*>EPA
United States
Environmental Protection
Agency
EPA/600/R-17/351 November 2017 www.epa.gov/ord
Risk Analysis of Near-
Coastal Species of the
U.S. Pacific Coast:
Case Study Comparing
Risks Associated with
Two Future Climate
Scenarios
Henry Lee II, Christina Folger,
Patrick Clinton, Deborah Reusser
and Rene Graham
Office of
Research and Development
National Health and
Environmental Effects
Research Laboratory
Western Ecology Division
Pacific Coastal Ecology Branch
-------
3EPA/600/R-17/351
November, 2017
Risk Analysis of Near-Coastal Species
of the U.S. Pacific Coast:
Case Study Comparing Risks Associated with
Two Future Climate Scenarios
By
Henry Lee II, U.S. EPA, Western Ecology Division
Christina Folger, U.S. EPA, Western Ecology Division
Patrick Clinton, U.S. EPA, Western Ecology Division
Deborah Reusser, USGS (Emeritus)
Rene Graham, CSS
Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Western Ecology Division
11
-------
Disclaimer
This document has been reviewed by the U.S. Environmental Protection Agency, Office of Research
and Development, and is 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. Risk Analysis of Near-Coastal
Species of the U.S. Pacific Coast: Case Study Comparing Risks Associated with Two Future Climate
Scenarios. EPA/600/R-17/351. 64 pages.
111
-------
Acknowledgements
Special thanks to Robert Reusser for programming the risk assessment algorithms and to the following
people who provided programming support for CBRAT: Marshall Hanshumaker, Dylan McCarthy, and
Rachel Nehmer. Also special thanks Carol DeLong, our technical editor, who spent long hours assisting
with data entry and editorial reviews. Emily Saarinen, Melanie Frazier and Katie Marko provided
insights into the early development of the risk 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. The authors would also
like to acknowledge Tim Counihan and Jill Hardiman of the USGS Western Fisheries Research Center
for their assistance in collating information on 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 Cadi en, 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
assessments. 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 Don Cadien, 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 provided insightful suggestions on the risk assessment approach in their
reviews of previous CBRAT documents: Rebecca Flitcroft (USFS), Thomas Hurst (NOAA), Walter
Nelson (EPA), Tony Olsen (EPA), Steve Rumrill (ODF&W), James Markwiese (EPA), Maggie Dutch
(Washington Dept. of Ecology), Valerie Partridge (WA DOE), Dean Pasko (Dancing Coyote
Environmental), Larry Lovell (County Sanitation Districts of Los Angeles County), and Don Cadien
(Los Angeles County Sanitation Districts). Also, we wish to thank the reviewers of this current case
study for their insights: Dr. Walter Nelson and Valerie Partridge. Finally, Dr. Lee would like to
acknowledge the continued support of EPA's Air, Climate, and Energy (ACE) research program.
iv
-------
Change Log
11/17/2017: 508 compliant version.
11/17/2017: 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 Ecoregion and Oregon,
Washington, Vancouver Coast and Shelf). The likelihood of northern colonization was recalculated
using these average air values, and the text and figures modified in Section 4.
v
-------
Table of Contents
Disclaimer iii
Recommended Citation iii
Acknowledgements iv
Change Log v
Acronyms and Abbreviations x
Executive Summary xii
Section 1. Introduction 1
1.1 Objectives 1
Section 2. Risk Assessment Methods 3
2.1 Overview of Methods 3
2.2 Geographic and Taxonomic Scope of Current Risk Assessment 3
2.3 Risk Categories and Overall Vulnerability 4
2.4 CBRAT and Algorithm-Based Risk Analysis 5
2.5 Calculation of Temperature Risks 5
2.5.1 Ecoregional Thermal Windows (ETW) Approach 6
2.5.2 Biogeographical Thermal Limit Approach (BTL) 6
2.6 Northern Colonization 8
2.7 Ocean Acidification Risks 8
2.8 Sea Level Rise Risks 11
2.9 Baseline/Status Risks 12
Section 3. Comparison of Risks with RCP 4.5 vs. RCP 8.5 Scenarios 13
3.1 Introduction 13
3.2 Coastal Risk Patterns: Sum-Occupied Ecoregions with Species at Moderate or High Risks 15
3.3 Patterns of Risk Across Ecoregions 16
Section 4. Northern Colonization 20
4.1 Introduction 20
4.2 Northern Colonization Projections 20
Section 5. Discussion 26
5.1 Introduction 26
5.2 Case Study - Efficacy of CBRAT in Conducting Risk Assessments 26
5.3 Case Study - Adding New Taxon 26
5.4 Risk Reduction. Comparing the RCP 4.5 Scenario to the RCP 8.5 Scenario 28
5.5 Northern Colonization 29
5.6 Overall Assessment of Risk Reduction under the RCP 4.5 Scenario 29
Section 6. Uncertainty Analysis and Quality Assurance/Quality Control 31
6.1 Example of Uncertainty Analysis - Aragonite Saturation State Thresholds 31
6.2 Overview of Uncertainty Related to Risk Reduction 33
6.3 EPA/ORD's Quality Assurance/Quality Control 34
vi
-------
Appendix A.Baseline Temperatures and Predicted Increases 36
Appendix B.Baseline pH/Aragonite Saturation State Values and Predicted Decreases 44
Appendix C. Sea Level Rise Rates 53
Glossary of Terms 54
Bibliography 58
List of Tables
Table 2-1. Example of temperature thresholds & risks calculated using the Biogeographic Thermal Limit (BTL)
approach 7
Table 3-1. Number of species per ecoregion, total number of species in the U.S. Arctic and Northeast Pacific
ecoregions, and total number of occupied ecoregions 19
Table 3-2. Comparison of the ranges of percent of At-Risk species and median percent of At-Risk species with
RCP 4.5 vs. 8.5, and number of ecoregions with reduced percent of At-Risk species with RCP 4.5 19
Table A-1. Projected annual, summer, and winter increases in SST by ecoregion under the RCP 8.5 scenario for
the 2050-2099 timeframe 36
Table A-2. Projected annual, summer, and winter increases in SST by MEOW ecoregion under the RCP 4.5
scenario for the 2050-2099 timeframe 36
Table A-3. Annual mean SST temperature thresholds by ecoregion 37
Table A-4. Summer mean SST temperature thresholds by ecoregion 37
Table A-5. Winter mean SST temperature thresholds by ecoregion 38
Table A-6. Historical, predicted increase, and projected annual air temperature by ecoregion under RCP 8.5 38
Table A-7. Historical, predicted increase, and projected annual air temperature by ecoregion under RCP 4.5 39
Table A-8. Historical, predicted increase, and projected summer air temperature by ecoregion under RCP 8.5. ..39
Table A-9. Historical, predicted increase, and projected summer air temperature by ecoregion under RCP 4.5. ..40
Table A-10.Historical, predicted increase, and projected winter air temperature by ecoregion under RCP 8.5 40
Table A-11. Historical, predicted increase, and projected winter air temperature by ecoregion under RCP 4.5 41
Table A-12. Historical, predicted increase, and projected SST temperature by ecoregion under RCP 8.5 41
Table A-13. Historical, predicted increase, and projected SST temperature by ecoregion under RCP 4.5 41
Table A-14. Historical, predicted increase, and projected 30 m depth temperature by ecoregion under RCP 8.5. 42
Table A-15. Historical, predicted increase, and projected 30 m depth temperature by ecoregion under RCP 4.5. 42
Table A-16. Historical, predicted increase, and projected 100 m depth temperature by ecoregion under
RCP 8.5 43
Table A-17. Historical, predicted increase, and projected 100 m depth temperature by ecoregion under
RCP 4.5 43
Table
Table
B-1. Historical, predicted decreases, and projected annual mean pH by ecoregion under RCP 8.5.
B-2. Historical, predicted decrease, and projected annual mean pH by ecoregion under RCP 4.5..
.44
.45
-------
Table B-3. Historical, predicted decrease, and projected summer mean pH by ecoregion under RCP 8.5 45
Table B-4. Historical, predicted decrease, and projected summer mean pH by ecoregion under RCP 4.5 45
Table B-5. Historical, predicted decrease, and projected winter mean pH by ecoregion under RCP 8.5 46
Table B-6. Historical, predicted decrease, and projected winter mean pH by ecoregion under RCP 4.5 46
Table B-7. Historical, predicted decrease, and projected annual mean aragonite saturation state (Qa) by
ecoregion under RCP 8.5 47
Table B-8. Historical, predicted decrease, and projected annual mean aragonite saturation state (Qa) by
ecoregion under RCP 4.5 47
Table B-9. Minimum Acceptable Toxicant Concentrations (MinATC) of pH for fish 48
Table B-10. Minimum Acceptable Toxicant Concentrations (MinATC) of aragonite saturation state (Qa) for
bivalves 50
Table B-11. High sensitivity thresholds for pH for decapods and fish 52
Table B-12. Moderate sensitivity thresholds for pH for decapods and fish 52
Table B-13. Low sensitivity thresholds for pH for decapods and fish 52
Table B-14. High, moderate, and low sensitivity thresholds of aragonite saturation state (Qa) for bivalves 52
Table C-1. Net sea level rise by ecoregion with an eustatic rate of 12 mm/yr and default isostatic rates 53
Table C-2. Net sea level rise by ecoregion with an eustatic rate of 5 mm/yr and default isostatic rates 53
List of Figures
Figure 2-1. MEOW ecoregions comprising the Northeast Pacific and U.S. Arctic 4
Figure 2-2. Conceptual relationship between increasing climate risk level and likelihood of adverse impacts 5
Figure 2-3. Distribution of Chionoecetes bairdi illustrating WOE, NWUE, COE, and NCUE ecoregions 7
Figure 2-4. Cumulative distribution of the most sensitive MinATCs for aragonite saturation state (Qa) for each
bivalve species 11
Figure 3-1. Example of the number of Sum-Occupied ecoregions based on two species 14
Figure 3-2. Differences in the percent of Sum-Occupied ecoregions with species at moderate risk, high risk, and
At-Risk under RCP 4.5 compared to RCP 8.5 15
Figure 3-3. Percent of brachyuran species within each ecoregion At-Risk with RCP 4.5 and 8.5 17
Figure 3-4. Percent of lithodid species within each ecoregion At-Risk with RCP 4.5 and 8.5 17
Figure 3-5. Percent of Sebastes species within each ecoregion At-Risk with RCP 4.5 and 8.5 18
Figure 3-6. Percent of bivalve species within each ecoregion At-Risk under RCP 4.5 and 8.5 18
Figure 4-1. Percent of the species in the U.S. Arctic and NEP classified as potential northern colonists under RCP
4.5 and RCP 8.5 21
Figure 4-2. Brachyura - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios 22
Figure 4-3. Lithodoidea - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios 23
-------
Figure 4-4. Sebastes- northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios 24
Figure 4-5. Bivalves - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios 25
Figure 6-1. Risks to bivalves from ocean acidification at different aragonite saturation state thresholds (Qa) under
RCP 4.5 and RCP 8.5 33
IX
-------
Acronyms and Abbreviations
Qa Aragonite saturation state m
ACE Air, Climate and Energy (EPA/ORD MATC...
research program)
BTL Biogeographic Thermal Limit MEOW.
approach
MinATC
°C Degrees Celsius
CBRAT Coastal Biodiversity Risk Analysis mm
Tool
NCUE...
cm Centimeter(s)
ND
CMIP5 Coupled Model Intercomparison
Project Phase 5 NEP
COE Coolest occupied ecoregion NOAA...
CSV Comma-separated values file
NOAEL.
EPA Environmental Protection Agency
NWUE..
ESLR Eustatic sea level rise
OA
ETW Ecoregional Thermal Window
approach PDF
GIS Geographic information system RCP
HOAEL Highest observed adverse effects
level RSLR...
hr Hour SLAMM
IPCC Intergovernmental Panel on Climate SLR
Change
sp
IUCN International Union for Conservation
of Nature spp
km Kilometer(s) SST
Lat. & Long Latitude and Longitude USGS...
LOAEL Lowest observed adverse effects W
level
WOE ....
Meter(s)
Maximum acceptable toxicant
concentration
Marine Ecoregions of the World
Minimum acceptable toxicant
concentration
Millimeter(s)
Next coolest unoccupied ecoregion
No data
Northeast Pacific
National Oceanic and Atmospheric
Administration
No observed adverse effects level
Next warmest unoccupied ecoregion
Ocean acidification
Portable Document Format
Representative concentration
pathways
Relative sea level rise
Sea Level Affecting Marsh Model
Sea level rise
Single species
Multiple species
Sea surface temperature
United States Geological Survey
Watts
Warmest occupied ecoregion
x
-------
THIS PAGE INENTIONALLY BLANK
xi
-------
Executive Summary
Within the scientific community, climate change is widely recognized as a major ecological threat. What
is less well understood is how these risks change under different emission scenarios. To address this
need, we developed a rule-based framework to predict relative climate change risks for near-coastal
species (0-200 m depth) at regional scales. To capture the effects of multiple climate drivers, the risk
assessments include temperature increases, sea level rise, ocean acidification, and baseline biotic traits
and status related to climate vulnerability. Risks are evaluated at the scale of the marine ecoregions
established by the Marine Ecosystems of the World (MEOW). This risk framework is described in detail
in a companion document (Lee et al., 2017) and implemented in the ecoinformatics website Coastal
Biodiversity Risk Analysis Tool (CBRAT, fattp://www.cbrat.ore/).
In our first case study, we evaluated the efficacy of CBRAT in conducting scenario analyses by
comparing the risks associated with two climate scenarios as defined by Representative Concentration
Pathway (RCP) 4.5 and 8.5. RCP 8.5 represents an unregulated, "business as usual" scenario while RCP
4.5 represents an intermediate scenario. In this analysis, we evaluated risks for all the species of four
taxa occurring in the ten contiguous coastal ecoregions from Southern California to the Beaufort Sea.
We had previously synthesized information on brachyuran crabs (true crabs: 135 species), lithodid crabs
(king crabs: 21 species), and Sebastes (rockfish: 68 species). To evaluate the practicality of synthesizing
the biotic trait information required for a risk analysis, we summarized information on the 470 bivalve
species occurring in these ecoregions. Based on this effort, we concluded that while practical, addition
of a new taxon requires a dedicated effort, especially if it contains a large number of species. As a rough
estimate, one-quarter to one-half year time of a person(s) familiar with taxonomy and natural history
would be required to add a new taxon equivalent to the bivalves.
Changes in risk under RCP 4.5 were evaluated at two spatial scales:
• Coastal scale, based on changes summed across all ten ecoregions covering the contiguous US
Pacific Coast.
• Ecoregion scale, comparing geographical patterns of change within individual ecoregions.
Coastal scale. At the coastal scale, risk was based on the Sum-Occupied ecoregions, which is the sum of
the number of the ten ecoregions inhabited by any of the species within a taxon. For example, the 470
bivalve species occupied 1584 ecoregions in total. Changes in risk level were calculated as the
difference between the percent of the Sum-Occupied ecoregions predicted to be at high risk versus
moderate risk. At this scale, the RCP 4.5 scenario showed substantial decreases in the percent of
ecoregions with species at high risk compared to RCP 8.5. The reduction in high risks ranged from a
32% decrease for lithodid crabs to a 93% decrease for bivalves. Concurrently, there was an increase in
the number of ecoregions with species at moderate risk with the lithodid crabs, rockfish, and bivalves.
Xll
-------
This pattern indicates that under the RCP 4.5 scenario, the risk was reduced from high to moderate for
many species.
Ecoregion scale. Geographical patterns of risk reduction were evaluated as the differences in the percent
of At-Risk species in each of the ecoregions, where At-Risk is defined as species at high risk or
moderate risk. The results showed no consistent pattern in risk reduction with RCP 4.5, neither across
the four taxa or geographically. In each of the taxa, there was at least one ecoregion with no or only
minor (<4%) reductions in the percent of At-Risk species. Conversely, for each taxon there was at least
one ecoregion with a substantial (>25%) reduction in the percent of At-Risk species. In three of the taxa,
more than half of the occupied ecoregions showed a reduction in the percentage of At-Risk species.
However, the lithodid crabs showed a risk reduction in only four of its nine occupied ecoregions.
Another effect of climate change is the potential for southern species to colonize currently unoccupied
northern ecoregions as temperatures warm. This was evaluated by assessing the thermal suitability of
currently unoccupied northern ecoregions for southern species. As with the risks, there was considerable
geographic variation in the reduction of potential colonists per ecoregion under RCP 4.5. However,
when evaluated at a coastal scale, there were substantial decreases in the number of potential colonists in
each of the four taxa under RCP 4.5.
Generating simple conclusions regarding risk reduction with the RCP 4.5 scenario is complicated by the
multiple ways that risk reductions can be quantified and the lack of consistent taxonomic and
geographical patterns. Nonetheless, relying on the coastal analysis of risks as the single best measure,
the substantial decrease in high-risk species for all four taxa provides a clear indicator of risk reduction
under RCP 4.5. Additionally, each ecoregion showed moderate to substantial reductions in risk in at
least one of the four taxa. Concurrent with the reductions in risk, fewer species were predicted to
colonize northern ecoregions under the RCP 4.5 scenario. Overall, we conclude that the RCP 4.5
scenario would substantially reduce ecological risks to brachyuran crabs, lithodid crabs, rockfish, and
bivalves along the North American Pacific Coast.
Finally, we evaluated the efficacy of CBRAT in conducting uncertainty analyses by varying the default
aragonite saturation state thresholds for bivalves from -50% to +25%. In the risk framework, pH and
aragonite saturation state thresholds determine a species' sensitivity to ocean acidification. Based on the
percent of Sum-Occupied ecoregions with At-Risk species, ocean acidification risks were robust to
changes in the aragonite thresholds under RCP 8.5. In comparison, ocean acidification risks showed
greater sensitivity to changes in the thresholds under RCP 4.5. While CBRAT is well suited to
conducting uncertainty analyses, the large number of parameters that can be varied singly and in
combination limits the uncertainty analyses to key issues.
-------
Section 1. Introduction
1.1 Objectives
Climate change is a major threat to marine and estuarine species and to the ecosystem services they
provide (e.g., Hannah, 2012; Portner et al., 2014; Molinos et al., 2016). Responding to this threat in a
scientifically-sound manner, whether through prevention or through adaptive management, requires a
basic knowledge of: 1) severity of the threat and how the threat varies among different species; 2)
geographic patterns in the nature and severity of the threat; 3) relative importance of different climate
drivers; and 4) changes in the relative risks under different climate scenarios. To address these issues for
near-coastal (0-200 m depth) species and resources, we developed a climate risk assessment framework
to assess the regional-scale risks associated with temperature increases, ocean acidification, sea level
rise (SLR), and baseline/status risks, i.e., the biotic traits of species and non-climate related
circumstances that make them more or less vulnerable to climate change. The risk assessment
framework, along with key assumptions, are detailed in a companion document, "Predicting Patterns of
Vulnerability to Climate Change in Near Coastal Species Using an Algorithm-Based Risk Framework"
(Lee et al., 2017). The framework is implemented in an online ecoinformatics tool, the "Coastal
Biodiversity Risk Analysis Tool" (CBRAT, http://www.cbrat.oref). Users are also referred to Lee et al.,
2015 which details the baseline and status traits in CBRAT.
While analysis of the effects of climate change will never be entirely turnkey, a major objective in
designing CBRAT was to provide a practical tool for informed researchers and managers to explore
different aspects of the risks associated with climate change. To gage our success in achieving this
objective, we evaluated the efficacy of CBRAT in addressing one component of risk. Specifically, in
this case study we evaluated the general geographic and taxonomic patterns of risk reduction with the
RCP 4.5 versus RCP 8.5 scenarios (see Sidebar #1) for coastal species from Southern California to the
Beaufort Sea. We focused this first case study on comparing RCPs because of the policy implications
and because such a comparison captures many components of a regional risk assessment, thus providing
a rigorous test of the ability of CBRAT to inform management on climate risks. As part of this
evaluation, we also added bivalves to test the practicality of expanding CBRAT to include a new taxon.
Finally, since uncertainty is inherent in any climate model, we evaluated the utility of CBRAT as a tool
to assess the uncertainty in aragonite saturation state thresholds, one component of the risk assessments.
1
-------
Sidebar #1. Representative Concentration Pathways
Current assessments of climate change utilize Representative Concentration Pathways (RCPs) rather
than the previous emission scenarios (e.g., A1B) (Van Vuuren et al., 2011; Snover et al., 2013; Wayne,
2013; IPCC, 2014), where RCPs are measured in W/m2 The four standard RCPs are (modified from
Snover et al., 2013):
RCP 2.6: Low pathway that "reflects aggressive greenhouse gas reduction and sequestration
efforts." Likely to maintain global temperature increases below 2°C in the 21st Century (IPCC,
2014).
RCP 4.5: Intermediate pathway where radiative forcing is stabilized by mid-century and then
decreases into 2100 with no overshoot of the 4.5 W/m2 value. More unlikely than likely to
maintain global temperature increases below 2°C in the 21st Century (IPCC, 2014).
RCP 6.0: Intermediate pathway where radiative forcing increases until stabilizing in the final
decades of the 21st century. More unlikely than likely to maintain global temperature increases
below 3°C in the 21st Century (IPCC, 2014).
RCP 8.5: Increasing greenhouse gas emissions throughout the 21st century. A "business as
usual" pathway. More unlikely than likely to maintain global temperature increases below 4°C
in the 21st Century (IPCC, 2014). Currently used as the default in CBRAT.
In this analysis, we compare the risks associated with RCP 4.5 versus RCP 8.5. A practical reason for
focusing on these pathways is that the NOAA Climate Change Web Portal
(https://www.esrl.noaa.gov/psd/ipcc/ocn/). our primary source for climate projections, does not include
the RCP 2.6 or RCP 6.0 pathways. As additional climate projections become available, in particular the
RCP 2.6 pathway, CBRAT can be used to evaluate the associated risks on near-coastal species.
On a policy level, the RCP 8.5 approaches a "worst case" with no or minimal policy efforts to control
emissions and other factors such as deforestation (see Riahi et al., 2011). Under this pathway, CO2
equivalents are predicted to increase by 74% to 178% by 2100 (IPCC, 2014). In contrast, the RCP 4.5
pathway represents a moderate pathway with policy actions that result in a 21% to 134% reduction in
CO2 equivalents by 2100 (IPCC, 2014).
2
-------
Section 2. Risk Assessment Methods
2.1 Overview of Methods
The specifics of the risk assessment methods and assumptions are detailed in our climate risk assessment
framework document (Lee et al., 2017); this section briefly summarizes the approaches while
Appendices A, B, and C document the projected environmental values under the RCP 4.5 and 8.5
pathways. The general approach to predicting climate risk is to first overlay predicted changes in
temperature, pH, or sea level rise on baseline values to generate projected values at an ecoregion scale
under different climate scenarios. Relative risk is then calculated by comparing the projected
environmental values to temperature, pH/aragonite, and sea level rise effects thresholds for individual
species.
2.2 Geographic and Taxonomic Scope of Current Risk Assessment
The Marine Ecoregions of the World (MEOW) (Spalding et al., 2007) provides the biogeographic
schema for evaluating species' distributions and climate risks. The present effort analyzes species in ten
MEOW ecoregions, (Figure 2-1):
• Southern California Bight
• Northern California
• Oregon, Washington, Vancouver Coast and Shelf
• Puget Trough/ Georgia Basin
• North American Pacific Fijordland (note: spelling as used in Spalding et al., 2007)
• Gulf of Alaska
• Aleutian Islands
• Eastern Bering Sea
• Chukchi Sea
• Beaufort Sea - continental coast and shelf
Though they are part of the Northeast Pacific (NEP) region, we did not attempt to analyze the species in
the Magdalena Transition or Cortezian (Gulf of California) ecoregions at this time because of
insufficient information. Thus, when we refer to the Northeast Pacific (NEP) in this document, we are
referring to the seven ecoregions ranging from the Aleutians through Southern California.
Climate risks were analyzed for all the Brachyura (true crabs: 135 species), Lithodoidea (king crabs: 21
species), Sebastes (rockfish: 68 species), and Bivalvia (clams/mussels: 470 species) that occur in these
ten ecoregions. The risk analysis includes estuarine species and offshore species that occur within depths
of 0 - 200 meters.
3
-------
g
in
i ret i
c Ocean • • • - - ,
N
ChukchiSea J _
Beaufort Sea ^ \
^ u.s.A.
g
S
- fattem
Bering
im
•N-tn. Canada
GulfofAbUi
Aleutian bland*
North J Pu»«Trou«h/
Americjn oron* Bjixi
Pacrtk FjordUnd ~
g
•n
¦
Pacific Ocean
°*'WA' * v.,a,
VhMttM
Co«1 and
simM
1 u s A
Northern
California ^Uo»Aa»d»
Z
t>
S
Ktwai't
. JkivMu
1 \ ¦ \
CWWP \
Bight
Mjftdjfera Ma/a4an
Tramtlfon
Corteiian
z
in
* ¦
180» 160CTW 140 OW 120 OW 100 ow
Figure 2-1. 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 (NEP) 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. The current effort analyzes climate risk with species occurring from the Southern
California Bight through the Beaufort Sea - Continental Coast and Shelf ecoregions. Hawaii is not part of the NEP
and is not included in this analysis. (From Lee et al., 2017)
2.3 Risk Categories and Overall Vulnerability
Climate risk levels are classified as minor, low, moderate, or high. As the risk level increases, the
likelihood, severity, and types of adverse impacts are predicted to increase, with population declines
considered likely under high-risk scenarios (Figure 2-2). An attempt was made to standardize risks
across the various climate drivers (e.g., temperature increases versus SLR) so that a high risk for one
climate driver is approximately equivalent to a high risk for another driver. This proved challenging and
parity among different climate drivers was only approximated. In particular, ocean acidification was
problematic because much of the data are from laboratory exposures using physiological and behavioral
endpoints that are not readily related to population viability, which was the focus for temperature and
SLR effects.
We concluded that the most scientifically defensible approach to assigning a single climate risk
("Overall Vulnerability") within an ecoregion is to define the overall risk as the single greatest risk
4
-------
among the temperature, ocean acidification, sea level rise, and baseline/status risks (see Section 2.3 in
Lee et al., 2017).
Minor Risk
Low Risk
Likelihood of
Moderate Risk
Likelihood of Other Ecologically Adverse Impacts
Severity of Ecologically Adverse Impacts
Number of Ecologically Adverse Impacts
Figure 2-2. Conceptual relationship between increasing climate risk level and likelihood of adverse impacts.
Not every type of impact is necessarily expected in all cases, though population declines are likely with a high
risk. (Modified from Lee et al., 2017)
2.4 CBRAT and Algorithm-Based Risk Analysis
The climate risks presented in this document were generated using CBRAT (ittp://www.cbrat. org,). As
detailed in Lee et al. (2017), CBRAT is a platform to synthesize biotic traits and species' distributions
and to calculate climate risks using this knowledge base and a set of explicit rules that draw upon
projected climate changes for temperature, pH/aragonite saturation state, and sea level rise. This
algorithm-based approach is in contrast to the use of expert opinion to evaluate climate risks (e.g.,
Moyle et al., 2013; Morrison et al., 2015; see Secion 1.5 in Lee et al., 2017). Advantages of algorithm-
based approaches compared to expert opinion include, but are not limited to:
• Reduces possible biases associated with expert opinion;
• Promotes transparency in the data and the logic;
• Promotes consistency in risk analyses across different taxa and geographical regions;
• Simplifies evaluating results from different climate scenario analyses and
assumptions;
• Simplifies updating risks with new information.
Consistency, the ease of adding new information and the simplification of evaluating results for different
scenarios were especially important in the comparison of ecological responses to climate change under
different RCP scenarios.
2.5 Calculation of Temperature Risks
Two methodologies were developed to predict risks associated with temperature increases: The
Ecoregional Thermal Windows (ETW) approach and the Biogeographical Thermal Limit (BTL)
approach.
5
-------
2.5.1 Ecoregional Thermal Windows (ETW) Approach
The Ecoregional Thermal Windows approach (ETW) compares the projected sea surface temperature
(SST) in each ecoregion to the historic range of SST values in the "warmest occupied ecoregion"
(WOE) (Figure 2-3). Temperatures in the WOE are assumed to represent the upper ecological thermal
limit for species to maintain a viable population (see Appendix D of Lee et al., 2017 for a discussion of
warm edge distributions). The ecoregion-scale historic SSTs were derived from an analysis of 28 years
of "advanced very high resolution radiometer" (AVHRR) remote sensing data (Payne et al., 2011,
2012a, 2012b, and unpublished). Future projections were extracted from the CMIP5 model (Taylor et
al., 2012) served through NOAA's Climate Change Web Portal
("https://www.esrl.noaa.gov/psd/ipcc/ocnA). The risk level is determined by comparing the projected SST
in each ecoregion to the historic mean plus a number of standard deviations (SD) in the WOE. Risk
assessment values for the annual mean, summer, and winter are individually calculated. Moderate risk is
defined as a projected SST in an ecoregion that is greater than the WOE mean + 2 SDs. High risk is
defined as a projected temperature that is greater than the WOE mean + 3 SDs.
Table A-l and Table A-2 document the projected increases in SST used in the ETW analysis. Tables A-
3 through Table A-5 document the annual, summer, and winter thermal thresholds used in the ETW
analysis.
2.5.2 Biogeographical Thermal Limit Approach (BTL)
Though SSTs are frequently used to evaluate impacts of climate change (e.g., Molinos et al., 2016), they
may not adequately capture risks to subtidal species. To address this issue, we developed the
Biogeographical Thermal Limit (BTL) approach which predicts risks at different depth classes
appropriate for the particular species. The BTL approach predicts risks for intertidal species using the
projected mean annual, summer, and winter air temperatures. Subtidal species fall into two classes:
shallow subtidal (>0 m to 30 m) and deep subtidal (>30 m to 200 m). For shallow subtidal species, we
use projected temperatures at 30-m depth; for deep subtidal species, we use projected temperatures at
100-m depth. Additionally, risks associated with SSTs, based on the CMIP5 model, are evaluated for all
species. For species that span more than one depth class, risks are calculated independently for each
depth class that the species occupies, plus SST.
The BTL generates risks by comparing the projected temperatures in the target ecoregion to temperature
thresholds based on four equal bins between the historic temperatures in the WOE and the "next
warmest unoccupied ecoregion" (NWUE) (Figure 2-3). The NWUE is assumed to be too warm to
maintain a viable population of the target species, thus the closer the projected temperature in the target
ecoregions comes to the NWUE temperature, the greater the risk. Risk were calculated independently
for each depth class and the SSTs. Moderate risk is defined as a projected temperature that lies within
the third bin between the WOE and NWUE. High risk is defined as a projected temperature greater than
the 3rd bin between the WOE and NWOE. For all depth classes, the risk is reduced by one risk class if
the depth class is classified as Observed rather than Preferred (see Section 5.4.2 in Lee et al., 2017). An
example of a species occurring in two depth classes is given in Table 2-1.
6
-------
Table 2-1. Example of temperature thresholds & risks calculated using the Biogeographic Thermal Limit (BTL)
approach.
In this example, historic annual air temperature in the WOE is 20 °C and 18 °C in the shallow subtidal. The NWUE
has an annual air temperatures of 24 °C and 20 °C in the shallow subtidal. This results in risks levels based on the
four bins between the WOE and NWUE for annual air temperature and the shallow subtidal. As an example, a
target species occurs in both the intertidal and shallow subtidal (>0-30 m) and both depth ranges are classified as
preferred. Assuming projected temperatures in the target ecoregion of 21.5 °C for annual air temperature andl 9.7
°C in shallow subtidal, the final risks are Moderate for air temperature and High for subtidal temperatures. The bins
that the projected temperatures fall into are highlighted in red.
Bins (Risk Level)
WOE - NWUE Annual Air
Temperature Range
(20 to 24 °C)
WOE - NWUE Shallow
Subtidal Temperature Range
(18 to 20 °C)
Bird (Minor Risk)
20 to 21 °C
18 to 18.5 °C
Bin 2 (Low Risk)
>21 to 22 °C
>18.5 to 19.0 °C
Bin 3 (Moderate Risk)
>21 to 22 °C
>19.0 to 19.5 °C
Bin 4 (High Risk)
>23 to 24 °C
>19.5 to 20 °C
With the BTL approach, both the baseline and future projections were based on the CMIP5 model
downloaded from NOAA's Climate Change Portal. The baseline temperatures, predicted increases, and
projected temperatures used in the BTL approach are documented in Appendix A in Table A-14 through
Table A-16.
NEXT Coolest UNOCCUPIED Ecoregion - Chukchi Sea
Coolest OCCUPIED Ecoregion - Eastern Bering Sea
Warmest OCCUPIED Ecoregion - Northern California
NEXT Warmest UNOCCUPIED Ecoregion - Southern California
Figure 2-3. 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
NWUE is considered too warm to maintain a viable population while the NCUE is considered too coid to maintain
a viable population.
7
-------
2.6 Northern Colonization
Temperature increases associated with climate change may result in sufficient warming of cooler,
unoccupied northern ecoregions to allow colonization by southern species. To evaluate the potential for
northern colonization, we reversed the logic of the BTL approach and derived temperature thresholds
based on temperature bins between the "coolest occupied ecoregion" (COE) and the "next coolest
unoccupied ecoregion" (NCUE) (Figure 2-3) for SST and intertidal, shallow subtidal, and deep subtidal
depths classes. Instead of risks, the likelihood ("suitability") of northern colonization is evaluated by
how close the projected temperature in the unoccupied ecoregion comes to that in the COE. Analogous
to calculating risk (Table 2-1), the suitability of the projected temperatures is calculated independently
for SSTs and each of the depth classes, with the suitability in a depth class reduced by one class if it is
Observed versus Preferred. Moderate suitability for colonization is defined as the projected temperature
in the unoccupied ecoregion falling in the 3rd temperature bin between the COE and NCUE. High
suitability is defined as the projected temperature greater than the 3rd bin. The overall suitability across
SSTs and depth classes is determined as the lowest suitability score, which is a more stringent
requirement than predicting northern colonization based on SSTs only. In the current analysis, we define
"potential colonists" as species with a suitability of moderate or high.
Temperature data used in the BTL analysis for northern colonization are documented in Table A-6
through Table A-17 in Appendix A.
2.7 Ocean Acidification Risks
Ocean acidification is a complicated process at both local (Wallace et al., 2014) and regional (e.g.,
Wootton et al., 2008) scales. Nonetheless, we suggest there is sufficient information to generate a first-
order metric applicable to assessing risk patterns across multiple species at regional scales. Our
approach is to treat effects of pH reductions as if it were a contaminant. Specifically, we derive the
"minimum acceptable toxicant concentration" (MinATC), which is based on the "maximum acceptable
toxicant concentration" (MATC) used with pollutants and pesticides (see Sidebar #2). Because aragonite
saturation state is considered the primary stressor for bivalves (see Barton et al., 2015; Waldbusser et al.,
2015), MinATCs were generated for aragonite saturation state for bivalves, while MinATCs based on
pH were generated for fish and decapods.
Within each taxon, MinATCs were calculated for each species from a synthesis of exposure experiments
and the single most sensitive MinATC for each species was identified regardless of the specific endpoint
or life history stage (Table B-9 for fish, Table B-10 for bivalves, see Table 6-6 in Lee et al., 2017 for
decapods). Species within a taxon vary greatly in their sensitivity to reduced pH or aragonite saturation
state. For example, the most sensitive MinATC for aragonite saturation state with bivalves ranged from
0.07 to 4.72 among species. To capture this variation, a cumulative frequency distribution curve was
generated from the most sensitive MinATC for each species within a taxon. Based on breaks in the
frequency distribution, high, moderate, and low sensitivity threshold classes of species were identified
for each taxon (Figure 2-4). Then, cutpoints were calculated within each sensitivity threshold class for
8
-------
high, moderate, low, and minor risks. The pH outpoints for risks associated with the high, moderate, and
low pH sensitivity threshold classes for decapods and fish are given in Table B-l 1, Table B-12, and
Table B-13. The cutpoints for risks associated with the three sensitivity threshold classes for aragonite
saturation state for bivalves are given in Table B-14.
Which sensitivity threshold class a species is assigned to can have a major effect on the predicted risk.
Because of the lack of information on the sensitivity of most species, we used the moderate sensitivity
threshold classes as the defaults for all taxa, with two exceptions. The first was that congeners of species
used in ocean acidification experiments were assigned the sensitivity threshold class of the test species,
with the exception oiMytilus which spanned two sensitivity threshold classes. The second exception
was that species with brood protection or lecithotrophic larvae were assigned to the low sensitivity
threshold class (e.g., Lucey et al., 2015; see Section 6.4 in Lee et al, 2017). After assigning a species to a
sensitivity threshold class, risk was calculated by overlaying the pH or aragonite saturation state effects
cutpoints on the projected values for each ecoregion. Another factor affecting risk is the interaction
between ocean acidification and elevated temperatures. To account for this interaction, moderate ocean
acidification risks were elevated to high risk under moderate and high temperature risks (see Section 6.4
in Lee et al., 2017).
The CMIP5 model served through NOAA's Climate Change Web Portal was used as the source for the
historical and projected pH values for surface waters. The aragonite saturation state values were
generated from projections developed by Cao and Caldeira, which were derived from the University of
Victoria Earth System Climate Model version 2.8 (see Cao and Caldeira, 2008). The historical and
projected values for pH under RCP 4.5 and 8.5 are documented in Appendix B (Table B-l through
Table B-6). The historical and projected values for aragonite saturation state for both RCPs are
documented in Table B-7 and Table B-8.
9
-------
Sidebar #2. Derivation of the "Minimum Acceptable Toxicant Concentration" Metric
We draw upon the extensive history of assessing contaminant effects to derive our approach to
assessing the regional-scale effects of ocean acidification. Specifically, we derived our approach from
the "maximum acceptable toxicant concentration" (MATC), which is used to estimate the
"acceptable" concentration of pollutants and pesticides. MATCs have been used in ecological risk
assessments (U.S. EPA, 1998), in evaluating soil contamination for Superfund (EPA, 2003),
evaluating chemicals under the Toxic Substances Control Act (TSCA) (Nabholz, 1991), pesticides
under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) (EPA, 2004; Fairchild et al.,
2009), and as a freshwater sediment toxicity benchmark (Nowell et al., 2016).
The MATC is calculated as the geometric mean of the "no observed adverse effects level" (NOAEL),
the highest test level for 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 are significantly different than the controls or non-significant exposure concentration (U.S.
EPA Risk Assessment Forum, 1998). MATCs are generally based on chronic tests, and have been
referred to as "chronic values" (CV) (U.S. EPA, 1998).
Because the effects of ocean acidification result from reductions in pH, the objective of the ocean
acidification metric is to calculate the minimum acceptable pH or aragonite saturation state value
rather than the maximum acceptable concentration as with pollutants. Thus, to avoid confusion, we
refer to the ocean acidification metric as the "minimum acceptable toxicant concentration" (MinATC)
and replace the LOAEL with the "highest observed adverse effects level" (HOAEL) for pH and
aragonite saturation state (note that this terminology was not used in Lee et al., 2017). For
consistency, we use the same basic formula for MinATCs as for MATCs. However, there are slight
differences in how it is calculated for pH versus aragonite saturation state:
MinATC - pH: Because pHs are reported in logio units, the geometric mean is calculated on
the antilogs of the pHs associated with the NOAEL and HOAEL. The resulting geometric
mean is then back transformed into a pH by taking the logio.
MinATC - aragonite saturation state: The geometric mean is calculated directly using the
NOAEL and HOAEL values.
The MinATC are a formal way to integrate pH/aragonite levels with an observed effect versus no
effect. Given the state of the science, we suggest that the MinATCs are a viable first-order approach
to identifying geographical risk patterns and differences in relative risks among taxa. However, as
discussed in the text, they have limitations from being based on both sublethal and lethal effects. We
caution that MinATCs are not exactly equivalent to MATCs for pollutants/pesticides. If, in the future,
a consistent set of chronic effects studies become available, and researchers use a single pH
measurement (see "pH" in the Glossary), it should be possible to generate MinATCs equivalent to
MATCs.
10
-------
Cumulative Percent Most Sensitive MinATCs
Bivalves-Aragonite Saturation State
Aragonite Saturation State (OJ
Figure 2-4. Cumulative distribution of the most sensitive MinATCs for aragonite saturation state (Da) for each
bivalve species.
The MinATCs are based on all endpoints and life history stages. Three groups are identified: high sensitivity
species (red), moderate sensitivity species (orange) and low sensitivity species (green). Minor, low, moderate,
and high risks are then generated from the values in each of the sensitivity groups.
2.8 Sea Level Rise Risks
Sea level rise risks are generated from estimates of the extent of loss of key intertidal habitats, assuming
that the extent of population decline in intertidal species occupying these habitats is proportional to the
habitat loss (Section 7 in Lee et al., 2017). The risk calculation integrates four steps:
1. Estimate a net ecoregion sea level rise value (mm) for each ecoregion based on the global
eustatic rate and regional rates of isostatic adjustment.
2. Generate "habitat thresholds" for each of the major intertidal habitat types based on the literature
and SLR models. Examples of major habitats include lower marsh, tide flats, and rocky intertidal
(see Table 7-3 in Lee et al., 20107). 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
11
-------
used for Puget Sound through Southern California and the unconstrained thresholds used for the
less developed ecoregions.
3. Generate risk values for the target species based on the habitat thresholds and species' depth
preferences.
4. Because many near-coastal species occupy multiple habitats, the final step assigns the greatest
SLR risk across all observed and preferred habitats occupied by the species.
We assigned an eustatic rate of 12 mm/yr (1.2 m in 100 years) as the sea rise associated with RCP 8.5.
This is considered an "Intermediate-High" scenario (Parris et al., 2012). For the RCP 4.5 scenario, we
used an eustatic rate of 5 mm/yr (0.5 m in 100 years). The ecoregion-specific isostatic rates and net sea
level rises associated with both RCPs are documented in Appendix C.
2.9 Baseline/Status Risks
Baseline risks are risks resulting from inherent biotic traits of species that increase (or decrease) their
vulnerability to climate change, such as habitat specialization, symbiotic relationships, or endemicity.
Status risks are ecosystem-specific changes in a species' population viability (e.g., population decline)
due to external factors, such as overfishing, that increase vulnerability to climate change. We generated
17 rules predicting vulnerability or resilience to climate change, with many of the rules modified by the
relative abundance of the species in an ecoregion (Section 4 in Lee et al., 2017). At this time, only the
rules predicting increased risk are incorporated into the calculation of climate risk, though the resilience
traits are included in the outputs. Since these baseline/status rules are meant to capture the additional
risk under climate stress, they are only invoked when there is moderate or high risk from temperature
increases, ocean acidification, and/or sea level rise.
Both the baseline traits and ecoregion-specific relative abundances for the brachyuran and lithodid crabs,
Sebastes, and bivalves can be downloaded using the "Basic Search" function in CBRAT (see Section IX
in the CBRAT User's Guide, Lee et al., 2015).
12
-------
Section 3. Comparison of Risks with RCP 4.5
vs. RCP 8.5 Scenarios
3.1 Introduction
The effects of climate change can be quantified using a variety of metrics, depending upon objectives
and spatial scale. Because the objectives of the current case study are to evaluate differences in the
geographic and taxonomic risk patterns between the RCP 4.5 and RCP 8.5 scenarios, we sum the risks
across individual species. Users can output the risks for individual species by following the procedures
outlined in Appendix B in Lee et al., 2017. The current case study evaluates risk reductions at two
spatial scales. The first is the coastal scale which sums the risks across all species in a taxon across the
ten ecoregions of the U.S. Arctic and NEP. The second is at the scale of the individual ecoregions. The
following metrics were used in the assessments:
Overall Vulnerability. The single risk assigned to a species based on the greatest risk among the
climate drivers: temperature, ocean acidification, SLR, and the baseline/status traits (see Section
2.3 in Lee et al., 2017). Overall vulnerability is not modified if a species has the same risk level
for more than one climate driver. Overall vulnerability is distinct from risks associated with
individual climate drivers, such as ocean acidification.
At-Risk. Species are classified as At-Risk if one or more of the climate drivers indicates a
moderate risk or a high risk. We consider it likely that At-Risk species will be affected by
climate change. By combining moderate and high risks into this single metric we can identify the
broad patterns as well as address the issue of different sensitivities in the ETW and BTL
approaches (see Section 5.4.3 in Lee et al., 2017). The At-Risk metric can be applied at the
ecosystem or coastal scale.
Sum-Occupied Ecoregions: Sum of the number of ecoregions within the U.S. Arctic and NEP
occupied by the species within a taxon (Figure 3-1). This is equal to the sum of the number of
species in each ecoregion across the ten ecoregions; the number of Sum-Occupied ecoregions for
each taxon is given in Table 3-1. The effects of climate change are measured as the total number
of Sum-Occupied ecoregions by all the species within a taxon with high risks, moderate risks, or
At-Risk. The Sum-Occupied ecoregions combine both the number of vulnerable species and
their geographic distributions, and is a coastal scale metric.
Percent of At-Risk Species: At a coastal scale, the percent of the total number of species within a
taxon occurring in the U.S. Arctic and NEP at moderate or high risk. The percent species at
moderate or high risk is also calculated at an ecoregion scale based on the number of species
occurring within each of the ecoregions. The total number of species in the U.S. Arctic and NEP
and the numbers of species per ecoregion are given in Table 3-1.
13
-------
Percent decrease: The decrease in risk from the RCP 8.5 scenario to the RCP 4.5 scenario is
calculated as:
value at 4,5 - value at 8.5
value at 8.5
X 100
For example, if the number of brachyuran species with high risks under RCP 4.5 is 48 compared
to 67 under RCP 8.5, the percent reduction under RCP 4.5 is -28% =
48 - 67
67
X 100
This measure is scaled from 0% (no difference) to -100% decrease (no species at risk under RCP
4.5), and can be applied at a coastal or ecoregion scale. It is undefined if there are no species at
risk under the RCP 8.5 scenario.
Species 1
"Sum-Occupied Ecoregions"
Metric Integrating # Species & Distributions
Risk Levels for Species 1:
1 High Risk ecoregion
1 Moderate Risk ecoregion
2 Minor Risk ecoregions
: 4 Sum-Occupied Ecoregions
I = Minor Climate Risk
= Low Climate Risk
= Moderate Climate Risk
| = High Climate Risk
Species 1 + Species 2
COMBINED:
1 High Risk Ecoregion
1 Moderate ecoregion
2 Minor ecoregions
2 High Risk ecoregion
2 Minor Risk ecoregions
= 8 Species X Ecoregions
Sum-Occupied Ecoregions =
3 High Risk
1 Moderate Risk
4 Minor Risk
Species 2
2 High Risk ecoregions
2 Minor Risk ecoregions
= 4 Sum-Occupied ecoregions
Figure 3-1. Example of the number of Sum-Occupied ecoregions based on two species.
Combining the two species, the Sum-Occupied ecoregions equals eight ecoregions, with three at high risk, one at
moderate risk, and four at minor risk. For a taxon, the Sum-Occupied ecoregions is the number of occupied
ecoregions in the U.S. Arctic and NEP summed across all species.
14
-------
3.2 Coastal Risk Patterns: Sum-Occupied Ecoregions with Species at Moderate or High
Risks
The first analysis at the coastal scale evaluated the decrease in risk with RCP 4.5 compared to the 8.5
scenario based on the Sum-Occupied ecoregions with species classified At-Risk using Overall
Vulnerability (Figure 3-2). This metric summarizes the risks from Southern California through the
Beaufort Sea, and provides an overview of the reduction in risk with RCP 4.5. The brachyuran crabs and
bivalves showed substantial reductions in the percent of Sum-Occupied ecoregions classified as At-Risk,
-69% and -57%, respectively. In comparison, the lithodid crabs and Sebastes showed moderate
reductions, -19% and -17%, respectively.
The second analysis, which evaluated the percent changes in Sum-Occupied ecoregions with moderate
versus high risks, is more illuminating (Figure 3-2). There were substantial decreases in the percent of
ecoregions at high risk, ranging from a 32% decrease for lithodid crabs to a 93% decrease for bivalves.
However, in three cases (lithodid crabs, Sebastes, and bivalves) the percent of ecoregions at moderate
risk increased by 38% to 125%. Though there wasn't an increase in the moderate risks with the
brachyurans, the decrease in the percent of ecoregions with moderate risk was about a third smaller than
the decrease with high risks (Figure 3-2). Thus, a substantial number of species in each tax on were
predicted to have their risk level reduced from high under RCP 8.5 to moderate under RCP 4.5.
-125
Differences between RCP 4.5 and RCP 8.5
Percent Sum-Occupied Ecoregions
Brachyurans Lithodid Sebastes
Bivalves
125
« At Risk1
High
Moderate
Figure 3-2. Differences in the percent of Sum-Occupied ecoregions with species at moderate risk, high risk, and
At-Risk under RCP 4.5 compared to RCP 8.5.
Values are based on the sums across all ten ecoregions. At-Risk = moderate or high risk.
15
-------
3.3 Patterns of Risk Across Ecoregions.
To evaluate geographic patterns of risk reduction, we use the percent of At-Risk species in each
ecoregion based on Overall Vulnerability (Figure 3-3 through Figure 3-56). Because the percent of At-
Risk species is partially dependent upon the number of species in an ecoregion, the numbers of species
per ecoregion are summarized in Table 3-1 for context.
All taxa displayed substantial risks to climate change in several ecoregions, though there was wide
variation in the percent of At-Risk species both within and among taxa. For example, under RCP 8.5, the
percent of At-Risk species ranged across ecoregions from 19% to 100% with brachyuran crabs and from
6% to 86% with Sebastes (Table 3-2). In addition to the variation, there was no consistent geographical
pattern as to where the greatest risk occurred. Brachyuran crabs showed the greatest risk in the Beaufort
Sea (Figure 3-4), though this is based on only three species (Table 3-1). In comparison, the lithodid
crabs showed the greatest risk in the Chukchi Sea and in Southern California Bight ecoregions (Figure
3-5) while Sebastes showed a strong spike in the percent of At-Risk species in the Southern California
Bight Ecoregion (Figure 3-6). Bivalves showed a substantial percentage of At-Risk species across all
ecoregions with RCP 8.5 (Figure 3-6).
The assessment of the extent of risk reduction at the ecoregion scale depends upon how the data are
analyzed. The ranges of the percent At-Risk species across the ecoregions showed little change under
RCP 4.5 compared to RCP 8.5 in three of the four taxa (Table 3-2). Thus, for these taxa, the RCP 4.5
scenario did not substantially change the lower or upper limits of the percent of At-Risk species within
the ecoregions. Only the bivalves showed a substantial reduction in both the upper and lower limits of
the range. When evaluated as the median reduction in the percent of At-Risk species across the
ecoregions, both the lithodid crabs and rockfish showed small reductions with RCP 4.5 (Table 3-2). In
comparison, the brachyuran crabs and bivalves showed moderate and substantial reductions,
respectively. The next analysis was to calculate the number of ecoregions in which the taxon showed a
reduction in the percent of At-Risk species (Table 3-2). With three of the four taxa, the majority of the
ecoregions showed a reduction in the percent of At-Risk species, with the bivalves showing a reduction
in every ecoregion. The exception was the lithodid crabs which showed a reduction in only four of the
nine occupied ecoregions.
When evaluated along the U.S. Pacific Coast, the extent of risk reduction with RCP 4.5 across
ecoregions failed to show a simple geographical pattern. For example, the bivalves showed substantial
reductions in the percent of At-Risk species in the three Arctic ecoregions (Beaufort, Chukchi, Eastern
Bering) while the brachyuran and lithodid crabs showed no reductions in the Arctic ecoregions. The one
spatial trend was for the North Pacific Fijord, Puget-Georgia Basin, and Oregon, Washington,
Vancouver Coast and Shelf ecoregions to have reduced At-Risk species with all four taxa, though the
percent reduction varied among the taxa (Figure 3-3 through Figure 3-6).
16
-------
% Brachyura "At-Risk" - Overall Vulnerabilty
under RCP 4.5 and RCP 8.5
(f)
<0
^
^ ^ OP o*
* ^ J-*
"o
e
<1)
O
<1)
CL
80%
70%
60%
50%
40%
30%
20%
10%
0%
na na
.CT
75 75
38
38
35
13 13 13
25
47
22
33
25
50
42
80
42
80
^ s/
aV-
-------
-SC
in
ro
w
0)
'o
0)
Q.
(/)
-4—'
c
a>
o
<5
0-
% Sebastes "At-Risk" - Overall Vulnerabilty
under RCP 4.5 and RCP 8.5
80%
50%
40%
30%
20%
10%
0%
86
na na na na 6 6
14
10
13
15
86
y / y /¦ >°
* * ^
RCP 4.5 ¦ RCP 8.5
f /
Figure 3-5. Percent of Sebastes species within each ecoregion At-Risk with RCP 4.5 and 8.5.
Based on Overall Vulnerability. At-Risk = moderate or high risk.
% Bivalves "At-Risk" - Overall Vulnerabilty
under RCP 4.5 and RCP 8.5
<0
TO
W
O
"o
<1)
CL
V)
c
0)
o
1—
<5r
&
J* /
RCP 4.5 ¦ RCP 8.5
Figure 3-6. Percent of bivalve species within each ecoregion At-Risk under RCP 4.5 and 8.5.
Based on Overall Vulnerability. At-Risk = moderate or high risk.
18
-------
Table 3-1. Number of species per ecoregion, total number of species in the U.S. Arctic and Northeast Pacific ecoregions, and total number of occupied
ecoregions.
Transient species were excluded from the species counts. Total number of occupied ecoregions is the sum of the number of ecoregions in the U.S. Arctic
and NEP occupied by the species in the target taxon, which is equal to the sum of the number of species in each ecoregion.
Ecoregion/
Taxon
Beaufort
Chukchi
E.
Bering
Aleutian
Gulf
Alaska
N.
Pac.
Fijord
Puget-
Georgia
OR-WA-
VANC
N.
California
S.
California
Total #
Species
Total #
Occupied
Ecoregions
Brachyura
3
7
14
18
26
35
27
43
59
115
135
347
Lithodoidea
0
4
8
15
16
17
9
12
12
10
21
103
Sebastes
0
0
18
22
37
39
26
52
62
58
68
314
Bivalvia
66
88
115
96
144
168
139
197
221
350
470
1584
Table 3-2. Comparison of the ranges of percent of At-Risk species and median percent of At-Risk species with RCP 4.5 vs. 8.5, and number of ecoregions
with reduced percent of At-Risk species with RCP 4.5.
# Occupied Ecoregions = number of the ten U.S. Arctic and NEP ecoregions occupied by the taxon. At-Risk = moderate or high risk.
Taxon
RCP 4.5
Range (%)
RCP 8.5
Range (%)
RCP 4.5
Median (%)
RCP 8.5
Median (%)
Reduction in
Medians
With RCP 4.5
# Ecoregions with
Reduced %
Species At-Risk
(# Occupied Ecoregions)
Brachyura
19.2-100
19.2-100
36.1
54.7
18.6%
9
(10)
Lithodoidea
12.5-80
13.3-80
35.3
41.7
6.4%
4
(9)
Sebastes
0 - 86.2
5.6-86.2
7.69
11.9
4.2%
5
(8)
Bivalvia
7.7-75
77.1 - 100
44.4
83.6
39.2%
10
(10)
19
-------
Section 4. Northern Colonization
4.1 Introduction
Another climate effect is the colonization of northern ecoregions by southern species in response to
ocean warming. Several studies have reported northern migration of marine species in response to recent
temperature increases (e.g., Pinsky et al., 2013; Hale et al., 2017) and such colonization would
presumably increase in response to further warming. These northern colonizers may be predators or
competitors with the existing species, potentially disrupting food webs and key ecosystem services much
like invasive species (e.g., Nichols et al., 1990; Byers, 2000; Katsanevakis et al., 2014). Conversely, the
introduction of a commercial species could potentially increase fishing opportunities, such as with the
deliberate introduction of the red king crab into the Barents Sea (Orlov and Ivanov, 1978), though the
establishment of this crab may be negatively affecting other ecosystem services (Jorgensen and
Primicerio, 2007).
To evaluate the possibility of northern colonization, we reversed the logic of the BTL approach and
assessed whether cooler, unoccupied ecoregions would become sufficiently warm to allow southern
species to survive (Section 2.6; Section 5.5 in Lee et al., 2017). The northern colonization analysis
evaluates the suitability of projected temperatures in cooler ecoregions. However, the predictions do not
currently take into account other factors that could limit a species' ability to become established in a
northern ecoregion, such as lack of suitable habitat.
4.2 Northern Colonization Projections
The BTL approach was utilized to assess thermal suitability for all the depth ranges a species occupies,
including air temperature, SST, 30-m temperature, and 100-m temperature. The suitability of the
projected temperatures was evaluated by taking the lowest suitability score for all occupied depth
classes. Thus, for a species that occurred from the intertidal to a 27 m depth, the annual, summer, and
winter air temperatures, SST, and 30-m temperature would all be evaluated and the lowest suitability
score used. Evaluating thermal suitability at all depths is more stringent than relying only on SSTs. In
the present analysis, both moderate and high temperature suitability are combined as "potential
colonizers" (analogous to At-Risk). Note, however, that suitability in the Puget Sound/Georgia Basin
Ecoregion is based only on air temperatures and SST since the subsurface temperature projections are
not available.
At the coastal scale, the percent of species classified as potential northern colonizers relative to the
existing species in the U.S. Arctic and NEP ranged from more than 45% of all bivalve species to 70% of
the brachyuran crab species under RCP 8.5 (Figure 4-1). The percent of northern colonists declined
under the RCP 4.5 scenario from, ranging from a 6% reduction for rockfish to a 48% reduction for
brachyuran crabs. Even with these reductions, 21% of the brachyuran crabs to 56% of the Sebastes were
classified as potential colonizers under RCP 4.5.
20
-------
There is a wide vari ation in the percent of potential colonists at the ecoregion scale (Figure 4-2 through
Figure 4-5). Among the true crabs, king crabs, and rock fish, one or more ecoregions have no predicted
colonists under the RCP 8.5 scenario. Alternatively, with both the true crabs and king crabs there were
ecoregions where the percent of potential colonists is predicted to equal or exceed the number of
existing species under RCP 8.5, and every taxon has at least one ecoregion where the potential colonists
exceed 40% of the current species. There is also a wide variation in the reduction in potential colonists
under RCP 4.5. With the brachyuran crabs in the Beaufort Sea and Sebastes in the Aleutians, reducing
emissions to the RCP 4.5 scenario did not result in any reduction. In other cases, however, there was a
50% or greater reduction; for example, no brachyuran or bivalve colonists were predicted in the
Northern and Southern California ecoregions with RCP 4.5.
OA
% Species as Potential Colonists - RCP 4.5 &
8.5
70
1
6,
so
1 56
o
d) Af\
_
ID
Q.
2 30
38
O)
(J Ofl
30
i_
0)
Q_ in
¦
A
¦ 1
¦ ¦
¦ ¦
Brachyura Lithodoidea Sebastes Bivalves
¦ RCP 8.5 RCP 4.5
Figure 4-1. Percent of the species in the U.S. Arctic and NEP classified as potential northern colonists under RCP
4.5 and RCP 8.5.
Based on moderate and high temperature suitability for all depths using the BTL approach.
21
-------
Northern Colonists as Percent of Existing Species -
Brachyura RCP4.5 & 8.5
140
¦ RCP8.5 RCP4.5
Figure 4-2. Brachyura - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios.
Based on moderate and high temperature suitability for all depths using the BTL approach. The numbers of
species per ecoregion are given in Table 3-1.
22
-------
Northern Colonists as Percent of Existing Species -
Lithodoidea RCP 4.5 & 8.5
v>
OJ
o
0)
Q.
(f)
c
a>
o
£
CL
100
90
80
70
60
50
40
30
20
10
0
J
?f>
0 0 0 0
g|-f| ^ 6 00 00 00 00 00
JP
~ ^
,xO
SP <
S*
S°
tO
i RCP 8.5 RCP 4.5
Figure 4-3. Lithodoidea - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios.
Based on moderate and high temperature suitability for all depths using the BTL approach. The numbers of
species per ecoregion are given in Table 3-1.
23
-------
Northern Colonists as Percent of Existing Species -
Sebastes RCP4.5 & 8.5
100
90
¦ RCP8.5 RCP4.5
Figure 4-4. Sebastes - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios.
Based on moderate and high temperature suitability for all depths using the BTL approach. The numbers of
species per ecoregion are given in Table 3-1.
24
-------
Northern Colonists as Percent of Existing Species -
Bivalves RCP 4.5 & 8.5
50
45
40
¦ RCP 8.5 RCP 4.5
Figure 4-5. Bivalves - northern colonists as percentage of the existing species in each ecoregion with RCP 4.5
and 8.5 scenarios.
Based on moderate and high temperature suitability for all depths using the BTL approach. The numbers of
species per ecoregion are given in Table 3-1.
25
-------
Section 5. Discussion
5.1 Introduction
The current case study had two main objectives. The first was to analyze the taxonomic and geographic
patterns in risk reduction between the RCP 4.5 and 8.5 scenarios. The second was to evaluate the
efficacy of CBRAT in conducting regional-scale climate risk assessments for multiple taxa. In turn, this
second objective consisted of three facets: 1) user friendliness of conducting risk assessments and
comprehensiveness of the assessments; 2) practicality of adding information required to analyze a new
taxon; and 3) utility of CBRAT in conducting uncertainty assessments. The last of these, conducting
uncertainty analyses, is discussed in Section 6.1.
5.2 Case Study - Efficacy of CBRAT in Conducting Risk Assessments
In terms of conducting risk assessments, CBRAT calculated the risks for each of the four taxa under
both RCPs with no known issues. After changing input values or taxon, a new risk analysis could be
finalized in a reasonable time, about 15-20 minutes. One challenge, however, was the need to replace all
the RCP 8.5 default projections with the RCP 4.5 values by hand, which was both time consuming and
susceptible to error. In response, we added a new functionality to automatically load either the RCP 4.5
or 8.5 ecoregion-specific values for temperature, ocean acidification, and eustatic sea level rise.
CBRAT facilitates both a broad and detailed analysis of risk patterns by outputting a CSV file that
summarizes the risks for each species in each occupied ecoregion. The broad view provided by the
Overall Vulnerability score was the metric used in this case study. The output also includes the single
greatest risk value for temperature, ocean acidification, SLR, and baseline/status traits. An example of
using risks associated with one of the climate drivers is the uncertainty analysis that evaluated only
ocean acidification risks. In addition, the CSV file includes each risk incorporated in each of these
climate drivers. For example, the output includes the individual risks associated with annual, summer,
and winter air temperatures, 30-m depth temperature, 100-m depth temperature, and the SST from the
BTL analysis. Relatively straightforward spreadsheet manipulations of the CSV output allow users to
address a range of detailed questions. A limitation of the spreadsheet analysis, especially for less
technical users, is "seeing" the overall geographical risk patterns. To provide a gestalt view, we will add
the functionality to produce PDF maps of each species' risks by ecoregion in the next version of
CBRAT.
5.3 Case Study - Adding New Taxon
To evaluate the process and practicality of conducting risk assessments on additional taxa, we added the
required biotic information for bivalves as part of this case study. Bivalves were chosen as the test taxon
because of the large number of species, range of life history traits, and economic importance. One
challenge with adding any new taxon is the extensive range of biotic data that needs to be synthesized to
calculate climate risks: 1) biogeographic distributions; 2) regional relative abundance patterns; 3)
preferred depth range; 4) habitat preferences; 5) any specialized habitat requirements; 6) any symbiotic
26
-------
relationships; 7) any specialized trophic relationships; 8) breeding traits (e.g., brooding); 9) larval type
(e.g., lecithotrophic); 10) ecoregion-scale population trends; and 11) productivity metrics (for fish only).
Synthesizing this information requires a dedicated effort and access to a wide range of information
sources including peer-reviewed literature, agency reports, and online databases, and potentially
interactions with local and national experts. An additional challenge to adding a new taxon is the need to
synthesize ocean acidification literature to generate pH and/or aragonite saturation state thresholds. Our
guideline has been to synthesize experiments for at least 25 different species to generate such thresholds,
which is not a trivial exercise (see Table B-9 and Table B-10).
One approach to expediting the data synthesis is to create a PDF library of pertinent literature. We have
a library of over 67,000 publications. Using file searching software, such as dtSearch™, it is possible to
search the literature for key traits or sensitivity to ocean acidification. Another labor-saving approach is
to summarize biotic traits at higher taxonomic levels, such as at the family level. Currently, there is a
beta version of the "biotic matrix" in CBRAT that populates individual species' traits from traits entered
at the genus level or above. A key feature of the biotic matrix is the functionality to accommodate
exceptions to traits at higher levels (e.g., all sponges are suspension feeders except species in the family
Cladorhizidae which are predators), thus capturing both general and specific traits.
From our experience with the bivalves, we conclude that that with the exception of problematical groups
(e.g., nematodes), it is practical to add new taxa. In general, the greater the number of species in the
taxon, the greater the effort. As a rough guide, we estimate it would take 6 months to a year to add the
required information for a new taxon equivalent to the bivalves. This would require one-quarter to one-
half year time of a technician and/or researcher conversant with CBRAT, taxonomy, and natural history.
We found an efficient approach is to have one contractor working with one EPA researcher. In addition,
consultation with experts may be an effective way of capturing required information; for example, we
held three workshops on the bivalves. While not highly expensive, such workshops usually require at
least funding for travel by the experts. Adding a new location (e.g., U.S. East Coast) is analogous to
adding a new taxon in terms of synthesizing the biotic traits of the species in each new taxon in the new
ecoregions. One savings is that the thresholds for pH/aragonite do not need to be redone. However,
climate projections would have to be generated for each new ecoregion, requiring several months of
analysis by a GIS practitioner.
27
-------
5.4 Risk Reduction. Comparing the RCP 4.5 Scenario to the RCP 8.5 Scenario
At the coastal scale, the percent of Sum-Occupied ecoregions with species At-Risk under the RCP 4.5
scenario was substantially reduced compared to RCP 8.5 for two of the taxa (brachyuran crabs and
bivalves) and moderately reduced with the other two (lithodid crabs and rockfish). These reductions in
At-Risk species indicate that climate impacts would occur with fewer species and fewer locations along
the U.S. Pacific Coast. When the risks were parsed out to high and moderate risks, a key pattern is the
substantial reductions in the percent of Sum-Occupied ecoregions with high risks with all four taxa
under RCP 4.5. These reductions in high risks ranged from -32% for the lithodid crabs to -93% for the
brachyuran crabs. For three of the taxa, there was a corresponding increase in the percent of ecoregions
with a moderate risk, indicating that risks were being reduced from high to moderate. Thus, under RCP
4.5, many species and locations would be at a lower risk to climate change.
At an ecoregion scale, all four taxa showed reduced percents of At-Risk species under RCP 4.5 but there
were no consistent taxonomic or geographical patterns. The bivalves showed the greatest reductions, as
indicated by having the greatest predicted decrease in the median percent of At-Risk species within
ecoregions (-39%) and by having predicted reductions in the percent of At-Risk species in all ten
ecoregions. In comparison, the lithodid crabs and rockfish showed slight decreases (-4% to -6%) in the
median percent of At-Risk species, and the lithodid crabs showed only a decrease in the At-Risk species
in four of the nine occupied ecoregions. Geographically, there was no pattern as to where the risk
reductions occurred. The bivalves and brachyuran crabs showed moderate to substantial decreases in the
Southern California Bight Ecoregion, while the lithodid crabs and rockfish showed no decrease. The one
weak pattern was that all four taxa generally showed reductions in the percent of At-Risk species in the
North Pacific Fijord, Puget-Georgia Basin, and Oregonian ecoregions, though the percent reduction
varied among the taxa.
On reflection, it is not surprising that there are no simple geographic or taxonomic patterns because of
the multitude of factors affecting climate vulnerability. Each of the following factors can influence the
extent of risk within an ecoregion and hence the geographical and taxonomic patterns of risk as well as
the extent of risk reduction with RCP 4.5:
• Geographical patterns in the projected changes in temperature, ocean acidification, and sea level
rise, each of which have different regional patterns.
• Differences in the sensitivity among taxa to changes in temperature, sea level rise, and ocean
acidification.
• Differences in the regional abundance patterns among taxa (e.g., cold-water vs. warm-water
taxa).
• Differences in the southern range limit of species across taxa.
• Differences among taxa in latitudinal gradients in species richness, which in turn affect
calculations based on the percentage of species in an ecoregion.
• Differences in the geographical patterns of population trends.
28
-------
• Differences in the geographical patterns in the number of habitat specialists and symbionts
across taxa.
• Interactions between temperature and ocean acidification risks as well as between climate
stressors and baseline/status traits, both of which can result in complex patterns.
The lack of simple patterns in risk reduction implies that managers need to evaluate each major taxon, or
species, in each ecoregion of concern. We contend that CBRAT offers a practical tool to conduct first-
order regional assessments either for a taxon or select groups of species. In addition to the risk within a
single ecoregion, evaluating regional patterns of risk reduction would inform managers where adaptation
efforts would be most efficient. Though not addressed in this case study, evaluating which climate
drivers are most important in an ecoregion would help inform the design of effective adaptation
measures.
5.5 Northern Colonization
Northern colonization is likely to result in major, but poorly understood, ecological alterations along the
U.S. Pacific Coast. The number of species in the four taxa classified as potential northern colonists
under RCP 8.5 ranged from 44% to 70% of the total species in the NEP and U. S. Arctic (Figure 4-1).
These numbers were substantially reduced (ranging from -6% to -48%) under the RCP 4.5 scenario.
Nonetheless, the total number of potential colonist species remained high, with 21% to 56% of the
species within the four taxa predicted to potentially colonize currently unoccupied ecoregions.
As with the risks, there is considerable geographic variation in the number of potential colonists. The
predicted number of northern colonists exceeds the number of native brachyuran and lithodid crab
species in one ecoregion under RCP 8.5 (Figure 4-2 and Figure 4-3). In addition, the predicted number
of potential northern colonists is equal to at least 42% of the native species in one or more ecoregions
with the rockfish and bivalves under the RCP 8.5 scenario (Figure 4-4 and Figure 4-5). Conversely,
potential colonists were predicted to constitute less than 10% of the native species in a number of
ecoregions for each taxon under the RCP 8.5 scenario. The predicted reductions in the number of
colonists per ecoregion under RCP 4.5 also varies across taxa and ecoregions. Comparing the two RCPs,
there is at least one ecoregion with each of the taxa that has no predicted colonists under RCP 4.5.
Conversely, there is at least one ecoregion with each of the taxa in which there would be no change in
the number of potential colonists under RCP 4.5. Thus, as with risks, the RCP 4.5 scenario predicted
reduced numbers of northern colonizers but does not obviate all potential colonizers.
5.6 Overall Assessment of Risk Reduction under the RCP 4.5 Scenario
The multiple ways risk reduction can be quantified, the lack of consistent taxonomic and geographical
patterns, and the uncertainty inherent in climate predictions complicate coming to a simple conclusion
regarding the benefits of the RCP 4.5 scenario. Nonetheless, relying on the predicted changes in risks in
Sum-Occupied ecoregions at a coastal scale as the single best metric, the substantial reduction in high
risks is a strong indicator of reduced severity of impacts with the RCP 4.5 scenario (Figure 3-2).
Additionally, at least a moderate risk reduction is predicted in each of the ten ecoregions for one or more
of the taxa that occur in the ecoregion. Coupled with the reduction in risks, the predicted number of
29
-------
potential northern colonists was substantially reduced in each taxon (Figure 4-1). Based on these
predictions, we conclude that achieving the RCP 4.5 scenario would result in a substantial reduction in
the ecological risks to brachyuran crabs, lithodid crabs, rockfish, and bivalves along the U.S. Pacific
Coast. These four taxa encompass a wide range of life history traits, biogeographic distributions, and
sensitivities. Thus, we suggest that achieving RCP 4.5 is likely to reduce risks with other bottom-
associated taxa along the U.S. Pacific Coast as well.
30
-------
Section 6. Uncertainty Analysis and Quality
Assurance/Quality Control
6.1 Example of Uncertainty Analysis - Aragonite Saturation State Thresholds
As with all climate models, CBRAT is subject to a variety of uncertainties. Recognizing the importance
of evaluating uncertainty, CBRAT was designed to facilitate analysis of different climate projections,
effects thresholds, and assumptions. To evaluate this functionality, we conducted an uncertainty analysis
related to the aragonite saturation state (Qa) thresholds for bivalves. We chose this as the test parameter
because aragonite thresholds were classified as having "High" uncertainty in our earlier assessment
(Table 8-1 in Lee et al., 2017) and because ocean acidification is predicted to be a major stressor on
bivalves.
The pH and aragonite saturation state thresholds are based on results from experimental exposures that
include a range of behavioral and physiological response, as well as mortality (see Table B-9 and Table
B-10). In contrast, the temperature and sea level rise thresholds are more directly related to population
viability (see Section 3.3.3 in Lee et al., 2017). Inclusion of sublethal responses in the synthesis of
MinATCs is likely to generate more sensitive pH/aragonite thresholds than thresholds more directly
related to population viability. Potentially, this could result in an overestimation of ocean acidification
risks relative to temperature increases and sea level rise. To evaluate the sensitivity of risks to this
parameter, the high, moderate, and low sensitivity thresholds for aragonite saturation state were changed
from the default values by the following percentages: -50%, -25%, -20%, -15%, -10%, -5%, +5%,
+10%), and +25%. The default thresholds are given in Table B-14. After changing the threshold values,
risk analyses were conducted using the RCP 4.5 or RCP 8.5 climate projections for temperature,
aragonite saturation state, and eustatic sea level rise. For this analysis, we focused on the ocean
acidification risks only rather than Overall Vulnerability based on all climate drivers.
We evaluated sensitivity to changes in the thresholds by using the number of occupied ecoregions
classified At-Risk to ocean acidification compared to the number using default thresholds (Figure 6-1.).
With RCP 8.5, there were no or only minor differences in the number of ecoregions At-Risk resulting
from a -20% decrease to a +25% increase. Even with a -25% decrease, the percent of ecoregions At-
Risk was within two-fold of the default value. In comparison, there was a greater sensitivity with the
RCP 4.5 scenario. There was slightly more than a two-fold range resulting from a -15% decrease to a
10%) increase in the thresholds, and a three-fold or greater difference with larger percentage increases or
decreases in the thresholds.
As a measure of the extent of uncertainty, we use the guidelines of being correct 8 out of 10 times as low
uncertainty and 5 out of 10 as moderate uncertainty (Mastrandrea et al. 2010; Table 8-2 in Lee et al.,
2017). In the present context, we interpret this as the Sum-Occupied ecoregions with At-Risk species
being within 0.8X to 1.25X of the value generated using the default thresholds for low uncertainty and
31
-------
within 0.5X to 2X of the values for moderate uncertainty. Based on these criteria, there is low
uncertainty in ocean acidification risk predictions with RCP 8.5 resulting from likely ranges in the
aragonite saturation state thresholds. With RCP 4.5, the approximate two-fold range in risks indicates a
low uncertainty over moderate changes in the thresholds, but moderate or even high uncertainty with
greater changes. This difference in sensitivity between the two RCPs complicates making simple
conclusions regarding uncertainty, but points out the need to evaluate uncertainty under different
emission scenarios.
In terms of using CBRAT to conduct uncertainty analyses, running multiple risk assessments was
straightforward and relatively rapid, and a total or 20 risk analyses were conducted as part of the
uncertainty analysis. The limitation in conducting such analyses is related less to the functionality of
CBRAT than to the large number of potential scenarios, including different baselines, projections, and
thresholds for both the ETW and BTL temperature predictions, the pH and aragonite saturation state,
and sea level rise as well as different RCPs, all of which can be evaluated singly or in combination.
Additionally, uncertainty can be quantified at different scales using several different metrics. All of this
analyzed by each taxon. Because of the extensive number of possibilities, we recommend that
uncertainty analysis be focused on specific issues.
32
-------
Percent Sum-Occupied Ecoregions At-Risk to Ocean
Acidification
_ Bivalves - Uncertainty Analysis on Aragonite Thresholds
w 90%
B 80%
0%
Cl
Percent Change in Qa Thresholds (%)
~ % At-Risk 4.5 ¦ % At-Risk 8.5
Figure 6-1. Risks to bivalves from ocean acidification at different aragonite saturation state thresholds (Qa) under
RCP 4.5 and RCP 8.5.
Aragonite saturation state thresholds were generated as percent changes from the default values (in purple).
Default aragonite saturation thresholds changed by -50% to +25%, with the default thresholds in Table B-14.
Ocean acidification risks reported as the percent of Sum-Occupied ecoregions based on At-Risk species.
6.2 Overview of Uncertainty Related to Risk Reduction
A list of the potential sources of uncertainty and a characterization of the uncertainties related to climate
risk assessments are given in our framework document (Table 8-1 in Lee et al., 2017), and are not
repeated here. Rather, we provide a qualitati ve assessment of uncertainty as it primarily relates to an
assessment of risk reduction in a non-technical summary, as recommended by Salway and Shaddick (no
date):
Using the same set of risk rules coupled with a single biotic trait database promotes consistency
in risk predictions among taxa and ecoregions.
There is less uncertainty in the biogeographical patterns of risk for a taxon than for an individual
species.
75 75 75 75 75 77 77 75
56
80
47
11
17
17
7 7 I 7
11 rl n
-50 -25 -20 -15 -10 -5 0 5 10 25
25
25
33
-------
There is less uncertainty in evaluation of relative changes in risk than absolute risks.
The greatest uncertainty for risks is associated with ocean acidification.
The least uncertainty is for risk associated with temperature.
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.
Predictions are sufficient to identify the scope and patterns of risk and for regional-scale
adaptation planning.
Predictions are sufficient to flag vulnerable versus low risk commercial/recreational species but
not for fisheries management.
6.3 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 Procedures:
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.
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/ORD/R-17/052. 303 pages.
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 -rO.
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. QMPNHEERL/WED/1995-01-r4.0.
34
-------
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.
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.
35
-------
Appendix A. Baseline Temperatures and
Predicted Increases
The baseline temperatures and predicted temperature increases based on the RCP 4.5 and 8.5 scenarios
are documented in this appendix. The details of processing the baseline and projected temperatures are
given in Lee et al., 2017.
Table A-l and Table A-2 give the projected increases in SST for the RCP 8.5 and 4.5 scenarios,
respectively. These projections were used in calculating temperature risks with the ETW approach
(Section 2.5). Table A-3, Table A-4, and Table A-5 give the annual, summer, and winter thermal
thresholds for SST used in the ETW approach.
Table A-6 through Table A-17 give the historical, predicted increase, and projected temperatures for air
temperatures, SSTs, 30-m depth, and 100-m depth for the RCP 4.5 and 8.5 scenarios. These values were
used in the BTL approach (Section 2.5). Note that the baseline temperatures are modeled and can be
different between the RCP 4.5 and 8.5 projections because of different data used in model generation.
Table A-1. Projected annual, summer, and winter increases in SST by ecoregion under the RCP 8.5 scenario for
the 2050-2099 timeframe.
Summer is the average temperature for July, August, and September. Winter is the average for January,
February, and March. Projections based on CMIP5 model downloaded from NOAA's Climate Change Web Portal
ChttDi//www,esrl,noaa,aov/
osd/iDcc/ocn/'). Temperatures are in °C.
Ecoregion
Annual Increase
Summer Increase
Winter Increase
Beaufort
2 29
5 55
0.16
Chukchi
2 6
5 13
0.61
E Bering
3 56
4.03
2 92
Aleutians
3 03
3 63
2 53
Gulf Alaska
3.1
3 53
2 79
N Pac Fijord
2.8
3 18
2 53
Puget-Georgia
2.15
3 12
18
OR-WA-Vanc
2 62
2 9
2 41
N California
2.54
2 83
2 34
S California
2.4
2 38
2 34
Table A-2. Projected annual, summer, and winter increases in SST by MEOW ecoregion under the RCP 4.5
scenario for the 2050-2099 timeframe.
Summer is the average temperature for July, August, and September. Winter is the average for January,
February, and March. Projections based on CMIP5 model downloaded from NOAA's Climate Change Web Portal.
Temperatures are in °C.
Ecoregion
Annual Increase
Summer Increase
Winter Increase
Beaufort
1 26
3 4
0 04
Chukchi
14
3.1
0.13
36
-------
Ecoregion
Annual Increase
Summer Increase
Winter Increase
E Bering
2 26
2 65
1 72
Aleutians
2.09
2 5
1.77
Gulf Alaska
2 07
2 3
1.92
N Pac Fijord
1.89
2.15
1.71
Puget-Georgia
1.34
1.91
1.16
OR-WA-Vanc
1.77
1 93
1 63
N California
1.67
1.81
1.56
S California
1.61
1.61
1.57
Table A-3. Annual mean SST temperature thresholds by ecoregion.
Ranges derived from analysis of an analysis of 28 years of "advanced very high resolution radiometer" (AVHRR)
remote sensing data (Payne et al., 2011, 2012a, 2012b, and unpublished). Lowest value under Minor Risk is the
ecoregion annual mean. Temperatures are in °C.
Ecoregion
Minor Risk
Low Risk
Moderate Risk
High Risk
(Historic mean SST
+ 1 SD in WOE)
(< Historic mean
SST + 2 SDs in
WOE)
(< Historic mean
SST + 3 SDs in
WOE)
(> Historic mean
SST + 3 SDs in
WOE)
Beaufort
0.03-0.46
0 47-0 9
0.91 - 1 34
>1.35
Chukchi
0.55-1.23
1 24- 1 92
1 93-2 61
>2.62
E Bering
3.75-4.32
4 33-4 9
4 91 -5 47
>5 48
Aleutians
5 67-6 06
6 07-6 47
6 48-6 87
>6.88
Gulf Alaska
7 42-7 89
7 9-8 38
8 39-8 87
>8.88
N Pac Fijord
9 47-9 92
9 93- 10 38
10 39- 10 84
>10 85
Puget-Georgia
10 44- 10 93
10 94- 11 43
11.44-11.94
>11 95
OR-WA-Vanc
11.51 - 12 06
12 07 - 12 61
12 62- 13 17
>13 18
N California
13 55- 14 16
14 17 - 14 78
14 79- 15 4
>15 41
S California
17.81 - 18 39
184- 18 99
19 0- 19 58
>19.59
Table A-4. Summer mean SST temperature thresholds by ecoregion.
Summer is the average temperature for July, August, and September. Ranges derived from an analysis of 28
years of "advanced very high resolution radiometer" (AVHRR) remote sensing data (Payne et al., 2011, 2012a,
2012b, and unpublished). Lowest value under Minor Risk is the ecoregion summer mean. Temperatures are in
°C.
Ecoregion
Minor Risk
Low Risk
Moderate Risk
High Risk
(Historic mean SST
+ 1 SD in WOE)
(< Historic mean
SST + 2 SDs in
WOE)
(< Historic mean
SST + 3 SDs in
WOE)
(> Historic mean
SST + 3 SDs in
WOE)
Beaufort
1 77-3 07
3 08-4 38
4 39-5 68
>5.69
Chukchi
3 35-4 81
4 82-6 27
6 28-7 74
>7 75
E Bering
8 44-9 02
9 03-9 62
9 63- 10 21
>10 22
Aleutians
8 67-964
9 65- 10 62
10 63 - 11 6
>11.61
Gulf Alaska
11 82- 1249
12 5- 13 17
13 18 - 13 85
>13.86
N Pac Fijord
13 22- 13 94
13 95- 14 66
14 67- 15 38
>15.39
Puget-Georgia
13 59- 14 92
14 93- 16 26
16 27- 17 6
>17 61
OR-WA-Vanc
14 12- 14 98
14 99- 15 85
15 86- 16 72
>16.73
37
-------
Ecoregion
Minor Risk
(Historic mean SST
+ 1 SD in WOE)
Low Risk
(< Historic mean
SST + 2 SDs in
WOE)
Moderate Risk
(< Historic mean
SST + 3 SDs in
WOE)
High Risk
(> Historic mean
SST + 3 SDs in
WOE)
N California
15.19 -15.96
15.97-16.74
16.75-17.52
>17.53
S California
20.53 -21.4
21.41 -22.29
22 3 -23 17
>23.18
Table A-5. Winter mean SST temperature thresholds by ecoregion.
Winter is the average for January, February, and March. Ranges derived from an analysis of 28 years of
"advanced very high resolution radiometer" (AVHRR) remote sensing data (Payne et al., 2011, 2012a, 2012b, and
unpublished). Lowest value under Minor Risk is the ecoregion winter mean. Temperatures are in °C.
Ecoregion
Minor Risk
Low Risk
Moderate Risk
High Risk
Beaufort
-1.53 to -1.31
-1.3 to to1.07
-1.06 to -0.83
>-0.82
Chukchi
-1.26 to -0.76
-0.75 to -0.25
-0.24 to 0.26
>0.27
E Bering
1.0 to 1.54
1.55 to 2.08
2 09 to 2 63
>2.64
Aleutians
3.68 to 4.15
4.16 to 4.64
4.65 to 5.13
>5.14
Gulf Alaska
4.12 to 4.73
4.74 to 5.34
5.35 to 5.95
>5.96
N Pac Fijord
6 68 to 7 28
7.29 to 7.9
7.91 to 8.52
>8.53
Pugetto-Georgia
7.51 to 7.94
7.95 to 8.38
8.39 to 8.82
>8.83
OR-WA-Vanc
9.34 to 10.04
10.05 to 10.76
10.77 to 11.47
>11.48
N California
12.64 to 13.52
13.53 to 14.41
14.42 to 15.3
>15.31
S California
15 72 to 16 56
16.57 to 17.42
17.43 to 18.27
>18.28
Table A-6. Historical, predicted increase, and projected annual air temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C.
* = 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 Annual
Historical Mean
Air Annual
Predicted Increase
Air Annual
Projected Temperature
Beaufort
-11.66
8.34
-3.32
Chukchi
-11.44
8.97
-2.47
E Bering
-0.67
5.56
4.89
Aleutians
4.67
3.52
8.19
Gulf Alaska
4.84
3.73
8.57
N Pac Fijord
7.7
3.32
11.02
Puget-Georgia*
9.41
3.19
12.06
OR-WA-Vanc
11.12
3.05
14.17
N California
15.32
2.88
18.2
S California
19.38
2.78
22.1
38
-------
Table A-7. Historical, predicted increase, and projected annual air temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C.
* = 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 Annual
Historical Mean
Air Annual
Predicted Increase
Air Annual
Projected Temperature
Beaufort
-12.34
5.42
-6 92
Chukchi
-11.92
5.89
-6 03
E Bering
-1.04
3 83
2 79
Aleutians
4.62
2 32
6.94
Gulf Alaska
4.8
2.46
7 26
N Pac Fijord
7 96
2.11
10.07
Puget-Georgia*
9 63
2.0
11.63
OR-WA-Vanc
11.29
1.89
13.18
N California
14.84
1.76
16 6
S California
18.38
1.74
20.12
Table A-8. Historical, predicted increase, and projected summer air temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. Ecoregions are ordered by the lowest
historical mean summer temperature.
* = 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
1.36
5.04
6.4
Beaufort
1.69
5.1
6.79
E Bering
6.87
4.28
11.15
Aleutians
9.24
3.88
13.12
Gulf Alaska
10.17
3.79
13.96
N Pac Fijord
12.77
3.52
16.29
Puget-Georgia*
14.24
3.49
17.73
OR-WA-Vanc
15.7
3.46
19.16
N California
19.19
3.15
22.34
S California
23.02
2.81
25.83
39
-------
Table A-9. Historical, predicted increase, and projected summer air temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. Ecoregions are ordered by the lowest
historical mean summer temperature.
* = 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
1.0
2.99
3.99
Beaufort
1.17
3.01
4.18
E Bering
6.85
2.68
9.53
Aleutians
9.37
2.42
11.79
Gulf Alaska
10.39
2.32
12.71
N Pac Fijord
12.91
2.12
15.03
Puget-Georgia*
14.11
2.05
16.16
OR-WA-Vanc
15.31
1.98
17.29
N California
17.81
1.84
19.65
S California
21.38
1.71
23.09
Table A-10.Historical, predicted increase, and projected winter air temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. Ecoregions are ordered by the lowest
historical mean temperature.
* = 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
-26.14
10.96
-15.18
Chukchi
-25.52
12.56
-12.96
E. Bering
-8.56
7.73
-0.83
Gulf Alaska
0.52
3.89
4.41
Aleutians
1.09
3.33
4.42
N Pac Fijord
3.45
3.33
6.78
Puget-Georgia*
5.45
3.10
8.14
OR-WA-Vanc
7.44
2.86
10.3
N California
12.12
2.66
14.78
S California
16.31
2.71
19.02
40
-------
Table A-11. Historical, predicted increase, and projected winter air temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C.
* = 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
-27.03
6.67
-20.36
Chukchi
-26.31
7.73
-18.58
E Bering
-9.36
5.64
-3.72
Aleutians
0.25
2.68
2.93
Gulf Alaska
0.91
2.3
3.21
N Pac Fijord
3.88
2.17
6.05
Puget-Geogia*
5.95
2.03
7.97
OR-WA-Vanc
8.01
1.88
9.89
N California
12.29
1.72
14.01
S California
15.82
1.73
17.55
Table A-12. Historical, predicted increase, and projected SST temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C.
Ecoregion
SST
Historical mean
SST
Predicted Increase
SST
Projected Temperature
Beaufort
-1.07
2.29
1.22
Chukchi
-0.73
2.6
1.87
E Bering
3.02
3.56
6.58
Aleutians.
5.95
3.03
8.98
Gulf Alaska
6.95
3.1
10.05
N Pac Fijord
9.77
2.8
12.57
Puget-Georgia
11.34
2.15
13.49
OR-WA-Vanc
12.39
2.62
15.01
N California
16.37
2.54
18.91
S California
20.67
2.4
23.07
Table A-13. Historical, predicted increase, and projected SST temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C.
Ecoregion
SST
Historical mean
SST
Predicted Increase
SST
Projected Temperature
Beaufort
-1.15
1.26
0.11
Chukchi
-0.78
1.4
0.62
E Bering
2.76
2.26
5.02
Aleutians
5.79
2.09
7.88
41
-------
Ecoregion
SST
Historical mean
SST
Predicted Increase
SST
Projected Temperature
Gulf Alaska
6.78
2.07
8.85
N Pac Fijord
9.63
1.89
11.52
Puget-Georgia
11.34
1.34
12.68
OR-WA-Vanc
12.24
1.77
14.01
N California
15.74
1.67
17.41
S California
19.81
1.61
21.42
Table A-14. Historical, predicted increase, and projected 30 m depth temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. No data (ND) are available for the
Puget Trough/Georgia Basin Ecoregion.
Depth 30 m
Depth 30 m
Depth 30 m
UUvl Cy 1U1 I
Historical Mean
Predicted Increase
Projected Temperature
Beaufort
-1.4
0.33
-1.07
Chukchi
-1.09
0.33
-0.76
E Bering
1.75
2.82
4.57
Aleutians
4.59
2.54
7.13
Gulf Alaska
5.27
3.07
8.34
N Pac Fijord
8.11
2.89
11.0
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
10.35
2.74
13.09
N California
14.94
2.51
17.45
S California
19.55
2.51
22.06
Table A-15. Historical, predicted increase, and projected 30 m depth temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. No data (ND) are available for the
Depth 30 m
Depth 30 m
Depth 30 m
UUvl Cy 1U1 I
Historical Mean
Predicted Increase
Projected Temperature
Beaufort
-1.45
0.69
-0.76
Chukchi
-1.12
0.77
-0.35
E Bering
1.33
2.06
3.39
Aleutians
4.2
2.04
6.24
Gulf Alaska
4.99
1.98
6.97
N Pac Fijord
7.93
1.78
9.71
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
10.54
1.61
12.15
N California
14.56
1.5
16.06
S California
18.8
1.5
20.3
42
-------
Table A-16. Historical, predicted increase, and projected 100 m depth temperature by ecoregion under RCP 8.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. No data (ND) are available for the
Puget Trough/Georgia Basin Ecoregion. Ecoregions are ordered by the lowest historical mean temperature.
Depth 100 m
Depth 100 m
Depth 100 m
UUvl Cy 1U1 I
Historical Mean
Predicted Increase
Projected Temperature
Chukchi
-1.43
0.75
-0.68
Beaufort
-1.28
0.82
-0.46
E Bering
1.89
3.1
4.99
Aleutians
3.74
2.33
6.07
Gulf Alaska
4.95
2.77
7.72
N Pac Fijord
7.32
2.36
9.68
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
9.23
1.95
11.18
N California
11.96
1.54
13.5
S California
15.15
1.14
16.29
Table A-17. Historical, predicted increase, and projected 100 m depth temperature by ecoregion under RCP 4.5.
Predicted increases are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. Temperatures are in °C. No data (ND) are available for the
Puget Trough/Georgia Basin Ecoregion.
Depth 100 m
Depth 100 m
Depth 100 m Projected
UUvl Cy 1U1 I
Historical Mean
Predicted Increase
Temperature
Beaufort
-1.45
0.36
-1.09
Chukchi
-1.38
0.47
-0.91
E Bering
1.67
1.97
3.64
Aleutians
3.05
1.8
4.85
Gulf Alaska
4.35
1.81
6.16
N Pac Fijord
6.75
1.49
8.24
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.65
1.29
9.94
N California
11.05
1.16
12.21
S California
14.41
1.11
15.52
43
-------
Appendix B. Baseline pH/Aragonite
Saturation State Values and Predicted
Decreases
The baseline pH and aragonite saturation state values, predicted decreases based on the RCP 4.5 and
RCP 8.5 scenarios, and effects thresholds are documented in this appendix. The details of processing the
baseline and projected pH and aragonite saturation values are given in Lee et al., 2017. As discussed in
Section 2.7, ocean acidification risks are calculated from pH for decapods and fish and from aragonite
saturation state for bivalves.
Table B-l to Table B-6 give the historical, predicted pH decline, and projected pH values by ecoregion
for RCP 4.5 and RCP 8.5. pH values are given for the annual mean, summer, and winter. Table B-7 and
Table B-8 give the historical, predicted decline, and projected values for aragonite saturation state. The
synthesis of the most sensitive "minimum acceptable toxic concentrations" (MinATCs) of pH for
decapods is given in Table 6-6 of Lee et al., 2017 (referred to as MATC in that document). The
corresponding synthesis of the most sensitive MinATCs of pH for fish and of aragonite saturation state
for bivalves are given in Table B-9 and Table B-10, respectively. The high, moderate, and low pH
sensitivity thresholds for decapods and fish are given in Table B-l 1, Table B-12, and Table B-13,
respectively. The high, moderate, and low sensitivity thresholds of aragonite saturation state for bivalves
are given in Table B-14.
Table B-1. Historical, predicted decreases, and projected annual mean pH by ecoregion under RCP 8.5.
Predicted declines are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal (http://www.esrl.noaa.qov/psd/ipcc/ocn/). No data (ND) are
available for the Puget Trough/Georgia Basin Ecoregion, though the average of the N Pac Fijord and OR-WA-
Vanc ecoregions could be used as an approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.12
-0.35
7.77
Chukchi
8.11
-0.33
7.78
E Bering
8.11
-0.3
7.81
Aleutians
8.09
-0.28
7.81
Gulf Alaska
8.11
-0.3
7.81
N Pac Fijord
8.1
-0.3
7.8
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.11
-0.3
7.81
N California
8.1
-0.27
7.83
S California
8.09
-0.27
7.82
44
-------
Table B-2. Historical, predicted decrease, and projected annual mean pH by ecoregion under RCP 4.5.
Predicted declines are for the 2050-2099 timeframe. Predictions are based on the CMIP5 climate models
downloaded from the NOAA Climate Web Portal. No data (ND) are available for the Puget Trough/Georgia Basin
Ecoregion, though the average of the N Pac Fijord and OR-WA-Vanc ecoregions could be used as an
approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.13
-0.29
7.84
Chukchi
8.12
-0.27
7.85
E Bering
8.12
-0.25
7.87
Aleutians
8.1
-0.23
7.87
Gulf Alaska
8.12
-0.25
7.87
N Pac Fijord
8.11
-0.25
7.86
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.12
-0.25
7.87
N California
8.11
-0.23
7.88
S California
8.09
-0.22
7.87
Table B-3. Historical, predicted decrease, and projected summer mean pH by ecoregion under RCP 8.5.
Summer is the average pH for July, August, and September. Predicted declines are for 2050-2099 timeframe.
Predictions are based on the CMIP5 climate models downloaded from the NOAA Climate Web Portal. No data
(ND) are available for the Puget Trough/Georgia Basin Ecoregion, though the average of the N Pac Fijord and
OR-WA-Vanc ecoregions could be used as an approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.15
-0.38
7.77
Chukchi
8.15
-0.36
7.79
E Bering
8.1
-0.3
7.8
Aleutians
8.1
-0.28
7.82
Gulf Alaska
8.09
-0.29
7.8
N Pac Fijord
8.07
-0.28
7.79
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.08
-0.28
7.8
N California
8.07
-0.27
7.8
S California
8.04
-0.26
7.78
Table B-4. Historical, predicted decrease, and projected summer mean pH by ecoregion under RCP 4.5.
Summer is the average pH for July, August, and September. Predicted declines are for the 2050-2099 timeframe.
Predictions are based on the CMIP5 climate models downloaded from the NOAA Climate Web Portal. No data
(ND) are available for the Puget Trough/Georgia Basin Ecoregion, though the average of the N Pac Fijord and
OR-WA-Vanc ecoregions could be used as an approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.18
-0.27
7.91
Chukchi
8.15
-0.23
7.92
E Bering
8.08
-0.18
7.9
Aleutians
8.11
-0.16
7.95
Gulf Alaska
8.09
-0.18
7.91
N Pac Fijord
8.07
-0.17
7.9
45
-------
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.06
-0.16
7.9
N California
8.06
-0.16
7.9
S California
8.04
-0.15
7.89
Table B-5. Historical, predicted decrease, and projected winter mean pH by ecoregion under RCP 8.5.
Winter is the average pH for January, February, and March. Predicted declines are for the 2050-2099 timeframe
Predictions are based on the CMIP5 climate models downloaded from the NOAA Climate Web Portal. No data
(ND) are available for the Puget Trough/Georgia Basin Ecoregion, though the average of the N Pac Fijord and
OR-WA-Vanc ecoregions could be used as an approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.1
-0.33
7.77
Chukchi
8.09
-0.31
7.78
E Bering
8.1
-0.3
7.8
Aleutians
8.07
-0.28
7.79
Gulf Alaska
8.11
-0.3
7.81
N Pac Fijord
8.12
-0.31
7.81
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.13
-0.3
7.83
N California
8.13
-0.28
7.85
S California
8.12
-0.27
7.85
Table B-6. Historical, predicted decrease, and projected winter mean pH by ecoregion under RCP 4.5.
Winter is the average pH for January, February, and March. Predicted declines are for the 2050-2099 timeline.
Predictions are based on the CMIP5 climate models downloaded from the NOAA Climate Web Portal. No data
(ND) are available for the Puget Trough/Georgia Basin Ecoregion, though the average of the N Pac Fijord and
OR-WA-Vanc ecoregions could be used as an approximation.
Ecoregion
Historical pH
Predicted pH Decline
Projected pH
Beaufort
8.08
-0.21
7.87
Chukchi
8.09
-0.2
7.89
E Bering
8.11
-0.18
7.93
Aleutians
8.07
-0.16
7.91
Gulf Alaska
8.12
-0.18
7.94
N Pac Fijord
8.12
-0.18
7.94
Puget-Georgia
ND
ND
ND
OR-WA-Vanc
8.13
-0.17
7.96
N California
8.12
-0.16
7.96
S California
8.12
-0.16
7.96
46
-------
Table B-7. Historical, predicted decrease, and projected annual mean aragonite saturation state (Qa) by
ecoregion under RCP 8.5.
Baseline values and predictions are based on Cao and Caldeira (2008) from the University of Victoria Earth
System Climate Model version 2.8.
Ecoregion
Historical Saturation
Value
Predicted Saturation
Value Decline
Projected Saturation
Value
Beaufort
1.49
-0.84
0.65
Chukchi
1.44
-0.77
0.67
E Bering
2.33
-0.88
1.45
Aleutians
1.86
-0.85
1.01
Gulf Alaska
1.97
-0.87
1.1
N Pac Fijord
2.22
-1.00
1.22
Puget-Georgia
2.44
-1.02
1.42
OR-WA-Vanc
2.4
-1.03
1.37
N California
2.48
-1.11
1.37
S California
2.63
-1.18
1.45
Table B-8. Historical, predicted decrease, and projected annual mean aragonite saturation state (Qa) by
ecoregion under RCP 4.5.
Baseline values and predictions are based on Cao and Caldeira (2008) from the University of Victoria Earth
System Climate Model version 2.8.
Ecoregion
Historical Saturation
Value
Predicted Saturation
Value Decline
Projected Saturation
Value
Beaufort
1.49
-0.60
0.89
Chukchi
1.44
-0.55
0.89
E Bering
2.33
-0.58
1.75
Aleutians
1.86
-0.57
1.29
Gulf Alaska
1.97
-0.59
1.38
N Pac Fijord
2.22
-0.68
1.54
Puget-Georgia
2.44
-0.69
1.75
OR-WA-Vanc
2.40
-0.69
1.71
N California
2.48
-0.74
1.74
S California
2.63
-0.78
1.85
47
-------
Table B-9. Minimum Acceptable Toxicant Concentrations (MinATC) of pH for fish.
The pH MinATC for each fish species is based on the most sensitive sublethal or mortality endpoint for any life history stage. NOAEL = no observed
adverse effects level; HOAEL = highest observed adverse effects level; DPH = days post hatch. DPF = days post fertilization. DPS = days post spawning.
ELS = early life stage. The NOAEL is either the control pH or the lowest non-significant exposure, which are identified in the table by an asterisk (*).
MinATC is calculated as the geometric mean of the NOAEL and HOAEL. The procedure was to take the antilog of each pH, take the geometric mean of
the two values, then convert back to pH by taking logio of the geometric mean. Only species with significant effects are included.
Species
PH
NOAEL
PH
HOAEL
Most Sensitive
pH MinATC
Duration
of
Exposure
Life
Stage
End Point
Response
Type
Citation
Mustelus manazo
6.18*
6.02
6.1
72 hrs
Adult
Survival
Mortality
Hayashi et al.,
2004
Paralichthys olivaceus
6.41*
6.18
6.30
48 hrs
Adult
Survival
Mortality
Hayashi et al.,
2004
Seriola quinqueradiata
6.41*
6.18
6.30
8 hrs
Adult
Survival
Mortality
Hayashi et al.,
2004
Paralichthys dentatus
7.81
7.47
7.64
2-7 dpf
Embryo
(ELS)
Survival
Mortality
Chambers et al.,
2014
Notothenia rossii
7.914
7.493
7.70
4 weeks
Adult
Increased COX Activity
Physiology
Strobel et al., 2013
Menidia menidia
8.06
7.42
7.74
100 dph
Larval
(ELS)
Weight
Physiology
Murray et al., 2016
Clupea harengus
8.08
7.45
7.77
25 days
Larval
Organ Damage
Physiology
Frommel et al.,
2014
Sparus aurata
8.07
7.52
7.80
15 days
Larval
Growth
Development
Pimental et al.,
2016
Argyrosomus regius
8.09
7.53
7.81
10 days
Larval
Time Spent Swimming
Behavior
Pimental et al.,
2016
Oryzias latipes
8.05
7.6
7.83
2-5 dpf
Embryo
Development Time
Development
Tseng et al., 2013
Lepidopsetta polyxystra
8.07
7.63
7.85
33-87 dph
Post-
flexion
Larval
Length
(Longer Than Control)
Physiology
Hurst et al., 2017
Gasterosteus aculeatus
8.08
7.65
7.87
43 days
Adult
Survival
Mortality
Jutfelt et al., 2013
48
-------
Species
PH
NOAEL
PH
HOAEL
Most Sensitive
pH MinATC
Duration
of
Exposure
Life
Stage
End Point
Response
Type
Citation
Thunnus albacares
8.06
7.69
7.88
<1 day
Embryo
Time to Hatch
Development
Bromhead et al.,
2015
Amphiprion melanopus
8*
7.81
7.91
6-8 dps
Embryo
Egg Survival
Mortality
Miller et al., 2015
Sebastes diploproa
8.1
7.75
7.93
7 days
Juvenile
Time Spent in Dark Vs
Light Area
Behavior
Hamilton et al.,
2014
Pomacentrus amboinensis
7.97*
7.89
7.93
4 days
Juvenile
Visual Anti-Predator
Response
Behavior
Ferrari et al., 2011
Menidia beryllina
8.003*
7.875
7.94
8 days
Embryo
& Larval
(ELS)
Survival
Mortality
Baumann et al.,
2012
Amphiprion percula
8.15
7.8
7.98
11 dph
Larval
Ability of Larvae to Detect
Olfactory Cues from Adult
Habitats
Behavior
Munday et al., 2009
Ostorhinchus cyanosoma
8.15
7.8
7.98
7 days
Adult
Aerobic capacity
Physiology
Munday et al., 2009
Ostorhinchus doederleini
8.15
7.8
7.98
7 days
Adult
Aerobic capacity
Physiology
Munday et al., 2009
Theragra chalcogramma
8.02
7.97
7.99
6 weeks
Juvenile
Mean Otolith Increment
Width (MIW)
Development
Hurst et al., 2012
Plectropomus leopardus
8.04*
7.97
8.01
28 days
Juvenile
Predator Avoidance
Behavior
Munday et al., 2013
Sillago japonica
8.17
7.85
8.01
3 dph
Larval
(ELS)
Survival
Mortality
Yona et al., 2016
Pseudochromis fuscus
8.16
8.03
8.1
4-7 days
Adult
Response to Injured Prey
Behavior
Cripps et al., 2011
Cheilodipterus quinquelineatus
8.14
8.06
8.1
4 days
Adult
Homing Ability
Behavior
Devine et al., 2012
49
-------
Table B-10. Minimum Acceptable Toxicant Concentrations (MinATC) of aragonite saturation state (Qa) for bivalves.
The aragonite MinATC for each bivalve species is based on the most sensitive sublethal or mortality endpoint for any life history stage. NOAEL = no
observed adverse effects level; HOAEL=highest observed adverse effects level; HPF = hours post fertilization; DPF = days post fertilization; ELS = early
life stage; PDII =prodissoconch II. The NOAEL is either the control aragonite saturation state or the lowest non-significant exposure, which are identified in
the table by an asterisk (*). MinATC is calculated as the geometric mean of the NOAEL and HOAEL. Only species with significant effects are included.
Species
Qa
NOAEL
Qa
HOAEL
Most
Sensitive
Qa MinATC
Duration
of
Exposure
Life
Stage
End Point
Response
Type
Citation
Cerastoderma edule
0.1*
0.04
0.06
-80 d
Adult
Duration of Shell Growth
Development
Milano et al., 2016
Macoma balthica
1.02
0.38
0.62
20 d
Larval
Abundance
Mortality
Jansson et al., 2013
Mytilus trossulus
0.8*
0.5
0.63
20 d
Adult
Force to Break Byssal Plaques
Physiological
O'Donnell et al.,
2013
Saccostrea glomerata
1.15
0.64
0.86
24 h
Larval
Mortality (%)
Survival
Watson et al., 2009
Laternula elliptica
1.09
0.74
0.90
29 d
Larval
% Malformed
Development
Bylenga et al., 2017
Scrobicularia plana
1.83
0.45
0.91
7 d
Adult
Survivorship
Survival
Conradi et al., 2016
Ruditapes philippinarum
1.38*
0.67
0.96
84 d
Adult
Spawning Rate
Reproduction
Xu et al., 2016
Pecten maximus
1.13*
0.88
1.00
7 d
Veliger
Larvae
(ELS)
Survival
Mortality
Andersen et al.,
2013
Saxidomus gigantea
1.54
0.74
1.07
14 d
Juvenile
Shell Dissolution (PDII)
Physiological
Kelley et al., 2017
Mercenaria mercenaria
1.87
0.65
1.10
24 d
Larval
Growth (Shell Diameter)
Growth
Gobler and Talmage,
2013
Donax trunculus
1.6*
0.87
1.18
9 d
Embryo
(ELS)
Hatching Rate
Development
Pereira et al., 2016
Argopecten irradians
1.71
1.02
1.32
20 d
Larval
Growth (Shell Diameter)
Growth
Gobler and Talmage,
2013
Mytilus californianus
2.4
0.81
1.39
8 d
Larval
Shell Length
Growth
Frieder et al., 2014
Ostrea lurida
2.023
1.3
1.62
16 d
Larval
Percent Metamorphosis/Settlement
Growth
Hettinger et al., 2013
Tegillarca granosa
2.25
1.34
1.74
40 d
Adult
Clearance Rate
Physiological
Zhao et al., 2017
Area inflata
2.38*
1.29
1.75
144 hpf
Larval
Shell Length
Development
Li et al., 2014
Crassostrea gigas
2.1
1.6
1.83
48 h
Juvenile
Mid-Growth
Growth
Barton et al., 2012
Mytilus chilensis
2.38
1.51
1.90
70 d
Juvenile
Scope for Growth
Development
Navarro et al., 2013
50
-------
Species
Qa
NOAEL
Qa
HOAEL
Most
Sensitive
Qa MinATC
Duration
of
Exposure
Life
Stage
End Point
Response
Type
Citation
Mytilus edulis
2.75
1.38
1.95
2 d
D-
Veliger
Average Length of D-Shape Shell
Growth
Gazeau et al., 2010
Mytilus coruscus
2.93
1.33
1.97
72 h
Adult
Total Hemocyte Count
Physiological
and Immune
Functions
Sui et al., 2016
Pinctada fucata
3.15
1.24
1.98
96 h
Adult
Nacrein Expression
Physiological
Liu et al., 2012
Crassostrea angulata
2.9
1.91
2.35
24 hpf
Larval
% of Deformed D-Larvae
Development
Guo et al., 2015
My a are n ah a
2.63
2.11
2.36
60 d
Adult
Net Calcification (+)/Dissolution (-)
Physiological
Ries et al., 2009
Mytilus galloprovincialis
3.58
1.96
2.65
78 d
Juvenile
Tissue Weight
Growth
Fernandez-Reiriz et
al., 2012
Crassostrea hongkongensis
3.99
1.76
2.65
30 dpf
Larval
Reached Competent Pediveliger
Stage
Development
Dineshram et al.,
2013
Ruditapes decussatus
5.64
2.91
4.05
87 d
Juvenile
0:N Index
Physiological
Fernandez-Reiriez et
al., 2011
Crassostrea virginica
5.85
3.58
4.58
70 d
Juvenile
Mortality (%)
Survival
Dickinson et al.,
2012
Cerastoderma edule
0.1*
0.04
0.06
-80 d
Adult
Duration of Shell Growth
Development
Milano et al., 2016
51
-------
Table B-11. High sensitivity thresholds for pH for decapods and fish.
Taxa
Minor Risk
Low Risk
Moderate Risk
High Risk
Decapods
>7.96
7.91 to 7.95
7.88 to 7.9
<7.87
Fish
>8.1
8.02 to 8.09
7.98 to 8.01
<7.97
Table B-12. Moderate sensitivity thresholds for pH for decapods and fish.
Taxa
Minor Risk
Low Risk
Moderate Risk
High Risk
Decapods
>7.8
7.77 to 7.79
7.76 to 7.76
<7.75
Fish
>7.94
7.88 to 7.93
7.75 to 7.87
<7.74
Table B-13. Low sensitivity thresholds for pH for decapods and fish.
Taxa
Minor Risk
Low Risk
Moderate Risk
High Risk
Decapods
>7.6
7.5 to 7.59
7.36 to 7.49
<7.35
Fish
>7.7
7.58 to 7.69
6.62 to 7.57
<6.61
Table B-14. High, moderate, and low sensitivity thresholds of aragonite saturation state (Qa) for bivalves.
Aragonite saturation state thresholds are only used for bivalves.
Taxa / Sensitivity Group
Minor Risk
Low Risk
Moderate Risk
High Risk
Bivalves -
High sensitivity
>4.72
2.83 to 4.71
2.38 to 2.82
<2.37
Bivalves -
Moderate sensitivity
>2.2
1.86 to 2.19
1.62 to 1.85
<1.61
Bivalves -
Low sensitivity
>1.37
1.03 to 1.36
0.08 to 1.02
< 0.07
52
-------
Appendix C. Sea Level Rise Rates
This appendix documents the net sea level rise projections by ecoregion based on the ecoregion-specific
isostatic rates. A eustatic rate of 12 mm/yr was used in the RCP 8.5 risk analysis (Table C-l). In
comparison, a eustatic rate of 5 mm/yr was used in the RCP 4.5 risk analysis (Table C-2). Other
components of the sea level rise risk assessment, such as habitat thresholds, are detailed in Lee et al.,
2017.
Table C-1. Net sea level rise by ecoregion with an eustatic rate of 12 mm/yr and default isostatic rates.
The default ecoregion-scale isostatic rates were derived from data in NOAA's Tides and Currents
(httpsi//ttdesandcurrents.noaa.gov/sltrends/sltrends.htmI) and Proshutinsky et al., 2004. Net SLR projections are
based on a 100-hundred-year duration.
Ecoregion
Isostatic Rate
(mm/yr)
Eustatic Rate
(mm/yr)
Net SLR Projection
(mm)
Beaufort
-0.55
12
1145
Chukchi
0.77
12
1277
E Bering
0.97
12
1297
Aleutians
-5.86
12
614
Gulf Alaska
-8.16
12
384
N Pac Fijord
-10.08
12
192
Puget-Georgia
-0.79
12
1121
OR-WA-Vanc
-1.22
12
1078
N California
-0.55
12
1145
S California
0.13
12
1213
Table C-2. Net sea level rise by ecoregion with an eustatic rate of 5 mm/yr and default isostatic rates.
The default ecoregion-scale isostatic rates were derived from data in NOAA's Tides and Currents
(httpsi//ttdesandcurrents.noaa.gov/sltrends/sltrends.html') and Proshutinsky et al., 2004. Net SLR projections are
based on a 100-hundred-year duration.
Ecoregion
Isostatic Rate
(mm/yr)
Eustatic Rate
(mm/yr)
Net SLR Projection
(mm)
Beaufort
-0.55
5
445
Chukchi
0 77
5
577
E Bering
0.97
5
597
Aleutians
-5.86
5
-86
Gulf Alaska
-8.16
5
-316
N Pac Fijord
-10.08
5
-508
Puget-Georgia
-0.79
5
421
OR-WA-Vanc
-1 22
5
378
N California
-0.55
5
445
S California
0.13
5
513
53
-------
Glossary of Terms
Term
Algorithm-based risk
assessment
Aragonite
Aragonite saturation
state (Qa)
At-Risk
Baseline/Status Risk
Brachyuran crabs
Brooded
Calcite
CBRAT
Climate-adjusted
baseline/status risk
Coastal
Coastal
Biogeographic Risk
Analysis Tool
(CBRAT)
Coolest Occupied
Ecoregion (COE)
Constrained
Coupled Model Inter-
comparison Project
Phase 5 (CMIP5)
Definition
Risk assessment based on a knowledge base (database) and a rule set, with no expert
intervention in calculating risks. Used to avoid the limitations of expert solicitations, including
potential sources of bias.
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.
Species are classified as At-Risk if one or more of the climate drivers has a moderate risk or
a high risk. At-Risk species are assumed to be impacted by climate change.
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.
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.
The larval or juvenile phase is brooded within the adult or tube of the adult; ovoviviparous.
Carbonate mineral CaCCb.
Coastal Biodiversity Risk Analysis Tool (http://www.cbrat.om).
Greatest baseline/status risks weighted by the greatest climate risk.
In the context of the current case study, coastal refers to estuaries and the open ocean out
to 200 m depth.
Ecoinformatic tool synthesizing life history, habitat, distributional, and abundance data on
near-coastal species. Predicts vulnerability to climate change, including temperature
increases, ocean acidification, and sea level rise. Available at http://www.cbrat.org.
In CBRAT, the COE is the coolest ecoregion in which the species maintains a viable
population. Usually the northernmost occupied ecoregion. The temperature associated with
a COE will have different values depending on whether it is based on air temperatures,
SSTs, or subsurface temperatures.
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 (httpi//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).
54
-------
Term
Deep subtidal
Direct development
Eco region
Endemic
Eustatic sea level rise
(ESLR)
Highest observed
adverse effects level
(HOAEL)
Isostatic adjustment
Lecithotrophic larvae
Lithodid crabs
Lowest observed
adverse effects level
(LOAEL)
Marine Ecoregions of
the World (MEOW)
Maximum acceptable
toxicant concentration
(MATC)
Minimum acceptable
toxicant concentration
(MinATC)
Near coastal
Definition
> 30 - 200 m depth.
Development without a larval phase.
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.org/publications/marine-ecoreqions-of-the-world-a-
bioregionalization-of-coastal-and-shelf-areas.
Species only located in a restricted location. In CBRAT, defined as species occurring in only
one ecoregion.
Worldwide change in sea level primarily caused by thermal expansion of sea water and
melting of glaciers and ice sheets.
The highest pH or aragonite saturation state value resulting in an adverse effect. Analogous
to the lowest observed adverse effects level (LOAEL) for pollutants and pesticides.
Vertical movement of the earth's plates resulting in local uplift or subsidence and resulting in
the raising or lowering of sea level. Either mitigates or increases effects of sea level rise.
Larvae that derive nourishment from yolk (vs. planktotrophic larvae).
Crabs of the families Lithodidae and Hapalogastridae of the Infraorder Anomura. King
crabs; not "true" crabs of the Infraorder Brachyura.
With pollutants and pesticides, this is the lowest concentration resulting in an adverse effect.
See "highest observed adverse effects level" (HOAEL).
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/publicatioris/marine-ecoregions-of-the-world-a-
bioregionalizatlon-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). For pH and aragonite saturation state, see MinATC..
The lowest acceptable pH or aragonite saturation state value. It is calculated as the
geometric mean of the NOAEL and the HOAEL; with pH, the values are first transformed by
taking the anti-log. Analogous to maximum acceptable toxicant concentration (MATC) for
po 11 uta nts/pesticid es.
As used in CBRAT, the region from the supratidal down to 200 m depth. Includes both
estuaries and offshore areas.
55
-------
Term
Definition
Next Coolest
Unoccupied
Ecoregion (NCUE)
Next Warmest
Unoccupied
Ecoregion (NWUE)
Northeast Pacific
(NEP)
Occupied ecoregions
Ocean acidification
(OA)
Overall vulnerability
Ovoviviparous
PH
Planktotrophic larvae
Radiative Forcing
Relative sea level rise
(RSLR)
In CBRAT, the 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. The
temperature associated with a NCUE will have different values depending on whether it is
based on air temperatures, SSTs, or subsurface temperatures.
In CBRAT, the 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. The
temperature associated with a NWUE will have different values depending on whether it is
based on air temperatures, SSTs, or subsurface temperatures
As used in CBRAT, the near-coastal region from the Aleutians Islands through the Gulf of
California. However, in this case study, limited to the Aleutians through Southern California.
Ecoregions occupied by the target species ortaxon. In the current case study, only the
ecoregions from the Beaufort through Southern California are included.
A reduction in the pH of the ocean caused primarily by uptake of carbon dioxide (CO2) from
the atmosphere.
Single risk assigned to a species based on the greatest risk among the climate drivers:
temperature, ocean acidification, SLR, and baseline/status traits
Eggs develop within the female, or male in some cases, but the embryo derives no
nourishment from the parent. A brooder.
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 feed on other organisms.
The difference between insolation (sunlight) absorbed by the Earth and energy radiated
back to space.
The net change in sea level at a particular location due to both eustatic SLR and local
factors such as isostatic adjustments.
56
-------
Term
Representative
Concentration
Pathways (RCP)
Rule-based system
Scenario modelling
Shallow subtidal
Sum-Occupied
Summer temperatures
(months used)
Symbiont
Temperature-adjusted
ocean acidification
risk
Transient
Uncertainty Analysis
Unconstrained
Warmest occupied
ecoregion (WOE)
Winter temperatures
(months)
Definition
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).
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.
Evaluation of how risks change under different climate scenarios.
> 0 - 30 m depth.
Sum of the number of ecoregions within the U.S. Arctic and NEP inhabited by the species
within a taxon
July, August and September: used in modeling the effects of summer temperature
increases.
Organisms living in direct contact or close physical proximity with another organism,
including mutualistic (+/+), commensal (+/0), neutral (0/0), and negative (-/+ or-/0)
relationships.
Risk due to reduced pH or aragonite saturation state incorporating interaction with
enhanced temperatures.
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.
In CBRAT, the WOE is the warmest ecoregion in which the species maintains a viable
population. Usually the southernmost occupied ecoregion. The temperature associated with
a WOE will have different values depending on whether it is based on air temperatures,
SSTs, or subsurface temperatures
January, February, and March: used in modeling the effects of winter temperature increases
in CBRAT.
57
-------
Bibliography
Andersen, S., Grefsrud, E.S., and Harboe, T. 2013. Effect of increased pCC>2 level on early shell development in
great scallop (Pecten maximus Lamarck) larvae. Biogeosciences 10: 6161-6184.
Barton, A., Hales, B., Waldbusser, G.G., Langdon, C., and Feely, R.A. 2012. The Pacific oyster, Crassostrea
gigas, shows negative correlation to naturally elevated carbon dioxide levels: Implications for near-term
ocean acidification effects. Limnology and Oceanography 57: 698-710.
Barton, A., Waldbusser, G.G., Feely, R.A., Weisberg, S.B., Newton, J.A., Hales, B., Cudd, S., Eudeline, B.,
Langdon, C.J., Jefferds, I., King, T., Suhrbier, A., and McLaughlin, K. 2015. Impacts of coastal
acidification on the Pacific Northwest shellfish industry and adaptation strategies implemented in
response. Oceanography 28:146-159.
Baumann, H., Talmage, S.C., and Gobler, C.J. 2012. Reduced early life growth and survival in a fish in direct
response to increased carbon dioxide. Nature Climate Change 2 :38-41.
Bromhead, D., Scholey, V., Nicol, S., Margulies, D., Wexler, J., Stein, M., Hoyle, S., Lennert-Cody, C.,
Williamson, J., Havenhand, J., Ilyina, T., and Lehodey, P. 2015. The potential impact of ocean
acidification upon eggs and larvae of yellowfin tuna (Thunnus albacares). Deep Sea Research Part II:
Topical Studies in Oceanography 113: 268-279.
Byers, J.E. 2000. Competition between two estuarine snails: implications for invasions of exotic species. Ecology
81: 1225-1239.
Bylenga, C.H., Cummings, V.J., and Ryan, K.G. 2017. High resolution microscopy reveals significant impacts of
ocean acidification and warming on larval shell development in Laternula elliptica. PLoS ONE 12(4):
e0175706. https://doi.orq/10.1371/ioymal.poiie.0175706.
Cao, L. and Caldeira, K. 2008. Atmospheric CO2 stabilization and ocean acidification. Geophysical Research
Letters 35: 1-5.
Chambers, R.C., Candelmo, A.C., Habeck, E.A., Poach, M.E., Wieczorek, D., Cooper, K.R., and Greenfield, C.E.
2014. Effects of elevated CO2 in the early life stages of summer flounder, Paralichthys dentatus, and
potential consequences of ocean acidification. Biogeosciences: 1613-1626.
Conradi, M., Riba, I., Almagro-Pastor, V., and DelValls, T.A. 2016. Lethal and sublethal responses in the clam
Scrobicularia plana exposed to different C02-acidic sediments. Environmental Research 151: 642-652.
Cripps, I.L., Munday, P.L., and McCormick, M.I. 2011. Ocean acidification affects prey detection by a predatory
reef fish. PLoS ONE 6(7): e22736. doi:10.1371/journal.pone.0022736.
Devine, B.M., Munday, P.L., and Jones, G.P. 2012. Homing ability of adult cardinalfish is affected by elevated
carbon dioxide. Oecologia, 269-276.
Dickinson, G.H., Ivanina, A.V., Matoo, O.B., Portner, H.O., Lannig, G., Bock, C. Beniash, E., and Sokolova, I.M.
2012. Interactive effects of salinity and elevated CO2 levels on juvenile eastern oysters, Crassostrea
virginica. The Journal of Experimental Biology 215: 29-43.
Dineshram, R., Thiyagarajan, V., Lane, A., Ziniu, Y., Xiao, S., and Leung, P.T.Y. 2013. Elevated CO2 alters larval
proteome and its phosphorylation status in the commercial oyster, Crassostrea hongkongensis. Marine
Biology 160: 2189-2205.
58
-------
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.
Fernandez-Reiriz, M.J., Range, P., Alvarez-Salgado, X.A., Espinosa, J., and Labarta, U. 2012. Tolerance of
juvenile Mytilus galloprovincialis to experimental seawater acidification. Marine Ecology Progress Series
454: 65-74.
Ferrari, M.C.O., Dixson, D.L., Munday, P.L., McCormick, M.I., Meekan, M.G., Sih, A., and Chivers, D.P. 2011.
Intrageneric variation in antipredator responses of coral reef fishes affected by ocean acidification:
implications for climate change projections on marine communities. Global Change Biology 17: 2980-
2986.
Frieder, C.A., Gonzalez, J.P., Bockmon, E.E., Navarro, M.O., and Levin, L.A. 2014. Can variable pH and low
oxygen moderate ocean acidification outcomes for mussel larvae? Global Change Biology 20: 754-764.
Frommel, A.Y., Maneja, F., Lowe, D., Pascoe, C.K., Geffen, A.J., Folkvord, A., Piatkowski, U., and Clemmensen,
C. 2014. Organ damage in Atlantic herring larvae as a result of ocean acidification. Ecological
Applications 24:1131-1143.
Gazeau, F., Gattuso, J.P., Dawber, C., Pronker, A.E., Peene, F., Peene, J., Heip, C.H.R., and Middelburg, J.J.
2010. Effect of ocean acidification on the early life stage of the blue mussel Mytilus edulis.
Biogeosciences 7: 2051-2060.
Gobler, C.J. and Talmage, S.C. 2013. Short- and long-term consequences of larval stage exposure to constantly
and ephemerally elevated carbon dioxide for marine bivalve populations. Biogeosciences 10: 2241-2253.
Guo, X., Huang, M., Pu, F., You, W., and Ke, C. 2015. Effects of ocean acidification caused by rising CO2 on the
early development of three mollusks. Aquatic Biology 23: 147-157.
Hale, S.S., Buffum, H.W., Kiddon, J.A., and Hughes, M.M. 2017. Subtidal benthic invertebrates shifting northward
along the US Atlantic Coast. Estuaries and Coasts DOI 10.1007/s12237-017-0236-z
Hamilton, T.J., Holcombe, A., and Tresguerres, M. 2014. CC>2-induced ocean acidification increases anxiety in
Rockfish via alteration of GABAA receptor functioning. Proceedings of the Royal Society B 281:
20132509.
Hannah, L. (Ed.) 2012. Saving a Million Species: Extinction Risk from Climate Change. Island Press, Washington.
432 pages.
Hayashi, M., Kita, J., and Ishimatsu, A. 2004. Acid-base responses to lethal aquatic hypercapnia in three marine
fishes. Marine Biology 144: 153-160.
Hettinger, A., Sanford, E., Hill, T.M., Hosfelt, J.D., Russell, A.D., and Gaylord, B. 2013. The influence of food
supply on the response of Olympia oyster larvae to ocean acidification. Biogeosciences 10: 6629-6638.
Hurst, T.P., Fernandez, E.R., Mathia, J.T., Miller, J.A., Stinson, C.M., and Ahgeak, E.F. 2012. Resiliency of
juvenile walleye pollock to projected levels of ocean acidification. Aquatic Biology 17: 247-259.
Hurst., T.P., Laurel, B.J., Hanneman, E., Haines, S.A., and Ottmar, M.L. 2017. Elevated CO2 does not exacerbate
nutritional stress in larvae of a Pacific flatfish. Fisheries Oceanography 26(3): 336-349.
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.
https://www.ipcc.ch/pdf/assessment-report/arS/svr/S INAL full.pdf.
59
-------
Jansson, A., Norkko, J., and Norkko, A. 2013. Effects of reduced pH on Macoma balthica larvae from a system
with naturally fluctuating pH-dynamics. PLoS ONE 8: e68198.
Jorgensen, L.L. and Primicerio, R. 2007. Impact scenario for the invasive red king crab Paralithodes
camtschaticus (Tilesius, 1815) (Reptantia, Lithodidae) on Norwegian, native epibenthic prey.
Hydrobiologia 590: 47-54.
Jutfelt, F., de Souza, K.B., Vuylsteke, A., and Sturve, J. 2013. Behavioural disturbances in a temperate fish
exposed to sustained high CO2 levels. PLoS One 8:1-6. doi:10.1371/journal.pone.0065825
Katsanevakis, S., Wallentinus, I., Zenetos, A., Leppakoski, E., Cinar, M.E., Ozturk, B., Grabowski, M., Golani, D.,
and Cardoso, A.C. 2014. Impacts of marine invasive alien species on ecosystem services and
biodiversity: a pan-European review. Aquatic Invasions 9: 391-423.
Kelley, A.L., Miller, C.A., Hetrick, J., and Ramsay, J. 2017. Examining the effects of ocean acidification on a
native Alaskan bivalve. Saxidomus gigantea, the butter clam. Presentation by Ocean Acidification
Research Center, University of Alaska Fairbanks, AK; Alutiiq Pride Shellfish Hatchery; Hakai Institute.
http://alutiiqpridehatcherv.com/wp-content/uploads/2017/06/QA PI meeting 12312016.pdf
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/ORD/R-
17/052. 304 pages, http://www.cbrat.org/pclfs/aerieral/vylrierabilitv framework ndf.
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/601 /B—15/001. 123 pages.
http://www.cbrat.org/pdfs/gerieral/cbrat users guide and metadata.pdf.
Li, J., Jiang, Z., Zhang, J., Mao, Y., Bian, D., and Fang, J. 2014. The potential of ocean acidification on
suppressing larval development in the Pacific oyster Crassostrea gigas and blood cockle Area inflata
Reeve. Chinese Journal of Oceanology and Limnology 32: 1307-1313.
Liu, W., Huang, X., Lin, J., and He., M. 2012. Seawater acidification and elevated temperature affect gene
expression patterns of the pearl oyster Pinctada fucata. PLoS ONE 7(3): e33679.
doi: 10.1371 /journal, pone. 0033679.
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.
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/sypportl erial/uncertainty-
guida te.pdf.
Milano, S., Schone, B.R., Wang, S., and Muller, W.E. 2016. Impact of high PCO2 on shell structure of the bivalve
Cerastoderma edule. Marine Environmental Research 119: 144-155.
Miller, G.M., Kroon, F.J., Metcalfe, S., and Munday, P.L. 2015. Temperature is the evil twin: effects of increased
temperature and ocean acidification on reproduction in a reef fish. Ecological Applications 25: 603-620.
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.
60
-------
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.
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.
Munday, P.L., Dixson, D.L., Donelson, J.M., Jones, G.P., Pratchett, M.S., Devitsina, G.V., and D0ving, K.B. 2009.
Ocean acidification impairs olfactory discrimination and homing ability of a marine fish. Proceedings of the
National Academy of Sciences 106: 1848-1852.
Munday, P.L., Pratchet. M.S., Dixson, D.L., Donelson, J.M., Endo, G.G.K., Reynolds, A.D., and Knuckey, R.
2013. Elevated CO2 affects the behavior of an ecologically and economically important coral reef fish.
Marine Biology 160: 2137-2144.
Murray, C.S., Fuiman, L.A., and Baumann, H. 2016. Consequences of elevated CO2 exposure across multiple life
stages in a coastal forage fish. ICES Journal of Marine Science 74: 1051-1061.
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.
Navarro, J.M., Torres, R., Acuna, K., Duarte, C., Manriquez, P.H., Lardies, M., Lagos, N.A., Vargas, C., and
Aguilera, V. 2013. Impact of medium-term exposure to elevated PCO2 levels on the physiological
energetics of the mussel Mytilus chilensis. Chemosphere 90: 1242-1248.
Nichols, F.H., Thompson, J.K., and Schemel, L.E. 1990. Remarkable invasion of San Francisco Bay (California,
USA) by the Asian clam Potamocorbula amurensis. II. Displacement of a former community. Marine
Ecology Progress Series 66: 95-101.
Nowell, L.H., Norman, J.E., Ingersoll, C.G., and Moran, P.W. 2016. Development and application of freshwater
sediment-toxicity benchmarks for currently used pesticides. Science of the Total Environment 550: 835-
850.
O'Donnell, M.J., George, M.N., and Carrington, E. 2013. Mussel byssus attachment weakened by ocean
acidification. Nature Climate Change 3: 587-590.
Orlov, Yu. I. and Ivanov, B.G. 1978. On the introduction of the Kamchatka king crab Paralithodes camtschatica
(Decapod: Anomura: Lithodidae) into the Barents Sea. Marine Biology 48: 373-375.
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 forthe U.S. National Climate
Assessment. NOAA Technical Report OAR CPO-1. 37 pages.
https://scenarios.qlobalchanqe.gov/sttes/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 forthe Arctic Ocean ecoregions: U.S. Geological Survey Open-File Report
2011-1246, 20 pages, http://pubs.ysas.aov/of/2011/1246/.
Payne, M.C., Reusser, D.A., Lee II, H., and Brown, C.A. 2011. Moderate-resolution sea surface temperature data
forthe nearshore North Pacific: U.S. Geological Survey Open-File Report: 2010-1251.
http://pubs.ysas.gov/of/2010/1251/.
61
-------
Pereira, A.M., Range, P., Campoy, A., Oliveira, A.P., Joaquim, S., Matias, D., Chicharo, L., and Gaspar, M.B.
2016. Larval hatching and development of the wedge shell (Donax trunculus L.) under increased CO2 in
southern Portugal. Regional Environmental Change 16: 855-864.
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.
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/.
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.
Riahi, K., Krey, V., Rao, S., Chirkov, V., Fischer, G., Kolp, P., Kindermann, G., Nakicenovic, N., and Rafai, P.
2011. RCP 8.5 - A Scenario of Comparatively High Greenhouse Gas Emissions. Climatic Change 109:
33-57
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.
Salway, R. and Shaddick, G. No date. Implementation of Qualitative Uncertainty Guidance: A Worked Example.
http ://www .integrated-
assessment.eu/eu/sttes/default/files/Example%20of%20aualitative%20uncertaintv%20analvsis.pdf.
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. 130 pages.
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.
Strobel, A., Leo, E., Portner, H.O., and Mark F.C. 2013. Elevated temperature and PCO2 shift metabolic pathways
in differentially oxidative tissues of Notothenia rossii. Comparative Biochemistry and Physiology, Part B
166: 48-57.
Sui, Y., Kong, H., Huang, X., Dupont, S., Hu, M., Storch, D., Portner, H.-O., Lu, W., and Wang, Y. 2016.
Combined effects of short-term exposure to elevated CO2 and decreased 02 on the physiology and
energy budget of the thick shell mussel Mytilus coruscus. Chemosphere 155: 207-216.
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.
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 offish (Oryzias
latipes). Comparative Biochemistry and Physiology, Part A 165: 119-130.
62
-------
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). 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.
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.
Waldbusser, G. G., Hales, B., Langdon, C. J., et al. 2015. Ocean acidification has multiple modes of action on
bivalve larvae. PLoS One 10(6): e0128376.
Wallace, R.B., Baumann, H., Grear, J.S., Aller, R.C., and Gobler, C.J. 2014. Coastal ocean acidification: the other
eutrophication problem. Estuarine, Coastal and Shelf Science 148:1-13.
Watson, S.-A., Southgate, P.C., Tyler, P.A., and Peck, L.S. 2009. Early larval development of the Sydney rock
oyster Saccostrea glomerata under near-future predictions of C02-driven ocean acidification. Journal of
Shellfish Research 28: 431-437.
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.
Xu, X., Yang, F., Zhao, L., and Yan, X. 2016. Seawater acidification affects the physiological energetics and
spawning capacity of the Manila clam Ruditapes philippinarum during gonadal maturation. Comparative
Biochemistry and Physiology, Part A 196: 20-29.
Yona, G., Madalla, N., Mwandya, A., Ishimatsu, A., and Bwathondi, P. 2016. Effects of elevated carbon dioxide
and temperature on survival and morphology of Japanese whiting Sillago japonica. International Journal
of Fisheries and Aquatic Studies 4: 48-55.
Zhao, X., Shi, W., Han, Y., Lui, S., Guo, C. Fu, W., Chai, X., and Lui, G. 2017. Ocean acidification adversely
influences metabolism, extracellular pH and calcification of an economically important marine bivalve,
Tegillarca granosa. Marine Environmental Research 125: 82-89.
63
-------
SCIENCE
*>EPA
United States
Environmental Protection
Agency
Office of Research and Development (8101R)
Washington, DC 20460
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Official Business
Penalty for Private Use
$300
Recycled/Recyclable Printed on paper that contains a minimum of
50% postconsumer fiber content processed chlorine free
u~r
------- |