
SLR) X
Duration
Use Net SLR
projections
to determine
a risk class
for each
habitat the
species
occupies
A
Use a spp.
Depth
preference
to
determine
the risk for
each
habitat
J V
The overall
species risk is
determined by
the sp.'s
habitat
preference.
The highest
risk across
multiple
habitats is the
final risk
A
R
e
I
a
t
¦
i
v
e
V_
Figure 7-1. Generalized sea level rise approach to calculating relative risk.
Step 1 determines the ecoregion-specific relative sea level rise in mm. Step 2 determines the risk class
for each occupied habitat based on the percent of habitat lost by comparing habitat threshold values to
the predicted SLR. Different habitat thresholds are used in ecoregions where inland migration of habitats
is limited due to barriers (constrained) versus ecoregions where there are few barriers to inland migration
("unconstrained"). Step 3 generates the risk values for the target species based on the species' depth
preferences. Step 4 determines the final risk value based on the species' habitat preferences, with the
final risk factor based on the greatest risk value across habitats.
7.3 Eustatic Rates
The first input into the SLR risk analysis is the eustatic sea level rise rate (mm/yr). There is
considerable uncertainty regarding future levels of sea level rise. As pointed out by Parris et aL
2012, "Scenarios do not predict future changes, but describe future potential conditions in a
manner that supports decision-making under conditions of uncertainty." As reasonable default
values, we use the rates generated by the NOAA (Parris et al., 2012; Table 7-1) with the
modification that the lowest scenario from Parris et al. (2012) was increased from 2 mm/yr to 3.3
mm/yr based on more current estimates of recent SLR (Fiissel, 2009; NRC, 2012).
157
-------
Table 7-1. Eustatic sea-level rise scenarios used as default values for ecoregion SLR risk analysis.
Table modified from Parris et al. (2012). Parris et al. used a value 0.2 m SLR by 2100 for the lowest
scenario, however we use a value of 0.33 m based on recent sea level data from tide gauges and
satellites (Fiissel, 2009; NRC, 2012).
Scenarios
SLR by 2100 (m)
(mm/yr assuming
100 years)
Source / Applications
Highest
2.0 m
(20 mm/yr)
"Our Highest Scenario of global SLR by 2100 is derived from a
combination of estimated ocean warming from the IPCC AR4 global SLR
projections and a calculation of the maximum possible glacier and ice
sheet loss by the end of the century. The Highest Scenario should be
considered in situations where there is little tolerance for risk".
Intermediate-High
1.2 m
(12 mm/yr)
"based on an average of the high end of semi-empirical, global SLR
projections. ... The Intermediate-High Scenario allows experts and
decision makers to assess risk from limited ice sheet loss."
Intermediate-Low
0.5 m
(5 mm/yr)
"based on the upper end of IPCC Fourth Assessment Report (AR4) global
SLR projections resulting from climate models using the B1 emissions
scenario. ... The Intermediate Low Scenario allows experts and decision
makers to assess risk primarily from ocean warming."
Lowest
0.33 m
(3.3 mm/yr)
"based on a linear extrapolation of the historical SLR rate derived from
tide gauge records beginning in 1900 (1.7 mm/yr). The Lowest Scenario
should be considered where there is a great tolerance for risk." [Note: We
suggest a value of 0.33 m versus the 0.2 m in Parris et al., 2012 based on
recent observed SLR rates.]
7.4 Regional Isostatic Rates
The eustatic rate of sea level rise is modified locally by a number of factors, the most important
of which are isostatic adjustments. To account for isostatic adjustments, we generated average
isostatic rates (mm/yr) for each ecoregion. The ecoregion-specific isostatic rate input into
CBRAT is multiplied by the duration being modeled, and then this ecoregion-specific adjustment
(mm) is added to the eustatic sea level to generate a projected relative sea level rise (mm) for the
ecoregion.
For 10 of the 12 ecoregions, the average isostatic value was calculated by first determining the
observed historic sea level rise trend at all the sites within the ecoregion. Then, the historic
eustatic sea level rise rate was subtracted from each of the observed trends, with the difference
assigned as the isostatic rate for that location. These values were averaged for all of the sites
within an ecoregion to generate the ecoregion-specific isostatic rate (Table 7-2). For these ten
ecoregions, the observed historic sea level trends were downloaded from NOAA's Sea Level
Trends site (http://tidesandcurrents.noaa.gov/sltrends/sltrends.html; see Zervas, 2009). For the
historic eustatic rate, we use a rate of 1.7 mm/yr, the global average between 1901 and 2010
158
-------
reported by the IPCC (Church et al., 2013), which has a reported range of 1.5 to 1.9 mm/yr.
Because the NOAA Portal does not report sea level trend values for the Magdalena Transition
Ecoregion, we took the average of the locations to the north and south, Ensenada and Cabo San
Lucas, as an approximation. Sea level rise trends are also not available for the Chukchi Sea from
the NOAA Portal. For this ecoregion, we subtracted the eustatic rate (1.7 mm/yr) from the
average observed sea level trend for five sites on the Russian side of the Chukchi from
Proshutinsky et al. (2004; their Table 3).
Several of the ecoregions had sites with uplift and others with subsidence. Since our objective is
to predict habitat loss at the ecoregion level, such within-ecoregion variation should not
introduce a substantial error in terms of estimating the overall habitat available. As an example,
the Oregonian ecoregion has about a quarter of the sites show uplift and the others showing
subsidence. We provide the average isostatic values for these two groups of sites (Table 7-2),
which can be used to model the range of possibilities of isostatic adjustment within this
ecoregion.
159
-------
Table 7-2. Derivation of ecoregion-specific isostatic rates.
Except as noted, historical sea level trend values were extracted for each site within an ecoregion from
the NOAA Sea Level Trends portal (httpi//tidesandcurrents.noaa.gov/sltrends/slrmap.htm). The
ecoregion-scale isostatic rate was calculated by averaging the historic SLR trends across all the sites
within each ecoregion and then subtracting the historic eustatic sea level rise rate (1.7 mm/yr, Church et
al., 2013). The value for the Magdalena is the average of the values for the two closest sites, Ensenada
and Cabo San Lucas, Mexico. The Chukchi value is the average observed SLR trend from Proshutinsky
et al., 2004 ("Observations" in their Table 3) minus the 1.7 mm/yr eustatic rate. Negative isostatic values
indicate uplift, while positive values indicate subsidence. For the Oregonian ecoregion, the average
isostatic rates are also given separately for sites experiencing uplift versus subsidence (italicized).
NA= no data.
ECOREGION
# Sites
Average Historical RSLR
(Not Isostatically Corrected)
(mm/yr)
Average Ecoregion
Isostatic Rate
(mm/yr)
# Sites in Ecoregion
Showing Historic Uplift
(Negative Values)
Beaufort
1
1.20
-0.50
1
Chukchi
5
1.90
0.20
1
Bering
2
2.72
1.02
0
Aleutians
2
-4.11
-5.81
2
Gulf of Alaska
9
-6.41
-8.11
9
Pacific Fjords
5
-6.44
-8.14
5
Puget
7
0.83
-0.87
6
Oregonian
11
0.33
-1.37
8
Oregonian
3
2.87
1.17
0
Oregonian
8
-0.62
-2.32
8
N. California
8
1.20
-0.50
5
S. California
8
1.88
0.18
3
Magdalena
0
2.01
0.31
ND
Cortezian
2
2.89
1.19
1
160
-------
7.5 Duration
Duration, in years, is input to convert the eustatic and isostatic rates (mm/yr) into a total relative
sea level rise value (mm). While it is possible to input any duration, the model was designed and
calibrated for 100 years, and we caution about using other durations at this time. Some studies
have predicted an initial increase in low wetlands over approximately fifty years but then a
decline by 100 years (e.g., Stralberg et al., 2011; Thorne et al., 2015). The default habitat
thresholds for emergent marshes are based on the effects over 100 years, and in this case would
overestimate the loss of lower wetlands over 50 years. If durations other than 100 years are
modeled, it is important to check the applicability of the habitat thresholds.
7.6 Constrained Versus Unconstrained Habitats
The effect of sea level rise on intertidal habitats is alleviated if the habitat can migrate inland as
the water level rises. Conversely, habitats surrounded by anthropogenic barriers like rip-rap,
dikes, armoring, and seawalls, or natural barriers, like cliffs, are more vulnerable to SLR. The
prevention of intertidal habitats to migrate inland with SLR is referred to as "coastal squeeze"
(e.g., Short et al., 2016). Coastal squeeze is most important for vegetated and unvegetated soft-
sediment habitats while anthropogenic barriers do not appear to be a major factor limiting rocky
intertidal assemblages. SLR models vary in how coastal squeeze is parametrized; some studies
analyzed just the intertidal land seaward of dikes or other barriers while other studies allowed the
model to consider land behind dikes or barriers as potential area for habitat expansion. Because
these considerations result in large differences in potential habitat expansion, we analyze the
'unconstrained' and 'constrained' scenarios separately.
Several researchers conducted their analyses using both constrained and unconstrained GIS
layers to demonstrate restoration potential if barriers were removed (e.g. Stralberg et al. 2011;
Warren Pinnacle Consulting, Inc. 2011). These side-by-side comparisons provide an opportunity
to compare projected differences in habitat area loss with and without physical barriers. In
Stralberg et al. 2011 the difference between constrained and unconstrained (based on dike
removal) ranged from a decrease in 24% of low marsh with the constrained layer to an increase
40% low marsh under the unconstrained data layer.
Besides the effects at a local scale, the concept of constrained versus unconstrained can be
applied at a regional scale. The coastal shorelines and estuaries in the Puget, Oregonian,
Northern California, and Southern California ecoregions are subject to moderate to extensive
shoreline modifications (e.g., Dugan and Hubbard, 2010; Hubbard et al., 2014; Washington State
Department of Natural Resources, no date; Myers, 2010). Approximately a third of Puget
Sound's shoreline has been anthropogenically modified (Washington State ShoreZone
Inventory). In addition to these anthropogenic barriers, natural cliffs are a common shoreline
feature in Washington, Oregon and California. Because of the frequency of these barriers, we
apply the constrained habitat thresholds for soft-sediment habitats to these ecoregions (Table
161
-------
7-3). However, with their lower population densities and extents of coastal development, we
apply the unconstrained habitat thresholds to the other eight ecoregions. The consequence of
using these different habitat thresholds is that the model predicts a greater risk for the same net
sea level rise in the four developed ecoregions compared to the less developed ecoregions.
A different type of limitation to migration is that some models are "bounded". That is, they
model changes in the relative percent change of different habitat types within a defined, or
bounded, area. As such, these model do not incorporate the possibility of landward migration. As
discussed below, one such bounded model is that of Thorne et al. (2015).
7.7 High and Low Exposure Habitats
Intertidal habitats that will experience the effects of sea level rise are termed "high exposure"
habitats. There is another suite of habitats for which SLR will have no or only a trivial impact,
which we refer to as "low exposure" habitats. From the Level I habitats in CBRAT, we identify
Terrestrial, Pelagic Ecosystems, and Specialized Systems as low exposure. From the Level II
habitats, Unvegetated Subtidal, Kelp, Coral, Subtidal Rocky, Non-coral reefs, Solitary sponge,
Bryozoan mats, and Rhodoliths/Maerl are classified as low exposure. A minor risk is assigned to
these low exposure habitats by setting the habitat thresholds to the maximum projected sea level
rise (80.32 m) resulting from the melting of all the glaciers in the Antarctic, Greenland, and other
ice fields (Poore et al., 2000). These habitat thresholds set the extent of habitat loss to zero
except with the most extreme SLR projections.
While a few of the low exposure habitats may occasionally occur in the intertidal, such as
coralline algal mats, the vast majority of them occur subtidally (e.g., see Chenelot et al., 2008 for
subtidal coralline mats in the Aleutians). Thus, functionally assigning them a minor risk by
applying the upper bound SLR thresholds should introduce a negligible underestimation of the
impacts. The low exposure habitats are based on the NEP and U.S. Arctic and it may be
necessary to modify the list for other geographical areas. For example, corals are subtidal in the
Gulf of California but may form intertidal assemblages in other regions (Richards et al., 2015).
7.8 Habitat Thresholds
7.8.1 Introduction
A key step to predicting SLR impacts on populations of intertidal species is to approximate the
percent habitat loss for each of the habitats the species occupy. We approximate these losses by
generating "habitat thresholds" for each of the Level II habitat types in CBRAT (Table 7-3),
where habitat thresholds are the net sea level rise, in mm, that result in different ranges of
percent losses of specific habitats when averaged across the ecoregion. The concept of the
habitat thresholds is that different habitat types vary in their vulnerability to SLR due to factors
such as geomorphology and coastal slope, which are similar to the factors Thieler and Hammar-
162
-------
Klose (2000) used in developing their coastal vulnerability index. As pointed out by these
authors, "the relative vulnerability of different coastal environments to sea-level rise may be
quantified at a regional to national scale using basic information on coastal geomorphology, rate
of sea-level rise, past shoreline evolution, and other factors." The concept of habitat thresholds is
also similar in spirit to SLR estimates used by other authors to identify critical levels of SLR rise
for different habitats (e.g., Morris et al., 2002; Blankespoor et al., 2012).
Table 7-3. Habitat thresholds associated with different levels of percent habitat loss.
Habitat thresholds are the levels of net SLR (mm) for major habitats that define different percent habitat
loss classes. Habitat threshold classes are: minor (<10% loss), low (11% to 29% loss), moderate (30% to
49% loss), and high (>50% loss). Minor is expressed as a loss but under some scenarios habitat area
may increase with these levels of SLR. As appropriate, both constrained and unconstrained habitat
thresholds are presented. Constrained habitats are impeded from inland migration due to artificial and
natural barriers, while unconstrained habitats are not impeded. High exposure systems are intertidal
habitats that would be affected by SLR. Low exposure systems are primarily subtidal and pelagic habitats
that are essentially immune to all but the most extreme sea level rise. Maximum sea level rise values from
Poore et al. (2000) are assigned to the low exposure habitats. Blue values = SLR values equal to or
greater than NOAA's Intermediate-High scenario; Red values = SLR values equal to or greater than
NOAA's Highest scenario (Table 7-1).
High Exposure Habitat Classes
Habitat
Minor
(<10% loss)
Low
(11 to 29% loss)
Moderate
(30 to 49% loss)
High
(>50% loss)
Con-
strained
(mm)
Uncon-
strained
(mm)
Con-
strained
(mm)
Uncon-
strained
(mm)
Con-
strained
(mm)
Uncon-
strained
(mm)
Con-
strained
(mm)
Uncon-
strained
(mm)
Oyster Beds
340
390
690
1000
770
2250
>770
>2250
Tide Flats
340
390
690
1000
770
2250
>770
>2250
Low Marsh
160
2500
790
2750
1420
3000
>1420
>3000
Rocky
Intertidal
-
400
-
800
-
1400
-
>1400
Mussel Beds
-
400
-
800
-
1400
-
>1400
SAV
540
1080
720
1440
900
1800
>900
>1800
Coastal
Beaches
550
650
600
800
800
1000
>800
>1000
Mangrove
-
750
-
1150
-
1600
-
>1600
Low Exposure Habitat Classes
Low
Exposure
Habitats
81000
82000
83000
>83000
163
-------
The literature synthesized for the derivation of the habitat thresholds by major habitat type is
summarized below (Sections 7.8.2 to 7.8.7). There is considerable uncertainty in several of the
thresholds, such as for the unconstrained lower marsh. Such divergent results can be attributed,
in part, to the differences in the modeling assumptions, which were influenced by the goals of
each analyses. If, for example, the model is allowed to expand into the adjacent grids based on
elevation alone, without incorporating real world barriers (e.g., dikes, levees, roads and
seawalls), the habitats will increase continuously until elevation limits landward migration.
Conversely, if the models use a predefined area (bounded) without any manmade structures, as in
Thorne et al. (2015), the future habitat change more likely approaches reality.
Many of the unconstrained modeling efforts that show large increases in habitat do not consider
future changes that are likely to exacerbate impacts from sea level rise such as increased
armoring and land subsidence due to aquifer depletion with ever increasing coastal populations
(California Natural Resources Agency, 2014). We contend that including current and potential
future blockages to habitat expansion will provide a more realistic ('pessimistic") scenario of
habitat changes with sea level rise. Accordingly, as detailed under emergent marshes (Section
7.8.4) and tide flats (Section 7.8.6), we derive the habitat thresholds based only on the modeling
results predicting habitat losses. Further, we note that the purpose of the habitat thresholds is as a
metric to help approximate the population loss of the associated species, and not as a habitat
model per se. Thus, the uncertainty in the specific value of the thresholds has less of an impact
on the risk values, in particular for the high threshold values exceeding likely SLR scenarios
(Table 7-1).
Table 7-3 summarizes both constrained and unconstrained habitat thresholds using four classes
based on the percent of habitat loss: minor (<10% loss), low (11% to 29% loss), moderate (30%
to 49% loss), and high (>50% loss). To the extent practical, the percentages defining the habitat
threshold classes were harmonized with the percent population changes used in the population
trends as shown in Table 7-4.
164
-------
Table 7-4. Habitat threshold classes based on the percentage of habitat lost to sea level rise.
The Minor class captures cases of minor decreases (to 10%), as well as the possibility of habitat
increases with SLR. The corresponding population trends classes for individual invertebrate and fish
species are shown in the last column (see Section 4.3.4).
Habitat Threshold
Class
Percent Loss in
Habitat Area
Corresponding Population Trend Classes
Minor
0% to -10%
Order of Magnitude Increase (>10x increase)
to
No Apparent Trend (-29% to +29%)
Low
-11% to -29%
No Apparent Trend (-29% to +29%)
Moderate
-30% to -49%
Moderate Decrease (-30 to -49%)
High
> -50%
Substantial Decrease (-50 to -79%)
and
Extreme Decline (-80 to -99%)
7.8.2 Rocky Intertidal and Mussel Beds
Rocky intertidal habitats will be inundated with a rising ocean, however there is no consensus on
the extent of the impact or how to approach the problem. In part, this reflects that less attention
has been paid to the effects of SLR on rocky shores compared to marshes and other soft-
sediment habitats. In addressing the vulnerability of this habitat, a key question is whether there
is suitable hard substrate upward of the existing rocky intertidal assemblage for upward
migration with SLR. Based on limited information, it has been suggested that in areas where the
seashore is mostly uniform, the risk to coastal squeeze is low (Kendall et al., 2004) but greater in
rocky seashores that have steep inclines and/or are backed by hard cliffs (Jackson and
Mcllvenny, 2011).
We found only a handful of studies assessing the role of SLR on rocky intertidal habitats on the
Pacific Coast. In their development of a "coastal vulnerability index" for the Pacific Coast,
Thieler and Hammar-Klose (2000) listed "rocky, cliffed coasts" as very low vulnerability,
"medium cliffs" as low vulnerability, and "low cliffs" as moderate vulnerability. In comparison,
estuaries, mud flats, salt marshes, mangroves, and other soft-sediment habitats were assigned
"high" or "very high" vulnerabilities. Glick et al. (2007) used SLAMM 5.0 to predict SLR
effects on multiple habitat types in Puget Sound and along the coast of SW Washington and NW
Oregon. For the rocky intertidal, they predicted an average 13% loss with a SLR of 0.28m, a
34% loss with a 0.69 m SLR, and a 70% loss with a 1.5 m SLR (Table 7-5). However, our
evaluation of SLAMM Ver. 5 indicated that it does not incorporate the area of the hard substrate
above the existing rocky intertidal assemblage (see Clough, 2008). Thus, these are losses with no
upward migration potential (bounded predictions) and thereby represent upper estimates of rocky
intertidal habitat loss. A Pacific Northwest study that did account for upward migration was an
165
-------
analysis of the loss of rocky intertidal foraging habitat for the black oystercatcher (Haematopus
bachmani) at Rabbit Rock, Oregon (Hollenbeck et al., 2014). Using terrestrial laser scanning
(TLS), the authors predicted that 10.9% of the rocky intertidal will disappear with aim SLR
and 57.3% will disappear with a 2 m SLR (Hollenbeck et al., 2014).
Because of the paucity of information for the Pacific Coast, we developed a GIS-based approach
using LIDAR derived digital elevation models (DEMs) to estimate the potential area of the hard
substrate upward of the existing intertidal assemblage (Clinton and Lee, 2016). The required data
were available for the Southern California Bight, Northern California, Oregonian, and Puget
Trough/Georgia Basin ecoregions. In this initial analysis, we include the rocky intertidal mussel
beds with the general rocky intertidal habitat thresholds. The GIS methodology and the metadata
for this analysis are available in Appendix C-4.
Using this model, the percent of rocky intertidal habitat loss is estimated independently for each
of the four ecoregions (Table 7-5). However, until this approach is further evaluated, we believe
that it is more appropriate to use the average of the four ecoregions rather than ecoregion-
specific values. Based on the averages, the rocky intertidal habitat thresholds are: minor <0.4 m,
low > 0.4 m, moderate > 0.8 m, and high > 1.4 m (Table 7-3). Though we advocate using the
averages, the ranges across the four ecoregions can be used as an estimate of uncertainty in
evaluating different scenarios. Because the required GIS data are not available in other locations,
we use the averages of these four ecoregions as first-order estimates for the other eight
ecoregions.
To compare our results from the LIDAR analysis, we evaluated rocky intertidal studies in
Scotland and Australia. Jackson and Mcllvenny (2011), using a modeling study, stated that with
a 0.3 m SLR 10-27%) of the rocky intertidal habitat in Scotland would be lost and at 1.9 m SLR,
26-50%) would disappear. Thorner et al. (2014) combined LIDAR with high-resolution digital
imagery in a study of five rock reefs in Australia. It is difficult to directly compare our regional-
scale results with this localized study, in part, because they evaluated vulnerability in seven
categories of rocky shore habitats (Table 7-5). Nonetheless, it appears that some of the
Australian habitats are more sensitive to SLR than our analysis suggests. In particular, their
"deep pools" habitat disappear at four of the five sites at 1 m SLR, a SLR that our LIDAR
analysis would classify as moderate (30%> to 49% habitat loss). Others of their habitats, such as
the upper boulder field and lower platform, have generally similar responses to those predicted
by our analysis.
Given the general agreement of previous studies with our analysis, we consider our habitat
thresholds based on LIDAR analysis sufficient for a first-order analysis of rocky intertidal
habitats on a regional scale. Our approach does not assess risk to specific types of rocky
intertidal habitats, such as tide pools, which are likely to be more vulnerable (Thorner et al.,
2014). Both tide pools and supralittoral splash pools are identified as specialized habitats in
166
-------
CBRAT, and we recommend that the habitat preferences of rocky intertidal species be evaluated
for utilization of these habitats to identify which species may have a greater risk than predicted.
Table 7-5. Studies predicting percentage loss of rocky intertidal habitat due to sea level rise.
For Glick et al., 2007, we present both the individual values and the combined values for different SLR
scenarios for the five sites. The combined values are those reported by Glick et al., 2007. ForThorner et
al., 2014, the predicted percent lost was estimated from their Figure 2 that depicted the percentage of
area covered for seven habitat types at five different study sites. Their graphs were divided into quartiles
and the results were summarized by counting the number of sites that fell within each quartile. These
results are displayed as the number of habitats in each quartile with the highest percent habitat loss on
top and the lowest percent habitat loss on the bottom. The habitat threshold classes are Minor: 0-10%,
Low: 11-29%, Moderate: 30-49%, and High: >50% habitat loss. The average values for the EPA LIDAR
analysis and Glick et al. combined results are highlighted. NI=No Information. TLS = terrestrial laser
scanning. Sources: 1 = EPA LIDAR Analysis, 2 = Glick et al., 2007, 3 = Hollenbeck et al., 2014, 4 =
Jackson and Mcllvenny, 2011,5 = Thorner et al., 2014.
Source
Location
Ecoregion
Habitat
SLR in
2100
(mm)
Percent
Habitat
Loss
Type of
Study
1
Puget
Puget Trough/
Georgia Basin
Rocky
Intertidal
200
12.8
Modeling
(LIDAR)
1
Oregonian
Oregonian
Rocky
Intertidal
900
13.4
Modeling
(LIDAR)
1
N. CA
Northern California
Rocky
Intertidal
400
13.7
Modeling
(LIDAR)
1
S. CA
S. California Bight
Rocky
Intertidal
200
17.6
Modeling
(LIDAR)
1
Average Low Habitat
Loss
EPA LIDAR
Puget Trough thru
S. California
Rocky
Intertidal
425
14.4
Modeling
(LIDAR)
1
Puget
Puget Trough/
Georgia Basin
Rocky
Intertidal
500
29.7
Modeling
(LIDAR)
1
Oregonian
Oregonian
Rocky
Intertidal
1400
32.6
Modeling
(LIDAR)
1
N. CA
Northern California
Rocky
Intertidal
800
31.2
Modeling
(LIDAR)
1
S. CA
S. California Bight
Rocky
Intertidal
400
34.2
Modeling
(LIDAR)
1
Average Moderate Habitat
Loss
EPA LIDAR
Puget Trough thru
Southern California
Rocky
Intertidal
775
31.9
Modeling
(LIDAR)
1
Puget
Puget Trough/
Georgia Basin
Rocky
Intertidal
Nl
Nl
Modeling
(LIDAR)
167
-------
Source
Location
Ecoregion
Habitat
SLR in
2100
(mm)
Percent
Habitat
Loss
Type of
Study
1
Oregonian
Oregonian
Rocky
Intertidal
2000
50.1
Modeling
(LIDAR)
1
N. CA
Northern California
Rocky
Intertidal
1500
50.9
Modeling
(LIDAR)
1
S. CA
Southern California
Bight
Rocky
Intertidal
800
50.7
Modeling
(LIDAR)
1
Average High Habitat
Loss
EPA LIDAR
Oregonian thru
Southern California
Rocky
Intertidal
1433
50.6
Modeling
(LIDAR)
2
Site 1: Nooksack Delta,
Lummi Bay, & Bellingham
Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
280
13
SLAMM
2
Site 2: Padilla Bay, Skagit
Bay, & Port Susan Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
280
4
SLAMM
2
Site 6: Dyes Inlet, Sinclair
Inlet, & Bainbridge Island
Puget Trough/
Georgia Basin
Rocky
Intertidal
280
4
SLAMM
2
Site 7: Elliott Bay to the
Duwamish Estuary
Puget Trough/
Georgia Basin
Rocky
Intertidal
280
28
SLAMM
2
Site 11: Willapa Bay,
Columbia River Estuary, &
Tillamook Bay
Oregonian
Rocky
Intertidal
280
22
SLAMM
2
Combined
280 mm SLR
Puget and
Oregonian
Rocky
Intertidal
280
13
SLAMM
2
Site 1: Nooksack Delta,
Lummi Bay, & Bellingham
Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
690
41
SLAMM
2
Site 2: Padilla Bay, Skagit
Bay, & Port Susan Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
690
12
SLAMM
2
Site 6: Dyes Inlet, Sinclair
Inlet, & Bainbridge Island
Puget Trough/
Georgia Basin
Rocky
Intertidal
690
6
SLAMM
2
Site 7: Elliott Bay to the
Duwamish Estuary
Puget Trough/
Georgia Basin
Rocky
Intertidal
690
37
SLAMM
2
Site 11: Willapa Bay,
Columbia River Estuary, &
Tillamook Bay
Oregonian
Rocky
Intertidal
690
62
SLAMM
2
Combined
690 mm SLR
Puget and
Oregonian
Rocky
Intertidal
690
34
SLAMM
168
-------
Source
Location
Ecoregion
Habitat
SLR in
2100
(mm)
Percent
Habitat
Loss
Type of
Study
2
Site 1: Nooksack Delta,
Lummi Bay, & Bellingham
Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
1500
81
SLAMM
2
Site 2: Padilla Bay, Skagit
Bay, & Port Susan Bay
Puget Trough/
Georgia Basin
Rocky
Intertidal
1500
27
SLAMM
2
Site 6: Dyes Inlet, Sinclair
Inlet, & Bainbridge Island
Puget Trough/
Georgia Basin
Rocky
Intertidal
1500
53
SLAMM
2
Site 7: Elliott Bay to the
Duwamish Estuary
Puget Trough/
Georgia Basin
Rocky
Intertidal
1500
44
SLAMM
2
Site 11: Willapa Bay,
Columbia River Estuary, &
Tillamook Bay
Oregonian
Rocky
Intertidal
1500
93
SLAMM
2
Combined
1500 mm SLR
Puget and
Oregonian
Rocky
Intertidal
1500
70
SLAMM
3
Rabbit Rock, OR
Oregonian
Rocky
Intertidal
1000
10.9
Modeling
(TLS data)
3
Rabbit Rock, OR
Oregonian
Rocky
Intertidal
2000
57.3
Modeling
(TLS data)
4
Scotland
North Sea
Rocky
Intertidal
300
10-27
Modeling
(EDINA
DIGIMAP)
4
Scotland
North Sea
Rocky
Intertidal
1900
26-50
Modeling
(EDINA
DIGIMAP)
5
Australia
Tweed-Moreton
Upper
Boulder
Field
300
2 sites >25
2 sites <25
1 site = 0
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Boulder
Field
300
4 sites>50
1 site >25
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Platform
300
2 sites >25
3 sites <25
Modeling
(LIDAR &
digital
imagery)
169
-------
Source
Location
Ecoregion
Habitat
SLR in
2100
(mm)
Percent
Habitat
Loss
Type of
Study
5
Australia
Tweed-Moreton
Lower
Platform
300
1 site >25
2 sites <25
2 sites
increased
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Shallow
Pool
300
1 site >75
3 sites = 50
1 site >25
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Shallow
Pool
300
2 sites > 50
1 site <25
2 sites
increased
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Deep Pool
300
2 sites >50
2 sites <25
1 sites = 0
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Boulder
Field
500
1 site >50
3 sites >25
1 site = 0
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Boulder
Field
500
1 site = 100
3 sites >75
1 site >50
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Platform
500
2 sites >50
2 sites >25
1 site <25
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Platform
500
1 site = 50
1 site >25
1 site <25
2 sites
increased
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Shallow
Pool
500
3 sites >75
2 sites >50
Modeling
(LIDAR &
digital
imagery)
170
-------
Source
Location
Ecoregion
Habitat
SLR in
2100
(mm)
Percent
Habitat
Loss
Type of
Study
5
Australia
Tweed-Moreton
Lower
Shallow
Pool
500
1 site >75
2 sites >50
2 sites
increased
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Deep Pool
500
3 sites = 100
1 site >75
1 site = 0
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Boulder
Field
1000
1 site = 100
1 site >75
2 sites >50
1 site >25
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Boulder
Field
1000
4 sites >75
1 site >50
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Platform
1000
4 sites >75
1 site >25
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Platform
1000
1 site = 75
1 site >50
2 sites <25
1 site
increased
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Upper
Shallow
Pool
1000
2 sites = 100
2 sites > 75
1 site >50
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Lower
Shallow
Pool
1000
4 sites >75
1 sites >50
Modeling
(LIDAR &
digital
imagery)
5
Australia
Tweed-Moreton
Deep Pool
1000
4 sites = 100
1 site = 0
Modeling
(LIDAR &
digital
imagery)
171
-------
7.8.3 Open Coastal Beaches, Backshore Beach Zones and Algal Beach Wrack
Open coast beaches and high intertidal backshore zones surrounding estuaries are often
characterized by algal wrack deposited along the driftline. Since these habitats are found at
approximately the same elevation, we combine them for the purpose of estimating SLR risk.
These habitats are vulnerable due to a number of natural factors and anthropogenic disturbances.
Coastal headlands are common along the U.S. West Coast and create littoral cells which
contribute to spatial isolation of invertebrate species living in these habitats, making these
species particularly vulnerable to habitat loss and fragmentation (Hubbard et al. 2014). This
vulnerability is exacerbated by coastal development and human activities (armoring, regular
beach grooming and sand nourishment). Southern California has disproportionately degraded
beach zones, which has had a substantial impact on the invertebrate community structure,
including local extirpations and regional declines of endemic isopods (Hubbard et al. 2014).
Coastal beaches also face threats from SLR and shoreline erosion. As the ocean rises, the narrow
bluff-backed beaches where many species of upper beach invertebrate populations persist will
have little potential to expand into adjacent habitats. The natural supply of fluvial sediment has
been greatly reduced due to upriver dams retaining sediment that historically resupplied eroding
beaches and provided some vertical resistance to SLR. In their Coastal Vulnerability Index
(CVI), Thieler and Hammar-Klose (2000) classify outer coast sand beaches under the most
vulnerable category citing high erosion, low coastal slope and a high rate of SLR as the
contributing factors. Hubbard et al. (2014) predicts that only a small fraction (<10%) of the 450
km of Southern California coast will have the potential to provide suitable upper beach habitat
under a scenario of 1400 mm of sea level rise by 2100 based on the predictions by Revell et al.
(2011) and NOAA (2012). Similarly, Glick et al. (2007) reports an average loss of 98% of
Pacific Northwest coastal beaches by 2100 under a 1500 mm SLR scenario.
Both Hubbard et al. (2014) and Glick et al. (2007) predict >90% habitat loss with a SLR of
1400-1500 mm. Because our high habitat threshold is based on a >50% habitat loss, we reduced
the SLR values to >800 mm under a constrained scenario and >1000 mm under an unconstrained
scenario (Table 7-3). For the low habitat threshold, Glick et al. (2007) predicted a 6% habitat
loss of coastal beaches for the Pacific Northwest with a SLR of 690 mm by 2100. From these
results, and considering the natural vulnerability of these habitats, we set the constrained and
unconstrained SLR minor habitat thresholds at 550 and 650 mm, respectively. We then
interpolated between the minor and high habitat thresholds to generate the low and moderate
habitat thresholds.
7.8.4 Emergent Marsh
For the purposes of establishing SLR risk, we limited our analysis to the "low" marsh habitat
where species such as land crabs, high intertidal amphipods, etc. are commonly found. Other
marsh vegetation zones (mid, high or transitional marsh) are not currently considered. Thorne et
172
-------
al. (2015) defined low marsh as, "the range of elevations from the lowest extent of vegetation at
a site to the elevation reached by at least one daily high tide on average". Low marsh in the NEP
is characterized by salt tolerant plants including Sarcocorniaperennis (previously referred to as
Salicornia virginica in the NEP), Distichlis spicata, Jaumea carnosa, Agrostis stolonifera, Car ex
lyngbyei, and Triglochin maritima (e.g., Janousek and Folger, 2012).
In the six papers reviewed (Table 7-6), five of the studies have a similar approach in that they
considered land seaward of dikes (constrained) or land seaward of dikes as well as land behind
dikes (unconstrained) and let elevation determine the extent of potential intertidal habitat area
under different SLR scenarios. One modeling effort, Thorne et al. (2015), created defined marsh
areas (ranging from 5 to 97 hectares) and allowed the model to predict habitat expansion or
constriction according to site-specific DEMs within each predetermined area. With such a
bounded model, a habitat type can expand only if another habitat type is reduced. Thorne et al.
(2015) reported their model projections in percent habitat type (total of habitat types =100 %).
To compare results across studies, we transformed the percent habitat type into percent habitat
change using Equation 2 (dates specific to scenario):
Eq. 2: Percent change in marsh area = ((% of low marsh habitat in 2110 - % low marsh habitat in
2010)/% low marsh habitat in 2010) X 100
Expansion and contractions of marsh vegetation zones are highly variable depending on specific
marsh geomorphology, gradient, degree of human development, freshwater flow and other
physical parameters. Even with these local effects, most models predict that low marshes will
experience a net gain in area by 2100 with low to moderate levels of SLR. In comparison, high
marsh habitat will likely experience net losses especially under high rates of SLR. Adding to the
complexity, changes in marsh habitat may not vary consistently with the rate of SLR. For low
marsh habitat (results averaged across nine estuaries), Thorne et al. (2015) projects a decrease in
area under the National Research Council (NCR) low SLR rate (1.2 mm/yr), a very large
increase under the mid (6.3 mm/yr) rate, and a more substantial loss of low marsh under the high
SLR rate (14.2 mm/yr). Under the NRC high SLR rate, sediment accretion rates will not be able
to keep pace resulting in a rapid decline in the low marsh habitat. Eventually, most lower
elevation marsh habitat will convert to intertidal mudflats as low marshes are no longer able to
sustain themselves through natural feedbacks (e.g., sediment accretion, vegetation growth and
organic matter accumulation).
Other analyses, such as Warren Pinnacle Consulting, Inc. (2011) predicted large expansions of
low marsh at low, mid and high rates of sea level rise for both their constrained and
unconstrained analyses. The unconstrained scenario, in particular, resulted in very large increases
in low marsh habitat and consequently large losses in adjacent high marsh and upland habitats.
173
-------
Table 7-6. Summary of low marsh percent habitat change under different SLR rates.
Compiled from six studies and 21 estuaries. Thorne et al. (2015) modeled changes in habitats within a
bounded area. Stralberg et al. (2011) evaluated removal of diked areas as did Warren Pinnacle
Consulting, Inc. (2011) and Ducks Unlimited (no date) evaluated the effects of SLR maintaining dikes
(constrained) and with the removal of dikes (unconstrained). Sources: 1 - Warren Pinnacle Consulting,
Inc. 2011; 2 - Ducks Unlimited, no date; 3 - Galbraith et al., 2002; 4 - Glick et al., 2007; 5 - Stralberg et al.,
2011; 6 - Thorne et al. 2015.
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Alsea River,
OR
390
2100
29.0
expansion
SLAMM 6
Constrained
1
Alsea River,
OR
690
2100
35.0
expansion
SLAMM 6
Constrained
1
Alsea River,
OR
1000
2100
56.0
expansion
SLAMM 6
Constrained
1
Alsea River,
OR
1500
2100
150.0
expansion
SLAMM 6
Constrained
1
Alsea River,
OR
2000
2100
145.0
expansion
SLAMM 6
Constrained
1
Alsea River,
OR
390
2100
61.0
expansion
SLAMM 6
Unconstrained
1
Alsea River,
OR
690
2100
80.0
expansion
SLAMM 6
Unconstrained
1
Alsea River,
OR
1000
2100
114.0
expansion
SLAMM 6
Unconstrained
1
Alsea River,
OR
1500
2100
201.0
expansion
SLAMM 6
Unconstrained
1
Alsea River,
OR
2000
2100
180.0
expansion
SLAMM 6
Unconstrained
1
Chetco River,
OR
390
2100
363.0
expansion
SLAMM 6
Constrained
1
Chetco River,
OR
690
2100
426.0
expansion
SLAMM 6
Constrained
174
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Chetco River,
OR
1000
2100
684.0
expansion
SLAMM 6
Constrained
1
Chetco River,
OR
1500
2100
1592.0
expansion
SLAMM 6
Constrained
1
Chetco River,
OR
2000
2100
3287.0
expansion
SLAMM 6
Constrained
1
Chetco River,
OR
390
2100
363.0
expansion
SLAMM 6
Unconstrained
1
Chetco River,
OR
690
2100
427.0
expansion
SLAMM 6
Unconstrained
1
Chetco River,
OR
1000
2100
685.0
expansion
SLAMM 6
Unconstrained
1
Chetco River,
OR
1500
2100
1595.0
expansion
SLAMM 6
Unconstrained
1
Chetco River,
OR
2000
2100
3294.0
expansion
SLAMM 6
Unconstrained
1
Coos Bay, OR
390
2100
64.0
expansion
SLAMM 6
Constrained
1
Coos Bay, OR
690
2100
92.0
expansion
SLAMM 6
Constrained
1
Coos Bay, OR
1000
2100
112.0
expansion
SLAMM 6
Constrained
1
Coos Bay, OR
1500
2100
175.0
expansion
SLAMM 6
Constrained
1
Coos Bay, OR
2000
2100
193.0
expansion
SLAMM 6
Constrained
1
Coos Bay, OR
390
2100
270.0
expansion
SLAMM 6
Unconstrained
1
Coos Bay, OR
690
2100
242.0
expansion
SLAMM 6
Unconstrained
1
Coos Bay, OR
1000
2100
205.0
expansion
SLAMM 6
Unconstrained
175
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Coos Bay, OR
1500
2100
219.0
expansion
SLAMM 6
Unconstrained
1
Coos Bay, OR
2000
2100
231.0
expansion
SLAMM 6
Unconstrained
1
Nehalem Bay,
OR
390
2100
20.0
expansion
SLAMM 6
Constrained
1
Nehalem Bay,
OR
690
2100
28.0
expansion
SLAMM 6
Constrained
1
Nehalem Bay,
OR
1000
2100
50.0
expansion
SLAMM 6
Constrained
1
Nehalem Bay,
OR
1500
2100
140.0
expansion
SLAMM 6
Constrained
1
Nehalem Bay,
OR
2000
2100
187.0
expansion
SLAMM 6
Constrained
1
Nehalem Bay,
OR
390
2100
123.0
expansion
SLAMM 6
Unconstrained
1
Nehalem Bay,
OR
690
2100
181.0
expansion
SLAMM 6
Unconstrained
1
Nehalem Bay,
OR
1000
2100
249.0
expansion
SLAMM 6
Unconstrained
1
Nehalem Bay,
OR
1500
2100
354.0
expansion
SLAMM 6
Unconstrained
1
Nehalem Bay,
OR
2000
2100
370.0
expansion
SLAMM 6
Unconstrained
1
Nestucca Bay,
OR
390
2100
25.0
expansion
SLAMM 6
Constrained
1
Nestucca Bay,
OR
690
2100
32.0
expansion
SLAMM 6
Constrained
1
Nestucca Bay,
OR
1000
2100
54.0
expansion
SLAMM 6
Constrained
176
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Nestucca Bay,
OR
1500
2100
151.0
expansion
SLAMM 6
Constrained
1
Nestucca Bay,
OR
2000
2100
124.0
expansion
SLAMM 6
Constrained
1
Nestucca Bay,
OR
390
2100
132.0
expansion
SLAMM 6
Unconstrained
1
Nestucca Bay,
OR
690
2100
163.0
expansion
SLAMM 6
Unconstrained
1
Nestucca Bay,
OR
1000
2100
177.0
expansion
SLAMM 6
Unconstrained
1
Nestucca Bay,
OR
1500
2100
240.0
expansion
SLAMM 6
Unconstrained
1
Nestucca Bay,
OR
2000
2100
180.0
expansion
SLAMM 6
Unconstrained
1
Rogue River,
OR
390
2100
1721.0
expansion
SLAMM 6
Constrained
1
Rogue River,
OR
690
2100
2082.0
expansion
SLAMM 6
Constrained
1
Rogue River,
OR
1000
2100
2883.0
expansion
SLAMM 6
Constrained
1
Rogue River,
OR
1500
2100
6225.0
expansion
SLAMM 6
Constrained
1
Rogue River,
OR
2000
2100
11795.0
expansion
SLAMM 6
Constrained
1
Rogue River,
OR
390
2100
1723.0
expansion
SLAMM 6
Unconstrained
1
Rogue River,
OR
690
2100
2085.0
expansion
SLAMM 6
Unconstrained
1
Rogue River,
OR
1000
2100
2890.0
expansion
SLAMM 6
Unconstrained
177
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Rogue River,
OR
1500
2100
6244.0
expansion
SLAMM 6
Unconstrained
1
Rogue River,
OR
2000
2100
11835.0
expansion
SLAMM 6
Unconstrained
1
Siuslaw R.,
OR
390
2100
54.0
expansion
SLAMM 6
Constrained
1
Siuslaw R.,
OR
690
2100
73.0
expansion
SLAMM 6
Constrained
1
Siuslaw R.,
OR
1000
2100
113.0
expansion
SLAMM 6
Constrained
1
Siuslaw R.,
OR
1500
2100
128.0
expansion
SLAMM 6
Constrained
1
Siuslaw R.,
OR
2000
2100
118.0
expansion
SLAMM 6
Constrained
1
Siuslaw R.,
OR
390
2100
143.0
expansion
SLAMM 6
Unconstrained
1
Siuslaw R.,
OR
690
2100
127.0
expansion
SLAMM 6
Unconstrained
1
Siuslaw R.,
OR
1000
2100
137.0
expansion
SLAMM 6
Unconstrained
1
Siuslaw R.,
OR
1500
2100
149.0
expansion
SLAMM 6
Unconstrained
1
Siuslaw R.,
OR
2000
2100
134.0
expansion
SLAMM 6
Unconstrained
1
Umpqua R.,
OR
390
2100
62.0
expansion
SLAMM 6
Constrained
1
Umpqua R.,
OR
690
2100
92.0
expansion
SLAMM 6
Constrained
1
Umpqua R.,
OR
1000
2100
145.0
expansion
SLAMM 6
Constrained
178
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
1
Umpqua R.,
OR
1500
2100
232.0
expansion
SLAMM 6
Constrained
1
Umpqua R.,
OR
2000
2100
217.0
expansion
SLAMM 6
Constrained
1
Umpqua R.,
OR
390
2100
304.0
expansion
SLAMM 6
Unconstrained
1
Umpqua R.,
OR
690
2100
355.0
expansion
SLAMM 6
Unconstrained
1
Umpqua R.,
OR
1000
2100
365.0
expansion
SLAMM 6
Unconstrained
1
Umpqua R.,
OR
1500
2100
349.0
expansion
SLAMM 6
Unconstrained
1
Umpqua R.,
OR
2000
2100
286.0
expansion
SLAMM 6
Unconstrained
2
Grays Harbor,
OR
690
2100
35.0
expansion
SLAMM 6
Constrained
2
Lower
Columbia
690
2100
-18.9
loss
SLAMM 6
Constrained
2
Lower
Columbia
690
2100
134.1
expansion
SLAMM 6
Unconstrained
2
North Puget
Sound
690
2100
2.2
expansion
SLAMM 6
Constrained
2
Willapa Bay,
WA
690
2100
-6.4
loss
SLAMM 6
Constrained
3
Humboldt Bay,
CA
200
2100
72.6
expansion
SLAMM 4
Constrained
3
Humboldt Bay,
CA
340
2100
175.6
expansion
SLAMM 4
Constrained
3
Humboldt Bay,
CA
770
2100
1886.0
expansion
SLAMM 4
Constrained
179
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
3
Northern San
Francisco Bay
200
2100
0.0
no change
SLAMM 4
Constrained
3
Northern San
Francisco Bay
340
2100
0.0
no change
SLAMM 4
Constrained
3
Northern San
Francisco Bay
770
2100
-18.1
loss
SLAMM 4
Constrained
3
Southern San
Francisco Bay
200
2100
-50.7
loss
SLAMM 4
Constrained
3
Southern San
Francisco Bay
340
2100
-63.2
loss
SLAMM 4
Constrained
3
Southern San
Francisco Bay
770
2100
-82.9
loss
SLAMM 4
Constrained
3
Willapa Bay,
WA
200
2100
12.8
expansion
SLAMM 4
Constrained
3
Willapa Bay,
WA
340
2100
10.5
expansion
SLAMM 4
Constrained
3
Willapa Bay,
WA
770
2100
12.8
expansion
SLAMM 4
Constrained
4
Annas Bay
and
Skokomish
estuary
690
2100
49.0
expansion
SLAMM 5
Constrained
4
Annas Bay
and
Skokomish
estuary
1500
2100
48.9
expansion
SLAMM 5
Constrained
4
Dyes &
Sinclair Inlet
and
Bainbridge Is.
690
2100
4388.0
expansion
SLAMM 5
Constrained
180
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
4
Dyes &
Sinclair Inlet
and
Bainbridge Is.
1500
2100
5250.0
expansion
SLAMM 5
Constrained
4
Nooksack,
Lummi &
Bellingham
Bays
690
2100
469.0
expansion
SLAMM 5
Constrained
4
Nooksack,
Lummi &
Bellingham
Bays
1500
2100
1786.8
expansion
SLAMM 5
Constrained
4
Nooksack,
Lummi &
Bellingham
Bays
690
2100
3927.0
expansion
SLAMM 5
Unconstrained
4
Olympia, Budd
Inlet &
Nisqually
Delta
690
2100
422.0
expansion
SLAMM 5
Constrained
4
Olympia, Budd
Inlet &
Nisqually
Delta
1500
2100
501.9
expansion
SLAMM 5
Constrained
4
Olympia, Budd
Inlet &
Nisqually
Delta
690
2100
1059.0
expansion
SLAMM 5
Unconstrained
4
Padilla, Skagit
& Port Susan
Bays
690
2100
96.0
expansion
SLAMM 5
Constrained
4
Padilla, Skagit
& Port Susan
Bays
1500
2100
41.2
expansion
SLAMM 5
Constrained
4
Padilla, Skagit
& Port Susan
Bays
690
2100
1115.0
expansion
SLAMM 5
Unconstrained
181
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
4
Port Angeles,
Dungeness
Spit & Sequim
Bay
690
2100
65.0
expansion
SLAMM 5
Constrained
4
Port Angeles,
Dungeness
Spit & Sequim
Bay
1500
2100
190.5
expansion
SLAMM 5
Constrained
4
Snohomish
estuary &
Everett
690
2100
1522.0
expansion
SLAMM 5
Constrained
4
Snohomish
estuary &
Everett
1500
2100
1431.0
expansion
SLAMM 5
Constrained
4
Snohomish
estuary &
Everett
690
2100
7548.0
expansion
SLAMM 5
Unconstrained
4
Whidbey Is.,
Port
Townsend,
Admiralty Inlet
690
2100
814.0
expansion
SLAMM 5
Constrained
4
Whidbey Is.,
Port
Townsend,
Admiralty Inlet
1500
2100
496.5
expansion
SLAMM 5
Constrained
4
Willapa,
Columbia &
Tillamook Bay
690
2100
6.0
expansion
SLAMM 5
Constrained
4
Willapa,
Columbia &
Tillamook Bay
1500
2100
29.9
expansion
SLAMM 5
Constrained
5
San Francisco
Bay (Tidal +
diked lands)
520
2110
40.0
expansion
Marsh98
Constrained
(with dike
removal)
182
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper
habitat
constrained
in model?
5
San Francisco
Bay (Tidal +
diked lands)
1650
2110
108.0
expansion
Marsh98
Constrained
(with dike
removal)
5
San Francisco
Bay (Tidal
only)
520
2110
-24.0
loss
Marsh98
Constrained
5
San Francisco
Bay (Tidal
only)
1650
2110
53.3
expansion
Marsh98
Constrained
6
Bandon
Marsh, OR
120
2110
2.0
expansion
WARMER
Constrained
6
Bandon
Marsh, OR
630
2110
59.1
expansion
WARMER
Constrained
6
Bandon
Marsh, OR
Bando
n
Marsh,
OR
2110
-100.0
loss
WARMER
Constrained
6
Coos Bay, OR
120
2110
-100.0
loss
WARMER
Constrained
6
Coos Bay, OR
630
2110
-100.0
loss
WARMER
Constrained
6
Coos Bay, OR
1420
2110
-100.0
loss
WARMER
Constrained
6
Grays Harbor,
WA
120
2110
-100.0
loss
WARMER
Constrained
As is apparent from Table 7-6, the predicted responses of lower marshes to SLR are complex. To
extract SLR thresholds from these "messy" data, we made two simplifying assumptions. Firstly,
at this stage, we ignore non-linear responses (i.e., greater habitat loss at a lower SLR). This
assumption may underestimate habitat loss at the minor and low habitat thresholds. Secondly, we
derive the habitat thresholds only from the sites with habitat losses. Evaluating sites only with
losses is a "pessimistic" scenario resulting from reduced potential for landward migration as a
result of increased construction of barriers, such as rip rap, to protect against sea level rise and/or
land subsidence. In these cases, sites that were predicted to increase may not be able to migrate
183
-------
landward as they would have done historically, resulting in a habitat decline with SLR. Such a
scenario is certainly not out of the realm of possibilities as social and financial pressures mount
to protect infrastructure and shoreline development, especially with increased population
densities in coastal areas (King et al., 2011; California Natural Resource Agency, 2014).
At the highest SLR values, it is also possible that the sediment load would be insufficient for
sediment accumulation to keep pace with SLR (Stralberg et al. 2011). Much of the West Coast is
predicted to experience longer periods of summer drought further reducing the supply of
sediment from upriver sources. Further uncertainty stems from the possibility of increased winter
storms generating wave erosion at the lower end of marshes and additional erosion from seawalls
at the upper end of marshes (Stralberg et al. 2011; CA Natural Resource Agency, 2014). Thus,
use of the pessimistic scenario may better capture future impacts. However, in the absence of
increased barriers and/or insufficient sediment load, deriving the thresholds from the sites with
losses ignores the increases in other marshes, overestimating the extent of habitat loss at an
ecoregion scale.
In evaluating the constrained lower marshes (Figure 7-2), some sites showed some minor loss
(<10%) of marsh at 120 mm SLR but then a quarter of the sites experienced high (>50%) habitat
loss at 200 mm SLR. This indicates that the minor habitat threshold occurs between these two
values. Taking the average, we set the cut point between minor and low thresholds at 160 mm.
Based on the spike of almost 78% of the sites with high habitat loss at a SLR value of 1420 mm,
we use this value as the cut point between moderate and high habitat thresholds. The cut point
between the low and moderate thresholds is then generated by taking the average of these values,
resulting in a value of 790 mm.
Model results for the unconstrained lower marshes all predict that low marsh habitat will expand
continuously with SLR based on the assumption that there are adequate adjacent lands of similar
elevation to expand into. Because the modelling results do not identify the cutpoints, we
tentatively suggest setting the minor threshold at 2500 mm based on the potential for limited
sediment accretion. For the other thresholds, we tentatively added cumulative increments of 250
mm to estimate the higher cutpoints. With all the models predicting increases up to the maximum
tested of 2000 mm, it is reasonable to assume that associated species would be at 'minor' or
'low' risk at most likely levels of sea level rise in unconstrained lower marshes (see Table 7-1).
Therefore, the exact cut points are not as critical in this habitat compared to a habitat that is
expected to decrease.
184
-------
% of Sites in Each Habitat Loss class - Low Marsh - Constrained
data
uJJ LJ.
120 200 340 390 520 630 690 770 1000 1420 1500 1650 2000
Net SLR
¦ Minor, %<=-10 ¦ Low, % 11 -29 loss ¦ Moderate, % 30-49 loss ¦ High, %>=50 loss
Figure 7-2. Low marsh habitat - Constrained.
This values represent the percentage of sites in the six papers analyzed that fell within each of
the habitat threshold categories (Minor, Low, Moderate, High). Only sites with predicted habitat
losses are included in the analysis.
7.8.5 Submerged Aquatic Vegetation
Submerged aquatic vegetation (SAV) is a term used to describe a suite of rooted, vascular plants
that grow completely underwater except for periods of exposure at low tides. Species of SAV are
often referred to as seagrasses. Eight species of seagrasses occur on the Pacific Coast (Wyllie-
Echeverria and Ackerman, 2003), including the nonindigenous Zostera japonica (Kaldy, 2006),
however we focus on the native Z. marina. Z. marina is the most abundant seagrass in estuaries
in the NEP (Lee and Brown, 2009) and ranges from the Bering Sea into the Gulf of California
(Wyllie-Echeverria and Ackerman, 2003; Shaughnessy et al., 2012). Though SLR has been
considered a threat to seagrasses for almost two decades (Short and Neckles, 1999), there are
relative few studies compared to marshes. Here we synthesize two published models on SLR and
the research conducted by EPA. The reason for fewer studies on seagrasses appears to be a
result, at least in part, that earlier versions of the often used SLAMM model did not predict
effects on SAV. We addressed this limitation by creating a module to SLAMM Ver. 6 that
allows users to predict SLR effects on Zostera (Lee et al., 2014). Unfortunately, we have not had
the resources to use this tool in a regional study of sea level rise effects.
Shaughnessy et al. (2012) modeled SLR effects on seagrass area for a period of 100 years for
seven estuaries located from Alaska to Mexico. Modeling combinations of low and high bottom
100
90
SO
70
60
50
40
30
20
10
0
185
-------
change (estuarine specific sediment and tectonic rates) with three levels of SLR produced six
scenarios at each location. The low SLR rate scenario (2.8 mm/yr) was from the current
estimated SLR rate for the Pacific Ocean basin based on satellite altimetry, the moderate SLR
rate scenario (6.3 mm/yr) was derived from the mean rate estimated for the period 2090-2099 for
IPCC AR4 scenario AIFI (Meehl et al., 2007), and the high SLR rate scenario (12.7 mm/yr) was
based on a study linking global sea level rise to projections of global mean surface temperature
(Rahmstorf, 2007).
Shaughnessy et al. (2012) predicted that seagrass habitat would increase or only show minor
change with SLR at most locations (Table 7-7). Five estuaries that were not topographically
constrained and had sufficient available upslope area for migration, experienced a greater
increase in seagrass area under moderate and high SLR than with low SLR. However, seagrass
declined at the topographically constrained Morro Bay under moderate and high SLR. In north
Humboldt Bay, SAV showed declines at "high bottom change" with both low and moderate SRL
and with "low bottom change" with low SLR. When averaged across all locations and scenarios,
these models predicted a 15.2% increase in SAV with SLR. However, the authors pointed out
that barriers to landward migration of SAV had not been encountered in the 100 year simulations
in several of the estuaries. Once the water level rises to the point of encountering these barriers,
Z. marina would likely decline in the subtidal portion of the population due to light extinction.
Thus, habitat loss may increase in the longer term.
In a detailed modeling study by Kairis and Rybczyk (2010) of Padilla Bay, WA, seagrass area
increased (7.9 to 43.7%) with increasing SLR until leveling off (37.4%) at the highest SLR rate.
The authors modeled eight different SLR rates, ranging from 1.714 to 18.182 mm/yr over a 100-
year period. In this shallow bay, Z. marina was predicted to migrate from the center of the bay
shoreward, colonizing the extensive mudflats. We note that while this model appears to be
constrained, by colonizing the existing mudflats the expansion of Z. marina was not limited by
dikes or other barriers.
The EPA constructed a SLR model for SAV based on geomorphological features that was
applied to the Yaquina, Tillamook, and Alsea estuaries in Oregon (Clinton et al., 2012). Based
on the topobathy, these models allowed landward migration of intertidal habitats. However, since
the model did not incorporate sediment accumulation, we consider the results as more closely
approximating a constrained condition. The model predicted an increase in Z. marina in the
Alsea with SLR, but it should be noted that there currently is very little Zostera in the Alsea. In
comparison, seagrass is an important habitat in both the Yaquina and Tillamook estuaries (Lee
and Brown, 2009), and Z. marina is predicted to decline by 31% and 68% with aim SLR in
these estuaries, respectively.
Using the EPA results as a guide for constrained seagrass, a >50% loss was observed in
Tillamook at SLR of 750 and 1000 mm but not in Yaquina. This range suggests, that on the
186
-------
average, a >50% loss would occur at about a meter increase in depth, and we tentatively set a
high constrained threshold at >900 mm. The moderate constrained threshold thus becomes <900
mm. At 500 mm, Yaquina and Tillamook showed a 21% and 40% loss, though Alsea showed a
35% increase. The average of these three sites is an 8.6% loss. From these rates, we set the minor
constrained threshold at 540 mm. The low constrained threshold was then set as the average of
the minor and moderate thresholds, or 720 mm.
Generating the unconstrained values are even more challenging since the current models predict
that they will increase in most cases. Thus, there is likely to be a mosaic of effects, with some
areas within an ecoregion increasing and other declining due to factors such as sediment
accretion not keeping up with sea level rise. Additionally, there may be some bias in these
studies toward areas with few barriers to inland migration, such a Padilla Bay versus other
portions of Puget Sound which show extensive armoring (Washington State Department of
Natural Resources, no date; Myers, 2010). Given these uncertainties, as a tentative first step, we
double the constrained threshold values, resulting in unconstrained thresholds for SAV of 1080
mm for minor, 1440 mm for low, 1800 mm for moderate, and >1800 mm for high.
187
-------
Table 7-7. Submerged Aquatic Vegetation (Zostera marina), summary of seagrass percent habitat
change under different SLR values.
Compiled from three studies and eleven estuaries. The percent area of SAV in Clinton et al. (2012)
were converted to percent SAV loss. Sources: 1 = Kairis and Rybczyk, 2010; 2 = Clinton et al., 2012; 3
= Shaughnessy et al., 2012.
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model Type
Upper habitat
constrained
in model?
1
Padilla Bay,
WA
170
2102
8
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
330
2102
22
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
560
2102
34
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
560
2102
34
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
640
2102
37
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
860
2102
41
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
1270
2102
44
expansion
Spatial Relative Elev.
Model
Constrained
1
Padilla Bay,
WA
1820
2102
-37
loss
Spatial Relative Elev.
Model
Constrained
2
Alsea, OR
250
2100
9.6
expansion
Geomorphological
topobathy model
Constrained
2
Alsea, OR
500
2100
35.4
expansion
Geomorphological
topobathy model
Constrained
2
Alsea, OR
750
2100
41.7
expansion
Geomorphological
topobathy model
Constrained
2
Alsea, OR
1000
2100
39.6
expansion
Geomorphological
topobathy model
Constrained
2
Tillamook,
OR
250
2100
-18
loss
Geomorphological
topobathy model
Constrained
2
Tillamook,
OR
500
2100
-40
loss
Geomorphological
topobathy model
Constrained
2
Tillamook,
OR
750
2100
-59
loss
Geomorphological
topobathy model
Constrained
2
Tillamook,
OR
1000
2100
-68
loss
Geomorphological
topobathy model
Constrained
2
Yaquina, OR
250
2100
-14
loss
Geomorphological
topobathy model
Constrained
2
Yaquina, OR
500
2100
-21
loss
Geomorphological
topobathy model
Constrained
2
Yaquina, OR
750
2100
-28
loss
Geomorphological
topobathy model
Constrained
188
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model Type
Upper habitat
constrained
in model?
2
Yaquina, OR
1000
2100
-31
loss
Geomorphological
topobathy model
Constrained
3
Bahia San
Quintin,
Mexico
280
2112
1
no change
Low accretion,
low SLR
Unconstrained
3
Bahia San
Quintin,
Mexico
630
2112
5
no change
Low accretion,
moderate SLR
Unconstrained
3
Bahia San
Quintin,
Mexico
1270
2112
25
expansion
Low accretion, high
SLR
Unconstrained
3
Bahia San
Quintin,
Mexico
280
2112
11
expansion
High accretion,
low SLR
Unconstrained
3
Bahia San
Quintin,
Mexico
630
2112
0
no change
High accretion,
moderate SLR
Unconstrained
3
Bahia San
Quintin,
Mexico
1270
2112
15
expansion
High accretion, high
SLR
Unconstrained
3
Izembek
Lagoon, AK
280
2112
8
no change
Low accretion,
low SLR
Unconstrained
3
Izembek
Lagoon, AK
630
2112
19
expansion
Low accretion,
moderate SLR
Unconstrained
3
Izembek
Lagoon, AK
1270
2112
16
expansion
Low accretion, high
SLR
Unconstrained
3
Izembek
Lagoon, AK
280
2112
1
no change
High accretion,
low SLR
Unconstrained
3
Izembek
Lagoon, AK
630
2112
14
expansion
High accretion,
moderate SLR
Unconstrained
3
Izembek
Lagoon, AK
1270
2112
21
expansion
High accretion, high
SLR
Unconstrained
3
Morro Bay,
CA
280
2112
-1
loss
Low accretion,
low SLR
Unconstrained
3
Morro Bay,
CA
630
2112
22
expansion
Low accretion,
moderate SLR
Unconstrained
3
Morro Bay,
CA
1270
2112
-45
loss
Low accretion, high
SLR
Unconstrained
3
Morro Bay,
CA
280
2112
-64
loss
High accretion,
low SLR
Unconstrained
3
Morro Bay,
CA
630
2112
2
no change
High accretion,
moderate SLR
Unconstrained
3
Morro Bay,
CA
1270
2112
-6
no change
High accretion, high
SLR
Unconstrained
189
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model Type
Upper habitat
constrained
in model?
3
North
Humboldt
Bay, CA
280
2112
-30
loss
Low accretion,
low SLR
Unconstrained
3
North
Humboldt
Bay, CA
630
2112
18
expansion
Low accretion,
moderate SLR
Unconstrained
3
North
Humboldt
Bay, CA
1270
2112
87
expansion
Low accretion, high
SLR
Unconstrained
3
North
Humboldt
Bay, CA
280
2112
-63
loss
High accretion,
low SLR
Unconstrained
3
North
Humboldt
Bay, CA
630
2112
-5
no change
High accretion,
moderate SLR
Unconstrained
3
North
Humboldt
Bay, CA
1270
2112
64
expansion
High accretion, high
SLR
Unconstrained
3
Padilla Bay
complex,
WA
280
2112
10
no change
Low accretion, low
SLR
Unconstrained
3
Padilla Bay
complex,
WA
630
2112
16
expansion
Low accretion,
moderate SLR
Unconstrained
3
Padilla Bay
complex,
WA
1270
2112
15
expansion
Low accretion, high
SLR
Unconstrained
3
Padilla Bay
complex,
WA
280
2112
4
no change
High accretion, low
SLR
Unconstrained
3
Padilla Bay
complex,
WA
630
2112
11
expansion
High accretion,
moderate SLR
Unconstrained
3
Padilla Bay
complex,
WA
1270
2112
17
expansion
High accretion, high
SLR
Unconstrained
3
South
Humboldt
Bay, CA
280
2112
6
no change
Low accretion,
low SLR
Unconstrained
3
South
Humboldt
Bay, CA
630
2112
27
expansion
Low accretion,
moderate SLR
Unconstrained
3
South
Humboldt
Bay, CA
1270
2112
68
expansion
Low accretion, high
SLR
Unconstrained
3
South
Humboldt
Bay, CA
280
2112
-6
no change
High accretion,
low SLR
Unconstrained
190
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model Type
Upper habitat
constrained
in model?
3
South
Humboldt
Bay, CA
630
2112
14
expansion
High accretion,
moderate SLR
Unconstrained
3
South
Humboldt
Bay, CA
1270
2112
64
expansion
High accretion, high
SLR
Unconstrained
3
Willapa Bay,
WA
280
2112
12
expansion
Low accretion,
low SLR
Unconstrained
3
Willapa Bay,
WA
630
2112
46
expansion
Low accretion,
moderate SLR
Unconstrained
3
Willapa Bay,
WA
1270
2112
106
expansion
Low accretion, high
SLR
Unconstrained
3
Willapa Bay,
WA
280
2112
-6
loss
High accretion,
low SLR
Unconstrained
3
Willapa Bay,
WA
630
2112
27
expansion
High accretion,
moderate SLR
Unconstrained
3
Willapa Bay,
WA
1270
2112
94
expansion
High accretion, high
SLR
Unconstrained
7.8.6 Tide Flats - Unvegetated Sand/Mud & Oyster Beds & Macroalgal Mats
The beaches and tidal flats of the Pacific Northwest are vulnerable to the rising sea-level over the
next century based on the four studies we reviewed (Table 7-8). Under a 693 mm global average
sea-level rise scenario, about 65 percent of estuarine beaches and 44 percent of tidal flats are
predicted to be lost across all eleven study sites by 2100 according to the 2007 analysis by Glick
et al. This degree of loss will likely cause significant changes in the coastal landscape. For
example, Dungeness Spit, WA is predicted to be subject to inundation, erosion, and overwash
due to storm events, leading to major losses of beach and tidal flat habitats (Glick et al. 2007). A
reduction in intertidal habitat corresponds to reduced yields of commercial oyster production as
well as recreationally harvested bivalve species. Such declines in shellfish harvesting may have
significant impacts on small coastal economies in Oregon and Washington (Norman et al., 2007).
Galbraith et al. 2002, using a constrained dike data layer, predicts major intertidal habitat loss at
all four of their West Coast study sites. Willapa Bay, Humboldt Bay, and northern and southern
San Francisco Bay are predicted to lose between 20% and 70% of their current intertidal habitat.
They predict the most severe losses are likely to occur in the areas where the coastline is unable
to move inland because of steep topography or seawalls. In sharp contrast to Galbraith et al.
(2002), Thorne et al. (2015), whose study was also considered constrained, predicts that more
than one-half of their study sites under a high SLR scenario resulted in very large (near 100%)
191
-------
increases in mudflat habitat as a result of low tidal marsh transitioning into mudflat; suggesting
that rates of net accretion cannot keep pace with rising sea levels.
Table 7-8. Summary of tide flat percent habitat change under different SLR rates.
Compiled from four studies and 23 estuaries. Ducks Unlimited (no date) evaluated the effects of dike
removal in the lower Columbia. Warren Pinnacle Consulting, Inc. (2011) evaluated the effects of SLR with
dikes (constrained) and with dike removal (unconstrained). Sources: 1 - Ducks Unlimited, no date; 2 -
Galbraith et al., 2002; 3 - Glick et al., 2007; 4 - Thorne et al., 2015; 5 - Warren Pinnacle Consulting, Inc.
2011. NA's indicate that the original habitat area was zero thus making the Equation 2 undefined.
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
1
Grays Harbor, WA
690
2100
35
expansion
SLAMM 6
Constrained
1
Willapa Bay, WA
690
2100
-6
loss
SLAMM 6
Constrained
1
Lower Columbia
690
2100
-19
loss
SLAMM 6
Constrained
1
North Puget Sound
690
2100
2
expansion
SLAMM 6
Constrained
1
Lower Columbia
690
2100
134
expansion
SLAMM 6
Unconstrained
2
Willapa Bay, WA
historic
2100
-1
loss
SLAMM 4
Constrained
2
Humboldt Bay
historic
2100
0
loss
SLAMM 4
Constrained
2
Northern San
Francisco Bay, CA
historic
2100
-4
loss
SLAMM 4
Constrained
2
Southern San
Francisco Bay, CA
historic
2100
-54
loss
SLAMM 4
Constrained
2
Humboldt Bay
340
2100
-29
loss
SLAMM 4
Constrained
2
Northern San
Francisco Bay, CA
340
2100
-39
loss
SLAMM 4
Constrained
2
Southern San
Francisco Bay, CA
340
2100
-70
loss
SLAMM 4
Constrained
2
Willapa Bay, WA
340
2100
-18
loss
SLAMM 4
Constrained
2
Humboldt Bay, CA
770
2100
-91
loss
SLAMM 4
Constrained
2
Northern San
Francisco Bay, CA
770
2100
-81
loss
SLAMM 4
Constrained
2
Southern San
Francisco Bay, CA
770
2100
-83
loss
SLAMM 4
Constrained
2
Willapa Bay, WA
770
2100
-62
loss
SLAMM 4
Constrained
3
Nooksack, Lummi &
Bellingham Bays
690
2100
22
expansion
SLAMM 5
Constrained
192
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
3
Padilla, Skagit & Po.
Susan Bays
690
2100
613
expansion
SLAMM 5
Constrained
3
Whidbey Is., Po.
Townsend, Admiralty
Inlet
690
2100
1425
expansion
SLAMM 5
Constrained
3
Snohomish estuary &
Everett
690
2100
411
expansion
SLAMM 5
Constrained
3
Port Angeles,
Dungeness Spit &
Sequim Bay
690
2100
-81
expansion
SLAMM 5
Constrained
3
Dyes & Sinclair Inlet
and Bainbridge Is.
690
2100
1916
expansion
SLAMM 5
Constrained
3
Elliot Bay & the
Duwamish estuary
690
2100
319
expansion
SLAMM 5
Constrained
3
Annas Bay and
Skokomish estuary
690
2100
67.3
expansion
SLAMM 5
Constrained
3
Commencement Bay,
Tacoma & Gig
Harbor
690
2100
12.5
expansion
SLAMM 5
Constrained
3
Olympia, Budd Inlet &
Nisqually Delta
690
2100
50.7
expansion
SLAMM 5
Constrained
3
Willapa, Columbia &
Tillamook Bay
690
2100
-63
loss
SLAMM 5
Constrained
3
Nooksack, Lummi &
Bellingham Bays
690
2100
75
expansion
SLAMM 5
Unconstrained
3
Padilla, Skagit & Port
Susan Bays
690
2100
1559
expansion
SLAMM 5
Unconstrained
3
Snohomish estuary &
Everett
690
2100
2422
expansion
SLAMM 5
Unconstrained
3
Olympia, Budd Inlet &
Nisqually Delta
690
2100
NA
expansion
SLAMM 5
Unconstrained
3
Nooksack, Lummi &
Bellingham Bays
1500
2100
95
expansion
SLAMM 5
Constrained
3
Padilla, Skagit & Po.
Susan Bays
1500
2100
869
expansion
SLAMM 5
Constrained
3
Whidbey Is., Port
Townsend &
Admiralty Inlet
1500
2100
1565
expansion
SLAMM 5
Constrained
3
Snohomish estuary &
Everett
1500
2100
706
expansion
SLAMM 5
Constrained
3
Port Angeles,
Dungeness Spit &
Sequim Bay
1500
2100
-82
loss
SLAMM 5
Constrained
3
Dyes & Sinclair Inlet
and Bainbridge Is.
1500
2100
1411
expansion
SLAMM 5
Constrained
193
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
3
Elliot Bay & the
Duwamish estuary
1500
2100
611
expansion
SLAMM 5
Constrained
3
Annas Bay and
Skokomish estuary
1500
2100
58.2
expansion
SLAMM 5
Constrained
3
Commencement Bay,
Tacoma & Gig
Harbor
1500
2100
22.7
expansion
SLAMM 5
Constrained
3
Olympia, Budd Inlet &
Nisqually Delta
1500
2100
64.6
expansion
SLAMM 5
Constrained
3
Willapa, Columbia &
Tillamook Bay
1500
2100
-63
loss
SLAMM 5
Constrained
4
Bandon Marsh, OR
120
2110
0
no change
WARMER
Constrained
4
Coos Bay, OR
120
2110
0
no change
WARMER
Constrained
4
Grays Harbor, WA
120
2110
0
no change
WARMER
Constrained
4
Nisqually National
120
2110
0
no change
WARMER
Constrained
4
Padilla Marsh, WA
120
2110
0
no change
WARMER
Constrained
4
Port Susan Bay, WA
120
2110
0
loss
WARMER
Constrained
4
Siletz Bay, OR
120
2110
0
no change
WARMER
Constrained
4
Skokomish Estuary,
WA
120
2110
0
no change
WARMER
Constrained
4
Willapa Bay, WA
120
2110
0
no change
WARMER
Constrained
4
Bandon Marsh, OR
630
2110
100
expansion
WARMER
Constrained
4
Coos Bay, OR
630
2110
100
expansion
WARMER
Constrained
4
Grays Harbor, WA
630
2110
0
no change
WARMER
Constrained
4
Nisqually National,
WA
630
2110
0
no change
WARMER
Constrained
4
Padilla Marsh, WA
630
2110
0
no change
WARMER
Constrained
4
Port Susan Bay, WA
630
2110
-50
loss
WARMER
Constrained
4
Siletz Bay, OR
630
2110
0
no change
WARMER
Constrained
4
Skokomish Estuary,
WA
630
2110
67
expansion
WARMER
Constrained
4
Willapa Bay, WA
630
2110
0
no change
WARMER
Constrained
194
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
4
Bandon Marsh, OR
1420
2110
100
expansion
WARMER
Constrained
4
Coos Bay, OR
1420
2110
100
expansion
WARMER
Constrained
4
Grays Harbor, WA
1420
2110
0
no change
WARMER
Constrained
4
Nisqually National,
WA
1420
2110
100
expansion
WARMER
Constrained
4
Padilla Marsh, WA
1420
2110
100
expansion
WARMER
Constrained
4
Port Susan Bay, WA
1420
2110
96
expansion
WARMER
Constrained
4
Siletz Bay, OR
1420
2110
100
expansion
WARMER
Constrained
4
Skokomish Estuary,
WA
1420
2110
99
expansion
WARMER
Constrained
4
Willapa, OR
1420
2110
100
expansion
WARMER
Constrained
5
Alsea River, OR
390
2100
-3
loss
SLAMM 6
Constrained
5
Alsea River, OR
690
2100
-4
loss
SLAMM 6
Constrained
5
Alsea River, OR
1000
2100
-9
loss
SLAMM 6
Constrained
5
Alsea River, OR
1500
2100
-41
loss
SLAMM 6
Constrained
5
Alsea River, OR
2000
2100
-53
loss
SLAMM 6
Constrained
5
Alsea River, OR
390
2100
-3
loss
SLAMM 6
Unconstrained
5
Alsea River, OR
690
2100
-3
loss
SLAMM 6
Unconstrained
5
Alsea River, OR
1000
2100
-9
loss
SLAMM 6
Unconstrained
5
Alsea River, OR
1500
2100
-39
loss
SLAMM 6
Unconstrained
5
Alsea River, OR
2000
2100
-48
loss
SLAMM 6
Unconstrained
5
Chetco River, OR
390
2100
-21
loss
SLAMM 6
Constrained
5
Chetco River, OR
690
2100
-26
loss
SLAMM 6
Constrained
5
Chetco River, OR
1000
2100
-32
loss
SLAMM 6
Constrained
5
Chetco River, OR
1500
2100
-40
loss
SLAMM 6
Constrained
195
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
5
Chetco River, OR
2000
2100
-46
loss
SLAMM 6
Constrained
5
Chetco River, OR
390
2100
-21
loss
SLAMM 6
Unconstrained
5
Chetco River, OR
690
2100
-26
loss
SLAMM 6
Unconstrained
5
Chetco River, OR
1000
2100
-32
loss
SLAMM 6
Unconstrained
5
Chetco River, OR
1500
2100
-40
loss
SLAMM 6
Unconstrained
5
Chetco River, OR
2000
2100
-46
loss
SLAMM 6
Unconstrained
5
Coos Bay, OR
390
2100
-9
loss
SLAMM 6
Constrained
5
Coos Bay, OR
690
2100
-14
loss
SLAMM 6
Constrained
5
Coos Bay, OR
1000
2100
-14
loss
SLAMM 6
Constrained
5
Coos Bay, OR
1500
2100
-14
loss
SLAMM 6
Constrained
5
Coos Bay, OR
2000
2100
-11
loss
SLAMM 6
Constrained
5
Coos Bay, OR
390
2100
-1
loss
SLAMM 6
Unconstrained
5
Coos Bay, OR
690
2100
11
expansion
SLAMM 6
Unconstrained
5
Coos Bay, OR
1000
2100
14
expansion
SLAMM 6
Unconstrained
5
Coos Bay, OR
1500
2100
2
expansion
SLAMM 6
Unconstrained
5
Coos Bay, OR
2000
2100
-1
loss
SLAMM 6
Unconstrained
5
Nehalem Bay, OR
390
2100
-7
loss
SLAMM 6
Constrained
5
Nehalem Bay, OR
690
2100
-9
loss
SLAMM 6
Constrained
5
Nehalem Bay, OR
1000
2100
-13
loss
SLAMM 6
Constrained
5
Nehalem Bay, OR
1500
2100
-20
loss
SLAMM 6
Constrained
5
Nehalem Bay, OR
2000
2100
-19
loss
SLAMM 6
Constrained
5
Nehalem Bay, OR
390
2100
-4
loss
SLAMM 6
Unconstrained
5
Nehalem Bay, OR
690
2100
-4
loss
SLAMM 6
Unconstrained
196
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
5
Nehalem Bay, OR
1000
2100
-5
loss
SLAMM 6
Unconstrained
5
Nehalem Bay, OR
1500
2100
-2
loss
SLAMM 6
Unconstrained
5
Nehalem Bay, OR
2000
2100
13
expansion
SLAMM 6
Unconstrained
5
Nestucca River, OR
390
2100
-6
loss
SLAMM 6
Constrained
5
Nestucca River, OR
690
2100
-4
loss
SLAMM 6
Constrained
5
Nestucca River, OR
1000
2100
-4
loss
SLAMM 6
Constrained
5
Nestucca River, OR
1500
2100
-1
loss
SLAMM 6
Constrained
5
Nestucca River, OR
2000
2100
19
expansion
SLAMM 6
Constrained
5
Nestucca River, OR
390
2100
-5
loss
SLAMM 6
Unconstrained
5
Nestucca River, OR
690
2100
-3
loss
SLAMM 6
Unconstrained
5
Nestucca River, OR
1000
2100
4
expansion
SLAMM 6
Unconstrained
5
Nestucca River, OR
1500
2100
20
expansion
SLAMM 6
Unconstrained
5
Nestucca River, OR
2000
2100
47
expansion
SLAMM 6
Unconstrained
5
Rogue River, OR
390
2100
-18
loss
SLAMM 6
Constrained
5
Rogue River, OR
690
2100
-18
loss
SLAMM 6
Constrained
5
Rogue River, OR
1000
2100
-20
loss
SLAMM 6
Constrained
5
Rogue River, OR
1500
2100
-24
loss
SLAMM 6
Constrained
5
Rogue River, OR
2000
2100
-30
loss
SLAMM 6
Constrained
5
Rogue River, OR
390
2100
-18
loss
SLAMM 6
Unconstrained
5
Rogue River, OR
690
2100
-18
loss
SLAMM 6
Unconstrained
5
Rogue River, OR
1000
2100
-20
loss
SLAMM 6
Unconstrained
5
Rogue River, OR
1500
2100
-24
loss
SLAMM 6
Unconstrained
5
Rogue River, OR
2000
2100
-30
loss
SLAMM 6
Unconstrained
197
-------
Source
Location
SLR
(mm)
Projection
Year
%
Habitat
Change
Expansion
/ Loss
Model
Type
Upper habitat
constrained
in model?
5
Siuslaw River, OR
390
2100
-6
loss
SLAMM 6
Constrained
5
Siuslaw River, OR
690
2100
-10
loss
SLAMM 6
Constrained
5
Siuslaw River, OR
1000
2100
-15
loss
SLAMM 6
Constrained
5
Siuslaw River, OR
1500
2100
-9
loss
SLAMM 6
Constrained
5
Siuslaw River, OR
2000
2100
-5
loss
SLAMM 6
Constrained
5
Siuslaw River, OR
390
2100
2
expansion
SLAMM 6
Unconstrained
5
Siuslaw River, OR
690
2100
6
expansion
SLAMM 6
Unconstrained
5
Siuslaw River, OR
1000
2100
7
expansion
SLAMM 6
Unconstrained
5
Siuslaw River, OR
1500
2100
8
expansion
SLAMM 6
Unconstrained
5
Siuslaw River, OR
2000
2100
4
expansion
SLAMM 6
Unconstrained
5
Umpqua River, OR
390
2100
0
expansion
SLAMM 6
Constrained
5
Umpqua River, OR
690
2100
3
expansion
SLAMM 6
Constrained
5
Umpqua River, OR
1000
2100
-5
loss
SLAMM 6
Constrained
5
Umpqua River, OR
1500
2100
-24
loss
SLAMM 6
Constrained
5
Umpqua River, OR
2000
2100
-8
loss
SLAMM 6
Constrained
5
Umpqua River, OR
390
2100
7
expansion
SLAMM 6
Unconstrained
5
Umpqua River, OR
690
2100
17
expansion
SLAMM 6
Unconstrained
5
Umpqua River, OR
1000
2100
29
expansion
SLAMM 6
Unconstrained
5
Umpqua River, OR
1500
2100
43
expansion
SLAMM 6
Unconstrained
5
Umpqua River, OR
2000
2100
50
expansion
SLAMM 6
Unconstrained
The tide flats were similar to the marshes in that the models predicted increases at a number of
sites. To address this, we treat the tide flats, and associated oyster beds, like the low marshes and
focus on the sites with predicted losses (Figure 7-3 and Figure 7-4). At a SLR of 390 mm with
the unconstrained tide flats, 50% of the sites had predicted declines of <10% (minor habitat
threshold) and 25% of the sites had losses between 11% and 29% (low habitat threshold). No
198
-------
sites had moderate habitat losses (>30%). Given the number of sites predicted to have a less than
a 10% loss, we use 390 mm as the minor habitat threshold. At 1000 mm, both the minor and low
percent losses each equaled 25% of the sites but moderate losses now constitute 12.5% of the
sites. The onset of sites with losses >30% suggests that a SLR value of 1000 mm is a justifiable
threshold between the low and moderate habitat threshold classes. The proportion of sites with
moderate habitat losses increases up to a SLR of 2000 mm, but no sites have predicted losses
>50%, indicating the high habitat threshold is greater than this value. To approximate the break
between moderate and high thresholds, we increase the 2000 mm value to 2250 mm to capture
substantial effects on tide flats, though this value should be evaluated in SLAMM or other
models.
% of Sites in Each Habitat Loss Threshold Class - Intertidal
Sand/Mud Flats Unconstrained data
50
45
40
(/) 35
O) "
5 30
¦D 25
K 20
o 15
^ 10
1
390 690 1000 1500 2000
Net SLR
Minor, %<= -10 ¦ Low, % 11 -29 loss ¦ Moderate,% 30-49 loss ¦ High, %>=50 losss
Figure 7-3. Tide flats - Unconstrained.
The values represent the percentage of sites that fall within each habitat threshold categories (Minor,
Low, Moderate, High) under various SLR amounts. Only sites with predicted habitat losses are included
in the analysis.
199
-------
% of Sites in Each Habitat Loss Threshold Class - Intertidal
Sand/Mudflats Constrained data
111 ,i, ll r.N
340 390 630 690 770 1000 1500 2000
Net SLR (mm)
Minor, %<- -10 ¦ Low, % 11 -29 loss Moderate,% 30-49 loss ¦ High, %>=50 losss
Figure 7-4. Tide flats - Constrained.
The values represent the percentage of sites that fall within each habitat threshold categories (Minor,
Low, Moderate, High) under various SLR amounts. Only sites with predicted habitat losses are included
in the analysis.
7.8.7 Mangroves
Mangrove forests are critical intertidal ecosystems occurring throughout the subtropics and
tropics (FAO, 2007; Spalding et al., 2010). Within the NEP, mangroves are abundant in both the
Gulf of California and the Magdalena ecoregions. The northern limit of mangroves in the Eastern
Pacific is Rhizophora mangle found just north of Laguna San Ignacio at the northern tip of the
Magdalena Transition Ecoregion, while Laguncularia racemosa is first found just to the south
(Spalding et al., 2010). Compared to the Indo-West Pacific (see Polidoro et al., 2010), the
diversity of mangrove species is low in the NEP, with only four species: Avicennia germinans,
Rhizophora mangle (referred to as Rhizophora samoensis by Polidoro et al., 2010), Laguncularia
racemosa, and Conocarpus erect us. While recognizing that the relative abundances of these four
species will vary among individual mangrove forests, we evaluate mangroves in toto without
reference to the individual species.
The mangroves in the Gulf of California are already under stress, decreasing at an annual rate of
about 2% because of sedimentation, eutrophication and deforestation (Aburto-Oropeza et al.,
2008, Lopez-Medellin et al., 2011). Over an 8-year period between 1973 and 1981 there was a
V)
175
>
T3
3
4->
cn
v»-
o
£
100
90
80
70
60
50
40
30
20
10
0
200
-------
23% decline in the mangroves in La Paz due to development (Aburto-Oropeza et al., 2008). Sea
level rise is now added as an additional stressor. As with other intertidal vegetated habitats, the
sediment accretion rate is a key factor determining the long-term effects of SLR (Gilman et al.,
2008). Accordingly, the two most vulnerable mangrove forests are low-relief carbonate islands
with low rates of sedimentation and little available upland space and arid, semi-arid, and dry sub-
humid regions which also have limited sediment inputs (Webber et al., 2016). Examples of
vulnerable mangrove forests are those found on low lying keys composed of carbonate
sediments. The least vulnerable are those occurring along macrotidal coastlines with significant
riverine inputs and high accretion rates (Webber et al., 2016). Under these conditions, mangroves
can keep pace with SLR. Mclvor et al. (2013) listed locations of mangroves around the world
along with the rates of SLR, ranging between 0.85 mm/yr to 10 mm/yr, in which each mangrove
forest was able to "keep pace with" SLR via accretion. Though semi-arid, the mangroves in Baja
and the Gulf of California appear to tend more towards the less vulnerable category based on the
reports of inland migration.
Mangroves have shown an ability to migrate inland with rising water levels. In Magdalena Bay,
the landward margins of mangroves have shown a significant increase with SLR (Lopez-
Medellin et al., 2011). There has been more than a 20% increase in the canopy cover in
Magdalena Bay, with mangrove saplings now growing in the landward mudflats (Lopez-
Medellin et al., 2011). A consequence of this inland migration is the loss of marshes, salt flats,
and mudflats. In another study, a SLAMM model predicted reductions in salt marsh and
oligohaline marsh areas with increased mangrove areas in Tampa Bay, Florida (Meyer, 2013).
However, these increases are for mangroves as a group and individual species respond
differently. Specifically, the species on the seaward side may be more vulnerable (Lopez-
Medellin et al. (2011).
Table 7-9 summarizes 16 papers on how mangroves respond to SLR. One set of these studies
evaluated the historical response to sea level rise during the Holocene while the other studies
modeled future changes in response to SLR. The Holocene studies relate changes to rates of SLR
(mm/yr) and do not report the total SLR (mm), which are the basis of the habitat thresholds.
Additionally, some of the initial Holocene studies were alarming. Based on a reconstruction of
mangrove responses to SLR during the Holocene, Ellison and Stoddart (1991) concluded that
mangroves in low islands would not persist with a SLR rise of 12 cm per 100 years (1.2 mm/yr).
However, it has been pointed out that mangroves survived in Key West, Florida at a rate of 19
cm per 100 years (Mclovr et al., 2013) and recent reviews (e.g., Alongi, 2015; Godoy and de
Lacerda, 2015; Woodroffe et al., 2016) indicated that many, if not most, mangrove systems are
relatively robust to sea level rise.
To set habitat threshold values, we focused on the modeling studies and used the Holocene
studies with approximately similar environments to Baja and the Gulf of Mexico as a check.
201
-------
Only two of the studies pushed their models to determine the SLR associated with a >50%
decline in mangroves. Geselbracht et al. (2015) had a 59% reduction in mangrove forest at 2000
mm SLR while Warren Pinnacle Consulting, Inc. (2014) predicted a 50%, 62%, and 64%
reduction in mangrove cover at 1200 mm, 1500 mm, and 2000 mm SLR, respectively. To
generate the high habitat threshold, we averaged the 2000 mm from Geselbracht et al. (2015)
with the 1200 mm result from Warren Pinnacle Consulting, Inc. (2014), for a value of >1600
mm. This then sets the moderate threshold at 1600 mm (Table 7-3). The minor habitat threshold
is the net SLR that results in <10% loss, and possible gains, which we generated from the lower
end of the modeling results. Three studies showed declines of 5 to 13% at SLR of 290 mm to 880
mm. Another two showed increases of 35% at SLR of 640 mm and 700 mm. Based on this
range, we set the minor habitat threshold at 750 mm, a value that should not result in substantial
effects on mangroves not starved of sediment. To estimate the low habitat threshold, we
interpolated between the minor and moderate thresholds to generate a value of 1150 mm.
Table 7-9. Summary of mangroves percent habitat change under different SLR values and rates.
Compiled from 16 studies. Modeling studies reported results in sea level rise (mm) while the Holocene
studies reported sea level rise rates (mm/yr). ND = no data. Sources: 1 - Geselbracht et al., 2015; 2 -
Warren Pinnacle Consulting, Inc. 2014; 3 - Traill et al., 2011; 4 - Di Nitto et al, 2014; 5 - Seddon et al.,
2011; 6 - McKee et al., 2007; 7 - Ellison, 2000; 8 - Fujimoto et al., 1996; 9 - Woodroffe and Mulrennan,
1993; 10 - Woodroffe, 1990; 11 - Woodroffe, 1990 (in Woodroffe, 1995); 12 - Maul and Martin, 1993 (in
Snedaker et al., 1994); 13- Parkinson, 1989 (in Snedaker et al., 1994); 14 - Woodroffe, 1995 (in Mclvor
et al., 2013); 15 - Woodroffe, 1990 (in Mclvor et al., 2013); 16 - Ellison and Stoddart, 1991.
Source
Location
Year
Projected to
/
Time Period
SLR
(mm)
SLR
Rates
(mm/yr)
% Habitat
Change
Expansion /
Loss
Model /
Scenario
1
Estuaries in
Florida's Gulf
Coast
2100
700
ND
+35
expansion
SLAMM
1
Estuaries in
Florida's Gulf
Coast
2100
1000
ND
+40
expansion
SLAMM
1
Estuaries in
Florida's Gulf
Coast
2100
2000
ND
-59
loss
SLAMM
2
US Coastline
of Gulf of
Mexico
2100
500
ND
-10
loss
SLAMM
2
US Coastline
of Gulf of
Mexico
2100
1000
ND
-39
loss
SLAMM
202
-------
2
2
2
3
3
3
4
4
4
4
5
6
Location
Year
Projected to
I
Time Period
SLR
(mm)
SLR
Rates
(mm/yr)
% Habitat
Change
Expansion I
Loss
US Coastline
of Gulf of
Mexico
2100
1200
ND
50
loss
US Coastline
of Gulf of
Mexico
2100
1500
ND
¦62
loss
US Coastline
of Gulf of
Mexico
2100
2000
ND
¦64
loss
Southeast
Queensland,
Australia
2100
290
ND
loss
Southeast
Queensland,
Australia
2100
640
ND
+ 35
expansion
Southeast
Queensland,
Australia
2100
1790
ND
19
loss
Gazi Bay,
Kenya, East
Africa
2100
90
ND
+ (0-?)
expansion
Gazi Bay,
Kenya, East
Africa
2100
200
ND
(0-?)
expansion
Gazi Bay,
Kenya, East
Africa
2100
480
ND
(0-?)
expansion
Gazi Bay,
Kenya, East
Africa
2100
880
ND
13
loss
Diablas
lagoon,
Isabela Island,
Galapagos
Since 2700
years BP
ND
5.7
10
loss
Twin Cays,
Caribbean
Since 8000
years BP
ND
3.5
(0-?)
expansion
203
-------
Source
Location
Year
Projected to
/
Time Period
SLR
(mm)
SLR
Rates
(mm/yr)
% Habitat
Change
Expansion /
Loss
Model /
Scenario
6
Twin Cays,
Caribbean
Since 7600
years BP
ND
5
-100
loss
Holocene
Analysis
7
Low islands
Holocene
ND
1.2
-(0-10)
loss
Holocene
analysis
7
High Islands
Holocene
ND
4.5
-(0-10)
loss
Holocene
analysis
8
Kosrae Island,
Micronesia
Since 5000
years BP
ND
<2
-(0-10)
loss
Holocene
Analysis
8
Kosrae Island,
Micronesia
Since 5000
years BP
ND
2-10
-(10-50)
loss
Holocene
Analysis
8
Kosrae Island,
Micronesia
Since 5000
years BP
ND
> 10
-(50-100)
loss
Holocene
Analysis
9
Australia
Since 6790
years BP
ND
10
-(0-10)
loss
Holocene
analysis
10
Multiple
studies
Since 6500
years BP
ND
5-8
-(0-10)
loss
Holocene
analysis
10
Multiple
studies
Since 6500
years BP
ND
8-10
-(10-50)
loss
Holocene
analysis
10
Multiple
studies
Since 6500
years BP
ND
10-15
-(50-100)
loss
Holocene
analysis
11
Northern
Australian
Estuaries
Not Specified
ND
5-8
-(0-10)
loss
Holocene
analysis
12
Key West,
Florida
1925-1992
ND
2.3
-(0-10)
loss
1846-1992 data
analysis (in
Literature
Review)
13
Ten Thousand
Island region
of Florida
Not Specified
ND
2.7
-(0-10)
loss
Holocene
analysis (in
Literature
Review)
14
Northern
Australian
estuaries
Not Specified
ND
8-10
-(0-10)
loss
Holocene
analysis
204
-------
Source
Location
Year
Projected to
/
Time Period
SLR
(mm)
SLR
Rates
(mm/yr)
% Habitat
Change
Expansion /
Loss
Model /
Scenario
15
South Alligator
tidal river
Not Specified
ND
0.2-6
-(0-10)
loss
Holocene
analysis
16
Worldwide,
low islands?
Not Specified
ND
0.8-0.9
-(0-10)
loss
Holocene
analysis
16
Worldwide,
low islands?
Not Specified
ND
0.9-1.2
-(10-50)
loss
Holocene
analysis
16
Worldwide,
low islands?
Not Specified
ND
>1.2
- (0-50)
loss
Holocene
analysis
16
South Florida
last 4000-
5000 years
ND
0.46
-(0-10)
loss
Holocene
analysis
16
South Florida
last 4000-
5000 years
ND
0.98
-(0-10)
loss
Holocene
analysis
7.9 Sea Level Rise Risks for Invertebrate and Fish Species
The habitat thresholds predict the percent loss of major habitat types with SLR. The next step is
to translate these habitat losses into impacts on the populations of the invertebrate and fish
species associated with the various habitats. As a first-order assumption, we assume that the
population decline in the target species is proportional to the loss in its habitat area. This is
easiest to visualize with sedentary species, such as barnacles, where the species attaches to the
habitat. But even with mobile species, over sufficient time periods, we assume the population
equilibrates to the available area, especially at regional scales. This is not a necessary assumption
and it would be possible to incorporate "habitat multipliers" to adjust the percent change in the
population as a function of the percent of habitat loss for specific habitat types. However, we are
unware of information indicating that such non-linear responses to habitat loss are an important
general phenomenon with near-coastal species.
Two biotic traits used to assess the relative importance of a habitat to a species are the species'
depth preferences and habitat preferences (Figure 7-1). Combined with the concept of high and
low exposure habitats (Section 7.7), we produced a SLR risk matrix (Table 7-10). The values in
Table 7-10 are global and apply to all ecoregions, but users have the option to modify these
values in CBRAT on an ecoregion-by-ecoregion basis.
205
-------
Table 7-10. Risk values assigned to each combination of depth, habitat and exposure classes for each
habitat threshold.
Risks in this table are global values meaning that the same risks are applied to all ecoregions. Risk
classifications are: minor risk = 0; low risk = -1; moderate risk = -2; high risk = -3. Obs. = observed depth
or habitat, Pref. = preferred depth or habitat.
Habitat
Threshold
Pref. High
Exposure
&
No Low
Exposure
Depth
Pref. High
Exposure
&
Obs. Low
Exposure
Depth
Pref. High
Exposure
&
Pref. Low
Exposure
Depth
Obs. High
Exposure
&
No Low
Exposure
Depth
Obs. High
Exposure
&
Obs. Low
Exposure
Depth
Obs. High
Exposure
&
Pref. Low
Exposure
Depth
No High
Exposure
&
Obs. Low
Exposure
Depth
No High
Exposure
&
Pref. Low
Exposure
Depth
Minor
0
0
0
0
0
0
0
0
Low
-1
-1
-1
-1
-1
-1
0
0
Moderate
-2
-1
-1
-1
-1
-1
0
0
High
-3
-2
-2
-1
-1
-1
0
0
7.9.1 Depth Preferences
Depth preferences of the target species are used to approximate what proportion of species'
population occurs intertidally, and thus is potentially vulnerable to sea level rise. CBRAT uses a
three-level classification system for depth classification (Lee et al., 2015). For the current SLR
risk algorithm, it is only necessary to use the Level II classifications (e.g., intertidal and neritic)
for benthic species and the Level I classification for pelagic species. However, classifying these
depth classes as observed versus preferred (see Section 4.1.1) has a major influence on the
assigned risk since it is assumed that only a relatively small proportion of the population occurs
at depths classified as observed versus the majority of the population occurring in depth(s)
classified as preferred.
Depth is divided into high exposure and low exposure depth classes (Section 7.7), with the high
exposure classes directly affected by SLR and the low exposure classes minimally affected.
Because most species occur in multiple depth classes, a set of rules is used to approximate the
extent of the population captured by different combinations of observed and preferred depth
classes (Table 7-10). An example is a species with preferred high exposure class(es) and no low
exposure class(es). These are exclusively intertidal species, the highest SLR risk scenario. At the
other end of the spectrum are species with preferred low exposure classes and no high exposure
classes. These are subtidal or pelagic species, and they are assigned a minor (0) risk across all
habitat thresholds. One intermediate scenario is a species with preferred high exposure depth
classes and an observed low exposure depth classes. Since the majority of the population is
predicted to be in an exposed depth, we do not downgrade the risk compared to exclusively
206
-------
intertidal species. The other intermediate scenario is a species with a preferred low exposure
class and just an observed high exposure class. In this case, the assumption is that only a small
portion of the population would be vulnerable to SLR, and thus the risk would be downgraded to
low (-1).
7.9.2 Habitat Preferences
As with depth preferences, habitat preferences are used to predict where the majority of the
population occurs. In this iteration of the SLR model, organisms are keyed to the Level I or
Level II Ecosystem/Habitat classes in CBRAT (see Lee et al., 2015). For example, under
Unconsolidated Ecosystems (Level I), an organism can be linked to SAV or to mangroves
(Level II), but the SLR analysis does not consider the Level III habitats, such as the specific
species of SAV or mangrove.
7.9.3 Final SLR Risk
If a species only occurred in one habitat, the risk would be determined from depth and the habitat
threshold. However, many, if not most, coastal species occupy more than one habitat type. For
example, of the 366 brachyuran crabs, 310 occurred in at least two distinct 2nd level habitat
classifications. To address occupancy of multiple habitat types, the risk based on habitat
thresholds and depth preferences are modified by the combination of observed and preferred
habitat classifications of the target species (Table 7-10). The rules are based on the assumption
that the majority of the population occurs in the preferred habitat(s). In cases with occupation of
multiple habitats, risk is calculated for each habitat independently with the final SLR risk
assigned as the greatest risk among the preferred habitats.
Assessing risk in poorly studied species for which it is not possible to identify preferred habitats
or depths is more complicated. Of the four cases (Table 7-10):
In the case of "Observed High Exposure & No Low Exposure Depth", the species has
only been reported only from intertidal habitat(s) but it is not known whether this is a
preferred habitat or what the preferred depths are. Because of the uncertainty, the risk is
downgraded by one class compared to the "Preferred High Exposure & No Low
Exposure Depth" (exclusively intertidal) scenario. This results in a moderate risk (-2)
with high habitat thresholds and low risk (-1) at moderate to low habitat thresholds.
In the case of "Observed High Exposure & Observed Low Exposure Depth", it is again
known that the species occurs in an intertidal habitat, but it is also known that the species
extends into the subtidal (or pelagic). Based on the available information, it is assumed
that a smaller portion of the population occurs in the intertidal and is thus less vulnerable
to sea level rise. For these poorly known intertidal and subtidal species, we assign a low
risk (-1) for low to high habitat thresholds.
207
-------
The "Observed High Exposure & Preferred Low Exposure Depth" captures species that
are primarily subtidal but also occur in the intertidal. Because the available information
indicates that the majority of the population would not be vulnerable to SLR, the risks are
set to low (-1) for low to high habitat thresholds.
The last case, "No High Exposure & Observed Low Exposure Depth" captures species
that are only known from the subtidal, though it is not known if this is the preferred depth
range. Given the uncertainty, these species are assigned a minor (0) risk.
While these cases with only observed habitats are necessary for completeness, only 34 of the 366
brachyuran crabs currently do not have a preferred habit
208
-------
Section 8. Uncertainty Analysis and Quality
Assurance/Quality Control
8.1 Uncertainty Analysis - Overview
Uncertainty analysis is a key component of risk assessments. In this section, we address the
strategy for conducting an uncertainty analysis for our climate framework. The full uncertainty
analysis will be conducted as part of the climate risk analysis for crabs, bivalves, and rockfish
(Lee et al., in progress). Our approach to uncertainty draws from several sources, including the
IPCC (Mastrandrea et al., 2010), Integrated Environmental Health Impact Assessment System
(http://www.integrated-assessment.eu/eu/index.html. Salway and Shaddick, no date), and
Planque et al. (2011). Though the specifics differ slightly among these sources, the key steps for
a qualitative uncertainty analysis based on Salway and Shaddick are:
1. Identification of uncertainty sources
2. Qualitative characterization of uncertainty in terms of:
a. direction and magnitude of uncertainty on the results
b. knowledge about the uncertainty source
3. Reporting of qualitative uncertainties in a non-technical summary
We address identification and characterization of uncertainties in Section 8.2 and the reporting of
uncertainties in Section 8.3. Section 8.4 provides the documentation required under the Western
Ecology Division's Quality Management Plan.
8.2 Sources and Levels of Uncertainty
Table 8-1 lists the major sources of uncertainty related to the biotic traits and an estimate of their
level of uncertainty. We also list examples of major uncertainties related to the numerical climate
values and the risk assessment model, specifically the climate thresholds and model assumptions.
A full analysis of thresholds and assumptions will be provided with the risk assessment.
209
-------
Table 8-1. Preliminary analysis of the major sources of uncertainty in the climate risk framework.
Uncertainty levels for biotic traits are derived from estimated percentages of misclassifications. Uncertainty levels for climate projections,
thresholds, and major assumptions are preliminary qualitative estimates; quantitative evaluations from sensitivity studies will replace these as part
of the formal risk assessment. The "Directionality of Uncertainty" are estimates of whether errors in the parameter are more likely to overestimate
risk, underestimate risk, or are random. Sections indicates the primary sections in this document where the parameter is discussed in terms of
generating risks.
Parameter
Level of
Uncertainty
Directionality
of Uncertainty
Sections
Comments
Biotic Traits
Global distributions
Low
Overestimates
risks
4.2
Errors most likely underestimate global distributions, especially
in tropics.
Abundance classifications
Low
Random
3.3
Only need to identify abundant or rare species, depending
upon the rule.
Depth preferences
Low
Random
5.4, 7.9.1
-
Habitat preferences
7.8
-
Breeding type
Low
Random
6.4
-
Nonindigenous species
Moderate
Underestimates
risk
4.2.8
Biggest source of uncertainty is whether an NIS is established
in Asia.
Population trends
Moderate
Underestimates
risks
4.3.4
In absence of evidence, defaulted to "no apparent trend",
which likely underestimated the number of species with
population declines.
Transients
Low
Random
4.3.6
Uncertainty in distinguishing rare species vs. a non-established
vagrant.
Habitat specialization
Low
Random
4.4.3
-
Trophic specialization
Low
Random
4.4.4
-
Symbiotic relationship
Moderate
Overestimates
risks
4.4.2
The major source of uncertainty is the unknown response of
the hosts to climate change; assume that hosts impacted.
Anadromous / Catadromous
Low
Random
4.4.5
-
210
-------
Parameter
Level of
Uncertainty
Directionality
of Uncertainty
Sections
Comments
Productivity parameters
Low
Random
4.4.6
Fish only.
Historical Climate Values and Projections
Historical SSTs-AVHRR
(Cortezian to Eastern Bering)
Low
(Estuaries Moderate)
Random
5.3
Values in Puget Trough/Georgia Basin Ecoregion may not be
as accurate. SST values for offshore, not in estuaries though at
regional scales there should be general correspondence.
Historical SSTs-AVHRR (Chukchi
and Beaufort)
Moderate
Underestimates
risks?
5.3
Substantial loss of winter data results in higher annual mean.
Small differences in mean temperatures between Chukchi and
Beaufort are susceptible to small errors.
Historical SST-CMIP5
Low?
(Estuaries Moderate)
Random
5.4
CMIP5 is at too coarse a resolution for estuaries.
Historical 30-m and 100-m depth
temperatures - CMIP5
Low?
(Estuaries Moderate
to High?)
Random?
5.4
Subsurface temperatures only very generally applicable to
estuaries.
Projected SST and subsurface
temperature increases
Moderate
(Estuaries Moderate
to High)
Random?
5.4
Will evaluate min and max models in risk assessment.
Subsurface temperatures only very generally applicable to
estuaries.
Historical temperatures - Air
Low
Random
5.4
Moderate to High for the Puget Trough/Georgia Basin because
of averaging temperatues to the north and south.
Projected temperature - Air
Moderate
Random?
5.4
Will evaluate min and max models in risk assessment.
Moderate to High for the Puget Trough/Georgia Basin because
of averaging projections to the north and south.
Historical aragonite saturation state
values
Low?
(Estuaries Moderate
to High?)
Random?
6.2.2
-
Projected aragonite saturation
values
Moderate
(Estuaries Moderate
to High)
Random?
6.2.2
-
Historical pH values
Low?
(Estuaries
Moderate?)
Random
6.2.1
CMIP5 is at too coarse a resolution for estuaries.
211
-------
Parameter
Level of
Uncertainty
Directionality
of Uncertainty
Sections
Comments
Projected pH values
Moderate
(Estuaries Moderate
to High?)
Random
6.2.1
-
Projected sea level rise
Moderate?
Underestimates
risks?
7.3
Several models projected greater sea level rise. Limited
isostatic rates in several ecoregions.
Climate Thresholds
ETW thresholds based on SDs
around mean in WOE
Low?
Random
5.3.3
-
BTL thresholds based on bins
between WOE and NWUE
Moderate?
(High for Magdalena
and Cortezian)
Underestimates
risks
5.4.2
From S. California north, the BTL underestimated risk
compared to the ETW in about 13% of the cases.
Overestimated risk in Magdalena and Cortezian.
pH & aragonite saturation state
High
Random?
6.3
Relatively small changes in the MATC values can have large
impacts on risk. Changes in classification of species as high,
moderate, or low sensitivity can have large impacts on risk.
SLR habitat thresholds
Moderate?
Overestimates
risk
7.8
In many localities, in absence of barriers, SAV, lower marsh,
and mangrove can stay up with SLR by migrating inland.
Major Assumptions (Examples)
Warm edge limits are determined by
direct and indirect effects of
temperature.
Low
Overestimates
risks
Appendix D
When temperature is not the direct/indirect cause for the
absence of a species in a warmer ecoregion, assigning
temperature as the cause overestimates the temperature risk.
Warm genotypes from southern
ecoregions will colonize northern
ecoregions as they warm.
Moderate
Underestimates
risks
5.3.2
Violation results in greater thermal risk in northern ecoregions.
Ocean projections for pH and
aragonite saturation are indicative of
estuarine risk.
High
Unknown
6.1
Current regional-scale models are at too coarse a resolution to
generate estuarine projections.
212
-------
The IPCC provides guidance for quantitatively calibrated levels of confidence (Mastrandrea et
al., 2010) that we adapted to evaluate levels of uncertainty (Table 8-2). For abundance and biotic
trait classifications, we can estimate the confidence levels based on the likely number of
misclassifications for the corresponding classes. For example, there are 704 crab species X
ecoregion combinations that have an abundance classification. Based on our interactions with
experts at several workshops, we estimate the error rate for abundance classifications is less than
20% (<140 misclassifications) and potentially less than 10% (<70 misclassifications). Thus, we
assign a low uncertainty to this parameter, especially considering it is only necessary to identify
the rare or abundant species, depending upon the rule. There may be greater uncertainty in the
assignment of depth range as observed versus preferred, but we estimate that this parameter has
less than a 20% error rate, and is also assigned a low level of uncertainty. A more detailed
analysis, as will be conducted with the risk assessments, would evaluate whether different
confidence levels should be applied to abundant species versus rare species, which are less well
known.
Table 8-2. Level of confidence adapted from the IPCC.
Adapted from Mastrandrea et al. (2010). We derived the "Uncertainty Levels in Current Analysis" based
on the IPCC guidance.
Level of
Confidence
Degree of Confidence
in Being Correct
Uncertainty Level in
Current Analysis
Very high
At least 9 out of 10 chance
Low
High
About 8 out of 10
Low
Medium
About 5 out of 10
Moderate
Low
About 2 out of 10
High
Very low
Less than 1 out of 10
High
The translation of the criteria in Table 8-2 to numerical parameters and model structure is less
clear than for the biotic traits. Rather, we use the criteria listed in Salway and Shaddick for
qualitative risk assessments:
"The magnitude of uncertainty is rated low when it is judged that large changes within
the source of uncertainty would have only a small effect on the assessment results and
when the values of the data sets needed for the assessment are known. A designation of
medium implies that a change within the source of uncertainty is likely to have a
moderate effect on the results and the values of the data sets needed for the assessment
are unknown (completely or partially). A characterization of high implies that a small
change in the source would have a large effect on results and the values of the data sets
needed for the assessment are unknown."
For the numerical parameters and thresholds, determining whether a "small" change in their
values would have minor or large effects on the results will be addressed by conducting
213
-------
simulation studies as part of the risk assessment. In the interim, we provide preliminary
qualitative assessments of uncertainty in Table 8-1.
8.3 Reporting of Uncertainty
The detailed reporting of uncertainties will be presented in the risk assessment report (Lee et al.,
in progress). However, we provide here a preliminary report on the overall assessment of the
uncertainties:
• There is less uncertainty in the biogeographical patterns of risk for a taxon overall and the
taxonomic patterns of risks among taxa than for an individual species.
• Greatest uncertainty for risks is associated with ocean acidification.
• Least uncertainty is for risk associated with temperature (other than Magdalena and
Cortezian ecoregions).
• The lack of sufficient spatial resolution in the available regional-scale climate predictions
of temperature and pH changes in estuaries increases the uncertainty associated with
estuarine organisms.
• The habitat thresholds are the greatest source of uncertainty in the SLR algorithm.
• For some species, the actual risks associated with low and moderate risk classifications
may be underestimated because of stressor interactions and/or unmodeled effects (e.g.,
disease).
• Predictions are sufficient to identify the scope and patterns of risk and for regional-scale
adaptation planning.
• Predictions are sufficient to flag high risk versus low risk commercial/recreational
species but not sufficient for fisheries management.
8.4 EPA/ORD's Quality Assurance/Quality Control
This research falls under ORD's quality assurance Category B. The research presented in this
report was conducted under the following Quality Assurance/Quality Control documents:
Standard Operating Procedure:
Lee II, H., Marko, K., Hanshumaker, M., Folger, C., and Graham, R. 2015. User's Guide &
Metadata to Coastal Biodiversity Risk Analysis Tool (CBRAT): Framework for the
Systemization of Life History and Biogeographic Information. EPA Report. EPA/601/B-15/001.
123 pages.
214
-------
Quality Assurance Project Plans (OAPPs):
Secondary Data Collection and Analysis for Estuarine Ecosystem Services Research Project:
Multi-Scalar Benthic Indicators, Estuary Scale, Regional Scale, and Estuarine Global Climate
Change Tasks. QAPP-NHEERL/WED/PCEB/HL/2009-01-r0.
Coastal Biodiversity Risk Assessment Tool (CBRAT): Assessing impacts of individual and
multiple climate stressors on near-coastal species at a regional scale Air Climate and Energy
(ACE) Program. E-WED-0030833.
Quality Management Plan:
Quality Management Plan, Western Ecology Division (WED), National Health and
Environmental Effects Research Laboratory (NHEERL), Office of Research and Development,
United States Environmental Protection Agency. Corvallis, Oregon. QMP-
NHEERL/WED/1995-01 -r4.0.
Electronic Notebook:
This project archives significant project documents in an electronic notebook (MS OneNote) in
accordance with Office of Research and Development (ORD) PPM 13.6, Scientific
Recordkeeping: Electronic. The One Note electronic notebook is not intended to be inclusive of
all electronic records used in the project but rather is seen as a starting point for an electronic
records structure for consistency and as a valuable resource for all researchers involved with the
project.
Quality Objectives and Criteria for Existing Measurement Data:
Nearly all the data entered into CBRAT will be existing information available from the scientific
literature, published books, and scientific databases. Existing or secondary data is defined as
information previously collected for other projects or intended applications. Potential limitations
on the use of the existing data for CBRAT are best appreciated with respect to their original
intended application. To facilitate this understanding, the source(s) of the information is
documented for each species in CBRAT in the 'Comments' section including full references for
each original data source. Every species also has a 'References' page that lists the papers,
reports, databases, and personal communications that are linked to that particular species. Other
data, including species relative abundance and population trends, are generated from key
literature specific to each taxonomic group and potentially augmented by expert opinion. The
result of this quality assurance effort is a transparent presentation of existing data sources and
any limitations on their use in the context of the original study.
215
-------
Audit Records:
CBRAT was the subject of an external audit review in 2015. Auditors external to the EPA
reviewed the project's electronic notebook and interviewed Project staff and WED's QA
Manager. The auditors found no deficiencies in the project and noted several best practices such
as documentation of records.
216
-------
Appendix A.
Under The Hood - Hardware, Software,
Access Levels, & Backups
A-1 Servers
CBRAT is a database backed website developed with Ruby on Rails web application framework.
The biological information is stored in a PostgreSQL database. PostGIS provides spatial,
geographic objects for the PostgreSQL database and interacts with MapServer to display species'
geographic distributions and abundances.
The CBRAT Information system has three servers in its configuration: 1) maintenance server, 2)
development server, and 3) public server. Each server plays an important role in the tools
development cycle as described below. Each of the CBRAT servers has a Linux based operating
system. On the development and public servers, Apache2 is installed as the web hosting software
to host the development and public websites. These servers are currently located within the
Northwest Knowledge Network (NKN) at the University of Idaho
(https://www.northwestknowledge.net/).
Maintenance Server: The maintenance server provides the repository for backups (described
below). It also has Git server software installed to provide project management for software
development between multiple users. When changes to the code base are made, the changes are
documented and pushed to the Git server. The Git server maintains version control so that code
changes can be reversed if the changes create unresolvable errors. Once code changes are
documented and pushed to the Git server, the code is then pushed to the development server.
Development Server: The development server provides a staging area to test new functionality
and tools being designed to enhance CBRAT. The data on the development server can be
changed or manipulated to test different risk scenarios without affecting the final data in the
public server. It also provides a mirror copy of the public-facing CBRAT web server to evaluate
compatibility between current versions and new software releases prior to upgrades on the
public-facing server. The database on the development server is periodically updated with the
most recent public database version, but the data on the development server is not backed up, and
is never used to update the public server. It exists strictly for testing and training purposes.
217
-------
Public Server: The public-server hosts the website and current database, and is the version used
by managers and researchers to conduct the risk assessments. The public server is the repository
of all the biological, environmental, and geographical data as well as the rule sets used to
calculated risk. Different access levels are managed by a website security certificate maintained
by Digicert, as described below.
A-2 Software
• Operating System: Ubuntu 12.04.2 LTS (GNU/Linux 3.5.0-23-generic x86_64).
• Apache2 Server version: Apache/2.2.22.
• Ruby: Ruby 1.9.3p49 (2013-05-15 revision 40747 [x86_64-linux],
• Rails: Rails 3.0.11.
• Git: git version 1.7.9.5.
• PostgreSQL: PostgreSQL 9.1.9 on x86_64-unknown-linux-gnu, compiled by gcc
(Ubuntu/Linaro 4.6.3-lubuntu5) 4.6.3, 64-bit.
• POSTGIS: POSTGIS="2.0.0 r9605" GEOS="3.3.8-CAPI-1.7.8" PROJ="Rel. 4.8.0, 6
March 2012" LffiXML="2.8.0".
• MapServer: MapServer 6.2.1.
A-3 Access Levels
Depending on the user's expertise and interest, the administrator will assign an access level to
each user's account from one of the following categories with the corresponding privileges. A
summary of these privileges is displayed in Table A-l.
Public: The first level is public access. No login is required to view biological or environmental
information that has been reviewed and released by the U.S. EPA. Information on individual
species and/or taxonomic groups of species as well as risk assessments will be released to the
public as the information is finalized through the review process. The public cannot change any
data or run new risk assessments.
Manager: This level requires a user name and password to gain access to biological and
environmental information that has not been released to the public. Users with this access level
are able to view all information, test tools, and temporarily modify inputs into the risk analyses.
Users with this access level are not able to edit information contained in the database but they do
218
-------
have the ability to submit issues related to any bugs encountered, incorrect information about a
species, or suggestions on how to improve the website.
Expert User: This third level of access also requires a user name and password and provides the
additional ability to add/edit all the information contained in CBRAT except user accounts and
deleting or combing species. Experts can enter "Master" records, though they can be overwritten
by Gatekeepers and Administrators.
Gatekeepers: This access level also requires a user name and password for access. Gatekeepers
have all the access privileges of expert user/taxonomic experts and have the added ability to
review information that has been entered by other expert users. Gatekeepers also have the ability
to import data from spreadsheets with a linked PDF as well as delete or combine species.
Administrator: This access level also requires a user name and password for access.
Administrators have all the access privileges of gatekeepers. Administrators can view user's
statistics (e.g., hours logged-in) and can approve species for public viewing. Administrators also
have access to the user management tools to edit and change access levels of all user accounts.
A-4 Backup Strategy
Cron is a time-based job scheduler software utility used to automate system administration tasks
on Unix and Linux operating systems. A Linux cron job has been created to automatically back
up the website database every evening at midnight and at the end of every month at midnight to
the maintenance server. Each of the last seven days of backups is stored using the day of the
week naming convention with each day overwriting the backup made seven days previous.
Likewise, the monthly backups are stored using a month naming convention with each month
overwriting the backup made the end of the month one year ago. This cron job is located on the
public server and manages all the weekly, monthly and annual backups as outlined. Offsite
backup storage of the database and website code is updated regularly in the event of a
catastrophic hardware failure at University of Idaho.
219
-------
Table A-1. Summary of the privileges associated with each level of access in CBRAT.
Privileges
Public
Manager
Expert
Gatekeeper
Administrator
Login Required
no
X
X
X
X
View Public Species
X
X
X
X
X
Generate Spreadsheet Summaries of Abundance and Life History
Data for Public Species
X
X
X
X
X
Generate PDF Profile for a Single Public Species
X
X
X
X
X
View Non-Public Species
no
X
X
X
X
(species that have not gone through final review)
Generate Spreadsheet Summaries of Abundance and Life History
Data for Non-Public Species
no
X
X
X
X
Generate PDF Profiles for Multiple Public
& Non-Public Species
no
X
X
X
X
Test Tools
no
X
X
X
X
Submit Issues
no
X
X
X
X
Enter Abundance and Trait Data
no
X
X
X
Export Summaries from the Biotic Matrix
X
X
X
X
X
Modify Data in Biotic Matrix
no
X
X
X
View Results from Default Risk Assessments
X
X
X
X
X
Modify Inputs into Risk Assessments
no
X
X
X
X
Run New Risk Assessments
no
X
X
X
X
Create Master Records for Abundance
no
no
X
X
X
Site Management: Modify or Approve User Accounts
no
no
no
no
X
220
-------
Privileges
Public
Manager
Expert
Gatekeeper
Administrator
Site Management: Manage Species
(Delete, Combine, or Switch Species with Synonym)
no
no
no
X
X
Site Management: View User Statistics
no
no
no
no
X
Site Management: Data Imports: Species Name Check
X
X
X
X
X
Site Management: Data Imports: Import PDFs, Link PDFs, Delete
Links
no
no
no
X
X
Site Management: Output Data Reports
(who entered specific data and when)
no
no
no
X
X
Site Management: Approve Species Public Viewing
no
no
no
no
X
221
-------
Appendix B.
Outputting Risk Assessment Results
As described in this section, all users of CBRAT are able to download detailed results from the
last risk assessment. Additionally, users with an access level of "manager" or above (see Section
A-3) can modify biotic input data and conduct new risk assessments on all stressors (Section B-
1) or individual climate stressors (Section B-2). Users are referred to the CBRAT User's Guide
(Lee et al., 2015) for an overview of CBRAT.
B-1 Vulnerability Summary Output
The Output Vulnerability Summary page is available as part of the Risk Analysis tab in CBRAT
(Figure B-1). Clicking on "Generate Results" generates a screen output of the risks, from which
it is possible to output a CSV or XLS file. These output files list all the species within the chosen
taxon, location, and depth range with all their risk and resilience factors listed by ecoregion
(Figure B-2 and Figure B-3). The XLS file has color coded risks but it is easier to manipulate the
numerical values and perform mathematical operations (e.g., summation of risks) with the CSV
file.
The "Vulnerability" column in the output (Figure B-2) is the overall climate risk for the species
within the ecoregion based on the greatest risk (lowest number) for all climate risks, using the
temperature-adjusted ocean acidification risk for pH and aragonite saturation, and the climate-
adjusted baseline/status risk. As detailed in Section 2.3, we contend that the single greatest risk is
the most defensible approach to assigning overall climate vulnerability. However, with the CSV
output, users can explore other approaches to setting overall vulnerability such as basing it on the
number of high and moderate risks. Additionally, with basic spreadsheet manipulations, users
can remove any particular risk factor to evaluate its importance.
The vulnerability summary output lists the risks generated from the last climate scenario
analyzed. If any of the abundance classifications have been changed since the last risk analysis, it
is necessary to click the "Make Abundance Reports", which updates the abundance
classifications. This update takes approximately 15 minutes with the crabs. Then click on the
"Run Vulnerability Summaries" to update the risks, which also takes approximately 15 minutes.
If none of the abundances have been changed but there have been changes to any of the biotic
traits, effects thresholds, or climate input values it is only necessary to click on the "Run
Vulnerability Summaries", which also takes about 15 minutes. It is not necessary to click on
either of these if there haven't been any changes to abundances, biotic traits, effects thresholds,
or climate input values.
222
-------
After updating as needed, click on "Generate Results". To avoid the possibility of one user
changing values while another user runs a risk assessment, CBRAT does not allow more than
one user to run a risk assessment at a time. Further, it is not possible to change abundances,
biotic traits, or climate values while a risk calculation is underway.
CBRAT - Coastal Biodiversity Risk Analysis Tool
Home Search Risk Analysis Data Export Documents Site Management
Output Vulnerability Summary
Vulnerability Summary information is updated nightly (04:00am (PDT)).
Make Abundance Reports Run Vulnerability Summaries
Taxonomic Ranks: Infraorder
Values: | Brachyura
Locations:
Single Ecoregion * Ecoregion Group
Groups:
N.E. Pacific & U.S. Arctic
< Filter to Benthic (0-200m) values
I Generate Results Reset
Additional values to filter by:
~ Vulnerabilities
Figure B-1. Output Vulnerability Summary screen.
The Output Vulnerability Screen is accessed via the Risk Analysis tab. Choose the
Taxonomic Rank and corresponding taxonomic Value. Then choose a single
ecoregion or all the species in the NEP and U.S. Arctic. The default is to output
results for benthic species that occur from 0 to 200 m; unclicking the filter box will
output species at all depths. Clicking on Generate Results will output a screen with
the risks, from which a CSV or XLS file can be generated (Figure B-2 and Figure
B-3) Reset clears the input values on the page. Click on Make Abundance Reports
if any abundance classifications have been changed followed by the Run
Vulnerability Summaries. Click on the Run Vulnerability Summaries if any biotic
traits, thresholds, or climate values have been changed. The Vulnerabilities
checkbox allows outputs of individual risk or resilience factors (Figure B-4).
The Vulnerabilities checkbox at the bottom of the page allows users to output only species that
have a specific risk or resilience factor (non-null, including 0 values) associated with it (Figure
B-4). This function is useful for checking results for a specific stressor and to find species
missing a particular risk because of missing data. The major difference between this function and
outputting individual climate risks (Section B-2) is that the vulnerability checkbox under the
Vulnerability Summary still calculates all climate and baseline/status risks and so is slower than
outputting individual risks.
223
-------
B-2 Outputting Individual Climate Risks
Risks associated with an individual climate driver can be analyzed separately using the "Test"
function under the Temperature Increases, Ocean Acidification, and Sea Level Rise Risks links,
which are under the Risk Analysis tab. The test screen for sea level rise is shown in Figure B-5.
As with the vulnerability summary, the "Make Abundance Report" is clicked if abundance
classifications have been changed. The "ReCalc SLR Values" is clicked to update the risks
following the abundance update or if any of the biotic traits, thresholds or climate values have
been changed. A portion of an output for the ETW SST risks is illustrated in Figure B-6. The test
function is considerably faster than the vulnerability summary since only one family of risks is
calculated. The increase in speed is particularly useful when conducting scenario modelling on a
particular climate stressor and during quality assurance checks.
224
-------
Species
Family
location
Taxa-
code
Vulner-
ability
Abundance
Class
Sea
Level
Rise
Sea Surface
Temperature
Annual
Upper Mean
Sea Surface
Temperature
Summer
increase
Sea Surface
Temperature
Winter
Increase
Within
Ecoregion
SST Risk
air_summer
temp
air_winter
temp
air_annual
temp
depth_100
temp
depth_30
temp
surface
temp
annual
_ph
summer
Ph
winter aragon-
ph ite
Greatest
Temperature
Risk
Greatest
OA Risk
Temperature
Adjusted OA
Risk
Herbstia
parvrfrons
Epfaltidae
Southern
California
Bight
DEC
-2
tow
Moderate
-2
0
0
0
¦1
0
0
0
0
0
0
0
•1
0
0
-1
-1
Heteractaea
lunata
Xanthidae
Southern
California
Bight
DEC
-2
Very Rare
-2
0
0
0
-2
0
0
0
0
0
0
-1
0
0
-1
-1
Hexapanopeus
rubicundus
Southern
Fanopeidae
California
Bight
DEC
¦1
Very Rare
0
Q
Q
0
•2
0
0
0
0
0
0
-1
0
0
-1
-1
Hoplocypode
occidentals
Ocypodidae
Southern
California
Bight
Southern
DEC
-1
Very Rare
-1
0
0
0
-2
0
0
0
0
0
-1
0
0
-1
-1
inachoides iaevis
Inachoididae
California
Bight
DEC
-1
Rare
-i
0
0
0
-3
Q
0
0
0
0
0
0
-1
0
0
-1
-1
Latulambrus
occidentals
Parthenopldae
Southern
California
Bight
Southern
DEC
-1
Nigh
Moderate
-1
0
0
0
-2
0
0
0
0
0
0
0
-1
0
0
-1
-1
Ubinia setosa
Epialtidae
California
Bight
DEC
-1
Rare
0
0
0
0
-2
0
0
0
0
-1
0
0
-1
-1
Lophopanopeus
helius
Panopeidae
Southern
California
Bight
DEC
-3
Rare
-2
-3
-2
-2
-3
-2
0
-2
-1
-2
-2
0
-1
0
-3
-1
-1
Lophopanopeus
diegensis
Fanopeidae
Southern
California
Bight
DEC
-3
low
Moderate
-2
-3
-2
-2
-3
-2
0
-2
-2
-2
-2
0
-1
0
-3
-1
-1
Figure B-2. Portion of Vulnerability Summary CSV - Output for climate risks.
Each species is listed by each ecoregion. "Vulnerability" is the overall climate risk calculated as the single greatest risk from the sea level rise risk,
greatest temperature risk, temperature-adjusted ocean acidification risk, and the climate-adjusted baseline risk (see Figure B-3). The "Within
Ecoregion SST Risk" is a worst-case scenario and is not included in calculating the overall risky. Climate risks are classified from minor (0) to high
(-3). Aragonite is null in this case because pH and not aragonite saturation state was chosen as the major stressor for decapods. These are test
data and do not represent the final risk assessment for these species.
225
-------
Species
Family
Location
lode
Vulner-
Abundance
Class
Greatest
Climate
Risk
Greatest
Baseline
Risk
Climate
Adjusted
Baseline
Risk
Endemic
Restricted
Distrib.
Wide
Distribution
Arctic
Endemic
Small
Island
Distrib.
Nonin-
digenous
Species
Rare
Rare
Every-
where
Abundant
Some-
where
Population
S.Rare-N.
Mod/Abun.
Transient
Symbiotic
Specializat
ion
Habitat
Specializa-
tion
Specializa-
tion
Anadromous /
Catadromous
Slow
Repro. /
Long Lived
Herbstia
parvifrons
Epialtidae
Southern
California
Bight
DEC
-2
Moderate
-2
0
2
Heteractaea
lunata
Xanthidae
Southern
California
Bight
DEC
-2
Very Rare
-2
-3
-3
2
-2
-3
Hexapanopeus
rubicundus
Panopeidae
Southern
California
Bight
DEC
Very Rare
-1
0
-1
Hoplocypode
occidentalis
Ocypodidae
Southern
California
Bight
DEC
Very Rare
0
2
Inachoides laevis
Inachoididae
Southern
California
Bight
DEC
Rare
0
Latulambrus
occidentalis
Parthenopidae
Southern
California
Bight
DEC
High
Moderate
0
2
Libinia setosa
Epialtidae
Southern
California
Bight
DEC
Rare
-1
0
-1
Lophopanopeus
bellus
Panopeidae
Southern
California
Bight
DEC
-3
Rare
-3
0
Lophopanopeus
diegensis
Panopeidae
Southern
California
Bight
DEC
-3
Low
Moderate
-3
-1
-1
-1
Figure B-3. Portion of Vulnerability Summary CSV - Output for baseline/status risks.
Each species is listed by each ecoregion. Both risks, classified from 0 to -3, and resilience factors, classified from 1 to 3, are output for the
baseline/status traits. Null values indicate either that the risk is minor (0) or that there is missing data to calculate the risk. The overall value of the
baseline/status risks are weighted for each species by the extent of climate risk, given in the "Greatest Climate Risk" column, with the "Climate
Adjusted Baseline Risk" used in calculating the overall risk. These are test data and do not represent the final risk assessment for these species.
226
-------
CBRAT - Coastal Biodiversity Risk Analysis Tool
Home Search Risk Analysis Data Export Documents Site Management
Output Vulnerability Summary
Vulnerability Summary information is updated nightly (04:00am (PDT)),
| Make Abundance Reports | j Run Vulnerability Summaries |
Taxonomic Ranks: I Infraorder w ^
„ , „ Locations: Single Ecoregion " Ecoregion Group
Values: Brachyura ~ - —
Groups: • N.E. Pacific & U.S. Arctic
* Filter to Benthic (0-200m) values
| Generate Results j | Reset
Additional values to filter by:
@ Vulnerabilities
Only include Species with Vulnerability values of:
Select all:
Abundant Somewhere
Anadromous /
Catadromous
Arctic Endemic
C Endemic
Habitat Specialization
Hyper-Rare
Nonindigenous Species
Ocean Acidification
Declines
Population Decline
Rare Everywhere
Restricted Distrib.
Sea Level Rise
* Sea Surface Temperature
Summer Increase
Sea Surface Temperature
Winter Increase
Sea Surface Temperature
Yearly Upper Mean
Slow Repro. / Long Lived
Small Island Distrib.
S.Rare-N.Mod/Abun.
Symbiotic Specialization
Air and Subsurface
Temperature
Transient
Trophic Specialization
Wide Distribution
Figure B-4. Outputting vulnerability summary limited to species with a specific
risk or resilience factor.
Clicking on "Vulnerabilities" displays a list of the individual risk and resilience
factors. Choose a single factor and click on "Generate Results". This will
generate a vulnerability summary (Figure B-2 and Figure B-3) but only for
species with a non-null value for the chosen factor. In this case, only species
with a non-null risk for sea level rise are output.
227
-------
CBRAT - Coastal Biodiversity Risk Analysis Tool
Home Search Risk Analysis Data Export Documents Site Management About Terms of Use
Habitat Thresholds
Sea Level Rise Risks
Sea Level Rise Input Variables
Taxonomic Ranks: | Infraorder ~
Values:| Brae hy ura
Groups:
-------
species_id
species
taxa_cocfe
meow_class
increment
incremental
increase
projected_
increase
sst_range
2343
Pugettia gracilis
DEC
Gulf of Alaska
SUMMER
3.53
15,35
minor
2343
Pugettia gracilis
DEC
Aleutian Islands
SUMMER
3.63
12.07
minor
2343
Pugettia gracilis
DEC
North American Pacific
Fjordland
SUMMER
3.18
16,4
low
2343
Pugettia gracilis
DEC
Puget Trough/Georgia Basin
SUMMER
3,12
16,71
low
2343
Pugettia gracilis
DEC
Oregon, Washington,
Vancouver Coast and Shelf
SUMMER
2.9
17,02
moderate
2343
Pugettia gracilis
DEC
Northern California
SUMMER
2,83
18,02
high
2343
Pugettia gracilis
DEC
Puget Trough/Georgia Basin
WINTER
1.8
9,31
minor
2343
Pugettia gracilis
DEC
Northern California
WINTER
2.34
14.98
moderate
2343
Pugettia gracilis
DEC
Oregon, Washington,
Vancouver Coast and Shelf
WINTER
2.41
11,75
minor
2343
Pugettia gracilis
DEC
Aleutian Islands
WINTER
2,53
6,21
minor
2343
Pugettia gracilis
DEC
Gulf of Alaska
WINTER
2.79
6.91
minor
2.343
Pugettia gracilis
DEC
North American Pacific
Fjordland
WINTER
2.53
9.21
minor
2343
Pugettia gracilis
DEC
Oregon, Washington,
Vancouver Coast and Shelf
ANNUAL
2.62
14,13
minor
2343
Pugettia gracilis
DEC
Northern California
ANNUAL
2.54
16.09
high
2343
Pugettia gracilis
DEC
Aleutian Islands
ANNUAL
3,03
8,7
minor
2343
Pugettia gracilis
DEC
Gulf of Alaska
ANNUAL
3.1
10,52
minor
2343
Pugettia gracilis
DEC
North American Pacific
Fjordland
ANNUAL
2.8
12,27
minor
2343
Pugettia gracilis
DEC
Puget Trough/Georgia Basin
ANNUAL
2.15
12,59
minor
Figure B-6. Portion of the output from an individual climate risk output.
Example for ETW temperature risks for the crab Pugettia gracilis. Full output includes each species listed
by ecoregion. For ETW, the risk level is listed under "sst_range". The "incrementaljncrease" is the future
increase in temperature while the "projectedjncrease" is the projected future temperature. The variables
(columns) in the output are specific to each climate stressor. The pivot table function in spreadsheets can
be used to organize the data. These are test data and do not represent the final risk assessment for these
species.
B-3 Outputting Results for Northern Colonization
The Northern Colonization Test function (Figure B-7), located under the "Temperature
Increases" page in CBRAT, outputs an analysis of whether currently unoccupied northern
(cooler) ecoregions will become sufficiently warm to allow colonization (Section 5.5). The
suitability of temperatures in northern ecoregions is determined using the BTL approach (Section
5.4). A portion of an output is shown in Figure B-8.
229
-------
CBRAT - Coastal Biodiversity Risk Analysis Tool
Home Search Risk Analysis Data Export Documents Site Management About Terms of Use
Predicted SST Increases
SST Thresholds
SST Test
Air and Subsurface
Increases
Air and Subsurface Test
Make Abundance Reports
Northern Colonization values
Taxonomic Ranks: Infraorder
Values: Brachy
Temperature Increases
Northern Colonization
Locations: Single Ecoregion
Ecoregion Group
Groups: • N.E. Pacific & U.S. Arctic
< Filter to Benthic (0-200ni) values
Export CSV
Figure B-7. Northern Colonization Test screen.
Screen to evaluate suitability of temperatures in unoccupied northern ecoregions using the BTL approach.
Choose the Taxonomic Rank and corresponding Value. Then choose a single ecoregion or the NEP and
U.S. Arctic. The default is to output results for benthic species that occur from 0 to 200 m; unclicking the
filter box will output all species at all depths. Click on Make Abundance Reports if abundance values have
been changed, and then the ReCalc SLR Values to update the risks. Click on the ReCalc button if any
biotic traits, thresholds, or climate values have been changed. Clicking on Export CSV will generate a
CSV file with an analysis of the temperature suitability of all species within the chosen taxon in currently
unoccupied northern (cooler) ecoregions. An example output is shown in Figure B-8.
species_id
species
taxa_code
eco_region
depth_class
suitability
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
air_summer
high
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
a ir_ winter
moderate
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
air_yearly
high
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
depth_100
low
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
depth_30
minor
121
Metacarcinus magister
DEC
Beaufort Sea - continental coast and shelf
sst
moderate
121
Metacarcinus magister
DEC
Chukchi Sea
air_summer
high
121
Metacarcinus magister
DEC
Chukchi Sea
air_winter
moderate
121
Metacarcinus magister
DEC
Chukchi Sea
air_vearly
high
121
Metacarcinus magister
DEC
Chukchi Sea
depth_100
minor
121
Metacarcinus magister
DEC
Chukchi Sea
depth_30
minor
121
Metacarcinus magister
DEC
Chukchi Sea
sst
moderate
Figure B-8. Portion of the output from Northern Colonization Test.
The temperature suitability for northern colonization, as calculated using the BTL approach, is listed for all
species for all northern (cooler) unoccupied ecoregions. The pivot table function in spreadsheets can be
used to organize the data. Suitability is classified from minor (not suitable) to high (high suitability). These
are test data and do not represent the final risk assessment for these species.
230
-------
Appendix C.
Near-Coastal Habitat Areas and GIS
Metadata
C-1 Introduction
Over the last decade there has been a dramatic increase in the availability of georeferenced
marine/estuarine landscape data for California, Oregon, and Washington. These GIS layers have
been generated for a variety of purposes, including evaluating essential fish habitat, assessing
potentially threatened habitat types (e.g., marshes, seagrasses, and kelps), and for assessing
tsunami risks. We synthesized a number of these layers to generate estimates of major offshore
and estuarine habitats for the Southern California Bight, Northern California, and Oregon,
Washington, Vancouver Coast, and Puget Trough/Georgia Basin ecoregions. These areal
estimates were then used as inputs in assigning relative abundance of species in these four
ecoregions (Section 3) as well as a guide to the relative areas of major habitats in other
ecoregions.
C-2 Near-Coastal Habitat Areas - Patterns of Offshore and Estuarine Habitats
Offshore habitats were split into those that occur from the shoreline to 30 m deep and those from
>30 m to 200 m depth, which correspond to our shallow and deep subtidal depth classes. As our
primary concern was evaluating broadly across habitat patterns, these two depth classes were
combined for the current analyses. The resolution of the offshore data varied among ecoregions,
with the greatest detail off the coasts of Oregon and Washington. At this time, we do not have
areal estimates for coastal beaches, while the rocky intertidal was analyzed separately. The
estuarine data were primarily derived from the National Wetland Inventory (NWI; USFWS,
2009; httj) ://www.fws.gov/wetlands/) with additional sources for seagrass layers. Total estuary
areas included unvegetated sediments, emergent marsh, submerged aquatic vegetation (SAV),
tidal riverine areas, marine areas at the mouth of estuaries, hard substrates, and woody
vegetation, which is the definition of estuary area used in Lee and Brown (2009). Offshore and
estuarine data were available only for the United States so that the areas for the Puget
Trough/Georgia Basin Ecoregion do not include Canada while the Southern California Bight
Ecoregion areas do not include Mexico.
The key points relating to habitat area at an ecoregion scale are:
• Estuarine area is several-fold smaller in the Southern California Bight Ecoregion than in
the other three ecoregions (Figure C-1). Inclusion of the Mexican portion of the
231
-------
ecoregion would increase the estuarine area (e.g., Bahia San Quintin), though it is likely
that the total estuarine area would still be substantially less than in the other three
ecoregions. Puget Trough/Georgia Basin Ecoregion had the second smallest estuarine
area; inclusion of Canada (e.g., Fraser River) would likely increase the area to be more
comparable to Northern California and Oregon, Washington, Vancouver Coast and Shelf
ecoregions.
• Intertidal and subtidal unconsolidated habitats combined constitute the major estuarine
habitat types in all ecoregions (Figure C-2). The relative contribution of these
unconsolidated habitats was smallest in the Southern California Bight Ecoregion. The
area of estuarine hard substrates (not shown) was trivial in all ecoregions.
• The greatest area of emergent marshes occurs in Northern California and the smallest in
the Southern California Bight (Figure C-2). Inclusion of Mexico would increase the area
of marsh in the Southern California Bight, though it is likely that it would still be smaller
than the other ecoregions because of the relatively smaller estuarine area in Southern
California and northern Baja.
• Areas of intertidal and subtidal submerged aquatic vegetation (SAV) are substantially
less than the areas of unvegetated unconsolidated sediments in the three northern
ecoregions while more similar in the Southern California Bight Ecoregion (Figure C-2).
Zostera marina is relatively abundant in Puget Sound, in which 200 km2 (20,000
hectares) have been estimated (Mumford, 2007), though this is substantially smaller than
the unconsolidated habitats in Puget Sound.
• Estuarine unconsolidated habitats constitute a small fraction (1% - 3.1%) of the area of
the offshore unconsolidated habitat (Figure C-3).
• Across the three ecoregions with offshore area estimates, unconsolidated habitat
constitutes the greatest area, with the Southern California Bight showing a reduced
percentage compared to the other ecoregions (Figure C-4 through Figure C-6).
• Rocks/boulders and rocks mixed with other substrate types occupied a moderate offshore
area in Northern California and the Oregon, Washington, Vancouver Coast and Shelf
ecoregions (Figure C-5 and Figure C-6). Rocks occupied both a greater absolute area and
a greater proportion of the offshore area in the Southern California Bight Ecoregion
(Figure C-4).
• Kelp occupied a small percentage of the area (0.09% - 0.6%) in all three ecoregions with
offshore data layers (Figure C-4 through Figure C-6). The area of floating kelp ranged
from approximately 1900 to 4700 hectares in the Strait of Juan de Fuca, one of the
232
-------
primary locations for floating kelp in Puget Sound (Puget Sound Action Team, 2007),
representing a small percentage of unconsolidated habitats in Puget Sound.
• In the more detailed analysis off Oregon and Washington, cobble/gravel and shell
substrates composed a small percentage of the total offshore area (Figure C-l).
200
Southern CaliforuialSorttaem California Onegouiau Puget
Figure C-1. Total estuarine area in the Southern California Bight, Northern California, Oregon,
Washington, Vancouver Coast and Shelf, and Puget Trough/Georgia Basin ecoregions.
Values for Southern California do not include the Mexican portion of the ecoregion while values for Puget
do not include the Canadian portion of the ecoregion. Estuarine areas are derived from the NWI.
233
-------
~ S. CA
N. CA
~ OR
~ Puget
-------
10,000,000
1,000,000
100,000
10,000
1,000
100
10
1
Estuary ¦ Offshore
96,9%
98.9° o
Southern California
Northern California
Oregon
Figure C-3. Areas of offshore versus total estuarine unconsolidated habitats by ecoregion.
Most of the substrate type around islands presumably consists of rock/boulders. Values for Southern
California do not include the Mexican portion of the ecoregion.
1000000
100000
10000
W
2 iooo
(O
+->
o
u 100
6&5%
16° o
Unconsolidated Rock/Boulders
Kelp
Islands
ND
J
Figure C-4. Area of major habitat types from 0-200 m offshore in the Southern California Bight Ecoregion.
Most of the substrate type around islands presumably consists of rock/boulders. ND = no data. Values for
Southern California do not include the Mexican portion of the ecoregion.
235
-------
1000000
100000
yj 10000
o
iS iooo
o
-------
C-3 Near-Coastal Habitat Areas - Geospatial Analysis
The following summarizes the data sources and geospatial methodologies used to estimate near-
coastal habitat areas.
C-3.1 Deepwater (Offshore) Marine Habitat Data Compilation Method
Geographic Information Systems (GIS) software ArcMap v. 10.1 tools were used to overlay
geospatial data including surficial geological habitats (Romsos et al, 2007), canopy kelp forests
(PSMFC, 2004), MEOW marine ecoregions (Spalding et al, 2007), and bathymetric depths
(PFMC, 2004). These data were then cross-tabulated and summarized in a Microsoft Excel™
pivot table.
Sources: The surficial geological habitat geospatial data layer was downloaded from the Pacific
Coast Ocean Observing System (PaCOOS) West Coast Habitat Server maintained by Oregon
State University's Active Tectonics & Seafloor Mapping Lab. The canopy kelp forest geospatial
data were downloaded from the Oregon Ocean Information website maintained by the State of
Oregon to support marine spatial planning in the Oregon Territorial Sea. The marine ecoregion
geospatial data were downloaded from The Nature Conservancy's TNCMAPS website that
provides The Nature Conservancy's core conservation datasets. Bathymetric depth geospatial
data were downloaded from the Pacific Fishery Management Council's Pacific Coast Marine
Habitat Information website. National Wetland Inventory (NWI) geospatial data downloaded
from the U.S. Fish & Wildlife Survey's National Wetland Inventory (NWI) website in 2011 were
also used in the process (Cowardin et al., 1979).
GIS Overlay and Extraction: The geospatial data layers above were downloaded in a variety of
cartographic projections. All were projected into the Albers projection for compatibility in the
overlay process. These data were all in the ESRI shapefile vector polygon format. The
bathymetric data consisted of polygons representing 10-meter depth zones. The bathymetric data
consisted of five shapefiles, one each for Oregon and Washington waters and three for California
waters. The five layers were merged and polygons less than or equal to 200 m were extracted.
The "identity" overlay tool was used to simultaneously overlay and extract surficial geologic
habitat polygons less than or equal to 200 m deep. The specific vertical datum was not provided
but it is assumed to be mean sea level. The "identity" overlay tool was again used to
simultaneously overlay and extract marine ecoregion polygons. Canopy kelp polygons were
similarly overlain. The final GIS overlay procedure was to use NWI polygons to erase estuarine
polygons from the compiled shapefile in order to avoid duplication with a parallel West Coast
estuarine habitat data compilation effort. The last step in the geospatial data compilation was to
calculate the area of each resulting polygon in hectares.
237
-------
C-3.2 Nearshore Marine, Estuarine and Tidal Riverine Habitat Data Compilation
Method
Geographic Information Systems (GIS) software ArcMap v. 10.1 tools were used to overlay
geospatial data including NWI habitats (Cowardin et al. 1979), seagrass habitat (PFMC, 2005),
and marine ecoregions (Spalding et al, 2007). These data were then cross-tabulated and
summarized in a Microsoft Excel™ pivot table.
Sources: Marine, estuarine and tidal riverine NWI polygons for Washington, Oregon and
California were downloaded from the U.S. Fish & Wildlife Survey's NWI website in 2011 and
compiled as part of a West Coast estuarine classification study (Lee and Brown, 2009). Seagrass
habitat geospatial data were downloaded from the Pacific Fishery Management Council's
(PFMC) Pacific Coast Marine Habitat Information website. These data were compiled in support
of an Environmental Impact Statement (EIS) to consider the designation and conservation of
Essential Fish Habitat (EFH) for Pacific Coast Groundfish. The marine ecoregion geospatial data
were downloaded from The Nature Conservancy's TNCMAPS website described above.
GIS Overlay and Extraction: The wetlands and deepwater habitats classification codes termed
"Attributes" by the NWI are alpha-numeric codes that provide detailed habitat descriptions for
polygon areas in the shapefile. The codes, developed by Cowardin et al. (1979), represent a
complex hierarchical classification of ecological taxa. Only three of the highest levels of
classification were selected and extracted; marine, estuarine and riverine. Both subtidal and tidal
polygons were used for the marine and estuarine systems, however for the riverine system only
the tidal riverine subsystem polygons were used. The next levels of classification, classes and
subclasses, are based on substrate material, flooding regime, or vegetation class. Special
modifiers are also used. In order to enhance the usefulness of these data, a table was constructed
and joined to the geospatial data that parses the NWI codes into 101 human-readable classes. An
online NWI classification decoder tool was used to aid the construction of the join table. Because
seagrasses or submerged aquatic vegetation may not be consistently mapped due the limitations
of aerial imagery as the primary data source used to detect wetlands in the NWI mapping process
(USFWS 2004), the PFMC seagrass shapefile was incorporated using the 'identity' overlay tool.
The 'identity' overlay tool was again used to simultaneously overlay and clip marine ecoregion
polygons to the extent of the extracted NWI polygons. The last step in the geospatial data
compilation was to calculate the area of each resulting polygon in hectares.
238
-------
C-3.3 GIS Data Links for Geospatial Analysis of Near-Coastal Habitats (Accessed
08/09-14/2013)
• Pacific Coast Ocean Observing System/ West Coast Habitat Server
http://pacoos.coas.oregonstate.edu/index.htm
• Oregon Ocean Information
http://www.oregonocean.info/index.php/ocean-data-and-resources
• Pacific Coast Marine Habitat Information http://marinehabitat.psmfc.org/
• U.S.FWS National Wetlands Inventory http://www.fws.gov/wetlands/Data/Data-
Download.html
• NWI online Classification Code Decoder
https://www.fws.gov/wetlands/Data/Wetland-Codes.html
C-4 Calculation and Metadata for Computing Habitat Thresholds for West Coat
Intertidal Rocky Habitats due to Sea Level Rise using LiDAR Topobathy
GIS Methodology:
Within each ecoregion, topobathy LIDAR digital elevation models (DEMs) downloaded from
NOAA's Digital Coast GIS data repository (Dept. Commerce, 2016) were used to estimate the
percent change in the area of rocky intertidal habitat in 10 cm increments with different levels of
eustatic sea level rise. Puget Sound topobathy LIDAR was released concurrent to this study and
required conversion to digital elevation model (DEM) from LAZ compressed LIDAR point
format (Isenburg, 2011). Environmental Sensitivity Index (ESI) Map Shoreline data were used to
identify rocky shorelines (Dept. Commerce, 2013). Such stretches of shoreline were extracted for
each of the four ecoregions and buffered by 100 m to include the intertidal and evaluate the
potential area for upland habitat migration. All available LIDAR topobathy DEMs from Digital
Coast were extracted using the resulting polygons and two rasters were synthesized from the
results, a 10 cm increment zone raster and a non-planimetric surface area raster (Jenness, 2004)
for zonal summation. Current rocky intertidal non-planimetric surface areas for each ecoregion
were computed between Mean Higher High Water (MHHW) and Mean Lower Low Water
(MLLW) levels established from published datum sheets for tidal stations central to each
MEOW ecoregion (Gill and Schultz, 2000). Percent change in non-planimetric surface area for
the same relative ranges were calculated in 10 cm incremental steps of eustatic SLR from the
zonal summation.
The sources and steps for each parameter are given below.
239
-------
Marine Ecoregions of the World (MEOW):
fattp ://www.marinereeion s. ore/ sources. php#meow
Marine Ecoregions of the World polygon shapefiles were downloaded. The four MEOW
ecoregions on the US west coast were selected in ArcGIS 10.2.2 and exported to new
shapefiles. They are: Puget Trough/Georgia Basin; Oregon, Washington, Vancouver
Coast and Shelf; Northern California and; Southern California Bight.
Environmental Sensitivity Index (ESI) Map Shoreline data:
http://response.restoration.noaa.eov/maps-and-spatial-data/environmental-sensitivity-
ESI shoreline data were downloaded. All shorelines with the term "Rocky" in their
attributes were selected in ArcGIS 10.2.2 and exported to a new shapefile. The rocky
shoreline was then buffered by 100 m using the ArcGIS 10.2.2 Analysis tool 'Buffer' to
create a polygon shapefile. This shapefile then clipped by each ecoregion using the
ArcGIS 10.2.2 Analysis tool 'Clip'.
Coastal Topobathy Lidar (JALBTCX) Digital Elevation Models:
https://cGast.noaa.gov/datareeistry/search/dataset/C10406A4~FB7D~4D30~96D7~
E036F6942EB6
Topobathy LiDAR digital elevation models were downloaded from NOAA's Digital
Coast GIS data repository and mosaicked by ecoregion. Sub-dataset rasters were
extracted by the buffered rocky shoreline for the ecoregions using the using the ArcGIS
10.2.2 Spatial Analyst tool 'Extract by Mask'. A second raster was derived from these
data using the ArcGIS 10.2.2 Spatial Analyst tool 'Raster Calculator' by multiplying z-
values by ten and integerizing to serve as 10 cm elevation "zones". A third raster of non-
planimetric surface area was derived from these data by using the surface area tool in the
DEM Surface Tools v. 2.1.375 created by Jenness Enterprises.
http://wwwjenn.essent.com/arceis/arcgis extensions.htm
Non-planimetric surface areas were summarized in tabular format for each 10 cm
elevation "zone" using the ArcGIS 10.2.2 Spatial Analyst tool 'Zonal Statistics as Table'.
Puget Sound topobathy LiDAR was downloaded in .laz compressed LiDAR point format.
These data were uncompressed into xyz ASCII format using the open source LASzip tool
developed by Martin Isenburg.
https://rapidlasso.com/
These data were opened in ArcMap 10.2 and displayed as point events. The point events
were turned into shapefiles using the ArcGIS 10.2.2 Analysis tool Clip using the buffered
rocky shoreline and merged using the ArcGIS 10.2.2 Data Management tool 'Merge'.
These data were then converted to raster via the ArcGIS 10.2.2 Spatial Analyst tool
interpolation tool 'IDW' with the three nearest points at a maximum variable distance of
30 meters. Raster calculation and surface area procedures described above were then
applied.
240
-------
Tidal Datums:
Mean Higher High Water (MHHW) and Mean Lower Low Water (MLLW) levels
relative to NAVD88 (the LiDAR elevation datum) were computed from published datum
sheets for tide stations representative of each MEOW ecoregion. The stations selected
are: Santa Monica, CA (9410840) for Southern California Bight; Point Reyes, CA
(9415020) for Northern California; South Beach, OR (9435380) for Oregon, Washington,
Vancouver Coast and Shelf and; Port Townsend, WA (9444900) for the Puget
Trough/Georgia Basin.
https://tidesandciirrents.noaa.eov/stations.html
Habitat Threshold Calculation:
Tabular data of the sum of non-planimetric surface area for each 10 cm elevation 'zone'
was opened in an Excel spreadsheet. A new column was added and the zone number was
multiplied by 10. This step restores elevation relative to NAVD88. A second sheet was
added with the first column labeled MSL (mean sea-level rise). The second column is
labeled 'Risk' and is formatted as percentage. Tidal datum sheets are consulted and a
formula is entered as follows: =
1 -(SUM(LiDAR!E398:E423))/(SUM(LiDAR!E$398:E$423))
where 'LiDAR!' is the summary sheet 'E398' is the sum of the non-planimetric surface
area at MLLW and 'E423' is the sum of the non-planimetric surface area at MHHW. This
provides the percentage of the current non-planimetric surface area of the intertidal for
each 0.1 m rise in sea level. Habitat Thresholds were assigned as follows: 0 - 10% loss =
Minor; 11 - 29% loss = Low; 30 - 49% loss = Moderate; >50% loss = High.
Results for percent habitat loss of rocky intertidal by ecoregion:
Figure C-7 shows the results of the GIS model by each ecoregion.
241
-------
A)
B)
Seal Level Rise (m)
PugetUDAR PugetRisk OrWaUDAR OrWa Risk NorCal LiDAR NorCal Risk SoCal LiDAR SoCal Risk
Seal Level Rise (m) Average risk
Percent Loss
0
0.0%
0.0% Minor
0.0% Minor
0.0% Minor
Minor
0-10%
0.1
6.7%
-0.8% Minor
3.0% Minor
4.3% Minor
0.4 Low
11-29%
0.2
15.0% Low
-1.3% Minor
5.7% Minor
17.6% Low
0.8 Moderate
30-49%
0.3
22.6% Low
0.8% Minor
9.4% Minor
26.9% Low
1.4 Hign
>=50%
0.4
27.7%
Low
1.3% Minor
13.7% Low
34.2% Moderate
0.5
30.8%
Moderate
2.3% Minor
18.2% Low
39.6% Moderate
0.6
33.4%
Moderate
4.0% Minor
22.4%
Low
43.9% Moderate
C)
0.7
34.9%
Moderate
6.4% Minor
26.8%
Low
47.7% Moderate
Ecoregion
Tidal Range (m)
High
Moderate
Low
0.8
35.5%
Moderate
9.6% Minor
31.2%
Moderate
50.7%
Hign
Puget
2.60
1.4
0.5
0.2
0.9
36.3%
Moderate
13.4% Low
35.0%
Moderate
51.5%
Hign
OrWa
2.54
2.0
1.4
0.9
1
38.2%
Moderate
17.5% Low
38.4%
Moderate
52.1%
Hign
NorCal
1.79
1.5
0.8
0.4
1.1
40.7%
Moderate
21.5% Low
41.3%
Moderate
51.5%
Hign
SoCal
1.65
0.8
0.4
0.3
1.2
43.2%
Moderate
25.3% Low
44.2%
Moderate
49.2%
Hign
1.3
47.0%
Moderate
29.0%
Low
46.8%
Moderate
47.9%
Hign
1.4
50.0%
Hign
32.6%
Moderate
49.0%
Moderate
47.4%
Hign
1.5
52.4%
Hign
35.9%
Moderate
50.9%
Hign
46.9%
Hign
1.6
54.5%
Hign
38.9%
Moderate
52.6%
Hign
46.9%
Hign
1.7
56.5%
Hign
42.0%
Moderate
54.1%
Hign
47.7%
Hign
1.8
57.9%
Hign
44.9%
Moderate
55.5%
Hign
48.6%
Hign
1.9
59.3%
Hign
47.7%
Moderate
56.7%
Hign
49.4%
Hign
2
60.8%
Hign
50.1%
Hign
57.9%
Hign
50.1%
Hign
2.1
61.8%
Hign
52.2%
Hign
59.1%
Hign
50.8%
Hign
2.2
62.6%
Hign
54.1%
Hign
60.2%
Hign
51.5%
Hign
2.3
63.3%
Hign
55.9%
Hign
61.3%
Hign
52.8%
Hign
2.4
64.1%
Hign
57.4%
Hign
62.2%
Hign
53.3%
Hign
2.5
64.7%
Hign
59.0%
Hign
63.2%
Hign
53.8%
Hign
2.6
65.3%
Hign
60.5%
Hign
64.1%
Hign
54.8%
Hign
2.7
65.7%
Hign
62.0%
Hign
65.0%
Hign
57.8%
Hign
2.8
66.0%
Hign
63.5%
Hign
65.9%
Hign
60.5%
Hign
2.9
66.3%
Hign
65.0%
Hign
66.7%
Hign
64.1%
Hign
3
66.3%
Hign
66.6%
Hign
67.5%
Hign
69.1%
Hign
Figure C-7. Calculation of habitat thresholds by ecoregion for rocky intertidal habitats.
In Part A), the first column lists the average sea level rise for US CONUS West Coast while the columns
labeled LIDAR are the predicted percent habitat loss by SLR within each ecoregion. The color coded
columns are the habitat threshold class for each SLR value. Part B) shows the percent habitat losses
used to define each habitat threshold class. Part C) first lists the tide range within each ecoregion and
then the SLR values for each habitat threshold class calculated by each ecoregion.
242
-------
Appendix D.
Evaluation of Temperature as Determinant
for Warm-Edge Range Limits of Marine
Species
The logic to predict the species' risk to increased temperatures inherently assumes that species
do not exist in southern (warmer) unoccupied ecoregions because they are too warm (Section 5).
It is known that factors other than temperature can affect species' range limits (Gaston, 2003),
including food supply, interspecific competition, and interactions between biotic and abiotic
variables (e.g., Helmuth et al., 2006; Sexton et al., 2009; Gaston, 2009). Temperature, however,
is the most important determinant of species' warm-edge range limits. Strong support for this
contention comes from Cahill et al. (2013) who reported that temperature was supported 68.8%
of the time as the factor limiting the warm-edge of distributions (i.e., southern limits in the
northern hemisphere and northern limits in the southern hemisphere).
Cahill et al.'s review included 48 marine species. To further evaluate marine species, we
reviewed studies not included in Cahill et al. (Table D-l). Based on this review, we identified
four lines of evidence that support the critical role of temperature in setting warm-edge range
limits:
• Physiological
• Range Shifts
• Impaired Fecundity/Recruitment
• Trophic Dynamic Shift
D-1 Physiological
Physiological limits to temperature affect species' distributions directly and were the most
supported proximate cause of warm-edge range limits in the Cahill et al. review. We reviewed
nine papers that found that increases in temperature affect aerobic scope, growth, and protein
synthesis/denaturation (Table D-l). Increases in water temperature reduce the capacity of water
to hold oxygen and other dissolved gases. The resulting combination of high temperature and
low oxygen concentration is very stressful to many fishes and aquatic invertebrates because high
temperatures also cause elevated metabolic rates and increased demand for oxygen (Lomolino et
243
-------
al., 1953). InMaja squinado, Zoarces viviparous, Gadus morhua, Ostorhinchus cyanosoma and
Ostorhinchus doederleini, limited circulation and ventilation at high temperatures caused
insufficient oxygen supply, thus limiting aerobic scope, thermal tolerance, and even growth
performance (Frederich and Portner, 2000; Portner and Knust, 2007; Portner et al., 2008; Nilsson
et al., 2009). Studies have shown that growth performance for Acanthochromispolyacanthus and
Cheilodactylus spectabilis declined at the warmest sea surface temperatures experienced at their
warm range boundary, respectively (Munday et al., 2008; Neuheimer et al., 2011). High
temperatures in the natural habitat of Mytilus edulis caused protein denaturation, suggesting that
the species' distribution is restricted by its thermal limit (Chappie et al., 1998).
D-2 Range Shifts
Range shifts are a key example of temperature's effect at the population level. They appear as
direct evidence, displayed by contractions of lower latitude limits, and suggestive evidence,
displayed by expansions of higher latitude limits. In marine ectotherms, species' ranges conform
closely to their limits of thermal tolerance, thus both range boundaries have been equally
responsive to warming temperatures (Sunday et al., 2012). Physiological limitations to rising
temperatures are the likely cause of such range shifts (Somero, 2012; Cahill et al., 2013).
Eleven papers reviewed in Table D-l support this line of evidence. Two-thirds of North Sea fish
species' distributions have shifted in response to increased temperatures (Perry et al., 2005). The
Atlantic cod's expansion poleward is likely due to temperature's effects on reproductive
performance and reduced food availability (Portner et al., 2008). In addition, intertidal
communities on the California coast have shifted poleward in response to elevated temperatures
(Barry et al., 1995). Studies have shown that the lower latitude range boundary of Semibalanus
balanoides has shifted poleward 350 km (Jones et al., 2012); transplant experiments and thermal
modelling revealed mortality in transplanted barnacles due to high temperatures during aerial
exposure, suggesting that temperature is driving contraction of the lower latitude range boundary
(Jones et al., 2012; Wethey and Woodin, 2010). A special case of a range shift is the northern
colonization of the southern wood borer Teredo bartschi near a thermal discharge (Hoagland and
Turner, 1980). Marine species shift at different rates because they follow local climate velocities
(Pinsky et al., 2013, also see Molinos et al., 2016).
In addition to long-term changes, extreme temperature events affect population distribution,
potentially resulting in range shifts (Wethey et al., 2011). It is clear that temperature plays a
major role in defining the southern range limits, though the pattern may be "messy" if the
number and severity of extreme events increases along with changes in mean temperatures.
244
-------
D-3 Impaired Fecundity/Recruitment
The direct effect of temperature on recruitment is also an important line of evidence. We
reviewed six papers which supported this line of evidence (Table D-l). In Atlantic cod, it has
been found that as temperatures warm, recruitment decreases in stocks inhabiting the uppermost
part of their temperature range (Planque and Fredou, 1999; Sundby, 2000). This indicates that
temperature is an important factor limiting the southern distribution of this species. The same
effect was observed in Macoma balthica in which increased temperatures resulted in a decrease
in reproductive output, recruitment, and growth, with effects detected in populations
approximately 1000 km poleward of the warm edge of the species' range (Beukema et al., 2009).
Studies of this clam have also found that rising seawater temperature affected recruitment by a
decrease in reproductive output and by spring advancement of bivalve spawning (Philippart et
al., 2003). In flatfish, temperature during gonadal maturation affects recruitment and distribution
through its influence on the rate of gonadal maturation and spawning time (Lange and Greve,
1997). There is a negative correlation between temperature and abundance of plaice, suggesting
that as temperature rises, recruitment decreases (Lange and Greve, 1997). Additionally, it has
been found in fish that high temperatures shorten the time before the larva experiences
irreversible starvation, or the time it takes before the larva exhausts all its yolk reserves and
becomes too weak to feed on exogenous food supplies, negatively affecting recruitment (Blaxter,
1992; McGurk, 1984). It appears that warm temperatures are generally important to successful
larval and juvenile development (Sundby, 2000; Rutherford and Houde, 1996), but past a certain
threshold, high temperatures are detrimental to recruitment and limit a species' distribution.
D-4 Trophic Dynamic Shifts
Changes in trophic dynamics of marine ecosystems are an indirect effect of temperature on
warm-edge range boundaries. Marine organisms exist in a thermal niche and interact in a food
web specific to that niche. When warming temperatures shift ranges of species, the trophic
dynamics are potentially affected. We reviewed seven articles (Table D-l) that support this line
of evidence. For example, in the eastern North Atlantic Sea and European shelf seas, warm-water
copepod species have experienced a northward extension by more than 10 degrees latitude while
colder-water species have decreased in numbers (Beaugrand et al., 2002). Because copepods are
prey for many larger marine organisms, these shifts could have substantial effects on the entire
ecosystem, especially on fish abundances, with a possible decline or collapse in the stock of
boreal species such as cod (Beaugrand et al., 2002; Sundby, 2000). Temperature can also impact
trophic dynamics because each species has different temperature sensitivities; thus, when
temperatures rise, predator-prey interactions can be altered based on the thermal sensitivities of
the species involved. Studies have found that in temperate estuaries, crustaceans are more readily
able to cope with an increase in temperature, even increasing their growth potential, than their
predators (fish species) and bivalve prey. As a result, bivalve recruitment could be negatively
affected due to higher predation pressure (Freitas et al., 2007).
245
-------
Table D-1. Summary of studies supporting the assumption that temperature sets the warm-edge range limits of marine species.
Studies are limited to references not included in the Cahill et al. (2013) review.
Species
Taxon
Experiment/Approach
Evidence Type
Species' Response/Main finding
Citation
Mytilus edulis
Bivalve
Tested levels of stress-70
protein isoforms of 70, 72
and 78 kDa
Physiological/Biochemica
I: Protein synthesis and
denaturation
High temperature caused protein
denaturation, suggesting that the mussels'
distribution is restricted by temperature
Chappie et al.,
1998
Maja squinado
Crustacean
Measured physiological
limitations of thermal
tolerance
Physiological: Aerobic
scope
Limited circulation and ventilation at high
temperatures caused insufficient oxygen
supply, thus limiting aerobic scope and
thermal tolerance
Frederich and
Portner, 2000
Five coral reef
fishes
Fish
Tested the effect of
increased water
temperatures on the
resting and maximum
rates of oxygen
consumption
Physiological: Aerobic
scope
Aerobic scope decreased in all species due
to increased temperature increases of 2-4 °C,
but varied across species, suggesting
changes in community composition with
climate change
Nilsson et al.,
2009
Zoarces
viviparus
Fish
Examined thermally
limited oxygen delivery in
southernmost distribution
area
Physiological: Aerobic
scope/growth
Growth performance decreased and
heat-induced mortality occurred as a result of
thermally limited oxygen delivery in higher
temperatures
Portner and
Knust, 2007
Gadus morhua,
Zoarces
viviparus
Fish
Reviewed and analyzed
temperature-dependent
metabolic adaptation,
energy budgets,
biogeography, and fitness
Physiological: Aerobic
scope/Range shift:
Expansion of higher
latitude limit/Trophic
effects
Population declined due to high temperatures
as a result of oxygen limitation, range shifted
poleward, loss of larger copepod species,
changing trophic dynamics
Portner et al.,
2008
Acanthochromis
polyacanthus
Fish
Tested the effect of
temperature on growth
Physiological: Growth
Growth performance of juveniles and adults
declined at maximum sea surface
temperature experienced at location
Munday et al.,
2008
NA
Marine
invertebrates
Examine the effect of
temperature on larval
development
Physiological: Growth
Temperature best explained latitudinal
differences in developmental rates in marine
invertebrates, increased temperature affected
developmental rates and time to hatching
Hoegh-
Guldberg and
Pearse, 1995
246
-------
Species
Taxon
Experiment/Approach
Evidence Type
Species' Response/Main finding
Citation
Engraulis
japonicus,
Sardinops
melanostictus
Fish
Examined the optimal
temperature for growth of
larva
Physiological: Growth
Japanese anchovy and Japanese sardine
regime shift occurred due to different optimal
temperature for larval growth (optimal growth
rate for anchovy larvae occurred at 22.0 °C,
whereas that for sardine larvae occurred at
16.2 °C)
Takasuka et
al., 2007
Cheilodactylus
spectabilis
Fish
Examined effect of
warming sea water on
growth and metabolism
using changes in otoliths
over 90 years
Physiological: Growth
Reduced growth as a result of high
temperatures at the warm boundary,
suggesting increased metabolic costs
Neuheimer et
al., 2011
360 species or
species groups
Fish and
invertebrates
Measured range shifts to
understand how marine
species respond to
climate velocity
Range shift: Climate
velocities
Marine taxa follow climate velocities, thus
variation in species range shifts can be
explained by local variation in temperature,
species tended to shift deeper when
experiencing increased sea surface
temperature
Pinsky et al.,
2013
Semibalanus
balanoides
Crustacean
Transplant experiments
and thermal modelling to
investigate role of climate
on poleward contraction of
southern range
Range shift: Contraction
of lower latitude limit
Southern limit contracted 350 km northward,
mortality occurred in transplanted barnacles
in the sun due to high temperatures during
aerial exposure
Jones et al.,
2012
NA
Fish
Examined shifts in
species' boundaries and
centers of distribution in
response to increased
temperature
Range shift: Contraction
of lower latitude limit and
expansion of higher
latitude limit
>2/3 of species' distributions shifted in
response to climatic warming (shifted
poleward or moved deeper in the water
column), southern boundaries of 1/4 of the fish
shifted north
Perry et al.,
2005
Semibalanus
balanoides,
Diopatra
neapolitana
Crustacean,
polychaete
Comparing distribution
changes with temperature
changes
Range shift: Contraction
of lower latitude limit and
expansion of higher
latitude limit
Southern geographic limit of the barnacle
retreated 300 km, northern geographic limit of
the polychaete shifted poleward 300 km
Wethey and
Woodin, 2010
247
-------
Species
Taxon
Experiment/Approach
Evidence Type
Species' Response/Main finding
Citation
NA
Marine
ectotherms
and terrestrial
species
Test predictions of thermal
tolerance in relation to
range and range shifts
Range shift: Contraction
of lower latitude limit and
expansion of higher
latitude limit
Marine ectotherm ranges expanded
northward and contracted at the southern
boundary due to thermal tolerance, latitudinal
ranges correspond to thermal tolerance
Sunday et al.,
2012
Arctica islandica,
Spisula
solidissima
Bivalves
Examined effects of rising
temperature on
distribution
Range shift: Contraction
of lower latitude limit
Lower growth rate and tissue weight occurred
at high density at southern edge of range,
predicted contraction of southern range limit
Weinberg et
al., 2002
9 species
Crustacean,
polychaete
Examined the effect of
extreme temperature
events and tested
mechanistic geographic
hypothesis on the factor
that sets range
Range shift: Extreme
temperature event
Extreme temperature events affect population
distributions, climate change is punctuated by
extreme episodes and the rate of change of
temperature is highly variable, thus the
spatial pattern of range shifts varies
Wethey et al.,
2011
9 species
categories
Coral
Examined 80 years of
SST data and range
changes of coral
Range shift: Expansion
of higher latitude limit
Four major coral species categories shifted
poleward at 14 km/year since 1930
Yamano et al.,
2011
Teredo bartschi
Bivalve
Evaluated species in the
thermal effluent of a
nuclear generating station
Range shift: Expansion
of higher latitude limit
The southern wood borer, Teredo bartschi,
was found in New Jersey only in the thermal
effluent. The northern distributions of another
wood borer, Teredo furcifera, and polychaete,
Ficopomatus enigmaticus, also appear to be
related to the thermal effluent.
Hoagland and
Turner, 1980
45 species
Invertebrate
intertidal
species
Reported changes in
abundance/distribution of
45 species over 61 years
Range shift:
Expansion of higher
latitude limit
Ranges shifted northward and community
structure was altered due to relative changes
of abundances of species
Barry et al.,
1995
NA
Crustacean
Examined copepod range
shifts and ecosystem
changes due to
temperature increases
Range shift: Expansion
of higher latitude limit
/Trophic effects
Northward extension occurred of more than
10° latitude of warm-water species
associated with a decrease in the number of
colder water species, negatively affects
boreal species due to food web changes
Beaugrand et
al., 2002, 2009
Gadus morhua
Fish
Examined the effect of
temperature on
recruitment
Impaired
fecundity/Recruitment
Negative relationship between temperature
and recruitment of Atlantic cod stocks located
in warm water
Planque and
Fredou, 1999
248
-------
Species
Taxon
Experiment/Approach
Evidence Type
Species' Response/Main finding
Citation
Limanda,
Microstomus kitt,
Pleuronectes
platessa
Fish
Examined the effect of
temperature on spawning
time, recruitment, and
distribution offish
Impaired
fecundity/Recruitment
Temperature during gonadal maturation
affects recruitment and distribution through
influence of rate of gonadal maturation and
spawning time, negative correlation between
temperature and abundance
Lange and
Greve, 1997
Macoma balthica
Bivalve
Studied the population
responses to warmer than
average temperatures
Impaired
fecundity/Recruitment
Warming temperatures caused a reduction of
reproductive output and recruitment,
decreased growth, and increased mortality
due to low BMI condition values
Beukema et
al., 2009
NA
Fish
Examined the effect of
temperature on
recruitment, growth, and
trophic dynamics
Impaired
fecundity/Recruitment
/Trophic effects
Temperature affected body size, growth,
differentiation of muscle and meristic
characters, predicted a mismatch of larvae
with their food supply
Blaxter, 1992
Macoma balthica
Bivalve
Examined temperature-
induced effects on
reproduction, onset of
spawning, and juvenile
mortality rate
Impaired
fecundity/Recruitment
/Trophic effects
Rising seawater temperature affected
recruitment by a decrease in reproductive
output and by spring advancement of bivalve
spawning, rising temperature causes a
mismatch of spawning, phytoplankton bloom,
and settlement of juvenile shrimp
Philippart et
al., 2003
Calanus
finmarchicus,
Gadus morhua
Fish,
crustacean
Examined the effect of
temperature on
recruitment and trophic
dynamics
Impaired
fecundity/Recruitment
/Trophic effects
Atlantic cod in the uppermost part of the
temperature range show a decrease in
recruitment with increasing temperature due
to effects on vital rates and food web
Sundby, 2000
NA
Phytoplankton
Examined relationships
between temperature and
biomass of primary
producers
Trophic effects
Temperature explained 73% of variance in
the relative contribution of small cells to total
phytoplankton biomass, predicting a shift
toward smaller primary producers in warmer
ocean, changing trophic dynamics
Moran et al.,
2010
NA
Benthic
intertidal
species
Examined temperature
sensitivity of predators
versus prey
Trophic effects
Different thermal tolerances in predator/prey
changed trophic dynamics (more predator
pressure) and negatively affected prey
recruitment
Freitas et al.,
2007
249
-------
Appendix E. Metadata of GIS Analysis of
Temperature and Ocean Acidification Values
E-1 Aragonite Saturation State Projections by MEOW Ecoregion GIS Process
Aragonite saturation state projections to 2100 (Cao and Caldeira, 2008) in netCDF format were obtained
from Long Cao on 7/19/2014. These data were outputs from a coarse resolution model coupled with a
climate-carbon cycle model that had a horizontal resolution of 1.8° latitude and 3.6° longitude
(approximately 200 km X 400 km at the equator).
The ArcGIS 10.2.2 Multi-dimension tool MAKE NetCDF RASTER LAYER was used to convert
aragonite saturation state netCDF formatted data into Arclnfo grid format rasters for RCP8.5 scenarios
for 2010 (baseline), 2050 and 2100. In order to overlay these data with the marine and estuarine portions
of the Marine Ecoregions of the World (MEOW) ecoregion polygons, the ArcGIS 10.2.2 Raster
Projection SHIFT tool was used to shift their x coordinates by -180 degrees so that the Prime Meridian
was -180 degrees (International Date Line). The ArcGIS 10.2.2 Spatial Analyst tool EXTRACT BY
MASK was then used to 'clip' the aragonite saturation state grids to the MEOW marine ecoregions.
The values in the resulting grid cells were multiplied by 100, integerized and converted to vector
polygon shapefiles. The resulting shapefiles were overlaid with the marine and estuarine portion of the
MEOW polygons using the INTERSECT tool and projected into an Albers equal area projection. Items
were created in the resulting shapefile's attribute tables, to: 1) calculate area in square meters, 2) to
divide the GRIDCODE item by 100 to restore the projected aragonite saturation state values and 3) to
calculate the product of area and aragonite saturation values.
These attribute tables (.dbf files) were then opened in Excel and saved as Excel files. Pivot tables were
inserted in each file and the area values and area X aragonite saturation values were summarized by
MEOW ecoregion. The final step to calculate the mean aragonite saturation state projected values by
ecoregion for 2010, 2050, and 2100 was to divide the sum of aragonite saturation state projected values
by the sum of the areas of each ecoregion.
E-2 NOAA Climate Projections by MEOW Ecoregion GIS Process
Historical climate (1956-2005) and Anomaly (2050-2099) data were downloaded from NOAA's
Climate Data Portal (Table E-1) as netCDF files using the "Average of all Models" data selection
variable. The data variables available when the statistic selected is anomaly includes: anomaly, histclim,
histstddev, and varratio. We used the 'histclim', which represents the average historical 1956-2005
values as stated in http://www.esrl.noaa.gov/psd/ipcc/ocn/ccwp.html:
"If the user selects "Anomaly" as the statistic: The climate change panel (upper right) will show the
difference in the mean climate in the future time period (RCP8.5) compared to the historical reference
period. The climate variability panel (lower left) will show the average inter-annual (de-trended)
250
-------
standard deviation for the historical reference period (1956-2005) (or just a single model's historical
variability). The change in variability (lower right panel) is expressed as a ratio of the de-trended
variance (average or single model) in the future, divided by the past."
The Marine Geospatial Ecology Toolkit (MGET) and zonal statistics tools were used to streamline the
GIS process that was used to extract aragonite saturation data from the netCDF format files. These tools
are available from these two websites:
http: //m gel .en v. duke. edu/m get
and
http://desktop.arcgis.eom/en/arcmap/10.3/tools/spatial-analvst-toolbox/zonal-statistics.htm.
Headers for each netCDF file were extracted in ArcMap 10.2.2 using the MGET tool 'Find netCDFs and
Extract Headers' then copied and pasted into the Header Files Tab of the MGET tool 'Convert 2D
Variable in NetCDF to ArcGIS raster.
Data were summarized from the resulting rasters using the marine and estuarine portion of the Marine
Ecoregions of the World (MEOW) polygons by the Spatial Analyst tool 'Zonal Statistics as table' which
calculated the following statistics: AREA, MIN, MAX, RANGE, MEAN, STD and SUM. Thirteen
MEOW ecoregions were analyzed from the Mexican Tropical Pacific to the Beaufort - continental coast
and shelf.
Table E-1. Historical climate (1956-2005) and Anomaly (2050-2099) data.
Data were downloaded from NOAA's Climate Change Web Portal at: http://www.esrl.noaa.gov/psd/ipcc/ocn/. The
downloaded data were analyzed by MEOW ecoregion
CALCULATED VALUES
SEASON
21st CENTURY PERIOD
Average historical SST AND projected SST by ecoregion
Entire Year
2005-2099
Average historical SST AND projected SST by ecoregion
July-Aug.-Sept.
2005-2099
Average historical SST AND projected SST by ecoregion
Jan.-Feb.-March
2005-2099
Average historical air temp. AND projected air temp, by ecoregion
Entire Year
2005-2099
Average historical air temp. AND projected air temp, by ecoregion
July-Aug.-Sept.
2005-2099
Average historical air temp. AND projected air temp, by ecoregion
Jan.-Feb.-March
2005-2099
Average historical pH AND projected pH by ecoregion
Entire Year
2005-2099
Average historical pH AND projected pH by ecoregion
July-Aug.-Sept.
2005-2099
Average historical pH AND projected pH by ecoregion
Jan.-Feb.-March
2005-2099
Average historical Chi AND projected Chi by ecoregion
Entire Year
2005-2099
Average historical Chi AND projected Chi by ecoregion
July-Aug.-Sept.
2005-2099
Average historical Chi AND projected Chi by ecoregion
Jan.-Feb.-March
2005-2099
Average historical 30-m temp. AND projected 30-m temp, by ecoregion
Entire Year
2005-2099
Average historical 30-m temp. AND projected 30-m temp, by ecoregion
July-Aug.-Sept.
2005-2099
Average historical 30-m temp. AND projected 30-m temp, by ecoregion
Jan.-Feb.-March
2005-2099
Average historical 50-m temp. AND projected 50-m temp, by ecoregion
Entire Year
2005-2099
Average historical 100-m temp. AND projected 100-m temp, by ecoregion
Entire Year
2005-2099
251
-------
CALCULATED VALUES
SEASON
21st CENTURY PERIOD
Average historical 200-m temp. AND projected 200-m temp, by ecoregion
Entire Year
2005-2099
252
-------
Glossary of Terms
Term
Absent
Abyssal
Albers projection
Algorithm-based risk
assessment
Anadromous
Aragonite
Aragonite saturation
state (Qa)
Arctic endemic
Area of occupancy
(AOO)
Baseline/Status Risk
Bathyal
Benthic larvae
Benthopelagic
Binary fission
Brachyuran crabs
Brackish
Broadcast spawner
Brooded
Definition
Term previously used in CBRAT to indicate the Error/Extinct classification.
A vertical depth zone in the ocean between >2000 to 6000 m.
Albers equal-area conic projection. A conic, equal area map projection that uses two
standard parallels. Although scale and shape are not preserved, distortion is minimal
between the standard parallels. See
http://desktop.arcqis.com/en/arcmap/latest/map/proiections/albers-egual-area-conic.htm.
Risk assessment based on a knowledge base (database) and a rule set, with no expert
intervention in calculating final risks. Used to avoid the limitations of expert solicitations,
include potential sources of bias.
Species that spend most of their lives in saltwater and migrate to freshwater to breed.
A highly soluble form of calcium carbonate.
The ratio of the concentration of aragonite present in sea water compared to the total
amount of aragonite that sea water could hold when saturated, symbolized by Qa. When Qa
< 1, the seawater is undersaturated with respect to aragonite, and aragonite shells will tend
to dissolve.
Native to the Arctic region and not occurring naturally anywhere else.
Area of the outermost limits over which a species actually occurs; total area of all patches
occupied by a species.
Baseline risks are inherent biotic traits of species that increase vulnerability to climate
change. Status risks are changes in a species' viability (e.g., population decline) due to
external factors, such as overfishing that increase vulnerability to climate change.
> 200 - 2000m. This benthic zone is below the euphotic zone and extends down the
continental slope.
Larvae that remain on the bottom or within the tubes of adults.
Animals living all or part of their life in the water column directly above but not on the
bottom.
Reproduction by splitting into two approximately equal parts.
Decapod crustaceans of the infraorder Brachyura. True crabs not to be confused with
similarly named animals such as hermit crabs, king crabs, porcelain crabs, or horseshoe
crabs.
Salinity 0.5 - < 30 psu.
Both males and females discharge gametes into the water column.
The larval or juvenile phase is brooded within the adult or tube of the adult; ovoviviparous.
253
-------
Term
Definition
Budding and
fragmentation
Calcite
Catadromous
CBRAT
Chemoautotrophic
Climate-adjusted
baseline/status risk
Coastal acidification
Coastal
Biogeographic Risk
Analysis Tool
(CBRAT)
Coolest Occupied
Ecoregion (COE)
Cold temperate
province
Constrained
Coupled Model Inter-
comparison Project
Phase 5 (CMIP5)
Cryptic species
Cryptofauna
Decomposer
Deep subtidal
Deposit feeder
Splitting into unequal parts. Buds may form on the body of the "parent".
Carbonate mineral CaCCb.
Species that spend most of their lives in freshwater and migrate to saltwater to breed.
Coastal Biodiversity Risk Analysis Tool (http://www.cbrat.org).
Organisms, typically bacteria, that derive their energy from inorganic sources, including
sulfides and ferrous iron. Chemoautotrophic bacteria live symbiotically with certain
organisms, providing nutrients to their host. Chemosynthetic.
Greatest baseline/status risks weighted by the greatest climate risk.
Reduction in pH in near-coastal waters, including estuaries, as opposed to reductions in pH
in ocean waters.
Ecoinformatic tool synthesizing life history, habitat, distributional, and abundance data on
near-coastal species. Predict vulnerability to climate change, including temperature
increases, ocean acidification, and sea level rise. Available at http://www.cbrat.oro.
In CBRAT, the COE is the coolest ecoregion in which the species maintains a viable
population. Different ecoregions may be defined as the COE depends upon the specific
temperature measurement (air, SST, subsurface).
In the MEOW biogeographical schema, provinces are the unit larger than ecoregions and
smaller than realms. In the NEP, the Cold Temperate Northeast Pacific Province is
composed of the Aleutian Islands, Gulf of Alaska, North American Pacific Fjordland, Puget
Trough/Georgia Basin, Oregon, Washington, Vancouver Coast and Shelf, and Northern
California ecoregions.
As used in CBRAT, SLR predictions of habitat loss in which the habitat is not allowed to
migrate inland due to anthropogenic or natural barriers. See Unconstrained.
A climate model based on an international effort (http://cmip-pcmdi.llnl.gov/). CMIP5 was
used in IPCC Fifth Assessment. Results are served by the NOAA's Climate Web Portal
(http://www.esrl.noaa.gov/psd/ipcc/ocn/ccwp.html).
Two or more distinct species classified as a single species.
Sessile and vagile organisms living in the interstices and crevices formed by epibenthic
organisms or their structures, such as formed by mussel beds, living corals, and coral
rubble.
Organisms that breakdown and digest dead organisms. Bacteria and fungi are major
decomposer groups.
> 30 - 200 m depth.
Animal that ingests sediment particles, feeding on the associated detritus, microflora, and
microorganisms.
254
-------
Term
Detritivore
Direct development
Dominance
normalized relative
abundance (DNRA)
Eco region
Ectoparasite
Endemic
Endoparasite
Epibiotic
Epiphytic
Epizoic
Error/Extinct
Eustatic sea level rise
(ESLR)
Expert solicitation
Extent of occurrence
(EOO)
Folivore
Free scale pH (pHf)
Freecast spawners
Gonochoristic /
Dioecious
Grazer
Definition
Animals that feed on small detritus (i.e., plant and animal remains). Q.v. scavenger.
Development without a larval phase.
As defined in this document, abundance of a species divided by the average abundance of
the dominant species in the sample, where dominant species are defined as those
constituting >75% of the individuals.
In the Marine Ecoregions of the World (MEOW) biogeographic schema, ecoregions are the
smallest coastal unit. They are defined as areas "of relatively homogeneous species
composition, clearly distinct from adjacent systems." Globally, there are 232 ecoregions.
See http://www.worldwildlife.orq/publications/marine-ecoreqions-of-the-world-a-
bioregionalization-of-coastal-and-shelf-areas.
External parasite, including gill parasites.
Species only located in a restricted location. In CBRAT, defined as species occurring in only
one ecoregion.
Internal parasite.
Organisms living on surface of a living or dead organism. Relationship may be mutualistic,
parasitic, or commensal.
Living on surface of living or dead plant.
Living on surface of a living or dead animal.
Species that have been reported to occur in an ecoregion but do not actually occur, either
because they were incorrectly reported or because they went extinct in the ecoregion See
Hierarchical abundance classification schema.
Worldwide change in sea level primarily caused by thermal expansion of sea water and
melting of glaciers and ice sheets.
A formal or informal synthesis of opinions from experts on a designated topic. Also referred
to as expert opinion.
Distance or area between the outermost limits of the occurrence of a species. The broad
range of a species.
Feeds on leaves.
See pH
In animals, males and/or females discharge gametes directly into the water column.
Having separate sexes. In plants, male and female flowers are produced on different
individuals.
An organism that feeds by rasping benthic algae from sediment, rocks, or leaf surfaces.
May consume some smaller benthic organisms, but if animals are dominant food source,
the species is classified as a predator.
Hadal
> 6000 m. The deepest areas of the sea, including ocean trenches.
255
-------
Term
Definition
Haploid/diploid
phases
Hermaphrodite/
Monoecious
Heterogamy
Hierarchical
abundance
classification schema
Holoplankton
Hybrid approach to
estimating relative
abundance
Hyperbenthos
Hyper-rare species
Intrinsic rate of
increase (r)
Island ecoregion
Isostatic adjustment
Kleptoparasite
Lecithotrophy
LIDAR
In plants, fungi, kelp, and some microorganisms, an alternation of multicellular haploid and
diploid phases.
Organisms having both male and female sexual organs.
Alternation between sexual and asexual (parthenogenetic) reproductive phases.
In CBRAT, a classification schema for species' relative abundance within an ecoregion. The
abundance classifications are arranged in three levels according to the amount of available
data:
Level I: Present, Not Reported, Error/Extinct, and Transient.
Level II: Abundant, Moderate, Rare.
Level III: Very Abundant, Moderately Abundant; High Moderate, Low Moderate; Moderately
Rare, Very Rare, Hyper-rare.
Species that are planktonic for their entire life cycle.
A systematic approach to combining multiple lines of evidence to assign relative
abundances. The following sources are listed in order of the weight assigned them:
Regional scale, randomized surveys such as the previous EMAP surveys.
Regional-scale, non-randomized quantitative surveys, such as NOAA's RACE surveys.
Expert opinion addressing the ecoregional abundance of a taxon.
Local randomized and non-randomized surveys.
Natural history and taxonomic texts.
Frequency of occurrence data such as from OBIS and GBIF.
Benthic animals that make periodic forays from the bottom into the water column, such as
some of the corophiid amphipods.
Species that have not been observed in >50 years assuming at least a minimal sampling
effort.
The theoretical maximum rate of increase of a population per individual, assuming no
density-dependent effects.
In CBRAT, a MEOW ecoregion that is surrounded by water with no direct contact with the
mainland. Small island ecoregions have a land area of <200 km2.
Vertical movement of the earth's plates resulting in local uplift or subsidence and the raising
or lowering of sea level.
Parasites that feed on the food items that the host has collected; symbionts that "steal" food
from their hosts.
Larvae that derive nourishment from yolk.
Light Detection and Ranging; a remote sensing method that uses light from a pulsed laser
to accurately measure distance to the Earth, from which precise three-dimensional maps
can be generated. See http://oceanservice.noaa.qov/facts/lidar.html.
256
-------
Term
Lithodid crabs
Marine Ecoregions of
the World (MEOW)
Maximum acceptable
toxicant concentration
(MATC)
Medusa/polyp phases
Meroplankton
Mixed fines
Mixed sediments
Monoecious (plants)
Near coastal
Negative evidence
Neritic
Network Common
Data Form (netCDF)
Next Coolest
Unoccupied
Ecoregion (NCUE)
Next Warmest
Unoccupied
Ecoregion (NWUE)
Nonindigenous
species (NIS)
Northeast Pacific
(NEP)
Definition
Crabs of the families Lithodidae and Hapalogastridae. King crabs, not "true" crabs of the
Infraorder Brachyura.
The Marine Ecoregions of the World (MEOW) is a global biogeographic system for coastal
and shelf areas consisting of a nested system of 12 realms, 62 provinces, and 232
ecoregions. See http://www.worldwildlife.org/publications/marine~ecoreqions-of~the~wor1d~a~
bioregionalization~of~coastal~and~shelf~areas.
In toxicology, the MATC is the greatest acceptable concentration, calculated as geometric
mean of the "no observed adverse effects level" (NOAEL) and the "lowest observed adverse
effects level" (LOAEL). When applied to pH and aragonite saturation state, the MATC is the
lowest acceptable level.
In Cnidaria, an alternation between a polypoid benthic stage and a free-living medusoid
stage.
Species that are planktonic for only part of their life cycle, usually the larval phase.
Combination of mud and sand, where the two classes constitute >95% of the weight. Do not
confuse with "mixed sediments", a mixture of mud/sand and cobble/gravel/rock.
Unconsolidated sediment composed of both sand and mud with gravel or cobble, where
gravel and cobble constitute >5% but <75% of the sediment weight. Do not confuse with
"mixed fines".
Plants having separate male and female flowers on the same individual plant.
As used in CBRAT, the region from the supratidal down to 200 m depth. Includes both
estuaries and offshore areas.
Evidence based on not observing an expected event. In CBRAT, absence or a limited
number of reports of a species is used as potential evidence for rarity in an ecoregion.
> 0 - 200m. Subtidal zone extending from the low water mark to the approximate edge of
the continental shelf. Also referred to as the sublittoral zone or coastal waters.
Data formats that support the creation, access, and sharing of array-oriented scientific data.
Often used for oceanographic data. See http://www.unidata.ucar.edu/software/netcdf/.
In CBRAT, the MEOW ecoregion that is not occupied by the target species and is the next
coolest ecoregion compared to the coolest occupied ecoregion (COE). Assumed that the
temperature in the NCUE is too cool for the species is to maintain a viable population.
In CBRAT, the MEOW ecoregion that is not occupied by the target species and is the next
warmest ecoregion compared to the warmest occupied ecoregion (WOE). Assumed that the
temperature in the NWUE is too warm for the species is to maintain a viable population.
Species introduced outside of their natural range via anthropogenic vectors, such as ballast
water discharges. Also referred to as exotic species or invaders.
As used in CBRAT, the near-coastal region from the Aleutians Islands through the Gulf of
California.
257
-------
Term
Definition
Ocean acidification
(OA)
Oceanic
Osmotrophy
Oviparous
Ovoviviparous
Parthenogenesis /
Agamospermy
PH
Planktonic larvae
Planktotrophic larvae
Primary producer
P rota n dry
Protogyny
Rare
Red List
A reduction in the pH of the ocean caused primarily by uptake of carbon dioxide (CO2) from
the atmosphere. Q.v. Coastal acidification.
As used in CBRAT, > 200 m depth. Includes the benthos and water above the continental
slope and ocean floor. Also includes deeper portion of inland seas like Puget Sound and
Gulf of California.
Uptake of dissolved organic compounds by osmosis for nutrition. Can be the sole source of
nutrition or a supplemental source.
Eggs are laid by the female and develop outside of either parent. Crabs are considered
oviparous rather than ovoviviparous.
Eggs develop within the female, or male in some cases, but the embryo derives no
nourishment from the parent. A brooder.
In animals, parthenogenesis is the development of an unfertilized egg. In plants,
agamospermy (apomixes) is the production of fertile seeds without pollination.
Measure of the acidity (pH <7) or basicity (pH >7) of a solution. Theoretically, the negative
of the logarithm to base 10 of the activity of the hydrogen ion. Operationally, pH in seawater
has been measured by four different scales that can differ by more than 0.1 pH unit.
Free scale pH (pHf): pH = -logio[H+], This measures the free H+ ion concentration,
which corresponds to the theoretical definition of pH. However, it is difficult to measure
free ion concentration in seawater. Further, it does not include other ions, such as
sulfate, that affect the "acidity" of seawater.
NBS scale pH (pHnbs): pH obtained with glass electrodes when calibrated against an
NBS or NIST buffer. NBS buffers have a low ionic strength (ca. 0.1 mol kg-1) compared
to full-strength seawater (ca. 0.72 mol kg-1), and the use of such dilute buffers are not
generally recommended for seawater.
Total scale pH (pHt): The total scale pH includes both hydrogen ions and sulfate ions
in the calculation.
Seawater scale pH (pHsws): The seawater scale pH includes hydrogen ions, sulfate
ions, and fluoride ions in the calculation.
Larvae that spend at least part of the larval phase in the water column.
Larvae that feed on other organisms.
Organism whose metabolic energy is derived from sunlight or chemosynthesis in contrast to
consumption of other organisms.
Initially a male and changes to a female.
Initially a female and changes to a male.
See "Hierarchical abundance classification schema".
List of threatened and endangered plant and animal species produced by the IUCN. See
https]//www.iucn.orq/resources/conservation-tools/iucn-red-list-threatened-species.
258
-------
Term
Relative abundance
Relative sea level rise
(RSLR)
Representative
Concentration
Pathways (RCP)
Rhodoliths / Maerl
Rule-based system
Scavenger
Scenario modelling
Seawater scale pH
(pHsws):
Sequential
hermaphrodite
Shallow subtidal
Specialized systems
Spermcast spawner
Sporogenesis
Subsurface deposit
feeder
Summer temperatures
(months used)
Supra littoral
Definition
Abundances normalized to some measure of the abundances of other species in the taxon
or guild. Values depend upon what taxon or guild is used to relativize the abundances. In
CBRAT, quantitative abundances are normalized to the average abundance of the dominant
species within major taxa in a sample (see Dominance normalized relative abundance). See
"Hierarchical abundance classification schema" for relative abundance classes used in
CBRAT.
The net change in sea level at a particular location due to both eustatic SLR and local
factors.
A set of four climate pathways (scenarios) expressed in radiative forcing value (W/m2). RCP
2.6 reflects the lowest emissions while RCP 8.5 reflects continuing emission increases
through the 21st century as a result of both high population growth and a slower rate of
technology development (van Vuuren et al., 2011).
Free-living masses of coralline algae forming a hard substrate. Large aggregations of
rhodoliths can form beds covering hectares. Referred to as maerl in Europe.
A system of representing human expert knowledge in an automated system by coding
logical assertions as IF-THEN statements. Approach used in CBRAT to automatically
calculate risks to climate change as an alternative to expert solicitation.
Feeds on dead organic material. Usually used for species feeding on larger particles or
animal remains.
Evaluation of how risks change under different climate scenarios.
See pH.
Animals that change their sex, from male to female or from female to male.
> 0 - 30 m depth.
As used in CBRAT, ecosystems composed of benthic and pelagic habitats with physical
and/or chemical characteristics distinct from surrounding ecosystems (e.g., saline lagoons,
hydrothermal vents).
Only male discharges gametes into the water column.
Reproduction and dispersal through formation of spores. Spores differ from seeds in having
little food reserves. Most spores are haploid and may be part of an alternation of haploid
and diploid life history stages. Red algae have both diploid and haploid spores.
Deposit feeder that ingests subsurface particles.
July, August and September: used in modeling the effects of summer temperature
increases.
Area above the high water level that is periodically wetted by breaking waves or during
extreme storms. The splash zone.
259
-------
Term
Surface deposit
feeder
Suspension feeder
Symbiont
Symbiotic algae
Synchronous
hermaphrodite
Temperature-adjusted
ocean acidification
risk
TopoBathy
Total scale pH (pHt)
Transient
Uncertainty Analysis
Unconstrained
Vegetative
propagation
Viviparous
von Bertalanffy growth
coefficient (k)
Warm temperate
province
Warmest occupied
ecoregion (WOE)
Winter temperatures
(months)
Definition
Animals that ingests particles at the sediment interface.
Feeds on phytoplankton, zooplankton, and/or suspended particles in the water column.
Organisms living in direct contact or close physical proximity with another organism,
including commensal (+/0), neutral (0/0, and negative (-/+) relationships.
Microflora living in association with other organisms, supplying nutrition to the host (e.g.,
hermatypic corals).
Animals having both male and female sexual organs at the same time (i.e., simultaneous
hermaphrodites).
Risk due to reduced pH or aragonite saturation state incorporating interaction with
enhanced temperatures.
GIS layers that combine topography (land elevation) and bathymetry (water depths),
see pH.
A species that temporarily inhabits an ecoregion beyond its normal range due to unusual
climatic or oceanographic events. By definition, transients are unable to maintain a long-
term viable population in the new ecoregion under present conditions.
Evaluation of how risks change with different effects thresholds and/or model assumptions.
Less formal than a sensitivity analysis.
In CBRAT, SLR predictions of habitat loss in which the habitat is allowed to migrate inland;
absence of anthropogenic or natural barriers to landward migration of intertidal habitats.
See Constrained.
Formation of new individuals in plants without the production of spores or seeds by stolons
(runners) or formation of bulbs.
Development takes place within the female and embryo derives nourishment from the
mother.
In the von Bertalanffy growth equation, k is the rate (1/year) at which the asymptotic length
(size) is approached.
In the MEOW biogeographical schema, provinces are the unit larger than ecoregions and
smaller than realms. In the NEP, the Warm Temperate Northeast Pacific Province is
composed of the Southern California Bight, Magdalena Transition and Cortezian
ecoregions.
In CBRAT, the WOE is the warmest ecoregion in which the species maintains a viable
population. Different ecoregions may be defined as the WOE depends upon the specific
temperature measurement (air, SST, subsurface).
January, February, and March: used in modeling the effects of winter temperature increases
in CBRAT.
260
-------
Term Definition
Wrack Phytodetritus, including kelp, other macroalgae and SAV, deposited in the upper intertidal
on both coastal shores and in estuaries.
261
-------
THIS PAGE INTENTIONALLY BLANK
262
-------
Bibliography
Abele, L.G. 1976. Comparative species composition and relative abundance of decapod crustaceans in marine
habitats of Panama. Marine Biology 38: 263-278.
Aburto-Oropeza, O., Ezcurra, E., Danemann, G., Valdez, V., Murray, J. and Sala, E. 2008. Mangroves in the Gulf of
California increase fishery yields. Proceedings of the National Academy of Sciences 105: 10456-10459.
/Egisdottir, S., White, M.J., Spengler, P.M., Maugherman, A.S., Anderson, L.A., Cook, R.S., Nichols, C.N.,
Lampropoulos, G.K., Walker, B.S., Cohen, G., and Rush, J.D. 2006. The meta-analysis of clinical judgment
project: Fifty-six years of accumulated research on clinical versus statistical prediction. The Counseling
Psychologist 34: 341-382.
Agnalt, A.L., Grefsrud, E.S., Farestveit, E., Larsen, M., and Keulder, F. 2013. Deformities in larvae and juvenile
European lobster (Homarus gammarus) exposed to lower pH at two different temperatures.
Biogeosciences 10: 7883-7895.
Akgakaya, H.R., Butchart, S.H.M., Mace, G.M., Stuart, S.N., and Hilton-Taylor, C. 2006. Use and misuse of the
IUCN Red List Criteria in projecting climate change impacts on biodiversity. Global Change Biology
12: 2037-2043.
Allen, J.D. and Pernet, B. 2007. Intermediate modes of larval development: bridging the gap between
planktotrophy and lecithotrophy. Evolution and Development 9: 643-653.
Allen, L.G., Yoklavich, M.M., Cailliet, G.M., and Horn, M.H. 2006. Bays and Estuaries. The Ecology of Marine
Fishes - California and Adjacent Waters. (Eds. Allen, L.G., Pondella, D.J., and Horn, M.H.) 119-148.
University of California Press. Berkeley 671 pages.
Alongi, D.M. 2008. Mangrove forests: resilience, protection from tsunamis, and responses to global climate
change. Coastal and Shelf Science. 76: 1-13.
Alongi, D.M. 2015. The impact of climate change on mangrove forests. Current Climate Change Reports 1:30-39.
Angulo, E.A., Deves, L., Saint Jalmes, M., and Courchamp, F. 2009. Fatal attraction: rare species in the spotlight.
Proceedings of the Royal Society of London B: Biological Sciences 276: 1331-1337.
Anlauf, H., D'Croz, L., O'Dea, A. 2011. A corrosive concoction: the combined effects of ocean warming and
acidification on the early growth of a stony coral are multiplicative. Journal of Experimental Marine
Biology and Ecology 397: 13-20.
Appelhans, Y.S., Thomsen, J., Pansch, C., Melzner, F., and Wahl, M. 2012. Sour times: seawater acidification
effects on growth, feeding behaviour and acid-base status of Asterias rubens and Carcinus maenas.
Inter-Research Marine Ecology Progress Series 459: 85-98.
263
-------
Arnberg, M., Calosi, P., Spicer, J.I., Tandberg, H.S., Nilsen, M., Westerlund, S., and Bechmannm, R.K. 2013.
Elevated temperature elicits greater effects than decreased pH on the development, feeding and
metabolism of northern shrimp (Pandalus borealis) larvae. Marine Biology 160: 2037-2048.
Arnold, K.E., Findlay, H.S., Spicer, J.I., Daniels, C.L., and Boothroyd, D. 2009. Effect of C02-related acidification on
aspects of the larval development of the European lobster, Homarus gammarus (L.). Biogeosciences
6:1747-1754.
Babyak, M.A. 2004. What you see may not be what you get: a brief, nontechnical introduction to overfitting in
regression-type models. Psychosomatic Medicine 66: 411-421.
Balint, M., Domisch, S., Engelhardt, C.H.M., Haase, P., Lehrian, S., Sauer, J., Theissinger, K., Pauls, S.U., and
Nowak, C. 2011. Cryptic biodiversity loss linked to global climate change. Nature Climate Change 1: 313-
318.
Barboza, F.R. and Defeo, O. 2015. Global diversity patterns in sandy beach macrofauna: a biogeographic analysis.
Scientific Reports 5: 14515, doi: 10.1038/srepl4515.
Barry, J.P., Baxter, C.H., Sagarin, R.D., and Gilman, S.E. 1995. Climate-related, long-term faunal changes in a
California rocky intertidal community. Science 267: 672-675.
Beaugrand, G., Luczak, C., and Edwards, M. 2009. Rapid biogeographical plankton shifts in the North Atlantic
Ocean. Global Change Biology 15: 1790-1803.
Beaugrand, G., Reid, P.C., Ibanez, F., Lindley, J.A., and Edwards, M. 2002. Reorganization of North Atlantic
marine copepod biodiversity and climate. Science 296: 1692-1694.
Bechmann, R.K., Taban, I.C., Westerlund, S., Godal, B.F., Arnberg, M., Vingen, S., Ingvarsdottir, A., and Baussant,
T. 2011. Effects of ocean acidification on early life stages of shrimp (Pandalus borealis) and mussel
(Mytilus edulis). Journal of Toxicology and Environmental Health, Part A 74: 424-438.
Belan, T.A. 2003. Benthos abundance pattern and species composition in conditions of pollution in Amursky Bay
(the Peter the Great Bay, the Sea of Japan). Marine Pollution Bulletin 46: 1111-1119.
Bellard, C., Bertelesmeier, C., Leadley, P., Thuiller, W., and Courchamp, F. 2012. Impacts of climate change on
the future of biodiversity. Ecology Letters 15: 365-377.
Berkeley, S.A., Chapman, C., and Sogard, S.M. 2004. Maternal age as a determinant of larval growth and survival
in a marine fish, Sebastes melanops. Ecology 85: 1258-1264.
Bernhardt, J.R. and Leslie, H.M. 2013. Resilience to climate change in coastal marine ecosystems. Annual Review
of Marine Science 5: 371-392.
Beukema, J.J., Dekker, R., and Jansen, J.M. 2009. Some like it cold: populations of the tellinid bivalve Macoma
balthica (L.) suffer in various wars from a warming climate. Marine Ecology Progress Series 384: 135-
145.
264
-------
Bickford, D., Lohman, D.J., Sodhi, N.S., Ng, P.K.L., Meier, R., Winker, K., Ingram, K.K., and Das, I. 2006. Cryptic
species as a window on diversity and conservation. Trends in Ecology & Evolution 22: 148-155.
Blackburn, T.M. and Gaston, K.J. 1997. Who is rare? Artifacts and complexities in rarity determination. The
Biology of Rarity. (Eds. Kunin, W.E. and Gaston, K.J.) Chapman & Hall, London. Pages 48-60.
Blankespoor, B., Dasgupta, S., and Laplante, B. 2012. Sea-level rise and coastal wetlands: impacts and costs.
Policy Research working paper number WPS 6277. Washington, DC: World Bank Group. 27 pages.
http: //documents.worldbank.org/curated/en/375251468323964212/Sea-level-rise-and~coastal-
wetlands-impacts-and-costs.
Blaxter, J.H.S. 1992. The effect of temperature on larval fishes. Netherlands Journal of Zoology 42: 336-357.
Blunden, J. and Arndt, D.S. (eds.) 2016. State of the climate in 2015. Bulletin of the American Meteorological
Society 97: S1-S275.
Bopp, L., Resplandy, L., Orr, J.C., Doney, S.C., Dunne, J.P., Gehlen, M., Halloran, P., Heinze, C., llyina, T., Seferian,
R., Tjiputra, J., and Vichi, M. 2013. Multiple stressors of ocean ecosystems in the 21st century:
projections with CMIP5 models. Biogeosciences 10: 6225-6245.
Branch, T.A., Dejoseph, B.M., Ray, L.J., and Wagner, C.A. 2013. Impacts of ocean acidification on marine seafood.
Trends in Ecology & Evolution 28: 178-186.
Breitburg, D.L., Salisbury, J., Bernhard, J.M., Cai, W.-J., Dupont, S., Doney, S.C., Kroeker, K.J., Levin, L.A., Long, C.,
Milke, L.M., Miller, S.H., Phelan, B., Passow, U., Seibel, B.A., Todgham, A.E., and Tarrant, A.M. 2015. And
on top of all that...: Coping with ocean acidification in the midst of many stressors. Oceanography
28: 48-61.
Brown, C.J., Schoeman, D.S., Sydeman, W., Brander, K., Buckley, L., Burrows, M., Duarte, C.M., Moore, P.J.,
Pandolfi, J.M., Poloczanska, E., Venables, W., and Richardson, A.J. 2011. Quantitative approaches in
climate change ecology. Global Change Biology 17: 3697-3713.
Buckley, L.B., Urban, M.C., Angilletta, M.J., Crozier, L.G., Rissler, L.J., and Sears, M.W. 2010. Can mechanism
inform species' distribution models? Ecology Letters 13: 1041-1054.
Bulling, M.T., Hicks, N., Murray, L., Paterson, D.M., Raffaelli, D., White, P.C.L., and Solan, M. 2010. Marine
biodiversity-ecosystem functions under uncertain environmental futures. Philosophical Transactions of
the Royal Society B: Biological Sciences 365: 2107-2116.
Busch, D.S. and McElhany, P. 2016. Estimates of the direct effect of seawater pH on the survival rate of species
groups in the California current ecosystem. PLoS ONE 11(8): e0160669.
https://doi.org/10.1371/iournal.pone.0160669.
Bustamante, R.H., Branch, G.M., and Eekhout, S., 1995. Maintenance of an exceptional intertidal grazer biomass
in South Africa: subsidy by subtidal kelps. Ecology 76: 2314- 2329.
265
-------
Byrne, M. 2011. Impact of ocean warming and ocean acidification on marine invertebrate life history stages:
Vulnerabilities and potential for persistence in a changing ocean. Oceanography and Marine Biology: An
Annual Review (Eds. Gibson, R.N., Atkinson, R.J.A., Gordon, J.D.M., Hughes, R.N., Hughes, D.J., and
Smith, I.P.) 49: 1-42.
Cadien, D.C. and Lovell, L.L. 2012. Expert Workshops: An Approach to CBRAT Data Acquisition, Correction, and
Finalization. Los Angeles, CA: Southern California Association of Marine Taxonomists. 14 pages.
Cahill, A.E., Aiello-Lammens, M.E., Fisher-Reid, M.C., Hua, X., Karanewsky, C.J., Ryu, H.Y., Sbeglia, G.C., Spagnolo,
F., Waldron, J.B., and Wiens, J.J. 2013. Causes of warm-edge range limits: systematic review, proximate
factors and implications for climate change. Journal of Biogeography 41: 429-442.
California Natural Resources Agency. 2014. Safeguarding California: Reducing Climate Risk. An Update to the
2009 California Climate Adaptation Strategy.
http://resources.ca.gov/docs/climate/Final Safeguarding CA Plan July 31 2014.pdf
Cao, L. and Caldeira, K. 2008. Atmospheric C02 stabilization and ocean acidification. Geophysical Research
Letters 35: 1-5.
Carter, H.A., Ceballos-Osuna, L., Miller, N.A., and Stillman, J.H. 2013. Impact of ocean acidification on
metabolism and energetics during early life stages of the intertidal porcelain crab Petrolisthes cinctipes.
The Journal of Experimental Biology 216: 1412-1422.
Ceballos-Osuna, L., Carter, H.A., Miller, N.A., and Stillman, J.H. 2013. Effects of ocean acidification on early life-
history stages of the intertidal porcelain crab Petrolisthes cinctipes. Journal of Experimental Biology
216: 1405-1411.
Chan, N. and Connolly, S.R. 2013. Sensitivity of coral calcification to ocean acidification: a meta-analysis. Global
Change Biology 19: 282-290.
Chappie, J.P., Smerdon, G.R., Berry, R.J., and Hawkins, A.J.S. 1998. Seasonal changes in stress-70 protein levels
reflect thermal tolerance in the marine bivalve Mytilus edulis L. Journal of Experimental Marine Biology
and Ecology 229: 53-68.
Charrette, N.A., Cleary, D.F.R., and Mooers, A.O. 2006. Range restricted, specialist Bornean butterflies are less
likely to recover from ENSO-induced disturbance. Ecology 87: 2330-2337.
Chavez, F.P., Collins, C.A., Huyer, A., and Mackas, D.L. 2002. El Nino along the west coast of North America.
Progress in Oceanography 54: 1-5.
Chenelot, H., Jewett, S., and Hoberg, M. 2008. Invertebrate communities associated with various substrates in
the nearshore Eastern Aleutian Islands, with emphasis on thick crustose coralline algae. In: Brueggeman,
P., Pollock, N.W. (eds). Diving for science 2008. Proc. Am. Acad. Underwater Sci. 27th Annual Diving
Symp., Dauphin Island, Ala., pp 13-36.
266
-------
Cheung, W.W.L., Lam, V.W.Y. and Pauly, D. 2008. Modelling present and climate-shifted distribution of marine
fishes and invertebrates. Fisheries Centre Research Reports. Vancouver, BC: University of British
Columbia 16. 72 pages.
Cheung, W.W.L., Lam, V.W.Y., Sarmiento, J.L., Kearney, K., Watson, R., and Pauly, D. 2009. Projecting global
marine biodiversity impacts under climate change scenarios. Fish and Fisheries 10: 235-251.
Cheung, W.W.L., Sarmiento, J.L., Dunne, J., Frolicher, T.L., Lam, V.W.Y., Deng Palomares, M.L., Watson, R., and
Pauly, D. 2012. Shrinking of fishes exacerbates impact of global ocean changes on marine ecosystems.
Nature Climate Change 3: 254-258.
Christmas, A.-M.F. 2013. Effects of ocean acidification on dispersal behavior in the larval stage of the Dungeness
crab and the Pacific Green Shore crab. Western Washington University, Master Thesis. 72 pages.
Chown, S.L. 2012. Trait-based approaches to conservation physiology: forecasting environmental change risks
from the bottom up. Philosophical Transactions of the Royal Society of London Series B: Biological
Sciences 367: 1615-1627.
Chu-Agor, M.L., Munoz-Carpena, R., Kikera, G.A., Aiello-Lammens, M.E., Akgakaya, H.R., Convertino, M., and
Linkov, I. 2012. Simulating the fate of Florida Snowy Plovers with sea-level rise: Exploring research and
management priorities with a global uncertainty and sensitivity analysis perspective. Ecological
Modeling 224: 33^7.
Church, J.A., Clark, P.U., Cazenave, A., Gregory, J.M., Jevrejeva, S., Levermann, A., Merrifield, M.A., Milne, G.A.,
Nerem, R.S., Nunn, P.D., Payne, A.J., Pfeffer, W.T., Stammer, D., and Unnikrishnam, A.S. 2013. Sea level
change. Climate Change 2013: The Physical Basis, Contribution of Working Group I to The Fifth
Assessment Report of the Intergovernmental Panel On Climate Change (Eds. Stocker, T.F., Qin, D.,
Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P.M.).
Cambridge. Pages 1029-1136.
Clark, K.B. 1994. Ascoglossan (=Sacoglossa) molluscs in the Florida Keys: Rare marine invertebrates at special
risk. Bulletin of Marine Science 54: 900-916.
Clinton, P.J. and Lee II, H. 2016. Computing risk to West Coast intertidal rocky habitat due to sea level rise using
LiDAR topobathy. ESRI Ocean GIS Forum, Redlands, California, November 01 - 03, 2016.
https://cfpub.epa.gov/si/si public record report.cfm?dirEntryld=331151.
Clinton, P.J., Frazier M.R., McCoy, L., Reusser, D., and Lee II, H. 2012. Predicting change in eelgrass distribution
due to sea level rise in three Pacific Northwest estuaries. Presented at ESRI Users Conference, San Diego,
CA, July 23-27, 2012. https://cfpub.epa.gov/si/si public record report.cfm?dirEntrvld=240860.
Clough, J.S., 2008. SLAMM 5.0.2 Technical Documentation. Warren Pinnacle Consulting, Inc. 37 pages.
Coan, E.V. and Valentich-Scott, P. 2012. Bivalve Seashells of Tropical West America: Marine Bivalve Mollusks
from Baja California to Northern Peru. Santa Barbara, Ca. Santa Barbara Museum of Natural History.
1258 pages.
267
-------
Coan, E.V., Valentich-Scott, P., and Bernard, F.R. 2000. Bivalve Seashells of Western North America: Marine
Bivalve Mollusks from Arctic Alaska to Baja California. Santa Barbara, Ca: Santa Barbara Museum of
Natural History. 764 pages.
Cohen, A. and Holcomb., M. 2009. Why corals care about Ocean Acidification: uncovering the mechanism.
Oceanography 22: 118-127.
Collins, M., Knutti, R., Arblaster, J., Dufresne, J.-L., Fichefet, T., Friedlingstein, P., Gao, X., Gutowski Jr., W.J.,
Johns, T., Krinner, G., Shongwe, M., Tebaldi, C., Weaver, A.J., and Wehner, M. 2013. Long-term climate
change: projections, commitments and irreversibility. Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of the International Panel on Climate
Change (Eds. Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y.,
Bex, V., and Midgley, P.M.), Cambridge University Press, Cambridge. Pages 1029-1136.
http: //www.climatechange2013.org/images/report/WGlAR5 Chapterl2 FINAL.pdf.
Committee on Assessing Numeric Limits for Living Organisms in Ballast Water; National Research Council. 2011.
Assessing the relationship between propagule pressure and invasion risk in ballast water. Washington,
DC: National Academies Press. 144 pages, http: //www, nap.edu/catalog/13184/assessing-the~
relationship-between-propagule~pressure~and-invasion-fisk-in-ballast-water.
Cooper, N., Bielby, J., Thomas, G.H., and Purvis, A. 2008. Macroecology and extinction risk correlates of frogs.
Global Ecology and Biogeography 17: 211-221.
Cornwall, C.E. and Hurd, C.L. 2016. Experimental design in ocean acidification research: problems and solutions.
ICES Journal of Marine Science 73: 572-581.
Costello, M.J., Bouchet, P., Boxshall, G., Fauchald, K., Gordon, D., et al. 2013. Global coordination and
standardization in marine biodiversity through the World Register of Marine Species (WoRMS) and
related databases. PLoS ONE 8(1): e51629. doi:10.1371/journal.pone.0051629.
Cowardin, L.M., Carter, V., Golet, F.C., and LaRoe, E.T. 1979. Classification of wetlands and deepwater habitats of
the United States. Washington, D.C: U.S. Fish and Wildlife Service Report FWS/OBS/-79/31.
http://mawwg.psu.edu/docs/resources/ClassificationWetlandsDeepwaterHabitatsUS.pdf.
Craft, C., Clough, J., Ehman, J., Joye, S., Park, R., Pennings, S., Guo, H., and Machmuller, M. 2009. Forecasting the
effects of accelerated sea-level rise on tidal marsh ecosystem services. Frontiers in Ecology and the
Environment 7: 73-78.
Crisp, D.J. and Southward, A.J. 1958. The distribution of intertidal organisms along the coasts of the English
Channel. Journal of the Marine Biological Association of the United Kingdom 37: 157-208.
Darwin, C. R. 1854. A monograph on the sub-class Cirripedia, with figures of all the species. The Balanidae, (or
sessile cirripedes); the Verrucidae, etc. etc. etc. London: The Ray Society. Volume 2.
268
-------
Dasher, D., Lomax, T., Jewett, S., Norcross, B., Holladay, B., and Blanchard, A. 2015. Alaska Monitoring and
Assessment Program 2010 and 2011 Chukchi Sea Coastal Survey Environmental Status. Alaska
Department of Environmental Conservation Division of Water, Water Quality Standards, Assessment
and Restoration, Anchorage, AK, DEC AKMAP Chukchi Sea/2015. 43 pages.
Davies, K.F., Margules, C.R., and Lawrence, J.F. 2000. Which traits of species predict population declines in
experimental forest fragments? Ecology 81: 1450-1461.
Davies, K. F., Margules, C. R., and Lawrence, J. F. 2004. A synergistic effect puts rare, specialized species at
greater risk of extinction. Ecology 85: 265-271.
Dawes, R.M., Faust, D., and Meehl, P.E. 1989. Clinical versus actuarial judgment. Science 243: 1668-1674.
De la Haye, K.L., Spicer, J., Widdicombe, S., and Briffa, M. 2011. Reduced sea water pH disrupts resource
assessment and decision making in the hermit crab Pagurus bernhardus. Animal Behaviour 82: 495-501.
Delefosse, M., Banta, G.T., Canal-Verges, P., Penha-Lopes, G., Quintana, C.O., Valdemarsen, T., and Kristensen,
E. 2012. Macrobenthic community response to the Marenzelleria viridis (Polychaeta) invasion of a
Danish estuary. Marine Ecology Progress Series 461: 83-94.
Denney, N.H., Jennings, S., and Reynolds, J.D. 2002. Life-history correlates of maximum population growth rates
in marine fishes. Proceedings of the Royal Society B- Biological Sciences 269: 2229-2138.
Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), National Ocean Service
(NOS), Office of Response and Restoration (ORR), Emergency Response Division (ERD). 2013. NOAA
Shoreline Assessment Manual 4th edition. Office of Response and Restoration, Emergency Response
Division, Seattle, Washington. 154 pages.
http: //response.restoration.noaa.gov/sites/default/files/manual shore assess aue2013.pdf.
Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), National Ocean Service
(NOS), Office for Coastal Management (OCM). 2016. United States Interagency Elevation Inventory
(USIEI). NOAA's Ocean Service, Office for Coastal Management (OCM), Charleston, SC.
https: //coast.noaa.gov/digitalcoast/tools/inventory.
Di Nitto, D., Neukermans, G., Koedam, N., Defever, H., Pattyn, F., Kairo, J.G., and Dahdouh-Guebas, F. 2014.
Mangroves facing climate change: landward migration potential in response to projected scenarios of
sea level rise. Biogeosciences 11: 857-871.
Diamond, S.E., Frame, A.M., Martin, R.A., and Buckley, L.B. 2011. Species' traits predict phenological responses
to climate change in butterflies. Ecology Letters 92: 1005-1012.
Dijkstra, J.A., Lambert, W.J., and Harris, L.G. 2013. Introduced species provide a novel temporal resource that
facilitates native predator population growth. Biological Invasions 15: 911-919.
269
-------
Dissanayake, A. and Ishimatsu, A. 2011. Synergistic effects of elevated C02 and temperature on the metabolic
scope and activity in a shallow-water coastal decapod (Metapenaeus joyneri; Crustacea: Penaeidae).
ICES Journal of Marine Science 68: 1147-1154.
Dodd, L.F., Grabowski, J.H., Piehler, M.F., Westfield, I., and Ries, J.B. 2015. Ocean acidification impairs crab
foraging behavior. Proceedings of the Royal Society B 282: 1-9.
Doney, S.C., Fabry, V.J., Feely, R.A., and Kleypas, J.A. 2009. Ocean acidification: The other C02 problem. Annual
Review of Marine Science 1: 169-192.
Donohue, P.J.C., Calosi, P., Bates, A.H., Laverock, B., Rastrick, S., Mark, F.C., Strobel, A., and Widdicombe, S.
2012. Impact of exposure to elevated pC02 on the physiology and behavior of an important ecosystem
engineer, the burrowing shrimp Upogebia deltaura. Aquatic Biology 15: 73-86.
Drake, J.S., Berntson, E.A., Cope, J.M., Gustafson, R.G., Holmes, E.E., Levin, P.S., Tolimieri, N., Waples, R.S.,
Sogard, S.M., and Williams, G.D. 2010. Status review of five rockfish species in Puget Sound,
Washington: bocaccio (Sebastes paucispinis), canary rockfish (5. pinniger), yelloweye rockfish (5.
ruberrimus), greenstriped rockfish (5. elongatus), and redstripe rockfish (5. proriger). U.S. Department of
Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service. 247
pages.
Drescher, M., Perera, A.H., Johnson, C.J., Buse, L.J., Drew, C.A., and Burgman, M.A. 2013. Toward rigorous use of
expert knowledge in ecological research. Ecosphere 4: 83. http: //dx.doi.org/10.1890/ES12.-00415.1.
Drolet, D., Locke, A., Lewis, M.A., and Davidson, J. 2015. Evidence-based tool surpasses expert opinion in
predicting probability of eradication of aquatic nonindigenous species. Ecological Applications 25: 441-
450.
Ducks Unlimited. No date. Sea Level Rise Providing Nature A-Right-of Way. Online presentation,
http://www.ecv.wa.gov/programs/sea/shorelines/smp/toolbox/docs/winterll marshes model.pdf.
Dugan, J.E. and Hubbard, D.M. 2010. Ecological effects of coastal armoring: a summary of recent results for
exposed sandy beaches in southern California. Puget Sound Shorelines and the Impacts of Armoring—
Proceedings of a State of the Science Workshop. May, 2009. (Eds. Shipman, H., Dethier, M.N.,
Gelfenbaum, G., Fresh, L., and Dinicola, R.S.), U.S. Geological Survey Scientific Investigations Report
2010. Pages 187-194.
Dulvy, N.K., Ellis, J.R., Goodwin, N.B., Grant, A., Reynolds, J.D., and Jennings, S. 2004. Methods of assessing
extinction risk in marine fishes. Fish and Fisheries 5: 255-276.
Dulvy, N.K., Rogers, S.I., Jennings, S., Stelzenmiiller, V., Dye, S.R., and Skjoldal, H.R. 2008. Climate change and
deepening of the North Sea fish assemblage: A biotic indicator of warming seas. Journal of Applied
Ecology 45: 1029-1039.
Dulvy, N.K., Sadovy, Y., and Reynolds, J.D. 2003. Extinction vulnerability in marine populations. Fish and Fisheries
4: 25-64.
270
-------
Dumbauld, B.R., Chapman, J.W., Torchin, M.E., and Kuris, A.M. 2011. Is the collapse of mud shrimp (Upogebia
pugettensis) populations along the Pacific coast of North American caused by outbreaks of a previously
unknown bopyrid isopod parasite (Orthione griffensis)? Estuaries and Coasts 34: 336-350.
Duncan, R.P. and Young, J.R. 2000. Determinants of plant extinction and rarity 145 years after European
settlement of Auckland, New Zealand. Ecology 81: 3048-3061.
Dupont, S., Lundve, B., and Thorndyke, M. 2010. Near future ocean acidification increases growth rate of the
lecithotrophic larvae and juveniles of the sea star Crossaster papposus. Journal of Experimental Zoology
Part B, Molecular and Developmental Evolution 314: 382-389.
Edgar, G.J., Bates, A.E., Bird, T.J., Jones, A.H., Kininmonth, S., Stuart-Smith, R.D., and Webb, T.J. 2016. New
approaches to marine conservation through the scaling up of ecological data. Annual Review of Marine
Science 8: 435-461.
Edmunds, P. J., Brown, D., and Moriarty, V. 2012. Interactive effects of ocean acidification and temperature on
two scleractinian corals from Moorea, French Polynesia. Global Change Biology, 18: 2173-2183.
Eisenlord, M.E., Groner, M.L., Yoshioka, R.M., Elliot, J., Maynard, J., Fradkin, S., Turner, M., Pyne, K., Rivlin, N.,
van Hooidonk, R., Harvell, C.D. 2016. Ochre star mortality during the 2014 wasting disease epizootic:
role of population size structure and temperature. Philosophical Transactions of the Royal Society B
Biological Sciences 371: http://dx.doi.org/10.1098/rstb.2015.0212.
Elith, J. and Leathwick, J.R. 2009. Species distribution models: ecological explanation and prediction across space
and time. Annual Review of Ecology, Evolution, and Systematics 40: 677-697.
Ellison, J., 2000. How South Pacific mangroves may respond to predicted climate change and sea level rise. In:
Gillespie, A., Burns, W. (Eds.), Climate Change in the South Pacific: Impacts and Responses in Australia,
New Zealand, and Small Islands States. Kluwer Academic Publishers, Dordrecht, Netherlands, (Chapter
15), pp. 289-301.
Ellison, J.C. and Stoddart, D.R. 1991. Mangrove ecosystem collapse during predicted sea level rise - Holocene
analogs and implications. Journal of Coastal Research 7: 151-165.
Engle, J.M., and Richards, D.V. 2001. New and unusual marine invertebrates discovered at the California Channel
Islands during the 1997-1998 El Nino. Bulletin of the Southern California Academy of Sciences 100: 186-
198.
Fairchild, J.F., Allert, A.L., Feltz, K.P., Nelson, K.J., and Valle, J.A. 2009. An ecological risk assessment of the acute
and chronic effects of the herbicide Clopyralid to rainbow trout (Oncorhynchus mykiss). Archives of
Environmental Contamination and Toxicology 57: 725-731.
Fehsenfeld, S., Kiko, R., Appelhans, Y., Towle, D.W., Zimmer, M., and Melzner, F. 2011. Effects of elevated
seawater pC02 on gene expression patterns in the gills of the green crab Carcinus maenas. BMC
Genomics 12: 488.
271
-------
Flather, C.H. and Seig, C.H. 2007. Species rarity: Definition, causes and classification. Conservation of Rare or
Little-Known Species. Biological, Social, and Economic Considerations (Eds. Raphael, M.G. and Molina,
R.) 40-66. Washington D.C: Island Press. 392 pages.
Foden, W., Mace, G., Vie, J.-C., Angulo, A., Butchart, S., DeVantier, L., Dublin, H., Gutsche, A., Stuart, S., and
Turak, E. 2008. Species susceptibility to climate change impacts. An analysis of the IUCN Red List of
Threatened Species (Eds: Vie, J.-C., Hilton-Taylor, C., and Stuart, S.) 77-87. Gland, Switzerland: IUCN.
180 pages.
Foden, W.B., Butchart, S.H.M., Stuart, S.N., Vie, J.-C., Akgakaya, H.R., Angulo, A., DeVantier, L.M., Gutsche, A.,
Turak, E., Cao, L., Donner, S.D., Katariya, V., Bernard, R., Holland, R.A., Hughes, A.F., O'Hanlon, S.E.,
Garnett, W.T., §ekercioglu, £.H., and Mace, G.M. 2013. Identifying the World's Most Climate Change
Vulnerable Species: A Systematic Trait-Based Assessment of all Birds, Amphibians and Corals. PLoS ONE
8(6): e65427. doi: 10.1371/journal.pone.0065427.
Foo, S.A. and Byrne, M. 2016. Acclimatization and adaptive capacity of marine species in a changing ocean.
Advances in Marine Biology 74: 69-116.
Food and Agriculture Organization of the United Nations (FAO). 2007. The world's mangroves 1980-2005. FAO
Forestry Paper 153. Rome, 77 pages.
Frederich, M. and Portner, H.O. 2000. Oxygen limitation of thermal tolerance defined by cardiac and ventilatory
performance in spider crab, Maja squinado. American Journal of Physiology - Regulatory, Integrative
and Comparative Physiology 279: R1531-R1538.
Freitas, V., Campos, J., Fonds, M., and Van der Veer, H.W. 2007. Potential impact of temperature change on
epibenthic predator-bivalve prey interactions in temperate estuaries. Journal of Thermal Biology
32: 328-340.
Fujimoto, K., Miyagi, T., Kikuchi, T., and Kawana, T. 1996. Mangrove habitat formation and response to Holocene
sea-level changes on Kosrae Island, Micronesia. Mangroves and Salt Marshes 1: 47-57.
Fussel, H.-M. 2009. An updated assessment of the risks from climate change based on research published since
the IPCC Fourth Assessment Report. Climatic Change 97: 469-482.
Galbraith, H., Jones, R., Park, R., Clough, J., Herrod-Julius, S., Harrington, B. and Page, G. 2002. Global climate
change and sea level rise: potential losses of intertidal habitat for shorebirds. Waterbirds 25: 173-183.
Galbraith, H. and Price, J., 2009. A Framework for Categorizing the Relative Vulnerability of Threatened and
Endangered Species to Climate Change. EPA/600/R-09/011. National Center for Environmental
Assessment, Washington, D.C. 113 pages.
Gallagher, A.J., Kyne, P.M., and Hammerschlag, N. 2012. Ecological risk assessment and its application to
elasmobranch conservation and management. Journal of Fish Biology 80: 1727-1748.
Gaston, K.J. 1994. Rarity. London, UK: Chapman and Hall. 205 pages.
272
-------
Gaston, K.J. 1997. What is rarity? The Biology of Rarity (Eds. Kunin, W.E. and Gaston, K.J.) 30-47. London, UK:
Chapman & Hall. 280 pages.
Gaston, K.J. 2003. The Structure and Dynamics of Geographic Ranges. Oxford Series in Ecology and Evolution.
New York, Oxford University Press Inc. 266 pages.
Gaston, K.J. 2009. Geographic range limits: achieving synthesis. Proceedings of the Royal Society B 276: 1395-
1406.
Gaston, K.J. 2011. Common ecology. Bioscience 61: 354-362.
Geselbracht, L.L., Freeman, K., Birch, A.P., Brenner, J. and Gordon, D.R. 2015. Modeled sea level rise impacts on
coastal ecosystems at six major estuaries on Florida's Gulf Coast: Implications for adaptation planning.
PLoS ONE 10(7): e0132079. doi:10.1371/journal.pone.0132079.
Gill, S.K. and Schultz, J.R. (eds). 2000. Tidal Datums and their Applications. Department of Commerce (DOC),
National Oceanic and Atmospheric Administration (NOAA), National Ocean Service (NOS), Center for
Operational Oceanographic Products and Services, Silver Springs, Maryland. 132 pages.
http: //tidesandcurrents.noaa.gov/publications/tidal datums and their applications.pdf.
Gilman, E., Ellison, J.C., Duke, N.C., Field, C., and Fortuna, S. 2008. Threats to mangroves from climate change
and adaptation options: a review. Aquatic Botany 89: 237-250.
Glenn, E.P., Nagler, P.L., Brusca, R.C., and Hinojosa-Huerta, O. 2006. Coastal wetlands of the northern Gulf of
California: inventory and conservation status. Aquatic Conservation: Marine and Freshwater Ecosystems
16: 5-28.
Glick, P., Clough, J., and Nunley, B. 2007. Sea-level Rise and Coastal Habitats in the Pacific Northwest: An Analysis
for Puget Sound, Southwestern Washington, and Northwestern Oregon. National Wildlife Federation,
Seattle, Washington. 106 pages.
Glynn, P.W. 2012. Global warming and widespread coral mortality: evidence of first coral reef extinctions. Saving
a Million Species: Extinction Risk from Climate Change. (Ed. Hannah, L.) Island Press, Washington, DC.
Pages 103-119.
Godoy, M.D.P. and de Lacerda, L.D. 2015. Mangroves response to climate change: a review of recent findings on
mangrove extension and distribution. Anais da Academia Brasileira de Ciencias 87: 651-67.
Gollasch, S. 2006. Assessment of the introduction potential of aquatic alien species in new environments.
Assessment and Control of Biological Invasion Risks (Eds. Koike, F., Clout, M., Kawamichi, M., De Poorter,
M., and Iwatsuki, K.) Kyoto, Japan and Gland, Switzerland: Shoukadoh Book Sellers, IUCN. Pages 88-91.
Graham, N.A.J. 2007. Ecological versatility and the decline of coral feeding fishes following climate driven coral
mortality. Marine Biology 153: 119-127.
273
-------
Gray, J.S., Bj0rgesaeter, A., Ugland, K.I. 2005. The impact of rare species on natural assemblages. Journal of
Animal Ecology 74: 1131-1139.
Greene, K.E., Zimmerman, J.L., Laney, R.W., and Thomas-Blate, J.C. 2009. Atlantic coast diadromous fish habitat:
A review of utilization, threats, recommendations for conservation, and research needs. Washington,
DC: Atlantic States Marine Fisheries Commission. 464 pages.
Griffen, B.D. 2009. Effects of a newly invasive parasite on the burrowing mud shrimp, a widespread ecosystem
engineer. Marine Ecology Progress Series 391: 73-83.
Grove, W.M., Zald, D.H., Lebow, B.S., Snitz, B.E., and Nelson, C. 2000. Clinical versus mechanical prediction: A
meta-analysis. Psychological Assessment 12: 19-30.
Gutowska, M.A. and Melzner, F. 2009. Abiotic conditions in cephalopod (Sepia officinalis) eggs: embryonic
development at low pH and high pC02. Marine Biology 156: 515-519.
Gynther, I., Waller, N., and Leung, L.K.-P. 2016. Confirmation of the extinction of the Bramble Cay melomys
Melomys rubicola on Bramble Cay, Torres Strait: results and conclusions from a comprehensive survey in
August-September 2014. Unpublished report to the Department of Environment and Heritage
Protection, Queensland Government, Brisbane, https: //www.ehp.qld.gov.au/wildlife/threatened-
species/documents/bramble-cav-melomvs-survev-report.pdf.
Haer, T., Kalnay, E., Kearney, M., and Moll, H. 2013. Relative sea-level rise and the conterminous United States:
Consequences of potential land inundation in terms of population at risk and GDP loss. Global
Environmental Change 23: 1627-1636.
Haigh, R., lanson, D., Holt, C.A., Neate, H.E., and Edwards, A.M. 2015. Effects of ocean acidification on temperate
coastal marine ecosystems and fisheries in the Northeast Pacific. PLoS ONE 10(2): e0117533.
doi:10.1371/journal.pone.0117533.
Hale, S.S. and Heltshe, J.F. 2008. Signals from the benthos: development and evaluation of a benthic index for
the nearshore Gulf of Maine. Ecological Indicators 8: 338-350.
Halpern, B.S., Walbridge, S., Selkoe, K.A., Kappel, C.V., Micheli, F., D'Agrosa, C., Bruno, J.F., Casey, K.S., Ebert, C.,
Fox, H.E., Fujita, R., Heinemann, D., Lenihan, H.S., Madin, E.M.P., Perry, M.T., Selig, E.R., Spalding, M.,
Steneck, R., and Watson, R. 2008. A global map of human impact on marine ecosystems. Science
319: 948-952.
Hannah, L. (Ed.) 2012. Saving a Million Species: Extinction Risk from Climate Change. Island Press, Washington.
432 pages.
Hans, S., Fehnsefeld, S., Treberg, J.R., and Weihrauch, D. 2014. Acid-base regulation in the Dungeness crab
(Metacarcinus magister). Marine Biology 161: 1179-1193.
274
-------
Hare, J.A., Morrison, W.E., Nelson, M.W., Stachura, M.M., Teeters, E.J., Griffis, R.B., Alexander, M.A., Scott, J.D.,
Alade, L., Bell, R.J., Chute, A.S., Curti, K.L., Curtis, T.H., Kircheis, D., Kocik, J.F., Lucey, S.M., McCandless,
C.T., Milke, L.M., Richardson, D.E., Robillard, E., Walsh, H.J., McManus, M.C., Marancik, K.E., and
Griswold, C.A. 2016. A vulnerability assessment of fish and invertebrates to climate change on the
northeast U.S. continental shelf. PLoS ONE 11(2): e0146756. doi:10.1371/ journal.pone.0146756.
Harley, C.D., Randall Hughes, A., Hultgren, K.M., Miner, B.G., Sorte, C.J., Thornber, C.S., Rodriguez, L.F.,
Tomanek, L., and Williams, S.L. 2006. The impacts of climate change in coastal marine systems. Ecology
Letters 9: 228-241.
Harrell, F.E., Lee, K.L., and Mark, D.B. 1996. Multivariable prognostic models: Issues in developing models,
evaluating assumptions and adequacy, and measuring and reducing errors. Statistics in Medicine
15:361-387.
Hart, J. 1982. Crabs and their relatives of British Columbia. Victoria, British Columbia, Canada. British Columbia
Provincial Museum Handbook 40. 267 pages.
Hartley, S. and Kunin, W. 2003. Scale dependency of rarity, extinction risk, and conservation priority.
Conservation Biology 17: 1559-1570.
Harvey, B.P., Gwynn-Jones, D., and Moore, P. J. 2013. Meta-analysis reveals complex marine biological
responses to the interactive effects of ocean acidification and warming. Ecology and Evolution 3: 1016-
1030.
Hawkins, S.J. and Jones, H.D. 1992. Marine Field Course Guide 1. Rocky Shores. Marine Conservation Society,
Immel Publishing. 112 pages.
Heinrich, D.D.U., Watson, S.-A., Rummer, J.L., Brandl, S.J., Simpfendorfer, C.A., Heupel, M.R., and Munday, P.L.
2016. Foraging behaviour of the epaulette shark Hemiscyllium ocellatum is not affected by elevated C02.
ICES Journal of Marine Science 73: 633-640.
Helmuth, B. and Hofmann, G.E. 2001. Microhabitats, thermal heterogeneity, and patterns of physiological stress
in the rocky intertidal zone. Biological Bulletin 201: 374-384.
Helmuth, B., Broitman, B.R., Yamane, L., Gilman, S.E., Mach, K., Mislan, K.A.S., and Denny, M.W. 2010.
Organismal climatology: analyzing environmental variability at scales relevant to physiological stress.
Journal of Experimental Biology 213: 995-1003.
Helmuth, B., Mieszkowska, N., Moore, P., and Hawkins, S.J. 2006. Living on the edge of two changing worlds:
Forecasting the responses of rocky intertidal ecosystems to climate change. Annual Review of Ecology,
Evolution, and Systematics 37: 373-404.
Hendriks, I.E., Duarte, C.M., Alvarez, M. 2010. Vulnerability of marine biodiversity to ocean acidification: a meta-
analysis. Estuarine, Coastal and Shelf Science 86: 157-164.
275
-------
Herborg, L.M., O'Hara, P., and Therriault, T.W. 2009. Forecasting the potential distribution of the invasive
tunicate Didemnum vexillum. Journal of Applied Ecology 46: 64-72.
Hercos, A.P., Sobansky, M., Queirozl, H.L., and Magurran, A.E. 2012. Local and regional rarity in a diverse tropical
fish assemblage. Proceedings of the Royal Society B Biological Sciences 280: 20122076.
http://dx.doi.org/10.1098/rspb.2012.2076.
Hewitt, J.E., Ellis, J.L., and Thrush, S.F. 2016. Multiple stressors, nonlinear effects and the implications of climate
change impacts on marine coastal ecosystems. Global Change Biology 22: 2665-2775.
Hiddink, J. G., Burrows, M.T. and Molinos, J.G. 2015. Temperature tracking by North Sea benthic invertebrates in
response to climate change. Global Change Biology 21: 117-129.
Hines, A.H. 1982. Coexistence in a kelp forest: Size, population dynamics, and resource partitioning in a guild of
spider crabs (Brachyura, Majidae). Ecological Monographs 52: 179-182.
Hoagland, K.E. and Turner, R.D. 1980. Range extensions of teredinids (shipworms) and polychaetes in the vicinity
of a temperate-zone nuclear generating station. Marine Biology 58: 55-64.
Hobbs, J.-P.A., Jones, G.P., and Munday, P.L. 2011. Extinction risk in endemic marine fishes. Conservation
Biology 25: 1053-1055.
Hoegh-Guldberg, O. and Pearse, J.S. 1995. Temperature, food availability, and the development of marine
invertebrate larvae. American Zoologist 35: 415-425.
Hofmann, G.E. and Todgham, A.E. 2010. Living in the now: physiological mechanisms to tolerate a rapidly
changing environment. Annual Review of Physiology 72: 127-145.
Hollenbeck, J.P., Olsen, M.J., and Haig, S.M. 2014. Using terrestrial laser scanning to support ecological research
in the rocky intertidal zone. Journal of Coastal Conservation 18: 701-714.
Horn, H.G., Boersma, M., Garzke, J., Loder, M.G., Sommer, U., and Aberle, N. 2016. Effects of high C02 and
warming on a Baltic Sea microzooplankton community. ICES Journal of Marine Science 73: 772-782.
Hubbard, D.M., Dugan, J.E., Schooler, N.K., and Viola, S.M. 2014. Local extirpations and regional declines of
endemic upper beach invertebrates in southern California. Estuarine, Coastal and Shelf Science 150: 67-
75.
Hyland, J.L., Balthis, W.L., Posey, M.H., Hackney, C.T. and Alphin, T.D. 2004. The soft-bottom macrobenthos of
North Carolina estuaries. Estuaries 27: 501-514.
Inger, R., McDonald, R., Rogowski, D., Jackson, A., Parnell, A., Preston, S., Harrod, C., Goodwin, C., Griffiths, D.,
Dick, J., Elwood, R., Newton, J., and Bearhop, S. 2010. Do non-native invasive fish support elevated
lamprey populations? J. Appl. Ecol. 47: 121-129.
276
-------
IPCC, 2007. Summary for Policymakers. Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change
(Eds. Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., and Miller, H.L.)
Cambridge, UK and New York, NY, USA: Cambridge University Press. 18 pages.
http: //www.ipcc.ch/pdf/assessment-report/ar4/wgl/ar4-wgl-spm.pdf.
IPCC, 2014: Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-
Madruga, Y. Sokona, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B.
Kriemann, J. Savolainen, S. Schlomer, C. von Stechow, T. Zwickel and J.C. Minx (eds.)]. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change. (Eds Pachauri, R.K. and Meyer,
L.). IPCC, Geneva, Switzerland, 151 pages. http://ar5~
syr.ipcc.ch/ipcc/ipcc/resources/pdf/IPCC SynthesisReport.pdf.
Isenburg, M. 2011. LASzip: lossless compression of LiDAR data. European LiDAR Mapping Forum. Salzburg,
Austria. 9 pages, http: //www.cs.unc.edu/~isenbyre/lastools/download/laszip.pdf.
IUCN Species Survival Commission. 2001. International Union for Conservation of Nature red list categories and
criteria Version 3.1. Second edition. Gland, Switzerland: International Union for Conservation of Nature.
33 pages, http: /'/sB.amazonaws.com/iucnredlist-
newcms/staging/public/attachments/3097/redlist cats crit en.pdf.
IUCN Standards and Petitions Subcommittee 2016. Guidelines for using the IUCN Red List Categories and
Criteria. Version 12. 101 pages, http: //cmsdocs.s3.amazonaws.com/RedListGuidelines.pdf.
Iyengar, E.V. 2005. Seasonal feeding-mode changes in the marine facultative kleptoparasite Trichotropis
cancellata (Gastropoda, Capulidae): trade-offs between trophic strategy and reproduction. Canadian
Journal of Zoology 83: 1097-1111.
Jackson, A.C. and Mcllvenny, J. 2011. Coastal squeeze on rocky shores in northern Scotland and some possible
ecological impacts. Journal of Experimental Marine Biology and Ecology 400: 314-321.
Janousek, C.N. and Folger, C.L. 2012. Patterns of distribution and environmental correlates of macroalgal
assemblages and sediment chlorophyll A in Oregon tidal wetlands. Journal of Phycology 48: 1448-1457.
Jenness, J.S. 2004. Calculating landscape surface area from digital elevation models. Wildlife Society Bulletin
32: 829-839.
Jensen, G.C. 1995. Pacific Coast Crabs and Shrimps. Monterey, California: Sea Challengers. 87 pages.
Jensen, G.C. 2014. Crabs and Shrimps of the Pacific Coast: A Guide to Shallow-Water Decapods from
Southeastern Alaska to the Mexican Border. Mola Marine. 240 pages.
277
-------
Jiguet, F., Gadot, A.-S., Julliard, R., Newson, S.E., Couvet, D. 2007. Climate envelope, life history traits and the
resilience of birds facing global change. Global Change Biology 13: 1672-1684.
Johansen, J.L. and Jones, G.P. 2011. Increasing ocean temperature reduces the metabolic performance and
swimming ability of coral reef damselfishes. Global Change Biology 17: 2971-2979.
Jones, M.C., Dye, S.R., Fernandes, J.A., Frolicher, T.L., Pinnegar, J.K. et al. 2013. Predicting the Impact of Climate
Change on Threatened Species in UK Waters. PLoS ONE 8(1): e54216.
doi: 10.1371/journal.pone.0054216.
Jones, M.J., Fielding, A., and Sullivan, M. 2006. Analysing extinction risk in parrots using decision trees.
Biodiversity and Conservation 15: 1993-2007.
Jones, S.J., Southward, A.J., and Wethey, D.S. 2012. Climate change and historical biogeography of the barnacle
Semibalanus balanoides. Global Ecology and Biogeography 21: 716-724.
J0rstad, K.E., Farestveit, E., Rudra, H., Agnalt, A.-L., and Olsen, S. 2002. Studies on red king crab (Paralithodes
camtschaticus) introduced to the Barents Sea. Crabs in Cold Water Regions: Biology, Management and
Economics; Alaska Sea Grant Collage Program. AK-SG-02-01. Pages 425-437.
Josefson, A.B. and Mokievsky, V., 2013. Marine invertebrates. Arctic Biodiversity Assessment. Status and Trends
in Arctic Biodiversity. (Ed. Meltofte, H.). Conservation of Arctic Flora and Fauna, Akureyri, Iceland. Pages.
276-309.
Jumars, P.A., Dorgan, K.M., Lindsay, S.M. 2015. Diet of worms emended: an update of polychaete feeding guilds.
Ann. Rev. Mar. Sci. 7: 497-520.
Kairis, P.A. and Rybczyk, J.M. 2010. Sea level rise and eelgrass (Zostera marina) production: A spatially explicit
relative elevation model for Padilla Bay. Ecological Modeling 221: 1005-1016.
Kaldy, J.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.
Keller, V. and Bollmann, K. 2004. From red lists to species of conservation concern. Conservation Biology
18:1636-1644.
Kendall, M.A., Burrows, M.T., Southward, A.J., and Hawkins, S.J. 2004. Predicting the effects of marine climate
change on invertebrate prey of the birds of rocky shores. Ibis 146: 40-47.
Keppel, E.A., Scrosati, R.A., and Courtenay, A.C. 2012. Ocean acidification decreases growth and development in
American lobster (Homarus americanus) larvae. Journal Northwest Atlantic Fish Science 44: 61-66.
Kim, K.C. and Byrne, L.B. 2006. Biodiversity loss and the taxonomic bottleneck: emerging biodiversity science.
Ecological Research, 21: 794-810.
Kim, T.W., Taylor, J., Lovera, C., and Barry, J.P. 2015. C02 driven decrease in pH disrupts olfactory behavior and
increases individual variation in deep-sea hermit crabs. ICES Journal of Marine Science 73: 613-619.
278
-------
Kinlan, B.P. and Gaines, S.D. 2003. Propagule dispersal in marine and terrestrial environments: a community
perspective. Ecology 84: 2007-2020.
King, P.G., McGregor, A.R. and Whittet, J.D., 2011. The economic costs of sea-level rise to California beach
communities. Report from the California Department of Boating and Waterways and San Francisco State
University. California Department of Boating and Waterways, San Francisco State University. 82 pages.
Kleypas, J.A., Feely, R.A., Fabry, V.J., Langdon, C., Sabine, C.L., and Robbins, L.L. 2006. Impacts of ocean
acidification on coral reefs and other marine calcifiers: A guide for future research. Report of a workshop
sponsored by National Science Foundation, National Oceanic and Atmospheric Administration, and the
US Geological Survey and held April, 2005, St. Petersburg, Florida. Contribution No. 2897 from
NOAA/Pacific Marine Environmental Laboratory. 97 pages.
https: //www.researchKate.net/publication/248700866 Impacts of Ocean Acidification all Ree
fs and Other Marine Calcifiers A Guide for Future Research.
Ko, G.W.K., Dineshram, R., Camoanati, C., Chan, V.B.S., Havenhand, J. and Thiyagarajan, V. 2015. Interactive
effects of ocean acidification, elevated temperature, and reduced salinity on early-life stages of the
Pacific oyster. Environ. Sci. Technol. 48:10079-10088.
Kohl, W.T., McClure, T.I., and Miner, B.G. 2016. Decreased temperature facilitates short-term sea star wasting
disease survival in the keystone intertidal sea star Pisaster ochraceus. PLoS ONE 11(4): e0153670.
doi:10.1371/journal.pone.0153670.
Kriegler, E., Hall, J.W., Held, H., Dawson, R., and Schellnhuber, H.J. 2009. Imprecise probability assessment of
tipping points in the climate system. Proceedings of the National Academy of Sciences 106: 5041-5046.
Kroeker, K.J., Kordas, R.L., Crim, R., Hendriks, I.E., Ramajo, L., Singh, G.S., Duarte, C.M., Gattuso, J.P. 2013.
Impacts of ocean acidification of marine organisms: quantifying sensitivities and interaction with
warming. Global Change Biology 19: 1884-1896.
Kuhnert, P.M., Hayes, K., Martin, T.G., and McBride, M.F. 2009. Expert opinion in statistical models. In 18th
World IMACS Congress and MODSIM09 International Congress on Modelling and Simulation. Modelling
and Simulation Society of Australia and New Zealand and International Association for Mathematics and
Computers in Simulation: 4262-4268. http: //www.mssanz.org.au/modsim09/J2/kuhnert J2.pdf.
Kuncel, N.R., Klieger, D.M., Connelly, B.S., and Ones, D.S. 2013. Mechanical versus clinical data combination in
selection and admissions decisions: a meta-analysis. Journal of Applied Psychology 98: 1060-1072.
Kurihara, H., Matsui, M., Furukawa, H., Hayashi, M., and Ishimatsu, A. 2008. Long-term effects of predicted
future seawater C02 conditions on the survival and growth of the marine shrimp Palaemon pacificus.
Journal of Experimental Marine Biology and Ecology 367: 41-46.
Kynn, M. 2008. The 'heuristics and biases' bias in expert elicitation. Journal of the Royal Statistical Society: Series
A (Statistics in Society) 171: 239-264.
279
-------
Laffoley, D. and Baxter, J.M. (Eds.) 2016. Explaining ocean warming: Causes, scale, effects and consequences.
IUCN, Gland, Switzerland Full report: 456 pages,
http: //www.obs-vlfr.fr/~gattuso/files/Magnan_Gattuso_2016.pdf.
Landes, A. and Zimmer, M., 2012. Acidification and warming affect both a calcifying predator and prey but not
their interaction. Inter-Research Marine Ecology Progress Series 450: 1-10.
Lange, U. and Greve, W. 1997. Does temperature influence the spawning time, recruitment and distribution of
flatfish via its influence on the rate of gonadal maturation? German Journal of Hydrography 49: 251-
263.
Lawrence, A.J. and Soame, J.M. 2004. The effects of climate change on the reproduction of coastal
invertebrates. Ibis 146: 29-39.
Lea, R.N., and Rosenblatt, R.H. 2000. Observations on fishes associated with the 1997-98 El Nino off California.
CalCOFI Report 41:117-129.
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, Newport OR. EPA/600/R-
09/140. 225 pages, http: //nepis.epa.gov/Adobe/PDF/P1006Q2H.PDF.
Lee II, H. and Reusser, D.A., 2012. Atlas of nonindigenous marine and estuarine species in the north Pacific.
Office of Research and Development, National Health and Environmental Effects Research Laboratory,
EPA/600/R/12/631. 1915 pages.
Lee II, H., Marko, K., Hanshumaker, M., Folger, C., and Graham, R. 2015. User's guide & metadata to Coastal
Biodiversity Risk Analysis Tool (CBRAT): Framework for the systemization of life history and
biogeographic information. EPA Report. EPA/601/B-15/001. 123 pages.
http: //www.cbrat.org/pdfs/general/cbrat users guide and metadata.pdf.
Lee II, H., Reusser, D.A., Frazier, M.R., McCoy, L.M., Clinton, P.J., and Clough, J.S. 2014. Sea Level Affecting
Marshes Model (SLAMM) - New functionality for predicting changes in distribution of submerged
aquatic vegetation in response to sea level rise. Version 1.0. U.S. EPA, Office of Research and
Development, National Health and Environmental Effects Research Laboratory, Western Ecology
Division. EPA/600/R-14/007. 50 pages.
http: //warrenpinnacle.com/prof/SLAMM6/SLAMM 6.3 final release.pdf.
Loarie, S.R., Carter, B.E., Hayhoe, K., McMahon, S., Moe, R., Knight, C.A., and Ackerly, D.D. 2008. Climate change
and the future of California's endemic flora. PLoS ONE 3(6): e2502.doi:10.1371/journal.pone.0002502.
Lomolino, M.V. and Weiser, M.D. 2001. Towards a more general species-area relationship: diversity on all
islands, great and small. Journal of Biogeography 28: 431-445.
Lomolino, M.V., Riddle, B.R., and Brown, J.H. 1953. Biogeography, Third Edition. Massachusetts: Sinauer
Associates, Inc. 845 pages.
280
-------
Long, E.R., MacDonald, D.D., Smith, S.L., and Calder, F.D. 1995. Incidence of adverse biological effects within
ranges of chemical concentrations in marine and estuarine sediments. Environmental Management 19:
81-97.
Long, W.C., Swiney, K.M., Harris, C., Page, H.N., and Foy, R.J. 2013. Effects of ocean acidification on juvenile red
king crab (Paralithodes camtschaticus) and Tanner crab (Chionoecetes bairdi) growth, condition,
calcification and survival. PLoS ONE 8(4): e60959. doi:10.1371/journal.pone.0060959.
Long, W.C., Swiney, K.M., and Foy, R.J. 2016. Effects of high pC02 on Tanner crab reproduction and early life
history, Part 2: carryover effects on larvae from oogenesis and embryogenesis are stronger than direct
effects. ICES Journal of Marine Science 73: 836-848.
Lopez-Medellfn, X., Castillo, A. and Ezcurra, E. 2011. Contrasting perspectives on mangroves in arid
Northwestern Mexico: Implications for integrated coastal management. 2011. Ocean & Coastal
Management 54: 318-329.
Lord, J.P., Barry, J.P. and Graves, D. 2017. Impact of climate change on direct and indirect species interactions.
Marine Ecology Progress Series 571:1-11.
Love, M.S. 2011. Certainly more than you want to know about the fishes of the Pacific coast. San Francisco, CA:
Really Big Press. 672 pages.
Love, M.S., Mecklenburg, C.W., Mecklenburg, T.A., and Thorsteinson, L.K. 2005. Resource inventory of marine
and estuarine fishes of the West Coast and Alaska: a checklist of North Pacific and Arctic Ocean species
from Baja California to the Alaska-Yukon border. Dept. Int., USGS, Biol. Res. Div., Seattle, WA, OCS Study
MMS 2005-030 and USGS/NBII 2005-001. 276 pages.
Love, M.S. and Yoklavich, M.M. 2006. Deep rock habitats. The ecology of marine fishes: California and adjacent
waters (Eds. Allen, L.G., Pondella, D.J., and Horn, M.H.) 253-266. University of California Press. Berkley,
California. 670 pages.
Love, M.S., Yoklavich, M., Thorsteinson, L., and Butler, J. 2002. The rockfishes of the northeast Pacific. Berkley
and Monterey, California: University of California Press and Monterey Bay Aquarium Press. 416 pages.
Lovell, L.L. and Cadien, D.C. 2013. Coastal Biogeographic Risk Analysis Tool (CBRAT): Further contributions to
CBRAT data acquisition and correction: Second Expert Workshop 14-16 May, 2013. Los Angeles, CA:
Southern California Association of Marine Taxonomists. 13 pages.
Lucey, N.M., Lombardi, C., DeMarchi, L., Schulze, A., Gambi, M.C., and Calosi, P. 2015. To brood or not to brood:
are marine invertebrates that protect their offspring more resilient to ocean acidification? Scientific
Reports 5: 1-7 12009; doi: 10.1038/srepl2009.
Maclean, I.M.D. and Wilson, R.J. 2012. Recent ecological responses to climate change support predictions of
high extinction risk. Proceedings of the National Academy of Sciences 108: 12337-12342.
281
-------
Malcolm, J.R., Liu, C., Neilson, R.P., Hansen, L., and Hannah, L. 2006. Global warming and extinctions of endemic
species from biodiversity hotspots. Conservation Biology 20: 538-548.
Marcot, B.G. and Molina, R. 2007. Special considerations for the science, conservation, and management of rare
or little-known species. Conservation of Rare or Little-Known Species: Biological, Social, and Economic
Considerations. (Eds. Raphael, G.M. and Molina, R.). 93-124. Island Press, Washington D.C. 389 pages.
Mastrandrea, M.D., Field, C.B., Stocker, T.F., Edenhofer, O., Ebi, K.L., Frame, D.J., Held, H., Kriegler, E., Mach, K.J.,
Matschoss, P.R., Plattner, B.-K., Yohe, G.W., and Zwiers, F.W. 2010. Guidance note for lead authors of
the IPCC fifth assessment report on consistent treatment of uncertainties. Intergovernmental Panel on
Climate Change (IPCC), 7 pages.
https: //www.ipcc.ch/pdf/supporting-material/uncertaintv-guidance-note.pdf.
Mathis, J.T., Cooley, S.R., Lucey, N., Colt, S., Ekstrom, J., Hurst, T., Hauri, C., Evans, W., Cross, J.N., and Feely, R.A.
2015. Ocean acidification risk assessment for Alaska's fishery sector. Progress in Oceanography 136: 71-
91.
Matozzo, V., Chinellato, A., Munari, M., Finos, L., Bressan, M., and Marin, M.G., 2012. First evidence of
immunomodulation in bivalves under seawater acidification and increased temperature. PLoS ONE 7(3):
e33820. doi:10.1371/journal.pone.0033820
Maul, G.A. and Martin, D.M. 1993. Sea level rise at Key West, Florida, 1846-1992: America's longest instrument
record? Geophysical Research Letters 20: 1955-1958.
McAfee, A. 2013. Big data's biggest challenge? Convincing people NOT to trust their judgment. Harvard Business
Review: 9. https: //hbr.org/2013/12/big-datas-biggest-challenge~convincing-people-not-to-trust-
their-iudgment#comment-1156190677.
McBride, M.F., Fidler, F., and Burgman, M.A. 2012. Evaluating the accuracy and calibration of expert predictions
under uncertainty: predicting the outcomes of ecological research. Diversity and Distributions 18: 782-
794.
McCune, B. 2011. Nonparametric multiplicative regression for habitat modeling.
http: //www.pcord.com/NPMRintro.pdf.
McDowell, W.G., McDowell, W.H., and Byers, J.E. 2017. Mass mortality of a dominant invasive species in
response to an extreme climate event: implications for ecosystem function. Limnology and
Oceanography 62: 177-18.
McGurk, M.D. 1984. Effects of delayed feeding and temperature on the age of irreversible starvation and on the
rates of growth and mortality of Pacific herring larvae. Marine Biology 84: 13-26.
282
-------
Mclvor, A.L., Spencer, T., Moller, I. and Spalding. M. 2013. The response of mangrove soil surface elevation to
sea level rise. Natural Coastal Protection Series: Report 3. Cambridge Coastal Research Unit Working
Paper 42. Published by The Nature Conservancy and Wetlands International. 59 pages.
https://repository.tudelft.nI/islandora/obiect/uuid:6655e6c9~14c8~466Q~92.62~
7d42e6bf73el?collection=research.
McKee, K.L., Cahoon, D.R, and Feller, I.C. 2007. Caribbean mangroves adjust to rising sea level through biotic
controls on change in soil elevation. Global Ecology and Biogeography 16: 545-556.
McKinney, M.L. 1997. How do rare species avoid extinction? A paleontological review. The Biology of Rarity (Eds.
Kunin, W.E. and Gaston, K.J.): 110-129. Chapman & Hall, London, UK. 280 pages.
McLaughlin, J.F., Hellmann, J.J., Boggs, C.L., and Ehrlich, P.R. 2002. Climate change hastens population
extinctions. Proceedings of the National Academy of Sciences of the United States of America 99: 6070-
6074.
Mecklenburg, C.W., Mecklenburg, T.A., and Thorsteinson, L.K. 2002. Fishes of Alaska. American Fisheries Society,
Bethesda, Maryland. 1116 pages.
Meehl, G.A., Covey, C., Delworth, T., Latif, M., McAvaney, B., Mitchell, J.F.B., Stouffer, R.J., and Taylor, K.E. 2007.
The WCRP CMIP3 multimodel dataset - a new era in climate change research. Bulletin of the American
Meteorological Society 88: 1383-1394.
Mellin, C., Russell, B.D., Connell, S.D., Brook, B.W., and Fordham, D.A. 2012. Geographic range determinants of
two commercially important marine molluscs. Diversity and Distributions 18: 133-146.
Meseck, S.L., Alix, J.H., Swiney, K.M., Long, W.C., Wikfors, G.H., and Foy, R.J. 2016. Ocean acidification affects
hemocyte physiology in the Tanner Crab (Chionoecetes bairdi). PLoS ONE 11(2):
e0148477.doi:10.1371/journal. pone.0148477.
Metzger, R., Sartoris, F.J., Langenbuch, M., and Portner, H.O. 2007. Influence of elevated C02 concentrations on
thermal tolerance of the edible crab Cancer pagurus. Journal of Thermal Biology 32: 144-151.
Meyer, C.A. 2013. Evaluating habitat vulnerability and sustainability of urban seagrass resources to sea level rise.
Graduate Theses and Dissertations, University of South Florida, 147 pages.
http: //scholarcommons.usf.edu/etd/4918.
Miller, J.J., Maher, M., Bohaboy, E., Friedman, C.S., and McElhany, P. 2016. Exposure to low pH reduces survival
and delays development in early life stages of Dungeness crab (Cancer magister). Marine Biology
163: 118.
Miller, S.H., Zarate, S., Smith, E.H., Gaylord, B., Hosfelt, J.D., and Hill, T.M. 2014. Effect of elevated pC02 on
metabolic responses of porcelain crab (Petrolisthes cinctipes) larvae exposed to subsequent salinity
stress. PLoS ONE 9(10): el09167. doi: 10.1371/journal.pone.0109167.
283
-------
Molinos, J.G., Halpern, B.S., Schoeman, D.S., Brown, C.J., Kiessling, W., Moore, P.J., Pandolfi, J.M., Poloczanska,
E.S., Richardson, A.J., and Burrows, M.T. 2016. Climate velocity and the future global redistribution of
marine biodiversity. Nature Climate Change 6: 83-88.
Molnar, J.L., Gamboa, R.L., Revenga, C., and Spalding, M.D. 2008. Assessing the global threat of invasive species
to marine biodiversity. Frontiers in Ecology and the Environment 6: 485-492.
Monte-Luna, P., Uuch-Belda, D., Serviere-Zaragoza, E., Carmona, R., Reyes-Bonilla, H., Aurioles-Gamboa, D.,
Castro-Aguirre, J.L., del Proo, S.A.G., Trujillo—Milan, O., and Brook, B.W. 2007. Marine extinctions
revisited. Fish and Fisheries 8: 107-122.
Moran, X.A.G., Lopez-Urrutia, A., Calvo-Dfaz, A., and Li, W.K.W. 2010. Increasing importance of small
phytoplankton in a warmer ocean. Global Change Biology 16: 1137.
Morissette, S. and Himmelman, J.H. 2000. Subtidal food thieves: interactions of four invertebrate
kleptoparasites with the sea star Leptasterias polaris. Animal Behavior 60: 531-543.
Morris, J.T., Sundareshwar, P.V., Nietch, C.T., Kjerfve, B., and Cahoon, D.R. 2002. Responses of coastal wetlands
to rising sea level. Ecology 83: 2869-2877.
Morrison, W., Nelson. M., Howard, J., Teeters, E., Hare, J.A., Griffis, R., Scott, J.D., and Alexander, M.A. 2015.
Methodology for assessing the vulnerability of marine fish and shellfish to a changing climate. NOAA
Technical Memorandum NMFS-OSF-3. U.S. Department of Commerce, National Oceanic and
Atmospheric Administration, National Marine Fisheries Service. 48 pages.
Morueta-Holme, N., Fl0jgaard, C., and Svenning, J.-C. 2010. Climate change risks and conservation implications
for a threatened small-range mammal species. PLoS ONE 5(4): el0360.
doi:10.1371/journal.pone.0010360.
Moyle, P.B., Kiernan, J.D., Crain, P.K., and Quinones, R.M. 2013. Climate change vulnerability of native and alien
freshwater fishes of California: A systematic assessment approach. PLoS ONE 8(5): e63883
doi: 101371/journal.pone.0063883.
Mumford, T.M. 2007. Kelp and eelgrass in Puget Sound: Seattle, Wash., U.S. Army Corps of Engineers, Puget
Sound Nearshore Partnership Report No. 2007-05. 27 pages.
http://www.pueetsoundnearshore.org/technical papers/kelp.pdf.
Munday, P.L., Crawley, N.E, and Nilsson, G.E. 2009. Interacting effects of elevated temperature and ocean
acidification on the aerobic performance of coral reef fishes. Marine Ecology Progress Series 388:235-
242.
Munday, P.L., Kingsford, M., O'Callaghan, M., and Donelson, J.M. 2008. Elevated temperature restricts growth
potential of the coral reef fish Acanthochromis polyacanthus. Coral Reefs 27: 927-931.
Musick, J.A. 1999. Criteria to define extinction risk in marine fishes. Fisheries 24: 6-14.
284
-------
Musick, J.A., Harbin, M.M., Berkeley, S.A., Burgess, G.H., Eklund, A.M., Findley, L., and Wright, S.G. 2000. Marine,
estuarine, and diadromous fish stocks at risk of extinction in North America (exclusive of Pacific
salmonids). Fisheries 25: 6-30.
Myers, D., 2010, Shoreline development on Puget Sound. Puget Sound Shorelines and the Impacts of
Armoring—Proceedings of a State of the Science Workshop, May 2009. (Eds. Shipman, H., Dethier, M.N.,
Gelfenbaum, G., Fresh, K.L., and Dinicola, R.S.). 43-48. U.S. Geological Survey Scientific Investigations
Report 2010-5254. 262 pages.
Nabholz, J.V. 1991. Environmental hazard and risk assessment under the United States Toxic Substances Control
Act. The Science of the Total Environment 109/110 649-665.
National Research Council. 2012. Sea-level Rise for the Coasts of California, Oregon, and Washington: Past,
Present, and Future. The National Academies Press, Washington D.C. 201 pages.
Nelson, W.G., Hyland, J.L., Lee II, H., Cooksey, C.L., Lamberson, J.O., Cole, F.A., and Clinton, P.J. 2008. Ecological
Condition of Coastal Ocean Waters along the U.S. Western Continental Shelf: 2003. EPA 620/R-08/001,
U.S. EPA, Office of Research and Development, National Health and Environmental Effects Research
Laboratory, Western Ecology Division, Newport OR, 97365; and NOAA Technical Memorandum NOS
NCCOS 79, NOAA National Ocean Service, Charleston, SC 29412-9110. 137 pages.
Nelson, W.G., Lee II, H., and Lamberson, J.O. 2005. Condition of Estuaries of California for 1999: A Statistical
Summary. EPA/620/R-05/004. U.S. Environmental Protection Agency, Office of Research and
Development, Washington, DC. 154 pages.
Nelson, W.G., Lee II, H., Lamberson, J.O., Cole, F.A., Weilhoefer, C.L., and Clinton, P.J. 2007. The condition of
tidal wetlands of Washington, Oregon and California -2002. U.S. Environmental Protection Agency,
Washington D.C. EPA620/R-07/002. 102 pages.
Nelson, W.G., Lee II, H., Lamberson, J.O., Engle, V., Harwell, L., and Smith, L.M. 2004. Condition of estuaries of
western United States for 1999: a statistical summary. U.S. Environmental Protection Agency,
Washington D.C EPA/620/R-04/200. 154 pages.
Neuheimer, A.B., Thresher, R.E., Lyle, J.M., and Semmens, J.M. 2011. Tolerance limit for fish growth exceeded by
warming waters. Nature Climate Change 1: 110-113.
Nicholls, R.J., Hanson, S.E., Lowe, J.A., Warrick, R.A., Lu, X., Long, A.J. and Carter, T.R. 2011: Constructing sea-
level scenarios for impact and adaptation assessment of coastal area: a guidance document. Supporting
Material, Intergovernmental Panel on Climate Change Task Group on Data and Scenario Support for
Impact and Climate Analysis (TGICA). 47 pages.
Nilsson, G.E., Crawley, N., Lunde, I.G., and Munday, P.L. 2009. Elevated temperature reduced the respiratory
scope of coral reef fishes. Global Change Biology 15: 1405-1412.
285
-------
NOAA. 2012. Status of stocks 2012 annual report to congress on the status of U. S. fisheries. National Oceanic
and Atmospheric Association. 8 pages.
http: //www.nmfs.noaa.gov/sfa/statusoffisheries/2012/2012 SOS RTC.pdf.
NOAA. 2013. Shoreline Assessment Manual. 4th Edition. U.S. Dept. of Commerce. Seattle, WA: Emergency
Response Division, Office of Response and Restoration, National Oceanic and Atmospheric
Administration. 73 pp + appendices.
Noisette, F., Comtet, T., Legrand, E., Bordeyne, F., Davoult, D., and Martin, S. 2014. Does encapsulation protect
embryos from the effects of ocean acidification? the example of crepidula fornicata. PLoS ONE 9(3):
e93021. doi:10.1371/journal.pone.0093021
Norman, K., Sepez, J., Lazrus, H., Milne, N., Package, C., Russell, S., Grant, K., Lewis, R.P., Primo, J., Springer, E.,
Styles, M., Tilt, B., and Vaccaro, I. 2007. Community profiles for West Coast and North Pacific fisheries-
Washington, Oregon, California, and other U.S. states. U.S. Department of Commerce, NOAA Tech.
Memo. NMFS-NWFSC-85. 602 pages.
https://www.nwfsc.noaa.gov/assets/25/499 01082008 153910 CommunityProfilesTM85WebFinalSA.p
Ml
NRC. 2012. National Research Council. Sea-Level Rise for the Coasts of California, Oregon, and Washington:
Past, Present, and Future. The National Academies Press, Washington, DC. 201 pages,
http: //www.ecy.wa.gov/climatechange/docs/ipa_slr_nrcfullreport.pdf.
Nygren, A. 2014. Cryptic polychaete diversity: a review. Zoologica Scripta 43: 172-183.
Olden, J.D., Poff, N.L., and Bestgen, K.R. 2008. Trait synergisms and the rarity, extirpation, and extinction risk of
desert fishes. Ecology 89: 847-856.
O'Leary, R.A., Choy, S.L., Murray, J.V., Kynn, M., Denham, R., Martin, T.G., and Mengersen, K. 2008. Comparison
of three expert elicitation methods for logistic regression on predicting the presence of the threatened
brush-tailed rock-wallaby Petrogale penicillata. Environmetrics 20: 379-98.
Pacific States Marine Fisheries Commission (PSMFC). 2004. Risk Assessment for the Pacific Groundfish FMP.
Page 15. ftp://ftp.streamnet.org/pub/marinehabitat/DraftEFHEIS/AppendixA/crav4.pdf
Paganini, A.W., Miler, N.A., and Stillman, J.H. 2014. Temperature and acidification variability reduce
physiological performance in the intertidal zone porcelain crab Petrolisthes cinctipes. Journal of
Experimental Biology 217: 3974-3980.
Pansch, C., Schaub. I., Havenhand, J., and Wahl, M. 2014. Habitat traits and food availability determine the
response of marine invertebrates to ocean acidification. Global Change Biology 20: 765-777.
Parkinson, R.W. 1989. Decelerating Holocene sea level rise and its influence on southwest Florida coastal
evolution: A transgressive/regressive stratigraphy. Journal of Sedimentary Petrology 59: 960-972.
286
-------
Parker, S.J., Berkeley, S.A., Golden, J.T., Gunderson, D.R., Heifetz, J., Hixon, M.A., Larson, R., Leaman, B.M., Love,
M.S., Musick, J.A., O'Connell, V.M., Ralston, S., Weeks, H.J., and Yoklavich, M.M. 2000. Management of
Pacific Rockfish. Fisheries 25: 22-30.
Parris, A., Bromirski, P., Burkett, V., Cayan, D., Culver, M., Hall, J., Horton, R., Knuuti, K., Moss, R., Obeysekera, J.,
Sallenger, A., and Weiss, J. 2012. Global Sea Level Rise Scenarios for the U.S. National Climate
Assessment. NOAA Technical Report OAR CPO
https://scenarios.globalchange.gov/sites/default/files/NOAA SLR r3 O.pdf
Payne, M.C., Brown, C.A., Reusser, D.A., and Lee II, H. 2012a. Ecoregional analysis of nearshore sea-surface
temperature in the North Pacific. PLoS ONE 7(1): e30105. doi:10.1371/journal.pone.0030105.
Payne, M.C., Reusser, D.A., and Lee II, H. 2012b. Moderate-resolution sea surface temperature data and
seasonal pattern analysis for the Arctic Ocean ecoregions: U.S. Geological Survey Open-File Report
2011-1246, 20 pages, http: //pubs.usgs.gov/of/2011/1246/.
Payne, M.C., Reusser, D.A., Lee II, H., and Brown, C.A. 2011. Moderate-resolution sea surface temperature data
for the nearshore North Pacific: U.S. Geological Survey Open-File Report: 2010-1251.
http: //pubs.usgs.gov/of/201Q/12.51/.
Pearson, R.G. 2010. Species' Distribution Modeling for Conservation Educators and Practitioners. Synthesis.
American Museum of Natural History, Lessons in Conservation. Pages 54-89.
https://pdfs.semanticscholar.org/66db/947eela6ab91c408f489dl7cfb6eQ68931a6.pdf.
Pearson, R.G., Stanton, J.C., Shoemaker, K.T., Aiello-Lammens, M.E., Ersts, P.J., Horning, N., Fordham, D.A.,
Raxworthy, C.J., Ryu, H.Y., McNees, J., and Akgakaya, H.R. 2014. Life history and spatial traits predict
extinction risk due to climate change. Nature Climate Change 4: 217-221.
Pechenik, J.A. 1999. On the advantages and disadvantages of larval stages in benthic marine invertebrate life
cycles. Marine Ecology Progress Series 177: 269-297.
Pedersen T.M., Hansen J.L., Josefson, A.B., and Hansen, B.W. 2008. Mortality through ontogeny of soft-bottom
marine invertebrates with planktonic larvae. Journal of Marine Systems 73:185-207.
Perry, A.L., Low, P.J., Ellis, J.R., and Reynolds, J.D. 2005. Climate change and distribution shifts in marine fishes.
Science 308: 1912-1915.
PFMC (Pacific Fishery Management Council). 2004. 10-meter Bathymetry Polygons for Washington, Oregon and
California. Downloadable data: Data Source Information: The Active Tectonics and Seafloor Mapping
Lab, College of Oceanic and Atmospheric Sciences, Oregon State University developed the data for
Oregon and Washington. The Center for Habitat Studies, Moss Landing Marine Laboratories developed
the data for California. Data compiled by TerraLogic GIS, Inc. http: //marinehabitat.psmfc.org/.
287
-------
PFMC (Pacific Fishery Management Council). 2005. Pacific Coast Groundfish Fishery Management Plan for the
California, Oregon, and Washington groundfish fishery. Appendix B, Part 1. Assessment methodology for
groundfish essential fish habitat. Pacific Fishery Management Council. Portland, OR.
http://www.westcoast.fisheries.noaa.gov/publications/nepa/groundfish/groyndfish efh eis/fmp-
appendix-bl.pdf.
Philippart, C., van Aken, H., Beukema, J., Bos, O., Cadee, G., and Dekker, R. 2003. Climate-related changes in
recruitment of the bivalve Macoma balthica. Limnology and Oceanography 48: 2171-2185.
Piepenburg, D., Archambault, P., Ambrose Jr., W.G., Blanchard, A.L., Bluhm, B.A., Carroll, M.L., Conlan, K.E.,
Cusson, M., Feder, H.M., Grebmeier, J.M., Jewett, S.C., Levesque, M., Petryashev, V.V., Sej'r, M.K.,
Sirenko, B.I., Wtodarska-Kowalczuk, M. 2011. Towards a pan-Arctic inventory of the species diversity of
the macro- and megabenthic fauna of the Arctic shelf seas. Marine Biodiversity 41: 51-70.
Pilgrim, E.M., Blum, M.J., Reusser, D.A., Lee II, H., and Darling, J.A. 2013. Geographic range and structure of
cryptic genetic diversity among Pacific North American populations of the non-native amphipod
Grandidierella japonica. Biological Invasions 15: 2415-2428.
Pinsky, M.L., Worm, B., Fogarty, M.J., Sarmiento, JL., and Levin, S.A. 2013. Marine taxa track local climate
velocities. Science 341: 1239-1242.
Planque, B. and Fredou, T. 1999. Temperature and the recruitment of Atlantic cod (Gadus morhua). Canadian
Journal of Fisheries and Aquatic Sciences 56: 2069-2077.
Planque, B., Bellier, E., and Loots, C. 2011. Uncertainties in projecting spatial distributions of marine populations.
ICES Journal of Marine Science 68: 1045-1050.
Polidoro, B.A., Carpenter, K.E., Collins, L., Duke, N.C., Ellison, A.M., et al. 2010. The Loss of species: Mangrove
extinction risk and geographic areas of global concern. PLoS ONE 5(4): el0095.
doi:10.1371/journal.pone.0010095.
Polidoro, B.A., Brooks, T., Carpenter, K.E., Edgar, G.J., Henderson, S., Sanciangco, J. and Robertson, D.R. 2012.
Patterns of extinction risk and threat for marine invertebrates and habitat-forming species in the
Tropical Eastern Pacific. Marine Ecology Progress Series 448: 93-104.
Pondella, D.J. and Allen, M.J. 2001. Proceedings of special symposium: new and rare fish and invertebrate
species to California during the 1997-1998 El Nino sponsored by the Southern California Academy of
Sciences May 20, 2000. Bulletin Southern California Academy of Science 100: 129-130.
Poore, R.Z., Williams, R.S., Jr., and Tracey, C. 2000. Sea level and climate: U.S. Geological Survey Fact Sheet 002-
00, 2 pages, https://pubs.usgs.gov/fs/fs2-00/.
Portner, H.O. 2008. Ecosystem effects of ocean acidification in times of ocean warming: A physiologist's view.
Marine Ecology Progress Series 373: 203-217.
288
-------
Portner, H.O. and Knust, R. 2007. Climate change affects marine fishes through the oxygen limitation of thermal
tolerance. Science 315: 95-97.
Portner, H.O., Bock, C., Knust, R., Lannig, G., Lucassen, M., Mark, F.C., and Sartoris, F.J. 2008. Cod and climate in
a latitudinal cline: physiological analyses of climate effects in marine fishes. Climate Research 37: 253-
270.
Portner, H.O., Karl, D.M., Boyd, P.W., Cheung, W., Lluch-Cota, S.E., Nojiri, Y., Schmidt, D.N., Zavialov, P.O.,
Alheit, J., Aristegui, J., Armstrong, C., Beaugrand, G., Belkovich, V., Bowler, C., Brewer, P., Church, M.,
Cooley, S.R., del Monte-Luna, P., Edwards, M., Flint, M., Follows, M.J., Frolicher, T., Fulton, E.A., Gattuso,
J.P., Hoegh-Guldberg, O., Hofmann, E.E., Knoll, A.H., Levin, L.A., Menzel, L., Moloney, C.L., Perry, R.I.,
Poloczanska, E.S., Roberts, J.M., Rost, B., Sarmiento, J.L., Sedlacek, J., Storch, D., Wiencke, C., and
Wittmann, A.C. 2014. Ocean systems. Climate Change 2014: Impacts, Adaptation and Vulnerability. Part
A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change (Eds. Field, C., Barros V., Dokken D., Mach K., Mastrandrea,
P., and White, L.). Cambridge University Press, Cambridge, UK and New York, NY, USA. Pages 411-484.
http: //epic.awi.de/37501/.
Portner, H.O., Karl, D.M., Boyd, P.W., Cheung, W.W.L., Lluch-Cota, S.E., Nojiri, Y., Schmidt, D.N., and Zavialov,
P.O. 2014. Ocean systems. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global
and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change (Eds. Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J.,
Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S.,
Levy, A.N., MacCracken, S., Mastrandrea, P.R. and White, L.L.) Cambridge, UK and New York, NY:
Cambridge University Press. 411-484. https://www.ipcc.ch/pdf/assessment-report/ar5/wg2/WGIIAR5-
Chap6 FINAL.pdf.
Prather, C.M., Pelini, S.L., Laws, A., Rivest, E., Woltz, M., Bloch, C.P., Del Toro, I., Ho, C.-K., Kominoski, J.,
Newbold, T.A.S., Parsons, S. and Joern, A. 2013. Invertebrates, ecosystem services and climate change.
Biological Reviews 88: 327-348.
Proshutinsky, A.I., Ashik, M., Dvorkin, E.N., Hakkinen, S., Krishfield, R.A., and Peltier, W.R. 2004. Secular sea level
change in the Russian sector of the Arctic Ocean. Journal of Geophysical Research: Oceans 109.
doi: 10.1029/2003JC002007.
PSMFC (Pacific States Marine Fisheries Commission). 2004. Risk Assessment for the Pacific Groundfish FMP. p.
15. ftp: //ftp.streamnet.org/pub/marinehabitat/DraftEFHEIS/AppendixA/crav4.pdf.
Puget Sound Action Team. 2007. 2007 Puget Sound Update: Ninth Report of the Puget Sound Assessment and
Monitoring Program. Olympia, Washington: Puget Sound Action Team. 260 pages.
http: //wdfw.wa.gov/publications/01051/wdfw01051.pdf.
Punt, A.E., Foy, R.J., Dalton, M.G., Long, W.C., and Swiney, K.M. 2016. Effects of long-term exposure to ocean
acidification conditions on future southern Tanner crab (Chionoecetes bairdi) fisheries management.
ICES Journal of Marine Science 73: 849-864.
289
-------
Ramsar. 2008. Resolution X.2.0. Biogeographic regionalization in the application of the strategic framework for
the List of Wetlands of International Importance: scientific and technical guidance. Ramsar, Changwon,
Republic of Korea, http://archive.ramsar.ore/pdf/res/kev res x 20 e.pdf.
Raphael, M.G. and Molina, R. (Eds.) 2007. Conservation of Rare or Little Known Species. Biological, Social, and
Economic Considerations. Island Press. 375 pages.
Rastrick, S.P.S., Calosi, P., Calder-Potts, R., Foggo, A., Nightingale, G., Widdicombe, S., and Spicer, J.I. 2014. Living
in warmer, more acidic oceans retards physiological recovery from tidal emersion in the velvet
swimming crab, Necora puber. Journal of Experimental Biology 217: 2499-2508.
Reusser, D.A. and Lee II, H. 2008. Predictions for an invaded world: a strategy to predict the distribution of
native and non-indigenous species at multiple scales. ICES Journal of Marine Science 65: 742-745.
Reusser, D.A. and Lee II, H. 2011. Evolution of natural history information in the 21st century - developing an
integrated framework for biological and geographical data. Journal of Biogeography 38: 1225-1239.
Reusser, D.A., Frazier, M.L., Loiselle, R.A., Lee II, H., and Thorsteinson, L.K. 2016. Chapter 5. Arctic climate
change—a tale of two cods. Alaska Arctic Marine Fish Ecology Catalog (Thorsteinson, L.K., and Love,
M.S., Eds.). U.S. Geological Survey Scientific Investigations Report 2016-5038 (OCS Study, BOEM 2016-
048). Pages 659-677.
Revell, D.L., Battalio, R., Spear, B., Ruggiero, P., and Vandever, J. 2011. A methodology for predicting future
coastal hazards due to sea-level rise on the California Coast. Climatic Change 109: 251-276.
Reynolds, J.D., Dulvy, N.K., Goodwin, N.B., and Hutchings, J.A. 2005. Biology of extinction risk in marine fishes.
Proceedings of the Royal Society B- Biological Sciences 272: 2337-2344.
Rahmstorf, S. 2007. A semi-empirical approach to projecting future sea-level rise. Science. 315: 368-370.
Richards, Z.T., Garcia, R.A., Wallace, C.C., Rosser, N.L., and Muir, P.R. 2015. A diverse assemblage of reef corals
thriving in a dynamic intertidal reef setting (Bonaparte Archipelago, Kimberley, Australia). PLoS ONE
10(2): e0117791. doi:10.1371/journal.pone.0117791.
Ries, J.B. 2011. Skeletal mineralogy in a high-C02 world. Journal of Experimental Marine Biology and Ecology
403:54-64.
Ries, J.B., Cohen, A.L., and McCorkle, D.C., 2009. Marine calcifiers exhibit mixed responses to C02-induced ocean
acidification. Geology 37: 1131-1134.
Roberts, C.M. and Hawkins, J.P. 1999. Extinction risk in the sea. Trends in Ecology & Evolution 14: 241-246.
Robertson, D.R. 2008. Global biogeographical data bases on marine fishes: caveat emptor. Diversity and
Distributions 14: 891-892.
Rodrigues, A.S.L. and Gaston, K.J. 2002. Rarity and conservation planning across geopolitical units. Conservation
Biology 16: 674-682.
290
-------
Romsos, C., Goldfinger, C., Robison, R., Milstein, R.L., and Wakefield, W.W. 2007. Development of a regional
seafloor surficial geologic habitat map for the continental margins of Oregon and Washington, USA.
Mapping the Seafloor for Habitat Characterization (Eds. Todd, B.J. and Greene, H.G.) 229. Geological
Association of Canada, Special Paper 47: 219-243.
Ross, P.M., Parker, L., and Byrne, M. 2016. Transgenerational responses of molluscs and echinoderms to
changing ocean conditions. ICES Journal of Marine Science 73: 537-549.
Ross, P.M., Parker, L., O'Connor, W.A., and Bailey, E.A. 2011. The impact of ocean acidification on reproduction,
early development and settlement of marine organisms. Water 3: 1005-1030.
Rowe, G. and Wright, G. 2011. The Delphi technique: Past, present, and future prospects—introduction to the
special issue. Technological Forecasting and Social Change 78: 1487-1490.
Rudnick, D., Veldhuizen, T., Tullis, R., Culver, C., Hieb, K., and Tsukimura, B. 2005. A life history model for the San
Francisco estuary population of the Chinese mitten crab, Eriocheir sinensis (Decapoda: Grapsoidea).
Biological Invasions 7: 333-350.
Ruiz, G.M., Fofonoff, P.W., Carlton, J.T., Wonham, M.J., and Hines, A.H. 2000. Invasion of coastal marine
communities in North America: apparent patterns, processes, and biases. Annual Review of Ecology and
Systematics 31: 481-531.
Rumrill, S.S., 1990. Natural mortality of marine invertebrate larvae. Ophelia 32 (1-2), 163-198.
Rutherford, E.S. and Houde, E.D. 1996. The influence of temperature on cohort-specific growth, survival, and
recruitment of striped bass, Morone saxatilis, larvae in Chesapeake Bay. Oceanographic Literature
Review 43: 73.
Salway, R. and Shaddick, G. No date. Implementation of Qualitative Uncertainty Guidance: A Worked Example.
http://www. integrated-
assessment.eu/eu/sites/default/files/Example%20of%20qualitative%20uncertaintv%20analvsis.pdf.
Scott, J.D., Alexander, M.A., Murray, D.R., Swales, D., and Eischeid, J. 2016. The climate change web portal - a
system to access and display climate and earth system model output from the CMIP5 archive. Bulletin of
the American Meteorological Society 97: 523-530.
Seddon, A.W., Froyd, C.A., Leng, M.J., Milne, G.A., and Willis, K.J. 2011. Ecosystem resilience and threshold
response in the Galapagos Coastal Zone. PLoS ONE E 6(7): e22376. doi:10.1371/journal.pone.0022376.
Selkoe, K.A. and Toonen, R.J. 2011. Marine connectivity: a new look at pelagic larval duration and genetic
metrics of dispersal. Marine Ecology Progress Series 436: 291-305.
Sexton, J.P., Mclntyre, P.J., Angert, A.L., and Rice, K.J. 2009. Evolution and ecology of species range limits. Annual
Review of Ecology, Evolution, and Systematics 40: 415-436.
291
-------
Shaughnessy, F.J., Gilkerson, W., Black, J.M., Ward, D.H., and Petrie, M. 2012. Predicted eelgrass response to sea
level rise and its availability to foraging black Brant in Pacific Coast estuaries. Ecological Applications 22:
1743-1761.
Schiffer, M., Harms, L., Lucassen, M., Mark, F.C., Portner, H.O., Storch, D. 2014. Temperature tolerance of
different larval stages of the spider crab Hyas araneus exposed to elevated pC02. Frontiers in Zoology
11:87 DOI 10.1186/sl2983-014-0087-4.
Sheridan, J.A. and Bickford, D. 2011. Shrinking body size as an ecological response to climate change. Nature
Climate Change 1: 401-406.
Short, F.T. and Neckles, H.A. 1999. The effects of global climate change on seagrasses. Aquatic Botany 63: 169-
196.
Short, F.T., Kosten, S., Morgan, P.A., and Malone, S. 2016. Impacts of climate change on submerged and
emergent wetland plants. Aquatic Botany 135: 3-17.
Small, D., Calosi, P., White, D., Spicer, J.I., and Widdicombe, S. 2010. Impact of medium-term exposure to C02
enriched seawater on the physiological functions of the velvet swimming crab Necora puber. Aquatic
Biology 10: 11-21.
Small, D.P., Calosi, P., Boothroyd, D., Widdicombe, S., and Spicer, J.I. 2016. The sensitivity of the early benthic
juvenile stage of the European lobster Homarus gammarus (L.) to elevated pC02 and temperature.
Marine Biology 163: 53. doi:10.1007/s00227-016-2834-x
Snedaker, S.C., Meeder, J.F., Ross, M.S., and Ford, R.G 1994. Discussion of Ellison, Joanna C. and Stoddart, David
R., 1991. Mangrove ecosystem collapse during predicted sea-level rise: Holocene analogues and
implications. Journal of Coastal Research, 7(1), 151-165. Journal of Coastal Research 10: 497-498.
Snover, A.K., Mauger, G.S., Binder, L.W., Krosby, M., and Tohver, I. 2013. Climate change impacts and adaptation
in Washington State: Technical summaries for decision makers. State of Knowledge Report prepared for
the Washington State Department of Ecology. Climate Impacts Group, University of Washington, Seattle.
Sodhi, N.S. and Ehrlich, P.R. 2010. Conservation Biology for All. Oxford University Press. 350 pages.
Somero, G.N. 2012. The physiology of global change: linking patterns to mechanisms. Annual Review of Marine
Science 4: 39-61.
Sorte, C.J.B., Williams, S.L., and Carlton, J.T. 2010. Marine range shifts and species introductions: comparative
spread rates and community impacts. Global Ecology and Biogeography 19: 303-316. Snover, A.K, G.S.
Mauger, L.C. Whitely Binder, M. Krosby, and I. Tohver. 2013. Climate Change Impacts and Adaptation in
Washington State: Technical Summaries for Decision Makers. State of Knowledge Report prepared for
the Washington State Department of Ecology. Climate Impacts Group, University of Washington, Seattle.
292
-------
Spalding, M., Fox, H.H., Allen, G.R., Davidson, N., Ferdana, Z.A., Finlayson, M., Halpern, B.S., Jorge, M.A.,
Lombana, A., Lourie, S.A., Martin, K.D., McManus, E., Molnar, J., Recchia, C.A., and Robertson, J. 2007.
Marine ecoregions of the world: a bioregionalization of coastal and shelf areas. Bioscience 57: 573-583.
Spalding, M., Kainuma, M., and Collins, L. 2010. World Atlas of Mangroves. Washington DC: Earthscan. 336
pages.
Spicer, J.I., Raffo, A. and Widdicombe, S. 2007. Influence of C02-related seawater acidification on extracellular
acid-base balance in the velvet swimming crab Necora puber. Marine Biology 151: 1117-1125.
Steyerberg, E.W., Eijkemans, M.J., Harrell Jr., F.E., and Habbema, J.D.F. 2000. Prognostic modelling with logistic
regression analysis: a comparison of selection and estimation methods in small data sets. Statistics in
Medicine 19: 1059-1079.
Stralberg, D., Brennan, M., Callaway, J.C., Wood, J.K., Schile, L.M., Jongsomjit, D., Kelly, M., Parker, V.T., and
Crooks, S. 2011. Evaluating tidal marsh sustainability in the face of sea-level rise: a hybrid modeling
approach applied to San Francisco Bay. PLoS ONE 6(11): e27388. doi:10.1371/journal.pone.0027388.
Streftaris, N., Zenetos, A., and Papathanassiou, E. 2005. Globalisation in marine ecosystems: the story of non-
indigenous marine species across European seas. Oceanography and Marine Biology: An Annual Review
43: 419-453.
Styf, H.K., Skold, H.N., and Eriksson, S.P. 2013. Embryonic response to long-term exposure of the marine
crustacean Nephrops norvegicus to ocean acidification and elevated temperature. Ecology and Evolution
3:5055-5065.
Sumaila, U.R., Cheung, W., Lam, V.W.Y., Pauly, D., and Herrick, S. 2011. Climate change impacts on the
biophysics and economics of world fisheries. Nature Climate Change 1: 449-456.
Sunday, J.M., Bates, A.E., and Dulvy, N.K. 2012. Thermal tolerance and the global redistribution of animals.
Nature Climate Change 2: 686-690.
Sundby, S. 2000. Recruitment of Atlantic cod stocks in relation to temperature and advection of copepod
populations. Sarsia 85: 277-298.
Sutherland, W.J. 2006. Predicting the ecological consequences of environmental change: a review of the
methods. Journal of Applied Ecology 43: 599-616.
Swartz, R.C., Cole, F.A., and Deben, W.A. 1986. Ecological changes in the Southern California Bight near a large
sewage outfall: benthic conditions in 1980 and 1983. Marine Ecology Progress Series 31: 1-13.
Swiney, K.M., Long, W.C., and Foy, R.J. 2016. Effects of high pC02 on Tanner crab reproduction and early life
history- part 1: long-term exposure reduces hatching success and female calcification, and alters
embryonic development. ICES Journal of Marine Science 73: 825-835.
293
-------
Takasuka, A., Oozeki, Y., and Aoki, I. 2007. Optimal growth temperature hypothesis: why do anchovy flourish
and sardine collapse or vice versa under the same ocean regime? Canadian Journal of Fisheries and
Aquatic Sciences 64: 768-776.
Taylor, J.R.A., Gilleard, J.M., Allen, M.C., and Deheyn, D.D. 2015. Effects of C02-induced pH reduction on the
exoskeleton structure and biophotonic properties of the shrimp Lysmata californica. Scientific Reports
5: 1-12.
Taylor, K.E., Stouffer, R.J., and Meehl, G.A. 2012. An Overview of CMIP5 and the experiment design. Bulletin of
the American Meteorological Society 93: 485-498.
Telford, M. 1982. Echinoderm spine structure, feeding and host relationships of four species of Dissodactylus
(Brachyura: Pinnotheridae). Bulletin of Marine Science 32: 584-594.
Thieler, E.R., and Hammar-Klose, E.S., 2000. National Assessment of Coastal Vulnerability to Sea-Level Rise: U.S.
Pacific Coast. U.S. Geological Survey, Open-File Report 00-178. https://pubs.usES.gov/of/2000/of00-
178/.
Thorne, K.M., Dugger, B.D., Buffington, K.J., Freeman, C.M., Janousek, C.N., Powelson, K.W., Gutenspergen, G.R.,
and Takekawa, J.Y. 2015. Marshes to mudflats—effects of sea-level rise on tidal marshes along a
latitudinal gradient in the Pacific Northwest. U.S. Geological Survey Open-File Report 2015-1204. 54
pages plus appendices, http: //dx.doi.org/10.3133/ofr20151204.
Thorner, J., Kumar, L., and Smith, S.D. 2014. Impacts of climate-change-driven sea level rise on intertidal rocky
reef habitats will be variable and site specific. PLoS ONE 9(1): e86130.
doi:10.1371/journal.pone.0086130.
Thresher, R.E., Koslow, J.A., Morison, A.K., and Smith, D.C. 2007. Depth-mediated reversal of the effects of
climate change on long-term growth rates of exploited marine fish. Proceedings of the National
Academy of Sciences of the United States of America 104: 7461-7465.
Tomanek, L. 2010. Variation in the heat shock response and its implication for predicting the effect of global
climate change on species' biogeographical distribution ranges and metabolic costs. Journal of
Experimental Biology 213: 971-979.
Toropova, C., Meliane, I., Laffoley, D., Matthews, E., and Spalding, M. (eds.) 2010. Global ocean protection:
present status and future possibilities. Brest, France: Agence des aires 230 marines protegees, Gland,
Switzerland, Washington, DC and New York, USA: IUCN WCPA, Cambridge, UK: UNEP-WCMC, Arlington,
USA: TNC, Tokyo, Japan: UNU, New York, USA: WCS. 96 pages.
Traill, L.W., Perhans, K., Lovelock, C.E., Prohaska, A., McFallan, S., Rhodes, J.R., and Wilson, K.A. 2011. Managing
for change: wetland transitions under sea-level rise and outcomes for threatened species. Biodiversity
and Distributions 17: 1225-1233.
Triantis, K.A., Guilhaumon, F., and Whittaker, R.J. 2012. The island species-area relationship: biology and
statistics. Journal of Biogeography 39: 215-231.
294
-------
Trowbridge, C.D. 1994. Defensive responses and palatability of specialist herbivores: predation on the NE Pacific
ascoglossan gastropods. Marine Ecology Progress Series 105: 61-70.
Tseng, Y.C., Hu, M.Y., Stumpp, M., Lin, L.Y., Melzner, F., and Hwang, P.P. 2013. C02-driven seawater acidification
differentially affects development and molecular plasticity along life history of fish (Oryzias latipes).
Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 165: 119-130.
U.S. Environmental Protection Agency (EPA). 1998. Guidelines for Ecological Risk Assessment. U.S. EPA Risk
Assessment Forum. Washington D.C. EPA/630/R095/002F, 1998. 171 pages.
https://www.epa.gov/risk/framework-ecological-risk-assessment.
U.S. Environmental Protection Agency (EPA). 2003. Guidance for developing ecological soil screening levels.
Office of Solid Waste and Emergency Response (US EPA), OSWER Directive 9285.7-55. 87 pages.
https: //rais.ornl.gov/documents/ecossl.pdf.
U.S. Environmental Protection Agency (EPA). 2004. Overview of the ecological risk assessment process in the
Office of Pesticide programs, U.S. Environmental Protection Agency - endangered and threatened
species effects determinations. Office of Prevention, Pesticides and Toxic Substances, Office of Pesticide
Programs Washington, D.C. 92 pages, https: //www.epa.gov/sites/production/files/2014-
11/documents/ecorisk-overview.pdf.
U.S. Environmental Protection Agency (EPA). 2008. Combined benthic database for development of Pacific Coast
indicators. U.S. EPA Office of Research and Development, National Health and Environmental Effects
Laboratory, Western Ecology Division, Pacific Coastal Ecology Branch. Newport, OR.
U.S. Environmental Protection Agency (EPA). 2009. A framework for categorizing the relative vulnerability of
threatened and endangered species to climate change. Washington D.C: National Center for
Environmental Assessment, U.S. Environmental Protection Agency. EPA/600/R-9/011.
https: //www.fws.gov/uploadedFiles/508 Climate%20change%20report%20-
%2.0GCWA%2.0example%5bl%5d.pdf.
U.S. Environmental Protection Agency (EPA). 2012. Freshwater Traits Database. Global Change Research
Program, National Center for Environmental Assessment, Washington, DC; EPA/600/R-11/038F.
Available from the National Technical Information Service, Springfield, VA. http://www.epa.gov/ncea.
https://cfpub.epa.gov/ncea/global/recordisplay.cfm?deid=241813.
U.S. Environmental Protection Agency (EPA), 2015. Office of Water and Office of Research and Development.
National Coastal Condition Assessment 2010. Washington, DC. December, 2015.
EPA 841-R-15-006. https://www.epa.gov/national-aquatic-resource-survevs/national-coastal-condition-
assessment-2010-report.
U.S. Fish and Wildlife Service. 2004. National standards and quality components for wetlands, deepwater and
related habitat mapping. Arlington, VA: U.S. Fish and Wildlife Service, Division of Habitat and Resource
Conservation, Branch of Habitat Assessment. 18 pages.
http: //www.fws.gov/stand/standards/dl wetlands National%20Standards.doc.
295
-------
U.S. Fish and Wildlife Service. 2009. Data collection requirements and procedures for mapping wetland,
deepwater and related habitats of the United States. Arlington, VA: U.S. Fish and Wildlife Service,
Division of Habitat and Resource Conservation, Branch of Habitat Assessment. 96 pages.
http: //www.fws.gov/wetlands/Documents/Data-Collection-Requirements-and-Procedures-for-
Mapping-Wetland-Deepwater-and-Related-Habitats-of-the-United-States.pdf.
USGS/EPA RACE Data 1977-2006. 2013. USGS/EPA Consolidated RACE (Resource Assessment Conservation and
Engineering) Data 1977- 2006. Seattle, WA: National Oceanic and Atmospheric Administration (NOAA),
Alaska Fisheries Science Center (AFSC), Resource Assessment Conservation and Engineering (RACE)
Division.
Uthicke, S., Schaffelke, B., and Byrne, M. 2009. A boom-bust phylum? Ecological and evolutionary consequences
of density variations in echinoderms. Ecological Monographs 79: 3-24.
Valentine, J.W. and Jablonski, D. 1986. Mass extinctions: sensitivity of marine larval types. Proceedings of the
National Academy of Sciences 83: 6912-6914.
Van Soest, R.W.M., Boury-Esnault, N., Vacelet, J., Dohrmann, M., Erpenbec, D., De Voogd, N.J., Santodomingo,
N., Vanhoorne, B., Kelly, M., and Hooper, J.N.A. 2012. Global diversity of sponges (Porifera) PLoS ONE
7(4): e35105. doi: 10.1371/journal.pone.0035105.
van Vuuren, D.P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., Hurtt, G.C., Kram, T., Krey, V.,
Lamarque, J.-F., Masui, T., Meinshausen, M., Nakicenovic, N., Smith, S.J., and Rose, S.K. 2011. The
representative concentration pathways: an overview. Climatic Change 109: 5-31.
Vazquez, L.B. and Gaston, K.J. 2004. Rarity, commonness, and patterns of species richness: the mammals of
Mexico. Global Ecology and Biogeography. 13: 535-542.
Visser, M.E. 2008. Keeping up with a warming world; assessing the rate of adaptation to climate change.
Proceedings of the Royal Society: Biological Sciences 275: 649-659.
Waldbusser, G.G. and Salisbury, J.E. 2014. Ocean acidification in the coastal zone from an organism's
perspective: multiple system parameters, frequency domains, and habitats. Annual Review of Marine
Science 6: 221-247.
Waldbusser, G.G., Brunner, E.L., Haley, B.A., Hales, B., Langdon, C.J., and Prahl, F.G. 2013. A developmental and
energetic basis linking larval oyster shell formation to ocean acidification. Geophysical Research Letters
40:2171-2176.
Waldbusser, G.G., Gray, M.W., Hales, B., Langdon, C.J., Haley, B.A., Gimenez, I., Smith, S.R., Brunner, E.L., and
Hutchinson, G. 2016. Slow shell building, a possible trait for resistance to the effects of acute ocean
acidification. Limnology and Oceanography 61: 1969-1983.
Waldbusser, G.G., Hales, B., Langdon, C.J., Haley, B.A., Schrader, P., Brunner, E.L., Gray, M.W., Miller, C.A.,
Gimenez, I., and Hutchinson, G. 2015. Ocean acidification has multiple modes of action on bivalve
larvae. PLoS ONE 10(6): e0128376. doi:10.1371/journal.pone.0128376.
296
-------
Walther, G.R., Roques, A., Hulme, P.E., Sykes, M.T., Pysek, P., Kiihn, I., Zobel, M., Bacher, S., Botta-Dukat, Z.,
Bugmann, H, Czucz, B., Dauber,J., Hickler, T., Jarosfk, V., Kenis, M., Klotz, S., Minchin, D., Moora, M.,
Nentwig, W., Ott,J., Panov, VE., Reineking, B., Robinet, C., Semenchenko, V., Solarz, W., Thuiller, W.,
Vila, M., Vohland, K., and Settele, J. 2009a. Alien species in a warmer world: risks and opportunities.
Trends in Ecology & Evolution 24: 686-693.
Walther, K., Anger, K. and Portner, H.O. 2010. Effects of ocean acidification and warming on the larval
development of the spider crab Hyas araneus from different latitudes (54° vs. 79°N). Marine Ecology
Progress Series 417: 159-170.
Walther, K., Sartoris, F.J., Bock, C., and Portner, H.O. 2009b. Impact of anthropogenic ocean acidification on
thermal tolerance of the spider crab Hyas araneus. Biogeosciences 6: 2207-2215.
Warren Pinnacle Consulting, Inc. 2011. SLAMM Analysis of the Oregon Coast for Ducks Unlimited, Vancouver,
WA. 224 pages.
http://warrenpinnacle.com/prof/SLAMM/DU/SLAMM Ducks Oregon Coast 1 13 2012.pdf.
Warren Pinnacle Consulting, Inc. 2014. Evaluation of Regional SLAMM Results to Establish a Consistent
Framework of Data and Models. Gulf Coast Prairie Land Conservation Cooperative. Pages 1-46.
http://warrenpinnacle.com/prof/SLAMM/GCPLCC/WPC GCPLCC Final Report.pdf.
Warren, M.S., Hill, J.K., Thomas, J.A., Asher, J., Fox, R., Huntley, B., Roy, D.B., Telfer, M.G., Jeffcoate, S., Harding,
P., and Jeffcoate, G. 2001. Rapid responses of British butterflies to opposing forces of climate and
habitat change. Nature 414: 65-69.
Washington State Department of Natural Resources. No Date. The Washington State ShoreZone Inventory
website:
http://www.dnr.wa.gov/programs-and-services/aquatics/aqyatic-science/nearshore-habitat-inventorv.
Wayne, G.P. 2013. The Beginner's Guide to Representative Concentration Pathways. Skeptical Science. Version
1.0. 24 pages, https: //www.skepticalscience.com/docs/RCP Guide.pdf .
Webber, M., Calumpong, H., Ferreira, B., Granek, E., Green, S., Ruwa, R., and Soares, M. 2016. Chapter 48.
Mangroves. The First Global Integrated Marine Assessment World Ocean Assessment I (eds. Inniss, L.
and Simcock, A.). United Nations. 18 pages.
http://www.un.org/depts/los/global reporting/WOA RPROC/Chapter 48.pdf.
Weinberg, J.R., Dahlgren, T.G., and Halanych, K.M. 2002. Influence of rising sea temperature on commercial
bivalve species of the U.S. Atlantic coast. American Fisheries Society Symposium. Pages 131-140.
Wernberg, T., Bennett, S., Babcock, R.C., de Bettignies, T., Cure, K., Depczynski, M., Dufois, F., Fromont, J.,
Fulton, C.J., Hovey, R.K., Harvey, E.S., Holmes. T.H., Kendrick, G.A., Radford, B., Santana-Garcon, J.,
Saunders, B.J., Smale, D.A., Thomsen, M.S., Tuckett, C.A., Tuya, F., Vanderklift, M.A., Wilson, S. 2016.
Climate-driven regime shift of a temperate marine ecosystem. Science 353: 169-172.
297
-------
Wethey, D.S. and Woodin, S.A. 2010. Ecological hindcasting of biogeographic responses to climate change in the
European intertidal zone. Hydrobiologia 606: 139-151.
Wethey, D.S., Woodin, S.A., Hilbish, T.J., Jones, S.J., Lima, F.P., and Brannock, P.M. 2011. Response of intertidal
populations to climate: effects of extreme events versus long term change. Journal of Experimental
Marine Biology and Ecology 400: 132-144.
Whiteley, N.M. 2011. Physiological and ecological responses of crustaceans to ocean acidification. Marine
Ecology Progress Series 430: 257-271.
Whittaker, R.J. and Fernandez-Palacios, J.M. 2007. Island Biogeography: Ecology, Evolution, and Conservation,
Second Edition. Washington, DC: Smithsonian Institution Press. 416 pages.
Wicks, L.C. and Roberts, J.M. 2012. Benthic invertebrates in a high C02 world. Oceanography and Marine
Biology: An Annual Review (Eds. Gibson, R.N., Atkinson, R.J.A., Gordon, J.D.M., Hughes, R.N., Hughes,
D.J., and Smith, I.P.) Vol. 50: 127-188.
Wittmann, A.C. and Portner, H.O. 2013. Sensitivities of extant animal taxa to ocean acidification. Nature Climate
Change 3: 995-1001.
Wong, P.P., Losada, I.J., Gattuso, J.-P., Hinkel, J., Khattabi, A., Mclnnes, K.L., Saito, Y., Sallenger, A. 2014. Coastal
systems and low-lying areas. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B:
Regional Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. (Eds. Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J.,
Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., Girma, B., Kissel, E.S.,
Levy, A.N., MacCracken, S., Mastrandrea, P.R., and White, L.L.). Cambridge University Press, Cambridge,
UK, and New York, NY USA. pp. 361-409.
Wood, H.L., Spicer, J.I., Lowe, D.M., and Widdicombe, S. 2010. Interaction of ocean acidification and
temperature; the high cost of survival in the brittlestar Ophiura ophiura. Marine Biology 157, 2001-
2013.
Woodroffe, C.D. 1990. The impact of sea-level rise on mangrove shorelines. Progress in Physical Geography
14:483-520.
Woodroffe, C.D. 1995. Response of tide-dominated mangrove shorelines in northern Australia to anticipated
sea-level rise. Earth Surface Processes and Landforms 20: 65-85.
Woodroffe, C.D. and Mulrennan, M.E. 1993. Geomorphology of the lower Mary River plains, Northern Territory.
North Australia Research Unit and the Conservation Commission of the Northern Territory. 152 pages.
Woodroffe, C.D., Rogers, K., McKee, K.L., Lovelock, C.E., Mendelssohn, I.A., and Saintilan, N. 2016. Mangrove
sedimentation and response to relative sea-level rise. Annual Review of Marine Science 8: 243-246.
298
-------
World Register of Marine Species, 2013. (Eds. Boxshall, G., Mees, J., Costello, M.J., Hernandez, F., Vandepitte, L.,
Gofas, S., Hoeksema, B.W., Klatau, M., Kroh, A., Poore, G.C.B., Read, G., Stohr, S., de Voogd, N.J., Walter,
C.T., De Broyer, C., Horton, T., and Kennedy, M.). http://www.marinespecies.org/index.php.
Wyllie-Echeverria, S. and Ackerman, J.D. 2003. Seagrasses of the Pacific Coast of North America. World Atlas of
Seagrasses; Present status and Future Conservation (Eds. Green, E. and Short, F.). 217-225. University of
California Press: Berkeley, CA. 260 pages.
Yamane, L. and Gilman, S.E. 2009. Opposite responses by an intertidal predator to increasing aquatic and aerial
temperatures. Marine Ecology Progress Series 393: 27-36.
Yamano, H., Sugihara, K., and Nomura, K. 2011. Rapid poleward range expansion of tropical reef corals in
response to rising sea surface temperatures. Geophysical Research Letters 38: L0460.
Zervas, C. 2009. Sea Level Variations of the United States 1854-2006. National Oceanic and Atmospheric
Administration, US Department of Commerce, National Ocean Service, Center for Operational
Oceanographic Products and Services. NOAA Technical Report NOS CO-OPS 053. 194 pages.
Zhang, H., Shin, P.K., and Cheung, S.G. 2016. Physiological responses and scope for growth in a marine
scavenging gastropod, Nassarius festivus (Powys, 1835), are affected by salinity and temperature but
not by ocean acidification. ICES Journal of Marine Science 73: 814-824.
Zippay, MK.L. and Hofmann, G.E. 2010. Effect of pH on gene expression and thermal tolerance of early life
history stages of red abalone (Haliotis rufescens). Journal of Shellfish Research 29: 429-439.
299
-------
SEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
Recycled/Recyclable Printed on paper that contains a minimum of
50% postconsumer fiber content processed chlorine free
------- |